A Practical Approach, Second Edition=Ronald D. Ho.pdf

Second EditionDEVELOPMENTALandREPRODUCTIVETOXICOLOGYA Practical Approach© 2006 by Taylor & Francis Group, LLC

Second EditionDEVELOPMENTALandREPRODUCTIVETOXICOLOGYA Practical ApproachEdited byRonald D. HoodBoca Raton London New YorkA CRC title, part of the Taylor & Francis imprint, a member of theTaylor & Francis Group, the academic division of T&F Informa plc.© 2006 by Taylor & Francis Group, LLC

Published in 2006 byCRC PressTaylor & Francis Group6000 Broken Sound Parkway NW, Suite 300Boca Raton, FL 33487-2742© 2006 by Taylor & Francis Group, LLCCRC Press is an imprint of Taylor & Francis GroupNo claim to original U.S. Government worksPrinted in the United States of America on acid-free paper10 9 8765 4321International Standard Book Number-10: 0-8493-1254-X (Hardcover)International Standard Book Number-13: 978-0-8493-1254-0 (Hardcover)Library of Congress Card Number 2005043988This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted withpermission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publishreliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materialsor for the consequences of their use.No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, orother means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any informationstorage or retrieval system, without written permission from the publishers.For permission to photocopy or use material electronically from this work, please access www.copyright.com(http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. Fororganizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only foridentification and explanation without intent to infringe.Library of Congress Cataloging-in-Publication DataDevelopmental and reproductive toxicology : a practical approach / edited by Ronald D. Hood. -- 2nd ed.p. cm.First ed. published in 1997 under title: Handbook of developmental toxicology.ISBN 0-8493-1254-X1. Developmental toxicology--Handbooks, manuals, etc. 2. Reproductive toxicology--Handbooks,manuals, etc. I. Hood, Ronald D. II. Handbook of developmental toxicology.RA1224.2.H36 2005615.9--dc22 2005043988Taylor & Francis Groupis the Academic Division of T&F Informa plc.Visit the Taylor & Francis Web site athttp://www.taylorandfrancis.comand the CRC Press Web site athttp://www.crcpress.com© 2006 by Taylor & Francis Group, LLC

CONTENTSPART IPRINCIPLES AND MECHANISMS .............................................................................................1Chapter 1 Principles of Developmental Toxicology Revisited ..................................................3Ronald D. HoodChapter 2Chapter 3Chapter 4Chapter 5Experimental Approaches to Evaluate Mechanisms of DevelopmentalToxicity.....................................................................................................................15Elaine M. Faustman, Julia M. Gohlke, Rafael A. Ponce,Thomas A. Lewandowski, Marguerite R. Seeley, Stephen G. Whittaker,and William C. GriffithPathogenesis of Abnormal Development.................................................................61Lynda B. Fawcett and Robert L. BrentMaternally Mediated Effects on Development........................................................93Ronald D. Hood and Diane B. MillerPaternally Mediated Effects on Development.......................................................125Barbara F. Hales and Bernard RobaireChapter 6 Comparative Features of Vertebrate Embryology .................................................147John M. DeSessoPART IIHAZARD AND RISK ASSESSMENT AND REGULATORY GUIDANCE........................199Chapter 7 Developmental Toxicity Testing — Methodology ................................................201Rochelle W. Tyl and Melissa C. MarrChapter 8 Nonclinical Juvenile Toxicity Testing ...................................................................263Melissa J. Beck, Eric L. Padgett, Christopher J. Bowman,Daniel T. Wilson, Lewis E. Kaufman, Bennett J. Varsho,Donald G. Stump, Mark D. Nemec, and Joseph F. HolsonChapter 9Significance, Reliability, and Interpretation of Developmental andReproductive Toxicity Study Findings .................................................................329Joseph F. Holson, Mark D. Nemec, Donald G. Stump, Lewis E. Kaufman,Pia Lindström, and Bennett J. VarshoChapter 10 Testing for Reproductive Toxicity .........................................................................425Robert M. ParkerChapter 11The U.S. EPA Endocrine Disruptor Screening Program: In Vitro and In VivoMammalian Tier I Screening Assays.....................................................................489Susan C. Laws, Tammy E. Stoker, Jerome M. Goldman, Vickie Wilson,L. Earl Gray, Jr., and Ralph L. Cooper© 2006 by Taylor & Francis Group, LLC

Chapter 12Chapter 13Chapter 14Chapter 15Chapter 16Role of Xenobiotic Metabolism in Developmental and Reproductive Toxicity.......525Arun P. KulkarniUse of Toxicokinetics in Developmental and Reproductive Toxicology..............571Patrick J. WierCellular, Biochemical, and Molecular Techniques in DevelopmentalToxicology..............................................................................................................599Barbara D. Abbott, Mitchell B. Rosen, and Gary HeldFunctional Genomics and Proteomics in Developmental andReproductive Toxicology .......................................................................................621Ofer Spiegelstein, Robert M. Cabrera, and Richard H. FinnellIn Vitro Methods for the Study of Mechanisms of DevelopmentalToxicology..............................................................................................................647Craig Harris and Jason M. HansenChapter 17 Statistical Analysis for Developmental and Reproductive Toxicologists .............697James J. ChenChapter 18Chapter 19Chapter 20Chapter 21Quality Concerns for Developmental and Reproductive Toxicologists................713Kathleen D. Barrowclough and Kathleen L. ReedPerspectives on the Developmental and Reproductive Toxicity Guidelines.........733Mildred S. Christian, Alan M. Hoberman, and Elise M. LewisHuman Studies — Epidemiologic Techniques in Developmental andReproductive Toxicology .......................................................................................799Bengt KällénDevelopmental and Reproductive Toxicity Risk Assessmentfor Environmental Agents......................................................................................841Carole A. Kimmel, Gary L. Kimmel, and Susan Y. EulingChapter 22 Principles of Risk Assessment — FDA Perspective .............................................877Thomas F. X. Collins, Robert L. Sprando, Mary E. Shackelford,Marion F. Gruber, and David E. MorseAppendix A Terminology of Anatomical Defects.....................................................................911Appendix B Books Related to Developmental and Reproductive Toxicology.........................967Appendix C Postnatal Developmental Milestones ....................................................................969© 2006 by Taylor & Francis Group, LLC

PrefaceThe areas of developmental and reproductive toxicology are becoming increasingly important.Developmental toxicology encompasses the study of hazard and risk associated with exposure totoxicants during prenatal development and has been expanded in recent years to include effects onthe developmental process until the time of puberty, i.e., until the completion of all developmentalprocesses. It is well known that a considerable percentage of newborns have significant anatomicaldefects, that birth defects are a major cause of hospitalization of infants, that “spontaneous” abortionand perinatal death are common, and that numerous individuals suffer from congenital functionaldeficits such as mental retardation. Although considerable progress has been made in determiningthe cause of these defects, the etiology of the majority of birth defects is not yet known or onlypoorly established. We still must learn much about the mechanisms involved in eliciting congenitaldefects and the genetic and environmental factors and their interactions that trigger these mechanisms.Reproductive toxicology is the study of adverse effects of chemical or physical agents onthe reproductive processes of both sexes and the causes of such effects. Concern has been voicedregarding reproductive issues, such as the possibility that human sperm counts have decreased inat least some geographic areas in recent times and that early menarche and precocious breastdevelopment may have environmental causes. Thus it is certain that both mechanistic studies ofknown developmental and reproductive toxicants and the toxicological assessment of pharmaceuticalagents, food additives, pesticides, industrial chemicals, environmental pollutants, and the liketo which humans may be exposed will be of importance for the foreseeable future. This situationpoints to the need for useful references in the field of developmental toxicology, teratology, andreproductive toxicology, and provides the impetus for the current volume.The purpose of this book is to provide a practical guide to the practice of developmental andreproductive toxicology; inclusion in a single volume of material from these areas will be of valueto the many individuals with professional responsibilities in both. Further, this book providesinformation in one source that is currently scattered through the literature or has not been readilyavailable; it provides much of that information in considerable detail. Although the current workis primarily oriented toward research designed to establish the likelihood of harm to humans, itshould also prove useful to those who are primarily interested in effects on other organisms. Thisbook should be especially useful for individuals working in industry who are responsible for testingchemical agents for developmental or reproductive toxicity and for those who manage such endeavors.It should be helpful to regulatory scientists at all levels of government who must evaluate theadequacy of studies and who must interpret data on hazard and establish the potential risk fromexposures. This volume will also be useful in training students and technicians, and for individualsactive in other areas who find the need to become familiar with the principles and practice ofdevelopmental or reproductive toxicology.Developmental and Reproductive Toxicology: A Practical Approach is intended to be a practicalguide as well as informative, providing insights gained from hands-on experience along with atheoretical foundation. Although this book is intended to be a relatively comprehensive guide tothe fields of developmental and reproductive toxicology, there were practical limits to the numberand scope of areas that it could address. It should also be noted that mention of vendors, tradenames, or commercial products does not constitute an endorsement or recommendation for use.The editor wishes to especially thank the contributing authors, whose efforts and expertise madethis project a success. Thanks also go to Taylor & Francis and the following individuals in itsemploy: Stephen Zollo, who proposed that the current work be published as a follow-up to theHandbook of Developmental Toxicology, which I had previously edited, and who has providedinvaluable advice and encouragement on the project, and to Patricia Roberson, who has providedessential technical guidance during the process of bringing the book to publication.© 2006 by Taylor & Francis Group, LLC

The EditorRonald D. Hood, Ph.D., is Professor Emeritus of Biological Sciences, having retired from thefaculty of the Cell, Molecular, and Developmental Biology Section of the Department of BiologicalSciences of The University of Alabama, Tuscaloosa, Alabama, in June, 2000. Dr. Hood remainsactive in research and consulting, and has retained his office and laboratory at the university. Heis also principal of Ronald D. Hood and Associates, Toxicology Consultants, and Adjunct Professorof Public Health in the School of Public Health of the University of Alabama at Birmingham. Dr.Hood received his B.S. and M.S. degrees from Texas Tech University and his Ph.D. in reproductivephysiology from Purdue University (1969). He joined the faculty of The University of Alabama in1968 as assistant professor, advanced to the rank of full professor in 1978, and served as interimdepartment chair in 1996 to 1997. Dr. Hood was also Consultant in Environmental Medicine, U.S.Veterans Administration, Office of Medicine and Surgery, Agent Orange Special Projects Office(off site) during 1983 and Special Consultant, Science Advisory Board, U.S. Environmental ProtectionAgency, from 1983 through 1993. In addition, he has served as a consultant to industrialclients, trade associations, federal and state agencies, and law firms since 1978, and as a grant ordocument reviewer for the Environmental Protection Agency, the Agency for Toxic Substances andDisease Registry, the Congressional Office of Technology Assessment, ICCVAM, and the NationalResearch Council.Dr. Hood is a member of the Teratology Society, the Society of Toxicology, and the Reproductiveand Developmental Toxicology Specialty Section of the Society of Toxicology (RDTSS), and heis the current editor of the RDTSS Newsletter. Dr. Hood was also a charter member of the Societyfor the Study of Reproduction and of the Neurobehavioral Teratology Society. He has been particularlyactive in the Teratology Society, where he has chaired the society’s Membership, Education,and Constitution/Bylaws committees, was a member of the society’s Ad Hoc Committees on Ethics,Warkany Lecturer Selection, Expert Testimony, and Web Site. Dr. Hood has served as a memberof the editorial boards of Fundamental and Applied Toxicology, Toxicological Sciences, andInScight, and he is a current member of the EPA’s Food Quality Protection Act Science ReviewBoard (FQPA/SRB).At the University of Alabama, Dr. Hood taught courses on teratology, developmental toxicology,both general and environmental toxicology, developmental biology, human embryology, reproductivephysiology, endocrinology, and general physiology.Dr. Hood’s current research interests include investigation of mechanisms and development ofassays for developmental toxicity. His recent research has involved assessment of the ability ofindole-3-carbinol to protect against developmental toxicity, determination of whether the developmentaltoxicity of chromium picolinate is due to the picolinate alone, detection of toxicant-inducedDNA strand breaks in organogenesis stage mammalian embryos, influence of maternal diet andbiotransformation by methylation on arsenical-induced developmental toxicity, maternal stress andteratogen interactions, and mechanistic studies of mitochondrial poisons as teratogens. He has alsoinvestigated the teratogenic potential of environmental toxicants, such as mycotoxins and arsenicals;he has conducted research on teratogen-teratogen and gene-teratogen interactions, the use ofdeveloping Drosophila melanogaster as an “in vitro” screening system for developmental toxicants,and arsenic biotransformation in vivo and in vitro. Dr. Hood has participated in numerous workshopsand expert review panels on developmental toxicity and risk assessment. He has also authored oredited some 92 publications in print or in press (research articles, reviews, books, and bookchapters), including Handbook of Developmental Toxicology, the predecessor to the current volume,in addition to numerous unpublished reports.© 2006 by Taylor & Francis Group, LLC

Contributors ListBarbara D. AbbottU.S. Environmental Protection AgencyResearch Triangle Park, North CarolinaKathleen D. BarrowcloughDuPontNewark, DelawareMelissa J. BeckWIL Research LaboratoriesAshland, OhioChristopher J. BowmanWIL Research LaboratoriesAshland, OhioRobert L. BrentDuPont Hospital for ChildrenWilmington, DelawareRobert M. CabreraTexas A & M University System Health ScienceCenterHouston, TexasJames J. ChenU.S. Food and Drug AdministrationJefferson, ArkansasMildred S. ChristianArgus ResearchHorsham, PennsylvaniaThomas F.X. CollinsU.S. Food and Drug AdministrationLaurel, MarylandRalph L. CooperU.S. Environmental Protection AgencyResearch Triangle Park, North CarolinaJohn M. DeSessoMitretek SystemsFalls Church, VirginiaSusan Y. EulingU.S. Environmental Protection AgencyWashington, D.C.Elaine M. FaustmanUniversity of WashingtonSeattle, WashingtonLynda B. FawcettJefferson Medical CollegeWilmington, DelawareRichard H. FinnellTexas A & M University System Health ScienceCenterHouston, TexasJulie M. GohlkeUniversity of WashingtonSeattle, WashingtonJerome M. GoldmanU.S. Environmental Protection AgencyResearch Triangle Park, North CarolinaEarl Gray, Jr.U.S. Environmental Protection AgencyResearch Triangle Park, North CarolinaWilliam C. GriffithUniversity of WashingtonSeattle, WashingtonMarion F. GruberU.S. Food and Drug AdministrationRockville, MarylandBarbara F. HalesMcGill UniversityMontreal, Quebec, CanadaJason H. HansenEmory UniversityAtlanta, GeorgiaCraig HarrisThe University of MichiganAnn Arbor, MichiganGary HeldU.S. Environmental Protection AgencyResearch Triangle Park, North Carolina© 2006 by Taylor & Francis Group, LLC

Alan M. HobermanArgus ResearchHorsham, PennsylvaniaJoseph F. HolsonWIL Research LaboratoriesAshland, OhioRonald D. HoodThe University of AlabamaTuscaloosa, AlabamaBengt KällénTornblad InstituteLund, SwedenLewis E. KaufmanWIL Research LaboratoriesAshland, OhioCarole A. KimmelU.S. Environmental Protection AgencyWashington, D.C.Currently atKimmel and AssociatesSilver Spring, MarylandGary L. KimmelU.S. Environmental Protection AgencyWashington, D.C.Currently atKimmel and AssociatesSilver Spring, MarylandArun P. KulkarniUniversity of South FloridaTampa, FloridaSusan C. LawsU.S. Environmental Protection AgencyResearch Triangle Park, North CarolinaThomas A. LewandowksiUniversity of WashingtonSeattle, WashingtonElise M. LewisArgus ResearchHorsham, PennsylvaniaPia LindströmMaxim PharmaceuticalsSan Diego, CaliforniaMelissa C. MarrCenter for Life Sciences and ToxicologyResearch Triangle Park, North CarolinaDiane B. MillerCenters for Disease Control and PreventionMorgantown, West VirginiaDavid E. MorseU.S. Food and Drug AdministrationRockville, MarylandMark D. NemecWIL Research LaboratoriesAshland, OhioEric L. PadgettWIL Research LaboratoriesAshland, OhioRobert M. ParkerHoffmann-LaRoche Inc.Nutley, New JerseyRafael A. PonceUniversity of WashingtonSeattle, WashingtonKathleen L. ReedDuPontNewark, DelawareBernard RobaireMcGill UniversityMontreal, Quebec, CanadaMitchell B. RosenU.S. Environmental Protection AgencyResearch Triangle Park, North CarolinaMarguerite R. SeeleyGradient CorporationCambridge, MassachusettsMary E. ShacklefordU.S. Food and Drug AdministrationCollege Park, Maryland© 2006 by Taylor & Francis Group, LLC

Ofer SpiegelsteinTexas A & M University System Health ScienceCenterHouston, TexasRobert L. SprandoU.S. Food and Drug AdministrationLaurel, MarylandTammy E. StokerU.S. Environmental Protection AgencyResearch Triangle Park, North CarolinaDonald G. StumpWIL Research LaboratoriesAshland, OhioRochelle W. TylCenter for Life Sciences and ToxicologyResearch Triangle Park, North CarolinaBennett J. VarshoWIL Research LaboratoriesAshland, OhioStephen G. WhittakerUniversity of WashingtonSeattle, WashingtonPatrick J. WierGlaxoSmithKline PharmaceuticalsUpper Merion, PennsylvaniaDaniel T. WilsonWIL Research LaboratoriesAshland, OhioVickie WilsonU.S. Environmental Protection AgencyResearch Triangle Park, North Carolina© 2006 by Taylor & Francis Group, LLC

APPENDIX ATerminology of Anatomical DefectsGLOSSARY OF DEVELOPMENTAL DEFECTSRonald D. Hood and Kit Kellerablepharia reduction or absence of the eyelids, with continuous skin covering the eyesabrachia absence of the arms (forelimbs)acampsia rigidity or inflexibility of a joint [ankylosis]acardia absence of the heartacaudia, acaudate without a tail [anury]acephaly agenesis of the head [acephalia]acheiria congenital absence of one or both hands (forepaws)achondroplasia cartilage defect causing inadequate bone formation and resulting in short limbsand other defectsacorea absence of the pupil of the eyeacrania partial or complete absence of the craniumacystia absence of the urinary bladderadactyly absence of digitsagenesis lack of development of an organaglossia absence of the tongueagnathia absence of the lower jaw (mandible)agyria characterized by a small brain, lacking the normal convolutions of the cerebral cortex[lissencephaly]amastia absence of the mammae (breasts)amelia absence of a limb or limbs (see also ectromelia)ametria absence of the uterusanasarca generalized edemaanencephaly absence of the cranial vault, with the brain missing or greatly reducedanephrogenesis absence of kidney(s)aniridia absence of the irisanisomelia inequality between paired limbsankyloglossia partial or complete adhesion of the tongue to the floor of the mouth911© 2006 by Taylor & Francis Group, LLC

912 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONankylosis abnormal fixation of a joint; implies bone fusionanodontia absence of some or all of the teethanonchia absence of some or all of the nailsanophthalmia absent or vestigial eye(s)anorchism uni- or bilateral absence of the testes [anorchia]anostosis defective development of bone; failure to ossifyanotia absence of the external ear(s) (i.e., pinnae, auricles)anovarism absence of the ovaries [anovaria]anury see acaudiaaphakia absence of the eye lensaphalangia absence of a digit or of one or more phalangesaplasia see agenesisapodia absence of one or both of the feet (paws)aprosopia partial or complete absence of the facearachnodactyly abnormal length and slenderness of the digitsarrhinia absence of the nosearthrogryposis persistent flexure or contracture of a jointarthrogryposis multiplex congenita syndrome distinguished by congenital fixation of the jointsand muscle hypoplasiaasplenia absence of the spleenastomia absence of the opening of the mouthatelectasis incomplete expansion of a fetal lung or a portion of the lung at birthathelia absence of the nipple(s)athymism absence of the thymus gland [athymia]atresia congenital absence of a normally patent lumen or closure of a normal body openingatresia ani agenesis or closure of the anal opening [imperforate anus]atrial septal defect postnatal communication between the atriabifid tongue cleft tonguebipartite division of what is normally a single structure into two parts; usually refers to areas ofskeletal ossificationbrachydactyly abnormal shortness of digitsbrachygnathia abnormal shortness of the mandiblebrachyury abnormally short tailbuphthalmos enlargement and distension of the fibrous coats of the eye; congenital glaucoma[buphthalmia, buphthalmus]camptodactyly permanent flexion of one or more digits [camptodactylia]carpal flexure abnormal flexion of the fetal carpus (wrist); most often seen in the rabbit; maybe transient or permanentceloschisis congenital fissure of the abdominal wallcelosomia fissure or absence of the sternum, with visceral herniationcephalocele protrusion of part of the brain through the craniumcleft face clefting due to incomplete fusion of the embryonic facial primordiacleft lip cleft or defect in the upper lip [hare lip]cleft palate fissure or cleft in the bony palate [palatoschisis]clinodactyly permanent lateral or medial deviation of one or more fingersclub foot see talipescoarctation stricture or stenosis, usually of the aorta© 2006 by Taylor & Francis Group, LLC

TERMINOLOGY OF ANATOMICAL DEFECTS 913coloboma fissure or incomplete development of the eyeconceptus everything that develops from the fertilized egg, including the embryo or fetus andthe extraembryonic membranescraniorachischisis fissure of the skull and vertebral columncranioschisis congenital cranial fissurecraniosynostosis premature ossification of cranial suturescraniostenosis premature cessation of cranial growth because of craniosynostosiscryptorchidism failure of one or both testes to enter the scrotum [cryptorchism]cyclopia fusion of the orbits into a single orbit, with the nose absent or present as a tubularappendage (proboscis) above the orbitdelayed ossification incomplete mineralization of an otherwise normal ossification centerdextrocardia abnormal displacement of the heart to the rightdiaphragmatic hernia protrusion of abdominal viscera through a defect in the diaphragmdicephalus conjoined twins with two heads and one bodydiplomyelia complete or incomplete doubling of the spinal cord due to a longitudinal fissurediplopagus symmetrically duplicated conjoined twins with largely complete bodies that mayshare some internal organsdiprosopus fetus with partial duplication of the facedysarthrosis malformation of a jointdysgenesis defective developmentdysplasia abnormal tissue developmentdysraphism failure of fusion, especially of the neural folds [dysraphia]dystocia prolonged, abnormal, or difficult deliveryectopia displacement or malposition of an organ or body partectopia cordis displacement of the heart outside of the thoracic cavityectopic kidney abnormal position of one or both kidneysectopic pregnancy pregnancy located outside the uterine cavityectrodactyly absence of all or only part of one or more digitsectromelia hypoplasia or absence of one or more limbsectropion abnormal eversion of the margin of the eyelidencephalocele herniation of part of the brain, encased in meninges, through an opening in theskull [encephalomeningocele, meningoencephalocele]endocardial cushion defect heart defects resulting from incomplete fusion of the embryonicendocardial cushionsentropion abnormal inversion of the margin of the eyelidepispadias absence of the upper wall of the urethra; more common in males, where the urethraopens on the dorsal surface of the penisethmocephalus form of cyclopia in which the eyes are closely set and the nose is hypoplasticand may be displaced upward [ethmocephaly]eventration protrusion of bowels through the abdominal wallexencephaly absence of part or the entire cranium, leaving the brain exposedexomphalos see omphalocele [umbilical hernia]exophthalmos abnormal protrusion of the eyeball [exophthalmus]exostosis abnormal bony growth projecting outward from the surface of a boneexstrophy congenital eversion of a hollow organ, e.g., of the bladderfetal wastage postimplantation death of embryo or fetusfetus developing mammal from the completion of major organogenesis to birth© 2006 by Taylor & Francis Group, LLC

914 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONgastroschisis fissure of the abdominal wall, not involving the umbilicus and usually involvingprotrusion of visceragonadal dysgenesis general term for various abnormalities of gonadal development,e.g., gonadal aplasia, hermaphroditism, “streak gonads”hamartoma benign nodular or tumorlike mass resulting from faulty embryonal development ofcells and tissues natural to the parthemimelia absence of all or part of the distal half of a limbhemivertebra incomplete development of one side of a vertebrahermaphrodite individual with both male and female gonadal tissue (see alsopseudohermaphrodite)heterotopia development of a normal tissue in an abnormal locationholocardius grossly defective monozygotic twin whose circulation is dependent upon the heartof a more perfect twinholoprosencephaly failure of division of the prosencephalon, resulting in a deficit in midlinefacial development, with hypotelorism; cyclopia occurs in the severe formhydramnion, hydramnios see polyhydramnioshydranencephaly complete or almost complete absence of cerebral hemispheres, which havebeen replaced by cerebrospinal fluidhydrocele collection of fluid in the tunica vaginalis of the testis or along the spermatic cordhydrocephalus marked dilation of the cerebral ventricles with excessive fluid, usuallyaccompanied by a vaulted or dome-shaped head [hydrocephaly, hydrencephaly,hydrencephalus]hydronephrosis distension of the pelvis and calyces of the kidney with fluid, as a result ofobstruction of urinary outflowhydroureter distension of the ureter with urine because of obstruction of the ureterhypermastia presence of one or more supernumerary mammary glands [polymastia]hypertelorism abnormally great distance between paired parts or organs, e.g., between the eyeshypospadias abnormal opening of the urethra on the underside of the penis, on the perineum inmales, or into the vagina in femalesichthyosis developmental skin disorders characterized by excessive or abnormal keratinization,with dryness and scalinginiencephaly anomaly of the brain and neck characterized by an occipital bone defect, spinabifida of the cervical vertebrae, and fixed retroflexion of the head on the cervical spinekyphosis abnormal dorsal convexity in the curvature of the thoracic spine (humpback, hunchback)lissencephaly see agyrialordosis abnormal anterior convexity in the curvature of the spine (swayback)macroglossia abnormally large tongue, often protrudingmacrophthalmia abnormally large eye(s)macrosomia abnormally large bodymacrostomia abnormally wide mouthmalformation permanent structural deviation that generally is incompatible with or severelydetrimental to normal postnatal development or survivalmeningocele herniation of the meninges through a defect in the cranium [cranial m.] or spinalcolumn [spinal m.]meningoencephalocele herniation of meninges and brain tissue through a defect in the cranium[encephalocele, encephalomeningocele]© 2006 by Taylor & Francis Group, LLC

TERMINOLOGY OF ANATOMICAL DEFECTS 915meningomyelocele herniation of meninges and spinal cord through a defect in the spinal column[myelomeningocele]meromelia absence of part of a limbmicrencephaly abnormally small brainmicrocephaly abnormally small head [microcephalia; microcephalus]microglossia tongue hypoplasiamicrognathia abnormally small jaw (usually the mandible)micromelia having abnormally small or short limb(s)microphthalmos abnormal smallness of one or both eyes [microphthalmia]microstomia hypoplasia of the mouthmicrotia hypoplasia of the pinna, with an absent or atretic external auditory meatusmisaligned sternebrae when the two ossification centers of each sternebra are not aligned in thetransverse planemyelocele herniation of spinal cord through a vertebral defectmyelomeningocele (see meningomyelocele)myeloschisis cleft spinal cord caused by failure of neural tube closureneural tube defect any malformation (e.g., anencephaly, exencephaly, spina bifida) resultingfrom failure of closure of the neural tube during embryonic developmentnevus circumscribed skin malformation, usually hyperpigmented or with abnormalvascularizationoligodactyly having fewer than the normal number of digitsoligohydramnios abnormally reduced quantity of amniotic fluidomphalocele herniation of intestine covered with peritoneum and amnion through a defect in theabdominal wall at the umbilicus [exomphalos, umbilical hernia]otocephaly extreme underdevelopment of the mandible, allowing close approximation or unionof the ears on the anterior aspect of the neckoverriding aorta displacement of the aorta to the right, so that it appears to arise from bothventriclespagus combining form (suffix) indicating conjoined twinspalatine rugae alteration misaligned or otherwise abnormal palatal ridgespatent ductus arteriosus an open communication between the pulmonary trunk and aorta,persisting postnatallypatent foramen ovale failure of adequate postnatal closure of the foramen ovale, the atrial septaldefect allowing interchange of blood between the atriaphocomelia absence of the proximal part of a limb(s), the distal part attached to the trunk by asmall, irregularly shaped boneplagiocephaly asymmetrical condition of the cranium resulting from premature closure of thecranial sutures on one sidepolydactyly supernumerary digitspolyhydramnios abnormally increased quantity of amniotic fluidpolymastia see hypermastiaproboscis cylindrical protuberance of the facepseudohermaphroditism partial masculinization or partial feminization, with gonadal tissue ofonly one sex presentptosis drooping of the upper eyelid because of abnormal muscle or nerve developmentrachischisis congenital fissure of the vertebral column© 2006 by Taylor & Francis Group, LLC

916 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONresorption conceptus that died after implantation and is being, or has been, reabsorbed into themother’s bloodstreamrhinocephaly possessing a proboscis-like nose above partially or completely fused eyes[rhinocephalus]runt normally developed fetus or newborn significantly smaller than the remainder of a litterscoliosis lateral deviation of the vertebral columnsirenomelia any of several degrees of side-to-side fusion of the lower extremities and concomitantmidline reduction of the pelvis; soft tissues and long bones, lower paws (feet), and visceraof the pelvis tend to be reduced or absent; anus and external genitalia are often absent[symmelia]situs inversus lateral transposition of the visceraspina bifida localized defective closure of the vertebral arches, through which the spinal cordand/or meninges may protrudespina bifida aperta spina bifida in which the neural tissue is exposedspina bifida cystica spina bifida with herniation of a cystic swelling containing the meninges(meningocele), spinal cord (myelocele), or both (myelomeningocele)spina bifida occulta spina bifida with intact skin and no herniationstillbirth birth of a dead fetussympodia fusion of the feetsyndactyly webbing or fusion between adjacent digitstalipes congenital deformity of the foot, which is twisted out of shape or position [clubfoot]talipes equinovalgus talipes in which the heel is elevated and turned outward from the midlinetalipes equinovarus talipes in which the foot is plantarflexed and turned inward from the midline;the most common talipesteratogen agent that can cause abnormal development of the embryo or fetusteratogenicity ability to cause defective development of the embryo or fetustetralogy of Fallot combination of cardiac defects consisting of pulmonary or infundibularstenosis, interventricular septal defect, overriding aorta, and right ventricular hypertrophytracheoesophageal fistula abnormal connection between trachea and esophagusumbilical hernia see omphaloceleventricular septal defect persistent communication between the ventricleswavy ribs extra bends in one or more ribs[Adapted in part, with modifications, from the following sources: Anderson, D. M., Keith, J., Novak, P. D., and Elliott,M.A., Dorland’s Illustrated Medical Dictionary, 28th ed., W. B. Saunders, Philadelphia, 1983; MARTA Committee, Glossaryof fetal anomalies, in Handbook of Developmental Toxicology, R. Hood, Ed., CRC Press, Boca Raton, 1997, 697 pp.; Keller,K. unpublished glossary; Moore, K.L. and Persaud, T. V. N., The Developing Human, 6th ed., W. B. Saunders, Philadelphia,1998; O’Rahilly, R. and Müller, F., Human Embryology & Teratology, 2nd ed., Wiley-Liss, New York, 1996; Schardein,J.L., unpublished glossary (1995); Spraycar, M., Ed., Stedman’s Medical Dictionary, 26th ed. Williams & Wilkins, Baltimore,1995; U.S. EPA, Health Effects Division, Standard Evaluation Procedure, Developmental Toxicity Studies. Health EffectsDivision, Office of Pesticide Programs, Washington, DC, 1993; Wise, L.D., Beck, S.L., Beltrame, D., Beyer, B.K., et al.Terminology of developmental abnormalities in common laboratory mammals (version 1), Teratology 55, 249, 1997.]© 2006 by Taylor & Francis Group, LLC

TERMINOLOGY OF ANATOMICAL DEFECTS 917TERMINOLOGY OF DEVELOPMENTAL ABNORMALITIESIN COMMON LABORATORY MAMMALS*L. David Wise, 1,19,20,21,* Sidney L. Beck, 2,22 Diana Beltrame, 3,19,† Bruce K. Beyer, 4,22 IbrahimChahoud, 5,19 Robert L. Clark, 6,20 Ruth Clark, 7,19,†† Alice M. Druga, 8,19 Maureen H. Feuston, 9,20Pierre Guittin, 10,19,††† Susan M. Henwood, 11,22 Carole A. Kimmel, 12,19 Pia Lindström 13, 21, 23 ,Anthony K. Palmer 14, 19 , Judith A. Petrere 15, 22 , Howard M. Solomon, 16,20,21 Mineo Yasuda, 17,19and Raymond G. York 18,201Merck Research Laboratories, Safety Assessment, West Point, Pennsylvania, 194862DePaul University, Department Biological Sciences, Chicago, Illinois, 606143Pharmacia & Upjohn, Worldwide Toxicology, Milan, Italy4Bristol-Myers Squibb, Pharmaceutical Research Institute, Evansville, Indiana, 477215Freie Universitat, Institut fur Toxikologie and Embryo Pharmakologie, 14195 Berlin 33, Germany6Rhône-Poulenc Rorer Research and Development, Department of Toxicology, Collegeville, Pennsylvania, 19426-09947Ruth Clark Associates Ltd., North Lincolnshire, DN17 3JB, United Kingdom8Institute for Drug Research Ltd., Division of Safety Studies Budapest, Hungary, H-13259Sanofi Research, Division of Sanofi Pharmaceuticals, Inc., Malvern, Pennsylvania, 1935510Rhône-Poulenc Rorer, Drug Safety, General and Reproductive Toxicology Department, 94403, Vitry-sur-Seine, France11Covance Laboratories Inc., Madison, Wisconsin, 5370712U.S. Environmental Protection Agency, NCEA, Washingtom, DC, 20460; currently on detail to the Food and DrugAdministration, NCTR, Rockville, Maryland, 2085713Quintiles, Inc., Regulatory Affairs, Research Triangle Park, North Carolina, 2770914Huntingdon, PE18 7XJ, United Kingdom15Warner-Lambert/Parke-Davis Pharmaceutical Research Division, Pathology/Experimental Toxicology Department,Ann Arbor, Michigan, 4810516SmithKline Beecham Pharmaceuticals, Research & Development Division, King of Prussia, Pennsylvania, 19406-093917Hiroshima University School of Medicine, 734, Hiroshima, Japan18Argus Research Laboratories, Horsham, Pennsylvania, 1904419International Federation of Teratology Societies Committee on International Harmonization of Nomenclature inDevelopmental Toxicology20Middle Atlantic Reproduction and Teratology Association Nomenclature Committee21Teratology Society Terminology Committee22Midwest Teratology Association Nomenclature Committee23Food and Drug Administration Standardized Nomenclature Program† Representing the Italian Nomenclature Working Group. Other members include Dr. Rita Bussi, RBM, Ivrea; Prof.Erminio Giavini, University of Milan; and Dr. Alberto Mantovani, Instituto Superiore di Sanità, Rome.†† Representing the UK Foetal Pathology Terminology Group. Other members include Mr. Robert Bramley, ZenecaPharmaceuticals, Macclesfield; Mrs. Elizabeth J. Davidson, Medeva Group Development, Surrey; Mr. Keith P.Hazelden, Inveresk Research Ltd., Edinburgh (currently Huntingdon Life Sciences, Eye); Mr. David M. John, HuntingdonLife Sciences, Eye; Mrs. Mary Moxon, Zeneca Central Toxicology Laboratory, Macclesfield; Ms. Meg M.Parkinson, Glaxo Wellcome Research and Development, Ware; and Mrs. Sheila A. Tesh, Tesh Consultants International,Saxmundham.††† Representing the French Teratology Association’s Nomenclature Working Group. Other members include Dr. PaulBarrow, Chrysalis International, L’Arbresle; Dr. Catherine Boutemy, Synthelabo Recherche, Gargenville; Dr. DidierHiss, Ruffey les Echirey; Dr. Andre Morin, Faculte de Medecine, Lyon; Dr. Charles Roux, CHU St-Antoine, Paris;and Dr. Jeanne Stadler, Pfizer, Amboise.* Correspondence to: L. David Wise, Merck Research Laboratories, W45-1, West Point, PA 19486* This article was previously published as: Wise, L.D. et al., Terminology of Developmental Abnormalities in CommonLaboratory Mammals (Version 1). Teratology 55:249-292, 1997. With permission.© 2006 by Taylor & Francis Group, LLC

918 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONABSTRACTThis paper presents the first version of an internationally-developed glossary of terms for structuraldevelopmental abnormalities in common laboratory animals. The glossary is put forward by theInternational Federation of Teratology Societies (IFTS) Committee on International Harmonizationof Nomenclature in Developmental Toxicology, and represents considerable progress toward harmonizationof terminology in this area. The purpose of this effort is to provide a common vocabularythat will reduce confusion and ambiguity in the description of developmental effects, particularlyin submissions to regulatory agencies world-wide. The glossary contains a primary term or phrase,a definition of the abnormality, and notes, where appropriate. Selected synonyms or related terms,which reflect a similar or closely related concept, are noted. Nonpreferred terms are indicated wheretheir usage may be incorrect. Modifying terms used repeatedly in the glossary (e.g., absent,branched) are listed and defined separately, instead of repeating their definitions for each observation.Syndrome names are generally excluded from the glossary, but are listed separately in anappendix. The glossary is organized into broad sections for external, visceral, and skeletal observations,then subdivided into regions, structures, or organs in a general overall head to tail sequence.Numbering is sequential, and not in any regional or hierarchical order. Uses and misuses of theglossary are discussed. Comments, questions, suggestions, and additions from practitioners in thefield of developmental toxicology are welcomed on the organization of the glossary as well on asthe specific terms and definitions. Updates of the glossary are planned based on the commentsreceived.INTRODUCTIONNomenclature used to describe observations of fetal and neonatal morphology often varies considerablyamong laboratories, investigators, and textbooks in the fields of teratology and developmentaltoxicology. Standard medical, veterinary, and anatomical dictionaries often differ in the definitionsof commonly used terms. This lack of a common vocabulary leads to confusion and uncertaintywhen communicating scientific findings, and can be a particular problem in submissions to regulatoryagencies worldwide. Thus, there is a recognized need for a common nomenclature to usewhen describing such observations.Over the years numerous efforts have been undertaken to provide a common nomenclature.Various groups in different countries have made attempts to provide solutions, and internationalorganizations, such as the International Programme on Chemical Safety (IPCS) and the Organizationfor Economic Cooperation and Development (OECD) have shown interest in promoting internationalharmonization. For various reasons, previous efforts have not been adopted on a widespreadbasis. In an attempt to provide an internationally acceptable common vocabulary, the InternationalFederation of Teratology Societies (IFTS*) in May 1995 appointed a Committee on InternationalHarmonization of Nomenclature in Developmental Toxicology. The first priority for this committeewas to collect and review existing glossaries and ongoing efforts in several countries aimed atdeveloping atlases of fetal abnormalities, and to attempt to develop a glossary and atlas that reflectedinternational harmonization. Several groups, notably in France, Germany, Japan, and the UnitedKingdom (UK) have begun developing atlases of normal and/or abnormal fetal morphology eitheras books, CD-ROM, or both. Although these image-based atlases promise to provide the best meansof portraying specific observations with photographs and drawings, they are time-consuming toprepare. The IFTS committee soon decided to concentrate first on harmonization of the terminology.In this way the image-based products could incorporate this harmonized terminology, avoid confusionfrom one atlas to another, and be used in further refinement of the terminology.* The IFTS is composed of members of the teratology societies of Australia, Europe, Japan, and North America© 2006 by Taylor & Francis Group, LLC

TERMINOLOGY OF ANATOMICAL DEFECTS 919The most convenient material available for initiating the process at the time the IFTS committeebegan its work was the glossary previously developed and distributed in the U.S.A. by the MiddleAtlantic Reproduction and Teratology Association (MARTA). A revision of this glossary wasunderway, in part, due to a request by the U.S. Food and Drug Administration for assistance intheir project to standardize submissions of data on new drugs. The effort to revise the MARTAglossary was initiated with the involvement of the Midwest Teratology Association (MTA). Aftera review by the IFTS Nomenclature Committee, the three organizations agreed to work jointly ondeveloping an international glossary. The glossary was revised in response to hundreds of commentsreceived from the initial request as well as comments from the IFTS committee and associatednomenclature committees in the UK, France, and Italy. The document presented herein representsconsiderable progress towards harmonization. Although not the product of a more formal internationalharmonization process such as ICH guideline development, this document is the first versionof an international glossary that will continue to evolve over the coming years with input fromworldwide IFTS members as well as other users of the glossary.We invite and would welcome questions, comments, additions, and suggestions of either a specificor general nature. Based on these comments and proposed modifications, an update of the glossarywill be initiated by the nomenclature committees. Specific or detailed comments and proposedchanges to the glossary may be directed to any author listed above, or, if not convenient, to Dr. L.David Wise (Merck Research Laboratories, W45-1, West Point, PA 19446; phone 215-652-6974;fax 215-652-3423; e-mail david_wise@merck.com). Comments or information regarding the overallIFTS initiative may be addressed to Dr. Carole A. Kimmel (U.S. Food and Drug Administration,NCTR [HFT-10] Rm. 16B-06, 5600 Fishers Lane, Rockville, MD 20857; phone 301-827-3403;fax 301-443-3019; e-mail ckimmel@nctr.fda.gov).Organization of the GlossaryThe terms included in the glossary describe morphological changes observed grossly or with theaid of a dissecting microscope in the most common laboratory animals used for developmentalstudies (mainly rats, mice, and rabbits). The term ‘abnormalities’ is used here as a collective termto mean differences from normal in the specimen under examination relative to the perceived normof control specimens for a particular species. It is perhaps not an ideal term since normality isrelative and may change over time, but is considered an acceptable term for this list.Terms are included which describe observations in fetal as well as neonatal animals. Thus, forexample, hypospadias and ocular coloboma, which are not readily detectable at birth in theseanimals, but are observable at some later time, are included in the glossary. On the other hand,cryptorchidism is not included as an observation, since the applicable event (descent of the testesinto the scrotum) does not occur until after weaning in these species. We sought to describeobservations of abnormalities with mainly descriptive terms and to avoid or limit terms that weremore diagnostic or perhaps implied mechanisms. Some observations may require follow-up studiesor examinations to determine more detail.The glossary is organized into broad sections for external, visceral, and skeletal observations,which reflect how the data are typically collected in developmental toxicity studies. The observationsare subdivided into regions, structures, or organs in a general overall head to tail sequence; theobservations within each are listed in alphabetical order. Some external observations are betterdefined with a subsequent visceral and/or skeletal examination, and these terms should then beassociated to reflect a single observation. The notes for external observations sometimes indicatewhere this is the case.Each term is sequentially numbered in this version in order to facilitate tracking of futurerevisions. These code numbers are not intended to provide any regional or hierarchical information.© 2006 by Taylor & Francis Group, LLC

920 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe first digit of all terms in this version is “1”, and corresponds to the present version number.Observation entries that are revised in subsequent versions will be indicated by changing the firstdigit to correspond to the revision number of the document. Totally new observations added insubsequent versions will have a first digit corresponding to that version number, and will benumbered sequentially from the preceding version. Thus, the first new term added in version 2 willbe numbered 20869.An observation is represented by a primary term or phrase, and definition of the observationis included along with notes, where appropriate. Among other uses, the notes indicate when analteration is part of a common syndrome. Selected synonyms or related terms, which reflect asimilar or closely related concept, are then noted. Nonpreferred terms are indicated where theirusage may be incorrect. In some cases, the nonpreferred term may be synonymous with theobservation but it is not recommended for general use since other terms are more descriptive ormore widely accepted.Certain modifying terms used repeatedly in the glossary (e.g., absent, branched) are listed anddefined separately in Appendix I, instead of repeating their definitions for each observation. Inaddition, syndrome names are generally excluded from the glossary even if the name of thesyndrome is common. For example, there is no entry for ectrosyndactyly, since this can be describedby using the separate entries of ectrodactyly and syndactyly. Common syndromes omitted fromthe glossary are listed in a separate section (Appendix II). Various options for the organization ofthe glossary were considered so comments from users would be very welcomed.The list of terms is formatted so that it can be handled electronically. One of the designatedobjectives of the list was to facilitate transfer of data within and between computer databases. Also,the IFTS intends to make the glossary with an associated atlas of images accessible on the internetfor easy access by the wider scientific community.Uses and Misuses of the GlossaryDuring the course of creating and revising this glossary a number of issues arose regarding itspurpose and uses.1. Not all observations in this relatively extensive list are necessarily intended to be used in thesummarization of data for the purpose of reporting and comparing group incidences. Initially, thelist of terms may be used so that separate but coincident abnormalities are enumerated as separatefindings. Then, multiple coincidental abnormalities can be described using a single combiningterm. As a simple example, domed head may be described externally, but upon visceral examination,if hydrocephaly is found, only this term is used on a summary table. Another example would bewhere multiple fetuses have different digits that are variably short. Instead of enumerating thevalues for absent and short phalanges, the summary table could use brachydactyly as the generalcombining term.2. To reiterate, the terms are to be used to describe morphological observations outside the normalrange. Thus, for example, it is intended that the use of the term “small” is understood to describea structure that is morphologically similar to the norm but small compared to control specimensof the same general age.3. A ranking or classification of terms into categories, such as malformation and variation, does notappear in the glossary since a given observation may be a malformation in one species but avariation in another species, or the classification may change depending on the gestational day ofexamination. In addition, there is no consensus at present as to which classification scheme is mostrelevant. Classification of each observation into a certain category is a common practice, and mustbe left as an option to the user.4. It is beyond the scope of this glossary to specify the normal number of skeletal elements for eachskeletal region as separate observations; although, for general reference, these numbers have beenlisted in the notes for various vertebral regions. This omission does NOT negate the common© 2006 by Taylor & Francis Group, LLC

TERMINOLOGY OF ANATOMICAL DEFECTS 921practice of reporting number of ossified elements in specific regions (e.g., sacral and/or caudalvertebrae), for this approach does offer certain advantages.5. In general, observations are represented by simple descriptive terms or phrases, except in caseswhere a common, generally accepted medical term is available.6. Terminology in the glossary typically reflects the basic observation. For example, open eye includesobservations of partially open eyelid and completely open eyelid, and cervical rib includes unilateraland bilateral. Modifiers, such as those relating to symmetry (e.g., right, left, medial, unilateral,bilateral) and degree (e.g., slight, moderate, marked), are not included in the glossary, except wheresidedness is an integral part of the specific observation (e.g., right-sided aortic arch). Some usersmay wish to further define abnormalities using these modifiers (e.g., “slightly dilated ureter” and“markedly dilated ureter”).7. Analysis of data on treatment-related effects may be more useful when glossary terms are combined.For example, many of the individual observations could be combined into general categories(e.g., tail abnormalities or thoracic vertebral abnormalities). Alternatively, some terms may besubdivided into relevant categories based on modifying terms (e.g., median versus unilateral cleftlip). Thus, tables used to present the individual or summary results of the examinations of fetusesor offspring may use the individual terms, names for groups of terms, or subsets of the terms.Finally, this glossary represents the start of a process that is working toward harmonization ofterminology among scientists from many countries. The intent is for it to be a basic support fordescribing developmental abnormalities. However, scientific judgment is required to use the observationsappropriately to allow an accurate interpretation of the data.The following dictionary was used as the primary reference for the glossary when inconsistencieswere found in other reference sources: International Dictionary of Medicine and Biology (3volumes), 1986, Wiley Medical Publications, John Wiley & Sons, NY (ISBN 0-471-01849-X).Terms covering the field of teratology within this dictionary were authored by the late Dr. JamesG. Wilson, an internationally recognized expert in the field of teratology. Modifications to definitionsmay have been made in some cases in order to reach an agreed upon, workable definition.© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Extrernal abnormalities 1Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition NoteGeneral 10001 Anasarca Generalized edema Generalized edema10002 Conjoined twins Double monster Monozygotic twins with variableincomplete separation into two duringcleavage or early stages ofembryogenesis10003 Cutis aplasia Localized region of no skin development10004 Hemorrhage Petechia, Purpura,An accumulation of extravasated bloodEcchymosis, Hematoma10005 Local edema Localized accumulation of fluid10006 Oligohydramnios Less than normal amount of amnioticfluid10007 Polyhydramnios Excessive amount of amniotic fluidCranium 10008 Acephalostomia Acrania Absence of head but with the presenceof mouth-like orifice in the neck region10009 Acephaly Absence of the head10010 Anencephaly Acrania Absence of the cranial vault, with thebrain missing or reduced to smallmass(es)Site and extent of fusionmay be describedAcrania is a skeletal term10011 Cranioschisis Fissure of the cranium [Craniorachischisis]10012 Domed head Cranium appears more elevated androunded than normalMay or may not beassociated withhydrocephaly10013 Exencephaly Acrania The brain protrudes outside the skull dueto absence of the cranial vault10014 Iniencephaly Exposure of occipital brain and upperspinal cord tissue. Involves extremeretroflection of the head.10015 Macrocephaly Disproportionate largeness of head10016 Meningocele Herniation of meninges through defect inskull10017 Meningoencephalocele Encephalomeningocele Cephalocele,Craniocele,Encephalocele10018 Microcephaly Leptocephaly,NanocephalyHerniation of brain and meningesthrough a cranial openingSmall craniumErosion of brain tissue hasnot occurred as inAnencephalyMay or may not be coveredby skin922 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermEar 10019 Anotia Agenesis, Aplastic Absence of pinna10020 Macrotia Large pinna10021 Malpositioned pinna Displaced, Ectopic, Lowset10022 Microtia Ear tab Small pinna10023 Misshapen pinna Abnormally shaped,Irregularly shaped10024 Synotia Fusion or abnormal approximation of [Otocephaly]pinnae below the faceEye 10025 Absent eye bulge May indicate micro- oranophthalmia10026 Cryptophthalmia Cryptophthalmos Skin continuous over eye(s) withoutformation of eyelid(s)Usually associated withmicrophthalmia10027 Cyclopia Monophthalmia,SynophthalmiaNon-preferredTerm Definition NoteSingle orbit; eye(s) can be absent,completely or incompletely fused10028 Enlarged eye bulge10029 Exophthalmos Proptosis Pop-eye Excessive protrusion of the eyeball10030 Malpositioned eye Displaced, Ectopic10031 Microblepharia Short vertical dimension of eyelid10032 Ocular coloboma A family of eye defects generallyinvolving a cleft or fissure of an eyestructure10033 Open eye Ablepharia Partial or complete deficiency of theeyelidNose may be absent orappear as a frontonasalappendage (proboscis)above the orbitNot readily apparent infetuses or soon after birthEyeball usually visible10034 Palpebral coloboma A notch or fissure of the eyelid10035 Small eye bulge Reduced May indicatemicrophthalmiaFace 10036 Cleft face Prosoposchisis Fissure of the face and jaw [Cheilognathoschisis,Cheilognathopalatoschisis]10037 Short snout RudimentaryNose 10038 Arhinia Agenesis, Aplastic Absence of the nose10039 Malpositioned naris Displaced, Ectopic10040 Malpositioned nose Displaced, Ectopic10041 Misshapen nose Abnormally shaped,Irregularly shaped10042 Naris atresia Imperforate10043 Proboscis Tubular projection replaces the nose [Rhinocephaly]10044 Single naris One naris instead of two naresTERMINOLOGY OF ANATOMICAL DEFECTS 923© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Extrernal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10045 Small nose Hypoplastic, Reduced,RudimentaryMouth/Jaw 10046 Aglossia Agenesis, Aplastic Absence of the tongue10047 Agnathia Agenesis, Aplastic Absence of the lower jaw (mandible)10048 Ankyloglossia Shortness or absence of the frenum ofthe tongue; tongue fused to the floor ofthe mouth10049 Anodontia Edentia Partial or complete absence of one ormore teeth10050 Astomia Agenesis, Aplastic Absence of the mouth10051 Cleft lip Cheiloschisis Harelip Fissure of the upper lip10052 Cleft palate Palatoschisis,Fissure of the palateUranoschisis10053 High-arched palate A higher than normal palatal arch10054 Macroglossia Large tongue May be protruding10055 Macrostomia Abnormal elongation of the mouth10056 Mandibular cleft10057 MandibularmacrognathiaPrognathism Exognathia Enlarged or protruding lower jaw(mandible)10058 MandibularBrachygnathia,Small lower jaw (mandible)micrognathiaMicromandible10059 Maxillary macrognathia Prognathism Exognathia Enlarged or protruding upper jaw(premaxilla/maxilla)10060 Maxillary micrognathia Micromaxilla Small upper jaw (premaxilla/maxilla)10061 Microglossia Small tongue10062 Microstomia Small mouth10063 Misaligned palate Asymmetricrugae10064 Misshapen palaterugaeAbnormally shaped,Irregularly shaped10065 Protruding tongueLimb 10066 Amelia Agenesis, Aplastic,Complete absence of one or more limbs Fleshy tab may be presentEctromelia10067 Bowed limb Usually seen as an outwardbending of the limb924 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated Term10068 Hemimelia Ectromelia Absence or shortening of the distal twosegments of limbMay be further defined asbeing fibular, radial, tibial,or ulnar10069 Limb hyperextension Arthrogryposis The excessive extension or straighteningof a limb or a joint10070 Limb hyperflexion Flexed The excessive flexion or bending of alimb or a joint10071 Macromelia An excessive size of one or more limbs10072 Malrotated limb Limb turned toward the center (i.e.,inward rotation) or the periphery (i.e.,outward rotation)10073 Micromelia Nanomelia Acromelia Disproportionately short or small limb(s)10074 Phocomelia Ectromelia Reduction or absence of proximalportion of limb, with the paws beingattached to the trunk of the body10075 Sirenomelia Symelia Fusion of lower limbs May involve lower torsoPaw/Digit 10076 Absent claw Agenesis, Aplastic Refers to distal-most tip,nail10077 Adactyly Agenesis, Aplastic Absence of all digits10078 Apodia Acheiria, Acheiropodia Absence of one or more paws10079 Brachydactyly Microdactyly Shortened digit(s) Expected skeletalalterations includeabsence or shortening ofphalanx(ges)10080 Ectrodactyly Agenesis, Aplastic,OligodactylyNon-preferredTerm Definition NoteAbsence of one or more digit(s)Expected skeletalalterations includeabsence of all phalangesin each affected digit10081 Enlarged digit Dactylomegaly,Macrodactyly10082 Malpositioned claw Displaced, Ectopic Refers to distal-most tip,nail10083 Malpositioned digit Clinodactyly,Camptodactyly,Displaced, EctopicDeflection of digit(s) from the central axis10084 Malrotated paw Clubbed paw, Talipes Paw turned toward the center (i.e.,inward) or the periphery (i.e., outward)10085 Misshapen digit Abnormally shaped,Irregularly shaped10086 Paw hyperextension Arthrogryposis The excessive extension or straighteningof a pawIncludes fixed flexiondeformity of digit(s)TERMINOLOGY OF ANATOMICAL DEFECTS 925© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Extrernal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservationSynonym orRelated Term10087 Paw hyperflexion The excessive flexion or bending of apaw10088 Polydactyly Supernumerary digit(s) Supernumerary digit(s)10089 Small claw Hypoplastic, Reduced,Rudimentary10090 Small paw Hypoplastic, Microcheiria,Smallness of paw(s)Reduced, Rudimentary10091 Syndactyly Ankylodactyly Partial or complete fusion of, or webbingbetween, digits10092 Sympodia Fusion of the hindlimb pawsTail 10093 Acaudate Agenesis, Anury, Aplastic Anurous Absence of tail10094 Bent tail Angulated, Bowed Shaped like an angle10095 Blunt-tipped tail Rounded or flat at the end, not tapered10096 Curled tail Curly Curved into nearly a full circle10097 Double-tipped tail Forked Duplication of the tail at the end10098 Fleshy tail tab10099 Hooked tail Approximately 180 degree bend or curveof the tail10100 Kinked tail Kinky Localized undulation(s) of the tail10101 Malpositioned tail10102 Narrowed tail Constricted Ring tail10103 Short tail Brachyury, Rudimentary10104 Thread-like tail Filamentous, FiliformTrunk 10105 Anal atresia Aproctia, Imperforate Absence or closure of the anal opening10106 Absent genital tubercle Agenesis, Aplastic10107 Decreased anogenital Reduced perineumShortened anogenital distance (AGD)distance10108 Ectopia cordis Heart displaced outside thoracic cavitydue to failure of ventral closure10109 Gastroschisis Laparoschisis,SchistoceliaNon-preferredTerm Definition NoteFissure of abdominal wall, not involvingthe umbilicus, and usuallyaccompanied by protrusion of viscerawhich may or may not be covered by amembranous sacRefers to distal-most tip,nailIncludes bony,cartilaginous, and/or softtissueMay be further defined asmedial (gastroschisis) orlateral fissure(laparoschisis)926 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservation10110 Holorachischisis Fissure of the entire spinal column [Craniorachischisis]10111 Hypospadias Urethra opening on the underside of thepenis or on the perineumNot readily apparent infetuses or soon after birth10112 Increased anogenitalIncreased anogenital distance (AGD)distance10113 Kyphosis Humpback,HunchbackIncreased convexity in the curvature ofthe thoracic spinal column as viewedfrom the side10114 Lordosis Hollowback,Swayback,SaddlebackAnterior concavity in the curvature of thelumbar and cervical spinal column asviewed from the side10115 Omphalocele Exomphalos Protrusion of intra-abdominal viscerainto the umbilical cord with visceracontained in a thin translucent sac ofperitoneum and amnion10116 Scoliosis Lateral curvature of the spinal column10117 Short trunk10118 Small anus10119 Small genital tubercle Hypoplastic, Reduced,Rudimentary10120 Spina bifida Meningocele,Meningomyelocele,Myelocele,Myelomeningocele,RachischisisA family of defects in the closure of thespinal columnThe condition may presentwith ruptured sacMay be covered with skin(spina bifida oculta); mayinvolve protrusion of spinalcord and/or meninges10121 Thoracogastroschisis Thoracoceloschisis Failure of closure leaving the thoracicand abdominal cavities, or major partsthereof, exposed ventrally10122 Thoracoschisis Fissure of thoracic wall Lung(s) may be herniated10123 Thoracostenosis Narrowness of the thoracic region10124 Umbilical hernia Protrusion of a skin-covered segment ofthe gastrointestinal tract and/or greateromentum through a defect in theabdominal wall at the umbilicus, theherniated mass being circumscribedand covered with skin; or protrusion ofskin-covered viscera through theumbilical ring with prominence of thenavel1For terms in brackets, see Appendix 1-B below.Synonym orRelated TermNon-preferredTerm Definition NoteTERMINOLOGY OF ANATOMICAL DEFECTS 927© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Visceral abnormalities 1Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition NoteGeneral 10125 Aneurysm Localized sac formed by distention of anartery or vein that is filled with blood10126 Fluid-filled abdomen Ascites, Hemorrhage Effusion and accumulation of watery fluid(ascites) or blood (hemorrhage)10127 Situs inversus Mirror-image transposition of theabdominal and/or thoracic visceraBrain 10128 Dilated cerebralventricleBulbousSlight to moderate dilatation of cerebralventricles, with no evidence of decreasein cortical thickness10129 Hemorrhagiccerebellum10130 Hemorrhagic cerebrum10131 Hydrocephaly Excessive cerebrospinal fluid within theskull10132 Misshapen cerebellum Abnormally shaped,Irregularly shaped10133 Misshapen cerebrum Abnormally shaped,Irregularly shaped10134 Small cerebellum Hypoplastic, Reduced,Rudimentary10135 Small cerebrum Hypoplastic, Reduced,RudimentaryEar 10136 Misshapen inner ear Abnormally shaped,Irregularly shapedEye 10137 Anophthalmia Agenesis, Aplastic Absence of eye(s)10138 Aphakia Agenesis, Aplastic Absence of lens10139 Cataract Opacity of the crystalline lens10140 Hemorrhagic eye10141 Macrophthalmia MegalophthalmiaLarge eye, eyeball(-mus,-mos)10142 Malpositioned eye Displaced, Ectopic10143 Microphthalmia Hypoplastic, Reduced,Small eyeRudimentary10144 Misshapen lens Abnormally shaped,Irregularly shapedMay be further defined bylocationIncludes absence ofinterhemispheric fissure928 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/Structure10145 Ocular coloboma A family of eye defects generallyinvolving a cleft or fissure of an eyestructure10146 Retina fold Undulation of retinal layers May be due to processingartifact10147 Small lens Hypoplastic, Reduced,RudimentaryNose 10148 Absent nasal conchae Agenesis, Aplastic10149 Absent nasal septum Agenesis, Aplastic10150 Enlarged nasal cavity Enlarged nasal cavity10151 Malpositioned nasal Displaced, Ectopicconchae10152 Malpositioned nasalseptumDisplaced, EctopicThyroidglandCodeNumberObservationSynonym orRelated Term10153 Small nasal cavity Reduced Reduced nasal cavity10154 Small nasal conchae Hypoplastic, Reduced,Rudimentary10155 Small nasal septum Hypoplastic, Reduced,Rudimentary10156 Absent thyroid gland Agenesis, Aplastic10157 Malpositioned thyroid Displaced, EctopicglandThymus 10158 Absent thymus Agenesis, Aplastic,Athymia, Athymism10159 Hemorrhagic thymus10160 Malpositioned thymus Displaced, Ectopic10161 Misshapen thymus Abnormally shaped,Irregularly shaped10162 Small thymus Hypoplastic, Reduced,Rudimentary, Trace,VestigialNon-preferredTerm Definition NoteReduced size, imperfect form, orremnant of thymusIncludes reduced numberof convolutions10163 Split thymus Bifid, Bipartite, Cleft Consisting of two separate parts10164 Supernumerary thymus Additional, ExtraHeart 10165 Absent aortic valve Agenesis, Aplastic10166 Absent cordaeAgenesis, Aplastictendinae10167 Absent left A-V valve Agenesis, Aplastic A-V = atrioventricular10168 Absent papillary Agenesis, Aplasticmuscle10169 Absent pulmonic valve Agenesis, AplasticTERMINOLOGY OF ANATOMICAL DEFECTS 929© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Visceral abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservation10170 Absent right A-V valve Agenesis, Aplastic A-V = atrioventricular10171 Atrial septum defect Postnatal communication between atria;includes defects of septa primum andsecundum; premature closure offoramen ovale apertum10172 Cardiomegaly Large heart10173 Cor biloculare Two-chambered heart with an atrium anda ventricle10174 Cor triloculare Three-chambered heart with two atriaand a ventricle or one atrium and twoventricles10175 Defect of A-V septum Inappropriate communication betweenatrium and ventricleA-V = atrioventricular10176 Dextrocardia Heart located on the right side of thethoracic cavityCan occur as isolatedalteration or inassociation with situsinversus10177 Enlarged aortic valve10178 Enlarged atrialchamber10179 Enlarged A-V ostium Enlargement of an atrioventricular orifice A-V = atrioventricular10180 Enlarged left A-V valve A-V = atrioventricular10181 Enlarged pulmonicvalve10182 Enlarged right A-VA-V = atrioventricularvalve10183 Enlarged ventricularchamber10184 Enlarged ventricular[Tetralogy of Fallot]wall10185 Hydropericardium Accumulation of fluid in the sac thatenvelopes the heart10186 Malpositioned heart Displaced, Ectopic Levocardia abnormal onlyin situs inversus10187 Membranousventricular septumdefectSynonym orRelated TermMembranous VSDNon-preferredTerm Definition NoteAn opening in the membranous septumbetween the ventricles[Tetralogy of Fallot]930 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated Term10188 Microcardia Hypoplastic, Reduced,Rudimentary10189 Misshapen aortic valve Abnormally shaped,Irregularly shaped10190 Misshapen left A-V Abnormally shaped,valveIrregularly shaped10191 Misshapen pulmonic Abnormally shaped,valveIrregularly shaped10192 Misshapen right A-V Abnormally shaped,valveIrregularly shapedNon-preferredTerm Definition NoteSmall heartUsed in absence ofstructural changesIncludes alterations innumber of cuspsA-V = atrioventricularIncludes alterations innumber of cuspsIncludes alterations innumber of cusps.A-V=atrioventricular[Tetralogy of Fallot]10193 Muscular ventricularseptum defectMuscular VSDAn opening in the muscular septumbetween the ventricles10194 Persistent A-V canal Defects of endocardial cushionsA-V = atrioventricularresulting in low atrial and highventricular septal defects10195 Small aortic valve Hypoplastic, Reduced,Rudimentary10196 Small left A-V valve Hypoplastic, Reduced,A-V = atrioventricularRudimentary10197 Small pulmonic valve Hypoplastic, Reduced,Rudimentary10198 Small right A-V valve Hypoplastic, Reduced,A-V = atrioventricularRudimentary10199 Small ventricular Hypoplastic, Reduced,chamberRudimentaryAorta 10200 Aortic atresia Imperforate Absence of patent communicationbetween the ascending aorta and theventricular mass which is usuallygrossly hypoplastic10201 Dilated aorta Bulbous10202 Double aorta10203 Malpositioned aorta Displaced, Ectopic10204 Narrowed aorta Coarctation, Constricted,Stenosis, Stricture10205 Overriding aorta Biventricular origin of aorta [Tetralogy of Fallot]Aortic arch 10206 Aortic arch atresia Imperforate10207 Dilated aortic arch Bulbous10208 Interrupted aortic arch Ascending aorta not connected todescending aorta10209 Narrowed aortic arch Coarctation, Constricted,Stenosis, StrictureTERMINOLOGY OF ANATOMICAL DEFECTS 931© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Visceral abnormalities 1 (continued)Region/Organ/StructureCarotidarteryDuctusarteriosusGreatvesselsInnominatearteryCodeNumberObservationSynonym orRelated Term10210 Retroesophageal aorticarch10211 Right-sided aortic arch10212 Absent carotid Agenesis, Aplastic10213 Dilated carotid Bulbous10214 Malpositioned carotid Displaced, EctopicbranchNon-preferredTerm Definition NoteIncludes right carotidoriginating from aorta(i.e., no innominate)10215 Narrowed carotid Coarctation, Constricted,Stenosis, Stricture10216 Retroesophagealcarotid10217 Absent ductusAgenesis, Aplasticarteriosus10218 Dilated ductusBulbousarteriosus10219 Malpositioned ductus Displaced, Ectopicarteriosus10220 Narrowed ductus Coarctation, Constricted,Normal occurrence afterarteriosusStenosis, Stricturebirth10221 Patent ductusPersistentOpen and unobstructed ductusNormal only in fetal periodarteriosusarteriosus postnatally10222 A-P septum defect Communication between ascending A-P = Aorticopulmonaryaorta and pulmonary trunk10223 Persistent truncus Truncus arteriosusA common aortic and pulmonary truncusarteriosuscommunis10224 Transposition of greatOrigin of aorta from right ventricle andvesselspulmonary trunk from left ventricle10225 Dilated innominate Bulbous Also known as brachiocephalic trunk10226 Elongated innominate Also known as brachiocephalic trunk10227 MalpositionedDisplaced, EctopicAlso known as brachiocephalic trunkinnominate10228 Narrowed innominate Coarctation, Constricted,Stenosis, StrictureAlso known as brachiocephalic trunk932 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructurePulmonaryarteryPulmonarytrunkSubclavianarteryCodeNumberObservationSynonym orRelated Term10229 Short innominate Rudimentary Also known as brachiocephalic trunk10230 MalpositionedDisplaced, Ectopicpulmonary arterybranch10231 Dilated pulmonary Bulboustrunk10232 Narrowed pulmonary Coarctation, Constricted,[Tetratology of Fallot]trunkStenosis, Stricture10233 Pulmonary trunk Imperforateatresia10234 Retroesophagealpulmonary trunk10235 Right-sided pulmonarytrunk10236 Absent subclavian Agenesis, Aplastic10237 Dilated subclavian Bulbous10238 MalpositionedDisplaced, Ectopicsubclavian branchNon-preferredTerm Definition Note10239 Narrowed subclavian Coarctation, Constricted,Stenosis, Stricture10240 RetroesophagealsubclavianTrachea 10241 Malpositioned trachea Displaced, Ectopic10242 Narrowed trachea Coarctation, Constricted,Stenosis, Stricture10243 TracheoesophagealfistulaEsophagus 10244 Absent esophagus Agenesis, Aplastic10245 Atresia of esophagus Imperforate10246 Diverticulum ofesophagus10247 MalpositionedDisplaced, Ectopicesophagus10248 Narrowed esophagus Coarctation, Constricted,Esophagostenosis,Stenosis, StrictureLung 10249 Abnormal lung lobation Fused, Absent fissure Fusion of lung lobesIncludes right subclavianartery originating fromaortaTERMINOLOGY OF ANATOMICAL DEFECTS 933© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Visceral abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservation10250 Absent lung Agenesis, Aplastic,ApulmonismAbsence of one or more lobes10251 Atelectasis Incomplete expansion of lungs orportions of lungs, due to collapse ofpulmonary alveoli during postnatal life10252 Discolored lung Mottled10253 Enlarged lung10254 Hemorrhagic lung10255 Infarct of lung Well demarcated macroscopic area oftissue discoloration due to insufficiencyof blood supply10256 Malpositioned lung Displaced, Ectopic10257 Misshapen lung Abnormally shaped,Irregularly shaped10258 Pale lung10259 Small lung Hypoplastic, Reduced,Rudimentary10260 Supernumerary lunglobeAdditional, ExtraVeins 10261 Absent anterior venacavaSynonym orRelated TermAgenesis, AplasticNon-preferredTerm Definition NoteRefers to one or morelobes. In rabbits mostoften the caudate lobe;see notes for code 10260Usually whitish in colorand the result ofnecrosis, which shouldbe confirmedhistologicallyNormal lobation numberfor rabbits: 3 lobes onright (diaphrag. lobefurther divided into largelateral and smallcaudate), and 2 on left.Normal number forrodents: 3 lobes on right,one medial lobe (fromright bronchus), and 1 onleftMay also be called cranialor superior vena cava10262 Absent azygos Right only normal in rabbit,left only normal in rat andmouse934 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservation10263 Absent posterior venacavaAgenesis, AplasticMay also be called caudalor inferior vena cava10264 Dilated anterior venacavaBulbousMay also be called cranialor superior vena cava10265 Dilated posterior venacavaBulbousMay also be called caudalor inferior vena cava10266 Interrupted anteriorvena cavaMay also be called cranialor superior vena cava10267 Interrupted posteriorvena cavaMay also be called caudalor inferior vena cava10268 Malpositioned anteriorvena cavaDisplaced, EctopicMay also be called cranialor superior vena cava10269 Malpositioned posteriorvena cavaDisplaced, EctopicMay also be called caudalor inferior vena cava10270 Narrowed anterior venacavaCoarctation, Constricted,Stenosis, StrictureMay also be called cranialor superior vena cava10271 Narrowed posteriorvena cavaCoarctation, Constricted,Stenosis, StrictureMay also be called caudalor inferior vena cava10272 Supernumerary azygos Additional, Extra Azygos veins on both sides Right only normal in rabbit,left only normal in rat andmouse10273 Transposed azygos Azygos vein on side other than normal Right only normal in rabbit,left only normal in rat andmouseDiaphragm 10274 Absent diaphragm Agenesis, Aplastic10275 Diaphragmatic hernia Absence of portion of the diaphragm withprotrusion of some abdominal viscera10276 Eventration ofdiaphragmSynonym orRelated TermAbnormal anterior protrusion of a part ofthe diaphragm which is thin and coversvariably displaced abdominal visceraLiver 10277 Abnormal liver lobation Fused, Absent fissure Fusion of liver lobes10278 Absent liver Agenesis, Aplastic Absence of one or more lobes10279 Discolored liver Mottled10280 Hemorrhagic liver Hepatorrhagia10281 Hepatomegaly Enlarged liver10282 Infarct of liver Well demarcated, macroscopic area oftissue discoloration10283 Malpositioned liver Displaced, EctopicNon-preferredTerm Definition NoteUsually whitish in colorand is the result of celldestruction which shouldbe confirmedhistologicallyTERMINOLOGY OF ANATOMICAL DEFECTS 935© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Visceral abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservationSynonym orRelated Term10284 Misshapen liver Abnormally shaped,Irregularly shaped10285 Pale liver10286 Small liver Hypoplastic,Microhepatia, Reduced,Rudimentary10287 Supernumerary liverlobeAdditional, ExtraNon-preferredTerm Definition NoteRefers to one or moreextra lobesGallbladder 10288 Absent bile duct Agenesis, Aplastic Not present in rats10289 Absent gallbladder Agenesis, Aplastic Not present in rats10290 Bilobed gallbladder Not present in rats10291 Elongated bile duct Not present in rats10292 Enlarged gallbladder Not present in rats10293 MalpositionedDisplaced, EctopicNot present in ratsgallbladder10294 Misshapen gallbladder Abnormally shaped,Not present in ratsIrregularly shaped10295 Short bile duct Rudimentary Not present in rats10296 Small gallbladder Hypoplastic, Reduced,Not present in ratsRudimentary10297 Supernumerary Additional, ExtraNot present in ratsgallbladderStomach 10298 Absent stomach Agastria, Agenesis,Aplastic10299 Atresia of stomach Atretogastria, Imperforate10300 Distended stomach10301 Diverticulum ofstomach10302 Enlarged stomach Gastromegaly10303 Malpositioned stomach Displaced, Ectopic, DextrogastriaGastroptosis10304 Narrowed stomach Coarctation, Constricted,Includes pylorusStenosis, Stricture10305 Small stomach Hypoplastic, Microgastria,Reduced, Rudimentary936 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition NotePancreas 10306 Absent pancreas Agenesis, Aplastic10307 Small pancreas Hypoplastic, Reduced,Rudimentary10308 Supernumerary Additional, ExtrapancreasSpleen 10309 Asplenia Agenesis, Aplastic Absence of the spleen10310 Cyst on spleen10311 Discolored spleen Mottled10312 Malpositioned spleen Displaced, Ectopic10313 Misshapen spleen Abnormally shaped,Irregularly shaped10314 Pale spleen10315 Small spleen Hypoplastic,Microsplenia, Reduced,Rudimentary10316 Splenomegaly Enlarged spleen10317 Supernumerary spleen Additional, ExtraIntestines 10318 Absent intestine Agenesis, Aplastic Includes any region fromduodenum to rectum10319 Atresia of intestine Imperforate Includes any region fromduodenum to rectum10320 Diverticulum ofintestineIncludes any region fromduodenum to rectum10321 Enlarged intestine Enteromegaly Includes any region fromduodenum to rectum10322 Fistula of intestine Includes any region fromduodenum to rectum10323 Malpositioned intestine Displaced, Ectopic Includes any region fromduodenum to rectum10324 Narrowed intestine Coarctation, Constricted,Stenosis, StrictureIncludes any region fromduodenum to rectum10325 Short intestine Rudimentary Includes any region fromduodenum to rectumKidney 10326 Absent kidney Agenesis, Aplastic10327 Absent renal papilla10328 Cyst on kidney10329 Dilated renal pelvis Bulbous Slight or moderate dilatation of renalpelvis10330 Discolored kidney Mottled10331 Enlarged kidney10332 Fused kidney HorseshoeTERMINOLOGY OF ANATOMICAL DEFECTS 937© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Visceral abnormalities 1 (continued)Region/Organ/StructureAdrenalglandCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10333 Hemorrhagic kidney10334 Hydronephrosis Marked dilatation of renal pelvis andcalices secondary to obstruction ofurine flow, usually combined withdestruction of the renal parenchyma10335 Infarct of kidney10336 Malpositioned kidney Displaced, Ectopic10337 Misshapen kidney Abnormally shaped,Irregularly shaped10338 Pale kidney10339 Small kidney Hypoplastic, Reduced,Rudimentary10340 Small renal papilla Hypoplastic, Reduced,Rudimentary10341 Small renal pelvis Hypoplastic, Reduced,Rudimentary10342 Supernumerary kidney Additional, Extra10343 Absent adrenal Agenesis, Aplastic10344 Enlarged adrenal10345 Extracapsular adrenaltissue10346 Fused adrenal10347 Hemorrhagic adrenal10348 Malpositioned adrenal Displaced, Ectopic10349 Misshapen adrenal Abnormally shaped,Irregularly shaped10350 Small adrenal Hypoplastic, Reduced,Rudimentary10351 Supernumerary Additional, ExtraadrenalBladder 10352 Acystia Agenesis, Aplastic Absence of the bladder10353 Cyst on bladder10354 Distended bladderDestruction should beconfirmed histologically.Often associated withdilated ureter938 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10355 Small bladder Hypoplastic, Reduced,RudimentaryUreter 10356 Absent ureter Agenesis, Aplastic10357 Convoluted ureter Coiled, Folded, Kinked,Folded, curved, and/or tortuous windingsTwisted10358 Dilated ureter Bulbous, Distended Slight or moderate distention of ureter(s)10359 Doubled ureter10360 Hydroureter Markedly dilated ureter(s) Compare to Dilated; mayaccompanyhydronephrosis10361 Retrocaval ureter Passing dorsally to the vena cavaGonad 10362 Absent gonads Agenesis, Agonad,Aplastic10363 Hermaphroditism Hermaphrodism Presence of both male and femalegonadal tissue10364 Small gonad Hypoplastic, Reduced,RudimentaryUsed when sex can not bedeterminedUsed when sex can not bedetermined10365 Supernumerary gonad Additional, Extra Used when sex can not bedeterminedTestis 10366 Anorchia Agenesis, Aplastic Absence of a testis If both gonads are absentuse “Absent gonad”10367 Enlarged testis10368 Hemorrhagic testis10369 Malpositioned testis Displaced, Ectopic Cryptorchidism Cryptorchidism =testis(es) neverdescended into scrotum10370 Misshapen testis Abnormally shaped,Irregularly shaped10371 Small testis Hypoplastic, Microrchidia,Reduced, Rudimentary10372 Supernumerary testis Additional, ExtraEpididymis 10373 Absent epididymis Agenesis, Aplastic10374 Hemorrhagicepididymis10375 MalpositionedDisplaced, Ectopicepididymis10376 Short epididymis Rudimentary10377 Small epididymis Hypoplastic, Reduced,RudimentaryVasdeferens10378 Absent vas deferens Agenesis, AplasticTERMINOLOGY OF ANATOMICAL DEFECTS 939© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Visceral abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservation10379 Atresia of vas deferens Imperforate10380 Malpositioned vas Displaced, Ectopicdeferens10381 Short vas deferens RudimentaryOvary 10382 Absent ovary Agenesis, Aplastic If both gonads are absentuse Absent gonads10383 Cyst on ovary10384 Enlarged ovary10385 Hemorrhagic ovary Oophorrhagia10386 Malpositioned ovary Displaced, Ectopic10387 Misshapen ovary Abnormally shaped,Irregularly shaped10388 Small ovary Hypoplastic, Reduced,Rudimentary10389 Supernumerary ovary Additional, ExtraOviduct 10390 Absent oviduct Agenesis, Aplastic10391 Cyst on oviduct10392 Enlarged oviduct10393 Hemorrhagic oviduct10394 Malpositioned oviduct Displaced, Ectopic10395 Misshapen oviduct Abnormally shaped,Irregularly shaped10396 Short oviduct RudimentaryUterus 10397 Absent uterine horn Agenesis, Ametria,AplasticAbsence of one uterine hornUse Absent uterus if bothhorns are absent10398 Atresia of uterus Hysteratresia, Imperforate Specify total or partial10399 Hemorrhagic uterus10400 Misshapen uterus Abnormally shaped,Irregularly shaped10401 Small uterus Hypoplastic, Reduced,Rudimentary1For terms in brackets, see Appendix 1-B.Synonym orRelated TermNon-preferredTerm Definition Note940 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition NoteSkull, General 10402 Acrania Absence of the calvarium and variablyother bones comprising the braincase10403 Craniofenestria The incomplete ossification of thebones of the cranial vault such thatthe calvarium appears fenestrated10404 Craniostenosis Premature closure of cranial suturesand fontanels resulting in smallmaldeveloped skullUsed to describe multipleskull bone fusions10405 Enlarged fontanel Specify anterior orposterior10406 Extra ossification site Additional, Bone island,Isolated site of ossificationSpecify locationFontanellar bone, SutureboneAlisphenoid 10407 Absent alisphenoid Agenesis, Aplastic10408 Alisphenoid hole(s)10409 Fused alisphenoid Specify structures that arefused10410 Incomplete ossification Reduced ossificationof alisphenoid10411 Misshapen alisphenoid Abnormally shaped,Irregularly shaped10412 Small alisphenoid Hypoplastic, Reduced,Rudimentary10413 Unossified alisphenoidAuditoryossicles10414 Absent auditoryossiclesAgenesis, AplasticIncludes incus, malleus,and stapes10415 Fused auditory ossicles Specify structures that arefused. Includes incus,malleus, and stapes10416 Misshapen auditoryossiclesAbnormally shaped,Irregularly shapedIncludes incus, malleus,and stapes10417 Unossified auditoryossiclesIncludes incus, malleus,and stapesBasioccipital 10418 Absent basioccipital Agenesis, Aplastic10419 Basioccipital hole(s)10420 Fused basioccipitalTERMINOLOGY OF ANATOMICAL DEFECTS 941© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservation10421 Incomplete ossificationof basioccipital10422 MisshapenbasioccipitalReduced ossificationAbnormally shaped,Asymmetric, Irregularlyshaped10423 Small basioccipital Hypoplastic, Reduced,Rudimentary10424 Unossified basioccipitalBasisphenoid 10425 Absent basisphenoid Agenesis, Aplastic10426 Basisphenoid hole(s)10427 Fused basisphenoid Specify structures that arefused10428 Incomplete ossificationof basisphenoidReduced ossification10429 MisshapenbasisphenoidSynonym orRelated TermAbnormally shaped,Asymmetric, IrregularlyshapedNon-preferredTerm Definition Note10430 Small basisphenoid Hypoplastic, Reduced,Rudimentary10431 UnossifiedbasisphenoidExoccipital 10432 Absent exoccipital Agenesis, Aplastic10433 Exoccipital hole(s)10434 Fused exoccipital Specify structures that arefused10435 Incomplete ossification Reduced ossificationof exoccipital10436 Misshapen exoccipital Abnormally shaped,Irregularly shaped10437 Small exoccipital Hypoplastic, Reduced,Rudimentary10438 Unossified exoccipitalFrontal 10439 Absent frontal Agenesis, Aplastic10440 Frontal hole(s)10441 Fused frontal Specify structures that arefused942 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10442 Incomplete ossification Reduced ossificationof frontal10443 Misshapen frontal Abnormally shaped,Asymmetric, Irregularlyshaped10444 Small frontal Hypoplastic, Reduced,Rudimentary10445 Unossified frontalHyoid 10446 Absent hyoid Agenesis, Aplastic10447 Bent hyoid Angulated, Bowed10448 Incomplete ossification Reduced ossificationof hyoid10449 Misshapen hyoid Abnormally shaped,Asymmetric, Irregularlyshaped10450 Small hyoid Hypoplastic, Reduced,Rudimentary10451 Unossified hyoidInterparietal 10452 Absent interparietal Agenesis, Aplastic10453 Bipartite ossification of Bifid ossificationinterparietal10454 Fused interparietal Specify structures that arefused10455 Incomplete ossification Reduced ossificationof interparietal10456 Interparietal hole(s)10457 Misshapen interparietal Abnormally shaped,Asymmetric, Irregularlyshaped10458 Small interparietal Hypoplastic, Reduced,Rudimentary10459 Unossified interparietalLacrimal 10460 Absent lacrimal Agenesis, Aplastic10461 Fused lacrimal Specify structures that arefused10462 Incomplete ossification Reduced ossificationof lacrimal10463 Misshapen lacrimal Abnormally shaped,Irregularly shaped10464 Small lacrimal Hypoplastic, Reduced,RudimentaryTERMINOLOGY OF ANATOMICAL DEFECTS 943© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10465 Unossified lacrimalMandible 10466 Absent mandible Agenesis, Aplastic10467 Fused mandible Specify structures that arefused10468 Incomplete ossification Reduced ossificationof mandible10469 Misshapen mandible Abnormally shaped,Asymmetric, Irregularlyshaped10470 Small mandible Hypoplastic, Reduced,Rudimentary10471 Unossified mandibleMaxilla 10472 Absent maxilla Agenesis, Aplastic10473 Fused maxilla Specify structures that arefused10474 Incomplete ossification Reduced ossificationof maxilla10475 Misshapen maxilla Abnormally shaped,Irregularly shaped10476 Small maxilla Hypoplastic, Reduced,Rudimentary10477 Unossified maxillaNasal 10478 Absent nasal Agenesis, Aplastic10479 Fused nasal Specify structures that arefused10480 Incomplete ossification Reduced ossificationof nasal10481 Misshapen nasal Abnormally shaped,Asymmetric, Irregularlyshaped10482 Nasal hole(s)10483 Small nasal Hypoplastic, Reduced,Rudimentary10484 Unossified nasalPalatine 10485 Absent palatine Agenesis, Aplastic944 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10486 Fused palatine Specify structures that arefused10487 Incomplete ossification Reduced ossificationof palatine10488 Misshapen palatine Abnormally shaped,Asymmetric, Irregularlyshaped10489 Small palatine Hypoplastic, Reduced,Rudimentary10490 Split palatine Bifid, Bipartite, Cleftpalate10491 Unossified palatineParietal 10492 Absent parietal Agenesis, Aplastic10493 Fused parietal Specify structures that arefused10494 Incomplete ossification Reduced ossificationof parietal10495 Misshapen parietal Abnormally shaped,Asymmetric, Irregularlyshaped10496 Parietal hole(s)10497 Small parietal Hypoplastic, Reduced,Rudimentary10498 Unossified parietalPremaxilla 10499 Absent premaxilla Agenesis, Aplastic10500 Fused premaxilla Specify structures that arefused10501 Incomplete ossification Reduced ossificationof premaxilla10502 Misshapen premaxilla Abnormally shaped,Asymmetric, Irregularlyshaped10503 Premaxilla hole(s)10504 Small premaxilla Hypoplastic, Reduced,Rudimentary10505 Unossified premaxillaPresphenoid 10506 Absent presphenoid Agenesis, Aplastic10507 Fused presphenoid Specify structures that arefused10508 Incomplete ossificationof presphenoidReduced ossificationTERMINOLOGY OF ANATOMICAL DEFECTS 945© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservation10509 MisshapenpresphenoidAbnormally shaped,Asymmetric, Irregularlyshaped10510 Presphenoid hole(s)10511 Small presphenoid Hypoplastic, Reduced,Rudimentary10512 UnossifiedpresphenoidSquamosal 10513 Absent squamosal Agenesis, Aplastic Bone may be calledtemporal10514 Fused squamosal Specify structures that arefused Bone may becalled temporal10515 Incomplete ossificationof squamosalReduced ossificationBone may be calledtemporal10516 Misshapen squamosal Abnormally shaped,Irregularly shapedBone may be calledtemporal10517 Small squamosal Hypoplastic, Reduced,RudimentaryBone may be calledtemporal10518 Squamosal hole(s)10519 Unossified squamosal Bone may be calledtemporalSupraoccipital 10520 Absent supraoccipital Agenesis, Aplastic10521 Bipartite ossification of Bifid ossificationsupraoccipital10522 Fused supraoccipital Specify structures that arefused10523 Incomplete ossificationof supraoccipitalReduced ossification10524 MisshapensupraoccipitalSynonym orRelated TermAbnormally shaped,Asymmetric, Irregularlyshaped10525 Small supraoccipital Hypoplastic, Reduced,Rudimentary10526 Supraoccipital hole(s)Non-preferredTerm Definition Note946 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10527 UnossifiedsupraoccipitalTympanic 10528 Absent tympanic Agenesis, Aplasticannulusannulus10529 Fused tympanicannulusSpecify structures that arefused10530 Incomplete ossification Reduced ossificationof tympanic annulus10531 Misshapen tympanicannulusAbnormally shaped,Irregularly shaped10532 Small tympanicannulusHypoplastic, Reduced,Rudimentary10533 Unossified tympanicannulusVomer 10534 Absent vomer Agenesis, Aplastic10535 Fused vomer Specify structures that arefused10536 Incomplete ossification Reduced ossificationof vomer10537 Misshapen vomer Abnormally shaped,Asymmetric, Irregularlyshaped10538 Small vomer Hypoplastic, Reduced,Rudimentary10539 Unossified vomerZygomatic 10540 Absent zygomatic Agenesis, Aplastic Bone may be called jugalor malar10541 Fused zygomatic Specify structures that arefused Bone may becalled jugal or malar10542 Incomplete ossificationof zygomaticReduced ossificationBone may be called jugalor malar10543 Misshapen zygomatic Abnormally shaped,Irregularly shapedBone may be called jugalor malar10544 Small zygomatic Hypoplastic, Reduced,RudimentaryBone may be called jugalor malar10545 Unossified zygomatic Bone may be called jugalor malarClavicle 10546 Absent clavicle Agenesis, Aplastic10547 Bent clavicle Angulated, Bowed10548 Incomplete ossificationof clavicleReduced ossificationTERMINOLOGY OF ANATOMICAL DEFECTS 947© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10549 Misshapen clavicle Abnormally shaped,Irregularly shaped10550 Small clavicle Hypoplastic, Reduced,Rudimentary10551 Thickened clavicle10552 Unossified clavicleScapula 10553 Absent scapula Agenesis, Aplastic10554 Bent scapula Angulated, Bowed10555 Incomplete ossification Reduced ossificationof scapula10556 Misshapen scapula Abnormally shaped,Forked spine, Irregularlyshaped10557 Thickened scapula10558 Unossified scapulaHumerus 10559 Absent humerus Agenesis, Aplastic10560 Bent humerus Angulated, Bowed10561 Fused humerus Specify structures that arefused10562 Incomplete ossification Reduced ossificationof humerus10563 Malpositioned humerus Displaced, Ectopic10564 Misshapen humerus Abnormally shaped,Irregularly shaped10565 Short humerus Rudimentary10566 Thickened humerus10567 Unossified humerusRadius 10568 Absent radius Agenesis, Aplastic, Radialhemimelia10569 Bent radius Angulated, Bowed10570 Fused radius Specify structures that arefused10571 Incomplete ossification Reduced ossificationof radius10572 Malpositioned radius Displaced, Ectopic948 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10573 Misshapen radius Abnormally shaped,Irregularly shaped10574 Short radius Rudimentary10575 Thickened radius10576 Unossified radiusUlna 10577 Absent ulna Agenesis, Aplastic, Ulnarhemimelia10578 Bent ulna Angulated, Bowed10579 Fused ulna Specify structures that arefused10580 Incomplete ossification Reduced ossificationof ulna10581 Malpositioned ulna Displaced, Ectopic10582 Misshapen ulna Abnormally shaped,Irregularly shaped10583 Short ulna Rudimentary10584 Thickened ulna10585 Unossified ulnaCarpal bone 10586 Absent carpal bone Agenesis, Aplastic10587 Fused carpal bone Specify structures that arefused10588 Incomplete ossificationof carpal bone10589 Malpositioned carpal Displaced, Ectopicbone10590 Misshapen carpal bone Abnormally shaped,Irregularly shaped10591 Small carpal bone Hypoplastic, Reduced,Rudimentary10592 Supernumerary carpal Additional, Extrabone10593 Unossified carpal boneMetacarpal 10594 Absent metacarpal Agenesis, Aplastic10595 Fused metacarpal Specify structures that arefused10596 Incomplete ossification Reduced ossificationof metacarpal10597 MalpositionedDisplaced, Ectopicmetacarpal10598 Misshapen metacarpal Abnormally shaped,Irregularly shapedTERMINOLOGY OF ANATOMICAL DEFECTS 949© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10599 Small metacarpal Hypoplastic, Reduced,Rudimentary10600 Supernumerary Additional, Extrametacarpal10601 Unossified metacarpalForepawphalanx10602 Absent phalanx Agenesis, Aphalangia,AplasticAbsence of one or more phalanges inone or more digits10603 Fused phalanx Specify structures that arefused10604 Incomplete ossification Reduced ossificationof phalanx10605 Malpositioned phalanx Displaced, Ectopic10606 Misshapen phalanx Abnormally shaped,Irregularly shaped10607 Small phalanx Hypoplastic, Reduced,Rudimentary10608 Supernumerary Additional, Extraphalanx10609 Thickened phalanx10610 Unossified phalanxSternebra 10611 Absent sternebra Agenesis, Aplastic10612 Bipartite ossification of Bifid ossificationsternebra10613 Extra sternebralossification siteAdditional Supernumerary Refers to advancedossification betweensternebral centers10614 Fused sternebra Specify structures that arefused10615 Incomplete ossification Reduced ossificationof sternebra10616 MalpositionedDisplaced, Ectopicsternebra10617 Misaligned sternebra Offset Checkerboard10618 Misshapen sternebra Abnormally shaped,Asymmetric, Irregularlyshaped950 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated Term10619 Sternoschisis Bifid, Bipartite, CleftsternumSplit sternum involving the cartilageand one or more sternebral centers10620 Unossified sternebraRib 10621 Absent rib Agenesis, Aplastic Generally the normalnumber of rib pairs is 13in rodent and 12 or 13 inrabbit; but must berelated to normal numberfor given species, strain,and time in eachlaboratory.10622 Bent rib Angulated, Bowed Shaped like an angle10623 Branched rib Bifurcated, Forked Split A rib that divides distally into two ribs10624 Branched rib cartilage Bifurcated, Forked A rib cartilage that divides distally intotwo10625 Cervical rib Supernumerary cervical Ossification lateral to a cervical arch Usually seen on lastcervical arch10626 Detached rib Floating rib Rib with no attachment to the vertebralcolumn10627 Discontinuous rib Interrupted ossification Absence of cartilage and Alizarin Reduptake in a central region of a rib10628 Full supernumerary rib Extra full, LongsupernumerarythoracolumbarNon-preferredTerm Definition NoteLumbar ribAn extra rib that is distally joined by acartilaginous portionGenerally the normalnumber of rib pairs is 13in rodent and 12 or 13 inrabbit; but must berelated to normal numberfor given species, strain,and time in eachlaboratory.10629 Fused rib Specify structures that arefused10630 Fused rib cartilage10631 Incomplete ossification Reduced ossificationof rib10632 Intercostal rib A non-articulating rib-like structurelocated between two other ribs10633 Knobby rib Focal enlargement,A rounded protuberance on a ribNodulated10634 Malpositioned rib Displaced, Ectopic10635 Misaligned rib10636 Misshapen rib Abnormally shaped,Asymmetric, IrregularlyshapedTERMINOLOGY OF ANATOMICAL DEFECTS 951© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureVertebra,GeneralCodeNumberObservation10637 Short rib Rudimentary Not used forsupernumerary rib;generally less than halfnormal length10638 Short supernumeraryribExtra rudimentary,Rudimentarysupernumerarythoracolumbar, ShortsupernumerarythoracolumbarLumbar ribAn extra rib that is usually roundeddistally and without cartilaginousextensionsGenerally the normalnumber of rib pairs is 13in rodent and 12 or 13 inrabbit; but must berelated to normal numberfor given species, strain,and time in eachlaboratory10639 Thickened rib Clubbed10640 Unossified rib10641 Wavy rib Undulation(s) along the length of a rib10642 Absent vertebra Agenesis, Aplastic Used when subregion notspecified10643 Supernumerary Additional, ExtraUsed when subregion notvertebraspecifiedCervical arch 10644 Absent cervical arch Agenesis, Aplastic Atlas = #1, Axis = #210645 Fused cervical arch Specify structures that arefused Atlas = #1,Axis = #210646 Incomplete ossification Reduced ossification Atlas = #1, Axis = #2of cervical arch10647 Malpositioned cervical Displaced, Ectopic Atlas = #1, Axis = #2arch10648 Misaligned cervicalarchAtlas = #1, Axis = #210649 Misshapen cervicalarchSynonym orRelated TermAbnormally shaped,Asymmetric, IrregularlyshapedNon-preferredTerm Definition NoteAtlas = #1, Axis = #210650 Small cervical arch Hypoplastic, Reduced,Atlas = #1, Axis = #2Rudimentary10651 Supernumerarycervical archAdditional, Extra Atlas = #1, Axis = #2952 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCervicalcentrumCodeNumberObservation10652 Unossified cervicalarchAtlas = #1, Axis = #210653 Absent cervical Agenesis, Aplastic Atlas = #1, Axis = #2centrum10654 Bipartite ossification of Bifid ossification Atlas = #1, Axis = #2cervical centrum10655 Dumbbell ossification Dumbbell-shaped Atlas = #1, Axis = #2of cervical centrum10656 Dumbbell-shapedAtlas = #1, Axis = #2cartilage of cervicalcentrum10657 Fused cervical centrum Specify structures that arefused Atlas = #1,Axis = #210658 Fused cervical centrumSpecify structures that arecartilagefused Atlas = #1,Axis = #210659 Hemicentric cervicalAtlas = #1, Axis = #2centrum10660 Incomplete ossification Reduced ossification Atlas = #1, Axis = #2of cervical centrum10661 Misaligned cervicalAtlas = #1, Axis = #2centrum10662 Misshapen cervicalcentrumSynonym orRelated TermNon-preferredTerm Definition NoteAbnormally shaped,Atlas = #1, Axis = #2Asymmetric, IrregularlyshapedBifid, Bipartite, Cleft Atlas = #1, Axis = #210663 Split cartilage ofcervical centrum10664 Supernumerary Additional, Extra Atlas = #1, Axis = #2cervical centrum10665 Unilateral cervicalAtlas = #1, Axis = #2centrum cartilage10666 Unossified cervicalAtlas = #1, Axis = #2centrumCervical 10667 Absent cervical Agenesis, AplasticNormal: 7 in rodent andvertebravertebrarabbit10668 Cervical hemivertebra10669 Malpositioned cervical Displaced, Ectopicvertebra10670 Supernumerary Additional, ExtraNormal: 7 in rodent andcervical vertebrarabbitThoracic arch 10671 Absent thoracic arch Agenesis, AplasticTERMINOLOGY OF ANATOMICAL DEFECTS 953© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureThoraciccentrumCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10672 Fused thoracic arch Specify structures that arefused10673 Incomplete ossification Reduced ossificationof thoracic arch10674 Malpositioned thoracicarch10675 Misaligned thoracicarch10676 Misshapen thoracicarchAbnormally shaped,Asymmetric, Irregularlyshaped10677 Small thoracic arch Hypoplastic, Reduced,Rudimentary10678 Supernumerary Additional, Extrathoracic arch10679 Unossified thoracicarch10680 Absent thoracic Agenesis, Aplasticcentrum10681 Bipartite ossification of Bifid ossificationthoracic centrum10682 Dumbbell ossification Dumbbell-shapedof thoracic centrum10683 Dumbbell-shapedcartilage of thoraciccentrum10684 Fused thoracic centrum Specify structures that arefused10685 Fused thoracic centrumSpecify structures that arecartilagefused10686 Hemicentric thoraciccentrum10687 Incomplete ossification Reduced ossificationof thoracic centrum10688 Misaligned thoraciccentrum954 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/Structure10689 Misshapen thoraciccentrumAbnormally shaped,Asymmetric, IrregularlyshapedBifid, Bipartite, Cleft10690 Split cartilage ofthoracic centrum10691 Supernumerary Additional, Extrathoracic centrum10692 Unilateral thoraciccentrum cartilage10693 Unossified thoraciccentrumThoracic 10694 Absent thoracic Agenesis, AplasticNormal: 13 in rodent,vertebravertebra12 or 13 in rabbit10695 Malpositioned thoracic Displaced, Ectopicvertebra10696 Thoracic hemivertebra10697 Supernumerary Additional, ExtraNormal: 13 in rodent,thoracic vertebra12 or 13 in rabbitLumbar arch 10698 Absent lumbar arch Agenesis, Aplastic10699 Fused lumbar arch Specify structures that arefusedLumbarcentrumCodeNumberObservation10700 Incomplete ossification Reduced ossificationof lumbar arch10701 Malpositioned lumbar Displaced, Ectopicarch10702 Misaligned lumbar arch10703 Misshapen lumbar arch Abnormally shaped,Asymmetric, Irregularlyshaped10704 Small lumbar arch Hypoplastic, Reduced,Rudimentary10705 Supernumerary lumbar Additional, Extraarch10706 Unossified lumbar arch10707 Absent lumbar centrum Agenesis, Aplastic10708 Bipartite ossification oflumbar centrum10709 Dumbbell ossificationof lumbar centrum10710 Dumbbell-shapedcartilage of lumbarcentrumSynonym orRelated TermBifid ossificationDumbbell-shapedNon-preferredTerm Definition NoteTERMINOLOGY OF ANATOMICAL DEFECTS 955© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureLumbarvertebraCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10711 Fused lumbar centrum Specify structures that arefused10712 Fused lumbar centrumcartilageSpecify structures that arefused10713 Hemicentric lumbarcentrum10714 Incomplete ossification Reduced ossificationof lumbar centrum10715 Misaligned lumbarcentrum10716 Misshapen lumbarcentrumAbnormally shaped,Asymmetric, Irregularlyshaped10717 Split cartilage of lumbar Bifid, Bipartite, Cleftcentrum10718 Supernumerary lumbar Additional, Extracentrum10719 Unilateral lumbarcentrum cartilage10720 Unossified lumbarcentrum10721 Absent lumbar vertebra Agenesis, Aplastic Normal: 6 in rodent, 6 or 7in rabbit10722 Lumbar hemivertebra10723 Malpositioned lumbar Displaced, Ectopicvertebra10724 Supernumerary lumbar Additional, Extra Normal: 6 in rodent, 6 or 7vertebrain rabbitSacral arch 10725 Absent sacral arch Agenesis, Aplastic10726 Fused sacral arch Specify structures that arefused10727 Incomplete ossification Reduced ossificationof sacral arch10728 Malpositioned sacral Displaced, Ectopicarch10729 Misaligned sacral arch956 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureSacralcentrumSacralvertebraCodeNumberObservationSynonym orRelated Term10730 Misshapen sacral arch Abnormally shaped,Asymmetric, Irregularlyshaped10731 Small sacral arch Hypoplastic, Reduced,Rudimentary10732 Supernumerary sacral Additional, Extraarch10733 Unossified sacral arch10734 Absent sacral centrum Agenesis, AplasticNon-preferredTerm Definition Note10735 Bipartite ossification of Bifid ossificationsacral centrum10736 Dumbbell ossification Dumbbell-shapedof sacral centrum10737 Dumbbell-shapedcartilage of sacralcentrum10738 Fused sacral centrum Specify structures that arefused10739 Fused sacral centrumcartilageSpecify structures that arefused10740 Hemicentric sacralcentrum10741 Incomplete ossification Reduced ossificationof sacral centrum10742 Misaligned sacralcentrum10743 Misshapen sacralcentrumAbnormally shaped,Asymmetric, Irregularlyshaped10744 Split cartilage of sacral Bifid, Bipartite, Cleftcentrum10745 Supernumerary sacral Additional, Extracentrum10746 Unilateral sacralcentrum cartilage10747 Unossified sacralcentrum10748 Absent sacral vertebra Agenesis, Aplastic Normal: 4 in rodent andrabbit10749 Malpositioned sacral Displaced, EctopicvertebraTERMINOLOGY OF ANATOMICAL DEFECTS 957© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservation10750 Sacral hemivertebra10751 Supernumerary sacralvertebraAdditional, ExtraNormal: 4 in rodent andrabbitCaudal arch 10752 Absent caudal arch Agenesis, Aplastic10753 Fused caudal arch Specify structures that arefused10754 Incomplete ossification Reduced ossificationof caudal arch10755 Malpositioned caudal Displaced, Ectopicarch10756 Misaligned caudal arch10757 Misshapen caudal arch Abnormally shaped,Asymmetric, Irregularlyshaped10758 Small caudal arch Hypoplastic, Reduced,Rudimentary10759 Unossified caudal archCaudal10760 Absent caudal centrum Agenesis, Aplasticcentrum10761 Bipartite ossification of Bifid ossificationcaudal centrum10762 Dumbbell ossification Dumbbell-shapedof caudal centrum10763 Fused caudal centrum Specify structures that arefused10764 Hemicentric caudalcentrum10765 Incomplete ossificationof caudal centrumReduced ossification10766 Misaligned caudalcentrum10767 Misshapen caudalcentrum10768 Unossified caudalcentrumSynonym orRelated TermAbnormally shaped,Asymmetric, IrregularlyshapedNon-preferredTerm Definition Note958 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition NoteCaudal10769 Absent caudal vertebra Agenesis, Aplasticvertebra10770 Caudal hemivertebra10771 Malpositioned caudal Displaced, Ectopicvertebra10772 Supernumerary caudal Additional, ExtravertebraIlium 10773 Absent ilium Agenesis, Aplastic10774 Bent ilium Angulated, Bowed10775 Fused ilium Specify structures that arefused10776 Incomplete ossification Reduced ossificationof ilium10777 Malpositioned ilium10778 Misshapen ilium Abnormally shaped,Irregularly shaped10779 Small ilium Hypoplastic, Reduced,Rudimentary10780 Thickened ilium10781 Unossified iliumIschium 10782 Absent ischium Agenesis, Aplastic10783 Bent ischium Angulated, Bowed10784 Fused ischium Specify structures that arefused10785 Incomplete ossification Reduced ossificationof ischium10786 Malpositioned ischium10787 Misshapen ischium Abnormally shaped,Irregularly shaped10788 Small ischium Hypoplastic, Reduced,Rudimentary10789 Thickened ischium10790 Unossified ischiumPubis 10791 Absent pubis Agenesis, Aplastic10792 Bent pubis Angulated, Bowed10793 Incomplete ossification Reduced ossificationof pubis10794 Malpositioned pubis10795 Misaligned pubis10796 Misshapen pubis Abnormally shaped,Irregularly shapedTERMINOLOGY OF ANATOMICAL DEFECTS 959© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10797 Small pubis Hypoplastic, Reduced,Rudimentary10798 Thickened pubis10799 Unossified pubisFemur 10800 Absent femur Agenesis, Aplastic10801 Bent femur Angulated, Bowed10802 Fused femur Specify structures that arefused10803 Incomplete ossification Reduced ossificationof femur10804 Malpositioned femur Displaced, Ectopic10805 Misshapen femur Abnormally shaped,Irregularly shaped10806 Short femur Rudimentary10807 Thickened femur10808 Unossified femurFibula 10809 Absent fibula Agenesis, Aplastic10810 Bent fibula Angulated, Bowed10811 Fused fibula Specify structures that arefused10812 Incomplete ossification Reduced ossificationof fibula10813 Malpositioned fibula Displaced, Ectopic10814 Misshapen fibula Abnormally shaped,Irregularly shaped10815 Short fibula Rudimentary10816 Thickened fibula10817 Unossified fibulaTibia 10818 Absent tibia Agenesis, Aplastic, Tibialhemimelia10819 Bent tibia Angulated, Bowed10820 Fused tibia Specify structures that arefused10821 Incomplete ossificationof tibiaReduced ossification960 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10822 Malpositioned tibia Displaced, Ectopic10823 Misshapen tibia Abnormally shaped,Irregularly shaped10824 Short tibia Rudimentary10825 Thickened tibia10826 Unossified tibiaCalcaneus 10827 Absent calcaneus Agenesis, Aplastic10828 Fused calcaneus Specify structures that arefused10829 Incomplete ossification Reduced ossificationof calcaneus10830 MalpositionedDisplaced, Ectopiccalcaneus10831 Misshapen calcaneus Abnormally shaped,Irregularly shaped10832 Small calcaneus Hypoplastic, Reduced,Rudimentary10833 Supernumerary Additional, Extracalcaneus10834 Unossified calcaneusTalus 10835 Absent talus Agenesis, Aplastic Bone may be calledastragalus10836 Fused talus Specify structures that arefused Bone may becalled astragalus10837 Incomplete ossificationof talusReduced ossificationBone may be calledastragalus10838 Malpositioned talus Displaced, Ectopic Bone may be calledastragalus10839 Misshapen talus Abnormally shaped,Irregularly shapedBone may be calledastragalus10840 Small talus Hypoplastic, Reduced,RudimentaryBone may be calledastragalus10841 Thickened talus Bone may be calledastragalus10842 Unossified talus Bone may be calledastragalusTarsal bone 10843 Absent tarsal bone Agenesis, Aplastic Excludes calcaneus andtalusTERMINOLOGY OF ANATOMICAL DEFECTS 961© 2006 by Taylor & Francis Group, LLC

Terminology of developmental abnormalities in common laboratory mammals (version 1): Skeletal abnormalities 1 (continued)Region/Organ/Structure10844 Fused tarsal bone Specify structures that arefused Excludescalcaneus and talus10845 Incomplete ossificationof tarsal bone10846 Malpositioned tarsalboneDisplaced, EctopicExcludes calcaneus andtalus10847 Misshapen tarsal bone Abnormally shaped,Irregularly shapedExcludes calcaneus andtalus10848 Small tarsal bone Hypoplastic, Reduced,RudimentaryExcludes calcaneus andtalus10849 Supernumerary tarsalboneAdditional, ExtraExcludes calcaneus andtalus10850 Thickened tarsal bone Excludes calcaneus andtalus10851 Unossified tarsal boneMetatarsal 10852 Absent metatarsal Agenesis, Aplastic10853 Fused metatarsal Specify structures that arefused10854 Incomplete ossification Reduced ossificationof metatarsal10855 MalpositionedDisplaced, Ectopicmetatarsal10856 Misshapen metatarsal Abnormally shaped,Irregularly shaped10857 Small metatarsal Hypoplastic, Reduced,Rudimentary10858 Supernumerary Additional, Extrametatarsal10859 Unossified metatarsalHindpawphalanxCodeNumberObservationSynonym orRelated TermNon-preferredTerm Definition Note10860 Absent phalanx Agenesis, Aphalangia,AplasticAbsence of one or more phalanges inone or more digits10861 Fused phalanx Specify structures that arefused10862 Incomplete ossificationof phalanxReduced ossification962 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Organ/StructureCodeNumberObservation10863 Malpositioned phalanx Displaced, Ectopic10864 Misshapen phalanx Abnormally shaped,Irregularly shaped10865 Small phalanx Hypoplastic, Reduced,Rudimentary10866 Supernumerary Additional, Extraphalanx10867 Thickened phalanx10868 Unossified phalanx1For terms in brackets, see Appendix 1-B.Synonym orRelated TermNon-preferredTerm Definition NoteTERMINOLOGY OF ANATOMICAL DEFECTS 963© 2006 by Taylor & Francis Group, LLC

964 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAppendix 1-ATermDescriptive terminology used more than once in the glossaryDefinitionSynonym orRelated termAbsentAn absolute failure of development Agenesis, Aplastic Missingof an organ or part; in the case ofbone, no cartilage model is presentAtresiaAbsence or closure of a normal body Imperforateorifice or tubular organBipartiteHaving two separate ossification Bifid ossification Splitossification sitesBentAbnormal curvature of a normally Angulated, Bowedstraight structureBranchedDeviation from the predominant Bifurcated, Forkedpattern due to displacement and/orduplication of one or more normallyoccurring structuresCystAny abnormal sac; usuallycontaining fluid or other materialDilatedWidened or expanded orifice or BulbousSwollenvesselDiscolored Not the normal color Mottled, PaleDistended Enlarged or expanded organ SwollenDiverticulum A localized sac or pouch formed inthe wall of a hollow viscus andopening into its lumenDumbbellossificationTwo roughly spherical ossificationsites attached at or near themid-line by an ossified bridgeDumbbell-shapedossificationNon-preferred TermBilobate, BilobedEnlarged Larger than normal HyperplasticFistulaAbnormal passage or communicationbetween two normallyunconnected structures, bodycavities, or the surface of the bodyFusedJoined or blended togetherHemicentric Absence of either hemicentrum of acentrumHemivertebra Absence of a lateral half (arch +hemicentrum), or major part of alateral half of a vertebraHemorrhagic Descriptive of any tissue into whichabnormal bleeding is observed(may be graded as petechia,purpura, ecchymosis, orhematoma)IncompleteReduced ossificationossificationMalpositionedMisalignedMisshapenNarrowedRetro-esophagealRetrotrachealPartial ossification (as assessed byAlizarin Red uptake) at a site which,in controls of the same age, usuallyhas a more advanced degree ofossificationNot occurring in the proper positionand/or orientationAbnormal relative position ofstructures on opposite sides of adividing line or about the center oraxisAbnormally shaped [Not to be usedto describe sites of incompleteossification]Constriction of a cylindricalstructure, such as the aorta or tail,or of a lumenPassing dorsal to the esophagusPassing dorsal to the tracheaDisplaced, EctopicAbnormally shaped,Asymmetric,Irregularly shapedCoarctation,Constricted,Stenosis, StrictureDelayed ossification,Retarded ossificationMisdirected, DislocatedUnaligned© 2006 by Taylor & Francis Group, LLC

TERMINOLOGY OF ANATOMICAL DEFECTS 965Appendix 1-ATermShortSmallSplitSupernumeraryThickenedUnossifiedDescriptive terminology used more than once in the glossary (continued)DefinitionLess than the normal or expectedlengthIncompletely developed structure, orless than normal in sizeDivision of a single structure (usuallyinto two parts) with no interveningstructure between the partsMore than the usual or expectednumberA widening of a skeletal elementrelative to normalAbsence of ossification (asassessed by Alizarin Red uptake)at a site which, in controls of thesame age, is usually at leastpartially ossifiedSynonym orRelated termRudimentaryAplastic, Hypoplastic,Reduced,RudimentaryBifid, Bipartite, CleftAdditional,ExtraNon-preferred TermUnderdevelopedAccessoryBulbous, ClubbedNonossifiedAppendix 1-BCaudal dysplasiaCheilognathopalatoschisisCheilognathoschisisCraniorachischisisEthmocephalyOtocephalyRhinocephalySirenomeliaTetralogy of FallotSyndromes and combining termsSevere reduction of caudal structures, including reduction of or absence ofhindlimbs, tail, and/or sacral area.Cleft lip, jaw, and palate; also called CheilognathouranoschisisCleft lip and jawFailure of the neural tube to close in regions of both the brain and spinal cord.Some degree of cyclopia in which the eyes may be closely set but the nose ishypoplastic (rudimentary).Extreme underdevelopment of the lower jaw, permitting close approximation orunion of the ears on the anterior aspect of the neck.Proboscis-like nose above partially of completely fused eyes.Any of several degrees of side-to-side fusion of lower extremities and concomitantmidline reduction of the pelvis. Soft tissues and long bones, lower paw (feet), andviscera of the pelvis tend to be reduced or absent; anus and external genitaliaare often absent.Defect of the heart which includes all of the following: pulmonary stenosis,interventricular septal defect, dextraposed aorta overriding the ventricular septum,and enlarged right ventricular wall.© 2006 by Taylor & Francis Group, LLC

APPENDIX BBooks Related to Developmental andReproductive ToxicologyRonald D. HoodPlease note: It is hoped that the following list will be useful and informative, but it contains onlythose references that I could readily discover. Therefore, it is virtually certain that potentially usefulbooks have inadvertently been omitted. Further, the list contains a number of books that I have notreviewed. Thus, I cannot vouch for their accuracy or merit. Only books published in 1989 or laterhave been included, and several are no longer in print, although they may still be available inlimited quantities.Abel, E. L., Behavioral Teratogenesis and Behavioral Mutagenesis: A Primer in Abnormal Development,Plenum, New York, 1989.Abel, E. L., Ed., Fetal Alcohol Syndrome: From Mechanism to Prevention, CRC Press, Boca Raton, FL, 1996.Boekelheide, K., Gandolfi, A.J., Harris, C., Hoyer, P.B., McQueen, C.A., Sipes, I. G., and Chapin, R., Eds.,Comprehensive Toxicology, vol. 10: Reproductive and Endocrine Toxicology, Pergamon Press, NewYork, 1997.Briggs, G. G., Freeman, R. K., and Yaffe, S. J., Eds., Drugs in Pregnancy and Lactation, 7th ed., LippincottWilliams & Wilkins, Philadelphia, PA, 2005.Chapin, R. E., Heindel, J. J., Tyson, C. A., Witschi, H. R., Eds., Methods in Toxicology, Part A: MaleReproductive Toxicology, Academic Press, San Diego, CA, 1993.Daston, G. P., Molecular and Cellular Methods in Developmental Toxicology, CRC Press, Boca Raton, FL,1996.Frazier, L. M. and Hage, M. L., Reproductive Hazards of the Workplace, John Wiley & Sons, New York, 1997.Golub, M. S., Metals, Fertility, and Reproductive Toxicity, CRC Press, Boca Raton, FL 2005.Harry, G. J., Ed., Developmental Neurotoxicology, CRC Press, Boca Raton, FL, 1994.Haschek, W. M., Rousseaux, C. G., and Wallig, M. A., Eds., Handbook of Toxicologic Pathology, 2nd ed.,Vol. 2, Academic Press, San Diego, CA, 2001.Holladay, S. D., Ed., Developmental Immunotoxicology, CRC Press, Boca Raton, FL, 2004.Hood, R. D., Ed., Developmental Toxicology — Risk Assessment and the Future, Van Nostrand Reinhold, NewYork, 1990.Hood, R. D., Ed., Handbook of Developmental Toxicology, CRC Press, Boca Raton, FL, 1997.Hoyer, P. B., Ed., Ovarian Toxicology (Target Organ Toxicology Series), CRC Press, Boca Raton, FL, 2004.Kavlock, R. J. and Daston, G. P., Eds., Drug Toxicity in Embryonic Development. I: Advances in UnderstandingMechanisms of Birth Defects: Morphogenesis and Processes at Risk (Handbook of ExperimentalPharmacology, vol. 142/2), Springer Verlag, Berlin, 1997.967© 2006 by Taylor & Francis Group, LLC

968 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONKimmel, C. A. and Buelke-Sam, J., Eds., Developmental Toxicology (Target Organ Toxicology Series), 2nded., Raven Press, New York, 1994.Kochhar, D. M., In Vitro Methods in Developmental Toxicology: Use in Defining Mechanisms and RiskParameters, CRC Press, Boca Raton, FL, 1989.Korach, K. S., Reproductive and Developmental Toxicology, Marcel Dekker, New York, 1998.Koren, G., Maternal-Fetal Toxicology: A Clinicians’ Guide, 2nd ed., Marcel Dekker, New York, 1994.Lumley, C. E. and Walker, S. R., Eds., Current Issues in Reproductive and Developmental Toxicology: Canan International Guideline Be Achieved? (CMR Workshop Series), Quay Publishing, Lancaster, 1991.McLachlan, J. A., Guillette, L. J., and Iguchi, T., Eds., Environmental Hormones: The Scientific Basis ofEndocrine Disruption (Annals of the New York Academy of Sciences, V. 948), New York Academyof Sciences, New York, 2001.NAS Committee on Hormonally Active Agents in the Environment, National Research Council, HormonallyActive Agents in the Environment, National Academy Press, Washington, DC, 1999.National Research Council, Committee on Developmental Toxicology, Board on Environmental Studies andToxicology, Commission on Life Sciences, Scientific Frontiers in Developmental Toxicology and RiskAssessment, National Academy Press, Washington, DC, 2000.National Research Council, Subcommittee on Reproductive and Developmental Toxicology, Committee onToxicology, Board on Environmental Studies and Toxicology, Commission on Life Sciences, EvaluatingChemical and Other Agent Exposures For Reproductive and Developmental Toxicity, NationalAcademy Press, Washington, DC, 2001.Needleman, H. L. and Bellinger, D., Prenatal Exposure to Toxicants: Developmental Consequences, JohnsHopkins University Press, Baltimore, MD, 1994.Neubert, D., Risk Assessment of Prenatally-Induced Adverse Health Effects, Springer Verlag, New York, 1993.Olshan, A. F. and Mattison, D. R., Eds., Male-Mediated Developmental Toxicity, Plenum, New York, 1994.Paul, M. and Paul, S., Eds., Occupational and Environmental Reproductive Hazards: A Guide for Clinicians,Lippincott, Williams & Wilkins, Philadelphia, PA, 1993.Rama Sastry, B.V., Ed., Placental Toxicology, CRC Press, Boca Raton, FL, 1995.Richardson, M., Ed., Reproductive Toxicology, VCH Publishing, Weinheim, 1993.Robaire, B. and Hales, B. F., Advances in Male Mediated Developmental Toxicity (Advances in ExperimentalMedicine and Biology, V.), Plenum, New York, 2003.Schardein, J. L., Chemically Induced Birth Defects, 3rd ed., Marcel Dekker, New York, 2000.Scialli, A. R. and Clegg, E. D., Reversibility in Testicular Toxicity Assessment, CRC Press, Boca Raton, FL,1992.Scialli, A. R. and Zimaman, M. J., Reproductive Toxicology and Infertility, McGraw Hill, New York, 1993.Scialli, A. R., A Clinical Guide to Reproductive and Developmental Toxicology, CRC Press, Boca Raton, FL,1992.Scialli, A. R., Lione, A., and Padgett, G. K. B., Reproductive Effects of Chemical, Physical, and BiologicAgents: Reprotox, Johns Hopkins University Press, Baltimore, MD, 1995.Slikker, W. and Chang, L. W., Eds., Handbook of Developmental Neurotoxicology, Academic Press, San Diego,CA, 1998.Sullivan, F. M., Watkins, W. J., and van der Venne, M. Th., Eds., The Toxicology of Chemicals: Series 2 -Reproductive Toxicology (Industrial Health and Safety), European Communities/Union, Office forOfficial Publications of the EC, Luxembourg, 1993.Thiel, R. and Klug, S., Methods in Developmental Toxicology and Biology, Blackwell Publishers, Boston, 1997.Thorogood, P., Embryos, Genes and Birth Defects, John Wiley & Sons, Chichester, 1997.Tyson, C. A., Heindel, J. J., and Chapin, R. E., Eds., Methods in Toxicology, Part B: Female ReproductiveToxicology, Academic Press, San Diego, CA, 1993.Witorsch, R. J., Ed., Reproductive Toxicology, 2nd ed., Raven Press, New York, 1995.World Health Organization, Principles for Evaluating Health Risks to Reproduction Associated with Exposureto Chemicals (Environmental Health Criteria), World Health Organization (WHO), Geneva, 2001.© 2006 by Taylor & Francis Group, LLC

APPENDIX CPostnatal Developmental MilestonesAPPENDIX C-1*SPECIES COMPARISON OF POSTNATALBONE GROWTH AND DEVELOPMENTTracey Zoetis, 1 Melissa S. Tassinari, 2 Cedo Bagi, 2 Karen Walthall, 1 and Mark E. Hurtt 2,31Milestone Biomedical Associates, Frederick, MD 21701, USA2Pfizer Global Research & Development, Groton, CT 06340, USA3Correspondence to: Dr. Mark E. Hurtt, Pfizer Global Research & Development, Drug Safety Evaluation, Eastern PointRoad, Mailstop 8274-1306, Groton, CT 06340. 860-715-3118. Fax (860) 715-3577. Email: mark_e_hurtt@groton.pfizer.com1 IntroductionThe purpose of this review is to identify the critical events in human postnatal bone growth and tocompare human bone development to that of other species. Bone is a dynamic connective tissue whosepostnatal development in all species reflects its multiple tasks within the body. The two major functionsof the skeleton in all mammals are 1) to provide mechanical support as a part of the locomotor apparatus,and 2) to ensure the safe environment of the vital organs of the body. In addition, bones are involvedin metabolic pathways associated with mineral homeostasis serving as a reservoir for calcium andphosphorus. Also, bone matrix is a repository for many growth factors and cytokines that are releasedlocally and systemically during the process of bone resorption (autocrine, paracrine and endocrinefunction). Finally, bones provide safe and steady environment for bone marrow.Important characteristics of postnatal bone growth are expressed in different types of bone.Each bone in our body has a unique shape that reflects a specific function of that bone and timewhen this function starts to develop. Growth patterns of specific bones that exhibit importantcharacteristics of postnatal bone development are used to illustrate these characteristics. Thus, thisdiscussion focuses on the postnatal growth of one bone of the upper limb (humerus – non weightbearing bone), one bone of the lower limb (femur – weight bearing bone) and one bone from thecraniofacial region (mandible – non weight bearing bone with intermittent mechanical load).* Source: Zoetis, T., Tassinari, M. S., Bagi, C., Walthall, K., and Hurtt, M. E., Species comparison of postnatal bone growthand development, Birth Defects Research, Part B: Developmental and Reproductive Toxicology, 68, 86-110, 2003.969© 2006 by Taylor & Francis Group, LLC

970 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe following events are critical milestones in postnatal bone growth and development:• The appearance of secondary ossification centers• Longitudinal bone growth at the epiphyseal growth plate• The fusion of secondary ossification centers• Diametric bone growth• Bone vascularityThis review begins with a general discussion of the composition and structure of bone, discussesthe appearance and fusion of secondary ossification centers, details the role of the epiphyseal growthplate and metaphysis in bone growth, discusses diametric increases in the diaphysis as well asformation and remodeling of osteons in human compact bone, and emphasizes the importance ofbone vascularity in postnatal growth and development.1.1 Overview of Postnatal Bone GrowthThe earliest evidence of bone development in humans is marked by the formation of mesenchymalcell clusters during gestation weeks 4 and 5 (1). In general, bone is formed either by endochondralbone formation or ossification (arising from cartilage) and by intramembranous bone formation orossification (arising from membranous tissue) (2).For example, cranial bones such as the mandible are formed by intramembranous bone formation.During mandibular development, soft connective tissue is replaced directly by bone tissue.Similarly, during prenatal development, long bones develop by intramembranous formation; however,the soft connective tissue is first replaced by cartilage, which is replaced by bone throughendochondral bone formation during postnatal development (2).Long bones grow by means of a structure called the epiphyseal or cartilagenous growth platethat is formed after the appearance of secondary ossification centers. The growth plate is a thindisk of cartilage that develops between the metaphysis and epiphysis. Under normal conditions,the growth plate facilitates lengthening of long bones throughout childhood and adolescence andstops functioning when adult height is attained. At the growth plate, endochondral bone formationoccurs, i.e. layers of cartilage are successively replaced by new layers of bone. As a result, the growthplate and epiphysis grow away from the central portion of the bone depositing calcified tissue into thetransitional metaphyseal region which gradually increases the length of the diaphyseal shaft (2).As bones grow in length, they must also grow in diameter to support the increased load.Diametric increases in bone occur by periosteal apposition. Unlike longitudinal bone growth,periosteal apposition can occur at any time during development and is a form of intramembranousbone formation that involves the osteogenic fibrous membrane called the periosteum.Remodeling of immature bone also occurs during postnatal development. Initially, the boneformed in newborns consists primarily of woven or immature bone. As development continues, thewoven bone is remodeled into mature adult bone. In compact bone of humans and other largemammals, primary osteons are characteristically found in immature bone whereas secondaryosteons are found in maturing bone.Normal postnatal bone growth and development can only occur with an adequate vascularsupply. Bone vascularity plays a vital role in the successful growth and development of both humanand animal bones. Thus, to ensure proper nutrition of bone, specialized blood vessels supply variousareas of a growing bone.2 Bone Structure and CompositionBone is a highly organized structure at both the macroscopic and microscopic levels. Macroscopicobservation of bone reveals two different forms of tissue, cortical or compact bone and spongy or© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 971OsteonCompact bone(Osteon, Haversian system)Cement lineHaversian canalConcentric lamellaeCollagen fibersResorptioncavitiesInterstitiallamellaeInnercircumferentiallamellaePeriostealvesselLacunaCanaliculiCancellousbone(Trabeculae)OutercircumferentiallamellaeVolkmann's canalConcentric lamellaeFigure 1 (6) Schematic view of a part of the proximal shaft of a long bone containing cortical (left) and trabecular(right) bone. The cortical bone is composed mainly of Haversian systems or osteons of concentric lamellae withinner and outer circumferential lamellae at the periphery. Many completed osteons can be seen, each with itscentral Haversian canal, cement line, concentric lamellae, canaliculi, and lacunae. Note the orientation of thecollagen fibers in successive lamellae in the protruding osteon. Cross connections between osteons are establishedby Volkmann’s canals. Volkmann’s canals connect with periosteal vessels and the marrow cavity and,unlike the Haversian canals, lack a covering of concentric lamellae. Interstitial lamellae are seen betweencompleted Haversian systems. Three resorption cavities are also depicted.cancellous bone (3, 4, 5, 6). In general, cancellous bone is composed of a lattice of thin bony rods,plates, and arches that form the spaces occupied by bone marrow. Macroscopically, compact boneappears as a solid continuous mass; however, microscopic examination reveals that compact boneis made up of densely packed tubular structures called osteons (3). In the adult human skeleton,compact bone makes up about 80% of total bone mass.Mature compact and cancellous bone are both made up of lamellae, or stacks of parallel orconcentrically curved sheets (Figure 1). Each lamella is about 3 to 7 µm thick and is composed ofhighly organized, densely packed collagen fibers (6, 4). There are 4 general types of lamellar bone(trabecular lamellae, inner and outer circumferential lamellae, and lamellae of osteons), three ofwhich are found only in compact bone (4).Cancellous or spongy bone is made up of trabecular lamellae. Trabecular lamellae form thespaces in spongy bone that are filled with bone marrow (Figure 1). Also housed in trabeculae areosteocytes that are connected by tiny branching tubular passages called canaliculi. The canaliculiprovide a channel for the exchange of metabolites between osteocytes and the extracellular space.Outer and inner circumferential lamellae are found only in compact bone (Figure 1). Outercircumferential lamellae are located directly underneath the periosteum extending around the entirecircumference of the bone shaft. Inner circumferential lamellae are located on the inner surface of© 2006 by Taylor & Francis Group, LLC

972 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONEpiphysisGrowth plateCancellousboneMetaphysisCortical boneEndosteumCiaphysisPenosteumFused growthplateFigure 2 (6) Schematic diagram of a tibia. The interior of a typical long bone showing the growing proximalend with a growth plate and a distal end with the epiphysis fused to the metaphysis.the shaft adjacent to the endosteum (6). Finally, in mature compact bone circular rings of lamellaeare arranged concentrically around longitudinal vascular channels called Haversian canals. Thevascular channel and the concentric lamellae form an osteon or Haversian system. The outer limitsof osteons are defined by distinct borders called cement lines. Osteons communicate with oneanother, the periosteum, and the bone marrow via oblique channels called Volkman’s canals (3, 6).Compact and cancellous bone also contain regularly spaced small cavities called lacunae thateach contain an osteocyte (Figure 1). Extending from each lacuna in a 3 dimensional fashion arebranching tubular passages called canaliculi that connect with adjacent lacuna (6, 3).2.1 Long BonesThe cylindrical shaft of a long bone is called the diaphysis (Figure 2). The diaphysis is made upof a thick wall of compact bone surrounding a central medullary cavity made up of cancellousbone (3). The bulbous ends of long bones in growing humans and animals are called the epiphyses.In mature long bones, the epiphysis is primarily composed of cancellous bone covered by a thinlayer of compact bone. In addition, articular cartilage can be found covering the compact bone atthe uppermost portion of the epiphysis. During postnatal bone growth, epiphyses are separatedfrom the diaphysis by a cartilaginous epiphyseal plate that is connected to the diaphysis by an areaof spongy bone called the metaphysis. Growing bones increase in length in the growth zone thatis formed by the epiphyseal growth plate and the metaphysis.In general, the external surfaces of bone are covered by a layer of fibrous connective tissuecalled periosteum that has the ability to produce bone (Figure 3). The periosteum is made up of 2layers, an outer dense fibrous layer, and an inner vascular cellular layer (4). The inner layer containsosteoblasts and osteoclasts. The periosteum is anchored to bone by bundles of collagen fibers calledSharpey’s fibers.The inner cavities of bone are lined with a layer of osteogenic connective tissue called endosteum(Figure 3). Endosteum is made up of a single layer of flattened osteoprogenitor cells and vascular© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 973InterstitiallamellaeInnercircumferentiallamellaeTrabeculaeof cancellousboneOuter circumferentialIamellaeHaversian systems(osteons)PeriosteumBlood vesselsSharpey'sfibersEndosteumHaversiancanalsFigure 3 (3) Diagram of a sector of the shaft of a long bone, showing the disposition of the lamellae in theosteons, the interstitial lamellae, and the outer and inner circumferential lamellae.connective tissue and is thinner than the periosteum. All of the cavities of bone, including themedullary cavities and haversian canals are lined with endosteum.The long bones include the clavicle, humerus, radius, ulna, femur, tibia, fibula, metacarpals,metatarsals, and phalanges (7).2.2 Short BonesUnlike long bones, short bones do not consist of a diaphysis, metaphysis, and epiphysis. Shortbones are made up of a core of cancellous bone that is surrounded by a thin layer of compact bonethat is covered with periosteum.The short bones include the carpus, tarsus and sometimes the patellae (7).2.3 Flat BonesMany of the flat bones are found in the skull, these include the occipital, parietal, frontal, nasal,lacrimal, and vomer (7). The flat bones of the skull are made up of two layers of compact bonecalled inner and outer plates or tables. A layer of cancellous bone called the diploe separates theseinner and outer plates from one another (8, 3). These flat bones are lined by connective tissue thatis similar to the periosteum and endosteum of long bones. The outer surfaces of the flat bones ofthe skull are covered with a layer of vascular and osteogenic connective tissue called the pericranium,whereas the inner surface of the skull flat bones are lined with a vascular connective tissuecalled the dura mater (3).Other flat bones include the scapula, os coxae, sternum, and ribs (7). These bones are also madeup of two thin layers of compact bone separated by cancellous bone and are covered with periosteum.2.4 Irregular BonesVolkmann'scanalsThe irregular bones of the skeleton are those that due to their shape cannot be grouped with thelong, short, or flat bones. These bones are made up of cancellous tissue that is enclosed in a thinlayer of compact bone covered by the periosteum. Irregular bones include the vertebrae, sacrum,© 2006 by Taylor & Francis Group, LLC

974 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONcoccyx, temporal, sphenoid, ethmoid, zygomatic, maxilla, mandible, palatine, inferior nasal concha,and hyoid (7).2.5 The Cells of BoneMicroscopically, actively growing bone is composed of intercellular calcified material called bonematrix, and four types of cells (osteoprogenitor cells, osteoblasts, osteocytes, and osteoclasts) (3,9, 10, 11, 12). The functions of each of these cells are discussed below.Osteoprogenitor cells are found on or near all surfaces of bone. These cells have the potentialto differentiate into chondroblasts (cartilage cells) or osteoblasts (bone cells). Osteoprogenitor cellsare most active in the postnatal period during the growth of bones, but are also active in adult lifeduring repair of bone injury (3).Osteoblasts originate from local mesenchymal stem cells. Although the primary function ofosteoblasts is the production of bone, these cells have been shown to have a variety of other roles.For example, osteoblasts are involved in aiding bone resorption, organizing collagen fibrils, andsynthesizing matrix components such as type I collagen, proteoglycans, and glycoproteins (3, 10).During embryonic and postnatal growth the periosteum contains an inner layer of osteoblasts, thatare in direct contact with bone.Osteocytes have an important role as a mechanosensors and are associated with local activationof bone turnover. Osteocytes are formed from osteoblasts that become confined within the newlyformed bone matrix during bone formation (10, 11, 9). The exact function of osteocytes is unknown(9). In mature bone, osteocytes reside within cavities called lacunae that are found throughout theinterstitial spaces of bone. Osteocytes have numerous cytoplasmic projections that extend fromcanaliculi into the bone matrix and contact adjacent cells (3, 11).Osteoclasts arise from hematopoietic mononuclear cells in the bone marrow. Osteoclasts arelarge mutinucleated cells that adhere to the surface of bone for the sole purpose of resorption (13,3). Osteoclasts secrete products that lower the pH of the bone surface to aid in resorption ofmineralized bone and cartilage (13).2.6 Bone MatrixThe matrix of bone, occupies more volume than the cells. Bone matrix is composed of approximately65% inorganic material (calcium, phosphorus, sodium, magnesium), > 20% organic material(type I, V and XII collagen, glycoproteins, bone proteoglycans), and > 10% water (3, 4).Bone matrix is formed in 2 stages, deposition and mineralization. During matrix deposition,osteoblasts secrete the initial matrix called osteoid, that is made up of type I collagen, variousproteins, and sulfated glycoaminoglycans. During mineralization, calcium phosphates and carbonatesthat were previously stored in vessicles of cells are released into the matrix where they aredeposited onto collagen fibrils with the help of glycoproteins. Osteoblasts also secrete alkalinephosphatase that is important for mineralization of osteoid.3 OssificationThe process by which bone tissue replaces its precursor connective tissue is called ossification.Ossification can either be intramembranous or endochondral. Intramembranous ossification involvesthe replacement of membranous fibrous tissue by bone, whereas endochondral ossification involvesthe replacement of cartilage by bone (2).In general, ossification begins and is concentrated in a focal area that subsequently expands insize until the previously existing tissue is completely replaced by bone. The initial site of ossificationis called the primary ossification center (2). Most primary ossification centers develop during theembryonic and early fetal period; however, a few also develop postnatally. In some bones (e.g.,© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 975Table 1Age of Appearance and Fusion of Secondary Ossification Centers in the Humerus and FemurAgeProximalEpiphysisAppearanceDistalEpiphysisProximalEpiphysisFusionDistalEpiphysisHumerusHuman Gestation week 36–4 years 6 months–10 years 12–20 years 11–19 yearsMonkey Birth Birth–1 month 4–6 years 1.75–4.5 yearsDog 1–2 weeks 2–9 weeks 10–12 months 6–8 monthsRabbit 1 day 1 day 32 weeks 32 weeksRat 8 days 8–30 days 52–181 weeks 31–158 daysMouse 5–10 days 5–19 days 6–7 weeks 3 weeksFemurHuman 1–12 years Gestation week 36–40 11–19 years 14–19 yearsMonkey Birth–6 months Birth 2.25–6 years 3.25–5.75 yearsDog 1 week–4 months 2–4 weeks 6–13 months 8–11 monthsRabbit 1–5 days 1 day 16 weeks 32 weeksRat 21–30 days 8–14 days 78–156 weeks 15–162 weeksMouse 14–15 days 7–9 days 13–15 weeks 12–13 weekslong bones), the primary ossification center does not expand into the entire precursor tissue area.Secondary centers of ossification develop in regions where primary centers do not extend. Thesesecondary ossification centers are generally formed postnatally (2).Shortly after secondary centers of ossification appear in the epiphyseal region of long bones,the center grows in all directions and results in development of the epiphyseal growth plate. Theepiphyseal growth plate creates a barrier between the epiphysis and the diaphysis. As bone is formedat the growth plate, the epiphysis moves away from the diaphysis and bone is deposited in thetransitional region called the metaphysis located directly below the epiphyseal growth plate (Figure2). As bone is formed at the growth plate, lengthening of the diaphysis occurs. When the longbones have reached their adult length, the epiphyseal growth plate fuses with the diaphysis thusremoving the barrier between these two regions of the bone. Once fusion of the epiphyseal growthplate occurs, longitudinal bone growth can no longer occur. Union of the epiphysis with thediaphysis is also called fusion of secondary ossification centers, or fusion of the epiphyseal growthplate.The number, location, and time of appearance of secondary centers of ossification varies betweenspecies. Also, within species, there is some variation regarding the timing of ossification (14).The following sections include information regarding the ossification of the humerus, femur,and mandible in humans, monkeys, dogs, rabbits, rats, and mice. Table 1 provides an overview ofthe ages at which appearance and fusion of secondary ossification centers occur in the humerusand the femur.3.1 HumanOssification centers that develop prenatally in humans include those in the skull, vertebral column,ribs, sternum, the primary centers in the diaphysis of major long bones, their girdles, and thephalanges of the hands and feet (2). In addition, some primary centers in the ankle and secondarycenters around the knee also develop during the last few weeks of gestation.In humans, postnatal ossification centers develop over a timespan ranging from immediatelyafter birth through early adulthood. Ossification centers that fuse during adolescence include the© 2006 by Taylor & Francis Group, LLC

976 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 2 Timing of Appearance and Fusion of Secondary Ossification Centersin the Human Humerus (2, 19, 22)Region of Humerus Time of Appearance Time of FusionProximal EpiphysisHead Gestation week 36–40or2–6 months after birthGreater TubercleLesser Tubercle3 months–3 years4+ years2–7 yrs: composite proximal epiphysis formed12–19 yrs: proximal epiphysis fused with diaphysisin females15.75–20 yrs: proximal epiphysis fuses with diaphysisin malesDistal EpiphysisCapitulum 6 months–2 years 10–12 yrs: composite distal epiphysis is formedLateral EpicondyleTrochleaMedial Epicondyle10 yearsBy 8 years4+ years11–15 yrs: distal epiphysis fuses with diaphysis in females12–17 yrs: distal epiphysis fuses with diaphysis in males11–16 yrs: fusion with diaphysis in females14–19 yrs: fusion with diaphysis in malesepiphyses of the major long bones of the limbs, those of the hands and feet, and the spheno-occipitalsynchondrosis of the skull. Following adolescence, fusion occurs in the jugular growth plate of theskull, and in the secondary ossification centers of the vertebrae, scapula, clavicle, sacrum, and pelvis (2).The appearance and fusion of ossification centers in the humerus and femur are detailed below.Pre-and postnatal growth of the mandible is also discussed below.3.1.1 HumerusPrimary ossification centers appear in the diaphysis of the humerus during gestation. At birth, 79%of the human humerus is made up of an ossified diaphyseal shaft and 21% is made up of nonossifiedcartilagenous material found primarily in the proximal and distal epiphyses (2).In the newborn infant, the humerus has a rounded proximal end, and a triangular distal region,separated from each other by the diaphyseal shaft. The proximal epiphysis of the humerus developsfrom 3 separate secondary ossification centers: one in the head, one in the greater tubercle, andone in the lesser tubercle. The time of appearance for each of the secondary ossification centers ofthe human humerus is listed in Table 2. Studies have shown that the ossification center in the headusually appears by 6 months after birth, but sometimes develops during weeks 36-40 of gestation(15, 16, 2). Early appearance of the ossification center in the head of the humerus has been shownto be related to birth weight, sex, nationality, maternal history, size, and maturity (2, 15, 16, 17,18). The timing of appearance for the secondary ossification center of the greater tubercle rangedfrom 3 months of age to 3 years of age. In general, it appears earlier in girls than in boys. Theexistence of a third center in the lesser tubercle has been debated in the literature. Some studiesnote that only 2 centers of ossification develop in the proximal epiphysis (head and greater tubercle),others note a third at the lesser tubercle (19, 20). Appearance of the lesser tubercle was noted tooccur between the ages of 4 and 5 years (2).The ages reported for unification of the secondary ossification centers in the proximal epiphysisof the humerus range from 2 to 7 years. Radiologic studies report that each of the secondary© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 977ossification centers of the proximal epiphysis join to form a single compound proximal epiphysisbetween the ages of 5 and 7 years (2). Histological studies demonstrate that a compound proximalepiphysis forms as early as 2 or 3 years of age (19, 2).The proximal epiphysis of the humerus is responsible for up to 80% of the growth in lengthof the diaphyseal shaft (21). When diaphyseal growth is complete, fusion of the proximal epiphysisoccurs. Fusion of the proximal epiphysis has been reported to occur at ages ranging from 12 to 19years in females and 15.75 to 20 years in males (2).Ossification in the distal epiphysis of the humerus occurs via 4 separate secondary centers thatdevelop in the capitulum, medial epicondyle, lateral epicondyle, and trochlea. By the age of 2years, the secondary ossification center of the capitulum is evident; however, this center may alsoappear as early as 6 months after birth (2). The secondary ossification center of the medialepicondyle is usually visible by age 4 but develops slowly thereafter (2). Development of thesecondary ossification center in the trochlea begins with the appearance of multiple foci at the ageof 8 years. Soon after appearance, the trochlear epiphysis becomes joined to the capitulum.Ossification in the lateral epicondyle is evident by the age of 10 years.The secondary centers of the capitulum, trochlea, and lateral epicondyle join with each otherbetween the ages of 10 and 12 years. Fusion of these structures with the diaphyseal shaft beginsposteriorly and leaves an open line above the capitulum, lateral trochlea, and proximal lateralepicondyle which becomes fused at approximately the age of 15 years.The ossification center in the medial epicondyle does not fuse with the capitulum, trochlea, andlateral epicondyle prior to uniting with the diaphyseal shaft. In females and males, fusion has beenseen to occur at ages ranging from 11 to 16 and 14 to 19 years, respectively (22, 2).3.1.2 FemurIn the femur, primary ossification centers of the diaphysis appear during the prenatal period.The epiphysis that is found at the distal end of the femur is the largest and fastest growingepiphysis in the body (2). This secondary ossification center is the first long bone epiphysis thatappears during skeletal development and one of the last to fuse (2). This distal epiphysis thatdevelops from a single center of ossification usually appears during prenatal weeks 36-40; however,some variation in the time of appearance exists. For example, in premature infants, this center isnot always present and some studies have found that the distal femoral epiphysis is sometimesvisible as early as prenatal week 31. Even so, the distal epiphysis is always visible by the postnatalage of 3 months (15, 20, 23, 22).At birth, the distal femoral epiphysis of females is about 2 weeks more advanced than that ofmales. By the time of puberty, girls are about 2 years more developmentally advanced than boys (2).During postnatal months 6-12, the distal epiphyseal plate begins to develop and the epiphysistakes on an ovoid shape (2). During postnatal years 1 to 3, the width of the epiphysis increasesrapidly as ossification spreads throughout the epiphyseal region. When females and males reachages 7 and 9, respectively, the epiphysis is as wide as the metaphysis (2).The distal epiphysis is responsible for approximately 70% of the longitudinal growth of thefemur (24, 25, 2). When growth of the femur is complete, fusion of the distal femoral epiphysisoccurs. Fusion of the distal femoral epiphysis occurs in females and males between the ages of 14to 18 and 16 to 19 years, respectively (2).At the proximal end of the femur, 3 to 4 separate secondary centers of ossification develop.Unlike those in the proximal epiphysis of the humerus, these centers develop and fuse independentlywith either the neck or shaft of the femur. At birth, the proximal epiphyseal growth plate is dividedinto 3 sections, medial, subcapital, and lateral subtrochanteric portions. By the age of 2 years, theneck has grown and divided the epiphyseal region into the head and the greater trochanter. Thelesser trochanter lies below and medial to the epiphyseal region. The head, greater trochanter andlesser trochanter each develop a separate secondary ossification center. The center in the head of© 2006 by Taylor & Francis Group, LLC

978 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 3 Timing of Appearance and Fusion of SecondaryOssification Centers in the Human Femur (2, 22)Region of Femur Time of Appearance Time of FusionProximal EpiphysisHead By year 1 11–16 years: females14–19 years: malesGreater Trochanter 2–5 years 14–16 years: females16–18 years: malesLesser Trochanter 7–12 years 16–17 yearsDistal Epiphysis Gestation weeks 36-40 14–18 years: females16–19 years: malesthe femur appears in most infants between 6 months to 1 year after birth (15, 16, 2). Fusion of thefemoral head has been reported to occur in females and males between the ages of 11 to 16 and14 to 19 years, respectively (22, 2). The secondary center of ossification of the greater trochanterappears between the ages of 2 and 5 years. Fusion occurs at about 14 to 16 years in girls and 16to 18 years in boys (2). Appearance of a secondary ossification center in the lesser trochanter ofthe femur is reported to range from age 7 to age 11. Fusion of the lesser trochanter with the femoralshaft is seen at the age of 16 to 17 years.A summary of the ages at which secondary ossification centers in the femur appear and fuseis provided in Table 3.3.1.3 MandibleThe mandible, or lower jaw, is the second bone in the body to begin ossification. Ossification ofthe human mandible primarily takes place during the prenatal period. The mandible is made up ofa curved, horizontal portion called the body, and two perpendicular portions called the rami (7).In general, the body of the mandible includes the alveolar region which includes deep pockets thathouse teeth later in development, and the symphyseal region that connects the 2 two portions ofthe bone early in development. Each ramus is made up of a condylar process and a coronoid process.Postnatal growth and development of the human mandible — Of all the facial bones, themandible undergoes the most postnatal variation in size and shape. As a result, the perinatalmandible differs greatly in size and shape from the mature mandible.At birth, the mandible consists of two halves that are connected by a fibrous region called thesymphysis. During the first year of life (usually by 6 months), the right and left halves of themandibular body fuse at the midline in the symphyseal region (2). At birth, both the mandible andmaxilla bone are equal in size, but the mandible is located posterior to the maxilla. As rapid postnatalgrowth continues, the mandible assumes its normal position in line with the maxilla (achievingnormal occlusion) (2).Two types of bone growth occur in the mandible, appositional/resorptive and condylar growth.The posterior border of the ramus is an active site of bone apposition and the anterior border ofthe mandibular body is an active site of bone resorption (26). Deposition of bone on the posteriorend of the ramus and resorption of bone on the anterior portion allows the mandibular body tolengthen and make space for developing teeth (2).The condyle has been shown to play a primary role in development of the mandible. Growthof the condyle essentially leads to a downward and forward “displacement of the mandible.” Thisgrowth occurs as cartilage in the condyle is replaced by bone. Condylar growth plays a role indecreasing the mandibular or gonial angle. During the perinatal period, the mandibular angle rangesfrom 135° to 150°; however, soon after birth, it decreases to 130° to 140° (2). In the adult mandible,the gonial angle measures between 110° and 120° (2, 7).© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 979Table 4 Growth and Ossification of the Human Mandible (2)AgePrenatal week 6Prenatal week 7Prenatal week 8About Prenatal week 10Prenatal weeks 12–14Prenatal weeks 14–16BirthPostnatal year 1Infancy and childhoodBy 12–14 yearsEventIntramembranous ossification center develops lateral to Meckel’s cartilageCoronoid process begins differentiatingCoronoid process fuses with main mandibular massCondylar and coronoid processes are recognizable. The anterior portionof Meckel’s cartilage begins to ossifySecondary cartilages for the condyle, coronoid, and symphysis appearDeciduous tooth germs start to formMandible consists of separate right and left halvesFusion of the right and left halves of the mandible at the symphysisIncrease in size and shape of the mandibleEruption and replacement of teethAll permanent teeth emerged except third molarsIn the perinatal mandible, the condyle is essentially at the same level as the superior border ofthe mandibular body. After birth a rapid increase in the height of the ramus occurs. This increasein ramus height results in the condyle being situated on a higher plane than the alveolar surface(the top portion of the mandibular body that is hollowed into cavities for teeth) of the mandibularbody. In addition, during growth, an increase in the height of the mandibular body occurs as aresult of alveolar bone growth (26).As the cranial width increases, resorption and deposition of bone also results in widening ofthe mandibular body. Other changes occurring in the mandible during growth include changes inthe horizontal and vertical position of the mental foramen. Initially, the mental foramen is positionedbelow and between the canine and first molar. Once dentition (cutting of teeth) begins, the foramenmoves under the first molar and is later is positioned between the first and second molar. Theforamen also changes vertical position as the depth of the alveolar process increases (2).The mental proturbance or chin region also changes during postnatal development. Growth ofthe mental proturbance has been shown to occur rapidly from birth until the age of 4 years old.During this time, the depth of the chin increases to make room for the developing roots of theincisor teeth (27, 2).The sequence of growth and ossification in the human mandible is shown in Table 4.3.2 MonkeyWhen the timing of appearance and fusion of secondary ossification centers in rhesus monkeyswas studied, it was noted that more ossification centers were present in monkeys at birth than inman. In general, bone ossification in the newborn rhesus monkey, resembles that of a 5 to 6 yearold human (28). Ossification in the limbs of male and female rhesus monkeys is complete by 5.25and 6.5 years, respectively.In the chimpanzee, the average onset of postnatal ossification in long and short bone centersoccurs about 12 to 20 months earlier, than ossification in man (28).3.2.1 HumerusAt birth in male rhesus monkeys, two separate ossification centers are present in the proximalepiphysis (29). In the epiphyseal region of the distal humerus, one secondary ossification center ispresent at birth (29). Another 2 centers develop in the distal humerus during the first postnatalmonth. At 9 months after birth, the separate centers in the proximal and distal humeral epiphysesunite to form single secondary ossification centers in their respective regions (29).Fusion of the proximal epiphysis to the diaphysis in male and female cynomolgus monkeysoccurs at approximately 6 and 4.75 to 5.5 years, respectively (30, 31). Union of the distal epiphysis© 2006 by Taylor & Francis Group, LLC

980 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 5a Timing of Secondary Ossification Center Appearance and Epiphyseal Fusionin the Rhesus Monkey (29)MaleFemaleBone Time of Appearance Time of Fusion Time of Appearance Time of FusionHumerusProximal Epiphysis Birth 4–6 years Birth 4–5.25 yearsDistal Epiphysis Birth–1 month 2 years Birth 1.75–2 yearsFemurProximalEpiphysis:6 months 2.75–3.75 years 6 months Information notavailableLesser TrochanterHead Birth 3–3.75 years Birth 2.25–3.25 yearsMedial Epicondyle Information not 3–3.75 years Information not 2.25–2.5 yearsavailableavailableDistal Epiphysis Birth 4–5.75 years Birth 3.25–4.25 yearsTable 5b Timing of Secondary Ossification Center Appearance and Epiphyseal Fusionin the Cynomologus Monkey (30, 31)MaleFemaleBone Time of Appearance Time of Fusion Time of Appearance Time of FusionHumerusProximal Epiphysis < 5 months 6 years < 4 months 4.75–5.5 yearsDistal Epiphysis < 5 months 3.4–4.5 years < 4 months 2.25 yearsFemurProximal Epiphysis < 5 months 6 years < 4 months 4.75 yearsDistal Epiphysis < 5 months 5.25 years < 4 months 4.75 yearsto the diaphysis is said to occur in male monkeys between 3.4 and 4.5 years after birth (30, 31).Fusion of the distal epiphysis occurs in female monkeys at approximately 2 years after birth (31).The timing of appearance and fusion of the epiphyses in the humerus of rhesus monkeys andcynomolgus monkeys is shown in Tables 5a and 5b, respectively.3.2.2 FemurIn rhesus monkeys at birth, an epiphyseal ossification center is present in the femoral head (29).Epiphyseal ossification centers in the distal femur are also present at birth. In addition, at 6 monthsafter birth, secondary ossification centers develop in the lesser trochanter of the femur (29).When the femur of female cynomolgus monkeys was examined at 4 months after birth, ossificationcenters were already present in the proximal and distal epiphyses. In male monkeys,ossification centers were present in the proximal and distal epiphyses when examined at 5 monthsafter birth (31).In male cynomolgus monkeys, fusion the proximal and distal epiphyses to the respectivediaphysis occurred in the femur at 6 and 5.25 years of age, respectively. Fusion of both the proximaland distal epiphyses occurred in females at 4.75 years of age (31).The timing of appearance and fusion of the epiphyses in the femur of rhesus monkeys andcynomolgus monkeys is shown in Tables 5a and 5b, respectively.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 9813.2.3 MandibleDuring postnatal growth in the mandible of monkeys, a large increase in overall size of the mandible,occurs as a result of extensive growth and remodeling processes that take place from infancy toadulthood (32). Essentially, bone growth occurs on all surfaces of the primate mandible duringpostnatal development (33, 34).The following discussion includes information taken primarily from a study conducted byMcNamara and Graber (32) to investigate the postnatal development of the mandible in rhesusmonkeys. In this study, 4 groups of monkeys (at various ages) were evaluated for 6 consecutivemonths. The groups included infant monkeys ages 5.5 to 7 months, juvenile monkeys ages 18 to24 months, adolescent monkeys ages 45 to 54 months, and young adult monkeys over the age of72 months (32).In rhesus monkeys, the mandible grew most rapidly during infancy (32). Growth at the mandibularcondyle was most rapid in infant monkeys and successively slowed as animals aged.Specifically, during the 6-month observation period, growth of the condyle in the infant, juvenile,adolescent, and young adult monkeys was 5.92, 4.47, 3.00, and 1.07 mm, respectively.During postnatal development, deposition of bone occurs along the posterior border of theramus and resorption of bone occurs along the anterior border of the ramus. The largest increasein the width of the ramus was seen in infant monkeys (5.5 to 13 months). When compared toanimals over the age of 18 months, the mandibular ramus of infant monkeys had 4 times as muchbone deposited along the posterior border than was resorbed from the anterior border. Bonedeposition on the posterior border of the ramus decreased as the animals aged. Changes in thewidth of the mandibular ramus in the adult monkey were slight (32). In addition, bone depositionoccurred along the posterior border of the condyle (32).In the infant, juvenile, and adolescent animals, remodeling of the mandibular angle was noted;however, remodeling of the mandibular angle was not seen in the adult monkeys (32). During the6-month observation period, the largest change in the mandibular angle occurred in infant monkeys.The average decrease in mandibular angle over the 6 month observation period in infant, juvenile,and adolescent monkeys was 6.2, 2.4, 1.7 degrees, respectively.3.3 DogAppearance of secondary ossification centers and fusion of the proximal and distal epiphyses withthe diaphysis is summarized in Table 6.3.3.1 HumerusThe humerus of the dog develops from five secondary centers of ossification: one found in the diaphysisthat is ossified at birth, one in the head and tubercles of the proximal epiphysis, and three in the distalepiphysis (one each for the medial condyle, lateral condyle, and medial epicondyle) (35, 36).A secondary ossification center appears in the head of the humerus at 1 to 2 weeks after birth (37,35, 38). At 5 months after birth, partial fusion of the epiphysis to the diaphysis begins. Closure of theepiphyseal plate occurs at approximately 10 - 12 months of age; however, in some dogs, the fusionline may still be evident at 12 months, but usually disappears completely by 14 months (35, 37, 30).In the distal epiphysis of the humerus, the ossification centers in the medial and lateral condylesappear during the 2 nd , 3 rd , or 4 th postnatal week (35, 37, 36). The center in the medial epicondyleappears during postnatal weeks 5, 6, 7, 8, or 9 (36, 38, 35). The center of the medial epicondyleunites with the medial condyle during postnatal months 4, 5, or 6, and the whole distal extremityfuses with the diaphyseal shaft during the 6 th , 7 th , or 8 th month after birth (35, 37, 36). At approximately10 months of age, the fusion line in the distal epiphysis of the humerus is barely visible (37).© 2006 by Taylor & Francis Group, LLC

982 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 6 Appearance and Fusion of Ossification Centers in the Dog (36, 35, 38, 25, 37)Bone Age at Appearance Age at FusionHumerusProximal EpiphysisHead 1 to 2 weeks 1,2,3,4 10 to 12 months 1,3Distal EpiphysisMedial condyle 2, 3 or 4 weeks 1,2,3 To lateral condyle 6 to 9 weeks 1Lateral condyle 2 to 3 weeks 1,2,3 See aboveEpicondyle (medial and lateral) 5 to 9 weeks 1,2,3 To condyles 4 to 6 months 1,3 complete toshaft 6 to 8 months 1,3FemurProximal EpiphysisHead 2 to 3 weeks; 1 1 to 3 weeks 2 6 to 10 months 1Greater Trochanter 9 to 10 weeks; 1 2 months 2 8 to 11 months 1Lessor Trochanter 3 to 4 months; 1 7 to 11 weeks 2 10 to 13 months 1Distal EpiphysisTrochlea 2 weeks 1 Trochlea to condyles 3 to 4 months 1Medial condyle 3 weeks; 1,5 2 to 4 weeks 2 Complete to diaphysis at 8 to 11 months; 1Lateral condyle 3 weeks; 1 2 to 4 weeks 2250 to 325 days 51362383354375253.3.2 FemurA slight amount of variation exists in the scientific literature regarding the number of ossificationcenters in the femur of the dog. Andersen and Floyd (25) reports that the canine femur developsfrom the six centers of ossification: one in the diaphysis that is ossified at birth, one each in thehead, greater trochanter, and lessor trochanter of proximal region, and one each in the medial andlateral condyle of the distal epiphysis. Parcher and Williams (36) lists seven centers of ossificationby an additional center in the trochlea noting one in the trochlea. Hare (38) reports five centers ofossification in the femur of the dog, listing only one center in the distal epiphysis.In spite of the variation in the number of ossification centers reported in the literature, reasonableconsistency exists concerning the age at which secondary ossification centers appear. At 7-10 daysafter birth, ossification is evident in the diaphysis of the femur (25). By postnatal week 2 and 3,ossification centers develop in the head of the femur (36, 38). In the greater trochanter and lessortrochanter, secondary ossification centers do not develop until 8 to 10 weeks and 2, 3 or 4 monthsafter birth, respectively (36, 38). Parcher and Williams (36) list that fusion of the epiphyseal platesin the femoral head, greater trochanter, and lessor trochanter occurs at 6 to 10, 8 to 11, and 10 to13 months after birth, respectively.By 21 days after birth in the distal region of the femur, epiphyseal growth plates develop inthe medial and lateral epicondyle (25, 36). Similarly, Hare (38) reports appearance of the epiphysealgrowth plate in the distal epiphysis of the femur at 2, 3, or 4 weeks. Parcher and Williams (36)report that the secondary ossification center of the trochlea appeared one week earlier at 2 weeksafter birth. By 3 and 4 months after birth, it is reported that the epiphyseal growth plate in the© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 983Table 7 Appearance and Fusion of OssificationCenters in the Rabbit (41, 42)Bone Age at Appearance Age at FusionHumerusProximal Epiphysis 1 day 32 weeksDistal Epiphysis 1 day 32 weeksFemurProximal Epiphysis 1–5 days 16 weeksDistal Epiphysis 1 day 32 weekstrochlea begins to fuse with those of the medial and lateral condyles (36). By 8 to 11 months, theepiphyseal growth plate in the distal region of the femur fuses with the diaphysis (36, 25).3.3.3 MandibleDuring prenatal development, the mandible of the dog, is formed by intramembranous ossification.During postnatal development, endochondral bone formation occurs at the condyle; however,elsewhere in the mandible, bone growth occurs by the apposition of bone along the mandibularsurfaces (39).The rate at which bone is added to the mandible is most rapid during the early postnatal period.After dogs reach about 40 to 60 days of age, the rate of bone formation in the mandible plateaus(39, 40). By the age of 6 to 7 months, even though bone apposition and resorption continues inthe mandible, the volume of the mandible remains constant because of a balance between the twoprocesses (39).During postnatal development, the mandible increases in overall size, length, width, and height.The increase in the length that occurs in the mandible of the dog is primarily due to bone formationon the rear portions of both the ramus and the mandibular body (39). The mandibular ramusincreases in width as more bone is deposited at the caudal border of the mandible than is beingresorbed at the rostral border. Resorption at the rostral border of the mandible occurs to makesroom for eruption of the molars. The height of the mandible increases by apposition of bone onthe ventral surface of the mandible, and by growth of alveolar bone (39). A change in the mandibularangle occurs as bone is laid down on the caudo-ventral portion of the mandible (39). Growth ofthe mandible in a downward and backward direction is caused by endochondral bone formation atthe mandibular condyle (39).3.4 RabbitA summary of the appearance of secondary ossification centers and the fusion of the epiphyses ofthe humerus and femur is provided in Table 7.3.4.1 HumerusWhen maturation of secondary ossification centers was studied in the Japanese white rabbit, nosignificant differences between male and female rabbits were noted (41). As early as postnatal Day1, secondary ossification centers were present in the proximal and distal epiphyses of the humerus(41). By 1 week after birth, all secondary ossification centers had appeared in the long bones withthe exception of those in the proximal epiphysis of the fibula, which appeared at 2 weeks of age (41).At 32 weeks of age, the proximal and distal epiphyses of the humerus fused with the diaphyses (41).© 2006 by Taylor & Francis Group, LLC

984 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION3.4.2 FemurAccording to Heikel (42), secondary ossification centers in the head of the femur appear in therabbit by the 4 th or 5 th postnatal day. However, Fukuda and Matsuoka reported that secondaryossification centers were present in the head and distal epiphysis of the femur at 1 day after birth (41).Growth in the rabbit femur occurs rapidly until 3 months after birth at which time growth slowsuntil completely stopping at about 6 months (43). This suggests that fusion of the distal and proximalepiphyseal growth plates occur at approximately 6 months after birth. However, Fukuda andMatsuoka reports fusion of the femoral head and distal epiphysis of the femur occur at approximately16 and 32 weeks after birth, respectively (41).3.4.3 MandibleUnlike the human mandible, the rabbit mandible does not contain a prominent coronoid process,nor do the right and left half of the mandibular body fuse during development (44). The two portionsof the mandibular body remain separated at the junction between the right and left half.In the rabbit, postnatal craniofacial growth occurs most rapidly immediately after birth. Thelength of the rabbit mandible increases rapidly from birth to 16 weeks at which time approximately90% of the adult length is achieved. The largest increase in mandibular length occurs during thefirst 2-6 weeks after birth (45).During postnatal growth in the rabbit mandible, several combined growth processes worktogether to move the ramus in a upward and backward direction. First, the ramus grows in anupward and backward direction via endochondral bone formation that occurs within the condylarcartilage (44). Second, the combination of resorption and deposition on opposite sides of themandible contribute to movement of the ramus in a backward and upward direction (44). Forexample, the lingual side of the ramus (area facing the tongue or oral cavity) mainly exhibits bonedeposition whereas the buccal surface (area nearest to the check) primarily experiences resorptionof bone (44). Thirdly, as bone is being deposited on the posterior border of the ramus, resorptionis taking place on the anterior border of the ramus. This combination of resorption and depositionalso contributes to posterior growth of the mandibular ramus (44).As the ramus moves posteriorly via the growth mechanisms discussed above, elongation of themandibular body occurs. Areas of bone that were once part of the ramus are remodeled and becomepart of the mandibular body (44). In addition, as bone is deposited along the lower border of themandibular angle the angular process enlarges (44).3.5 Rat3.5.1 HumerusJohnson (46) reported that secondary ossification centers appear in the head and greater tubercleof the humerus on postnatal day 8. This account corresponds with that of Fukuda and Matsuoka(47) who reported that secondary ossification centers in the head of the humerus appear duringpostnatal week 2. However, considerable variation exists regarding the timing of epiphyseal fusionin the proximal region of the humerus. Fukuda and Matsuoka (47) reported that fusion of thesecondary ossification centers in the proximal epiphysis of the humerus is complete by 52 weeksafter birth; however, Dawson (48) noted that the epiphysis in the head of the humerus did not unitewith its diaphyseal shaft until 162-181 weeks after birth.Fukuda and Matsuoka (30, 47) acknowledge the existence of other reports in the literatureregarding incomplete epiphyseal fusion in aged mice, rats, and rabbits. This incomplete fusionleaves the epiphyseal line visible for quite some time after fusion has begun and progressed farenough to ensure that longitudinal growth of the bone is no longer occurring. Fukuda and Matsuoka© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 985Table 8 Appearance and Fusion of Postnatal Secondary OssificationCenters in the Rat (47, 48, 46)BoneHumerus:Time of Appearance(day/week after birth)Time of FusionProximal EpiphysisGreater Tubercle 8 days 1 52 weeks 2Head 8 days 1 , 4 weeks 2 52 weeks 2 , 162–181 weeks 3Distal EpiphysisCapitellum 8 days 1 4 weeks 2 , 31–42 days 3Trochlea 8 days 1 4 weeks 2 , 31–42 days 3Medial Condyle 30 days 1 4 weeks 2 , 31–42 days 3Lateral Condyle 30 days 1 4 weeks 2 , 31–42 days 3Epicondyle Information not available 130–158 days 3FemurProximal EpiphysisGreater Trochanter 30 days 1 , 4 weeks 2 78 weeks 2 , 143–156 weeks 3Lesser Trochanter 21 days 1 Information not availableHead 21 days 1 , 4 weeks 2 104 weeks 2 , 143–156 weeks 3Distal EpiphysisMedial Condyle 8 days 1 , 2 weeks 2 15–17 weeks 2 , 162 weeks 3Lateral Condyle 8 days 1 , 2 weeks 2 15–17 weeks 2 , 162 weeks 3146247348(41) noted that the incomplete fusion seen in some long bones of aged mice, rats, and rabbits mightbe a common occurrence. He notes that in these cases, if the incomplete fusion remains for anextended period of time, then the earlier timepoint of partial fusion would be accepted as the“complete” fusion of the epiphysis (47, 30).In the distal epiphysis of the rat humerus, secondary ossification centers appear in the capitulum,and trochlea during week 2 after birth, specifically, on postnatal day 8 (46, 47). Dawson (48) reportsthat by postnatal day 15, a single well-defined secondary center of ossification can be seen in thecapitulum and trochlea. Fukuda and Matsuoka (47) report that fusion is noted in the distal epiphysisof the humerus by week 4. The epiphysis that develops in the distal region of the humerus is thefirst long bone epiphysis to unite with its corresponding diaphyseal shaft (48). By day 42, theepiphyseal growth plate of the distal humerus is no longer visible, indicating that epiphyseal fusionhas occurred. Dawson (48) reported that in the epiphysis of the capitulum and trochlea, thecharacteristic columnar arrangement of cartilage cells within the growth plate was not seen. As aresult it was postulated that in this case, the typical sequence of events is not followed due to therapid rate of epiphyseal fusion (48).The appearance and fusion of secondary ossification centers in the humerus is summarized inTable 8.3.5.2 FemurSecondary ossification centers appear in the greater trochanter and head of the femur at 4 weeksafter birth (47). According to Fukuda and Matsuoka (47) and Dawson (48), fusion occurred in the© 2006 by Taylor & Francis Group, LLC

986 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONgreater trochanter at 78 and 143-156 weeks respectively, and in the head of the femur at 104 and143-156 weeks, respectively.At the distal end of the femur, secondary ossification centers appear at 2 weeks after birth (47).Fusion of the distal epiphysis occurs between 15 and 17 weeks as reported by Fukuda and Matsuoka(47) and at 162 weeks as reported by Dawson (48). The discrepancy noted here regarding the timeof fusion could be due to the fact that by 15 weeks, all of the secondary ossification centers in thefore and hind limbs of the rat had begun fusion, but all had not completed the fusion process (47).For example, it was also reported that fusion of secondary ossification centers in the distal epiphysisof the radius, both epiphysis of the ulna, the proximal epiphyses of the tibia and fibula had notcompleted fusion even at 134 weeks of age (47).The appearance and fusion of secondary ossification centers in the femur is summarized inTable 8.3.5.3 MandiblePrenatal — Ossification of the rat mandible begins in the fetus during prenatal week 15 or 16.The mandible of the rat becomes ossified from a single center of ossification that allows bonegrowth in various directions and results in formation of the mandible. This ossification centersappears lateral to Meckel’s cartilage, in the same general area as the first molar tooth germ (49).Postnatal — Bone deposition and resorption occurs rapidly in the rat immediately after birth andthroughout the early postnatal period. These activities play an important role in the growth anddevelopment of the mandible. At one week after birth, bone is rapidly being deposited in variousareas of the rat mandible (50). For example, resorption is occurring at the incisor socket makingway for the incisor tooth that will erupt at 2 weeks.At birth, the mandibular body of the rat consists of 2 separate halves joined by the symphysis.At one day after birth a wedged shaped cartilaginous bar is present in the symphyseal region ofthe rat mandible (51). At seven days after birth, the cartilage elongates posteriorly. At 14 days afterbirth, the symphyseal cartilage is still present; however, endochondral bone has begun to form inthis region (50). By 15 days after birth, sealing of the cartilaginous border begins. By 21 days afterbirth, the cartilaginous border of the symphysis is completely sealed (51).During the first postnatal week, bone begins to be remodeled in the developing mandibularramus (50). Growth that occurs at the mandibular condyle contributes to the normal growth of themandible, particularly to the growth of the ramus (52). Specifically, increases in mandibular lengthand height that occur during postnatal growth are primarily due to the directional growth of themandibular condyle and the angular process (50). At birth, the condyle is positioned above andslightly lateral to the angular process (50). During the first 3 days after birth, both the condylarand angular processes grow primarily in a backward direction (49). This backward growth contributesto the lengthening of the mandibular ramus. However, growth of the condylar cartilage has aslight dorsal and lateral component and angular cartilage growth also has a slight ventral component.These components not only facilitate increases in length, but also cause the divergence of thecondyle and angular process from one another in the vertical and horizontal planes. As a result ofthis divergence, from postnatal day 3 forward, the condylar cartilage changes its direction of growth,and begins to primarily grow upward, though also continuing backward and lateral growth (49).Up to about 12 days after birth, the increased in the width of the ramus is due to progressiveincreases in the circumference of the condylar and angular cartilages. From postnatal day 12 to15, the portion of the ramus that lies just behind the molar region increases rapidly in width dueto the enlargement of the incisal socket and the posterior growth of the incisor. From 15 to 30 daysafter birth, growth continues ventromedially and backward (49).At approximately 2 weeks after birth, the coronoid process of the rat mandible is very thin andbone begins to be deposited on its surfaces. As bone is deposited, the coronoid process grows© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 987Table 9 Appearance and Fusion of PostnatalSecondary Ossification Centers in the Mouse (30, 46)Bone Appearance FusionHumerusProximal EpiphysisGreater TubercleHeadDistal EpiphysisCapitellumTrochlearMedial CondyleLateral CondyleFemurProximal EpiphysisGreater TrochanterLesser TrochanterHeadDistal EpiphysisMedial CondyleLateral Condyle5 days7–10 days5 days5 days9 days19 days14 days14 days15 days7 days9 days6–7 weeks3 weeks13, 14, or 15 weeks12 or 13 weeksbackward, and increases both in size and height. By 10 weeks after birth, the coronoid process iscomposed of compact bone and exhibits little bone forming activity (50).By three months after birth, bone is no longer being deposited on the surfaces of the mandiblebut internal bone remodeling continues. Specifically, bone growth at the condylar cartilage decreasessubstantially at 6 weeks after birth and by 3 months after birth the angular process is composedof dense inactive bone (50).By 14 weeks after birth, growth of the rat mandible has essentially ceased (50).3.6 MouseWhen compared to the rat or human, ossification in the mouse skeleton occurs over a relativelyshort period of time (14). A summary of the maturation of secondary ossification centers in thehumerus and femur of mice is provided in Table 9.3.6.1 HumerusIn the proximal epiphysis of the humerus, secondary ossification centers develop in the greatertubercle, on postnatal day 5 (46). Shortly thereafter, these secondary ossification centers unite andform a single broad plate of bone in the epiphysis (46). On postnatal day 7, an ossification centerappears in the head of the humerus. By postnatal day 17, the centers in the head and in the greatertubercle unite and form a single secondary ossification center in the proximal epiphysis (46). Fusionof the proximal epiphysis to the diaphysis occurs at approximately 6 to 7 weeks after birth (30).In the distal epiphysis, the medial and lateral condyles develop separate centers of ossificationat 9 and 19 days after birth, respectively. In addition, a secondary center of ossification first appearsin the trochlea and capitulum at 5 days after birth (46). Union of the distal epiphysis of the humeruswith the diaphysis usually occurs in the mouse at 3 weeks after birth (30).© 2006 by Taylor & Francis Group, LLC

988 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION3.6.2 FemurIn the proximal region of the femur, by day 14 after birth ossification centers develop in the greaterand lesser trochanters. The head of the femur exhibits a small ossification center on the 15 dayafter birth (46).In the femur, at 7 days after birth, a secondary ossification center appears in the medial condyle atthe distal portion of the femur. At 9 days after birth, the lateral condyle develops a secondary ossificationcenter. By postnatal day 13, these centers form a single broad plate of bone in the distal epiphysis (46).At the end of the first postnatal month, the proximal and distal epiphyses of the femur arealmost completely ossified; however, they have not yet fused with the diaphysis (46). Fusion ofthe proximal epiphysis occurs approximately between 13 and 15 weeks after birth (30). Fusion ofthe distal epiphysis occurs prior to postnatal week 15, approximately between week 12 and 13 (30).3.6.3 MandibleOssification — In the mouse, a mandibular ossification center develops by day 15 of gestation(46). By gestation day 17, the alveolar process of the mandible is relatively well developed andthe mental foramen is fully developed (46). At birth, the coronoid process, the condyloid process,and the mandibular (gonial) angle are apparent. As late as postnatal week 5, cartilage can still beseen in the condylar and coronoid processes and fusion of the anterior end of the mandible has notyet occurred in the symphysis (46).Postnatal growth — During the first 3 postnatal weeks, morphological and histochemical changesare noted in the condylar cartilage of the mouse (53). In addition, the height of the condylar cartilageincreases dramatically during the first 2 postnatal weeks. As a result, the height of the mandibularramus also increases during this period. The width of the condylar cartilage also increases, especiallyduring the 2 nd and 3 rd weeks of life (53).The cartilagenous growth center in the mouse condyle reaches skeletal maturity by postnatalweek 8 and a well-developed region of primary spongiosa is visible below the growth center (53).However, by 8 weeks after birth, the primary spongiosa in the condyle no longer exists, insteadlamellar bone is now in direct contact with the undersurface of the mature cartilage (53).Livne et al., (54) noted that the period of most rapid and active growth of the mouse mandibleoccurs from 0 to 8 weeks after birth and results in appositional growth of cartilage followed byendochondral ossification (54). These results were consistent with those of Silberman and Livne(53) who reported that the cartilagenous growth center in the mouse condyle reaches skeletalmaturity by postnatal week 8. As maturity progresses a well-developed area of primary spongiosais visible below the condylar growth center (53). From this point forward, the mature condylarcartilage no longer serves as a growth center but begins to serve as an articulating surface for thesquamoso-mandibular joint (54).4 Epiphyseal Growth PlateAs briefly discussed in Section 3, soon after the appearance of the secondary ossification centersin the epiphyses of developing long bones, a region containing cartilage, bone, and fibrous tissuedevelops between the epiphysis and the diaphysis. This region is called the epiphyseal growth plateand is responsible for increases in diaphyseal length that occur during postnatal development (55,6). As bone is produced at the epiphyseal growth plate a transitional region between the diaphysisand the epiphysis is created called the metaphysis (6, 2).In mammals, during postnatal development, longitudinal growth of long bones occurs at theepiphyseal growth plate by the process of endochondral ossification (56, 57, 58, 59, 60, 61). Ingeneral, this process involves production of cartilagenous matrix by proliferating chondrocytes,mineralization of the cartilage matrix, removal of calcified and uncalcified matrix, and the replacement© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 989Reserve ZoneProliferative ZoneZone of MaturationZone of DegenerationHypertrophicZoneZone ofProvisionalCalcificationLast intact transverse septumFigure 4a (70)Drawing depicting the various zones of the cartilaginous portions of the growth plate.of matrix by bone (61). When the epiphyseal growth plate is active, the above-mentioned processesare continually occurring in the growth plate and results in deposition of bone in the metaphysis.When the rate of bone formation begins to exceed the rate of cartilage proliferation, the epiphysealplate begins to narrow, and epiphyseal fusion occurs and results in the disappearance of theepiphyseal growth plate and its replacement with bone and marrow. In both humans and animals,epiphyseal fusion marks the end of longitudinal bone growth (62, 2).The epiphyseal growth plates of human bone are structurally and functionally similar to theepiphyseal growth plates that develop in the bones of other mammals (63, 64, 60, 65, 57, 66, 67,68, 69). Thus, the following description of the structure of the growth plate is applicable to mammalsin general.The growth plate is divided into several functional zones, the names of which have not beenstandardized throughout the scientific literature (Figure 4a-b). The zone closest to the epiphysisand farthest from the diaphysis is called the resting, germinal, or reserve zone (2, 70). Chondrocytesin this area are small, quiescent, and randomly distributed. It is thought that the cells in the restingzone store lipids, glycogen, and other material (70, 71). Adjacent to the germinal zone is theproliferative zone. This zone is dedicated to rapid, ordered cell division and matrix production. Inthe proliferative zone, chondrocytes increase in size (accumulating glycogen in their cytoplasm),exhibit mitotic division, synthesize and secrete matrix, and become arranged in columns. Directlybeneath the proliferative zone is the zone of cartilage transformation, which is sometimes furtherdivided into upper and lower hypertrophic or maturation and degeneration zones. In the zone ofcartilage transformation, chondrocytes hypertrophy (swell), preparing for their replacement by bone,and continuing their synthesis of matrix (70, 71, 2). Many of the chondrocytes in the zone ofcartilage transformation progressively degenerate and their intracellular connections are removedby the advancing metaphyseal sinusoidal loops (2). In the zone of ossification also called the zoneof provisional calcification, hyaline matrix becomes calcified. Calcification of the matrix results inrestricted diffusion of nutrients and the eventual death of hypertrophic chondrocytes. In this finalregion of the epiphyseal growth plate, osteoblasts form a layer of bone on the remaining mineralizedcartilage (2, 70, 71). Additionally, invading metaphyseal blood vessels bring nutrients and osteogeniccells into the zone of ossification to aid in the formation of bone.© 2006 by Taylor & Francis Group, LLC

990 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONProliferatingcartilage cellsHypertrophiccartilage cellsProvisionalcalcificationInvasion orcartilagePrimaryspongiosaSecondaryspongiosaABFigure 4b (3) Endochondral ossification in longitudinal sections of the zone of epiphyseal growth of the distalend of the radius in a puppy. (A) Neutral formalin fixation, no decalcification, von Kossa and hematoxylin-eosinstains. All deposits of bone salt are stained black; thus, bone and calcified cartilage matrix stain alike. (B) Zenkerformolfixation, specimen decalcified, and stained with hematoxylin-eosin-azure II.Discussed below is the rate of longitudinal growth in both humans and animals during postnataldevelopment as a result of cell proliferation at the epiphyseal growth plate.4.1 HumanIn humans, skeletal growth is initiated by the formation of a cartilage template that is subsequentlyreplaced by bone (72). Linear growth, particularly of the long bones, commences with the proliferationof epiphyseal cartilage cells and ceases with epiphyseal closure at puberty. In general anincrease in rate of growth of the long bones occurs between the ages of 9 and 14 (43).4.2 DogIn dogs, during postnatal development, increases in diaphyseal length occur by proliferation andmaturation of chondrocytes at the epiphyseal growth plate followed by calcification of matrix andendochondral bone formation (69, 60).In general, rapid growth of the limb bones has been shown to cease by approximately 5 monthsafter birth (37).4.3 RabbitThe structure and function of the epiphyseal growth plate in the rabbit is similar to that of humans(73, 74, 75). Thus, in the rabbit, increases in the length of long bones also occur as a result ofcellular activity at the epiphyseal growth plate.Fukuda and Matsuoka (41) report that in the rabbit, increases in limb length occur rapidly from0 to 8 weeks after birth. Once animals reach 8 weeks of age, longitudinal growth of the limbsslows until 32 weeks after birth at which time growth ceases completely (41). Khermosh et al.,(43) report that a steady increase in growth rate of the femur was seen from a few days after birth© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 991until 3 months. After 3 months, there is a gradual decrease in growth rate up to 6 months whenthe growth ceases. In later studies, Rudicel et al., (76) report that the length of the rabbit femurrapidly increases between 2 and 4 weeks of age. Growth rate begins to plateau at 8 weeks of ageand by 10 to 14 weeks of age growth essentially ceases.As is true in all mammals, the growth rate of long bones in the rabbit decreases with increasingage (77). When compared to rats and humans, rabbits have the highest longitudinal growth rate/day(77). The rate of longitudinal growth in the rabbit ages 20-60 days is greater than that seen in therat ages 20-60 days, and than that seen in humans ages 2-8 years (77). At 20 days old, the growthrate in rabbits and rats were 554 and 375 µm/day, respectively, when measured in the proximaltibias (77). The growth rate in humans (femur) at 2 years of age was 55 µm/day. By 60 days oldthe growth rate in the rabbit and rat had decreased to 378 and 159 µm/day, respectively. The growthrate in humans at age 5-8 years decreased to 38 µm/day (77).4.4 RatThe structure of the epiphyseal growth plate in the rat is similar to that of the human. In the rat,at about the time of weaning, the epiphyseal growth plate has been formed (67, 68).A rapid increase in length of each long bone of the fore and hind limb has been seen up untilthe rats reach the age of 8 weeks (47). The growth of the long bones in male and female rats slowsalmost to the point of stopping between the age of 15 to 17 weeks (47).5 MetaphysisThe metaphysis is a transitional region of bone that develops between the epiphyseal growth plateand the diaphysis during postnatal growth of the long bones. The structure and function of themetaphysis is similar among mammals (60, 78, 79, 64, 80). The metaphyseal region is one of theactive sites of bone turnover in the long bone of growing mammals (79, 55).In long bones, the metaphyseal region consists of an area of spongy bone directly beneath theepiphyseal growth plate. Together, the epiphyseal growth plate and the metaphysis form the growthzone (3). During postnatal bone growth and development, the composition of the metaphysischanges significantly. Early in development, calcified cartilage is the predominant hard tissue inthe metaphysis; however, as development progresses the metaphyseal tissue is converted to bonewith minimal amounts of cartilage (79).During postnatal development, the cancellous bone of the metaphysis is divided into twodifferent regions, the primary spongiosa and the secondary spongiosa (81, 79, 6). As a long bonegrows in length, the primary spongiosa fills the areas that were previously occupied by the growthplate (6). The primary spongiosa contains many osteoblasts and osteoprogenitor cells and ischaracterized by thin trabeculae made up of bone covered calcified cartilage. Capillary loops at thefront of the primary spongiosa, play a role in introducing osteoprogenitor cells that later formosteoblasts and osteoclasts (60).The tissue outside the primary spongiosa is called the secondary spongiosa. The secondaryspongiosa is mainly composed of bone with small amounts of calcified cartilage (79). As boneformation proceeds at the growth plate, most of the primary spongiosa is converted into secondaryspongiosa (6).Debates have been raised regarding whether rat long bone metaphysis is an appropriate modelto simulate human cancellous bone remodeling. In humans the two processes, mini-modeling andremodeling, are responsible for developing and maintaining cancellous bone mass and architecture.Mini modeling changes trabeculae by removing old bone from one surface and adding new boneon the opposite surface. Remodeling removes pieces of old lamellar bone and refills the space(lacuna) with new lamellar bone (78). Mini-modeling is the mostly occurs before skeletal maturitywhile remodeling occurs mostly after skeletal maturity has occurred.© 2006 by Taylor & Francis Group, LLC

992 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSimilarly, in the metaphysis of rats, the remodeling of primary bone results in secondary lamellarbone (78). Specifically, in the rat proximal tibial metaphysis cancellous bone remodeling is similarto that described in humans once a rat reaches the age when longitudinal bone growth plateaus (78).6 DiaphysisDuring postnatal growth, the diaphysis of long bones increase in length by the process of endochondralbone formation whereas the diaphyseal cortex increases in diameter by intramembranousbone formation (55). In mammals, longitudinal growth must occur prior to fusion of the epiphysiswith the diaphysis; however, bones can increase or decrease in width at any time during life byappositional growth.6.1 Growth in DiameterDuring postnatal growth, in both humans and animals, as bones grow in length, they also growcircumferentially to accommodate the increase in load (82). Long bones increase in diameter byappositional bone growth i.e. deposition of new bone on the periosteal surface (i.e. beneath theperiosteum) (55, 3). During appositional growth, osteoblasts under the periosteum secrete matrixon the outside surfaces of bone. As bone is deposited on the periosteal surface, osteoclasticresorption on the endosteal surface regulates the thickness of the bone forming the cortex of thediaphyseal shaft. The amount of bone in the cortex remains fairly constant throughout growth duethis coupled deposition and resorption. Thus, the overall diameter of the diaphysis increases rapidlywhereas the thickness of the cortex increases slowly (3).During postnatal growth, the shape of a growing bone is maintained by continuous remodelingof the bone surface. This involves bone deposition on some periosteal surfaces and absorption onother sometimes-adjacent periosteal surfaces (3).6.1.1 HumanAt birth, the diameter of the diaphysis in long bones is rapidly increasing in diameter. Kerly (83)reports that in humans at birth resorption by osteoclasts is highest in the mid region of the diaphysealcortex. By four years of age, osteoclastic resorption is occurring throughout the cortex in preparationfor the formation of secondary osteons. Between the ages of 10 and 17 years, as the long bonesundergo rapid growth, osteoclastic resorption and osteoblastic apposition occur extensively at theperiosteal surface of bone to preserve bone shape (83, 3). Once the bone achieves adult size,resorption is greatly decreased and is balanced by osteoblastic deposition of bone in the form ofosteons.Parfitt et al., (84) report that when appositional bone growth was studied in the iliac bone ofhumans ages 1.5 to 23 years, significant increases in the diameter of the diaphysis was seen tooccur with age. The results indicated that between 2 and 20 years of age, the ilium increased inwidth by 3.8 mm by periosteal apposition, and 1.0 mm by endosteal bone formation. During thistime period, resorption at the endosteal and periosteal surfaces accounted for removal of 3.2 and0.4 mm of bone, respectively. As a result of coupled apposition and resorption at the periosteal andendosteal surfaces, the width of the diaphyseal cortex increased from 0.52 mm at 2 years of ageto 1.14 mm by the age of 20 years (84).6.1.2 DogIn the beagle dog at approximately 100 days of age, the medullary cavity of the diaphysis reachesits adult diameter; however, the cortex or outside diameter of the shaft does not stop increasinguntil the animal is approximately 180 days of age (25).© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 9936.1.3 RatIn the femur of Norwegian male rats, growth of the long bones occurred by both diametric andlongitudinal growth coupled with remodeling (85). During the early stages of growth (at approximately30 days old), widening of the marrow cavity occurred by resorption of bone at the endostealsurface of the diaphysis, and an increase in cortical thickness occurred by apposition of new boneat the periosteal surface. Endosteal resorption was highest in the posterior wall of the femur wherethe cortex is visibly curved.6.2 Osteon Formation and Remodeling6.2.1 HumanBone that is added to the cortex of the diaphysis during appositional bone growth is immature orwoven bone (3). Much of the diaphyseal bone of newborns is made up of this nonlamellar wovenbone (86). During development, as the diaphysis expands diametrically, blood vessels becometrapped in the cavities of the bone matrix. In these cavities, woven bone fills the spaces around theblood vessel forming primary osteons or Haversian systems (55).In humans, remodeling of osteons begins shortly after birth and continues throughout life (55,3). Immediately after birth, the substantia compacta of growing bones consist only of primary(immature woven) bone. As growth progresses, the primary bone is completely or partly replacedby secondary, Haversian bone by the process of remodeling (87).During bone growth and remodeling, primary osteons are resorbed and replaced by secondaryosteons in which lamellar bone is formed around central vascular canals (86, 3). The formation ofsecondary osteons begins when osteoclasts form a cutting cone that advances through the bonecreating a resorption cavity (Figure 5). Osteoblasts lining the resorption cavity lay down osteoidfilling the resorption cavity with concentric lamellae from the outside of the cavity inwards (6).Osteoblastic apposition of concentric lamellae continues in the resorption cavity until a normalHaversian canal diameter has been achieved. The end result of this process is completion or closureof a new secondary Haversian system or osteon (6).Fawcett (3) reports that in humans aged 1 year and older, only organized lamellar bone isdeposited in the long bone shaft and that all primary bone is eventually replaced by secondaryHaversian bone. In, addition Fawcett (3) notes that internal bone resorption and reconstruction doesnot cease once primary bone is replaced by secondary bone. Resorption and remodeling take placethroughout life. Thus, in adult bone the following can be seen; mature osteons, forming osteons,and new absorption cavities. In addition, portions of former osteons that are not destroyed becomeinterstitial lamellae and are situated between mature osteons (55).In humans, approximately 1 µm of bone is deposited in developing Haversian systems per day(3). As osteons near completion, the rate of appositional bone formation slows (3). Osteon completionor closure occurs in 4 to 5 weeks.6.2.2 AnimalsThe fact that rodent models of human skeletal disorders lack Haversian system like those of largermammals, and that systemic internal remodeling of cortical and cancellous bone is dissimilar toone in humans was the reason for use of so called “second species” (88). Skeletal maturity (sealedgrowth plates) as well as presence of Haversian system and intra-cortical type of bone remodelingis required in order to more completely assess the effect of novel therapies or procedures aimedto ameliorate bone disorders of various etiologies. It is worth mentioning here that quadrupeds,non-primates have different bone biomechanical characteristics than bipeds, the fact to keep inmind when animal bone studies are extrapolated to humans (89, 90).© 2006 by Taylor & Francis Group, LLC

994 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONCuttingConeReversalZoneClosing ConeOABCDEBFGI II III IVGGGABEFDOForming ResorptionCavityResorption CavityForming HaversianSystemCompletedHaversian SystemFigure 5 (6) Diagram showing a longitudinal section through a cortical remodeling unit with correspondingtransverse sections below. A. Multinucleated osteoclasts in Howship’s lacunae advancing longitudinally fromright to left and radially to enlarge a resorption cavity. B. Perivascular spindle-shaped precursor cells. C. Capillaryloop. D. Mononuclear cells lining reversal zone. E. Osteoblasts apposing bone centripetally in radial closure andits perivascular precursor cells. F. Flattened cells lining Haversian canal of completed Haversian system.Transverse sections at different stages of development: (I) resorption cavities lined with osteoclasts; (II) completedresorption cavities lined by mononuclear cells, the reversal zone; (III) forming Haversian system or osteons linedwith osteoblasts that had recently apposed three lamellae; and (IV) completed Haversian system with flattenedbone cells lining canal. Cement line (G); osteoid (stippled) between osteoblast (O) and mineralized bone.As reported by Hert et al., (91), the diaphysis of long bones in adult humans and large adultmammals (longer living) such as dogs and sheep is composed of Haversian, secondary bone thatreplaces the original primary bone. However, the compact bone of smaller adult mammals such asmice and rats is not made up of Haversian systems (87). In fact, Bellino (92) notes that the rabbitis the smallest species known to undergo Haversian bone remodeling.Jowsey (93) reports that secondary osteons are not present in all species with the same degreeof frequency that is seen in mature human compact bone. He states that in some animals, primarybone is replaced at an early age, while in others, primary bone persists for quite some time. Forexample, it was noted that in the 2-year-old human, cortical bone of the femur consists mostly ofsecondary osteons whereas the cortical bone in the femur of a 7-year-old rabbit only has “occasional”osteons (93).When the tibia and fibula of rats, hamsters, guinea pigs, and rabbits were investigated at agesranging from 6 months to 5 years, it was noted that in varying degrees, primary bone is present incompact bone throughout the life of each of these animals (87). Specifically, in guinea pigs andrats, secondary bone in the form of osteons was found in the compact bone of only a few “veryold” animals. In the adult rabbit, however, a greater degree of “Haversian transformation” i.e.remodeling of primary osteons to secondary osteons was seen to occur throughout the life of theanimal (87).Enlow (94) also noted that there was a characteristic absence of Haversian systems in thecompact bone of many vertebrate species, specifically naming the white rat. However, Enlow (94)reported that the appearance of primary osteons is quite common in the young growing dog and© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 995monkey. Albu et al., (95) noted that Haversian systems do exist in the compact diaphyseal bone oflarger mammals such as the dog, pig, cow, and horse. In fact, Georgia et al., (96) reports that indogs, the density of Haversian canals in the compact bone of the femoral diaphysis is higher thanthat of humans whereas the diameter of the Haversian canals are smaller in dogs than they are inman.The rate of osteoblastic apposition of bone in the resorption cavity of a forming secondaryosteon has been extensively studied in dogs (97, 98, 99). Lee (97) noted that age plays an importantrole in the rate of formation of osteons. In dogs, as age increased, the rate of appositional bonegrowth in osteons decreased (97). The mean growth rate of appositional growth in forming osteonsin a 3 month old, 1 to 2 year old, and an adult dog of unknown age was 2.0 µ/day, 1.5 µ/day, and1.0 µ/day, respectively (97). In addition, osteons in the dog have been shown to grow to maturitywithin 4 to 8 weeks (99).In dogs, the appositional rate of bone formation was highest during the early stages of osteonformation when compared to the later stages. It was postulated that the rates of appositional bonegrowth were most rapid initially to smoothe over absorption lacunae and to convert the resorptioncavity into a concentric canal. As the Haversian canal cavity grew smaller due to addition ofconcentric lamellae by osteoblasts, the appositional bone formation decreased (97, 100). Thus, therate of appositional bone formation decreased as the cavity resorption cavity was closing and nearedcompletion (100).7 VascularitySuccessful growth and development of human bone is vitally dependent upon a vascular bloodsupply (2). The vascular supply to a typical long bone is strikingly similar among mammals (55).In general, long bones, as well as many flat and irregular bones are vascularized by nutrientarteries and veins that pass through the compact bone acting as channels for the entry and exit ofblood (101). These vessels gain entry to the compact bone via openings called nutrient foramensand canals. When a nutrient artery reaches the marrow cavity, it divides and branches proximallyand distally (101).During postnatal development, the diaphysis, metaphysis, and epiphysis of long and short bonesare separated by a cartilagenous growth plate. While the growth plate persists, the terminal branchesof nutrient arteries do not cross through the growth plate into the epiphysis (101, 102). Instead,the capillary ends of these branches terminate just below the growth plate. In the adult, the growthplate has fused which allows the terminal branches of nutrient arteries to extend unhindered intothe epiphyseal region where they connect with arteries arising from the periosteum (101). It isimportant to note that the functional importance of the vessels that cross into the epiphysis duringadulthood is uncertain as they cannot supply sufficient blood to the entire epiphysis in the eventof epiphyseal artery damage (4). It is also important to note that while the growth plate persists,blood is supplied to the region by 3 different sources.In addition to the nutrient artery, the blood supply of many flat bones is also derived fromperiosteal vessels. With age, long bones can sometimes also become increasingly dependent onperiosteal vessels (103).The blood supply of the metaphysis, diaphysis, epiphysis, and epiphyseal growth plate (Section7.2) is discussed below.7.1 Vessels of Long BonesSpecific groups of arterial networks are responsible for supplying blood to the various portions oflong bones (Figure 6a). These networks include the diaphyseal arteries, the metaphyseal arteries,and the epiphyseal arteries (2). In addition, vessels arising from the periosteum, can also providea source of vasculature to bones (55).© 2006 by Taylor & Francis Group, LLC

996 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPeri-articular arcadeLevel ofgrowth plateEpiphysealvesselsMetaphyseal vesselsAscending branchesNutrient arteryPeriosteal boneDescending branchesLevel ofgrowth plateCortical networkMetaphysealvesselsEpiphysealvesselsFigure 6a (2)A summary of the main arterial supply to a developing long bone.Postnatally, nutrient arteries supply the diaphysis and central portion of the metaphysis withblood while the epiphyseal arteries supply only the epiphyses. The metaphyseal arteries supply theperipheral regions of the metaphysis (103).In addition, during early postnatal development in the epiphyseal region, cartilage canalspersisting from the fetal period are a source of nutrition for the epiphyses. Cartilage canals mayalso play a role in the appearance of the secondary center of ossification (55, 103).7.1.1 Diaphyseal ArteriesDiaphyseal arteries gain entry to the diaphysis of long bones through nutrient foramens that leadinto nutrient canals (2). Once inside of the medullary cavity, the diaphyseal arteries branch proximallyand distally (2, 104). The ascending and descending branches of the vessel travel throughoutthe marrow cavity and end in helical loops near the metaphysis. In the metaphyseal region thesevessels connect with both the metaphyseal and periosteal arteries (104).The main nutrient artery is surrounded by several venules that merge into 1 or 2 large nutrientveins. In fact, in bone, many openings or foramina give exit to veins alone so that the number ofnutrient veins draining a bone usually exceeds the number of the nutrient arteries supplying it (103).Long bones are sometimes supplied by more than one nutrient diaphyseal artery (103). Forexample, the rabbit tibia has 2 diaphyseal arteries and the artery of the trochanteric fossa in the ratfemur can develop into a second afferent vessel to the diaphysis. On occasion, the diaphysis of thehuman humerus may have 3 nutrient arteries and the diaphysis of the human femur may have upto 2 nutrient arteries (103).In the scientific literature conflicting views exist regarding the direction and source of arterialblood flow in the cortex of the diaphysis. Historically, scientists believed that the compact bone ofthe diaphyseal cortex was fed by periosteal vessels, and the inner portion of the shaft received itsblood supply from the nutrient artery (102, 2). However, modern research has demonstrated thatin young bones (< 35 years old), the diaphyseal cortex primarily receives its blood supply fromthe nutrient artery of the diaphysis, but as bones age, periosteal arteries may begin to supply bloodto the cortex (2).© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 997EpiphysealArteryPerichondrialArteryMetaphysealArteryNutrientArteryMetaphysealArteryPerichondrial(“Central”)ArteryFigure 6b (107)Drawing showing the blood supply of a typical growth plate7.1.2 Metaphyseal ArteriesThe blood supply of the metaphysis comes from both the diaphyseal nutrient artery and themetaphyseal arteries (Figure 6a-b).In early fetal development, the blood supply to the metaphysis comes only from branches ofthe diaphyseal nutrient artery (103, 70). However, as development continues, the metaphysealarteries become an additional source of nutrition (103). As a result, during postnatal bone growth,the active metaphysis receives its blood supply from both the branches of the diaphyseal nutrientartery and from the metaphyseal arteries that pierces the substantia compacta in the vicinity of themetaphysis (103, 105). When these 2 nutrient sources are present, the central region of the metaphysisreceives blood from the diaphyseal arteries, while the peripheral portions of the metaphysisare supplied by metaphyseal arteries (102, 103). In the adult, the metaphyseal arteries branch intothe entire (non-growing) metaphysis and supply blood to the whole region (103).The metaphyseal arteries enter bone from the periosteum next to the metaphysis and join withbranches of the nutrient artery (106). In the absence of a diaphyseal artery, the metaphyseal arteriescan supply blood to the entire diaphyseal shaft (103).7.1.3 Epiphyseal ArteriesEarly in development (fetal and early postnatal periods), cartilage canals provide nutrition to thecartilage of the epiphysis; however, during postnatal bone growth, it is the epiphyseal arteries takethe lead role in providing a vascular supply to the growing epiphysis.In humans and large vertebrates such as dogs and rabbits, cartilage canals are present in theepiphyses of developing bone months before the formation of secondary ossification centers. Thesecartilage canals contain arterioles, venules, and intervening capillaries in a connective tissue matrix(108, 103, 55). The purpose of these cartilage canals is to provide nutrition to the cartilage. Laterin development cartilage canals may play a role in the formation of secondary ossification centers(108, 55). In the late embryonic and early postnatal periods, cartilage canals can been seen crossingpartly or completely through the region where the growth plate later develops (108). Postnatally,the vessels that cross through the region where the epiphyseal growth plate develops are destroyed© 2006 by Taylor & Francis Group, LLC

998 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONleaving this region avascular. These changes occur in the rabbit a few weeks after birth, and inhumans several months after birth (108).The role of cartilage canals in the formation of secondary ossification centers has not been clearlyestablished. However, it is believed that during postnatal development, osteoprogenitor cells from theconnective tissue covering the canals differentiate into the osteoblasts that begin ossification (55).Epiphyseal arteries enter long bones through nutrient foramens in the non-articular region ofthe epiphysis between the articular cartilage and the cartilagenous growth plate. During postnatalbone growth, epiphyseal arteries are responsible for supplying blood to the epiphysis and to thezone of resting cells at the top of the growth plate (see Section 7.2, Vascularity of the EpiphysealGrowth Plate) (103, 104).While the growth plate persists, diaphyseal nutrient arteries do not cross the epiphyseal growthplate and thus the epiphysis relies solely on epiphyseal arteries for its blood supply (104, 2).7.1.4 Periosteal VesselsThe periosteum, which covers the external surface of bone is an extremely vascular connectivetissue. This tissue provides an additional source of vasculature to bones. Specifically, blood vesselsfrom the periosteum form vascular networks that penetrate the outer surface of the bone and providenutrition to the outer portions of the compact bone. If the blood supply to the medullary cavity isdestroyed, the periosteum can provide blood to the entire diaphyseal cortex through Volkmannsand Haversian canals (55).The periosteum is also a major source of vasculature in flat bones (103).7.2 Vascularity of the Epiphyseal Growth PlateThe vascular supply of the growth plate has been shown to come from 3 sources: epiphyseal arteries,metaphyseal arteries, and perichondrial arteries (Figure 6b) (105, 107, 70). Perichondrial arteriesare derived from the fibrous connective tissue called the perichondrium the covers the peripheralregions of the cartilagenous growth plate (107, 70).From the epiphyseal side, terminal branches of the epiphyseal arteries provide blood to theuppermost regions of the cartilagenous growth plate. Once the inside the epiphysis, smaller arteriolesbranch from the main epiphyseal artery and pass through small cartilage canals in the reserve orresting zone to end at the top of the columns of cells in the proliferative zone (70). As a result, theproliferative zone is well supplied with blood from the epiphyseal arteries. It is important to notethat these branching epiphyseal arteries do not penetrate into the hypertrophic zone.From the metaphyseal side, the surface of the growth plate is supplied by metaphyseal arteries(105, 107, 70). Terminal branches from the nutrient and metaphyseal arteries enter the zone ofossification and end in vascular loops at the base of the cartilage portion of growth plate (55). Itis here that the vessels loop back so that the venous branches can descend to drain (70).Perichondrial arteries supply the peripheral regions of the growth plate (65, 107, 70, 103).7.3 Vascularity of the Flat Bones of the SkullLike the long bones, growth of flat bones in the skull (calvaria) is dependent upon a vascular supply.The vascularity of the calvaria is derived from pericranial vessels, dural vessels, sutural vessels,meningeal vessels, and calvaria veins (103).7.3.1 Pericranial VesselsThe outer surface of the brain is covered with a layer of connective tissue that is called thepericranium (3). The pericranium has a vascular network that is similar to that of long bone© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 999periosteum. The vessels of the pericranium are connected to the internal vessels of the cranial bonesby arteries and veins that pass through the juxtasutural foramina (103).In addition, capillaries of the outer table connect the diploic veins with those of the pericranium(103).7.3.2 Dural VesselsThe dura mater is the outer and most fibrous of the 3 membranes that surround the brain. The duramater is made up of two fused layers; an endosteal outer layer called the endocranium and theinner meningeal layer. Located between these 2 layers are the venous sinuses (103).Two distinct vascular networks are found within the dura mater. The inner vascular plexus, andthe outer vascular plexus. The inner vascular plexus is made up of arteries, veins, and a capillarynetwork called the primary dural plexus. The arteries are derived from the meningeal vessels, andthe veins communicate with those of the outer vascular plexus (103).Nutrient arteries derived from dural vessels penetrate the inner tables of flat bones. Branchesof these arteries extend peripherally from the ossification centers the develop during the fetal periodto the sutural boundaries of the frontal, parietal, and occipital bones (103).7.3.3 Sutural VesselsIn the skull, many bones such as the frontal, parietals, squamous occipital, squamous temporal,and greater wing of sphenoid are connected together by junctions called sutures (2). During postnataldevelopment, sutures are composed of dense connective tissue which later becomes ossified (109).During postnatal skull growth, a plethora of small foramina that contain arteries and veins canbe seen near the sutural borders (103). These “juxta-sutural vessels” play an important role in aidingthe growth in surface area of individual flat bones (103). In addition, the fibrous tissue of the suturecontains sutural veins. Trans-sutural connections of veins can be found which unite the veins ofone bone with those of the neighboring bones. The sutures also connect the blood vessels of thepericranium (the connective tissue membrane that surrounds the skull) with the vessels of the dura(the outermost and most fibrous of 3 membranes surrounding the brain). The edges of bonesconnected by sutures contain many holes for the passage of blood vessels (103).After ossification and closure of sutures, the skull maintains vascularity as the surface of bonesare covered with small foramina filled with capillaries that serve to join the capillaries of the diploewith those of the pericranium (103). Blood vessels can also be seen passing through the inner tablejoining with those of the diploe and the dura mater (103).7.3.4 Meningeal ArteriesThe cranial meningeal arteries supply the dura mater and the 2 other membranes covering the brain(leptomeninges) with blood; however, the main function of the meningeal arteries is to supplyblood to the cranial bones. (103).7.3.5 Calvarial VeinsIn the developing calvaria small arteries and large venules join together without capillary mediation(arteriovenous anastomoses). In addition, in the adult skull, venous sinuses of the diploe areconnected to vessels of the pericranium and dura mater by capillaries from the inner and outertables. Large diploic sinuses (the veins of Breschet) are proximal and parallel to the sutures. Thediploic sinuses exit the outer table of the skull through the frontal, anterior temporal, posteriortemporal, and occipital diploic foramina (103).© 2006 by Taylor & Francis Group, LLC

1000 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONConclusionsGrowth and weight gain within normal ranges is viewed as part of the assessment of the overallhealth of young children and animals. Bone growth and development occurs in a similar fashionin all mammalian species studied during the course of this literature review. Bone growth has beenassessed in animal studies by direct measure of selected bones at necropsy (110) and by noninvasivetechniques such as dual energy X-ray absorptiometry (DEXA) (111, 112) and autoradiography.Using direct measurements of body and femur and cranial length, Schunior et al. (110) were ableto study the relationship between postnatal growth and body weight. Bone growth kinetics over amatter of hours has been studied using tetracycline as a marker followed by histopathology (113).Bone quality is also assessed bone mineral density (BMD) and bone mineral content (BMC)analyses.This review summarizes available literature regarding important processes in postnatal bonegrowth. The timing of important developmental milestones was compared between humans andlaboratory animal species. These milestones included ossification, epiphyseal growth plate, metaphysis,as well as diaphysis development, osteon formation, and vascularization.In conclusion, postnatal skeletal growth can be assessed in appropriately designed animal studiessince the growth and developmental patterns are similar for both laboratory animals and humans.REFERENCES(1) Moore KL and Persaud TVN. The Developing Human: Clinically Oriented Embryology (Philadelphia:W.B. Saunders Company, 1998), 405-424(2) Scheuer, Louise and Black, Sue, Developmental Juvenile Osteology (San Diego: Academic Press,2000).(3) Fawcett DW. Bloom and Fawcett A Textbook of Histology, Twelfth Edition. (New York: Chapman &Hall, 1994), 194-259(4) Buckwalter JA, Glimcher MJ, Cooper RR and Recker R. “Bone Biology I: Structure, Blood Supply,Cells, Matrix, and Mineralization.” In: Instructional Course Lectures. Ed. Pritchard, DJ, vol. 45.(Rosemont, IL: American Academy of Orthopaedic Surgeons, 1996), 371-86(5) Price JS, Oyajobi BO and Russell RGG. The Cell Biology of Bone Growth. Eur. J. Clin. Nutr. 48(Suppl.1):S131-149 (1994)(6) Jee WSS. “The Skeletal Tissues.” In: Cell and Tissue Biology: A Textbook of Histology, Sixth Edition.Ed. Leon Weiss, M.D., with 35 contributors. Baltimore: Urban & Schwarzenberg, Inc., 1988.(7) Gray, H. Anatomy of the Human Body. Philadelphia: Lea & Febiger, 1918; Bartleby.com, 2000.www.bartleby.com/107/. [10-12-01](8) Junqueira LC, Carneiro J and Kelley RO. “Bone.” In: a LANGE Medical Book Basic Histology, NinthEdition. (Stamford, CT: Appleton & Lange, 1998), 134-151 (1998)(9) Zerwekh JE. Bone Metabolism. Seminars in Nephrology. 12(2):79-90 (1992)(10) Freemont AJ. Basic Bone Cell Biology. Review. Int. J. Exp. Pathol. 74(4):411-6 (1993)(11) Marks SC and Hermey DC. The Structure and Development of Bone. In Principles of Bone Biology.J.P. Bilezekian, L.G. Raiz, G.A. Rodan (Eds.). Academic Press,1996, 3-14.(12) Boskey AL and Posner AS. Bone Structure, Composition, and Mineralization. Orthop. Clin. NorthAm. 15(4):597-612 (1984)(13) Marks SC and Popoff SN. Bone Cell Biology: The Regulation of Development Structure and Functionin the Skeleton. Am. J. Anat. 183:1-44 (1988)(14) Patton JT and Kaufman MH. The Timing of Ossification of the Limb Bones, and Growth Rates ofVarious Long Bones of the Fore and Hind Limbs of the Prenatal and Early Postnatal LaboratoryMouse. J. Anat. 186(Pt. 1):175-85 (1995)(15) Davies DA and Parsons FG. The Age Order of the Appearance and Union of the Normal Epiphysesas Seen by X-rays. J. of Anat. 62:58-71 (1927)© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1001(16) Menees TO and Holly LE. The Ossification in the Extremities of the New-Born. Am. J. Roentgenology.28(3):389-90 (1932)(17) Kuhns LR, Sherman MP, Poznanski AK and Holt JF. Humeral-Head and Coracoid Ossification in theNewborn. Radiology. 107:145-149 (1973)(18) Kuhns LR and Finnstrom O. New Standards of Ossification of the Newborn. Radiology. 119:655-660(1976)(19) Ogden JA, Conlogue GJ and Jensen P. Radiology of Postnatal Development: The Proximal Humerus.Skeletal Radiol. 2:153-160 (1978)(20) Paterson RS. A Radiological Investigation of the Epiphyses of the Long Bones. J. Anat. 64:28-46(1929)(21) Pritchett JW. Growth-plate Activity in the Upper Extremity. Clin. Orthop. and Related Res. 268:235-242 (1991)(22) Hansman CF. Appearance and Fusion of Ossification Centres in the Human Skeleton. Am. J. Roentgenology.88(3):476-482 (1962)(23) Christie A. Prevalence and Distribution of Ossification Centers in the Newborn Infant. Am. J. Dis.Child. 77:355-361 (1949)(24) Gill GG and Abbott LC. Practical Method of Predicting the Growth of the Femur and Tibia in theChild. Arch. Surg. 45:286-315 (1942)(25) Anderson AC and Floyd M. Growth and Development of the Femur in the Beagle. Am. J. Vet. Res.24:348-351 (1963)(26) Sarnat BG. Growth Pattern of the Mandible: Some Reflections. Am. J. Orthod. Dentofac. Orthop.90(3):221-33 (1986)(27) Brodie AG. On the Growth Pattern of the Human Head from the Third Month to the Eighth Year ofLife. Am. J. Anat. 68:209-262 (1941)(28) Michejda M. Growth Standards in the Skeletal Age of Rhesus Monkey (M. mulatta) Chimpanzee(Pan tryglodytes) and Man. Dev. Biol. Stand. 45(2):45-50 (1980)(29) van Wagenen G and Asling CW. Roentgenographic Estimation of Bone Age in the Rhesus Monkey(Macaca Mulatta). Am. J. Anat. 103:163-186 (1958)(30) Fukuda S and Matsuoka O. Comparative Studies on Maturation Process of Secondary OssificationCenters of Long Bones in the Mouse, Rat, Dog, and Monkey. Exp. Anim. 29(3):317-26 (1980)(31) Fukuda S, Cho F and Honjo S. Bone Growth and Development of Secondary Ossification Centers ofExtremities in the Cynomolgus Monkey (Macaca fascicularis). Exp. Anim. 27(4):387-97 (1978)(32) McNamara JA Jr and Graber LW. Mandibular Growth in the Rhesus Monkey (Macaca mulatta). Am.J. Phys. Anthrop. 42:15-24 (1975)(33) Baume LJ. The Postnatal Growth of the Mandible in Macaca Mulatta. A Metric, Roentgenographicand Histologic Study. Am. J. Orthod. 39:228-229 (1953)(34) Turpin DL. Growth and Remodeling of the Mandible in the Macava Mulatta Monkey. Am. J. Orthod.54:251-271 (1968)(35) Hare WCD. Radiographic Anatomy of the Canine Pectoral Limb. Part II. Developing Limb. J.A.V.M.A.305-310 (1959)(36) Parcher JW and Williams JR. “Ossification.” In: The Beagle as an Experimental Dog. Ed., AndersenAC, Technical Ed., Good LS (Ames: The Iowa State University Press Inc., 1970), 158-161(37) Yonamine H, Ogi N, Ishikawa T and Ichiki H. Radiographic Studies on Skeletal Growth of the PectoralLimb of the Beagle. Jpn. J. Vet. Sci. 42(4):417-25 (1980)(38) Hare WCD. The Ages at Which the Centers of Ossification Appear Roentgenographically in the LimbBones of the Dog. Am J of Vet. Res. 22(90):825-835 (1961)(39) Hennet PR and Harvey CE. Craniofacial Development and Growth in the Dog. Reviewed Article –Anatomy. J Vet Dent. 9(2):11-18 (1992)(40) Sullivan PG. On Growth of the Canine Mandible. Brit. J. Orthod. 2(3):159-163 (1975)(41) Fukuda S and Matsuoka O. Radiographic Studies on Maturation Process of Secondary OssificationCenters in Long Bones of the Japanese White Rabbit. Exp. Anim. 30(4):497-501 (1981)(42) Heikel HVA. On Ossification and Growth of Certain Bones of the Rabbit; with a Comparison of theSkeletal Age in the Rabbit and in Man. Acta Orthop. Scand. 29(39):171-184 (1959-1960)(43) Khermosh O, Tadmor A, Weissman SL, Michels CH and Chen R. Growth of the Femur in the Rabbit.Am. J. Vet. Res. 33(5):1079-82 (1972)© 2006 by Taylor & Francis Group, LLC

1002 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(44) Bang S and Enlow DH. Postnatal Growth of the Rabbit Mandible. Arch. Oral. Biol. 12(8):993-8 (1967)(45) Masoud I, Shapiro F and Moses A. Longitudinal Roentgencephalometric Study of the Growth of theNew Zealand White Rabbit: Cumulative and Biweekly Incremental Growth Rates for Skull andMandible. J. Craniofac. Genet. & Dev. Biol. 6:259-87 (1986)(46) Johnson ML. The Time and Order of Appearance of Ossification Centers in Albino Mouse. Am. J.Anat, 52(2):241-271 (1933)(47) Fukuda S and Matsuoka O. Maturation Process of Secondary Ossification Centers in the Rat andAssessment of Bone Age. Exp. Anim. 28(1):1-9 (1979)(48) Dawson AB. The Age Order of Epiphyseal Union in the Long Bones of the Albino Rat. Anat. Rec.31(1):1-17 (1925)(49) Bhaskar SN. Growth Pattern of the Rat Mandible from 13 Days Insemination Age to 30 Days afterBirth. Am. J. Anat. 92(1):1-53 (1953)(50) Manson JD. A Comparative Study of the Postnatal Growth of the Mandible. (London: Henry Kimpton,1968)(51) Bernick S and Patek PQ. Postnatal Development of the Rat Mandible. J. Dent. Res. 48(6):1258-63(1969)(52) Meikle MC. The Role of the Condyle in the Postnatal Growth of the Mandible. Am. J. Orthod.64(1):50-62 (1973)(53) Silbermann M and Livne E. Skeletal Changes in the Condylar Cartilage of the Neonate MouseMandible. Biol Neonate. 35:95-105 (1979)(54) Livne E, Weiss A and Silbermann M. Changes in Growth Patterns in Mouse Condylar CartilageAssociated with Skeletal Maturation and Senescence. Growth, Dev. & Aging. 54(4):183-93 (1990)(55) Kincaid SA and Van Sickle DC. Bone Morphology and Postnatal Osteogenesis. Potential for Disease.Vet. Clin. North Am.: Small Anim. Pract. 13(1):3-17 (1983)(56) Sissons HA. Experimental Determination of Rats of Longitudinal Bone Growth. J Anat., Lond. 87:228-236 (1953)(57) Hert J. Growth of the Epiphyseal Plate in Circumference. Acta Anat. 82:420-36 (1972)(58) Hansson LI. Determination of Endochondral Bone Growth in Rabbit by Means of Oxytetracycline.Acta Univ. Lund. II(1):2-10 (1964)(59) Wallis GA. Bone Growth: Coordinating Chondrocyte Differentiation. Curr. Biol. 6(12):1577-80 (1996)(60) Nap RC and Hazewinkel HAW. Growth and Skeletal Development in the Dog in Relation to Nutrition;a Review. Vet. Quart. 16(1):50-9 (1994)(61) Rosati R, Horan GSB, Pinero GJ, Garofalo S, Keene DR, Horton WA, Vuorio E, de Crombrugghe Band Behringer RR. Normal Long Bone Growth and Development in Type X Collagen-null Mice. Nat.Genet. 8(2):129-35 (1994)(62) Haines RW. The Histology of Epiphyseal Union in Mammals. J. Anat. 120(1):1-25 (1975)(63) Thyberg J and Friberg U. Ultrastructure of the Epiphyseal Plate of the Normal Guinea Pig. Z.Zellforsch. 122:254-272 (1971)(64) Ingalls TH. Epiphyseal growth: Normal Sequence of Events at the Epiphyseal Plate. Endocrinology.29:710-719 (1941)(65) Hansson LI. Daily Growth in Length of Diaphysis Measure by Oxytetracycline in Rabbit Normallyand after Medullary Plugging. Acta Orthop. Scand. Suppl 101:9-199 (1967)(66) Miralles-Flores C, Delgado-Baeza E. Histomorphometric Differences Between the Lateral Regionand Central Region of the Growth Plate in Fifteen-Day-Old Rats. Acta Anat. 139:209-13 (1990)(67) Kember NF. Comparative Patterns of Cell Division in Epiphyseal Cartilage Plates in the Rat. J. Anat.111(1):137-142 (1972)(68) Moss-Salentijn L. Studies of Long Bone Growth. I. Determination of Differential Elongation in PairedGrowth Plates of the Rat. Acta Anat. 90:145-60 (1974)(69) Guthrie S, Plummer JM, Vaughan LC. Post natal development of the canine elbow joint: a light andelectron microscopical study. Res Vet Sci. 52(1):67-71 (1992)(70) Brighton CT. The growth Plate. Orthop. Clin. North Am. 15(4):571-95 (1984)(71) Zaleske DJ. Cartilage and Bone Development. AAOS Instr. Course Lect. 47:461-8 (1998)(72) Raisz LG and Kream BE. Hormonal control of skeletal growth. Ann. Rev. Physiol. 43:225 (1981)(73) Ring PA. Ossification and Growth of the Distal Ulnar Epiphysis of the Rabbit. J Anat., Lond. 89:457-464 (1955)© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1003(74) Seinsheimer F III and Sledge CB. Parameters of longitudinal growth rate in rabbit epiphyseal growthplates. J. Bone Joint Surg. AM63-A(4):627-30 (1981)(75) McCormick MJ, Lowe PJ and Ashworth MA. Analysis of the relative contributions of the proximaland distal epiphyseal plate to the growth in length of the tibia in the New Zealand white rabbit.Growth. 36(2):133-44 (1972)(76) Rudicel S, Lee KE and Pelker RR. Dimensions of the rabbit femur during growth. Am. J. Vet. Res.46(1):268-9 (1985)(77) Thorngren K-G and Hansson LI. Cell Production of Different Growth Plates in the Rabbit. Acta. Anat.110:121-7 (1981)(78) Li XJ, Jee WSS, Ke HZ, Mori S and Akamine T. Age-related Changes of Cancellous and CorticalBone Histomorphometry in Female Sprague-Dawley Rats. Cells and Materials. Suppl 1:25-35 (1991)(79) Kimmel DB and Jee WS. A Quantitative Histologic Analysis of the Growing Long Bone Metaphysis.Calcif. Tissue Int. 32(2):113-22 (1980)(80) Lozupone E. A Quantitative Analysis of Bone Tissue Formation in Different Regions of the Spongiosain the Dog Skeleton. Anat. Anz. 145(5):425-52 (1979)(81) Stump CW. The Histogenesis of Bone. J. Anat. 59:136-154 (1925)(82) Parfitt AM. The Two Faces of Growth: Benefits and Risks to Bone Integrity. Osteoporosis Int. 4:382-98 (1994)(83) Kerley ER. Age Determination of Bone Fragments. J. Forensic Sci. 14(1):59-67 (1969)(84) Parfitt AM, Travers R, Rauch F and Glorieux FH. Structural and Cellular Changes During BoneGrowth in Healthy Children. Bone. 27(4):487-94 (2000)(85) Tomlin DH, Henry KM and Kon SK. Autoradiographic Study of Growth and Calcium Metabolismin the Long Bones of the Rat. Brit. J. Nutr. 7:235-253 (1953)(86) Cooper RR, Milgram JW and Robinson RA. Morphology of the Osteon. An Electron MicroscopicStudy. J. Bone Jt. Surg. 48-A(7):1239-1271 (1966)(87) Fiala P. Age-related Changes in the Substantial compacta of the Long Limb Bones. Folia Morphologica.26(4):316-20 (1978)(88) Rogers, MJ, Nicholson, PHF, Davie, MWJ, Lohmann, G, and Turner, AS. The pathophysiology ofosteoporous – species differences. In Current Research in Osteoporosis and Bone Mineral Measurement2. Eds: Ring, EFJ, Elvins, DM and Bhalla, AK, pp. 50-57 (1993)(89) Biewener, AA., Biomechanics of mammalian terresterial locomotion. Science 250: 1097-1103 (1990)(90) Einhorn, TA., Bone strength: The bottom line. Calcif Tissue Int 51: 333-339 (1992)(91) Hert J, Fiala P and Petrtyl M. Osteon Orientation of the Diaphysis of the Long Bones in Man. Bone.15(3):269-77 (1994)(92) Bellino FL. Nonprimate animal models of menopause: workshop report. Menopause: J. North Am.Menopause Soc. 7(1):14-24 (2000)(93) Jowsey J. Studies of the Haversian Septums in Man and Some Animals. J. Anat. 100(4):857-864 (1966)(94) Enlow DH. Functions of the Haversian System. Amer. J. Anat. 110: 269-305 (1962)(95) Albu I, Georgia R and Georoceanu M. The Canal System in the Diaphysial Compacta of the Femurin Some Mammals1. Anat. Anz., Jena. 170:181-187 (1990)(96) Georgia R, Albu I, Sicoe M and Georoceanu M. Comparative Aspects of the Density and Diameterof Haversian Canals in the Diaphyseal Compact Bone of Man and Dog. Morphol.-Embryol. 28(1):11-4 (1982)(97) Lee WR. Appositional Bone Formation in Canine Bone: A Quantitative Microscopic Study UsingTetracycline Markers. J Anat., Lond. 98(4):665-677 (1964)(98) Marotti G. The Dynamics of Osteon Formation in Inert Bones. Calcif Tissue Res. 2(Suppl):86-88(1968)(99) Vanderhoeft PJ, Kelly PJ and Peterson LFA. Determination of Growth Rates in Canine Bone by Meansof Tetracycline-labeled Patterns. Lab. Invest. 11(9):714-726 (1962)(100) Anderson C and Danylchuck KD. Appositional Bone Formation Rates in the Beagle. Am. J. Vet. Res.40(7):907-10 (1979)(101) Miller ME, Christensen GE and Evans HE. Anatomy of the Dog (Philadelphia: W.B. SaundersCompany, 1965, 1964), 1-94(102) Lewis OJ. The Blood Supply of Developing Long Bones with Special Reference to the Metaphyses.J. Bone Joint Surg. 38(4):928-933 (1956)© 2006 by Taylor & Francis Group, LLC

1004 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(103) Brookes M and Revell WJ. “Nutrient Vessels in Long Bones,” In: Blood Supply of Bone. ScientificAspects. (Great Britain: Springer-Verlag London Limited, 1998)(104) Buckwalter JA and Cooper RR. “Bone structure and function.” In: Instructional Course Lectures.American Academy of Orthopaedic Surgeons, Ed. Griffin PP, vol. 36. Easton, PA: Mack PrintingCompany, 1987, 27-48(105) Irving MH. The Blood Supply of the Growth Cartilage in Young Rats. J Anat., London. 98(4):631-9(1964)(106) Spira E and Farin I. The Vascular Supply to the Epiphyseal Plate Under Normal and PathologicalConditions. Acta Orthop. Scandinav. 38: 1-22 (1967)(107) Brighton CT. Structure and Function of the Growth Plate. Clin. Orthop. 136:22-32 (1978)(108) Shapiro F. Epiphyseal and Physeal Cartilage Vascularization: A Light Microscopic and TritiatedThymidine Autoradiographic Study of Cartilage Canals in Newborn and Young Postnatal Rabbit Bone.Anat. Rec. 252(1):140-8 (1998)(109) Zimmerman B, Moegelin A, de Souza P and Bier J. Morphology of the Development of the SagittalSuture of Mice. Anat. Embryol. 197:155-65 (1998)(110) Schunior A, Zengel AE, Mullenix PJ, Tarbell NJ, Howes A and Tassinari MS. An Animal Model toStudy Toxicity of Central Nervous System Therapy for Childhood Acute Lymphoblastic Leukemia:Effects on Growth and Craniofacial Proportion. Cancer Research. 50:6455-6460 (1990)(111) Arikoske, P., Komulainen, J., Riikonen, P., Voutilainen, R., Knip, M., and Kroger, H. Alterations inbone turnover and impaired development of bone mineral density in newly diagnosed children withcancer: A 1-year prospective study. J. Clin. Endocrinol. Metabolism 84(9):3174-3181 (1999)(112) Juhn A., Weiss A., Mendes D., and Silbermann M. Non-invasive assessment of bone mineral densityduring maturation and aging of wistar female rats. Cells and Materials. Suppl 1: 19-24 (1991)(113) Tam CS, Reed R, Campbell JE and Cruickshank B. Bone Growth Kinetics II. Short-term Observationson Bone Growth in Sprague-Dawley Rats. J. Pathol. 113(1):39-46 (1974)© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1005APPENDIX C-2*SPECIES COMPARISON OF ANATOMICAL ANDFUNCTIONAL RENAL DEVELOPMENTTracey Zoetis 1 , and Mark E. Hurtt 2,31Milestone Biomedical Associates, Frederick, MD 21701, USA2Pfizer Global Research & Development, Groton, CT 06340, USA3Correspondence to: Dr. Mark E. Hurtt, Pfizer Global Research & Development, Drug Safety Evaluation, Eastern PointRoad, Mailstop 8274-1306, Groton, CT 06340. 860-715-3118. Fax (860) 715-3577. Email: mark_e_hurtt@groton.pfizer.comIntroductionRenal development involves both anatomical and functional aspects that occur in predictabletimeframes. Xenobiotics can interrupt either anatomic or functional development or both. Thepurpose of this paper is to identify critical time frames for anatomical and functional developmentof the kidney in the humans and to compare these events to other species. This comparison shouldresult in data that will be useful in designing and interpreting studies of the possible prenatal and/orpostnatal developmental effects of chemicals on the maturing kidney.Kidney maturation has been assessed by the size and distribution of the tubules and by thehistologic appearance of glomeruli (1). Major renal anatomic developmental events occur in humansprenatally, with functional development continuing into the postnatal periods. A brief descriptionof human anatomical renal development is presented followed by a description of comparativespecies. Similarly, a description of human functional renal maturation is presented using the majorrenal functions of glomerular filtration, urine concentration, acid base equilibrium, and urine volumecontrol, followed by a description of comparative species for each function.Anatomical DevelopmentHumansThe human kidney begins to develop through a process of reciprocal inductive interaction duringweek 5 of gestation. The interaction occurs between the primitive ureteric bud and the metanephricmesenchyme (a “temporary” kidney derived from the pronephros). Beginning at gestation week 5,the diverticulum of the mesonephric duct (primitive ureter) comes into contact with the caudalmeasenchyme of the nephrogenic cord and induces it to undergo epithelial transformation that givesrise to the upper urinary tract or ureteric (or metanephric) bud. The mesenchyme then induces theureteric bud to grow, differentiate, and branch into it. The mesenchymal cells condense around thetip of each branch of the developing ureter. These condensates develop into vesicles, and then intoa comma-shaped body that develops into an S-shaped glomerulus. Endothelial cells collect in theS-shaped body and form a glomerular capillary loop. Nephrogenesis and kidney vasculature developduring the same timeframe and are complete about week 35 of gestation (2, 3, 4, 5, 6). A diagramof this process is presented below.* Source: Zoetis, T. and Hurtt, M. E., Species comparison of anatomical and functional renal development, Birth DefectsResearch, Part B: Developmental and Reproductive Toxicology, 68, 111-120, 2003.© 2006 by Taylor & Francis Group, LLC

1006 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONLoose mesenchymeCondensation14 32epithelialureter budmesenchymeComma-ShapeDistalS-ShapeTubule elongationproximalPodocyte FoldingdistalProximalPodocyte CapsuleFigure 1 Schematic representation of nephrogenesis. The branching ureteric epithelium interaction with loosemetanephric mesenchyme (A) results in condensation of the mesenchyme (B). Cell lineages shown include: 1)ureteric epithelium, 2) vasculature, 3) undifferentiated mesenchyme, and 4) condensed mesenchyme differentiatinginto epithelia. These stages are followed by infolding of the primitive glomerular epithelium to form commaandS-shaped bodies (C and D). Elongation of the proximal and distal tubular elements subsequently occur (E)along with further infolding of the glomerular epithelium and vascular structures to form the mature glomerularcapillary network (F). The initial phases of glomerular vascularization are believed to occur during the earlystages of glomerular differentiation (C and D). From Ekblom, (1984), with permission.By the fifth month there are 10 to 12 branchings (7). Approximately 20% of nephrons areformed by 3 months of gestation, 30% by 5 months, and nephrogenesis is complete with a totalof approximately 800,000 nephrons by 34 weeks of gestation (1, 7, 8). Juxtamedullary nephronsare formed earliest (by 5 months gestation) and superficial cortical nephrons form later (by 34weeks gestation). Postnatal maturation of the nephrons and elongation of the tubules continuesduring the first year of life (7). An illustration of the branching process is presented below.Three developmental periods have been defined based on the rate of glomerular proliferation(9). The first period begins during gestation week 10 and is marked by a slow increase in thenumber of glomeruli. The second period begins during gestation week 17 or 18 when glomerularproliferation increases abruptly and occurs rapidly until gestation week 32. The period betweengestation weeks 18 – 32 is a critical time point in renal development; it is at this time thatnephrogenetic development reaches its peak (9). The third period lasts from gestation week 32 on,© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1007Collecting tubulesOutgrowing collecting tubulesMajor calyxMetanephric blastemaPelvisMinor calyxureterRenal pelvisA B C DFigure 11-5. Schematic drawings showing the development of the renal pelvis, calyces and collecting tubulesof the metanephros. A, At 6 weeks; B, end of sixth week; C, 7 weeks; D, newborn. Note the pyramid form ofthe collecting tubules entering the minor calyx. From: Langman, 1975. (4)when no increase in the number of glomeruli is observed, the nephrogenic blastema disappears,and the nephrogenetic process is complete.Low birth weight premature infants exhibit nephrogenesis at a level commensurate with agerather than body weight, and maturation continues during early postnatal weeks (1). The numberof nephrons in a given species is constant; however, cellular growth can be influenced by exposureto environmental factors, renal blood flow and glomerular filtration, capacity for excretion of sodiumand water, renal prostaglandin production and urinary excretion of calcium (5, 7).Comparative SpeciesMammalian kidneys follow similar developmental pathways; however, the time frame with regardto birth varies between species. Cell culture studies using glomeruli from monkey, sheep, dog,rabbit, and rat kidneys demonstrate that the pattern of growth and morphologic features of eachcell type, including numbers present and rates of division were the same for all species studied(10). Cell culture studies using nephrons from various species have demonstrated similar geneticexpression of patterns that can be related to time scales in morphogenesis (3). An electron microscopicstudy of granulated glomerular epithelial cells from 17 mammalian species demonstratedsimilarities in morphology between all species studies (11). Thus, animal studies provide criticalinformation that aids in the understanding of renal development across species, including primatessuch as humans. The following table marks the gestation day of the first appearance of metanephroscells that differentiate and proliferate to form the kidney for several species.Onset of Kidney Development as Evidenced by MetanephrosSpeciesMetanephros(gestation day)Total gestation period(days)Man 35–37 267Macaque 38–39 167Guinea Pig 23 67Rabbit 14 32Rat 12.5 22Mouse 11 19Hamster 10 16Chick 6 21From: Evan, et. al., 1984 (12)© 2006 by Taylor & Francis Group, LLC

1008 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe following table illustrates the timing of completion nephrogenesis for various species, assummarized in Kleinman (13), Fouser and Avner (14), and Gomez et. al, (2). More detaileddescriptions of individual species follow.Species Timing of Nephrogenesis CompletionMan35 weeks gestationSheepBefore birthGuinea PigBefore birthDog Postnatal week 2Pig Postnatal week 3MouseBefore birthRat Postnatal week 4–6RatsIn the rat, nephrogenesis was observed to occur at a rapid pace between birth and 8 days and iscomplete by 11 days of age (15), tubular differentiation continues until the time of weaning, andfunctional maturity occurs even later (16).One of the factors involved in nephrogenesis is the nutritional status of the mother. In rats fedlow protein diets during days 8 – 14 or 15 – 22 of gestation, offspring were observed to have lowernumbers of nephrons and low renal size. This observation persisted until 19 weeks of age, at whichtime glomerular filtration rate (GFR) was normal (17).Maturation of the loop of Henle has been studied in rats, with regard to succinic dehydrogenaseand acid phosphatase activity (18). Neonatal loops of Henle are relatively short loops without thinascending limbs. As maturation occurs, apoptic deletion of thick ascending cells and transformationinto thin ascending limb cells yields a well-defined boundary between inner and outer medullas bypostnatal day 21 in the rat (18).MiceMice and humans share, among renal histological characteristics, timing of onset of nephrogenesis(19). This process occurs prenatally in both mice and humans. In mice nephrogenesis begins ongestation day 11 and is complete by birth (14).DogsThe puppy kidney is immature in both structure and function at birth (20). Functionally, the puppyhas a lower glomerular filtration rate, renal plasma flow, and filtration fraction compared to theadult dog (20). In a study of casts of renal vessels from puppies age 1 – 21 days after birth,investigators found evidence of immaturity of the intrarenal vascular system and the proximal tubulewhen compared to adult dog. Nephrogenesis continues for at least 2 weeks postnatally in the dog(20).Glomerular blood flow and thus maturation was studied using microsphere injection in 26puppies ranging in age from 5h to 42 days and in 5 adult dogs (21). The renal cortex was divided© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1009in to four equal zones progressing from the outer to inner cores, and glomeruli were counted ineach. Results confirmed that nephrogenesis is still underway at the time of birth. Positive identificationof the differentiation of the primitive metanephric vesicle from the vascularized glomeruliwas not always possible since the glomeruli become more separated as the dog ages and tubuleselongate. In newborns, the glomeruli of the inner zone are larger than those in the outer zone, butthis difference gradually disappears as the kidney grows (21). A similar study showed that plasmaflow is an important factor in this process in maturing puppies (22).SheepNephrogenesis is complete in sheep at birth (23, 24).PrimatesNephrogenesis is complete in monkeys at birth (23, 24).RabbitsNephrogenesis is completed 2 – 3 weeks postnatally in rabbits (24). Maturation of the loop ofHenle has been studied in rabbits, with regard to succinic dehydrogenase and acid phosphataseactivity, and carbonic anhydrase IV activity (18). Succinic dehydrogenase and acid phosphatasehave high levels of activity in the neonate and decrease to adult levels by about postnatal day 28in the rabbit. Neonatal loops of Henle are relatively short loops without thin ascending limbs. Asmaturation occurs, apoptic deletion of thick ascending cells and transformation into thin ascendinglimb cells yields a well-defined boundary between inner and outer medullas. The expression ofcarbonic anhydrase IV has been correlated with this maturation process in the rabbit (18).Functional DevelopmentCompared to the adult renal function, the human infant has decreased renal blood flow, glomerularfiltration rate, tubular secretion, and a more acidic urinary pH (7). Urine production begins duringthe 10 th week of gestation (6, 2). Maturation of glomerular filtration, concentrating ability, acidbaseequilibrium, and urine volume control are described below.Glomerular FiltrationHumansThe primary function of the glomeruli is to act as a filter before plasma reaches the proximal tubule.Glomerular filtration rate (GFR) continues to increase after birth and reaches adult levels at 1 – 2years of age. After one month of age, creatinine clearance is used to measure GFR; however, GFRis commonly estimated in units of ml/min/1.73m 2 , based on the formula K L/S Cr , where K is aconstant*, L is length (or height in cm) and S Cr is serum creatinine (mg/dl). Serum creatinine levelsduring the postnatal days 1 and 2 reflect maternal values and decrease to 0.2 – 0.4 mg/dl by 3months. An increase in creatinine level indicates a decrease in GFR regardless of gestational age(2). Normal values for GFR are presented in the following table.* The K constant is 0.55 in children and adolescent girls, 0.70 in adolescent boys, 0.45 in term neonates, and 0.33 inpreterm infants.© 2006 by Taylor & Francis Group, LLC

1010 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONNormal Values of Glomerular Filtration Rate for Humans by AgeAge GFR (ml/min/1.73m 2 )Preterm infant (gestation weeks 25–28)1 week 11.0 ± 5.42–8 weeks 15.5 ± 6.2Preterm infant (gestation weeks 29–34)1 week 15.3 ± 5.62–8 weeks 28.7 ± 13.8Term infant5–7 days 50.6 ± 5.81–2 months 64.6 ± 5.83–4 months 85.8 ± 4.85–8 months 87.7 ± 11.99–12 months 86.9 ± 8.42–12 years 133 ± 27From: Gomez et al. 1999 (2)The kidney is a well-perfused organ, receiving approximately 20% of cardiac output at rest.Glomerular filtration delivers filtered plasma to the proximal tubule. Small molecules (~5000daltons) pass through the glomerular barrier and restriction increases as molecular mass increasesto where molecules the size of albumin (~68,000 daltons) do not pass the barrier. Glomerularfiltration rate (GFR) is determined using clearance studies comparing urine volume excreted withamount of excreted inulin or some other metabolically inert material not reabsorbed or secreted bythe renal tubules. A normal GFR for newborn infants is less than 50 mL/min/1.73m 2 , rises to 100to 140 mL/min/1.73m 2 in humans over 1 year of age, and reaches an adult level by the age of 2years (1, 7, 8). After birth the rapid rise in glomerular filtration rate is attributed to several factorsincluding increased mean arterial blood pressure and glomerular hydraulic pressure, a sharp declinein renal vascular resistance, with a redistribution of intrarenal blood flow from the juxtamedullaryto the superficial cortical nephrons (25).Immature glomeruli are present for months after birth and glomerular maturation and filtrationincreases during early infancy. Glomerular size follows a regular growth curve in childhood, withan average diameter of 100 µm at birth, progressing to 300 µm (1). However, glomerular filtrationrate does not increase proportionally with body size. Prior to gestation week 34, body size increasesbut glomerular filtration rate does not; after that time glomerular filtration rate increases morerapidly than body weight (26). Immature glomeruli are functional once capillary function has beenestablished (1).Comparative SpeciesThree stages of glomerular filtration rate development have been identified by studying variousspecies (13). The first stage is characterized by equivalent rates of increase in GFR and kidneymass. The second stage is characterized by a greater rate of increase in GFR than kidney growth.In the final stage, GFR increases at the same rate as that of kidney mass. These three stages do notnecessarily correlate with anatomical development across species. In addition to empirical measurements,allometric scaling has been used to predict GFR for various species (27).Changes in glomerular filtration rate parallel those for renal blood flow. Intrarenal blood flowoccurs at different rates for the inner and outer cortexes. Thus glomerular filtration developmentoccurs during different time periods as intra renal blood flow distributes between inner and outercortexes (13).Dogs — Inner cortical blood flow begins to increase with age about postnatal Day 12 in dogs.This results in a sharp decline in the ratio between inner and outer cortical blood flow, that has© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1011been reported as ~0.95 on Day 1 to 0.3 on Day 12 and throughout adulthood (13). In the case ofthe neonatal dog the developmental change in intra renal blood flow distribution correlates withthe period of nephrogenesis. An illustration of the timing of maturation of the inner cortical andouter cortical blood flow in dogs is presented below.130110IC/OC Flow Ratio90705030102 6 10 14 18 22 26 30 34 38Age (days)From: Kleinman, 1982. (13)Glomerular filtration rate increases with age postnatally in dogs. Based on clearance studiesusing small to large molecules, an increase in glomerular capillary surface area and pore densityoccurs between postnatal weeks 1 and 6 in dogs (28).Rats — Glomerular filtration rate sharply increases in rats during the first six weeks of postnatallife (29). A comparison of early postnatal glomerular filtration rates in rats and dogs is presentedbelow.Glomerular Filtration Rate(ml.min −1 /g Kidney)0.90.60.31 2 3Postnatal Age (months)Glomerular filtration rate in rat () and canine () kidney during postnatal maturation. From: Horster, 1977. (29)© 2006 by Taylor & Francis Group, LLC

1012 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSheep — The first stage of GFR development has been observed in fetal lambs. Blood pressureand renal vascular resistance effect renal blood flow. In the early postnatal period, there is little orno change in renal blood flow per gram kidney (13).In fetal sheep, renin-angiotensin type 1 receptor antagonist (10-mg/kg losartan) was dosed ongestation days 125 – 132. Although renal blood flow increased, glomerular filtration rate decreased,resulting in decreased filtration fraction. However, glomarulotubular balance was maintained sincethere was no change in sodium reabsorption (30).Rabbit — Rabbit studies indicate that body temperature of the neonate can to have an effect onglomerular filtration rate. In newborn rabbits, a 2 o C decrease in body temperature causes renalvasoconstriction accompanied by a decrease in glomerular filtration rate (25).Concentrating AbilityHumansThe newborn human infant is incapable of excreting concentrated urine at birth and this functionreaches maturity by the first year of life.The ability of the kidney to concentrate urine is controlled by mechanisms that control waterbalance, including antidiuretic hormone, short loops of Henle, low NaCl transport in the thickascending limb, and decreased tubular response to arginine vasopressin. Normal osmolality isregulated by antidiuretic hormone (ADH) controlled by the hypothalamus. The mechanism thatregulates ADH begins to operate 3 days after birth and the kidney becomes responsive to ADH atthat time (8). A series of active transports and passive diffusions occur along the tubule, withchloride actively transported out of the ascending tubule and remaining water is transported bydiffusion. The primary function of the tubules is to reabsorb glomerular filtrate and the loop ofHenle functions to concentrate and dilute urine. Tubular length and volume increases postnatally,allowing for increased capacity for transport and metabolism (1). The loop of Henle forms a hairpinturn as it enters and exits the medulla. At a specific point in the loop of Henle, urea enters into thedescending limb and plays an important role in generating a high sodium concentration in the fluid.Importantly for the neonate, quantities of urea in breast milk or formula are not sufficient formaximal urine concentration. Thus the neonate cannot excrete highly concentrated urine, but hasno difficulty in excreting dilute urine (1). Fluid from a higher level is progressively concentratedas water diffuses out of the descending limb (1). Maturational changes in the ability to concentrateurine are illustrated in the following table.Maximal Urine Osmolality by AgePostnatal Age mOsm/L3 days 151 ± 1726 days 663 ± 13310–30 days 896 ± 17910–12 months 1118 ± 15414–18 years 1362 ± 109From: Gomez et. al., 1999 (2)Sodium excretion also changes in the maturing infant, as the ability to respond to aldosteronematures. Premature infants excrete more sodium than do full term infants. At gestation week 31,the fractional excretion of sodium is about 5% and declines to about 1% by the second postnatalmonth (2). The term infant retains about 30% of dietary sodium, which is necessary to supportnormal growth; however, they have difficulty excreting an acute load of sodium and water, whichcan lead to edema. Low GFR and enhanced distal tubular sodium reabsorbtion is responsible for© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1013sodium retention. The ability to excrete a sodium load is fully developed by the end of the firstyear of life (2).Comparative SpeciesRats — Neonatal rats do not excrete concentrated urine at birth, but the concentration increasesdramatically with age (15). As single-nephron GFR increases with maturation, proximal tubularsodium reabsorption increases proportionally (13).Dogs — The fractional reabsorption of water in dogs is constant during postnatal maturation andsimilar in adults (13). Sodium excretion was studied in 1, 2, 3, and 6 week old puppies by expandingintravascular volume with either isotonic saline or isoncotic albumin in saline (31). Glomerularfiltration rate, sodium excretion, fractional excretion of sodium, and plasma volume measurementswere made. Of these parameters the sodium excretion was increased over controls at all ages. Thehighest level of sodium excretion was observed in 3-week-old puppies. The authors state that themechanism underlying the difference between the response to isotonic saline and isoncotic albuminin saline is already operative at birth in dogs (31). A graph depicting the timing of urine concentratingability in dogs is presented below.Cumulative Sodium Excretion (m Eq/kg)5.04.54.03.53.02.52.01.51.00.51 2 3 6Age (weeks)Cumulative sodium excretion in developing animals sustaining volume expansion with either saline (open bars)or isoncotic albumin (hatched bars). From: Aladjem, et al. 1982. (31)Results from other sodium loading studies in neonatal and adult dogs indicate that the pressurenaturesis occurs in the proximal tubules and the newborn proximal tubule is more sensitive to renalarterial blood pressure changes when compared to the adult (32).Rabbits — Using a polyclonal antibody directed at rabbit carbonic anhydrase IV, Schwartz et al.(18) noted the maturation pattern for the expression of the enzyme in the medulla of the maturingkidney paralleled that of the urine concentrating system. These investigators note that the localizationof the carbonic anhydrase IV expression within the kidney is important to its function. Forexample, in rabbits carbonic anhydrase IV is expressed in the outer medullar collecting ducts, butnot in rats. With the regard to function, the rabbit medulla matures after 21 days of age, and cleardistinction between inner an outer medulla is not visible at 3 weeks of age, but is at 5 weeks.Further investigation demonstrated that the maturational pattern observed in the inner medulla was© 2006 by Taylor & Francis Group, LLC

1014 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONsimilar to carbonic anhydrase IV expression that was approximately one-fourth adult levels at 2weeks of age and surged during postnatal weeks 3 and 4. The ability to concentrate urine developedalong this same timeline.Sheep — Renal water and electrolyte reabsorbtion have been studied in fetal sheep. The percentageof water, sodium and chloride reabsorbed increased throughout the later stages of gestation.However, electrolyte reabsorbtion exceeded water reabsorbtion to a degree that resulted in hypotonicurine until the last 2 weeks of gestation, at which time hypertonic urine was produced (8).Guinea Pigs — The fractional reabsorption of water in guinea pigs is constant during postnatalmaturation and similar in adults (13).Acid-Base EquilibriumHumansGrowth rate and the composition of intake determine renal acid-base control in infants. Excretedphosphate levels prior to birth and for the first two days after birth are very low, resulting in lowtitratable acidity (8). The intake of the fetus is relatively constant and regulated by the placenta,which limits the renal ability to contribute to acid-base equilibrium (8). As newborn feeding begins,the acidity varies directly in relation to the amount of protein, sulfate, and phosphate in the diet,and inversely with the rate of body growth. Walker (8) reports disturbances in equilibrium can beproduced by changes in intake in the two to three week old child, where these same changes wouldnot affect a 2-month-old child.Comparative SpeciesStudies in animals have demonstrated the role of enzymes in establishing and maintaining acidbase equilibrium. Carbonic anhydrase is a zinc metalloenzyme that catalyzes the hydration of CO 2and the dehydration of carbonic acid. Carbonic anhydrase activity has been detected in rat proximalconvoluted tubules and inner medullary collecting duct, and rabbit outer medullary collecting duct(18). Inhibition of luminal carbonic anhydrase has been found to diminish renal acid excretion andreduce HCO3- reabsorption (H + excretion) in the proximal tubule, suggesting an important role forcarbonic anhydrase IV in maintaining acid-base equilibrium. The authors hypothesized that the lowlevels of carbonic anhydrase in the neonatal kidney may help to explain the difficulty in maintainingacid-base homeostasis.Rats — In the rat carbonic anhydrase IV mRNA is expressed in the 20-day rat fetal kidney andincreases dramatically by postnatal day 17 (18).Rabbits — In rabbits, the expression of carbonic anhydrase was one-fourth adult levels at 2 weekspostnatally and surged to adult levels during postnatal weeks 3 and 4 (18).Dogs — In a study using mongrel dogs, postnatal excretion of uric acid decreased from 83% atbirth to 51% at 90 days of age (33). A direct correlation was observed between uric acid and sodiumclearance during early development. This study indicates acid-base homeostasis develops postnatallyin dogs.Urine Volume ControlUrine volume control in response to water diuresis is demonstrable in the human infant on Day 3after birth, and this capacity increases over a period of weeks (8). The diuretic response in infants© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1015differs from that observed in adults in that it is controlled by an increased GFR accompanied witha minimal reduction in specific gravity in the infant, whereas the adult GFR remains constant.Normal values for neonatal renal function are presented in the following table.Normal Values of Neonatal Renal FunctionPrematureFirst 3 daysTerm infantFirst 3 days2 WeeksDaily Excretion (ml/kg/24 hr.) 15–75 20–75 25–125Max. osmolality (mosmos/kg H 2 O) 400–500 600–800 800–900GFR (ml/min x 1.73m 2 ) 10–15 15–20 35–45Hentschel et al. 1996 (34)Renin-Angiotensin SystemThe role of angiotensin converting enzyme (ACE) activity differs between mature and immaturesystems. ACEs play a role in renal anatomical and functional development and maturation. Anillustration of the differences of angiotensin effects between the neonate and adult is presentedbelow.FoetusAIIBirthAdultAIIVascular Tubule Adrenal Vascular Tubule AdrenalBPGFRNo effect orinhibits NareabsorptionNo effectBPGFRStimulatesNareabsorptionAldosteroneSalt losingSalt RetainingDifferences between the renal actions of fetal and adult renin-angiotensin systems. From: Lumbers, 1995. (23)The developing kidney has different periods of susceptibility to either functional or anatomicalinjury. The normal development of the kidney can be further understood by studying abnormaldevelopment that has occurred in the presence of known xenobiotics.An example of an agent that interferes with normal renal development is the ACE inhibitor.The effects of this drug class on renal development are briefly reviewed as a case study below.ACE is a peptidyl dipeptidase that catalyzes the conversion of angiotensin I to angiotensin II,which in turn, acts as a vasoconstrictor. Angiotensin II also stimulates aldosterone secretion by theadrenal cortex. Inhibition of ACE results in decreased plasma levels of angiotensin II, and subsequentvasopressor activity and decreased aldosterone secretion. ACE inhibitors are prescribed asantihypertensives.© 2006 by Taylor & Francis Group, LLC

1016 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONHumansThe role of ACE inhibitors in infant renal failure has been well documented in the literature (35,36, 37, 38). Udwadia-Hegde et al. (38) present a review of case histories of mothers taking ACEinhibitors during the second and third trimesters of pregnancy. Oligohydramnios was observedwhich resulted in preterm delivery of neonates with intrauterine growth retardation, severe hypotension,and anuria. Biopsy of the kidneys generally showed renal tubular dysplaisia. Major anomaliesinduced by ACE inhibitors in humans include oligohydramnios, neonatal anuria/renal tubulardysgenesis, pulmonary hypoplasia, intrauterine growth retardation, persistent patent ductus arteriosus,calvarial hypoplasia/acalvaria, fetal or neonatal death (35). Since 1992, ACE inhibitorsmarketed in the United States carry a black box warning in the label regarding the use of this drugclass during pregnancy, as in the following label for Lotrel.When used in pregnancy during the second and third trimesters, angiotensin converting enzymeinhibitors can cause injury and even death to the developing fetus. When pregnancy is detected,Lotrel should be discontinued as soon as possible (Physician’s Desk Reference, 2000).Comparative SpeciesMice — Renin angiotensin is essential for the development of the mammalian kidney and urinarytract (19). Using mutant mice carrying a targeted null mutation of the angiotensin I or II receptor,Miyazaki and Ichikawa (19) demonstrated that both mutants have distinct phenotype in the kidneyand urinary tract system. Angiotensin II is involved in multiple aspects of the early morphogenesisof the kidney and urinary tract. Angiotensin I receptor induces the development of the renal pelvis,which promotes the removal of urine from the renal parenchyma. Failure of the angiotensin receptorsto operate properly during specific developmental stages results in congenital anomalies of thekidney and urinary tract in utero and hydronephrosis ex utero (19).Other investigators have noted the important role of renin-angiotensin in nephrogenesis, vascularization,and architectural and functional development of the kidney (39). Renal developmentwas studied using ACE inhibitors in neonatal rats and ACE mutant mice and similar renal pathologywas observed in both cases. The authors hypothesized that the primary lesion was a disturbance inrenal vessel development, with tubular pathology due to the close temporal and spatial relationshipof the tubules to the vessels during late stages of development (39).Rats — Rat studies illustrate the importance of the timing of exposure to ACE inhibitors as acritical factor in the development of altered renal morphology. The rat is susceptible to altered renalmorphology mainly during the last 5 days of pregnancy and the first two weeks of life (2, 39).Treatment of newborn rats with an ACE inhibitor for the first 12 days of postnatal life resulted inmarked renal abnormalities (2). Microscopic findings included relatively few and immature glomeruli,distorted and dilated tubules, relatively few and short thick arterioles that resulted in lessbranching and arrested maturation. The changes did not resolve after termination of treatment after23 days of age (2). Similarly, when rats were treated with enalapril on postnatal days 3 – 13,abnormalities in renal morphology were correlated with functional abnormalities (16). Functionalabnormalities included impairment of urinary concentrating ability, which correlated with the degreeof papillary atrophy. When losartan treatment began in 21-day-old rats or enalapril treatment beganin 14-day-old rats, no changes in renal morphology were observed (16, 36, 39). The timeframe inwhich rats are susceptible to renal injury induced by ACE inhibitor exposure correlates with thecritical period for nephrogenesis and marked tubular growth and differentiation.The human kidney is more mature than the rat kidney at birth, and therefore the adverse effectsof renin-angiotensin blockers on human kidneys are more likely to occur when exposure occursduring the last few weeks of pregnancy (39).© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1017Rabbits — A single oral dose of 30-mg/kg enalapril to pregnant rabbits on Day 26 resulted in100% fetal death (34). At doses within the human therapeutic range, fetal deaths occurred in middleto late gestation with a peak effect on gestational day 26 (of 31 in the rabbit) (36).The mechanism of fetal death has been postulated as decreased fetal-placental blood flow insheep and rabbits (35).Sheep — The fetal renin-angiotensin system in sheep (gestation days 125 – 132) helps to regulatefetal renal blood flow and is essential in the maintenance of fetal glomerular function (30, 40).Differences between fetal and adult renin-angiotensin systems have been documented using sheep(23). During gestation, the fetal renin-angiotensin system maintains glomerular filtration rate, thusthe author proposed that the renal excretion of sodium and water into the amniotic cavity, ensuringadequate amniotic fluid volume to support normal growth and development, was impaired. However,in the adult, the glomerular filtration rate is not regulated by angiotensin II. Angiotensin stimulatestubule sodium reabsorption in the adult but not in the fetus (23).Baboons — Baboons were treated with enalapril at a level comparable to a clinical dose used toachieve moderate but sustained ACE inhibition, and below doses generally selected for toxicologicstudy (41). Treatment began prior to mating and continued throughout pregnancy. Eight out of 13had adverse outcomes (fetal death or intrauterine growth retardation) compared with 0 of 13 in theplacebo group; no histopathologic evaluation was performed to determine the cause of death andno fetal malformations were observed (41). A direct effect on the fetal renin-angiotensin systemand placental ischemia has been postulated to be the mechanism of toxicity in baboons (35).ConclusionBoth anatomical and functional development of the kidney must be considered when makingcomparisons between species. This literature review has shown that the end of the anatomicaldevelopment of the kidney is marked by completion of nephrogenesis. Nephrogenesis is completeprior to birth in humans, monkeys, mice, sheep, and guinea pigs, and after birth in rats, dogs, andpigs. Maturation of renal function occurs during different time frames for different species. Thefocus of this paper was on the maturation of major renal functions including glomerular filtration,concentrating ability, acid-base equilibrium, and urine volume control. Glomerular filtration canbe detected as early as the first trimester of pregnancy and is important in tubular reabsorption ofNa + and Cl that helps to maintain sodium balance in amniotic fluid. Concentrating ability developspostnatally in humans, rats, rabbits, and sheep, and prenatally in dogs and guinea pigs. Acid baseequilibrium develops postnatally in all species reported which includes humans, rats, rabbits, anddogs. Control of urine volume also develops postnatally.In conclusion, design and interpretation of studies in prenatal and juvenile animals regardingrenal development should include careful consideration of the variability in time points of maturationof both anatomical and functional developmental milestones between species.REFERENCES1. Bernstein J. “Morphologic Development and Anatomy.” In: Rudolph’s Pediatrics. Eds. Rudolph, AM;coeditors, Hoffman, JIE, Rudolph, CD; assistant editor, Sagan P; associate editor, Travis LB, Chapter.25. The Kidneys and Urinary Tract. (Norwalk, CT: Appleton & Lange, 1991), pp.1223-1224.2. Gomez, R.A, Maria Luisa S. Sequeira Lopez, Lucas Fernadez, Daniel R. Chernavvksy, and VictoriaF. Norwood, “The Maturing Kidney: Development and Susceptibility,” Renal Failure, vol.21, no.3&41999: 283-91.3. Horster, MF, Gerald S. Braun, and Stephan M. Huber, “Embryonic Renal Epithelia: Induction,Nephrogenesis, and Cell Differentiation,” Physiological Reviews, vol.79, no.4 Oct. 1999: pp.1157-91.© 2006 by Taylor & Francis Group, LLC

1018 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION4. Langman J. “Urogenital Systems.” Chapt. 11 in Medical Embryology; Human Development-Normaland Abnormal. Baltimore: Williams & Wilkins, 1975.5. Larsson SH, and Aperia A. “Renal Growth in infancy and childhood – experimental studies ofregulatory mechanisms,” Pediatr Nephrol, vol.5 1991: pp.439-42.6. Norwood, Victoria F., Scott G. Morham, and Oliver Smithies, “Postnatal development and progressionof renal dysplasia in cyclooxygenase-2 null mice,” Kidney International, vol.58 2000: pp2291-2300.7. Witte MK, Stork JE, Blumer JL. Diuretic Therapeutics in the Pediatric Patient. Am J Cardiol. 1986;57: 44A-53A.8. Walker DG. “Functional Differentiation of the Kidney.” In: Intra-Uterine Development. Ed. Barnes,AC, Chapt. 13 (Philadelphia, PA: Lea & Febiger, 1968), 245-252.9. Gasser, B., Y. Mauss, J.P. Ghnassia, R. Favre, M. Kohler, O. Yu, J.L. Vonesch, “A Quantitative Studyof Normal Nephrogenesis in the Human Fetus: Its Implication in the Natural History of KidneyChanges due to Low Obstructive Uropathies,” Fetal Diagn Ther, vol.8 1993: pp.371-84.10. Holdsworth, SR, Glasgow EF, Atkins RC and Thomson NM, “Cell Characteristics of CulturedGlomeruli from Different Animal Species,” Nephron, vol.22, no.4-6 1978: pp.454-9.11. Gall, JA, Daine Alcorn, Aldona Butkus, John P. Goghlan, and Graeme B. Ryan, “Distribution ofglomerular peripolar cells in different mammalian species,” Cell Tissue Res, vol.244, no.1 1986:pp.203-8.12. Evan, AP, Vincent H. Gattone, II, and Philip M. Blomgren, “Application of scanning electron microscopyto kidney development and nephron maturation,” Scan Electron Microsc, pt.1 1984: pp.455-73.13. Kleinman LI. “Developmental Renal Physiology,” Physiologist. 1982 Apr; 25(2): 104-10.14. Fouser, M.D., Laurie and Ellis D. Avner, M.D., “Normal and Abnormal Nephrogenesis,” AmericanJournal of Kidney Disease, vol.21, no.1 Jan. 1993: pp.64-70.15. Kavlock, RJ, and Jacqueline A. Gray, “Evaluation of Renal Function in Neonatal Rats,” Biol Neonate,vol.41 1982: pp.279-88.16. Guron, Gregor, Niels Marcussen, Annika Nilsson, Birgitta Sundelin, and Peter Friberg, “PostnatalTime Frame for Renal Vulnerability to Enalapril in Rats,” J Am Soc Nephrol, vol. 10 1999: pp.1550-60.17. Langley-Evans, S.C., Simon JM Welham, and Alan A. Jackson, “Fetal exposure to a maternal lowprotein diet impairs nephrogenesis and promotes hypertension in the rat,” Life Sciences, vol.64, no.111999: pp.965-74.18. Schwartz GL, Olson J, Kittelberger AM, Matsumoto T, Waheed A, Sly WS. Postnatal developmentof carbonic anhydrase IV expression in rabbit kidney. Am J Physiol. 1999 Apr; 276(4 Pt 2): F510-20.19. Miyazaki Y, Ichikawa I. Role of the angiotensin receptor in the development of the mammalian kidneyand urinary tract. Comp Biochem Physiol A Mol Integr Physiol. 2001 Jan; 128(1): 89-97.20. Evan, AP, James A. Stoeckel, Vickie Loemaker, and Jeffrey T. Baker, “Development of the vascularsystem of the puppy kidney,” Anat Rec, vol.194, no.2 Jun 1979: pp.187-99.21. Olbing H, M. Donald Blaufox, Lorenzo C. Aschinberg, Geraldine I. Silkalns, Jay Bernstein, AdrianSpitzer, and Chester M. Edelmann, Jr., “Postnatal Changes in Renal Glomerular Blood Flow Distributionin Puppies,” J Clin Invest, vol.52, no.11 Nov. 1973: pp.2885-95.22. Tavani, N Jr, Philip Calcagno, Steve Zimmet, Walter Flamenbaum, Gilbert Eisner, and Pedro Jose,“Ontogeny of Single Nephron Filtration Distribution in Canine Puppies,” Pediatr Res, vol.14, no.6Jun 1980: pp.799-802.23. Lumbers ER. Functions of the renin-angiotensin system during development. Clin Exp PharmacolPhysiol. 1995 Aug; 22(8): 499-505.24. Seikaly MG, Billy S. Arant, Jr., “Development of Renal Hemodynamics: Glomerular Filtration andRenal Blood Flow,” Clin Perinatol, vol.19, no.1 1992: pp.1-13.25. Toth-Heyn P, Drukker A, Guignard JP. The stressed neonatal kidney: from pathophysiology to clinicalmanagement of neonatal vasomotor nephropathy. Pediatr. Nephrol. 2000 Mar; 14(3): 227-39.26. Arant, BS. Jr., The Newborn Kidney. In Rudolph’s Pediatrics. Eds. Rudolph, AM; coeditors, Hoffman,JIE, Rudolph, CD; assistant editor, Sagan, P; associate editor,27. Singer MA. Of Mice and Men and Elephants: Metabolic Rate Sets Glomerular Filtration Rate. Am JKidney Dis. 2001 Jan; 37(1): 164-178.28. Goldsmith, DI, Roberto A. Jodorkovsky, Julius Sherwinter, Stuart R. Kleeman, and Adrian Spitzer,“Glomerular capillary permeability in developing canines,” Am J Physiol, vol.251, no.3, pt.2 1986:pp.F528-31.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 101929. Horster, M, “Nephron Function and Perinatal Homeostatis,” Ann Rech Vet., vol.8, no.4 1977: pp.468-82.30. Stevenson KM, Gibson KJ, Lumbers ER. Effects of losartan on the cardiovascular system, renalhaemodynamics and function and lung liquid flow in fetal sheep. Clin Exp Pharmacol Physiol. 1996Feb; 23(2): 125-33.31. Aladjem, Mordechai, Adrain Spitzer, and David I. Goldsmith, “The Relationship between IntravascularVolume Expansion and Natriuresis in Developing Puppies,” Pediatr Res, vol.16, no.10 Oct 1982:pp.840-5.32. Kleinman LI, and Robert O. Banks, “Pressure natriuresis during saline expansion in newborn andadult dogs,” Am J Physiol, vol.246, no.6, pt.2 Jun 1984: pp.F828-34.33. Stapleton FB, and Billy S. Arant, Jr., “Ontogeny of Renal Uric Acid Excretion in the Mongrel Puppy,”Ped Res, vol.15, no.12 Dec 1981: pp.1513-6.34. Hentschel R, Lodige B, Bulla M. Renal insufficiency in the neonatal period. Clin Nephrol. 1996 Jul;46(1): 54-8. Review.35. Buttar HS. An overview of the influence of ACE inhibitors on fetal-placental circulation and perinataldevelopment. Mol Cell Biochem. 1997 Nov; 176(1-2): 61-71.36. Sedman AB, Kershaw DB, Bunchman TE. Recognition and management of angiotensin convertingenzyme inhibitor fetopathy. Pediatr Nephrol. 1995 Jun; 9(3): 382-5.37. Sorensen AM, Christensen S, Jonassen TE, Andersen D, Petersen JS. [Teratogenic effects of ACEinhibitorsand angiotensin II receptor antagonists]. Ugeskr Laeger. 1998 Mar 2; 160(10): 1460-4.Danish.38. Udwadia-Hegde A, Parekji S, Ali US, Mehta KP. “Angiotensin converting enzyme inhibitor fetopathy,”Indian Pediatr. 1999 Jan; 36(1): 79-82.39. Hilgers KF, Norwood VF, Gomez RA. Angiotensin’s role in renal development. Semin Nephrol. 1997Sep; 17(5): 492-501.40. Lumbers ER, Bernasconi C, Burrell JH. Effects of inhibition of the maternal renin-angiotensin systemon maternal and fetal responses to drainage of fetal fluids. Can J Physiol Pharmacol. 1996 Aug;74(8): 973-82.41. Harewood, W.J., Andrew F. Pippard, Geoffrey G. Duggin, John S. Horvath, and David J. Tiller,“Fetotoxicity of angiotensin-converting enzyme inhibition in primate pregnancy: A prospective, placebo-controlledstudy in baboons (Papio hamadryas),” Am J Obstet Gynecol, vol.171, no.3 Sep. 1994:pp.633-42ADDITIONAL RELATED REFERENCESAirede A, Bello M, Weerasinghe HD. Acute renal failure in the newborn: incidence and outcome. J PediatricChild Health. 1997 Jun; 33(3): 246-9.Bernardini N, Mattii L, Bianchi F, Da Prato I, Dolfi A. TGF-Alpha mRNA Expression in Renal Organogenesis:A Study in Rat and Human Embryos. Exp Nephrol. 2001 Mar; 9(2): 90-98.Capulong MC, Kimura K, Sakaguchi N, Kawahara H, Matsubara K, Likura Y. Hypoalbuminemia, oliguriaand peripheral cyanosis in an infant with severe atopic dermatitis. Pediatr Allergy Immunol. 1996May; 7(2): 100-2.Casellas D, Bouriquest N, Artuso A, Walcott B, Moore LC. New method for imaging innervation of the renalpreglomerular vasculature. Alterations in hypertensive rats. Microcirculation. 2000 Dec; 7(6 Pt 1):429-37.Dawson R Jr., Liu S, Jung B, Messina S, Eppler B. Effects of high salt diets and taurine on the developmentof hypertension in the stroke-prone spontaneously hypertensive rat. Amino Acids. 2000; 19(3-4): 643-65.De Heer E, Sijpkens YW, Verkade M, den Dulk M, Langers A, Schutrups J, Bruijn JA, van Es La. Morphometryof interstitial fibrosis. Nephrol Dial Transplant. 2000; 15 Suppl 6: 72-3.Forhead AJ, Gillespie CE, Fowden AL. Role of cortisol in the ontogenic control of pulmonary and renalangiotensin-converting enzyme in fetal sheep near term. J Physiol 2000 Jul 15; 526 Pt. 2: 409-16.Gomez, M.D., R.A., and Victoria F. Norwood, M.D., “Recent advances in renal development,” Current Opinionin Pediatrics, vol.11 1999: pp.135-40.© 2006 by Taylor & Francis Group, LLC

1020 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONGriffet J, Bastiani-Griffet F, Jund S, Moreigne M, Zabjek KF. Duplication of the leg-renal agenesis: congenitalmalformation syndrome. J Pediatr Orthop B. 2000 Oct; 9(4): 306-8.Hamar P, Peti-Peterdi J, Szabo A, Becker G, Flach R, Rosivall L, Heemann U. Interleukin-2-DependentMechanisms are involved in the development of the glomerulosclerosis after partial renal ablation inrats. Exp Nephrol. 2001 Mar; 9(2): 133-141.Hegde AU, Parekji S, Ali US, Mehta KP. Angiotensin converting enzyme inhibitor fetopathy. Indian Pediatr.1999 Jan; 36(1): 79-82.Ibrahim SH, Bhutta ZA, Khan IA. Haemolytic uraemic syndrome in childhood: an experience of 7 years atthe Aga Khan University. JPMA J Pak Med Assoc. 1998 Apr; 48(4): 100-3.Jones, MD, DP, and Russell W. Chesney, MD, “Development of Tubular Function,” Clin Perinatol, vol.19,no.1 Mar 1992: pp.33-57.Karnak I, Muftuoglu S, Cakar N, Tanyel FC. Organ growth and lung maturation in rabbit fetuses. Res ExpMed (Berl). 1999 Mar; 198(5): 277-87.Kleinman LI, and John H. Reuter, “Maturation of Glomerular Blood Flow Distribution in the New-Born Dog,”J. Physiol, vol.228 1973: pp.91-103.Kleinman, LI, “Developmental Renal Physiology,” Physiologist, vol.25, no.2 Apr 1982: pp.104-10.Komhoff, Martin, Jun-Ling Wang, Hui-Fang Cheng, Robert Langenbach, James A. McKanna, Raymond C.Harris, and Matthew D. Breyer, “Cyclooxygenase-2-selective inhibitors impair glomerulogenesis andrenal cortical development,” Kidney International, vol.57 2000: pp.414-22.Kusuda S, Kim TJ, Miyagi N, Shishida N, Litani H, Tanaka Y, Yamairi T. Postnatal change of renal arteryblood flow velocity and its relationship with urine volume in very low birth weight infants during thefirst month of life. J Perinat Med. 1999; 27(2): 107-11.Landau D, Shelef I, Polacheck H, Marks K, Holcberg G. Perinatal vasoconstrictive renal insufficiency associatedwith material nimesulide use. Am J Perinatol. 1999; 16(9): 441-4.Lankin VZ, Sherenesheva NI, Konovalova GG, Tikhaze AK. Beta-carotene-containing preparation carinatinhibits lipid peroxidation and development of renal tumors in rats treated with chemical carcinogen.Bull Exp Biol Med. 2000 Jul; 130(7): 694-6.Lavoratti G, Seracini D, Fiorini P, Cocchi C, Materassi M, Donzelli G, Pela I. Neonatal anuria by ACEinhibitors during pregnancy. Nephron 1997; 76(2): 235-6.Ludders JW, G.F. Grauer, R.R. Dubielzig, G.A., Ribble, J.W. Wilson, “Renal microcirculatory and correlatedhistologic changes associated with dirofilariasis in dogs,” Am J Vet Res, vol.49, no.6 Jun 1988: pp.826-30.McCracken GH, Ginsberg C, Chrane DF, Thomas ML, Horton LJ. Clinical pharmacology of penicillin in newborn infants. J Ped. 1973 April; 82(4): 692-698.McDonald MC, Mota-Filipe H, Paul A, Cuzzocrea S, Abdelrahman M, Harwood S, Plevin R, Chatterjee PK,Yaqoob MM, Thiemermann C. Calpain inhibitor I reduces the activation of nuclear factor-{kappa}Band organ injury/dysfunction in hemorrhagic shock. FASEB J. 2001 Jan 1; 15(1): 171-186.Neuhuber WL, Eichhorn U, Worl J. Enteric co-innervation of striated muscle fibers in the esophagus: Just a“hangover”? Anat Rec. 2001 Jan 1; 262(1): 41-46.O’Brien KL, Selanikio JD, Hecdivert C, Placide MF, Louis M, Barr BD, Barr JR, Hospedales CJ, Lewis MJ,Schwartz B, Philen RM, St. Victor S. Espindola J, Needham, LL, Denerville K. Epidemic of pediatricdeaths from acute renal failure caused by diethylene glycol poisoning. Acute Renal Failure InvestigationTeam. JAMA. 1998 Apr 15; 279(15): 1175-80.Okada T, Iwamoto A, Kusakabe K, Mukamot M, Kiso Y, Morioka H, Kodama H, Sasaki F, Morikawa Y.Perinatal Development of the Rat Kidney: Proliferative Activity and Epidermal Growth Factor. BiolNeonate. 2001 Jan; 79(1): 46-53.O’Rourke, Dawn A., Hiroyuki Sakurai, Katherine Spokes, Crystal Kjelsberg, Masahide Takahashi, SanjayNigam, and Lloyd Cantley, “Expression of c-ret promotes morphogenesis and cell survival in mIMCD-3 cells,” American Physiological Society 1999: pp.F581-88.Peneyra RS, and Roger S. Jaenke, “Functional and Morphologic Damage in the Neonatally Irradiated CanineKidney,” Radiat Res, vol.104, no.2, pt.1 Nov 1985: pp.166-77.Peruzzi, Licia, Bruno Gianoglio, Maria Gabriella Porcellini, and Rosanna Coppo, “Neonatal end-stage renalfailure associated with maternal ingestion of cyclo-oxygenase-type-1 selective inhibitor nimesulideas tocolytic,” The Lancelet, vol.354, no.9190 Nov. 6, 1999: pp.1615.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1021Querfeld U, Ortmann M, Vierzig A, Roth B. Renal tubular dysgenesis: a report of two cases. J Perinatol.1996 Nov-Dec; 16(6): 498-500.Ramirez O, Jimenez E. Opposite transitions of chick brain catalytically active cytosolic creatine kinaseisoenzymes during development. Int J Dev Neurosci. 2000 Dec; 18(8): 815-23.Sandberg K, Ji H. Kidney angiotensin receptors and their role in renal pathophysiology. Semin Nephrol. 2000Sep; 20(5): 402-16.Sitdikov FG, Gil’mutdinova RI, Minnakhmetov RR, Zefirov TL. Asymmetrical effects of vagus nerves onfunctional parameters of rat heart in postnatal ontogeny. Bull Exp Biol Med. 2000 Jul; 130(7): 620-3.Strehl, R., Will W. Minuth, “Nephron induction-the epithelial mesenchymal interface revisited,” PediatrNephrol, vol.16 2001: pp.38-40.Suzuki T, Kimura M, Asano M, Fujigaki Y and Hishida A. Role of Atrophich Tubules in Development ofInterstitial Fibrosis in Microembolism-Induced Renal Failure in Rat. Am J Pathol. 2001 Jan; 158(1):75-85.Taylor HA, Delany ME. Ontogeny of telomerase in chicken: impact of downregulation of pre- and postnataltelomere length in vivo. Dev Growth Differ. 2000 Dec’ 42(6): 613-21.Traebert, Martin, Marius Lotscher, Ralph Aschwanden, Theresia Ritthaler, Jurg Biber, Heini Murer, and BrigitteKaissling, “Distribution of the Sodium/Phosphate Transporter during Postnatal Ontogeny of the RatKidney,” J Am Soc Nephrol, vol.10 1999: pp.1407-15.Travis LB. The Kidneys and Urinary Tract. Rudolph’s Pediatrics. 1991; 19th Ed., Chapt. 25: 1223-1236.Tucker LB, Stehouwer DJ. L-DOPA-induced air-stepping in the preweaning rat: electromyographic andkinematic analyses. Behav Neurosci. 2000 Dec; 114(6): 1174-82.Vesna, Lackovic, and Mujovic Spomenka, “Postnatal development of the kidney juxtaglomerular appartus inrats,” Acta anat., vol.108 1980: pp.281-87.White, DVM JV, D.R. Finco, DVM, Ph.D., W.A. Crowell, DVM, S.A. Brown, DVM, Ph.D., D.A. Hirakawa,Ph.D., “Effect of dietary protein on functional, morphologic, and histolgic changes of the kidneyduring compensatory renal growth in dogs,” Am J Vet Res, vol.52, no.8 Aug 1991: pp.1357-65.Zhang G, Oldroyd SD, Huang LH, Yang B, Li Y, Ye R, El Nahas AM. Role of Apoptosis and Bcl-2/Bax inthe Development of Tubulointerstitial Fibrosis during Experimental Obstructive Nephropathy. ExpNephrol. 2001 Mar; 9(2): 71-80.Zhang SL, To C, Chen X, Filep JG, Tang SS, Ingelfinger JR, Carriere S, Chan JS. Effect of Renin-AngiotensisSystem Blockade on the Expression of the Angiotensinogen Gene and Induction of Hypertrophy inRat Kidney Proximal Tubular Cells. Exp Nephrol. 2001 Mar; 9(2): 109-117.© 2006 by Taylor & Francis Group, LLC

1022 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAPPENDIX C-3*SPECIES COMPARISON OF LUNG DEVELOPMENTTracey Zoetis 1 , and Mark E. Hurtt 2,31Milestone Biomedical Associates, Frederick, MD 21701, USA2Pfizer Global Research & Development, Groton, CT 06340, USA3Correspondence to: Dr. Mark E. Hurtt, Pfizer Global Research & Development, Drug Safety Evaluation, Eastern PointRoad, Mailstop 8274-1306, Groton, CT 06340. 860-715-3118. Fax (860) 715-3577. Email: mark_e_hurtt@groton.pfizer.comIntroductionThe purpose of this paper is to identify critical time frames for development of the lung in humansand to compare these events to other species. This comparison should result in data that will beuseful in designing and interpreting studies of the possible prenatal and/or postnatal developmentaleffects of chemicals and drugs on the developing lung. A brief description of growth and developmentof the human lung is presented as a baseline for comparison with other species. This isfollowed by an inter-species comparison of lung development.Growth and Development of the Human LungThe development of the human lung is a relatively steady and continuous process, arbitrarily dividedinto the following 6 stages: embryonic, pseudo-glandular, canalicular, saccular, alveolar, and vascularmaturation (1). The first 4 developmental stages are complete during fetal development. Atbirth, the human neonate has entered the alveolar stage of development. Over 80% of alveoli areformed after birth by a process of air space septation (1). The developmental stages are illustratedin Figure 1.Structural changes occur on a continuum with increases in the length of the respiratory tractand in number of alveoli as the child ages (3). Lung surface area increases as a result of the increasein the alveoli numbers. Lung surface area increases during late developmental stages may simplyrepresent an expansion of airspace of the lung. A schematic representation of human lung developmentat various stages was developed by Reid (3) as is presented in Figure 2.Alveoli increase in number and surface area with increasing age and begin to level off betweenthe ages of 2 to 4 years (4, 3). This is accompanied by decreasing interstitial tissue. The followinggraph illustrates the rate of development. Data are inconclusive regarding the timing that alveolarformation is complete in humans, and range from 2 years (1, 4) to 8 years (3).Accurate interpretation of this graph requires knowledge of the criteria used to determine thecompletion of alveolization. Although all authors did not explain this, interpretation of the endpointcould range from the age of detection of the last known immature alveolus to determination of thecritical timepoint for lung function. Another confounding factor in the interpretation of this graphis the fact that some of the data were obtained from post mortem examinations of patients whosealveolar development may have been compromised by disease. Dietert et al. (2000)(6) report thealveolar proliferation takes place during the first 2 years of life while expansion takes place duringthe ages of 2 – 8 years.* Source: Zoetis, T. and Hurtt, M. E., Species comparison of Lung development, Birth Defects Research, Part B:Developmental and Reproductive Toxicology, 68, 121-124, 2003.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1023EmbryonicperiodStage ofmicrovascularmaturationAlveolarstageCanaucularstagePseuooglandularstageSaccularstageNormal growthperiodLUNG DEVELOPMENTLUNG GROWTHFertilization10 20 30 3 6 9 1 2 3 4 5 6 7Weeks Months YearsAGEBirthFigure 1 Timing of stages of human lung development. (Reproduced from Zeltner and Burri, 1987(1)).Inter-species Comparison of Lung Growth and DevelopmentDevelopmental stages identified for the human lung are also found in other mammalian speciesreported in the literature. In this section, the growth and development of the pulmonary system incommon laboratory species will be discussed and compared with that of humans.The timing of each developmental stage and the degree of lung development at birth varieswidely between species (3). As an example, at birth the opossum lung is very primitive; rat andmouse lungs have no alveoli; kittens, calf and humans have relatively few alveoli; and the lamblungs are quite well developed. Postnatal development of the lung has also been studied in otherspecies: guinea pig, hamster, dog, monkey and humans. A comparative table of the timing of eachstage for several species is presented below (7).Species Glandular Canalicular Saccular AlveolarMouse 14–16 16.5–17.4 17.4– PD 5–Rat 13–18 19–20 21–PD PD 7–21Rabbit 19–24 24–27 27– —Sheep –95 95–120 120– —Human 42–112 112–196 196–252 252–childhoodRat Lung Growth and DevelopmentRats are perhaps the most extensively studied laboratory species for lung development (2). Manyhypotheses regarding human lung development are based on research conducted in rats. Postnatalchanges in the rat lung occur in 3 phases: short expansion; proliferation; and equilibration. Thefirst phase lasts from birth to day 4 of postnatal age. During this period, the lung transforms fromthe saccular stage into the alveolar stage. During the proliferating phase, alveolar numbers increasesmore rapidly than body weight. So does the lung volume and lung weight. During the last phase,© 2006 by Taylor & Francis Group, LLC

1024 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAgeLength fromTB to pleuraa16 wk gest 0–1 mmTBpleura0–1 mmb19 wk gest 0–2 mmTBRB 3RB 1 RB 2c28 wk gest 0–6 mmTBRB 1 RB 2RB 3TOS 1S 2S 2S 3S 3defbirth 1–1 mm2 months 1–75 mm7 years 4 mm TBRB 3TBRB 1RB 2RB 3TBRBRB 2 1RB 1RB RB 2 3TSTOS 1TSAD 1ADAD 2 AD 43ASAD 3 AD 4AD 1 AD 2AD 6AD 5 AtAtFigure 2 A schematic presentation of human lung development at various ages. TB = Terminal bronchiole; Rb i= respiratory bronchioles; TD = transitional duct; Ad i = Alveolar duct; At = Atrium; AS = alveolar sac. FromThurlbeck (3).the equilibrated growth phase, growth of the lung parallels overall body growth. This phase maylast until very late in age.The rat appears to be an acceptable model for study of juvenile populations because of similaritiesbetween rats and humans regarding the stages of lung development. The following table isa comparison of lung parameters between rats and humans.Lung Growth in Humans and Rats (NumbersRepresent the Fold Change from Newborn to Adult)Pulmonary Parameter Human RatLung volume (ml) 23.4 23.5Parenchyma airspace volume (ml) 30.2 26.9Septal volume (ml) 13.5 13.6Alveolar surface area (m 2 ) 21.4 20.5Capillary surface area (m 2 ) 23.3 19.2Source: Zeltner et al (1987)(4).Limitations exist, but can be overcome to some degree with careful planning for timing andduration of dosing. Nevertheless, there are differences between rats and humans in lung development.For example, rats are born during the saccular developmental stage and humans are bornduring the alveolar developmental stage. Rat lungs reach alveolar stage at the age of 4 days. Rat© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1025600500DunnillWeibelAngusDaviesHieronymi400?3002001000( )0 2 4 6 8 10 12 14 16 18 20Figure 3 Postnatal development of lung alveoli in humans. Data with a wide range on the right side was fromAngus and Thurlbeck (1972) (5). The solid (regression) line was calculated from Dunnill’s data only. Source:W.M. Thurlbeck, 1975(3).lungs develop rapidly, with most lung development complete within the first 2 weeks after birth.Furthermore, the rat lung continues to proliferate at a slow rate through out its life span. Bycomparison, human lung development continues until about age 3 to 8, with functional parameterspeaking at the age of 18 – 25 years (8).Dog Lung Growth and DevelopmentThe few studies that specifically address postnatal lung development in dogs present conflictingfindings. Boyden and Tompsett (1961)(9) reported that morphologically, dog lungs are slightly lessmature at birth than human lungs. Mansell and colleagues (1995)(10) report that lung developmentis more mature at birth in dogs than in humans. Biological variation may contribute significantlyto this difference. The degree of lung maturity at postnatal day 11 in dogs appears to be comparableto humans at birth (9). Evidence of lung development and growth activity was noted at the age of8 weeks. One investigator estimates the maximal functional efficiencies of the lung are reached atabout the age of 1 year in dogs compared to 20 years in humans (8). The dog is considered anacceptable species for testing the safety of inhaled drugs intended for pediatric populations.Other Species ConsideredOther species considered to date include the rabbit, sheep, pig, and monkey. These species werenot considered to be acceptable models of postnatal lung development because of their advanceddevelopment at birth when compared to humans (3, 10, 11, 12, 13, 14, 15, 16 and 17).© 2006 by Taylor & Francis Group, LLC

1026 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONConclusionsThe literature does not provide definitive evidence for the timing of the completion of alveolardevelopment, and a weight-of-evidence approach to address the safety of inhaled drugs intendedfor pediatric populations is needed. Inhaled drugs intended to treat children over 2 years of agegenerally do not require extensive testing in juvenile animals. For children over the age of 2 years,the pulmonary safety of the drug is demonstrated in clinical trials by measuring lung functionparameters during the course of the trial. The issue of local postnatal developmental toxicity forinhaled drugs is most important in children under 2 years of age. For the purposes of testing thesafety of inhaled agents in pediatric populations under 2 years of age, the rat and dog appear tobe acceptable models.REFERENCES1. Zeltner, T.B. and Burri, P.H. (1987). The postnatal development and growth of the human lung. II.Morphology. Respirat. Physiol. 67: 269-282.2. Burri, P.H. (1996). Structural Aspects of Prenatal and Postnatal Development and Growth of the Lung.3. Thurlbeck, W.M. (1975). Postnatal Growth and Development of the Lung American Review ofRespiratory Disease. Vol. 111.4. Zeltner, T.B., Cauduff, J.H., Gehr, P., Pfenninger, J. and Burri, P.H. (1987). The postnatal developmentand growth of the human lung. I. Morphometry. Respiration Physiology. 67:247-267.5. Angus, G.E., and Thurlbeck, W.M. 1972. Number of alveoli in the human lung. J Appl Physiol 32(4):483-5.6. Dietert, R.R., Etzel, R.A., Chen, D., Halonen, M., Holladay, S.D., Jarabek, A.M., Landreth, K., Peden,D.B., Pinkerton, K., Smialowicz, R.J., Zoetis, T. (2000). Workshop to Identify Critical Windows ofExposure for Children’s Health: Immune and Respiratory Systems Work Group Summary. EnvironmentalHealth Perspectives. Vol. 108, Supplement 3:483-90.7. Lau, C. and Kavlock, R.J. (1994). Functional Toxicity in the Development Heart, Lung, and Kidney.Developmental Toxicity, 2nd ed. 119-188.8. Mauderly, J.L., Effect of age on pulmonary structure and function of immature and adult animals andman. (1979), Fed. Proc. Vol. 38, No. 2. February 173-177.9. Boyden, E.A., and Tompsett, D.H. (1961). The Postnatal Growth of Lung in the Dog. Acta Anat. Vol.47, No. 3: 185 – 21510. Mansell, A.L., Collins, M.H., Johnson, E., Jr. and Gil, J. (1995) Postnatal Growth of Lung Parenchymain the Piglet: Morphometry Correlated With Mechanics. Anat. Rec. 241: 99-104.11. Winkler, G.C. and Cheville, N.F. (1985). Morphometry of Postnatal Development in the Porcine Lung.Anat. Rec. 211(4): 427-433.12. Mills, A.N., Lopez-Vidriero, M.T., and Haworth, S.G. 1986. Development of the airway epitheliumand submucosal glands in the pig lung: changes in epithelial glycoprotein profiles. Br J Exp Pathol67(6): 821-9.13. Zeilder, R.B. and Kim, H.D. (1985). Phagocytosis, chemiluminescence, and Cell volume of AlveolarMacrophages From Neonatal and Adult Pigs. J Leukoc Biol. 37(1): 29-43.14. Rendas, A., Branthwaite, M. and Reid, L., (1978). Growth of pulmonary circulation in normal pig –structural analysis and cardiopulmonary function. J. Appl. Physiol. 45(5): 806-817.15. Kerr, G.R., Couture, J., and Allen, J.R. (1975). Growth and Development of the Fetal Rhesus Monkey.VI. Morphometric Analysis of the Development Lung. Growth. 39:67-84.16. Boyden, E.A. (1977). Development and Growth of the Airways. In: Development of the Lung. W.A.Hodson, Ed. Marcel Decker, New York, pp. 3-35.17. Hislop, A., Howard, S. and Fairweather, D.V.I. (1984). Morphometric studies on the structural developmentof the lung in Macaca fascicularis during fetal and postnatal life. J. Anat. 138(1):95-112.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1027ADDITIONAL RELEVANT REFERENCESAdamson, I.Y.R. and King, G.M. (1986). Epithelial-Interstitial Cell Interactions in Fetal Rat Lung DevelopmentAccelerated by Steroids. Laboratory Investigations. Vol. 55, No. 2, 145.Barry, B.E., Mercer, R.R., Miller, F.J., and Crapo, J.D., (1988). Effects of Inhalation of 0.25 ppm Ozone onthe Terminal Bronchioles of Juvenile and Adult Rats. Experimental Lung Research. 14:225-245.Bisgaard, H., Munck, S.L., Nielsen, J.P., Petersen, W., and Ohlsson, S.V. (1990). Inhaled budesonide oftreatment of recurrent wheezing in early childhood. The Lancet. Vol. 336:649-651.Blanco, L.N. and Frank, L. (1993). The Formation of Alveoli in Rat Lung during the Third and Fourth PostnatalWeeks: Effect of Hyperoxia, Dexamethasone, and Deferoxamine. International Pediatric ResearchFoundation, Inc. Vol. 34, No. 3.Boyden, E.A., (1977). The Development of the Lung in the Pig-tail Monkey (Macaca nemestrina, L.). Anat.Rec. 186:15-38.Burri, P.H. (1974). The Postnatal Growth of the Rat Lung. III. Morphology. Anat. Rec. 180:77-98.Burri, P.H., Dbaly, J., and Weibel, E.R. (1973). The Postnatal Growth of the Rat Lung. I. Morphometry. Anat.Rec. 278:711-730.Carson, S.H., Taeusch, H.W, Jr., Avery, M.E., Inhibition of Lung cell division after hydrocortisone injectioninto fetal rabbits. Journal of Applied Physiology. Vol. 34, No. 5, May 1973.Chang, L-Y., Graham, J.A., Miller, F.J., Ospital, J.J., Crapo, J.D. (1986). Effects of Subchronic Inhalation ofLow Concentrations of Nitrogen Dioxide. Toxicology and Applied Pharmacology. 83: 46-61.Ellington, B., McBride, J.T., and Stokes, D.C. (1990). Effects of corticosteriods on postnatal lung and airwaygrowth in the ferret. J. Appl. Physiol. 68(5): 2029-2033.Hyde, D.M., Bolender, R.P., Harkema, J.R., Plopper, C.G. (1994). Morphometric Approaches for EvaluatingPulmonary Toxicity in Mammals: Implications for Risk Assessment. Risk Analysis. Vol 14, No. 3:293-302.Kamada, A.K., Szefler, S.J., Martin, R.J., Boushey, H.A., Chinchilli, V.M. Drazen, J.M., Fish, J.E., Israel, E.,Lazarus, S.C., Lemanske, R.F. (1996) Issue in the Use of Inhaled Glucocorticoids. Am. J. Respir. Crit.Care Med. Vol. 153. 1739-1748.Massaro, D., Teich, N., Maxwell, S., Massaro, G.D., Whitney, P. Postnatal development of Alveoli; Regulationand Evidence of a Critical Period in Rats. J. Clin. Invest. Vol. 76, October 1985, 1297-1305.Merkus, P.J.F.M., Have-Opbroek, A.A.W., and Quanjer, P.H. (1996). Human Lung Growth: A Review. PediatricPulmonology 21:383-397.Moraga, F.A., Riquelme, R.A., PharmD, López, A.A., Moya, F.R., Llanos, A.J. (1994). Maternal administrationof glucocorticoid and thyrotropin-releasing hormone enhances fetal lung maturation in undisturbedpreterm lambs. Am J. Obstet. Gynecol. 171(3): 729-734.Morishige, W.K. (1982). Influence of Glucocorticoids on Postnatal Lung Development in the Rat: PossibleModulation by Thyroid Hormone. Endocrinology Vol. 111, No. 5:1587-1594.Murphy, S. and Kelly, H.W. (1992). Evolution of Therapy for Childhood Asthma. Am. Rev. Respir. Dis.146:544-575.Odom, M.W., Ballard, P.L. Developmental and Hormonal Regulation of the Surfactant System. 495-575.Ogasawara, Y., Kuroki, Y., Tsuzuke, A., Ueda, S., Misake, H., Akino, T. Pre-and Postnatal StimulationSurfactant Protein D by In vivo Dexamethasone Treatment of Rats. Life Sciences, Vol. 50:1761-1767.Picken, J., Lurie, M. and Kleinerman, J. (1974). Mechanical and Morphologic Effects of Long-Term CorticosteroidAdministration on the Rat Lung. Am Rev. Respa. Dis. Vol. 110:746-753.Robinson, D.S., Geddes, D.M. (1996). Review Article Inhaled Corticosteroids: Benefits and Risks. J. Asthma.33(1): 5-16.Rooney, S.A., Dynia, D.W., Smart, D.A., Chu, A.J., Ingleson, L.D., Wilson, C.M. and Gross, I. (1986).Glucocorticoid stimulation of choline-phosphate cytidylyltransferase activity in fetal rat lung: receptorresponserelationships. Biochem. Biophys. Acta. 888(2): 208-216.Sahebjami, H. and Domino, M. (1989). Effects of Postnatal Dexamethasone Treatment of Development ofAlveoli in Adult Rats. Exp. Lung Res. 15(6): 961-973.Schellenberg, J-C., Liggins, G.C., (1987). New approaches to hormonal acceleration of fetal lung maturation.J. Perinat. Med. 15(5): 447-451.Sindhu, R.K., Rasmussen, R.E. and Kikkawa, Y. (1996). Exposure to Environmental Tobacco Smoke Resultsin an Increased Production of (+)-anti-Benzo[a]pyrene-7,8-Dihydrodiol-9,10-Eposide in Juvenile FerretLung Homogenates. J. Toxicol. Environ. Health 46 (6): 523-534.© 2006 by Taylor & Francis Group, LLC

1028 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTabor, B.L., Lewis, J.F., Ikegami, M., Polk, D. and Jobe, A.H. (1994). Corticosteroids and Fetal InterventionInteract to Alter Lung Maturation in Preterm Lambs. Pediatr Res. 35(4): 479-483.Tough, S.C., Green, F.H.Y., Paul, J.E., Wigle, D.T., Butt, J.C. (1996). Sudden Death from Asthma in 108Children and Young Adults. J. Asthma. 33(3): 179-188.Van Essen-Zandvliet, E.E., Hughes, M.D., Waalens, H.J., Duiverman, E.J., Pocock, S.J., Kerrebijn, K.F. andThe Dutch Chronic Non-Specific Lung Disease Study Group. (1992). Am. Rev. Respir. Dis. 146: 547-554.Wallkens, H.J., Vanessen-Zandvliet, E.E., Hughes, M.D., Gerritsen, L., Duiverman, E.J., Knol, K., Kerrebijn,K.F., and The Dutch CNSLD Study Group (1993). Am. Rev. Respir. Dis. 148:1252-1257.Ward, R.M. (1994). Pharmaoclogic Enhancement of Fetal Lung Maturation. Clin Perinatol. 21(3): 523-542.Weiss, S.T., Tosteson, T.D., Segal, M.R., Tager, I.B., Redline, S. and Speizer, F. (1992). Effects of Asthma ofPulmonary Function in Children. Am. Rev. Respir. Dis. 145: 58-64.Yeh, H-C., Hulbert, A.J., Phalen, R.F., Velasques, D.J., Harris, T.D. (1975). A Stereoradiographic Techniqueand Its Application to the Evaluation of Lung Casts. Investigative Radiology. Vol. 10, July-August:351-357.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1029APPENDIX C-4*DEVELOPMENT AND MATURATIONOF THE MALE REPRODUCTIVE SYSTEMM. Sue Marty, 1,5 Robert E. Chapin, 2 Louise G. Parks, 3 and Bjorn A. Thorsrud 41Dow Chemical Company, Midland, MI 48674, USA2Pfizer Global Research & Development, Groton, CT 06340, USA3Merck & Company, West Point, PA 19486, USA4Springborn Laboratories, Inc., Spencerville, OH 45887, USA5Correspondence to: Dr. M. Sue Marty, Dow Chemical Company, Toxicology Research Laboratory, 1803 Building,Midland, MI 48674, 517-636-6653, Fax 517-638-9863, Email: martym@dow.comIntroductionThis review briefly describes some of the key events in the postnatal development of the malereproductive system in humans, non-human primates, rats and dogs. Topics discussed includedevelopment of the testes, epididymides, the blood-testis barrier, anogenital distance, testiculardescent, preputial separation, accessory sex glands (prostate and seminal vesicles), and the neuroendocrinecontrol of the reproductive system. The objective of this work is merely to allow thereader to make initial comparisons of the developmental processes and timing of these events inhuman versus animal models. This review is not intended to be comprehensive, but merely providesan initial overview of these processes. In some cases, information was not available for all species.Available information is summarized in Table 1.1 Reproductive Organs1.1 Testes1.1.1 HumanIn the human, spermatogenesis does not begin until puberty; however, the prenatal, early postnataland prepubescent testis plays a critical role in hormone production. In the early postnatal testis,immature Sertoli cells are the most common cell type 1,2 with limited numbers of germ cells in arelatively undifferentiated state. According to Cortes et al. 2 total Sertoli cell number increases fromthe fetal period through childhood, puberty and early adulthood. In contrast, Lemasters et al. 3reported that Sertoli cells proliferate after birth, ceasing at 6 months when the adult number ofSertoli cells are achieved. Sertoli cells secrete inhibin B until 2-4 years of age and anti-Müllerianhormone during the entire prepubescent period. 4 In humans, there are three known testosteronesurges, one from 4-6 weeks of gestation, one from 4 months of gestation to 3 months of age, andthe last from 12 to 14 years of age. 5 Consistent with the early postnatal increase in testosterone,there is a biphasic increase in Leydig cell number, which includes an early increase in Leydig cells,followed by a decrease to the lowest level at 1.5 years of age, 6 then a continuous increase in Leydigcells until their numbers plateau in adulthood. In infant boys, an increase in total germ cell numberoccurs with peak numbers achieved between 50 and 150 days of age, followed by a decrease inolder boys. 7 During the early proliferative period, there is an overall decrease in germ cell density* Source: Marty, M. S., Chapin, R. E., Parks, L. G., and Thorsrud, B. A., Development and maturation of the malereproductive system, Birth Defects Research, Part B: Developmental and Reproductive Toxicology, 68, 125-136, 2003.© 2006 by Taylor & Francis Group, LLC

Table 1Human Non-human Primate Rat Dog Mouse0-7 days NeonatalDevelopmental Stages 0-1 month Neonatal1 m – 2 yr Infantile2 –12 yr children12-16 yr Adolescents 127 8-21 days Infantile21-35 days juvenile35- (55-60) daysperipubertal dependingon sex 20Anogenital DistanceSimilar to rat 572.5 X > in males(AGD)compared to females 49Preputial Separation Begins during lateSprague Dawley PND(PPS)gestation 6642-46 58Complete from 9 monthsAndrogens play key roleto 3 years of age 67,68,69.Androgens play keyrole 70Puberty 12-14 yrs 9,10,11,13 2.5 years of age 19 Early puberty begins atPPS (~PND 43) 58Prostate StructureNot sharply demarcatedand appears as a singlegland with several zones.It is the middle lobe thatobstructs the urethra inmen with enlargedprostate. Lateral lobe,dorsal (or posterior) lobe,and median (or middle)lobe 74Discrete lobes (ventral,lateral, dorsal and thepaired anterior lobes,also known as thecoagulating glands).Dorsal and lateral losedistinct borders atadulthoodForm lobes between PND1-7, tubular lumenbetween PND 7-14,secretory granules PND14-21 and shows adultcytology PND 28-35Secretory activityapproaches adult levelsca. PND 43-46 73,883 phases of testisgrowth 52 : 1: 0-22 wks2: 22-36 wks3: 36-46 wks34-36 weeks, as defined PND 35by presence ofejaculated sperm 41, 52The only well-developedaccessory sex gland inthe dog. Completelysurrounds the urethra 73and divided into right andleft lobes with middlelobe either poorlydeveloped or absent.Dogs are the only labspecies thatspontaneously developsbenign prostatichypertrophy (BPH).There are differencesbetween dogs andhumans in theirresponse toantiandrogentreatment. 84,85 Prostatesecretion starts at ~4monthsVentral prostate anddorsolateral prostateincrease dramaticallybetween PND 1-15,reach adult secretorylevels by ~PND 30. Formore details seeSugimura et al 89,90,911030 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Human Non-human Primate Rat Dog MouseSeminal Vesicles (SV)SpermatozoaLeydig CellsThe SV are present bygestational month 6 andhave attained adult formby the 7th month ofgestation. Developmentof the muscular wall ismediated by estrogenstimulation. Growth ofthe SV continues slowlyuntil puberty. 66 Secretoryactivity of the SV isandrogen dependent. 95Spermatogonia increase6-fold between birth and10 years of age, thenincrease at puberty.Mean age at spermarcheis 13.4 years; ejaculationpossible during middle tolate puberty. 1,8,13,16Testosterone (T)producing cells, Leydigcells (LC), beginproducing T ~7-8 weeksof gestation 96 andultimately come undercontrol of placentalgonadotrophin (humanchorionic gonadotrophin,hCG). 97 Pituitarygonadotrophin synthesisbegins at ~ 12 weeks ofgestation after Tproduction begins. 98In the testis,spermatogonia becomemore numerous by theend of the 1 st year.Spermatozoa in thetestis appear as early as3 yrs with fertility startingat approximately 3.5yrs. 17,19LC are prominent in fetallife but decrease innumber during the 1st yrand dedifferentiate. Bythe end of 3 yrs the LCsredifferentiate. T levels inimmature males is ~30-250 ng/100 ml rising to230-1211 ng/100ml inmature males. 18The basic pattern of theSV are present at PND10, with lumen formationfrom PND 2-15 andmarkedly increasing insize between PND 11-24with adult appearanceand secretory propertiesbetween PND 40-50. 84Secretory granules areevident at PND 16.In seminiferous tubulesPND 45, 31 in vasdeferens PND 58-59. 32Spermatogenesis isstimulated by increasedtestosterone andgonadotrophinproduction.Testosterone productionbegins during lategestation and decreasesjust prior to birth. Duringthe infantile-juvenileperiod (PND 8-35) theprimary androgensproduced includeandrostenedione, 5alpha-androstanediol,and dihydrotestosteroneNOT testosterone 105,106but by ~PND 25 andonward testosteronebecomes the primaryandrogen 105,107 Differentandrogen levels areprimarily due to changesin steroidogenic enzymelevels or activity.The dog has neither SVnor bulbourethralglands 73In testis at 26-28 wks ofage. First visible in theepididymis 26-28 wks(beagles 41 ); first inejaculate 32-34 weeks(fox terriers 52 )First visible histologicallyGD 36-46. 39 TesticularLH receptors begin toincrease at 2 mos., reachmax. 12-24 mos. Testic.T and DHT levels beginto increase at 6 mos.,plateau 12-24 mos(beagles 128 )LCs proliferation isdependent ongonadotrophin and ends~ PND 21-33 33POSTNATAL DEVELOPMENTAL MILESTONES 1031© 2006 by Taylor & Francis Group, LLC

Table 1 (continued)Sertoli CellsHuman Non-human Primate Rat Dog MouseSertoli cell (SC)differentiation andproduction ofantimullerian hormone(MIS) begins during theend of the first trimester 18and is triggered by anunknown mechanismmediated by the Ychromosome.Testis DescentTestis descent occursprenatally 62,63The testes descend atbirth but soon after birthascend into the inguinalcanal (postnatalregression). At ~3 yrs thetestes descend whileincreasing in size 64,65Epididymal Ontogeny Undifferentiated PND 0-15, differentiation PND16-44 and expansionIncrease in number at First visible GD 36-46~ GD 16 with peak (Beagles 39 ) Divide updivision at GD19 and until 8 wks post-partum;cease division at ~PND stable thereafter 4114-16 23,24,25,26 . Folliclestimulating hormone(FSH) receptors on SCsincreases markedlybefore birth 27,28,22 andpeaks ~PND 18 with adecline until adult levelsmet ~PND 40-50 29Testes attached to internal Passage of testes throughinguinal ring Gestational inguinal canal beginsday 20-21PND 3 or 4. DescentTestes into scrotum complete ca. PND~ PND 15 60 35-42 61Postnatally: low columnarepithelium in allsegments from birth to> PND 44 46 20 wks post-partum.Diameter of luminaincrease slowly untilweek 20, after whichthere is a burst of growth,which levels off ca. week48 (caput), 36 (corpus)or >48 (cauda). 41Division is under theinfluence of pituitarygonadotrophin and endsat PND-17. 36,37,381032 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1033partially due to increased testicular volume. Spermatogonia increase in number 6-fold betweenbirth and 10 years of age 1,8 and this number increases exponentially at puberty along with anincrease in testicular volume.Puberty signals the trigger for the daily production of millions of spermatozoa. Spermarche,the age at which spermatogenesis begins, occurs early in puberty and can be verified by theappearance of sperm in the urine. Increasing levels of testosterone play a key role in the initiationof spermatogenesis at puberty. With the initiation of the spermatogenic cycle, the development ofthe tubular lumen and the formation of the blood-testis barrier are critical events. The start ofpubertal testicular growth in healthy boys (one-sided testicular volume 3-4 ml) occurs between11.8 and 12.2 years. 9,10,11 Precocious puberty is before the age of 9 years. 12 Nielson et al. 13 reportedthe mean age at spermarche as 13.4 years (range: 11.7 to 15.3 years). Approximately two yearsafter spermarche, adult levels of testosterone are achieved. 14 From puberty onset, 3.2 ± 1.8 yearsare required to attain adult testicular volume. 15 Although spermatozoa are produced early in puberty,ejaculation is not possible until middle or late puberty. This signifies the onset of male fertility.Unlike spermatogonia, Sertoli cells and Leydig cells do not proliferate in the adult. 161.1.2 Non-Human Primate (Rhesus; Macaca mulatta)The steroid hormone producing cells of the testes are the leydig cells (LC) and are prominentduring fetal life. During the first year the LCs decrease in number and dedifferentiate entering aperiod of suspended development. By the end of the third year the LCs redifferentiate and beginto produce the primary steroid hormone testosterone. 17 The testosterone levels in immature malesis approximately 30 to 250 ng/100ml and in the mature male range from 230 to 1211 ng/100ml. 18In the Yale colony, puberty began at approximately 2.5 years of age. 19 Spermatogonia in thetestis become more numerous by the end of the first year but the earliest appearance of spermatozoais approximately 3 years of age.1.1.3 RatIn contrast to humans and primates, rats do not exhibit a period of testicular quiescence in whichthere is a sustained interruption in gonadotrophin secretion. In rats, postnatal testicular developmentbegins early and maturation steadily progresses. Ojeda and colleagues described male rat postnatalsexual development in 4 stages: 20 a neonatal period from birth to 7 days of age, an infantile periodfrom postnatal days 8-21, a juvenile period which extends to approximately 35 days of age, and aperipubertal period until 55 to 60 days of age, ending when mature spermatozoa are seen in thevas deferens. Spermatocytes, which arise from gonocytes in the fetal testis, cease dividing ongestation day 18 and many degenerate between postnatal days 3 and 7. The remainder (often asfew as 25% of those present at birth) begin to divide to form the first spermatogonia. 21 Testicularluteinizing hormone (LH) receptors, interstitial cells and testicular testosterone content increaseduring gestation, reaching a maximum at about the time of birth. 22 Concentrations of testosteronedecline soon after birth. Sertoli cells within the seminiferous tubules begin to increase in numberat approximately GD 16, reach a peak of division near GD19, and cease at approximately postnatalday 14-16. 23,24,25,26 Similarly, Sertoli cell follicle stimulating hormone (FSH) receptors increasemarkedly before birth. ,27,28,22 Sertoli cells do not proliferate after postnatal day 16. 16 FSH responsivenesspeaks postnatally about PND 18, and declines thereafter until it reaches adult levels atabout PND 40-50. 29 There is a rapid proliferation of Leydig cells between days 14 and 28 in therat with a second cell division between days 28 and 56. Similarly, androgen production increasesslowly until 28 days of age, then increases rapidly until 56 days of age. 30 Leydig cells typically donot proliferate in adult rats, although this can occur to replace damaged or destroyed cells.© 2006 by Taylor & Francis Group, LLC

1034 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIn male rats, the first spermatozoa appear in the lumen of the seminiferous tubules at 45 daysof age 31 and transit through the epididymis to the vas deferens, where they can be detected at 58-59 days of age. 32 Testicular sperm reach a plateau by 77 days of age. 30In contrast to these values, Maeda et al. cited the following postnatal developmental time linesfor Wistar rats: 33 testicular spermatids first appear in the testis at day 20-30 after birth and 100%of animals contain testicular sperm at day 70. In the tail of the epididymis, some males containsperm at approximately day 40 and almost all animals do so by day 90. The relative weights ofthe testis and epididymis reach their peak value at around day 70. Testicular weight is around 1%of body weight at 50 days after birth, the time of puberty.It is important to note that the first wave or two of spermatogenesis in rats is quite inefficient,and there is much greater cell loss peripubertally than there is in adult rats. 34 Thus, studies evaluatinganimals peripubertally should expect to see greater levels of cell death and structural abnormalitiesin the controls, which will complicate the identification of such effects in treated animals. Thetestes continue to increase in size after puberty due to increased total sperm production, althoughproduction per unit weight of the testis rapidly reaches a maximal value. 211.1.4 MousePrenatally, germ cells and Sertoli cells migrate from the mesonephric ridge much as in the rat. 35Postnatally, the earliest Type A spermatogonia can be observed at PND 3, and these are proliferating.Sertoli cells proliferate until PND17 under the influence of pituitary gonadotrophins. 36,37,38 Leydigcells proliferate much later, closer to puberty (PND 21-33), 15 and this is dependent on gonadotrophinin serum. 38 Puberty in the mouse is approximately PND 35.1.1.5 DogTestis differentiation was observed at GD 36 in Schnauzers and beagles; 39 this was followed closelyby regression of the Müllerian ducts. Postnatally, the seminiferous tubules are composed of immatureSertoli cells and gonocytes at 2 weeks of age. During the first 3 postnatal weeks, the Leydigcells appear to be mature, then they appear to regress from weeks 4 to 7, although there is noappreciable change in androgen levels. At 8 weeks of age, Leydig cells appear to be active andmitotic figures can be found in both Sertoli cells and spermatogonia. Between 16 and 20 weeks,the number of germ cells per cross section of cord decreases and the amount of lipid increases. At16 weeks, evidence for the leptotene stage of meiotic prophase is evident as condensed spermatogonialchromatin. 40 At wk 18-20, the germ cells begin their rapid division, and tubule cellularity anddiameter increase rapidly. 41 Round spermatids begin to appear at wk 22, long spermatids at wk 26,and by wk 28 in beagles, diameter is nearly at adult values, when all cell types are represented.1.2 Epididymides1.2.1 HumanAs with other mammals, development of the epididymis is dependent upon androgens from thetesticular Leydig cells. Furthermore, differentiation of the epididymal epithelium is dependent uponconstituents of luminal fluid from the testis or proximal epididymis. 421.2.2 RatLike the testis, the epididymis arises embryologically from the middle of three nephroic regionsin the fetus. The most caudal region gives rise to the kidney. The cells in the mesonephros, incontrast, migrate caudally, and forms a diffuse networks of ducts. In the presence of Müllerian© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1035inhibiting substance, the cranial portion of the Wolffian duct forms the epididymis. These cellsdevelop and proliferate under the influence of testosterone, and is formed into a single tubule bylate gestation. 43 Prenatal exposure to compounds that reduce testosterone synthesis androgenicsignal interfere with epididymal development, frequently resulting in the absence of whole sectionsof the organ. 44,45Sun and Flickinger 46 name three periods of rat epididymal ontogeny: an undifferentiated period(birth to PND15), a period of differentiation (PND16-44), and a period of expansion (>PND44).DeLarminat et al 47 found the time of greatest cell division was PND25. This picture becomes morecomplex when cell division is considered by region, 48 but still, the vast majority of cell divisionoccurs prior to PND30. 49 This is reasonably consistent with the picture presented by Limanowskiet al. 50 Both testosterone and the arrival of germ cells and their bathing fluid from the rete testisare believed to contribute to epididymal differentiation. 16 Spermatozoa appear in the epididymis at49 days of age and reach their highest levels in the epididymis at 91 days of age. This periodcorresponds with the intervals with the greatest increases in epididymal weight between days 49-63 and 77-91. 30Another critical element in epididymal development is the formation of the blood-epididymisbarrier, which is complete in rats by postnatal day 21, prior to the appearance of spermatozoa inthe epididymal lumen. 511.2.3 DogKawakami 41 found that beagle epididymal epithelial cell height and duct diameter both rose veryslowly postnatally until ca. postnatal wk 22, whereupon both measures increased sharply and thenleveled off. Visible epididymal sperm apparent density was zero at wk 24, “a small number” at wk26, more at wk 28, and apparent sperm density in the epididymis plateaued at and after wk 30.Mailot 52 (using fox terriers) followed ejaculated sperm measures from postnatal wks 30-57, andfound a continual increase in count; obvious influences on this number could be the underlyingmaturation of spermatogenesis, confounded by the dog’s accommodation to the sample collectionprocess.1.3 The Blood-Testis Barrier1.3.1 HumanAt 5 years of age, some junctional particles are visible in freeze-fractured preparations of the humantestis. By 8 years of age, rows and plaques of communication junctions are present. Lanthanumreadily penetrates into the intercellular space. The blood-testis barrier is completed at puberty, whena continuous belt of junctional particles can be seen at the onset of spermatogenesis. 531.3.2 RatPermeability of the blood-testis barrier decreases markedly between 15 and 25 days of age. 54,55 TheSertoli cell tight junctions form between postnatal days 14 and 19, concurrent with the cessationof Sertoli cell divisions, and the movement of the first early spermatocytes up to the luminal sideof the barrier. The blood-testis barrier in adults excludes even smaller molecules than those restrictedfrom entering the testis in prepubescent animals. 561.3.3 DogAt 8 weeks of age, there are no occluding junctions between Sertoli cells in the beagle testis, butseptate junctions are present that partially limit the penetration of lanthanum. At 13 to 17 weeks© 2006 by Taylor & Francis Group, LLC

1036 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof age, Sertoli cells exhibit incompletely formed tight junctions, which appear in patches. By 20weeks of age, these tight junctions appear linearly arranged and connected with the septate junctions.At this time, lanthanum can no longer penetrate into the adluminal compartment. 401.4 Anogenital Distance1.4.1 RatIn rats, gender is often determined by external examination of anogenital distance (AGD) duringearly neonatal periods. AGD, defined as the distance between the genital tubercle and the anus, isapproximately 2.5X greater for males than females. A similar sex difference in AGD has beenreported for rhesus monkey fetuses. 57 It is not clear whether a change in AGD in laboratory animalscorresponds to an adverse effect in humans. There are no reports of decreased AGD in pseudohermaphroditicmen, who lack 5-reductase, and finasteride, a 5-reductase inhibitor, failed to alter theAGD of monkeys at doses causing hypospadias. 57 Age-related changes in AGD are cited in Clark. 29Mean AGD for control groups of Sprague-Dawley rats on postnatal day 0 was 3.51 (3.27-3.83)mm for males and 1.42 (1.29-1.51) mm for females. Absolute AGD is affected by body size. 58,59Note that there may be some interlaboratory variation in AGD measurements.1.5 Testicular DescentThe process of testicular descent is similar in most species, although there are some timing andtopographical differences (e.g., gubernacular outgrowth, mesenchymal regression and developmentof the cremaster muscle 60 ). Basically, the testicles are attached to the mesenephros intra-abdominallyduring gestation. During testicular descent, the mesenephros degenerates and the gubernaculumproprium increases in size, thereby dilating the inguinal canal and allowing the testicle to passthrough. Once testicular descent is complete, the gubernaculum shortens, allowing the testicle tomove into the scrotum. 611.5.1 HumansHogan et al. 62 cites gestation weeks 7 to 28 as the period for testicular descent in humans whereasGondos 63 identifies the seventh month of gestation as the time the processus vaginalis grows andthe inguinal canal increases in diameter for passage of the testis. Descent occurs due to degenerationof part of the gubernaculum. Thus, Gondos 63 cites the latter part of gestation as the period whenthe testis completes its descent into the scrotum.1.5.2 Non-Human PrimateTesticular descent occurs at birth. 64 However, soon after birth testes ascend into the inguinal canaland precedes into a postnatal regression. 65 At approximately 3 years of age the testes descend againwhile also increasing in size.1.5.3 DogTesticular descent is not complete in dogs at the time of birth. Most of the mesonephros degenerationis complete prior to initiation of testicular descent and there is little increase in testicular size whilethe testes remain in the abdomen. Passage of the testicle through the inguinal canal occurs onpostnatal day 3 or 4. Regression of the gubernaculum begins after testicular passage and is completeby 5-6 weeks of age. 61© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 10371.5.4 RatOn gestation day 20-21, the testicle in the rat is attached to the internal inguinal ring with its caudalpole and the cauda epididymis located within the canal. Enlargement of the gubernaculum occurspostnatally with descent of the testicles occurring on approximately postnatal day 15. Unlike manyspecies, the inguinal canal remains wide in the adult rat, allowing the male rat to lift its testiclesinto its abdomen. 601.6 Preputial Separation1.6.1 HumanIn humans, PPS begins during late gestation 66 and is generally completed postnatally between 9months and 3 years of age. 67,68,69 Androgens are assumed to play a role in PPS in humans. 701.6.2 RatAn external sign of puberty onset in the male rat is balano-preputial separation (PPS) when theprepuce separates from the glans penis. 71 Initially, the rat penis looks similar to the clitoris of thefemale and the glans of the penis is difficult to expose by 30 days after birth. 33 The shape of thetip of the penis changes from a V-shape to a W-shape (20-30 days after birth) to the U-shape, whichis seen in 100% of animals by day 70 after birth. 33 Mean age at PPS for Sprague-Dawley rats was43.6 ± 0.95 (X + SD) days of age (range: 41.8-45.9 58 ). There is a complex relationship betweenthe onset of puberty and body weight. 722 Accessory Sex Glands2.1 ProstateAlthough serving a similar function, the prostate gland varies among mammals, making selectionof an animal model problematic. For example, latent cancer of the prostate is a high incidencedisease in older men, yet it is rare among animals. In animal models, neither prostate cancer(spontaneous or induced) nor prostatic hypertrophy parallels exactly the human disease in morphology,biochemistry, response to hormonal manipulation and metastatic spread. 732.1.1 HumanAlthough present, the lobes of the human prostate are not sharply demarcated; the prostate appearsas a single gland with several zones. During the fetal period, the prostate appears as a few widelyspacedtubules supported by stromal cells. 74 Prior to birth, the number and proximity of prostatictubules increases markedly, and proliferation and secretion of the tubule epithelium occurs. Themiddle and lateral prostate lobes exhibit prominent squamous metaplasia of tubular epithelium.Eventually, desquamation occurs and epithelial cells in the prostatic tubules are shed into the lumen.Thus, for a short period after birth (~1 month), the histology of the prostate does not changeappreciably; however, metaplastic changes that occurred during the fetal period, regress after birthleaving empty tubules with only a few remnants of the previous metaplasia by 3 months of age. 74The tubule epithelium also regresses. 74 During the period of regression, the state of the prostatictubules may vary somewhat from being filled with metaplastic cells to being predominantly emptywith some debris from desquamated cells. 74 It should be noted that squamous metaplasia is limitedto the fetal period, except for recurrence during pathological conditions in the adult. 74 The remaininglobes (anterior and posterior lobes) exhibit little or no metaplastic changes. 74© 2006 by Taylor & Francis Group, LLC

1038 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONProstatic secretion is present at the time of birth. Androgen-stimulated secretion from theprostate gland initially occurs at approximately gestation week 14 74 and increases in the incidenceand degree of secretion throughout the remainder of the fetal period. Up to the last month ofgestation, secretion is limited to small tubules at the periphery of the lateral and anterior lobes ofthe fetal prostate. During this time, an intermediate zone separates the secreting tubules from themore central parts, which are still undergoing varying degrees of squamous metaplasia. Oncesquamous metaplasia has subsided and the metaplastic epithelial cells have been shed, secretoryactivity may begin in these areas too. 74 After birth, when squamous metaplasia has entirely subsided,secretion will continue for some time. Strong secretion can still be noted 1 month after birth, butbecomes more variable thereafter. 74,75The transition of the prostate from metaplasia to desquamation to secretion is a hormonallymediatedprocess. In utero estrogen stimulation induces metaplasia in the prostatic utricle and thesurrounding glands. As gestation progresses, the metaplastic cells become squamous and areeventually shed, thus leading to lumen formation and its sac-like structure. After birth, metaplasticchanges regress and estrogen stimulation ends, resulting in a cessation of the metaplastic processand a decrease in utricle distension. 42 As estrogen levels decline, the changing hormonal balancefavors increased androgen levels and stimulation of prostatic secretion. Because squamous metaplasiaand secretion are controlled by different hormones, these processes are localized to differentareas of the prostate; thus, secretion is initiated in the periphery and extends into the central areasof the prostate once squamous metaplasia has been completed in these areas. 742.1.2 DogThe prostate is the only well-developed accessory sex gland in the dog. It is relatively large,completely surrounding the urethra. 73 The prostate is divided into a right and a left lobe with themiddle lobe either poorly developed or absent. In contrast, it is the middle lobe that obstructs theurethra in men with enlarged prostate. 76 In the canine prostate, branching secretory acini and ductsradiate from each side of the urethra. According to O’Shea, 77 the size and weight of the adult canineprostate are as follows: 1.9-2.8 cm long, 1.9-2.7 cm wide, 1.4-2.5 cm high and 4.0-14.5 g (0.21-0.57 g/kg) in weight.In newborn puppies, the prostate is comprised primarily of stroma with some discernableparenchyma. During adolescence, the parenchyma proliferate more quickly than the stroma, makingparenchymal cells the primary cell type in the sexually mature adult dog. Parenchyma are notequally distributed; the stroma still predominates in the central area of the prostate with littleparenchyma visible anterior to the colliculus seminalis and dorsal to the urethra, even in the adultdog. 78The prostate of the adult dog passes through 3 stages: normal growth in the young animal,hyperplasia during the middle of adult life, and senile involution. 73 Similar to other species, prostatedevelopment and function are controlled by androgens; however, the histological structure of thecanine prostate varies with age. Up to 1 year of age, prostatic growth is slow until pubertyapproaches, then growth is rapid and associated with the development of structural and functionalmaturity. 73 As androgen levels rise during puberty, androgens reach a sufficient level to completenormal prostatic growth and maturation. As androgenic stimulation continues, a phase of hyperplasticgrowth occurs which is manifested as a loss of normal histology and onset of glandularhyperplasia. Cysts may form during this period. 79,73 Growth of the prostate in the adult dog appearsto proceed at a steady rate up to about 11 years of age. Senile involution occurs from ~11 yearsonward. During this stage, there is a steady decline in prostate weight, probably due to decreasedandrogen production. The prostate may or may not exhibit histological evidence of atrophy. 73Evidence for prostatic secretion by the alveolar epithelium can be detected at ~4 months of agein dogs. Under normal conditions, secretory epithelial cells originate from differentiated basalreserve cells rather than through a process involving metaplasia. Similar to the parenchyma in the© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1039Table 2RatManVentral prostate —Lateral prostate Lateral lobesDorsal prostate Dorsal (or posterior) lobeAnterior prostate Median (or middle) lobeperiphery of the prostate, some glandular tissue is present in the suburethral submucosa and in thecolliculus seminalis and surrounding tissue at 9 months of age. This glandular tissue does not seemto increase significantly in number or size throughout life. 73Aside from man, dogs are the only common mammals that spontaneously develop benignprostatic hypertrophy (BPH). 80,81 BPH in these two species is similar with respect to older age atonset (5 years or 5th decade), requirement for normal testicular function and prevention by earlyIn both species, dihydrotestosterone levels are elevated in hyperplastic prostate tissue. 82,83 However,human and canine BPH differ in histology, 76,77,79,81 symptomatology and magnitude of their responseto antiandrogen treatment. 84,85Canine prostate glands can respond to both androgenic and estrogenic signals, having receptorsfor both steroids with a prevalence of estrogen receptors. 86 Estradiol increases cytosolic androgenbindingprotein and thereby, stimulates androgen-mediated prostate growth. 87 Within the prostate,dihydrotestosterone is the predominant hormone at various ages, including in immature, matureand hypertrophic prostate glands. 822.1.3 RatUnlike the human prostate, the rat prostate is comprised of discrete lobes (ventral, lateral, dorsaland the paired anterior lobes that are also known as the coagulating glands). In immature and youngrats, separation of the lobes of the prostate, particularly the dorsal and lateral lobes, is possible,whereas in the adult rat, the dorsolateral prostate is often examined due to the lack of a distinctborder between these glands. 73 Table 2 lists probable homologies between the rat and humanprostates. 73,88 Note that there is no embryological corresponding structure in the adult man to therat ventral prostate.Much of rat prostate development occurs postnatally. The prostatic lobes form between postnataldays 1-7 with prostatic tubular lumen forming between postnatal days 7-14. Secretory granules areevident in the developing prostate between postnatal days 14-21. The prostate has attained its adultappearance by postnatal days 28-35, which parallels the postnatal increase in testosterone levels.Several other distinctions appear between rat and human prostates. The rat prostate lacks astrong, well developed fibromuscular stroma. 73 Furthermore, there are differences in enzyme activities,zinc uptake and concentration, and antibacterial factor distribution (present throughout theprostate in humans, but only in dorsolateral prostate in the rat). 732.1.4 MouseThe postnatal development of the murine prostate was evaluated in a series of papers by Sugimuraet al. 89,90,91 These studies found that the number of branch points and tips in both the ventral prostateand the dorsolateral prostate increased dramatically in the first 15 days postpartum. There weresignificant differences between lobes in terms of the number of branch points, and morphologies(both gross and microscopic). The numbers of main ducts and tips of ducts reached adult levelsby ≈ PND30. The differences between lobes noted here have been reported in numerous otherstudies, mostly in rats. 92,93,94© 2006 by Taylor & Francis Group, LLC

1040 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION2.2 Seminal Vesicles2.2.1 HumanThe seminal vesicles and the ductus deferens are present by the 6th month of fetal developmentin an arrangement similar to that seen in adults. The seminal vesicles have a larger lumen andthicker, stronger muscular wall than the ductus deferens. Prenatal development of the muscularwall is mediated by estrogenic stimulation. By the 7th month of gestation, the seminal vesicleshave attained their adult form, although it is not until term that the mucous membrane surroundingthe lumen begins to arrange itself in folds. 42 The seminal vesicles continue to grow slowly untilpuberty. 42Secretory activity is present in the seminal vesicles by the 7th month of gestation and slowlyincreases thereafter, persisting for a considerable time after birth (detected at 17 months). At 4years of age, seminal vesicle secretion was no longer detected. 74 Secretion by the seminal vesiclesis androgen dependent. 952.2.2 DogThe dog has neither seminal vesicles nor bulbourethral glands. 732.2.3 RatIn rats, the basic pattern for seminal vesicle formation is present at postnatal day 10. Lumenformation occurs over a relatively protracted period from postnatal days 2-15. Secretory granulesbecome evident at postnatal day 16. The seminal vesicles markedly increase in size betweenpostnatal days 11-24, and continues to grow until adult appearance and secretory properties areattained between PND 40 and 50. 84 Thus, proliferation and differentiation of the seminal vesiclesparallels the postnatal increase in testosterone levels.3 Neuroendocrine Control of the Reproductive SystemIn mammals, the pituitary and gonads are capable of supporting gametogenesis prior to puberty;however, events in the brain are required to change the hypothalamic-pituitary-gonadal (HPG) axisand trigger maturational changes.3.1 HumanIn humans, much of the process controlling testicular androgen production is present at the timeof birth. In the fetal testis, production of testosterone and antimüllerian hormone begins at the endof the first trimester of pregnancy. 18 Sertoli cells, triggered through an unknown mechanismmediated by the Y chromosome, differentiate and produce antimüllerian hormone. Subsequently,Leydig cells differentiate at approximately 7-8 weeks of gestation and begin producing androgens 96that ultimately come under the control of the placental gonadotrophin, human chorionic gonadotrophin(hCG). 97 Pituitary gonadotrophin synthesis begins around week 12 of gestation with initiallyhigh levels that decline towards the end of gestation, the likely period for the onset of negativefeedback regulation. 98 Thus, the hypothalamic-pituitary-gonadal axis is fully functional in the fetusand neonate. Neonatal exposure of the hypothalamus to androgen is need for sexual differentiationof the LH release mechanism, allowing LH secretion to be modified by either androgen or estrogen. 16Thus, transient elevations of FSH, LH and testosterone have been noted in boys during the first 6months of life, 7 then pulsatile secretion of gonadotrophin declines, reaching its lowest point at 6years of age. 99 Thereafter, pulsatile gonadotrophin secretion begins to increase.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1041Both LH and FSH are involved in the initiation of spermatogenesis. Pulses of LH elicit increasesin androgen concentrations. As age increases, pulsatile gonadotrophin secretion increases in frequencyand amplitude in response to pulsatile GnRH secretion. During puberty, inhibin from Sertoli cellsis the primary negative feedback agent to control FSH release. Testosterone regulates both FSHand LH at the level of the hypothalamus. From puberty onward, androgen increases libido.3.2 RatA comprehensive review of rat neuroendocrine development for the control of reproduction hasbeen compiled by Ojeda and Urbanski. 100Pituitary-gonadal maturation occurs later in rats than in humans, although the sequence ofevents is similar. 98 Gondatrophin secretion and testicular androgen production begin during the lastthird of gestation and continue to gradually decline during the first two weeks of postnatal life. 101In neonates, testicular androgen production is needed during the first few days of life to imprintmale sexual behavior. Unless exposure to steroids occurs, the rat hypothalamus will show a femalepattern of discharge exhibiting cyclical activity. 21Postnatal development of the reproductive system requires hormonal signals from the hypothalamic-pituitary(HP) axis, a subsequent testicular response and feedback from the testis to the HPaxis to modulate gonadotrophin release. Initially during the neonatal period, serum gonadotrophinsare high in male rats, but decline rapidly within a few days. 102 Leydig cells undergo rapid proliferationfrom postnatal days 14 to 28 and another cell division between 28 and 56. Similarly,androgen production increases slowly until 28 days of age with a pronounced increase between 50and 60 days of age. 102,103,104 During this peripubertal period, hormone production by the testischanges due to changes in steroidogenic enzyme levels. During the prepubescent period, androstenedione,5-androstanediol, and dihydrotestosterone are the primary hormones produced by thetestis; 105,106 however, after 40 days of age, testosterone becomes the primary testicular androgen. 107Adult testosterone levels are achieved at approximately 56 days of age. 30 Increased testosteroneproduction, coupled with gonadotrophins, stimulates spermatogenesis and development and maintenanceof the accessory sex organs.Maturation of the reproductive system is initiated by a centrally-mediated process. Gonadotrophinrelease is negatively controlled by testosterone and inhibin. As early as the neonatal period,testosterone-induced negative feedback to the HP is present. 108,109,110 During postnatal life and intoadulthood, concentrations of GnRH in the hypothalamus continue to increase. 111,72,112 Coincidentwith this, LH and FSH levels in the pituitary rise postnatally and the pituitary response to GnRHstimulation also is enhanced. 113,114 FSH supports spermatogenesis within the seminiferous tubulesand stimulates the formation of gonadotrophin receptors in the testis, 115,116 while LH triggersinterstitial cells to produce and secrete testosterone. Increases in serum FSH promote the productionof steroidogenic enzymes 117 and overall testicular growth. 118 As the animal matures, the HP becomesprogressively less sensitive to negative feedback allowing for puberty onset. At this time, pulsatileGnRH release is enhanced, resulting in increased circulating levels of LH and FSH from thepituitary. 119,120,121 Thus, testicular maturation and puberty onset occur secondary to changes in thesecretion of pituitary gonadotrophins.After puberty has occurred, gonadotrophin levels stabilize. Thus, after maximum serum FSHlevels have been achieved (30-40 days), serum testosterone rises and FSH decreases to relativelylow adult levels. 102,122,120,123 The maximum responses to FSH and LH occur between 25-35 and 35-45 days of age, respectively, 124,125 then decline to adult levels. Adult responses are achieved between60 and 80 days of age. 126,80© 2006 by Taylor & Francis Group, LLC

1042 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONConclusionOverall, there is evidence of similar patterns of male postnatal reproductive system developmentbetween humans and experimental animal models. However, a more detailed examination of thedevelopmental processes reveals pertinent cross-species differences. As research progresses, morethorough comparisons on the cellular and molecular level will become possible and the specificpathways targeted by chemicals can be identified. This knowledge will assist in the selection ofsensitive and predictive animal models and further reduce uncertainty when extrapolating animaldata to human risk.REFERENCES1. Müller J. and Skakkebaek, N.E. (1992). The prenatal and postnatal development of the testis.Bailliere’s Clin. Endo. Metab. 6, 251-271.2. Cortes, D., Müller, J. and Skakkebaek, N.E. (1987). Proliferation of Sertoli cells during developmentof the human testis assessed by stereological methods. Int. J. Androl. 10, 589-596.3. Lemasters, G.K., Perreault, S.D., Hales, B.F., Hatch, M., Hirshfield, A.N., Hughes, C.L., Kimmel,G.L., Lamb, J.C., Pryor, J.L., Rubin, C. and Seed, J.G. (2000). Workshop to identify critical windowsof exposure for children’s health: reproductive health in children and adolescents work group summary.Environ. Health Perspect. 108 (Suppl. 3), 505-509.4. Rey, R. (1999). The prepubertal testis: a quiescent or a silently active organ? Histol. Histopathol. 14,991-1000.5. Claudio, L., Bearer, C.F. and Wallinga, D. (1999). Assessment of the U.S. Environmental ProtectionAgency methods for identification of ;hazards to developing organisms, part I: the reproduction andfertility testing guidelines. Am. J. Ind. Med. 35, 543-553.6. Clements, J.A., Reyes, F.I. Winter, J.S.D. and Faiman, C. (1976). Studies on human sexual development.III: fetal pituitary and serum and amniotic fluid concentrations of LH, CG, and FSH. J. Clin.Endocrinol. Metab. 42, 9-19.7. Müller, J. and Skakkebaek, N.E. (1984). Fluctuations in the number of germ cells during the latefoetal and early postnatal periods in boys. Acta Endocrinol. 105, 271-274.8. Müller, J. and Skakkebaek, N.E. (1983). Quantification of germ cells and seminiferous tubules bystereological examination of testicles from 50 boys who suffered from sudden death. J. Androl. 6,143-156.9. Largo, R.H. and Prader, A. (1983). Pubertal development in Swiss boys. Helv. Paediatr. Acta 38, 211-228.10. Biro, F.M., Lucky, A.W. and Huster, G.A. (1995). Pubertal staging in boys. J. Pediatr. 127, 100-102.11. Nysom, K., Pedersen, J.L., Jorgensen, M., Nielsen, C.T., Müller, J., Keiding, N. and Skakkebaek,N.E. (1994). Spermaturia in two normal boys without other signs of puberty. Acta Paediatr. 83, 520-521.12. Partsch, C.-J. and Sippell, W.G. (2001). Pathogenesis and epidemiology of precocious puberty. Effectsof exogenous oestrogens. Hum. Reprod. Update 7, 292-302.13. Nielsen, C.T., Skakkebaek, N.E., Richardson, D.W., Darling, J.A., Hunter, W.M., Jorgensen, M.,Nielsen, A., Ingerslev, O., Keiding, N. and Müller, J. (1986). Onset of the release of spermatozoa(spermarche) in boys in relation to age, testicular growth, pubic hair, and height. J. Clin. Endocrinol.Metab. 62, 532-535.14. Nielsen, C.T., Skakkebaek, N.E., Darling, J.A.B., Hunter, W.M., Richardson, D.W., Jorgensen, M.and Keiding, N. (1986). Longitudinal study of testosterone and luteinizing hormone (LH) in relationto spermarche, pubic hair, height and sitting height in normal boys. Acta Endocrinol. Suppl. 279, 98-106.15. Buchanan, C.R. (2000). Abnormalities of growth and development in puberty. J. R. Coll. PhysiciansLond. 34, 141-145.16. Pryor, J.L., Hughes, C., Foster, W., Hales, B.F. and Robaire, B. (2000). Critical windows of exposurefor children’s health: the reproductive system in animals and humans. Environ. Health Perspect. 108(Suppl. 3), 491-503.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 104317. Catchpole, H.R. and Van Wagenen, G., Reproduction in the Rhesus Monkey, Macaca mulatta. In: TheRhesus Monkey Vol. II; Management, Reproduction and Pathology (ed. Geoffrey H. Bourne), AcademicPress, New York, Pgs. 132-138, 1975.18. Bennett, W.I, Dufau, M.L., Catt, K.J. and Tullner, W.W. (1973). Effect of human menopausalgonadotropin upon spermatogenesis and testosterone production in juvenile rhesus monkeys.Endocrinology 92, 813-812.19. Van Wagenen, G. and Catchpole, H.R. (1956). Amer. J. Phys. Anthropol. 14, 245-274.20. Ojeda, S.R., Andrews, W.W., Advis, J.P., and Smith-White, S. (1980). Recent advances in the endocrinologyof puberty. Endocr. Rev. 1, 228-257.21. Setchell, B.P. (1978). Spermatogenesis. In: The Mammalian Testis. Cornell University Press, Ithaca,New York, pp. 181-232.22. Warren, D.W., Huhtaniemi, I.T., Tapanainen, J., Dufau, M.L., and Catt, K.J. (1984). Ontogeny ofgonadotropin receptors in the fetal and neonatal rat testis. Endocrinology 114, 470-476.23. Steinberger, A., and Steinberger, E. (1971). Replication pattern of Sertoli cells in maturing rat testisin vivo and in organ culture. Biol. Reprod. 4, 84-87.24. Orth, J. (1982). Proliferation of Sertoli cells in fetal and postnatal rats: A quantitative autoradiographicstudy. Anat. Rec. 203, 485-492.25. Wang, Z.X., Wreford, N.G.M., and de Kretser, D.M. (1989). Determination of Sertoli cell numbersin the developing rat testis by stereological methods. Int. J. Androl. 12, 58-64.26. Van den Dungen, H.M., van Dieten, J.A.M.J., van Rees, G.P., and Schoemaker, J. (1990). Testicularweight, tubular diameter and number of Sertoli cells in rats are decreased after early prepubertaladministration of an LHRH-antagonist; the quality of spermatozoa is not impaired. Life Sci. 46, 1081-1089.27. Steinberger, A., Thanki, K.H., and Siegal, B. (1974). FSH binding in rat testes during maturation andfollowing hypophysectomy. Cellular localization of FSH receptors. In: Current Topics in MolecularEndocrinology. Vol. 1. (M.L. Dufau, and A.R. Means, eds.) Plenum Press, New York, pp. 177-192.28. Hodgson, Y., Robertson, D.M., and de Kretser, D.M. (1983). The regulation of testicular function.Int. Rev. Physiol. 27, 275-327.29. Waites, G.M.H., Speight, A.C., and Jenkins, N. (1985). The functional maturation of the Sertoli celland Leydig cell in the mammalian testis. J. Reprod. Fertil. 75, 317-326.30. Scheer, H. and Robaire, B. (1980). Steroid 4 -5-reductase and 3-hydroxysteroid dehydrogenase in therat epididymis during development. Endocrinology 107, 948-953.31. Clermont, L., and Perey, B. (1957). Quantitative study of the cell population of the seminiferoustubules in immature rats. Am. J. Anat. 100, 241-267.32. Clegg, E.J. (1960). The age at which male rats become fertile. J. Reprod. Fertil. 1, 119-120.33. Maeda, K., Ohkura, S., and Tsukamura, H. (2000). Physiology of Reproduction. In: The LaboratoryRat. (G. J. Krinke, ed.) Academic Press, New York, pp. 145-176.34. Russell, L.D., Alger, L.E., and Nequin, L.G. (1987). Hormonal control of pubertal spermatogenesis.Endocrinol. 120, 1615-1632.35. Anderson,R., Garcia-Castro, M., Heasman, J., and Wylie, C. (1998). Early stages of male germ celldifferentiation in the mouse. APMIS 106, 127-133.36. Vergouwen, R.P.F.A., Jacobs, S.G.P.M., Huiskamp, R., Davids, J.A.G., and de Rooij, D.G. (1991)Proliferative activity of gonocytes, Sertoli cells, and interstitial cells during testicular development inmice. J. Reprod. Fertil. 93, 233-243.37. Vergouwen, R.P.F.A., Huiskamp, R., Bas, R.J., Roepers-Gajadien, H.L., Davids, J.A.G., and de Rooij,D.G. (1993) Postnatal development of testicular cell populations in mice. J. Reprod. Fertil. 99, 479-485.38. Baker, P.J. and O’Shaughnessy, P.J. (2001) Role of gonadotrophins in regulating numbers of Leydigand Sertoli cells during fetal and postnatal development in mice. Reproduction 122, 227-234.39. Myers-Wallen, V.N., Manganaro, T.F., Kuroda, T., Concannon, P.W., MacLaughlin, D.T., and Donahoe,P.K. (1991) The critical period for Mullerian duct regression in the dog embryo. Biol. Reprod. 45,626-633.40. Connell, C.J. (1980). Blood-testis barrier formation and the initiation of meiosis in the dog. In:Testicular Development, Structure, and Function. (A. Steinberger and E. Steinberger, eds.) RavenPress, New York, pp. 71-78.© 2006 by Taylor & Francis Group, LLC

1044 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION41. Kawakami, E., Tsutsui, T, and Ogasa, A (1991) Histological observations of the reproductive organsof the male dog from birth to sexual maturity. J. Vet. Med. Sci. 53, 241-248.42. Rogriguez, C.M., Kirby, J.L. and Hinton, B.T. (2002). The development of the epididymis. In: TheEpididymis. From Molecules to Clinical Practice. (B. Robaire and B.T. Hinton, eds.), Kluwer Academic/PlenumPublishers, New York, pp. 251-267.43. Büskow, A.G., and Hoyer, P.E. (1988) The Physiology of Reproduction. Vol. 1. Raven Press, NewYork, pp. 265-302.44. Mylchreest, E., D. G. Wallace, R. C. Cattley and P. M. D. Foster: Dose-Dependent Alterations inAndrogen-Regulated Male Reproductive Development in Rats Exposed to Di(n-butyl) Phthalate duringLate Gestation. Toxicol. Sci. 2000, 55, 143-151.45. McIntyre, B. S., N. J. Barlow and P. M. D. Foster: Androgen-mediated development in male ratoffspring exposed to flutamide in utero: Permanence and correlation of early postnatal changes inanogenital distance and nipple retention with malformations in androgen-dependent tissues. Toxicol.Sci. 2001, 62, 236-249.46. Sun, E.L. and Flickinger, C.J. (1979). Development of cell types and of regional differences in thepostnatal rat epididymis. Am. J. Anat. 154, 27-56.47. DeLarminat, M.A., Cuasnicu,P.S., and Blaquier, J.A. (1981). Changes in trophic and functionalparameters of the rat epididymis during sexual maturation. Biol. Reprod. 25, 813-819.48. Sun, E.L. and Flickinger, C.J. (1982). Proliferative activity in the rat epididymis during postnataldevelopment. Anat. Record.203, 273-284.49. Ramirez, R., Martin, R., Martin, J.J., Ramirez, J.R., Paniagua, R., and Santamaria, L (1999). Changesin the number, proliferation rates, and bcl-2 protein immunoexpression of epithelial and periductalcells from rat epididymis during postnatal development. J. Androl. 20, 702-712.50. Limanowski, A., Miskowiak, B., Otulakowski, B., Partyka, M., and Konwerska, A. (2001). Morphometricstudies on rat epididymis in the course of postnatal development (computerised image analysis).Folia Histochemica et Cytobiol. 39, 201-20251. Agarwal, A. and Hoffer, A.P. (1989). Ultrastructural studies on the development of the blood-epididymisbarrier in immature rats. J. Androl. 10, 425-431.52. Mailot, J.P., Guerin, C, and Begon, D (1985) Growth, testicular development, and sperm output inthe dog from birth to post-pubertal period. Andrologia 17, 450-460.53. Camatini, M., Franchi, E. and DeCurtis, I. (1981). Differentiation of inter-Sertoli junctions in humantestis. Cell Biol. Int. Rep. 5, 109.54. Vitale, R., Fawcett, D.W., and Dym, M. (1973). The normal development of the blood-testis barrierand the effects of clomiphene and estrogen treatment. Anat. Rec. 176, 333-344.55. Setchell, B.P., Laurie, M.S., and Fritz, I.B. (1980). Development of the function of the blood-testisbarrier in rats and mice. In: Testicular Development, Structure, and Function. (A. Steinberger and E.Steinberger, eds.) Raven Press, New York, pp. 65-69.56. Setchell, B.P., Voglmayr, J.K., and Waites, G.M.H. (1969). A blood-testis barrier restricting passagefrom blood into rete testis fluid but not into lymph. J. Physiol. 200, 73-85.57. Prahalada, S., Tarantal, A.F., Harris, G.S., Ellsworth, K.P., Clarke, A.P., Skiles, G.L., MacKenzie, K.I.,Kruk, L.F., Ablin, D.S., Cukierski, M.A., Peter, C.P., vanZwieten, M.J., and Hendrickx, A.G. (1997).Effects of finasteride, a type 2 5-alpha reductase inhibitor, on fetal development in the rhesus monkey(Macaca mulatta). Teratology 55, 119-131.58. Clark, R.L. (1999). Endpoints of reproductive system development. In: An Evaluation and Interpretationof Reproductive Endpoints for Human Risk Assessment. International Life Sciences Institute/Healthand Environmental Science Institute, Washington, D.C. pp. 27-62.59. Wise, L.D., Vetter, C., Anderson, C.A., Antonello, J.M., and Clark, R.L. (1991). Reversible effects oftriamcinolone and lack of effects with aspirin or L-656,224 on external genitalia of male Sprague-Dawley rats exposed in utero. Teratology 44, 507-520.60. Wensing, C.J. (1986). Testicular descent in the rat and comparison of this process in the rat with thatin the pig. Zoology 423-434.61. Wensing, C.J.G. (1973). Testicular descent in some domestic mammals. I. Anatomical aspect oftesticular descent. Zoology 423-434.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 104562. Hogan, M.D., Newbold, R.R. and McLachlan, J.A. (1987). Extrapolation of teratogenic responsesobserved in laboratory animals to humans: DES as an illustrative example. Cold Spring HarborLaboratory, Cold Spring Harbor, New York.63. Gondos, B. (1985). Development of the reproductive organs. Ann. Clin. Lab. Sci. 15, 363-373.64. Wislocki, G.B. (1933). Anat. Rec. 57, 133-148.65. Catchpole, H.R. and Van Wagenen, G., Reproduction in the Rhesus Monkey, Macaca mulatta. In: TheRhesus Monkey Vol. II; Management, Reproduction and Pathology (ed. Geoffrey H. Bourne), AcademicPress, New York, Pgs. 132-138, 1975.66. Deibert, G.A. (1933). The separation of the prepuce in the human penis. Anat. Rec. 57, 387-393.67. Ben-Ari, J., Merlob, P., Mimouni, F., and Reisner, S.H. (1985). Characteristics of the male genitaliain the newborn: penis. J. Urol. 134, 521-522.68. Oster, J. (1968). Further fate of the foreskin. Incidence of preputial adhesions, phimosis, and smegmaamong Danish schoolboys. Arch. Dis. Child 43, 200-203.69. Gairdner, D. (1949). The fate of the foreskin: a study of circumcision. Br. Med. J. 2, 1433-1437.70. Burrows, H. (1944). The union and separation of living tissue as influenced by cellular differentiation.Yale J. Biol. Med. 17, 397-407.71. Korenbrott, C.C., Huhtaniemi, I.T., and Weiner, R.L. (1977). Preputial separation as an external signof pubertal development in the male rat. Biol. Reprod. 17, 298-303.72. Cameron, J.L. (1991). Metabolic cues for the onset of puberty. Horm. Res. 36, 97-103.73. Sandberg, A.A., Karr, J.P., and Müntzing, J. (1980). The prostates of dog, baboon and rat. In: MaleAccessory Sex Glands. Biology and Pathology. (E. Spring-Mills and E.S.E. Hafez, eds.). Elsevier/NorthHolland Biomedical Press, New York, p. 565-608.74. Zondek, L.H., and Zondek, T. (1980). Congenital malformations of the male accessory sex glands inthe fetus and neonate. In: Male Accessory Sex Glands. Biology and Pathology. (E. Spring-Mills andE.S.E. Hafez, eds.). Elsevier/North Holland Biomedical Press, New York, p. 17-37.75. Zondek, L.H., and Zondek, T. (1979). Effects of hormones on the human fetal prostate. In: FetalEndocrinology. (L.H. Zondek and T. Zondek, eds.) Karger, New York.76. Schlotthauer, C.F. (1932). Observations on the prostate gland of the dog. J. Amer. Vet. Med. Assoc.81, 645.77. O’Shea, J.D. (1962). Studies on the canine prostate gland. I. Factors influencing its size and weight.J. Comp. Path. 72, 321.78. Berg, O.A. (1958a). The normal prostate gland of the dog. Acta Endocrinol. 27, 129.79. Berg, O.A. (1958b). Parenchymatous hypertrophy of the canine prostate gland. Acta Endocrinol. 27, 140.80. Huggins, C., and Moulder, P.V. (1945). Estrogen production by Sertoli cell tumors of the testis. CancerRes. 5, 510.81. Ofner, P. (1968). Effects and metabolism of hormones in normal and neoplastic prostate tissue. Vitam.Horm. 26, 237-291.82. Gloyna, R.E., Siiteri, P.K., and Wilson, J.D. (1970). Dihydrotestosterone in prostatic hypertrophy. II.The formation and content of dihydrotestosterone in the hypertrophic canine prostate and the effectof dihydrotestosterone on prostate growth in the dog. J. Clin. Invest. 49, 1746-1753.83. Siiteri, P.K., and Wilson, J.D. (1970). Dihydrotestosterone in prostatic hypertrophy. I. The formationand content of dihydrotestosterone in the hypertrophic prostate of man. J. Clin. Invest. 49, 1737-1745.84. Brooks, J.E., Busch, R.D., Patanelli, D.J., and Steelman, S.L. (1973). A study of the effects of a newanti-androgen on the hyperplastic dog prostate. Proc. Soc. Exp. Biol. Med. 143, 647-655.85. Scott, W.W., and Wade, J.C. (1969). Medical treatment of benign nodular prostatic hyperplasia withcyproterone acetate. J. Urol. 101, 81-85.86. Robinette, C.L., Blume, C.D., and Mawhinney, M.G. (1978). Androphilic and estrophilic moleculesin canine prostate glands. Invest. Urol. 15, 425-431.87. Moore, R.J., Gazak, J.M., and Wilson, J.D. (1979). Regulation of cytoplasmic dihydrotestosteronebinding in dog prostate by 17-estradiol. J. Clin. Invest. 63, 351-357.88. Price, D. (1963). Comparative aspects of development and structure in the prostate. Nat. Cancer Inst.Monogr. 12, 1.89. Sugimura, Y., Cunha, G.R., and Doncajour, A.A. (1986). Morphogenesis of ductal networks in themouse prostate. Biol. Reprod. 34, 961-971.© 2006 by Taylor & Francis Group, LLC

1046 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION90. Sugimura, Y., Cunha, G.R., and Doncajour, A.A. (1986). Morphological and histological study ofcastration-induced degeneration and androgen-induced regeneration in the mouse prostate. Biol.Reprod. 34, 973-983.91. Sugimura, Y., Cunha, G.R., Doncajour, A.A., Bigsby, R.M., and Brody, J.R. (1986). Whole-mountautoradiography study of DNA synthetic activity during postnatal development and androgen-inducedregeneration in the mouse prostate. Biol. Reprod. 34, 985-995.92. Lee, C., Sensibar, J.S., Dudek, S.M., Hiipakka, R.A., and Liao, S. (1990). Prostatic ductal system inrats: regional variation in morphological and functional activities. Biol. Reprod. 43, 1079-1086.93. Prins, G.S. (1989). Differential regulation of androgen receptors in the separate rat prostate lobes:androgen-independent expression in the lateral lobe. J. Steroid Biochem. 33, 319-326.94. Higgins, S.J., Smith, S.E., and Wilson, J. (1982). Development of secretory protein synthesis in theseminal vesicles and ventral prostate of the male rat. Mol. Cell. Endocrinol. 27, 55-65.95. Mann, T., and Lutwak-Mann, C. (1976). Evaluation of the functional state of male accessory glandsby the analysis of seminal plasma. Andrologia 8, 237-242.96. Wilson, J.D., George, F.W., and Griffin, J.E. (1981). The hormonal control of sexual development.Science 211, 1278-1284.97. Huhtaniemi, I.T., Korenbrot, C.C., and Jaffe, R.B. (1977). HCG binding and stimulation of testosteronebiosynthesis in the human fetal testis. J. Clin. Endocrinol. Metab. 44(5), 963-96798. Huhtaniemi, I., Pakarinen, P., Sokka, T., and Kolho, K.-L. (1989). Pituitary-gonadal function in thefetus and neonate. In: Control of the Onset of Puberty III. (H.A. Delemarre-van de Waal, ed.) ElsevierScience Publishers, New York, pp. 101-109.99. Brook, C.G. (1999). Mechanism of puberty. Horm. Res. 51 (Suppl. 3), 52-54.100. Ojeda, S.R. and Urbanski, H.F. (1994). Puberty in the Rat. In: The Physiology of Reproduction. SecondEdition. Vol. 2. (E. Knobil and J.D. Neill, eds.), Raven Press, New York, pp.363-409.101. Tapanainen, J., Kuopio, T., Pelliniemi, L.J., and Huhtaniemi, I. (1984). Rat testicular endogenoussteroids and number of Leydig cells between the fetal period and sexual maturity. Biol. Reprod. 31,1027-1035.102. Döhler, K.D., and Wuttke, W. (1974). Serum LH, FSH, prolactin and progesterone from birth topuberty in female and male rats. Endocrinology 94, 1003-1008.103. Piacsek, B.E., and Goodspeed, M.P. (1978). Maturation of the pituitary-gonadal system in the malerat. J. Reprod. Fertil. 52, 29-35.104. .Resko, J.A., Feder, H.H., and Goy, R.W. (1968). Androgen concentrations in plasma and testis ofdeveloping rats. J. Endocrinol. 40, 485-491.105. Podesta, E.J., and Rivarola, M.A. (1974). Concentration of androgens in whole testes, seminiferoustubules and interstitial tissue of rats at different stages of development. Endocrinology 95, 455-461.106. Wiebe, J.P. (1976). Steroidogenesis in rat Leydig cells: Changes in activity of 5-ane and 5-ene 3-hydroxysteroid dehydrogenase during sexual maturation. Endocrinology 98, 505-513.107. Gupta, D., Zarzycki, J., and Rager, K. (1975b). Plasma testosterone and dihydrotestosterone in malerats during sexual maturation and following orchidectomy and experimental bilateral cryptorchidism.Steroids 25, 33-42.108. Goldman, B.D., Grazia, Y.R., Kamberi, I.A., and Porter, J.C. (1971). Serum gonadotropin concentrationsin intact and castrated neonatal rats. Endocrinology 88, 771-776.109. Ramirez, V.D., and McCann, S.M. (1965). Inhibitory effect of testosterone on luteinizing hormonesecretion in immature and adult rats. Endocrinology 76, 412-417.110. Negro-Vilar, A., Ojeda, S.R., and McCann, S.M. (1973). Evidence for changes in sensitivity totestosterone negative feedback on gonadotropin release during sexual development in the male rat.Endocrinology 93, 729-735.111. Desjardin, C. (1981). Endocrine signaling and male reproduction. Biol. Reprod. 24, 1-21.112. Brabant, G., Ray, A.K., Wagner, T.O.F., Krech, R., and von zur Mühlen, A. (1985). Peripubertalvariation in GnRH-release from in vitro superfused hypothalami of male rats. Neuroendocrinol. Lett.7, 197-202.113. Chiappa, S.A., and Fink,G. (1977). Releasing factor and hormonal changes in the hypothalamicpituitary-gonadotrophinand adrenocorticotrophin systems before and after birth and puberty in male,female and androgenized female rats. J. Endocrinol. 72, 211-224.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1047114. Chappel, S.C., and Ramaley, J.A. (1985). Changes in the isoelectric focusing profile of pituitaryfollicle-stimulating hormone in the developing male rat. Biol. Reprod. 32, 567-573.115. Odell, W.D., and Swerdloff, R.S. (1976). Etiologies of sexual maturation: A model system based onthe sexually maturing rat. Recent Prog. Horm. Res. 32, 245-288.116. Huhtaniemi, I.T., Nozu, K., Warren, D.W., Dufau, M.L., and Catt, K.J. (1982). Acquisition of regulatorymechanisms for gonadotropin receptors and steroidogenesis in the maturing rat testis. Endocrinology111, 1711-1720.117. Murono, E.P., and Payne, A.H. (1979). Testicular maturation in the rat. In vivo effects of gonadotropinson steroidogenic enzymes in the hypophysectomized immature rat. Biol. Reprod. 20, 911-917.118. Ojeda, S.R., and Ramirez, V.D. (1972). Plasma levels of LH and FSH in maturing rats: response tohemigonadectomy. Endocrinology 90, 466-472.119. Chowen-Breed, J.A., and Steiner, R.A. (1989). Neuropeptide gene regulation during pubertal developmentin the male rat. In: Control of the Onset of Puberty III. (H.A. Delemarre-van de Waal, ed.)Elsevier Science Publishers, New York, pp. 63-77.120. Payne, A.H., Kelch, R.P., Murono, E.P., and Kerlan, J.T. (1977). Hypothalamic, pituitary and testicularfunction during sexual maturation of the male rat. J. Endocrinol. 72, 17-26.121. Araki, S., Toran-Allerand, C.D., Ferin, M., and Vande Wiele, R.L. (1975). Immunoreactive gonadotropin-releasinghormone (Gn-RH) during maturation in the rat: ontogeny of regional hypothalamicdifferences. Endocrinology 97, 636-697.122. Döhler, K.D., and Wuttke, W. (1975). Changes with age in levels of serum gonadotropins, prolactin,and gonadal steroids in prepubertal male and female rats. Endocrinology 97, 898-907.123. Nazian, S.J., and Cameron, D.F. (1992). Termination of the peripubertal FSH increase in male rats.Am. J. Physiol. 262, E179-E184.124. Dullaart, J. (1977). Immature rat pituitary glands in vitro: Age- and sex-related changes in luteinizinghormone releasing hormone-stimulated gonadotrophin release. J. Endocrinol. 73, 309-319.125. Debeljuk, L., Arimura, A., and Schally, A.V. (1972). Studies on the pituitary responsiveness toluteinizing hormone-releasing hormone (LH-RH) in intact male rats of different ages. Endocrinology90, 585-588.126. Dalkin, A.C., Bourne, G.A., Pieper, D.R., Regiani, S., and Marshall, J.C. (1981). Pituitary and gonadalgonadotropin-releasing hormone receptors during sexual maturation in the rat. Endocrinology 108,1658-1664.127. Federal Register, Part Two, Vol 63, no 231, Wednesday, Dec 2, 1998 Rules and regulations, Departmentof Health and Human Services, Food and Drug Administration 21 CFR Parts 201, 312, 314 and 601,pg 66650.128. Inaba, T., Matsuoka, S., Kawate, N., and Mori, J. (1994). Developmental changes in testicularluteinizing hormone receptors and androgens in the dog. Res. Vet. Sci. 57, 305-309.© 2006 by Taylor & Francis Group, LLC

1048 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAPPENDIX C-5*LANDMARKS IN THE DEVELOPMENTOF THE FEMALE REPRODUCTIVE SYSTEM † †David A. Beckman, PhD, and Maureen Feuston, PhDAuthor affiliations:David A. Beckman, PhD, Novartis Pharmaceuticals Corporation, Preclinical Safety, Toxicology, 59 Route 10, Building406-142, East Hanover, NJ 07936 USAMaureen H. Feuston, PhD, Sanofi-Synthelabo Research, Toxicology, 9 Great Valley Parkway, P.O. Box 3026, Malvern,PA 19355 USAAddress correspondence to:David A. Beckman, PhD, Novartis Pharmaceuticals Corporation, Preclinical Safety, Toxicology, 59 Route 10, Building406-142, East Hanover, NJ 07936 USA, E-mail: david.beckman@pharma.novartis. com, Tel: 862-778-3490, Fax: 862-778-5489Drugs and environmental chemicals have the potential to adversely affect the developing femalereproductive system. The consequences of an exposure are influenced by the magnitude and durationof the exposure, the mechanism of action for the drug/chemical, the sensitivity of target tissue, andthe critical windows in the development of the reproductive system. For the purposes of this review,Table 1 is presented to define the postnatal age for phases of sexual development in the femaleorganism and to indicate equivalent ages in selected common laboratory species, specifically therat, Beagle dog, and non-human primates. ‡ It is important to point out that the range of ages for aparticular phase can vary somewhat depending on the organ systems under consideration and the speciesselected for the comparison. This follows from the observation that tissues and organs do not necessarilymature in the same sequence across all species or by the same overall stage of development.Table 1Summary of Locomotor DevelopmentHuman Rat DogCrawling ≈ PND 270(9 months)Walking ≈ PND 396(13 months)PND = Postnatal DayaBipedal locomotionNon-HumanPrimate (Rhesus)PND 3-12 PND 4-20 PND 4-49PND 12-16 PND 20-28 PND 49 aAlthough the number of oocytes is determined by birth, the postnatal maturation of the reproductivetissues, steroid hormone production, external genitalia, sexual behavior and cyclic signalingevents enables the female to reach her reproductive potential. Furthermore, although puberty maybe defined as the time at which the generative organs mature and reproduction may occur, it doesnot signify full or normal reproductive capacity.Focusing on the rat first because it is the most widely used species in pharmaceutical/chemicalindustry research for toxicity assessments, plasma follicle-stimulating hormone (FSH) levels start* Source: Beckman, D. A. and Feuston, M., Landmarks in the development of the female reproductive system, Birth DefectsResearch, Part B: Developmental and Reproductive Toxicology, 68, 137-143, 2003.† The categories presented by the U.S. Department of Health and Human Services (95) to aid in the design of clinicalinvestigations of medicinal products are not included in Table 1 since they are not consistent with those used for thelaboratory species.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1049to increase shortly after birth, reaching maximum concentrations on day 12, then declining graduallyto approximately 20% of the day-12 values by the end of the juvenile period (6,7,8). Plasmaluteinizing hormone (LH) levels are also higher in neonatal-infantile rats than in juvenile rats andare characterized by sporadic surges (6,7,8,10). These surges disappear completely by the juvenilephase and LH levels remain low during this period.During the infantile period, estradiol exerts relatively little negative feedback, due to highaffinitybinding to the high levels of alpha-fetoprotein (11), and aromatizable androgens play aleading role in the steroid negative feedback control of gonadotropin secretion. This changes froma predominantly androgenic control to a dual estrogenic-androgenic control during the juvenileperiod. During the juvenile period, there is an increase in secretion of dehydroepiandrosterone(DHEA) and DHEA sulfate associated with the maturation of a prominent adrenal zona reticularis.Occurring at day 20 in the rat (12), adrenarche, the maturation of a prominent zona reticularis, ischaracterized by a prepubertal rise in adrenal secretion of DHEA and DHEA sulfate that isindependent of the gonads or gonadotropins. The rise in DHEA and DHEA sulfate productionresults from the increased presence of 17alpha-hydroxylase/17,20-lyase and DHEA sulfotransferaseand the decreased presence of 3beta-hydroxysteroid dehydrogenase (13,14).Also during the juvenile period, morphological maturation of neurons in the hypothalamuscoincides with changes in the diurnal pattern of LH release and the time at which the hypothalamicpituitaryunit becomes fully responsive to estradiol positive feedback (15).The ovary is relatively insensitive to gonadotropin stimulation during the first week after birth,coming under strong gonadotropin control during the second week (16,17). Waves of folliculardevelopment and atresia occur during the juvenile period, but these follicles are not ovulated. Atthe end of the juvenile phase, the mode of LH release begins to change (9,10), animals are olderthan 30 days of age, their uteri are small (wet weight less than 100 mg), no intrauterine fluid canbe detected and vaginal patency is not yet achieved (18). Animals progressing to the next phasehave larger uteri with intrauterine fluid and the vagina remains closed (19). On the day of the firstproestrus, there is a large amount of intrauterine fluid, the wet weight of the uterus is greater than200 mg, the ovaries have large follicles, and the vagina is closed in most animals (18). On the dayof the first ovulation, uterine fluid is no longer present, fresh corpora lutea have been formed, thevagina is open and vaginal cytology shows a predominance of cornified cells. A common parameterof sexual maturity in rats is vaginal opening. Vaginal opening occurs on the day after the firstpreovulatory surge of gonadotropins, i.e. day 36-37 (20,21,22,13) but can range from 32 to 109days (24). Female rats may maintain their full reproductive potential to 300 days (1), however, thisis influenced by the strain of rat.As in the rat, there is an early postnatal rise in gonadotropin secretion in the human and monkey,followed by a reduction lasting from late infancy to the end of the prepubertal period (25,26,27,28).Until about 12-24 months in girls, the ovaries respond to the increased follicle-stimulating hormone(FSH) by secreting estradiol reaching levels that are not again achieved before the onset of puberty(29,30). The secretion of estradiol then decreases with a nadir, in the human, at about 6 years ofage (29). During this time, there is a reduction in the pulsatile release of gonadotrophin releasinghormone (GnRh) which appears to involve both gonadal and central nervous system restraints. Inthe human, the timing of adrenarche may vary from 5 to 8 years (31,32). However, several reportssuggest that adrenarche may be a gradual maturational process beginning at 7-8 years that continuesto 13-15 years and is not the result of sudden rapid changes in adrenal enzyme activities or adrenalandrogen concentrations (10,17,33,34). Interestingly, adrenarche appears to be absent in non-humanprimates (5,35,36). Although adrenal androgen concentrations are similar to those in humans,adrenal androgen levels throughout development in rhesus and cynomolgus monkeys and baboonsare similar to those in adulthood (37,38).Menarche, the onset of menstrual cyclicity, is regarded as an overt sign of the initiation ofpuberty, although endocrine profiles indicate significant hormonal changes years/months prior. Thefirst ovulation occurs sometime after the first cycle. This supports the “adolescent sterility” hypothesis© 2006 by Taylor & Francis Group, LLC

1050 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(the interval between menarche and first ovulation) proposed by Young and Yerkes (39). In mostyoung women, ovulation does not occur until 6 or more months after menarche and a regularrecurrence of ovulatory menstrual cycles does not appear to become established until several yearslater (40,41,42,43). The postmenarcheal phase of development in the rhesus monkey is the resultof a high incidence of anovulatory and short luteal-phase cycles (44,45).Estradiol concentrations increase at initial stages of puberty at 2.5 to 3 years in the non-humanprimate (46,47) and at 8 to 10 years in the human (48,49,50,51). Gonadotropes acquire the capacityto respond to the stimulatory action of estradiol on LH secretion after menarche (25,26,27). The onsetof puberty in the human is marked by an increase in the amplitude of LH pulses, which is taken toindicate an increase in the amplitude of GnRH pulses. Studies in the rhesus monkey (28,46,52,53)suggest that the central nervous system mechanisms responsible for the inhibition of the GnRH pulsegenerator during childhood involve primarily gamma-aminobutyric acid (GABA) and GABAergicneurons. With the onset of puberty, the reactivation of the GnRH pulse generator is associated with afall in GABAergic neurotransmission and an associated increase in the input of excitatory amino acidneurotransmitters (including glutamate) and possibly astroglial-derived growth factors (30,54).It appears that adequate levels of leptin and a leptin signal are required to achieve puberty andto maintain cyclicity and reproductive function in the rodent (55,56,57,58,59,60,61), non-humanprimate and human (30,62,63,64). However, leptin does not appear to have a direct effect on thecentral control of the pulsatile release of GnRh (53). The available evidence suggests that leptinacts as one of several permissive factors but alone is not sufficient to initiate sexual maturation.Prolactin has a luteotropic role in rats. The prolactin content and prolactin-containing cells ofthe anterior pituitary increase with postnatal age (65,66) but prolactin levels remain low until theprepubertal period (67,68). Prolactin is not leuteotropic in monkeys or humans (69).In order to aid in pinpointing the timing of landmarks in the postnatal development of thefemale reproductive tract and for cross-species comparisons, Table 2 presents the timing of specificlandmarks for the rat, dog, non-human primate and human, including endocrine status, folliclematuration, ovulation, reproductive tract development, estrous cyclicity/menarche, adrenarche,puberty and fertility.The information in this review may be useful in the interpretation of results from investigationsinto the potential effects of chemical/compound exposures during the development of younganimals. Although this information may also aid in the selection of the most appropriate speciesfor an investigation, it cannot be separated from other issues in the design of a study that impactthe final species selection, including:Study purpose: Is the investigation a primary or screening study, with the primary purpose to identifypotential adverse effects in juvenile animals that are not seen in adults of the same species, or isit secondary/focused study, with the primary purpose of investigating, for example, the scope,severity, potential for recovery, etc., for adverse effects on a specific organ system during postnataldevelopment? What regulatory guidelines must be followed?Exposure window and duration: Based on the purpose of the study, will the exposure be limited toa defined stage in postnatal development or should it include several stages, possibly from neonatalto sexual maturity? Also, do results from previous studies in adults or in young animals, if available,demonstrate changes in no-observed-effect levels over time and/or progression of target organtoxicities with chronic administration? Do considerations of enzyme induction, immature metabolism,and the desired stage(s) of development for exposure limit the species selection?Compound specific considerations: Are there species-specific characteristics for the compound withrespect to the mechanism of action, pharmacokinetics, target organ toxicity, etc., that limit potentialspecies for consideration?Concordance between animal and human toxicity: Are results available (most probably in the adult)that support concordance of target organ toxicity for the compound from animal and human studies?While some of the framing of the preceding questions may be more appropriate for pharmaceuticaldevelopment, they nevertheless provide useful discussion for other types of investigations.© 2006 by Taylor & Francis Group, LLC

Table 2 Landmarks in the development of the female reproductive tract for the rat, dog, non-human primate and humanParameterEndocrine statusEstrogen/estradiolLuteinizing hormone, LHFollicle-stimulating hormone,FSHProlactinRat (Wistar or SpragueDawley, if possible)Range of free estradiolconcentration during earlyneonatal period is similar tothat during cycling adult (70).LH levels increase shortly afterbirth to a maximum on day 12then declines to about one-fifthof the day 12 values by the endof the juvenile period (6,7,8).Prepubertal increase in LH 8-9days before the expected dayof first proestrus (9,10).FSH levels increase shortlyafter birth to a maximum onday 12 then declines to aboutone-fifth of the day 12 valuesby the end of the juvenileperiod (6,7,8).Luteotropic in rats and miceAnterior pituitary prolactincontent and prolactincontainingcells increase withpostnatal age (65,66) butprolactin levels remain low untilthe prepubertal period (67,68).Dog(Beagle, if possible)Not availableGonadotropes acquire thecapacity to respond to thestimulatory action of estradiolon LH secretion by day 20 andfull capacity by day 28 (77).By 4 months, FSHconcentrations are similar tothose in adult anestrus (3).By 4 months, FSHconcentrations are similar tothose in adult anestrus (3).Luteotropic role is questionable(78,79).SpeciesNon-human primate(Macaca mulatta, if possible)2.5 to 3 years, increasedestradiol concentrations atinitial stages of puberty(46,47).Gonadotropes acquire thecapacity to respond to thestimulatory action of estradiolon LH and FSH secretion aftermenarche (25,26,27).GnRH neurosecretory system isactive in neonatal period thenenters a dormant state. Anincrease in pulsatile release isessential for onset of puberty(28).Gonadotropes acquire thecapacity to respond to thestimulatory action of estradiolon LH and FSH secretion aftermenarche (25,26,27,46).No luteotropic effect inmonkeys.Human8 to 10 years, increasedestradiol concentrations atinitial stages of puberty(48,49,50,51).Gonadotropes acquire thecapacity to respond to thestimulatory action of estradiolon LH and FSH secretion,probably after menarche(25,26,27).GnRH neurosecretory system isactive in neonatal period thenenters a dormant state. Anincrease in pulsatile release isessential for onset of puberty(28). Also see reference 51.Gonadotropes acquire thecapacity to respond to thestimulatory action of estradiolon LH and FSH secretionprobably after menarche(25,26,27). Also see reference51.No luteotropic effect in humans.Also see reference 69.POSTNATAL DEVELOPMENTAL MILESTONES 1051© 2006 by Taylor & Francis Group, LLC

Table 2 (continued) Landmarks in the development of the female reproductive tract for the rat, dog, non-human primate and human (continued)ParameterRat (Wistar or SpragueDawley, if possible)LeptinNecessary for maintenance ofestrous cyclicity (60).Adequate levels essential butnot sufficient for onset ofpuberty (57).Adequate levels essential foronset of puberty andmaintenance of cyclicity andreproductive function (59).Follicle maturationOvarian follicles becomesubjected to stronggonadotropin control duringthe second week of postnatallife (16,17).Ovulation 1 st ovulation on about day 38 (1).1 st ovulation on day 29 (19).1 st ovulation on day 37 (71).Dog(Beagle, if possible)Present (80) but role inreproduction not defined.5-6 months: primary folliclesshow antrum formation (81).8-12 months (3).9-14 months (81).SpeciesNon-human primate(Macaca mulatta, if possible)In “monkey” adequate levelsessential but not sufficient foronset of puberty (57).Adequate levels essential foronset of puberty andmaintenance of cyclicity andreproductive function (59).Adequate levels essential foronset of puberty (53).Not availableThe postmenarcheal phase ofdevelopment in the rhesusmonkey is the result of a highincidence of anovulatory andshort luteal-phase cycles(44,45).HumanAdequate levels essential foronset of puberty andmaintenance of cyclicity andreproductive function (59).Follicles were seen in 86% ofprepubertal girls and in 99% ofpubertal girls (85).In most young women, ovulationdoes not occur until 6 or moremonths after menarche and aregular recurrence of ovulatorymenstrual cycles does notappear to become establisheduntil several years later(40,41,42,43,86,87).12-24 months post menarche(38).Ovarian function last ~30 years(51).1052 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

ParameterReproductive tractmaturationUterine maturationAnogenital distanceRat (Wistar or SpragueDawley, if possible)Differentiation of the muscularand glandular epithelium:uterine glands first appear onpostnatal day 9 and increaseduntil day 15 (71).Prepubertal: Small uterus, wetweight less than 100mg withno intrauterine fluid (18).Postpubertal: Larger uterus, wetweight greater than 200 mg(18).Increase in uterine weight onday 27 (19).Anogenital distance is 3.5 mmfor males and approx one halfthat for females and (73).GD21: females, 1.29-1.41 mm;3.15-4.44 mm for males (74).PND 0: females, 1.29-1.51 mm;males, 3.27-3.83 mm (74).Anogenital distance varies withbody weight – normalization(75).Vaginal opening 32-109 days (24).Wistar 35.6 days ± 0.8 SE (76).Occurs on the day after the firstpreovulatory surge ofgonadotropins, i.e. day 36-37(20,21,22,23).Crl SD: 31.6-35.1 days; Mean,33.4 days (74).Dog(Beagle, if possible)Non-human primate(Macaca mulatta, if possible)HumanPuberty (81). Not available Uterine growth begins beforepuberty (85).Not determined Not determined Not determinedNot determined Not determined Not determinedPOSTNATAL DEVELOPMENTAL MILESTONES 1053© 2006 by Taylor & Francis Group, LLC

Table 2 (continued) Landmarks in the development of the female reproductive tract for the rat, dog, non-human primate and human (continued)ParameterAdrenarche Day 20 (12). 11 weeks of age (82). Absence of adrenarche innon-human primates(5,35,36).“In primates, adrenal androgenconcentrations are similar tohumans. However, rhesus andcynomegalus monkeys andbaboons exhibit adrenalandrogen levels throughoutdevelopment that are nodifferent than those duringadulthood” (37,38).PubertyEstrous cyclicity/Menarche(1 st estrus/mensus)Rat (Wistar or SpragueDawley, if possible)Vaginal opening and firstovulation at ~5 weeks (1,70).~5 weeks (24).36.4 days ± 1.2 SE (74).Dog(Beagle, if possible)8-12 months (3). 5-13 months(mean, 9 months) in 8-15 kgbeagles (83).Sexual maturity/Fertility 50 ± 10 days (24). 6-12 months (2).8-12 months (3).SpeciesNon-human primate(Macaca mulatta, if possible)2.5-3 years, nipple growth,increase in perineal swellingand coloration (46,47).5 years (32).6-8 years (31).Onset 7-8 years (6-8 yearsskeletal age) that continues to13-15 years (10,17,34).8-10 years: appearance of labialhair, initiation of breastenlargement (48,49,50,88).Menarche at 8-14 yr(89,90,91,92,93,94)Menses at 12-13 years (88)8-13 years (34)8-12 months (3). 2-3 years (84). ~13 years (90,91,92,94).12-13 years (93).8-14 years (89).10-16.5 years; Mean, 13.4 years(34).2.6-3.5 years (84).3-4 years (2).Human11-16 years (2).1054 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1055REFERENCES1 Ojeda SR, Andrews WW, Advis JP, Smith-White S. Recent advances in the endocrinology of puberty.Endocr Rev 1980;1:228-57.2 Altman PL, Dittmer DS. Biology Data Book, 2nd ed., Vol. 1. Bethesda: Fed Am Soc Exp Biol, 19723 Concannon PW. Biology of gonadotrophin secretion in adult and prepubertal female dogs. J ReprodFertil Suppl 1993;47:3-27.4 Andersen AC. Reproduction. In: Andersen AC, editor. The Beagle as an Experimental Dog. Ames,Iowa: The Iowa State University Press, 1970:31-39.5 Meusy-Dessolle N, Dang DC. Plasma concentrations of testosterone, dihydrotestosterone, delta 4-androstenedione, dehydroepiandrosterone and oestradiol-17 beta in the crab-eating monkey (Macacafascicularis) from birth to adulthood. J Reprod Fertil 1985;74:347-59.6 Dohler KD, Wuttke W. Serum LH, FSH, prolactin and progesterone from birth to puberty in femaleand male rats. Endocrinology 1974;94:1003-8.7 Kragt CL, Dahlgen J. Development of neural regulation of follicle-stimulating hormone (FSH) secretion.Neuroendocrinology 1972;9:30-408 Ojeda SR, Ramirez VD. Plasma levels of LH and FSH in maturing rats: response to hemigonadectomy.Endocrinology 1972;9066-72.9 Andrews WW, Ojeda SR. A detailed analysis of the serum LH secretory profiles of conscious freemovingfemale rats during the time of puberty. Endocrinology 1981;109:2032-9.10 MacDonald J, Dees WL, Ahmed CE, Ojeda SR. Biochemical and immunocytochemical characterizationof neuropeptide Y in the immature rat ovary. Endocrinology 1987;120:1703-10.11 Meijs-Roelofs HMA, Kramer P. Maturation of the inhibitory feedback action of oestrogen on folliclestimulatinghormone secretion in the immature female rat: a role for alphafoetoprotein. J Endocrinol1979;81:199-208.12 Pignatelli D, Bento MJ, Maia M, Magalhaes MM, Magalhaes MC, Masaon JI. Ontogeny of 3betahydroxysteroiddehydrogenase expression in the rat adrenal gland as studied by immunohistochemistry.Endocr Res 1999;25:21-7.13 Gell JS, Carr BR, Sasano H, Atkins B, Margraf L, Mason JI, Rainey WE. Adrenarche results fromdevelopment of a 3beta-hydroxysteroid dehydrogenase-deficient adrenal reticularis. J Clin EndocrinolMetab 1998;83:3695-701.14 Parker CR Jr. Dehydroepiandrosterone and dehydroepiandrosterone sulfate production in the humanadrenal during development and aging. Steroids 1999;64:640-7.15 Urbanski HF, Ojeda SR. The juvenile-peripubertal transition period in the female rat: establishmentof a diurnal pattern of pulsatile luteinizing hormone secretion. Endocrinology 1985;117:644-9.16 Hage AJ, Groen-Klevant AC, Welschen R. Follicle growth in the immature rat ovary. Acta Endocrinol1978;88:375-82.17 Lunenfeld B, Kraiem Z, Eshkol A. The function of the growing follicle. J Reprod Fertil 1975;45:567-74.18 Advis JP, Andrews WW, Ojeda SR. Changes in ovarian steroidal and prostaglandin E responsivenessto gonadotropins during the onset of puberty in the female rat. Endocrinology 1979;104:653-8.19 Tamura K, Abe Y, Kogo H. Phenytoin inhibits both the first ovulation and uterine development ingonadotropin-primed immature rats. Eur J Pharmacol 2000;398:317-22.20 Critchlow V, Elswers Bar-Sela M. Control of the onset of puberty. In: Martini L, Ganong WF, editors.Neuroendocrinology, Vol. II. New York: Academic Press, Inc., 1967:101-6221 Meijs-Roelofs HMA, Uilenbroek JThJ, de Greef WJ, de Jong FH, Kramer D. Gonadotropin andsteroid levels around the time of first ovulation in the rat. J Endocrinol 1975;67:275-82.22 Oheda SR, Wheaton JE, Jameson HE, McCann SM. The onset of puberty in the female rat. I. Changesin plasma prolactin, gonadotropins, LHRH levels and hypothalamic LHRH content. Endocrinology1976;98:630-8.23 Ramirez VD. Endocrinology of puberty. In: Greep RO, Astwood EB, editors. Handbook of Physiology,Vol. II, Section 7. Washington, DC: American Physiological Society, 1973:1-28.24 Kohn DF, Barthold SW. Biology and diseases of rats. In: Fox JG, Cohen, BJ, Loew, FM, editors.Laboratory Animal Medicine, Chapt. 4. San Diego: Academic Press, Inc., 1984:91-122.© 2006 by Taylor & Francis Group, LLC

1056 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION25 Dierschke DJ, Weiss G, Knobil E. Sexual maturation in the female rhesus monkey and the developmentof estrogen-induced gonadotropic hormone release. Endocrinology 1974;94:198-206.26 Reiter EO, Kulin HE, Hamwood SM. The absence of positive feedback between estrogen and luteinizinghormone in sexually immature girls. Pediatr Res 1974;8:740-5.27 Presl J, Horejsi J, Stroufova A, Herzmann J. Sexual maturation in girls and the development ofestrogen-induced gonadotropic hormone release. Ann Biol Anim Biochem Biophys 1976;16:377-83.28 Terasawa E, Fernandez DL. Neurological mechanisms of the onset of puberty in primates. EndocrRev 2001;22:111-5129 Brook CG. Mechanism of puberty. Horm Res 1999;51 Suppl 3:52-4.30 Grumbach MM. The neuroendocrinology of human puberty revisited Horm Res 2002;57 Suppl 2:2-14.31 Dardis A, Saraco N, Rivarola MA, Belgorosky A. Decrease in the expression of the 3beta-hydroxysteroiddehydrogenase gene in human adrenal tissue during prepuberty and early puberty: implicationsfor the mechanism of adrenarche. Pediatr Res 1999;45:384-8.32 Suzuki T, Sasano H, Takeyama J, Kaneko C, Freije WA, Carr BR, Rainey WE. Developmental changesin steroidogenic enzymes in human postnatal adrenal cortex: imunohistochemical studies. Clin Endocrinol2000;53:739-47.33 Palmert MR, Hayden DL, Mansfield MJ, Crigler JF, Crowley WF, Chandler DW, Boepple PA. Thelongitudinal study of adrenal maturation during gonadal suppression: evidence that adrenarche is agradual process. J Clin Endocrinol Metab 2001;86:4536-42.34 Styne, DM, and MM Grumbach. Puberty in the male and female: Its physiology and disorders. In:Yen SS, Jaffe RB, editors. Reproductive Endocrinology, Chapter 12. Philadelphia: WB SaundersCompany, 1986:313-384.35 Crawford BA, Harewood WJ, Handelsman DJ. Growth and hormone characteristics of pubertaldevelopment in the hamadryas baboon. J Med Primatol 1997;26:153-63.36 Mann DR, Castracane VD, McLaughlin F, Gould KG, Collins DC. Developmental patterns of serumluteinizing hormone, gonadal and adrenal steroids in the sooty mangabey (Cercocebus atys). BiolReprod 1983; 28:279-84.37 Gonzalez, F. Adrenarch. In: Knobil E, Neill JD, editors. Encyclopedia of Reproduction, Volume 1.San Diego: Academic Press, 1998:51-61.38 Hatasaka, H. Menarche. In: Knobil E, Neill JD, editors. Encyclopedia of Reproduction, Volume 3.San Diego: Academic Press, 1998:177-83.39 Young WC, Yerkes RM. Factors influencing the reproductive cycle in the chimpanzee: the period ofadolescent sterility and related problems. Endocrinology 1943;33:121-5440 Treloar AE, Boynton RE, Behn BG, Brown BW. Variation of the human menstrual cycle throughreproductive life. Int J Fertil 1967;12:77-126.41 Doring GK. The incidence of anovular cycles in women. J Reprod Fertil 1969;6(Suppl.):77-81.42 Apter D, Viinikka L, Vihko R. Hormonal pattern of adolescent menstrual cycles. J Clin EndocrinolMetab 1978;47:944-54.43 Metcalf, MG, Skidmore DS, Lowry GF, Mackenzie JA. Incidence of ovulation in the years after themenarche. J Endocrinol 1983;97:213-9.44 Foster DL. Luteininzing hormone and progesterone secretion during sexual maturation of the rhesusmonkey: short luteal phases during the initial menstrual cycles. Biol Reprod 1977;17:584-90.45 Resko JA, Goy RW, Robinson JA, Norman RL. The pubescent rhesus monkey: some characteristicsof the menstrual cycle. Biol Reprod 1982;27:354-61.46 Terasawa E, Nass TE, Yeoman RR, Loose MD, Schultz NJ. Hypothalamic control of puberty in thefemale rhesus macaque. In: Norman RL, editor. Neuroendocrine Aspects of Reproduction. New York:Academic Press Inc., 1983:149-82.47 Wilson ME, Gordon TP, Collins DC. Ontogeny of luteinizing hormone secretion and first ovulationin seasonal breeding rhesus monkeys. Endocrinology 1986;118:293-301.48 Bidlingmaier F, Wagner-Barnach M, Butenandt O, Kron D. Plasma estrogen in childhood and pubertyunder physiologic and pathologic conditions. Pediatr Res 1973;7:901-7.49 Jenner MR, Kelch RP, Kaplan SL, Grumbach MM. Hormonal changes in puberty: IV. plasma estradiol,LH, and FSH in prepubertal children, pubertal females, and in precocious puberty, premature thelarche,hypogonadism and in a child with a feminizing ovarian tumor. J Clin Endocrinol Metab 1972;34:521-30.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 105750 Winter JSD, Faiman C. Pituitary-gonadal relations in female children and adolescents. Pediatr Res1973;7:948-53.51 Yen SS. The Human Menstrual Cycle. In: Yen SS, Jaffe RB, editors. Reproductive Endocrinology,Chapter 8. Philadelphia: WB Saunders Company, 1986:200-36.52 Kasuya E, Nyberg CL, Mogi K, Terasawa E. A role of gamma-amino butyric acid (GABA) andglutamate in control of puberty in female rhesus monkeys: effect of an antisense oligodeoxynucleotidefor GAD67 messenger ribonucleic acid and MK801 on luteinizing hormone-releasing hormonerelease. Endocrinology 1999;140:705-12.53 Plant TM. Neurological bases underlying the control of the onset of puberty in the rhesus monkey:a representative higher primate. Front Neuroendocrinol 2001;22:107-39.54 Ma YJ, Ojeda SR. Neuroendocrine control of female puberty: glial and neuronal interactions. J InvestigDermatol Symp Proc 1997;2:19-22.55 Cheung CC, Thornton JE, Kuijper JL, Weigle DS, Clifton DK, Steiner RA. Leptin is a metabolic gatefor the onset of puberty in the female rat. Endocrinology 1997;138:855-8.56 Cheung CC, Thornton JE, Nurani SD, Clifton DK, Steiner RA. A reassessment of leptin’s role intriggering the onset of puberty in the rat and mouse. Neuroendocrinology 2001;74:12-21.57 Cunningham MJ, Clifton DK, Steiner RA. Leptin’s actions on the reproductive axis: perspectives andmechanisms. Biol Reprod 1999;60:216-22.58 Urbanski HF. Leptin and puberty. Trends Endocrinol Metab 2001;12:428-9.59 Mantzoros CS. Role of leptin in reproduction. Ann N Y Acad Sci 2000;900:174-8360 Carro E, Pinilla L, Seoane LM, Considine RV, Aguilar E, Casanueva FF, Dieguez C. Influence ofendogenous leptin tone on the estrous cycle and luteinizing hormone pulsatility in female rats.Neuroendocrinology 1997; 66:375-7.61 Watanobe H, Schioth HB. Postnatal profile of plasma leptin concentrations in male and female rats:relation with the maturation of the pituitary-gonadal axis. Regul Pept 2002;105:23-8.62 Chan JL, Mantzoros CS. Leptin and the hypothalamic-pituitary regulation of the gonadotropin-gonadalaxis. Pituitary 2001;4:87-92.63 Farooqi IS. Leptin and the onset of puberty: insights from rodent and human genetics. Semin ReprodMed 2002;20:139-4464 Bouvattier C, Lahlou N, Roger M, Bougneres P. Hyperleptinaemia is associated with impairedgonadotrophin response to GnRH during late puberty in obese girls, not boys. Eur J Endocrinol1998;138:653-8.65 Watanabe H, Takebe K. Involvement of postnatal gonads in the maturation of dopaminergic regulationof prolactin secretion in female rats. Endocrinology 1987;120:2212-9.66 Yamamuro Y, Aoki T, Yukawa M, Sensui N. Developmental profile of prolactin cells in the anteriorpituitary of postnatal rats during the lactational stage. Proc Soc Exp Biol Med 1999;220:94-9.67 Natagani S, Guthikonada P, Foster DL. Appearance of a nocturnal peak of leptin secretion in thepubertal rat. Horm Behav 2000; 37:345-52.68 Ojeda SR, Advis JP, Andrews WW. Neuroendocrine control of the onset of puberty in the rat. FedProc 1980;39:2365-71.69 Yen SS. Prolactin in human reproduction. In: Yen SS, Jaffe RB, editors. Reproductive Endocrinology,Chapter 9. Philadelphia: WB Saunders Company, 1986:237-63.70 Montano MM, Welshons WV, vom Saal FS. Free estradiol in serum and brain uptake of estradiolduring fetal and neonatal sexual differentiation in female rats. Biol Reprod 1995; 53:1198-207.71 Rice VM, Limback SD, Roby KF, Terranova PF. Changes in circulating and ovarian concentrationsof bioactive tumour necrosis factor alpha during the first ovulation at puberty in rats and in gonadotrophin-treatedimmature rats. J Reprod Fertil 1998;113:337-41.72 Branham WS, Sheehan DM, Zehr DR, Ridlon E, Nelson CJ. The postnatal ontogeny of rat uterineglands and age-related effects of 17 beta-estradiol. Endocrinology 1985;117:2229-37.73 Tyl RW, Marr MC. Developmental toxicity testing – methodology, p. 204. In: Hood RD, editor.Handbook of Developmental Toxicology. Boca Raton: CRC Press, 1996:175-225.74 Clark, RL. Endpoints of reproductive system development. In: G Daston and C Kimmel, editors. AnEvaluation and Interpretation of Reproductive Endpoints for Human Health Risk Assessment. Washington,D.C.: ILSI Press, 1999.© 2006 by Taylor & Francis Group, LLC

1058 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION75 Gallavan RH Jr, Holson JF, Stump DG, Knapp JF, Reynolds VL. Interpreting the toxicologicalsignificance of alterations in anogenital distance: potential for confounding effects of progeny bodyweights. Reprod Toxicol 1999; 13(5):383-90.76 Eckstein B, Golan R, Shani J. Onset of puberty in the immature female rat induced by 5α-androstane-3B,17B-diol. Endocrinology 1973;92:941-5.77 Andrews WW, Mizejewski GJ, Ojeda SR. Development of estradiol positive feedback on LH releasein the female rat: a quantitative study. Endocrinology 1981;109:1404-13.78 Okkens AC, Kooistra HS, Dieleman SJ, Bevers MM. Dopamine agonistic effects as opposed toprolactin concentration in plasma as the influencing factor on the duration of anoestrus in bitches. JReprod Fetil Suppl 1997;51:55-8.79 Onclin K, Verstegen JP. Secretion patterns of plasma prolactin and progesterone in pregnant comparedwith nonpregnant dioestrous beagle bitches. J Reprod Fertil Suppl 1997;51:203-8.80 Le Bel C, Bourdeau A, Lau D, Hunt P. Biologic response to peripheral and central administration ofrecombinant human leptin in dogs. Obes Res 1999;7:577-85.81 The Beagle as an Experimental Dog. AC Andersen, Editor. Iowa State University Press. 1970.82 Perez-Fernandez R, Facchinetti F, Beira A, Lima L, Gaudiero GJ, Genazzani AR, Devesa J. Morphologicaland functional stimulation of adrenal reticularis zone by dopaminergic blockade in dogs. JSteroid Biochem 1987; 28:465-70.83 Concannon, PW. Reproduction in dog and cat. In: Reproduction in Domestic Animals. PT Cupps,Editor. Academic Press, 1991.84 Hendrickx AG, Dukelow WR. Reproductive biology. In: Bennett BT, Abee CR, Henrickson R, editors.Nonhuman Primates in Biomedical Research. San Diego: Academic Press, 1995:147-191.85 Holm K, Laursen EM, Brocks V, Muller J. Pubertal maturation of the internal genitalia: an ultrasoundevaluation of 166 healthy girls. Ultrasound Obstet Gynecol 1995;6:175-81.86 Chabbert Buffet N, Djakoure C, Maitre SC, Bouchard, P. Regulation of the human menstrual cycle.Front Neuroendocrinol 1998;19:151-86.87 Hartman CG. On the relative sterility of the adolescent organism. Science 1931;74:226-7.88 Herman-Giddens ME, Slora EJ, Wasserman RC, Bourdony CJ, Bhapkar MV, Koch GG, HasemeierCM. Secondary sexual characteristics and menses in young girls seen in office practice: a study fromthe Pediatric Research in Office Settings network. Pediatrics 1997;99:505-12.89 Blondell RD, Foster MB, Dave KC. Disorders of puberty. Am Fam Physician 1999;60:209-4.90 Chowdhury S, Shafabuddin AK, Seal AJ, Talukder KK, Hassan Q, Begum RA, Rahman Q, TomkinsA, Costello A, Talukder MQ. Nutritional status and age at menarche in a rural area of Bangladesh.Ann Hum Biol 2000; 27:249-56.91 Graham MJ, Larsen U, Xu X. Secular trend in age at menarche in China: a case study of two ruralcounties in Anhui Province. J Biosoc Sci 1999;31:257-67.92 Marrodan MD, Mesa MS, Arechiga J, Perez-Magdaleno A. Trend in menarcheal age in Spain: ruraland urban comparison during a recent period. Ann Hum Biol 2000;27:313-9.93 Marshall WA. Puberty. In: Falkner F, Tanner JM, editors. Human Growth, Volume 2: Post-natal Growth.New York: Plenum Press, 1978:141-81.94 Pasquet P, Biyong AM, Rikong-Adie H, Befidi-Mengue R, Garba MT, Froment A. Age at menarcheand urbanization in Cameroon: current status and secular trends. Ann Hum Biol 1999;26:89-97.95 U.S. Department of Health and Human Services, Food and Drug Administration. Guidance forindustry. E11. Clinical investigation of medicinal products in the pediatric population. www.fda.gov/cber/guidelines.htm, December 2000.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1059APPENDIX C-6*POSTNATAL ANATOMICAL AND FUNCTIONALDEVELOPMENT OF THE HEART:A SPECIES COMPARISONKok Wah Hew 1,3 and Kit A. Keller 21Purdue Pharma L.P., Nonclinical Drug Safety Evaluation, Ardsley, New York2Consultant, Washington DC3Correspondence to: Kok Wah Hew, PhD, Purdue Pharma L.P., Nonclinical Drug Safety Evaluation, 444 Saw MillRiver Road, Ardsley, NY 10502. E-mail: kok-wah.hew@pharma.comIntroductionThe purpose of this review is to summarize the postnatal development, growth and maturation ofthe human heart and its vascular bed, and explain how the infant/juvenile systems differ, bothmorphologically and functionally, from that seen in the adult. A species comparison with commonlaboratory animals, where available, is also included to aid in the extrapolation to human riskassessment. The available data is mostly obtained from smaller laboratory animals. The postnatalchanges observed in the heart are relatively similar across the mammalian species, but do differ inthe timing of events. As a general rule, the functional and morphological changes in the heart occurfaster, according to the developmental rate, in small laboratory animals compared to humans dueto the faster growth and maturation rate in these animals. This review is by no means a comprehensivecompilation of all of the literature available on the human and animal postnatal heartdevelopment, but rather a basic outline of key events.Postnatal Anatomical Growth and Maturation of the HeartImmediate Postnatal Changes in Cardiac Morphology and VascularizationThe transition from prenatal to postnatal circulation involves three primary steps: 1) cessation ofumbilical circulation, 2) the transfer of gas exchange function from the placenta to the lungs, and3) closure of the prenatal shunts, at first functionally and then structurally (Rakusan, 1980, 1984;Friedman and Fahey, 1993; Smolich, 1995).The switch from fetal to adult type of circulation is not immediate. During the neonatal period,a transient intermediate period occurs, characterized by decreasing pulmonary vascular resistance,increasing pulmonary blood flow, increasing left atrium blood volume, and increasing systemicvascular resistance. During this period there is constriction of the ductus arteriosus and closure ofthe foramen ovale. In addition, the heart function changes from acting as two parallel pumps (rightand left) to acting as two pumps performing in series.HumanUmbilical Vasculature — Elimination of the vascular lumina of the umbilical vessels takes fromseveral weeks to months after birth (O’Rahilly and Muller, 1996). The umbilical vein between the* Source: Hew, K. W. and Keller, K., Postnatal anatomical and functional development of the heart: A species comparison,Birth Defects Research, Part B: Developmental and Reproductive Toxicology, 68, 309-320, 2003.© 2006 by Taylor & Francis Group, LLC

1060 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONumbilicus and liver becomes the ligamentum teres in the falciform ligament. The proximal partsof the umbilical arteries remain as the internal iliac arteries.Closure of the Foramen Ovale — Functional closure of the foramen ovale occurs very rapidlyin association with the first breath and occurs as a result of the hemodynamic changes that holdthe valve of the foramen ovale closed (Walsh et al., 1974). Anatomical closure of the foramen ovaleis a slow process, which normally does not occur before the end of the first year (Walsh et al.,1974). Anatomical closure is complete when the valve becomes fixed to the interatrial septum.Closure of the Ductus Arteriosus — Following birth, functional closure of the ductus arteriosus,which is a direct connection of the pulmonary trunk into the dorsal aorta, is brought about bycontraction of the smooth muscle in the wall of the ductus arteriosus. Anatomical closure is producedby structural changes and necrosis of the inner wall of the ductus arteriosus, followed by connectivetissue formation, fibrosis, and permanent sealing of the lumen (Broccoli and Carinci, 1973; Clyman,1987). In full-term infants, the ductus arteriosus begins to constrict soon after the first breath.Functional closure is complete in about 15 hours after birth (Heymann and Rudolph, 1975).Anatomical closure occurs in about one-third of infants by 2-3 weeks after birth, in nearly 90%by 2 months of age and in 99% by 1 year of age (Broccoli and Carinci, 1973; O’Rahilly and Muller,1996). The remaining fibrous cord following anatomical closure is known in adult anatomy as theligamentum arteriosum.The biochemistry behind the initial functional closure of the ductus arteriosus is believed toinvolve a balance between the opposing actions of oxygen and prostaglandin E 2 (PGE 2 ). PGE 2 hasa dilating effect and acts to keep the duct open in utero, while increased oxygen concentration afterbirth alters pulmonary and systemic arterial pressure causing the duct to constrict (Clyman, 1987).This is supported by reports that closure of the ductus arteriosus is delayed or absent in prematureinfants (who have high PGE 2 levels) and in neonates exposed to low oxygen environments, includinghigh altitudes (Moss et al., 1964; Penaloza et al., 1964).Closure of the Ductus Venosus — The ductus venosus, which carries portal and umbilicalblood to the inferior vena cava, is functionally closed within hours of birth. Permanent closurebegins a few days after birth and is complete by 18-20 days of age (Meyer and Lind, 1966; Fugelsethet al., 1997; Kondo et al., 2001). The duct becomes the ligamentum venosum in a fissure on theback of the liver.Comparative SpeciesClosure of the Foramen Ovale — In the rat, the foramen ovale is diminished in the first twodays after birth and is completely closed three days after birth (Momma et al., 1992a).Closure of the Ductus Arteriosus — In the rat and mouse, functional closure of the ductusarteriosus is complete by 2-5 hours after birth (Jarkovska et al., 1989; Tada and Kishimoto, 1990).Functional closure is complete in a few minutes after birth in guinea pigs and rabbits (Clyman,1987). In rats, rabbits, guinea pigs and lambs, anatomical closure is complete in 1-5 days afterbirth (Jones et al., 1969; Fay and Cooke, 1972; Heymann and Rudolph, 1975). In dogs, anatomicalclosure of the ductus arteriosus occurs at about 7-8 days after birth (House and Enderstrom, 1968).As in the human infant, the biochemistry behind the initial functional closure of the ductusarteriosus is believed to involve a balance between the opposing actions of oxygen and PGE 2(Clyman, 1987). Studies in guinea pigs and lambs suggest that this contractile response to oxygenis the results of interaction between oxygen and a cytochrome P 450 - catalyzed enzymatic process,that can be inhibited by carbon monoxide (Fay and Jobsis, 1972; Coceani et al., 1984). Specifically,cytochrome a 3 becomes oxidized with subsequent generation of adenosine triphosphate (ATP) and© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1061muscle cell contraction. In full-term animals, the loss of responsiveness to PGE 2 shortly after birthprevents the duct from reopening once it has constricted (Clyman, et al., 1985a). This change inresponsiveness may involve the presence of thyroid hormones (Clyman, et al., 1985b).Closure of the Ductus Venosus — In the rat, the ductus venosus narrows rapidly after birth andcloses completely in two days (Momma and Ando, 1992b). In the dog and lamb, functional closureof the ductus venosus occurs between 2-3 days after birth (Lohse and Suter, 1977; Zink and VanPetten, 1980).A study in lambs suggests that thromboxane may play a role in the closure of the ductus venosusafter birth (Adeagbo et al., 1985).Shape, Position and Size of the Heart During Postnatal DevelopmentThe rate of cardiac growth during the postnatal period is species dependent (Hudlicka and Brown,1996). In humans, the relative heart weight does not significantly change from infancy to adulthoodand the highest variability in weights occurs in newborns. In contrast, the relative heart weights inmost experimental animals decrease significantly with age and the highest variability in heart weightoccurs in adults. In humans and laboratory animals, the right ventricular weight fraction is higherat birth than in the adult and this value changes over to a left ventricular dominance during thepostnatal period. Postnatal changes in heart shape and dimensions vary according to species.HumanIn humans, the heart is relatively large in proportion to the chest at birth and is usually oval orglobular in shape. At birth the length is approximately 75% of its width (Rakusan, 1984). By theend of the first year, the length increases to approximately 80% of its width. By adulthood, thelength and width are almost the same.At birth, the heart is positioned higher and more “transverse” than in the adult. The characteristicadult oblique lie is attained between 2 and 6 years of age (O’Rahilly and Muller, 1996).The greatest rate of increase in absolute heart weight occurs during the first postnatal year. Theheart doubles its birth weight by 6 months of age and triples by 1 year (Smith, 1928; Rakusan,1980). It should be noted that heart weights in infants and small children are highly variable andavailable mean data show a deviation factor of 2 to 3 (Rakusan, 1980). After a brief surge duringthe early postnatal period, the increase in cardiac weight is proportional to the increase in bodyweight, resulting in a constant relative cardiac weight throughout the first half of life.Human Heart Weight and Heart Weight as % of Body Weight (Smith, 1928)Age in yearsMean Heart Weight (g) % Body WeightMale Female Male FemaleBirth to 1 year 45 35 0.62 0.552–4 69 55.5 0.43 0.375–7 87.7 94 0.35 0.378–12 149 120 0.44 0.4213–17 225 180 0.43 0.3818–21 308 245 0.42 0.42Part of this growth in mass in early development involves a modest increase in right ventricularwall thickness and a marked increase in the left ventricular wall thickness (Eckner et al., 1969;Graham et al., 1971; Oberhansli et al., 1981). At birth, volumes of the right and left ventricularcavities are approximately the same. In infants, the right ventricular volume is about twice that ofthe newborns, whereas the left ventricular volume does not change. In the second year, the ratio© 2006 by Taylor & Francis Group, LLC

1062 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof right ventricle:left ventricle volumes reaches 2:1 and does not change with age (Kyrieleis, 1963).Studies on the relation of the right and left ventricular weight at birth show high variability (Rakusan,1980).Comparative SpeciesThe basic geometry of the heart does not change with age in the rabbit and guinea pig (Lee et al.,1975). In the rat and dog, the left ventricle is less spherical in the postnatal period, compared toadults, due to growth-related changes in ventricular dimensions (House and Ederstrom, 1968;Grimm et al., 1973; Lee et al., 1975). Heart weight, in relation to body weight, is greater in thenewborn than the adult in most common laboratory animals (House and Ederstrom, 1968; Lee etal., 1975; Rakusan, 1980). Subsequently, the heart:body weight ratio decreases with age.Heart Weight as % of Body Weight (Rakusan, 1980)NewbornAdultRat 0.52–0.57 0.17–0.22Mouse 0.46 0.44Rabbit 0.49 0.20Guinea Pig 0.39–0.51 0.23–0.19Dog 0.93 0.76In the rat, the heart grows in proportion to the weight of the body during the early postnatalperiod (first month of age). Later the cardiac growth rate is slower than that of the body, resultingin a gradual decrease in relative cardiac weight (Addis and Gray, 1950; Rakusan et al., 1963;Mattfeldt and Mall, 1987). In the Wistar rat, the rate of increase of heart weight, compared to thatof the body weight, is greater from 1-5 postnatal days and less from 5-11 postnatal days (Anversaet al., 1980). Throughout the first 11 days of growth, the left ventricular weight increases at a fasterrate than that of the right ventricle, being 16%, 51%, 113% greater at 1, 5, and 11 days, respectively.The 6.2-fold increase in left ventricular weight is accompanied by a 2.7-fold increase in wallthickness. In contrast, no significant change in right ventricular wall thickness is observed despitea 3.4-fold weight gain. A 400% increase in the thickness of the primum septum occurs in the first2 postnatal days (Momma et al., 1992a). The septum secundum also grows rapidly during thistime, with a 69% increase in length and width.Postnatal Heart Growth in Wistar Rats (Anversa et al., 1980; Rakusan et al., 1965)Age (postnatal day)Mean Values1 5 11 23 40 68 120Body weight, g 5.35 11.75 26.34 36 81 216 328Heart weight, mg 21.44 60.6 104.9 181 318 641 778Left ventricle 8.86 28.3 55.1 NA NA NA NARight ventricle 7.64 18.7 25.9 NA NA NA NALeft ventricle wall thickness, mm 0.470 0.869 1.256 NA NA NA NARight ventricle wall thickness, mm 0.383 0.335 0.354 NA NA NA NAHeart growth rate in the mouse is proportional to body growth from birth to 24 weeks of age(Wiesmann et al., 2000). Left ventricular mass is reported to increases with age.Reported relative heart weights in the guinea pig range from 0.39% to 0.51% at birth to 0.19%to 0.23% in adults, with decreasing relative fraction of right ventricular weight from 57% to 55%of the left ventricle (Webster and Liljegren, 1949; Lee et al., 1975).Relative heart weight in the rabbit is reported to decrease from 0.49% at birth to 0.20% inadults (Lee et al., 1975; Boucek, 1982; Fried et al., 1987). Investigators also reported a relative© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1063decrease in right ventricular weight from 86% of the left ventricular at birth to 54% in adults, withpreferential growth of the left ventricle between postnatal days 1 and 7.For the beagle dog, Deavers et al. (1972) reported a small but significant decrease in relativeheart weights from birth to adulthood. The ratio of left ventricular volume and weight did notchange with age (Lee et al., 1975).Postnatal Development of Heart Cellular ConstituentsAt the cellular level, cardiac growth is due to both cellular proliferation and an increase in cellularvolume (Hudlicka and Brown, 1996). Myocyte hypertrophy is accompanied by an increase incellular DNA. Cardiac myocytes at birth contain one diploid nucleus. The number of binucleatedcells rapidly increases during the early postnatal stages. In human myocytes, the majority of thesenuclei subsequently fuse, leading to an increase in polyploidy. On the other hand, binucleatedcardiac myocytes remain predominant throughout life in rodents. Hearts from other experimentalanimals contain both binucleated and polyploid myocytes. Other significant intracellular maturationalchanges include a progressive increase in sarcoplasmic reticulum and myofibrils.HumanThe newborn heart contains about half the total number of myocytes present in the adult heart.Adult values are reached probably before the age of 4 months (Linzbach, 1950, 1952; Hort, 1953).At birth, the majority of cardiac myocytes are mononucleated. During postnatal development, thepercentage of binucleated cells increases up to 33% during late infancy or early childhood, with asubsequent decrease to adult values [5-13% of cells in the left ventricle are binucleated and 7% ofthe cells in the right ventricle are binucleated] (Schneider and Pfitzer, 1973). In infants, almost allnuclei of the cardiac myocyte are diploid. In adults, 60% are diploid, 30% are tetraploid and 10%are octoploid (Eisentein and Wied, 1970). Myocyte size is also reported to increase with age fromapproximately 5 µm at birth, 8 µm at 6 weeks, 11 µm at 3 years, 13 µm at 15 years to 14 µm inadults (Ashley, 1945; Rakusan, 1980).An increase in myocardial contractile function during the early postnatal period is accompaniedby intracellular changes such as increased sarcoplasmic reticulum and myofibrils, the organellesthat regulate and utilize calcium to produce cardiac contraction (Fisher and Towbin, 1988).Comparative SpeciesIn the rat, as in humans, myocyte cell volume increases nearly 25-fold and myocyte cell number3-4 fold from birth to 2 months of age, (Anversa et al., 1986; Englemann et al., 1986; Mattfeldtand Mall, 1987;Vliegen et al., 1987). The increase in cardiac mass is due to cell proliferation untilpostnatal day 3-6, hyperplasia and hypertrophy until postnatal day 14, and solely by hypertrophythereafter (Chubb and Bishop, 1984; Anversa et al., 1986, 1992; Batra and Rakusan, 1992; Li etal., 1996). Most of the muscle cell nuclei are diploid in the rat heart (Korecky and Rakusan, 1978;Rakusan, 1984). The frequency of polyploidy, which is rather low at birth, even decreases slightlywith age in the rat heart (Grove et al., 1969). Mast cells, which are postulated to play a role incapillary angiogenesis, are present in very low numbers in early rat pups, but start to increase inthe postnatal weeks 3 and 4 to maximum numbers by 100 days of age (Rakusan et al., 1990).In mice, the volume of myocytes remains relatively constant despite a concomitant increase inheart weight, indicating growth due to cell division during the first four postnatal days (Leu et al.,2001). After postnatal day 5, the volume of myocytes increases markedly until postnatal day 14,when the increase slows down. Myocytes reach their adult volume at around 3 months of age.In rabbits, myocyte diameter increases progressively (from ~ 5 µm) towards adult values(14 µm), starting on postnatal day 8 (Hoerter et al., 1981). Rabbit neonate studies show that, as in© 2006 by Taylor & Francis Group, LLC

1064 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthe case of human infants, cellular sarcoplasmic reticulum changes dramatically during the perinatalperiod (Seguchi et al., 1986). Such changes suggest that the early perinatal hearts are moredependent on transsarcolemmal Ca 2+ influx in excitation-contraction coupling, when compared tothe adult heart, which depends on Ca 2+ release and uptake from sarcoplasmic reticulum. This maybe due to the immaturity in early neonates of the transverse tubular system that interconnects withthe sarcoplasmic reticulum (Nakamura et al., 1986).In dogs, the number of myocytes remains relatively constant as the heart enlarges with increasingage, but myocyte shape and dimensions and intercellular connections changed dramatically (Legato,1979).Average Diameter of Cardiac Myocytes(Rakusan, 1980; Hoerter et al., 1981)AgeDiameter of Myocytes (µm)Human Birth 56 weeks 83 years 1115 years 13Adult 14Rat 5–14 days 5.530 days 10.5–11.8Young adult 15.0–16.0Rabbit Day 8 5Adult 14Dog < 100 days 6.50.5–0.9 year 13.31–4.2 years 13.8An increase in left ventricular mass, due to an increase in myocyte size rather than proliferation,is also observed in the neonatal pig (Satoh et al., 2001).In addition to changes in myocyte diameter, a progressive change in the organization and patternof association between gap junctions and cell adhesion junctions are also observed within the first20 postnatal days in rats and dogs (Gourdie et al., 1992; Angst et al., 1997).Postnatal Cardiac VasculatureThere are marked species differences in the development of the coronary vascular bed, dependingon the degree of maturity of the species at birth (Hudlicka and Brown, 1996). Rapid developmentalheart growth is accompanied by a proportional growth of capillaries but not always of larger vessels,and thus coronary vascular resistance gradually increases. Growth of adult hearts can be enhancedby thyroid hormones, catecholamines and the renin-angiotensin system hormones, but these do notalways stimulate growth of coronary vessels. Likewise, chronic exposure to hypoxia leads to growth,mainly of the right ventricle and its vessels but without vascular growth elsewhere in the heart. Onthe other hand, ischaemia is a potent stimulus for the release of various growth factors involved inthe development of collateral circulation. In humans, the coronary vascular bed is well establishedat birth, although some capillary angiogenesis and vessel maturation occurs postnatally. In comparison,rats show a greater immaturity in the development of the coronary vascular bed at birth,with marked development of the coronary arteries and capillaries postnatally. In all species, capillarydensity decreases with age into adulthood.HumanThe coronary vascular bed in humans is established well before birth, but some capillary angiogenesisoccurs postnatally. Rakusan et al. (1994) report similar arteriolar density and vessel thickness in hearts© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1065from infants, children and adults. While larger branches are already formed at birth, there are reportsthat smaller arteriole vessels continue to proliferate postnatally (Ehrich et al., 1931; Rakusan etal., 1992). The rate of cardiac growth is higher than the growth of capillaries. The number ofcapillaries per arteriole decreases from 500 in infant hearts to 397 in children and 347 in adulthearts. During postnatal development, final vessel maturation is characterized by the establishmentof definitive morphological and biochemical features in the vascular walls and changes in the spatialorientation of the capillaries (Rakusan et al., 1994). In newborns, the average diameter of coronaryarteries is 1 mm. It doubles during the first postnatal year, and it reaches maximal values aroundthe age of 30 years (Vogelberg, 1957; Rabe, 1973).Comparative SpeciesIn rats, development of both the coronary arteries and capillaries continues after birth (Olivetti et al.,1980a; Mattfeldt and Mall, 1987; Batra and Rakusan, 1992; Rakusan et al., 1994; Tomanek et al.,1996; Heron et al., 1999). Formation and maturation of coronary arterioles, including thickening ofthe media due to intensive production of connective tissue and hypertrophy, is complete within the firstpostnatal month (Looker and Berry, 1972; Olivetti et al., 1980b; Heron et al., 1999). Following birth,the coronary capillaries also continue to proliferate and mature, with close to half of all capillaries inthe adult heart forming during the early postnatal period (Rakusan et al., 1994). During the first fewweeks, capillaries grow 2-3 times more rapidly than heart mass. The volume fraction of capillaries inthe rat heart increases from values around 4% at birth to 16% by postnatal day 28, with subsequentdecrease to adult values of 8-9%. Recent studies of early postnatal vascularization in the rat report thatboth vascular endothelial growth factor (VEGF) and beta-fibroblast growth factor (bFGF) modulatecapillary growth and bFGF facilitates arteriolar growth (Tomanek et al., 2001).In rabbits, there is a rapid growth of the heart capillaries during the first postnatal week, butno growth was detected in adult (Rakusan et al., 1967).During the postnatal period in dogs, the degree of myocyte surface area in close contact tocapillary walls increases dramatically (Legato, 1979).Postnatal Cardiac InnervationThere are some species differences in the development of cardiac innervation, depending on thedegree of maturity of the heart of the species at birth, but the general processes appear to be similar.HumanInnervation of the human heart is both morphologically and functionally immature at birth. Thenumber of neurons gradually increases and reaches a maximum density in childhood, at whichtime an adult pattern in innervation is achieved (Chow et al., 2001).Comparative SpeciesIn rats, adrenergic innervation patterns in cardiac tissues and the vasculature reach adult levels bythe third postnatal week, while the thickness and density of nerve fibers reach adult levels by thefifth postnatal week (De Champlain et al., 1970; Scott and Pang, 1983; Schiebler and Heene, 1986;Horackova et al., 2000). Histochemical studies suggest that ventricular innervation develops laterthan that of the atria in the rat (De Champlain et al., 1970). The cholinergic innervation is alsostructurally and functionally immature at birth in the rat heart (Truex et al., 1955; Winckler, 1969;Vlk and Vincenzi, 1977; Horackova et al., 2000). Neuropeptide Y and tyrosine hydroxylase positivenerve fibers are reported to increase rapidly during the first two weeks postnatally and reach adultlevels by the third week (Nyquist-Battie et al., 1994).© 2006 by Taylor & Francis Group, LLC

1066 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONDevelopment and maturation in adrenergic innervation patterns are reported to occur earlier inthe guinea pig, compared to rats (De Champlain et al., 1970; Friedman, 1972).Friedman et al. (1968) reported comparable cholinergic density in neonate and adult rabbits.In the dog, the adrenergic and cholinergic systems are incomplete at birth, and their developmentcontinues during the first 2 months of age (Ursell et al., 1990; Cua et al., 1997). The quantities ofmonoaminergic nerve endings are similar to adult levels in the atria by 6 weeks of age and in theventricles by 4 months of age (Dolezel et al., 1981).Postnatal Functional and Physiological Development of the HeartBiochemistry of the Heart during Postnatal DevelopmentA large amount of information, primarily in experimental animals, is available on the developmentalchanges in cardiac metabolism that is beyond the scope of this review (see Rakusan, 1980 forreview). The major characteristic change during early development is the change from primarilycarbohydrate energy metabolism to lipid energy metabolism, increasing concentrations of cytochromes,creatinine, and phosphocreatinine and increasing activity in mitochondrial enzymes activity.HumanWater and Mineral Content — Widdowson and Dickerson (1960) reported a small but significantreduction in water content in the heart during postnatal development. No significant changesin calcium, magnesium, sodium and potassium were observed in heart autopsies in subjects rangingfrom newborn to 90 years of age (Eisenstein and Wied, 1970).Energy Metabolism — The switching energy metabolism from carbohydrate (glucose) to fattyacids involves changes in phosphofructokinase isozymes. In humans, this isozyme switch occursbefore birth (Davidson et al., 1983). In addition, the concentrations of cytochromes and mitochondrialprotein in cardiac cells increase during the early postnatal period (Rakusan, 1980).Comparative SpeciesWater and Mineral Content — Solomon et al. (1976) reported a rapid postnatal decline in watercontent of the rat heart up to 23 days of age. Thereafter, a smaller decline toward adult levels wasobserved. Sodium (Na) and potassium (K) levels also increased to the maximum value by 16 daysof age with a subsequent marked decrease by 40 days of age (Hazlewood and Nichols, 1970;Solomon et al., 1976). The ratio of Na/K is very high at birth and declines to its minimum at 40days of age. Chloride and calcium content declined with age.Lipid and Protein Content — Immature rats are reported to have lower myocardial myoglobinconcentrations compared to adults (Rakusan et al., 1965; Dhindsa et al., 1981). A slight increasein heart phospholipid levels was reported in the early postnatal rat (Carlson et al., 1968).In hamsters, heart lipid and phospholipids increase during the first 4 months postnatally (Barakatet al., 1976).Energy Metabolism — The heart in postnatal animals shows a higher dependence on glucose asan energy source than the adult heart (see Rakusan, 1980 for further review). Unlike humans, inwhich phosphofructokinase isozyme changes responsible for the switching of myocardial fuelsfrom carbohydrate (glucose) to fatty acids occur before birth, rat isozyme changes occur duringthe first 2 postnatal weeks (Davidson et al., 1983). Carbohydrate dependence may also be partlydue to the limited ability to oxidize long chain fatty acids, compared to adults.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1067Adenylate cyclase and guanylate cyclase activities, as well as cyclic GMP levels, are higher inheart homogenates of weanling compared to adult rats (Dowell, 1984). Cyclic AMP levels weresimilar in weanlings and adult rats.The concentrations of cytochromes and mitochondrial proteins as well as enzyme activity incardiac cells increase during the early postnatal period in rats and dogs (Rakusan, 1980; Schonfeldet al., 1996). In comparison, the heart of the guinea pig, which is more mature at birth, showed nodifferences in cardiac mitochondrial enzyme activity between newborns and adults (Barrie andHarris, 1977).Creatinine and phosphocreatine increased postnatally in the dog, rabbit and guinea pig duringthe early postnatal period (Rakusan, 1980). In the rat and mouse, creatinine phosphokinase levelsin the heart increase after birth and reach adult values and isozyme types by 25 - 30 days of age(Hall and De Luca, 1975; Baldwin et al., 1977).Postnatal Electrophysiology of the HeartIn general, cardiac innervation, especially the adrenergic component, is functionally immature at birth.There are age-dependent alterations in the myocardial alpha 1-adrenergic, beta-adrenergic and muscarinicsignal transduction cascades during postnatal development (Rakusan, 1980, 1984; Robinson, 1996;Garofolo et al., 2002). An inhibitory alpha 1-adrenergic response appears in early maturation, whichdiffers from the pre-existing excitatory response both with respect to the specific receptor subtypeinvolved and its G protein coupling. Likewise, sympathetic innervation appears to be involved in theloss of an excitatory muscarinic response during development. The role of innervation in developmentalregulation of the beta-adrenergic response is not fully understood. Functional innervation during thepostnatal period favors excitation (chronotropic and/or inotropic) over inhibition. In addition, heartsfrom younger animals have higher stimulation thresholds than those from adults.HumanNerve Functional Development — Innervation of the human heart is both morphologically andfunctionally immature at birth (Rakusan, 1980, 1984). Human infants show a paucity of neuronswith acetylcholinesterase (AChE), but a substantial amount of pseudocholinesterase activity, comparedto adults (Chow et al., 1993, 2001). During maturation into adulthood, a gradual loss inpseudocholinesterase activity occurs with increasing numbers of AChE-positive nerves. This coincideswith initial sympathetic dominance in the neural supply to the human heart in infancy, andits gradual transition into a sympathetic and parasympathetic codominance in adulthood.Electrocardiograpic and Vectorcardiographic Measurements — In general, the duration ofECG deflection and intervals increase with age, reflecting a postnatal decline in basal heart rate.The amplitudes of ECG waves are also age dependent, but vary widely. Early changes in QRS-Trelationships from birth to 4 days of age reflect the immediate transition from in utero to postnatallife (Rautaharju et al., 1979). Over a number of years, the horizontal QRS loops change from acomplete clockwise rotation in newborns into a figure-of-eight loop that eventually unwinds intoa counterclockwise loop orientation towards the left, as normally observed in adults (Namin andMiller, 1966; Rautaharju et al., 1979). The magnitude of the ventricular gradient vector changesincreases from 3 weeks of age until about 7 years of age (Rautaharju et al., 1979). The spatialangle between QRS and STT vectors reaches its minimum at 1.5 to 4.5 years of age.Comparative SpeciesNerve Functional Development — There are marked species differences in the development offunctional innervation of the heart in the perinatal period (Vlk and Vincenzi, 1977). At birth, the© 2006 by Taylor & Francis Group, LLC

1068 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrat heart is quite immature functionally and requires several weeks to mature. On the other hand,guinea pig and rabbit heart innervation functions are relatively mature at birth, with cholinergicfunction more mature than adrenergic function.In the rat, final functional sympathetic and parasympathetic innervation of the heart occurs afterbirth, with the parasympathetic system maturing first (Lipp and Rudolph, 1972; Mackenzie andStanden, 1980; Quigley et al., 1996). Adrenergic - and -receptors are present in the myocardiumat birth in the rat; however, positive inotropic responses to sympathetic nerve stimulation do notdevelop until between 2 and 3 weeks of age (Mackenzie and Standen, 1980; Metz et al., 1996).Inotropic responsiveness in the rat atria shows significant levels at one week of age and reach 70%of adult values at two weeks of age. Receptor expression during this period appears to be regulatedby thyroid hormone (Metz et al., 1996). Concomitantly, opposing A 1 adenosine receptor density isreported to be twice as numerous than adults and decrease to adult levels by two weeks of age(Cothran et al., 1995). It is believed that catecholaminergic stimulation plays an important role inmaintaining basal heat rate during this early neonatal period in the rat (Tucker, 1985). After birth,sensitivity of the heart to catecholamines increases with postnatal age and generally shows highersensitivity to catecholamines, compared to adult hearts (Penefsky, 1985; Shigenobu et al., 1988).Finally, it has been suggested that the cardiac renin-angiotensin system may also play a role inheart growth and in the adaptation of the heart to postnatal circulatory conditions (Sechi et al.,1993; Charbit et al., 1997).Seven-day old rabbits and guinea pigs display adult cholinergic responsiveness to stimulation(Vlk and Vincenzi, 1977). However, neonatal rabbits appear to display some functional deficiencyof adrenergic innervation as manifested by lack of blood pressure increase and heart rate decreaseafter asphyxiation.Cardiac adrenergic and cholinergic innervation is also structurally and functionally immatureat birth in the dog (Truex et al., 1955; Winckler, 1969; Vlk and Vincenzi, 1977; Haddad and Armour,1991; Pickoff and Stolfi, 1996). Both the sympathetic and parasympathetic systems become fullyfunctional by about 7 weeks of age (Haddad and Armour, 1991).Cardiac Neurotransmitters — The concentration of norepinephrine in the heart is very low atbirth but increases rapidly during the first postnatal weeks in the rat, rabbit, hamster, pig and sheep(Friedman et al. 1968; De Champlain et al., 1970; Heggenes et al., 1970; Friedman, 1972; Sole etal., 1975).In the rat and rabbit, concentrations and uptake of catecholamine in the heart are very low at birthbut increase rapidly during the first postnatal weeks (Glowinski et al., 1964; Friedman et al., 1968).Tissue levels of acetylcholine in the atria of 3-4 day old rats are reported to be only one-sixththe levels found in the adult (Vlk and Vincenzi, 1977). Cholinesterase first appears in neurons at4 days of age and is present in nerve fibers of the sinus node by 15 days of age.Electrocardiographic and Vectorcardiographic Measurements — In the rat, the PQ intervaldecreases between 6 and 14 days of age and reaches adult values at 53 days of age (Yamori et al.,1976; Diez and Schwartze, 1991). The QT interval also decreases during the first 3 week of lifeand reaches adult values later than 53 days of age.As in the case in humans, shorter QRS intervals are noted and horizontal QRS loops changefrom clockwise, figure-of-eight to counterclockwise loop oriented towards the left, as observed inthe adult, during the early postnatal period in rats (Diez and Schwartze, 1991). QRS intervals areconsiderably shorter than in the adult, even at day 53. At this age, rats have attained the adultconfiguration of the QRS loop. From day 30 onward, QRS frequency merges with the T wave.Similar findings are reported in the guinea pig where the QRS interval increases from 15.5 msin the newborn to 32 ms in the adult (Yamori et al., 1976). The lengthening of the conduction© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1069course over the ventricles in adults, compared to infant and juvenile animals, might explain thisincreasing excitation period.In mice, the entire J junction-S-T segment-T wave complex, R-R interval and Q-T intervalswere shorter during early postnatal development and matured by postnatal days 7 to 14 (Wang etal., 2000). Age-dependent ECG responses to K + channel blockers were also reported in mice.Postnatal Cardiac Output and HemodynamicsAt birth, the cardiovascular system changes dramatically; arterial blood pressure, heart rate, andcardiac output increase, and blood flow distribution undergoes regional changes. During the earlypostnatal period, increasing heart rate, rather than stroke volume, is the primary mechanism bywhich cardiac output is increased during the neonatal period.HumanThe mean basal heart rate at birth is approximately 120 to 130 beats/minute. A steep increase inheart rate occurs between the 5 th and 10 th postnatal day, reaching maximum levels of ~140-150beats/minute, which then gradually decrease to ~120 beats/minute over the first 100 postnatal daysas parasympathetic restraints develop. Adult heart beat rates (55-85 beats/min) are achieved around12 years of age (Oberhansli et al., 1981; Mrowka et al., 1996).Mean Values for Heart Functional Parameters in Human Infants (Oberhansli et al., 1981)Heart Parameter1day3days6daysAge1month2month6–11months12–14monthsHeart Rate (beats/min) 133 129 135 155 150 140 124RV Pre-ejection Period (msec) 71 63 59 51 55 55 61RV Ejection Time (msec) 199 203 203 193 204 232 243LV Pre-ejection Period (msec) 65 61 59 55 59 59 65LV Ejection Time (msec) 197 193 192 184 192 200 228Mean Velocity 1.67 1.72 1.75 1.79 1.73 1.97 1.51Ejection Fraction 70 70 71 70 70 77 71RV- right ventricle; LV- left ventricleInfants and young children are reported to have significantly smaller ventricular end-diastolicvolumes and stroke indexes than adults, slightly smaller ejection fractions, but no differences incardiac index (Graham et al., 1971; Mathew et al., 1976). Oh et al. (1966) reported marked increasesin the distribution of cardiac output to the kidney during the first days after birth, but still a smallfraction of output, when compared to adults.The average blood pressure values (systolic/diastolic) increased rapidly from 62.1/39.7 mm Hgat age day one to up to 72.7/46.9 mm Hg at age one week and up to 85.0/47.4 mmHg at age twomonths (Hwang and Chu, 1995). Blood pressure changes relatively little between the ages of 6months and 10 years. Systolic blood pressure rose from a mean of 88.5 mm Hg at age 6 monthsto 96.2 mm Hg at 8 years, and diastolic blood pressure rose from 57.8 mm Hg at 5 years to 61.8mm Hg at 10 years (de Swiet et al., 1992). Blood pressure was correlated with weight, weightadjusted for height, height, and arm circumference, at all ages studied. Blood pressure values(systolic/diastolic) increased rapidly from 68.0/43.7 mm Hg in children weighing less than 5 kgup to 87.6/41.8 mm Hg in those weighing 5 to 10 kg. Subsequently, these values increased graduallywith body weight (Hwang and Chu, 1995).© 2006 by Taylor & Francis Group, LLC

1070 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAssociations between birth weight, childhood growth and blood pressure are reported bynumerous investigators (Georgieff et al., 1996; Uiterwaal et al., 1997; Walker et al., 2001). Therewas a significant correlation between systolic and diastolic blood pressure and both weight andlength at birth to 4 months of age (Georgieff et al., 1996). Greater weight gain between ages 7 and11 was associated with a greater increase in systolic blood pressure. The relation between growthand blood pressure is complex and has prenatal and postnatal components (Walker et al., 2001).Systolic blood pressure increased, with height, from 111 to 120 mm Hg in girls (body height 120to 180 cm), as well as in boys from 112 to 124 mm Hg (Steiss and Rascher, 1996). However, meandiastolic blood pressure did not change during maturation and was 72 to 74 mm Hg, irrespectiveof height and sex.The correlation coefficient between systolic and diastolic blood pressure increased steadily withage from 0.28 at 2 years to 0.59 at 10 years, but do not reach adult levels during this period (Levineet al., 1979; de Swiet et al., 1992).In adults, arterial baroreceptors are the major sensing elements of the cardiovascular system.In the majority of species, including humans, baroreceptor reflexes at birth exhibit depressedsensitivity with a gradual postnatal maturation to adult levels (Gootman et al., 1979; Vatner andManders, 1979).Comparative SpeciesThe heart rate in neonatal rats increases during the early postnatal period and then remains relativelyconstant into adulthood (Kyrieleis, 1963; Diez and Schwartze, 1991; Quigley et al., 1996). Thisincrease in rate appears to be inversely proportional to the decrease in QT interval seen during thisperiod, and may be explained by the increasing influence of the sympathetic system.Heart Rate of Postnatal Wistar Rats(Diez and Schwartze, 1991)Age (days)Heart Rate (beats/min)6 296 ± 6014 405 ± 3321 433 ± 4830 436 ± 3638 444 ± 2453 435 ± 27Hemodynamics in neonatal rats is characterized by a high cardiac output and low systemicresistance (Prewitt and Dowell, 1979). Shortly before puberty, there is a rise of systemic resistance,accompanied by a decrease in cardiac output (Smith and Hutchins, 1979). This is primarily due tostructural maturation of the vasculature. Developmental changes in vascular sensitivity to vasoactiveagents play a less important role for the maturational rise in systemic resistance. Marked maturationof the baroreceptor reflex also occurs during the postnatal period (Andresen et al., 1980). Normalsystolic blood pressure in the young rat is more than double that of the neonate and reaches adultlevels by 10 weeks of age (Litchfield, 1958; Gray, 1984; Rakusan et al., 1994).Blood Pressure in RatsAgeBlood Pressure (mm Hg)Newborn 15–253-4 Weeks of Age 80–95© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1071Estimation of coronary blood flow indicated no significant differences between immature andadult rats (Vizek and Albretch, 1973; Rakusan and Marcinek, 1973). However, marked changeshave been reported in the organ distribution of cardiac output throughout postnatal development,especially during the first four weeks (Rakusan and Marcinek, 1973). The most prominent observationis the very low renal blood flow, which gradually increases during the postnatal period.Wiesmann et al. (2000) reported increases in cardiac output and stroke volume in mice with age,but no significant change in heart rate, ejection fraction, and cardiac index with age. On the otherhand, Tiemann et al. (2003) recently reported an increase in heart rate during maturation from 396beats/min at 21 days of age to 551 beats/min at 50 days of age. Mean arterial blood pressure alsoincreased in parallel from 86 to 110 mm Hg and remained constant thereafter. Baroreflex heart ratecontrol matures at around 2 weeks after birth in the mouse (Ishii, 2001).Cardiac Output and the Heart Rate of C57bl/6 mice (Wiesmann et al., 2000)Age3 days 10 days 3 weeks 4 weeks 5 weeks 10 weeks 16 weeksLV cardiac output (ml/min) 1.1 5.3 8.7 9.3 11.2 15.7 14.3LV stroke volume (µl) 3.2 15.0 20.8 23.9 30.5 35.6 40.2Heart rate (beats/min) 372 418 422 390 366 442 360LV-left ventricleCardiac output fractions to the kidneys and intestines are markedly lower in newborn rabbits,when compared to adults (Boda et al., 1971).In dogs, a significant increase in blood pressure and a decrease in heart rate occurred withgrowth from 1 week to 6 months (Adelman and Wright, 1985). These changes are qualitativelysimilar to those observed in young infants and children. Baroreceptor reflex control in the neonateis less developed than in the adult heart (Hageman et al., 1986).The heart rate in the pig does not change significantly with age (Satoh et al., 2001). Systolicand diastolic blood pressures increase (from 57 to 94 mm Hg and from 38 to 55 mm Hg,respectively) during the first 6 days postnatally in the pig (Satoh et al., 2001).In the sheep, heart rate, cardiac output, ventricular output, and stoke volume are higher inneonates, when compared to adults (Klopfenstein and Rudolph, 1978; Berman and Musselman,1979). Cardiac output in the newborn lamb is four times greater than in the adult. At birth,baroreceptor reflexes exhibit depressed sensitivity, with a gradual postnatal maturation to adultlevels (Dawes et al., 1980; Segar et al., 1992).Cardiac Parameters in SheepHeart Parameter Term Fetus Newborn AdultHeart Rate (beats/min) 150 200 100Cardiac Output (ml/min/kg) 450 400 100Ventricular Output (ml/min/kg) - Left 150 400 100Ventricular Output (ml/min/kg) - Right 300 400 100Stoke Volume (ml/kg) - Left 1 2 1Stoke Volume (ml/kg) - Right 2 2 1© 2006 by Taylor & Francis Group, LLC

1072 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSummary Table – Postnatal Heart Development: A Species ComparisonParameter Human Rat DogHeart Size/ShapeHeart WeightCardiac CellsCoronary VasculatureCardiac InnervationHeart position higher andmore transverse;oblique lie attainedbetween 2 and 6 yearsof age.At birth, ventricularvolumes are equal.Right ventricular volumedoubles by 12 months,left ventricular volumeunchanged. R/L ratioreaches 2:1 at 2 years.Relative heart weightconstant.Absolute weight doubledby 6 months and tripledby 1 year; achievedadult relative to bodyweight by about 21years.Myocyte count at birth is50% of adult, withproliferation, adult valuereached by 4 months.Growth thereafter due tomyocyte hypertrophy: 5µm at birth, 8 µm at 6wks, 11 µm at 3 yrs., 13µm at 15 yrs and 14 µmin adults.Myocytes diploid at birth,compared to 60%diploid, 40% polyploid inadults.Primary arteriesestablished before birth,diameter of coronaryarteries doubled at 1styear reaching maximumat 30 years of age.Some capillaryangiogenesis occurspostnatally. Capillarydensity decreases withage.Morphologically andfunctionally immature atbirth. Number ofneurons graduallyincrease and reachmaximum density andadult patterns inchildhood.Age-related changes inventricular dimensions.Becomes morespherical with age.Decreasing relative heartweight with age.High weight increases onpostnatal days 1-5 andslower growthafterwards.Myocyte proliferation (3-4fold increase) andhypertrophy from birth to2 months of age.Myocyte diameter 5.5µm at 14 days, 10.5-11.8 µm at 30 days and15-16 µm in adults.Myocytes primarilydiploid in both infant andadult.More immature at birth.Capillary and arterioleangiogenesis occurspostnatally. Arterialmaturation by onemonth.Volume fraction ofcapillaries reachesmaximum of 16% by day28. Capillary densitysubsequently decreaseswith age.Morphologically andfunctionally immature atbirth. Adrenergicpatterns complete by 3weeks and nervedensity complete by 5weeks. Cholinergic andother nerve types alsomatured postnatally.Age-related changes inventricular dimension.Becomes morespherical with age.Decreasing relative heartweight with age.Myocardial cell numbersrelatively constant.Growth primarilythrough myocytehypertrophy: 7 µm at100 days, 13 µm at0.5-0.9 year and 14 µmat 1-4.2 years.No information availableMinimal informationavailable. Capillaryangiogenesis occurspostnatally.Morphologically andfunctionally immature atbirth and continuesdevelopment during thefirst 2 –4 months of age.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1073Summary Table – Postnatal Heart Development: A Species ComparisonParameter Human Rat DogElectrophysiologyCardiac BiochemistryShorter QRS intervals inneonates with horizontalQRS loops changingfrom a completeclockwise rotation into afigure-of-eight loop thateventually unwinds intoa counterclockwise looporientation towards theleft as normallyobserved in adults. Themagnitude of theventricular gradientvector changesincreases from age 3weeks until about 7years of age. The spatialangle between QRS andSTT vectors reaches itsminimum at 1.5 to 4.5years of age.Small decrease in watercontent but electrolyteconcentrations relativelyconstant.Shorter QRS intervals inneonates with horizontalQRS loops changingfrom clockwise, figureof-eighttocounterclockwise looporiented towards the left.PQ interval decreasesbetween 6 and 14 daysand reaches adultvalues at 53 days ofage. The QT intervalalso decreases duringthe first 3 weeks. Fromday 30 onward, QRSfrequency merges withthe T wave.Rapid decline in watercontent up to 23 days ofage. Increased Na andK with maximum levelsat 16 days. Chloride andCa levels decrease withage.No information available.No information available.Concentrations ofcytochromes andmitochondrial proteinsincrease during earlypostnatal period.Concentrations ofcytochromes,mitochondrial proteinsand enzymes increaseduring early postnatalperiod.Concentrations ofcytochromes,mitochondrial proteinsand enzymes increaseduring early postnatalperiod.Cardiac Output andHemodynamicsPhosphofructokinaseisozyme switch occursbefore birth.Decreasing basal heartrate: 138 beats/min atbirth to 55-85 beats/minin adults.Infants and youngchildren have smallerventricular volumes,stroke index andejection fractions, but nodifferences in cardiacindex compared toadults.Phosphofructokinaseisozyme switch occursduring first 2 postnatalweeks.Early increase in heartrate, then remainsrelatively constant intoadulthood.High cardiac output andlow systemic resistancepostnatally.Significant increase inblood pressure anddecrease in heart ratefrom 1 week to 6months.Rapid increase in bloodpressure from birth totwo months(systolic/diastolic –62/40 to 85/47);relatively constant from6 months to 8 years ofage (diastolic 58 to 62).Systolic blood pressuredoubles from neonate toyoung adults andreaches adult levels by10 weeks of age.No information available.© 2006 by Taylor & Francis Group, LLC

1074 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONREFERENCESAddis, T, Gray, H. 1950. Body size and organ weight. Growth 14:49-80.Adeagbo AS, Bishai I, Lees J, Olley PM, Coceani F. 1985. Evidence for a role of prostaglandin I2 andthromboxane A2 in the ductus venosus of the lamb. Can J Physiol Pharmacol 63:1101-1105.Adelman RD, Wright J. 1985. Systolic blood pressure and heart rate in the growing beagle puppy. DevPharmacol Ther 8:396-401.Andresen MC, Kuraoka S, Brown AM. 1980. Baroreceptor function and changes in strain sensitivity innormotensive and spontaneously hypertensive rats. Circ Res 47:821-828.Angst BD, Khan LU, Severs NJ, Whitely K, Rothery S, Thompson RP, Magee AI, Gourdie RG. 1997.Dissociated spatial patterning of gap junctions and cell adhesion junctions during postnatal differentiationof ventricular myocardium. Circ Res 80:88-94.Anversa P, Capasso JM, Olivetti G, Sonnenblick EH. 1992. Monocyte cellular hypertrophy and hyperplasiain the heart during postnatal maturation and ageing. Acta Paediatr 383:29-31.Anversa P, Olivetti G, Loud AV. 1980. Morphometric study of early postnatal development in the left andright ventricular myocardium of the rat. I. Hypertrophy, hyperplasia, and binucleation of myocytes.Circ Res 46:495-502.Anversa P, Ricci R, Olivetti G. 1986. Quantitative structural analysis of the myocardium during physiologicgrowth and induced cardiac hypertrophy: a review. J Am Coll Cardiol 7:1140-1149.Ashley LM. 1945. A determination of the diameters of myocardial fibers in man and other mammals. J Anat77:325-347.Baldwin KM, Cooke DA, Cheadle WG. 1977. Enzyme alterations in neonatal heart muscle during development.J Mol Cell Cardiol 9:651-60.Barakat HA, Dohm GL, Loesche P, Tapscott EB, Smith C. 1976. Lipid content and fatty acid composition ofheart and muscle of the BIO 82.62 cardiomyopathic hamster. Lipids 11:747-51.Barrie SE, Harris P. 1977. Myocardial enzyme activities in guinea pigs during development. Am J Physiol233:H707-H710.Batra S, Rakusan K. 1992. Capillary network geometry during postnatal growth in rat hearts. Am J Physiol262: H635-H640.Berman W, Musselman J. 1979. Myocardial performance in the newborn lamb. Am J Physiol 237:H66-H70.Boda D, Belay M, Eck E, Csernay L. 1971. Blood distribution of the organs examined by 86 Rb uptake underintrauterine conditions and in the newborn, in normal and hypoxic rabbits. Biol Neonate 18:71-77.Boucek RJ Jr. 1982. Polyamine metabolism during the perinatal development of the rabbit right and leftventricle. Pediatr Res 16:721-727.Broccoli F, Carinci P. 1973. Histological and histochemical analysis of the obliteration processes of ductusarteriosus Botalli. Acta Anat 85:69-83.Carlson LA, Froberg SO, Nye ER. 1968. Effect of age on blood and tissue lipid levels in the male rat.Gerontologia 14:65-79.Charbit M, Blazy I, Gogusev J, et al. 1997. Nitric oxide and the renin angiotensin system: contributions toblood pressure in the young rat. Pediatr Nephrol 11:617-22.Chow LT, Chow SS, Anderson RH, Gosling JA. 1993. Innervation of the human cardiac conduction systemat birth. Br Heart J 69:430-5.Chow LT, Chow SS, Anderson RH, Gosling JA. 2001. Autonomic innervation of the human cardiac conductionsystem: changes from infancy to senility--an immunohistochemical and histochemical analysis. AnatRec 264:169-82.Chubb FJ, Bishop SP. 1984. Formation of binucleated cells in the neonatal rat. Lab Invest 50:571-577.Clyman RI. 1987. Ductus arteriosus: current theories of prenatal and postnatal regulation. Seminars inPerinatology 11:64-71.Clyman RJ, Campbell D, Heymann M., et al. 1985a. Persistent responsiveness of the neonate ductus arteriosusin immature lambs: A possible cause for reopening of patent ductus arteriosus after indomethacininducedclosure. Circulation 71:141-145.Clyman RJ, Breall J, Maher P., et al. 1985b. Thyroid hormones and the ductus arteriosus. Pediatr Res 19:338A(abstr).Coceani F, Hamilton NC, Labuc J., et al. 1984. Cytochrome P 450-linked monooxygenase: involvement inthe lamb ductus arteriosus. Am J Physiol 246: H640-H643.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1075Cothran DL, Lloyd TR, Taylor H, Linden J, Matherne GP. 1995. Ontogeny of rat myocardial A 1 adenosinereceptors. Biol Neonate 68:111-118.Cua M, Shvilkin A, Danilo P, Rosen MR. 1997. Developmental changes in modulation of cardiac repolarizationby sympathetic stimulation: The role of beta- and alpha-adrenergic receptors. J Cardiovasc Electrophysiol8:865-871.Dawes GS, Johnston BM, Walker DW. 1980. Relationship of arterial pressure and heart rate in fetal, newbornand adult sheep. J Physiol 309:405-417.Davidson M, Collins M, Byrne J, Vora S. 1983. Alterations in phosphofructokinase isoenzymes during earlyhuman development. Establishment of adult organ-specific patterns. Biochem J 214:703-10.Deavers S, Huggins RA, Smith EL. 1972. Absolute and relative organ weights of the growing beagle. Growth36:195-208.De Champlain J, Malmfors T, Olson L, Sachs C. 1970. Ontogenesis of peripheral adrenergic neurons in therat: pre- and postnatal observations. Acta Physiol Scand 80: 276-288.de Swiet M, Fayers P, Shinebourne EA. 1992. Blood pressure in first 10 years of life: The Brompton study.Brit Med J 304:23-26.Diez U, Schwartze H. 1991. Quantitative electrocardiography and vectorcardiography in postnatally developingrats. J Electrocardiol 24:53-62.Dhindsa DS, Metcalf J, Blackmore DW, Koler RD. 1981. Postnatal changes in oxygen affinity of rat blood.Comp Biochem Physiol 69A:279-283.Dolezel S, Gervoa M, Gero J, Vasku J. 1981. Development of sympathetic innervation of the coronary arteriesand the myocardium in the dog. Morphologica 29:189-191.Dowell RT. 1984. Metabolic and cyclic nucleotide enzyme activities in muscle and non-muscle cells of ratheart during perinatal development. Can J Physiol Pharmaco 63:78-81.Eckner FAO, Brown BW, Davidson DL, Glagov S. 1969. Dimension of normal human hearts. Arch Pathol88:497-507.Ehrich W, De La Chapelle C, Cohn AE. 1931-1932. Anatomical ontogeny. B. Man (a study of the coronaryarteries). Am J Anat 49:241-282.Eisenstein R, Wied GL. 1970. Myocardial DNA and protein in maturity and hypertrophied hearts. Proc SocExp Biol Med 133:176-179.Englemann GL, Vitullo JC, Gerrity RG. 1986. Morphometric analysis of cardiac hypertrophy during development,maturation, and senescence in spontaneously hypertensive rats. Circ Res 60:487-494.Fay FS, Cooke PH. 1972. Guinea pig ductus arteriosus. II. Irreversible closure after birth. Am J Physiol222:841-849.Fay FS, Jobsis FF. 1972. Guinea pig ductus arteriosus. III. Light absorption changes during response to O 2 .Am J Physiol 223: 588-595.Fisher DJ, Towbin J. 1988. Maturation of the heart. Clin Perinat 15:421-446.Friedman AH, Fahey JT. 1993. The transition from fetal to neonatal circulation: Normal responses andimplications for infants with heart disease. Sem Perinat 17:106-121.Friedman WF. 1972. The intrinsic physiologic properties of the developing heart. Prog Cardiovasc Dis 15:87-111.Friedman WF, Pool PE, Jacobowitz D, Seagren SC, Braunwald E. 1968. Sympathetic innervation of thedeveloping rabbit heart. Biochemical and histochemical comparisons of fetal, neonatal, and adultmyocardium. Circ Res 23:25-32.Fried R, Jolesz FA, Lorenzo AV, Francis H, Adams DF. 1987. Developmental changes in proton magneticresonance relaxation times of cardiac and skeletal muscle. Invest Radiol 23:289-293.Fugelseth D, Lindemann R, Liestol K, Kiserud T, Langslet A. 1997. Ultrasonographic study of ductus venosusin healthy neonates. Arch Dis Child Fetal Neonatal Ed 77:F131-F134.Garofolo MC, Seidler FJ, Auman JT, Slotkin TA. 2002. beta-Adrenergic modulation of muscarinic cholinergicreceptor expression and function in developing heart. Am J Physiol Regul Integr Comp Physiol282:R1356-63.Georgieff MK, Mills MM, Gomez-Marin O, Sinaiko AR. 1996. Rate of change of blood pressure in prematureand full term infants from birth to 4 months. Pediatr Nephrol 10:152-155.Glowinski J, Axelrod J, Kopin IJ, Wurtman RG. 1964. Physiological disposition of H 3 -norepinephrine in thedeveloping rat. J Pharm Exp Ther 146: 48-53.Gootman PM, Buckley NM, Gootman N. 1979. Postnatal maturation of neural control of the circulation. RevPerinat Med 3:1-72.© 2006 by Taylor & Francis Group, LLC

1076 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONGourdie RG, Green CR, Severs NJ, Thompson RP. 1992. Immunolabelling patterns of gap junction connexinsin the developing and mature rat heart. Anat Embryol 185:363-378.Graham TP, Jarmakani JM, Canent RV, Morrow MN. 1971. Left heart volume estimation in infancy andchildhood. Reevaluation of methodology and normal values. Circulation 43:895-904.Gray SD. 1984. Pressure profiles in neonatal spontaneously hypertensive rats. Biol Neonate 45:25-32.Grimm AF, Katele KV, Klein SA, Lin HL. 1973. Growth of the rat heart: left ventricular morphology andsarcomere lengths. Growth 37:189-201.Grove D, Nair KG, Zak R. 1969. Biochemical correlates of cardiac hypertrophy. III. Changes in DNA content;the relative contributions of polyploidy and mitotic activity. Circ Res 25:463-471.Haddad C, Armour JA. 1991. Ontogeny of canine intrathoracic cardiac nervous system. Reg Integ CompPhysiol 30:R920-R927.Hageman GR, Neely BH, Urthaler F. 1986. Cardiac autonomic efferent activity during baroreflex in puppiesand adult dogs. Am J Physiol 251:H443-H447.Hall N, De Luca M. 1975. Developmental changes in creatine phosphokinase isoenzymes in neonatal mousehearts. Biochem Biophys Res Commun 66:988-94.Hazlewood CF, Nichols BL. 1970. Johns Hopkins Med J 127:136-145 (as cited in Rakusan, 1980).Heggeness FW, Diliberto J, Distefano V. 1970. Effect of growth velocity on cardiac norepinephrine contentin infant rats. Proc Soc Exp Biol Med 133:1413-1416.Heron MI, Kuo C, Rakusan K. 1999. Arteriole growth in the postnatal rat heart. Microvasc Res 58:183-186.Heymann MA, Rudolph AM. 1975. Control of the ductus arteriosus. Physiol Review 55:62-78.Hoerter J, Mazet F, Vassort G. 1981. Perinatal growth of the rabbit cardiac cell: Possible implications for themechanism of relaxation. J Mol Cell Cardiol 13:725-740.Horackova M, Slavikova J, Byczko Z. 2000. Postnatal development of the rat intrinsic cardiac nervous system:A confocal laser scanning microscopy study in whole-mount atria. Tissue Cell 32:377-388.Hort W. 1953. Quantitative Histologische Untersuchungen an Wachsenden Herzen. Virchows Arch [PatholAnat] 323:223-242.House EW, Ederstrom HE. 1968. Anatomical changes with age in the heart and ductus arteriosus in the dogafter birth. Anat Rec 160:289-296.Hudlicka O, Brown MD. 1996. Postnatal growth of the heart and its blood vessels. J Vasc Res 33:266-87.Hwang B, Chu NW. 1995. Normal oscillometric blood pressure values in Chinese children during their firstsix years. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 36:108-12.Ishii T, Kuwaki T, Masuda Y, Fukuda Y. 2001. Postnatal development of blood pressure and baroreflex inmice. Auton Neurosci 94:34-41.Jarkovska D, Janatova T, Hruda J, Ostadal B, Samanek M. 1989. The physiological closure of ductus arteriosusin the rat, an ultrastructural study. Anat Embryol 180:497-504.Jones M, Barrow MV, Wheat MW. 1969. An ultrastructural evaluation of the closure of the ductus arteriosusin rats. Surgery 66:891-898.Kondo M, Itoh S, Kunikata T, Kusaka T, Ozaki T, Isobe K, Onishi S. 2001. Time of closure of ductus venosusin term and preterm neonates. Arch Dis Child Fetal Neonatal Ed 85:F57-F59.Klopfenstein MS, Rudolph AM. 1978. Postnatal changes in the circulation and responses to volume loadingin sheep. Circ Res 42:839-845.Korecky B, Rakusan K. 1978. Normal and hypertrophic growth of the rat heart: Changes in cell dimensionsand number. Am J Physiol 234:H123-H128.Kyrieleis C. 1963. Die Formveränderungen des menschlichen Herzens nach der Geburt. Virchows Arch [PatholAnat] 337:142-163.Lee JC, Taylor JFN, Downing SE. 1975. A comparison of ventricular weights an geometry in newborn, youngand adult mammals. J Appl Physiol 38:147-150.Legato MJ. 1979. Cellular mechanisms of normal growth in the mammalian heart. I. Qualitative and quantitativefeatures of ventricular architecture in the dog from birth to five months of age. Circ Res 44:250-262.Leu M, Ehler E, Perriard JC. 2001. Characterisation of postnatal growth of the murine heart. Anat Embryol204:217-24.Levine RS, Hennekens CH, Klein B, et al., 1979. A longitudinal evaluation of blood pressure in children. AmJ Public Health 69:1175-1177.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1077Li F, Wang X, Capasso JM, Gerdes AM. 1996. Rapid transition of cardiac myocytes from hyperplasia tohypertrophy during postnatal development. J Mol Cell Cardiol 28:1737-1746.Linzbach AJ. 1950. Die Muskelfaserkonstante und das Wachstumgesetz der menschlichen Herzkammer.Virchows Arch [Pathol Anat] 318:575-618.Linzbach AJ. 1952. Die Anzahl der Herzmuskelkerne in normalen, überlasteten, atrophen und Corhormonbehandelten Herzkammer. Z Kreislaufforsch 41:641-658.Lipp J, Rudolph A. 1972. Sympathetic nerve development in the rat and guinea-pig heart. Biol Neonate 21:76-82.Litchfield JB. 1958. Blood pressure in infant rats. Physiol Zool 31:1-6.Looker T, Berry CL. 1972. The growth and development of the rat aorta. II. Changes in nucleic acid andscleroprotein content. J Anat 113:17-34.Lohse CL, Suter PF. 1977. Functional closure of the ductus venosus during early postnatal life in the dog.Am J Vet Res 38:839-844.Mackenzie E, Standen NB. 1980. The postnatal development of adrenoreceptor responses in isolated papillarymuscles from rat. Pfluegers Arch 383:185-187.Mathew R, Thilenius OG, Arcilla RA. 1976. Comparative response of right and left ventricles to volumeoverload. Am J Cardiol 38:209-17.Mattfeldt T, Mall G. 1987. Growth of capillaries and myocardial cells in the normal rat heart. J Mol CellCardiol 19:1237-46.Metz LD, Seidler FJ, McCook EC, Slotkin TA. 1996. Cardiac -adrenergic receptor expression is regulated bythyroid hormone during a critical developmental period. J Mol Cell Cardiol. 28:1033-1044.Meyer WW, Lind J. 1966. The ductus venosus and the mechanism of its closure. Arch Dis Child 41:597-605.Momma K, Ito T, Ando M. 1992a. In situ morphology of the foramen ovale in the fetal and neonatal rat.Pediatr Res 32:669-672.Momma K, Ito T, Ando M. 1992b. In situ morphology of the ductus venosus and related vessels in the fetaland neonatal rat. Pediatr Res 32:386-389.Moss AJ, Emmanouilides GC, Adams FH, et al. 1964. Response of ductus arteriosus and pulmonary andsystemic arterial pressure to changes in oxygen environment in newborn infants. Pediatrics 33:937-944.Mrowka R, Patzak A, Schubert E, Persson PB. 1996. Linear and non-linear properties of heart rate in postnatalmaturation. Cardio Res 31:447-454.Nakamura S, Asai J, Hama K. 1986. The transverse tubular system of rat myocardium: Its morphology andmorphometry in the developing and adult animal. Anat Embryol 173:307-315.Namin EP, Miller RA. 1966. The normal electrocardiogram and vectorcardiogram in children. In: CassetlsDE and Ziegler RF (Eds) Electrocardiography in Infants and Children. Grune & Stratton, Orlando,Florida.Nyquist-Battie C,Cochran PK, Sands SA, Chronwall BM. 1994. Development of neuropeptide Y and tyrosinehydroxylase immunoreactive innervation in postnatal rat heart. Peptides 15:1461-9.Oberhansli I, Brandon G, Friedli B. 1981. Echocardiographic growth patterns of intracardiac dimensions anddetermination of function indices during the first year of life. Helv Paedatr Acta 36:325-340.Oh W, Oh MA, Lind J. 1966. Renal function and blood volume in newborn infants related to placentaltransfusion. Acta Paediat 56:197-210.Olivetti G, Anversa P, Loud AD. 1980a. Morphometric study of early postnatal development in the left andright ventricular myocardium of the rat. II. Tissue composition, capillary growth, and sarcoplasmicalterations. Circ Res 46:503-512.Olivetti G, Anversa P, Melissari M, Loud AV. 1980b. Morphometry of medial hypertrophy in the rat thoracicaorta. Lab Invest 42:559-565.O’Rahilly R, Müller F. 1996. Human Embryology and Teratology. 2nd Edition, Wiley-Liss, New York, pp.159-206.Penaloza D, Arias-Stella J, Sime F, et al. 1964. The heart and pulmonary circulation in children at highaltitudes. Pediatrics 34:568-582.Penefsky ZJ. 1985. Regulation of contractility in developing heart. In: Legato MJ (Ed) The Developing Heart.Martinus Nijhoff Publishing, Boston, p. 113.Pickoff AS, Stolfi A. 1996. Postnatal maturation of autonomic modulation of heart rate. Assessments ofparasympathetic and sympathetic efferent function in the developing canine heart. J Electrocardiol29:215-222.© 2006 by Taylor & Francis Group, LLC

1078 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPrewitt RL, Dowell RF. 1979. Structural vascular adaptations during the developmental stages of hypertensionin the spontaneously hypertensive rat. Bibl Anal 18:169-173.Quigley KS, Shair HN, Myers MM. 1996. Parasympathetic control of heart during early postnatal developmentin the rat. J Auto Nervous Sys 59:75-82.Rabe D. 1973. Kalibermessungen an den Herzkranzarterien und dem Sinus coronarius. Basic Res Cardiol68:356-379.Rakusan, K. 1980. Postnatal development of the heart. In: Hearts and Heart-Like Organs, Vol. 1, AcademicPress, New York, pp. 301-348.Rakusan, K. 1984. Cardiac growth, maturation, and aging. In: Zak R (Ed) Growth of the Heart in Health andDisease, Raven Press, New York, pp. 131-164.Rakusan K, Cicutti N, Flanagan MF. 1994. Changes in the microvascular network during cardiac growth,development and aging. Cell Mol Biol Res 40:117-122.Rakusan K, Du Mesnil De Rochemont W, Braasch W, Tschopp H, Bing RJ. 1967. Capacity of the terminalvascular bed during normal growth, in cardiomegaly and in cardiac atrophy. Circ Res 21:209-214.Rakusan K, Flanagan MF, Geva T, et al. 1992. Morphometry of human coronary capillaries during normalgrowth and the effect of age in left ventricular pressure-overload hypertrophy. Circulation 86:38-46.Rakusan K, Korecky B, Roth Z, Poupa O. 1963. Development of the ventricular weight of the rat heart withspecial reference to the early phases of postnatal ontogenesis. Physiol Bohemoslov 12:518-525.Rakusan K, Marcinek H. 1973. Postnatal development of the cardiac output distribution in rat. Biol Neonate22:58-63.Rakusan K, Radl J, Poupa O. 1965. The distribution and content of myoglobins in the heart of rat duringpostnatal development. Physiol Bohem 14:317-319.Rakusan K, Sarkar K, Turek Z, Wicker P. 1990. Mast cells in the rat heart during normal growth and in cardiachypertrophy. Circ Res 66:511-516.Rautaharju PM, Davignon A, Soumis F, Boiselle E, Choquette A. 1979. Evolution of QRS-T relationship frombirth to adolescence in Frank-lead orthogonal electrocardiograms of 1492 normal children. Circ60:196-204.Robinson RB. 1996. Autonomic receptor--effector coupling during post-natal development. Cardiovasc Res31:E68-E76.Satoh K, Shirota F, Tsunajima T, Beinlich CJ, Morgan HE, Ichihara K. 2001. Limitation of heart growth inneonatal piglets by simvastatin and atorvastatin: Comparison with pravastatin. Am J Physiol HeartCirc Physiol 280:H2746-H2751.Schiebler TH, Heene R. 1986. Histochemie 14: 28-334 (as cited in Rakusan, 1980).Schneider R, Pfitzer P. 1973. Die Zahl der Kerne in isolierten Zellen des menschlichen Myokards. VirchowsArch [Pathol Anat] 12:238-258.Schonfeld P, Schild L, Bohnensack R. 1996. Expression of the ADP/ATP carrier and expansion of themitochondrial (ATP + ADP) pool contribute to postnatal maturation of the rat heart. Eur J Biochem241:895-900.Scott TM, Pang SC. 1983. The correlation between the development of sympathetic innervation and thedevelopment of medial hypertrophy in jejunal arteries in normotensive and spontaneously hypertensiverats. J Auton Nerv Syst 8: 25-32.Sechi LA, Sechi G, De Carli S, Griffin CA, Schambelan M, Bartoli E. 1993. Angiotensin receptors in the ratmyocardium during pre- and postnatal development. [Studio dei recettori dell’angiotensina nel miocardiodi ratto durante lo sviluppo pre- e post-natale.] Cardiologia 38:471-476.Segar JL, Hajduczok G, Smith BA, et al. 1992. Ontogeny of baroreflex control of renal sympathetic nerveactivity and heart rate. Am J Physiol 263:H1819-H1826.Seguchi M, Harding JA, Jarmakani JM. 1986. Developmental change in the function of sarcoplasmic reticulum.J Mol Cell Cardiol 18:189-195.Shigenobu K, Tanaka H, Kasuya Y. 1988. Changes in sensitivity of rat heart to norepinephrine and isoproterenolduring pre- and postnatal development and its relation to sympathetic innervation. Dev PharmacolTher 11:226-236.Smith HL. 1928. The relation of the weight of the heart to the weight of the body and of the weight of theheart to age. Am Heart J 4:79-93.Smith TL, Hutchins PM. 1979. Central hemodynamics in the developing stage of spontaneous hypertensivein the unanesthetized rat. Hypertension 1:508-517.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1079Smolich JJ. 1995. Ultrastructural and functional features of the developing mammalian heart: a brief overview.Reprod Fertil Dev 7:451-61.Sole MJ, Lo CM, Laird CW, et al. 1975. Norepinephrine turnover in the heart and spleen of the cardiomyopathicSyrian hamster. Circ Res 37:855-862.Solomon S, Wise P, Ratner A. 1976. Proc Soc Exp Biol Med 153:359-362 (as cited in Rakusan, 1980).Steiss JO, Rascher W. 1996. Automated blood pressure measurement in children. Z Kardiol 85 (Suppl 3):81-84.Tada T, Kishimoto H. 1990. Ultrastructural and histological studies on closure of the mouse ductus arteriosus.Acta Anat (Basel) 139:326-34.Tiemann K, Weyer D, Djoufack PC, et al. 2003. Increasing myocardial contraction and blood pressure inC57BL/6 mice during early postnatal development. Am J Physiol Heart Circ Physiol 284:H464-H474.Tomanek RJ, Haung L, Suvarna PR, O’Brien LC, Ratajska A, Sandra A. 1996. Coronary vascularization duringdevelopment in the rat and its relationship to basic fibroblast growth factor. Cardiovasc Res 31:E116-E126.Tomanek RJ, Sandra A, Zheng W, Brock T, Bjercke RJ, Holifield JS. 2001. Vascular endothelial growth factorand basic fibroblast growth factor differentially modulate early postnatal coronary angiogenesis. CircRes 88:1135-41.Truex RC, Scott JC, Ong DM, Smythe MQ. 1955. Anat Rec 123:201-225 (as cited in Rakusan, 1980).Tucker DC. 1985. Components of functional sympathetic control of heart rate in neonatal rats. Am J Physiol248:R601-R610.Ursell PC, Ren CL, Danilo P. 1990. Anatomic distribution of autonomic neural tissue in the developing dogheart: I. Sympathetic innervation. Anat Rec 226:71-80.Uiterwaal CS, Anthony S, Launer LJ, et al. 1997. Birth weight, growth, and blood pressure: An annual followupstudy of children aged 5 through 21 years. Hypertension 30:267-271.Vatner SF, Manders WT. 1979. Depressed responsiveness of the carotid sinus reflex in conscious newbornanimals. Am J Physiol 237:H40-H43.Vizek M, Albretch I. 1973. Development of cardiac output in male rats. Physiol Bohemoslov 22:573-580.Vliegen HW, van der Laarse S, Huysman JAN, et al. 1987. Morphometric quantification of myocyte dimensionsvalidated in normal growing rat hearts and applied to hypertrophic human hearts. Cardiovasc Res21:352-357.Vlk J, Vincenzi FF. 1977. Functional autonomic innervation of mammalian cardiac pacemaker during theperinatal period. Biol Neonate 31:19-26.Vogelberg K. 1957. Die Lichtungsweite der Koronarostien an normalen und hypertrophen Herzen. Z Kreislaufforsch46:101-115.Walker SP, Gaskin P, Powell CA, et al. 2001. The effects of birth weight and postnatal linear growth retardationon blood pressure at age 11-12 years. J Epidemiol Community Health 55:394-398.Walsh SZ, Myer WW, Lind J. 1974. The Human Fetal and Neonatal Circulation. Charles C. Thomas,Springfield, Illinois.Wang L, Swirp S, Duff H. 2000. Age-dependent response of the electrocardiogram to K+ channel blockersin mice. Am J Physiol Cell Physiol 278:C73-C80.Webster SH, Liljegren EJ. 1949. Am J Anat 85:199-230 (as cited in Rakusan, 1980).Winckler J. 1969. Z Zellforsch Mikosk Anat 98:106-121 (as cited in Rakusan, 1980).Widdowson EM, Dickerson JWT. 1960. J Mol Cell Cardiol 9:859-866 (as cited in Rakusan, 1980).Wiesmann F, Ruff J, Hiller KH, Rommel E, Haase A, Neubauer S. 2000. Developmental changes of cardiacfunction and mass assessed with MRI in neonatal, juvenile, and adult mice. Am J Physiol Heart CircPhysiol 278:H652-H657.Yamori Y, Ohtaka U, Nara Y. 1976. Vectorcardiographic study on left ventricular hypertrophy in spontaneouslyhypertensive rats. Jpn Circ J 40:315.Zink J, Van Petten GR. 1980. Time course of closure of the ductus venosus in the newborn lamb. Pediatr Res14:1-3.© 2006 by Taylor & Francis Group, LLC

1080 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAPPENDIX C-7*SPECIES COMPARISON OF ANATOMICAL ANDFUNCTIONAL IMMUNE SYSTEM DEVELOPMENTMichael P. Holsapple 1,4 , Lori J. West 2 and Kenneth S. Landreth 31ILSI Health and Environmental Sciences Institute (HESI), Washington, DC;2The Hospital for Sick Children, University of Toronto, Toronto, ON, Canada;3Department of Microbiology, Immunology and Cell Biology, West Virginia University, Morgantown, WV.4Address All Correspondence to: Michael P. Holsapple, Ph.D., Executive Director, ILSI Health and EnvironmentalSciences Institute, One Thomas Circle, NW, Ninth Floor, Washington, DC 20005-5802, PH: 202-659-3306, FAX:202-659-3617. Email: mholsapple@ilsi.orgIntroductionIn recent years, there has been increasing regulatory pressure to protect children’s health becauseit is suspected that immature populations may be at greater risk for chemically induced toxicity.Congress enacted two statutes in 1996, the FOOD QUALITY PROTECTION ACT (FQPA) andthe SAFE DRINKING WATER ACT (SDWA). FQPA stated that “when establishing, . . . a tolerance. . . of a pesticide residue on food, EPA must perform s separate assessment for infants and children. . .”; and the SDWA “requires that EPA conduct studies to identify subpopulations, such as infants,children, pregnant women . . . that may be more susceptible than the general population . . .”. In1997, presidential Executive Order #13045 was issued which emphasized the following, “eachFederal Agency shall make it a high priority to identify and assess environmental health risks andsafety risks that may disproportionately affect children and ensure that its policies, programs,activities and standards address disproportionate risk to children that result from environmentalhealth risks or safety risks . . . ”.Interestingly, a review of the available literature in virtually any area of health-related toxicologyindicated that an overwhelming proportion of previous research and testing had been directed towardexposure of adults as opposed to children (Dietert et al., 2000). As the number of studies in younganimals has begun to increase to address this regulatory pressure, it is already clear that there aremany challenges to the design, conduct and interpretation of these new data.Components of the immune system have not traditionally been emphasized as potential targetorgans in standard developmental and reproductive toxicity (DART) protocols. Moreover, althoughimmunotoxicology has evolved as a science to the point where several regulatory agencies havecrafted guidelines to address the immunotoxic potential of both drugs and non-drug chemicals,these protocols are generally performed in young adult animals, principally either rats or mice.Developmental immunotoxicology is predicated around the possibility that the immune system mayexhibit unique susceptibility during its ontological development that may not be seen if data is onlyacquired in adults. It is important to emphasize at the onset, that there are currently no validatedor widely accepted methods for evaluating the effects of a drug or chemical on the developingimmune system. Nonetheless, because of concerns over children’s health issues, specifically thepossibility that the very young are uniquely susceptible to chemical perturbation, governmentalregulators are beginning to ask for information about potential effects on the developing immunesystem.* Source: Holsapple, M. P., West, L. J., and Landreth, K. S., Species comparison of anatomical and functional immunesystem development, Birth Defects Research, Part B: Developmental and Reproductive Toxicology, 68, 321-334, 2003.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1081Conception Birth WeaningDays621–28RATSmall No’sB- & T-cellsDemarcation ofSplenic ArchitectureGerminalCentersConceptionBirthWeeks14–18 26 6–8Weaning104Small No’sB- & T-cellsDemarcation ofSplenic ArchitectureGerminalCentersHUMANFigure 1A number of workshops were recently organized to examine scientific questions that underliedevelopmental immunotoxicity tests, and the interpretation of results as they relate to human riskassessment. One of the most important questions considered in the workshops was how to determinethe most appropriate species and strains to model the developing human immune system. Aworkshop organized by ILSI HESI in 2001 considered the following animal models in a series ofplenary presentations: mice (Landreth, 2002; Herzyk et al., 2002 and Holladay and Blalock, 2002),rats (Smialowicz, 2002; Chapin, 2002), pigs (Rothkotter et al., 2002), dogs (Felsburg, 2002) andnonhuman primates (Hendrickx et al., 2002; Neubert et al., 2002), in addition to humans (West,2002; Neubert et al., 2002). A subsequent panel discussion, with input from all participants, offereda number of important conclusions (Holsapple, 2002a). Although rabbits were emphasized as apreferred animal model for developmental toxicity studies, they have been rarely used in immunotoxicitystudies. Although pigs and mini-pigs may offer advantages for mechanistic questions indevelopmental immunotoxicology, the consensus of the participants was that these species werenot appropriate for screening. Although there are known differences between the development ofthe immune system in mice/rats and humans, rodents were judged to be the most appropriate modelfor screening for developmental immunotoxicology. The importance of the differences betweenrats and humans are highlighted in Figure 1, which illustrates three developmental landmarks. First,while small numbers of B- and T-lymphocytes can be detected in the spleen at the beginning ofthe second trimester of pregnancy in humans, these cells are only detectable in rats at birth. Second,the demarcation of the spleen into recognizable red and white pulp areas also occurs in utero forhumans, but not until after parturition in rats. Finally, while germinal centers can be detected inhumans very early during postnatal development, this landmark does not occur in rats until afterweaning. These differences indicate that the maturation of the immune system in rats is delayedrelative to humans and will be discussed in greater detail below.The objective of this review is to compare the anatomical and functional development of theimmune system in a number of species important to either preclinical studies for drug developmentand /or safety assessments for chemicals, with what is known in humans. Our current understanding© 2006 by Taylor & Francis Group, LLC

1082 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof the development of the immune system in rodents has come almost exclusively from mechanisticexperiments in mice. We know far less about the development of immunocompetent cells in rats.However, the general outline and timing of immune cell development in rats appear to closelyparallel studies in mice where data are available. These points will be highlighted below. There islittle doubt that we know the most about the immune systems of rodents and humans; but thisreview will also include what is known about the developing immune systems in dogs and subhumanprimates. Dogs play an important role in the investigation of new drugs since they are one of themajor species used in preclinical studies, including the effects of investigational drugs on theimmune system (Felsberg, 2002). In the past, the major limitation of the use of dogs as anexperimental model in immunologic research has been the paucity of reagents available to dissectthe canine immune system. While still limited when compared to reagents for the murine immunesystem, great strides have been made in recent years to develop reagents to study dogs. Nonhumanprimates have played an increasingly important role as a test species in preclinical testing due tothe phylogenetic and physiologic similarities to man, especially in assessing the effects of immunomodulatoryagents (Hendrickx et al., 2000).Background — Immune System DevelopmentThe establishment of a functional immune system in all mammals, including humans, requires asequential series of carefully timed and coordinated developmental events, beginning early inembryonic/fetal life and continuing through the early postnatal period. The immune system developsfrom a population of pluripotential hematopoietic stem cells that are generated early in gestationfrom uncommitted mesenchymal stem cells in the intraembryonic splanchnopleure surrounding theheart. This early population of hematopoietic stem cells gives rise to all circulating blood celllineages, including cells of the immune system, via migration through an orderly series of tissues,and a dynamic process that involves continual differentiation of lineage restricted stem cells.Establishment of these populations of lymphoid-hematopoietic progenitor cells involves the migrationof these cells from intraembryonic mesenchyme to fetal liver and fetal spleen, and ultimately,the relocation of these cells in late gestation to bone marrow and thymus. The latter two organsare the primary sites of lymphopoiesis and appear to be unique in providing the microenvironmentalfactors necessary for the development of functionally immunocompetent cells. The lineagerestrictedstem cells expand to form a pool of highly proliferative progenitor cells that are capableof a continual renewal of short-lived functional immunocompetent cells, and that ultimately providethe necessary cellular capacity for effective immune responsiveness and the necessary breadth ofthe immune repertoire (Good, 1995). It is important to realize that immune system developmentdoes not cease at birth, and that immunocompetent cells continue to be produced from proliferatingprogenitor cells in the bone marrow and thymus. Mature immunocompetent cells leave theseprimary immune organs and migrate via the blood to the secondary immune organs - spleen, lymphnodes, and mucosal lymphoid tissues. The onset of functional competence depends on the specificparameter being measured and is also species-specific. Senescence of the immune responses is notwell understood, but it is clear that both innate and acquired immune responses to antigens aredifferent in the last quartile of life. This failure of the immune response is due, in part, to a continualreduction in the production of newly formed cells, and to the decreased survival of long-lived cellsin lymphoid tissues.The concept that any of a number of dynamic changes associated with the developing immunesystem may provide periods of unique susceptibility to chemical perturbation has been previouslyreviewed (Barnett, 1996; Dietert et al., 2000; Holladay and Smialowicz, 2000). In particular, anunderstanding of these developmental landmarks has prompted some to speculate about the existenceof five critical windows of vulnerability in the development of the immune system (Dietertet al., 2000). The first ‘window’ encompasses a period of hematopoietic stem cell formation from© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1083undifferentiated mesenchymal cells. Exposure of the embryo to toxic chemicals during this periodcould result in failures of stem cell formation, abnormalities in production of all hematopoieticlineages, and immune failure. The second ‘window’ is characterized by migration of hematopoieticcells to the fetal liver and thymus, differentiation of lineage-restricted stem cells, and expansion ofprogenitor cells for each leukocyte lineage. This developmental window is likely to be particularlysensitive to agents that interrupt cell migration, adhesion, and proliferation. The critical developmentalevents during the third ‘window’ are the establishment of bone marrow as the primaryhematopoietic site and the establishment of the bone marrow and the thymus as the primarylymphopoietic sites for B-cells and T-cells, respectively. The fourth ‘window’ addresses the criticalperiods of immune system functional development, including the initial period of perinatal immunodeficiency,and the maturation of the immune system to adult levels of competence. The final‘window’, addresses the subsequent period during which mature immune responses are manifest,and functional pools of protective memory cells are established.Emergence of Hematopoietic Stem CellsHematopoietic stem cells (HSC) are a population of multipotential stem cells that retain the capacityto self-renew and which have the capacity to differentiate to form all subclasses of leukocytes thatparticipate in immune responses, as well as megakaryocytic and erythrocytic cells (Weissman,2000).HumansImmune system development in the human fetus generally begins with HSC formation in the yolksac, which first appear to migrate at approximately 5 weeks, and is followed by seeding of lymphoidand myeloid lineage progenitor cells (Haynes et al., 1988; Migliaccio et al., 1986).MiceHSC first appear developmentally in intraembryonic splanchnopleuric mesenchyme surroundingthe heart (or the aorto-gonadomesonephros, AGM) at approximately 8 days of gestation in mice(Cumano and Godin, 2001). These cells are found at essentially the same developmental stage inthe extraembryonic blood islands of the yolk sac. Embryonic circulation is established by gestationalday 8.5 in the mouse and it remains unclear to what extent there is exchange of cells fromintraembryonic hematopoietic tissues to extraembryonic sites in any rodent species. However, recentevidence clearly demonstrates that the population of intraembryonic stem cells, but not those whichappear in the yolk sac, contribute to sustained intraembryonic blood cell development and theemergence of the immune system in postnatal rodents (Cumano et al., 2001).RatsNo information available.DogsNo information available.PrimatesNo information available.© 2006 by Taylor & Francis Group, LLC

1084 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFetal Liver HematopoiesisThe fetal liver serves as the initial hematopoietic tissue in early gestation (in some species it servesas the primary hematopoietic tissue throughout gestation) and is characterized by the emergenceand rapid expansion of lineage restricted progenitor cells for all types of leukocytes.HumansLymphocyte progenitors appear in human fetal liver at approximately the 7-8th week followingconception (Migliaccio et al., 1986). Thus, an early period of susceptibility occurs during cellmigration and early lymphohematopoiesis during which stem cells and lineage-committed progenitorsare at risk (~7-10 weeks post-conception).MiceHSC migrate to the developing fetal liver by gestation day 10 in mice (Cumano and Godin, 2001and Cumano et al., 2001). Differentiation of HSC to form lineage-restricted subpopulations of stemcells for the lymphoid and myeloid cell lineages has not been demonstrated prior to gestationalday 10 in mice (Godin et al., 1999). Interestingly, the period of rapid hematopoietic progenitor cellexpansion in fetal liver is not accompanied by wholesale differentiation to mature immunocompetentcell phenotypes (Godin et al., 1999). In fact, mature lymphocytes are not found in the developingliver until day 18 of gestation in the mouse, and after the initiation of hematopoiesis in embryonicbone marrow (Kincade, 1981).RatsNo information available.DogsNo information available.PrimatesNo information available.Development of the SpleenThe role of the spleen in the immune system varies both across the timeline of development andaccording to a distinct species-specificity. The development of other relevant immune organs (e.g.,lymph nodes, Peyer’s patches, etc.) will be discussed as appropriate. The development of the thymusis discussed below.HumansSites of lymphopoiesis begin to develop in the human fetus over the last half of the first trimester,starting with thymic stroma, which forms during the 6th week post-conception (Haynes et al.,1988). Stem cells and T-cell progenitors begin migration from fetal liver to thymus during the 9thweek, while B-cell progenitors appear in blood by about week 12 (Loke, 1978; vonGaudecker,1991; Kendall, 1991; Royo et al., 1987). Gut-associated lymphoid tissue develops from week 8onward, beginning with the lamina propria during weeks 8-10, followed by Peyer’s patches and© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1085the appendix from weeks 11-15. Lymph nodes begin development from 8-12 weeks post-conception,followed by tonsils and spleen from 10-14 weeks. Smith (1968) reported a similar sequence in thatlymphocytes appear first in the fetal thymus, then in peripheral blood shortly after their appearancein the thymus and then in the fetal spleen. Although the spleen never completely loses hematopoieticfunction in the mouse, as described below, the human spleen has largely ceased hematopoiesis bythe time of birth (Tavassoli, 1995).MiceHSC and lineage restricted hematopoietic progentior cells are found in fetal spleen at approximatelygestational day 13 (Landreth, 2002). The spleen continues to contain limited reserves of hematopoieticcells throughout gestation and into postnatal life, and to support myeloid and erythroid celldevelopment, particularly if the bone marrow is damaged. However, lymphopoiesis does not occurin the spleen of postnatal mice under any experimental conditions tested (Paige et al., 1981),suggesting that the bone marrow hematopoietic microenvironment is unique and required forlymphocyte production in mice.RatsAs depicted in Figure 1, only small numbers of B- and T-lymphocytes are found in the spleens ofnewborn rats (Marshall-Clarke et al., 2000). The demarcation of the spleen into recognizable whiteand red pulp areas occurs at around postnatal day 6. B-cell follicles are not seen until about twoweeks of age and the ability to form germinal centers does not develop until weeks three to four(Figure 1; Dijkstra and Dopp, 1983). For comparative purposes in Figure 1, although T- and B-cells can be found as early as week 14-18 of gestation and white pulp areas of the spleen are clearlydemarcated by week 26, germinal centers are not observed until several weeks after birth (Namikawaet al., 1986). In addition, as discussed further below, the immune system of the neonate is immatureand the phenotype of the cells that are found in the marginal zone of the spleen remains immatureuntil around two years of age (Timens et al., 1989).DogsBetween days 27 and 28 of gestation, the primordia of the spleen is evident (Kelly, 1963; Felsberg,1998). Lymphocytic infiltration of the spleen and lymph nodes with evidence of T-cell dependentzones is evident between days 45 to 52 (Bryant and Shifrine, 1972; Felsberg, 1998). Peyer’s patchesare present in the small intestine at about the same time, gestation days 45 to 55 (HogenEsch etal., 1987). Germinal centers and plasma cells appear in the spleen and lymph nodes shortly afterbirth (Yang and Gawlack, 1989; Felsberg, 1998).PrimatesSubstantial lymphocyte differentiation occurs in fetal macaques by gestational day 65 and isaccompanied by increasing lymphoid tissue organization (e.g., demarcation) into specific T-celland B-cell areas that are apparent in peripheral lymphoid organs by gestational day 80 (Hendrickxet al, 2002). In the fetal spleen at gestational day 75, a large proportion of the lymphocytes wereCD20 + B-cells and there were low numbers of T-cells. By gestational day 145, the ratio of whitepulp to red pulp was 1:1, similar to that seen in the mature primate spleen. At gestational day 80,large numbers of B- and T-cells were scattered throughout the parenchyma of lymph nodes, withearly evidence of organization into the characteristic compartments. In the last stage examined (gd145), both the follicle areas containing B-cells and the paracortex containing T-cells had expanded.In the small intestine at gestational day 80, both CD20 + B-cells and CD3 + T-cells were frequently© 2006 by Taylor & Francis Group, LLC

1086 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONencountered in the lamina propia. In the latter stages of fetal development (between gd 100 andgd 145), lymphoid aggregates were well defined within the lamina propia with B-cells orientedtoward the luminal side and T-cells common on the muscularis side.Development of the ThymusUnder the influence of thymic epithelial cells, developing thymic lymphocytes initiate expressionof the T-cell receptor (TcR), undergo a series of selection events which remove autoreactive cells,and ultimately migrate out of that tissue expressing the TcR and a set of linage specific membraneglycoproteins that define phenotype and predict function of these cells (e.g., the so-called ‘clusterdesignation’ or CD markers). Interference during these important stages of T-cell development canalter the evolution of ‘self’ vs. ‘non-self’ recognition, leading either to autoimmunity or to immunodeficienciesresulting in increased susceptibility to infections (Jenkins et al., 1988). Thymic cellproduction wanes rapidly following sexual maturity in all vertebrates.HumansThe seeding of the immune microenvironment during human fetal thymus development has beenreviewed (Lobach and Haynes, 1987; Haynes and Hale, 1998). As noted above, the thymic stromaforms in the human fetus during the 6th week post-conception (Haynes et al., 1988), and T-cellprogenitors begin migration from fetal liver to thymus during the 9th week (Kay et al., 1962; Royoet al., 1987; vonGaudecker, 1991; Kendall, 1991). T-lymphocyte development proceeds in thethymus, which divides into cortex and medulla at 10-12 weeks post-conception (Loke, 1978). Thethymic medulla is fully formed by 15-16 weeks, after which there is an orderly progression of T-cell development beginning in the cortex as thymocytes proceed from the cortex towards the medulla(Royo et al., 1987; vonGaudecker, 1991; Kendall, 1991). Gene re-arrangement in developingthymocytes begins at approximately week 11, leading first to expression of the γδ-TcR, then theαβ-TcR (Royo et al., 1987; Penit and Vasseur, 1989; Kendall, 1991). Sequential expression of coreceptorsfollows: CD3, CD4 and CD8 (corresponding to expression of major histocompatibilitycomplex class I and II antigens by thymic epithelial and dendritic cells) (Royo et al., 1987; Penitand Vasseur, 1989; vonGaudecker, 1991; Kendall, 1991). Lobach and Haynes (1987) reported thathuman fetal thymocytes express the T-cell markers, CD3, CD4, CD5 and CD8, at 10 weeks (Lobachand Haynes, 1987). The complex process of ‘thymic education’ continues to progress throughoutthe mid-trimester of gestation, with positive and then negative selection of ‘double positive’ T-lymphocytes co-expressing the CD4 and CD8 surface molecules (Kay et al., 1970; Loke, 1978;Gale, 1987). This selection process culminates eventually in an enormous reduction in total lymphocytepopulations emerging from the thymus. Export of ‘single positive’ CD4 + and CD8 + T-lymphocytes begins after week 13 (Berry et al., 1992). Single positive T-cells are detectable inspleen by about week 14 and in cord blood by week 20 (Peakman et al., 1992). Thereafter followsan intense expansion of T-cell numbers occurring during weeks 14-26 (Kay et al., 1970; Loke,1978; Royo et al., 1987; Gale, 1987; Berry et al., 1992; Peakman et al., 1992). Exposures duringthis period that alter the emerging T-cell repertoire, such that T-cells are more promiscuous in termsof antigen recognition, may result in wider ‘holes’ in the available T-cell response to certain antigensencountered later in life.MiceIn mice, the thymus anlage develops from the 3 rd and 4 th pharyngeal pouches on gestational day10 and is colonized by immigrating HSC by gestational day 11 (Shortman et al., 1998). PluripotentHSC continue to be detectable in the thymus throughout gestation, however, the majority ofimmigrating cells differentiate within the thymic microenvironment to form immature proliferating© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1087lymphoid thymocytes which express the TcR and the TcR-specific cell surface proteins, CD3, CD4and CD8. Thymic cell production wanes rapidly following sexual maturity resulting in the virtualabsence of thymocyte production at one year of life in mice (Shortman et al., 1998).RatsLess is known about thymus development in the rat. The thymus anlage in rats is invaded byhematopoietic stem cells by day 13 of gestation, and these cells have differentiated to express bothCD4 and CD8 (e.g., double positive cells) by day 18 (Aspinall et al., 1991). Mature single positivethymocytes are not detected until day 21 in the rat.DogsThe primordia of the thymus is evident in dogs between days 27 and 28, and it descends from thecervical region into the anterior thoracic cavity on day 35 (Kelly, 1963; Felsberg, 1998). At this time,the thymus is composed of epithelial lobules and mesenchymal stroma only. Between days 35-40, thefetal thymus becomes actively lymphopoietic and shows corticomedullary demarcation; by day 45 thethymic microenvironment has assumed its normal postnatal histologic appearance (Miller and Benjamin,1985; Snyder et al., 1993; Felsberg, 1998). Fetal and postnatal thymopoiesis has been recentlyevaluated in dogs and the results indicate that normal thymopoiesis is occurring by day 45 of gestationwith a distribution of thymocyte subsets virtually identical to that seen in the postnatal thymus, althoughwith a markedly reduced total cellularity (Somberg et al., 1994; Felsberg, 1998). The thymus undergoesrapid postnatal growth and reaches maximum size at 1 to 2 months of age, as the percentage of bodyweight, and at 6 months of age in absolute terms (Yang and Gawlak, 1989).PrimatesFetal thymic histogenesis in primates (rhesus) has been demonstrated to be gradual and continuous,as is the case in humans (Tanimura and Tanioka, 1975). The thymic anlage separates from thepharyngeal pouches and engages in early stages of proliferation at gestation days 37 – 48 (Hendrickxet al., 1975). These same studies showed that the differentiation of lymphoid elements occurredwithin the thymus at gestation days 50 – 73; and the maturation of the fetal lymphoid systemoccurred at gestation days 100 – 133. Later studies by the same laboratory using another speciesof primates (cynomolgus) indicated that the cortex and medulla of the thymus were distinguishableby gestational day 65, and most of the thymocytes were CD3 + and localized throughout both regionsof the fetal thymus, while CD20 + B-cells were dispersed in the corticomedullary junction. In olderfetuses (gd 100 and gd 145), CD20 + B-cells increased in the medulla and corticomedullary junction,while the number of CD3 + thymocytes remained similar throughout gestation (Hendrickx et al.,2002). In the same study, flow cytometric analysis of lymphocyte subsets indicated a slight increasein the single positive (CD4 + /CD8 - and CD4 - /CD8 + ) thymocytes and a slight decrease (86% to 76%)in double positive (CD4 + /CD8 + ) thymocytes between gestational days 80 and 145.Bone Marrow HematopoiesisThe final critical stage in the embryonic development of the mammalian immune system is theestablishment of the bone marrow as the primary hematopoietic organ.HumansBone marrow lymphopoiesis begins in the human fetus at approximately week 12 (West, 2002).B-lymphocyte development begins at approximately the same time as T-cells in humans. B-cells© 2006 by Taylor & Francis Group, LLC

1088 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONbearing sIgM and sIgG are first found in the liver at 8 weeks of gestation and in the spleen at about12 weeks. IgD- and IgA-bearing cells are also found at 12 weeks of gestation, and Anderson et al.(1981) reported that adult levels of B-cells bearing sIg of all classes are reached by 14-15 weeks.MiceThe formation of the bone marrow cavity as long bones are mineralized late in gestation, and theimmigration of hematopoietic cells into that tissue site, occurs on gestational day 17.5 in mice(Kincade et al., 1981). The bone marrow rapidly assumes primary hematopoietic function aftergestational day 18 in mice, and this persists throughout postnatal life. The relationship of boneformation and emergence of hematopoietic tissue is particularly interesting, and it is unclear whetherthe migration of hematopoietic cells into this tissue actually initiates the process of bone ossification.The evacuated bone marrow cavity is colonized by HSC and these pluripotential cells expand toestablish the primary hematopoietic reserves of stem cells for postnatal life. In fetal mice, B-cellsexpressing surface immunoglobulin (sIg) appear in the liver, spleen and bone marrow on approximatelygestational day 17 (Verlarde and Cooper, 1984).RatsNo information available.DogsBone marrow in fetal dogs becomes heavily cellular, including abundant hematopoietic stem cellsbetween days 45 and 52 (Bryant and Shifrine, 1972; Felsberg, 1998).PrimatesNo information available.Postnatal Development of the Immune SystemBecause birth occurs at various stages of fetal maturity, the significance of parturition as a landmarkin the development of the immune system can vary from species to species. As such, directcomparison of immune functional development between humans and animals is complicated bydifferences in the maturity of the immune system before and after parturition. This difference hasbeen linked to the length of gestation (Holladay and Smialowicz, 2000) in that animals with shortgestation periods (e.g., mice, rats, rabbits and hamsters) have relatively immature immune systemsat birth compared to humans. However, this speculation is not absolute, because, as noted byFelsburg (2001), dogs are like humans in that their immune systems are essentially fully developedat birth even though their gestation period is markedly shorter than humans.Certainly, the event of birth marks an emergence from intra-uterine maternal influences, includingthe maternal immune ‘suppression’ associated with pregnancy (Oldstone and Tishon, 1977;Barrett et al., 1982; Papdogiannakis et al., 1985; Papadogiannakis and Johnsen, 1988; Loke andKing, 1991; Sargent et al., 1993 ). It is important to keep in mind that maternal influences continueto contribute throughout lactation and other maternal behaviors. After birth, infectious and similarantigenic exposures become more significant, including colonization with microbes through thegastro-intestinal tract and other sites. The importance of the early postnatal exposure to ‘antigens’is clearly demonstrated in studies where there is a delay in lymphoid development in animals raisedfrom birth in a germ-free environment (Thorbecke, 1959). More recent studies by Anderson et al.(1981) indicated that germ-free animals have significantly reduced levels of Ig, due to a 2-5-fold© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1089lower rate of Ig synthesis, with IgA and IgM more affected than IgG, and have approximately 10-fold lower numbers of plasma cells.HumansAt the time of human birth, the proportion of lymphocytes represented by T-cells is lowest andincreases over time (Semenzato et al., 1980; Series et al., 1991; Berry et al., 1992; Erkellar-Yukselet al., 1992; Hannet et al., 1992; Plebani et al., 1993; Hulstaert et al., 1994). Specifically, absolutenumbers of both CD4 + and CD8 + subsets increase, while the CD4/CD8 ratio has been variablyreported to differ, or not, with time after birth (Foa et al., 1984; Hannet et al., 1992; Plebani et al.,1993). Usage of TcR Vβ subsets likewise has been variably reported to differ, or not, betweennewborn and adult αβ-T-cells (Foa et al., 1984; Hayward and Cosyns, 1994). The proportion of Tcells expressing the γδ-TCR is high during fetal life, decreases at term, but remains higher innewborns than later in life (Peakman et al., 1992). Some reports have described circulating ‘doublepositive(CD4 + CD8 + ) immature’ T-cells; however generally positive and negative selection areprobably complete by birth (Foa et al., 1984; Griffiths-Chu et al., 1984; Solinger, 1985; Reason etal., 1990; Hayward and Cosyns, 1994). Nonetheless, T-cells do have phenotypic and functionalfeatures of ‘naïve’ cells. In particular, markers of ‘immature’ or antigenically ‘naïve’ cells such asCD45RO - /RA + on CD4 + cells are much higher than in adults (Clement et al., 1990; Denny et al.,1992; Erkellar et al., 1992; Hannet et al., 1992; Hulstaert et al., 1994; Igegbu et al., 1994; Jenningset al., 1994). Expression of some surface molecules (e.g., integrins, adhesion molecules) is reportedto be low at birth and increases post-natally, while expression of others is high and decreases(Clement et al., 1990; Hannet et al., 1992; Hayward and Cosyns, 1994).MiceFollowing birth, there is an immediate disappearance of hematopoietic cells from the liver as thatorgan assumes postnatal function. In the postnatal mouse, leukocytes are produced in the bonemarrow and, except for T lymphocytes, complete their maturation in that tissue. It is known thatsplenic hematopoiesis persists for several weeks after birth in rodents (Marshall-Clarke et al., 2000).Spear et al. (1973) observed an increase in B-cells and a decrease in the T-cell to B-cell ratio inmice between 2 and 3 weeks of age, which coincided, with the onset of antigen responsiveness intheir studies. As discussed below, similar results were presented in 21-day old rats by Ladics et al.(2000).One of the clearest indications that the immune system continues to develop in mice is the factthat perinatal (

1090 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(Ladics et al., 2000). Subsequent histological analysis of the spleens from 10-day old pups indicatedthat there were no germinal centers present either in untreated rats or in rats immunized with theT-dependent antigen, sheep erythrocytes (SRBC). As discussed below, these same studies indicatedthat no anti-SRBC antibody response could be detected in 10-day old rat pups. Taken together,these results suggested that the immune system of a 10-day rat pup is too immature to elicit aprimary response. The phenotypic analysis of the spleen from a 21-day old rat indicated that thenumber of B-cells was comparable to adults, and that the number of T-cells and T-cell subsets wasreduced relative to an adult (Ladics et al., 2000). As discussed below, a number of studies havedemonstrated that rats at approximately 21 days of age are capable of an immune response, butthat it is always less than that seen in an adult animal.DogsShortly after birth in dogs, germinal centers and plasma cells appear in the spleen and lymph nodes(Yang and Gawlack, 1989; Felsberg, 1998), and the thymus undergoes rapid postnatal growth (Yangand Gawlak, 1989; Somberg et al., 1994). The phenotype of the lymphocyte subpopulations inneonatal dogs differs significantly from that observed in adult dogs (Felsberg, 1998, Somberg etal., 1994; Felsberg, 2002). During the first 16 weeks, there is a gradual decline in the proportionof peripheral B-cells and an increase in the proportion of peripheral T-cells to normal adult levels,which occurs between 2 and 4 months of age. Thereafter, the proportion of B- and T-cells remainsfairly constant throughout the life of the dog. Other age-related differences in lymphocyte subsetsinclude the following: considerably higher proportions of CD4+ T-cells during the first six months,a resultant high CD4:CD8 ratios during the same time period, a decline in the proportion of CD4+at 10-12 months; CD8+ T-cells increasing to normal adult levels during the same time with aresultant normalization of the CD4:CD8 ratios (1.5 – 2.0). Finally, during the neonatal and immediatepostnatal period, greater than 90% of the peripheral T-cells are CD45RA+ (naïve) T-cells(Somberg et al., 1996; Felsberg, 2002). After 4 months of age, the relative frequency of CD45RA+T-cells declines such that only 40-50% of the peripheral T-cells in adults are CD45RA+. Interestingly,very similar age-related changes were observed in healthy children and adults (Denny et al.,1992; Erkellar et al., 1992), an observation consistent with the recognition that the immune systemsof dogs and humans are at a similar stage of development at birth.Although most indicators suggest that the development of the immune systems of dogs andhumans are very comparable, there is one important distinction that does impact on the respectivepostnatal development of the immune systems (Felsberg, 2002). Humans possess a hemochorialplacenta in which the blood of the mother is in direct contact with the trophoblast, thereby permittingdirect entry of maternal IgG into the fetal bloodstream. In contrast, dogs have an endotheliochorialplacenta with four structures separating the maternal and fetal blood – the endothelium of theuterine vessels, the chorion, mesenchyme (connective tissue) and the endothelium of fetal tissues.These four layers of tissue separating the maternal and fetal circulation in the dog limit the in uterotransfer of maternal IgG to the fetus. Ultimately, only 5 to 10% of maternal antibody in the dogis obtained in utero through the placenta with the majority being obtained via the colostrums duringthe first 24 hours after birth (Felsberg, 1998). The levels of serum IgG in newborn puppies thatreceive colostrums approach those levels found in adults. It is important to emphasize that the halflifeof maternal antibody in the dog is approximately 8.4 days, so the average protection frommaternal antibody in the neonate is 8 to 16 weeks. Following the decline in maternal IgG, there isa gradual increase of all three major classes of immunoglobulin, with IgM and IgG reaching normaladult levels by 2-3 months of age, and 6 to 9 months of age, respectively. As in other species, thesynthesis of IgA lags behind the other isotypes and doesn’t reach normal adult levels until approximatelyone year of age.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1091PrimatesA study by Neubert et al. (1996) compared a number of phenotypic markers in peripheral bloodlymphocytes between adults and newborns in both humans and primates (marmosets). Whileperipheral blood was used for adult humans, adult marmosets and newborn marmosets, umbilicalcord blood was used as a surrogate for newborn humans. Their results indicated no significantdifferences between newborns and adults in either humans or primates and only trends will bediscussed.* % lymphocytes: humans: umbilical = adultmarmosets: newborn ≤ adult* CD4 + (helper): humans: umbilical = adultmarmosets: newborn = adult* CD4 + /CD29 + (memory): humans: umbilical < adultmarmosets: newborn < adult* CD4 + /CD45RA (naïve): humans: umbilical > adultmarmosets: newborn >> adult* CD8 + (suppressor): humans: umbilical ≤ adultmarmosets: newborn ≥ adult* CD 56 + (NK cells): humans: umbilical = adultmarmosets: newborn = adult* CD 20+ (B-cells): humans: umbilical = adultarmosets: newborn = adultThe biggest differences between newborns and adults observed in this study were for memoryT-cells (e.g., CD4 + /CD29 + ), which were lower in newborns, and for naïve T-cells (CD4 + /CD45RA),which were higher in newborns. Both of these observations are consistent with, and do not detractfrom the interpretation that the immune systems of humans and primates are devoid of any antigenicexposure, but are effectively mature at birth. Neubert et al. (2002) emphasized that the immunesystem of newborns (e.g., marmosets and humans) is rather ‘immature’; that these deficiencies arenot so much an indication of a lack of important components of the immune system, but rather arethe result of little intrauterine contact with environmental antigens. The biggest difference betweenhumans and marmosets was in the number of white blood cells (e.g., measured either as total WBCsor as % lymphocytes) being considerably higher in the primates.Acquisition of Functional ImmunocompetenceAs noted above, the onset of functional immunocompetence varies across species and is strikinglydifferent between rodents and humans. Exposure to a specific antigen during the perinatal periodresults in a rapidly expanding accumulation of lymphocyte specificities in the pool of memory cellsin secondary lymphoid tissues. As thymic function wanes and thymocytes are no longer producedin that tissue, it is this pool of memory B- and T-cells that maintains immunocompetence for thelife of the individual.HumansThymocytes derived from human fetal tissue of less than 11 weeks gestation show no demonstrableresponse to mitogen stimulation (Kay et al., 1970; Sites et al., 1974; Royo et al., 1987). Functionalcapability of human fetal T-lymphocytes begins to develop between the end of the first trimesterand the end of the second trimester (~14-26 weeks). At 12-14 weeks, fetal thymocytes respond toPHA only, while by 13-14 weeks (after thymic colonization with stem cells), fetal thymocytesrespond to most mitogens (Kay et al., 1970; Sites et al., 1974; Royo et al., 1987). Mitogen-responsive© 2006 by Taylor & Francis Group, LLC

1092 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONT-cells can be demonstrated in spleen and peripheral blood by 16-18 weeks gestation (Kay et al.,1970). Similar results were reported by Mumford et al. (1978), who observed responses to Con Awith thymus cells at 13-14 weeks of gestation, and with spleen cells at approximately 18 weeksof gestation; while August et al. (1971) observed that fetal thymocytes responded to mitogenicstimulation at 12 weeks of gestation and splenic T-cells responded between 14 and 16 weeks ofgestation. Reactivity of fetal lymphocytes in the MLR has been variably reported as beginningduring weeks 10-14 (Ohama and Kaji, 1974; Loke, 1978; Royo et al., 1987), followed by functionaleffector reactivity in cell-mediated lympholysis (CML) assays subsequently, proceeding in variablefashion depending on the lymphoid tissue tested (Granberg and Hirvonen, 1980; Murgita andWigzell, 1981). Fetal thymocytes have been shown to demonstrate a positive MLR by 12 weeksgestation, and fetal splenocytes by about 19 weeks, although to a consistently-low degree until 23weeks (Granberg and Hirvonen, 1980). Lymphocytes derived from cord blood have been demonstratedto generate a positive CML response at 18-22 weeks gestation; however high individualvariability was found (Rayfield et al., 1980). The functionality of T cells from human neonates hasbeen demonstrated to be reasonably well-developed (Hayward, 1983). Thus, neonatal T cells canproliferate in MLR and in response to most mitogens, and show weak proliferation and cytokineproduction with stimulation by endogenous antigen presenting cells or anti-CD3 monoclonalantibody (representative ‘physiologic’ stimuli) (Granberg and Hirvonen, 1980; Rayfield et al., 1980;Loke and King, 1991). Furthermore, with TcR-independent stimulation, neonatal T-cells showequivalent proliferation and cytokine production to adult T cells (Splawski and Lipsky, 1991;Demeure et al., 1994; Tsuji et al., 1994). However, the cytokine profiles of the human neonateshow a deficient and/or delayed production of interleukin (IL)-2, interferon-γ, IL-4 and IL-6, thuspossibly skewed toward a Th2 phenotype (Splawski and Lipsky, 1991; Demeure et al., 1994; Tsujiet al., 1994). In some reports, the cytokine networks appear less efficient, requiring increasedstimulation and/or receptor maturation (Cairo et al., 1991; Tucci et al., 1991). T-cell reactivity canbe demonstrated in CML assays, but at lower levels than adult T cells (Granberg and Hirvonen,1980; Rayfield et al., 1980; Barret et al., 1980; Loke and King, 1991). Demonstrating in vivofunctionality, newborns can reject organ and tissue allografts, however, they remain very sensitiveto clinical immunosuppression (Demeure et al., 1994; Webber, 1996; Pietra and Boucek, 2000).Indeed, it has been demonstrated that a survival advantage of approximately 10-12% is maintainedmore than 10 years post-transplant if heart transplantation is performed within the first 30 days oflife compared to 1-6 months of age, and that this survival advantage is due to decreased immunerelatedevents (Pietra and Boucek, 2000).The development of functional NK cells in the human fetus occurs at 28 weeks of gestation,with full-term newborns displaying peripheral blood NK activity at approximately 60% of adultlevels (Toliven et al., 1981). NK cells are fewer in number at birth than later, and have been reportedto be less active and less responsive to stimulation, with diminished cytotoxicity capacity (Kohl etal., 1984; Rabatic et al., 1990). However, levels of NK cells during early fetal life have beendemonstrated to be significantly higher than during neonatal life, suggesting that certain aspectsof innate immunity may play a more important role in the fetal immune response than adaptiveimmunity (Erkellar et al., 1992; Hulstaert et al., 1994).As noted above, adult levels of B-cells bearing sIg of all classes are reached by 14-15 weeksgestation in humans (Anderson et al., 1981). Circulating B lymphocytes are generally at high levelsin the neonate, with immature markers demonstrable, and these decline with age (Tucci et al., 1991;Erkeller-Yuksle et al., 1992; Hannet et al., 1992; Peakman et al., 1992; Plebani et al., 1993; Hulstaertet al., 1994; Nahmias et al., 1994). The development of mature plasma cells in the bone marrowis incomplete at birth (Loke, 1978; Durandy et al., 1990). Isotype switching is defective, withsimultaneous surface expression of different isotypes, and immunoglobulin production is low (Loke,1978; Gathings et al., 1981; Hayward, 1983; Nahmias et al., 1994). Human neonatal B-cells arefunctionally defective in their capacity to generate antibody-producing cells in vitro, compared to© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1093B-cells from adults, and in general, human B-cells are assumed to be inherently immature at birth.As described above for neonatal T-cells, deficiencies in cytokine networks likely play an importantrole in diminished functionality of the humoral response (Splawsky and Lipsky, 1991; Tsuji et al.,1994). Although a small number of IgM-producing cells are detected, no IgG- or IgA-producingcells can be identified. The ability of human B-cells to produce either IgG or IgA antibodiesincreases with age, with adult levels being reached by 5 and 12 years of age, respectively (Miyawakiet al., 1981). These results prompted speculation that the greater susceptibility of human newbornsto certain bacterial infections is due to deficient B-cell function, especially the delay in productionof IgG and IgA antibodies. In contrast with what is observed in rodent, infants appear to respondwell to T-dependent antigens primarily through adequate production of IgM. However, they respondpoorly or not at all to polysaccharide antigens such as the antigens found on the cell walls of severalinfectious bacteria, which also contributed to increased susceptibility to infections (Garthings etal., 1981). Decreased responsiveness to antigen stimulation, particularly to non-protein ‘T cellindependent’antigens, continues well into the second year of life. Examples of this deficiency arewell-recognized clinically, such as the inability of the newborn to mount an immune response toStreptococcus pneumoniae and Haemophilus influenza - leading to increased susceptibility ofnewborns to infection with these pathogens - and to respond effectively to vaccines (Cadoz, 1998;Ahmad and Chapnick, 1999). As discussed below, it is interesting to note the distinction betweenhumans and mice/rats in terms of the maturation of antibody responses to T-dependent and T-independent antigens. To date, there is no explanation for this dichotomy.After the early neonatal period, there is continued acquisition of immune competence concomitantwith increased antigen exposure throughout years one and two of childhood. It is importantto consider the fact that attainment of immunocompetence does not necessarily mean closure of‘windows’. Lactation and other nutritional modalities make variable contributions depending onthe immune compartment and on continued risks of particular exposures. During late childhoodand adolescence, continuous growth processes play an important role in susceptibility to toxicexposures, as do the influences of hormonal fluctuations and the physiologic changes of adolescence.Furthermore, ongoing changes in social activities and behavioral modifications can increase susceptibilityduring these years. Importantly, there is need for continued tracking during childhoodand adolescence of the effects of exposures occurring earlier in life.MiceMouse thymocytes begin to respond to phytohemagglutinin (PHA), Concanavalin A (Con A) andin a mixed lymphocyte reaction (MLR) by gestation day 17 (Mosier, 1977). Although fetal thymocytesrespond to PHA and in the MLR similar to adults, the response to Con A does not reachadult levels until 2-3 weeks after birth. It is also during the immediate postnatal period that acquiredimmune function is first detectable in mice (Ghia et al., 1998). Functional B- and T-lymphocytesare produced in the bone marrow and thymus, respectively, and migrate to the spleen, lymph nodes,and mucosal associated lymphoid tissues (MALT). However, a mature pattern of immune responseto antigen is not achieved until approximately one month of age in rodents. During the first monthof postnatal life in mice, the immune system remains immature, fails to produce antibodies tocarbohydrate antigens, and is predominated by IgM mediated immune responses to antigen (Landreth,2002). Holladay and Smialowicz (2000) reported that in mice, antibody responses to T-independent antigens occurs soon after birth and reach near-adult levels by 2 or 3 weeks. In contrast,T-cell dependent antibody responses in mice begin after about two weeks and do not reach adultlevels until 6-8 weeks of age. Similarly, natural killer (NK) cell activity is absent in mice at birthand does not begin to appear until about three weeks of age (Santoni et al., 1982).During the first six months of life in rodents, there is enormous production of T- and B-lymphocytes from primary lymphoid tissues (Kincade, 1981).© 2006 by Taylor & Francis Group, LLC

1094 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONRatsAs discussed above, a recent study by Ladics et al. (2000) compared the antibody response toSRBC in 10-day and 21-day old rat pups. There was no antibody response to SRBC in 10-day ratpups. The results with the antibody response in 21-day old weanlings were mixed. The use of anELISA to measure antibody titer was precluded due to a high background. No high backgroundwas seen when the response was measured as the number of antibody-producing B-cells; and themagnitude of the response was at the low end of the historical range of responses seen in adultrats. The latter results were consistent with histological analysis of the spleens in that prominentgerminal centers were observed in 21-day old rat pups immunized with SRBC. Similar results wereseen when another T-dependent antigen, keyhole limpet hemocyanin (KLH) was used in that whilethe immune status of rat weanlings could be measured, the response was less than that seen inadults (Bunn et al., 2001). The observed profile of age-dependent antibody responses in rat pupsand weanlings is also consistent with previous studies by other laboratories. For example, Spearet al. (1973) found that consistent antibody responses to SRBC could not be detected until after 2weeks of age. Kimura et al. (1985) reported a steady age-related increase in the antibody responseto SRBC beginning around postnatal day 12. Interestingly, this same investigation showed a similarage-related kinetics in the antibody response to the T-independent antigens, TNP-Ficoll and TNPdextran,but a markedly accelerated antibody response to another T-independent antigen, TNP-Brucella abortus (e.g., measurable at birth; and nearly at adult levels by postnatal day 8).DogsOur knowledge of the ontogeny of immune responses in dogs is limited (Felsberg, 2002). Fetaldogs are capable of responding to various antigens, including an antibody response to bacteriophage,a T-dependent antigen, on gestational day 40, proliferation to PHA by lymphocytes from fetalspleen and lymph node on gestational day 45, an antibody response to RBC on gestational day 48,proliferation to PHA by fetal thymocytes and antibody response to vaccination with Brucella canison gestational day 50 (Bryant et al., 1973; Shifrine et al., 1971; Klein et al., 1983). Nonetheless,it is generally considered that dogs become immunologically mature close to, or at parturition(Felsberg, 2002). Jacoby et al. (1969) demonstrated that colostrums-deprived, gnotobiotic puppiesdeveloped both primary and secondary specific antibody responses when immunized with bacteriophagewithin the first 24 hours after birth. These results indicated that neonatal dogs possess afunctional B-cell and T-cell system at birth. However, it is important to note that these results didindicate that the humoral immune response matured with age, in that the titers were

POSTNATAL DEVELOPMENTAL MILESTONES 1095Table 1Summary of Immune System Development across Multiple SpeciesParameterMouse Dog Primate HumanGestation 20 days 60–63 d 1 155–165 d 2 40 weeksStem cells appear gd 8 (40) 3 — — gw 5 (12.5)Fetal liver hematopoiesis gd 10 (50) — — gw 7 (17.5)Splenic primordia gd 13 (65) gd 28 (45) —- gw 10 (25)Spleen demarcation 4 — gd 45 (72) gd 80 (50) gw 26 (65)Thymic primordia gd 10 (50) gd 28 (45) gd 35 (22) gw 6 (15)Bone marrow hemat. gd 18 (90) gd 45 (72) — gw 12 (30)Mitogen proliferation 5 gd 17 (85) gd 50 (81) — gw 12 (30)1Assume a gestational period of 62 days for later calculations.2Length of gestation affected by strain of primate; use 160 days for calculation.3Number in parentheses is the % of total gestational length.4Demarcation = organization of spleen into distinct red and white pulp areas.5Proliferation in response to mitogens (PHA, Con A) by fetal thymocytes.SummaryAn attempt has been made here to capture comparisons of the development of the immune systemacross five mammalian species: mouse, rat, dog, primate and human (Table 1). Because it isgenerally accepted that the development of the immune system in mice and rats is comparable andbecause much more information is available for the mouse, only the results for one rodent speciesare depicted in the summary table. Developmental landmarks that are described in greater detailin the text above, are shown in the context of the known length of gestation for each species. Thisdata base is incomplete and it is important to avoid over-interpreting these results. However, theydo allow speculation about this comparison. The results clearly depict that we know the most aboutthe immune systems of the mouse and human. Moreover, across every developmental landmarkdepicted, the mouse is developmentally delayed relative to humans. As emphasized above, thisdelay is ultimately manifested as a dramatic difference in the functional capabilities of the immunesystems of mice and humans at birth. A comparison between rats and humans for some selectdevelopmental landmarks is shown in Figure 1 and indicates a similar trend in that the developmentin this rodent species is delayed. It is generally assumed that the development of the immunesystems of rats and mice are largely comparable. Depicting the comparative development along atime-line illustrates major differences that must be taken into account in developmental immunotoxicologystudies – e.g., some phases of the development of the human immune system that occurin utero, occur after birth in rodents; and some phases of the development of the human immunesystem that occur during the lactational neonatal period, occur after weaning in rodents. It isimportant to emphasize that these differences do not necessarily obviate the selection of rodentsas a test species; but they must be factored into the interpretations of any results, especially in thecontext of what is known about transplacental transfer and/or transfer via the mother’s milk of thedrug or chemical being studied.As emphasized above, the dog is like the human in that it is born with an essentially functionalimmune system. In that regard, the results in Table 1 are quite interesting. With an admittedly verylimited database for the comparison, the results seem to suggest that the in utero development ofthe immune system in the dog is delayed, as compared to humans, and only slightly more advancedthan what is seen with rodents. This observation may become important to the design of a developmentalimmunotoxicology study in the dog where exposure to pregnant bitches would occur atdifferent times during gestation. Another major consideration in such studies relates to the differencebetween human and dog placentas. This distinction was discussed above in the context of thelimited transfer of maternal antibodies to the fetus and the possible impact on the postnataldevelopment of the immune system in dogs. However, the endotheliochorial placenta of the dog© 2006 by Taylor & Francis Group, LLC

1096 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmust be factored into preclinical studies for potential impacts on the transplacental transfer of agiven chemical. As with the distinctions noted for the rat, although these differences should notobviate the selection of the dog as the test species, they must be considered in the interpretationof any results.Unfortunately, there was an insufficient number of studies characterizing the in utero developmentof the immune system in primates to make comparisons to the other species. Defining withmore precision the developmental landmarks in the various test species is essential to maximizingthe interpretation of preclinical studies, and ensuring that the parameters we extrapolate from animalmodels are physiologically and clinically relevant to the developing human. As emphasized above,the sequence of developmental landmarks described here has led some investigators to think interms of ‘windows’ of vulnerability (Holladay and Smialowicz, 2000). In fact, one of the cornerstonesof the evolution of developmental immunotoxicology is that it is assumed that some chemicalsmay have a greater effect during one or more of these ‘windows’ than another. Stated anotherway, the premise is that the developing immune system may be more or less sensitive to toxicinsult than the mature adult immune system. However, this premise should consider the physiologyof the immune system during the various stages of development. As noted above, both the developingimmune system and the mature immune system are characterized by a number of physiologicalprocesses, including cellular proliferation, cell migration, cell-to-cell interactions, expressionof surface markers (e.g., for activation, differentiation, adhesion, etc.). As such, a chemical with amode of action that targets one of these processes (e.g., such as an antiproliferative agent) wouldbe expected to affect all ‘windows’ where that process (e.g., proliferation) played an important role.One feature of the developing immune system that clearly distinguishes it from the matureimmune system, especially during gestation, is the role played by organogenesis. Defects in thedevelopment of the immune system due to heritable changes in the lymphoid elements have providedclinical and experimental examples of the devastating consequences of impaired immune development(Rosen et al., 1995). Therefore, the effects of chemicals on the genesis of critical componentsof the immune system in the developing fetus may be more important than effects on these tissuesafter they have been populated by hematopoietic and lymphoid cells. However, a chemical thatinduces the formation of this kind of developmental abnormality in the fetus would legitimatelybe classified as a ‘teratogen’. Interestingly, studies by Hendrickx et al. (2000) characterized theeffects of several well-known teratogenic pharmaceutical agents on the developing immune systemsin primates. Treatment with triamcinolone acetonide, 13-cis-retinoic acid and experimentallyinducedzinc deficiency all were shown to trigger an adverse effect on fetal thymus in macaquesand baboons. Moreover, one of the most widely studied environment contaminants, 2,3,7,8-tetrachlorodibenzo-p-dioxin(TCDD or ‘dioxin’), that has been characterized by multiple laboratoriesas a developmental immunotoxicant (Barnett, 1996; Holladay and Smialowicz, 2000; Holladay andBlalock, 2002; Neubert et al., 1996; Neubert et al., 2002; and Smialowicz, 2002) is also a knownanimal teratogen (Couture et al., 1990). These results prompt the question as to how many knownteratogens would have an impact on the thymus or other critical immune organs and whether thesetypes of agents should be more appropriately classified as ‘immunoteratogens’, as opposed todevelopmental immunotoxicants.Results using clinical immunosuppressants, which are devoid of teratogenic effects, in pregnantwomen are relevant to this discussion. Recent advances in organ transplant success have resultedin increasing numbers of women becoming pregnant after an organ or tissue transplant procedure.These women require therapeutic immunosuppression throughout pregnancy to prevent allograftrejection. Studies of pregnant women receiving the required polytherapy (e.g., prednisone pluseither azathioprine or cyclosporine) following renal, heart or liver transplants indicate that the majorrisk is fetal growth restriction and that there is little evidence that exposure is associated with anincrease in congenital anomalies (Hou, 1999). In contrast, another study reported that a transientbut severe B-cell depletion was seen in neonates born to renal transplant mothers receiving cyclosporine,azathioprine and methylprednisolone (Takahashi et al., 1994). One report also showed an© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1097association between azathioprine and a slightly increased risk of nonspecific congenital anomaliesas well as prematurity (Penn et al., 1980), and another study described neonatal lymphopenia andthrombocytopenia in children following maternal azathioprine therapy (Davidson, 1994). No malformationshave been noted (Hou, 1999); but there have been reports of reduced birth weight andtransient neonatal immunocompromise (Jain, 1993; Laifer et al., 1994) following the use of tacrolimus,a cyclosporine-like immunosuppressant, during pregnancy. Taken together, the availableepidemiologic reports to date provide little evidence of increased malformations associated withthe use of clinical immunosuppressants during pregnancy, and the major risks appear to be adverseeffects on fetal growth and the fetal immune system following exposure to these agents during thesecond and third trimesters (Hendrickx et al., 2000).This discussion is relevant to subsequent decisions about how to test for adverse effects on thedeveloping immune system. Organs essential to the normal functioning of the immune system, likethe thymus or spleen or bone marrow, are not typically assessed in developmental and reproductivetoxicology (DART) studies, and this failure to assess developmental damage to the immune systemshould be reevaluated. The EPA has suggested that immune organs be weighed in rat pups beingculled from a standard two-generation reproductive toxicity study (OPPTS 870-3800). While thisis a logical first step, there is little rationale to support the predictive value of weighing neonatalimmune organs.In order to address this issue properly, it will be necessary to establish a more appropriate panelof tests to evaluate the effect of chemical toxicity on the developing immune system and to establishwhether or not critical windows of vulnerability to potentially immunotoxic compounds exist atspecific developmental stages in mammals. The primary conclusion from a recent workshop organizedby ILSI HESI was that the best approach to developmental immunotoxicology testing wasto address all of the ‘windows’ at once (e.g., using an exposure regimen that extended from earlygestation through lactation and weaning and into young adulthood), and then go back and dissectspecific windows if an effect is seen (Sandler, 2002; Holsapple, 2002b). Importantly, this approachhas the potential to distinguish between highly transient effects and those that are genuinelypersistent. This approach would also address the fact that the immune system continues to developpostnatally. This point is particularly relevant when rodents are used in studies because their immunesystem undergoes such considerable anatomical development after birth. However, an approachthat addresses all of the critical windows is also important in species where it is recognized thatthe immune system is essentially functional at birth. Even in humans, there is considerable developmentof functional competence during the neonatal period, and virtually the entire scope ofimmunological memory is established after birth.REFERENCESAhmad H. and Chapnick E.K. 1999. Conjugated polysaccharide vaccines. Infectious Disease Clinics of NorthAmerica. 13:113.Anderson, U., Bird, A.G., Britton, S. and Palacios, R. 1981. Humoral and cellular immunity in humans studiedat the cell level from birth to two years of age. Immunol. Rev., 57:1.Aspinall, R., Kampinga, J. and Bogaerde, J. 1991. T cell development in the fetus and the invariant hypothesis.Immunol. Today, 12:7.August, C.S., Berkel, A.I., Driscoll, S. and Merler, E. 1971. Onset of lymphocyte function in the developinghuman fetus. Ped. Res., 5:539.Barnett, J.B. 1996. Developmental Immunotoxicology. In: Experimental Immunotoxicology. (Smialowicz, R.J.and Holsapple, M.P., eds.) CRC Press, Boca Raton, FL. Pg. 47.Barrett D.S., Rayfield L.S. and Brent L. 1982. Suppression of natural cell-mediated cytotoxicity in man bymaternal and neonatal serum. Clin. Exp. Immunol., 47: 742.Berry S.M., Fine N., Bichalski J.A., Cotton D.B, Dombrowski M.P. and Kaplan J. 1992. Circulating lymphocytesubsets in second- and third-trimester fetuses: comparison with newborns and adults. Am. J. Obstet.Gynecol., 167: 895.© 2006 by Taylor & Francis Group, LLC

1098 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONBryant, B.J. and Shifrine, M. 1972. Histiogenesis of lymph nodes during the development of the dog. J. Ret.Soc., 12:96.Bryant, B.J., Shifrine, M. and McNeil, C. 1973. Cell-mediated immune response in the developing dog. Int.Arch. All. Appl. Immunol., 45:937.Bunn, T.L., Dietert, R.R., Ladics, G.S. and Holsapple, M.P. 2001. Developmental immunotoxicology assessmentin the rat: Age, gender and strain comparisons after exposure to lead. Toxicol. Methods, 11:41.Cadoz, M. 1998. Potential and limitations of polysaccharide vaccines in infancy. Vaccine 16:1391.Cairo, M.S., Suen Y., Knoppel E., VandeVen C., Nguyen A. and Sender L. 1991. Decreased stimulated GM-CSF production and GM-CSF gene expression but normal numbers of GM-CSF receptors in humanterm newborns compared with adults. Pediatr. Res., 30:362.Chapin, R.E. 2002. The use of the rat in developmental immunotoxicology studies. Hum. Exp. Toxicol. 21:521.Chen, S., Golemboski, K.A., Sanders, F.S. and Dietert, R.R. 1999. Persistent effect of in utero meso-2,3-dimercaptosuccininic acid (DMSA) on immune function and lead-induced immunotoxicity. Toxicology,132:67.Clement, L.T., Vink P.E. and Bradley G.E. 1990. Novel immunoregulatory functions of phenotypically distinctsubpopulations of CD4 + cells in the human neonate. J. Immunol., 145 (1):102.Couture, L.A., Abbott, B.D. and Birnbaum, L.S. 1990. A critical review of the developmental toxicity andteratogenicity of 2,3,7,8-tertachlorodibenzo-p-dioxin: recent advances toward understanding the mechanism.Teratology, 42:619.Cumano, A, and I Godin. 2001. Pluripotent hematopoietic stem cell development during embryogenesis. CurrOpin Immunol 13:166.Cumano, A, C Ferraz, M Klaine, J Santo, I Godin. 2001. Intraembryonic, but not yolk sac hematopoietic precursors,isolated before circulation, provide long-term multilineage reconstitution. Immunity 15:477.Davidson, J.M. 1994. Pregnancy in renal allograft recipients: problems, prognosis and practicalities. Balliere’sClin. Obst. Gynecol., 8:501.Demeure C.E., Wu C.Y., Shu U., et al. 1994. In vitro maturation of human neonatal CD4 T lymphocytes: II.Cytokines present at priming modulate the development of lymphokine production. J. Immunol.,152:4775.Denny, T., Yogev, R., Gellman, R., Skuza, C., Oleske, J., Chadwick, E., Cheng, S.C. and Connor, E. 1992.Lymphocyte subsets in healthy children during the first 5 years of life. J. Am. Med. Assoc., 267:1484.Dietert, R.R., Etzel, R.A., Chen, D., Halonen, M., Holladay, S.D., Jarabek, A.M., Landreth, K., Peden, D.B.,Pinkerton, K., Smialowicz, R.J. and Zoetis, T. 2000. Workshop to identify critical windows of exposurefor children’s health: Immune and Respiratory Systems Work Group Summary. Environ. HealthPerspec. 108, Suppl. 3:483.Dietert, R.R., Lee, J.-E. and Bunn, T.L. 2002. Developmental immunotoxicology: Emerging issues. Hum. Exp.Toxicol. 21:479.Dijkstra, C. and Dopp, E. 1983. Ontogenetic development of T- and B-lymphocytes and non-lymphoid cellsin the white pulp of the rat. Cell Tissue Res., 229:351.Durandy, A., Thuillier, L., Forveille, M. and Fischer, A. 1990. Phenotypic and functional characteristics ofhuman neonatal B lymphocytes. J. Immunol., 144 (1):60.Erkeller-Yuksel, F.M., Deneys, V., Yuksel, B., Hannet, I., Hulstaert, F., Hamilton, C., Mackinnon, H., Stokes,L.T., Munhyeshuli, V., Vanlangendonck, F., DeBruyere, M., Bach, B.A. and Lydyard, P.M. 1992. Agerelatedchanges in human blood lymphocyte subpopulations. J. Pediatr., 120 (2): 216.Felsberg, P.J. 1998. Immunology of the dog. In: Handbook of Vertebrate Immunology. (Pastoret, P.P., Griebel,P., Bazin, H. and Govearts, A. eds.) Academic Press, New York, pg. 261.Felsburg, P.J. 2002. Overview of immune system development in the dog: Comparison with humans. Hum.Exp. Toxicol. 21:487.Foa, R., Giubellino, M.C., Fierro, M.T., Lusso, P. and Ferrando, M.L. 1984. Immature T lymphocytes inhuman cord blood identified by monoclonal antibodies: A model for the study of the differentiationpathway of T cells in humans. Cell. Immunol., 89:194.Gale, R.P. 1987. Development of the immune system in human fetal liver. Thymus, 10: 45.Gathings, W.E., Kubagawa, H. and Cooper, M.D. 1981. A distinctive pattern of B-cell immaturity in perinatalhumans. Immunol. Rev., 57:107.Ghia, P, E Boekel, A Rolink, F Melchers. 1998. B-cell development: a comparison between mouse and man.Immunol Today 19(10):480.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1099Godin, I, J Garcia-Porrero, F Dieterlen-Lievre, A Cumano. 1999. Stem cell emergence and hemopoietic activityare incompatible in mouse intraembryonic sites. J. Exp. Med., 190(1):43.Good, R A. 1995. Organization and development of the immune system. Relation to its reconstruction. NYAcad. Sci., 77:8.Granberg, C. and Hirvonen, T. 1980. Cell-mediated lympholysis by fetal and neonatal lymphocytes in sheepand man. Cell. Immunol., 51:13.Griffiths-Chu, S., Patterson, J.A.K., Berger, C.L., Edelson, R.L. and Chu, A.C. 1984. Characterization ofimmature T cell subpopulations in neonatal blood. Blood, 64(1):296.Hannet, I., Erkeller-Yuksel, F., Lydyard, P., Deneys, V. and DeBruyere, M. 1992. Developmental and maturationalchanges in human blood lymphocyte subpopulations. Immunol. Today, 13(6):215.Haynes, B.F., Martin, M.E., Kay, H.H. and Kurtzberg, J. 1988. Early events in human T cell ontogeny. J. Exp.Med., 168:1061.Haynes, B.F. and Hale, L.O. 1998. The human thymus. A chimeric organ comprised of central and peripherallymphoid components. Imm. Res., 18:61.Hayward, A. and Cosyns, M. 1994. Proliferative and cytokine responses by human newborn T cells stimulatedwith staphylococcal enterotoxin B. Pediatr. Res., 35(1):293.Hayward, A.R. 1983. Development of immunity mechanism. In: Pædiatric Immunology. (Soothill, J.F., Hayward,A.R. and Wood, C.B.S., eds.) Oxford, Blackwell, pg. 48.Hendrickx et al. 1975. Teratogenic effects of triamcinolone on the skeletal and lymphoid systems in nonhumanprimates. Fed. Proc., 34:1661.Hendrickx, A.G., Makori, N. and Peterson, P. 2000. Nonhuman primates: their role in assessing developmentaleffects of immunomodulatory agents. Hum. Exp. Toxicol., 19:219.Hendrickx, A.G., Makori, N. and Peterson, P. 2002. The nonhuman primate as a model of developmentalimmunotoxicology. Hum. Exp. Toxicol. 21:537.Herzyk, D.J., Bugelski, P.J., Hart, T.K. and Wier, P.J. 2002. Practical aspects of including functional endpointsin developmental toxicity studies. Case study: Immune function in HuCD4 transgenic mice exposedto anti-CD4 Mab in utero. Hum. Exp. Toxicol. 21:507.HogenEsch, H., Housman, J.M. and Felsberg, P.J. 1987. Canine Peyer’s patches: macroscopic, light microscopic,scanning electron microscopic and immunohistochemical investigations. Adv. Exp. Med. Biol.,216A:249.Holladay, S.D. and Smialowicz, R.J. 2000. Development of the murine and human immune system: Differentialeffects of immunotoxicants depend on time of exposure. Environ. Health Perspec. 108, Suppl. 3:463.Holladay, S.D. and Blaylock, B.L. 2002. The mouse as a model for developmental immunotoxicology. Hum.Exp. Toxicol. 21:525.Holsapple, M.P. 2002a. Developmental immunotoxicology and risk assessment: A workshop summary. Hum.Exp. Toxicol. 21:473.Holsapple, M.P. 2002b. Developmental immunotoxicity testing: A review. Toxicology 185:193.Hou, S. 1999. Pregnancy in chronic renal insufficiency and end-stage renal disease. Am. J. Kid. Dis., 33:235.Hulstaert, F., Hannet, I., Deneys, V., et al. 1994. Age-related changes in human blood lymphocyte subpopulations.Clin. Immunol. Immunopath., 70:152.Ibegbu, C., Spira, T.J., Nesheim, S., et al. 1994. Subpopulations of T and B cells in perinatally HIV-infectedand noninfected age-matched children compared with those in adults. Clin. Immunol. Immunopath.,71:27.Jacoby, R.O., Dennis, R.A. and Griesemer, R.A. 1969. Development of immunity in fetal dogs: humoralresponses. Am. J. Vet. Res., 30:1503.Jain, A. 1993. FK506 and pregnancy in liver transplant patients. Transplantation, 56:751.Jenkins, M.K., Schwartz, R.H. and Pardoll, D.M. 1988. Effects of cyclosporine A on T cell development andclonal deletion. Science, 241:1655.Jennings, C., Rich, K., Siegel, J.N. and Landay, A. 1994. A phenotypic study of CD8 + lymphocyte subsetsin infants using three-color flow cytometry. Clin. Immunol. Immunopath., 71:8.Kay, H.E.M., Playfair, J.H., Wolfenade, M. and Hopper, P.K. 1962. Development of the human thymus duringfetal development. Nature, 196:238.Kay, H.E.M., Doe, J. and Hockley, A. 1970. Response of human fœtal thymocytes to phytohæmagglutinin(PHA). Immunology, 18:393.Kelly, W.D. 1963. The thymus and lymphoid morphogenesis in the dog. Fed. Proc., 20:600.© 2006 by Taylor & Francis Group, LLC

1100 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONKendall, M.D. 1991. Functional anatomy of the thymic microenvironment. J. Anat., 177:1.Kimmel, C. 2001. Developmental immunotoxicity considerations in testing and risk assessment for children’shealth (abstract). The Toxicologist, 60:65.Kimurs, S., Eldridge, J.H., Michalek, S.M., Morisaki, I., Hamada, S. and McGhee, J.R. 1985. Immunoregulationin the rat: Ontogeny of B cell responses to types 1, 2 and T-dependent antigens. J. Immunol.,134:2839.Kincade, P.W. 1981. Formation of lymphocytes in fetal and adult life. Adv. Immunol., 31: 177.Klein, A.K., Dyck, J.A., Stitzel, K.A., Shimizu, M.J., Fox, L.A. and Taylor, N. 1983. Characterization ofcanine lymphohematopoiesis. Exp Hematol., 11:263.Kohl, S., Loo, L.S. and Gonik, B. 1984. Analysis in human neonates of defective antibody-dependent cytotoxicityand natural killer cytotoxicity to herpes simplex virus infected cells. J. Infect. Dis., 150:14.Ladics, G.S., Smith, C., Bunn, T.L., Dietert, R.R., Anderson, P.K., Wiescinski, C.M. and Holsapple, M.P.2000. Characterization of an approach to developmental immunotoxicology assessment in the ratusing SRBC as the antigen. Toxicol. Methods, 10:283.Laifer, S.A., Yeagley, C.J., Armitage, J.M. 1994. Pregnancy after cardiac transplantation. Am. J. Perinatol.,11:217.Landreth, K.S. 2002. Critical windows in development of the rodent immune system. Hum. Exp. Toxicol.21:493.Lobach, D.F. and Haynes, B.F. 1987. Ontogeny of the human thymus during fetal development. J. Clin.Immunol. 7:81.Loke, Y.W. and King, A. 1991. Immunology of human pregnancy. In: Scientific Foundation of Obstetrics andGynæcology. (Philipp, E., Setchell, M. and Ginsburg, J., eds.) Oxford, Butterworth-Heinemann Ltd.,pg. 55.Loke, Y.W. 1978. Development of immunocompetence in the human foetus. In: Immunology and Immunopathologyof the Human Foetal-Maternal Interaction. (Loke, Y.W., ed.) Amsterdam, Elsevier North-Holland Biomedical Press, pg. 51.Marshall-Clarke, S., Reen, D., Tasker, L. and Hassan, J. 2000. Neonatal immunity: How well has it grownup? Immunol. Today, 21:35.Migliaccio, G., Migliaccio, A.R., Petti, S., et al. 1986. Human embryonic hemopoiesis - Kinetics of progenitorsand precursors underlying the yolk sac --> liver transition. J. Clin. Invest., 78:51.Miller, G.K. and Benjamin, S.A. 1985. Radiation-induced quantitative alterations in prenatal thymic developmentin the beagle dog. Lab. Invest., 52:224.Miyawaki, T., Moriya, N., Nagaoki, T. and Tonigachi, N. 1981. Maturation of B-cell differentiation abilityand T-cell regulatory function in infancy and childhood. Immunol. Rev., 57:69.Mosier, D.E. 1977. Ontogeny of T-cell function in the neonatal mouse. In: Development of Host Defenses.(Cooper, M.D. and Dayton, D.H. eds.). New York, Raven Press, pg. 115.Mumford, D.M., Sung, J.S., Wallis, J.O. and Kaufmann, R.H. 1978. The lymphocyte transformation responseof fetal hemolymphatic tissue to mitogens and antigens. Pediatr. Res. 12:171.Murgita, R.A. and Wigzell, H. 1981. Regulation of immune functions in the fetus and newborn. Prog. Allergy,29:54.Nahmias, A., Ibegbu, C., Lee, F. and Spira, T. 1994. The development of the immune system - importance inthe ascertainment of immunophenotypic changes in perinatal HIV infection. Clin. Immunol. Immunopath.,71:2.Namikawa, R., et al. 1986. Ontogenetic development of T- and B-cells and non-lymphoid cells in the whitepulp of the human spleen. Immunology, 57:61.Neubert, R., Helge, H. and Neubert, D. 1996. Nonhuman primates as models for evaluating substance-inducedchanges in the immune system with relevance for man. In: Experimental Immunotoxicology. (Smialowicz,R.J. and Holsapple, M.P., eds.) CRC Press, Boca Raton, FL. pg. 63.Neubert, R.T., Webb, J.R. and Neubert, D. 2002. Feasibility of human trials to assess developmental immunotoxicity,and some comparison with data on New World monkeys. Hum. Exp. Toxicol. 21:543.Ohama, K. and Kaji, T. 1974. Mixed cultures of fetal and adult lymphocytes. Am. J. Obstet. Gynecol., 119:552.Oldstone, M.B.A. and Tishon, A. 1977. Active thymus-derived suppressor lymphocytes in human cord blood.Nature, 269:333.Paige, C.J., P.W. Kincade, L.A. Shinefield, and V.L. Satio. 1981. Precursors of murine B lymphocytes: physicaland functional characterization and distinctions from myeloid stem cells. J. Exp. Med., 153:154.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1101Papadogiannakis, N., Johnsen, S.A. and Olding, L.B. 1985. Strong prostaglandin associated suppression ofthe proliferation of human maternal lymphocytes by neonatal lymphocytes linked to T versus T cellinteractions and differential PGE2 sensitivity. Clin. Exp. Immunol., 61:125.Papadogiannakis, N. and Johnsen, S.A. 1988. Distinct mitogens reveal different mechanisms of suppressoractivity in human cord blood. J. Clin. Lab. Immunol., 26:37.Peakman, M., Buggins, A.G.S., Nicolaides, K.H., Layton, D.M. and Vergani, D. 1992. Analysis of lymphocytephenotypes in cord blood from early gestation fetuses. Clin. Exp. Immunol., 90:345.Penit, C.and Vasseur, F. 1989. Cell proliferation and differentiation in the fetal and early postnatal mousethymus. J. Immunol., 142:3369.Penn, I., Makowski, E.L. and Harris, P. 1980. Parenthood following renal transplantation. Kidney International,18:221.Pietra, B.A. and Boucek, M.M. 2000. Immunosuppression for pediatric cardiac transplantation in the modernera. Progress in Pediatric Cardiology, 11:115.Plebani, A., Proserpio, A.R., Guarneri, D., Buscaglia, M., Cattoretti, G. 1993. B and T lymphocyte subsets infetal and cord blood: age-related modulation of CD1c expression. Biol. Neonate, 63:1.Rabatic, S., Benic, L., Mazuran, R. and Dekaris, D. 1990. Polymorphonuclear leucocyte and natural killercell functions in mature and premature newborns. Biol. Neonate, 58:252.Rayfield, L.S., Brent, L. and Rodeck, C.H. 1980. Development of cell-mediated lympholysis in human fœtalblood lymphocytes. Clin. Exp. Immunol., 42:561.Reason, D.C., Ebisawa, M., Saito, H., Nagakura, T. and Iikura, Y. 1990. Human cord blood lymphocytes donot simultaneously express CD4 and CD8 cell surface markers. Biol. Neonate, 58:87.Rosen, F.S., Cooper, M.D. and Wedgwood, R.J.P. 1995. The primary immunodeficiencies. N. Engl. J. Med.,333:431.Rothkotter, H.J., Sowa, E. and Pabst, R. 2002. The pig as a model of developmental immunotoxicolology.Hum. Exp. Toxicol. 21:533.Royo, C., Touraine, J.-L. and DeBouteiller, O. 1987. Ontogeny of T lymphocyte differentiation in the humanfetus: Acquisition of phenotype and function. Thymus, 10:57.Sandler, J.D. 2002. Executive summary. Hum. Exp. Toxicol. 21:469.Santoni, A., Riccardi, C., Barlozzari, T. and Herberman, R.B. 1982. Natural suppressor cells for murine NKactivity. In: NK Cells and Other Natural Effector Cells. (Herberman, R.B., ed.). New York, AcademicPress, pg. 443.Sargent, I.L., Redman, C.W.G. and Starkey, P.M. 1993. The placenta as a graft. In: The Human Placenta.(Redman, C.W.G., Sargent, I.L. and Starkey, P.M., eds.) Oxford, Blackwell Scientific Publications,pg. 334.Semenzato, G., Piovesan, A., Amadori, G., Colombatti, M., Gasparotto, G. and Rubaltelli, F.F. 1991. T cellimmune function in newborn infants. Biol. Neonate, 37:8.Series, I.M., Pichette, J., Carrier, C., et al. 1991. Quantitative analysis of T and B cell subsets in healthy andsick premature infants. Early Hum. Devel., 26:143.Shifrine, M.N., Smith, J.B., Bulgin, M.S., Bryant, J.B., Zec, Y.C. and Osborn, B.I. 1971. Response of caninefetuses and neonates to antigenic stimulation. J. Immunol., 107:965.Shortman K, Vremec D. Cocoran IM, Georgopolos K, Lukas K, Wu L. 1998. The linkage between T-cell anddendritic cell development in the mouse thymus. Immuno. Rev. 165:39.Smialowicz, R.J. 2002. The rat as a model in developmental immunotoxicology. Hum. Exp. Toxicol. 21:513.Smith, R.T. 1968. Development of fetal and neonatal immunological function. In: Biology of Gestation. VolumeII: The Fetus and Neonate. (Asal, N.S., ed.). New York Academic Press, pg. 321.Snyder, P.W., Kazacos, E.A. and Felsberg, P.J. 1993. Histologic characterization of the thymus in canine X-linked severe combined immunodeficiency. Clin. Imm. Immunpath., 67:55.Solinger, A.M. 1985. Immature T lymphocytes in human neonatal blood. Cell. Immunol., 92:115.Somberg, R.L., Robinson, J. and Felsberg, P.J. 1994. T-lymphocyte development and function in dogs withX-linked severe combined immunodeficiency. J. Immunol., 153:4006.Somberg, R.L., Tipold, A., Hartnett, B.J., Moore, P.F., Henthorn, P.S. and Felsberg, P.J. 1996. Postnataldevelopment of T-cells in dogs with X-linked severe combined immunodeficiency. J. Immunol.,156:1431.Spear, P.G., Wang, A., Rutishauser, U. and Edelman, G.M. 1973. Characterization of splenic lymphoid cellsin fetal and newborn mice. J. Exp. Med., 138:557.© 2006 by Taylor & Francis Group, LLC

1102 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSplawski, J.B. and Lipsky, P.E. 1991. Cytokine regulation of immunoglobulin secretion by neonatal lymphocytes.J. Clin. Invest., 88(3):967.Stites, D.P., Carr, M.C. and Fudenberg, H.H. 1974. Ontogeny of cellular immunity in the human fetus:development of responses to phytohemagglutinin. Cell. Immunol., 11:257.Takahashi, N., Nishida, H. and Hoshi, J. 1994. Severe B-cell depletion in newborns from renal transplantmothers taking immunosuppressive agents. Transplantation, 57:1617.Tanimura, T. and Tanioka, Y. 1975. Comparison of embryonic and fetal development in man and rhesusmonkey. In: Breeding Simians for Developmental Biology, Laboratory Animal Handbooks. LaboratoryAnimals Limited, London.Tavassoli, M. 1995. Ontogeny of hemapoiesis. In: Handbook of Human Growth and Development, Vol. III.(Meisami, S., ed.) New York-CRC Press.Thorbecke, G.J. 1959. Some histological and functional aspects of lymphoid tissue in germfree animals. I.Morphological studies. Ann. NY Acad. Sci., 78:107.Timens, W. et al. 1989. Immaturity of the human splenic marginal zone in infancy. Possible contribution tothe deficient infant immune response. J. Immunol., 143:3200.Toivanen, P., Uksila, J., Leino, A., Lassila, O., Hirnonen, T. and Ruuskanen, O. 1981. Development of mitogenresponding T cells and natural killer cells in the human fetus. Immunol. Rev., 57:89.Tsuji, T., Nibu, R., Iwai, K., et al. 1994. Efficient induction of immunoglobulin production in neonatal naiveB cells by memory CD4 + T cell subset expressing homing receptor L-selectin. J. Immunol., 152:4417.Tucci, A., Mouzaki, A., James, H., Bonnefoy, J.-Y. and Zubler, R.H. 1991. Are cord blood B cells functionallymature? Clin. Exp. Immunol., 84:389.Verlarde, A. and Cooper, M.D. 1984. An immunofluorescence analysis of the ontogeny of myeloid, T, and Blineage cells in mouse hematopoietic tissues. J. Immunol., 133:672.VonGaudecker, B. 1991. Functional histology of the human thymus. Anat. Embryol., 183:1.Watts, A.M., Stanley, J.R., Shearer, M.H., Hefty, P.S. and Kennedy, R.C. 1999. Fetal immunization of baboonsinduces a fetal-specific antibody response. Nat. Med., 5:427.Webber, S.A. 1996. Newborn and infant heart transplantation. Current Opinion in Cardiology, 11:68.Weissman, I. 2000. Stem cells: Units of development, units of regeneration and units in evolution. Cell 100:157.West, L.J. 2002. Defining critical windows in the development of the human immune system. Hum. Exp.Toxicol. 21:499.Yang, T.J. and Gawlack, S.L. 1989. Lymphoid organ weights and organ:body weight ratios of growing beagles.Laboratory Animals, 23:143.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1103APPENDIX C-8*SPECIES COMPARISON OF POSTNATAL CNS DEVELOPMENT:FUNCTIONAL MEASURESSandra L. Wood 1,4 , Bruce K. Beyer 2 and Gregg D. Cappon 31Merck Research Laboratories, West Point, PA 19486, USA2Sanofi-Synthelabo Research, Malvern, PA 19355, USA3Pfizer Global Research & Development, Groton, CT 06340, USA4Correspondence to: Dr. Sandra L. Wood, Merck Research Laboratories, WP45-103, Safety Assessment, West Point,PA 19486. Phone (215) 652-6334. Fax (215) 652-7758. Email: sandra_wood2@merck.comIntroductionThe purpose of this review is to identify key events in human postnatal CNS development andcompare them, when possible, to other species. Specifically, this review will focus on the behavioralmeasures of CNS development, but will not include reviews of neuroanatomical or neurochemicalmaturation. This review is part of the initial phase of a project undertaken by the Developmentaland Reproductive Toxicology Technical Committee of the ILSI Health and Environmental SciencesInstitute (HESI) to bring together information on a selected number of organ systems and comparetheir postnatal development across several species (Hurtt and Sandler, 2003). This review is notintended to propose effective study designs and testing strategies since this is the final objective ofthe HESI project. It is rather to be used as a resource to aid researchers when beginning to planthe necessary nonclinical safety evaluations for pediatric drugs (Department of Health and HumanServices (FDA), 1998; FDA, 1998).There is general recognition that the developing nervous system is qualitatively different fromthe adult system. With respect to the central nervous system (CNS), the human is an altricial species,i.e., born in a relatively immature state and has a prolonged postnatal dependency. The human CNSmust undergo considerable postnatal development in order to reach its adult state of maturation. Alarge percentage of laboratory animals, such as mice, rats, cats, dogs and rabbits, are like humansand must undergo a prolonged period of neurologic development postnatally. However, somespecies, like guinea pigs and sheep, are born with very mature nervous systems, which undergolittle CNS development postnatally. Therefore, these precocial species may be of little practical useas a comparative model to the human CNS.Few would argue with the premise that the ultimate manifestation of a neural process is behavior.However, the links between the biological processes occurring during CNS development and thebehavioral capacities exhibited are not easily established. Because behavior is the aspect of postnatalCNS development that is clinically monitored in humans as well as preclinically in many laboratoryspecies, behavioral measures of neurologic function will be the focus of this review. The measuresselected have no special status, but it is the hope of the authors to provide a broad overview of thedeveloping features that are under CNS control. The maturation of the CNS has been organized ineight categories: 1) Locomotor Development, 2) Ontogeny of Fine Motor Development and Dexterity,3) Sensory and Reflex Development, 4) Cognitive Development, 5) Communication, 6) SocialPlay, 7) Development of the Fear Response and 8) Development of Sleep Cycles.It should be noted that the review will not include detailed descriptions of all behavioralparadigms mentioned. The reader is encouraged to refer to the cited text for the specific details of* Source: Wood, S.L., Beyer, B.K., Cappon, G.D., Species comparison of postnatal CNS development: functional measures.Birth Defects Research, Part B: Developmental and Reproductive Toxicology, 68, 321-334, 2003.© 2006 by Taylor & Francis Group, LLC

1104 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONa behavioral assessment and Chapter 4 of “The Handbook of Behavioral Teratology” for a generaldescription of the methods most often utilized in behavioral evaluations (Adams, 1986).Ontogeny of Locomotor DevelopmentComparisons of locomotor development among humans, non-human primates, rats and dogs arecomplicated since these later two species never walk upright. In addition, human infants cannotreadily swim like rats or dogs. However, a comparison of motor develop in all these species isbeneficial since humans, rats and dogs all gradually development from a rather immature state oflocomotion (i.e., crawling) to a mature state of walking. Non-human primates also develop locomotorabilities postnatally, but at a faster pace than their human counterparts. This section of the review willcompare the early ontogeny of locomotion between humans and common laboratory species.HumansBy 12 to 14 weeks, an infant can hold her chin and shoulders off the floor with her forearms(Illingworth, 1983). By 24 weeks, an infant can hold her chest and upper part of her abdomen offthe floor and maintain her weight by her hands with extended arms. By 28 weeks she can bear herweight on one hand, and by 9 months she can usually crawl. Interestingly, the first stage of crawlingis typically progression backward. Around 10 months, she can creep on her hands and knees withher abdomen off the floor.The last stage before walking around 13 months of age typically involves intermittently placingone foot flat on the floor and creeping like a bear on hands and feet. Interestingly, an infant canbear her full weight (i.e., stand) while being held by her hands by 24 to 28 weeks and can walkwhile holding onto a piece of furniture by 48 weeks. When an infant does begin walking on herown, her arms are held outward, elbows are flexed and her base is broad. In addition, the stepsoften vary in length and direction and are by no means completely adult-like. By 18 to 24 months,the toddler learns to walk with legs closer together, arms moved inward and a heel-to-toe gaitdevelops. By 2.5 years, she can jump with both feet and walk on her tiptoes. By 3 years, the childbegins to display a more adult-like motor pattern.RatsLocomotion in the rat matures during the first few weeks after birth (Altman and Sudarshan, 1975;Westerga and Gramsbergen, 1990). Rat pups are able to ambulate through the use of their forelimbs,upper torso and head beginning around Postnatal Day (PND) 3-4. This “crawling” behavior peaksaround PND 7 and disappears around PND 15 (Altman and Sudarshan, 1975; Westerga andGramsbergen, 1990). It is not until around PND 8-10 that rat pups can stand with their abdomenscompletely off the floor (Altman and Sudarshan, 1975; Bolles and Woods, 1964). Around PND12-13, rat pups can walk while supporting their full weight, but the hindlimbs are typically rotatedoutward. Adult-like locomotion appears around PND 15-16 which is shortly after eye-opening inrats (Westerga and Gramsbergen, 1990). Fast running appears around PND 16 (Altman and Sudarshan,1975). Beginning around PND 16, rats can stand on their hind-legs while leaning their forelegsagainst a wall support (Altman and Sudarshan, 1975; Bolles and Woods, 1964). Around PND18, rats can rear for short periods of time without foreleg support.An additional issue regarding rat locomotion is a peak period of hyperactivity in rats observedaround PND 15-20, during which pups change from one activity to another very quickly (Bollesand Woods, 1964), e.g., switching from grooming to running to fighting in a matter of seconds.After this period, there is a slight decrease in their activity level then a gradual increase to adultactivity levels around PND 50-60.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1105SwimmingAnother type of rat “ambulation” that should be mentioned because of its importance in manybehavioral paradigms (e.g., Morris Water maze, Biel Water maze) is the ontogeny of swimming.The stages of swimming have been well documented (Bekoff and Trainer, 1979; Butcher andVorhees, 1979; Schapiro et al., 1970; Vorhees, 1986). When PND 6 rats are placed in water, theyoften display an asynchronous movement of the limbs which results in little forward motion butsome small turns. Over the next few days, rats will begin to propel themselves forward by paddlingin a coordinated fashion with all four limbs. Around PND16, paddling with all limbs begins todisappear, and what emerges is the more adult-like swimming style in which the hindlimbs areused to propel forward. The forelimbs are held stationary under the jaw, except during turning.DogsLocomotion in the canine develops during the first month or so after birth (Fox, 1964a; Fox, 1971;Thomas, 1940). Crawling forward (or a rooting reflex) occurs when there is bilateral stimulationof the head with the hand cupped around the muzzle (Fox, 1964a). This reflex is demonstratedbeginning around PND 4. This reflex becomes weaker and more variable around PND 14 andonward. Between PND 6 and 10, supporting of the forelimbs is first observed. Supporting of thehindlimbs begins around PND 11 to 15. Sitting upright on the forelimbs is seen around PND 20while standing is observed around PND 21. The hindlimbs appear to tremor at this time indicatinga weakness in muscle tone. There is a marked transition from variable locomotor responses toadult-like postures and equilibrium from PND 28 onward. The primitive reflexes that were anintegral part of neonatal behavior disappear around PND 28, and the ability of the dog to remainupright when subjected to lateral tilting is well developed by PND 35-42.Non-human PrimatesIt is normal for the young macaque to cling to its mother for the first 2 to 3 days of life (Hines,1942). Shortly after this period, the infant can raise her chest off the surface of the floor by extendingher arms and head. Surprisingly, the infant monkey begins to stand on all fours by 4 days afterbirth (Golub, 1990; Hines, 1942). Around approximately 3 weeks (in some cases as early as theend of first week), the position of the trunk becomes more horizontal to the floor surface (i.e., moreadult-like), and between the 5 th and 11 th weeks, the extremities become more extended (Hines,1942). Bipedal standing is observed as early as 7 weeks of age.The infant macaque develops locomotor capabilities in both the horizontal plane (i.e., the floorsurface) and the vertical plane (i.e., climbing and jumping). During the first two weeks of life,forward progression in the horizontal plane begins with reaching and grasping of the extremitiesin a rather abducted position (i.e., extremities extended away from the midline of the body). Theextension of the extremities becomes more exaggerated between the 5 th and 11 th weeks of life, andthe knee is rather extended compared to the adult form. Easy flexion-extension of the knee andadduction of the extremities (i.e., extremities extended toward the midline of the body) beginsaround 7 weeks, but is not seen with consistency until approximately 12 weeks of age.The infant macaque begins to develop climbing capabilities as early as 2 days of age. However,the ability to return either by backing down or turning around does not begin to emerge until theend of the 1 st week and the 5 th week, respectively. Jumping across a space was not attempted byeven the most adventurous macaque until 5 weeks of age.It is interesting that adult standing is obtained by approximately the 3 rd month, but an adultsitting posture is not obtained until 9 months of age. Specifically, by the 4 th week of life, the infantmacaque can squat on two feet with initially some aid from the tail. The infant begins to sporadically© 2006 by Taylor & Francis Group, LLC

1106 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONsit erect by 7 weeks, but the consistent use of the adult posture of sitting was not observed routinelyuntil approximately 9 months of age.Conclusions for Locomotor DevelopmentIn humans, rats, dogs and non-human primates, locomotor capabilities develop postnatally. In rats,dogs and humans, there appears to be a more gradual rostrocaudal maturation that leads to theshoulders in a raised position prior to the pelvis. In contrast, non-human primates develop motorabilities rather quickly after birth compared to humans. Therefore, assessments of motor developmentin juvenile toxicity studies would be beneficial in any of these species. However, studies innon-humans need to be carefully designed in order to accommodate the rapid development oflocomotion in this species.Table 1Summary of Locomotor DevelopmentHuman Rat DogCrawling ≈ PND 270(9 months)Walking ≈ PND 396(13 months)PND = Postnatal DayaBipedal locomotionNon-HumanPrimate (Rhesus)PND 3–12 PND 4–20 PND 4–49PND 12–16 PND 20–28 PND 49 aOntogeny of Fine Motor Development and DexterityFine motor skills and dexterity are generally associated with primates, especially humans. However,rats possess dexterity in both their fore- and hindlimbs. Although a reflexive ability to grasp objectsmay be present early in the postnatal life of many species, fine motor skills and other coordinatedmotor movements develop over time.HumansThe reflex grasping of an object placed in the palm of the newborn (i.e., the palmar grasp reflex)is present at birth and begins to disappear around 2 months of age (Herschkowitz et al., 1997;Rosenblith and Sims-Knight, 1904; Volpe, 1995). The amount of pre-reaching, i.e., opening of handwhile approaching an object, decreases at around 7 weeks of age. The infant does not seem to loseinterest in the object at this age, but attending to the object inhibits pre-reaching activity in someway. The form of pre-reaching also changes at this age. Initially, instead of opening the hand duringthe forward extension of the arm, the hand will be fisted. After this age, the amount of pre-reachingactivity increases and the hand starts opening during the forward extension, but only when theinfant looks at the object (von Hofsten and Fazel-Zandy, 1984).Onset of reaching occurs in the fifth postnatal month, and infants reliably grasp for objectswithin their reach 3-4 months after this milestone (Konczak and Dichgans, 1997; Lantz et al.,1996). However, most kinematic grasping parameters do not reach adult-like levels before age 2years. Between ages 2 and 3, improvements in hand- or joint-related movements are marginal. Theformation of a consistent interjoint synergy between shoulder and elbow motion is not achievedwithin the first year of life. Stable patterns of temporal coordination across arm segments begin toemerge at 12-15 months and continue to develop to age 3 (Konczak and Dichgans, 1997).Younger children (4-5 years) had relatively wider open grips than older (12 years), thus graspingwith a higher safety margin. Movement duration and peak spatial velocity of the reaching handwas similar for children 4-12 years of age. In addition, hand trajectory straightened, and coordination© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1107between hand transport and grip formation improved by 12 years of age. Only the oldest children(12 years) were able to accurately adjust their grip according to the various sizes of target objects.Thus development of prehensile skills during childhood lasts until the end of first decade of life(Kuhtz-Buschbeck et al., 1998).Sensorimotor memory coordinates the forces used to grip objects, and maturation of this systemis believed to underlay the development of precision grip. Young children have an immature capacityto adapt to the frictional condition, e.g., silk vs. sandpaper. During the first 5 years of age, childrenshow improvement in their ability to apply the correct grip to various frictional conditions. At 18months, infants could adjust to the frictional condition in a series of lifts when the same surfacestructure was presented in blocks of trials, but failed when the surface was changed unpredictablybetween subsequent lifts (Forssberg et al., 1995). This suggests a poor capacity to form sensorimotormemory. Older children (12-13 years) required a few lifts and adults only one lift to update theirforce coordination to a new friction. Therefore, young children use excessive grip force, a strategyto avoid frictional slips to compensate for an immature tactile control of the precision grip (Forssberget al., 1995).RatsFine motor and dexterity development is difficult to measure in the rat, and there is little data availabledescribing its development. When rats are suspended from their forepaws on a wire, their hindlimbsprovide support to prevent falling and aid progression along the wire. Wistar rats demonstrate hindlimbgrasping abilities as early as PND2 and can support their weight for about 10-20 seconds up until PND16 (Petrosini et al., 1990). Beginning around PND 16, Wistar rats can remain suspended on the wirefor 70-120 seconds with 100% of the animals mastering this task by PND 32.DogsNo information found.Non-human primatesPrimates demonstrate considerable postnatal development of fine motor movements and dexterity.The ontogeny of relatively independent finger movements in the newborn macaque monkey requiresthe development of functional cortico-motorneuronal (CM) connections (Lemon, 1999). Althoughcorticospinal fibers reach all levels of the spinal cord before birth, and there may be substantialinnervation of the spinal grey matter of the intermediate zone, significant CM connections to themotor nuclei innervating hand muscles are not present at birth.The first appearance of relatively independent finger movements occurs around 3 months ofage in monkeys (Lemon, 1999). Thereafter, both the anatomical and physiological development ofthe corticospinal system follows a protracted time course which, in general, parallels the maturationof motor skill. It is believed that fine finger movements are not observed before functional CMconnections are well established, and that rapid changes in the physiological properties of thecorticospinal system coincide with the period in which precision grip is known to mature (3-6months) (Lemon, 1999). However, corticospinal development continues long after simple measuresof dexterity indicate functional maturity, and these changes may contribute to the improved speedand coordination of skilled hand tasks (Olivier et al., 1997). Striking differences have been foundbetween the strong CM projections to the hand motorneuron pools in the dextrous capuchin monkeycompared with the much weaker projection in the far less-dextrous squirrel monkey. Similar findingshave been reported for the macaque, which is also very dextrous (Lemon, 1999).By 6 months of age, all basic forms of manipulation seen in adult capuchin monkeys haveappeared (Adams-Curtis and Fragaszy, 1994). Actions that occur frequently in the first 8 weeks© 2006 by Taylor & Francis Group, LLC

1108 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONare gentle and involve sustained visual orientation and aimed reaching. Later actions are morevigorous and involve grasping. Large increases in the rate of activity are evident over the 6-monthperiod of development studied. The increase from the first 8 weeks to the second 8 weeks may bedue to a) an increase in the amount of time spent alert and active, b) a decrease in the amount oftime spent in a ventral position on the mother, c) improvements in postural control and stamina,and d) the onset of independent locomotion. Changes in form can be attributed primarily to posturalfactors and to neuromuscular development, resulting in precisely aimed and controlled movementsappearing in the 5 th and 6 th months.Fine motor abilities develop rapidly in infant monkeys. By 14 weeks of age, infant macaquesdemonstrate well developed fine motor abilities including hand-to-mouth coordination, bimanualcoordination (evaluated using a task which involves lifting a transparent inverted beaker with onehand while retrieving a reward from under it with the other hand) and finger-thumb opposition(Golub et al., 1991b). These results may not be readily generalized to humans because of differencesin maturity of motor systems at birth. Monkey neonates are more advanced in motor function thanhuman newborns. For example, monkeys crawl within a few days of birth.Conclusions for Fine Motor and Dexterity DevelopmentFine motor and dexterity development is generally protracted. Human neonates do not possess thecapacity to perform fine finger movements, and these may not make their appearance for manymonths after birth. For example, precise coordination involved with grasping an object and liftingmay not be fully mature until the second decade of life. In contrast, much cruder grasp functionsare present at birth. Postnatal development of fine motor activity in humans and non-human primatesfollows a similar pattern, although primates develop much more rapidly than humans. However,there is a paucity of data in rats and dogs. Therefore, use of this measure in routine studies is notlikely to provide any benefit for risk assessment.Table 2Summary of Fine Motor Development and DexterityFine Motor DevelopmentHumanPalmar Grasping ReflexPrehensionPrecision GripApproximate Postnatal AgeAbility Emerges< 2 months5–9 months1.5–13 yearsRatsFine Motor Development PND 2–32DogNo Information FoundNo Information FoundNon-Human PrimateFine Motor DevelopmentPrecision GripManipulations11 weeks3–6 months6 monthsOntogeny of Sensory and Reflex DevelopmentThe various sensory systems develop in a sequential manner. Across a wide range of species, thefunctional onset of the auditory system precedes that of the visual system. Furthermore, thefunctional onset of tactile and olfaction usually precedes that of the auditory system (Alberts, 1984).Conditioned responses to various sensory modalities develop in the same sequence (Richardson etal., 2000), i.e., conditioned responses usually occur first to olfactory, then to auditory, and finallyto visual cues.While electrophysiological and histological research tools are very useful in determining theonset of functional sensory systems, the significance of sensory function is ultimately linked to© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1109behaviors that are measured as sensorimotor reflexes. Unfortunately, a limitation of this approachis that the development of sensorimotor system is dependent on the maturation of both the motorand sensory systems. Therefore, it is not always possible to completely separate the developmentof a given sensory system from the development of the motor system.One of the most common ways to assess sensory and reflex functionality in a developing animal,especially the rodent, is through the use of test batteries (e.g., Functional Observational Battery(FOB)) (Carlson, 2000; Geyer and Reiter, 1985; Spear, 1990). Test batteries have also beendeveloped to study behavioral functionality in monkeys and children (Paule, 2001).HumansThere are several standardized clinical tests for assessing the intactness of the nervous system ofinfants and young children (Winneke, 1995). One of the most commonly used batteries is theNeonatal Behavioral Assessment Scale (NBAS) (Connolly, 1985; Stengel, 1980; Stewart et al.,2000). During early infancy (0-2 months), reflexive development is most commonly used as adevelopmental milestone. Those reflexes that develop during the first month of postnatal life include:1) a response to touch [e.g., such as the in curving to the side of skin stimulation (Galant response)],2) grasping responses (Palmar finger grasp), and 3) stepping and placing reflexes (automatic walkingwhen held upright) (Connolly, 1985). The infantile oral reflexes, consisting of rooting, biting, mouthopening, lateral tongue movement and Babkin reflexes (Sheppard and Mysak, 1984) are commonlyevaluated in humans. These reflexes are generally apparent at birth and are active through 4-6months of age (Sheppard and Mysak, 1984). Onset of chewing behavior occurs around 5-8 months,and in normal development occurs during a waning of the infantile oral reflexes (Sheppard andMysak, 1984). In general, persistence of infantile oral reflexes beyond the age at which they arereported to occur in normal infants has been viewed as pathological.Somatosensory (touch/tactile) systems are fully developed during gestation. For example, atbirth the human neonate should respond to touch by directing attention to the body part that isbeing touched (Bower, 1979). Auditory function also develops prenatally. Physiologic evidence ofcochlear function after the 20 th week of gestation supports findings of fetal responses to sound. Atbirth, neonates should respond to loud sounds, but not to pure auditory stimulus (Eisenberg, 1970).Newborns contain a functional taste system; however, gradual transitions in electrical responsesto taste stimuli are shown from fetal to adult stage (Rao et al., 1997). The behavioral measure oftaste development is intake behavior. The ability to discriminate degrees of sweetness is welldeveloped at birth (Desor et al., 1975), bitter aversion develops within a few days of birth (Rao etal., 1997), but salt discrimination may not develop until 4 years of age (Desor et al., 1975; Kajiuraet al., 1992).The human visual system is functional at birth but undergoes significant postnatal development,and most components are fully operative by the end of the second year of life (Suchoff, 1979).Different functions emerge in different age ranges, for example, flicker sensitivity is mature by 3months postnatal, binocular functions emerge from 3-7 months of age and contrast sensitivity isnot adult-like until several years after birth (Boothe et al., 1985). Visual acuity develops from about20/600 at birth to 20/100 by 2 months and 20/20 by 4-6 months (Suchoff, 1979). Prior to the secondmonth of life, there is little evidence indicating selective attention. Following the second month oflife, human infants demonstrate clear fixation responses indicating that the perception of the infantis qualitatively similar to adults (Bond, 1972). Neonate are able to track slow linear movementsbut do not track fast rotary movements. This may indicate that while the visual-vestibular responseis functional very early after birth, it continues to develop postnatally.The eye blink response, a protective eye reflex that can be provoked by numerous stimuli, is ageneralized phenomenon in mammals (Esteban, 1999). In humans, the glabellar reflex refers to thebilateral eye blink elicited by gentle stimulation of the glabellar region (Zametkin et al., 1979).Following repeated stimulation, the glabellar eye blink response diminishes in adults. The mean© 2006 by Taylor & Francis Group, LLC

1110 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrate of spontaneous eye blink is less than 2 per minute in early infancy and increases steadily upuntil age 14 or 15 when adult levels of 10-16 blinks per minute are achieved (Zametkin et al., 1979).RatsSensorimotor behaviors (e.g., visual and tactile orientation) and righting reflexes develop in arostral-caudal pattern and achieve mature characteristics during the postnatal period (Almli andFisher, 1977). The most commonly evaluated reflexes during development of rats are righting(vestibular) and brainstem stimulus-response reflexes (e.g., auditory startle, pupillary light).The surface-righting reflex emerges soon after birth in the rat (PND1-3), and matures duringthe first week of development (Bignall, 1974; Smart and Dobbing, 1971). The organization of thisreflex emerges a few days postnatally in the low pons and medulla, and during development itbecomes dependent on, but not completely replaced by, the more rostral systems in the mescecephalon(Bignall, 1974). Soon after onset of the surface-righting reflex, rat pups achieve the ability toorient themselves upward on an inclined plane, a response known as negative geotaxis (achievedon PND 9-11) (Smart and Dobbing, 1971).The air-righting reflex, also known as free-fall righting, has been used extensively to assessreflex integrity in rodents (Cripps and Nash, 1983; Smart and Dobbing, 1971; Vorhees et al., 1994).In rats, and presumably other rodents, triggering of the reflex is not dependent on vision (Pellis etal., 1989) while in cats vision is required. The air righting reflex in rats can be detected as earlyas PND 9, and the response is fully present by PND 18 (Smart and Dobbing, 1971; Vorhees et al.,1994) with certain rat strains showing a fully functional air-righting response by PND 15 (Unis etal., 1991).Brainstem stimulus-response reflexes consist of motor responses to various sensory stimuli,such as smell, taste, touch or visual. Olfaction and taste are among the first sensory systems todevelop in rats with expression of olfactory receptor and transduction genes occurring just prior tobirth (Margalit and Lancet, 1993). While rats are likely born with a functional taste sense, theyshow increased sugar discrimination from birth through PND 15 and an adult-like salt preferenceby PND 21 (Hill and Mistretta, 1990).Development of nociceptive and tactile sensation varies based on the specific organ; however,in general the sensory nerve supply to the developing organ occurs immediately prior to partus inthe rat (Sisask et al., 1995). Visceral nociceptive receptors are absent at PND 5 and an adult-likeresponse does not develop until weaning (Bronstein-Goral et al., 1986). The tail flick response (aspinal reflex to heat stimulus on the tail) is present at PND 10 but does not mature to adult levelsuntil PND 25 (Ba and Seri, 1993). Vibrissa placing, a tactile response to a stimulus of the stiffhairs located about the nostrils that serve as a tactile organ, is developed by PND 12.5 in rats (Smartand Dobbing, 1971).At birth, rats are non-responsive to sound (Uziel and Marot, 1981). The acoustic or auditorystartle reflex (ASR) is a response elicited by an intense auditory stimulus. The ASR is functionalin rats by PND 12 (Smart and Dobbing, 1971). Inhibition of the ASR by brief auditory, cutaneousor visual stimulus (e.g., pre-pulse inhibition) is a brainstem-mediated mechanism (Davis andGendelman, 1977) that develops gradually beginning around the 3 rd postnatal week and continuesthrough the 6 th postnatal week, with tactile inhibition developing to an adult-like response at aslightly earlier age (Parisi and Ison, 1979; Parisi and Ison, 1981). The difference between the timesof onset of inhibition mediated by the different sensory modalities suggests that the shared centralmechanisms that modulate the ASR mature during the 3 rd week of postnatal development in rats.As with most mammals, the visual system is the last sensory modality to develop in rats(Gottlieb, 1971). Rats are born with their eyes closed, and eyelid opening does not occur untilapproximately PND 14-16 (Crozier and Pincus, 1937). However, phototaxic reactions are presentin rats as young as PND 5 (Crozier and Pincus, 1937; Routtenberg et al., 1979). Visual evokedpotentials can be obtained as early as PND 11 (Rose, 1968), and rat pups are able to perform a© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1111visual placing task by PND 17 (Smart and Dobbing, 1971). This suggests that although rats areborn with their eyes closed, the visual system of the rat may be at a developmental level roughlyequivalent to species born with their eyes open, including humans.DogsThe age when positive conditioned responses can first be seen is dependent on the modality of thestimulus, i.e., a conditioned response involving integration at a higher sensory level occurs later inlife than those at lower levels (e.g., rhinencephalic). The majority of the data available on theontogeny of sensory and reflexive development in dogs can be found in “Integrative Developmentof Brain and Behavior in the Dog” by Michael W. Fox (Fox, 1964a).Non-Human PrimatesA trend exists among primates for those species that reach higher levels of functioning to haveprolonged juvenile periods such that the duration for juvenile growth ranges from a period ofmonths in tree shrews to about 18 months in prosimians (e.g., slow loris), 2-3 years in monkeysand 8 years in great apes (Schultz, 1969).Primate infants demonstrate a rooting reflex at birth with a maximum response at PND 4, whichdisappears by PND 10 (Mowbray and Cadell, 1962). Other reflexes present at birth include handand foot grasping and clasping (Mowbray and Cadell, 1962). From birth, a rhesus monkey placeddorsal side down will quickly “right” by turning to achieve ventral contact (Hines, 1942). Squirrelmonkeys have the righting reflex during the first week post birth (Schusterman and Sjoberg, 1969),to orient up by 3 weeks of age (Elias, 1977) and to air-right at week 5 (King et al., 1974).Somatosensory systems are functional at birth; infant rhesus monkeys have been shown torespond to a pinprick at birth (Golub et al., 1991a). While these systems are functional at birth,they may still be refined during development. For example, macaques have been shown to discriminatetexture and size differences at the level of adult thresholds by postnatal week 10, indicatingthat tactile sensory systems, while functional at birth, continue to develop postnatally (Carlson,1984).In non-human primates, auditory responsiveness is observed at birth, indicating that a functionalauditory sensory system develops prenatally (Winter et al., 1973), and infant monkeys show a startleresponse to an abrupt sound on PND 10 (Mowbray and Cadell, 1962).Taste buds in the soft palate of the common marmoset increase in number after birth, reachinga maximum number at 2 months and then decrease until 9 years of age (Yamaguchi et al., 2001).Nonhuman primates can visually track their mothers at birth (Dolhinow and Murphy, 1982),can track a moving object within the first few days of birth (Mendelson, 21982), and orient to asmall object by PND 13 (Mowbray and Cadell, 1962). Optical quality (optical line spread function)for well-focused retinal images in nonhuman primates is very good at birth and improves rapidlyto adult levels by postnatal week 9 (Williams and Boothe, 1981). Visual acuity (argued to be afunction of maturation of neural elements) is low at birth and improves slowly during the firstpostnatal months (Neuringer et al., 1984) and reaches adult levels by about 1 year (Williams andBoothe, 1981).Sensory and Reflex ConclusionsThe sequence of development of sensory and reflex systems in rats, dogs and non-human primatescorrelate rather closely with those of the human infant. For example, the gustatory system innewborn animals is apparently important for selective ingestion, and olfaction and taste are amongthe first sensory systems to develop. Behavioral responses in humans and rats suggest the presenceof functional taste and olfaction systems at birth (Hudson and Distel, 1983; Teicher and Blass,© 2006 by Taylor & Francis Group, LLC

1112 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1977; Yamaguchi et al., 2001), and olfaction may even be functional in utero (Pedersen et al., 1983;Stickord et al., 1982). However, while the sequence of development of sensory and reflex systemsmay be similar, timing of development relative to birth varies for some systems. In general, rats,dogs and humans tend to develop slowly postnatally and have a clearly defined neonatal periodduring which they are dependent on the mother for shelter and food. In contrast, non-human primatestend to be more developed at birth. For example, in monkeys tactile, auditory and visual systemsare more developed at birth than in humans or rats. This likely reflects the need for rapid postnataldevelopment in non-human primates. Within 3 months of birth the infant macaque develops froma state of complete dependence on its mother to one of relative independence.Even though primates are precocial for reflexive and sensory behaviors, all of the commonlaboratory species are relatively developed at birth, with the exception of vision and taste. Therefore,evaluations of somatosensory development in juvenile toxicity studies are not likely to providesignificant insight because many of these systems are fully developed during gestation in humans(Table 3).Ontogeny of Cognitive DevelopmentThere are some learning abilities that appear extremely early in life in many species, e.g., classicalconditioning tasks. However, the development of more complex learning abilities (e.g., delayedresponseor spatial learning) is gradual and progressively improves with the developing animal.Learned responses are also instrumental in the ontogeny of social interactions across many species,including the development of play behavior and fear of strangers. In addition, normal sleep-wakerhythms also mature during the early periods of postnatal development.HumansGeneral UnderstandingIllingworth believed the first sign of understanding in a neonate can be observed in the first fewdays of life (Illingworth, 1983). A neonate will attentively watch her mother speak and will respondby opening and closing her mouth and/or bobbing her head up and down. By 4 to 6 weeks, aninfant will begin to smile, and about 2 weeks later will begin to vocalize. Between 12 and 16weeks, an infant opens her mouth when a bottle approaches, and by 24 weeks, stretches her armsout when she sees her mother is going to lift her. By 40 weeks, a toddler will pull her mother’sclothes to attract attention. Around 44 weeks, a toddler realizes she has to extend her arms in orderto put her coat on, and by 1 year may understand a phrase such as ‘where is your shoe?’ After the1 st birthday, her understanding is developing in numerous ways. A child can execute simple requests,has increased interest in toys and books, and is developing speech. By approximately 2 years ofage, she can put words together or form a simple sentence. The complexity of her language continuesto mature, and she will begin to count to 10 and name colors by 5 years of age.Specific Understanding: Learning & MemoryNeonates — A classical conditioning paradigm was administered to newborns by presenting themwith 50 conditioned-stimuli (CS) (i.e., a tone) that overlapped and terminated with an air-puff(Unconditioned Stimulus) which caused the neonate to blink (Little et al., 1984; Rovee-Collier andGerhardstein, 1997). Neonates were trained at 10, 20, or 30 days of age. Ten days later they receiveda retention session. The difference between the number of responses during the retention test andduring the initial training was the measure of retention. Significant retention was displayed inneonates trained at 20 and 30 days of age, but not those initially trained at 10-days of age.© 2006 by Taylor & Francis Group, LLC

Table 3Postnatal Sensory and Reflexive Development of Rat, Dog, Non-human Primate and Human (Approximate age at Demonstration of Characteristic)Human Rat Dog Non-human PrimateEating/Rooting Birth Birth Birth BirthSurface righting No Information Found PND 1–11 No Information Found BirthGrasping/clasping Birth–2 months No Information Found No Information Found BirthOrientation No Information Found PND 9–11 No Information Found 3 weeksAir righting No Information Found PND 15–18 No Information Found 5 weeksOlfaction No Information Found Birth Birth No Information FoundTasteBirthBirthBirthBirthDiscrimination Birth–4 yr Discrimination; Birth–PND 21 Discrimination Birth–PND14Tactile Birth Vibrissa placing: PND 12.5Nociceptive: PND 21Proprioceptive: PND1Air puff: PND 10Pinprick: BirthMature: 10 weeksTail flick: PND 10–25Nociceptive: PND 1–18Adult-like Nociceptive Response: PND 21Auditory Birth ASR: PND 12Inhibition of ASR: PND 21–42ASR: PND 19–25BirthASR: PND 14–25VisualVisual tracking: BirthPerception: mature in 2 monthsAcuity: mature in 4–6 monthsPhototaxis response: PND 5Eyes open: PND 14–16Visual placing: PND 17Visual orientation:PND 21–25ASR = Auditory Startle Response; PND = Postnatal Day; Mature: Age when animal demonstrates an adult like response.Visual tracking: BirthVisual orientation: PND 13Acuity: improves during first2 months of life; mature: 1yrPOSTNATAL DEVELOPMENTAL MILESTONES 1113© 2006 by Taylor & Francis Group, LLC

1114 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONInfants (1 month-2 years) — An operant conditioning task (i.e., a mobile conjugate reinforcementparadigm) has been used to assess “long-term” retention in infants 3 to 7 months of age(Rovee-Collier and Gerhardstein, 1997). In this task, the intensity of the reinforcement (mobilemovement) is proportional to the rate and vigor of the infant’s kick since a soft ribbon is connectedfrom the infant’s ankle to a mobile (acquisition) or an empty stand (nonreinforcement) (Rovee-Collier et al., 1978; Rovee and Rovee, 1969). All infants quickly learn that when the mobile isvisually present movement of their ankle will elicit a reward (i.e., a moving mobile) (Hill et al.,1988). However, the rate at which they forgot this association is different depending on the age ofthe infant during the initial learning of this task. Vander Linde et al. (Vander Linde et al., 1985)found that forgetting occurred within 3 days of training at 2 months of age. Sullivan et al. (Sullivanet al., 1979) found forgetting within 6-8 days of training at 3 months of age, and Hill et al. (Hillet al., 1988) observed a gradual forgetting over the first 2 weeks after training at 6 months of age.Learning and memory in infants between the ages of 6 and 18 months of age were evaluatedin a paradigm very similar to the mobile paradigm. Instead of their ankle moving a mobile, infantslearned to press a lever to move a miniature train around a circular track (Campos-de-Carvalho etal., 1993). Infants tested in this paradigm remember progressively longer between 6 and 18 monthsof age. Infants tested at 6 months of age gradually forgot the task over 2 weeks, and infants 18months of age remembered the association between the lever press and train movement for nearly12 weeks after the task was first administered.Fivuch and Hammond (Fivush and Hammond, 1989) engaged 24- and 28-month old toddlers in aseries of play events involving 4 different toy animals (zebra, duck, monkey and giraffe). One groupof children re-enacted the play event 2 weeks after the first session while another group did not. Allchildren were then tested for their retention 14 weeks later. Children who had repeated experiences inthe play event recalled significantly more items than children in the single experience group.Children/Juveniles (2 to 12 years) — Over the 1 st and 2 nd year of life, there is a steady increasein the length of the sequence that can be recalled accurately (Bauer, 1997). By 20 months of age,children were capable of reproducing three-act events (Bauer and Hertsgaard, 1993; Bauer et al.,1994; Bauer and Thal, 1990). At 24 months, children can reliably reproduce events five steps inlength (Bauer and Travis, 1993). And by 30 months, they can reliably reproduce an event involvingas many as 8 steps, such as “building a house” (Bauer and Fivush, 1992).There are few studies that have evaluated long-term recall in infancy and early childhood. Onestudy by Bauer et al. (Bauer et al., 1994) tested children 21, 24 and 29 months of age for theirrecall of a novel multi-step event sequence that they had experienced 8 months earlier, as 13-, 16-and 20-month-olds. The performance of the experienced children was significantly better comparedto age- and gender-matched control children who had never experienced the test event. These resultssuggest that even very young children can maintain organized representations of novel events overlong periods of time. Even though the long-term memory is present at 1 year of age, it does notimply that it is fully mature.Fivush (Fivush, 1991) tested the recall ability of children who visited Disney World around 3or 4 years of age. Half of the subjects were interviewed 6 months after the trip, the other half 12months later. All children recounted a great amount of information about their Disney Worldexperience, but there was no difference between retention interval groups. However, it was notedthat the older children recalled their trip with more detailed information, and the recall was morerapid than the younger children’s recall.The ontogenetic development of explicit (declarative) and implicit (procedural) learning werestudied in children between 5-10 years of age and compared to young adults (23 years of age)(Homberg et al., 1993). Explicit memory is the ability to recall words, dates, visual images or lifeevents;whereas, implicit memory is associated with a variety of nonconscious abilities, includingthe capacity for learning habits and skills and some forms of classical conditioning (Zola-Morganand Squire, 1993). Two tasks, story recall and pictorial memory, were used to assess explicit memory© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1115(Homberg et al., 1993). In the story recall task, individuals were read a story that comprised 11significant items. In the pictorial memory task, the subjects were presented with a series of 6pictures. In both explicit memory tasks, the number of recalled items gradually increased with age.Adult performance was not seen before 9 years of age in the picture recall task. Performance inthe story recall task was superior in the young adults compared to the oldest children tested (11years age). There were no differences in either implicit learning task between young adults andchildren (not described). These results demonstrated that implicit and explicit memory followdifferent maturational courses.The first signs of understanding can be observed within the first few days of life, and infantsvery quickly learn to associate certain responses with learned behaviors. At all ages, humans havememories that forget gradually, but the rate at which they forget becomes slower with increasingage. In addition, as age increases the ability to recall more complex sequences and detailedinformation continues to improve.RatsThe learning capabilities of the newborn rat are impressive. For example, Johanson and Hall(Johanson and Hall, 1979) demonstrated appetitive learning in 1-day-old rat pups. Specifically, 1-day-old pups were capable of discriminating between two paddles on the basis of odor and positionwhen they were provided milk as a reward for the correct choice. Rudy and Cheatle (Rudy andCheatle, 1977) demonstrated that when 2-day-old rats were exposed to a novel odor and injectedwith an illness-inducing drug, i.e., lithium chloride, these same pups, when retested at 8-days ofage, avoided the odor, whereas control pups did not. These data suggest rat pups are capable ofassociative learning at a very early age.Several investigators have suggested that immature animals have an enhanced pattern of forgettingcompared to adults (Campbell and Spear, 1972; Coulter et al., 1976). Neonatal rats 7, 9and 12 days of age were tested in a nondirectional active-avoidance task (Spear and Smith, 1978).Animals in the 12-day old group required fewer trials to reached criterion performance during thelearning phase and had superior retention 24-hrs later compared to the 9-day old rats. The 7-dayold rats never achieved criterion performance (5 out 6 avoidances) during the learning phase.Reactivation treatment (i.e., presentation of unconditioned stimulus (footshock) shortly before theretention test) was found to enhance retention over the 24-hr period for 12-day old rats, but onlysomewhat for 9-day old rats. The reactivation treatment seemed completely ineffective for rats 7days of age. These experiments supported the authors’ hypothesis that substantial deficits in younganimals are due, in part, to a deficiency in memory retrieval.The ontogeny of step-down passive avoidance was studied in Wistar rats 2-, 3-, 4-, 6-, 8- and13-weeks of age (Myslivecek and Hassmannova, 1991). During the retention trial, the step-downlatency was longest in 3-week old rats, but the most efficient learning was seen in rats 6 weeks ofage. In addition, passive avoidance learning was present in rats 10 days of age, but retentionimproved considerably in rats 15 days of age (Stehouwer and Campbell, 1980). These data suggestthat passive avoidance learning exists in early postnatal life, but the maturity of the storage and/orretrieval systems improves with age in the rat.Biel (Biel, 1940) trained white rats at various ages in a multiple T-maze, i.e., the Biel maze.Group A, B, C and D rats were trained in the Biel maze between the ages of 17 - 23, 20 - 26, 23– 29, and 30 - 36 days of age, respectively. His results showed that Group D rats were superior toand made fewer errors than Group C; Group C was better than Group B; and in turn, Group B wasmore proficient than Group A. In addition, Group A never reached the level of accuracy, i.e., asfew errors, as the other groups, and Group D learned the task most rapidly. Furthermore, the averagetime it took to complete the maze was shorter with the increasing age of the groups. The resultsof this study indicated that there are maturational differences that influence a rat’s ability to performthe Biel maze, and there is an improvement in maze performance with age.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1117every time the dog made 4 consecutive correct choices. Puppies 4- and 8-weeks of age made moreincorrect choices (of both the neutral and incorrect box) than the 12-week puppies. The 16-weekold puppies had an unexpected increase in the number of errors to the incorrect box. However,both the 12- and 16-week groups were capable of increasingly longer delayed responses. Aftersuccessfully achieving a delay time of 15 seconds, the 12-week group continued to increase theirdelay in a linear fashion (i.e., a 50 sec delay was possible on day 13 of testing), whereas the 16-week puppies were much slower (i.e., only a 35 sec delay was possible on day 13 of testing). Foxnoted that the apparent inferior abilities of the puppies at 16 weeks compared to 12 weeks mightbe due to the development of inhibition reflexes that appear around 16 weeks of age. This experimentdemonstrated that puppies 12- and 16-weeks of age can proficiently use both long-term memory(learned food-motor approach pattern) and short-term memory (remembering for a short time whichbox contained the reward) to solve a delayed response task. It is likely that the learning capabilitiesof dogs continue to improve with age based on the fact that in adult dogs the delayed responsetask can successfully be solved with as much as 120-sec delays.A spatial delayed nonmatching-to-sample task was evaluated in mongrel dogs divided into threegroups: 5 young (8 months-2.5 years), 3 middle-aged (5-8 years), and 5 aged dogs (Ω10 years)(Head et al., 1995). All ages performed equally well when a short delay was present, i.e., 20 seconds,but the young dogs had fewer errors compared to the older dogs when the delay was increased,i.e., 70 and 110 second delays. Identical results were seen by Milgram et al. (Milgram et al., 1994)in a delayed-nonmatching-to-sample paradigm using beagle dogs. In contrast, Fox (Fox, 1971)reported poorer performance of very young dogs relative to adults in a spatial delayed-responsetask. These studies emphasize the importance of selecting an appropriately aged animal whenassessing the impact of a compound on early cognitive development. It also points out the significantmaturation the CNS undergoes during early postnatal development in the canine.Non-Human PrimatesHarlow (Harlow, 1959) evaluated the development of learning in the rhesus monkey. A conditionedavoidance task was initiated on the third day of life. Infants showed improvement (i.e., learning)over the 5 consecutive days of testing. Interestingly, when a retention test was administered at 23days of age, it revealed considerable learning loss ranging from no indication of retention toconditioned responses on half of the test trials.Two groups of newborn macaques (one group started training at birth and the other on day 11)were trained on a black-white discrimination task, i.e., object discrimination. Learning of this taskby the neonatal macaques was rapid. In the group trained from birth, an accuracy of 90% wasattained beginning at 9 days of age. The second group was rewarded for climbing up either of twogray ramps for the first ten days of life. Then on day 11, the black and white ramps were substituted.By the second test day (i.e., day 12), these monkeys could solve the discrimination problem.Interestingly, during the reversal discrimination task, the infants made many errors when theproblem was first reversed and showed severe emotional disturbances (e.g., balking, vocalization,urination, and defecation).Even though 11-day old monkeys can solve a black-white discrimination task, the ability tosolve a triangle-circle discrimination problem requires greater CNS maturation. Monkeys 20 to 30days of age may require from 150 to 200 trials to solve a triangle-circle discrimination task. Thedevelopment of more complex object discrimination (i.e., three dimensional objects differing incolor, form, size and material) was evaluated in rhesus monkeys 30, 60, 90, 120, 150 and 366 daysof age. The animals were given 25 trials a day, 5-days a week for 4 weeks. The number of trialsrequired to have 10 consecutive correct responses decreased with increasing age and approachedadult performance around 120 to 150 days of age.In a delayed-response task, an animal was first shown a food reward, which was then concealedunder one of two identical containers during the delay period. The rhesus monkeys began testing© 2006 by Taylor & Francis Group, LLC

1118 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONat 60, 90, 120, or 150 of age. Over time, all four groups demonstrated an increasing ability to solvethe delayed response task. However, Harlow did note that the capacity to solve delayed-responseproblems was better in animals 125 to 135 days of age compared to 60-90 days of age. In addition,half of the monkeys at 200 to 250 days of age had essentially flawless performance in this task. Itis clear that the capacity to solve delayed-response problems matures later than the capacity tosolve object-discrimination tasks.Two broad conclusions can be derived from the monkey data presented thus far. First, certainlearning abilities appear extremely early in life in the rhesus monkey, e.g., conditioned avoidanceor black-white discrimination. The ability to solve a problem may be present as early as 1 day ofage, and within a matter of days, performance will change from chance levels to nearly 100%accuracy. Second, the development of certain, more complex, learning abilities is more gradual andprogressively improves, e.g., delay-response learning and some discrimination learning tasks (e.g.,color, pattern or shape discrimination). Interestedly, extensive early training on complex tasks (e.g.,delayed response task) did not facilitate the solution, but rather success was clearly a function ofage, i.e., even when young animals were trained more extensively than the older animals, theirperformance did not equal that of older monkeys (Zimmerman and Torrey, 1965).Conclusions for Cognitive DevelopmentThere is evidence that rats, dogs, monkeys and human infants have learning capacities that changeduring development. In all species evaluated, the capacity to mediate reflexive behaviors emergedbefore learned behavioral reactions. Vogt and Rudy (Vogt and Rudy, 1984) suggested that thesequence of behavioral development reflects a sequential caudal to rostral maturation of components.That is, the lower centers of a sensory system are sufficient to support primitive reflexivebehaviors, and it is well established that subcortical structures are sufficient to support manyinstances of Pavlovian conditioning (Girden et al., 1936; LeDoux et al., 1984). However, highercenters of learning (e.g., hippocampus and prefrontal cortex) are necessary for the acquisition ofmore complex behaviors.Therefore, in all common laboratory species, evaluations of cognitive function during juveniletoxicity studies can provide insight into the ontogeny of the maturation of these systems. The timerequired to learn a particular task is often greater in neonates and infants compared to adults. Inaddition, the accuracy of learning that task is often not equal to an adult. This implies that eventhough a task can be completed at a fairly young age, it is by no means functionally mature, i.e.,at adult levels of performance. Finally, it is important to remember that learning and memory areintimately associated with the maturation of many of the sensory systems described above (Table 4).Ontogeny of CommunicationEarly in postnatal development, many species use some form of communication to express fundamentalneeds (e.g., hunger and fear). However, the complexity of the communication often undergoesa period of prolonged postnatal development.HumansInfants have many methods of preverbal communication, e.g., crying, smiling, clinging, and kissing(Illingworth, 1983; Morley, 1972; Murray and Murray, 1980; Nelson, 2000). At about 2 months,the infant listens to voices and coos in response, and around 3 months, she begins to make sounds(e.g., “aah”, “ngah”). At 4 months, laughing out loud occurs, and the infant may show displeasureif social contact is broken or excitement at the sight of food. At 8 months, preference for the motheremerges, the infant begins to babble with new complexity and inflections that mimic the nativelanguage. Around 9 to 10 months a critical step occurs in speech development, an infant begins to© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1119Table 4Learning TaskSummary of Cognitive DevelopmentHumansClassical Conditioning (air puff)Operant Conditioning (task-dependent)Long-term recallImplicit LearningExplicit LearningRatsAppetitive LearningAssociative Learning (Olfactory)Active AvoidancePassive AvoidanceSequential Learning (Biel Maze)Proximal Cued Learning (Morris Water Maze)Spatial Learning (Morris Water Maze)DogsAssociative Learning (Olfactory)Visual DiscriminationDelayed Response TaskNon-Human PrimatesConditioned AvoidanceBlack-White Discrimination TaskColor, Pattern or Shape DiscriminationDelayed-Response TaskApproximate Postnatal Age Ability Emerges20–30 days old2 months–18 months≥13 months–20 months

1120 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONvocalizations predominate during the first 6 months of life and were uttered as single syllables orphrases. After 6 months of age, the number of distinct monosyllabic and bisyllabic sounds increased,and the sounds varied in pitch, timbre and inflection. Sounds of similar pitch were combineddifferently so that the phrase changed in response to definite situations in the laboratory. The numberof phrases increased to 4 or 5 during the first year, but never to 7 or 8 like was heard in the adult.Conclusions for the Development of CommunicationEvaluations of the development of communication in juvenile toxicity studies with rats and dogs(and to some degree in non-human primates) are not likely to provide significant insight becausecommunication in these species appears to function primarily as a means of expressing distress.In humans, the development of speech and language is a critical component in CNS maturation.Table 5CommunicationSummary of CommunicationHuman“Coos” in response to voicesBabble with inflectionImitates soundCan understand or say 2 or 3 words with meaningBegins to join words togetherRatsUltrasonic vocalizationsUSV (maximum)DogDistress vocalizationsWhining vocalizationsNon-Human PrimateEmotional vocalizations(1 or 2 syllables)Emotional vocalizations(3, 4 or 5 syllables)Approximate PostnatalAge Ability Emerges2 months8 months9–10 months12 months21–24 monthsPND 7–9PND 133–3.5 weeks4–5 weeks0–6 months6 months–1 yearOntogeny of Social PlaySocial play is one of the means by which an individual obtains information about his socialenvironment. By playing with peers, the individual learns norms of social communication, socialdominance, and social integration in a context where errors in performance usually have neitheraversive nor lethal consequences.HumansAt approximately 40 weeks of age, babies will begin to play simple games such as peek-a-boo andpat-a-cake; and beginning around their 1 st birthday, they will play simple ball games (Nelson, 2000).Around 2.5 years of age, children begin to pretend play. During the third year of life, childrendevelop a greater interest in others and will play simple games in parallel with other children.Around 4 years of age, children will play with several children and show the beginnings of socialinteractions, including role-playing. During the elementary school years (5-12), continued peergroup relationships are essential (Kaye et al., 1988). These relationships help the child develop thecapacity for socially acceptable aggressiveness, group cooperation and self-identity. During theearly teenage years, the adolescent may appear to behave with no rhyme or reason as she searches forindependence from her parents while looking for peers groups to achieve a sense of belonging.Closeness to the peer group is achieved by conformity of dress and frequent telephone communications.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1121Around 15 to 20 years of age, the adolescent begins to solidify her own ideals and values. She alsobegins to function as an independent thinking individual with a reasonable state of equilibriumamong instinctual drives, self-needs and the outside world.RatsIn rats, the earliest forms of social interactions are at birth with the competition to nurse and clustertogether to sleep (Bolles and Woods, 1964). After PND 14, a new interaction begins to emergewhere one pup may appear to spontaneously jump, and a chain reaction occurs involving all thelittermates. Panksepp (Panksepp, 1981) evaluated the development of play behavior in young ratsby observing the frequency and duration of pinning behavior. Play increased from 18-28 days,peaked during 32-40 days of age, and then gradually declined. He also found that social isolationof rats resulted in a marked increase in play behavior. One function of play may be to establishstable social relationships and help rats find their place in the existing social structure of a group.Grooming in rats may be thought of as a social behavior because it results in the intimateinteraction of siblings. Grooming consists primarily of three different kinds of behavior (Bollesand Woods, 1964). Face-washing and scratching in rudimentary forms can be observed as early asPND 2. By PND 10-12 both face-washing and scratching resemble the adult pattern and includethe characteristic licking of the paw. Around PND 13, the third component, fur-licking, begins. Inaddition, grooming of its littermates also begins around PND 13 and is quite common in PND 40rats. However, sibling grooming declines somewhat in adulthood. The younger animals spend aconsiderable amount of time washing their front paws and face and scratching with the hindpaws.However, licking the fur is the prevalent component of grooming in the adult. Grooming is theonly activity other than eating and sleeping in which young animals will persist at for 5 or moreminutes.DogsAround 3 to 3.5 weeks of age, pups begin the earliest forms of playful social interactions (Fox,1971). They will chew on each other’s ears and lick faces of one another, but will either avoid orelicit a distress vocalization if the stimulation becomes too intense. Between 4 to 5 weeks of age,play-fighting, scruff-holding and “prey-killing” (i.e., head shaking movements) appear, along withpouncing, snapping and aggressive vocalizations (e.g., growling and snarling with teeth bared)(James, 1955; James, 1961). Submissive postures and play-soliciting gestures (e.g., ears pricked,tail wagging, vocalization) also appear around this time. By 4-months of age, dominant-subordinaterelations are established between littermates (Fox, 1971; Pawlowski and Scott, 1956; Scott andMarston, 1948). In general, male pups also indulge in more rough-and-tumble play and biting thanfemales, but there are differences in aggressiveness depending on the breed of dog (Fuller andDuBuis, 1962).Non-Human PrimatesPlay behavior in the macaque has two forms: movement play and social play (Hines, 1942). Assoon as the infant monkey is able to move easily (i.e., about 3 weeks of age), she will repeatmovements such as jumping, running and climbing. Around 4 weeks of age, bouncing on a wiremesh is possible. Social play begins around 5 weeks of age and is described as dodging and chasinggames with older monkeys that were allowed to run loose in the room. Around 6 weeks of age,the infant monkey begins stealing food from another monkey’s cage in spite of the presence of anabundance of food in her own cage. Stealing food from another monkey’s cage is a well-establishedbehavior in the adult monkey. Around 7 to 9 weeks, infants will approach each other with armsextended and then lightly slap each other. This appeared to be a signal to initiate rough-and-tumble© 2006 by Taylor & Francis Group, LLC

1122 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrolling and biting. In infant rhesus monkeys, males generally indulge in more vigorous rough-andtumblewrestling and sham biting than do females (Harlow, 1962).Conclusions for Social Play DevelopmentEvaluations of the development of social play in juvenile toxicity studies may provide some insightbecause it appears to develop postnatally in all common laboratory species. However, evaluationsof the development of the social play would not be straightforward in routine screening studiessince methodologies across species are not easily comparable.Table 6Social PlaySummary of Social PlayApproximate Postnatal AgeAbility Begins to EmergesHuman40 weeks–1 yearRat PND 14–28Dog3–5 weeksNon-Human Primate 5–9 weeksDevelopment of Fear ResponseThe development of a fear response to a novel environment or stranger is found in many species,including rats, dogs, rhesus monkeys, and human infants.HumansTwo of the most prominent events that provoke fear in infants are the approach of unfamiliar peopleand temporary separation from the caregiver. Distress to both these events, even in different cultures,emerges between 7-9 months (Kaye et al., 1988; Nelson, 2000). While infants do not cry everytime they encounter a stranger or experience separation, the more unfamiliar the stranger or thecontext, the more likely behavioral signs of fear will occur. Since 3-month olds can discriminatea stranger from a parent with no fear of strangers, improvement in working memory at 7-9 monthsis thought to play a role in these reactions.The maturation of the dorsolateral prefrontal cortex and the hippocampus, as well as theintegration of limbic structures and the cortical network, are considered important in the developmentof the fear response process (Herschkowitz et al., 1997). Not surprisingly in humans, theamygdala is thought to participate in fear reactions to unfamiliarity like non-human primates. Theamygdala is linked reciprocally to the cerebral cortex by axons within the capsula interna, whichdevelops mature myelin at 10 months in human infants, a time when stranger and separation fearsare prominent (Herschkowitz et al., 1997).RatsIt was found that fearful behavior in response to strange situations begins to increase in rats around20 and 30 days of age (Candland and Campbell, 1962).DogsAt approximately 3 weeks of age (i.e., beginning of the socialization stage), young puppies willbegin to bark or whine when placed in strange places. Elliot and Scott (Elliot and Scott, 1961)showed that this reaction to isolation peaks around 6 to 7 weeks of age, and declines thereafter.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1123Puppies that had only social contact to other dogs during the critical periods of socialization (i.e.,5-7 weeks of age) exhibited extreme fear when introduced to human contact (Freedman et al., 1961).Non-Human PrimatesAround 2.5-3 months of age, but not before, monkeys separated from their mothers become agitated,emit frequent distress vocalizations, and show a distinct increase in adrenocorticotropic hormoneresponse (Harlow, 1959). Rhesus infants also begin to demonstrate fear reactions to threateningpictures, strange objects, and novel situations around this age. If a human enters a room with cagedmonkeys, the animals become silent and freeze, which are both signs of fear to unfamiliarity.Rhesus infants raised with an inanimate object do not show signs of fear to a novel location untilthey are 3-4 months old (Harlow, 1959).Much work has been done to investigate the CNS systems involved in development of fear. Thefreezing response appears to be influenced by gamma-aminobutyric acid systems (Herschkowitz et al.,1997), while the distress cries are influenced by opiate systems. Both the anterior cingulated cortexand the amygdala, which have high densities of opiate receptors, participate in the production of distresscalls. Bilateral ablation of the amygdala at 2.5 months produced a significant decrease in fear reactionsto novel situations. Control animals who showed fear to an unfamiliar place took three times longerthan lesioned animals to leave a familiar cage and enter an unfamiliar one. Amygdalectomized monkeysalso showed significantly fewer fear reactions to pictures of monkey faces expressing fear.The amygdala has been implicated in the mediation of emotional and species-specific socialbehavior. The amygdala in adults is involved in determining whether an object or organism ispotentially dangerous. If danger is detected, it coordinates a variety of other brain regions to producea species-typical response to avoid the danger. Therefore, the amygdala contributes to learningwhat is dangerous in the environment. Selective amygdala lesions in 2-week old macaques resultedin less fear of novel objects (e.g., rubber snakes) when the animals were tested at 6-8 months ofage. However, the lesioned animals displayed substantially more fear behavior than controls duringsocial interactions. These results suggest that neonatal amygdala lesions dissociate a system thatmediates social fear from one that mediates fear of inanimate objects. Thus, amygdala lesions earlyin development have different effects on social behavior than lesions produced in adulthood, whichdecrease fear of inanimate objects and increase fear during social behavior (Prather et al., 2001).Conclusions for Fear Response DevelopmentEvaluations of the development of a fear response in juvenile toxicity studies may provide someinsight because it appears to develop postnatally in all common laboratory species. However,methods to evaluate the development of the fear response are not widely applicable to screeningtoxicity studies.Table 7Fear ResponseSummary of Fear ResponseHuman7–9 monthsRat PND 20–30Dog3–7 weeksNon-Human Primate 2.5–4 monthsApproximate Postnatal AgeAbility Begins to EmergesDevelopment of Sleep CyclesThe sleep-wake cycle is developed postnatally in many species and is thought to be critical in thedevelopment of cognitive abilities.© 2006 by Taylor & Francis Group, LLC

1124 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONHumansIn humans, the periodicity of wakefulness and sleep is observed by between 10-11 weeks of age(de Roquefeuil et al., 1993; Hellbrügge, 1960). Between the ages of 18 to 21 weeks, the longestduration of sleep occurred between 11 pm and 5 am (Kleitman and Engelmann, 1953) (Moore andUcko, 1957). De Roquefeuil et al. (de Roquefeuil et al., 1993) examined the sleep-wake rhythmsof 12 children admitted to a day-care at 4-5 months of age until approximately 15 months of age.The circadian rhythm of wakefulness and sleep is well established by 4 months of age. In this samestudy, the results indicated that at 7, 11 and 15 months of age the number of sleep spans continuedto decrease in a 24-hr period (4, 3 and 2, respectively), and the duration of an individual sleep spanprogressively increased (200, 210 and 260 minutes, respectively).RatsUp until PND12, nursing and burrowing for preferred sleeping and nursing positions are thedominant activities in rats (Bolles and Woods, 1964). Any movement by one pup is likely to disturbthe entire litter. Gradually after PND 12, activities such as exploring and grooming occur withincreasing frequency. By PND 15, some rats prefer to bury their heads under the litter to sleepwhile others prefer to climb on top. By PND 19, sleeping becomes quieter with less twitching andmoving around. By PND 27, an individual rat can sleep soundly while littermates are playing ontop of him. Interestingly during the first 3 weeks of life, rat pups do not exhibit the normal diurnalsleep cycle of adult animals, i.e., sleep during daylight, and activity at night (Bolles and Woods,1964). Alföldi et al. (Alföldi et al., 1990) demonstrated that by PND 23 waking activities predominatedduring the 12-h dark period.DogsThe development of electrocortical activity as measured by EEG in the dog has been well describedby Fox (Fox, 1971). The results of the EEG analysis indicated changes in behavioral state (e.g.,sleep, alertness) were not accompanied by significant changes in the EEG during the first twoweeks of life. Up until 7 days of age, sleep in the puppy consisted of mainly body jerks and twitchesincluding occasional vocalizations, sucking, crawling and scratching movements (which couldeasily be mistaken for an alert state). After the second week of life, slow-wave activity began toemerge and was associated with the development of quiet sleep. Three weeks of age marks atransitional point when the relative durations of wakefulness (or attentiveness) and activated sleepare similar, and the duration of quiet sleep is increasing. The onset of the mature EEG occursbetween 3 and 4 weeks of age, which also marks a critical period of socialization (Scott, 1962).By 4-5 weeks of age, the EEG has developed the full range of activity that is present in the adult.Non-Human PrimatesInfant monkeys have a less well-defined sleep/wake pattern than adult animals. By 2 to 8 days ofage, neonatal monkeys already tend to sleep more frequently at night than during the day (Bowmanet al., 1970).Conclusions for Sleep-Wake Cycle DevelopmentWhile development of adult-like sleep-wake cycles occurs postnatally, utilizing sleep-wake cyclesas a measure in juvenile toxicity studies will provide limited insight because its development isnot well understood.© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1125Table 8“Semi-Adult-like”Sleep CycleSummary of Sleep-Wake CycleApproximate Postnatal AgeAbility Begins to EmergesHuman3–15 monthsRat PND 15–23Dog2–4 weeksNon-Human Primate 2–8 daysOverall ConclusionsThere is no species that provides a direct correlation to human neurological development for allfunctional measures discussed. However, when development is tracked within particular functionaldomains (e.g., sensory or locomotor development), the common laboratory species are born insimilar states of CNS development as the human, and there is an extended period of postnatal lifewhere these species develop similar to humans. Both the rat and dog are appropriate models forcomparative CNS development, with neither showing a clear advantage. Non-human primates aremore advanced at birth for several of the most commonly measured endpoints of CNS development.Because of their advanced cognitive evolution monkeys are more similar to humans in cognitivedevelopment. However, for most applications, the rat is likely to be the preferred species forevaluation of CNS function in juvenile toxicity testing, but should not be chosen as the defaultspecies for all studies. It is important to select the species that most appropriately correlates tohumans within the functional domain of concern.Additional Research NeedsThe ontogeny of memory is less well understood than adult memory systems for several reasons.First, unlike adults, infants do not typically suffer discrete brain lesions, which have helpedresearchers make associations between different types of memory and specific brain regions.Second, cognitive and linguistic limitations in infants provide additional challenges, which makestudies of the development of early memory challenging.Also, there is a strong need have additional reviews on postnatal development of the CNS whichinclude both neuroanatomical and neurochemical maturation. Further research on the ontogeny ofneurological development in all species is necessary in order to design better cognitive assessmentsof juvenile toxicity.REFERENCESAdams-Curtis LE, and Fragaszy DM (1994) Development of manipulation in capuchin monkeys during thefirst 6 months. Developmental Psychobiology 27: 123-136Adams J (1986) Methods of Behavioral Teratology. In: Riley EP, Vorhees CV (eds) Handbook of BehavioralTeratology. Plenum Press, New York, pp 67-97Alberts JA (1984) Sensory-perceptual development in the Norway rat: A view toward comparative studies.In: R.Kail, Spear NE (eds) Comparative Perspectives on the Development of Memory. Erlbaum,Hillsdale, NJ, pp 65-101Alföldi P, Tobler, and Borbely A (1990) Sleep regulation in rats during early development. American Journalof Physiology 258: R634-R644Almli CR, and Fisher RS (1977) Infant rats: sensorimotor ontogeny and effects of substantia nigra destruction.Brain Res. Bull. 2: 425-459Altman J, and Sudarshan K (1975) Postnatal development of locomotion in the laboratory rat. Anim. Behav.23: 896-920© 2006 by Taylor & Francis Group, LLC

1126 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONBa A, and Seri BV (1993) Functional development of central nervous system in the rat: ontogeny of nociceptivethresholds. Physiology & Behavior 54: 403-405Bauer PJ (1997) Development of memory in early childhood. In: Cowan N (ed) The Development of Memoryin Childhood. Psychology Press, pp 83-111Bauer PJ, and Fivush R (1992) Constructing event representations: Building on a foundation of variation andenabling relations. Cognitive Development 7: 381-401Bauer PJ, and Hertsgaard LA (1993) Increasing steps in recall of events: factors facilitating immediate andlong-term memory in 13.5- and 16.5-month-old children. Child Development 64: 1204-1223Bauer PJ, Hertsgaard LA, and Dow GA (1994) After 8 months have passed: long-term recall of events by 1-to 2-year-old children. Memory 2: 353-382Bauer PJ, and Thal DJ (1990) Scripts or scraps: reconsidering the development of sequential understanding.Journal of Experimental Child Psychology 50: 287-304Bauer PJ, and Travis LL (1993) The fabric of an event: Different sources of temporal invariance differentiallyaffect 24-month-olds’ recall. Cognitive Development 8: 319-341Bekoff A, and Trainer W (1979) The development of interlimb co-ordination during swimming in postnatalrats. The Journal of Experimental Biology 83: 1-11Biel WC (1940) Early Age Differences in Maze Performance in the Albino Rat. The Journal of GeneticPsychology 56: 439-453Bignall KE (1974) Ontogeny of levels of neural organization: the righting reflex as a model. Exp. Neurol. 42:566-573Bolles RC, and Woods PJ (1964) The ontogeny of behavior in the albino rat. Anim. Behav. XII: 427-441Bond E (1972) Perception of form by the human infant. Psychol Bull 77: 225-245Boothe RG, Dobson V, and Teller DY (1985) Postnatal development of vision in humans and nonhumanprimates. Ann. Rev. Neurosci. 8: 495-545Bower T (1979) Human Development. W.H. Freeman & Co., New York,Bowman RE, Wolf RC, and Sackett GP (1970) Circadian rhythms of plasma 17-hydroxycorticosteroids in theinfant monkey. Proceedings of the Society for Experimental Biology and Medicine 133: 342-344Bronstein-Goral H, Mitteldorf P, Sadeghi MM, Kirby K, and Lytle LD (1986) Visceral nociception in developingrats. Developmental Psychobiology 19: 473-487Butcher RE, and Vorhees CV (1979) A preliminary test battery for the investigation of behavioral teratologyof selected psychotropic drugs. Neurobehav. Toxicol. 1 Suppl 1: 207-212Campbell BA, and Spear NE (1972) Ontogeny of memory. Psychological Review 79: 215-236Campos-de-Carvalho M, Bhatt R, Wondoloski T, Klein P, and Rovee-Collier C (1993) Learning and Memoryat nine months. Paper presented at the meeting of the International Society for the Study of BehavioralDevelopment, Recife, BrazilCandland, D. K. and Campbell, B. A. Private Communication . 1962.Carden SE, and Hofer MA (1992) Effect of a social companion on the ultrasonic vocalizations and contactresponses of 3-day-old rat pups. Behavioral Neuroscience 106: 421-426Carlson M (1984) Development of tactile discrimination capacity in Macaca mulatta. Brain Res. 318: 69-82Carlson SE (2000) Behavioral methods used in the study of long-chain polyunsaturated fatty acid nutritionin primate infants. [Review] [45 refs]. Am. J. Clin. Nutr. 71: 268S-274SConnolly BH (1985) Neonatal Assessment: an overview. Physical Therapy 65: 1505-1513Coulter X, Collier AC, and Campbell BA (1976) Long-term retention of early Pavlovian fear conditioning ininfant rats. J Exp Psychol Anim Behav Process 2: 48-56Cripps MM, and Nash DJ (1983) Ontogeny and adult behavior of mice with congenital neural tube defects.Behavioral & Neural Biology 38: 127-132Crozier WJ, and Pincus G (1937) Photic stimulation of young rats. J Gen Psychol 17: 105-111Davis M, and Gendelman PM (1977) Plasticity of the acoustic startle reflex in acutely decerebate rats. J. Comp.Physiol. Psychol. 91: 549-563de Roquefeuil G, Djakovic M, and Montagner H (1993) New Data on the Ontogeny of the Child’s Sleep-Wake Rhythm. Chronobiology International 10: 43-53Department of Health and Human Services (FDA) (1998) 21 CFR Parts 201, 312, 314 and 601. FederalRegister 63: 66632-66672Desor JA, Maller O, and Andrews K (1975) Ingestive responses of human newborns to salty, sour and bitterstimuli. J. Comp. Physiol. Psychol. 89: 966-970© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1127Dolhinow P, and Murphy G (1982) Langur monkey (Presbytis entellus) development: the first 3 months oflife. Folia Primatol 39: 305-331Eisenberg R (1970) Development of hearing in man: an assessment of current status. J. Amer. Speech Hear.Assoc. 12: 199-223Elias MF (1977) Relative maturity of cebus and squirrel monkeys at birth and during infancy. DevelopmentalPsychobiology 10: 519-528Elliot O, and Scott JP (1961) Journal of Genetic Psychology 99:Esteban A (1999) A neurophysiological approach to brainstem reflexes. Blink reflex. [Review] [162 refs].Neurophysiologie Clinique 29: 7-38FDA. FDA, Federal Register Part II, Vol 63, No. 231. Wednesday, December 2, 1998; Rules and Regulations.12-2-1998.Fivush R (1991) The social construction of personal narratives. Merrill-Palmer Quarterly 37: 59-81Fivush R, and Hammond NR (1989) Time and Again: Effects of repetition and retention interval on 2-yearolds’event recall. Journal of Experimental Child Psychology 47: 259-273Foley Jr. JP (1934) First year development of a rhesus monkey, reared in isolation. Journal of GeneticPsychology 45: 39-105Forssberg H, Eliasson AC, Kinoshita H, Westling G, and Johansson RS (1995) Development of human precisiongrip. IV. Tactile adaptation of isometric finger forces to the frictional condition. Experimental BrainResearch 104: 323-330Fox MW (1964a) The ontogeny of behaviour and neurologic responses in the dog. In: Worden AN (ed) AnimalBehaviour. Bailliere, Tindall and Cox, Ltd., London, pp 301-310Fox MW (1964b) The Postnatal Growth of the Canine Brain and Correlated Anatomical and BehavioralChanges during Neuro-ontogenesis. Growth 28: 135-141Fox MW (1966) Neuro-behavioral ontogeny. A Synthesis of ethological and neurophysiological concepts.Brain Res. 2: 3-20Fox MW (1971) Integrative Development of Brain and Behavior in the Dog. The University of Chicago Press,Chicago,Freedman DG, King JA, and Elliot O (1961) Science 133:Fuller JL, DuBuis EM (1962) The behaviour of dogs. In: Hafez ESE (ed) The Behavior of DomesticatedAnimals. Bailliere, Tindall and Cox, London, pp 415-452Geyer MA, and Reiter LW (1985) Strategies for the selection of test methods. Neurobehav. Toxicol. Teratol.7: 661-662Girden E, Mettler FA, Finch G, and Culler E (1936) Conditioned responses in a decorticate dog to acoustic,thermal, and tactile stimulation. Journal of Comparative Psychology 21: 367-385Golub MS (1990) Use of monkey neonatal neurobehavioral test batteries in safety testing protocols. Neurotoxicology& Teratology 12: 537-541Golub MS, Eisele JH, and Donald JM (1991a) Effect of intrapartum meperidine on the behavioral consequencesof neonatal oxygen deprivation in Rhesus monkey infants. Developmental Pharmacology & Therapeutics4: 231-240Golub MS, Eisele LM, and Donald JM (1991b) Effect of intrapartum meperidine on the behavioral consequencesof neonatal oxygen deprivation in rhesus monkey infants. Developmental Pharmacology &Therapeutics 16: 231-240Gottlieb G (1971) Ontogenesis of sensory functionin birds and mammals. In: Tobach E, Aronson LR, ShawE (eds) The Biopsychology of Development. Academic Press, New York, pp 67-129Harlow HF (1959) The Development of Learning in the Rhesus Monkey. American Scientist Winter: 459-479Harlow HF (1962) Development of affection in primates. In: Bliss EL (ed) Roots of Behavior. Harper, New York,Head E, Mehta R, Hartley J, Kameka M, Cummings BJ, Cotman CW, Ruehl WW, and Milgram NW (1995)Spatial learning and memory as a function of age in the dog. Behavioral Neuroscience 109: 851-858Hellbrügge T (1960) The Development of circadian rhythms in infants. In: Anonymous Biological Clocks.Cold Spring Harbor Symp Quant Biol. pp 311-323Herschkowitz N, Kagan J, and Zilles K (1997) Neurobiological bases of behavioral development in the firstyear. Neuropediatrics 28: 296-306Hill DL, and Mistretta CM (1990) Developmental neurobiology of salt taste sensation. TINS 13: 188-195Hill WL, Borovsky D, and Rovee-Collier C (1988) Continuities in infant memory development. DevelopmentalPsychobiology 21: 43-62© 2006 by Taylor & Francis Group, LLC

1128 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONHines M (1942) The development and regression of reflexes, postures and progression in the young macaque.Contributions to Embryology 196: 155-209Hofer MA, Masmela JR, Brunelli SA, and Shair HN (1998) The ontogeny of maternal potentiation of theinfant rats’ isolation call. Developmental Psychobiology 33: 189-201Homberg V, Bickmann U, and Muller K (1993) Ontogeny is different for explicit and implicit memory inhumans. Neurosci Lett 150: 187-190Hudson R, and Distel H (1983) Nipple location by newborn rabbits: behavioral evidence for pheromonalguidance. Behavior 85: 260-275Hurtt ME, and Sandler JD (2003) Comparative Organ System Development: Introduction. Birth DefectsResearch (Part B) 68: 85-85Illingworth RS (1983) Normal Development. In: Anonymous Development of the Infant and Young Child.Churchhill Livingstone, New York, pp 120-162James WT (1955) Behaviors involved in expression of dominance among puppies. Psychological Reports 1:299-301James WT (1961) Preliminary observations on play behaviour in puppies. Journal of Genetic Psychology 98:273-277Johanson IB, and Hall WG (1979) Appetitive learning in 1-day-old rat pups. Science 205: 419-421Kajiura H, Cowhart BJ, and Beauchamp GK (1992) Early developmental changes in bitter taste responses inhuman infants. Developmental Psychobiology 25: 375-386Kaye R, Oski F, Barness L, (editors) (1988) Core Textbook of Pediatrics. J.B. Lippincott Company, Philadelphia,King JE, Fobes JT, and Fobes JL (1974) Development of early behaviors in neonatal squirrel monkeys andcotton-top tamarins. Developmental Psychobiology 7: 97-109Kleitman N, and Engelmann TG (1953) Sleep characteristics in infants. J. Appl. Physiol. 6: 269-282Konczak J, and Dichgans J (1997) The development toward stereotypic arm kinematics during reaching inthe first 3 years of life. Experimental Brain Research 117: 346-354Kuhtz-Buschbeck JP, Stolze H, Johnk J, Boczek-Funcke A, and Illert M (1998) Development of prehensionmovements in children: a kinematic study. Experimental Brain Research 122: 424-432Lantz C, Melen K, and Forssberg H (1996) Early infant grasping involves radial fingers. Developmentalmedicine-and-child-neurology38: 668-674Lashley KS, and Watson JB (1913) Notes on the development of a young monkey. Journal of Animal Behavior3: 114-139LeDoux JE, Sakaguchi A, and Reis DJ (1984) Subcortical efferent projections of the medial geniculate nucleusmediate emotional responses conditioned to acoustic stimuli. J. Neurosc. Meth. 4: 383-398Lemon RN (1999) Neural control of dexterity: what has been acheived? Exp. Brain Res. 128: 6-12Little AH, Lipsitt LP, and Rovee-Collier C (1984) Classical conditioning and retention of the infant’s eyelidresponse: effects of age and interstimulus interval. Journal of Experimental Child Psychology 37:512-524Margalit T, and Lancet D (1993) Expression of olfactory receptor and transduction genes during rat development.Brain Res. Developmental Brain Research. 73: 7-16Mendelson MJ (21982) Clinical examination of visual and social responses in infant rhesus monkeys. Dev.Psychol. 18: 658-664Milgram NW, Head E, Weiner E, and Thomas E (1994) Cognitive functions and aging in the dog: acquisitionof nonspatial visual tasks. Behavioral Neuroscience 108: 57-68Moore T, and Ucko C (1957) Night Waking in Early Infancy. Archives of disease in childhood 33: 333-342Morley ME (1972) The Development and Disorders of Speech in Childhood. Churchill Livingstone, Edinburgh,Mowbray JB, and Cadell TE (1962) Early Behavior patterns in rhesus monkeys. J. Comp. Physiol. Psychol.55: 350-357Murray T, Murray I (1980) Infant Communication, Cry and Early Speech. College Hill Press, Houston,Myslivecek J, and Hassmannova J (1991) Step-down passive avoidance in the rat ontogeny. Acta NeurobiolExp (Warsz) 51: 89-96Nelson WE (2000) The Newborn. In: Nelson WE (ed) Nelson Textbook of Pediatrics. W.B. Saunders Co,Philadelphia, pp 36-38Neuringer M, Coner WE, Petten CV, and Barstad L (1984) Dietary omega-3 fatty acid deficiency and visualloss ininfant rhesus monkeys. J. Clin. Invest. 73: 272-276© 2006 by Taylor & Francis Group, LLC

POSTNATAL DEVELOPMENTAL MILESTONES 1129Olivier E, Edgley SA, Armand J, and Lemon RN (1997) An electrophysiological study of the postnataldevelopment of the corticospinal system in the macaque monkey. J. Neurosc. 17: 267-276Panksepp J (1981) The ontogeny of play in rats. Developmental Psychobiology 14: 327-332Parisi T, and Ison J (1979) Development of the acoustic startle response in the rat: ontogenetic changes inmagnitude of inhibition by pre-pulse stimulation. Developmental Psychobiology 12: 219-230Parisi T, and Ison JR (1981) Ontogeny of control over the acoustic startle reflex by visual prestimulation inthe rat. Developmental Psychobiology 14: 311-316Paule MG (2001) Using Identical Behavioral Tasks in Children, Monkeys and Rats to Study the Effects inDrugs. Curr. Therap. Res. 62: 820-833Pawlowski AA, and Scott JP (1956) Hereditary differences in the development of dominance in litters ofpuppies. J. Comp. Physiol. Psychol. 49: 353-358Pedersen PE, Stewart WB, Green AG, and Shepherd GM (1983) Evidence of olfactory function in utero.Science 221: 478-480Pellis SM, Pellis VC, Morrissey TK, and Teitelbaum P (1989) Visual modification of vestibularly triggeredair-righting in the rat. Behav. Brain Res. 35: 23-26Petrosini, L. M., Molinari, M., and Gremoli, T. Hemicerebellectomy and motor behaviour in rats. I. Developmentof motor function after neonatal lesion. Experimental Brain Research 82(3), 472-482. 1990.Prather MD, Lavenex P, Mauldin-Jourdain M.L., Mason W.A., Capitanio JP, Mendoza SP, and Amaral DG(2001) Increased social fear and decreased fear of objects in monkeys with neonatal amygdala lesions.Neuroscience 106: 653-658Rao BS, Shankar N, and Sharma KN (1997) Ontogebny of taste sense. Indian Journal of Physiology andPharmacology 41: 193-203Richardson R, Paxinos G, and Lee J (2000) The ontogeny of conditioned odor potentiation of startle. BehavioralNeuroscience 114: 1167-1173Rose GH (1968) The development of visually evoked electro-cortical responses in the rat. DevelopmentalPsychobiology 1: 35-40Rosenblith JF, Sims-Knight JE (1904) In the Beginning. Brooks Cole, Monterey, CA,Routtenberg A, Strop M, and Jordan J (1979) Response of the infant rat to light prior to eyelid opening:mediation by the superior colliculus. Developmental Psychobiology 11: 469-478Rovee-Collier C, Gerhardstein P (1997) The development of infant memory. In: Cowan N (ed) The Developmentof Memory in Childhood. Psychology Press, pp 5-39Rovee-Collier C, Morrongiello BA, Aron M, and Kupersmidt J (1978) Topographical response differentiationin three-month-old infants. Infant Behavior and Devlopment 1: 323-333Rovee C, and Rovee DT (1969) Conjugate reinforcement of infant exploratory behavior. Journal of ExperimentalChild Psychology 8: 33-39Rudy JW, and Cheatle MD (1977) Odor-aversion learning in neonatal rats. Science 198: 845-846Rudy JW, Stadler-Morris S, and Albert P (1987) Ontogeny of spatial navigation behaviors in the rat: dissociationof “proximal”- and “distal”-cue-based behaviors. Behavioral Neuroscience 101: 62-73Schapiro S, Salas M, and Vukovich K (1970) Hormonal effects on ontogeny of swimming ability in the rat:assessment of central nervous system development. Science 168: 147-150Schenk F (1985) Development of place navigation in rats from weaning to puberty. Behavioral Neural Biology43: 69-85Schultz AH (1969) The Life of Primates. Weidenfeld and Nicolson, London,Schusterman RJ, Sjoberg A (1969) Early behavior patterns of squirrel monkeys (saimiri sciureus). In: CarpenterCR (ed) Behavior. Karger, Basel, pp 149-203Scott JP (1962) Critical Periods in Behavioral Development. Science 138: 949-958Scott JP, and Marston MV (1948) The development of dominance in litters of puppies. Anatom. Rec. 101:696-696Sheppard JJ, and Mysak ED (1984) Ontogeny of infantile oral reflexes and emerging chewing. Child Development55: 831-843Sisask G, Bjurholm A, Ahmed M, and Kreicbergs A (1995) Ontogeny of sensory nerves in the developingskeleton. Anatom. Rec. 243: 234-240Smart JL, and Dobbing J (1971) Vulnerability of developing brain. VI. Relative effects of foetal and earlypostnatal undernutrition on reflex ontogeny and development of behaviour in the rat. Brain Res. 33:303-314© 2006 by Taylor & Francis Group, LLC

1130 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSpear LP (1990) Neurobehavioral assessment during the early postnatal period. Neurotoxicology & Teratology12: 489-495Spear NE, and Smith GJ (1978) Alleviation of forgetting in preweanling rats. Dev Psychobiol 11: 513-529Stehouwer DJ, and Campbell BA (1980) Ontogeny of passive avoidance: role of task demands and developmentof species-typical behaviors. Dev Psychobiol 13: 385-398Stengel TJ (1980) The neonatal bhavioral assessment scale: description, clinical uses and research indications.Physical and Occupational Therapy in Pediatrics 1: 39-57Stewart P, Reihman J, Lonky E, Darvill T, and Pagano J (2000) Prenatal PCB exposure and neonatal behavioralasessment scale (NBAS) performance. Neurotoxicol. Teratol. 22: 21-29Stickord G, Kimble DP, and Smotherman WP (1982) In utero taste/odor aversion conditioning in the rat.Physiol Behav 28: 5-7Suchoff IB (1979) Visual development. J. Amer. Optom. Assoc. 50: 1129-1135Sullivan MH, Rovee-Collier C, and Tynes DM (1979) A conditioning analysis of infant long-term memory.Child Development 50: 162Teicher MH, and Blass EM (1977) First suckling response of the newborn albino rat: the roles of olfactionand amniotic fluid. Science 196: 635-636Thomas A (1940) Equilibre et equilibration. Masson, Paris,Unis AS, Petracca F, and Diaz J (1991) Somatic and behavioral ontogeny in three rat strains: preliminaryobservations of dopamine-mediated behaviors and brain D-1 receptors. Progress in Neuro-Psychopharmacology& Biological Psychiatry 15: 129-138Uziel ARR, and Marot M (1981) Development of cochlear potentials in rats. Audiology 20: 89-100Vander Linde E, Morrongiello BA, and Rovee-Collier C (1985) Determinants of retention in 8-week-oldinfants. Developmental Psychobiology 21: 601-613Vogt MB, and Rudy JW (1984) Ontogenesis of learning: IV. Dissociation of memory and perceptual-alteringprocesses mediating taste neophobia in the rat. Developmental Psychobiology 17: 601-611Volpe J (1995) The neurological examination normal and abnormal features. In: Anonymous Neurology of theNewborn. W.B. Saunders Co., Philadelphia, pp 101-101von Hofsten C, and Fazel-Zandy S (1984) Development of visually guided hand orientation in reaching. J ExpChild Psychol 38: 208-219Vorhees CV (1986) Methods for assessing the adverse effects of foods and other chemicals on animal behavior.Nutrition Reviews 44 Suppl: 185-193Vorhees CV, Acuff-Smith KD, Moran MS, and Minck DR (1994) A new method for evaluating air-rightingreflex ontogeny in rats using prenatal exposure to phenytoin to demonstrate delayed development.Neurotoxicology & Teratology 16: 563-573Westerga J, and Gramsbergen A (1990) The development of locomotion in the rat. Dev. Brain Res. 57: 163-174Williams R, and Boothe R (1981) Development of optical quality in the infant monkey (Macaca nemestrina)eye. Invest. Opthalmol. Visual Sci. 21: 728-736Winneke G (1995) Endpoints of developmental neurotoxicity in enviromentally exposed children. Toxicol.Lett. 77: 127-136Winter P, Handley P, Ploog D, and Schott D (1973) Ontogeny of squirrel monkey calls under normal conditionsand under acoustic isolation. Behaviour 47: 131-147Yamaguchi K, Harada S, Kanemaru N, and Kasahara Y (2001) Age-related alteration of taste bud distributionin the common marmoset. Chemical Senses 26: 1-6Zametkin AJ, Stevens JR, and Pittman R (1979) Ontogeny of spontaneous blinking and of habituation of theblink reflex. Ann. Neurol. 5: 453-457Zimmerman RR, Torrey CC (1965) Ontogeny of Learning. In: Schrier AM, Harlow HF, Stollnitz F (eds)Behavior of Nonhuman Primates: Modern Research Trends. Academic Press, New York and London,pp 405-447Zola-Morgan S, and Squire LR (1993) Neuroanatomy of memory. Annual Review of Neuroscience 16: 547-563© 2006 by Taylor & Francis Group, LLC

PART IPrinciples and Mechanisms© 2006 by Taylor & Francis Group, LLC

CHAPTER 1Principles of DevelopmentalToxicology RevisitedRonald D. HoodCONTENTSI. Introduction ............................................................................................................................3II. Basic Principles......................................................................................................................5A. Some Basic Terminology ..............................................................................................5B. Wilson’s Principles ........................................................................................................61. Susceptibility to Teratogenesis Depends on the Genotype of the Conceptusand the Manner in Which This Interacts with Adverse Environmental Factors....62. Susceptibility to Teratogenesis Varies with the Developmental Stageat the Time of Exposure to an Adverse Influence ..................................................73. Teratogenic Agents Act in Specific Ways (Mechanisms) on DevelopingCells and Tissues to Initiate Sequences of Abnormal DevelopmentalEvents (Pathogenesis)..............................................................................................84. The Access of Adverse Influences to Developing Tissues Dependson the Nature of the Influence (Agent)...................................................................85. The Four Manifestations of Deviant Development Are Death, Malformation,Growth Retardation, and Functional Deficit...........................................................96. Manifestations of Deviant Development Increase in Frequency and Degreeas Dosage Increases, from the No-Effect to the Totally Lethal Level.................10III. Who Will Conduct the Tests, and Who Will Interpret the Results?...................................11IV. Where Do We Go from Here?.............................................................................................12References ........................................................................................................................................12I. INTRODUCTIONDevelopmental toxicology has been evolving as a discipline for decades with only modest initialrecognition despite the early knowledge that an excess of certain nutrients (e.g., vitamin A) 1 oradministration of various chemicals could cause developmental defects in various animal species. 2–4As has been stated many times, it took the revelation in the early sixties that thalidomide, a drugpromoted as a relatively innocuous sedative and antiemetic, was a potent human teratogen 5 to arouse3© 2006 by Taylor & Francis Group, LLC

4 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONinterest in testing for potential developmental toxicants 6 and to toughen the Food, Drug, andCosmetic Act of 1938. 7 Since that time, testing protocols have slowly evolved, first under theguidance of regulatory agencies in individual countries and more recently as the result of jointefforts to harmonize test paradigms and reduce duplication of effort (cf. Chapter 19).Relatively early in this process, a series of three protocols was designed to evaluate test agentsfor their effects on developing mammals, with the intent of protecting humans exposed to pharmaceuticals,food additives, pesticides, workplace chemicals, and environmental pollutants. Theseprotocols were developed for the purpose of assessing (1) the outcomes on the conceptus of maternalexposures, beginning prior to mating and ending prior to implantation (Segment I: “Study of Fertilityand General Reproductive Performance”), (2) exposures during major organogenesis (Segment II:“Prenatal Developmental Toxicity Study” or “Teratological Study”), and (3) exposures during lategestation, parturition, and lactation (Segment III: “Perinatal and Postnatal Study”). The currentiterations of these protocols, descriptive terminology, and generated data are described and discussedelsewhere in this volume and particularly in Chapters 7, 9, and 19. It is of interest to note, however,that although the test procedures have evolved in specific aspects, they have not changed greatlysince they were first recommended by the U.S. Food and Drug Administration (FDA). 8Changes in test protocols have typically been modest, such as increases in the numbers of testfemales required or in the duration of treatment. Another example is the requirement for neurobehavioraltesting in certain cases. 9,10 Perhaps the greatest advances in developmental toxicity testinghave come not from improvements in the standard protocols but from our increased knowledge ofhow to interpret test outcomes, and how and when to modify the protocols. And it must be keptin mind that the standard testing protocols are necessarily compromises between the demands ofefficiency and cost effectiveness and the quality and completeness of the information the testsprovide. 11 The need to keep costs at a bearable level is in conflict with the needs of regulatoryagencies to obtain adequate data to serve as the basis for the required decisions. Thus, the SegmentII test protocol compromises by calling for treatment throughout organogenesis (or even throughoutgestation if the treatment is not expected to prevent implantation) until the day of palate closureor until the day prior to scheduled sacrifice at term, depending on the specific guideline. That isthe case even though the use of several groups of mated females treated at each dosage level duringbrief, discrete periods of organogenesis would be more effective at revealing a compound’s teratogenicitypotential. Interestingly, the latter methodology had been proposed initially. 12 Conversely,although smaller test groups of rabbits than of rats were once allowed, primarily to contain costs,today the minimum number of rabbits has been increased to provide more meaningful data.At some future date, cellular/molecular assays and quantitative structure-activity-relationships(QSAR) may become more routine to supplement (or even replace) the current whole-animaldevelopmental toxicity tests. However, that eventuality is likely to require major increases in ourunderstanding of both the mechanisms of developmental toxicity and the complex interplay of themolecular and physiological systems that govern and regulate both developmental processes andmaternal physiology and homeostasis.As is discussed in Chapter 2, developmental toxicants first act via specific mechanisms, i.e.,the initial event(s) in the germ cells or in the cells of the embryo or fetus that begin the series ofprocesses (i.e., pathogenesis, described in Chapter 3) leading to adverse outcomes. This is true oftoxic insults to adults as well, but such occurrences in immature systems are made more complexby the constantly changing nature of the developing organism, especially during the period fromconception through major organogenesis. Adding even further to the complexity in mammals isthe interplay between the developing conceptus and the supporting maternal “environment,” mediatedduring most of development by the extraembryonic membranes and with the eventual additionof the placenta.Our understanding of specific incidences of events such as abnormal development, functionaldeficit, or prenatal demise is further confounded by the likelihood that such manifestations may,at times, merely be sporadic failures of complex systems. It is likely that the genetic “blueprint”© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF DEVELOPMENTAL TOXICOLOGY REVISITED 5for the development of complex organisms is not failure proof, and that even in the absence ofdeleterious mutations or chromosomal anomalies, development can fail. This might happen if certaincritical gene alleles, which would ordinarily direct a robust developmental process, specified a moreerror prone process when present in a specific combination. Alternatively, some species or strainsof animals seem to have “weak points” in their development, such that some percentage of offspring,even if they had identical genomes and similar environments, would manifest an anomaly. In otherwords, the genetically specified plan for development is seldom if ever perfect. The myriad eventsit specifies must occur at just the right time, in the right location, and in a reasonably correctmanner, and in a small percentage of individuals one or more such processes may fail. This seemsa likely mechanism in cases where an inbred strain of mice exhibits a high “spontaneous” incidenceof some defect, such as cleft palate. The incidence of such anomalies can be further increased byexposure of the conceptus to a toxic agent or a compromised maternal environment, which presumablynudges borderline cases in the direction of abnormal development.Although much remains to be learned about the causation of adverse effects on developingoffspring, there are certain principles that should be considered by anyone seeking to plan, carryout, or interpret the results of tests for developmental toxicity, including epidemiologic studies. Anumber of these principles have been known for some time, and much of this chapter will bedevoted to such basic principles.A. Some Basic TerminologyII. BASIC PRINCIPLESAccording to Wilson, 13 teratology is “the science dealing with the causes, mechanisms, and manifestationsof developmental deviations of either structural or functional nature.” He also definedteratology as “the study of adverse effects of environment on developing systems, that is, on germcells, embryos, fetuses, and immature postnatal individuals.” Although it is recognized that a portionof developmental defects have a genetic causation, Wilson reckoned that, “Even hereditarydefects…were initiated as mutations at some time in the past,” and thus, “It is probable that allabnormal development has its causation in some aspect of environment.”In descriptions of the harmful effects of chemical or physical agents on developing systems,terms such as embryotoxicity and fetotoxicity have often been used. These are legitimate terms, butit should be recognized that they are properly applied only to toxic insults occurring during thespecified portion of the developmental process. They should not be used as all-encompassing termsto describe the effects of exposures throughout the entirety of development. Developmental toxicitycan be defined as the ability of a chemical or physical agent to cause any of the manifestations ofadverse developmental outcome (i.e., death, malformation, growth retardation, functional deficit),individually or in combination. Teratogenicity has been used to mean just the ability to produce“terata” or malformations. In the broader sense, it has been used in the same way as developmentaltoxicity, and the study of developmental toxicity is referred to as developmental toxicology. Malformationhas been defined as “a permanent structural change that may adversely affect survival,development, or function,” while a variation is “a divergence beyond the usual range of structuralconstitution that may not adversely affect survival or health.” 14 Distinguishing between these twocan be difficult, however, because they exist on a continuum between the normal and the abnormal.Although basic principles pertaining to developmental toxicology and teratology are presumablyknown to practitioners in the field, it can be useful to review some of these principles andexperimental support for them. Also, individuals new to the field can benefit from such accumulatedwisdom as an aid in designing, carrying out, and interpreting the results of experimental andepidemiological studies.© 2006 by Taylor & Francis Group, LLC

6 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONB. Wilson’s Principles 151. Susceptibility to Teratogenesis Depends on the Genotype of the Conceptusand the Manner in Which This Interacts with Adverse Environmental Factors 15It must also be kept in mind that in some cases there is a purely genetic cause, e.g., Down syndromein humans, and there can be purely environmental causes, such as x-rays at high doses. In stillother cases, however, a combination of environmental insult(s) and a susceptible genome in theconceptus is required for adverse effects to appear, a scenario that Fraser termed multifactorialcausation. 16It is now well established that the genetic makeup of the conceptus can significantly influencethe outcome of exposures to developmental toxicants, especially if the level of exposure is nearthe threshold for causing a particular adverse effect. This has been readily observed in studiesinvolving treatment of inbred mouse strains and crosses between them. However, one must beaware of the caveat that the developmental outcome may also be influenced by the genotype of thedam as a determinant of, for example, the rate or preferred pathway of biotransformation of thetoxicant in question and its peak level in the maternal blood or its area under the curve (AUC, thearea under the plasma [or serum or blood] concentration versus time curve). The outcome here canbe influenced by whether it is the parent compound or a metabolite that is developmentally toxic,and of course in some cases both can be toxic. Fetal alcohol syndrome may be a case where thematernal and/or fetal genotype can interact with the toxic agent (and probably with other environmentalinfluences, such as the maternal diet) to produce damage in some instances and few or noobvious effects in others. 17Recently, confirmation of the significant influence of specific genes has come from experimentsin which knockout mice lacking a specific gene have been found to be either enhanced or diminishedin susceptibility to exposure to a developmentally toxic agent. 18 Further, recent human studies haveprobed the potential influence of combinations of specific gene polymorphisms and dietary deficiencies,especially folate deficiency, on the incidence of neural tube defects. 19 And deficiency inthe activity of a detoxifying enzyme, epoxide hydroxylase, may be an important determinant ofthe manifestation of fetal hydantoin effects. 20Species differences in response to developmental toxicants may be due to differences in inherentsusceptibility of the conceptus to differences in maternal pharmacokinetics — including biotransformation— and maternal physiology, or to a combination of these. The same is true of strain andlitter differences in response to toxic insult, and the basis for strain differences can be determinedby use of reciprocal crosses and by embryo transfer. Typically, there are also individual differenceswithin litters. This is probably most commonly due, at least in part, to genotypic differences amongthe individual fetuses, though this should be less often true when inbred strains are involved.Differences in developmental stage at the time of exposure to toxic insults — those of shortduration or those that begin during organogenesis — may explain some of the differences inindividual response among fetuses within the same litter. In some cases, there are sex differencesin the response, i.e., males and females may be affected differently. Differences in the intrauterineenvironment of each conceptus may also contribute to nonuniformity of response within litters.For example, placental blood supply varies somewhat according to location in the uterus 21,22 andmay bring or remove greater or lesser amounts of a toxic substance or of nutrients, waste products,or respiratory gases. Female rodent fetuses that are found next to one male or (especially) betweentwo male fetuses can be altered in certain attributes, such as sexual attractiveness and estrous cyclelength, compared with those of females not found next to males. 23 Such effects are presumablycaused by the transfer of androgen from the male to the adjacent female fetuses.Even differences in fetal drug metabolizing capability may be influential, at least in fetuses inwhich enhanced levels of xenobiotic biotransforming enzymes had been induced. Nebert found© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF DEVELOPMENTAL TOXICOLOGY REVISITED 7this to be true for a strain of mice heterozygous at the AHH (aryl hydrocarbon hydroxylase) locus,where some fetuses lack the induction receptor. 24 This receptor responds to aryl hydrocarbons, suchas TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin), forming a complex involved in activating the genesfor cytochrome P-450 monooxygenases, a process thought to be involved in initiating the toxiceffects of TCDD.2. Susceptibility to Teratogenesis Varies with the Developmental Stageat the Time of Exposure to an Adverse Influence 15Development can be roughly divided into three periods with regard to susceptibility to toxic insult:a. PredifferentiationIf damaged during this period, the embryo typically either dies or completely repairs the damage,an “all or none” effect. There are some exceptions, however, such as the effects of x-rays or certainhighly potent genotoxic chemicals, and their effects may be genetically mediated. 25b. Early Differentiation or Early OrganogenesisThis occurs roughly from days 9 through 15 in the rat and days 9 through 18 in the rabbit (day onwhich mating was confirmed is day 0), with the timing varying by species (see Chapter 6, Table1 for others). Organ primordia, foundations for later development, are laid down at this time. Theembryo is most susceptible to induction of malformations during this period of development becauseonce the basic structures are formed, it becomes increasingly difficult to alter them structurally.Organs whose development is multiphasic, such as the eye and the brain, tend to have more thanone susceptible period.c. Advanced — or Late — OrganogenesisThis period is largely occupied with histogenesis and functional maturation. Insults during thisperiod mainly cause growth retardation, developmental delay, or functional disturbances, particularlyneurobehavioral problems, as the brain matures relatively late. However, even at this time,interrupted blood supply to a localized area or structure (e.g., because of an amniotic band) cancause degeneration of that area or structure, resulting in a malformation.A malformation may sometimes occur well after the initiating toxic insult, and this might bedue to alteration of biochemical events, such as gene activation and mRNA synthesis, that mayoccur prior to organ differentiation. This might also be due to alteration of an earlier event in asequence of events leading up to the formation of an organ — for example, interference with aninduction early in a “chain of inductions.”Organisms tend to be significantly more sensitive to many adverse environmental influencesduring early developmental stages, although this differential may not be quite as universallyapplicable in mammals as was once thought. And even though the human gestation period is long,the portion of that time spent in early organogenesis is fortunately relatively short; thus, the humanembryo is not at peak risk in comparison with its total gestation length. Increased sensitivity,especially during early organogenesis, apparently occurs because many complex and alterableevents are taking place. Many tissues are undergoing rapid cell division, and the embryo, and to aconsiderable extent, the fetus, has much less capacity to metabolize xenobiotics than does the adult.Some embryos, such as those of rodents, however, can be induced to biotransform a significantamount of certain developmental toxicants. Thus, metabolism can, in some cases, actually activatea toxicant, increasing its effects on the conceptus.© 2006 by Taylor & Francis Group, LLC

8 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION3. Teratogenic Agents Act in Specific Ways (Mechanisms) on DevelopingCells and Tissues to Initiate Sequences of Abnormal DevelopmentalEvents (Pathogenesis) 15One consequence of mechanistic specificity is the frequently observed “agent specificity” of malformations,malformation syndromes, and/or behavioral or other functional effects. Many teratogensproduce characteristic patterns of malformations and other effects as a result of their particularmechanisms of action and/or their tissue distribution. Although this specificity is most oftenmentioned in the context of evaluating human data, it is also important to consider it wheninterpreting animal test outcomes that appear somewhat ambiguous.It must also be appreciated that a given developmental toxicant may act by more than onemechanism at the same time in the same organism, although one or more such mechanisms maypredominate. Unfortunately, there has often been the tendency to interpret the outcomes of toxicologicaltesting of individual compounds in terms of a single mechanism proposed for the testagent. This has generally been the case even though it has long been known that many agents mayhave multiple actions in biological systems. In addition, most tests for developmental or reproductivetoxicity of necessity employ only one test agent and carefully control for other variables that mightinfluence test outcomes. This is true even though such conditions are never seen outside thelaboratory, as humans and other organisms are continually subjected to a variety of biologicallyactive agents and other environmental influences throughout the reproductive and developmentalprocess.With regard to pathogenesis, we must also consider that multiple defects are often seen in thesame individual. Thus the same or different mechanisms may initiate abnormalities that are actingconcurrently by pathogenetic pathways in different organs. Alternatively, as pointed out by Bixleret al., 26 expression of a given pathway of dysmorphogenesis may secondarily result in a defect insome other structure of the embryo or fetus. Interpretation of experimental findings is made morechallenging because, according to the concept of common pathogenetic pathways as proposed byWilson, 27 even like defects do not necessarily share the same initial causal mechanism.Mechanisms are discussed more fully in Chapter 2, and pathogenesis in Chapter 3, as theseimportant topics require a more complete treatment than would be appropriate here.4. The Access of Adverse Influences to Developing Tissues Dependson the Nature of the Influence (Agent) 15a. Placental TransferThe layer of cellular and extracellular material between the maternal and fetal bloodstreams hassometimes been called the placental barrier because at one time it was believed to afford greatprotection to the embryo and fetus. We now know that the degree of protection is often modest atbest, and that instead of being a barrier, the placental membrane acts more as an ultrafilter. 28 Also,a number of physiologically important substances, such as amino acids and glucose, are transferredacross the placental membrane by specific mechanisms. 29 This can sometimes allow xenobioticsto be transferred as well, if they can take advantage of the physiologic systems.A number of factors are known that can influence how readily a given substance crosses theplacenta:Molecular size — smaller molecules cross more readily, especially those under about 1000 in molecularweight.Charge — uncharged molecules cross more readily than charged compounds of the same size; negativelycharged molecules pass more readily than those that are positively charged.Lipid solubility — lipophilic compounds penetrate more quickly than more polar forms.© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF DEVELOPMENTAL TOXICOLOGY REVISITED 9Degree of ionization — less ionized substances pass more readily.Formation of complexes — complexed molecules are impeded in comparison with their “free” forms,especially if they are complexed with proteins.Existence of concentration gradients — substances present in high concentrations in the maternal bloodare likely to cross the placenta in larger amounts.Placental metabolism — the placenta can metabolize certain substances, but the influence of placentalmetabolism is thought to be minor for most compounds.From the foregoing, it can be seen that although the placenta only rarely acts as a completebarrier, it can greatly slow the passage of certain water soluble molecules, such as heparin andplasma proteins, that are large and/or highly charged. Also, certain physical agents, such as x-rays,gamma rays, ultrasound, and radiofrequency radiation, can readily reach the embryo or fetus, evenfrom outside the mother. And maternal metabolism and excretion may eliminate a portion of theabsorbed dose of a chemical agent or alter its nature prior to its reaching the conceptus. 29b. Relationship of the Placenta to TeratogenesisIt has been suggested that interference with the function of the yolk sac, particularly duringdevelopment of rodents and lagomorphs, may result in developmental defects because these speciesdepend for a considerable portion of gestation on an “inverted yolk sac placenta” 30 (see Chapter 6for more discussion and for comparisons with other species).c. Exposure of the ConceptusIf a potentially developmentally toxic agent reaches the embryo at a sufficiently high rate, it mayreach a sufficiently high concentration to cause an adverse effect; on the other hand, if the rate istoo low, it will not reach an effective concentration. This is because at very low exposure rates,cells may not be significantly affected, and at somewhat higher rates, damage that occurs may berepaired before irreparable harm is done. It should be noted that although mammalian embryosand fetuses can repair minor injuries, if they have developed past the blastocyst stage, they cannotregenerate lost parts. It is also of interest to note that in some cases a chemical may become moreconcentrated in the conceptus than in the mother, depending on the particular agent and exposureroute. 31-335. The Four Manifestations of Deviant Development Are Death, Malformation,Growth Retardation, and Functional Deficit 15In some cases, it is likely that malformation precedes and results in death. According to Kalter, 34as the dose is increased in such a scenario, an inverse relationship between malformation andprenatal mortality would be seen. This would occur as the severity of the induced defects increasedto the point of causing deaths among the offspring. A practical consequence is that an increasedincidence of prenatal deaths may obscure our ability to determine if an agent has caused malformationsbecause the most severely malformed offspring may not survive to be counted at cesarean sectionin a typical teratogenicity assay. However, malformation as the cause of prenatal demise can berevealed if examination of embryos or fetuses prior to birth reveals more malformations and fewerdeaths than would be expected at term. 15 In other cases, death and malformation may be due, atleast in part, to different causes, and thus it is sometimes possible to block or enhance one effectwithout blocking or enhancing the other. Independence of these two manifestations (malformationand death) would be suspected if their frequencies increased independently with increasing dose. 34The two additional possible consequences of developmental toxicity, growth retardation and functionaldeficit, may also be caused by malformations, or they may be induced independently by© 2006 by Taylor & Francis Group, LLC

10 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONeither the same or different mechanisms from those resulting in malformation or death. Further,the consequences of exposure to a developmental toxicant are strongly influenced by the timing ofexposure, as pointed out in Section II.B.2. and discussed elsewhere in this volume (e.g., Chapter 6).6. Manifestations of Deviant Development Increase in Frequency and Degreeas Dosage Increases, from the No-Effect to the Totally Lethal Level 15It is often assumed, although there are arguments on both sides of the issue, 35–37 that developmentallytoxic agents typically have a “no-effect” threshold because of the repair and regulative abilities ofthe embryo. Existence of or lack of such a threshold is virtually impossible to prove experimentally,however. Even if large group sizes are employed, it is difficult to decide how to fit a dose-responsecurve to the data at the lowest doses, as there is always a background level of adverse outcomes,such as malformation and prenatal demise. An argument against the existence of developmentaltoxicity thresholds in some cases states that when a hormone or other endogenous chemical isresponsible for adverse effects that make up a portion of the control (background) incidence, athreshold dose will not be observed for an exogenous chemical acting by the same mechanism.Even though a threshold may exist, it has already been exceeded by the endogenous chemical, andagents acting by the same mechanism merely add to the background incidence. 38The threshold concept has also been debated with regard to other forms of toxicity, such ascarcinogenicity and mutagenicity, but the existence of a threshold seems more likely to holdgenerally true for teratogenicity, with a few possible specific exceptions as described above. Andthe assumption of thresholds, along with no-observed-adverse-effect levels (NOAELs), has provenuseful for practical risk assessment, even in the absence of ability to confirm their existence absolutely.Teratogenesis dose-response curves are often quite steep, so that merely doubling the dose (orless) may extend the range from minimal to maximal effects, but there are exceptions (e.g.,thalidomide in humans). It is also possible to have a U-shaped (or J-shaped) dose response, wherean inadequate level of a compound is harmful, a higher level is beneficial, and a still higher levelis again harmful. Such responses are common in the case of vitamins and essential minerals. Forexample, vitamin A deficiency is teratogenic, appropriate levels are required for normal development,and excessive amounts are again teratogenic. 1 Another possibility is hormesis, which hasbeen described as producing a U- or J-shaped dose-response curve. In this case, if a low doseresults in a stimulatory effect on a normal function, e.g., growth, in comparison with no exposure,while higher doses are inhibitory, the response curve would appear as an “inverted U.” 39,40 Whenthe measured response shows some dysfunction (as would be expected with most toxicants), butlow doses are beneficial relative to controls while higher doses result in increased levels ofdysfunction, the dose-response is described as a “J” or a “noninverted U.” 40 The occurrence ofhormesis and mechanisms for such responses remain as areas of controversy, and whether hormesisoccurs at low doses of developmental or reproductive toxicants in laboratory animals or in humansis as yet unclear.It is not always appreciated that the current protocol for the developmental toxicity study is notterribly efficient at eliciting malformations, 41 and that, as stated by Palmer, 42 malformations can bea somewhat unreliable guide to developmental toxicity in Segment II tests. The practice of dosingthroughout much or all of gestation does severely stress the maternal/embryonic/fetal system,however, and is likely to elicit other useful dose-related indicators of adverse effects on theconceptus, such as decreased survival, diminished fetal weight, or elevated dose-related incidencesof developmental variations. But production of malformations often requires a relatively specificset of conditions, including the right species and appropriate treatment dose, mode, and timing. Itcan even be influenced by such factors as the vehicle for the test article. And it is well establishedthat dosing a pregnant animal once or twice on just one or two gestation days is generally moreeffective at eliciting malformations (and other manifestations of developmental toxicity) than isdosing for much of gestation. That is largely because if the dam receives only one or a few doses,© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF DEVELOPMENTAL TOXICOLOGY REVISITED 11higher doses can generally be used without causing maternal demise. Further, such targeted treatmentscan be administered during peak periods of sensitivity of the conceptus.The A/D ratio is a concept that was proposed by Marshall Johnson. 43,44 It is the ratio of the“adult toxic dose” to the “developmental toxic dose” (the dose harmful to the conceptus). If theconceptus is significantly more sensitive than the adult, the ratio is above 1, if the two are similarlysensitive, the ratio approximates 1, while if the mother is more sensitive, the ratio is less than 1.In practice, the teratogenic dose is often a level that is toxic or sometimes lethal to the mother, i.e.,the A/D ratio is typically close to 1, but some teratogens can be effective at doses that are apparentlyharmless to the mother. These agents with high A/D ratios have often been considered to beespecially significant in human risk assessment because exposures to toxicants at doses high enoughto be obviously harmful to adults are commonly avoided. Agents not obviously harmful to adults,however, might fail to cause sufficient concern about potential exposures. Nevertheless, pregnantwomen have frequently been exposed to developmental toxicants at doses harmful to both motherand offspring. Obvious examples include alcohol and cigarette smoke. In such a scenario, the likelylevel of maternal exposure vs. the margin of safety for the conceptus can be of more importancethan the relative sensitivities of adult and offspring. 45Other important considerations include the relative extent and duration of harm to the offspringvs. the mother. For example, a pregnant woman who is treated with Accutane during early pregnancyis not likely to experience a lasting harmful effect, while her baby may be irreparably harmed. 45 Itis also of interest to note that recent tests comparing the A/D ratios for the same compounds indifferent species suggest that the ratio is not at all constant across species. 46,47 Such findings indicatethat the A/D ratio is of relatively little use for risk extrapolation.One last comment on the issue of dose vs. effect is that the span of time during developmentwhen a defect can be produced by a toxic insult tends to widen as the dose level increases. 15III. WHO WILL CONDUCT THE TESTS,AND WHO WILL INTERPRET THE RESULTS?Graduate programs that are the source of the new generation of toxicologists have increasinglyemphasized research and course work in cellular and molecular toxicology, areas that are of obviousimportance (e.g., see Chapters 2, 11, 12, and 14 to 16). However, new personnel hired in otherareas of our disciplines often start out with inadequate training for the positions they are expectedto fill. For several years this state of affairs has become increasingly evident to many in the fieldsof developmental and reproductive toxicology, and it appears true of positions in both testinglaboratories and regulatory agencies. Students are no longer being trained in significant numbersin what might be termed “classical” testing methodology (e.g., the types of tests outlined in Chapters7 through 10). Lack of in-depth experience working with laboratory animals and insufficientknowledge of animal biology and principles of toxicology can hinder the ability of new professionalsto design, carry out, and interpret results of safety evaluation studies unless they are given considerable“on the job” training.The reasons for this apparent lack of classically trained graduates from institutions of higherlearning, including graduate programs in “toxicology,” seem clear. Research funding for academiarapidly shifted toward support of mechanistic studies conducted solely at the cellular, biochemical,and especially the molecular level, as investigators increasingly make use of the newly availabletools of molecular biology. Not only has there been a major shift in research funding, there hasalso been a change in the way certain types of research are viewed by many. Molecular approachesare often seen as the only “real science,” and work with whole animals may be viewed as outdated.New college students are told about the latest hot cellular or molecular research and often neverhear about other possibilities throughout their undergraduate and graduate training. Thus, the typicalstudent can complete a master’s or doctoral degree without ever touching a live mammal or having© 2006 by Taylor & Francis Group, LLC

12 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONto give much thought to the original source of the cells and molecules that he or she may encounterand manipulate in the laboratory. Nevertheless, despite advances in developmental and reproductivetoxicology at the cellular and molecular levels, the need for animal testing remains and will likelypersist for some time in the future. This situation poses a challenge to the developmental andreproductive toxicology profession. It puts increased pressure on the (hopefully) more knowledgeablesupervisors of newly hired professionals to mentor and monitor the activities of the new hiresuntil they can attain the appropriate levels of knowledge and experience. If this is not done, theresult will inevitably be less well-designed studies, poorly presented and interpreted data, andineffective or inappropriate regulatory decisions.IV. WHERE DO WE GO FROM HERE?Although, as stated above, developmental and reproductive toxicity testing in laboratory animalsremains a major and essential enterprise, especially in industry and in regulatory agencies, scientificadvances in related disciplines have opened doors to complementary areas of toxicologicalresearch. 18 Studies at the biochemical, cellular, and molecular levels (e.g., Adeeko et al. 48 ) havebrought the promise that we will increasingly understand the mechanistic bases of manifestationsof toxicity. They also bring the hope that we will become better able to predict toxicity and toextrapolate findings in laboratory animals to likely outcomes of human exposure with increasinglygreater accuracy. For example, as stated by MacGregor, 49 “It is clear that genomic technologies arealready being used to develop new screening strategies and biomarkers of toxicity, to determinemechanisms of cellular and molecular perturbations, to identify genetic variations that determineresponses to chemical exposure and sensitivity to toxic outcomes, and to monitor alterations in keybiochemical pathways.” Such advances in science bring on new regulatory challenges, however, inthat considerable understanding and wisdom will be required to make the best use of the escalatingwealth of new data. We must also take extreme care not to lose sight of our “roots.” An understandingof whole animal biology and toxicology will remain essential regardless of advances in the studyof mechanisms.REFERENCES1. Szabo, K.T., Congenital Malformations in Laboratory and Farm Animals, Academic Press, San Diego,1989, chap. 2.2. Gilman, J., Gilbert, C., and Gilman, G.C., Preliminary report on hydrocephalus, spina bifida, andother congenital anomalies in rats produced by trypan blue, South Afr. J. Med. Sci., 13, 47, 1948.3. Hoskins, D., Some effects of nitrogen mustard on the development of external body form in the rat,Anat. Rec., 102, 493, 1948.4. Ancel, P., La Chimioteratogenèse. Réalisation des Monstruosités par des Substances Chimiques Chezles Vertébrés, Doin, Paris, 1950.5. Lenz, W., A short history of thalidomide embryopathy, Teratology, 38, 203, 1988.6. Wilson, J.G., The evolution of teratological testing, Teratology, 20, 205, 1979.7. Kelsey, F.O., Thalidomide update: Regulatory aspects, Teratology, 38, 221, 1988.8. Goldenthal, E.I., Guidelines for Reproduction Studies for Safety Evaluation of Drugs for Human Use,Drug Review Branch, Division of Toxicological Evaluation, Bureau of Science, U. S. Food and DrugAdministration, 1966.9. Hass, U., Current status of developmental neurotoxicity: regulatory view, Toxicol. Lett., 140–141, 155,2003.10. Kaufman, W., Current status of developmental neurotoxicity: an industry perspective, Toxicol. Lett.,140–141, 161, 2003.© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF DEVELOPMENTAL TOXICOLOGY REVISITED 1311. Hood, R.D., Tests for developmental toxicity, in Developmental Toxicology: Risk Assessment and theFuture, Hood, R.D., Ed., Van Nostrand Reinhold, New York, 1990, chap. 15.12. Wilson, James G., Environment and Birth Defects, Academic Press, New York, 1973, chap. 10.13. Wilson, James G., Environment and Birth Defects, Academic Press, New York, 1973, chap. 1.14. U.S. Environmental Protection Agency, Guidelines for health assessment of suspect developmentaltoxicants, Federal Register, 51, 34028, 1986.15. Wilson, James G., Environment and Birth Defects, Academic Press, New York, 1973, chap. 2.16. Fraser, F.C., Relation of animal studies to the problem in man, in Wilson, J.G. and Fraser, F.C., Eds.,Handbook of Teratology, Vol. 1., General Principles and Etiology, Plenum, New York, 1977, chap. 3.17. Davidson, R.G. and Zeesman, S., Genetic aspects, in Maternal-Fetal Toxicology, A Clinician’s Guide,2nd ed., Koren, G., Ed., Marcel Dekker, New York, chap. 23.18. Committee on Developmental Toxicology, Board on Environmental Studies and Toxicology, Commissionon Life Sciences, National Research Council, Scientific Frontiers in Developmental Toxicologyand Risk Assessment, National Academy Press, Washington, DC, 2000.19. Botto, L.D. and Yang, Q., 5,10-Methylenetetrahydrofolate reductase gene variants and congenitalanomalies: A HuGE review, Am. J. Epidemiol., 151, 862, 2000.20. Finnell, R.H. Teratology: General considerations and principles, J. Allergy Clin. Immunol., 103, S339,1999.21. McLaren, A., Genetic and environmental effects on foetal and placental growth in mice, J. Reprod.Fertil., 9, 79, 1965.22. Bruce, N.W. and Abdul-Karim, R.W., Relationships between fetal weight, placental weight andmaternal circulation in the rabbit at different stages of gestation, J. Reprod. Fertil., 32, 15, 1973.23. Vom Saal, F.S. and Bronson, F.H., Variation in length of the estrous cycle in mice due to formerintrauterine proximity to male fetuses, Biol. Reprod., 22, 777, 1980.24. Shum, S., Jensen, N.M., and Nebert, D.W., The murine Ah locus: In vitro toxicity and teratogenesisassociated with genetic differences in benzo[a]pyrene metabolism, Teratology, 20, 365, 1979.25. Hood, R.D., Preimplantation effects, in Developmental Toxicology: Risk Assessment and the Future,Hood, R.D., Ed., Van Nostrand Reinhold, New York, 1990, chap. 6.26. Bixler, D., Daentl, D., and Pinsky, L., Panel discussion: Applied developmental biology, in Melnick,M. and Jorgenson, R., Eds., Developmental Aspects of Craniofacial Dysmorphology, Alan R. Liss,New York, 1979, p. 99.27. Wilson, James G., Environment and Birth Defects, Academic Press, New York, 1973, chap. 5.28. Faustman, E.M. and Ribeiro, P., Pharmacokinetic considerations in developmental toxicity, in DevelopmentalToxicology: Risk Assessment and the Future, Hood, R.D., Ed., Van Nostrand Reinhold, NewYork, 1990, chap. 13.29. Glazier, J.D., Harrington, B., Sibley, C.P., and Turner, M., Placental function in maternofetal exchange,in Fetal Medicine: Basic Science and Clinical Practice, Rodeck, C.H. and Whittle, M.J., Eds.,Churchill Livingstone, London, 1999, chap. 11.30. Hood, R.D., Mechanisms of developmental toxicity, in Developmental Toxicology: Risk Assessmentand the Future, Hood, R.D., Ed., Van Nostrand Reinhold, New York, 1990, chap. 4.31. Ranganathan, S. and Hood, R.D., Effects of in vivo and in vitro exposure to rhodamine dyes onmitochondrial function of mouse embryos, Teratogen. Carcinog. Mutagen., 9, 29, 1989.32. Ranganathan, S., Churchill, P.F., and Hood, R.D., Inhibition of mitochondrial respiration by cationicrhodamines as a possible teratogenicity mechanism, Toxicol. Appl. Pharmacol, 99, 81, 1989.33. Nau, H. and Scott, W.J., Drug accumulation and pH i in the embryo during organogenesis and structureactivityconsiderations. Mechanisms and models in toxicology, Arch. Toxicol., Suppl., 11, 128, 1987.34. Kalter, H., The relation between congenital malformations and prenatal mortality in experimentalanimals, in Porter, I.H. and Hook, E.B., Eds., Human Embryonic and Fetal Death, Academic Press,New York, 1980, p. 29.35. Brent, R.L., Editorial comment. Definition of a teratogen and the relationship of teratogenicity tocarcinogenicity, Teratology, 34, 359, 1986.36. Gaylor, D.W., Sheehan, D.M., Young, J.F., and Mattison, D.R., The threshold question in teratogenesis,Teratology, 38, 389, 1988.37. Giavini, E., Evaluation of the threshold concept in teratogenicity studies, Teratology, 38, 393, 1988.© 2006 by Taylor & Francis Group, LLC

14 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION38. Sheehan, D.M., Literature analysis of no-threshold dose-response curves for endocrine disruptors,Teratology, 57, 219, 1998.39. Calabrese, E.J. and Baldwin, L.A., The frequency of U-shaped dose responses in the toxicologicalliterature, Toxicol. Sci., 62, 330, 2001.40. Rodricks, J.V., Hormesis and toxicological risk assessment, Toxicol Sci., 71, 134, 2003.41. Johnson, E.M. and Christian, M.S., When is a teratology study not an evaluation of teratogenicity?J. Am. Coll. Toxicol., 3, 431, 1984.42. Palmer, A.K., The design of subprimate animal studies, in Wilson, J.G. and Fraser, F.C., Eds.,Handbook of Teratology, Vol. 4., Research Procedures and Data Analysis, Plenum, New York, 1978,chap. 8.43. Johnson, E.M., and Gabel, B.E.G., Application of the hydra assay for rapid detection of developmentalhazards, J. Am. Coll. Toxicol., 1, 57, 1982.44. Johnson, E.M., and Newman, L.M., The definition, utility and limitations of the A/D ratio concept inconsiderations of developmental toxicity, Teratology, 39, 461, 1989.45. Hood, R.D., A perspective on the significance of maternally-mediated developmental toxicity, Regulat.Toxicol. Pharmacol., 10, 144, 1989.46. Rogers, J.M., Barbee, B., Burkhead, L.M., Rushin, E.A., and Kavlock, R.J., The mouse teratogendinocap has lower A/D ratios and is not teratogenic in the rat and hamster, Teratology, 37, 553, 1988.47. Daston, G.P, Rogers, J.M., Versteeg, D.J., Sabourin, T.D., Baines, D., and Marsh, S.S., Interspeciescomparison of A/D ratios: A/D ratios are not constant across species, Fundam. Appl. Toxicol., 17,696, 1991.48. Adeeko, A., Li, D., Doucet, J., Cooke, G.M., Trasler, J.M., Robaire, B., and Hales, B.F., Gestationalexposure to persistent organic pollutants: Maternal liver residues, pregnancy outcome, and effects onhepatic gene expression profiles in the dam and fetus, Toxicol. Sci., 72, 242, 2003.49. MacGregor, J.T., Editorial: SNPs and Chips: Genomic data in safety evaluation and risk assessment,Toxicol. Sci., 73, 207, 2003.© 2006 by Taylor & Francis Group, LLC

CHAPTER 2Experimental Approaches to Evaluate Mechanismsof Developmental ToxicityElaine M. Faustman, Julia M. Gohlke, Rafael A. Ponce, Thomas A. Lewandowski,Marguerite R. Seeley, Stephen G. Whittaker, and William C. GriffithCONTENTSI. Introduction ..........................................................................................................................16A. Definitions....................................................................................................................16B. General Mechanistic Considerations...........................................................................16C. Guidelines for Evaluating Critical Events ..................................................................18D. Levels of Mechanistic Inquiry.....................................................................................20E. Genomic Conservation ................................................................................................21F. Processes of Organogenesis and Implication of Three-DimensionalContext of Evaluation..................................................................................................231. Cell-Signaling Pathways........................................................................................232. Cell-Cell Communication......................................................................................27II. Examples of Mechanistic Approaches.................................................................................28A. Mitotic Interference .....................................................................................................291. 5-Fluorouracil ........................................................................................................292. Radiation................................................................................................................303. Antitubulin Agents.................................................................................................304. Methylmercury.......................................................................................................30B. Altered Energy Sources...............................................................................................321. Inhibitors of Mitochondrial Respiration................................................................322. Cocaine ..................................................................................................................333. Rhodamine Dyes....................................................................................................33C. Enzyme Inhibition .......................................................................................................331. Methotrexate ..........................................................................................................332. Chlorpyrifos ...........................................................................................................343. 5-Fluorouracil ........................................................................................................354. Mevinolin...............................................................................................................35D. Nucleic Acids...............................................................................................................351. Hydroxyurea ..........................................................................................................352. Cytosine Arabinoside.............................................................................................36E. Mutations .....................................................................................................................371. Alkylating Agents ..................................................................................................382. Aromatic Amines...................................................................................................393. Radiation................................................................................................................3915© 2006 by Taylor & Francis Group, LLC

16 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONF. Alterations in Gene Expression...................................................................................401. Retinoic Acid .........................................................................................................402. Dioxin ....................................................................................................................413. New Directions in “Omic” Research ....................................................................42G. Programmed Cell Death ..............................................................................................421. Retinoic Acid .........................................................................................................432. Dioxin ....................................................................................................................433. Ethanol...................................................................................................................44III. Conclusions ..........................................................................................................................46Acknowledgments............................................................................................................................46References ........................................................................................................................................47I. INTRODUCTIONDaedalus, an architect famous for his skill, constructed the maze, confusing the usual marks ofdirection, and leading the eye of the beholder astray by devious paths winding in different directions.Thanks to the help of the princess Ariadne, Theseus rewound the thread he had laid, retraced his steps,and found the elusive gateway…The purpose of this chapter is to review methodological approaches for elucidating the mechanismsby which chemical and physical agents cause or contribute to dysmorphogenesis and teratogenicity.Emphasis is given to the approaches rather than to agent-specific mechanisms, and the focus is onhow molecular and cellular information is combined to evaluate mechanistic hypotheses.A. DefinitionsIn order to develop this paper, a few common definitions must be discussed. Mechanisms of toxicaction is used to refer to the detailed molecular understanding of how chemicals can impair normalphysiological processes and hence, produce developmental toxicity. Mechanistic information caninclude biochemical, genetic, molecular, cellular, and/or organ systems information. 1 Mode of actionfor developmental toxicants is frequently used to refer to the identification of critical steps that canexplain how an agent can produce developmental toxicity and usually refers to a less detailed butmore comprehensive description of the overall process of developmental toxicity. This chapter willinclude a discussion of approaches used for understanding mechanisms for all four endpoints ofdevelopmental toxicity: lethality, growth retardation, morphological defects (teratogenicity), andfunctional impacts. Throughout this chapter, the general term developmental dynamics is used todescribe the genetic, biochemical, molecular, cellular, organ, and organism level processes thatchange throughout development and that define and characterize the developing organism at eachlife stage. 2 The term kinetics is used to refer to the absorption, distribution, and metabolism ofchemicals because many of our discussions of developmental dynamics directly relate to the amountand form of the environmental or pharmacological agent that reaches the developing organism.B. General Mechanistic ConsiderationsFigure 2.1 shows an overall framework for assessing the effects of a toxicant on development. 1,3This figure illustrates how both kinetic and dynamic considerations are needed to link exposurewith developmental outcome. It provides a context for collecting mechanistic data and for orderingsequence of events data that structures a mode of action hypothesis. This framework has beenOvid© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 17Risk assessmentExposure assessment Toxicity assessment Risk characterizationExposure Toxico kinetics Toxico dynamics OutcomeInhalation Absorption Adolescent Normal parametersOralDevelopmental disorderDistributionChildDermal• LethalityMetabolismNewborn• Growth retardationEliminationConceptusOrgan, tissuesCellularOrganelleMolecular• Malformation• Altered functionCell signallingFigure 2.1Overall framework for assessing the effects of a toxicant on development. (Adapted from NationalResearch Council, Scientific Frontiers in Developmental Toxicity Risk Assessment, National AcademyPress, Washington, 2000, p. 354 and from Faustman, E. et al., IRARC Technical Report onDevelopmental Toxicity, Institute for Risk Analysis and Risk Communication, University of Washington,Seattle, WA, 2003. With permission.)Table 2.1 Example mechanisms for developmental toxicityMitotic interferenceAltered membrane function or signal transductionAltered energy sourcesEnzyme inhibitionAltered nucleic acid synthesisMutationsGene and protein expression changesAlterations in programmed cell deathmodified from the original National Academy of Sciences (NAS) framework to illustrate that theaffects of exposure during development can occur in utero, in newborns, in childhood, and in earlyadulthood. Manifestations of early exposures sometimes cannot be observed until adulthood. 4The study of mechanisms of toxicity is of vital importance not only for the insights providedinto the events underlying adverse developmental outcomes, but also for the information gainedconcerning the processes involved in normal development. Recently, there has been an increasedinterest in mechanistic information as a result of legislative actions. For example, in the FoodQuality Protection Act, exposure to agents that have common mechanisms of action should beconsidered in a cumulative manner. This has led to the joint evaluation of organophosphates inregard to their developmental neurotoxicity.Table 2.1 includes a list of potential mechanisms first proposed by Wilson 5 that included thefollowing general categories: mitotic interference, altered membrane function or signal transduction,altered energy sources, enzyme inhibition, altered nucleic acid synthesis, and mutations. Becausethese processes play essential roles in embryogenesis and normal development, it is logical toexpect that alterations may result in developmental toxicity, and the research literature is repletewith proof of this assumption. With our increased understanding of the molecular mechanisms© 2006 by Taylor & Francis Group, LLC

18 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 2.2Temporal associationDosage relationshipGuidelines for assessment of proposed mechanistic pathways in chemicallyinduced developmental toxicityStructure-activity relationshipsStrength of associationConsistency of associationCoherenceDoes developmental toxicity precede, occur simultaneously with, or follow theinitial event?Does the potential mechanistic event occur at or below those doses that resultin the developmental toxicity?Is there a dose-response relationship between exposure and severity of thedevelopmental outcome?Do chemicals with similar structures cause similar developmental outcomes?Is the proposed mechanistic process strongly or weakly linked to theappearance of the developmental outcome?Are the proposed mechanistic processes required for the appearance of thedevelopmental outcome?Does modification of the mechanistic event, or of one step in the process, alteror eliminate the adverse developmental outcome?Is there a molecular basis for the proposed mechanism of action by thechemical or physical agent that elicits the initial event?underlying normal development, we can now propose additional mechanisms, including perturbationsin gene and protein expression and programmed cell death. However, even with inspectionat this more basic level of action, only partial segments of the mechanistic path from initial insultto dysmorphogenesis are understood for even the most well-characterized developmental toxicants.In most circumstances, only phenomenological information is available. 1C. Guidelines for Evaluating Critical EventsThis chapter emphasizes the need to consider the underlying causes of developmental toxicity ratherthan relying on phenomenology. To accomplish this, a series of guidelines has been developed toascertain the significance of postulated critical events in the mechanistic pathway leading to anadverse developmental outcome. Such guidelines need to be considered when evaluating the validityof the relationship between an initial toxic insult and teratogenicity. These guidelines have beendeveloped from basic pharmacological principles of drug action 6 and from epidemiologicalapproaches, such as the Bradford Hill criteria of causality. 7 The general adaptability of theseapproaches is evidenced by their recent use in the International Programme on Chemical Safety(IPCS) Harmonization Approaches for Risk Assessment 8 and in EPA’s Carcinogen Risk AssessmentGuidelines. 9 Table 2.2 lists these assessment guidelines.The first guideline is the issue of temporal association. In this assessment, the question is thetemporal association between any altered development and the potential initial mechanistic event.Obviously, for this event to be a critical event in the process of developmental toxicity that eventmust precede or occur simultaneously with the pathology. Complex temporal relationships associatedwith development can often complicate analysis, and events that might be labeled as nontemporallyassociated are missed if the biology of developmental processes is not a prime factor inreviewing temporal associations.Because of the hierarchical nature of tissue organization in the developing organism, patternsof affected tissue can provide important temporal mechanistic clues. For example, Figure 2.2illustrates that if dysmorphogenic alterations are observed in cardiac cells and sensory and stomachepithelia (i.e., endoderm, mesoderm, and ectoderm), then events occurring during blastula andgastrula stages might be the first to be evaluated as potential critical events for the mechanism ofdysmorphogenesis linking these three responses. Nevertheless, later events occurring separately ineach of these tissue types (such as changes in cell proliferation or changes in receptor expression)could also explain the common responses in these three tissues. Thus, such temporal associationscan be used as initial clues, but mechanistic investigation must always be open to multiple explanationsfor the same response.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 19Blastula Gastrula Neurula Tailbud Cell types in adultEctodemNeural plateNeural crestEpidermisBrainSpinal cordAutonomic systemParts of skullMelanocytesSkinPlacodesNeurons, gliaCartilageMelanocytesEpidermis, gland cellsLens, sensory epitheliaCephalicMuscle, cartilage, fibroblastsDorsalNotochordChordocytesTrunkSomitesMuscle, cartilageEGGMesodermVentralLateral plateKidneyHaemopoietic systemLimbsHeartGutTubulesErythrocytes, lymphocytesMuscle, cartilage, fibroblastsCardiac muscleSmooth muscleBlood islandsErythrocytesPharynxLungsEndodermAlimentary canalStomachLiverIntestineCharacteristic epitheliaYolk cellsFigure 2.2The hierarchical embryonic origins of tissues and cells within the vertebrate embryo. (From Slack,J., From Egg to Embryo: Regional Specification in Early Development, 2nd ed., CambridgeUniversity Press, Cambridge, 1991, p. 348. With permission.)The second guideline concerns the establishment of a dose-response relationship for the proposedcritical mechanistic events. If exposure to a suspected developmental toxicant produces adose-related increase in malfunctions, then the possibility that the chemical is a developmentaltoxicant, causing the adverse outcome, is strengthened. Lack of a dose-response relationship,however, does not rule out the possibility that the suspected toxicant is a developmental toxicant.For example, Selevan and Lemasters 10 have dramatically illustrated the concept of competing risk.Although there might not be an observable, dose-related increase in malformations, the effects ofthe suspected developmental toxicant may be to increase embryolethality. Thus, at higher doses,fewer embryos survive to manifest increased malformations.The third guideline for assessing proposed mechanisms of action for developmental toxicantsrelates to structure activity relationships (SARs) and whether they exist for the compound underinvestigation. Good SAR examples for developmental toxicants are published in the literature foralkylating agents, 11–13 retinoic acid derivatives, 14,15 alkoxy acids, 16 short-chain carboxylic acids© 2006 by Taylor & Francis Group, LLC

20 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 2.3Intracellular eventsIntercellular eventsOrgan level eventsOrganism level eventsLitter responsesLevels of mechanistic inquiry for developmental toxicityBiochemical and molecular mechanisms of action define key intracellular events forboth normal and abnormal developmental responses.Specific cell-cell interactions and activities define activities for specialized cellpopulations.Specialized functions of organs define organ development and function.Embryonic and fetal responses are defined by the collective responses of organ andintra- and intercellular events.The combined embryo and fetal responses of a litter are defined within the singlematernal unit.(valproic acid derivatives), 17 and phenols. 18 Note that many of these SARs are determined using invitro as well as in vivo investigations to establish relationships.The fourth guideline concerns the strength and consistency of occurrence of the adverse outcomewith the postulated critical mechanistic event. For example, if the proposed mechanisms of actionfor compound X is that it inhibits neuronal cell division by 50% in the brain during early brainformation, causing microcephaly, then to determine the strength of this mechanistic association,two types of model experiments could be planned. First, the investigator could determine if otheragents that cause a comparable decrease in cell division at this time in development also causemicrocephaly. Secondly, the investigator could see if blocking the effects of compound X on celldivision would reverse the incidence of microcephaly. Cross-species extrapolation of results couldalso increase the consistency of these observations as key mechanistic processes. Although theseexample observations provide clues, failure to see these changes does not mean that the mechanistichypothesis must be abandoned.The last guideline relates to the importance of coherence in the overall mechanistic hypothesis. Ifa possible molecular or cellular basis can be described for the proposed mechanism of action, then thiscoherence provides a stronger degree of confidence in the postulated pathway. If no molecular basisis found, then the proposed mechanism may have a difficult battle for acceptance because it may be amechanism whose conception may have outpaced related molecular experimentation.D. Levels of Mechanistic InquiryOur current lack of understanding of the events underlying teratogenicity reflects the complexityof the developmental process. It may never be possible to describe every molecular or cellular eventthat ultimately leads to dysmorphogenesis. However, of the myriad of potential effects elicited bychemical and physical agents on embryonic development, it is probable that only a relative fewrepresent critical events responsible for developmental toxicity. Therefore, it is essential to identifythe key events, based on an understanding of the toxicological properties of the agent and thebiological processes involved. To gain an understanding of developmental toxicity, investigationsmust focus on multiple biological levels. The initial molecular and subcellular events must bedefined along with key processes occurring at the cellular, tissue, and organ level. Investigationsat the organ system and organism level are then required. Table 2.3 lists these levels of mechanisticinquiry for developmental toxicity. Recognition of these levels is critical for several reasons. First,mechanisms are frequently defined at only a single level. Thus, a cellular mechanism of action willbe defined in isolation from events occurring at higher levels of organization. Later investigatorsworking only at the fetal level may dismiss these mechanistic observations because strict temporalor dose-response relationships may be unclear at the higher level of investigation. However, if bothlevels are examined, both observations can be confirmed and a broader appreciation for the mechanisticcomplexities involved can be realized. The possibility of multiple mechanisms should berecognized; an adverse developmental outcome will probably not be attributable to a single eventbut rather to a cascade of events.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 21Table 2.4Critical intercellular signaling pathways important for development aPeriod during Developmentwhen Signaling Pathway Is UsedEarly (axis specification, germlayer specification, left-rightasymmetry) and continued in alllater stages.Middle (during organogenesis andcytodifferentiation) and continuedin all later stages.Late (after cell types havedifferentiated). Used in fetal,larval, and adult physiology.Pathway1. Wnt pathway2. Hedgehog pathway3. TGF receptor (ser/thr kinase) pathway4. Receptor tyrosine kinase (small G protein) pathway5. Notch/Delta pathway6. Cytokine receptor (cytoplasmic tyrosine kinases) JAK/STAT pathway7. IL1/Toll NFkB pathway8. Nuclear hormone receptor pathway9. Apoptosis pathway10. Integrin pathway11. Receptor phosphotyrosine phosphatase pathway12. Receptor guanylate cyclase pathway13. Nitric oxide receptor pathway14. G-protein coupled receptor (large G protein) pathway15. Cadherin pathway16. Gap junction pathway17. Ligand-gated cation channel pathwayaThe mammalian fetus uses all 17 pathways.Source: Adapted from National Research Council, Scientific Frontiers in Developmental Toxicity Risk Assessment,National Academy Press, Washington D.C., 2000, p. 354.E. Genomic ConservationInherent in most toxicological studies is the premise that chemically induced effects in animals arepredictive and instructive for understanding the potential for a chemical to alter development inhumans. Recent advances in developmental and molecular biology make these assumptions evenmore important. One of the most exciting advances is the determination that most of developmentis controlled by approximately 17 cell-signaling pathways and that these signaling pathways aregenomically conserved. 1 These pathways are listed in Table 2.4.A generalized signal transduction pathway is the basic mechanism underlying each of thesecell-signaling pathways. This multistep process is composed of a series of switchlike intermediatesthat are activated by a receptor-mediated signal, which ultimately activates a protein kinase. Thetarget protein is hence phosphorylated and is either activated or inactivated. Target proteins in thesesignaling cascades include proteins that are integral to processes of transcription, translation, cellcycling, cell migration, differentiation, etc. Fourteen of the cell-signaling pathways shown inTable 2.4 involve transmembrane receptors and two involve intracellular receptors. 1These findings on genomically conserved pathways have had some important implications forimproved approaches for mechanistic studies. First, the function of many of these pathways acrossmodel organisms has been determined by transgenic studies. Genes for the cell-signaling processeshave been cloned, and a variety of transgenic technologies have been applied to evaluate theirsignificance. Knockout and/or null mutations, overexpression, or miss-expression of genes can bestudied and used to evaluate the role that these specific genes and signaling pathways may play indevelopment. Table 2.5 shows examples of such transgenic studies, where the phenotypes of mousemutants lacking components of specific cell-signaling pathways are evaluated. For example, if theWnt-1 pathway is knocked out, the offspring are viable to adulthood, but no midbrain, cerebellum,or rhombomere 1 is present. Mice obviously also display behavioral defects. 1,19 Likewise, if Sonichedgehog (Shh) is knocked out, perinatal death is observed and embryos have evidence of cyclopiaand spinal cord, axial skeleton, and limb defects. 1,20Table 2.5 shows results from transgenic animal studies for selected key cell-signaling pathways;however, there are several concepts that are not captured in this table. One issue is that© 2006 by Taylor & Francis Group, LLC

22 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 2.5 Example phenotypes of mouse mutants lacking components of signalingpathwaysSignalingComponentVariability ofNull Mutant Phenotype of Null Mutant Ref.Wnt PathwayWnt-1 Adulthood No midbrain, cerebellum and rhombomere 1;behavioral deficits19, 320Transforming Growth Factor β PathwayTGFβ 1 Adulthood Immune defects, inflammation 297TGFβ 2 Perinatal death Defects of heart, lung, spine, limb andcraniofacial and spinal regions298,320,321BMP5 Adulthood Thin axial bones; abnormal lung, liver, ureter, 299and bladder; like a short ear mutantBMP7 Adulthood Defects in eye and kidney; skeletalabnormalities; hindlimb polydactyly300, 323Hedgehog PathwaySonic hedgehog Perinatal death Cyclopia, defects of spinal cord, axial 20skeleton, and limbsPatched receptor Homozygotes, Open neural tube 301early lethalityHeterozygotes,adulthoodRhabdomyosarcomas, hindlimb defects,large size. Like Gorlin syndrome in humans302, 324Nuclear Receptor PathwayProgesteronereceptorAdulthood Males normal; females: no ovulation, nomammary glands, uterine hyperplasia297,303,304Retinoic acid Adulthood Fertile, small size, transformations of cervical 305receptor βvertebraeRARα and RARβ Perinatal death Visceral abnormalities, reduced thymus and 305spleenSource: Adapted from National Research Council, Scientific Frontiers in Developmental ToxicityRisk Assessment, National Academy Press, Washington D.C., 2000, 354. With permission.as organisms become more complex, there is increasing redundancy in the downstream pathwaysof the 17 cell-signaling processes. If extensive redundancy exists, then interpretation of the significanceof specific pathways in transgenic animal studies is complicated. For example, in manycases where high levels of redundancy exist in a process, knockout animals will have minimalphenotypic changes. These may result from the fact that in rodent models there may be multipleligand genes (i.e., 24 TGFβ genes and 11 Wnt ligand genes in mice versus 3 to 5 TGFβ genes and1 to 3 Wnt genes in Drosophila). 1 In addition, when a single gene is knocked out in mice, thedevelopmental defect may be very subtle since that specific gene may only be expressed in a verynarrow temporal and tissue-specific context where no related genes are expressed to provideredundant function.These transgenic models are of methodological use for studying developmental toxicity inseveral ways. First, many researchers compare the phenotype of transgenic mouse models with thephenotype that can arise from treatment of animals with developmental toxicants. Such an approachwas useful for studies of Veratrum alkaloids, where cyclopamine produced cyclopia in livestockthat fed on plants containing Veratrum alkaloids. Molecular investigations of cyclopamine haverevealed that it can interfere with Shh signaling. 21–23 Mouse knockout studies showed that geneticmanipulation of the Shh pathway could result in the same types of defects.Other ways that developmental toxicants have been found to act through conserved cellsignalingpathways are portrayed in Table 2.6, which shows a list of examples of receptor-mediated© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 23developmental toxicity. These include both environmentally relevant examples (i.e., TCDD andcyclopamine) and examples of interest from medicinal chemistry (retinoic acid and diethylstilbestrol[DES]). In all of these cases, knowledge about the normal role of the receptor aided the mechanisticstudies of nonendogenous ligands.In some cases, a developmental toxicant can interact directly or indirectly with a receptor toactivate a receptor inappropriately (agonist) or inhibit signaling via the normal ligand (antagonist).If the response of a developmental toxicant or a receptor is to activate the receptor but to producea less than maximal response, that agent is known as a partial agonist. If the developmental toxicantcan cause a decrease in activity from baseline, then the agent can be considered a negative agonist.Hence, mechanistic approaches for evaluating receptor-ligand mediated developmental toxicity caninclude assessment of receptor binding, but more importantly, a quantitative characterization of thetype of response when the developmental toxicant binds is usually more informative. Pioneeringwork by Nebert showed the importance of Ah receptors (AhR) in mediating polycyclic aromatichydrocarbon-mediated developmental toxicity. 24,25 This was demonstrated with genetically differentstrains of Ah-responsive and nonresponsive mice.Other examples included in this table are the myriad of studies on retinoic acid receptors wheregood structure-activity relationships are available for developmental toxicity because of the productionof numerous candidate drugs. Depending upon receptor subtype, timing of exposure, anddose, almost all organ systems can be affected via this receptor interaction. 1F. Processes of Organogenesis and Implication of Three-DimensionalContext of Evaluation1. Cell-Signaling PathwaysThe importance of evaluating cell-signaling pathways within the overall context of organogenesisis elegantly illustrated in our knowledge of limb development from over 50 years of experimentalembryology and recent intense molecular developmental biology studies. 1Figure 2.3 shows the development of limb buds in vertebrates (tetrapods) and shows the complexinteractions of precise temporal and spatial signaling that is required for organ development. Thisfigure also shows the multitude of signaling pathways controlling proliferation that are required toestablish the three-dimensional morphology of this organ. 1,26,27 For example, to determine thepossible mechanistic ramifications of chemically induced changes in Shh for limb bud development,one would need to understand the impacts this signal would have initially on BMP2 signaling andcell proliferation activity in the bud mesoderm, and subsequently on the overall anterioposteriorgradient that determines bud extension. The dorsal epidermis secretes Wnt-7a onto the mesenchyme,where it induces expression of the LMX-1 gene and suppresses the engrailed-1 gene. If Wnt-7asignaling is defective, double ventral limbs can form. 1These changes would then need to be evaluated in terms of possible changes in the dorsalventral gradient and possible alterations in Wnt-7a/engrailed gene expression. Such three-dimensionalevaluations are difficult to impossible to glean from simple cell experiments. They requiremodel systems that have complex three-dimensional cell-signaling interactions. Mechanistic cluesto examine Shh or Wnt-7a cell-signaling pathways can come from more simplified systems whenrelevant dose-response relationships are established. It has been postulated that thalidomide canact by reducing cell proliferation in the progress zone (PZ), and this can result in prolonged contactof cells with fibroblast growth factor (FGF) secreted by the apical ectodermal ridge (AER). 1,28 Otherresearchers have postulated that thalidomide can interfere with integrin gene expression, henceinhibiting angiogenesis and subsequently proliferation. 29 These are just two of the myriad ofproposed hypotheses for the action of thalidomide; however, knowledge of normal developmentalbiology and of these complex signaling processes allows one to set up studies to methodologicallyand quantitatively determine the validity of such mechanisms. Using the guidelines presented in© 2006 by Taylor & Francis Group, LLC

Table 2.6Receptor(Official Name a )Examples of receptor-mediated developmental toxicityEndogenousLigandsBasic Helix-Loop-Helix Transcription SuperfamilyAryl hydrocarbon AHR Unknown Agonists: TCDD and relatedpolycyclicsNuclear Hormone ReceptorsAndrogenAR (NR3C4)EstrogenERa, ERb (NR3A1 and 2)GlucocorticoidGR (NR3C1)Retinoic acidRARα, β, and γ(NR1B1, 2, and 3)RXRα, β, and γ(NR2B1, 2, and 3)Thyroid hormoneTRα and β(NR1A1 and NR1A2)TestosteroneDihydro-testosteroneEstradiolCortisolAll trans and 9-cisretinoic acidsDevelopmentally ToxicLigand and Modifier Typical Effects Recent ReferencesAgonists: 17 a–ethinyltestosteroneand relatedprogesteronesAntagonists: b FlutamideAgonist: DESAntagonist: tamoxifen,clomiphene—weakAgonists: cortisone,dexamethasone, triamcinoloneAgonists: numerous natural andsynthetic retinoidsCleft palate, hydronephrosis 25, 31, 306, 325Agonists: Masculinization of femaleexternal genitalsAntagonists: Inhibition of Wolffian duct andprostate development and feminization ofexternal genitals in malesAgonist: various genital-tract defects inmales and females307308, 326Cleft palate 309, 327Almost all organ systems can be affected 226, 310–312, 3289-cis retinoic acid Antagonists: BMS493, AGN193109, and othersThyroxine (T4 and T3) Antagonist: nitrophen Lung, diaphragm, and harderian-glanddefects31324 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Hedgehog receptorPatchedMembraneEndothelin receptors Aand BSonic, Desert, andIndian HedgehogsEndothelins 1, 2, and 3Veratrum alkaloids:cyclopamine(mechanism unclear)Antagonists: L-753, 037, SB-209670, SB-217242Cyclopia, holoprosencephaly 21, 22, 329Craniofacial, thyroid, and cardiovasculardefects, intestinal aganglionosis(Hirschsprung’s disease)314–316, 330, 331Cation ChannelsDelayed-rectifying IKr Potassium ion Inhibitors: almokalant,Digit, cardiovascular, orofacial clefts 317, 318dofetilide, d-sotalolaNuclear Receptors Committee 1999.bAlso, 5-alpha reductase inhibitors (e.g., finasteride) affect prostate and external genitals. 319Source: Adapted from National Research Council, Scientific Frontiers in Developmental Toxicity Risk Assessment, National Academy Press, Washington D.C.,2000, 354. With permission.EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 25© 2006 by Taylor & Francis Group, LLC

26 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA.Locationof thelimb budsNeural tubeNotochordLimb budDorsal aortaCoelomB.ProximalAnteriorGremlinDistal BMP-2ShhFGFPosteriorProgress zone (PZ)of proliferating cellsApical ectodermal ridge(AER)Zone of polarizing activity(ZPA)C.DorsalWnt7aApical ectodermal ridge(AER)ProximalVentralDistalenenVentral expression of theengrailed gene blocksWnt7a gene expressionventrallyFigure 2.3Three-dimensional development of vertebrate limbs (tetrapod). This figure shows the three axesof limb development for four legged vertebrates: anteroposterior, proximodistal and dorsoventral.(A) Location of limb buds in relationship to an overall cross-sectional view of vertebrate development.(B) Cross-section of a limb bud showing the anteroposterior and dorsoventral axis. Thelocation of the zone of polarizing activity (ZPA) is shown, as is the apicalectodermal ridge (AER)and the progress zone (PZ) of proliferating cells. The signaling feedback between Sonic hedgehog(Shh), gremlin, BMP2, and FGF is illustrated. (C) Dorsoventral and proximodistal axis and therelationship of engrailed (en) gene and its inhibition of Wnt7a gene expression ventrally. (Adaptedfrom National Research Council, Scientific Frontiers in Developmental Toxicity Risk Assessment,National Academy Press, Washington, D.C., 2000, p. 354. With permission.)Table 2.2, researchers can determine, for example, if the temporal relationship, dose, and consistencyin changes in cell proliferation of the PZ correlate with the incidence of phocomelia.Likewise, researchers can determine the temporal relationship, dose, and consistency of specificlimb abnormalities by integrating gene expression changes induced with thalidomide. Because ofthe cross-species similarity in appendage or limb patterning, other approaches for mechanisticstudies could include looking at the consistency of these observations in Drosophila and chickembryos, as similar conserved gene domains exist. For example, similar domains exist for arthropodsand chordates for WG-HH-DPP and Wnt-Shh-BMP as do similar expressions of En, Ap, andLMX selector genes. 1 One caution that must be kept in mind when performing cross-species© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 27experiments with thalidomide is that although the cell-signaling processes controlling the developmentaldynamic processes may be conserved, dissimilar conservation of drug metabolism genes inthese various species may result in toxicokinetic differences. Hence, the use of “active metabolites”and correction for dose-to-target concentrations of thalidomide and its metabolites may be veryimportant for such mechanistic determinations. Such investigations would help to weave togethersignificant mechanistic clues for thalidomide effects on limb development.2. Cell-Cell CommunicationThe interaction between cells and their environment during fetal development plays a significantrole in both cell differentiation and morphogenesis. This interaction leads to information transferand can occur through direct cell-cell contact, through the activation of membrane-bound cellularreceptors, or through the associations created between a cell and the surrounding extracellularmatrix. Cell fate determination by specific cell interactions of the cell and its surrounding environmentis conditional specification and is dependent upon extracellular factors. 30 If the extracellularsignal causes a specific manifestation of one differentiation fate over others, the process is referredto as instructive induction. In contrast, the other form of conditional specification, permissiveinduction, results from a cell that already is committed to a specific differentiation path but onlyexpresses this differentiation phenotype after exposure to a signal. A good example of this is seenin limb development (see Figure 2.3) where a complex three-dimensional morphogenic signalinggradient is established across the limb bud. 1,30Appositional induction results from tissue interactions where signaling and responding tissuescome together and a common response is induced in the contact region. The fundamental importanceto normal development of this system was recognized by Wilson 5 in a discussion on the role ofaltered cell membrane function as a contributor to teratogenesis. Under the most severe chemicalexposure conditions, altered membrane integrity will likely result in cytolysis and cell death. Thisoccurs primarily as a result of the inability of a cell to maintain a normal physiologic ionic balanceand osmolarity due to an altered membrane permeability. Thus, at this extreme, monitoring cytotoxicityis an indirect and nonsubtle measure of altered membrane integrity. Less extreme examplescan result in functional changes, such as the excess cell proliferation seen at the fusion points inTCDD-exposed palatal shelves. 31–33Although much mechanistic information is lacking, there is some understanding of the role ofaltered membrane function and the importance of intracellular communication, cell adhesion, cellmigration, cell shape, and cellular receptors in developmental toxicology. Of particular interest arerecent investigations into the role of adhesion molecules on normal membrane function in cellmigration.A key cellular process occurring during differentiation is the migration of cells to new locations.Examples of this include: neural crest cells, which develop into a variety of cell types; precardiacmesodermal cells, which form the heart; and neurons, which migrate to various regions of thecerebrum from the ventricular zones following cell division. 34 In general, cell migration reliesheavily on cell-cell interaction, particularly cell adhesion. For example, the migrating neuron isbelieved to receive guidance cues from the extracellular environment and the extracellular matrix,as well as through contact between the neuron and supporting glial and neuronal cells. 35–46 Whenthese cues are altered by changing the chemical composition of the extracellular matrix, dramaticchanges in differentiation patterns occur; this is one technique that can be used to define the essentialnature of this matrix for developmental processes.Intracellular cytoskeletal components, such as microtubules and actin, provide a structural basisfor migration, 47,48 while extracellular cell adhesion molecules provide support and guidance for themigrating cell by offering a preferred substrata. 35,49 Extracellular adhesion molecules are also likelyto be involved in signal transduction, which results in directional guidance cues and cell motility. 48The proper functioning of the intracellular cytoskeleton and extracellular adhesion molecules is© 2006 by Taylor & Francis Group, LLC

28 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONcrucial for normal nervous system development; alterations in either may result in a variety ofcortical malformations. 43 For example, in cleft palate, faulty adhesion may underlie the failure ofepithelial fusion even though the palatal shelves may be in close apposition. 35,50Adhesion molecules are involved in numerous neurodevelopmental processes, including neuriteoutgrowth (integrin, neural cell adhesion molecule [N-CAM], N-cadherin, L1), peripheral nerveregeneration (L1, N-CAM), nerve target adhesion (N-CAM), regulation of intracellular and extracellularionic composition (adhesion molecule on glia, AMOG), cell-cell stabilization (N-CAM180), and others. 48 Alterations to these proteins, therefore, may play a role in toxicant-inducedcortical dysfunction. For example, N-CAM, an adhesion molecule involved in synaptogenesis andmigration, undergoes maturation during development from a sialic-acid–poor form in early embryogenesisto a sialic-acid–rich form, and finally to a sialic-acid–poor form again in the mature adult.Alterations to the sialylation state of N-CAM occur during cell migration and synaptogenesis. 51Low-level lead exposure has been associated with an inhibition of normal developmental N-CAMdesialylation, 52 and it has been proposed based on dose and time exposure studies that this leadinducedalteration to the normal maturation of N-CAM results in the observed abnormal synaptogenesisin the developing central nervous system. 51 Similar alterations to N-CAM maturation havealso been reported following exposure to other metals, such as methylmercury; 53 these findingssupport a role for faulty adhesion molecule function in methylmercury-induced neuronal ectopia.Another developmental process that relies heavily on a normal membrane function is theextension of neurites by the young neuron during the formation of a synaptic network. Neuronalfasciculation is accomplished through the activity of a motile growth cone found on the extremityof the neurite. As the growth cone advances away from a relatively stationary cell body, the axonis developed. Termination of growth cone activity may occur through contact inhibition involvingcell-to-cell communication, but this process is poorly understood. 54 Therefore, chemical agents thatalter cell-cell communication may disturb synaptogenesis and perhaps other developmental processes,such as cell proliferation and differentiation. For example, Trosko and colleagues 55 haveforwarded a series of arguments regarding the role of abnormal cell-cell communication in teratogenesis.In this work, they summarized evidence that gap junctional communication is involved innormal differentiation and development, and extend the hypothesis that a disruption in this communicationduring histo- or organogenesis will result in pathology. The identification of the integrin,cadherin, and gap junction pathways in the 17 key cell-signaling pathways that define developmentsupports these concepts and highlights the importance of these processes in normal development.In summary, the cell membrane is a key site where developmental toxicants may act. However,there is very little direct evidence implicating altered membrane function as the critical step in theetiology of developmental toxicity. Additional approaches to ascertain the role of cell communicationare needed.II. EXAMPLES OF MECHANISTIC APPROACHESThe following review is organized according to principal mechanisms by which chemical andphysical agents are thought to elicit dysmorphogenesis and developmental toxicity. Examples areprovided of chemical and physical agents that have been shown to be associated with the mechanismbeing discussed, and methodological approaches used to make those associations are highlighted.The purpose of this chapter is not to be all-inclusive in terms of the mechanistic knowledge forthe developmental toxicants. Rather, the purpose is to highlight several of the proposed mechanistichypotheses for developmental toxicity and to highlight examples of the types of experimentalapproaches employed and the results that were generated to test these hypotheses.In this review, we will examine approaches used for evaluating mitotic interference, altered cellsignaling, enzyme inhibition, mutation, alterations in gene expression, and programmed cell death.In a few cases where detailed kinetic and dynamic information is available, some explanation on© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 29how these considerations can add to our mechanistic evaluations has been included. Informationis provided that illustrates the guidelines discussed in this introductory section. This approach willprovide tools for the reader to critically evaluate mechanistic information.A. Mitotic InterferenceNormal fetal development is characterized by rapid and coordinated cell replication. Thus, it followsthat mitotic interference, defined as a change in the rate of cell proliferation, 5 is a potentialmechanism underlying chemically induced developmental defects. This rapid and specific cellproliferation confers a unique sensitivity of the fetus toward agents affecting cell division processes.Differential rates of cell division within developing tissues or organs may create specific subsetsof cells that are especially sensitive to chemical exposure. 56 Mitotic interference is most commonlyelicited by chemical or physical agents that delay or block cell cycling. However, an increase inthe cell cycle rate due to compensatory repair following an exposure might also contribute to ateratogenic outcome. 57 The nature of the pathological outcome and the ability of the organism tocompensate for the damage will depend upon, among other factors, the nature of the exposure, thedevelopmental stage of exposure, and the affected tissue or cell types. Because normal morphogenesisdepends upon a highly synchronized progression of events, a reduction in the total numberof cells or a delay in the production of cells as a result of cell cycle arrest or inhibition can havelong-term and irreversible consequences. For example, the rate of neuroblast proliferation may bean important determinant of cerebral organization and synaptic network formation, 58 and cell cycledelays may lead to altered synaptogenesis and an altered cortical cytoarchitecture. Additionally,the normal development of certain tissues may depend on attaining a critical cell mass for theprogress of normal cell differentiation and organogenesis; 59 an inhibition or delay in the proliferationof these tissues may result in developmental toxicity. In both models, normal morphogenesis relieson the normal progression of cell division.General mechanisms have been identified that result in an altered cell cycle rate. These include:(1) reduction of DNA synthesis; (2) interference in the formation or separation of chromatids; and(3) failure to form or maintain the mitotic spindle. 5 Critical cell-signaling pathways for controllingcell cycle pathways (i.e., p21 and p53) have also been identified. The mechanisms of mitoticinhibition for several example developmental toxicants have been described. Agents that affect cellcycling through these pathways are discussed.1. 5-FluorouracilDevelopmental toxicants have been identified that may act through inhibition of DNA synthesis.These include hydroxyurea, cytosine arabinoside, and 5-fluorouracil. 5-Fluorouracil (5-FU) exposureis associated with morphologic abnormalities in several animal species. 60,61 The S-phaseinhibition elicited by 5-FU exposure, 62 via inhibition of thymidylate synthetase, is well documented.However, because 5-FU is also incorporated into RNA, resulting in cell death, it is difficult toattribute 5-FU–induced teratogenicity solely to an inhibition of DNA synthesis. In fact, depletionof cytidine, rather than thymidine, may be the proximate cause of interrupted cell division. 63 It isunclear how the incorporation of 5-FU into DNA affects DNA fidelity or cell viability, and whetherthese effects are associated with teratogenicity (for a review of these mechanisms see Parker andCheng 64 ). The effects of 5-FU on cell cycling and viability are dependent on the cell type, cellcycle phase, and exposure conditions, 62 which implies a differential sensitivity among developingtissues. Coordinated biochemical, cellular, and morphological studies conducted by Shuey et al. 65demonstrate successful mechanistic approaches to understanding 5-FU–induced developmentaltoxicity. These studies revealed that the levels of incorporation of 5-FU derivatives into nucleicacids and direct interactions with DNA were too low to explain the resultant developmental toxicity.Subsequent mechanistic studies of 5-FU have focused on inhibition of thymidylate synthetase and© 2006 by Taylor & Francis Group, LLC

30 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONresultant cellular perturbations. (See discussion below on enzyme inhibition and how these mechanisticstudies were incorporated into a biologically based dose-response model.) These mechanisticstudies thus reveal the need for researchers to shift their investigations from one biological level— DNA and molecular level changes — to assessments at the cellular level. They might thusidentify critical rate-limiting processes that could more effectively explain the overall complexprocesses of nucleotide pool alterations and subsequent DNA synthesis perturbations.2. RadiationThe effects of radiation on fetal development, including the developmental effects of radiationinducedcell loss, have been well documented. 58,66–71 The effect of radiation on chromatid formationand cell cycling likely underlies radiation-induced cell cycle perturbations. 5 However, the variedeffects of radiation on DNA fidelity make it difficult to attribute radiation-induced teratogenicitysolely to effects on cycling cells. Among the DNA effects showing dose-response relationshipsfollowing radiation exposure are chromosome instability, single strand breaks, double strand breaks,DNA-protein cross-links, apoptosis, p53 activity, and mitotic inhibition. 72–75 The most sensitiveperiod for radiation-induced malformations is following day eight of organogenesis in the rat(corresponding to weeks 8 to 25 in the human fetus), during the period of maximal proliferationof neuronal precursors. Exposure (e.g., 100 Rd in the rat) prior to organogenesis can cause eitherfetal death or have no effect. Exposure during organogenesis results in decreased weight andthickness of cortical layers, formation of ectopic structures, and microcephaly. CNS effects arenoted if exposure takes place late in gestation, reflecting the extended period of sensitivity ofnervous system development (see reviews by Brent, 76 Beckman and Brent, 77 and Kimler 67 ).3. Antitubulin AgentsPerturbations in the mitotic spindle may result in cytoskeletal disruption, aneuploidy, micronuclei,alterations in cell division rate, cell cycle arrest, and/or cell death. The coincidence of these typesof toxicological manifestations is indicative of antimitotic agents that affect tubulin. 78 Classicantitubulin agents, such as benzimidazoles, carbamates, and colchicine, have been demonstrated toelicit aneuploidy 79–86 and developmental toxicity both in vivo (reviewed in Delatour and Parish, 87Ellis et al., 88 and Van Dyke 89 ) and in vitro. 78,90 These studies used dose-response relationships andstructure-activity information to strengthen the support for mitotic perturbations as the mechanismof action by which these agents cause developmental toxicity. Other methodological approachesused in these studies were comparisons across species that revealed significant cross-speciesdifferences in tubulin binding affinities that were related to their differing potency as toxicants indifferent species.4. MethylmercuryNumerous mechanisms have been proposed by which methylmercury (MeHg) may disrupt normalcell function, and these mechanisms are postulated to result in neurodevelopmental toxicity. Most,if not all, are associated with the exceptional affinity of MeHg for the thiol group, with theassociation constant of the Hg-SH pair being orders of magnitude greater than MeHg’s interactionwith any other ligand. 91 MeHg may interfere with the proper functioning of cells by disrupting thethiol bond–mediated structure and the function of key proteins or other molecules. While this ideais mechanistically a very simple concept, the numerous affected pathways may include (1) disruptionof protein synthesis, 92–94 (2) disruption of cellular energy production, 95–97 (3) disruption ofintracellular calcium levels, 98,99 particularly in the mitochondria, (4) disruption of microtubuleassembly and cellular division and transport, 100–104 and (5) induction of oxidative stress, either© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 31KineticsDynamicsBrainLiverKidneyFetal brainOther tissueCell deathExposureBloodµ 1 µν2XYCommitmentλ 1λ 2X X Y YDivision DivisionHealth riskFigure 2.4Example of a combined toxicokinetic and dynamic model for methyl mercury exposure duringpregnancy. Exposure may occur through ingestion, inhalation, or dermal absorption, through whichthe toxicant rapidly enters the bloodstream. Distribution throughout the body (toxicokinetics) determinesthe dose to the conceptal brain, affecting cell proliferation, differentiation, and death (toxicodynamics)and the risk of developmental neurotoxicity. The structure of the kinetic model isshown in more detail in Faustman, E., et al. Environ Toxicol and Pharmacol: 2005. (Adapted fromFaustman, E., et al. Inhalation Tox, 11, 101, 1999. With permission).through the depletion of the intracellular redox agent, glutathione, 88,105–107 or through the generationof reactive species. 108–110 Each of these mechanisms has been studied in considerable detail, and itis unclear whether any one mechanism is predominant. Many of these mechanisms, however, willaffect the cell cycle and appear as an antimitotic effect. To illustrate approaches for evaluating thistype of common “synthesized” endpoint, we have chosen in this discussion to focus on an evaluationof these mechanisms at a higher level of biological organizations, namely, how does MeHg causeneuronal cell loss during brain organogenesis?Studies in both humans and experimental animals have shown that MeHg causes developmentalCNS abnormalities, notably decreases in brain cell number and improper neuronal alignment. AsBurbacher et al. 111 noted, this phenomenon is observed consistently across a considerable doserange and across a variety of species, including humans, 112,113 rodents, 114–117 and monkeys. 111Although MeHg is also known to produce necrosis and apoptosis in neuronal cells, 118–120 alterationsin proliferative activity represent a more sensitive effect, and that is associated with low-dose humanexposures. Both in vivo and in vitro studies reveal that MeHg exposure can affect the dynamics ofcell cycling in the CNS. 101,114,121–123 In these assessments, DNA in actively proliferating cells isidentified in two ways: (1) mitotic figures are determined morphometrically or (2) actively proliferatingDNA is labeled with the thymidine analog, 5′ bromodeoxyuridine (BrdU), and BrdUincorporation and cell cycle progression is determined using bivariate flow cytometry. The in vitrostudies revealed that a G2/M cell cycle arrest occurred in the absence of direct cytolethality,supporting the hypothesis that cell cycle effects may be a more sensitive endpoint than cell death. 122In vivo studies have revealed that although both rats and mice are sensitive to cell cycle effects,mouse cells appear to be more sensitive. 123To put our in vivo and in vitro observations into context, we adapted our previous biologicallybased dose-response (BBDR) model to evaluate the effects of MeHg on CNS cell dynamics. 124,125Figure 2.4 shows a toxicokinetic and dynamic model framework for MeHg developmental toxicitythat builds from the general framework presented in Figure 2.1. Within the dynamic portion of thisframework is a dynamic model that evaluates the impacts of MeHg on the normal developmentalprocesses of proliferation, differentiation, and cell loss (apoptosis and necrosis). As shown in Figure2.4, our dynamic model was linked with a toxicokinetic model for MeHg exposures during pregnancyto assess neuronal cell exposures at realistic environmental exposures. 126 To evaluate brainconcentrations during development, we linked our kinetic model outputs in a stepwise manner toour dynamic model of midbrain development. 125,127© 2006 by Taylor & Francis Group, LLC

32 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe dynamic model framework was also used to evaluate the role that specific cell-signalingpathways, which control cell cycle checkpoints, might play in mediating MeHg cell cycle effectsand hence developmental toxicity. In these studies, the response of wild-type and p21 cell cyclegene knockout mouse cells to MeHg was evaluated. Whereas the G2/M accumulation induced byMeHg was independent of p21 status (as was cytotoxicity), a greater proportion of p21(-/-) cellswere able to complete one round of cell division in the presence of MeHg as compared with p21(+/-) or p21(+/+) cells. These data suggest an important role for p21 cell checkpoint pathways inmediating MeHg’s effects on the cell cycle. The importance of MeHg’s effects on cell cycle in ouranalysis was significantly enriched by making comparisons across species and across both in vivoand in vitro assessments. The modeling framework was critical for placing these mechanistic cluesinto a larger, more environmentally relevant context.In summary, agents that cause mitotic interference can be potent developmental toxicants. Doseresponseand structure-activity relationship have strengthened our understanding of common mechanismsof these agents causing delays and complete blockage of rapidly proliferating cells withinthe conceptus. Transgenic models for key cell cycle checkpoint pathways are also proving usefulin evaluating key signaling pathways.B. Altered Energy SourcesThe high replicative activity of cells during “biosynthesis and proliferation requires an uninterruptedsource of intrinsic energy generated in the developing tissues.” 5 Oxidative metabolism is essentialfor fetal development, and oxidative phosphorylation increases as gestation proceeds. 128,129 However,in only a few situations has altered mitochondrial function been associated with an adverse developmentaloutcome. Among these are achondroplasia and riboflavin deficiency. 130–134 In these studies,skeletal system malformations were associated with deficient mitochondrial activity. Chondrogenesismay be especially sensitive to agents that interfere with energy production because the growthplates of the long bones have “the lowest oxygen tension of any bodily organ undergoing activeproliferation.” 133 Although studies demonstrate that reduced mitochondrial function is associatedwith skeletal malformations, the relationship between reduced energy status and the appearance ofteratogenesis remains unclear. For example, other studies carried out by Mackler and Sheparddemonstrated that iron deficiency inhibited mitochondrial function (as measured by a 60% reductionin mitochondrial NADH oxidase activity) and produced a marked decreased fetal viability and size(but no congenital malformations). 135–137Few studies link chemically induced mitochondrial dysfunction with developmental toxicity.For example, classic inhibitors of mitochondrial respiration, such as rotenone or cyanide, have notbeen associated or are only weakly associated with teratogenic outcomes. While the high toxicityof these chemicals may preclude the observation of altered morphology because of fetal or embryonicdeath, there may be a narrow range of exposures where malformations are observed. 138Investigations into the effects of chemically induced inhibition of mitochondrial respiration andthe appearance of dysmorphogenesis have included studies of rhodamine dyes (rhodamine 6G andrhodamine 123), 139 diphenylhydantoin, chloramphenicol and sodium phenobarbital, 138 andcocaine. 140–142 More recent studies have investigated the role that mitochondria play in mediatingapoptotic signals (see Section II.G for more details on this mechanism).1. Inhibitors of Mitochondrial RespirationMackler et al. 138 investigated agents such chloramphenicol, phenobarbital, and malonate becauseof their known inhibitory effects on mitochondrial respiration. All of these agents were found toinhibit fetal growth, with the exception of malonate. Phenobarbital produced profound skeletalalterations, including cleft palate, edema, spinal retroflexion, and delayed ossification of the occiput© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 33and sternum. Diphenylhydantoin produced syndactyly and oligodactyly, and delayed ossificationof the occiput and sternum. Chloramphenicol produced edema, wavy ribs, and fused ribs. The lowpercentage of effects following exposure to malonate and chloramphenicol may be attributable tothe steep dose-response relationship between the production of dysmorphogenesis and lethality.While these chemicals were shown to decrease the specific activity of various enzymes involvedin electron transport in vitro, the authors found little inhibition of these enzymes when studyinghomogenates prepared from the exposed fetuses, with the exception of those treated with phenobarbital.Therefore, the relationship between inhibition of mitochondrial function and the productionof dysmorphogenesis was not determined.2. CocaineThe effects of cocaine on fetal dysmorphogenesis, while producing mitochondrial inhibition, have notbeen ascribed to mitochondrial dysfunction. Rather, cocaine-induced ischemia reperfusion has beenhypothesized to result in the production of reactive oxygen species, leading to focal tissue damage. 1433. Rhodamine DyesThe cationic rhodamine dyes, which include rhodamine 123 and rhodamine 6G, have been usedas mitochondrial-specific markers. The strongly negative charge potential across the mitochondrialmembrane causes accumulation of the positively charged dyes, while the neutral rhodamine dyesshow no specific localization to mitochondria. 139 The cationic rhodamine dyes also interfere withmitochondrial respiration, which has been observed to result in low ATP production followingeither in vivo or in vitro exposure. 139 Hood et al. 144 investigated the teratogenic effects of rhodamine123 in mice during gestation (7 to 10 days). Administration of rhodamine 123, in combination with2-deoxyglucose, an inhibitor of glycolytic ATP generation, led to elevated levels of both gross andskeletal malformations, as well as an increased incidence of early fetal death. 144The few studies presented here demonstrate that a compromised energy production capacityhas the potential to lead to adverse developmental outcomes that can range from relatively minorabnormalities to fetal death. However, at present, there is no clear understanding of the fundamentalcontribution of this pathway to teratogenesis. The investigations presented here support a modelwhere skeletal development is at highest risk from exposures to chemical agents that inhibitoxidative respiration.C. Enzyme InhibitionThe teratogenic effects of some compounds may be attributed to inhibition of specific enzymes.Enzymes critical for cell growth and proliferation, such as those involved in synthesis of DNA andRNA, are ones whose inhibition might have the greatest effect on developmental processes. Fourmodel developmental toxicants, methotrexate, chlorpyrifos, 5-fluorouracil, and mevinolin, are highlightedas examples of agents that cause enzyme inhibition.1. MethotrexateMethotrexate (MTX), a cancer chemotherapeutic agent, is a competitive inhibitor of dihydrofolatereductase (DHFR), the enzyme that converts folate to tetrahydrofolate. Tetrahydrofolate is subsequentlymetabolized to various coenzymes that participate in one-carbon metabolism (OCM), whichis critical for the synthesis of purines and amino acids and the conversion of deoxyuridylate tothymidylate. In utero exposure to MTX causes craniofacial defects, limb deformities involvingreduction in size, and decreased fetal weights. 145© 2006 by Taylor & Francis Group, LLC

34 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTo determine whether a chemical’s teratogenicity reflects inhibition of a specific enzyme, theeffects of administration of the inhibited enzyme’s product may be determined. By administeringleucovorin, a metabolic derivative of tetrahydrofolate, to animals treated with MTX DeSesso andGoeringer 146 demonstrated that MTX’s teratogenic effects are specifically due to inhibition ofDHFR: Treatment with leucovorin protected animals from the teratogenic effects of MTX.While it seems plausible that the teratogenic effects of MTX are due to its interference withOCM, owing to the importance of this pathway, it is still conceivable that MTX’s teratogenicity isa result of depletion of tetrahydrofolate, which may participate in an unidentified, yet crucial,metabolic pathway. This possibility was addressed by DeSesso and Goeringer 145 using 1-(p-tosyl)-3,4,4-trimethylimidazolidine (TTI), a functional analog of tetrahydrofolate, which participates inOCM. Since TTI is structurally dissimilar from tetrahydrofolate, yet still enables OCM to proceedas usual, the ability of TTI to prevent MTX-induced teratogenic effects can be attributed to itsrestoration of OCM. Results from this study provide strong evidence that the teratogenic effectsof MTX are related to its interference with OCM, in that TTI dramatically reduced both theincidence and severity of MTX-induced malformations. Although TTI did not completely alleviateMTX-associated teratogenicity, this may be partly due to the dosing regimen used by DeSesso andGoeringer.2. ChlorpyrifosChlorpyrifos (CP) and its active metabolite chlorpyrifos-oxon (CPO) act as nervous system toxicantsthrough their ability to inhibit acetylcholinesterase (AchE) activity, which can explain pesticideaction of CP and its ability to act as a poison at high doses in adults. Thus, a key question fordevelopmental toxicologists is whether similar mechanisms underlie the neurodevelopmental toxicityof chlorpyrifos in children.CP/CPO is capable of potent and irreversible inhibition of cholinesterase, and this inhibitionresults in subsequent accumulation of the neurotransmitter acetylcholine in the synaptic cleft. 147This may lead to multiple toxic effects, including cholinergic crisis and disruption of neurodevelopment.Acetylcholine is responsible for cholinergic neurotransmission. If this ligand is not removedfrom the synaptic cleft, overstimulation may occur at high doses, leading to cholinergic crisis. Suchpoisoning effects have been well documented in agricultural workers, 148 and monitoring of cholinesteraselevels is used as a method of exposure surveillance in California agricultural workers.Researchers are particularly interested in fetal and neonatal exposure because recent studieshave focused on adverse developmental effects associated with chlorpyrifos intake. Multiple mechanismsof chlorpyrifos toxicity have been proposed, and their relative importance appears to dependon developmental lifestage and dose. To identify critical modes of action that can explain specificadverse outcomes that can arise following CP exposures during development, consideration of dose,time of exposure, and target tissue is essential. Cholinesterases have been proposed to have distinctroles in multiple phases of neurogenesis, including neuronal differentiation, cell migration, neuriteoutgrowth, and synaptogenesis (reviewed by Small et al. 149 ). In vitro midbrain micromass studiesof differentiation suggest that many neuronal cells are cholinergic in nature. 150 There is also evidencethat neurodevelopment is influenced by AchE expression. In neuroblastoma cells, the ability ofneurites to extend and the appearance of other differentiation markers varied with AchE expression.151 In rabbits, AchE transcripts have been identified on embryonic day 12, 152 indicating anearly role for this enzyme. Thus, agents that disrupt these pathways would be hypothesized to havedevelopmental impacts. In vitro studies with primary cultured chick neurons indicate that cholinesteraseinhibitors induce growth cone collapse and inhibit neurite extension. 153 Thus, there is thepotential for CP to affect a variety of neurodevelopmental processes via AchE inhibition and manyin vitro and in vivo studies have been conducted that investigate these effects. However, it mustalso be noted that in vivo neurobehavioral effects by CP have been observed at levels below itseffects on AchE activity 154 and at time points prior to AchE dependent neurodevelopmental processes.In vitro studies 155 suggest other mechanisms such as the production of reactive oxidative© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 35stress may be significant. 155–161 Thus, AchE inhibition is thought by many to be one of severalmodes of CP neurotoxicity. 1623. 5-FluorouracilTeratogenic effects of 5-FU include cleft palate and limb and tail defects. 164,165 As mentioned earlier,5-FU can affect a variety of processes, including the direct incorporation of 5-FU into nuclearRNA, resulting in processing errors in forming mRNA and rRNA, 163 and incorporation of nucleotidebase residues into DNA. However, the resulting mitotic inhibition is insufficient to explain 5-FU’sdevelopmental toxicity. 63 Hence, other effects of 5-FU have been examined.Of particular interest for developmental effects is 5-FU’s ability to inhibit thymidylate synthetase(TS), which methylates deoxyuridylic acid to form thymidylic acid. This inhibition can lead to animbalance of nucleotide pools and alterations in cell proliferation and/or cell death.Evidence that the toxicity of 5-FU may be due to inhibition of TS was provided in a study byElstein et al. 62 in which 5-FU induced an accumulation of murine erythroleukemic cells (MELC)in early S-phase. Specifically, effects of 5-FU were greatest when cells are exposed during the S-phase of the cell cycle, when DNA synthesis occurs and when TS activity is highest. If 5-FU actedby misincorporation into RNA or DNA, effects would not necessarily be specific to the S-phaseof the cell cycle. A study by Abbott et al. 60 suggests that the teratogenicity of 5-FU was specificallydue to TS inhibition. In this study, inhibition of TS activity in palatal shelves of rat embryos wascorrelated with effects on growth and fusion of the palatal shelves. 604. MevinolinMevinolin is a competitive inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA).HMG-CoA participates in the mevalonate pathway, a precursor for the synthesis of isoprenoidsand cholesterol. Teratogenic effects of mevinolin include neural tube defects and rib and vertebralmalformations. 170 These effects may be due to depletion of isoprenoids, which, among other things,are required for the posttranslational farnesylation of the p21 ras protein.One indication that the teratogenicity of mevinolin results from inhibition of HMG-CoA reductaseis that the teratogenic effects of mevinolin can be diminished by mevalonate, the product ofHMG-CoA reductase. Additionally, the teratogenicity of mevinolin analogs correlates with theirability to inhibit HMG-CoA reductase. 170The approach used by Brewer et al. 171 to demonstrate that mevinolin’s teratogenicity may bedue to inhibition of HMG-CoA reductase is similar to that used for 5-FU by Abbott et al. 60 Usingin situ hybridization with a cRNA probe, Brewer et al. demonstrated that expression was high inthe neural tube, where mevinolin is known to produce developmental abnormalities. High expressionlevels of HMG-CoA reductase could indicate a requirement for high levels of mevalonate productsand likewise a high degree of sensitivity to depletion of mevalonate by HMG-CoA inhibitors.In summary, this section has highlighted three developmental toxicants that inhibit key enzymescritical to proper cellular function. As noted, inhibition of these enzymes critical to DNA and RNAsynthesis has dramatic effect on normal development.D. Nucleic AcidsAgents that interfere with the normal synthesis and functioning of DNA and RNA can be teratogenicbecause these processes are so vital to the rapidly proliferating cells of a developing embryo.1. HydroxyureaHydroxyurea (HU) blocks DNA synthesis by inhibiting ribonucleotide reductase, the enzymeresponsible for reducing uridine, cytidine, adenosine, and guanosine diphosphate to their correspondingdeoxyribonucleotides. 172 Teratogenic effects of HU in rats include malformations of limbs,© 2006 by Taylor & Francis Group, LLC

36 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONpalate, jaw, and tail, as well as mortality. 173 The morphological changes in embryonic murine cellstransplacentally exposed to HU are similar to apoptotic cell death because the observed effectsinvolve condensation of chromatin and shrinkage of cytoplasm. 174Evidence that HU depletes deoxyribonucleotides and interferes with DNA synthesis to elicitteratogenicity is provided by a study in which the teratogenic effects of HU in rats were eliminatedby coadministration of deoxycytidine monophosphate (dCMP) and HU. 173 Similar results wereobtained by Herken, 175 who found that coadministration of dCMP partially prevented cytotoxiceffects of HU on murine neuroepithelial cells isolated from transplacentally exposed embryos. Inaddition to abrogating some of HU’s cytotoxic effects, dCMP also reduced inhibition of DNAsynthesis. Since dCMP can be converted to dTTP as well as dCTP, the ability of dCMP to provideprotection from the effects of HU may indicate that the availability of pyrimidine deoxynucleotidesis more of a limiting factor on DNA synthesis then the availability of the purine deoxynucleotides.The inability of dCMP to provide complete protection from HU-induced cytotoxicity in theHerken 175 study may be due to a lack of the purine deoxynucleotides. However, simultaneousinjection of all four deoxynucleotides into mouse fibroblasts failed to completely restore DNAsynthesis. 176 This suggests the existence of an additional mechanism by which HU exerts effectson DNA synthesis and developmental toxicity.The dual nature of HU’s teratogenicity may be related to the presence of a hydroxylaminefunctional group on the molecule. This hydroxylamine group is capable of reacting with oxygento form hydrogen peroxide, which can ultimately generate highly reactive hydroxyl free radicals(discussed in DeSesso and Goeringer 177 ). To determine if any of the teratogenic effects of hydroxyureacan be attributed to generation of hydroxyl free radicals, DeSesso and Goeringer 177 pretreatedrabbits with either ethoxyquin or nordihydroguaiaretic acid, both of which are antioxidants thatcan terminate free radical reactions. Since pretreatment with either of these compounds delayedthe onset of embryonic cell death and lowered both the number of malformed fetuses and theincidence of specific malformations, while increasing body weight, DeSesso and Goeringer 177suggested that the developmental toxicity of hydroxyurea can be at least partly attributed to thegeneration of reactive oxygen species. Similar results were obtained with propyl gallate, whichdelayed onset of embryonic cell death without disrupting hydroxyurea’s inhibition of DNA synthesis.177,178 Taken together, these data are significant in that all three of these antioxidants arestructurally dissimilar. Thus, their ability to protect embryos from the teratogenic effects of hydroxyureais probably due to their antioxidant properties, rather than an unknown mechanism relatedto their structure. Since these three antioxidants didn’t completely prevent cell death completelyor hydroxyurea-induced developmental toxicity, it is reasonable to postulate that other propertiesof hydroxyurea, such as inhibition of DNA synthesis, also contribute to its teratogenicity.2. Cytosine ArabinosideCytosine arabinoside (Ara-C) inhibits DNA synthesis by functioning as a pyrimidine analog followingits intracellular phosphorylation to Ara-CTP. 179 Teratogenic effects of Ara-C in mice includeincreases in resorptions, decreased fetal weight, cleft palate, defects of the long bones, and oligodactyly.Higher doses can also cause fusion of vertebrate bodies and ribs. These effects occurfollowing treatment between gestation days (GD) 10.5 and 12.5. However, exposure to Ara-C afterGD 13 produced no discernable malformations, 180 possibly because Ara-C is particularly toxic torapidly proliferating cells in the S-phase of the cell cycle 181 while cells that have undergone somedifferentiation appear to be relatively insensitive to Ara-C. 182 One of the direct effects of Ara-C isits incorporation into DNA. DNA polymerase alpha can be moderately inhibited by Ara-C, but thismechanism probably does not account for all the observed inhibition of DNA synthesis. Ara-C alsoinhibits DNA ligase, the enzyme responsible for joining Okazaki fragments. Inhibition of DNAligase could contribute to the inhibition of DNA synthesis and consequently the cytotoxicity of© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 37Ara-C, and may account for the small size of DNA fragments seen with Ara-C treatment. However,cytotoxicity may be more related to formation of Ara-CTP. 183Ara-CTP may cause a deficiency in deoxycytidylic acid triphosphate (dCTP) by inhibiting thereduction of cytidylic acid diphosphate (CDP) to deoxycytidylic acid diphosphate (dCDP). 184 Ifthis mechanism occurs, then a restoration of normal levels of dCTP should protect embryos fromthe effects of Ara-C. This protection was achieved in a study in which deoxycytidine was administeredat a dose four times higher than that of Ara-C. 184 Deoxycytidine similarly protected mouseembryos from the cytotoxic and teratogenic effects of Ara-C when its dose was eight times greaterthan that of Ara-C. 180 However, these studies don’t rule out the possibility that DNA synthesis isinhibited by incorporation of Ara-CTP into DNA, because the high doses of deoxycytidine usedcould outcompete Ara-C for incorporation.The effect of HU and Ara-C on DNA synthesis have been highlighted in this section. Researchon these two agents provides excellent examples of mechanistic studies using strength and consistencyguidelines. (See also earlier comments on 5-FU for related discussions.)E. MutationsMutations are alterations of DNA nucleotide sequence. Such changes in the DNA sequence canresult from exchange of one base pair for another (transitions or transversions) or deletions orinsertions of a few bases, as well as from inversions, deletions, and translocations involving changesin long segments of DNA following strand breaks and errors in repair. Mutations generally arisefrom agents that damage DNA, including ionizing radiation and highly electrophilic substances.Because DNA replication is not 100% accurate, a low rate of spontaneously occurring DNA damagealso occurs. In addition, mutations can arise from inhibition or altered function of either DNArepair enzymes or the DNA polymerases involved in proofreading. Twenty percent of malformationsin humans are attributable to known genetic transmissions, and up to five percent are due tochromosomal aberrations. This section will not focus on these known genetic birth defects butrather will focus on malformations that are chemically or physically induced.There are a number of factors to consider when evaluating mutation as a mechanism ofchemically induced teratogenesis. One factor is the location of the mutation within the genome.For example, a loss-of-function mutation in a housekeeping gene that is required for cell survivalwould probably be cytotoxic, and therefore would not be passed on to future generations of cells.For a mutation to persist in the developing embryo, it would have to occur in a gene that is notrequired for cell survival; otherwise, the mutation would result in cell death. Examples of genesthat can be mutated and yet allow for cell survival are proto-oncogenes and tumor suppressor genes,many of which are involved with regulating cell growth and proliferation. Both proto-oncogenesand tumor suppressor genes are expressed at very specific times during the course of development.Additionally, mutated forms of these genes are found in a wide variety of tumors. Because protooncogenesand tumor suppressor genes have developmentally specific expression patterns, anymutation that alters either the timing or level of expression of one of these genes might be expectedto alter normal developmental processes.Another factor to consider is that the occurrence of a mutation in a gene that isn’t required forcell survival is probably a very rare event. For such a rare event to have a significant effect onembryogenesis, it would probably have to occur early, for instance, at or before the 22 blastocyststage, or would have to occur nonrandomly in the genome. Evidence of nonrandom distribution ofchemically induced genetic alterations has been reported. 185,186In this section, alkylating agents, aromatic amines, and ionizing radiation will be discussed asexamples of developmentally toxic agents that have both mutagenic and teratogenic properties.With all of these agents, it is not clear whether the teratogenic effects arise because of mutationsor whether the cytotoxicity of the agents determines the teratogenic outcome.© 2006 by Taylor & Francis Group, LLC

38 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1. Alkylating AgentsMonofunctional alkylating agents are highly electrophilic substances that form covalent alkyl bondswith nucleophilic sites on DNA and proteins. Commonly alkylated sites include the N-3 and N-7positions on purines, the O-6 positions on guanine, and the O-2 and O-4 positions on thymine. Themutagenic potential of alkylating agents is primarily due to alkylation at the exocyclic oxygens ofguanine and thymine.Teratogenic effects in rodents following in vivo gestational exposure to alkylating agents includemortality, growth retardation, and cephalic, CNS, palatal, and limb malformations, as well asanophthalmia. 12 Effects following in vitro exposure are similar to those seen following in vivoexposure and include mortality, growth retardation, abnormal neurulation, abnormal flexure, andoptic malformations. 11Although alkylating agents form DNA lesions that are both cytotoxic and mutagenic, studies byBochert et al. 187 strongly suggest that the teratogenic potential of alkylating agents may be relatedto their mutagenic potential. In this study, the teratogenic potencies of three different alkylatingagents (ethylmethane sulfonate, methylnitrosourea, and dimethylnitrosamine) in mouse embryoswere related to adduct levels at the O-6 position of guanine, a promutagenic lesion. In contrast,there was no correlation between teratogenic potency and adduct levels at the N-7 position of guanine,which is considered to be primarily a cytotoxic lesion. The importance of promutagenic lesions indevelopmental toxicity is further supported by the observation that similar adduct levels at the O-6position of guamine were observed at equally developmentally toxic levels of exposure of thesethree chemicals, despite significant physical and structural differences among these chemicals.The relative importance of the O 6 -alkylguanine adduct on cytotoxicity and inhibition of differentiationof primary embryonic rat midbrain (CNS) and limb bud (LB) cells was studied using O 6 -benzylguanine (O 6 -Bg), which is a potent and specific inhibitor of the protein that repairs O 6 -alkylguanine DNA adducts. 188 In these modulation studies, O 6 -Bg potentiated the effects of methylnitrosourea(MNU) to a greater extent than those of ethylnitrosourea (ENU), and for both compoundsinhibition of differentiation was potentiated to a greater extent than cytotoxicity. Theseresults provide further evidence that the promutagenic O 6 -alkylguanine adduct may be of particularimportance for developmental toxicity.Bifunctional alkylating agents, such as the cancer chemotherapeutic agent cyclophosphamide(CP), can also have both teratogenic and mutagenic properties. CP is bioactivated to a teratogenicmetabolite, 4-hydroperoxycyclophosphamide (4-OOH-CP). 4-OOH-CP spontaneously decomposesto phosphoramide mustard and acrolein, which is considered to be mutagenic. Exencephaly, cleftpalate, abnormal prosencephalon, as well as limb malformations have been observed in embryosfollowing in utero CP exposure. 189–192 CP has also been shown to inhibit DNA synthesis. 193Some investigators have hypothesized that CP-induced teratogenicity may be related to itsmutagenic potential. For example, treatment of male rats with CP altered growth and developmentof both second and third generation offspring. Effects included increases in mortality and malformationsand reductions in body weight, as well as learning disabilities. 194–196 CP-induced mutationswere also detected in transgenic mice containing shuttle vectors for detection of mutagenicity. 197–199However, the results of these studies do not exclude the possibility that CP could cause mutationsindirectly through epigenetic mechanisms, such as alteration of cellular redox status.To determine whether CP’s teratogenicity reflects the mutagenicity of the metabolite acrolein,investigators exposed D10 rat embryos in vitro to an analog of CP that breaks down into acroleinand dechlorophosphoramide. Dechlorophosphoramide does not have the DNA alkylating propertiesof phosphoramide. Consequently, any DNA damage may be attributable to binding of acrolein toDNA. This study revealed that DNA damage occurred only if embryos were cultured in serumfreemedia with buthionine sulphoximine (BSO) depletion of glutathione, and only at concentrationslethal to the embryo. DNA damage was not detected at concentrations that produced malformations© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 39but minimal lethality. 200 Thus, mutagenicity did not appear to be the mechanism by which CP exertsdevelopmental toxicity in vitro.2. Aromatic AminesAromatic amines are a class of industrially important chemicals that include potent mutagenic andcarcinogenic compounds. These agents have also been of interest as transplacental carcinogens anddevelopmental toxicants. 201–205 In particular, this class of arylating agents was investigated todetermine if mutagenic metabolites important in defining the carcinogenic potency of the arylatingagents were also important in the etiology of their developmental toxicity. It was determined thatmetabolites of 2-acetylaminofluorene (AAF), such as 7-hydroxy-acetylaminofluorene, that werenot important for carcinogenesis were important contributors to the developmental effects seen invitro. 204 In addition, Faustman-Watts et al. 203 were able to separate the mutagenic, teratogenic,cytolethal, and embryolethal effects of AAF metabolites, thus minimizing the strength and consistencyof any association between mechanisms of carcinogenesis and teratogenesis.3. RadiationIonizing radiation is another teratogen that can cause mutations. Ionizing radiation’s ability to causemutagenicity has been studied in many mammalian cell types, including those employed in specificlocus (such as the hypoxanthine guanine phosphoribosyltransferase locus) and shuttle vector systems.206,207 These studies reveal that ionizing radiation causes both point mutations (base changes,frameshifts, and small deletions) and large deletions, with the relative proportion of each typedependent on the type of radiation (either high or low linear energy transfer focus) dose and boththe cell type and the locus or vector used. 208–212 Detection of mutations in oncogenes and tumorsuppressor genes in radiation-induced tumors suggests that ionizing radiation could cause mutationsin specific genes. 213–215 However, ionizing radiation is also cytotoxic, and significant increases inmutations at specific genes are usually observed only at doses where there is also substantial cellkilling. 215 Prenatal exposure to ionizing radiation can impair fetal growth and cause structural,physiological, and behavioral abnormalites. 213,214,216–219 However, it is not clear whether the effectsseen following exposure to ionizing radiation are due to its cytotoxic potential, its mutagenicpotential, or a combination of both.Recent work on low dose effects of radiation has used microbeams, a unique type of tool thatcan irradiate either the nucleus or cytoplasm of selected cells in culture. These studies have shownthat mutational effects can occur when only the cytoplasm is irradiated or in adjacent cells thatwere not irradiated, which has led to the term “bystander effects” to describe these phenomena. 206Other types of special tools for radiation dose delivery have also demonstrated elevated mutationrates in vivo in nonirradiated tissues. 206 This suggests that such bystander effects may also providean important mechanism by which development could be disrupted.Techniques are now available to determine whether a mutation in a specific gene can alternormal developmental processes. One method that has been used to study effects of specificmutations on developmental processes involves transfection of cells with activated proto-oncogenes.Embryonal carcinoma cells, which are undifferentiated stem cells isolated from teratocarinomasand which can be induced to differentiate into a variety of cell types by manipulating cultureconditions, can be used for this purpose. 220 Aberrant expression of proto-oncogenes alters normaldifferentiation of these cells. For example, transfection of v-src into P19 embryonal carcinomacells alters morphology and causes loss of expression in stem cell markers, in addition to preventingnormal induction of differentiation along neuronal and mesodermal pathways. 221 Ectopic expressionof c-jun into P19 cells resulted in a cell population containing endodermal and mesodermal cells,© 2006 by Taylor & Francis Group, LLC

40 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONwhereas control cells remained as relatively undifferentiated, pluripotent cells. 222 The phenotypeof differentiated cells can also be transformed by transfection with proto-oncogenes. For example,adult human pigment epithelial cells acquire characteristics of neuronal cells when transfected withthe H-ras proto-oncogene. 223 Results from studies such as these suggest that mutations in developmentallyimportant genes could result in developmental abnormalities.Another useful method for studying effects of specific mutations is generation of transgenicmice carrying activated proto-oncogenes. Transgenic mice expressing high levels of the mos protooncogenetransgene in brain tissue exhibit degeneration of neurons, axons, and spiral ganglia, aswell as gliosis. These physiological abnormalities are accompanied by neurobehavioral anomalies,such as circling, head tilting, and head bobbing. 224 Expression of a mutated WT-1 tumor suppressorgene in transgenic mice caused abnormal development of kidney, mesothelium, heart, and lungs. 225Transgenic mice have been developed that lack functional HRas, KRas2, and N-Ras genes, resultingin tumors in all three mutants, with variable to midgestational deaths reported. In particular, KRastransgenic animals display CNS tissue effects. 1Although studies using transfected cells and transgenic mice are useful for demonstrating thatmutations in specific genes can cause developmental abnormalities, it would also be interesting todetermine if mutated proto-oncogenes are observed in animals or cells exposed to mutagens. Insummary, the role of mutation in the mechanistic process of developmental toxicants still remainsto be clarified. Questions regarding whether a significant level of mutation would occur in viableembryos and lead to teratogenic events are still at issue. Obviously, additional studies will berequired to separate correlation and phenomenology versus true mechanistic paths.F. Alterations in Gene Expression1. Retinoic AcidDevelopment occurs according to very specific and well-orchestrated patterns of gene expression.Hence, alterations in these expression patterns can result in serious adverse developmental consequences.Retinoic acids (RAs), the biologically active metabolites of Vitamin A, play an importantrole in controlling these synchronized expression patterns. Exogenous retinoids can be equallyeffective in disrupting these processes.Retinoids are well characterized developmental toxicants, with their potency being determinedby both kinetic and dynamic factors, including the timing of exposure, dose, and structural formof the retinoid under evaluation. 1 Structure-activity relationships established for retinoids havedemonstrated that an acidic polar terminus is essential for developmental toxicity 15,226 and thatgreater than a 1000-fold difference in potency between different retinoids can be observed. Manyof these differences appear to be due to kinetic differences resulting from changed elimination ratesand reduced affinity for cellular retinoid acid binding proteins. 227 Kinetic studies have shown thatthe area under the curve correlates better with RA teratogenicity than does the peak high dose. 228RA can produce many dysmorphogenic effects, such as truncation of the forebrain and posteriorizationof the hindbrain. Cross-species evaluations have revealed that these same effects are observedin mammals, birds, amphibia, and fishes.There are minimal strain and species differences when specific metabolites of RA (e.g., alltrans-RAor 13-cis-RA) are evaluated. However, species differences in metabolism of RA can affectthe form and type of retinoid at target sites and underscores the significance of kinetic studies indiscussing mechanistic differences in cross-species RA responses. 229,230Our knowledge of retinoic acid receptors indicates that they belong to the nuclear hormoneligand–dependent transcription-factor superfamily. This information has greatly facilitated ourmechanistic understanding of RA developmental toxicity 1 . Two classes of receptors have beenidentified: RARs and RXRs. 231 Each class of receptors has three receptor types: α, β, and γ; eachis encoded by a separate gene, and multiple isomers can be formed by differential promoter usage© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 41and alternative splicing. 1 The isoforms, which differ in domains, are responsible for conferring celltypeand tissue specificity.Much of our mechanistic knowledge about RA developmental toxicity has resulted from eleganttransgenic animal studies. Knockout mice have been generated for each receptor (RAR and RXR)and for most of the isoforms. These studies have revealed that most of the developmental toxicityof RA is mediated by RAR-RXR heterodimers. 1 This was determined by evaluating similarities inphenotype and from treatment studies in these genetically engineered mice. Null mutant mice havebeen used to correlate specific receptors with specific RA-induced teratogenic effects. For example,RXR appears to be responsible for RA-induced limb defects, 232 whereas RAR appears to beresponsible for RA-induced truncation of the posterior axial skeleton and partially responsible forcranial, facial, and neural tube defects. 1,233,234 Kochar and Kumawat 235 showed that RAR agonistswere potent teratogens and that RXR agonists were relatively inactive. Mixed agonists revealedintermediate developmental effects. In total, these observations from receptor binding studies andfrom transgenic and knockout mouse studies reveal the power of these two approaches for providingmechanistic clues for developmental toxicants.This research has been extended to molecular mechanisms of action by analyzing the DNAtarget sequences for RAR-RXR heterodimers. These RA response elements are found on Hox genesand are shown to be under transcriptional control by ras. 226,231 The Hox genes are downstreamtargets of RA in developmental toxicity. These transcription factors control developmental patterningof the CNS, limbs, skeleton, etc., and their expression encodes postural identity. 1 The Hoxexpression patterns following RA exposure can be expanded, reduced, or miss-expressed, resultingin abnormal cell fate and morphogenesis. 236 Treatment with RA can result in alternations inrhombomere expression in the brain and can result in altered patterning of expression domains. 236Small changes in gene expression patterns in RA treated embryos can result in alterations in cellmigration, differentiation, and proliferation. 2262. DioxinDioxin (2,3,7,8-tetrachloro-dibenzo-p-dioxin, TCDD) also alters gene expression by binding witha nuclear receptor. Dioxin has been hypothesized to cause developmental toxicity by interactingwith an endogenous cytoplasmic receptor, a basic helix-loop-helix DNA-binding receptor, andcausing gene expression changes in a host of genes. These include genes that regulate proliferation,differentiation, and the stress response. 1,5 In utero exposure to TCDD causes a wide range of adversedevelopmental outcomes, including mortality, growth retardation, behavioral abnormalities, andstructural defects such as cleft palate and hydronephrosis. Evidence from rodent studies using arylhydrocarbon (Ah) responsive and nonresponsive strains supports the hypothesis that activation ofthe Ah receptor is a critical process mediating TCDD’s developmental effects. 237–239 Both mRNAand protein expression levels for the Ah receptor correlate with sensitivity to TCDD-related developmentaltoxicity. 240 Abbott et al. 241,242 found that TCDD-induced gene expression changes are seenin levels of epidermal growth factor (EGF), transforming growth factor α (TGFα), EGF receptor,transforming growth factor β1 (TGFβ1), and TGFβ2 and also correlate with TCDD-induced cleftpalate. These investigators have examined expression changes in in vitro models of palatogenesis(mouse and human palate organ cultures), and these studies have informed species differences inthe TCDD response, as well as providing detailed mechanistic assessments. The induction of aTCDD responsive gene, CYP1A1, was used to compare mouse and human responsiveness to TCDD.Tissue-level dose and time-response assessments were made to quantitate species differences inAh receptor and Ah receptor nuclear translocator. By comparing the differences in response patterns,these investigators were able to explain quantitative differences in species response to TCDD byidentifying approximately 350 times fewer receptors in human tissues than in mouse tissues andhave estimated that approximately 200 times higher levels of TCDD are required to produceequivalent responses in human tissues versus mouse tissues.© 2006 by Taylor & Francis Group, LLC

Clearly, more research needs to be done in the area of teratogen-induced changes in gene expression.For instance, for genes whose expression is altered by exposure to teratogens, knowledge of theirprecise function during normal development would enable a prediction of whether alterations intheir expression could result in the observed teratogenic outcome. This chapter has highlightedsome of the recent advances in our understanding of 17 key cell-signaling pathways that can explaina significant portion of normal developmental processes (Table 2.4). A major focus of currentmolecular biology research is defining the downstream gene responses associated with theseprocesses. Increases in our knowledge about these normal expression patterns and function willgreatly aid our investigations of toxicant-altered gene expression patterns. Since this chapter wasfirst written an “omic” revolution has taken place in which our ability to evaluate changes in geneexpression is now at the level of simultaneously evaluating 10,000 genes or more. Such genomicanalysis of gene expression patterns via microarray analysis is just part of these “omic” assessments.In conjunction with proteomics (study of protein expression patterns) and metabolomics (study ofbiochemical functional changes), analyses of the dynamics of expression data can be linked withfunctional assessments. Monitoring normal and toxicant-induced changes in these patterns requiresan understanding of temporal, tissue, and even cell-specific expression changes. This need hasdriven the development of amplification techniques that allow for assessment of gene expressionchanges within a single cell, 243 and of laser capture in situ microdissection techniques that canallow for linkage of gene and protein expression patterns within an anatomical context. Internationaland national efforts to develop developmentally relevant databases of expression data are underway,244–247 as are efforts to develop databases for specific types of toxicant-induced expressionchanges (NIEHS Toxicogenomics Initiative). The generation of bioinformatic tools that will allowdevelopmental toxicologist to assess toxicant-induced changes in genes and proteins with functionalimpacts within the context of developmental life stage, biological level of assessment, and functionalconsequence represents both the promise and challenge for interpretation of “omic” data of ourmechanistic studies. 1 Additional material regarding such studies can be found in Chapter 15.42 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION3. New Directions in “Omic” ResearchG. Programmed Cell DeathProgrammed cell death (PCD), which is sometimes referred to as apoptosis, is a normal physiologicalprocess that occurs during development. PCD serves a number of very important functions,which include providing the embryo with the proper morphology and removing vestigial structures.248,249 PCD is an integral component of the development of the central nervous system (CNS).It is estimated that as much as half of the original cell population in the CNS may be eliminatedas a result of apoptosis. 250,251 Apoptosis is thought to optimize synaptic connections by removingunnecessary neurons. This is accomplished by the direct relationship between the extent of neuralconnections to a postsynaptic target and the survival of the trophic factor–dependent neuron. 252PCD is morphologically and biochemically distinct from necrosis.Although the exact sequence of events occurring during PCD depends on the cell type, acommon feature of many cells undergoing PCD is activation of key intracellular cysteinyl-aspartateproteases know as caspases, which cleave specific target substrates, such as poly (ADP-ribose)polymerase (PARP). Hallmarks of PCD include condensation of chromatin at the periphery of thenucleus, resulting in a pyknotic nucleus and shrinkage of the cytoplasm. Throughout this process,the cell membrane normally remains intact, although blebbing may occur. Thus, the appearance ofa cell undergoing PCD differs from that of a necrotic cell, where cell swelling and breakdown ofcellular membranes is observed. PCD also differs from necrosis in that it is an active process, oftenrequiring ongoing protein synthesis. Because PCD is critical for normal morphogenesis, alterations innormal patterns of PCD are therefore an important mechanism of teratogenicity. Observations that© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 43areas of the body with a high incidence of malformations coincide with areas where PCD 253 occurssupport the idea that disruptions in PCD could be teratogenic.1. Retinoic AcidOne example of a teratogen whose effects may be due to disruption of PCD is retinoic acid (RA).Under normal physiological conditions, RA appears to be involved in expression of homeoboxcontaininggenes that regulate pattern formation and specify positional identity in developingembryos. Some of these genes, such as GHox-8 (expressed in chick limb bud) may be involved inPCD, as this gene is expressed in zones of PCD. 254 A study by Coelho et al. 254 provides evidencethat addition of exogenous RA may alter the normal pattern of PCD. In this study, bead implantscoated with RA diminished cell death and inhibited expression of GHox-7, another chick limb budhomeobox gene that is normally expressed in necrotic zones of chick limb bud mesoderm but notin mutant limb buds, which lack such necrotic zones.PCD was also inhibited in palatal shelves of mice exposed to RA on day 10 of gestation (GD10). In exposed shelves, the medial epithelial cells continued to express EGF receptors and bindEGF at a time when EGFR expression and binding of EGF were decreasing in medial epithelialcells of control shelves. However, the effects of RA on palatal shelf cells may be secondary toother effects. The phenotype of medial epithelial cells depends on interactions with mesenchyme,basal lamina, extracellular matrix, and growth factors, all of which could be affected by RA. Forexample, RA decreases extracellular spacing between mesenchymal cells underlying the medialepithelium. 33Results from several additional studies suggest that under some conditions RA may increasethe extent of PCD. For example, human neuroblastoma cells exposed to RA show growth inhibition,neurite outgrowth, and PCD. 255 Vitamin A, the naturally occurring form of RA, can increase theextent of PCD in the interdigital necrotic zones that appear during limb morphogenesis. 256Sulik et al. 257 suggest that RA may cause both craniofacial and limb malformations by increasingcell death in areas of PCD. These investigators used supravital staining with Nile Blue Sulfate(NBS) as well as scanning electron microscopy (SEM). According to Sulik, uptake of NBS is mostintense in regions that have a high percentage of apoptotic, but not necrotic cells. 257,258 SEM plusNBS staining revealed that treatment of pregnant mice with a single oral dose of 13-cis-RA on GD11 causes excessive cell death in the apical ectodermal ridge of fetal limbs 12 h after treatment.When observed on GD 14, limbs from treated mice exhibit malformations, including oligodactylyand polydactyly. 258 Treatment of pregnant mice with all-trans-RA on GD 11 caused mesomeliclimb defects in fetuses observed on GD 18. Cell death patterns observed 12 h posttreatmentsuggested that cell death induced by RA was associated with zones of PCD as seen in controlembryos. 259 Treatment of dams with 13-cis-RA at 8 d, 14 h or 9 d, 6 h caused malformations ofthe secondary palate in areas that coincide with the PCD that normally occurs in the first visceralarch. 260 Alles and Sulik 259 suggest that cells in the vicinity of those undergoing PCD may be inducedto undergo PCD abnormally if perhaps an endogenous signal is stronger than usual or if anexogenous agent has a mechanism similar to that of the normal stimuli.2. DioxinTCDD is another teratogen that may act by altering normal patterns of PCD. In at least one exampleof exposure during embryogenesis, TCDD prevented PCD. 241 Between GD 14 and 16, the medialperidermal cells of mouse embryonic palatal shelves normally stop incorporating 3 H-thymidine,and both EGF binding and EGF receptor expression decrease. In control shelves, medial epithelialperidermal cells detached from basal cells, and a high percentage of cells contained dense cytoplasmand pyknotic nuclei (two features that are characteristic of PCD). In contrast, shelves exposed to© 2006 by Taylor & Francis Group, LLC

44 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTCDD in vivo or in vitro on GD 10 or in vitro on GD 12 continued expressing EGFR and bindingEGF and showed no peridermal cell degeneration. 32,2413. EthanolThe mechanism of action of ethanol-induced developmental toxicity has not been elucidated fully;however, oxidative stress, interference with growth factor regulation, and interference with retinoicacid synthesis are leading hypotheses. 261–263 The complex nature of ethanol-induced developmentalneurotoxicity provides a unique opportunity to evaluate the potential integration of numerousproposed mechanisms into a common mode of action. Of particular relevance to this discussionare the relative effects of ethanol on proliferation versus those on apoptosis, as both have beenproposed as key toxic impacts. 264–267Ethanol is a particularly prevalent and harmful developmental neurotoxicant. Numerous anomaliesare characteristic of gestational ethanol exposure, including general growth retardation, abnormalbrain development, microcephaly, mental retardation, and specific craniofacial malformations.261,268 Data from human and rat embryological and morphological studies identifying targetsof ethanol toxicity are consistent. Morphologically, the human and rat central nervous system arehighly susceptible to ethanol-induced growth retardation as manifested in microcephalic children andmicroencephalic rodent models. 261,269,270 Also, dissections of human FAS brains show similar deformitiesto those seen in rats, including decreased cerebral cortex, hippocampus, and cerebellum size.Many recent advances in design based on “unbiased” stereological methods in the field ofneuroscience have lead to a more complete picture of cellular composition and its relationship todevelopment and aging, and to a greater understanding of the effects of exogenous toxicants. 271Before the development of three-dimensional, unbiased stereological techniques to determine particlenumber in a structure, the field relied on two-dimensional “assumption-based” countingmethods to determine cellular density and on morphometrics for volume estimates. Earlier studiesmaking relative comparisons based on only density or volume estimates may be misleading, becausechanges in just one of these parameters may not reflect changes in cell number. To account forvariations in both of these parameters simultaneously, total cell number must be analyzed. Furthermore,earlier two-dimensional estimates of numerical density may be biased because of after-thefactcorrections for assumptions regarding particle size, shape, and orientation (e.g., Abercrombie,Weibul, and Gomez corrections). For example, larger nuclei will appear more often in sectionsthan will smaller nuclei, and nuclei will appear more often in sections when they are cut acrosstheir long axis. Assumptions about the size, shape, and orientation accompany corrections for splitnuclei, lost caps, and overprojection. Recent articles highlight this current controversy. 272–275The development of design-based stereological methods made possible statistically unbiasedestimation of volume, area, surface boundary, length, population size, and density. This is accomplishedby using an a priori combination of systematic random sampling and counting rules toeliminate the biases associated with size, shape, orientation, and spatial distribution of objectsinstead of applying corrections after the fact. This has provided developmental toxicologists withimproved tools to investigate pathological differences.However, it must be emphasized that although these new methods theoretically eliminatestatistical biases, the methods have no way of eliminating biases associated with the methodologies(e.g., sectioning and embedding) or observations (e.g., correct identification of a particle or structureboundary by the investigator). For example, bias introduced by counting poorly stained, thicksections used in optical methods or by counting sections by trying to align two separate sectionsmay well be much greater than the statistical bias introduced by two-dimensional methods. Furthermore,these new methods have not been rigorously compared with the two-dimensional estimatesor validated by three-dimensional reconstruction studies. Therefore, no independent, quantifiedmeasures of the benefits, whether in accuracy or efficiency, of these new methods over theold methods has been done. 272–275© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 45Stereological methods provide invaluable tools for estimation of cell number in various structures.Effects caused by perturbations can be measured precisely and efficiently to determine if anexposure affects the volume of a structure or if it also influences particle number. Stereologicalmethods provide quantitation of pathological states that can then be used in quantitative riskassessments. 271,276Investigations of ethanol-induced neuronal loss highlight the usefulness of stereology in toxicology.Investigators have documented that the most severe deficits in cell number may occur inthe neocortex, the hippocampus, or the cerebellum, depending on the stage of CNS developmentin which exposure to ethanol occurs. 264–267,277 Not only are distinct regions of the brain affecteddifferently by ethanol, but the timing and pattern of exposure plays a critical role in the finaloutcome of ethanol-induced cellular loss. 278,279 The neocortex is particularly sensitive to neuronalloss following a relatively low exposure (resulting in a peak blood ethanol concentration [BEC] of150 mg/dl which is roughly equivalent to a BEC in humans after imbibing three to five standarddrinks) during early developmental events, including neurogenesis and migration.Decreases in cell numbers can be caused by a decrease in proliferation or an increase in celldeath, and various studies have shown ethanol to be a potent inhibitor of cellular proliferation. Asingle dose of ethanol administered to female rats within 8 h after mating results in a dose-dependentretardation of cell division in the fertilized ova, which is sustained up to 42 h after the exposure. 280Within the cerebral cortex, ethanol-exposed rat fetuses generate 30% fewer neocortical neuronsbetween gestational day (GD) 12 and 19, the peak time of cortical neurogenesis in the rat. 281 When3H-thymidine incorporation into rat fetal brain and liver tissue after exposure to ethanol in uteroon GD 16 and 20 was compared, the brain tissue showed decreased incorporation, suggestingincreased susceptibility of proliferating neuronal and glial precursors as opposed to proliferatingliver cells. 282In-depth in vivo research on neocortical neurogenesis in the mouse model has been performed,relating functional data (cell cycle rates and migration) within an anatomical context. 289,290 Ethanolinducedeffects documented include a reduction in the proliferating cell population and an increasein the length of the cell cycle, both contributing to fewer numbers of neurons or glial cellsgenerated. 283–286 The development of a cumulative BrdU incorporation technique allowed determinationof the effect of moderate alcohol intake (peak BEC of 153 mg/dl) on the cell cycle lengthof the proliferating cells of the dorsal neocortices. 266,287 These studies have shown 30% increase incell cycle length (18 h compared with 11 h) during early neocortical neurogenesis (GD 13 to 16);however, the increase was not constant throughout neurogenesis. As normal neurogenesis proceeds,the cell cycling rate naturally becomes longer, whereas the ethanol-exposed cells showed the samecell cycling rate throughout cortical neurogenesis. No increase in pyknotic cells was detected,suggesting again that the cycling cell is the target. 288The postnatal period of neocortical synaptogenesis for the rat, which includes the brain growthspurt, is a highly sensitive period. In vitro and in vivo studies suggest that ethanol neurotoxicityduring this period may be orchestrated by mechanisms different from those governing the earlierperiod of neocortical neurogenesis. Recently, convincing evidence of increased cell death due topostnatal exposure to ethanol has been documented. The period of natural cell death for the cerebralcortex occurs between postnatal days (PD) 1 and 10 in the rat, with a peak on PD 7. 292 Ikonomidou 293showed that by blocking N-methyl-D-aspartate (NMDA) glutamate receptors and activating γ-aminobutyric acid (GABA) receptors, ethanol triggers widespread apoptosis (as great as 30 timesthe baseline rate) in the synaptogenesis period of many brain regions, including the hippocampus,thalamus, and frontal, parietal, cingulate, and retrosplenial cortex. In this study, cell death wasmeasured by DeOlmos silver staining and was confirmed with caspase 3 activation in a more recentstudy. 294 Furthermore, if blood concentrations exceeded 200 mg/dl for 4 h, apoptotic neurodegenerationwas significantly increased compared with controls. If this threshold was exceeded for morethan 4 h, the degenerative response became progressively more severe in proportion to the length© 2006 by Taylor & Francis Group, LLC

46 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof time the threshold had been exceeded. A discrete window of time, coinciding with the synaptogenesisperiod of each region tested and occurring anywhere from GD 19 to PD 14, dependingon the region, was the most susceptible period. 293 However, in comparison with the peak BEC of150 mg/dl shown to inhibit proliferation during the neurogenesis stage, the threshold of 200 mg/dlBEC for over 4 h is relatively high.In light of the above evidence of ethanol-induced cell death, it is interesting that no stereologicalstudies to date have shown reductions in neuronal cell number in the neocortex when exposureoccurs during the synaptogenesis period. However, this exposure scenario has been shown to causereduction in the volume and weight of cortical structures. 277,279 Conversely, Mooney et al. 277 showedthat no decrease in adult cortical volume or in neuronal cell number occurred after rat pups weresubjected to a bingelike exposure regimen from PD 4 to 9, acheiving a peak BEC of approximately300 mg/dl. Furthermore, no differences from controls were detected in adult cerebral cortex DNAcontent after postnatal exposure to ethanol. 278 When comparing results from prenatal and postnatalexposure paradigms, one must keep in mind that the exposure scenario may result in differentpharmacokinetics becoming a potential confounding factor. 295 The studies described in this sectionfor ethanol-induced neuronal deficits show some of the approaches available as well as complexitiesin linking observation of these morphological impacts with developmental dynamics. The differentialimpacts of dynamic mechanisms, such as insults to proliferation compared with cell death,can be quantitatively evaluated using BBDR models. 296III. CONCLUSIONSAs our initial quote by Ovid suggests, the mazelike path to discernment of mechanisms of developmentaltoxicity is indeed challenging. The aim of this chapter has been to suggest guidelines forevaluating these paths and to provide approaches for new mechanistic investigations. We haveemphasized the idea that mechanisms can be defined on many levels of organization, ranging fromthe molecular and cellular to the whole organism. We have also illustrated the importance of usingan evaluation framework that allows for examination of both kinetic as well as dynamic responsesin deciphering mechanistic clues. Although few complete mechanisms are known, recent advancesin molecular and cellular biology have provided new mechanistic “threads” and like Theseus, the“elusive gateway” beckons.A significant challenge for mechanism-based research is the need to identify critical rate-limitingchanges that are associated with an adverse outcome. In many cases, our ability to measure changes(for example, microarray gene changes) outstrips our ability to interpret the significance of subtlechanges. Thus, besides the criteria for causality that are discussed within this chapter; there is aneed to link quantitatively dynamic changes at any and/or all the levels of biological outcome withthe manifestations and the dose and temporal context of developmental toxicity. For example, inthe section in this chapter that discusses gene expression changes as proposed mechanisms ofdevelopmental toxicity, there is a need to understand the significance of subtle, transient changesin overall developmental outcomes over time and in the context of rapidly changing morphology.Although faced with somewhat similar issues in the past, such as assessments of subtle changesin fetal body weight, the sensitivity and ready availability of genomic tools will press the need tofrequently address and reevaluate this assessment issue. Presentation of a kinetic and dynamicframework for organizing our multilevel assessment is one step toward this evaluation. 1ACKNOWLEDGMENTSThis work was supported by the Institute for Risk Analysis and Risk Communication, the UWCenter for Child Environmental Health Risks research (EPA R826886 and NIEHS 1PO1ES09601),© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 47the UW NIEHS Center for Ecogenetics and Environmental Health (5 P30 ES07033), the Institutefor Evaluating Health Risks, DOE Low Dose Radiation Research Program Grant, and EPA ContractNo. W-2296-NATA.References1. National Research Council, Scientific Frontiers in Developmental Toxicity Risk Assessment, NationalAcademy Press (Committee on Developmental Toxicology, Board on Environmental Studies andToxicology, National Research Council), Washington D.C., 2000, p. 354.2. ILSI, I. L. S. I., Workshop to develop a framework for assessing risks to children from exposures toenvironmental agents, Environ. Health Perspect., 2003 (in press).3. Faustman, E.M., Griffith, W., Lewandowski, T., Gohlke, J., and Webb, T., IRARC Technical Reporton Developmental Toxicity, Institute for Risk Analysis and Risk Communication, University of Washington,Seattle, WA, 2003a.4. Barker, D., Fetal origins of coronary heart disease, BMJ, 311, 1995.5. Wilson, J.G., Current status of teratology: general principles and mechanisms derived from animalstudies, in Handbook of Teratology: General Principles and Etiology, Wilson, J.G. and Fraser, F.C.,Eds. Plenum Press, New York, 1977.6. Chabner, B., Allegra, C., Curt, G., and Calabresi, P., Antineoplastic agents, in Goodman and Gilman’sThe Pharmacological Basis of Therapeutics, Hardman, J., Limbird, L., Molinoff, P., et al., Eds.,McGraw-Hill, New York, 1996, p. 1233.7. Hill, A., The environment and disease: association or causation, Proc. Roy. Soc. Med., 58, 295, 1965.8 IPCS, I. P. o. C. S. IPCS workshop on developing a conceptual framework for cancer risk assessment,Lyon, France, 1999,9. U.S. EPA, Draft final guidelines for carcinogen risk assessment, U.S. EPA, Washington, D.C., 2003.10. Selevan, S.G. and Lemasters, G.K., The dose-response fallacy in human reproductive studies of toxicexposures, J. Occup. Med., 29, 451, 1987.11. Faustman, E.M., Kirby, Z., Gage, D., and Varnum, M., In vitro developmental toxicity of five directacting alkylating agents in rodent embryo cultures: Structure — activity patterns, Teratology, 40, 199,1989.12. Platzek, T., Bochert, G., Schneider, W., and Neubert, D., Embryotoxicity induced by alkylating agents:1. Ethylmethanesulfonate as a teratogen in mice — a model for dose response relationships ofalkylating agents, Arch. Toxicol., 51, 1, 1982.13. Platzek, T., Bochert, G., Meister, R., and Neubert, D., Embryotoxicity induced by alkylating agents:7. Low dose prenatal-toxic risk estimation based on NOAEL risk factor approach, dose-responserelationships, and DNA adducts using methylnitrosourea as a model compound, Teratog. Carcinog.Mutagen., 13, 101, 1993.14. Kistler, A., Limb bud cell cultures for estimating the teratogenic potential of compounds. Validationof the test system with retinoids, Arch. Toxicol., 60, 403, 1987.15. Willhite, C.C., Wier, P.J., and Berry, D.L., Dose response and structure-activity considerations inretinoid-induced dysmorphogenesis, Crit. Rev. Toxicol., 20, 113, 1989.16. Rawlings, S.J., Shuker, D.E., Webb, M., and Brown, N.A., The teratogenic potential of alkoxy acidsin post-implantation rat embryo culture: structure-activity relationships, Toxicol. Lett., 28, 49, 1985.17. Brown, N.A., Coakley, M.E., and Clarke, D.O., Structure-teratogenicity relationships of valproic acidcongeners in whole-embryo culture, in Approaches to Elucidate Mechanisms in Teratogenesis, Welsch,F., Ed., Hemisphere Publishing Corp., Washington, D.C., 1987, p. 17.18. Oglesby, L.A., Ebron-McCoy, M.T., Logsdon, T.R., Copeland, F., Beyer, P.E., and Kavlock, R.J., Invitro embryotoxicity of a series of para-substituted phenols: structure, activity, and correlation within vivo data, Teratology, 45, 11, 1992.19. McMahon, A.P., Joyner, A.L., Bradley, A., and McMahon, J.A., The midbrain-hindbrain phenotypeof Wnt-1-/Wnt-1-mice results from stepwise deletion of engrailed-expressing cells by 9.5 days postcoitum,Cell, 69, 581, 1992.20. Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H., et al., Cyclopia anddefective axial patterning in mice lacking Sonic hedgehog gene function, Nature, 383, 407, 1996.© 2006 by Taylor & Francis Group, LLC

48 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION21. Cooper, M.K., Porter, J.A., Young, K.E., and Beachy, P.A., Teratogen-mediated inhibition of targettissue response to Shh signaling, Science, 280, 1603, 1998.22. Incardona, J.P., Gaffield, W., Kapur, R.P., and Roelink, H., The teratogenic Veratrum alkaloid cyclopamineinhibits sonic hedgehog signal transduction, Development, 125, 3553, 1998.23. Roelink, H., Porter, J.A., Chiang, C., Tanabe, Y., Chang, D.T., Beachy, P.A., et al., Floor plate andmotor neuron induction by different concentrations of the amino-terminal cleavage product of sonichedgehog autoproteolysis, Cell, 81, 445, 1995.24. Nebert, D., Pharmacogenetics and pharmacogenomics: Why is this relevant to the clinical geneticist?,Clin. Genet., 56, 247, 1999.25. Couture, L.A., Abbott, B.D., and Birnbaum, L.S., A critical review of the developmental toxicity andteratogenicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin: recent advances toward understanding the mechanism,Teratology, 42, 619, 1990.26. Johnson, R. and Tabin, C., Molecular models for vertebrate limb development, Cell, 90, 979, 1997.27. Ferretti, P. and Tickle, C., The limbs, in Embryos, Genes and Birth Defects, Thorogood, P., Ed., JohnWiley & Sons, Chichester, 1997, p. 101.28. Tabin, C., A developmental model for thalidomide defects, Nature, 396, 322, 1998.29. Stephens, T.D. and Fillmore, B.J., Hypothesis: thalidomide embryopathy — proposed mechanism ofaction, Teratology, 61, 189, 2000.30. Brown, N., Normal Development: Mechanism of Early Embryogenesis, 2nd Ed., Raven Press, NewYork, 1994, p. 15.31. Abbott, B.D., Probst, M.R., Perdew, G.H., and Buckalew, A.R., AH receptor, ARNT, glucocorticoidreceptor, EGF receptor, EGF, TGF alpha, TGF beta 1, TGF beta 2, and TGF beta 3 expression inhuman embryonic palate, and effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), Teratology, 58,30, 1998.32. Abbott, B.D., Diliberto, J.J., and Birnbaum, L.S., 2,3,7,8-Tetrachlorodibenzo-p-dioxin alters embryonicpalatal medial epithelial cell differentiation in vitro, Toxicol Appl Pharmacol, 100, 119, 1989.33. Abbott, B.D., Harris, M.W., and Birnbaum, L.S., Etiology of retinoic acid-induced cleft palate varieswith the embryonic stage, Teratology, 40, 533, 1989.34. Gohlke, J., Griffith, W., Bartell, S., Lewandowski, T., and Faustman, E., A computational model forneocortical neuronogenesis predicts ethanol-induced neocortical neuron number deficits, Develop.Neurosci., 24, 467, 2002.35. Bernfield, M., Mechanisms of congenital malformations, in The Biological Basis of Reproductive andDevelopmental Medicine, Warshaw, J.B., Ed., Elsevier Biomedical, New York, 1983, p. 143.36. Chow, I., Cell-cell interaction during synaptogenesis, Physiol., Paris, 84, 121, 1990.37. Doherty, P. and Walsh, F.S., The contrasting roles of N-CAM and N-cadherin as neurite outgrowthpromotingmolecules, Cell Sci., 15, 13, 1991.38. Grumet, M., Structure, expression and function of Ng-CAM, a member of the immunoglobulinsuperfamily involved in neuron-neuron and neuron-glia adhesion., Neurosci. Res., 31, 1, 1992.39. Hatten, M.E. and Mason, C.A., Mechanisms of glial-guided neuronal migration in vitro and in vivo,Experientia, 46, 907, 1990.40. Lehmann, S., Kuchler, S., Theveniau, M., Vincendon, G., and Zanetta, J.P., An endogenous lectin andone of its neuronal glycoprotein ligands are involved in contact guidance of neuronal migration, Nat..Acad. Sci., 87, 6455, 1990.41. Liesi, P., Extracellular matrix and neuronal movement, Experientia, 46, 900, 1990.42. Pinto-Lord, M.C., Evrard, P., and Caviness, V.S., Jr., Obstructed neuronal migration along radial glialfibers in the neocortex of the reeler mouse: a golgi-EM analysis, Brain Res., 256, 379, 1982.43. Rakic, P., Defects of neuronal migration and the pathogenesis of cortical malformations, Prog. BrainRes., 73, 15, 1988.44. Tang, J., Landmesser, L., and Rutishauser, U., Polysialic acid influences specific pathfinding by avianmotoneurons, Neuron, 8, 1031, 1992.45. Walsh, F.S. and Doherty, P., Glycosylphosphatidylinositol anchored recognition molecules that functionin axonal fasciculation, growth and guidance in the nervous system, Cell Biol. Int. Rep., 15, 1151,1991.46. Wilkin, G.P. and Curtis, R., Cell adhesion molecules and ion pumps—Do ion fluxes regulate neuronalmigration?, Bioessays, 12, 287, 1990.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 4947. Kirschner, M. and Weber, K., Cytoplasm and cell motility, Current Sci., 1, 3, 1989.48. LoPachin, R.M. and Aschner, M., Glial-neuronal interactions: Relevance to neurotoxic mechanisms,Toxicol. Appl. Pharmacol., 118, 141, 1993.49. Yamada, K., Cell Morphogenic Movements, in Handbook of Teratology, Wilson, J. and Fraser, F.,Eds., Plenum Press, New York, 1977, p. 199.50. Pratt, R.M., Goulding, E.H., and Abbott, B.D., Retinoic acid inhibits migration of cranial neural crestcells in the cultured mouse embryo, J. Craniofac. Genet. Dev. Biol., 7, 205, 1987.51. Regan, C. M., Neural cell adhesion molecules, neuronal development and lead toxicity, Neurotoxicology,14, 69, 1993.52. Cookman, G.R., King, W., and Regan, C.M., Chronic low-level lead exposure impairs embryonic toadult conversion of the neural cell adhesion molecule, Neurochem., 49, 399, 1987.53. Graff, R. and KR, R., Neural cell adhesion molecules and molecules and microtubles are distincttargets of methylmercury in vitro, Toxicologist, 15, 1386, 1995.54. Bray, D. and Hollenbeck, P.J., Growth cone motility and guidance, Ann. Rev. Cell Biol., 4, 43, 1988.55. Trosko, J.E., Chang, C.C., and Netzloff, M., The role of inhibited cell-cell communication in teratogenesis,Teratog. Carcinog. Mutagen., 2, 31, 1982.56. Scott, W., Cell death and reduced proliferative rate, in Handbook of Teratology: Mechanisms andPathogenesis, Wilson, J. and Fraser, F., Eds., Plenum Press, New York, 1977, p. 81.57. Snow, M.H., Restorative growth in mammalian embryos, in Issues and Reviews in Teratology, Kalter,H., Ed., Plenum Press, New York, 1983, p. 251.58. Bayer, S.A., Cellular aspects of brain development, Neurotoxicology, 10, 307, 1989.59. Trasler, D.G. and Fraser, F.C., Time-position relationships, with particular reference to cleft lip andcleft palate, in Handbook of Teratology, Wilson, J.G. and Fraser, F. C., Eds., Plenum Press, New York,1977, p. 271.60. Abbott, B.D., Lau, C., Buckalew, A.R., Logsdon, T.R., Setzer, W., Zucker, R.M., et al., Effects of 5-fluorouracil on embryonic rat palate in vitro: fusion in the absence of proliferation, Teratology, 47,541, 1993.61. Chaube, S. and Murphy, M.L., The teratogenic effects of the recent drugs active in cancer chemotherapy,Advances Teratol., 3, 181, 1968.62. Elstein, K.H., Zucker, R.M., Shuey, D.L., Lau, C., Chernoff, N., and Rogers, J.M., Utility of themurine erythroleukemic cell (MELC) in assessing mechanisms of action of DNA-active developmentaltoxicants: application to 5-fluorouracil, Teratology, 48, 75, 1993.63. Shuey, D.L., Setzer, R.W., Lau, C., Zucker, R.M., Elstein, K.H., Narotsky, M.G., et al., Biologicalmodeling of 5-fluorouracil developmental toxicity, Toxicology, 102, 207, 1995.64. Parker, W.B. and Cheng, Y.C., Metabolism and mechanism of action of 5-fluorouracil, Pharmacol.Ther., 48, 381, 1990.65. Shuey, D.L., Buckalew, A.R., Wilke, T.S., Rogers, J.M., and Abbott, B.D., Early events followingmaternal exposure to 5-fluorouracil lead to dysmorphology in cultured embryonic tissues, Teratology,50, 379, 1994.66. Kimler, B.F. and Henderson, S.D., Cyclic responses of cultured 9L cells to radiation, Radiat. Res.,91, 155, 1982.67. Kimler, B., Prenatal irradiation: a major concern for the developing brain, Int. J. Radiat. Biol., 73,423, 1998.68. Brent, R.L. and Beckman, D.A., Environmental teratogens, Bull N. Y. Acad. Med., 66, 123, 1990.69. Jensh, R.P., Weinberg, I., and Brent, R.L., Teratologic studies of prenatal exposure of rats to 915 MHzmicrowave radiation, Radiat. Res., 92, 160, 1982.70. Rodier, P.M., Time of exposure and time of testing in developmental neurotoxicology, Neurotoxicology,7, 69, 1986.71. Schull, W., Norton, S., and Jensh, R., Ionizing radiation and the developing brain, Neurotox. Teratol.,12, 249, 1990.72. Bolaris, S., Bozas, E., Benekou, A., Philippidis, H., and Stylianopoulou, F., In utero radiation-inducedapoptosis and p53 gene expression in the developing rat brain, Int. J. Radiat. Biol., 77, 71, 2001.73. Brooks, A.L., Khan, M.A., Jostes, R.F., and Cross, F.T., Metaphase chromosome aberrations as markersof radiation exposure and dose, J. Toxicol. Environ. Health, 40, 277, 1993.© 2006 by Taylor & Francis Group, LLC

50 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION74. Frankenberg-Schwager, M., Induction, repair and biological relevance of radiation-induced DNAlesions in eukaryotic cells, Radiat. Environ. Biophys., 29, 273, 1990.75. Iliakis, G., The role of DNA double strand breaks in ionizing radiation-induced killing of eukaryoticcells, Bioessays, 13, 641, 1991.76. Brent, R.L., Radiations and other physical agents., in Handbook of Teratology, Wilson, J.G. and Fraser,F. C., Eds., Plenum Press, New York, 1977, p. 153.77. Beckman, D.A. and Brent, R.L., Mechanism of known environmental teratogens: drugs and chemicals,Clin. Perinatol., 13, 649, 1986.78. Whittaker, S.G. and Faustman, E.M., Effects of albendazole and albendazole sulfoxide on cultures ofdifferentiating rodent embryonic cells, Toxicol. Appl. Pharmacol., 109, 73, 1991.79. Albertini, S., Analysis of nine known or suspected spindle poisons for mitotic chromosome malsegregationusing Saccharomyces cerevisiae D61.M., Mutagenesis, 5, 453, 1990.80. Delatour, P. and Richard, Y., Proprietes embryotoxiques et antimitotiques en serie benzimidazole,Therapie, 31, 505, 1976.81. Eichenlaub-Ritter, U. and Boll, I., Nocodazole-sensitivity, age-related aneuploidy, and alterations inthe cell cycle during maturation of mouse oocytes, Cytogenet. Cell Genet., 52, 170, 1989.82. Generoso, W., Katoh, M., Cain, K., Hughes, L., Foxworth, L., Mitchell, T., et al., Chromosomemalsegregation and embryonic lethality induced by treatment of normally ovulated mouse oocyteswith nocodazole, Mutat. Res., 210, 313, 1989.83. Kappas, A., Georgopoulos, G., and Hastie, A.C., On the genetic activity of benzimidazole andthiophanate fungicides on diploid Aspergillus nidulans, Mutat. Res., 26, 17, 1974.84. Vanderkerken, K., Vanparys, P., Verschaeve, L., and Kirsch-Volders, M., The mouse bone marrowmicronucleus assay can be used to distinguish aneugens from clastogens, Mutagenesis, 4, 6, 1989.85. Whittaker, S.G., Moser, S.F., Maloney, D.H., Piegorsch, W.W., Resnick, M.A., and Fogel, S., Thedetection of mitotic and meiotic chromosome gain in the yeast Saccharomyces cerevisiae: effects ofmethyl benzimidazol-2-yl carbamate, methyl methanesulfonate, ethyl methanesulfonate, dimethylsulfoxide, propionitrile and cyclophosphamide monohydrate, Mutat. Res., 242, 231, 1990.86. Whittaker, S.G., Zimmermann, F.K., Dicus, B., Piegorsch, W.W., Fogel, S., and Resnick, M.A.,Detection of induced mitotic chromosome loss in Saccharomyces cerevisiae — an interlaboratorystudy, Mutat. Res., 224, 31, 1989.87. Delatour, P. and Parish, R., Benzimidazole anthelmintics and related compounds: Toxicity and evaluationof residues, in Drug Residues in Animals, Rico, A.G., Ed., Academic Press, Orlando, FL, 1986,p. 175.88. Ellis, W.E., Roos, F.D., Kavlock, R.J., and Zeman, F.J., Relationship of periventricular overgrowth tohydrocephalus in brains of fetal rats exposed to benomyl, Teratogen. Carcinog. Mutagen., 8, 377, 1988.89. Van-Dyke, J. and Ritchey, M., Colchicine influence during embryonic development in rats, Anat. Rec.,97, 375, 1947.90. Whittaker, S.G. and Faustman, E.M., Effects of benzimidazole analogs on cultures of differentiatingrodent embryonic cells, Toxicol. Appl. Pharmacol., 113, 144, 1992.91. Hughes, W., Ann. N.Y. Acad. Sci., 65, 454, 1957.92. Kajiwara, Y., Yasutake, A., Adachi, T., and Hirayama, K., Methylmercury transport across the placentavia neutral amino acid carrier, Arch. Toxicol., 70, 310, 1996.93. Brubaker, P., Kleing, R., Herman, S., Lucier, G., Alexander, L., and Long, M., DNA, RNA, and proteinsynthesis in brain, liver, and kidneys of asymptomatic methylmercury treated rats, Exp. Mol. Pathol.,18, 263, 1973.94. Syversen, T., Effects of methyl mercury on protein synthesis in vitro, Acta. Pharm. Tox., 49, 422, 1981.95. Salvaterra, P., Lown, B., Morganti, J., and Massaro, E., Alterations in neurochemical and behavioralparameters in the mouse induced by low doses of methyl mercury, Acta. Pharmacol. Tox., 33, 177, 1973.96. Massaro, E., The developmental cytotoxicity of mercurials, in Toxicology of Metals, Chang, L., Ed.,CRC Press, Boca Raton, FL, 1996.97. Lund, B.O., Miller, D.M., and Woods, J.S., Mercury-induced H 2 O 2 production and lipid peroxidationin vitro in rat kidney mitochondria, Biochem. Pharmacol., 42 Suppl, S181, 1991.98. Denny, M.F. and Atchison, W.D., Mercurial-induced alterations in neuronal divalent cation homeostasis,Neurotoxicology, 17, 47, 1996.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 5199. Hare, M.F., McGinnis, K.M., and Atchison, W.D., Methylmercury increases intracellular concentrationsof Ca++ and heavy metals in NG108-15 cells, J. Pharmacol. Exp. Ther., 266, 1626, 1993.100. Miura, K., Suzuki, K., and Imura, N., Effects of methylmercury on mitotic mouse glioma cells,Environ. Res., 17, 453, 1978.101. Sager, P., Doherty, R., and Rodier, P., Effects of methylmercury on developing mouse cerebellar cortex,Exp. Neurol., 77, 179, 1982.102. Vogel, D., Margolis, R., and Mottet, N., The effects of methyl mercury binding to microtubules, Tox.Appl. Pharm., 80, 473, 1985.103. Zucker, R., Elstein, K., Easterling, R., and Massaro, E., Flow cytometric analysis of the mechanismof methylmercury cytotoxicity, Am. J. Pathol., 137, 1187, 1990.104. Clarkson, T., Mercury — an element of mystery, N. Engl. J. Med., 323, 1137, 1990.105. Ali, S.F., LeBel, C.P., and Bondy, S.C., Reactive oxygen species formation as a biomarker of methylmercuryand trimethyltin neurotoxicity, Neurotoxicology, 13, 637, 1992.106. Woods, J. and Ellis, M., Upregulation of glutathione synthesis in rat kidney by methyl mercury:relationship to mercury-induced oxidative stress, Biochem. Pharmacol., 50, 1719, 1995.107. Ornaghi, F., Ferrini, S., Prati, M., and Giavini, E., The protective effects of N-acetyl-l-cysteine againstmethyl mercury embryotoxicity in mice, Fund. Appl. Tox., 20, 437, 1993.108. Yohana, M., Saito, M., and Sagai, M., Stimulation of lipid peroxidation by methyl mercury in rats,Life Sci., 32, 1507, 1983.109. Sarafian, T. and Verity, M., Oxidative mechanisms underlying methyl mercury neurotoxicity, Int. J.Devl. Neurosci., 9, 147, 1991.110. Verity, M.A., Torres, M., and Sarafian, T., Paradoxical potentiation by low extracellular Ca 2+ of acutechemical anoxic neuronal injury in cerebellar granule cell culture, Mol. Chem. Neuropathol., 15, 217, 1991.111. Burbacher, T.M., Rodier, P.M., and Weiss, B., Methylmercury developmental neurotoxicity: a comparisonof effects in humans and animals, Neurotoxicol. Teratol., 12, 191, 1990.112. Matsumoto, H., Koya, G., and Takeuchi, T., Fetal Minamata disease. A neuropathological study oftwo cases of intrauterine intoxication by a methyl mercury compound, J. Neuropathol. Exp. Neurol.,24, 563, 1965.113. Choi, B.H., Lapham, L.W., Amin-Zaki, L., and Saleem, T., Abnormal neuronal migration, derangedcerebral cortical organization, and diffuse white matter astrocytosis of human fetal brain: a majoreffect of methylmercury poisoning in utero, J. Neuropathol. Exp. Neurol., 37, 719, 1978.114. Chen, W.J., Body, R.L., and Mottet, N.K., Some effects of continuous low-dose congenital exposureto methylmercury on organ growth in the rat fetus, Teratology, 20, 31, 1979.115. Geelen, J.A., Dormans, J.A., and Verhoef, A., The early effects of methylmercury on the developingrat brain, Acta Neuropathol. (Berl.), 80, 432, 1990.116. Kakita, A., Wakabayashi, K., Su, M., Sakamoto, M., Ikuta, F., and Takahashi, H., Distinct pattern ofneuronal degeneration in the fetal rat brain induced by consecutive transplacental administration ofmethylmercury, Brain Res., 859, 233, 2000.117. Sager, P., Aschner, M., and Rodier, P., Persistent, differential alterations in developing cerebellar cortexof male and female mice after methylmercury exposure, Dev. Brain Res., 12, 1, 1984.118. Castoldi, A.F., Barni, S., Turin, I., Gandini, C., and Manzo, L., Early acute necrosis, delayed apoptosisand cytoskeletal breakdown in cultured cerebellar granule neurons exposed to methylmercury, J.Neurosci. Res., 59, 775, 2000.119. Miura, K., Koide, N., Himeno, S., Nakagawa, I., and Imura, N., The involvement of microtubulardisruption in methylmercury-induced apoptosis in neuronal and nonneuronal cell lines, Toxicol. Appl.Pharmacol., 160, 279, 1999.120. Nagashima, K., Fujii, Y., Tsukamoto, T., Nukuzuma, S., Satoh, M., Fujita, M., et al., Apoptotic processof cerebellar degeneration in experimental methylmercury intoxication of rats, Acta Neuropathol.(Berl.), 91, 72, 1996.121. Rodier, P., Aschner, M., and Sager, P., Mitotic arrest in the developing CNS after prenatal exposureto methylmercury, Neurobehav. Toxicol. Teratol., 6, 379, 1984.122. Ponce, R., Kavanagh, T., Mottet, N., Whittaker, S., and Faustman, E., Effects of methyl mercury onthe cell cycle of primary rat CNS cells in vitro, Toxicol. Appl. Pharmacol., 127, 83, 1994.© 2006 by Taylor & Francis Group, LLC

52 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION123. Lewandowski, T., Ponce, R., Charleston, J., Hong, S., and Faustman, E., Effect of methylmercury onmidbrain cell proliferation during organogenesis: Potential cross-species differences and implicationsfor risk assessment, Toxicol. Sci., 75(1), 124–133, 2003.124. Leroux, B.G., Leisenring, W.M., Moolgavkar, S.H., and Faustman, E.M., A biologically-based doseresponsemodel for developmental toxicology, Risk Anal., 16, 449, 1996.125. Faustman, E., Gohlke, J., Judd, N., Lewandowski, T., Bartell, S., and Griffith, W., Modeling developmentalprocesses in animals: applications in neurodevelopmental toxicology, Environ. Toxicol.Pharmacol., 19, 615, 2005.126. Lewandowski, T., Pierce, C., Pingree, S., Hong, S., and Faustman, E., Methylmercury distribution inthe pregnant rat and embryo during early midbrain organogenesis, Teratology, 66, 2002.127. Faustman, E., Lewandowski, T., Ponce, R., and Bartell, S., Biologically based dose-response modelsfor developmental toxicants: lessons from methlymercury, Inhalat. Toxicol., 11, 101, 1999.128. Fantel, A.G., Person, R.E., Burroughs, G.C., Shepard, T.H., Juchau, M.R., and Mackler, B., Asymmetricdevelopment of mitochondrial activity in rat embryos as a determinant of the defect patternsinduced by exposure to hypoxia, hyperoxia, and redox cyclers in vitro, Teratology, 44, 355, 1991.129. Mackler, B., Grace, R., Haynes, B., Bargman, G.J., and Shepard, T.H., Studies of mitochondrial energysystems during embryogenesis in the rat, Arch. Biochem. Biophys., 158, 662, 1973.130. Aksu, O., Mackler, B., Shepard, T.H., and Lemire, R.J., Studies of the development of congenitalanomalies in the ribloflavin-deficient, galactoflavin fed rats. II. Role of the terminal electron transportsystem, Teratology, 1, 93, 1968.131. Mackler, B., Davis, K.A., and Grace, R., Cytochrome a3 deficiency in human achondroplasia, Biochim.Biophys. Acta., 891, 145, 1987.132. Mackler, B., Grace, R., Davis, K.A., Shepard, T.H., and Hall, J.G., Studies of human achondroplasia:oxidative metabolism in tissue culture cells, Teratology, 33, 9, 1986.133. Mackler, B. and Shepard, T.H., Human achondroplasia: defective mitochondrial oxidative energymetabolism may produce the pathophysiology, Teratology, 40, 571, 1989.134. Shepard, T., Lemire, R.J., Aksu, O., and Mackler, B., Studies of the development of congenitalanomalies in the riboflavin-deficient, galactoflavin fed rats. I. Growth and embryologic pathology,Teratology, 1, 75, 1968.135. Mackler, B., Grace, R., Person, R., Shepard, T.H., and Finch, C.A., Iron deficiency in the rat:biochemical studies of fetal metabolism, Teratology, 28, 103, 1983.136. Sulik, K.K., Cook, C.S., and Webster, W.S., Teratogens and craniofacial malformations: relationshipsto cell death, Development, 103 (Suppl.) 213, 1988.137. Shepard, T.H., Mackler, B., and Finch, C.A., Reproductive studies in the iron-deficient rat, Teratology,22, 329, 1980.138. Mackler, B., Grace, R., Tippit, D.F., Lemire, R.J., Shepard, T.H., and Kelley, V.C., Studies of thedevelopment of congenital anomalies in rats. III. Effects of inhibition of mitochondrial energy systemson embryonic development, Teratology, 12, 291, 1975.139. Ranganathan, S., Churchill, P.F., and Hood, R.D., Inhibition of mitochondrial respiration by cationicrhodamines as a possible teratogenicity mechanism, Toxicol. Appl. Pharmacol., 99, 81, 1989.140. Fantel, A.G., Person, R.E., Burroughs, G.C.J., and Mackler, B., Direct embryotoxicity of cocaine inrats: effects on mitochondrial activity, cardiac function, and growth and development in vitro, Teratology,42, 35, 1990.141. Finnell, R.H., Toloyan, S., van-Waes, M., and Kalivas, P.W., Preliminary evidence for a cocaineinducedembryopathy in mice, Toxicol. Appl. Pharmacol., 103, 228, 1990.142. Plessinger, M.A. and Woods, J.R., Jr., Maternal, placental, and fetal pathophysiology of cocaineexposure during pregnancy, Clin. Obstet. Gynecol., 36, 267, 1993.143. Fantel, A.G., Barber, C.V., and Mackler, B., Ischemia/reperfusion: a new hypothesis for the developmentaltoxicity of cocaine, Teratology, 46, 285, 1992.144. Hood, R.D., Ranganathan, S., Jones, C.L., and Ranganathan, P.N., Teratogenic effects of a lipophiliccationic dye rhodamine 123, alone and in combination with 2-deoxyglucose, Drug Chem. Toxicol.,11, 261, 1988.145. DeSesso, J.M. and Goeringer, G.C., Methotrexate-induced developmental toxicity in rabbits is amelioratedby 1-(p-tosyl)-3,4,4-trimethylimidazolidine, a functional analogue for tetrahydrofolate mediatedone carbon transfer, Teratology, 45, 271, 1992.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 53146. DeSesso, J. and Goeringer, G., Amelioration by leucovorin of methotrexate developmental toxicityinrabbitss, Teratology, 43, 201, 1991.147. Costa, L.G., Organophosphorus compounds, in Recent Advances in Nervous System Toxicology,Spencer, P.S., Ed., Plenum, New York, 1988, p. 203.148. Levine, R., Recognized and possible effects of pesticides in humans, in Handbook of PesticidesToxicology, Hayes, W. and Lawa, E., Eds., Academic Press, New York, 1991, p. 275.149. Small, D.H., Michaelson, S., and Sberna, G., Non-classical actions of cholinesterases: role in cellulardifferentiation, tumorigenesis and Alzheimer’s disease, Neurochem. Int., 28, 453, 1996.150. Ribeiro, P.L. and Faustman, E.M., Embryonic micromass limb bud and midbrain cultures: differentcell cycle kinetics during differentiation, Toxic. In Vitro, 4/5, 602, 1990.151. Koenigsberger, C., Chiappa, S., and Brimijoin, S., Neurite differentiation is modulated in neuroblastomacells engineered for altered acetylcholinesterase expression, J. Neurochem., 69, 1389, 1997.152. Jbilo, O., L’Hermite, Y., Talesa, V., Toutant, J.P., and Chatonnet, A., Acetylcholinesterase and butyrylcholinesteraseexpression in adult rabbit tissues and during development, Eur. J. Biochem., 225, 115, 1994.153. Saito, S., Cholinesterase inhibitors induce growth cone collapse and inhibit neurite extension inprimary cultured chick neurons, Neurotoxicol. Teratol., 20, 411, 1998.154. Jett, D.A., Navoa, R.V., Beckles, R.A., and McLemore, G.L., Cognitive function and cholinergicneurochemistry in weanling rats exposed to chlorpyrifos, Toxicol. Appl. Pharmacol., 174, 89, 2001.155. Crumpton, T., Seidler, F., and Slotkin, T., Is oxidative stress involved in the developmental neurotoxicityof chlorpyrifos?, Dev. Brain. Res., 121, 189, 2000.156. Levin, E.D., Addy, N., Baruah, A., Elias, A., Christopher, N.C., Seidler, F.J., et al., Prenatal chlorpyrifosexposure in rats causes persistent behavioral alterations, Neurotoxicol. Teratol., 24, 733, 2002.157. Song, X., Violin, J.D., Seidler, F.J., and Slotkin, T.A., Modeling the developmental neurotoxicity ofchlorpyrifos in vitro: macromolecule synthesis in PC12 cells, Toxicol. Appl. Pharmacol., 151, 182,1998.158. Qiao, D., Seidler, F.J., and Slotkin, T.A., Developmental neurotoxicity of chlorpyrifos modeled invitro: comparative effects of metabolites and other cholinesterase inhibitors on DNA synthesis inPC12 and C6 cells, Environ. Health Perspect., 109, 909, 2001.159. Das, K. and Barone, S., Jr., Neuronal differentiation in PC12 cells is inhibited by chlorpyrifos andits metabolites: Is acetylcholinesterase inhibition the site of action?, Toxicol. Appl. Pharmacol., 160,217, 1999.160. Chanda, S.M. and Pope, C.N., Neurochemical and neurobehavioral effects of repeated gestationalexposure to chlorpyrifos in maternal and developing rats, Pharmacol. Biochem. Behav., 53, 771, 1996.161. Moser, V.C. and Padilla, S., Age- and gender-related differences in the time course of behavioral andbiochemical effects produced by oral chlorpyrifos in rats, Toxicol. Appl. Pharmacol., 149, 107, 1998.162. Slotkin, T.A., Developmental cholinotoxicants: nicotine and chlorpyrifos, Environ. Health Perspect.,107 Suppl, 1, 1999.163. Pinedo, H.M. and Peters, G.F., Fluorouracil: biochemistry and pharmacology, J. Clin. Oncol., 6, 1653,1988.164. O’Dwyer, P.J., King, S.A., Hoth, D.F., and Leyland-Jones, B., Role of thymidine in biochemicalmodulation: a review, Cancer Res., 47, 3911, 1987.165. Scheutz, J., Wallace, H., and Diasio, R., 5-Fluorouracil incorporation into DNA of CF-1 mouse bonemarrow cells as a possible mechanism of toxicity, Cancer Res., 44, 1358, 1984.166. Shuey, D.L., Lau, C., Logsdon, T.R., Zucker, R.M., Elstein, K.H., Narotsky, M.G., et al., Biologicallybased dose-response modeling in developmental toxicology: biochemical and cellular sequelae of 5-fluorouracil exposure in the developing rat, Toxicol. Appl. Pharmacol., 126, 129, 1994.167. Shuey, D.L., Zucker, R.M., Elstein, K.H., and Rogers, J.M., Fetal anemia following maternal exposureto 5-fluorouracil in the rat, Teratology, 49, 311, 1994.168. Lau, C., Mole, M.L., Copeland, M.F., Rogers, J.M., Kavlock, R.J., Shuey, D.L., et al., Toward abiologically based dose-response model for developmental toxicity of 5-fluorouracil in the rat: acquisitionof experimental data, Toxicol. Sci., 59, 37, 2001.169. Setzer, R.W., Lau, C., Mole, M.L., Copeland, M.F., Rogers, J.M., and Kavlock, R.J., Toward abiologically based dose-response model for developmental toxicity of 5-fluorouracil in the rat: amathematical construct, Toxicol. Sci., 59, 49, 2001.© 2006 by Taylor & Francis Group, LLC

54 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION170. Minsker, D.H., MacDonald, J.S., Robertson, R.T., and Bokelman, D.L., Mevalonate supplementationin pregnant rats suppresses the teratogenicity of mevinolinic acid, an inhibitor of 3-hydroxy-3-methylglutaryl-coenzyme a reductase, Teratology, 28, 449, 1983.171. Brewer, L.M., Sheardown, S.A., and Brown, N.A., HMG-CoA reductase mRNA in the post-implantationrat embryo studied by in situ hybridization, Teratology, 47, 137, 1993.172. Krakoff, I., Brown, N., and Reichard, P., Inhibition of ribonucleoside diphosphate reductase byhydroxyurea, Cancer Res., 28, 1559, 1968.173. Chaube, S. and Murhph, M.L., Protective effect of deoxycytidylic acid (CdMP) on hydrosyureainducedmalformations in rats, Teratoloy, 7, 79, 1973.174. Herken, R., Merker, H.J., and Krowke, R., Investigation of the effect of hydroxyurea on the cell cycleand the development of necrosis in the embryonic CNS of mice, Teratology, 18, 103, 1978.175. Herken, R., The influence of deoxycytidine monophosphate (cDMP) on the cytotoxicity of hydroxyureain the embryonic spinal cord of the mouse, Teratology, 30, 83, 1984.176. Wawra, E., Microinjected deoxynucleotides for the study of chemical inhibition of DNA synthesis,Nucleic Acids Res., 16, 5249, 1988.177. DeSesso, J.M. and Goeringer, G.C., Ethoxyquin and nordihydroguaiaretic acid reduce hydroxyureadevelopmental toxicity, Reprod. Toxicol., 4, 267, 1990.178. DeSesso, J.M., Amelioration of teratogenesis. I. Modification of hydroxyurea-induced teratogenesisby the antioxidant propyl gallate, Teratology, 24, 19, 1981.179. Kessel, D. and Shurin, S.B., Transport of two non-metabolized nucleosides, deoxycytidine and cytosinearabinoside, in a sub-line of the L1210 murine leukemia, Biochim. Biophys. Acta, 163, 179, 1968.180. Kochhar, D.M., Penner, J.D., and McDay, J.A., Limb development in mouse embryos., Teratology,18, 71, 1978.181. Bhuyan, B.K., Scheidt, L.G., and Fraser, T.J., Cell cycle phase specificity of antitumor agents, CancerRes., 32, 398, 1972.182. Kochhar, D.M. and Agnish, N.D., “Chemical surgery” as an approach to study morphogenetic eventsin embryonic mouse limb, Develop. Biol., 61, 388, 1977.183. Zittoun, J., Marquet, J., David, J.C., Maniey, D., and Zittoun, R., A study of the mechanisms ofcytotoxicity of Ara-C on three human leukemic cell lines, Cancer Chemother. Pharmacol., 24, 251,1989.184. Chaube, S., Kreis, W., Uchida, K., and Murphy, M.L., The teratogenic effect of 1--D- arabinofuranosylcytosinein the rat, Biochem. Pharmacol., 17, 1213, 1968.185. Faustman, E.M. and Goodman, J.I., Alkylation of DNA in specific hepatic chromatin fractionsfollowing exposure to methylnitrosourea or dimethylnitrosamine, Toxicol. Appl. Pharmacol., 58, 379,1981.186. Ohgaki, H., Eibl, R.H., Wiestler, O.D., Yasargil, M.G., Newcomb, E.W., and Kleihues, P., p53mutations in nonastrocytic human brain tumors, Cancer Res., 51, 6202, 1991.187. Bochert, G., Platzek, T., Rahm, U., and Neubert, D., Embryotoxicity induced by alkylating agents:6. DNA adduct formation induced by methylnitrosourea in mouse embryos, Arch. Toxicol., 65, 390,1991.188. Kidney, J. and Faustman, E., Modulation of nitrosourea toxicity in rodent embryonic cells by O 6– benzylguanine,a depletor of O 6 -methylguanine-DNA methyltransferase, Toxicol. Appl. Pharmacol., 133,1, 1995.189. Francis, B.M., Rogers, J.M., Sulik, K.K., Alles, A.J., Elstein, K.H., Zucker, R.M., et al., Cyclophosphamideteratogenesis: evidence for compensatory responses to induced cellular toxicity, Teratology,42, 473, 1990.190. Gebhardt, D.O., The embryolethal and teratogenic effects of cyclophosphamide on mouse embryos,Teratology, 3, 273, 1970.191. Gibson, J.E. and Becker, B.A., The teratogenicity of cyclophosphamide in mice, Cancer Res., 28,475, 1968.192. Mirkes, P.E., Fantel, A.G., Greenaway, J.C., and Shepard, T.H., Teratogenicity of cyclophosphamidemetabolites: phosphoramide mustard, acrolein, and 4-ketocyclophosphamide in rat embryos culturedin vitro, Toxicol. Appl. Pharmacol., 58, 322, 1981.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 55193. Chernoff, N., Rogers, J.M., Alles, A.J., Zucker, R.M., Elstein, K.H., Massaro, E.J., et al., Cell cyclealterations and cell death in cyclophosphamide teratogenesis. Teratogen, Carinogen, Mutagen., 9, 199,1989.194. Auroux, M.R., Dulioust, E.J., Nawar, N.N., Yacoub, S.G., Mayaux, M.J., Schwartz, D., et al., Antimitoticdrugs in the male rat. Behavioral abnormalities in the second generation, J. Androl., 9, 153,1988.195. Hales, B.F., Crosman, K., and Robaire, B., Increased postimplantation loss and malformations amongthe F2 progeny of male rats chronically treated with cyclophosphamide, Teratology, 45, 671, 1992.196. Dulioust, E.J., Nawar, N.Y., Yacoub, S.G., Ebel, A.B., Kempf, E.H., and Auroux, M.R., Cyclophosphamidein the male rat: new pattern of anomalies in the third generation, J. Androl., 10, 296, 1989.197. Hoorn, A.J., Custer, L.L., Myhr, B.C., Brusick, D., Gossen, J., and Vijg, J., Detection of chemicalmutagens using Muta Mouse: A transgenic mouse model, Mutagen., 8, 7, 1993.198. Kohler, S.W., Provost, G.S., Fieck, A., Kretz, P.L., Bullock, W.O., Putman, D.L., et al., Analysis ofspontaneous and induced mutations in transgenic mice using a lambda ZAP/lacI shuttle vector,Environ. Mol. Mutagen., 18, 316, 1991.199. Kohler, S.W., Provost, G.S., Fieck, A., Kretz, P.L., Bullock, W.O., Sorge, J.A., et al., Spectra ofspontaneous and mutagen-induced mutations in the lacI gene in transgenic mice, Proc. Natl. Acad.Sci. USA, 88, 7958, 1991.200. Little, J.B., Cellular, molecular, and carcinogenic effects of radiation, Hematol. Oncol. Clin. NorthAm., 7, 337, 1993.201. Faustman-Watts, E., Greenaway, J.C., Namkung, M.J., Fantel, A.G., and Juchau, M.R., Teratogenicityin vitro of 2-acetylaminofluorene: role of biotransformation in the rat, Teratology, 27, 19, 1983.202. Faustman-Watts, E., Yang, H., Namkun, M., Greenaway, J., Fantel, A., and Juchau, M., Mutagenic,cytotoxic, and teratogenic effects of 2-acetylaminofluorene and reactive metabolites in vitro, Teratogen.Carcinog. Mutagen., 4, 273, 1984.203. Faustman-Watts, E.M., Greenaway, J.C., Namkung, M.J., Fantel, A.G., and Juchau, M.R., Teratogenicityin vitro of two deacetylated metabolites of N-hydroxy-2-acetylaminofluorene, Toxicol. Appl.Pharmacol., 76, 161, 1984.204. Faustman-Watts, E.M., Namkung, M.J., Greenaway, J.C., and Juchau, M.R., Analysis of metabolitesof 2-acetylaminofluorene generated in an embryo culture system. Relationship of biotransformationto teratogenicity in vitro, Biochem. Pharmacol., 34, 2953, 1985.205. Juchau, M.R., Giachelli, C.M., Fantel, A.G., Greenaway, J.C., Shepard, T.H., and Faustman-Watts,E.M., Effects of 3-methylcholanthrene and phenobarbital on the capacity of embryos to bioactivateteratogens during organogenesis, Toxicol. Appl. Pharmacol., 80, 137, 1985.206. Morgan, W., Non-targeted and delayed effects of exposure to ionizing radiation: I. Radiation-inducedgenomic instability and bystander effects in vitro, Rad. Res., 159, 567, 2003.207. Morgan, W., Non-targeted and delayed effects of exposure to ionizing radiation: II. Radiation-inducedgenomic instability and bystander effects in vivo, Rad. Res., 159, 2003.208. Fuscoe, J.C., Zimmerman, L.J., Fekete, A., Setzer, R.W., and Rossiter, B.J., Analysis of X-ray-inducedHPRT mutations in CHO cells: insertion and deletions, Mutat. Res., 269, 171, 1992.209. Grosovsky, A.J., de Boer, J.G., de Jong, P.J., Drobetsky, E.A., and Glickman, B.W., Base substitutions,frameshifts, and small deletions constitute ionizing radiation-induced point mutations in mammaliancells, Proc. Natl. Acad. Sci. USA, 85, 185, 1988.210. Hutchinson, F., Molecular biology of mutagenesis of mammalian cells by ionizing radiation, Semin.Cancer Biol., 4, 85, 1993.211. Lutze, L.H. and Winegar, R.A., pHAZE: a shuttle vector system for the detection and analysis ofionizing radiation-induced mutations, Mutat. Res., 245, 305, 1990.212. Skandalis, A., Grosovsky, A.J., Drobetsky, E.A., and Glickman, B.W., Investigation of the mutagenicspecificity of X-rays using a retroviral shuttle vector in CHO cells, Environ. Mol. Mutagen., 20, 271,1992.213. Brachman, D.G., Hallahan, D.E., Beckett, M.A., Yandell, D.W., and Weichselbaum, R.R., p53 genemutations and abnormal retinoblastoma protein in radiation-induced human sarcomas, Cancer Res.,51, 6393, 1991.214. Garte, S.J. and Burns, F.J., Oncogenes and radiation carcinogenesis, Environ. Health Perspect., 93,45, 1991.© 2006 by Taylor & Francis Group, LLC

56 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION215. Little, J.B., Cellular, molecular, and carcinogenic effects of radiation, Hematol. Oncol. Clin. NorthAm., 7, 337, 1993.216. Jensh, R.P. and Brent, R.L., Effects of prenatal X-irradiation on the 14th-18th days of gestation onpostnatal growth and development in the rat, Teratology, 38, 431, 1988.217. Kiyono, S., Seo, M., and Shibagaki, M., Effects of environmental enrichment upon maze performancein rats with microcephaly induced by prenatal X-irradiation, Jpn. J. Physiol., 31, 769, 1981.218. Norton, S., Behavioral changes in preweaning and adult rats exposed prenatally to low ionizingradiation, Toxicol. Appl. Pharmacol., 83, 240, 1986.219. Tamaki, Y. and Inouye, M., Avoidance of and anticipatory responses to shock in prenatally x-irradiatedrats, Physiol. Behav., 22, 701, 1979.220. Rudnicki, M. and McBurney, M., Cell culture methods and induction of differentiation of embryonalcarcinoma cell lines, in Teratocarcinomas and Embryonic Stems Cells: A Practical Approach, Robertson,E. J., Ed., IRL Press, Oxford, 1987, p. 19.221. Boulter, C.A. and Wagner, E.F., The effects of v-src expression on the differentiation of embryonalcarcinoma cells, Oncogene, 2, 207, 1988.222. de Groot, R.P., Kruyt, F.A., van der Saag, P.T., and Kruijer, W., Ectopic expression of c-jun leads todifferentiation of P19 embryonal carcinoma cells, Embo. J., 9, 1831, 1990.223. Dutt, K., Scott, M., Sternberg, P.P., Linser, P.J., and Srinivasan, A., Transdifferentiation of adult humanpigment epithelium into retinal cells by transfection with an activated H-ras proto-oncogene, DNACell Biol., 12, 667, 1993.224. Propst, F., Rosenberg, M.P., Cork, L.C., Kovatch, R.M., Rauch, S., Westphal, H., et al., Neuropathologicalchanges in transgenic mice carrying copies of a transcriptionally activated Mos protooncogene,Proc. Natl. Acad. Sci. USA, 87, 9703, 1990.225. Kreidberg, J.A., Sariola, H., Loring, J.M., Maeda, M., Pelletier, J., Housman, D., et al., WT-1 isrequired for early kidney development, Cell, 74, 679, 1993.226. Collins, M. and Mao, G., Teratology of retinoids, Annu. Rev. Pharmacol. Toxicol., 39, 399, 1999.227. Pignatello, M.A., Kauffman, F.C., and Levin, A.A., Multiple factors contribute to the toxicity of thearomatic retinoid TTNPB (Ro 13-7410): interactions with the retinoic acid receptors, Toxicol. Appl.Pharmacol., 159, 109, 1999.228. Tzimas, G., Thiel, R., Chahoud, I., and Nau, H., The area under the concentration-time curve of alltrans-retinoicacid is the most suitable pharmacokinetic correlate to the embryotoxicity of this retinoidin the rat, Toxicol. Appl. Pharmacol., 143, 436, 1997.229. Nau, H., Chaoud, I., Dencker, L., Lammer, E., and Scott, W., Teratogenicity of vitamin A and retinoids,in Health and Diseases, Blomhoff, R., Ed., Marcel Dekker, New York, 1994.230. Kraft, J.C., Shepard, T., and Juchau, M.R., Tissue levels of retinoids in human embryos/fetuses,Reprod. Toxicol., 7, 11, 1993.231. Chambon, P., A decade of molecular biology of retinoic acid receptors, FASEB J., 10, 940, 1996.232. Sucov, H.M., Izpisua-Belmonte, J.C., Ganan, Y., and Evans, R.M., Mouse embryos lacking RXR alphaare resistant to retinoic-acid-induced limb defects, Development, 121, 3997, 1995.233. Lohnes, D., Kastner, P., Dierich, A., Mark, M., LeMeur, M., and Chambon, P., Function of retinoicacid receptor gamma in the mouse, Cell, 73, 643, 1993.234. Iulianella, A. and Lohnes, D., Contribution of retinoic acid receptor gamma to retinoid-inducedcraniofacial and axial defects, Dev. Dyn., 209, 92, 1997.235. Kochar, D.K. and Kumawat, B.L., Cerebral malaria or Plasmodium falciparum malaria with hypoglycaemia,Lancet, 347, 1549, 1996.236. Marshall, H., Morrison, A., Studer, M., Popperl, H., and Krumlauf, R., Retinoids and Hox genes,FASEB J., 10, 969, 1996.237. Chang, C., Smith, D.R., Prasad, V.S., Sidman, C.L., Nebert, D.W., and Puga, A., Ten nucleotidedifferences, five of which cause amino acid changes, are associated with the Ah receptor locuspolymorphism of C57BL/6 and DBA/2 mice, Pharmacogenetics, 3, 312, 1993.238. Lambert, G.H. and Nebert, D.W., Genetically mediated induction of drug-metabolizing enzymesassociated with congenital defects in the mouse, Teratology, 16, 147, 1977.239. Poland, A. and Glover, E., 2,3,7,8,-Tetrachlorodibenzo-p-dioxin: segregation of toxocity with the Ahlocus, E. Mol. Pharmacol., 17, 86, 1980.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 57240. Abbott, B., Perdew, G., and Birnbaum, L., Ah receptor in embryonic mouse palate and effects ofTCDD on receptor expression, Toxicol. Appl. Pharmacol., 126, 16, 1994.241. Abbott, B.D. and Birnbaum, L.S., TCDD alters medial epithelial cell differentiation during palatogenesis,Toxicol. Appl. Pharmacol., 99, 276, 1989c.242. Abbott, B.D. and Birnbaum, L.S., Retinoic acid-induced alterations in the expression of growth factorsin embryonic mouse palatal shelves, Teratology, 42, 597, 1990.243. Eberwine, J., Yeh, H., Miyashiro, K., Cao, Y., Nair, S., Finnell, R., et al., Analysis of gene expressionin single live neurons, Proc. Natl. Acad. Sci. USA, 89, 3010, 1992.244. Yu, Y., Zhang, C., Zhou, G., Wu, S., Qu, X., Wei, H., et al., Gene expression profiling in human fetalliver and identification of tissue- and developmental-stage-specific genes through compiled expressionprofiles and efficient cloning of full-length cDNAs, Genome Res., 11, 1392, 2001.245. Master, S.R., Hartman, J.L., D’Cruz, C.M., Moody, S.E., Keiper, E.A., Ha, S.I., et al., Functionalmicroarray analysis of mammary organogenesis reveals a developmental role in adaptive thermogenesis,Mol. Endocrinol., 16, 1185, 2002.246. Stuart, R.O., Bush, K.T., and Nigam, S.K., Changes in global gene expression patterns duringdevelopment and maturation of the rat kidney, Proc. Natl. Acad. Sci. USA, 98, 5649, 2001.247. Ko, M.S., Kitchen, J.R., Wang, X., Threat, T.A., Hasegawa, A., Sun, T., et al., Large-scale cDNAanalysis reveals phased gene expression patterns during preimplantation mouse development, Development,127, 1737, 2000.248. Shier, W.T., Why study mechanisms of cell death?, Methods Achiev. Exp. Pathol., 13, 1, 1988.249. Schwartz, L.M., The role of cell death genes during development, Bioessays, 13, 389, 1991.250. Nijhawan, D., Honarpour, N., and Wang, X., Apoptosis in neuronal development and disease, Annu.Rev. Neurosci., 23, 73, 2000.251. Oppenheim, R., Cell death during development of the nervous system, Annu. Rev. Neurosci., 14, 453, 1991.252. Cowan, W.M., Fawcett, J.W., O’Leary, D.D., and Stanfield, B. B., Regressive events in neurogenesis,Science, 225, 1258, 1984.253. Menkes, B., Sandor, S., and Ilies, A., Why study mechanisms of cell death?, Methods Achiev. Exp.Pathol., 13, 1, 1970.254. Coelho, C.N., Upholt, W.B., and Kosher, R.A., The expression pattern of the chicken homeoboxcontaininggene GHox-7 in developing polydactylous limb buds suggests its involvement in apicalectodermal ridge-directed outgrowth of limb mesoderm and in programmed cell death, Differentiation,52, 129, 1993.255. Melino, G., Stephanou, A., Annicchiarico, P.M., Knight, R.A., Finazzi, A.A., and Lightman, S.L.,Modulation of IGF-2 expression during growth and differentiation of human neuroblastoma cells:Retinoic acid may induce IGF-2, Neurosci. Lett., 151, 187, 1993.256. Schweichel, J., The influence of oral vitamin A doses on interdigital necrosis in the limb bud of therat, Teratology, 4, 502, 1971.257. Sulik, K.K., Johnston, M.C., Smiley, S.J., Speight, H.S., and Jarvis, B.E., Mandibulofacial dysostosis(Treacher Collins syndrome): a new proposal for its pathogenesis, Am. J. Med. Genet., 27, 359, 1987.258. Sulik, K.K. and Dehart, D.B., Retinoic-acid-induced limb malformations resulting from apical ectodermalridge cell death, Teratology, 37, 527, 1988.259. Alles, A.J. and Sulik, K.K., Retinoic-acid-induced limb-reduction defects: perturbation of zones ofprogrammed cell death as a pathogenetic mechanism, Teratology, 40, 163, 1989.260. Sulik, K.K., Smiley, S.J., Turvey, T.A., Speight, H.S., and Johnston, M.C., Pathogenesis of cleft palatein Treacher Collins, Nager, and Miller syndromes, Cleft Palate J., 26, 209, 1989.261. Clarren, S.K. and Smith, D.W., The fetal alcohol syndrome, New Eng. J. Med., 298, 1063, 1978.262. Miller, M.W., Development of the Central Nervous System: Effects of Alcohol and Opiates, Wiley-Liss, New York, 1992.263. Rodier, P., Developing brain as a target of toxicity, Environ. Health Perspect., 103, 73, 1995.264. Napper, R.M.A. and West, J.R., Permanent neuronal cell loss in the cerebellum of rats exposed tocontinuous low blood alcohol levels during the brain growth spurt: A stereological investigation, J.Comp. Neurol., 362, 283, 1995.265. Miller, M.W. and Potempa, G., Numbers of neurons and glia in mature rat somatosensory cortex:Effects of prenatal exposure to ethanol, J. Comp. Neur., 293, 92, 1990.© 2006 by Taylor & Francis Group, LLC

58 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION266. Miller, M.W. and Kuhn, P.E., Cell cycle kinetics in fetal rat cerebral cortex: effects of prenatal treatmentwith ethanol assessed by a cumulative labeling technique with flow cytometry, Alcohol Clin. Exp.Res., 19, 233, 1995.267. West, J.R., Hamre, K.M., and Cassell, M.D., Effects of ethanol exposure during the third trimesterequivalent on neuron number in rat hippocampus and dentate gyrus, Alcohol Clin. Exp. Res., 10, 190,1986.268. Coles, C.D., Prenatal alcohol exposure and human development, in Development of the CentralNervous System: Effects of Alcohol and Opiates, Miller, M.W., Ed., Wiley-Liss, Inc., New York, 1992,p. 9.269. Allebeck, P. and Olsen, J., Alcohol and fetal damage, Alcohol Clin. Exp. Res., 22, 329S, 1998.270. Samson, H.H., Microcephaly and fetal alcohol syndrome: human and animal studies, in Alcohol andBrain Dev., West, J. R., Ed.,1986, p. 167.271. Duffell, S.J., Soames, A.R., and Gunby, S., Morphometric analysis of the developing rat brain, Toxicol.Pathol., 28, 157, 2000.272. Benes, F.M., About assumptions in estimation of density of neurons and glial cells — Reply, Biolo.Psych., 51, 842, 2002.273. Benes, F.M. and Lange, N., Benes and Lange respond: reconciling theory and practice in cell counting,Neurosci., 24, 378, 2001.274. Benes, F.M. and Lange, N., Two-dimensional versus three-dimensional cell counting: a practicalperspective, Neurosci., 24, 11, 2001.275. von Bartheld, C.S., Counting particles in tissue sections: Choices of methods and importance ofcalibration to minimize biases, Histol. Histopath., 17, 639, 2002.276. Skoglund, T., Pascher, R., and Berthold, C., Aspects of the quantitative analysis of neurons in thecerebral cortex, J. Neurosci. Meth., 70, 201, 1996.277. Mooney, S.M., Napper, R.M.A., and West, J.R., Long-term effect of postnatal alcohol exposure onthe number of cells in the neocortex of the rat: A stereological study, Alcohol Clin. Exp. Res., 20,615, 1996.278. Miller, M.W., Effect of early exposure to ethanol on the protein and DNA contents of specific brainregions in the rat, Brain Res., 734, 286, 1996.279. Maier, S.E., Chen, W.-J.A., Miller, J.A., and West, J.R., Fetal alcohol exposure and temporal vulnerability:Regional differences in alcohol-induced microcephaly as a function of the timing of bingelikealcohol exposure during rat brain development, Alcohol Clin. Exp. Res., 21, 1418, 1997.280. Pennington, S.N., Taylor, W.A., Cowan, D.H., and Kalmus, G.W., A single dose of ethanol suppressesrat embryo development in vivo, Alcohol Clin. Exp. Res., 8, 326, 1984.281. Miller, M.W., Effects of alcohol on the generation and migration of cerebral cortical neurons, Science,233, 1308, 1986.282. Dreosti, I.E., Ballard, J., Belling, G.B., Record, I.R., Manuel, S.J., and Hetzel, B.S., The effect ofethanol and actealdehyde on DNA synthesis in growing cells and on fetal development in the rat,Alcohol Clin. Exp. Res., 5, 357, 1981.283. Miller, M.W., Effects of prenatal exposure to ethanol on neocortical development: II. Cell proliferationin the ventricular and subventricular zones of the rat, J. Comp. Neurol., 287, 326, 1989.284. Guizzetti, M. and Costa, L.G., Inhibition of muscarinic receptor-stimulated glial cell proliferation byethanol, J. Neurochem., 67, 2236, 1996.285. Miller, M.W., Effects of prenatal exposure to ethanol on cell proliferation and neuronal migration, inDevelopment of the Central Nervous System: Effects of Alcohol and Opiates, Miller, M. W., Ed.,Wiley-Liss, New York, 1992, p. 47.286. Miller, M.W., Effect of pre- or postnatal exposure to ethanol on the total number of neurons in theprincipal sensory nucleus of the trigeminal nerve: Cell proliferation and neuronal death, Alcohol Clin.Exp. Res., 19, 1359, 1995.287. Nowakowski, R., Lewin, S., and Miller, M.W., Bromodeoxyuridine immunohistochemical determinationof the lengths of the cell cycle and the DNA-synthetic phase for an anatomically definedpopulation, J. Neurocyt., 18, 311, 1989.288. Miller, M.W. and Muller, S., Structure and histogenesis of the principle sensory nucleus of thetrigeminal nerve: Effects of prenatal exposure to ethanol, J. Comp. Neurol., 282, 570, 1989.© 2006 by Taylor & Francis Group, LLC

EXPERIMENTAL APPROACHES TO EVALUATE MECHANISMS OF DEVELOPMENTAL TOXICITY 59289. Caviness, V., Takahashi, T., and Nowakowski, R., Numbers, time and neocortical neuronogenesis: ageneral developmental and evolutionary model., Trends Neurosci., 18, 379, 1995.290. Takahashi, T., Nowakowski, R., and Caviness, V., The leaving or Q fraction of the murine cerebralproliferative epithelium: A general model of neocortical neuronogenesis., J. Neurosci., 16, 6183, 1996.291. Climent, E., Pascual, M., Renau-Piqueras, J., and Guerri, C., Ethanol exposure enhances cell deathin the developing cerebral cortex: Role of brain-derived neurotrophic factor and its signaling pathways,J. Neurosci. Res., 68, 213, 2002.292. Ferrer, I., Bernet, E., Soriano, E., del Rio, T., and Fonseca, M., Naturally occurring cell death in thecerebral cortex of the rat and removal of dead cells by transitory phagocytes, Neuroscience, 39, 451,1990.293. Ikonomidou, C., Ethanol-induced apoptotic neurodegeneration and fetal alcohol syndrome, Science,287, 1056, 2000.294. Olney, J.W., Farber, N., Wozniak, D.F., Jevtovic-Todorovic, V., and Ikonomidou, C., Environmentalagents that have the potential to trigger massive apoptotic neurodegeneration in the developing brain,Environ. Health Perspect., 108, 383, 2000.295. Light, K.E., Kane, C.J., Pierce, D.R., Jenkins, D., Ge, Y., Brown, G., et al., Intragastric intubation:Important aspects of the model for administration of ethanol to rat pups during the postnatal period,Alcohol Clin. Exp. Res., 22, 1600, 1998.296. Gohlke, J.M., Griffith, W.C., and Faustman, E.M., A mechanism based model for evaluating neocorticaldevelopment: Applications for understanding ethanol neurodevelopmental toxicity, Teratology, 65,335, 2002.297 McCartney-Francis, N. L., Mizel, D. E., Frazier-Jessen, M., Kulkarni, A. B., McCarthy, J. B., andWahl, S. M., Lacrimal gland inflammation is responsible for ocular pathology in TGF-beta 1 nullmice, Am. J. Pathol., 151, 1281, 1997.298. Dunker, N. and Krieglstein, K., Targeted mutations of transfroming growth factor-beta genese revealimportant roles in mouse development and adult homeostasis, Eur. J. Biochem., 267(24), 6982, 2000.299. Mikic, B., Van der Meulen, M. C., Kingsley, D. M., and Carter, D. R., Mechanical and geometricchanges in the growing femora of BMP-5 deficient mice, Bone, 18, 601, 1996.300. Dudley, A. T. and Robertson, E. J., Overlapping expression domains of bone morphogenetic proteinfamily members potentially account for limited tissue defects in BMP7 deficient embryos, Dev. Dyn.,208, 349, 1997.301. Goodrich, L. V., Milenkovic, L., Higgins, K. M., and Scott, M. P., Altered neural cell fates andmedulloblastoma in mouse patched mutants, Science, 277, 1109, 1997.302. Hahn, H., Wojnowski, L., Zimmer, A. M., Hall, J., Miller, G., and Zimmer, A., Rhabdomyosarcomasand radiation hypersensitivity in a mouse model of Gorlin syndrome, Nat. Med., 4, 619, 1998.303. Lydon, J. P., DeMayo, F. J., Funk, C. R., Mani, S. K., Hughes, A. R., Montgomery, C. A., Jr., et al.,Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities, Genes Dev., 9,2266, 1995.304. Lydon, J.P., DeMayo, F.J., Conneely, O.M., and O’Malley, B.W., Reproductive phenotypes of theprogesterone receptor null mutant mouse, J. Steroid Biochem. Mol. Biol., 56(1–6 Spec No), 67, 1996.305. Wendling, O., Ghyselinck, N.B., Chambon, P., and Mark, M., Roles of retinoic acid receptors in earlyembryonic morphogenesis and hindbrain patterning, Development, 128(11), 2031–2038, 2001.306. Moriguchi, T., Motohashi, H., Hosoya, et al., Distinct response to dioxin in an arylhydrocarbonreceptor (AHR)-humanized mouse, Proc. Natl. Acad. Sci. U.S.A., 100(10), 5652–5657, 2003.307. Kassim, N. M., McDonald, S. W., Reid, O., Bennett, N. K., Gilmore, D. P., and Payne, A. P., Theeffects of pre- and postnatal exposure to the nonsteroidal antiandrogen flutamide on testis descent andmorphology in the albino Swiss rat, J. Anat., 190 (Pt 4): 577, 1997.308. Cunha, G. R., Forsberg, J. G., Golden, R., Haney, A., Iguchi, T., Newbold, R., et al., New approachesfor estimating risk from exposure to diethylstilbestrol, Environ. Health Perspect., 107 Suppl 4, 625,1999.309. Abbott, B.D., Perdew, G.H., Buckalew, A.R., and Birnbaum, L.S., Interactive regulation of Ah andglucocorticoid receptors in the synergistic induction of cleft palate by 2,3,7,8-tetrachlorodibenzo-pdioxinand hydrocortisone, Toxicol. Appl. Pharmacol., 128(1), 138–150, 1994.310. Chazaud, C., Chambon, P., and Dolle, P., Retinoic acid is required in the mouse embryo for left-rightasymmetry determination and heart morphogenesis, Development, 126, 2589, 1999.© 2006 by Taylor & Francis Group, LLC

60 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION311. Kochhar, D. M., Jiang, H., Penner, J. D., Johnson, A. T., and Chandraratna, R. A., The use of a retinoidreceptor antagonist in a new model to study vitamin A-dependent developmental events, Int. J. Dev.Biol., 42, 601, 1998.312. Elmazar, M., Ruhl, R., Reichert, U., Shroot, B., and Nau, H., RARalpha-mediated teratogenicity inmice is potentiated by an RXR agonist and reduced by an RAR antagonist: Dissection of retinoidreceptor-induced pathways, Toxicol. Appl. Pharmacol., 146, 21, 1997.313. Brandsma, A. E., Tibboel, D., Vulto, I. M., de Vijlder, J. J., Ten Have-Opbroek, A. A., and Wiersinga,W. M., Inhibition of T3-receptor binding by Nitrofen, Biochim. Biophys. Acta, 1201, 266, 1994.314. Spence, S., Anderson, C., Cukierski, M., and Patrick, D., Teratogenic effects of the endothelin receptorantagonist L-753,037 in the rat, Reprod. Toxicol., 13, 15, 1999.315. Treinen, K. A., Louden, C., Dennis, M. J., and Wier, P. J., Developmental toxicity and toxicokineticsof two endothelin receptor antagonists in rats and rabbits, Teratology, 59, 51, 1999.316. Gershon, M. D., Lessons from genetically engineered animal models. II. Disorders of enteric neuronaldevelopment: insights from transgenic mice, Am. J. Physiol., 277(2 Pt 1), G262, 1999.317. Webster, W. S., Brown-Woodman, P. D., Snow, M. D., and Danielsson, B. R., Teratogenic potentialof almokalant, dofetilide, and d-sotalol: drugs with potassium channel blocking activity, Teratology,53, 168, 1996.318. Wellfelt, K., Skold, A. C., Wallin, A., and Danielsson, B. R., Teratogenicity of the class III antiarrhythmicdrug almokalant. Role of hypoxia and reactive oxygen species, Reprod. Toxicol., 13, 93, 1999.319. Clarke, C., Clarke, K., Muneyyirci, J., Azmitia, E., and Whitaker-Azmitia, P. M., Prenatal cocainedelays astroglial maturation: immunodensitometry shows increased markers of immaturity (vimentinand GAP-43) and decreased proliferation and production of the growth factor S-100, Brain Res. Dev.Brain Res., 91, 268, 1996.320. Brault, V., Moore, R., Kutsch, S., et al., Inactivation of the beta-catenin gene by Wnt1-Cre-mediateddeletion results in dramatic brain malformation and failure of craniofacial development, Development,128(8), 1253–1264, 2001.321. Larsson, J., Goumans, M.J., Sjostrand, L.J., et. al., Abnormal angiogenesis but intact hematopoieticpotential in TGF-beta type I receptor-deficient mice, Embo. J., 20(7), 1663–1673, 2001.322. Miettinen, P.J., Chin, J.R., Shum, L., et al., Epidermal growth factor receptor function is necessaryfor normal craniofacial development and palate closure, Nat. Genet., 22(1), 69–73, 1999.323. Dudas, M., Sridurongrit, S., Nagy, A., Okazaki, K., and Kaartinen, V., Craniofacial defects in micelacking BMP type I receptor Alk2 in neural crest cells, Mech. Dev., 121(2), 173–182, 2004.324. Hahn, H., Wojnowski, L., Specht, K., et al., Patched target Igf2 is indispensable for the formation ofmedulloblastoma and rhabdomyosarcoma, J. Biol. Chem., 275(37), 28341–28344, 2000.325. Peters, J.M., Narotsky, M.G., Elizondo, G., et al., Amelioration of TCDD-induced teratogenesis inaryl hydrocarbon receptor (AhR)-null mice, Toxicol. Sci., 47(1), 86–92, 1999.326. Tanaka, M., Ohtani-Kaneko, R., Yokosuka, M., Watanabe, C., Low-dose perinatal diethylstilbestrolexposure affected behaviors and hypothalamic estrogen receptor-alpha-positive cells in the mouse,Neurotoxicol. Teratol., 26(2), 261–269, 2004.327. Abbott, B.D., Schmid, J.E., Brown, J.G., et al., RT-PCR quantification of AHR, ARNT, GR, andCYP1A1 mRNA in craniofacial tissues of embryonic mice exposed to 2,3,7,8-tetrachlorodibenzo-pdioxinand hydrocortisone, Toxicol. Sci., 47(1), 76–85, 1999.328. Elamazar, M.M. and Nau, H., Potentiation of the teratogenic effects induced by coadministration ofretinoic acid or phytanic acid/phytol with synthetic retinoid receptor ligands, Arch. Toxicol., 78(11),660–668, 2004.329. Nagase, T., Nagase, M., Osumi, N., et al., Craniofacial anomalies of the cultured mouse embryoinduced by inhibition of sonic hedgehog signaling: an animal model of holoprosencephaly, J. Craniofac.Surg., 16(1), 80–88, 2005.330. Gershon, M.D., Endothelin and the development of the enteric nervious system, Clin. Exp. Pharmacol.Physiol., 26(12), 985–988, 1999.331. Brand, M., Kempf, H., Paul, M., et al., Expression of endothelins in human cardiogenesis, J. Mol.Med., 80(11), 715–723, 2002.© 2006 by Taylor & Francis Group, LLC

CHAPTER 3Pathogenesis of Abnormal DevelopmentLynda B. Fawcett and Robert L. BrentCONTENTSI. Introduction ..........................................................................................................................61II. Manifestations of Developmental Toxicity..........................................................................62A. Structural Anomalies: Malformations, Deformations, and Disruptions .....................62B. Multiple Defects: Syndromes, Sequences, and Associations .....................................63III. Factors That Influence the Pathogenesis of Abnormal Development.................................63A. Stage of Development .................................................................................................63B. Tissue Specificity.........................................................................................................64C. Influence of Dose.........................................................................................................65IV. Pathogenesis at the Cellular Level ......................................................................................67A. Cell Death ....................................................................................................................68B. Cell Proliferation .........................................................................................................70C. Alterations in Cell-Signaling and Cell–Cell Interactions ...........................................71D. Cell Migration and Differentiation..............................................................................74V. Examples of Abnormal Pathogenesis ..................................................................................75A. Defects with Different Underlying Pathogenesis but Similar Morphology ...............761. Cleft Palate.............................................................................................................762. Neural Tube Defects..............................................................................................77B. Abnormalities with Similar Pathogenesis but Different Morphology........................78VI. Overview and Perspective....................................................................................................80References ........................................................................................................................................81I. INTRODUCTIONVertebrate development proceeds in a sequence of carefully timed events that progress from thecellular level to the formation of tissues, organ systems, and morphologic structures. Alterationsat any level may permanently alter later developmental processes. The initial adverse effects of anenvironmental agent on a developing organism most frequently occur at the molecular level. Sucheffects constitute the mechanisms by which the agent alters development. Subsequent events thatinclude deviations at the cellular, tissue, and organism levels constitute the pathogenesis of abnormaldevelopment. It is this sequence of abnormal developmental processes, i.e., events downstream ofthe mechanism, that is the focus of this chapter. Unfortunately it is not possible within the scope61© 2006 by Taylor & Francis Group, LLC

62 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof this chapter to include all that is known or suspected about the pathogenesis of the myriad agentsthat have been classified as developmental toxicants. Moreover, the pathogenesis involved in the genesisof many congenital anomalies has not yet been determined. Therefore, we instead cover the pathogenesisof selected toxicants and selected well-defined malformations as examples to illustrate the majorconsiderations involved in the study of the pathogenesis of abnormal development.II. MANIFESTATIONS OF DEVELOPMENTAL TOXICITYVertebrate development is hierarchical, composed of a series of events that are both temporally andspatially regulated. Perturbations in any of the normal developmental processes can potentially altersubsequent growth and morphogenesis and result in congenital malformations. Adverse fetal outcomesthat result from exposure to developmental toxicants are not limited to congenital malformationsbut also manifest as functional deficits, growth retardation, and death. Functional deficitsrefer to physiological or system deficits (e.g., immunological and hormonal) and, more commonly,to neurological and behavioral deficits. The potential adverse fetal outcomes are not mutuallyexclusive. In fact, they are frequently associated. For instance, anatomical defects are often accompaniedby growth retardation. 1The etiology and discrete pathogenesis responsible for the majority of congenital defects in humansis unknown. 2–4 For the most part, our understanding of the pathogenesis of congenital anomalies isbased on studies that use as models animals exposed to developmentally toxic agents, and on studiesof spontaneous mutants and transgenic mice. These studies have revealed that although there are someconsistent features associated with the induction of a given anomaly, the underlying pathogenesis maybe quite different. Thus, a particular type of anomaly, such as cleft palate, may be caused by diverseagents and mechanisms. While the early pathogenesis of the defect may differ, the pathogenic pathwaysultimately converge. Likewise, a particular agent or resulting pathogenesis may result in very differentoutcomes depending on factors such as dose, embryonic age, and genetic susceptibility. Because ofthis complexity, it is extremely difficult to ascertain the underlying pathogenesis of a given developmentalanomaly based on the final outcome, even to the extent of determining whether it is genetic orenvironmentally induced. Rather one must look earlier in the developmental process to determine whichtissues were initially affected and what the effects or consequences on the target tissue or cell populationwere. This knowledge, coupled with an understanding of the normal embryological processes thatwould govern later recovery and development of the structure, will provide insight into the overallpathogenesis leading to the defect.A. Structural Anomalies: Malformations, Deformations, and DisruptionsDysmorphology resulting from developmental toxicants can be described as resulting from processesof malformation, deformation, or disruption. 5 Malformations occur when the normal developmentalprocess is altered such that a given structure cannot form or forms improperly. In thiscase, the error is intrinsic to the morphogenetic process itself. Generally speaking, the embryo ismost susceptible to the adverse effects of toxicants that produce malformations during organogenesis,when cells are rapidly proliferating and differentiating, and extensive patterning is takingplace. Malformations such as limb defects, whether induced by chemical means during earlyorganogenesis or due to genetic defect, are usually (but not always) bilateral. 6–8Unlike malformations, deformations and disruptions involve alterations to already existingstructures. In these cases, the intrinsic developmental processes had proceeded normally, andextrinsic factors not related to the developmental processes lead to the dysmorphology observedat term. Deformations usually result from mechanical factors, such as uterine constraint or amnioticbands, and include alterations in body shape, form, or position. Well-known examples are congenitaldislocation of the hip, plagiocephaly, clubfoot, and some facial anomalies. Disruptions differ from© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 63deformations in that they are usually more severe, often involving extensive destruction of tissue,disruption of tissue function, and/or prevention of subsequent tissue or organ formation. Likedeformations, disruptions can arise from mechanical factors impinging on the fetal environment orfetal tissue directly, such as might occur in amniotic band syndrome, fetal vascular occlusion, orplacental emboli. Additionally, disruption can result from alterations in fetal physiology, especiallythose resulting in localized or general hypoxia and vascular insults. 9 It is important to rememberthat these classifications are not mutually exclusive. This is especially true when multiple structuraldefects are present. Thus, disruptions to another tissue may be caused by mechanical factors relatedto a previous defect, such as occurs with hydronephrosis. Excessive buildup of urine due to ureterblockage leads to oligohydramnios, damage to the kidneys, and pulmonary hypoplasia (reviewedby Coplen 10 ).B. Multiple Defects: Syndromes, Sequences, and AssociationsThe majority of developmental toxicants do not produce only a single anomaly. The effects of atoxicant on development will typically manifest as a distinguishable pattern of multiple anomaliesthat vary with the timing of exposure and the differing tissue susceptibilities of the embryo. Thereare some exceptions to this general rule, such as glucocorticoid-induced cleft palate in mice. 11,12Such outcomes can occur in the absence of other malformations and only the frequency of thedefect, and not the defect itself, changes with timing of administration. However, growth retardationusually accompanies the experimentally induced malformation. When multiple anomalies areobserved, it is often helpful to classify them according to whether the multiple defects arise intissues independently as a result of a similar pathogenesis, are secondary to a single anomaly orpathogenic mechanism, or represent other associations. A recognizable pattern of multiple structuralanomalies is generally referred to as a syndrome, a sequence, or an association.Multiple anomalies are considered a syndrome when they result from the same or similarunderlying mechanisms and thus are causally related. 5,13 Syndromes can result from genetic mutationor may be teratogen induced and thus usually have a known etiology. Examples of geneticsyndromes in humans include Down’s, Meckel, and Prader-Willi syndromes. Recognized teratogeninducedsyndromes include thalidomide syndrome, fetal alcohol syndrome, retinoic acid embryopathy,and diphenylhydantoin syndrome. A sequence represents a pattern of defects that are secondaryto a single defect, with a discrete known or suspected pathogenesis but often with an unknownetiology. The pathogenic events that result downstream from amniotic bands or from fetal hydronephrosiswould represent a sequence.A pattern of malformations that have not yet been classified as either a syndrome or a sequence,but that occur in a nonrandom fashion, is referred to as an association. 14 As the etiology andpathogenesis become more understood, associations are reclassified into one of the other twocategories. Examples of associations in the human include the VATER association (vertebraldefects–anal atresia–tracheoesophageal fistula–esophageal atresia–radial and renal dysplasia).When cardiac and limb defects are also present it is often termed the VACTERAL association. 14Because these defects tend to occur as a pattern of anomalies, there is a high likelihood that thereis a common, although as yet undiscovered, underlying mechanism and pathogenesis.A. Stage of DevelopmentIII. FACTORS THAT INFLUENCE THE PATHOGENESISOF ABNORMAL DEVELOPMENTDevelopmental toxicity induced by environmental agents usually results in a spectrum of morphologicanomalies or intrauterine death that varies in incidence depending on stage of exposure and© 2006 by Taylor & Francis Group, LLC

64 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONdose. The developmental period at which exposure occurs will determine which structures are mostsusceptible to the deleterious effects of the agent and to what extent the embryo can repair thedamage. The period of sensitivity may be narrow or broad, depending on the environmental agentand the malformation in question. Limb defects produced by thalidomide have a very short periodof susceptibility. Moreover, the type of limb defect that can be induced changes rapidly as embryonicage increases from 22 to 36 days postconception. 15,16 The mechanism of thalidomide’s action onthe cells, while still undetermined, very likely remains the same. Thus, the differing sensitivitiespresumably result from changes in the target cell population itself over time. Such changes couldinclude alterations in proliferation rate, stage in cell cycle, commitment to differentiation, orexpression of specific receptors. Other teratogens, and particularly those that cause a more nonspecificcytotoxicity, may have much longer periods of susceptibility. Radiation-induced microcephaly,which can be induced far into the fetal period, is an example. 17,18Numerous studies and observations have shown us that during the first period of embryonicdevelopment, from fertilization through the early postimplantation stage, the embryo is mostsensitive to the embryolethal effects of drugs and chemicals. Surviving embryos have malformationrates similar to the controls, not because malformations cannot be produced at this stage, but becausesignificant cell loss or chromosome abnormalities at these stages have a high likelihood of killingthe embryo. Because of the omnipotentiality of early embryonic cells, surviving embryos have amuch greater ability to have normal developmental potential. This trend of marked resistance tothe malforming consequences of teratogens has been termed the “all-or-none phenomenon.” Utilizingx-irradiation as the experimental teratogen, Wilson and Brent demonstrated that the all-ornonephenomenon disappears over a period of a few hours in the rat during early organogenesis. 19The term “all-or-none phenomenon” has been misinterpreted by some investigators to indicatethat malformations cannot be produced at this stage. On the contrary, it is likely that certain drugs,chemicals, and other insults during this stage of development can result in malformed offspring. 20,21But the nature of embryonic development at this stage will still reflect the basic characteristic ofthe all-or-none phenomenon, which is a propensity for embryolethality, rather than survivingmalformed embryos.The period of organogenesis encompasses rapid cell division, extensive patterning, and tissuedifferentiation. Thus, it is not surprising that this period represents the stage when the embryo ismost susceptible to dysmorphogenesis from developmental toxicants and that the majority ofcongenital malformations can be produced by alterations in developmental processes during thisperiod. Exceptions include malformations of the genitourinary system, palate, and brain, as wellas abnormalities produced by deformations and disruptions.During the fetal period, teratogenic agents may decrease the cell population by producing celldeath, inhibiting cell division, or interfering with cell differentiation. The resulting effects, such as celldepletion or functional abnormalities, may not be readily apparent at birth. Major structural anomaliescan be produced during the fetal period, and they usually result from disruptions or deformations dueto factors such as uterine constraint, hypoxia, or the action of vasoactive substances such as cocaine.Disruptions induced at this stage may involve limbs, major organs, including the brain, or other systems.Severe growth retardation in the whole embryo or fetus may also result in permanent deleterious effectson many organs or tissues, especially the brain and reproductive organs.B. Tissue SpecificityThe pathogenesis associated with a developmental toxicant depends not only on the nature of theteratogen itself but also on the susceptibility of the embryonic tissues. In general, rapidly dividingcell populations will be the most sensitive to developmental toxicants that result in cell death orreduced cell proliferation. This is not because other cell populations receive no exposure, butbecause depletion of the cell population in areas of rapid cell division is more likely to interfere© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 65with the normal developmental process. Exceptions include agents that act specifically on dividingcells, such as ionizing radiation, 5-aza-2-deoxycytadine (5-AZA), and 5-fluorouracil. In this caseonly dividing cells will be affected because the agents affect only active DNA synthesis. Sometissues such as the heart appear to be quite resistant to cytotoxic agents such as 5-AZA andcyclophosphamide metabolites. Other tissues, such as those in the limb bud and neural tube, arequite sensitive to these toxicants. It is interesting that the chorioplacenta, which is a proliferativeorgan, is very resistant to cytotoxic agents. 22Some developmental toxicants act only on very specific subpopulations of cells that have unique,tissue specific receptors. Teratogens that have this characteristic will only affect the target organand do not usually produce malformations at other sites. Examples of this class of toxicants includesex hormones, such as the progestins, estrogens, and androgens. In contrast to progesterone and17α-hydroxyprogesterone caproate, high doses of some of the synthetic progestins have beenreported to cause virilizing effects in humans. Exposure during the first trimester to large doses ofandrogen-related progestins, such as 17α-ethinyltestosterone, have been associated with masculinizationof the external genitalia of female fetuses. 23,24 Similar associations result from exposure tolarge doses of 17α-ethinyl-nortestosterone (norethindrone) and 17α-ethinyl-17-OH-5(10)estren-3-one (Enovid-R). 24,25 The preparations with androgenic properties may cause abnormalities in thegenital development of females only if present in sufficient amounts during the critical period ofdevelopment. 23,24,26 Grumbach and colleagues 25 point out that labial fusion could be produced withlarge doses if the fetuses were exposed before the 13th week of pregnancy, whereas clitoromegalycould be produced after this period. Such findings demonstrate that a specific form of maldevelopmentcan be induced only when the embryonic tissues are in a susceptible stage of development.The synthetic progestins, like progesterone, can influence only those tissues with the appropriatesteroid receptors. Because the steroid receptors that are necessary for naturally occurring andsynthetic progestin action are absent from nonreproductive tissues early in development, the evidenceis against the involvement of progesterone or its synthetic analogs in nongenital tissues. 6,27–31Only cells that bear the appropriate sex hormone receptors can be affected by sex steroids. Therefore,it follows that only those tissues bearing cells with the appropriate receptors can be developmentallymodified by the presence of these hormones.C. Influence of DoseThe dose or magnitude of the exposure to a developmental toxicant may greatly affect the ensuingpathogenesis and final outcome. Agents that have more generalized mechanisms of action, such asantiproliferative or cytotoxic drugs, may result in intrauterine growth retardation at lower dosesand structural anomalies or death at higher doses. On the other hand, agents that have receptormediatedtissue specificity may not manifest fetotoxicity at higher doses because tissues not bearingappropriate receptors would not be affected at any dose. However, one might expect a doseresponsive gradation in the severity of the insult to the specific tissues or organs affected.An important consideration is not only the dose but also the overall magnitude of the exposureto the toxicant. Prolonged exposure to doses below cytotoxic levels or below levels that inhibitcellular function would not be expected to induce perceptible changes to developmental processes.When designing toxicity studies to assess human reproductive safety, investigators need to carefullyconsider the effects from chronic exposure because gestation in the human is much longer than formost animal species used for these studies. The pathogenesis of some agents, such as the angiotensin-convertingenzyme (ACE) inhibitors, involves effects secondary to oligohydramnios in thefetal period, a consequence of fetal anuria. 32–39 Because of the much shorter gestation and fetalperiod of most experimental species, pathogenesis relating to these types of secondary effects maynot be observed in animal toxicology studies. In some instances the ACE-inhibitor syndrome wasnot found because the experimental animals were only exposed during early organogenesis.© 2006 by Taylor & Francis Group, LLC

66 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 3.1Stochastic and threshold dose-response relationships of diseases produced byenvironmental agentsPhenomenon Pathology Site Diseases Risk DefinitionStochasticThresholdDamage to asingle cell mayresult indiseaseMulticellularinjuryDNAMultiple;variableetiologyaffecting manycell and organprocessesCancer, germcell mutationMalformation,growthretardation,death,toxicity, etc.Source: Modified from Brent, R.L., Teratology, 34, 359, 1986. With permission.Some risk existsat all dosages;at low doses,risk may beless thanspontaneousriskNo increasedrisk below thethreshold doseThe incidence ofthe diseaseincreases butthe severity andnature of thedisease remainthe sameBoth the severityand incidence ofthe diseaseincrease withdoseDose Response Relationship of ReproductiveToxins as Compared to Mutagens and Carcinogens100Risk ofTeratogenesisPercent of Survivors withReproductive Toxicity30Risk of MutagenesisBackground Incidence of Clinically RecognizedHuman Reproductive Diseases (Genetic diseases,birth defects, spontaneous abortions)0Dose of Teratogen or MutagenFigure 3.1The dose response curve of environmental toxicants (drugs, chemicals, and physical agents) canreveal deterministic (threshold) and/or stochastic effects. Mutagenic and carcinogenic events arestochastic phenomena and theoretically do not have a threshold exposure below which no riskexists. At low exposures the risk still exists but is usually below the spontaneous risk of cancerand mutations. Whether the curve is linear or curvilinear for stochastic phenomena can be debated,but from a theoretical point of view it intersects zero. Toxicological phenomena, such as teratogenesis,that do not involve mutagenic and carcinogenic effects, usually follow an S-shaped curve witha threshold below which no effects are expected.A final but extremely important consideration regarding pathogenesis and dose involves theconcept of threshold. The threshold dose is the dosage below which the incidence of death,malformation, growth retardation, or functional deficit is not statistically greater than that ofcontrols. For drugs and other chemicals, the threshold level of exposure is usually from less thanone to three orders of magnitude below the teratogenic or embryopathic dose required to kill ormalform half the embryos. A teratogenic agent therefore has a no-effect dose, as compared withmutagens or carcinogens, which have a stochastic dose-response curve (Table 3.1, Figure 3.1). 40The severity and incidence of malformations produced by every exogenous teratogenic agent thathas been appropriately tested have exhibited threshold phenomena during organogenesis. 2© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 67Table 3.2Mechanisms of action of environmental teratogens1. Cell death or mitotic delay beyond the recuperative capacity of the embryo or fetus (e.x., ionizing radiation,chemotherapeutic agents, alcohol)2. Inhibition of cell migration, differentiation, and/or cell communication (e.x., retinoic acid, cyclopamine)3. Alterations in programmed cell death4. Interference with histogenesis by processes such as cell depletion, necrosis, calcification, or scarring5. Biologic and pharmacological receptor-mediated developmental effects (e.x., etretinate, isotretinoin, retinol,sex steroids, streptomycin, thalidomide)6. Metabolic inhibition (e.x., warfarin, anticonvulsants, nutritional deficiencies)7. Deformations (physical constraint from: uterine myoma, multiple pregnancy, oligohydramnios, amniotic bandsyndrome, etc.)8. Disruptions: vascular disruption, placental emboli, inflammatory lesions, amnionitis; (e.x., cocaine, chorionicvillus sampling, misoprostol)Source: Modified from Beckman, D.A. and Brent, R.L., Mechanism of known environmental teratogens: Drugsand chemicals, in Clinics in Perinatology, Brent, R.L. and Beckman, D.A., Eds., W.B. Saunders, Philadelphia,1986, p. 649. With permission.It is important to keep in mind that the threshold concept, as it is usually applied, concernsfetal outcome and malformation, not effects observed at the cellular level. Effects at the cellularlevel are probably still detectable below the observable threshold of an agent, but these effects maynot be deleterious and may be completely reversible. Cellular effects most likely also manifestthreshold phenomena because a certain number of receptors or ligands must be affected by achemical or drug before cellular processes themselves are diverted via second messenger pathwaysor by other means. At the tissue level, a threshold number of cells within the tissue must be divertedfrom normal processes or killed to result in dysmorphology, dysfunction, or embryonic death. Thus,another way to view threshold is to say that below the threshold, a developmental toxicant has nodiscernable final pathogenesis. However, that should not imply that there are no discernable earlycellular effects from the agent. Empirically, it means that the dose was low enough that the affectedcells did not permanently alter the developmental process, and recent experimental evidence supportsthis concept. Francis, Rogers, and colleagues 41,42 observed that cyclophosphamide and 5-AZA,agents that result in significant cell death and resultant developmental toxicity and dysmorphologyat higher doses, induced significant perturbations in the cell cycle at doses that did not produceovert teratogenesis. These studies illustrate the ability of the embryo to compensate or recover fromdamage of a certain level. It is expected that as techniques for studying early pathogenesis becomemore sensitive, we will have the increased ability to detect changes at the cellular level that do nottranslate into overt developmental toxicity or dysmorphogenesis.IV. PATHOGENESIS AT THE CELLULAR LEVELThere are myriad mechanisms by which developmental toxicants can affect cellular processes.Fortunately, from the standpoint of understanding pathogenesis, effects at the cellular level willusually manifest as one or more of the following: (a) cell death, (b) altered cell–cell interactions,(c) reduced biosynthesis, (d) impaired cell migration, or (e) reduced or impaired proliferation. Thepotential cellular outcomes from exposure to a developmental toxicant were first described byWilson 2 and have been modified and expanded by other investigators (Table 3.2). 43 For instance,while it has long been accepted that excessive cell death may result in dysmorphology, it is nowrecognized that a failure to induce appropriate programmed cell death can also result in dysmorphogenesis(reviewed by Mirkes 44 ). It is important to keep in mind that although alterations tocellular function can be described as a discrete outcome (e.g., cell death), the various cellularoutcomes may often occur concurrently within the organism. Moreover, because of the hierarchicalnature of pathogenesis, potentially all of these cellular events will eventually come into play asdevelopment proceeds.© 2006 by Taylor & Francis Group, LLC

68 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. Cell DeathA generalized mechanism of cell death is often suspected as an early event in the pathogenesis ofabnormalities when growth retardation is also present. It should also be suspected when the agentcauses manifestations of abnormal development in multiple tissues that, based on timing of exposure,correlate with the tissues’ rapid growth or development. However, the only way to determineif cell death represents a major early pathogenic event is to monitor for cell death soon afterexposure to the agent or event that results in the malformation. The toxicologist should then focusnot only on those tissues known to be adversely affected by the agent but also on other tissues thatare rapidly dividing and differentiating during that period of gestation.Pathogenesis of malformations resulting from exposure to cytotoxic agents is highly dependenton the stage of development, the number of cells affected, and the inherent ability of the tissue torecover. The cells of very early embryos (preorganogenesis) retain a great deal of multipotentiality.Thus, division of the remaining cells can often overcome cell loss, with little or no consequencesto later development. However, when the number of cells that die exceeds the capacity for theembryo to replace them, embryonic death or malformation may result. When organogenesis stageembryos are exposed to agents that cause cell death, there is a high likelihood of abnormaldevelopment. Pathogenesis will depend on both the nature and location of the cell type affectedand the extent of the cell death in that tissue. Many agents that produce developmental abnormalitieshave been associated with an increase in cell death in the affected structure. 45It is generally accepted that cell death can take one of two forms, i.e., apoptosis, often referredto as programmed cell death, and necrosis. Necrosis usually results from a more rapid and severeinsult to a cell, and has a greater potential for damage to surrounding cells and tissues. Cell deathmay take the form of necrosis when cells are irreversibly injured as a result of exposure to cytotoxicagents, extremes of pH or temperature, free radicals, or membrane disrupters. During the necroticprocess, there is loss of cell membrane integrity, with concomitant leakage of cytoplasm and markedenlargement of mitochondria (reviewed by Zakeri and Ahuja 46 ). Necrotic cell death culminates withcell lysis and release of the cellular contents into the surrounding environment, where nucleasesand proteases then degrade the cellular constituents. This rapid disintegration of the cell has injuriouseffects on surrounding tissues because of the release of hydrolytic (lysosomal) enzymes and theactivation of inflammatory mediators.Unlike necrosis, apoptosis is a carefully regulated and tightly controlled process that has fewor no direct adverse consequences on surrounding tissues. Apoptosis is a normal and essentialcomponent of development with well-known examples that include apoptosis in the vertebrateneural tube to facilitate neural tube fusion and removal of interdigital mesenchyme to form thedigits of the limb. 44 Recent studies using p53 knockout mice and 4-hydroperoxycyclophosphamide(4OOH-CPA) as the teratogen have provided evidence that whether the cellular response to acytotoxic agent follows a necrotic or an apoptotic pathway may be influenced by levels of p53.Wild type mice responded to 4OOH-CPA with cell death characteristic of apoptosis, while cells inmice null for p53 displayed characteristics of necrosis. Heterozygotes had cells that displayedcharacteristics of both apoptosis and necrosis. 47There are two major pathways for apoptosis: a receptor-mediated, or extrinsic, pathway, and amitochondrial, or intrinsic pathway. 44 Extrinsic apoptosis occurs in response to the binding ofligands, such as TNFα, to cellular receptors, while intrinsic apoptosis is initiated by factors thatultimately cause the release of cytochrome c from the mitochondria. It is this latter type of apoptosisthat is most frequently associated with exposure to developmentally toxic agents. 44 Both pathwaysare mediated by caspase cascades, especially the cleavage of procaspase 3 to caspase 3, and bothresult in similar dismantling of cellular constituents. However, in receptor-mediated apoptosis,caspases 8 and 10 are involved in the initiation of apoptosis, while in intrinsic apoptosis the cascadeis initiated by caspase 9. 44© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 69Table 3.3Methods to detect cell deathTechnique Methodology Comments Selected ReferencesAcridine orange Staining with vital dye Stains phagolysosomes 49, 64, 226associated with cell death;visual inspection usingfluorescence microscopyNeutral Red Staining with vital dye Same as above 52Nile Blue Sulfate Staining with vital dye Same as above 41, 42, 60, 61, 227Apoptosis — DNAladderApoptosis — TUNELDual staining forapoptosisDNA electrophoresisTerminal transferasedUTP nick end labeling;labels 3′ ends of cleavedDNACombines TUNEL withimmunohistochemistryCare must be taken with DNAsample to prevent extraneousdegradationDetects DNA strand breaksassociated with apoptosisRequires multiple wavelengthsfor fluorescence but can bemodified for colorimetricsubstrates47, 49, 5255, 65, 228, 22948Studies using a variety of agents and adverse stimuli have demonstrated reproducible differencesin the sensitivity of embryonic tissues to cell death. For instance, regions including the prosencephalicneuroepithelium and surrounding mesenchyme are especially sensitive to cell deathinduced by agents such as hyperthermia, while the heart, mesencephalic mesenchyme, and yolksac are resistant. 44,48 A similar pattern of tissue selectivity has been reported for cell death inducedby ethanol, 4OOH-CPA, 2-deoxyadenosine, staurosporine, and other developmental toxicants. 44,49–53This differential sensitivity appears to be related, at least in part, to the ability of cells in resistanttissues such as the heart to prevent the release of cytochrome c by the mitochondria and thus preventthe apoptotic cascade. 44,52–54 Recent evidence also indicates that the expression of p53 in response tocytotoxic agents may also differ among tissues and may contribute to differing tissue susceptibilities. 55Some agents, by their nature, target proliferating cells. This replication-associated cytotoxicitymost frequently results from agents that affect DNA synthesis. For example, cyclophosphamide(or its activated analogs) causes alkylation of DNA during the S-phase of the cell-cycle. 56 This isinitially detectable as an alteration of cell cycle, and eventually leads to cell death in the rapidlydividing cell populations of the developing embryo. 41,56 Similar findings have been reported with5-AZA. 42,57Unlike the case of agents that target proliferating cells, the pathogenesis of some developmentaltoxicants involves quantitative increases in cell death in regions of tissues that are already undergoingprogrammed cell death. Retinoic acid (RA) is a good example of a teratogen that displaysthese properties. Retinoic acid exposure results in a variety of defects involving diverse tissues,depending on dose and stage of exposure. Studies have demonstrated that exposure to RA resultsin increases in cell death in regions of normally occurring cell death. Excessive cell death in thesezones has been correlated to the formation of structural anomalies, including spina bifida, mesomelia,digit defects, and craniofacial malformations. 58–61It is important to differentiate between effects due to cell death and those caused by a failureof cellular proliferation or differentiation, as these have the potential to produce similar outcomes.In most instances, when excessive cell death is an initial component of pathogenesis, it occursrapidly and is detectable within the first few hours following exposure. Numerous techniques havebeen developed to measure cell death and to differentiate between necrosis and apoptosis, andmany of these have been applied to embryonic studies. These techniques are summarized in Table3.3. One of the most frequently applied methods involves the use of vital dyes, such as Nile BlueSulfate (NBS). NBS, like many vital dyes used to detect cell death, is accumulated in membranebound acidic compartments (lysosomes) that occur when neighboring cells or macrophages engulf© 2006 by Taylor & Francis Group, LLC

70 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONcellular debris from dead cells, and it also stains free apoptotic bodies. 62–64 More specific approachesused to detect apoptosis in embryonic studies include examination of DNA fragmentation anddetection of the presence of activated factors in the apoptotic cascade. Because apoptosis resultsin the cleavage of DNA into discrete lengths, electrophoresis of DNA from tissues undergoingapoptosis will reveal a “DNA ladder” effect. 47,49,52 Terminal transferase dUTP nick end labeling(TUNEL) can also detect DNA fragmentation. 65 This technique is based on the use of terminaldeoxynucleotidyl transferase (TdT) to label free 3-hydroxyl ends of DNA fragments with a labelednucleotide. This technique is relatively flexible with regard to the label used on the dUTP and hasincluded the use of peroxidase, alkaline phosphatase, biotin, and fluorescent markers. Cell deathcan also be examined by detecting specific molecules involved in apoptosis, such as cleaved caspase-3, by immunohistochemistry on histological sections or whole mounts. A recent and useful techniquehas been described that combines both immunostaining for cleaved caspase-3 and TUNEL, 48and allows for a more complete assessment of the presence of apoptosis in embryonic tissues.B. Cell ProliferationAgents that produce alterations in cell proliferation can produce fetal outcomes that are similar toagents that cause cell death. Like cell death, reduced cellular proliferation reduces the number ofcells available in a tissue for later growth and differentiation. Alterations in cellular proliferationmay be suspected as an early pathogenetic event when excessive cell death is not readily apparentor is insufficient to account for ensuing dysmorphogenesis. Reduced cell proliferation can resultfrom direct effects on progression through the cell cycle and mitosis, deficiencies in growth factors,deficiencies in signaling or induction factors, or impairment of the cellular response to thesemolecules.Tissues most likely affected by agents that directly impair cell-cycle mechanisms are those thatcontain the most rapidly dividing cells at the time of exposure. Replication-associated cytotoxicagents that affect DNA replication or mitosis, such as cyclophosphamide and 5-AZA, have beenshown to induce detectable but reversible alterations in the cell cycle at doses below those thatresult in cell death or dysmorphology. 41,42,56 Other teratogens, such as valproic acid (VPA), mayalter the cell cycle without directly interfering with DNA replication. For instance, VPA treatmentinduced a cell-cycle arrest in the mid-G1 phase in the C6 glioma cell line in vitro at doses belowthose producing overt cytotoxicity. 66,67 In vivo studies in mice have reported evidence that the VPAinducedcell-cycle arrest may involve alterations to the enzyme ribonucleotide reductase in specificregions of the neuroepithelium associated with neural tube closure. 68 Alterations in C6 glioma cellproliferation in vitro have also been associated with the ability of anticonvulsant agents to induceneural tube defects (NTDs) in vivo and thus have the potential to be a predictive indicator for NTDpathogenesis. 69In addition to direct effects on cell replication, failure of cell proliferation can also occur as aresult of a deficiency of the growth factors or signaling molecules that induce proliferation. Failureof cell proliferation by these means usually occurs downstream from other pathogenic events that affectthe cell populations responsible for the synthesis of the growth factor or signaling molecule involved.In these cases, measurement of cell proliferation in affected regions may show decreases in the numberof cycling cells but may not reveal discrete cell-cycle arrest. The effects of cell-signaling moleculeson tissue growth and differentiation will be discussed in greater depth later in this chapter.Numerous methods are employed to detect changes in cell proliferation and the cell cycle.Analysis of levels of cell proliferation can be performed on histological sections by analysis of thenumber of mitotic cells present or by assessment of the incorporation of labeled nucleotides. Sometraditional methods, such as incorporation of radiolabeled nucleotides into cells undergoing DNAsynthesis, have largely been replaced by similar techniques that utilize nonradioactive labels. Forexample, incorporation of bromodeoxyuridine (BrdU), a thymidine analog, can be detected by© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 71Table 3.4Methods to detect alterations in cell-cycleTechnique Methodology Comments Selected ReferencesFlow cytometryIncorporation of DNAanalogsIndicator cell lines;micromass cultureMarker of cell cycle usingpropidium iodide (or other)fluorescent DNA stainIncorporation of analogs intoDNA of replicating cellsIn vitro growth of cells ortissueRequires use of expensiveequipmentMost radioisotope techniqueshave been replaced by useof nonisotope labels, analogssuch as BrdUFollow by analyzingproliferation by use of flowcytometry, cell number, orDNA synthesis markers41, 42, 55, 5668, 70, 71, 22866–69, 230–234immunohistochemistry in tissue sections or by use of in vitro enzyme-linked immunosorbant assays(ELISAs) on cells following in vitro or in vivo exposure to potential toxicants. 68,70,71 A morequantitative measure for cell-cycle analysis utilizes flow cytometry on isolated nuclei stained withfluorescent labels, such as 4,6-diamidino-2-phenylindole (DAPI) or propidium iodide, that bindDNA. This type of analysis relies on distinguishing nuclei based on their DNA content and canthus differentiate cells in G0/G1, S, and G2/M phases. 41,42,56,72 It is important when using thistechnique to isolate the cell populations of interest, which may be a difficult task depending on thesize or stage of the embryo employed. Flow cytometric analysis of the cell cycle is also usefulwhen measuring the effects of a teratogen on indicator cell lines, such as C6 glioma cells, cellsisolated from embryos and grown in vitro as monolayers, or cells from micromass or organ culture.Selected methods for assessing cell proliferation are given in Table 3.4.C. Alterations in Cell-Signaling and Cell–Cell InteractionsCell-signaling pathways form the basis for embryonic patterning and differentiation and thus arean important consideration in determining the pathogenesis of a developmental toxicant. Signalingpathways control morphogenesis by the secretion of cell products that are both temporally andspatially distributed. The downstream effect of these developmental signaling molecules on a tissueusually occurs at the level of transcription and results in alterations of gene expression in the targetcells. The role of signaling molecules at the cellular and tissue level is discussed in the followingtext in the context of their involvement in both normal and abnormal development.Numerous cell-signaling pathways are involved in embryogenesis. Perhaps the best understoodinvolves the ligand Sonic Hedgehog (Shh), the receptors patched (Ptc) and smoothened (Smo), andthe GLi family of transcription factors that are involved in the transduction of the Shh signal. Thispathway has been highly conserved across organisms as diverse as drosophila, amphibians, andmammals (reviewed by Walterhouse et al. 73 ). Genetic alterations in Shh signaling result in similarand reproducible phenotypes in humans and in mice. For example, altered expression of Shh inhumans results in holoprosencephaly, 74–76 and in mutant mice lacking Shh (shh -/-) the resultingphenotype is also holoprosencephaly, with other axial patterning anomalies. 77 Human syndromesinvolving mutations in the GLi3 gene include Greig cephalopolydactyly and Pallister–Hall syndrome,both of which include postaxial polydactyly and craniofacial malformations. 78,79 The phenotypeof a mouse mutation of Gli3, extra-toes (Xt), also includes polydactyly and malformationssimilar to those observed in humans. 80 Such observations from genetic mutants and numerousembryological studies illustrate the importance of signaling molecules such as Shh for the developmentof numerous tissues and structures. For instance, in vertebrates the Shh pathway is crucialto dorsal-ventral patterning of the CNS, anterior-posterior specification in the limbs and somites,axial skeleton patterning, and gastrointestinal tract and lung development. 73,81–87© 2006 by Taylor & Francis Group, LLC

72 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe patterning or morphogenesis achieved by signaling pathways is based on the secretion ofdifferent signaling molecules from several signaling pathways in overlapping gradients and alongmultiple axis. Alterations in the strength of one signal, either by changes in the secretion or thetransduction of the signal, will substantially alter morphogenesis. Changes in transduction mayinvolve alterations in the expression of receptors, transcription factors, inhibitors, or downstreamsignaling factors, or in the distribution of gap junctions. 88–90 Cell signaling in the vertebrate limbhas been extensively studied and provides a good example of the effects of signaling molecules onmorphogenesis. In the vertebrate limb, as in other tissues, signaling molecules form overlappinggradients in three dimensions, such that cells within the limb bud attain a positional value. 86Signaling molecules for each dimensional specification are produced by discrete tissues within thelimb bud. Damage to these regions or alterations in the signaling pathways evoked from theseregions results in characteristic phenotypes that give insight to the identity of the affected tissueor pathway responsible. For example, dorsal-ventral patterning of the limb involves signalingpathways that originate in the overlying ectoderm of the limb bud and involves the expression ofWnt7a for dorsal specification and engrailed-1 for ventral patterning. 91 Mutations in engrailed-1result in both dorsal and ventral expression of Wnt7a and dorsal morphology on both dorsal andventral sides, while mutations in Wnt7a give the reverse scenario. 92Anterior-posterior patterning in the limb largely involves Shh, produced by cells in the zone ofpolarizing activity (ZPA), and transduction of the Shh signal by other proteins, such as BMP-2. 93–95Altered expression of Shh results in altered patterning of distal structures. 96,97 An increase in Shhsignaling, produced by beads soaked with Shh and placed on the anterior margin of the developinglimb bud, results in the development of additional digits. 95 In the mouse mutant extra-toes there isup-regulated signaling of Shh in the anterior regions of the limb bud and the addition of asupernumerary digit. Conversely, an absence of Shh results in the absence of distal structures (suchas digits), but proximal structures are generally unaffected. 96,97Absence of distal structures can also occur because of altered signaling involved in apical-distalgrowth. Apical-distal outgrowth requires signals produced from the apical ectodermal ridge (AER)that overlies the rapidly dividing progress zone of the limb bud. The signals produced by the AERthat appear responsible for apical-distal outgrowth are members of the FGF family of proteins. 86,98Shh from the ZPA also forms a positive feedback loop with FGF-4 from the AER, and this feedbackis also required for continued outgrowth of the limb bud and maintenance of the progress zone. 99–101Damage to the AER results in truncation of the limb, with failure of distal structures to form, 86 aphenotype similar to that observed with lack of Shh expression. The effect is stage dependent.Earlier removal of the AER leads to loss of all structures and later removal allows formation ofproximal features, such as long bones, but an absence of distal features, such as digits. 102,103In contrast to damage in the AER, depletion of cells in the progress zone by teratogens thatkill or impede cell proliferation, such as x-irradiation and 5-AZA, results in formation of distalstructures and failure to produce proximal structures, 71,104 a phenotype similar to that observed forthalidomide. Because the progress zone is a highly proliferative tissue, it is potentially susceptibleto teratogens that target mitotic cells. Excessive cell death or reduced proliferation in the progresszone may reduce the number of cells below the threshold number required to produce apical-distalstructures. Positional value is also influenced by the length of time cells spend in the progress zone,as well as by specification under the influence of Hox genes. 95,105,106Developmental toxicants can alter cell-signaling pathways involved in the morphogenesis ofstructures by causing a deficiency in either the appropriate inducer or target cells involved (due tofailures of differentiation or proliferation, or to excessive cell death), by inappropriate biosynthesisof signaling molecules, or by interference with the transduction of the signal to the target cellpopulation. It is probable that the pathogenesis of many defects induced by environmental agentsincludes altered cell-signaling patterns in the overall pathogenesis and that more than one signalingpathway will be altered, particularly by cytotoxic agents. However, some teratogens, and particularly© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 73those that are receptor mediated, may alter signaling pathways more specifically. This class ofteratogens includes the Veratrum alkaloids, plant toxins long known to cause cyclopia in sheep,that inhibit tissue responsiveness to Shh. 107,108 In studies with chicks, jervine, an especially toxicVeratrum alkaloid, also caused defects that were characteristic of holoprosencephaly. 108Another developmental toxicant that causes changes in cell signaling pathways is retinoic acid(RA). RA is a natural morphogen that is essential to many normal developmental processes. 109Investigators using animal models have shown that both a deficiency and an excess of RA can leadto malformations, including limb defects such as oligodactyly, polydactyly, and reduction defects,and in some cases, limb duplication. Differences in observed abnormalities depend on the timingof administration, dose, route, and species employed. 109–111 RA deficiency prevents normalAER–FGF-4–ZPA signaling in both chick and mouse embryos. 112,113 Studies using RA-deficientmice have demonstrated that RA is required for cells of the ZPA to acquire competence for Shhsecretion, most likely via induction of Hox genes in the mesenchyme and by secretion of FGF-4by the AER. 85,100 Mice deficient in RA lack normal ZPA patterning and exhibit oligodactyly. 109 Incontrast to RA deficiency, direct application of RA to the anterior domain of the chick limb budresults in extra digits or in mirror image limb duplication via the formation of a second ZPA. 93,114–116At higher doses of RA, skeletal elements are reduced or do not form.110, 117Although RA is teratogenic in humans 118,119 and in animals, 120–123 there are notable speciesdifferences in the effects observed. While similar defects exist in humans and rodents for craniofacial,CNS, cardiac, and thymic development, limb defects are seldom reported in humans exposedto RA in utero. 124 Like other teratogens, the effects of exogenous RA on development are highlystage dependent. Late administration during limb development in the mouse causes skeletaldefects, 125 whereas administration during preorganogenesis can induce the development of supernumeraryhind limbs. 126,127 Differences also occur with different methods of administration. Directapplication of RA leads to digit duplication, while systemic administration more often results inoligodactyly. Polydactyly can also occur, particularly at lower doses. 109,128,129 As is the case withseveral other teratogens, there appears to be a preferential loss of postaxial digits of the rightforelimb, 116 a phenotype also prevalent in Wnt7a null mice that display reduced Shh expression. 91Studies have also implicated alterations in the underlying mesenchyme and interference withectodermal-mesenchymal interactions in loss of AER function and signaling in RA pathogenesis,and it is of no small interest that cells of the underlying mesenchyme express receptors forRA. 116,130–133It is important to keep in mind that the preceding overview of signaling pathways in the limbis necessarily greatly simplified for the present discussion. Although discrete morphogenetic functionscan be assigned to regions such as the AER and ZPA, these regions are codependent for limbdevelopment, adding an additional layer of complexity to the system. For instance, maintenanceof the AER requires a functional progress zone and ZPA, while signaling from the AER and theZPA is required to maintain proliferation in the underlying mesenchyme and progress zone(reviewed by Johnson et al. 103 ). The complexity of this system, as well as species differences andeven differences between fore- and hind limb development within a species, will often makedetermining the initial pathological deviation from normal development quite difficult.Perhaps the preferred technique for detecting alterations in cell-signaling patterns followingtoxicant exposure is in situ hybridization using labeled probes to the mRNAs of the molecules ofinterest. There are several good reviews covering this highly valuable technique. 134–137 Patterns ofsignaling molecules in tissues can also be detected by whole mount immunostaining or immunohistochemistrywhen appropriate antibodies are available. For instance, Shh is a protein thatundergoes autocatalytic cleavage to yield a secreted amino (N) peptide and a carboxy (C) peptidethat remains cell associated. Specific antibodies are available to both peptides for several species.Antibodies to cellular receptors, such as cellular retinoic acid binding proteins (CRABPs), are alsouseful in identifying changes to populations of cells responsive to signaling molecules.© 2006 by Taylor & Francis Group, LLC

74 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIOND. Cell Migration and DifferentiationThe dynamic process of morphogenesis involves not only cell proliferation and specification asalready discussed but also cell migration and differentiation. Perhaps nowhere is the importanceof cell migration and differentiation better illustrated than in the changes that occur to the groupof cells collectively termed the neural crest. The neural crest initially develops from the neuroepitheliumon the dorsal aspect of the neural folds early in embryogenesis. Neural crest cells detachfrom the neuroepithelium and migrate as single cells in a process referred to as delamination andepithelial-mesenchymal transformation (EMT). Delamination involves specific changes to extracellularmatrix molecules that allow the neural crest cells to detach from the neuroepithelium. Thesechanges include down regulation of neural cell adhesion molecule (N-CAM) and N-cadherin, andup regulation of cadherins 7 and 11 (reviewed by Nieto 138 ). Neural crest cells then undergo extensivemigration throughout the embryo, whereupon they eventually differentiate to form tissues and structuresas diverse as the connective tissues of the face, elements of the peripheral nervous system, andmelanocytes. Because of the rapid dynamics of this cell population and the involvement of neural crestcells in the development of so many organ systems and tissues, it is not surprising that alterations inneural crest cell migration and differentiation often play a major role in the pathogenesis of developmentaltoxicants.It is generally accepted that early neural crest cells are multipotent and that specification ofneural crest cell lineages is based on signals encountered during cell migration, interactions withthe extracellular matrix, and tissue-specific gene expression of a variety of induction factors. 139However, evidence also suggests that some neural crest cells, and particularly those originating inthe hindbrain, may acquire some positional specification while the cells are still in the neural tube,possibly under the influence of Hox genes. 140,141 Neural crest cells that arise in the neuroepitheliumof the hindbrain form three discrete streams of migrating cells that invade the branchial arches toform the cranial ganglia and components of the craniofacial skeleton. Specifically, neural crest cellsadjacent to rhombomeres r2, r4, and r6 populate the first, second, and third branchial arches,respectively. 139 Misdirection of these discrete streams of cells in their allotted pathways, either byaltered prespecification or altered signals from the tissues through which the cells migrate, willresult in craniofacial anomalies.Neural crest cells arising in the trunk follow two defined pathways of migration. The firstmigration pathway occurs ventrally and leads to formation of the neurons and glia of the dorsalroot ganglia, the sympathetic ganglia, and the adrenal medulla. Later, trunk neural crest cells followa dorsolateral pathway that gives rise to the melanocytes. The formation of these discrete pathwaysfor neural crest cell migration is controlled to a large extent by interactions of the neural crest cellswith the extracellular matrix. These allow or impede migration in the tissue, depending on thepattern of expression of cell surface receptors, such as the ephrin receptors, on the neural crestcells themselves. 142,143Because the migration and differentiation of the neural crest cell population is necessary for thedevelopment of so many tissues and organs, it is intuitive that drugs and chemicals that alter themechanics of cell migration, for instance, those that cause alterations to the cytoskeleton, have thepotential to profoundly affect embryonic development. Furthermore, altered cell proliferation or celldeath, not only in the neural crest population but also in cells responsible for secretion of extracellularmatrix components and signaling factors required for neural crest migration and specification, mayalter the fate of the neural crest population and result in dysmorphology. For example, spinal nerveabnormalities in mouse embryos exposed to VPA coincided with alterations in somite morphogenesis,suggesting that the spinal nerve abnormalities resulted, at least in part, from anomalous patterns ofmigration and induction of the neural crest by the developing somites themselves. 144Altered patterns of neural crest cell migration may also arise from alterations in cell signalingthat result in changes in neural crest fate specification. For example, RA is involved in theprepatterning and specification process of hindbrain neural crest cells, 145,146 presumably by altering© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 75Table 3.5Methods to detect neural crest cells and derivativesMarker Method Comments Examples of UseNeurofilament proteinHNK-1CRABP1DiI (1,1-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine)Immunostaining forprotein found inneurons, craniofacialand dorsal rootganglia.Immunostaining forsialoprotein found onmigrating NC andderivativesImmunostaining for RAbinding protein foundin certain neural crestcell populations andcells of the hindbrainLipophilic fluorescentdye injected into cellsof interestCan be used either aswhole mount or onhistological sections. Avariety of labels isavailable for secondaryantibody. Staining is timeconsuming for wholemounttechnique.144, 147, 148, 227,235–237Same as above 145, 238, 239Same as above 133, 147, 148, 152, 153Requires microinjectionequipment andfluorescence microscopy145, 147, 238, 240Hox gene expression. 141 At high concentrations, RA appears to alter neural crest cell identity,resulting in altered migration patterns and subsequent dysmorphology in the target tissues. 145,147,148Similar effects have been noted across several species; however, there appear to be notable differenceswith respect to the resultant structures affected. 145,147,148 Some of the differences observedmay result from inherent differences between species in early neural crest cell development. Inchick embryos, for example, closure of the neural tube is a requirement for neural crest induction,while in mice the neural crest is formed and can migrate from the cranial region in the absence ofneural tube closure. 138,149,150 Differences in early neural crest development have also been noted inXenopus and zebra fish. 138In some cases neural crest cell migration may proceed normally but later differentiation of theneural crest is perturbed, resulting in dysmorphology similar to that induced by lack of neural crestcell migration. For instance, migration of the neural crest proceeds normally in mouse embryoslacking the endothelin A receptor on those cells. However, later signaling from the postmigratoryneural crest cells to surrounding ectomesenchymal cells in the arches for production of downstreaminductive factors is impaired. This results in cardiac malformations and in defects in branchial archderived craniofacial tissues. 151–153There are several markers that can be used to study the effects of a drug or chemical on themigration and fate of the neural crest cell population. Antibodies to the marker HNK-1 have provenexceptionally useful for detection of migrating neural crest cells in mammalian embryos. This cellsurface antigen is present on greater than 90% of migrating crest cells but is not expressed by themesectoderm. 154,155 Other markers include antibodies to neurofilament proteins that are expressedby neural crest derived neurons and ganglia, CD44 (a marker restricted to cranial neural crest inthe chick), 156 and cellular retinoic acid binding protein (CRABP1), which is expressed in hindbrainregions and the neural crest from r4 to r6, but not r1 or r3. 147,157,158 Selected markers and their usesare presented in Table 3.5.V. EXAMPLES OF ABNORMAL PATHOGENESISDeviations from normal developmental processes at the cellular level eventually result in alterationsin tissue and organ development. Alterations at the tissue level can manifest as hypoplasia, retardedor arrested growth, altered patterns of cellular differentiation, altered patterns of growth, altered© 2006 by Taylor & Francis Group, LLC

76 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 3.6Morphological observations: methods for external, visceral, and skeletal examinationsTechnique Methodology Comments Ref.Fresh visceral Dissection at termination Not appropriate to detect somemalformations; cephalicexamination requires fixationWilson’s dissectionFrozen cephalicsectionsSkeletalExternalSerial sections through fixed tissueto visualize gross internal organand tissue structuresFrozen sectioning for cephalicexamination combined with freshvisceral examAlizarin Red (bone) and Alcian Blue(cartilage) stainingStaining of fixed specimen to revealdetails of external morphology,especially somitesHazardous as Bouin’s fixativecontains picric acidAllows for faster analysis of cephalicmorphology compared to fixation inBouin’sDifferential staining of cartilage andbone.Can be done after whole mountimmunostaining or other techniques.Requires fluorescence microscopy241242, 243244245–250144, 251morphologic expression, tissue destruction, deformation, and/or altered morphogenetic movements(i.e., failure of elevation of palatal shelves or neural folds). Unlike alterations that occur on a cellularlevel, effects at the tissue level can be observed histologically or by gross observation. The remainderof this chapter focuses on features involved in the pathogenesis of several of the more commoncongenital defects to illustrate some of the ways in which alterations at the cellular level inducedby a toxicant may translate into a congenital defect recognizable at birth. Selected methods forgross morphological observations of fetal anomalies are listed in Table 3.6.A. Defects with Different Underlying Pathogenesis but Similar Morphology1. Cleft PalateOne example of a discrete defect that can result from diverse agents and differing underlyingpathogenesis is clefting of the secondary palate. Palatogenesis involves morphogenetic processesthat include cell migration, proliferation, cell signaling, alterations in extracellular matrix synthesis,and cell specification and differentiation. Changes to the normal developmental sequence involvingany of these processes can lead to a failure of palate fusion and a resulting cleft (reviewed by Younget al. 159 ). Cleft palate can occur as an isolated morphological defect, as observed with corticosteroidadministration in rodents, 11,12,160 as part of a multiple defect syndrome, as occurs with retinoic aciddysmorphogenesis, 111 or secondary to another anomaly, such as cleft lip. 161,162 Although the underlyingpathogenesis of a cleft palate resulting from exposure to a toxicant or environmental insultcan be quite different from that caused by another agent, the end result can be virtually indistinguishablein morphology.As with other anomalies, a key to determining the underlying pathogenesis of cleft palate is tofirst determine at what stage the normal developmental processes were initially affected. Forpractical purposes, the development of the secondary palate can be divided into three stages. 163 Thefirst of these is the formation and growth of the maxillary processes. These structures originatefrom neural crest–derived mesenchymal cells surrounded by an epithelium of ectodermal origin.The maxillary processes initially sit vertically in the oral cavity. Elevation of the shelves to ahorizontal position is the second stage of palatogenesis. The third and final stage comprises thefusion of the two shelves, degeneration of the medial edge epithelium (MEE), and formation of acontiguous mesenchyme. Identifying a defect in one of these major developmental events of palateformation can provide important clues to the underlying pathogenesis of the anomaly.Early changes to craniofacial morphogenesis that affect the formation of the maxillary processestend to involve perturbations in cell signaling, migration, and subsequent patterning. The connectivetissue of the various prominences that form the face and palate are derived from neural crest cells© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 77that originate from the neural tube in the midbrain and rostral hindbrain. 159 Failure of neural crestcell migration, differentiation, or proliferation, or extensive cell death in the developing primordiumresults in failure of the maxillary processes to grow to sufficient size for later palatal fusion.Patterning of the craniofacial structures may also be influenced by regional differences in theoverlying ectoderm. 159 An interesting feature of the craniofacial prominences that has direct implicationsfor ascertaining the pathogenesis of a teratogen is that growth and development of each ofthe facial primordia appear to be controlled by differing morphogenetic processes. 159 Thus, teratogensmay affect a specific portion of craniofacial morphogenesis while having no effect on others.The differing susceptibility of the facial prominences to teratogens such as RA may result, at leastin part, in differences in the expression of RA receptors and signaling pathways, such as the Shhpathway. 159Another frequent cause of cleft palate is a failure of palatal shelf elevation. Although thepersistence of the palatal shelves in the vertical position would be quite evident to gross visualinspection, it is often a delay in shelf elevation, rather than a complete lack of shelf movement,that results in cleft palate. When the shelves elevate too late in the developmental sequence, theother structures of the head have grown and changed in size and shape such that the two shelvescan no longer meet. Continued growth of the head further separates the shelves over time. 164Although the processes that govern shelf elevation are not well understood, there is substantialevidence that alterations in the extracellular matrix are essential to the process, possibly throughthe interactions with matrix metalloproteinases (reviewed by Kerrigan et al. 163 ). Studies also indicatethat the composition of the extracellular matrix molecules synthesized in the shelves changes withrespect to position along the shelf. This difference might impart a tension or force that is involvedin the process of elevation. 163 Others have suggested that the shelves are initially held in the verticalposition by the tongue. Movement of the tongue, which occurs developmentally at the same timeas shelf elevation, releases the shelves to attain the horizontal position. 163The final series of events in palatogenesis that are subject to perturbations brought about byexposure to developmental toxicants involves the fusion of the palatal shelves, the breakdown ofthe MEE, and the consolidation and fusion of the underlying mesenchyme. The degeneration ofthe MEE and its associated basement membrane is essential for fusion of the palate. Failure of theMEE to degenerate provides a barrier to the mesenchymal cells within the shelves, and a contiguousmesenchyme cannot form. As the craniofacial structures grow, the seam formed at the MEE is notsufficient to hold the shelves together, and clefting occurs. 159,163 A complete failure of fusion of themesenchyme of the two shelves following shelf elevation may leave a cleft that is virtuallyindistinguishable from that resulting from delayed shelf elevation.Although it was initially thought that the MEE was removed via programmed cell death, currentstudies have failed to identify apoptotic cells in the MEE. 165,166 Other studies suggest two possibilitiesfor the removal of the MEE. The first is that the MEE cells are transformed into mesenchymalcells that migrate into, and remain in, the fused shelves. 165–167 An alternative explanation for thefate of the MEE is that they migrate out of the seam and become incorporated into the oronasalepithelium. 168 Although little is known concerning the exact mechanisms involved in palate fusion,there is growing evidence for the involvement of growth factors, such as TGFβ3, in stimulating removalof the MEE and subsequent fusion. Mice that are null for TGFβ3 have incomplete palatal fusion, andantisense oligonucleotides or antibodies to TGFβ3 block palatal shelf fusion in vitro. 169–1712. Neural Tube DefectsThe vast majority of neural tube defects (NTDs) arise from a failure of neural tube closure. Failedclosure of the anterior neural tube results in anencephaly, while failed closure in the posterior neuraltube region results in spina bifida. It should be noted, however, that central nervous system defectsinvolving the neural tube may also result from improper closure, bifurcations in the neural tube orgroove, changes involving the overlying ectoderm, and alterations in vesicle formation. Aspects of© 2006 by Taylor & Francis Group, LLC

78 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthe pathogenesis of NTDs resulting from failed neural tube closure, the most frequent types ofNTDs, will be considered below. The processes governing neural tube closure are complex andinvolve diverse cell types and morphogenetic mechanisms. Because of this complexity, the pathogenesisof NTDs can vary widely. Although the initial events that lead to failed closure may differgreatly, the morphologic defect that results, once fusion has failed to occur, will proceed throughdevelopment along similar pathogenetic lines. The result is that NTDs resulting from differingunderlying pathogenesis may ultimately appear morphologically indistinguishable from each other.There are two major processes involved in formation of the neural tube. One is primaryneurulation, in which the neural plate is formed ectopically and subsequently fuses to form a tube.Following primary neurulation, there is secondary neurulation, a process that proceeds caudallyand forms the remainder of the tube by hollowing out the region below the ectoderm. NTDs suchas anencephaly and most spina bifida arise from aberrations in primary neurulation. Studies onhuman embryos suggest that the pathogenesis of the majority of human NTDs is limited to thoseprocesses involved in neural tube closure itself. 150 Other elements of neural tube development, suchas maturation of the neuroepithelium, appear to continue to occur despite lack of closure. Anencephalyand most myelomeningoceles result from failed neural tube closure. Other disorders suchas encephaloceles and a small percentage of myelomeningoceles, appear to be related to subsequentdevelopment of the nervous system and its coverings, rather than being directly related to neuraltube closure, and thus should be considered separately. 150Numerous in vivo and in vitro animal models have been used to study the process of neurulation,neural tube closure, and the pathogenesis of NTDs. Agents that are known or suspected to causeNTDs in humans, such as methotrexate, alcohol, hyperthermia, and others, also cause NTDs inanimal models, regardless of the species employed. However, it should be noted that there aredistinct differences between vertebrate species in the process of neural tube formation that maypotentially alter the effects of a toxicant. For example, the zones of primary and secondaryneurulation overlap in both humans and chicks, a feature not as evident in rodent embryos. Thisis of some significance, as observations of human infants with NTDs reveal that a large numberof myelomeningoceles occur at the site of overlap. 150 Moreover, differences among species appearto exist in the number and position of closure points that may also influence the location and severityof NTDs. Despite these differences, studies using animal models, chemical agents, and genetic mutantshave contributed a large body of information regarding the process of neural tube closure and thegenesis of NTDs. 172 The majority of studies have demonstrated a strong link between inhibited cellproliferation and increased cell death and failed neural fold elevation and closure. Because cell proliferationis so crucial for neural tube closure, it is probable that almost any agent that induces cell deathand is administered at the appropriate time and dose may result in open neural tube.Studies have repeatedly demonstrated a strong link between nutritional factors and the occurrenceof NTDs. In particular, it is now well established that supplemental folic acid lowers the riskof NTDs in the human population, a finding that may be true for other defects as well. 173–175 Folicacid is required for DNA synthesis, and thus cell replication, via its role in one-carbon metabolism.Methotrexate (MTX), a folic acid antagonist, causes NTDs in rodents, rabbits, and humans byinhibiting dihydrofolate reductase. 176 In rabbits, MTX-related defects were severely lessened whenan alternate one-carbon donor was provided concurrently with MTX treatment. 177 Other nutrientsthat may alter the incidence of or susceptibility to NTDs in experimental animal models includezinc, 150,178–183 methionine, 184–191 and inositol, 192–194 but the role of these nutrients in the etiology andpathogenesis of NTDS in humans has not been clearly determined. 195–201B. Abnormalities with Similar Pathogenesis but Different MorphologyAs discussed above, defects that are morphologically similar, such as closure defects, can becaused by different underlying pathologies. In the alternate scenario are those defects that differmorphologically at birth, and can involve dissimilar structures, but that have the same underlying© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 79pathogenesis. Many developmental toxicants induce different defects in an organism based onfactors such as dose and timing of administration. In this case, the underlying pathogenesis maybe similar but the tissue affected and the degree to which it is affected changes, resulting in differentmalformations at term. Disruption defects, and in particular vascular disruption defects, are anotherexample of a group of defects that have a shared pathogenesis but have the potential to result invery different morphologic outcomes.Unlike malformations, disruption defects result from the destruction of existing tissues orstructures that may have previously developed normally. Vascular disruption defects refer to thoseinvolving the interruption or destruction of some part of the fetal vasculature. 9 The pathogenesisof vascular disruption is extremely important to developmental toxicology because it can causedysmorphology at nearly any stage of gestation, affect almost any tissue or structure, and occur asthe primary cause of dysmorphology or secondary to existing malformations. The latter situationcan be extremely difficult to resolve because the secondary disruption can obliterate evidence forthe primary malformation, thus hampering determination of the underlying cause of the congenitalanomaly.Vascular disruption defects usually occur as a consequence of localized or general fetal hypoxiaresulting from direct effects on the fetal vasculature or secondary to changes to the maternal uterineor placental vasculature. Vasoconstrictive drugs, such as cocaine, may alter both maternal and fetalhemodynamics. This results in decreased uterine and placental blood flow resulting from constrictionof the uterine arteries, and brings about increased fetal blood pressure and heart rate. 202Vasodilating agents may decrease uterine blood flow secondary to diversion of blood to peripheralvasculature and yet have no direct effect on fetal hemodynamics. 203 In humans, defects consistentwith a vascular disruption pathogenesis have been associated with exposure to cocaine, 204–206antihypotensives, 203 and the abortifacient misoprostol. 207–209 Defects associated with such mechanicaldisturbances as chorionic villus sampling (CVS), uterine trauma, twin–twin transfusion syndrome,and amnion disruption sequence may also result from vascular disruption. 9,210–212 In animals,defects attributable to vascular disruption have been induced with a variety of treatments. Theseinclude physical methods, such as uterine vascular clamping in the rat, 213–215 and cocaine and othervasoactive substances. 202–204,216,217Studies have shown that the fetal response to hypoxia follows the same pattern of eventsregardless of the agent or mechanism responsible. Early histopathological changes in affectedtissues reveal that fetal hypoxia leads to edema and an increase in vessel diameter, particularly indistal structures. Loss of fetal blood vessel wall integrity leads to vessel rupture, hemorrhage, andformation of subcutaneous blebs or blood-filled blisters. 204,214–217 Necrosis of affected tissues occurswithin 24 to 48 h, and subsequent resorption results in distortion or loss of already formed structures.It has recently been shown that mechanisms involved in the rupture of fetal vessels may sharesimilarities with mechanisms associated with ischemia-reperfusion injury in adult tissues andspecifically may involve the detrimental action of oxygen free radicals. 218–220Like other teratogenic events, the types of anomalies associated with vascular disruption dependon both the stage of gestation and severity of the insult. However, vascular disruptive events, unlikemost teratogens, have been shown to induce congenital defects to some degree during a broad rangeof developmental stages, including postorganogenesis. 9 Studies in humans who had been exposedto cocaine indicate that the fetus is sensitive to vascular disruptive defects throughout the secondand third trimesters, 205,206,221 although genitourinary tract malformations and more severe malformationsmay be induced during the first trimester. 222Of the types of defects ascribed to vascular disruption pathogenesis, limb defects, includingadactyly, transverse reduction defects, syndactyly, polydactyly, and club foot appear to be the mostcommon. 9,205,206,213–217 The type of defect depends largely on the blood vessel affected and theseverity of the ensuing hemorrhage. In animals and humans, the most distal parts of the skeletonare the most easily affected by fetal hypoxia; thus, abnormalities of the phalanges as well as facialabnormalities are the most frequent morphologic manifestations of fetal hypoxia. 9,216,223 Other© 2006 by Taylor & Francis Group, LLC

80 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONdefects often ascribed to vascular disruption include limb body wall complex, gastroschisis, micrognathia,anencephaly, limb reduction defects, syndactyly, and orofacial malformations, such as cleftpalate. 9,212Discerning whether a given anomaly has an underlying pathogenesis of vascular disruption canbe difficult. For instance, a hematoma occurring in the early progress zone of the limb bud canlead to a truncation defect similar to those observed with other pathogenic phenomena, such as x-irradiation-induced cell death. However, substances that induce overt vascular disruption, such ascocaine, unlike many toxicants that cause anomalous development, are less likely to result in aconsistent pattern of malformations. With other developmental anomalies, such as RA-inducedoligodactyly, one finds consistent, bilateral changes to the limb buds, with increasing severity thatcorrelates to increasing dose. In contrast, agents that induce disruption defects produce defects thatare often asymmetrical and sporadic. A hallmark of vascular disruption is the occurrence ofsubcutaneous blebs or hematomas on the fetus, and hemorrhage in the affected structure early inthe pathogenetic sequence, features that are usually readily apparent if sought early followingexposure. 203,213–215,217,224,225 Moreover, the ability to induce defects in structures at a period ingestation beyond the normal period of susceptible development of that structure is also a goodindication that pathogenesis of the defects involves disruption.VI. OVERVIEW AND PERSPECTIVEAny discussion of the mechanisms that are operative and responsible for congenital malformationsfollowing exposure to toxic xenobiotics, including drugs, other chemicals, and physical agents,must include some aspects of the genetic control of normal development as well as the geneticchanges that are responsible for abnormal development. In the past, we studied the genetics ofabnormal development by studying genetic alterations that occurred spontaneously. But with theadvances in molecular biology resulting in deciphering the human genome and many animalgenomes, scientists have the ability to alter the genome of experimental animals to study the impactof individual genes or growth factors on development. It is not surprising that advances in geneticshave assisted teratological investigations in many ways. We have known for a long time thatenvironmental toxicants that are teratogenic can alter development and mimic some aspects ofgenetic malformations. In fact, clinicians have used the terms phenocopy and phenotype for almosta century to indicate that some environmentally produced malformations and genetic malformationscannot be distinguished from each other. Many genetic congenital malformations are due to anabnormality at one locus that results in a cascade of abnormal molecules or abnormal quantitiesof normal molecules that can affect many structures. Thus, genetic flaws can result in a singleisolated anatomical defect or a recognizable syndrome of malformations. Environmental toxicantsare less likely to produce a single malformation and more likely to result in a so-called teratogensyndrome. What is responsible for this difference between the effect of teratogens and genemutations on development?Joseph Warkany referred to teratogens as sledgehammer toxicants. In other words, their mechanismof action was due to their toxicity. Of course this label does not apply to the mechanismsof all teratogens. Warkany was probably referring to the many teratogens that are cytotoxic, e.g.,ionizing radiation, many cancer chemotherapy drugs, and many mutagenic drugs and chemicals.Geneticists may look upon teratology as an unsophisticated methodology to produce congenitalmalformation compared to the preparation of knockout mice or site-directed mutagenesis. Butunfortunately, human malformations can be produced by sledgehammer teratogens as well as byinherited genes, chromosome abnormalities, and new mutations that affect development. Onlycertain classes of teratogens can come close to the specificity of effect of an abnormal gene ondevelopment. The best examples are the receptor mediated teratogens, e.g., retinoic acid, sexsteroids, and possibly thalidomide. If environmental toxicants produce malformations that affect© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 81the genome, in most cases it will be only indirectly. And some teratogens, such as deformationsand vascular disruptive phenomena, are unrelated to any genetic alteration.The methods of molecular biology are used to study the structure and nature of the genes, andtheir products that result in congenital malformations. After we achieve an understanding of thebasic science of genes that produce congenital malformations, there is frequently a long and difficultroad before this information can be utilized to treat or prevent birth defects.Developmental toxicology is much different. If we can identify an environmental agent as ateratogen, our first task is to prevent exposure to the agent. In many instances that is a readilyattainable goal, especially if the agent is a drug or chemical over which society has control. Yetknowing that alcohol is a teratogen does not and has not solved the problem. Similarly environmentalcontaminants, such as mercury, are not readily controlled. Maternal disease states that are teratogenic,such as diabetes or teratogenic infections, represent more difficult problems, but they arestill solvable. The answer for the teratologist is epidemiological studies that identify developmentaltoxicants or animal studies that indicate or confirm the potential for harm. From the humanstandpoint, understanding the pathophysiology is irrelevant. Yet, as scholars it is important to studyand understand the mechanisms of teratogenesis and the pathophysiology from a scientific andclinical perspective. Drugs and chemicals with similar pathophysiological effects may represent arisk. Greater knowledge might thus permit chemists to prepare drugs that have the therapeuticbenefit sought, while eliminating the teratogenic risk, as has been attempted with valproic acidanalogs.The approach for solving the problem of genetically transmitted congenital malformations ismuch different. Basic science research into the nature of the normal and abnormal genes involvedin the genetically transmitted malformations will be the key to preventing or treating the geneticdefect. Since there is a cascade of gene products that can be unveiled once the gene is identified,understanding the totality of a particular gene’s role in development may permit the replacementof the deficient products attributable to the abnormal gene. There is even the possibility of insertingthe normal gene; although presently largely theoretical, it is a real possibility.So developmental toxicologists have to examine their field and recognize that it consists of thefollowing components:Epidemiological studies to identify agents that are causally associated with the production of birthdefects.Basic science research dealing with the mechanisms of teratogenesis and the pathophysiology ofabnormal development.Ecological interests to identify potential new environmental toxicants.Social and political actions to make certain that the information about environmental reproductive risksis acted upon promptly by governmental agencies.Responsibility for educating the public about environmentally induced birth defects as well as toeducate their scientific, clinical, and regulatory colleagues about these risks.REFERENCES1. Brent, R.L. and Jensh, R.P., Intrauterine growth retardation, Adv. Teratol., 2, 139, 1967.2. Wilson, J.G. Environment and Birth Defects, Academic Press, New York, 1973.3. Brent, R.L., Environmental factors: miscellaneous, in Prevention of Embryonic, Fetal and PerinatalDisease, Brent, R.L. and Harris, M.J., Eds., DHEW Pub. No. (NIH) 76-853, Department of Health,Education, and Welfare, Bethesda, 1976, p. 211.4. Brent, R.L., The magnitude of the problem of congenital malformations, in Prevention of Physicaland Mental Congenital Defects. Part A. Basic and Medical Science, Marois, M., Ed., Alan R. Liss,New York, 1985, p. 55.© 2006 by Taylor & Francis Group, LLC

82 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION5. Spranger, J., Benirschke, K., Hall, J.G., Lenz, W., Lowry, R.B., Opitz, J.M., Pinsky, L., Schwarzacher,H.G., and Smith, D.W., Errors of morphogenesis: concepts and terms. Recommendations of aninternational working group, J. Pediatr., 100, 160, 1982.6. Wilson, J.G. and Brent, R.L., Are female sex steroids teratogenic?, Am. J. Obstet. Gynecol., 114, 567,1981.7. Brent, R.L., Bendectin: Review of the medical literature of a comprehensively studied human nonteratogenand the most prevalent tortogen-litigen, Reprod. Toxicol., 9, 337, 1995.8. Brent, R.L., Review of the scientific literature pertaining to the reproductive toxicity of Bendectin, inModern Scientific Evidence: The Law and Science of Expert Testimony, Faigman, D.L., Kaye, D.H.,Saks, M.J., and Saunders, J.W., Eds., West Publishing Group, St. Paul, 1997, p. 373.9. Van Allen, M.I., Structural anomalies resulting from vascular disruption, Pediatr. Clin. N. Amer., 39,160, 1992.10. Coplen, D.E., Prenatal intervention for hydronephrosis, J. Urol., 157, 2270, 1997.11. Fraser, F.C. and Fainstat, T.D., Production of congenital defects in the offspring of pregnant micetreated with cortisone, Pediatrics, 8, 527, 1951.12. Fraser, F.C., Kalter, H., Walker, B.E., and Fainstat, T.D., The experimental production of cleft palatewith cortisone and other hormones, J. Cell. Comp. Physiol., 43, 237, 1954.13. Khoury, M.J., Moore, C.A., and Evans, J.A., On the use of the term “syndrome” in clinical geneticsand birth defects epidemiology, Am. J. Med. Genet., 49, 26, 1994.14. Lubinsky, M., Vater and other associations: historical perspectives and modern interpretations, Am. J.Med. Genet. Suppl., 2, 9, 1986.15. Brent, R.L. and Holmes, L.B., Clinical and basic science lessons from the thalidomide tragedy: Whathave we learned about the causes of limb defects? Teratology, 38, 241, 1988.16. Lenz, W., A short history of thalidomide embryopathy, Teratology, 38, 203, 1988.17. Jensh, R.P. and Brent, R.L., The effect of low level prenatal x-irradiation on postnatal growth in theWistar rat, Growth Dev. Aging, 52, 53, 1988.18. Jensh, R.P. and Brent, R.L., The effects of prenatal x-irradiation in the 14th-18th days of gestationon postnatal growth and development in the rat, Teratology, 38, 431, 1988.19. Wilson, J.G. and Brent, R.L., Differentiation as a determinant of the reaction of rat embryos to X-irradiation, Proc. Soc. Exp Biol. Med., 82, 67, 1953.20. Generoso, W.M., Rutledge, J.C., Cain, K.T., Hughes, L.A., and Downing, D.J., Mutagen-induced fetalanomalies and death following treatment of females within hours of mating, Mutat. Res., 199, 175,1988.21. Pampfer, S. and Streffer, C., Prenatal death and malformations after irradiation of mouse zygotes withneutrons or X-rays, Teratology, 37, 599, 1988.22. Brent, R.L., The indirect effect of irradiation on embryonic development. II. Irradiation of the placenta,Am. J. Dis. Child., 100, 103, 1960.23. Wilkins, L., Jones, H.W., Holman, G.H., and Stempfel, R.S., Masculinization of the female fetusassociated with the administration of oral and intramuscular progestins during gestation: nonadrenalfemale psuedohermaphrodism, J. Clin. Endrocrinol. Metab., 18, 559, 1958.24. Wilkins, L., Masculinization due to orally given progestins, JAMA, 172, 1028, 1960.25. Grumbach, M.M., Ducharine, J.R., and Moloshok, R.E., On the fetal masculinizing action of certainoral progestins, J. Clin. Endrocrinol. Metab., 19, 1369, 1959.26. Van Wyk, J. and Grumbach, M.M., Disorders of sex differentiation, in Textbook of Endocrinology,Williams, R.H., Ed., W.B. Saunders, Philadelphia, 1968, p. 537.27. Briggs, M.H. and Briggs, M., Sex hormone exposure during pregnancy and malformations, inAdvances in Steroid Biochemistry and Pharmacology, Briggs, M.H. and Corbin, A., Eds., AcademicPress, London, 1979, p. 51.28. Seegmiller, R.E., Nelson, B.W., and Johnson, C.K., Evaluation of the teratogenic potential of Delalutin(17a-hydroxyprogesterone caproate) in mice, Teratology, 28, 201, 1983.29. Hochner-Celinkier, D., Marandici, A., Iohan, F. and Monder, C., Estrogen and progesterone receptorsin the organs of prenatal Cynomolgus monkey and laboratory mouse, Biol. Reprod., 35, 633, 1986.30. Carbone, J.P., Figurska, K., Buck, S., and Brent, R.L., Effect of gestational sex steroid exposure onlimb development and endochondral bone ossification in the pregnant C57B1/6J mouse. I. Medroxyprogesteroneacetate, Teratology, 42, 121, 1990.© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 8331. Carbone, J.P. and Brent, R.L., Genital and non-genital teratogenesis of prenatal progesterone therapy:the effects of 17 alpha-hydroxyprogesterone caproate on embryonic and fetal development and endochondralbone ossification in the C57B1/6J mouse, Am. J. Obstet. Gynecol., 169, 1292, 1993.32. Guignard, J.P., Burgener, F., and Calame, A., Persistent anuria in neonate: a side effect of captopril,Int. J. Pediatr. Nephrol., 2, 1981.33. Guignard, J.P., Effect of drugs on the immature kidney, Adv. Nephrol. Neker Hosp., 22, 193, 1993.34. Hanssens, M., Keirse, M.J.N.C., Vankelecom, F., and Van Assche, F.A., Fetal and Neonatal effects oftreatment with angiotensin-converting enzyme inhibitors in pregnancy, Obstet. Gynecol., 79, 128,1991.35. Piper, J.M., Ray, W.A., and Rosa, F.W., Pregnancy outcome following exposure to angiotensinconverting enzyme inhibitors, Obstet. Gynecol., 80, 429, 1992.36. Martin, R.A., Jones, K.L., Mendoza, A., Barr, M., and Benirschke, K., Effect of ACE inhibition inthe fetal kidney: decreased renal blood flow, Teratology, 46, 317, 1992.37. Pryde, P.G., Sedman, A.B., Nugent, C.E., and Barr, M., Angiotensin-converting enzyme inhibitorfetopathy, J. Am. Soc. Nephrol., 3, 1575, 1993.38. Brent, R.L. and Beckman, D.A., Angiotensin-converting enzyme inhibitors, an embryopathic class ofdrugs with unique properties: information for clinical teratology counselors, Teratology, 43, 543, 1991.39. Beckman, D.A. and Brent, R.L., Teratogenesis: alcohol, angiotensin converting enzyme inhibitors,and cocaine, Curr. Opin. Obstet. Gynec., 2, 236, 1990.40. Brent, R.L., Editorial comment: definition of a teratogen and the relationship of teratogenicity tocarcinogenicity, Teratology, 34, 359, 1986.41. Francis, B.M., Rogers, J.M., Sulik, K.K., Alles, A.J., Elstein, K.H., Zucker, R.M., Massaro, E.J.,Rosen, M.B., and Chernoff, N., Cyclophosphamide teratogenesis: evidence for compensatory responseto induced cellular toxicity, Teratology, 42, 473, 1990.42. Rogers, J.M., Francis, B.M., Sulik, K.K., Alles, A.J., Massaro, E.J., Zucker, R.M., Elstein, K.H.,Rosen, M.B., and Chernoff, N., Cell death and cell cycle perturbation in the developmental toxicityof the demethylating agent 5-aza-2′-deoxycytidine, Teratology, 50, 332, 1994.43. Beckman, D.A. and Brent, R.L., Mechanism of known environmental teratogens: drugs and chemicals,in Clinics in Perinatology, Brent, R.L. and Beckman, D.A., Eds., W.B. Saunders, Philadelphia, 1986,p. 649.44. Mirkes, P.E., 2001 Warkany Lecture: To die or not to die, the role of apoptosis in normal and abnormaldevelopment, Teratology, 65, 228, 2002.45. Knudsen, T.B., Cell death, in Drug Toxicity in Embryonic Development., Kavlock, R.J. and Daston,G.P., Eds., Springer-Verlag, New York, 1997, p. 211.46. Zakeri, Z.F. and Ahuja, H.S., Cell death/apoptosis: normal, chemically induced, and teratogenic effect,Mutation Res., 396, 149, 1997.47. Moallem, S.A. and Hales, B.F., The role of p53 and cell death by apoptosis and necrosis in 4-hydroperoxycyclophosphamide-induced limb malformations, Development, 125, 3225, 1998.48. Umpierre, C.C., Little, S.A., and Mirkes, P.E., Co-localization of active caspase-3 and DNA fragmentation(TUNEL) in normal and hyperthermia-induced abnormal mouse development, Teratology, 63,134, 2001.49. Gao, X., Blackburn, M.R., and Knudsen, T.B., Activation of apoptosis in early mouse embryos by 2′-deoxyadensosine exposure, Teratology, 49, 1, 1994.50. Dunty, W.C., Chen, S.-Y., Zucker, R.M., Dehart, D.B., and Sulik, K.K., Selective vulnerability ofembryonic cell populations to ethanol-induced apoptosis: implications for alcohol related birth defectsand neurodevelopmental disorder, Alcohol Clin. Exp. Res., 25, 1523, 2001.51. Thayer, J.M. and Mirkes, P.E., PCD and N-acetoxy-2acetylamineofluorene-induced apoptosis in therat embryo, Teratology, 51, 418, 1995.52. Mirkes, P.E. and Little, S.A., Teratogen-induced cell death in postimplantation mouse embryos:differential tissue sensitivity and hallmarks of apoptosis, Cell Death Differ., 5, 592, 1998.53. Mirkes, P.E. and Little, S.A., Cytochrome c release from mitochondria of early postimplantationmurine embryos exposed to 4-dydroperoxycyclophosphamide, heat shock and staurosporine, Toxicol.Appl. Pharmacol., 162, 197, 2000.54. Little, S.A. and Mirkes, P.E., Caspase-3-induced augmentation of the intrinsic, mitochondrial apoptoticpathway, Teratology, 63, 248, 2001.© 2006 by Taylor & Francis Group, LLC

84 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION55. Torchinsky, A., Ivinisky, I., Savion, S., Shepshelovich, J., Gorivodsky, M., Fein, A., Carp, H., Schwartz,D., Frankel, J., Rotter, V., and Toder, V., Cellular events and the patterns of p53 protein expressionfollowing cyclophosphamide-initiated cell death in various organs of the developing embryo, TeratogenCarcinogen Mutagen., 19, 353, 1999.56. Little, S.A. and Mirkes, P.E., Effects of 4-hydroperoxycyclophosphamide (4-OOH-CP) and r-hydroperoxydechlorocyclophosphamide(4-OOh-deCICP) on the cell cycle of post implantation rat embryos,Teratology, 45, 163, 1992.57. Li, L.H., Olin, J., Fraser, T.J., and Bhuyan, B.K., Phase specificity of 5-azacytadine against mammaliancells in tissue culture, Cancer Res., 30, 2770, 1970.58. Sulik, K.K., Cook, C.S., and Webster, W.S., Teratogens and craniofacial malformations: Relationshipsto cell death, Development, 103, 213, 1988.59. Sulik, K.K. and Dehart, D.B., Retinoic acid-induced limb malformations resulting from apical ectodermalridge cell death, Teratology, 37, 527, 1988.60. Alles, A.J. and Sulik, K.K., Retinoic acid-induced limb-reduction defects: perturbation of zones ofprogrammed cell death as a pathogenic mechanism, Teratology, 40, 163, 1989.61. Alles, A.J. and Sulik, K.K., Retinoic acid-induced spina bifida: evidence for a pathogenic mechanism,Development, 108, 73, 1990.62. Saunders, J.W., Gasseling, M.T., and Saunders, L., Cellular death in morphogenesis of the avian wing,Dev. Biol., 5, 147, 1961.63. Pexieder, T., Cell death in the morphogenesis and teratogenesis of the heart, Adv. Anat. Embryol. CellBiol., 51, 1, 1975.64. Abrams, J.M., White, K., Fessler, L.I., and Steller, H., Programmed cell death during drosophilaembryogenesis, Development, 117, 29, 1993.65. Ansari, B., Coates, P.J., Greenstein, B.D., and Hall, P.D., In situ end-labeling detects DNA strandbreaks in apoptosis and other physical states, J. Pathol., 170, 1, 1993.66. Nau, H. and Hendrickx, A.G., Valproic acid teratogenesis: ISI atals of science, Pharmacology, 1, 52, 1987.67. Martin, M.L. and Regan, C.M., The anticonvulsant valproate teratogen restricts the glial cell cycle ata defined point in the mid-G1 phase, Brain Res., 554, 223, 1991.68. Craig, J.C., Bennett, G.D., Miranda, R.C., Mackler, S.A., and Finnell, R.H., Ribonucleotide reductasesubmit R1: a gene conferring sensitivity to valproic acid-induced neural tube defects in mice, Teratology,61, 305, 2000.69. Regan, C.M., Gorman, A.M., Larsson, O.M., Maguire, C., Martin, M.L., Schousboe, A., and Williams,D.C., In vitro screening for anticonvulsant-induced teratogenesis in neural primary cultures and celllines, Int. J. Dev. Neurosci., 8, 103, 1990.70. Miller, M.W. and Nowakowski, R.S., Use of bromodeoxyuridine-immunohistochemistry to examinethe proliferation, migration and time of origin of cells of the central nervous system, Brain Res., 457,44, 1988.71. Rosen, M.B. and Chernoff, N., 5-Aza-2′-deoxycytidine-induced cytotoxicity and limb reductiondefects in the mouse, Teratology, 65, 180, 2002.72. Chernoff, N., Rogers, J.M., Alles, A.J., Zucker, R.M., Elstein, K.H., Massaro, E.J., and Sulik, K.K.,Cell cycle alterations and cell death in cyclophosphamide teratogenesis, Teratogen CarcinogenMutagen,. 9, 199, 1989.73. Walterhouse, D.O., Yoon, J., and Iannaccone, P.M., Developmental pathways: Sonic hedgehogpatched-GLl,Environ. Hlth. Perspect., 107, 167, 1999.74. Roessler, E., Belloni, E., Gaudenz, K., Jay, P., Berta, P., Scherer, S.W., Tsui, L.C., and Muenke, M.,Mutations in the human Sonic hedgehog gene cause holoprosencephaly, Nat. Genet., 14, 357, 1996.75. Roessler, E., Belloni, E., Gaudenz, K., Vargas, F., Scherer, S.W., Tsui, L.C., and Muenke, M., Mutationsin the C-terminal domain of Sonic Hedgehog cause holoprosencephaly, Hum. Mol. Genet., 6, 1847, 1997.76. Wallis, D.E. and Muenke, M., Molecular mechanisms of holoprosencephaly, Mol. Genet. Metab., 68,126, 1999.77. Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H., and Beachy, P.A., Cyclopiaand defective axial patterning in mice lacking Sonic hedgehog gene function, Nature, 383, 407, 1996.78. Kang, S., Graham, J.M., Jr., Olney, A.H., and Biesecker, L.G., GLI3 frameshift mutations causeautosomal dominant Pallister-Hall syndrome, Nat. Genet., 15, 266, 1997.© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 8579. Wild, A., Kalff-Suske, M., Vortkamp, A., Bornholdt, D., Konig, R., and Grzeschik, K.H., Pointmutations in human GL13 cause Greig syndrome, Hum. Mol. Genet., 6, 1979, 1997.80. Hui, C.C. and Joyner, A.L., A mouse model of Greig cephalopolysyndactyly syndrome: the extra-toesJ mutation contains an intragenetic deletion of the Gli3 gene, Nat. Genet., 3, 241, 1993.81. Bumcrot, D.A. and McMahon, A.P., Sonic signals somites, Current Biology, 5, 612, 1995.82. Roberts, D.J., Johnson, R.L., Burke, A.C., Nelson, C.E., Morgan, B.A., and Tabin, C.J., Sonichedgehog is an endodermal signal inducing BMP-4 and Hox genes during induction and regionalizationof the chick hindgut, Development, 121, 3163, 1995.83. Motoyama, J., Liu, J., Mo, R., Ding, Q., Post, M., and Hui, C.C., Essential function of Gli2 and Gli3in the formation of lung, trachea, and oesophagus, Nat. Genet., 20, 54, 1998.84. Litingtung, Y., Lei, L., Westphal, H., and Chiang, C., Sonic hedgehog is essential to foregut development,Nat. Genet., 20, 58, 1998.85. Pearse, R.V. and Tabin, C.J., The molecular ZPA, J. Exp. Zool., 282, 677, 1998.86. Wolpert, L., Vertebrate limb development and malformations, Ped. Res., 46, 247, 1999.87. Warburton, D., Schwarz, M., Tefft, D., Flores-Delgado, G., Anderson, K.D., and Cardoso, W.V., Themolecular basis of lung morphogenesis, Mechan. Dev., 92, 55, 2000.88. Makarenkova, H. and Patel, K., Gap junction signaling mediated through connexin-43 is required forchick limb development, Dev. Biol., 207, 380, 1999.89. Coelho, C.N. and Kosher, R.A., A gradient of gap junctional communication along the anteriorposterioraxis of the developing chick limb bud, Dev. Biol. 148, 529, 1991.90. Coelho, C.N. and Kosher, R.A., Gap junctional communication during limb cartilage differentiation,Dev. Biol., 144, 47, 1991.91. Parr, B.A. and McMahon, A.P., Dorsalizing signal Wnt-7a is required for normal polarity or D-V andA-P axes of mouse limb, Nature, 374, 350, 1995.92. Cygan, J.A., Johnson, R.L., and McMahon, A.P., Novel regulatory interactions revealed by studies ofmurine limb pattern in Wnt-7a and En-1 mutants, Development Supplement 124, 5021, 1997.93. Riddle, R.D., Johnson, R.L., Laufer, E., and Tabin, C.J., Sonic hedgehog mediates the polarizingactivity of the ZPA, Cell, 75, 1401, 1993.94. Francis, P.H., Richardson, M.K., Brickell, P.M., and Tickle, C., Bone morphogenic proteins and asignaling pathway that controls patterning in the developing chick limb, Development 120, 209, 1994.95. Wada, N., Kawakami, Y., and Nohno, T., Sonic hedgehog signaling during digit pattern duplicationafter application of recombinant protein and expressing cells, Dev. Growth Differen. 41, 567, 1999.96. Chiang, C., Litingtung, Y., Harris, M.P., Simandl, B.K., Li, Y., Beachy, P.A., and Fallon, J.F., Manifestationof limb prepattern: limb development in the absence of Sonic Hedgehog function, Dev. Biol.236, 421, 2001.97. Litingtung, Y., Dahn, R.D., Li, Y., Fallon, J.F., and Chiang, C., Shh and Gli3 are dispensable for limbskeleton formation but regulate digit number and identity, Nature, 418, 979, 2002.98. Hara, K., Kimura, J., and Ide, H., Effects of FGFs on the morphogenic potency of the AER-maintenanceactivity of cultured progress zone cells of the chick limb bud, Int. J. Dev. Biol. 42, 591, 1998.99. Laufer, E., Nelson, C.E., Johnson, R.L., Morgan, B.A., and Tabin, C.J., Sonic hedgehog and FGF-4act through a signaling cascade and feedback loop to integrate growth and patterning of the developinglimb bud, Cell, 79, 993, 1994.100. Niswander, L., Jeffrey, S., Martin, G., and Tickle, C., Positive feedback signaling in vertebrate limbdevelopment, Nature, 371, 609, 1994.101. Tickle, C., Vertebrate limb development, Curr. Opin. Gen. Dev., 5, 478, 1995.102. Summerbell, D., A quantitative analysis of the effect of excision of the AER from the chick limb bud,J. Embryol. Morphol., 32, 651, 1974.103. Johnson, R.L., Riddle, R.D., and Tabin, C.J., Mechanisms of limb patterning, Curr. Opin. Gen. Dev.,4, 535, 1994.104. Wolpert, L., Tickle, C., and Sampford, M., The effect of cell killing by x-irradiation on patternformation in the chick limb, J. Embryol. Exp. Morphol., 50, 175, 1979.105. Nelson, C.E., Morgan, B.A., Burke, A.C., Laufer, E., DiMambro, E., Murtaugh, L.C., Gonzales, E.,L., T., Parada, L.F., and Tabin, C.J., Analysis of Hox gene expression in the chick limb bud, DevelopmentSupplement, 122, 1449, 1996.© 2006 by Taylor & Francis Group, LLC

86 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION106. Goff, D.J. and Tabin, C.J., Analysis of Hoxd-13 and Hoxd-11 misexpression in chick limb buds revealthat Hox genes affect both bone condensation and growth, Development Supplement 124, 627, 1997.107. Bryden, M.M., Evans, H.E., and Keeler, R.F., Cyclopia in sheep caused by plant teratogens, J. Anat.110, 507, 1971.108. Cooper, M.K., Porter, J.A., Young, K.E., and Beachy, P.A., Teratogen-mediated inhibition of targettissue response to Shh signaling, Science, 280, 1603, 1998.109. Niederreither, K., Vermot, J., Schuhbaur, B., Chambon, P., and Dolle, P., Embryonic retinoic acidsynthesis is required for forelimb growth and anteroposterior patterning in the mouse, Development,129, 3563, 2002.110. Summerbell, D., The effect of local application of retinoic acid to the anterior margin of the developingchick limb, J. Embryol. Exp. Morphol., 78, 269, 1983.111. Hendrickx, A.G., Teratogenicity of retinoids in rodents, primates and humans, J. Toxicol. Sci., 23,272, 1988.112. Power, S.C., Lancman, J., and Smith, S.M., Retinoic acid is essential for Shh/Hox signaling duringrat limb outgrowth but not for limb initiation, Dev. Dyn., 216, 469, 1999.113. Stratford, Abnormal anteroposterior and dorsoventral patterning of the limb bud in the absence ofretinoids, Mech. Dev., 81, 115, 1999.114. Tickle, C., Alberts, B., Wolpert, L., and Lee, J., Local application of retinoic acid to the limb budmimics the action of the polarizing region, Nature, 296, 564, 1982.115. Wanek, N., Gardiner, D.M., Muneoka, K., and Bryant, S.V., Conversion by retinoic acid of anteriorcells into ZPA cells in the chick wing bud, Nature, 350, 81, 1991.116. Bell, S.M., Schreiner, C.M., and Scott, W.J., Jr., Disrupting the establishment of polarizing activityby teratogen exposure, Mech. Dev., 88, 147, 1999.117. Kwasigroch, T.E. and Bullen, M., Effects of isotretinoin (13-cis-retinoic acid) on the development ofmouse limbs in vivo and in vitro, Teratology, 44, 605, 1991.118. Lammer, E.J., Chen, D.T., Hoar, R.M., Agnish, N.D., Benke, P.J., Braun, J.T., Curry, C.J., Fernhoff,P.M., Grix, A.W., Lott, I.T., Richard, J.M., and Sun, S.C., Retinoic acid embryopathy, N. Engl. J.Med., 313, 837, 1985.119. Rosa, F.W., Wilk, A.L., and Kelsey, F.O., Teratogen update: vitamin A cogeners, Teratology, 33, 355,1986.120. Kochhar, D.M., Limb development in mouse embryos. I. Analysis of teratogenic effects of retinoicacid, Teratology, 7, 289, 1973.121. Fantel, A.G., Shepard, T.H., Newell-Morris, L.L., and Moffett, B.C., Teratogenic effects of retinoicacid in pigtail monkeys (Macaca nemestrina). 1. General features, Teratology, 15, 65, 1976.122. Kochhar, D.M., Penner, J.D., and Tellone, C.J., Comparative teratogenic activities of two retinoids:effects of palate and limb development, Teratogen Carcinogen Mutagen., 4, 377, 1984.123. Webster, W.S., Johnston, M.C., Lammer, E.J., and Sulik, K.K., Isotretinoin embryopathy and thecranial neural crest: an in vivo study, J. Craniofacial Genet. Dev. Biol., 6, 211, 1986.124. Rizzo, R., Lammer, E.J., Parano, E., Pavone, L., and Argyle, J.C., Limb reduction defects in humansassociated with prenatal isotretinoin exposure, Teratology, 44, 499, 1991.125. Kwasigroch, T.E., and Kochhar, D.M., Production of congenital limb defects with retinoic acid:phenomenological evidence of progressive differentiation during limb morphogenesis, Anat. Embryol.,161, 105, 1980.126. Rutledge, J.C., Shourbaji, A.G., Hughes, L.A., Polifka, J.E., Cruz, Y.P., and Bishop, J.B., Limb andlower body duplications induced by retinoic acid in mice, Proc. Natl. Acad. Sci., USA, 91, 5436, 1994.127. Niederreither, K., Ward, S.J., Dolle, P., and Chambon, P., Morphological and molecular characterizationof retinoic acid-induced limb duplications in mice, Dev. Biol., 176, 185, 1996.128. Cusic, A.M. and Dagg, C.P., Spontaneous and retinoic acid-induced postaxial polydactyly in mice,Teratology, 31, 49, 1985.129. Kwasigroch, T.E., Vannoy, J.F., Church, J.K., and Skalko, R.G., Retinoic acid enhances and depressesin vitro development of cartilaginous bone anlagen in embryonic mouse limbs, Dev. Biol., 22, 150,1986.130. Maden, M., Ong, D.E., Summerbell, D., and Chytil, F., The role of retinoid-binding proteins in thegeneration of pattern in the developing limb, the regenerating limb and the nervous system, Dev.Suppl., 107, 109, 1989.© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 87131. Tickle, C., Crawley, A., and Farrar, J., Retinoic acid applications to chick wing buds leads to a dosedependentreorganization of the apical ectodermal ridge that is mediated by the mesenchyme, Development,106, 691, 1989.132. Dolle, P., Ruberte, E., Leroy, P., Morris-Kay, G., and Chambon, P., Retinoic acid receptor and cellularretinoid binding proteins. I. A systematic study of their differential pattern of transcription duringmouse organogenesis, Dev. Suppl., 110, 1133, 1990.133. Vaessen, M.J., Meijers, J.H., Bootsma, D., and Van Kessel, A.G., The cellular retinoic-acid-bindingprotein is expressed in tissues associated with retinoic-acid-induced malformations, Dev. Suppl., 110,371, 1990.134. Wilkinson, D.G. In Situ Hybridization: A Practical Approach, IRL Press, Oxford, 1992.135. Nieto, M.A., Patel, K., and Wilkinson, D.G., In situ hybridization analysis of chick embryos in wholemount and tissue sections, Methods Cell Biol., 51, 219, 1996.136. Lowe, L.A. and Kuehn, M.R., Whole-mount in situ hybridization to study gene expression duringmouse development, in Methods in Molecular Biology, Vol 137: Developmental Biology Protocols,Tuan, R.S., and Lo, C.W., Eds., Humana Press, Totowa, NJ, 1997, p. 125.137. Tuan, R.S., mRNA and protein co-localization on tissue sections by sequential, colorimetric in situhybridization and immunohistochemistry, in Methods in Molecular Biology, Vol 137: DevelopmentalBiology Protocols, Tuan, R.S. and Lo, C.W., Eds., Humana Press, Totowa, NJ, 1997, p. 117.138. Nieto, M.A., The early steps of neural crest development, Mech. Dev., 105, 27, 2001.139. Trainor, P.A. and Krumlauf, R., Hox genes, neural crest and branchial arch patterning, Curr. Opin.Cell. Biol., 13, 698, 2001.140. Noden, D.M., Interactions and fates of avian craniofacial mesenchyme, Development, 108, 121, 1988.141. Chisaka, O., Musci, T.S., and Capecchi, M.R., Developmental defects of the ear, cranial nerves andhindbrain resulting from targeted disruption of the mouse homeobox gene Hox 1.6, Nature, 355, 516,1992.142. Santiago, A. and Erickson, C.A., Ephrin-B ligands play a dual role in the control of neural crest cellmigration, Development, 129, 3621, 2002.143. Krull, C.E., Segmental organization of neural crest migration, Mech. Dev., 105, 37, 2001.144. Menengola, E., Broccia, M.L., Prati, M., and Giavini, E., Morphological alterations induced by sodiumvalproate on somites and spinal nerves in rat embryos, Teratology, 59, 110, 1999.145. Gale, E., Prince, V., Lumsden, A., Clarke, J., Holder, N., and Maden, M., Late effects of retinoic acidon neural crest and aspects of rhombomere identity, Development, 122, 783, 1996.146. Gale, E., Zile, M., and Maden, M., Hindbrain respecification in the retinoid deficient quail, Mechan.Dev., 89, 43, 1999.147. Lee, Y.M., Osumi-Yamashita, N., Ninomiya, Y., Moon, C.K., Eriksson, U., and Eto, K., Retinoic acidstage-dependently alters the migration pattern and identity of neural crest cells, Development, 121,825, 1995.148. Wei, X., Makori, N., Peterson, P.E., Hummler, H., and Hendrickx, A.G., Pathogenesis of retinoic acidinducedear malformations in a primate model, Teratology, 60, 83, 1999.149. Nichols, D.H., Neural crest formation in the head of the mouse embryo as observed using a newhistological technique, J. Embryol. Exp. Morphol., 64, 105, 1981.150. Campbell, L.R., Dayton, D.H., and Sohal, G.S., Neural tube defects: A review of human and animalstudies on the etiology of neural tube defects, Teratology, 34, 171, 1986.151. Clouthier, D.E., Hosoda, K., Richardson, J.A., Williams, S.C., Yanagisawa, H., Kuwaki, T., Kumada,M., Hammer, R.E., and Yanagisawa, M., Cranial and neural crest defects in endothelin-A receptordeficient mice, Development, 125, 813, 1998.152. Clouthier, D.E., Williams, S.C., Yanasigawa, H., Wieduwilt, M., Richardson, J.A., and Yanagisawa,M., Signaling pathways crucial for craniofacial development revealed by endothelin A receptordeficient mice, Dev. Biol., 217, 10, 2000.153. Yanagisawa, H., Hammer, R.E., Richardson, J.A., Williams, S.C., Clouthier, D.E., and Yanagisawa,M., Role of endothelin-1/endothelin-A receptor mediated signaling pathway in the aortic arch patterningin mice, J. Clin. Invest. 102, 22, 1998.154. Abo, T. and Balch, C.M., A differentiation antigen of human NK and K cells identified by a monoclonalantibody (HNK-1), Immunology, 127, 1024, 1981.© 2006 by Taylor & Francis Group, LLC

88 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION155. Tucker, G.C., Aoyama, H., Lipinski, M., Turz, T., and Thiery, J.P., Identical reactivity of monoclonalantibodies HNK-1 and NC-1: Conservation in vertebrates on cells derived from neural primordiumand on some leukocytes, Cell Differ., 14, 223, 1984.156. Corbel, C., Lehmann, A., and Davison, F., Expression of CD44 during early development of the chickembryo, Mech. Dev., 96, 111, 2000.157. Maden, M., Horton, C., Graham, A., Leonard, L., Pizzey, J., and Eriksson, U., Domains of cellularretinoic-acid-bindingprotein 1 (CRABP l) expression in the hindbrain and neural crest of the mouseembryo, Mech. Dev., 37, 13, 1992.158. Eriksson, U., Hansson, E., Nordlinder, H., Busch, C., Sundelin, J., and Peterson, P.A., Quantitationand tissue localization of the cellular retinoic acid binding protein, J. Cell Physiol., 133, 482, 1987.159. Young, D.L., Schneider, R.A., Hu, D., and Helms, J.A., Genetic and teratogenic approaches tocraniofacial development, Crit. Rev. Oral Biol. Med., 11, 304, 2000.160. Fawcett, L.B., Buck, S.J., Beckman, D.A., and Brent, R.L., Is there a no-effect dose for corticosteroidinducedcleft palate? The contribution of endogenous corticosterone to the incidence of cleft palatein mice, Ped. Res., 39, 856, 1996.161. Hackman, R.M. and Brown, K.S., Corticosterone-induced isolated cleft palate in A/J mice, Teratology,6, 313, 1972.162. Kalter, H., The history of the A family of inbred mice and the biology of its congenital malformations,Teratology, 20, 213, 1979.163. Kerrigan, J.J., Mansell, J.P., Sengupta, A., Brown, N., and Sandy, J.R., Palatogenesis and potentialmechanisms for clefting, J.R. Coll. Surg. Edinb., 45, 351, 2000.164. Fraser, F.C., The multifactorial/threshold concept — uses and misuses, Teratology, 14, 267, 1976.165. Griffith, C.M. and Hay, E.D., Epithelial-mesenchymal transformation during palatal fusion: carboxyfluoresencein traces cells at light and electron microscopic levels, Development, 116, 1087, 1992.166. Shuler, C.F., Halpern, D.E., Guo, Y., and Sank, A.C., Medial edge epithelium fate traced by celllineage analysis during epithelial-mesenchymal transformation in vivo, Dev. Biol., 154, 318, 1992.167. Fitchett, J.E. and Hay, E.D., Medial edge epithelium transforms to mesenchyme after embryonicpalatal shelves fuse, Dev. Biol., 131, 455, 1989.168. Carette, M.J.M. and Ferguson, M.W.J., The fate of medial edge epithelial cells during palatal fusionin vitro: an analysis of Dil labeling and confocal microscopy, Development, 114, 379, 1992.169. Brunet, C.L., Sharpe, P.T., and Ferguson, M.W.J., Inhibition of TGF-beta 3 (but not TGF1 beta 1 orTGF-beta 2) activity prevents normal mouse embryonic palate fusion, Int. J. Dev. Biol., 39, 345, 1995.170. Proetzel, G., Pawlowski, S.A., Wiles, M.V., Yin, M., Bovin, G.P., Howles, P.N., Ding, J., Ferguson,M.W.J., and Doetschman, T., Transforming growth factor beta 3 is required for secondary palatefusion, Nat. Genet., 11, 409, 1995.171. Kaartinen, V., Cui, X.M., Heisterkamp, N., Goffen, J., and Shuler, C.F., Transforming growth factorbeta3 regulates transdifferentiation of medial edge epithelium during palatal fusion and associateddegradation of the basement membrane, Dev. Dyn., 209, 255, 1997.172. Harris, M.J. and Juriloff, D.M., Toward understanding mechanisms of genetic neural tube defects inmice, Teratology, 60, 292, 1999.173. Medical Research Council (MRC) Vitamin Study Research Group, Prevention of neural tube defects:Results of the Medical Research Council Vitamin Study, Lancet, 338, 131, 1991.174. Czeizel, A.F. and Dudas, I., Prevention of the first occurrence of neural tube defects by periconceptionalvitamin supplementation, Arch. Dis. Child., 56, 911, 1992.175. Smithells, R.W., Sheppard, S., and Schorah, C.J., Apparent prevention of neural tube defects bypericonceptional vitamin supplementation, Arch. Dis. Child., 56, 911, 1981.176. Goldman, I.D., The mechanism of action of methotrexate: I. Interaction with a low-affinity intracellularsite required for maximum inhibition of deoxyribonucleic acid synthesis in L-cell mouse fibroblasts,Mol. Pharmacol., 10, 257, 1974.177. DeSesso, J.M. and Goeringer, G.C., Methotrexate-induced developmental toxicity in rabbits is amelioratedby 1-(p-tosyl)-3,4,4-trimethylimidizoline, a functional analog for tetrahydrofolate-mediatedone-carbon transfer, Teratology, 45, 271, 1992.178. Warkeny, J. and Petering, H.G., Congenital malformations of the central nervous system in ratsproduced by maternal zinc deficiency, Teratology, 5, 319, 1972.© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 89179. Keen, C.L., Taubeneck, M.W., Daston, G.P., Rogers, J.M., and Gershwin, M.E., Primary and secondaryzinc deficiency as factors underlying abnormal CNS development, Ann. New York Acad. Sci., 678, 37,1993.180. Meiden, G.D., Keen, C.L., Hurley, L.S., and Klein, N.W., Effects of whole rat embryos cultured onserum from zinc- and copper-deficient rats, J. Nutrition, 116, 2424, 1986.181. Apgar, J., Effects of zinc deprivation from day 12, 15, or 18 of gestation on parturition in the rat, J.Nutrition, 102, 343, 1972.182. Hurley, L.S., Gowen, J., and Swenerton, H., Teratogenic effects of short-term and transitory zincdeficiency in rats, Teratology, 4, 199, 1971.183. Hurley, L.S., Teratogenic aspects of manganese, zinc, and copper nutrition, Physiol. Rev., 61, 249,1981.184. Essein, F.B., Maternal methionine supplementation promotes the remediation of axial defects in Axdmouse neural tube mutants, Teratology, 45, 205, 1992.185. Essein, F.B. and Wannberg, S.L., Methionine but not folinic acid or Vitamin B-12 alters the frequencyof neural tube defects in Axd mutant mice, J. Nutrition, 123, 27, 1993.186. Fawcett, L.B., Pugarelli, J.E., and Brent, R.L., Effects of supplemental methionine on antiseruminduceddysmorphology in rat embryos cultured in vitro, Teratology, 61, 332, 2000.187. Flynn, T.J., Friedman, L., Black, T.N., and Klein, N.W., Methionine and iron as growth factors forembryos cultured on canine serum, J. Exp. Zool., 244, 319, 1987.188. Coelho, C.N.D., Weber, J.A., Klein, N.W., Daniels, W.G., and Hoagland, T.A., Whole rat embryosrequire methionine for neural tube closure when cultured on cow serum, Teratology, 42, 437, 1989.189. Chambers, B.J., Klein, N.W., Nosel, P.G., Khairallah, L.H., and Romanow, J.S., Methionine overcomesneural tube defects in rat embryos cultured on sera from laminin-immunized monkeys, J. Nutrition,125, 1587, 1995.190. Coelho, C.N. and Klein, N.W., Methionine and neural tube closure in cultured rat embryos: morphologicaland biochemical analysis, Teratology, 42, 437, 1990.191. Nosel, P.G. and Klein, N.W., Methionine decreases the embryotoxicity of sodium valproate in the rat:in vivo and in vitro observations, Teratology, 46, 499, 1992.192. Cockcroft, D.L., Brook, F.A., and Copp, A.J., Inositol deficiency increases the susceptibility to neuraltube defects of genetically predisposed (curly tail) mouse embryos in vitro, Teratology, 45, 233, 1992.193. Greene, N.D. and Copp, A.J., Inositol prevents folate-resistant neural tube defects in the mouse, Nat.Med,. 3, 60, 1997.194. Van Straaton, H.W. and Copp, A.J., Curly tail: a 50-year history of the mouse spina bifida model,Anat. Embryol,. 203, 225, 2001.195. Mukherjee, M.D., Sandstead, H.H., Ratnaparkhi, M.V., Johnson, L.K., Milne, D.B., and Stelling, H.P.,Maternal zinc, iron, folic acid and protein nutriture and outcome of human pregnancy, Am. J. Clin.Nutr., 40, 496, 1984.196. Velie, E.M., Block, G., Shaw, G.M., Samuels, S.J., Schaffer, D.M., and Kulldorff, M., Maternalsupplemental and dietary zinc intake and the occurrence of neural tube defects in California, Am. J.Epidemiol., 150, 605, 1999.197. Shaw, G.M., Todoroff, K., Schaffer, D.M., and Selvin, S., Periconceptional nutrient intake and riskfor neural tube defect-affected pregnancies, Epidemiology, 10, 711, 1999.198. Bower, C., Stanley, F.J., and Spickett, J.T., Maternal hair zinc and neural tube defects: no evidenceof an association from a case control study in western Australia, Asia-Pacific J. Pub. Health, 6, 156,1992-19993.199. Milunski, A., Morris, J.S., Jick, H., Rothman, K.J., Ulcickas, M., Jick, S.S., Shoukimas, P., and Willett,W., Maternal zinc and fetal neural tube defects, Teratology, 46, 341, 1992.200. Shaw, G.M., Velie, E.M., and Schaffer, D.M., Is dietary intake of methionine associated with areduction in risk for neural tube defect-affected pregnancies? Teratology, 56, 295, 1997.201. Shoob, H.D., Sargent, R.G., Thompson, S.J., Best, R.G., Drane, J.W., and Tocharoen, A., Dietarymethionine is involved in the etiology of neural tube defect-affected pregnancies in humans, J. Nutr.,131, 2653, 2001.202. Woods, J.R. and Plessinger, M.A., Maternal-fetal cardiovascular system: a target of cocaine, NIDAResearch Monograph 108, 7, 1991.© 2006 by Taylor & Francis Group, LLC

90 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION203. Danielsson, B.R.G., Danielson, M., Reiland, S., Rundqvist, E., Denker, L., and Regard, C.G., Histologicaland in vitro studies supporting decreased uteroplacental blood flow as explanation for digitaldefects after administration of vasodilators, Teratology, 41, 185, 1990.204. Webster, W.S. and Brown-Woodman, P.D.C., Cocaine as a cause of congenital malformations ofvascular origin: experimental evidence in the rat, Teratology, 41, 689, 1990.205. Jones, K.L., Developmental pathogenesis of defects associated with prenatal cocaine exposure: fetalvascular disruption, Clin. Perinatol., 18, 139, 1991.206. Gingras, J.L., Weese-Mayer, D.E., Hume, R.F., and O’Donnell, K.J., Cocaine and development:mechanisms of fetal toxicity and neonatal consequences of prenatal cocaine exposure, Early Hum.Dev., 31, 1, 1992.207. Gonzalez, C.H., Vargas, F.R., Perez, A.B.A., Kim, C.A., Marques-Dias, M.J., Leone, C.R., Neto, J.C.,Llerena, J.C., and Cabral de Almeida, J.C., Limb reduction defects with or without mobius sequencein seven Brazilian children associated with misoprostol use in the first trimester of pregnancy, Amer.J. Med. Genet., 47, 59, 1993.208. Brent, R.L., Congenital malformation case reports: The editor’s and reviewer’s dilemma, Am. J. Med.Genet., 47, 872, 1993.209. Genest, D.R., Di Salvo, D., Rosenblatt, M.J., and Holmes, L.B., Terminal transverse limb defectswith tethering and omphalocele in a 17 week fetus following first trimester misoprostol exposure,Clin. Dysmorphol., 8, 53, 1999.210. Burton, B.K., Schultz, C.J., and Burd, L.I., Limb anomalies associated with chorionic villus sampling,Obstet. Gynecol., 79, 726, 1992.211. Holmes, L.B., Chorionic villus sampling and limb defects, Prog. Clin. Biol. Res., 383A, 409, 1993.212. Stoler, J.M., McGuirk, C.K., Lieberman, E., Ryan, L., and Holmes, L.B., Malformations reported inchorionic villus sampling exposed children: a review and analytic synthesis of the literature, Genet.Med., 1, 315, 1999.213. Brent, R.L. and Franklin, J.B., Uterine vascular clamping: New procedure for the study of congenitalmalformations, Science, 132, 89, 1960.214. Franklin, J.B. and Brent, R.L., The effect of uterine vascular clamping on the development of ratembryos three to fourteen days old, J. Morphol., 115, 273, 1964.215. Leist, K.H. and Grauwiler, J., Fetal pathology in rats following uterine-vessel clamping on day 14 ofgestation, Teratology, 10, 55, 1974.216. Danielson, M.K., Danielsson, B.R.G., Marchner, H., Lundin, M., Rundqvist, E., and Reiland, S.,Histopathological and hemodynamic studies supporting hypoxia and vascular disruption as explanationto phenytoin teratogenicity, Teratology, 46, 485, 1992.217. Danielsson, B.R.G., Danielson, M., Rundqvist, E., and Reiland, S., Identical phalangeal defectsinduced by phenytoin and nifedipine suggest fetal hypoxia and vascular disruption behind phenytointeratogenicity, Teratology, 45, 247, 1992.218. Fantel, A.G., Barber, C.V., Carda, M.B., Tumbic, R.W., and Mackler, B., Studies of the role ofischemia/reperfusion and superoxide anion radical production in the teratogenicity of cocaine, Teratology,46, 293, 1992.219. Zimmerman, E.F., Potturi, R.B., Resnick, E., and Fisher, E.J., Role of oxygen free radicals in cocaineinducedvascular disruption in mice, Teratology, 49, 192, 1994.220. Fantel, A.G. and Person, R.E., Further evidence for the role of free radicals in the limb teratogenicityof L-NAME, Teratology, 66, 24, 2002.221. Brent, R.L., Editorial Comment: Relationship between uterine vascular clamping, vascular disruptionsyndrome, and cocaine teratogenicity, Teratology, 41, 757, 1990.222. Chavez, G.F., Mulinare, J., and Coredero, J., Maternal cocaine use during early pregnancy as a riskfactor for congenital urogenital anomalies, JAMA, 262, 795, 1989.223. Gaily, E., Distal phalangeal hypoplasia in children with prenatal phenytoin exposure, Am. J. Med.Genet., 35, 574, 1990.224. Webster, W.S., Lipson, A.H., and Brown-Woodman, P.D.C., Uterine vascular trauma and limb defects,Teratology, 35, 253, 1987.225. Fawcett, L.B., Buck, S.J., and Brent, R.L., Limb reduction defects in the A/J mouse strain associatedwith maternal blood loss, Teratology, 58, 183, 1998.© 2006 by Taylor & Francis Group, LLC

PATHOGENESIS OF ABNORMAL DEVELOPMENT 91226. Menkes, B., Prelipceanu, O., and Capalnasan, I., Vital fluorochroming as a tool for embryonic celldeath research, in Advances in the Study of Birth Defects, Persaud, T.V.N., Ed., University Park Press,Baltimore, 1979, p. 219.227. Kohlbecker, A., Lee, A.E., and Schorle, H., Exencephaly in a subset of animals heterozygous for AP-2a mutation, Teratology, 65, 213, 2002.228. Ivnitski, I., Elmaoued, R., and Walker, M.K., 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCCD) inhibitionof coronary development is preceded by a decrease in myocyte proliferation and an increase in cardiacapoptosis, Teratology, 64, 201, 2001.229. Takagi, T.N., Matsui, K.A., Yamashita, K., Ohmori, H., and Yasuda, M., Pathogenesis of cleft palatein mouse embryos exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCCD), Teratogen CarcinogenMutagen., 20, 73, 2000.230. Slavkin, H., Nuckolls, G., and Shum, L., Craniofacial development and patterning, in Methods inMolecular Biology, Vol 137: Developmental Biology Protocols, Tuan, R.S. and Lo, C.W., Eds., HumanaPress, Totowa, NJ, 1997, p. 45.231. Abbott, B., Palatal dysmorphogenesis, palate organ culture, in Methods in Molecular Biology, Vol 137:Developmental Biology Protocols, Tuan, R.S. and Lo, C.W., Eds., Humana Press, Totowa, NJ, 1997, p. 195.232. DeLise, A.M., Stringa, E., Woodward, W.A., Mello, M.A., and Tuan, R.S., Embryonic limb mesenchymemicromass culture as an in vitro model for chondrogenesis and cartilage maturation, in Methodsin Molecular Biology, Vol 137: Developmental Biology Protocols, Tuan, R.S. and Lo, C.W., Eds.,Humana Press, Totowa, NJ, 1997, p. 359.233. Renault, J.Y., Caillaud, J.M., and Chevalier, J., Ultrastructural characterization of normal and abnormalchodrogenesis in micromass rat embryo limb bud cell cultures, Toxicol. Appl. Pharmacol., 130, 177, 1995.234. Flint, O.P., In vitro tests for teratogens: desirable endpoints, test batteries and current status of themicromass teratogen test, Reprod. Toxicol., 7 Suppl 1, 103, 1993.235. Van Maele-Fabry, G., Gofflot, F., Clotman, F., and Picard, J.J., Alterations of mouse embryonicbranchial nerves and ganglia induced by ethanol, Neurotoxicol. Teratol., 17, 497, 1995.236. Van Maele-Fabry, G., Clotman, F., Gofflot, F., Bosschaert, J., and Picard, J.J., Postimplantation mouseembryos cultured in vitro, assessment with whole-mount immunostaining and in situ hybridization,Int. J. Dev. Biol., 41, 365, 1997.237. Mark, M., Lufkin, T., Vonesch, V.L., Ruberte, E., Olivo, J.C., Dolle, P., Gorry, P., Lumsden, A., andChambon, P., Two rhombomeres are altered in Hoxa-1 mutant mice, Development, 119, 319, 1993.238. Krull, C.E., Collazo, A., Fraser, S.E., and Bronner-Fraser, M., Segmental migration of trunk neuralcrest: time-lapse analysis reveals a role for PNA-binding molecules, Development, 121, 3733, 1995.239. Bhattacharyya, A., Brackenbury, R., and Ratner, N., Axons arrest the migration of schwann cellprecusors, Development, 120, 1411, 1994.240. Serbedzija, G.N., Bronner-Fraser, M., and Fraser, S.E., A vital dye analysis of the timing and pathwaysof neural crest migration, Development, 106, 809, 1989.241. Stuckhardt, J.L. and Poppe, S.M., Fresh visceral examination of rat and rabbit fetuses used interatogenicity testing, Teratogen Carcinogen Mutagen., 4, 181, 1984.242. Wilson, J.G., Methods for administering agents and detecting malformations in experimental animals,in Tertology: Principles and Technique., Wilson, J.G. and Warkany, J., Eds., University of ChicagoPress, Chicago, 1965, p. 262.243. Barrow, M.V. and Taylor, W.J., A rapid method for detecting malformations in rat fetuses, J. Morphol.,127, 291, 1969.244. Astroff, A.B., Ray, S.E., Rowe, L.M., Hilbish, K.G., Linville, A.L., Stutz, J.P., and Breslin, W.J.,Frozen-sectioning yields similar results as traditional methods for fetal cephalic examination in therat, Teratology, 66, 77, 2002.245. Inouye, M., Differential staining of cartilage and bone in fetal mouse skeleton by alcian blue andalizerin red, S. Cong. Anom., 16, 171, 1976.246. Webb, G.N. and Byrd, R.A., Simultaneous differential staining of cartilage and bone in rodent fetuses:an alcian blue and alizarin red S procedure without glacial acetic acid, Biotechnic Histochem., 69,181, 1994.247. Kelly, W.L. and Bryden, M.M., A modified differential stain for cartilage and bone in whole mountpreparations of mammalian fetuses and small vertebrates, Stain Technol., 58, 131, 1983.© 2006 by Taylor & Francis Group, LLC

92 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION248. Kimmel, C.A. and Trammell, C., A rapid procedure for routine double staining of cartilage and bonein fetal and adult animals, Stain Technol., 56, 271, 1981.249. Whitaker, J. and Dix, K.M., Double staining technique for rat foetus skeletons in teratological studies,Laboratory Animals, 13, 309, 1979.250. Young, D.L., Phipps, D.E., and Astroff, A.B., Large-scale double-staining of rat fetal skeletons usingAlizarin Red S and Alcian Blue, Teratology, 61, 273, 2000.251. Zucker, R.M., Elstein, K.H., Shey, D.L., Ebron-McCoy, M., and Rogers, J.M., Utility of fluorescencemicroscopy in embryonic/fetal topographical analysis, Teratology, 51, 430, 1995.© 2006 by Taylor & Francis Group, LLC

CHAPTER 4Maternally Mediated Effects on DevelopmentRonald D. Hood and Diane B. MillerCONTENTSI. Introduction ..........................................................................................................................94A. Maternally Mediated Effects Defined .........................................................................94B. Historical Background.................................................................................................94C. Causes of Maternal Stress ...........................................................................................951. Neurogenic Stress ..................................................................................................962. Systemic Stress ......................................................................................................973. An Alternative Categorization of Stressors...........................................................97D. Developmental Hazard Associated with Maternal Stress or Toxicity ........................971. Supernumerary Ribs ..............................................................................................982. Other Variations and Malformations .....................................................................983. Prenatal Mortality and Growth Inhibition.............................................................99E. Concern about Maternally Mediated Effects ..............................................................991. Distinguishing Maternally Mediated from Direct Effects ....................................992. Influence on Developmental Toxicity Test Outcomes ........................................1003. Influence on Interpretation of Developmental Toxicity Test Results .................1014. Potentiation of Chemical Teratogenesis by Maternal Stress ..............................1025. Possible Influence on Human Pregnancy Outcomes ..........................................103II.III.IV.Possible Mechanisms for Maternally Mediated Effects....................................................103Assessment of Maternal Stress or Toxicity.......................................................................104A. Biochemical Measures...............................................................................................1041. Measures Related to the Stress Cascade.............................................................1042. Catecholamines....................................................................................................1063. Stress Proteins......................................................................................................1064. Measures Related to Zinc Metabolism ...............................................................107B. Organ Weights ...........................................................................................................108C. Behavioral Measures .................................................................................................108D. Measures of Toxicity .................................................................................................109E. Overview of Stress and Toxicity Assessment ...........................................................110Experimental Assessment of Possible Maternally Mediated Effects................................110A. Factors to Be Considered in Experimental Design and Conduct.............................1101. Controlling the Animal’s Environment ...............................................................11093© 2006 by Taylor & Francis Group, LLC

94 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION2. Measurement of Stress ........................................................................................1113. Appropriate Controls ...........................................................................................1114. Developmental Timing ........................................................................................1115. Consistency of Methodology...............................................................................1126. Outcome Assessment and Interpretation of Results ...........................................112B. Induction of Maternal Restraint Stress as a Model System .....................................113V. Overview ............................................................................................................................115References ......................................................................................................................................115A. Maternally Mediated Effects DefinedI. INTRODUCTIONIn developmental toxicology, maternally mediated effects on development are those adverse consequencesthat occur secondarily, as a result of some effect on the pregnant mother. They differfrom direct effects on the conceptus primarily in their immediate source, rather than their end result.Since likely mechanisms for maternally mediated effects are more limited in number than are directactingmechanisms, the range of consequences to the offspring may also be more limited. Nevertheless,as pointed out by Daston, 1 since there are several mechanisms by which maternally mediatedeffects may occur, it is unlikely that they would all result in only a single limited spectrum ofeffects on the offspring.The potential for maternally mediated effects also makes the task of extrapolating from animaldata to potential human outcomes more problematic. 1,2 As stated by Kimmel and colleagues, 3 “Thedeveloping mammal and its maternal support system present a special situation in toxicology andrisk assessment.” They contend that because of the dependence of the conceptus on the nurturingmaternal environment, factors disturbing that environment may adversely affect the offspring’sdevelopment. They also recognize, however, that the maternal organism offers a degree of protectionagainst at least some environmental perturbations.It is well established that fetal disruptions can be maternally mediated, but questions remainregarding the kinds of effects produced, their prevalence, and their biological significance. 1,4 It is likelythat fetal variations, malformations, functional alterations, and deaths can result from direct effects onthe fetus, indirect (maternally mediated) effects, or a combination of the two. 5 Nevertheless, there hasoften been insufficient concern on the part of researchers for the potential of maternally mediatedeffects to affect the outcome of developmental toxicity studies. 3 The converse can also happen, however,in that some authors consider any effects on the conceptus that are seen only at maternally toxic dosesto be secondary to the maternal toxicity although there may be little or no evidence to substantiate thatsupposition. A more reasonable approach is that of Chernoff and coworkers, 2 who advocate concernabout manifestations of developmental toxicity, whether or not there is concurrent maternal toxicity,unless there is unequivocal evidence that humans would not be similarly affected.B. Historical BackgroundThe concept that stress and/or toxicity to the pregnant mother could indirectly result in adverseeffects on the developing conceptus is not new. Early studies investigated the potential of such“maternally mediated effects” on developing offspring. These studies generally used such stressorsas restraint, hypothermia, electric shock, noise, visual stimuli, shipping, or crowding on pregnantrodents. 6–17 Some investigators 14,18 used maternal food and water deprivation as a stressor for mice,but it is not entirely clear if alterations seen in the offspring were due to indirect effects or, at leastin part, to malnutrition of the conceptus. Nevertheless, food and water deprivation was correlated© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 95Figure 4.1 Schematic of the stress cascade. (Adapted from Colby, H.D., J. Am. Coll. Toxicol., 7, 45, 1988.)with increased maternal plasma corticosterone levels, 14,19 and resulted in an increased incidence ofcleft palate. Additional early studies suggested that maternal restraint stress in the rat couldpotentiate the effects of chemical teratogens. 20,21Although investigations of the role of maternal effects in causing growth retardation, death,behavioral effects, biochemical alterations, or dysmorphogenesis in the embryo and fetus continuedinto the 1980s, 22–30 the major impetus for further research came as a consequence of publicationsby Khera. 31–34 Khera reviewed published studies and proposed that a number of effects on theoffspring of mice, rats, and rabbits occurred merely as a consequence of maternal toxicity. Suchputative maternally mediated fetal effects included decreased body weights, certain malformationsand variations, and resorptions. Additional discussion of the significance of maternally mediatedeffects can be found in reviews by Hood, 4,35 Chernoff and colleagues, 2 and Daston. 1C. Causes of Maternal StressIn the broadest sense “stress” is any change in the organism’s environment that disturbs homeostasis.The resulting series of neural and endocrine adaptations is commonly referred to as the “stressresponse” or “stress-cascade” (Figure 4.1). Operationally defined, a “stressor” is any manipulationcapable of disturbing homeostasis. Many of the procedures utilized in the collection of data intoxicology, and teratology is no exception, would qualify as stressors under this definition (Table4.1). Thus, stress can be considered an adaptive mechanism that has evolved to protect the organismin times of crisis. 36The mammalian response to stress is a basic regulatory mechanism carried out in part by thelimbic-hypothalamo-pituitary-adrenal (LHPA) axis. 36 Release of catecholamines from the adrenalmedulla and the sympathetic nervous system; adrenocorticotropin (ACTH) from the anterior portionof the pituitary gland; corticotropin releasing factor (CRF) and arginine vasopressin (AVP) fromthe hypothalamus; and glucocorticoids from the adrenal cortex are all key components of thisaction. The stress cascade is considered an example of a classic negative feedback loop. Whenglucocorticoids, the terminal element in the cascade, reach a sufficiently high level in the circulation,they inhibit the release of the initiating components of the cascade, CRF and ACTH.© 2006 by Taylor & Francis Group, LLC

96 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 4.1Stress inducers in toxicologyDosingRouteInhalation (e.g., nose-only)Dermal (e.g., wearing jackets)Gavage (e.g., inexperienced technician)Maximum tolerated dose (MTD)DeprivationFood or waterMaternalRestraintDuring dosingDuring monitoringHousing ConditionsGroupSingleThe response of the body to stress can also result in alteration or release of other hormonesand biochemical substances, including prolactin, growth factors, prostaglandins and other arachidonicacids, as well as proteinases, lymphokines, and peptides. Many body systems, including endocrine,neural, renal, and immune systems, can be activated. 36–40A wide variety of stressors have been employed in attempts to understand the effects of stresson adult animals. Many of these same stressors have also been used in developmental studies toelucidate both the prenatal and postnatal consequences of stress. In a practical sense the broaddefinition of a stressor as any disrupter of homeostasis may not be the most useful in delineatingthe role of stress in maternally mediated effects on development. It is obvious that events disruptiveto homeostasis can result in a myriad of biochemical and physiological changes, and that eventsas diverse as moving an animal to a new cage or immobilizing it would qualify as stressors. It is,however, equally obvious that events categorized as stressors 41 do not all result in exactly the samespectrum of biochemical and physiological changes, nor do they result in the same spectrum ofdevelopmental changes when applied during the prenatal period.The developmental alterations engendered by stress depend to a large extent on the speciesstudied as well as the type and the duration of the stressor involved. For these reasons it may bemore useful to further subcategorize events capable of disturbing homeostasis. The scheme developedby Allen and coworkers, 42 in which stressors are considered as either neurogenic or systemstressors, may be useful in this regard. Stressors are subdivided into those directly affecting variouscomponents of the areas normally participating in the stress reaction (e.g., a test chemical directlyactivates the adrenal cortex) 43 and those where the action of the stressor is initiated at the neurallevel (e.g., an experimental procedure, such as injection, is interpreted as aversive by the organism,resulting in a release of CRF from the hypothalamus and initiation of the stress cascade). Of course,there may be instances in which a stressor could act as both a systemic and a neurogenic stressor.1. Neurogenic Stress“Neurogenic stress,” according to Allen et al., 42 is the type of stress that causes signals to be sentto the hypothalamus by neural pathways. For neurogenic stressors to be effective, the peripheralnervous system must interact with the central nervous system. Both aspects of the nervous system© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 97must thus be intact for a response, such as stimulation of ACTH production, to occur. The variousinputs indicative of stress are presumed to converge via a final common path, the neurons of thebrain’s medial parvocellular division of the paraventricular nucleus of the hypothalamus. 36 Neurogenicstressors are apparently “psychological stressors” that presumably do not affect the pregnantmother directly through toxicity, physical effects, or direct physiological alterations. Instead, suchstressors act by causing the animal to perceive its environment as potentially hazardous or annoying.These perceptions then result in changes in the pregnant animal’s physiology and/or biochemistrythat could potentially affect the conceptus. Examples of neurogenic stressors include noxiousstimuli, such as trauma, electric shock, and vibration, as well as annoying stimuli, such as noise,light, and crowding. Allen and coworkers 42 also list “psychological” or “emotional” stress in theneurogenic category. When using such stressors, appropriate control groups must be included, andcare must be taken to ensure that effects seen are not merely due to secondary consequences tothe mother, such as decreased food or water intake.2. Systemic StressThe second major category of stress listed by Allen et al. 42 is “systemic stress.” According to thoseauthors, systemic stressors may be pharmacologic agents, such as ether or endotoxin, that appearto act following transport to the hypothalamus and pituitary by the circulation. Other chemicaltoxicants may act by similar means, or they may elicit at least part of their effects by causing painand acting by neurogenic pathways. Allen et al. 42 listed other examples of systemic stressors, namely“hypotension, (forced) exercise, and starvation,” which were said to act by “humoral and/or neuralpathways,” and environmental factors, such as heat or cold. They also listed “forced immobilizationor restraint” as a systemic stressor, but it seems more likely that restraint stress is primarilypsychologically mediated (i.e., a neurogenic stressor), and forced immobilization or restraint canbe considered both a systemic and a neurogenic stressor. Other potential stressors, such as acidosisand alkalosis, would cause a physiologically based stress, but were not listed. It is not always cleara priori which stressors are primarily neurogenic, which are systemic, and which may act by bothpathways to stimulate responses, such as stress hormone release.3. An Alternative Categorization of StressorsPacák and Palkovits proposed a different classification of stressful stimuli. 41 They divided stressorsinto four categories: (1) physical stressors that have a negative (or, in some cases, a positive)psychological component and that include cold, heat, intense radiation, noise, vibration, etc.,chemical toxicants, and pain induced by chemical or physical means, (2) psychological stressorsthat reflect a learned response to previously experienced adverse conditions, and may evokeresponses that can be considered to reflect anxiety, fear, or frustration, (3) social stressors reflectingdisturbed interactions among individuals, e.g., an animal being placed in the home cage of adominant animal, or in the case of humans, unemployment or marital separation, and (4) stressorsthat challenge cardiovascular and metabolic homeostasis, for example, exercise, hypoglycemia, andhemorrhage. Some stressors would include components of more than one of the four categories ofstressors. Examples include handling, restraint, and anticipation of a painful stimulus.D. Developmental Hazard Associated with Maternal Stress or ToxicityA number of developmental alterations have been suggested as possible maternally mediated effects.Only a few, however, have been experimentally shown to be commonly inducible by at least sometypes of maternal stress and/or toxicity, and these are discussed in later sections. They form a subsetof the anomalies listed for rodents by Khera, 31,32 and even this subset is not invariably seen in© 2006 by Taylor & Francis Group, LLC

98 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONoffspring of stressed or intoxicated dams. Maternally mediated behavioral and physiological (especiallyendocrine) alterations in mammalian offspring, including those of humans, have also beennoted by a number of investigators, 6,9,25,29,30,44–53 but these will not be specifically addressed in thischapter.1. Supernumerary RibsAccording to Russell, 54 supernumerary (lumbar) ribs (SNR) result from “homeotic shifts” in theaxial skeleton, i.e., they reflect “evolutionarily based capabilities for alternative development” thatmay involve serially segmented structures. Presumably SNR reflect altered expression of homeoboxgenes, although the details remain to be elucidated, 55,56 and they can occur in untreated animals. 57In a study by Chernoff and collaborators, 58,59 fetuses from pregnant mice subjected to immobilizationstress responded with an increased incidence of SNR, while the offspring of rats showed no sucheffect. These results indicated that extra rib production in mice may be a general response tomaternal stress, and that stress alone (or at least stress in the presence of food and water deprivation)was adequate to induce such a response.Lumbar ribs were the only anomalies consistently found by Kavlock and coworkers 60 whenseveral test compounds were given to mice at maternally toxic doses; seven of the ten test agentsincreased the incidence of extra ribs. Kimmel and Wilson 61 differentiated between “extra” (supernumerary)ribs — at least half as long as the thirteenth rib — and “rudimentary” ribs (RR) —ossified structures shorter than extra ribs; they proposed that the former tended to be treatmentrelated, while the latter were more variable in incidence and did not appear to be related to testagent dose. Kavlock et al. 60 did not make a distinction between the two “rib” types, as they foundthe incidence of both to be concurrently increased.In a study 62 employing eight compounds at maternally toxic doses in the rat, only two testagents increased the incidences of SNR. Another study found that maternally toxic doses of theherbicide bromoxynil induced extra ribs in both mice and rats. 63 When Chernoff and coworkerstreated both species, the incidence of SNR was elevated by treatment, but those seen in the ratwere mainly rudimentary ribs, while mice exhibited both RR and SNR. 64 The persistence of theextra ossified structures through postnatal day 40 varied by rodent species and “rib” length.The studies in which maternally toxic doses of chemicals were associated with increasedincidences of SNR suggest that this endpoint may have occurred as a secondary effect of maternaltoxicity-associated stress, at least in the mouse. However, the increase in SNR may represent adirect effect of the compounds on the fetus, an indirect effect due to disturbance of maternalhomeostasis, or a combination of direct and indirect effects.2. Other Variations and MalformationsKhera’s literature surveys 31,32 identified a constellation of defects that often appeared in offspringof mice, rats, rabbits, and hamsters when their pregnant dams were given maternally toxic dosesof chemicals. The studies 60,62 described above that were done to test Khera’s hypothesis that maternaltoxicity would commonly result in specific defects did not find a consistent relationship for defectsother than SNR. Such data suggest that maternal toxicity, as usually defined, is not an effective orconsistent inducer of most developmental defects.Some data exist, however, that indicate a possible relationship between at least one type ofmaternal stress and exencephaly-encephalocele and fused ribs, in addition to the SNR discussedabove. This was first suggested in 1986 58 and 1987 59 in related reports describing the developmentaleffects of maternal immobilization in mice. The particular method of restraint utilized, however,appears to be crucial in inducing the malformations, a finding supported in a mouse restraint studyby Rasco and Hood. 65 The degree to which the movement of the dam is restricted may be thedefining factor. When employed later in gestation, maternal restraint induced cleft palate in mice. 18© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 99Another maternal stressor, audiogenic stress (objectionable noise), has been said to causeadverse effects in mice. Defects such as exencephaly, encephalocele, and fused sternebrae havebeen seen in fetuses from dams subjected to such stress, 23,66 although Kimmel and coworkers 15found no such effects in rats and only an increased incidence of resorptions in mice. More recently,gestational exposure to loud noise was reported to alter the development and responsiveness ofimmune system components in rats. 673. Prenatal Mortality and Growth InhibitionIn the studies 58,59,65,68 done to date with mice, significant decreases in prenatal survival or prenatalweight gain have not generally been noted as a result of maternal restraint stress on only one dayof gestation; the one exception was a statistically significant but numerically rather modest trendtoward embryo or fetal mortality noted across treatment days by Chernoff and coworkers. 69 Whenpregnant rats were stressed on multiple gestation days, however, litter size 13 or birth weight 70 wasdecreased. In the latter study, the decreased birth weight may have merely been due to food andwater deprivation as there was no mention of deprivation of the unstressed controls. In that study,restraint of young rats (70 to 120 days) both before and during gestation delayed parturition, butthe same was not true for a group of older (11 to 13 months) rats. 70Other stressors have been said to have deleterious effects on survival to term or fetal growth.Audiogenic stress 8,15,17,22,23,66 and anticipation of electric shock 11 have been reported to cause embryoor fetal mortality in rodents and decreased fetal weight at term. 11,17,23,66,71 Stress-induced immunologicalimbalances have been proposed as causes of abortion in both mice 73 and women. 74E. Concern about Maternally Mediated Effects1. Distinguishing Maternally Mediated from Direct EffectsKhera hypothesized that a number of common effects on the offspring of rodents and rabbits whosedams had received toxic doses of test agents during gestation were secondary effects of maternaltoxicity. 31,32,34 Khera also stated that such effects often were not dose related, tended to be speciesspecific, and were seldom seen at doses below the maternally toxic dose. Khera’s proposal precipitatedincreased interest in the potential of maternally mediated effects to influence the outcomesof developmental toxicity studies and has been supported by Black and Marks. 72 Nevertheless, hisproposal has also received criticism. His literature review approach indicates a possible associationbut does not establish a causal relationship. 59 Also, because of the retrospective nature of the studiesanalyzed by Khera and because negative data often remain unpublished, Khera’s reviews could notavoid a degree of selection bias. 75 As pointed out by Schardein, 76 the endpoints used in literaturereports to assess maternal toxicity are often ill defined and are sometimes ignored altogether.There are plausible alternative explanations for Khera’s findings. If both offspring and motherhave similar inherent sensitivities to the same or different mechanisms of toxicity, the conceptusmay be similar to the dam in susceptibility to many chemical agents. 3,59 Alternatively, lowerabsorbed dose or lack of activating enzymes in the embryo or fetus might increase its resistanceto the toxicity of certain compounds to a level approximating or exceeding that of the mother.Many developmental toxicants appear to have cytotoxic effects, and some may produce maternaland developmental effects at similar doses because of a lack of selectivity. Also, a given speciesor strain is likely to have specific points in its program of development that are the most sensitiveto toxicity. Thus, each species and strain is likely to exhibit a specific spectrum of commondevelopmental defects in response to a variety of toxic insults, 75 whether a specific effect on theoffspring is maternally mediated or direct.A critical consideration in any attempt to understand the relative role of maternally mediatedeffects is the ability to determine which effects are direct effects on the conceptus and which, if© 2006 by Taylor & Francis Group, LLC

100 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONany, are indirect (maternally mediated). As noted previously in this chapter, maternal stress alonecan cause certain adverse fetal outcomes. Although such findings are of interest, the more criticalquestion in interpreting results of developmental toxicity tests is how commonly is toxicity to themother translated into adverse effects on the offspring and under what circumstances is suchmaternal mediation likely to occur?In an attempt to test Khera’s proposal, 31 Kavlock et al. 60 exposed groups of pregnant mice tomaternally toxic doses of one of ten different chemicals. Although a number of effects on theoffspring were seen, only supernumerary ribs (SNR) were observed in a majority of the test groups(7 of the 10), and no such association was found for any of Khera’s other proposed maternallymediated effects. Expected effects, such as embryonic resorption, inhibited fetal growth, and grossmalformations, were seen in only a minority of the test groups. The study by Kavlock andcolleagues 60 does not rule out the possibility that tests with another set of compounds would yieldresults more compatible with Khera’s predictions, but it does indicate that maternal toxicity is byno means invariably associated with Khera’s predicted major effects. The authors concluded that“there clearly is no direct relationship between the induction of maternal toxicity and the productionof major abnormalities.” 60A more recent study with rats took another approach to examine maternal influence on themanifestation of developmental anomalies. 77 The timing of maternal toxicity, as indicated by clinicalsigns and effects on body weight, was correlated with the specific period of gestation during whichobserved fetal defects were most likely to have been elicited. In addition, maternal toxicity datafrom individual dams with and without affected litters were compared. Both approaches failed tosuggest a role for maternal toxicity in causing the observed malformations (eye defects, such asanophthalmia and microphthalmia) associated with exposure to the herbicide cyanazine.Examination of a representative sample of the large number of existing safety evaluation studiescould provide valuable insight into the maternal toxicity issue and perhaps settle any remainingdebate concerning its contribution to fetal defects. Another possible approach would be to conductparallel experiments with chemicals known to produce teratogenic effects only in the presence ofmaternal toxicity. One set of pregnant dams, probably rats or mice, would be given the test agentsby the usual routes, while another set would be treated via intraamniotic or intrauterine administration,with doses scaled to result in similar exposures to the conceptus. Presumably, the secondmethod would still allow direct effects, but should allow use of doses low enough to avoid maternaltoxicity. As stated by Schardein, 76 the literature has yet to show “an unequivocal relationshipbetween specific maternal and developmental toxicities,” and “developmental disruption appearsnot to result unconditionally from maternal toxicity.”2. Influence on Developmental Toxicity Test OutcomesThe possibility of maternally mediated adverse developmental effects in developmental toxicitytesting is of concern because if such secondary effects occur, they might produce positive resultsin animal tests that would not be seen in man (or other species of interest) at expected exposurelevels. According to Seidenberg and Becker, 78 testing for developmental toxicity at dose levels thatare maternally toxic is a somewhat controversial issue. This is because at times, “Investigators mayargue that the use of such high dose levels would lead to a large number of false positive resultsas a direct consequence of disturbance of the maternal-fetal homeostasis by the induced maternaltoxicity.” Although Seidenberg and Becker 78 were discussing interpretation of the results of aChernoff-Kavlock screen, which differs from the traditional Phase II (developmental toxicity) tests,the same principles are involved. The authors further stated that their data from 55 chemicalcompounds failed to indicate that maternal toxicity alone causes effects on the offspring detectablein such a screen. A similar conclusion was drawn by Chahoud et al., 79 who found that maternalbody weight change, as an indicator of maternal toxicity, was not always associated with embryoor fetal toxicity. Critical analyses by others 80–82 support the concept that maternal toxicity is not© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 101invariably followed by adverse prenatal effects, such as decreased survival or fetal weight or anincreased incidence of abnormal morphology. Black and Marks 72 note that in range-finding studies,where the range of doses typically includes some that are highly toxic to the dam, such doses donot often result in an increased incidence of malformation, “although dose-related increases in theincidence of [unspecified] fetal variations are almost always seen.”3. Influence on Interpretation of Developmental Toxicity Test ResultsWhether animal test results accurately predict human hazard potential, and especially whether thepredicted effects are biologically significant and irreversible, are important issues in developmentaltoxicology. 35 Thus, any potential confounders — such as maternal toxicity effects — are of importance,especially when they differ between the test species and humans. The argument that highdosetesting may produce misleading results has been forcefully expressed by Khera, 32 who statedthat, “Teratology-testing studies usually include apparently maternotoxic dose levels, and therepetitious and quite often predictable fetal outcome have [sic] incriminated a large number ofcompounds as potential teratogens thus making the testing methods a meaningless exercise.” It canalso be argued, however, that it is their interpretation, rather than the results themselves, that is ofconcern. Since it has been shown experimentally that even severe maternal toxicity is not necessarilyaccompanied by developmental toxicity, it appears that many of the fetal outcomes seen concurrentwith maternal toxicity are not necessarily secondary to maternal effects. A remaining and perhapsmore important possibility is that in at least some cases the maternal stress induced by toxicitymay exacerbate the effects of a chemical teratogen.The consensus at a workshop convened by the U.S. EPA was that if developmental toxicity isobserved it cannot routinely be ignored or discounted as secondary to maternal toxicity. 3 Theworkshop discussions also lead to the conclusion that hazard assessments should be conducted forall agents that elicit developmental effects, even if those effects “are seen only in the presence ofmaternal toxicity.” The effective experimental doses should then be compared with likely humanexposures to assess the risk to humans. The U.S. EPA Guidelines for Developmental Toxicity RiskAssessment 83 reflect this point of view. Those guidelines state that even if fetal effects are seen onlywhen maternally toxic doses have been administered, “the developmental effects are still consideredto represent developmental toxicity and should not be discounted as being secondary to maternaltoxicity.” They further state that, “Current information is inadequate to assume that developmentaleffects at maternally toxic doses result only from maternal toxicity.” Similarly, the position of theU.S. FDA was reflected in comments by Collins et al., 84 who stated, “Developmental effects thatoccur in the presence of minimal maternal toxicity are thus considered to be evidence of developmentaltoxicity, unless it can be established that the developmental effects are unquestionablysecondary to the maternal effects. In situations where developmental effects are observed only atdoses where there is a substantial amount of maternal toxicity, then the possible relationship betweenmaternal toxicity and the developmental effects should be evaluated in order to make a properassessment regarding the toxicity of the test substance.”The question of interpretation of fetal effects seen in the presence of maternal toxicity has alsobeen addressed by Schardein. 76 He pointed out that statements have been made to the effect that aspecific chemical was not actually “teratogenic,” even though it induced malformations, becausesuch effects occurred only in the presence of maternal toxicity. Schardein further mentionedcomments such as, “the chemical was teratogenic, but not a teratogenic hazard.” As he properlypoints out, an agent that causes malformations is teratogenic, regardless of whether the mechanismis direct or indirect; the important point is whether the teratogen is selective in its effects. A similarpoint of view was expressed by Hood, 35 who commented, “Once one ascertains whether there areeffects on the offspring, then it is important to determine as much as possible about the mechanism(s)involved, but if the same mechanism(s) may occur in man, it does not matter whether the effecton the embryo/fetus is direct or indirect. All that truly matters in such a case is the final outcome,”© 2006 by Taylor & Francis Group, LLC

102 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONand “In reality, it is of little consequence to the embryo/fetus whether it was harmed by a chemicalthat acted directly or through maternal mediation.” Thus, until we know much more about maternallymediated effects and their relation to developmental hazard, the critical factors should be thelikelihood of toxicity to the pregnant woman and the relative margin of safety between likelymaternal exposures and toxic doses.4. Potentiation of Chemical Teratogenesis by Maternal StressSeveral studies have attempted to determine if maternal stress can interact with or potentiate theeffects of teratogenic chemicals or radiation when given in combination with such agents. In anearly study, 20 Hartel and Hartel subjected pregnant rats to both intermittent loud noises and brightlights or to immobilization during midgestation, in combination with vitamin A. Only immobilizationincreased the incidence of retinoid-induced malformations, primarily cleft palate, and prenatalmortality. In contrast, Ishii and Yokobori 66 found that loud noise increased the incidence ofprenatal deaths and malformations in mice treated with trypan blue. Goldman and Yakovac 21 thencoadministered restraint stress and salicylate to pregnant rats and observed an enhancement of theteratogenic effects of the chemical. The finding that two central nervous system (CNS) depressants,chlorpromazine and pentobarbital, attenuated the ability of maternal restraint to potentiate salicylateteratogenicity in rats suggested that the restraint effect is mediated via the CNS. 85 In a later study,pregnant mice of two strains were either briefly restrained, injected with lucanthone, or both. 86Lucanthone treatment of NMRI mice caused a high incidence of malformations that was notincreased by maternal restraint, but restraint appeared to have increased the incidence of resorptions.Conversely, in the F/A strain, the chemical alone did not significantly increase the malformationrate but its teratogenic effect was potentiated by restraining the pregnant dams, with no increasein resorptions.More recently, combined exposure to cadmium and noise was assessed in mice. 87 The combinationwas said to have significantly increased the fetal malformation rate (gross plus skeletal), butthe data appeared equivocal, and the specific defects seen were not listed. In another study, femaleUje:WIST strain rats were dosed with lithium prior to and throughout pregnancy. 88 Their femaleoffspring were then mated at maturity, with no further lithium exposure, and half of them weresubjected to restraint on gestation days 6 through 20. The restrained females gained less weightduring pregnancy, and their offspring weighed less at birth than was true of the (unrestrained)controls. The lack of an appropriate food- and water-deprived control group limits the interpretationof the data because the restrained group was food and water deprived as well as restrained.Rasco and Hood exposed pregnant mice to restraint stress and low doses of teratogens todetermine if the apparent enhancement of teratogenic effects seen by others when combiningmaternal immobilization stress and chemicals was a common response or an anomaly. They foundthat maternal restraint concurrent with either sodium arsenate or all-trans retinoic acid enhancedthe teratogenicity of the chemical. 68,89 The timing of administration of the retinoid during therestraint period influenced the intensity of the potentiative effect. 90 Although Domingo’s laboratorysubsequently found relatively few apparent developmental effects of maternal restraint when combinedwith treatment with various metal salts and other chemicals, they most often used restraintperiods of only 1 to 2 h/d, repeated for several days. 91,92 Brief restraint periods may not be effective,and there is also the possibility of habituation to the repeated stress. 93The mechanisms by which maternal stress potentiates the effects of chemical teratogens remainto be investigated. A number of possibilities come to mind, including altered biotransformingcapability, 94 changes in gastrointestinal secretion and motility that may influence absorption of anorally administered compound, altered blood flow to the placenta and/or the maternal liver, alteredlevels of cell proliferation or thresholds for intracellular signaling due to effects of increasedmaternal serum hormone levels, altered target tissue receptor binding, altered gene or protein© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 103expression, 95 altered fetal programming, 96 altered lifestyle (in humans), 97,98 and altered body temperature.Note that some of these putative mechanisms are secondary effects of increases in maternalstress-related hormones and other endogenous compounds, while others may be caused, at least inpart, by neurogenically induced physiological changes.5. Possible Influence on Human Pregnancy OutcomesLittle is known of the potential of maternal stress or toxicity to cause adverse fetal outcomes or topotentiate chemical teratogenicity in humans, although several possibilities for such effects exist.According to early theories, exposure of a pregnant woman to a shocking, worrisome, or frighteningsituation could result in birth of a malformed child, but such theories have fallen into disfavor. 99Also, no association was shown between maternal “emotional upsets” and giving birth to a babywith cleft lip or palate, 100 and similar results were reported for maternal exposure to “airport noise”and malformation. 101 However, three later studies have reported an apparent connection betweenmaternal emotional stress and malformed offspring. 102–104 Although some studies have suggested arelationship between maternal stress and low birth weight, McAnarney and Stevens-Simon 105 foundthat the data were not definitive. More recently, Lou and coworkers 106 reported that maternal stresswas associated with decreased offspring head circumference, although the number of individualscompared was small, and Chen et al. 107 described birth weight reduction following maternaloccupational stress.Workplace stress has not generally been found to increase the likelihood of adverse pregnancyoutcomes, 108 but there are few studies of such possibilities, and epidemiologic studies typically canidentify only relatively strong developmental hazards or those with unusual outcomes. In a fewstudies some apparent stressors, such as working long hours and “psychosocial stress in theworkplace,” have been linked to preterm birth 109 or reduced fetal growth, 110 and both workplacestress and “negative life events” have been associated with spontaneous abortion. 111–113 Somecommon sources of stress, such as marital conflict, 114 appear not to have been investigated foreffects on fetal outcome since the work of Stott. 115 It should be kept in mind, however, that studiesseeking effects of stressors in women have been hampered by the difficulty of measuring psychologicalparameters in ways that allow for comparisons among groups. 116As pointed out by Daston, 1 the increased likelihood of neural tube defects associated withmaternal treatment with anticonvulsants might at least partly be a secondary effect of the concomitantdrug-induced folate deficiency. He also comments that certain drug treatments or toxic exposurescan adversely alter maternal zinc metabolism, which might affect offspring development, asmay the vitamin A depletion seen in alcoholic women. Further, as stated by Brent and Holmes, 117abnormal maternal metabolic states, such as diabetes mellitus and phenylketonuria, can contributeto abnormal development of the embryo or fetus. And a study of maternal stress response and fetalcardiac activity revealed that fetuses of high anxiety mothers had increases in heart rate duringperiods of maternal stress, suggesting that maternal stress can alter fetal physiology, 99a providinganother possible means for influencing birth outcome.II. POSSIBLE MECHANISMS FOR MATERNALLY MEDIATED EFFECTSNumerous possible mechanisms for secondary effects of maternal stress or toxicity have beenproposed, and no attempt will be made to review them in general here, as this topic has beenextensively reviewed by Daston 1 and by Khera. 33 Kalter and Warkany 119 and DeSesso, 5 also havereviewed the potential for maternal factors, including abnormal metabolic states, to influencedevelopmental outcome, and Khera recently described damage to maternal placental circulation asa possible contributor to adverse developmental effects in the mouse. 120 An interesting recent study© 2006 by Taylor & Francis Group, LLC

104 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof rats found that late gestational blockade of opioid receptors by administration of naltrexoneprevented reduction in male anogenital distance and altered certain behavioral outcomes in offspringof light and noise-stressed dams. 121 Further, the stress-induced reductions in pain induced byrestraint and other stressors can be blocked by opioid antagonists. 122 These data suggest that opioidreceptors may be involved in at least some maternally mediated effects of stress.At least one potential maternally mediated factor has apparently not been discussed in theabove-mentioned reviews. That is the possibility of inducing abnormal maternal biotransformationat high — often maternally toxic — doses of chemical agents that are typically metabolized to lesstoxic forms by the mother. Such high exposures may overwhelm the normal maternal biotransformingabilities and lead to metabolism by normally minor pathways, some of which may producehazardous metabolites. If such metabolites are produced in significant quantities, some may havethe potential to harm the conceptus. Further, maternal stress can inhibit normal xenobioticbiotransformation 123 and might result in exposure of the conceptus to higher levels of toxicantsthan would otherwise be the case.III. ASSESSMENT OF MATERNAL STRESS OR TOXICITYMultiple biochemical, physiological, and behavioral changes are engendered by stress and wouldappear to provide the toxicologist or teratologist with a variety of endpoints suitable for assessingmaternal stress. In studies specifically designed to evaluate the impact of stress on development,there is no doubt that both systemic and neurogenic stressors, 42 when applied to pregnant animalsin an experimental setting, will alter developmental outcome (see Chernoff et al., 2,59,60,62,124 Joffe, 125Hood, 4,35 and Weinstock et al., 126 for studies relevant to this issue). These studies, as well as studiesinvestigating the impact of stress in the adult, utilize a variety of methods to measure the manychanges associated with exposure to a stressor. In the following section, we briefly catalog a varietyof methods for assessing stress. However, the reader should be aware that their implementation ina standard teratology study may be quite difficult. It is one thing to determine the consequencesof an operationally defined stressor; it is quite another to determine if stress plays a role in aparticular developmental outcome. As Chernoff et al. 62 found in their investigation of overt maternaltoxicity and adverse developmental effects, the relationship between maternal toxicity, maternalstress, and developmental outcomes is not easy to define. One should, however, be aware of howvarious manipulations can act as neurogenic stressors (e.g., nose-only inhalation exposure procedures)and the impact these manipulations may have on development.A. Biochemical Measures1. Measures Related to the Stress CascadeThe biochemical indicators that are nearly synonymous with stress are those associated with thestress cascade; namely glucocorticoids (corticosterone in rodents and cortisol in primates, guineapigs, and hamsters). Selye 127 in his landmark studies first emphasized the use of this adrenal corticalproduct as a marker for stress. The levels of glucocorticoids in blood and their metabolites in urineare used much more often to gauge stress than are the levels of either ACTH or CRF. In mostexperimental studies that utilize circulating glucocorticoid levels as a stress marker, their elevationin blood is accepted as prima facie evidence that other components of the stress cascade have beenactivated. That is, CRF has been released from the hypothalamus into the hypophyseal portal vesselsleading to the pituitary gland. Subsequent release of ACTH from the corticotropes in this structureresults in the eventual stimulation of the adrenal cortex and release of glucocorticoids into thegeneral circulation. Generally, ACTH and CRF are evaluated after elevated glucocorticoid levels© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 105are noted and usually serve to provide knowledge about the functioning of the axis as a whole. Forexample, the direct action of a chemical at the adrenal cortex leading to the sustained secretion ofhigh levels of glucocorticoids over a period of weeks or months might be expected to result in adecreased secretion of ACTH (see Krieger and Liotta 38 for a discussion of ways to test thefunctioning of the HPA axis).Studies too numerous to list have measured plasma or serum levels of total blood glucocorticoidsas a primary indicator of stress. The total value includes free glucocorticoids as well as those boundto transcortin, the binding globulin found in blood. 128 Only unbound glucocorticoids are believedto have biological activity, but see Rosner and Hochberg 129 for a discussion of this issue. There hasbeen a recent emphasis, especially in humans, on utilizing saliva as an easily accessed body fluid thatcan be repeatedly collected without trauma and contains only unbound glucocorticoids. Saliva can beused to evaluate event-related activation of the HPA axis, as well as its general function. 130 Saliva canbe collected from small laboratory animals, but the limited volume may limit its use in gauging stress.Glucocorticoids can cross the placenta. Thus, quite a few investigators have attempted todelineate the role of elevated glucocorticoids in various morphological developmental abnormalities,and they have found that the ability of the human fetus to respond to stressors with a releaseof cortisol matures around the 20th week of gestation. 14,124,131–137 The role of glucocorticoids inpostnatal functional abnormalities, including altered sexual differentiation and sociosexualbehavior 25,126 and the response to stress by the neonatal, juvenile, and adult organism, 29,138–141 hasalso been investigated. However, the relationships between stress, maternal elevations in glucocorticoids,and developmental outcomes have remained elusive (see Hansen and coworkers 133 andMiller and Chernoff 124 for a discussion of this issue). Glucocorticoid elevations usually accompanystress, but this does not necessarily mean they are directly responsible for producing developmentalalterations, or that they play any role at all. Abnormalities may occur due to secondary alterationsin maternal and fetal physiology affected by elevated glucocorticoids. These abnormalities couldinclude altered uterine and placental blood flow, transient hypoxia due to limited oxygen or othersubstances in the blood, alterations in uterine contractility, and decreased production of essentialhormones (e.g., estrogen, progesterone, or DHEA)(see Schneider et al. 142 for a discussion).While elevated glucocorticoid levels may signal activation of other components of the axis, thisis not always the case, as a chemical may act directly on the adrenals. 43 Developmental toxicologystudies commonly involve the testing of chemicals, and a chemical may act as a systemic stressor,with direct activation of the HPA axis at various levels. Additional work would be needed todetermine if the elevation was due to an activation of the stress cascade or to a specific and directrelease of glucocorticoids by the compound.Many diverse agents can affect the adrenal glands, either through direct damage or by sustainedactivation, with a resultant increase in the circulating levels of its products (see Colby 143 for adiscussion). The susceptibility of the adrenal gland to toxic insult is documented by Ribelin 144 ina survey of the literature, where he noted that more than 90% of the citations describing toxicactions of chemicals on endocrine organs concern the adrenal and testes. While extensive documentationis not available to indicate direct chemical effects at other levels of the feedback loop,it is certainly possible (e.g., Bestervelt and coauthors 145,146 ).A primary difficulty in implementing measures of glucocorticoids, ACTH, CRF, or the catecholaminesas markers for stress is that blood must be collected. The collection process itself caninduce stress (e.g., see Reinhardt et al. 147 ), and to avoid this animals are anesthetized prior tocollection or receive indwelling catheters to facilitate blood sampling. 147,148 In some studies, groupsseparate from those used to assess toxic effects are included to determine the degree of stressinduced by various experimental conditions (e.g., Miller and Chernoff 124 ). In particular, care mustbe taken not to stress animals at the time of blood collection. Elevations in blood glucocorticoidlevels due to the manipulation of interest (whether chemical or procedural in nature) in subjectscould be masked by increased levels in inadvertently stressed control subjects.© 2006 by Taylor & Francis Group, LLC

106 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe small size of some species (e.g., mice) can preclude the collection of large amounts ofblood, limiting the number of end or time points that can be evaluated in a given sample. Investigatorsshould also be aware that there is a marked circadian pattern in the levels of blood glucocorticoids.149 In the rodent, the nadir (~1 g/dl) occurs during the lighted period, with a peak (~20g/dl) during the dark period. Thus, measuring blood glucocorticoids at a single time point can leadto difficulties in interpretation (see Kopf-Maier 135 and Hansen et al. 133 for a discussion of theseissues). There are also multiple reports of strain differences in the responsiveness of the HPA axisin many species. 150–154 Because of the many factors that affect glucocorticoid secretion, investigatorsshould not begin using glucocorticoids as markers for stress until they determine the basal circadianpattern of secretion obtained in their animal facility with the strains they normally use to evaluatechemicals for developmentally toxic effects. Finally, the investigator should be aware that in thepregnant mouse, as well as other species (e.g., rabbit, human), the total glucocorticoid level foundin blood increases dramatically at about gestational day 9, with the peak level observed betweendays 12 and 15. 133,155,156 The day and degree of peak elevation are strain dependent and may berelated to pregnancy-induced increases in the glucocorticoid binding protein in blood that occur ina number of species. 157–1592. CatecholaminesActivation of the HPA axis can also result in the release of catecholamines from the adrenal medulla,as well as from sympathetic nerve endings. Plasma norepinephrine is released primarily fromsympathetic nerve terminals, while epinephrine is secreted exclusively from the adrenal gland. 160These compounds have hemodynamic properties and can alter blood flow to the uterus. Exogenoustreatment of the dam with these substances, as well as other vasoactive agents, such as vasopressin,can cause developmental abnormalities (e.g., Chernoff and Grabowski 161 ). Thus, there has beensome interest in determining their role in maternal toxicity and stress in experimental animals aswell as in humans. 23,137 Cook and colleagues did not observe consistent alterations in blood oruterine catecholamine levels in response to noise, and the levels induced were much lower thanwould be obtained by exogenous treatment with these agents. 23 However, Gitau and colleagueshave observed impaired uterine flow in pregnant humans with documented high anxiety andattributed the impairment to possible elevations in catecholamines. 137There is some suggestion that plasma levels of catecholamines are more useful than glucocorticoidlevels for assessing stress because the former are more correlated with the intensity of thestressor. Indeed, blood catecholamine levels have been proposed as the best visceral indicator ofstress (e.g., see Natelson and coworkers 162 ). That is, catecholamine (rather than glucocorticoid)levels more accurately reflect levels of arousal. Both measures are similar in that a repeated exposureto the same neurogenic stressor results in a diminution of the response. 162 Again, these measuresrequire collection of blood, and care must be taken to avoid stressing animals during the collection.In many studies indwelling catheters are preferred for sampling, as release of catecholamines occursquickly in response to a stressor (e.g., within 5 min of the onset of restraint). Blood catecholaminelevels can be determined by high performance liquid chromatography. These methods have becomequite sophisticated in recent years and are able to detect quite small amounts (see Konarska andcoauthors 163 and Kopin 164 ).3. Stress ProteinsIn recent years there has been great interest in using those proteins known as “stress proteins” ascellular level markers for areas activated by stress or for monitoring the influence of differentenvironmental factors on animals. 165–168 The designation “stress proteins” usually includes theimmediate-early gene products (IEGPs) and the heat-shock proteins (HSPs). The HSPs were firstdiscovered in experiments involving hyperthermia but can be induced in response to other disruptive© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 107conditions. 169–171 Stress proteins include c-FOS, c-JUN, and the HSPs, which are named andclassified into different families on the basis of molecular weight (e.g., HSP 70, 90, or 100). Otherproteins, such as ornithine decarboxylase and metallothionein, as well as acute phase proteins, arealso sometimes included in the category of stress proteins. Stress proteins are highly conserved,rapidly inducible, and are synthesized in a variety of tissues (including brain, lung, heart, andlymphocytes) in response to conditions that alter homeostasis. 172–179 The areas where these proteinsare induced and their patterns of induction are dependent on the type of stressor (e.g., see Ceccatelliet al. 173 ). What is interesting is that these proteins are synthesized at a time when conditions (e.g.,stress) dictate a reduced synthesis of most other proteins.As stress proteins can be induced in response to chemical exposure, their measurement in theembryo has been suggested as a possible biomarker for teratogenic effects induced byhyperthermia 180 or more generally as a means of screening for or detecting teratogens in general. 181German 165 has also proposed the embryonic stress hypothesis of teratogenesis, but in general studieshave not supported his contention that induction of the heat-shock response is a common pathwaycapable of mediating developmental defects induced by diverse agents (see Finnell and colleagues 181or Mirkes and coworkers 182 ). It is more likely the induction of these proteins under certain conditionsassociated with developmental defects indicates the activation of particular cellular signaling pathwaysthat are parallel to those involved in development. The dam is not normally screened for theinduction of these proteins. However, as stress proteins appear to be induced in multiple organs andthe pattern of their induction is dependent on the disrupting condition, they may serve as a means foridentifying different kinds of stress (e.g., neurogenic vs. systemic). Conversely, they may identifystressors operating through common pathways because they are a biomarker for activation in a particularorgan. Screening for the induction of stress proteins may not be feasible in standard teratology studiesbecause tissue from the dam is required. However, in follow-up studies they may provide a way tobetter delineate the biochemical status of the dam when an anomaly is suspected to be due to stress.Use of stress proteins as a means of identifying target organs or transynaptic pathways activatedby certain conditions requires some means of identifying them. Some techniques that have beenused are two-dimensional gel electrophoresis, treatment of histological sections with antibodies,immunoblots of tissue homogenates, antibody-based detection, or general detection procedureswith the new proteomic platforms linked to mass spectrometry. 167,177,183–185 Because these proteinsare rapidly and transiently induced, time course determinations will provide the most information.4. Measures Related to Zinc MetabolismAdequate maternal zinc levels are needed for the appropriate development of the fetus. Alteredmaternal zinc metabolism may be a mediating factor in the abnormal development caused bydiverse toxicants. 1,186–188 Many of the same manipulations that are considered stressors (e.g.,hyperthermia, food and water deprivation) will cause a redistribution of maternal zinc, copper,and iron. This redistribution is suspected to occur as a consequence of an acute-phase responseto certain toxicants or manipulations, with a concomitant increased synthesis of certain serumand liver proteins.One of the proteins believed to figure prominently in the redistribution of zinc is the liverprotein, metallothionein (MT). Zinc, as well as copper, are bound to the newly synthesized MTand are therefore unavailable for distribution to the developing fetus. Glucocorticoids and catecholamines,both released in response to stress, can induce MT, and MT is sometimes classifiedas a stress protein. 189 As with other potential markers of maternal stress, time course considerationsare important. For example, an acute injection of α-hederin to the pregnant rat will induce MT;this induction reaches its peak in approximately 12 to 24 h. Serum from these dams was able tosupport embryo development in culture at 2 h but not 18 h after treatment. 186Whether repeated exposure to compounds capable of inducing MT and altering zinc distributionwould effectively lower blood zinc levels for a sufficient period to be detected at the usual tissue© 2006 by Taylor & Francis Group, LLC

108 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONcollection period used in a standard developmental toxicology testing protocol is unknown. Bloodshould be collected in zinc-free collection tubes to ensure accurate assessment of plasma zinc levels.Zinc concentrations in serum and other tissues as well as the concentration of other minerals canbe determined by flame atomic absorption spectrophotometry. The investigator should also be awarethere is a midgestational increase in the constitutive expressions of MT in mice but not in rats. 190This may serve to increase the vulnerability of mice to developmental insults mediated throughalterations in zinc metabolism. 186B. Organ WeightsDevelopmental toxicity protocols routinely evaluate maternal toxicity by monitoring weight andbody weight gain, food consumption, clinical signs of toxicity, and morbidity or mortality. Whileorgans are usually inspected at necropsy for gross toxicity, some protocols (e.g., reproductivetoxicity evaluation) require recording of organ weights. 191 Changes in certain organ systems shouldalert the investigator to possible maternal stress or activation of the stress cascade. For example,altered thymus or adrenal weight may suggest chronic adrenal axis activation or direct toxicity atsome level of the HPA axis. 192,193 As Akana and colleagues 192 have noted, both body and thymusweight are tightly regulated by glucocorticoids levels. Liver-to-body-weight ratio changes maysignal alterations in the liver, such as those associated with altered zinc metabolism. 186Chernoff et al. 62 investigated whether organ weight differences at necropsy define commonalitiesbetween diverse chemicals capable of inducing maternal toxicity. Organs (adrenals, thymus, spleen)expected to be affected by HPA axis activation were weighed at necropsies conducted at gestationaldays 8, 12, 16, or 20 from dams exposed to diverse chemicals agents on days 6 to 15. Maternaltoxicity was defined as alterations in body or organ weight and/or lethality. Consistent patterns inorgan or body weight changes were sought. All the evaluated chemicals induced maternal toxicityas indicated by lethality, reduced weight gain, and decreased relative thymus and spleen weight,as well as altered relative adrenal weight. Further, the changes were obvious throughout thetreatment period and persisted until term. As would be expected from the work of Akana andcolleagues, 192 the most sensitive organ was the thymus; the spleen and adrenal were less affected.Although all compounds induced maternal toxicity and affected organs associated with the stresscascade, there were no consistent developmental effects associated with maternal toxicity. A furtherinvestigation of the usefulness of organ weights for delineating the presence and consequences ofmaternal toxicity and stress would be helpful.C. Behavioral MeasuresMost of the measures outlined above necessitate extra groups of subjects because tissue must becollected close to the time of exposure rather than at the necropsy period utilized in the standarddevelopmental toxicology protocol. Thus, there is a place for noninvasive measures that will indicateif maternal stress is induced by certain conditions or compounds but that do not require terminationof the dam. 2 Along with glucocorticoids and catecholamines, stressors may also cause the releaseof endogenous opiates. As is the case with medicinal opiates (e.g., morphine) that are used to blockpain, one method of evaluating levels of endogenous opiates is to measure pain perception. Manyputative stressors, including restraint and electric shock, can rapidly diminish sensitivity to noxiousor painful events (i.e., stress-induced analgesia). 122 Altered pain perception has been utilized as anoninvasive gauge of stress in several studies evaluating the relationship between maternal stressand developmental outcome. 69,124Recently, Chernoff and colleagues 2 suggested that tail-flick assays, a measure of analgesia, maybe useful in determining if a compound or condition is stressful for species known to be responsiveto stress. It should be noted, however, that a test agent itself may have analgesic properties and beacting directly, rather than through the release of endogenous substances, to induce analgesia. It is© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 109true that analgesia assessment is a noninvasive procedure, but it does require skill as well as considerablehandling of the animal and, at the minimum, additional groups of nonhandled subjects. Again, thisprocedure, like others mentioned previously, may be the most useful in situations where the investigatorsuspects on the basis of preliminary information (e.g., 90-d toxicity studies) that maternal toxicity orstress may be a factor and where the study is designed specifically to address this issue.Analgesia can be detected with a variety of methods, but the most commonly used (e.g., inpharmaceutical studies) is the tail-flick assay. 194 This may have some benefits in testing pregnantanimals for stress-induced analgesia. For example, it has been suggested that endorphin systemsare activated during pregnancy, resulting in a form of endogenous analgesia, but there is somecontroversy regarding possible confounding of the analgesia measures employed. 195,196 Analgesiaincreased over the duration of pregnancy, but so did the weight of the rat, and the particular analgesiamethod used (flinch/jump test) can be influenced by the weight of the subject. The tail-flick assayinvolves little motor movement and is considered to be a polysynaptic reflex mediated within thespinal cord. 197Many manipulations considered to be stressful (e.g., restraint, shock, type of housing) can altergeneral activity for a significant period following their removal (see Miller 193 for a discussion ofthis in the rodent). Recently, Barclay and colleagues 198 suggested the use of the “Disturbance Index”(basically, movement in an open field) as a way of gauging distress in laboratory animals. Noticeablealterations in normal behaviors observed during preliminary toxicity tests or during cage-sideobservations during teratology testing should serve to alert the investigator that the compound understudy may cause maternal distress or stress.D. Measures of ToxicityEnd points of maternal toxicity were discussed at some length in a 1989 review by Chernoff andcoworkers 2 and more recently by the U.S. EPA. 83 The typical endpoints assessed during the courseof developmental toxicity assays are listed in Table 4.2. In general, these endpoints were chosenTable 4.2 End points of maternal toxicityMortalityMating index [(no. with seminal plugs or sperm/no. mated) X 100]Fertility index [(no. with implants/no. of matings) X 100]Gestation length (useful when animals are allowed to deliver pups)Body weight:Day 0During gestationDay of necropsyBody weight change:Throughout gestationDuring treatment (including increments of time within treatment period)Posttreatment to sacrificeCorrected maternal(body weight change throughout gestation minus gravid uterine weight or litter weight at sacrifice)Organ weights (in cases of suspected target organ toxicity and especially when supported by adversehistopathology findings):AbsoluteRelative to body weightRelative to brain weightFood and water consumption (where relevant)Clinical evaluations:Type, incidence, degree, and duration of clinical signsEnzyme markersClinical chemistriesGross necropsy and histopathologySource: From U.S. Environmental Protection Agency, 1991. 83© 2006 by Taylor & Francis Group, LLC

110 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONbecause they were readily assessable without adding greatly to the cost or length of studies; however,they are relatively crude, and their abilities to detect subtle or unusual toxicities are limited. Further,there has been some controversy over whether certain endpoints, such as enzyme induction orcertain physiological changes, should be considered to be manifestations of toxicity or merelyuseful adaptive responses. As stated by Chernoff and colleagues, 2 “maternal toxicity,” as it iscurrently determined, is often imprecisely defined, providing little that is of value to attempts tounderstand the potential impact of maternal effects on development.E. Overview of Stress and Toxicity AssessmentIt is clear from the preceding sections that there are a number of measures that may prove usefulin determining if certain conditions or agents induce “stress” in a pregnant dam. However, thesemeasures would be much more useful if there was a clearer understanding of the relationshipbetween different stressors and fetal outcome. As Schwetz and Harris 199 noted in their commentson the field of developmental toxicology, there is as yet no complete mechanistic understandingof the actions of various developmental toxicants. Maternal toxicity and stress is certainly noexception to this statement.IV. EXPERIMENTAL ASSESSMENT OF POSSIBLEMATERNALLY MEDIATED EFFECTSA. Factors to Be Considered in Experimental Design and ConductA number of factors must be considered in designing experiments to elucidate mechanisms ofdevelopmental toxicity. Some are typically needed in investigations of other types of toxicity, butothers are unique to developmental toxicity studies. This is true because in such a study one isassessing interactions between the maternal and the developing organism and because the conceptusis continually changing in its attributes and potential responses to toxicity over time. Thus, apparentfindings may have been confounded by maternal and/or developmental factors that must be takeninto account. Such factors may be unknown, and they may be difficult or impossible to control for,even if known. The remainder of this section deals with factors that should be considered indesigning experimental investigations of the potential influence of maternal factors or in interpretingresults of such studies.1. Controlling the Animal’s EnvironmentIt is not widely appreciated that the typical environment of laboratory animals is somewhat stressful,as measured by endogenous levels of stress-related hormones and other endpoints. According toreviews by Rowan 200 and by Barnard and Hou, 201 mere routine handling of animals that areunaccustomed to being handled is stressful, and acclimation to handling and to experimental housingor a test apparatus may decrease stress. Laboratory animals typically have an acute sense of hearing.Even routine animal room activities, such as changing cages, moving cage racks and feed containers,and running large cage or rack washers, can generate considerable levels of stressful noise, as canthe activities of some laboratory animals themselves (e.g., rabbits). 202 Barking of dogs housed ina nearby animal room may be heard by animals such as rats and causes stress. Even shipping,especially by air, is stressful. According to Brown et al., 12 a 48-h transport of pregnant A/J strainmice increased the incidence of cleft palate in their offspring.Merely moving the cages housing rats has been reported to cause altered hematologic valuesand heart function, as well as significant increases in serum levels of several hormones, such ascorticosterone, prolactin, and thyroxin. 203 Changing cages and introducing animals into a clean© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 111cage is also stressful, as it removes familiar odors. Riley designed a “low-stress” mouse facilitywith decreased noise levels and less cage changing. 204 He found that mice in such an environmenthad plasma corticosterone levels no higher than 35 ng/ml, while conventionally housed mice hadlevels ranging from 150 to 500 ng/ml.Also typically ignored is the possible stress from isolating experimental animals that arenormally somewhat social. For instance, although female rodents to be used in developmentaltoxicity research are commonly group housed initially, once they are mated they are often housedindividually. Whether this significantly affects the outcome of developmental toxicity tests is notknown. Excessive crowding can also be stressful, but this is unlikely to occur because it would bea violation of modern animal care standards.Further, few investigators consider that the effects of a given stressor may be quite complex.For example, restraint, in addition to inducing analgesia and activation of the HPA axis, can greatlyreduce body temperature in mice (e.g., 2 to 3 degrees in certain strains). 205 Such complex effectsmay complicate interpretation of experimental outcomes. 206 Conversely, restraint does not appearto cause hypothermia in rats. 207The researcher should also consider that samples of blood or solid tissue constituents can besignificantly influenced by stressful conditions and by hypoxia prior to sampling. For example,levels of metabolites, such as adenosine monophosphate, in rat liver were affected by the timebetween sacrifice of the animal and freezing of the liver, as well as by even normal levels of daytimelighting and animal room background noises and by pentobarbital anesthesia. 2082. Measurement of StressVarious measures of maternal stress, as described in Section III, can be employed and attemptsmade to correlate them with developmental outcome. An example is the use of the tail flick testby Chernoff et al. 69 to correlate stress and fetal outcome (i.e., incidence of supernumerary ribs) inmice. Although those authors achieved success in terms of measuring relative degrees of maternalstress, their measure of stress did not correlate significantly with fetal effects as manifested byincreases in supernumerary ribs.3. Appropriate ControlsControlling for confounders in studies dealing with maternal effects can require a great deal of thoughtand consideration of more experimental factors than is the case with more straightforward experiments,e.g., standard safety evaluation tests. For example, maternal stress (such as restraint) can prevent thedam from consuming food or water, thus requiring that controls be deprived of food and water. Instudies of interactions of stress and chemical toxicants, controls should also include dams that aresubjected to stress alone, the toxicant alone, and the toxicant plus food and water deprivation. A clevermethod of determining the relative role of corticosterone in mediating maternal effects was employedby Barbazanges and coworkers who compared effects of stressors on offspring of rats that were eitherintact or were adrenalectomized and given substitutive corticosterone therapy. 2094. Developmental TimingExperiments dealing with the mechanistic bases for adverse effects on development are greatlyinfluenced by aspects of timing. As outlined in Chapter 6 and the related tables, the embryo (andto a lesser extent the fetus) undergoes profound changes in its anatomy and extraembryonicmembranes as development progresses. These morphological alterations are accompanied byequally striking changes in biochemical, physiological, and genomic attributes. Thus, when subjectingthe conceptus to a toxic insult, one is always dealing with a “moving target,” and a differenceof only a few hours can be critical in determining the outcome. The practical consequences of this© 2006 by Taylor & Francis Group, LLC

112 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONare critically important. Obviously, treatments must be given to each test animal at the same time,or (if known) at as nearly the same developmental stage as possible. Since most experiments dealingwith maternally mediated effects will involve pregnant animals, treatment timing is generally usedas a surrogate for developmental stage. There are typically differences in developmental stagebetween individuals in the same litter in the common species of laboratory animals. 210 Thus, notall animals in the study are exposed at exactly the same stage in development, with a resultantincrease in the variability of the results. Also, the timing of fertilization in relation to treatmenttiming tends to be somewhat variable as well and might result in between-litter variability. Thismay not be a major factor, however, as Ishikawa et al. 211 have found that timing of ovulation is amore important determinant of developmental stage than is time of fertilization. Some researchershave advocated methods for reducing variability in developmental timing by restricting the timespan in which mating is allowed. 212 However, the most certain way to decrease intralitter stagevariability is to use embryo culture and select embryos at as nearly the same developmental stageas possible, an approach often incompatible with maternal effects research.A further consideration is the choice of controls for evaluating the effects of treatment timingon development. As the ability of the embryo or fetus to respond to a toxic insult changes overtime, the inappropriate choice of controls may result in data that cannot be interpreted.5. Consistency of MethodologyExperimental variability is inherent in all animal experimentation but is held to a minimum byrequiring a high degree of consistency and control in experimental technique. It is not alwaysappreciated, however, that even seemingly minor differences in methodology can result in significantlydifferent experimental outcomes. This is especially the case with highly complex interactingsystems such as the mammalian mother and conceptus as they respond to stress or toxicity.Adult rodents respond to diverse stress paradigms to different degrees. The rat responds torestraint stress by producing gastric ulcers, and ulcer production is further facilitated by cold, waterimmersion, or starvation. When more intense stress or multiple stressors are employed, less timeof exposure is needed for the same effect. 213 The intensity and nature of the stressor and the numberof episodes of stress determine the degree of stress perceived by the animal (or human) and thusthe type and degree of response. Such responses are mediated by a complex interplay of neuraland humoral factors, and it is not surprising that what seem to be modest changes in the applicationof stress can have significant effects on biological outcome. An example is seen in the results of arestraint stress study of pregnant mice by Rasco and Hood, 65 in which seemingly minor alterations inthe restraint method resulted in consistent differences in the incidence of restraint-induced rib fusion.6. Outcome Assessment and Interpretation of ResultsTraditional maternal and fetal endpoints — such as maternal clinical signs or body weight changesand fetal incidence of fused or supernumerary ribs — are in some cases adequate for establishingmaternally mediated effects of a test compound. More frequently the cause of a defect is obscureand more novel approaches are needed for determining the mechanism(s) responsible for theobserved defect. Often, in the case of chemical toxicity to the dam, it is difficult or impossible todetermine if specific effects seen in the offspring were direct or indirect.There are numerous ways in which the maternal physiology or biochemistry can be perturbed,and these can in turn adversely influence prenatal development. Thus it is essential to considerclues, such as knowledge of the target organs or possible mechanisms of maternal toxicity, thatmay suggest the proper endpoints to assess. Indeed, there have already been a number of caseswhere alert researchers were able to pinpoint such mechanisms and show that they were specificto the species involved or only occurred as a result of maternal toxicity. 24,26–28,132,214–217 Future mechanisticstudies may be aided by the availability of transgenic animals and knockout mice with defined© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 113defects in various components of pathways or systems activated by stressors or toxic substances (e.g.,the LHPA axis 218 ). Use of such animal models could allow inferences to be made regarding the possibleinfluences of alterations in maternal physiology on offspring development and survival.B. Induction of Maternal Restraint Stress as a Model SystemPhysical restraint, the partial or complete limitation of movement, has been used extensively as aninducer of stress in laboratory animals, and Paré, Glavin, and their colleagues have reviewed itsuse. 213,219 Restraint has advantages as a stress inducer: It can be applied relatively consistently, itcan be used in a manner that is not likely to cause pain to the experimental subjects, it does notrequire elaborate or expensive equipment, and it does not involve toxicity, so purely stress-relatedeffects can be separated from maternal toxicity effects. Nevertheless, results seem to differ dependingon the exact methodology employed for restraint (as this determines the degree to whichmovement is restricted) and the length of the restraint period.Chernoff and coworkers have established maternal restraint as a relatively consistent inducerof supernumerary ribs (SNR) in the CD-1 strain mice. 58,59,69 Exact timing is critical, as fetuses frommice restrained on gestation day 8 (copulation plug day is day 0 throughout this chapter) from9:00 a.m. to 9:00 p.m. had a significantly elevated incidence of SNR, while those from damsrestrained during the following 12 h did not. When pregnant mice were restrained for 12 h on oneof gestation days 6 through 14, increased incidences of SNR were found only in fetuses from damsrestrained on days 7 or 8. 69 Rib fusion was seen in a few fetuses from mice immobilized (i.e.,completely restricted in their movement) in a supine position by use of Johnson and JohnsonElastikon ® surgical tape, but not in those from dams restrained in padded conical holders, whichallow some movement. This suggestion that the exact method of restraint may be important wasconfirmed by Rasco and Hood, 65 who found that a slight modification of the taping procedure couldconsistently influence the incidence of fused ribs in CD-1 mice.The CD-1 mouse is an appropriate animal model for the investigation of effects of maternalimmobilization stress on offspring development. The restraint method must be applied consistently,and it should be employed on gestation day 7 or 8 if SNR are to be used as an endpoint. The initialstudies generally employed 12-h restraint periods; shorter restraint periods of 8 or 9 h may not beconsistently effective. The most acute stress effects on the conceptus appear to result from maternalrestraint in the supine position by some means such as immobilization with surgical tape, but othermethods may prove to be adequate.Plastic conical tipped screw cap 50-ml centrifuge tubes, such as the Falcon brand produced byBecton Dickinson Labware, act as inexpensive, easily obtained restraint devices. They have frostedareas useful for marking numbers on the tubes for identification of the restrained animals andassignment to treatment protocols. Holes for breathing should be drilled in the closed end andelsewhere as desired. The holes can be made using an ordinary electric power drill, but the bitshould be as sharp as possible to minimize melting the plastic from friction during the drilling, asthis leaves rough edges. The location of the holes should be kept consistent among different tubes.This can be facilitated by designating the appropriate locations with a marking pen prior to drilling.However, investigators should take care to use tubes from the same manufacturer to maintainconsistency within a given study or laboratory, as 50-ml centrifuge tubes from various supplierscan vary in their inner diameter.Pregnant mice can be introduced to the open end of the tube, and with a little encouragementwill generally enter it. They will quickly begin to attempt to back out, but they can be kept in byuse of the screw cap. Other means of keeping mice in the tubes include drilling two holes onopposite sides near the open end, placing a bolt through both holes, and securing it in place witha nut. Similarly, one can place a long hypodermic needle through both holes and stick the sharpend into a cork to secure the needle in place. Use of the screw cap allows the mouse to move backand forth in the tube, while use of the bolt or needle can either allow such movement or, if placed© 2006 by Taylor & Francis Group, LLC

114 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONnearer to the closed end, can inhibit such movement. The size and weight of the mouse are alsodeterminants of how much movement is afforded in the restraint method chosen. The latter twomethods utilizing the bolt or needle to prevent escape also allow access to the mouse for tail flicktests, obtaining blood samples, or measuring rectal temperatures with a suitable electronic thermometerand miniprobe. It should be kept in mind, however, that such additional manipulationsmay also stress the animals and may make experimental design and/or interpretation more difficult.If necessary, additional mated females, treated identically to the unsampled mice, can be used forcollecting such additional data, but their use does not allow for as accurate a correlation betweenthe data obtained and the fetal outcome.An alternative tube closure is the Identi-Plug plastic foam stopper, in the 35 to 45 mm size,made for closing Drosophila culture vials and similar containers. These are made by Jaece Industriesand marketed by major laboratory supply vendors. These resilient stoppers can be pushed into thetube behind the mouse and will typically stay in place, preventing the animal from moving backand forth. Care should be taken to avoid trapping the mouse’s tail between the stopper and the sideof the tube; twisting the stopper into place helps to avoid this.The ability of restraint to induce developmental effects in mice can be validated under one’schosen experimental conditions by findings of an increased incidence of SNR. Since most restraintmethods prevent access to food and water, use of a food- and water-deprived control group is ofvalue, although employment of appropriate controls can result in a relatively high number of testgroups. An example can be seen in a study by Rasco and Hood 68 that assessed the ability of maternalrestraint to enhance sodium arsenate teratogenicity. In addition to the experimental group givenarsenate alone, five control groups were used. Control mice were given one of the following: thevehicle, restraint, arsenate, food and water deprivation, or arsenate plus food and water deprivation.Similarly, in another study by the same authors, 90 timing of administration of all-trans-retinoic acid(tRA) during a period of maternal restraint was investigated by use of six control groups, alongwith five experimental groups. Controls included vehicle control, food and water deprived,restrained, tRA treated plus food and water deprived, and tRA given at one of two times concurrentwith the timing of restraint for the experimental (restrained plus tRA-treated) mice. Several additionalcontrols could have been used but were not considered to be critical.A further consideration in tests using incidence of SNR as one of the endpoints is the processof obtaining an accurate count of the fetal ribs. Thus, adequate clearing of the soft tissues isimportant. This is especially true for visualizing cervical ribs and extra thoracic ribs, which areinduced by some teratogens, but is somewhat less critical for seeing the lumbar ribs brought aboutby stress. The ribs may be counted beginning at the most caudal if one is concerned only withdetecting lumbar ribs. In this case, anything over 13 ribs is assumed to be a supernumerary rib. Ifdetection of cervical or thoracic ribs is desired, the cervical vertebrae (normally, seven) should becounted, beginning at the cranial end of the spinal column, to determine if there are any “ribs”associated with them. Then, the count can be continued through the thoracic vertebrae and theirassociated ribs, and finally, presence of lumbar ribs can be noted.Certain teratogens, e.g., all-trans-retinoic acid, may induce anteriorization of lumbar vertebrae,transforming them into replicas of thoracic vertebrae, with their associated ribs. These transformationscan be distinguished from typical lumbar ribs in that the transformed ribs are more similarin appearance to normal ribs. That is, tRA-induced ribs and normal ribs are similarly wide anduniform in width, with blunt ends, whereas lumbar ribs are generally more slender, have taperedends, and tend to point more ventrally. It is also possible for a fetus to have both one or twoteratogen-induced pairs of extra thoracic ribs and uni- or bilateral lumbar ribs (either stress inducedor “spontaneous,” which of course may possibly have been induced by the “normal” stress of lifein a typical noisy, well-lit laboratory animal facility). Finally, so-called rudimentary ribs 64 or“ossification sites” associated with lumbar vertebrae may be seen. These are also inducible bystress, but they are generally small and delicately formed and are commonly lacking in associatedcartilage. Thus rudimentary ribs are usually readily distinguishable from supernumerary ribs.© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 115V. OVERVIEWThe potential influence of changes in the maternal compartment on developmental outcome is notwell understood and is somewhat controversial, and continued interest in the potential for maternaltoxicity to affect the outcome and interpretation of developmental toxicity studies is evidenced bythe recent workshop dealing with the topic. 220 It is well established that maternal stress alone canresult in morphological or behavioral developmental effects on the offspring of laboratory rodents.Whether the outcomes of many developmental toxicity studies are greatly influenced by maternaltoxicity and/or stress is uncertain, although there are apparent examples of such effects in theliterature. Nevertheless, we must beware of improperly interpreting animal test results seen atmaternally toxic doses, and we should keep in mind that human exposures also sometimes occurat levels that are maternally toxic. It appears that maternal stress can influence the developmentaltoxicity of a variety of chemical agents, but it is again not known if this significantly influencesthe outcome of typical safety evaluation tests. It is also not understood whether the incidences ofdevelopmental defects, pre- or perinatal mortality, or growth retardation in human beings aresignificantly influenced by maternal stress. If such is the case, it would be useful to know if animalmodels can provide reliable information that can be extrapolated to the human situation.In brief, we know relatively little about maternally mediated effects on developing offspring,but we know they can occur, at least in certain laboratory animals. We need much more informationabout how this relates to human development, and we must be cautious about either over- orunderestimating the influence of maternal effects on the outcome of developmental toxicity assays.References1. Daston, G.P., Relationships between maternal and developmental toxicity, in Developmental Toxicology,2nd ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven, New York, 1994, p. 189.2. Chernoff, N., Rogers, J.M., and Kavlock, R.J., An overview of maternal toxicity and prenatal development:considerations for developmental toxicity hazard assessments, Toxicology, 59, 111, 1989.3. Kimmel, G.L., Kimmel, C.A., and Francis, E.Z., Implications of the Consensus Workshop on theEvaluation of Maternal and Developmental Toxicity, Teratogen. Carcinog. Mutagen., 7, 329, 1987.4. Hood, R.D., Maternal vs. developmental toxicity, in Developmental Toxicology: Risk Assessment andthe Future, Hood, R.D., Ed., Van Nostrand Reinhold, New York, 1990, p. 67.5. DeSesso, J.M., Maternal factors in developmental toxicity, Teratogen. Carcinog. Mutagen., 7, 225, 1987.6. Thompson, W.R., Influence of prenatal maternal anxiety on emotionality in young rats, Science, 125,698, 1951.7. Geber, W.F., Developmental effects of chronic maternal audiovisual stress on the rat fetus, J. Embryol.Exp. Morph., 16, 1, 1966.8. Geber, W.F. and Anderson, T.A., Abnormal fetal growth in the albino rat and rabbit induced by maternalstress, Biol. Neonat., 11, 209, 1967.9. Hutchings, D.E. and Gibbon, J., Preliminary study of behavioral and teratogenic effects of two “stress”procedures administered during different periods of gestation in the rat, Psych. Reports, 26, 239, 1970.10. Ward, C.O., Barletta, M.A., and Kaye, T., Teratogenic effects of audiogenic stress in albino mice, J.Pharmaceut. Sci., 59, 1661, 1970.11. Smith, D.J., Heseltine, G.F.D., and Corson, J.A., Pre-pregnancy and prenatal stress in five consecutivepregnancies: Effects on female rats and their offspring, Life Sci., 10, 1233, 1971.12. Brown, K.S., Johnston, M.C., and Niswander, J.D., Isolated cleft palate in mice after transportationduring gestation, Teratology, 5, 119, 1972.13. Euker, J.S. and Riegle, G.D., Effects of stress on pregnancy in the rat, J. Reprod. Fertil., 34, 343, 1973.14. Barlow, S.M., McElhatton, P.R., and Sullivan, F.M., The relation between maternal restraint and fooddeprivation, plasma corticosterone, and induction of cleft palate in the offspring of mice, Teratology,12, 97, 1975.15. Kimmel, C.A., Cook, R.O., and Staples, R.E., Teratogenic potential of noise in rats and mice, Toxicol.Appl. Pharmacol., 36, 239, 1976.© 2006 by Taylor & Francis Group, LLC

116 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION16. Barlow, S.M., Knight, A.F., and Sullivan, F.M., Prevention by diazepam of adverse effects of maternalrestraint stress on postnatal development and learning in the rat, Teratology, 19, 105, 1979.17. Fanghanel, J., and Schumacher, G.-H., Environmental effects on normogenesis and teratogenesis, withspecial regard to noise and vibration, in Teratological Testing, 2, Persaud, T.V.N., Ed., MTP Press,Lancaster, England, 1979, p. 325.18. Rosenzweig, S. and Blaustein, F.M., Cleft palate in A/J mice resulting from restraint and deprivationof food and water, Teratology, 3, 47, 1970.19. Hemm, R.D.L., Arslanoglau, L., and Pollock, J.J., Cleft palate following prenatal food restriction inmice: Association with elevated maternal corticosteroids, Teratology, 15, 243, 1977.20. Hartel, A., and Hartel, G., Experimental study of teratogenic effect of emotional stress in rats, Science,132, 1483, 1960.21. Goldman, A.S. and Yakovac, W.C., The enhancement of salicylate teratogenicity by maternal immobilizationin the rat, J. Pharmacol. Exptl. Therapeut., 142, 351, 1963.22. Nawrot, P.S., Cook, R.O., and Hamm, C.W., Embryotoxicity of broadband high-frequency noise inthe CD-1 mouse, J. Toxicol. Environ. Health, 8, 151, 1981.23. Cook, R.O., Nawrot, P.S., and Hamm, C.W., Effects of high-frequency noise on prenatal developmentaland maternal plasma and uterine catecholamine concentrations in the CD-1 mouse, Toxicol. Appl.Pharmacol., 66, 338, 1982.24. Clark, R.L., Robertson, R.T., Minsker, D.H., Cohen, S.M., Tocco, D.J., Allen, H.L., James, M.L., andBokelman, D.L., Diflunisal-induced maternal anemia as a cause of teratogenicity in rabbits, Teratology,30, 319, 1984.25. Ward, I.L. and Weisz, J., Differential effects of maternal stress on circulating levels of corticosterone,progesterone and testosterone in male and female fetuses and their mothers, Endocrinology, 114,1635, 1984.26. Clark, R.L., Robertson, R.T., Peter, C.P., Bland, J.A., Nolan, T.E., Oppenheimer, L., and Bokelman,D.L., Association between adverse maternal and embryo-fetal effects in norfloxacin-treated and fooddeprivedrabbits, Fund. Appl. Toxicol., 7, 272, 1986.27. Moriguchi, M. and Scott, W.J., Prevention of caffeine-induced limb malformations by maternaladrenalectomy, Teratology, 33, 319, 1986.28. Ugen, K.E. and Scott, W.J., Acetazolamide teratogenesis in Wistar rats: potentiation and antagonismby adrenergic agents, Teratology, 34, 195, 1986.29. Fride, E. and Weinstock, M., Prenatal stress increases anxiety related behavior and alters cerebrallateralization of dopamine activity, Life Sci., 42, 1059, 1988.30. Herrenkohl, L. R., Ribary, U., Schlumpf, M., and Lichtensteiger, W., Maternal stress alters monoaminemetabolites in fetal and neonatal rat brain, Experientia, 44, 457, 1988.31. Khera, K.S., Maternal toxicity — a possible factor in fetal malformations in mice, Teratology, 29,411, 1984.32. Khera, K.S., Maternal toxicity: a possible etiological factor in embryo-fetal death and fetal malformationof rodent-rabbit species, Teratology, 31, 129, 1985.33. Khera, K.S., Maternal toxicity of drugs and metabolic disorders — a possible etiologic factor in theintrauterine death and congenital malformation: a critique of the human data, CRC Critical Rev.Toxicol., 17, 345, 1987.34. Khera, K.S., Maternal toxicity in humans and animals: effects on fetal development and criteria fordetection, Teratogen. Carcinog. Mutagen., 7, 287, 1987.35. Hood, R.D., A perspective on the significance of maternally mediated developmental toxicity, Regulat.Toxicol. Pharmacol., 10, 144, 1989.36. Lopez, J.F., Akil, H., and Watson, S.J., Role of biological and psychological factors in early developmentand their impact on adult life, Biol. Psychiatry, 46, 1461, 1999.37. Hanoune, J., The adrenal medulla, in Hormones: From Molecules to Disease, Baulieu, E.-E. and Kelly,P.A., Eds., Chapman & Hall, New York, 1990, p. 309.38. Krieger, D. and Liotta, A., ACTH and related peptides, in Hormones: From Molecules to Disease,Baulieu, E.-E. and Kelly, P.A., Eds., Chapman & Hall, New York, 1990, p. 229.39. McEwen, B.S., DeKloet, E.R., and Rostene, W., Adrenal steroid receptors and actions in the nervoussystem, Physiol. Rev., 66, 1121, 1986.© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 11740. Munck, A., Guyre, P.M., and Holbrook, N.J., Physiological functions of glucocorticoids in stress andtheir relation to pharmacological actions, Endocrin. Rev., 5, 25, 1984.41. Pacák, K. and Palkovits, M. Stressor specificity of central neuroendocrine responses: implications forstress-related disorders, Endocrine Rev., 22, 502, 2001.42. Allen, J.P., Allen, C.F., Greer, M.A., and Jacobs, J.J., Stress-induced secretion of ACTH, in Brain-Pituitary-Adrenal Interrelationships, Brodish, A. and Redgate, E.S., Eds., S. Karger, Basel, Switzerland,1973, p. 99.43. Brunetti, L., Preziosi, P., Ragazzoni, E., and Vacca, M., Effects of vindesine on hypothalamic-pituitaryadrenalaxis, Tox. Lett., 75, 69, 1995.44. Herrenkohl, L.R., Prenatal stress may alter sexual differentiation in male and female offspring, Monogr.Neural Sci., 9, 176, 1983.45. Peters, D.A.V., Maternal stress increases fetal brain and neonatal cerebral cortex 5-hydroxytryptaminesynthesis in rats: a possible mechanism by which stress influences brain development, Pharmacol.Biochem. Behav., 35, 943, 1990.46. Lopez-Calderon, A., Ariznavarreta, C., and Chen, C.C.-L., Influence of chronic restraint stress on proopiomelanocortinmRNA and β-endorphin in the rat hypothalamus, J. Mol. Endocrinol., 7, 197, 1991.47. Ohkawa, T., Takeshita, S., Murase, T., Kambegawa, A., Okinaga, S., and Arai, K., Ontogeny of theresponse of the hypothalamo-pituitary-adrenal axis to maternal immobilization stress in rats, Endocrinol.Japon., 38, 187, 1991.48. Armario, A., Marti, O., Gavalda, A., Giralt, M., and Jolin, T., Effects of chronic immobilization stresson GH and TSH secretion in the rat — response to hypothalamic regulatory factors, Psychoneuroendocrinology,18, 405, 1993.49. Schmitz, C., Rhodes, M.E., Bludau, M., Kaplan, S., Ong, P., Ueffing, I., Vehoff, J., Korr, H., and Frye,C.A., Depression: reduced number of granule cells in the hippocampus of female, but not male, ratsdue to prenatal restraint stress, Mol. Psychiat., 7, 810, 2002.50. Frye, C.A. and Orecki, Z.A., Prenatal stress produces deficits in socio-sexual behavior of cycling, butnot hormone-primed, Long-Evans rats, Pharmacol. Biochem. Behav., 73, 53, 2002.51. Williams, M.T., Davis, H.N., McCrea, A.E., Long, S.J., and Hennessy, M.B. Changes in the hormonalconcentrations of pregnant rats and their fetuses following multiple exposures to a stressor during thethird trimester, Neurotox. Teratol., 21, 403, 1999.52. Koenig, J.I., Kirkpatrick, B., and Lee, P, Glucocorticoid hormones and early brain development inschizophrenia, Neuropsychopharmacology, 27, 309, 2002.53. Avishai-Eliner, S., Brunson, K.L., Sandman, C.A., and Baram, T.Z., Stressed-out, or in (utero)? TrendsNeurosci. 25, 518, 2002.54. Russell, L.B., Sensitivity patterns for the induction of homeotic shifts in a favorable strain of mice,Teratology, 20, 115, 1979.55. Gossler, A. and Balling, R., The molecular and genetic analysis of mouse development, Eur. J.Biochem., 204, 5, 1992.56. Kessel, M., Respecification of vertebral identities by retinoic acid, Development, 115, 487, 1992.57. Lansdown, A.B.G., Supplementary rib formation: spontaneous anomaly or teratogenic effect? Lab.Animals, 10, 353, 1976.58. Beyer, P.E. and Chernoff, N., The induction of supernumerary ribs in rodents: Role of maternal stress,Teratogen. Carcinog. Mutagen., 6, 419, 1986.59. Chernoff, N., Kavlock, R.J., Beyer, P.E., and Miller, D., The potential relationship of maternal toxicity,general stress, and fetal outcome, Teratogen. Carcinog. Mutagen., 7, 241, 1987.60. Kavlock, R.J., Chernoff, N., and Rogers, E.H., The effect of acute maternal toxicity on fetal developmentin the mouse, Teratogen. Carcinog. Mutagen., 5, 3, 1985.61. Kimmel, C.A. and Wilson, J.G., Skeletal deviation in rats: malformation or variation? Teratology, 8,309, 1973.62. Chernoff, N., Setzer, R.W., Miller, D.B., Rosen, M.B., and Rogers, J.M., Effects of chemically inducedmaternal toxicity on prenatal development in the rat, Teratology, 42, 651, 1990.63. Rogers, J.M., Francis, B.M., Barbee, B.D., and Chernoff, N., Developmental toxicity of bromoxynilin mice and rats, Fund. Appl. Toxicol., 17, 442, 1991.64. Chernoff, N., Rogers, J.M., Turner, C.I., and Francis, B.M., Significance of supernumerary ribs in rodentdevelopmental toxicity studies: postnatal persistence in rats and mice, Fund. Appl. Toxicol., 17, 448, 1991.© 2006 by Taylor & Francis Group, LLC

118 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION65. Rasco, J.F. and Hood, R.D., Differential effect of restraint procedure on incidence of restraint-stressinducedrib fusion in CD-1 mice, Tox. Lett., 71, 177, 1994.66. Ishii, H. and Yokobori, K., Experimental studies on teratogenic activity of noise stimulation, GunmaJ. Med. Sci., 9, 153, 1960.67. Sobrian, S.K., Vaughn, V.T., Ashe, W.K., Markovic, B., Djuric, V., and Jankovic, B.D., Gestationalexposure to loud noise alters the development and postnatal responsiveness of humoral and cellularcomponents of the immune system in offspring, Environ. Res., 73, 227, 1997.68. Rasco, J.F. and Hood, R.D., Effects of maternal restraint stress and sodium arsenate in mice, Reprod.Toxicol., 8, 49, 1994.69. Chernoff, N., Miller, D.B., Rosen, M.B., and Mattscheck, C.L., Developmental effects of maternalstress in the CD-1 mouse induced by restraint on single days during the period of major organogenesis,Toxicology, 51, 57, 1988.70. Glöckner, R. and Karge, E., Influence of chronic stress before and/or during gestation on pregnancyoutcome of young and old Uje:WIST rats, J. Exptl. Animal Sci., 34, 93, 1991.71. Smith, D.J., Joffe, J.M., and Heseltine, G.F.D., Modification of prenatal stress effects in rats byadrenalectomy, dexamethasone, and chlorpromazine, Physiol. Behav., 15, 461, 1975.72. Black, D.L. and Marks, T.A., Role of maternal toxicity in assessing developmental toxicity in animals:a discussion, Regulat. Toxicol. Pharmacol., 16, 189, 1992.73. Arck, P.C., Merali, F.S., Stanisz, A.M., Stead, R.H., Chaouat, G., Manuel, J., and Clark, D.A., Stressinducedmurine abortion associated with substance P-dependant alteration in cytokines in maternaluterine decidua, Biol. Reprod., 53, 814, 1995.74. Arck, P.C., Rose, M., Hertwig, K., Hagen, E., Hildebrandt, M., and Klapp, B.F., Stress and immunemediators in miscarriage, Human Reprod., 16, 1505, 2001.75. Palmer, A.K., Kavlock, R.J., et al., Consensus workshop on the evaluation of maternal and developmentaltoxicity. Work Group II Report, Teratogen. Carcinog. Mutagen., 7, 311, 1987.76. Schardein, J.L., Approaches to defining the relationship of maternal and developmental toxicity,Teratogen. Carcinog. Mutagen., 7, 255, 1987.77. Iyer, P., Gammon, D., Gee, J., and Pfeifer, K., Characterization of maternal influence on teratogenicity:an assessment of developmental toxicity studies for the herbicide cyanazine, Regulat. Tox. Pharm.,29, 88, 1999.78. Seidenberg, J.M. and Becker, R.A., A summary of the results of 55 chemicals screened for developmentaltoxicity in mice, Teratogen. Carcinog. Mutagen., 7, 17, 1987.79. Chahoud, I., Ligensa, A., Dietzel, L., and Faqi, A.S. Correlation between maternal toxicity andembryo/fetal effects, Reprod. Toxicol., 13, 375, 1999.80. Francis, E.Z. and Farland, W.H., Application of the preliminary developmental toxicity screen forchemical hazard identification under the Toxic Substances Control Act, Teratogen. Carcinog.Mutagen., 7, 107, 1987.81. Hardin, B.D., A recommended protocol for the Chernoff-Kavlock preliminary developmental toxicitytest and a proposed method for assigning priority scores based on results of that test, Teratogen.Carcinog. Mutagen., 7, 85, 1987.82. Palmer, A.K., An indirect assessment of the Chernoff/Kavlock assay, Teratogen. Carcinog. Mutagen.,7, 95, 1987.83. U.S. Environmental Protection Agency, Guidelines for developmental toxicity risk assessment, FederalRegister, 56, 63798, 1991.84. Collins, T.F.X., Sprando, R.L., Shackelford, M.E., Hansen, D.K., and Welsh, J.J., Food and DrugAdministration proposed testing guidelines for developmental toxicity studies, Regulat. Toxicol.Pharmacol., 30, 39, 1999.85. Goldman, A.S. and Yakovac, W.C., Prevention of salicylate teratogenicity in immobilized rats bycertain central nervous system depressants, Proc. Soc. Exptl. Biol. Med., 115, 693, 1964.86. Michel, C. and Fritz-Niggli, H., Radiation-induced developmental anomalies in mammalian embryosby low doses and interaction with drugs, stress and genetic factors, in Late Biological Effects ofIonizing Radiation, II, International Atomic Energy Agency, Vienna, 1978, p. 397.87. Murata, M., Takigawa, H., and Sakamoto, H., Teratogenic effects of noise and cadmium in mice:Does noise have teratogenic potential? J. Toxicol. Environ. Health, 39, 237, 1993.© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 11988. Glöckner, R., Schwarz, S., and Jähne, F., Enhanced effect of chronic stress on pregnancy outcome inUje:WIST rats by prenatal treatment with lithium, Exp. Toxicologic Pathol., 45, 35, 1993.89. Rasco, J.F. and Hood, R.D., Maternal restraint stress-enhanced teratogenicity of all-trans-retinoic acidin CD-1 mice, Teratology, 51, 57, 1995.90. Rasco, J.F. and Hood, R.D., Enhancement of the teratogenicity of all-trans-retinoic acid by maternalrestraint stress in mice as a function of treatment timing, Teratology, 51, 63, 1995.91. Colomina, M.E., Sanchez, D.J., Sanchez-Turet, M., and Domingo, J.L., Behavioral effects of aluminumin mice: influence of restraint stress, Biol. Psychol./Pharmacopsych., 40, 142, 1999.92. Torrente, M., Colomina, M.T., and Domingo, J.L., Effects of prenatal exposure to manganese on postnataldevelopment and behavior in mice: influence of maternal restraint, Neurotox. Teratol., 24, 219, 2002.93. Viau, V. and Sawchenko, P.E., Hypophysiotropic neurons of the paraventricular nucleus respond inspatially, temporally, and phenotypically differentiated manners to acute vs. repeated restraint stress,J. Comp. Neurol., 445, 293, 2002.94. Konstandi, M., Marselos, M., Radon-Camus, A.M., Johnson, E., and Lang, M.A., The role of stressin the regulation of drug metabolizing enzymes in mice, Eur. J. Drug Metab. Pharm., 23, 483, 1998.95. Cianfarani, S., Geremia, C., Scott, C.D., and Germani, D., Growth, IGF system, and cortisol in childrenwith intrauterine growth retardation: Is catch-up growth affected by reprogramming of the hypothalamic-pituitary-adrenalaxis? Pediat. Res. 51, 94, 2002.96. Edwards, C.R., Benediktsson, R., Lindsay, R.S., and Seckl, J.R., Dysfunction of placental glucocorticoidbarrier: link between fetal environment and adult hypertension? Lancet, 341, 355, 1993.97. Vogel, W. H., The effect of stress on toxicological investigations, Human Exptl. Toxicol., 12, 265, 1993.98. Slotkin, T.A., Lau, C., McCook, E.C., Lappi, S.E., and Seidler, F.J., Glucocorticoids enhance intracellularsignaling via adenylate cyclase at three distinct loci in the fetus: a mechanism for heterologousteratogenic sensitization?, Toxicol. Appl. Pharmacol., 127, 64, 1994.99. Warkany, J. and Kalter, H., Maternal impressions and congenital malformations, Plastic & Reconstr.Surg., 30, 628, 1962.99a. Monk, C., Fifer, W.P., Myers, M.M., Sloan, R.P., Trien, L., and Hurtado, A., Maternal stress responsesand anxiety during pregnancy: Effects on fetal heart rate, Dev. Psychobiol., 36, 67, 2000.100. Fraser, F.C. and Warburton, D., No association of emotional stress or vitamin supplement duringpregnancy to cleft lip or palate in man, Plastic & Reconstr. Surg., 33, 395, 1964.101. Edmonds, L.D., Layde, P.M., and Erickson, J.D., Airport noise and teratogenesis, Arch. Environ.Health, 34, 243, 1979.102. Nimby, G.T., Lundberg, L, Sveger, T., and McNeil, T.F., Maternal distress and congenital malformations:Do mothers of malformed fetuses have more problems?, J. Psychiat. Res., 33, 291, 1999.103. Hansen, D., Lou, H.C., and Olsen, J., Serious life events and congenital malformations: a nationalstudy with complete follow-up, Lancet, 356, 875, 2000.104. Carmichael, S.L. and Shaw, G.M., Maternal life event stress and congenital anomalies, Epidemiology,11, 30, 2000.105. McAnarney, E.R. and Stevens-Simon, C., Maternal psychological stress/depression and low birthweight. Is there a relationship? Am. J. Dis. Child., 144, 789, 1990.106. Lou, H.C., Hansen, D., Nordentoft, M., Pryds, O., Jensen, F., Nim, J., and Hemmingsen, R., Prenatalstressors of human life affect fetal brain development, Develop. Med. Child Neurol., 36, 826, 1994.107. Chen, D., Cho, S.-I., Chen, C., Wang, X., Damokosh, A.I., Ryan, L., Smith, T.J., Christiani, D.C.,and Xu X., Exposure to benzene, occupational stress, and reduced birth weight, Occupat. Environ.Med., 57, 661, 2000.108. U.S. Congress Office of Technology Assessment, Reproductive Health Hazards in the Workplace,U.S. Government Printing Office, Washington, DC, 1985.109. Luke, B., Mamelle, N., Keith, L., Munoz, F., Minogue, J., Papiernik, E., and Johnson, T.R.B., Theassociation between occupational factors and preterm birth: a United States nurses’ study, Am. J.Obstet. Gynecol., 173, 849, 1995.110. Hatch, M., Ji, B.-T., Shu, X.O., and Susser, M., Do standing, lifting climbing, or long hours of workhave an effect on fetal growth? Epidemiology, 8, 530, 1997.111. Fenster, L., Schaefer, C., Mathur, A., Hiatt, R.A., Pieper, C., Hubbard, A.E., Von Behren, J., and Swan,S.H., Psychologic stress in the workplace and spontaneous abortion, Am. J. Epidemiol., 142, 1176, 1995.© 2006 by Taylor & Francis Group, LLC

120 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION112. Neugebauer, R., Kline, J., Stein, Z. Shrout, P, Warburton, D., and Susser, M., Association of stressfullife events with chromosomally normal spontaneous abortion, Am. J. Epidemiol., 143, 588, 1996.113. Luke, B., Avni, M., Min, L., and Misiunas, R., Work and pregnancy: the role of fatigue and the“second shift” on antenatal morbidity, Am. J. Obstet. Gynecol., 181, 1172, 1999.114. Malarkey, W.B., Kiecoltglaser, J.K., Pearl, D., and Glaser, R., Hostile behavior during marital conflictalters pituitary and adrenal hormones, Psychosomatic Med., 56, 41, 1994.115. Stott, D.H., Follow-up study from birth of the effects of prenatal stresses, Develop. Med. Child.Neurol., 5, 770, 1973.116. Scialli, A.R., Is stress a developmental toxin? Reprod. Toxicol., 1, 163, 1988.117. Brent, R.L. and Holmes, L.B., Clinical and basic science lessons from the thalidomide tragedy: Whathave we learned about the causes of limb defects? Teratology, 38, 241, 1988.118. Monk, C., Fifer, W.P., Myers, M.M., Sloan, R.P., Trien, L, and Hurtado, A., Maternal stress responsesand anxiety during pregnancy: effects on fetal heart rate, Devel. Psychobiol., 36, 67, 2000.119. Kalter, H. and Warkany, J., Congenital malformations: Etiologic factors and their role in prevention,New Engl. J. Med., 308, 424, 1983.120. Khera, K.S., Mouse placenta: hemodynamics in the main maternal vessel and histopathologic changesinduced by 2-methoxyethanol and 2-methoxyacetic acid following maternal dosing, Teratology, 47,299, 1993.121. Keshet, G.I. and Weinstock, M., Maternal naltrexone prevents morphological and behavioral alterationsinduced in rats by prenatal stress, Pharmacol. Biochem. Behav., 50, 413, 1995.122 Miller, D.B., Restraint-induced analgesia in the CD-1 mouse: interactions with morphine and time ofday, Brain Res., 473, 327, 1988.123. Matamoros, R.A. and Levine, B.S., Stress response and drug metabolism in mice, Fundam. Appl.Toxicol., 30, 255, 1996.124. Miller, D.B. and Chernoff, N., Restraint-induced stress in pregnant mice — Degree of immobilizationaffects maternal indices of stress and developmental outcome in offspring, Toxicology, 98, 177, 1995.125. Joffe, J.M., Hormonal mediation of the effects of prenatal stress on offspring behavior, in Studies onthe Development of Behavior and the Nervous System: Early Influences, Gottlieb, G., Ed., AcademicPress, New York, 1978, p. 108.126. Weinstock, M., Fride, E., and Hertzberg, R., Prenatal stress effects on functional development of theoffspring, in Progress in Brain Research, Boer, G.J., Feenstra, M.G.P., Mirmiriam, M., Swaab, D.F.,and van Haaren, F., Eds., Elsevier, Amsterdam, vol. 73, 319, 1988.127. Selye, H., The general adaptation syndrome and disease of adaptation, J. Clin. Endocrinol. Metab.,6, 117, 1946.128. Hammond, G.L., Molecular properties of corticosteroid binding globulin and the sex-steroid bindingproteins, Endocrine Rev., 11, 65, 1990.129. Rosner, W. and Hockberg, R., Corticosteroid-binding globulin in the rat: isolation and studies of itsinfluence on cortisol action in vivo, Endocrinology, 91, 626, 1972.130. Kirschbaum C. and Hellhammer, D.H., Salivary cortisol in psychoneuroendocrine research: recentdevelopments and applications, Psychoneuroendocrinology, 19, 313, 1994.131. Barlow, S.M., Knight, A.F., and Sullivan, F.M., Diazepam-induced cleft palate in the mouse: the roleof endogenous maternal corticosterone, Teratology, 21, 149, 1980.132. Eldeib, M.M.R. and Reddy, C.S., Role of maternal plasma corticosterone elevation in the teratogenicityof secalonic acid D in mice, Teratology, 41, 137, 1990.133. Hansen, D.K., Holson, R.R., Sullivan, P.A., and Grafton, T.F., Alterations in maternal plasma corticosteronelevels following treatment with phenytoin, Toxicol. Appl. Pharmacol., 96, 24, 1988.134. Iida, H., Kast, A., Tsunenari, Y., and Asakura, M., Corticosterone induction of cleft palate in micedosed with orciprenaline sulfate, Teratology, 38, 15, 1988.135. Kopf-Maier, P., Glucocorticoid induction of cleft palate after treatment with titanocene dichloride?Toxicology, 37, 111, 1985.136. Uno, H., Lohmiller, L., Thieme, C., Kemnitz, J.W., Engle, M.J., Roecker, E.B., and Farrell, P.M.,Brain damage induced by prenatal exposure to dexamethasone in fetal rhesus macaques. I. Hippocampus,Devel. Brain Res., 53, 157, 1990.137. Gitau, R., Fisk, N.M., and Glover, V., Maternal stress in pregnancy and its effect on the human foetus:An overview of research findings, Stress, 4, 195, 2001.© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 121138. Ader, R. and Plaut, S.M., Effects of prenatal handling and differential housing on offspring emotionality,plasma corticosterone levels, and susceptibility to gastric erosion, Psychosomatic Med., 30, 277, 1968.139. Clarke, A.S., Wittner, D.J., Abbott, D.H., and Schneider, M.L., Long-term effects of prenatal stresson HPA axis activity in juvenile rhesus monkeys, Devel. Psychobiol., 27, 257, 1994.140. Fride, E., Dan, Y., Feldon, J., Halevy, G., and Weinstock, M., Effects of prenatal stress on vulnerabilityto stress in prepubertal and adult rats, Physiol. Behav., 37, 681, 1986.141. Peters, D.A., Prenatal stress: Effects on brain biogenic amine and plasma corticosterone levels,Biochem. Behav., 17, 721, 1982.142. Schneider, M.L., Coe, C.L., and Lubach, G.R., Endocrine activation mimics the adverse effects ofprenatal stress on the neuromotor development of the infant primate, Devel. Psychobiol., 25, 427, 1992.143. Colby, H.D., Adrenal gland toxicity: chemically induced dysfunction, J. Am. Coll. Toxicol., 7, 45, 1988.144. Ribelin, W.E., Effects of drugs and chemicals upon the structure of the adrenal gland, Fund. Appl.Toxicol., 4, 105, 1984.145. Bestervelt, L.L., Yong, C., Piper, W.N., Nolan, C., and Pitt, J.A., TCDD alters pituitary-adrenalfunction. I. Adrenal responsiveness to exogenous ACTH, Neurotoxicol. Teratol., 15, 365, 1993.146. Bestervelt, L.L., Pitt, J.A., Nolan, C.J., and Piper, W.N., TCDD alters pituitary-adrenal function II.Evidence for decreased bioactivity of ACTH, Neurotoxicol. Teratol., 15, 371, 1993.147. Reinhardt, V., Cowley, D., Scheffler, J., Vertein, R., and Wegner, F., Cortisol response of female rhesusmonkeys to venipuncture in home cage versus venipuncture in restraint apparatus, J. Med. Primatol.,19, 601, 1990.148. Wiersma, J. and Kastelijn, J., A chronic technique for high frequency blood sampling/transfusion inthe freely behaving rat which does not affect prolactin and corticosterone secretion, J. Endocrinol.,107, 285, 1985.149. Dallman, M.F., England, W.C., Rose, J.C., Wilkinson, C.W., Shinsako, J., and Siedenburg, F., Nyctohemeralrhythmic adrenal responsiveness to ACTH, Am. J. Physiol., 235 (Regulatory IntegrativeComp. Physiol. 4), R210, 1978.150. Clarke, A.S., Mason, W.A., and Moberg, G.P., Differential behavioral and adrenocortical responsesto stress among three macaque species, Am. J. Primatol., 14, 37, 1988.151. Dhabhar, F.S., McEwen, B.S., and Spencer, R.L., Stress response, adrenal steroid receptor levels, andcorticosteroid-binding globulin levels — a comparison between Sprague Dawley, Fischer 344, andLewis rats, Brain Res., 616, 89, 1993.152. Dhabhar, F. S., Miller, A. H., McEwen, B. S., and Spencer, R. L., Differential activation of adrenalsteroid receptors in neural and immune tissues of Sprague Dawley, Fischer 344, and Lewis rats, J.Neuroimmunol., 56, 77, 1995.153. Griffin, A.C. and Whitacre, C.C., Sex and strain differences in circadian rhythm fluctuation ofendocrine and immune function in the rat: Implications for rodent models of autoimmune disease, J.Neuroimmunol., 35, 53, 1991.154. Sternberg, E.M., Glowa, J.R., Smith, M.A., Calogero, A.E., Litwak, S.J., Aksentijevich, S., Chrousos,G.P., Wilder, R.L., and Gold, P.W., Corticotropin releasing hormone related behavioral and neuroendocrineresponses to stress in Lewis and Fisher rats, Brain Res., 570, 54, 1992.155. Barlow, S.M., Morrison, P.J., and Sullivan, F.M., Plasma corticosterone levels during pregnancy inthe mouse: the relative contributions of the adrenal gland and foeto-placental units, J. Endocrinol.,60, 473, 1974.156. Salomon, D.S., Gift, V.D., and Pratt, R.M., Corticosterone levels during midgestation in the maternalplasma and fetus of cleft palate-sensitive and resistant mice, Endocrinology, 104, 154, 1979.157. Ballard, P.L., Delivery and transport of glucocorticoids to target cells, in Glucocorticoid HormoneAction, Monographs on Endocrinology, 12, Baxter, J.D., and Rousseau, G G., Eds., Springer-Verlag,Berlin, 1979, p. 25.158. Doe, R.P., Zinneman, H.H., Flink, E.B., and Ulstrom, R.A., Significance of the concentration of nonproteinbound plasma cortisol in normal subjects, Cushing’s syndrome, pregnancy and during estrogentherapy, J. Clin. Endocrinol. Metab., 20, 1484, 1960.159. Seralini, G.-E., Smith, C.L., and Hammond, G.L., Rabbit corticosteroid-binding globulin: primarystructure and biosynthesis during pregnancy, Mol. Endocrinol., 4, 1166, 1990.160. McCarty, R. and Kopin, I.J., Stress-induced alterations in plasma catecholamines and behavior of rats:effects of chlorisodamine and bretylium, Behav. Biol., 27, 249, 1979.© 2006 by Taylor & Francis Group, LLC

122 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION161. Chernoff, N. and Grabowski, C., Responses of the rat foetus to maternal injections of adrenaline andvasopressin, Br. J. Pharmacol., 43, 270, 1971.162. Natelson, B.H., Creighton, D., McCarty, R., Tapp, W.N., Pitman, D., and Ottenweller, J.E., Adrenalhormonal indices of stress in laboratory rats, Physiol. Behav., 39, 117, 1987.163. Konarska, M., Stewart, R.E., and McCarty, R., Predictability of chronic intermittent stress: effects onsympathetic-adrenal medullary responses of laboratory rats, Behav. Neural Biol., 53, 231, 1990.164. Kopin, I. J., Catecholamine metabolism: basic aspects and clinical significance, Pharmacol. Rev., 37,333, 1985.165. German, J., Embryonic stress hypothesis of teratogenesis, Am. J. Med., 76, 293, 1984.166. Kochevar, D.T., Aucoin, M.M., and Cooper, J., Mammalian heat shock proteins: an overview with asystems perspective, Tox. Lett., 56, 243, 1991.167. Kohler, H.-R., Triebskorn, R., Stocker, W., Kloetz, P.-M., and Alberti, G., The 70 kD heat shockprotein (hsp 70) in soil invertebrates: a possible tool for monitoring environmental toxicants, Arch.Environ. Contam. Toxicol., 22, 334, 1992.168. Sagar, S.M., Sharp, F.R., and Curran, T., Expression of c-fos protein in brain: metabolic mapping atthe cellular level, Science, 240, 1328, 1988.169. Dienel, G.A., Kiessling, M., Jacewicz, M., and Pulsinelli, W.A., Synthesis of heat shock proteins inrat brain cortex after transient ischemia, J. Cerebral Blood Flow Metabol., 6, 505, 1986.170. Munro, S. and Pelham, H., What turns on heat shock genes? Nature, 317, 477, 1985.171. Schlesinger, M.J., Ashburner, M., and Tissieres, A., Heat Shock: From Bacteria to Man, Cold SpringHarbor Laboratory, Cold Spring Harbor, 1982.172. Brown, I.R. and Rush, S.J., Induction of a “stress” protein in intact mammalian organs after theintravenous administration of sodium arsenite, Biochem. Biophys. Res. Commun., 120, 150, 1984.173. Ceccatelli, S., Villar, M., Goldstein, M., and Hokfelt, T., Expression of c-Fos immunoreactivity intransmitter-characterized neurons after stress, Proc. Natl. Acad. Sci. USA, 86, 9569, 1989.174. Dragunow, M. and Robertson, H.A., Generalized seizures induce c-fos protein(s) in mammalianneurons, Neurosci. Lett., 82, 157, 1987.175. Hammond, G.L., Lai, Y.-K., and Markert, C.L., Diverse forms of stress lead to new patterns of geneexpression through a common and essential metabolic pathway, Proc. Natl. Acad. Sci. USA, 79, 3485,1982.176. Ewing, J.F., Haber, S.N., and Maines, M.D., Normal and heat-induced patterns of expression of hemeoxygenase-1 (HSP32) in rat brain: Hyperthermia causes rapid induction of mRNA and protein, J.Neurochem., 58, 1140, 1992.177. Kononen, J., Honkaniemi, J., Alho, H., Koistinaho, J., Iadarola, M., and Pelto-Huikko, M., Fos-likeimmunoreactivity in the rat hypothalamic-pituitary axis after immobilization stress, Endocrinology,130, 3041, 1992.178. White, F.P., The induction of “stress” proteins in organ slices from brain, heart, and lung as a functionof postnatal development, J. Neurosci., 1, 1312, 1981.179. Yang, G., Koistinaho, J., Zhu, S., and Hervonen, A., Induction of c-fos-like protein in the rat adrenalcortex by acute stress — immunocytochemical evidence, Mol. Cell. Endocrinol., 66, 163, 1989.180. Kimmel, C.A., Kimmel, G.L., Lu, C., Heredia, D.J., Fisher, B.R., and Brown, N.T., Stress proteinsynthesis as a potential biomarker for heat-induced developmental toxicity, Teratology, 43, 465, 1991.181. Finnell, R.H., Ager, P.L., Englen, M.D., and Bennett, G.D., The heat shock response: potential toscreen teratogens, Tox. Lett., 60, 39, 1992.182. Mirkes, P.E., Doggett, G., and Cornel, L., Induction of a heat shock response (HSP 72) in rat embryosexposed to selected chemical teratogens, Teratology, 49, 135, 1994.183. Nowak, T.S., Jr., Synthesis of a stress protein following transient ischemia in the gerbil, J. Neurochem.,45, 1635, 1985.184. Goering, P.L., Fisher, B.R., and Kish, C.L., Stress protein synthesis induced in rat liver by cadmiumprecedes hepatotoxicity, Toxicol. Appl. Pharmacol., 122, 139, 1993.185. Sawyer, T.K., Proteomics — structure and function. BioTechniques, 31, 156, 2001.186. Daston, G.P., Overmann, G.J., Baines, D., Taubeneck, M.W., Lehman-MeKeeman, L., Rogers, J.M.,and Keen, C.L., Altered Zn status by alpha-hederin in the pregnant rat and its relationship to adversedevelopmental outcome, Reprod. Toxicol., 8, 15, 1994.© 2006 by Taylor & Francis Group, LLC

MATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 123187. Keen, C.L. and Hurley, L.S., Zinc and reproduction: effects of deficiency on fetal and postnataldevelopment, in Zinc and Human Biology, Mills, C.F., Ed., Springer-Verlag, New York, 1989, p. 183.188. Taubeneck, M.W., Daston, G.P., Rogers, G.P., and Keen, C.L., Altered maternal zinc metabolismfollowing exposure to diverse developmental toxicants, Reprod. Toxicol., 8, 25, 1994.189. Klaassen, C.D. and Lehman-McKeeman, L.D., Induction of metallothionein, J. Am. Coll. Toxicol., 8,1315, 1989.190. Quaife, C., Hammer, R.E., Mottett, N.K., and Palmiter, R.D., Glucocorticoid regulation of metallothioneinduring murine development, Devel. Biol., 118, 549, 1986.191. Francis, E.Z., Testing of environmental agents for developmental and reproductive toxicity, in DevelopmentalToxicology, 2 nd Ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York, 1994,p. 403.192. Akana, S.F., Cascio, C.S., Shinsako, J., and Dallman, M.F., Cotricosterone: narrow range required fornormal body and thymus weight and ACTH, Am. J. Physiol., 249, R527, 1985.193. Miller, D.B., Caveats in hazard assessment: stress and neurotoxicity, in The Vulnerable Brain andEnvironmental Risks. Vol. 1. Malnutrition and Hazard Assessment, Isaacson, R.L. and Jensen, K.F.,Eds., Plenum Press, New York, 1992, p. 239.194. D’Armour, F.E. and Smith, D.L., A method for determining loss of pain sensations, J. Pharmacol.Exptl. Therapeut., 72, 74, 1941.195. Gintzler, A.R., Endorphin-mediated increases in pain threshold during pregnancy, Science, 210, 193, 1980.196. Dahl, J.L., Silva, B.W., Baker, T.B., and Tiffany, S.T., Endogenous analgesia in the pregnant rat: anartifact of weight-dependent measures?, Brain Res., 373, 316, 1986.197. Kirby, M.L., Alterations in fetal and adult responsiveness to opiates following various schedules of prenatalmorphine exposure, in Neurobehavioral Teratology, Yanai, J., Ed., Elsevier, New York, 1984, p. 235.198. Barclay, R.J., Herbert, W.J., and Poole, T.B., The Disturbance Index: A Behavioural Method ofAssessing the Severity of Common Laboratory Procedures on Rodents, Universities Federation forAnimal Welfare, Potters Bar, UK, 1988.199. Schwetz, B.A. and Harris, M.W., Developmental toxicology — status of the field and contribution ofthe National Toxicology Program, Environ. Health Perspect., 100, 269, 1993.200. Rowan, A.N., Refinement of animal research technique and validity of research data, Fund. Appl.Toxicol., 15, 25, 1990.201. Barnard, N. and Hou, S., Inherent stress. The tough life in lab routine, Lab. Animal, 17, 21, 1988.202. Milligan, S.R., Sales, G.D., and Khirnykh, K., Sound levels in rooms housing laboratory animals—Anuncontrolled daily variable, Physiol. Behav., 53, 1067, 1993.203. Gärtner, K., Büttner, D., Döhler, K., Friedel, R., Lindena, J., and Trautschold, I., Stress response ofrats to handling and experimental procedures, Lab. Animal, 14, 267, 1980.204. Riley, V., Psychoneuroendocrine influences on immunocompetence and neoplasia, Science, 212, 1100,1981.205. Miller, D.B. and O’Callaghan, J.P., Neurotoxicity of d-amphetamine in the C57BL/6J and CD-1mouse: interactions with stress and the adrenal system, Ann. N.Y. Acad. Sci., 801, 148, 1996.206. Johnson, E.A., Sharp, D.S., and Miller, D.B., Restraint as a stressor in mice: Against the dopaminergicneurotoxicity of d-MDMA, low body weight mitigates restraint-induced hypothermia and consequentneuroprotection, Brain Res., 875, 107, 2000.207. Wright, B.E. and Katovich, M.J., Effect of restraint on drug-induced changes in skin and coretemperature in biotelemetered rats, Pharmacol. Biochem. Behav., 55, 219, 1996.208. Faupel, R.P., Seitz, H.J., Tarnowski, W., Thiemann, V., and Weiss, C., The problem of tissue samplingfrom experimental animals with respect to freezing technique, anoxia, stress and narcosis. A newmethod for sampling rat liver tissue and the physiological values of glycolytic intermediates andrelated compounds, Arch. Biochem. Biophys., 148, 509, 1972.209. Barbazanges, A., Piazza, P.V., Le Moal, M., and Maccari, S., Maternal glucocorticoid secretionmediates long-term effects of prenatal stress, J. Neurosci., 16, 3943, 1996.210. Thiel, R., Chahoud, I., Jurgens, M., and Neubert, D., Time-dependent differences in the developmentof somites of four different mouse strains, Teratogen. Carcinog. Mutagen., 13, 247, 1993.211. Ishikawa, H., Omoe, K., and Endo, A., Growth and differentiation schedule of mouse embryos obtainedfrom delayed matings, Teratology, 45, 655, 1992.© 2006 by Taylor & Francis Group, LLC

124 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION212. Endo, A. and Watanabe, T., Interlitter variability in fetal body weight in mouse offspring fromcontinuous, overnight, and short-period matings, Teratology, 37, 63, 1988.213. Paré, W.P. and Glaven, G.B., Restraint stress in biomedical research: a review, Neurosci. Biobehav.Rev., 10, 339, 1986.214. Terry, R.D., Marks, T.A., Hamilton, R.D., Pitts, T.W., and Renis, H.E., Prevention of tilorone developmentaltoxicity with progesterone, Teratology, 46, 237, 1992.215. Sadler, T.W., Denno, K.M., and Hunter, E.S., Effects of altered maternal metabolism during gastrulationand neurulation stages of embryogenesis, in Maternal Nutrition and Pregnancy Outcome(Series: Ann. New York Acad. Sci., 678), Keen, C.L., Bendich, A., and Willhite, C.C., Eds., New YorkAcademy of Sciences, New York, 1993, p. 48.216. Tritt, S.H., Tio, D.L., Brammer, G.L., and Taylor, A.N., Adrenalectomy but not adrenal demedullationduring pregnancy prevents the growth-retarding effects of fetal alcohol exposure, Alcoholism: Clin.Exper. Res., 17, 1281, 1993.217. Marks, T.A., Black, D.L., and Terry, R.D., Counteraction of the embryolethal effects, but not thematernal toxicity, of bropirimine and tilorone by coadministration of indomethacin, J. Am. Coll.Toxicol., 13, 93, 1994.218. Bornstein, S.R., Böttner, A., and Chrousos, G.P., Knocking out the stress response, Mol. Psychiatry,4, 403, 1999.219. Glavin, G.B., Paré, W.P., Sandbak, T., Bakke, H.K., and Murison, R., Restraint stress in biomedicalresearch: an update, Neurosci. Biobehav. Rev., 18, 223, 1994.220. ECETOC, Influence of Maternal Toxicity in Studies of Developmental Toxicity, Workshop Report No. 4,European Centre for Ecotoxicology and Toxicology of Chemicals, Brussels, 2004.© 2006 by Taylor & Francis Group, LLC

CHAPTER 5Paternally Mediated Effects on DevelopmentBarbara F. Hales and Bernard RobaireCONTENTSI. Role of the Male in Mediating Developmental Toxicity ..................................................125A. Evidence from Epidemiological Studies...................................................................126B. Evidence from Animal Experimentation...................................................................127II.Potential Mechanisms Involved in Male-Mediated Developmental Toxicity...................128A. Drugs or Toxicants in Semen ....................................................................................128B. Drugs or Toxicants Affecting the Male Germ Cell ..................................................1281. Germ Cells in the Testis or the Posttesticular Excurrent Duct System .............1292. Reversibility.........................................................................................................1313. Heritability ...........................................................................................................132III. Methodological Approaches in Male-Mediated Developmental Toxicity ........................132A. Effects of Toxicants in Seminal Fluid.......................................................................132B. Effects on Sperm Quantity and Characteristics ........................................................132C. Effects on Sperm Quality ..........................................................................................1341. Sperm Chromatin Packaging and Function ........................................................1342. Sperm Genetic Integrity ......................................................................................1363. Epigenetic Changes .............................................................................................138IV. Summary and Conclusions ................................................................................................138Acknowledgments ..........................................................................................................................138References ......................................................................................................................................139I. ROLE OF THE MALE IN MEDIATING DEVELOPMENTAL TOXICITYIt is well established that there are risks to the progeny if the mother is exposed to a variety ofchemicals during pregnancy. The extent to which paternal exposures contribute to infertility andpregnancy loss is less evident. There is growing concern that paternal exposure to drugs, radiation,or environmental toxicants may result in a decrease in sperm count and an increase in male infertility,spontaneous abortions, birth defects, or childhood cancer, adversely affecting reproduction andprogeny outcome. Clear evidence that the exposure of males to xenobiotics can result in adverseeffects on progeny outcome has accumulated over the past two decades. 1 The recent publication ofthe proceedings of a multidisciplinary international conference on male-mediated developmentaltoxicity has highlighted the need for more research in this area. The goal of this research is to125© 2006 by Taylor & Francis Group, LLC

126 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 5.1Paternal exposures associated with adverse progeny outcomesAgent Pregnancy Loss a Birth Defects a Childhood Cancer aRadiation 0.9–1.5 1.4–5.6 1.8–6.7Solvents 0.9–2.3 NA b 1.7–7Anesthetic gases 1.5–1.8 NA NAHeavy metals 0.9–2.3 1.5–249 3.5–7Cigarette smoke 0.6–1.4 1.3 1.2–3.9Herbicides/pesticides NA 5.7–405 2.4–7.1Anticancer drugs NA 4.1 NAaValues represent the range of OR/RR (odds ratios/relative risk) found by differentstudiesbNA = not availableSource: Adapted from Aitken, R. J. and Sawyer, D., The human spermatozoon – notwaving but drowning, in Advances in Male-Mediated Developmental Toxicity, Advancesin Experimental Medicine and Biology, Vol. 518, Robaire, B. and Hales, B.F., Eds.Kluwer Academic/Plenum Press, New York, 2003. With permission.elucidate the extent to which the father contributes to the approximately 60% of all birth defectsthat are of unknown cause. 1,2Studies of the mechanisms by which paternal exposures adversely affect progeny outcome areessential to provide the scientific basis for the development of biomarkers for risk assessment. Thetwo major mechanisms by which exposure of a male to a drug or environmental toxicant mayadversely affect his progeny are: (1) direct exposure of the conceptus during mating to a chemicalpresent in the seminal fluid and (2) toxic effects on male germ cells.Two major approaches have been taken to identify instances in which paternal exposure to axenobiotic adversely affects progeny outcome, namely, epidemiological studies and animal experiments.A. Evidence from Epidemiological StudiesEpidemiological studies have focused largely on determining the consequences of paternal occupationalexposures on progeny. The effects of paternal occupational exposures on the offspringrange from early spontaneous abortion, which may be perceived as infertility, to delayed spontaneousabortions and stillbirth, malformations, preterm delivery, delivery of a small-for-gestationalage infant, childhood cancer, altered postnatal behavior, or changes in reproductive function. 1,3–9A variety of different paternal occupations have been associated with adverse progeny outcomes(Table 5.1). 3–9 Paternal exposures to radiation, solvents, heavy metals (e.g., mercury), pesticides,and hydrocarbons were associated with an increased incidence of spontaneous abortion or miscarriage,birth defects, or childhood cancer. 3–9 Paternal occupations that have been associated with anincreased risk of having a liveborn child with a birth defect include janitor, woodworker (forestryor logging, sawmill, carpenter), firefighter, electrical worker, and printer. 6,7 Men employed inoccupations associated with solvent exposures (with painters having the highest risk) were morelikely to have offspring with anencephaly. 5 Further, an increased risk of stillbirth, preterm delivery,or delivery of a small-for-gestational age infant was associated with paternal employment in theart (painters) or textile industries. 8,9Concerns about the potential consequences to offspring of exposure to herbicides were broughtto the forefront by male Vietnam veterans. The data from some of the studies conducted to addressthese concerns suggested that there was an increased incidence of birth defects among the childrenfathered by veterans, while other inquiries did not find a positive relationship. 10–13 Arbuckle andcolleagues reported recently that adverse progeny outcomes, including an increased risk of earlyabortions, were associated with the exposure of male farmers to pesticides that included carbamates,organophosphates, and organochlorines. 14© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 127Quantitative exposure estimates are available in very few studies. In one European study ofexposure levels and adverse effects, median sperm concentrations were reduced almost 50% inmen with a blood lead concentration ≥50 µg/dl. 15 Furthermore, in most studies, the specific chemicalexposures for each occupational group were not identified. Depending on the profession and the study,the likelihood that paternal occupational exposure is a factor in abnormal progeny outcome has rangedfrom 1.5-fold to as high as 4.5- to 5-fold. In addition, a wide range of paternal professions either werenot associated or were associated with a negative likelihood of abnormal progeny outcome. 4,7The possibility that “life style” or “recreational” exposures of the father to cigarette smoke,alcohol, caffeine, methadone, cannabinoids, cocaine, and other illicit agents may affect progenyoutcome also deserves consideration. Although some of these exposures have received at leastlimited attention, most studies have not shown any definitive evidence that paternal smoking oralcohol consumption causes birth defects in the offspring. 16 Several studies have established thatpaternal smoking is associated with increased miscarriages, while others have reported a decreasein birth weights. 17,18 Adverse effects of cigarette smoking on male reproductive functions includea decrease in sperm count, poor sperm motility, abnormal morphology, including retention ofcytoplasm, and ultrastructural abnormalities, as well as an increase in DNA adducts and reducedfertility. 19–23 Increased chromosome aneuploidy has been reported in the sperm of men who smokecompared with the sperm of those who do not. 19 At least four constituents in cigarette smoke havebeen shown to cause aneuploidy in different test systems. 24Therapeutic drug exposures are more readily documented than are “lifestyle” exposures. Despitethis, few studies have focused on the consequences to the progeny of paternal treatment with avariety of commonly used drugs, such as antihypertensives or antidepressants. One group of drugsthat has received some attention is the anticancer drugs. Treatment of men with these drugs is oftenassociated with transient or permanent infertility. In those instances when treated men have fatheredchildren, the limited studies available to date have found that the proportion of malformed childrenin the treatment group was not different from that in the control group or the general population. 25–27There was also no significant increase in cancer or genetic disease among the offspring of malecancer survivors. 28 At present, however, it is still difficult to assess the impact of the treatment ofmen with anticancer drugs because the number of patients in most studies is very low. Additionalcohorts of thousands of patients would be necessary to rule out a relative risk on the order of 1.5for a germ cell mutation. Thus, the true impact of the latest combination chemotherapeutic regimenson reproductive health, gamete genetic integrity, and more importantly, the genetic risks to futuregenerations, remains to be evaluated.To date, epidemiological studies have largely used reproductive outcome as the measure of apaternally mediated effect. In the future, as the knowledge gained from basic research is appliedto humans, it may be possible to analyze changes in spermatozoal DNA, gene expression, proteinproducts, or chromatin packaging as biomarkers of the impact of exposure to a male-mediateddevelopmental toxicant.B. Evidence from Animal ExperimentationAnimal studies have demonstrated that paternal exposures to a wide range of environmentalchemicals (e.g., carbon disulfide, lead, dibromochloropropane) and drugs (e.g., cyclophosphamide,chlorambucil, melphalan, ethanol) can result in abnormal progeny outcome. 1 Drugs or environmentalchemicals to which the male is exposed may be present in his seminal fluid and thus may havedirect effects on the sperm or ovulated egg, on the process of fertilization, or on the embryo itself.Alternatively, drugs or other chemicals may have adverse effects on the number of male germ cellsby affecting the hypothalamo-pituitary-testicular axis. Thus, they may alter the hormonal milieurequired to maintain spermatogenesis. The numbers of male germ cells may also be affected byinhibiting mitosis or meiosis, by triggering germ cell apoptosis, or by affecting the shedding of thegerm cells from the seminiferous epithelium. Insufficient numbers of functional sperm will result© 2006 by Taylor & Francis Group, LLC

128 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONin male infertility. Alternatively, the male germ cell may be functional but altered, with theconsequence that genetic defects or mutations are transmitted to the progeny, resulting in early orlate conceptal loss, malformations, or serious diseases after birth. Finally, epigenetic changes innoncomplementary or imprinted regions of the male genome may also adversely affect embryodevelopment in utero. Anticancer drugs are among the best studied male-mediated developmentaltoxicants. These drugs are present in the seminal fluid and are transferred to the female duringmating; they affect germ cell numbers and alter germ cell quality.II. POTENTIAL MECHANISMS INVOLVEDIN MALE-MEDIATED DEVELOPMENTAL TOXICITYA. Drugs or Toxicants in SemenA drug in the seminal fluid may be absorbed and distributed throughout the female, and may thusaffect the conceptus. A wide range of compounds has been shown to enter semen. 29–32 From animalexperiments, methadone, morphine, thalidomide, and cyclophosphamide are all examples of drugsreported to adversely affect progeny outcome by this mechanism. 33–36 Cyclophosphamide and/orits metabolites can enter all tissues of the male reproductive tract, including the seminal vesiclefluid; this drug is transmitted to the female partner, where it is absorbed through the vagina anddistributed to a large array of tissues. 36 When male rats were given a single dose (10 to 100 mg/kg)of cyclophosphamide immediately prior to cohabitation with females in proestrus, there was asignificant dose-dependent increase in preimplantation loss (the number of resorption sites in theuterus minus corpora lutea in both ovaries). 36 No significant increases in postimplantation loss (thetotal number of resorbed or dead fetuses per pregnant female) or abnormal fetuses were noted.The increase in preimplantation loss after the acute treatment of males with cyclophosphamidemay be due either to the presence of the drug in the seminal fluid or to an effect of the drug onspermatozoa in the testicular excurrent duct system. To distinguish between these possibilities,females in proestrus mated to control males were remated within 2 h to azoospermic, vasectomizedmales treated acutely with cyclophosphamide or saline. Once again, preimplantation loss wasincreased significantly in both drug-treated groups, suggesting that the increase in preimplantationloss was due to the presence of the drug in seminal fluid, rather than to an effect on spermatozoastored in the testicular excurrent duct system. 33 Thus, the presence of drugs in semen can modifyprogeny outcome; these effects are likely to occur early during gestation.B. Drugs or Toxicants Affecting the Male Germ CellThe second major mechanism by which drugs given to the male may affect progeny outcome isthat they alter male germ cell numbers or quality, either during spermatogenesis in the testis orduring spermatozoal maturation in the epididymis. Examples of male-mediated developmentaltoxicants thought to act via an effect on the male germ cell include lead, dibromochloropropane,vinyl chloride, 1,3-butadiene, acrylamide, and anticancer drugs. 1,37,38Lead exposure impairs reproductive capacity by inducing testicular toxicity, effects on spermatogenesis,and altered androgen metabolism. 39,40 Chronic low level exposures may compromisefertility in the absence of demonstrable effects on endocrine function and semen quality. 41,42 Effectson reproductive and learning behavior in the F 1 generation of rodents whose fathers were exposedto lead have been reported after exposure to low levels of lead, even in the absence of infertilityor overt testicular damage. 43,44Exposure to dibromochloropropane (DBCP), a nematocide, has been associated in men withazoospermia and oligospermia, as well as infertility, and in male rats with hepatic and renal effects,as well as testicular atrophy. 45,46 DBCP increased sister chromatid exchanges and chromosomal© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 129aberrations and tested positive in the dominant lethal test, 47 which involves mating treated andcontrol males with untreated females. The females are examined to determine the numbers ofcorpora lutea in the ovaries (an indication of the number of potential embryos) and implantationsin the uterus. Implants are classified as normal fetuses, dead fetuses, or early or late resorptions,and fetal weights may be monitored. Usually both pre- and postimplantation embryonic losses areconsidered to be indicative of dominant lethality. Vinyl chloride is a mutagen and carcinogen inmany test systems, but most studies found that it did not induce dominant lethality. 48 Dominantlethal mutations, congenital malformations, and heritable translocations were induced by the exposureof mice to butadiene. 49,50 Carbazole and cypermethrin were also positive in the dominant lethaltest when male germ cells were exposed to these insecticides. 51,52 The exposure of male rats toacrylamide in the drinking water for 10 weeks induced pre- and postimplantation loss in theirprogeny, as well as axonal fragmentation and/or swelling in the adult F 1 male progeny. 53Animal data reveal that anticancer drugs have a plethora of effects on male germ cells. Thatanticancer drugs affect germ cell numbers is not surprising, as they are almost without exceptiontoxic to rapidly dividing cells. Exposure to anticancer drugs may affect germ cell numbers byinhibiting mitosis or meiosis or by triggering germ cell death by apoptosis. 54 Multiple measures ofgerm cell quality (morphology, motility, DNA damage and genetic integrity, chromatin packaging)are also affected by anticancer drugs. 1,37 Alkylating agents, such as mechlorethamine, dacarbazine,or cyclophosphamide, were among the most potent germ cell mutagens, inducing heritable translocationsand dominant lethal and specific locus mutations; the predominant effects are usually inspermatids and early spermatozoa. 55,56 Cisplatin-induced dose-related increases in (1) DNA adductsand (2) chromatid and isochromatid breaks in leptotene and preleptotene spermatocytes and differentiatingspermatogonia; these germ cell effects were expressed as an increase in pre- andpostimplantation loss, as well as an increase in malformed and growth retarded fetuses. 57,58 Effectson stem cells represent the most serious threat to subsequent generations; these effects may differfrom those seen in post-stem cells. 59 Importantly, several alkylating agents (cyclophosphamide,procarbazine, melphalan, mitomycin C) induce chromosomal translocations in stem cells. 60 Therewas a significant increase in abnormal fetuses among the progeny of male mice mated to controlfemales 64 to 80 days after treatment with genotoxic chemicals, such as mitomycin C, ethylnitrosourea,or procarbazine. 60Doxorubicin and etoposide both inhibit topoisomerase II. Doxorubicin is cytotoxic at earlymeiotic stages and at all spermatogonial stages, and it has been demonstrated to affect spermmotility, sperm head structure, and sperm numbers, resulting in an increase in preimplantationloss. 61 Etoposide inhibits premeiotic DNA synthesis and induced micronuclei in spermatids afterexposure of diplotene-diakinesis and preleptotene spermatocytes. Like doxorubicin, bleomycin isan intercalating agent; it is postulated to induce DNA damage by generating reactive oxygen species.Interestingly, bleomycin induced specific locus mutations in spermatogonia but not in postspermatogonialstages. 62 The Vinca alkaloids vinblastine and vincristine interfere with the spindleapparatus, resulting in an arrest of mitosis and meiosis followed by cell death, as well as an increasein aneuploidy in surviving germ cells. Large doses of both drugs also affect Sertoli cells, destroyingmicrotubules and mitochondria. 63 Antimetabolites, such as 6-mercatopurine or 5-fluorouracil,induce dominant lethality and chromosomal aberrations in differentiating spermatogonia at stagesduring which rapid DNA synthesis is taking place. Thus, many drugs and other chemicals havebeen identified in animal experiments as male-mediated developmental toxicants.1. Germ Cells in the Testis or the Posttesticular Excurrent Duct SystemSpermatogenesis in the testis is a tightly regulated process. Spermatogonia undergo mitotic proliferationand then meiosis to form primary and secondary spermatocytes (spermacytogenesis). Aftercompletion of the second meiotic division, spermatocytes become spermatids and differentiate intospermatozoa (spermiogenesis), primarily by condensing nuclear elements, shedding most of the© 2006 by Taylor & Francis Group, LLC

130 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION100CPA5.1 mg/kg/day80∗∗∗∗∗Percent604030∗∗Post-Implantation lossPre-Implantation lossMalformations20100+ ++××1 2 3 4 5 6 7 8 9SpermatogenesisEpididymaltransitSpermatocytesSpermatidsSpermatogoniaFigure 5.1Adverse progeny outcomes after mating of control females to male rats treated by gavage withcyclophosphamide (5.1 mg/kg/d) for the indicated number of weeks. The observed effect dependson the phase of spermatogenesis in the germ cells when they are first exposed to cyclophosphamide.(Adapted from Trasler, J.M., Hales, B.F., and Robaire, B., Biol. Reprod., 34, 275,1986. Withpermission.)cytoplasm, and forming an acrosome and a flagellum. From the timing of the effect of an exposure,one can assess the stage specificity of the susceptibility of the germ cells proceeding throughspermatogenesis. For example, in mice, an effect on progeny outcome after exposure of males toa drug or x-rays for 1 to 5 days would represent an effect on spermatozoa, most probably thoseresiding in the epididymis. 64,65 Exposure to a toxicant 10 to 18 days prior to conception wouldrepresent an effect on spermatids, while long exposures (35 days or more) prior to conceptionshould represent an effect on spermatogonia. Germ cells at the spermatogonial stage in both miceand humans are very susceptible to x-ray exposure; sperm numbers are reduced, and the survivingsperm are morphologically abnormal. 64,65 Following exposure of mice to chlorambucil, a peak inmutation yield is observed when offspring are conceived from germ cells exposed as spermatids. 64,65In contrast, spermatozoa are the germ cells that are maximally sensitive to the specific locusmutations induced by acrylamide monomer. 65 Ethylnitrosourea has been shown to result in a highfrequency of embryo death (dominant lethality) in the offspring of mice after short treatment periodsin which only spermatozoa were exposed. 64,65Cyclophosphamide, a commonly used anticancer drug, remains one of the best studied examplesof a drug that affects male germ cells in a stage-specific manner (Figure 5.1). 66–70 Increased postimplantationloss was found after 2 weeks of chronic low dose cyclophosphamide treatment of male rats.This postimplantation loss rose dramatically to plateau at a level dependent on drug dose by 4 weeksof treatment and was reversed within 4 weeks of the termination of drug treatment. 67,71 Thus, cyclophosphamide-inducedpostimplantation loss was associated primarily with germ cell exposure duringspermiogenesis. Postmeiotic germ cells were also most susceptible to effects leading to the inductionof learning abnormalities in the progeny after paternal exposure to cyclophosphamide. 72–73 Heritabletranslocations were found in mice after exposure of spermatids and spermatozoa to cyclophosphamide. 74© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 131Figure 5.2Day 20 rat fetuses stained to reveal skeletal malformations were sired by control (left) and cyclophosphamide-treated(right, 5.1 mg/kg/day for 9 weeks) males.Interestingly, exposure of rat spermatocytes to cyclophosphamide resulted in increased preimplantationloss, as well as in synaptic failure, fragmentation of the synaptonemal complex, and altered centromericDNA sequences. 75 An increase in malformed (hydrocephaly, edema, micrognathia) and growth retardedfetuses was observed after 7 to 9 weeks’ treatment of male rats with a chronic low dose of cyclophosphamide(Figure 5.2), representing progeny sired by germ cells first exposed to cyclophosphamide asspermatogonia. 66,67 However, spermatogonia have been reported to be at low risk for the induction ofheritable translocations by cyclophosphamide. 74,76It is noteworthy that the malformations produced by exposure of male mice to urethane or x-rays or of male rats to cyclophosphamide are very similar; these malformations include dwarfism,open eyelids, and tail anomalies. 64,67 More than one mechanism may be involved in the developmentaltoxicity of such drugs. This is suggested by the spectrum of effects that are produced,ranging from pre- and postimplantation effects to growth retardation, malformations, and behavioralabnormalities. It is also indicated by the specificity of the susceptibility of germ cells at differentstages of development to insult with these agents.Many laboratories have concentrated their efforts on the effects of drugs on the male germ cellduring its differentiation in the testis. But drugs may also affect spermatozoa during their transitthrough the epididymis and vas deferens, or they may initiate drug-sperm interactions duringejaculation. As sperm transit through the epididymis, they become mature, i.e., able to fertilize anegg. 77 The administration of methyl chloride caused an increase in dominant lethal mutations as aconsequence of the selective inflammatory action of this agent on the epididymis; this increase inembryo loss was reversed by administering an anti-inflammatory agent. 78,79 Treatment of male ratswith cyclophosphamide resulted in an increase in postimplantation loss when the exposed spermatozoaoriginated in the caput or corpus, but not the cauda epididymidis. 80 The cysteine-rich protaminesbecome progressively more tightly compacted by the formation of disulfide bonds duringepididymal transit; the result of this may be that cauda epididymal spermatozoa are relativelyinaccessible to drugs such as cyclophosphamide.2. ReversibilityA major question that arises from studies on the consequences of exposure to male-mediateddevelopmental toxicants is the degree to which these effects are reversible. To determine whetherthe effects of chronic exposure of male rats to cyclophosphamide were reversible, investigators© 2006 by Taylor & Francis Group, LLC

132 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONtreated adult male rats with saline or cyclophosphamide for 9 weeks and then mated them at variousintervals posttreatment. 71 The high level of postimplantation loss observed after 9 weeks of exposureto cyclophosphamide was markedly decreased by 2 weeks post-treatment and had returned to thecontrol range by 4 weeks posttreatment. Thus, in rats, the male-mediated developmental toxicityof cyclophosphamide was reversible. Moreover, the rate of onset for the actions of cyclophosphamideon progeny outcome was parallel to the rate of reversal.3. HeritabilityAre there adverse effect(s) of paternal drug exposure that persist to subsequent generations in survivingoffspring? Significantly, after paternal cyclophosphamide treatment, increased postimplantation lossand malformations persisted to the F 2 generation; moreover, the malformations observed in the offspringof rats whose fathers were exposed to cyclophosphamide were similar to those found in the F 1 . 66,67,81Behavioral abnormalities caused by paternal cyclophosphamide treatment also persisted in subsequentgenerations. 72,73 Thus, this drug appears to affect spermatogonial stem cells, as well as postmeioticgerm cells, because effects are transmissible to the next generation. There is evidence that this is alsotrue for certain other male-mediated developmental toxicants. A number of the viable congenitalanomalies in the F 1 generation of urethane-treated males were expressed in the F 3 generation. 64 X-rayinduced anomalies have also been shown to be heritable. 64 Thus, germ-line alterations causing malformationscan be transmitted to the next generation as dominant mutations, often with reduced penetrance.No specific chromosomal aberrations were associated with these malformations.While it is clear from animal experiments that exposure of the father to a variety of chemicalscan result in a heritable alteration in his surviving “apparently normal” F 1 progeny, the limitedstudies done to date do not permit us to conclude whether it is valid to extrapolate between species,e.g., from rodents to humans.III. METHODOLOGICAL APPROACHESIN MALE-MEDIATED DEVELOPMENTAL TOXICITYA. Effects of Toxicants in Seminal FluidThe measurement of drugs or chemicals in the seminal fluid can be accomplished fairly readilywith the variety of chemical and physical techniques currently available. We know that drugs canbe transmitted to the female through the semen. One of the tasks facing investigators is theassessment of the consequences to progeny of the presence of drugs or chemicals in seminal fluid.Unless they are bound to spermatozoa, drugs or chemicals in the seminal fluid that enter the femaleare likely to be extensively diluted in the female reproductive tract before they reach the oocyte.Drugs bound either reversibly or irreversibly to spermatozoa may have greater access to theconceptus. Vasectomy experiments are useful in determining whether any effect on progeny outcomeis due to the presence of drug in the seminal fluid or the physical binding of the drug to thespermatozoon that fertilizes the egg. Females can be sequentially mated to a control (fertile) male,and then a drug-treated vasectomized male. Using this approach, it was shown that the effect ofcyclophosphamide after acute administration to male rats was mediated by metabolites of the drugin the seminal fluid, rather than by drug bound to spermatozoa. 33B. Effects on Sperm Quantity and CharacteristicsIdeally, we should like to use the male germ cell itself to evaluate the effects of toxicants with thepotential to cause developmental toxicity to the progeny. Most commonly, toxicants affect malegerm cell numbers, structure, motility, viability, or ability to fertilize the oocyte. 82© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 133The relationship between sperm numbers in the ejaculate and fertility and progeny outcomehas not been studied extensively. The administration of sustained release testosterone capsules torats reduced epididymal sperm reserves to less than 5% of control before fertility was affected, andno abnormal progeny outcome was noted with reduced sperm production. 83 While such directstudies have not been done in humans, contraceptive development studies indicate that sperm countsmay need to be reduced to less than 3 million per milliliter to produce effective contraception. 84Based on a large number of clinical studies of men seeking treatment for infertility, it would appearthat in spite of highly variable counts within a given individual, there is no correlation betweensperm count and the quality of progeny outcome. 85One of the striking characteristics of the spermatozoon is its potential for rapid progressivemotion. Computer assisted semen analysis allows for the determination of various parameters ofsperm movement, including forward progression rate and lateral head displacement. 86 Many drugsand environmental chemicals have been shown to modify sperm motility characteristics, while atleast one male-mediated developmental toxicant, cyclophosphamide, has been reported not to affectrat sperm motility. 86–88 Sperm motility may be an early and sensitive endpoint for the assessmentof the male reproductive toxicity of chemicals such as cadmium. 89 As this method becomes acceptedas a standard for the quantification of the effects of drugs on sperm motility, the impact of malemediateddevelopmental toxicants on this parameter of sperm function will become clearer.Sperm motility is needed not just for aiding the movement of sperm through the femalereproductive tract to the egg but also to allow the spermatozoon to “drill” its way through thevarious layers surrounding the egg. The first event in this process is the recognition by the zonapellucida (ZP-3 receptor protein) of specific proteins on the capacitated sperm membrane. 90 Theacquisition of this protein, as well as the ability of spermatozoa to undergo the acrosome reaction,can provide information about spermatozoal function. To date, this approach has not been used toassess the action of drugs on sperm function.It is perceived of as normal in humans that as many as 50% of spermatozoa have an abnormalappearance, including such features as misshapen nuclei and abnormal arrangements of axonemalelements in the tail. One of the consequences of such structural abnormalities may be the inabilityto fertilize. Among those sperm that are capable of fertilization, the consequences to progenyoutcome have not been well defined. However, with the introduction of intracytoplasmic sperminjection (ICSI), it has become possible to show that injection of abnormally shaped spermatozoacan result in progeny that produce spermatozoa with even greater defects. 91 The full range ofpotential effects of using ICSI remains to be determined, but this procedure does seem to beassociated with a higher rate of abnormal progeny outcome. Additional studies to ascertain theimpact of sperm with an abnormal appearance on pronuclear formation and early embryo developmentwill help to shed light on this possibility. 92The profile of proteins found in the sperm membrane changes almost continually during thetransit of spermatozoa through the epididymis. 93 Several of these proteins have been postulated toplay key roles in the process of sperm maturation and in the development of the components thatwill recognize the sperm receptor on the zona pellucida. 94 Simple tools, such as immunocytochemistryusing monoclonal antibodies or polyacrylamide gel electrophoresis, may be useful in identifyingdrug-dependent changes in sperm membranes and determining whether such changes havesignificant effects on sperm function and progeny outcome.During spermiogenesis, histones are removed and replaced with protamines, which are small,highly basic, sulfhydryl-rich proteins. This transition allows for a remarkably tight packaging ofchromatin in the sperm nucleus, the shape of which is species specific. 95 Interestingly, in humansthe absence of protamine 2 in sperm was associated with infertility, abnormal sperm penetrationrates, abnormal morphology, and decreased progressive motility, although fertilization after ICSIand early embryo development were unaffected. 96 Acrylamide is a male-mediated developmentaltoxicant that has been hypothesized to act, at least partially, by reacting directly with sperm nuclearprotamines. 97,98 Interaction with protamine may lead to chromosome breakage, possibly during© 2006 by Taylor & Francis Group, LLC

134 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONsperm chromatin remodeling after fertilization. Underprotaminated sperm bind to a fluorochrome,chromomycin A 3 (CMA 3 ). 99,100 Interestingly, the percentage of CMA 3 positive sperm was positivelycorrelated with the presence of endogenous nicks in sperm DNA and with low sperm counts, highpercentages of abnormal sperm, and low in vitro fertilization rates. 99,101 As spermatozoa progressfrom the testis to the cauda epididymidis, the extent to which chromatin is cross-linked by disulfidebonds increases remarkably. 102 The ability of drugs to interfere with the disulfide cross linkage ofchromatin may correlate with the loss of the ability of spermatozoa to produce normal offspring.C. Effects on Sperm QualityTo date, studies of male reproductive toxicity and male-mediated developmental toxicity in animalmodels have focused generally on “gross” abnormalities, such as pregnancy loss, growth retardation,or malformations, and effects on germ cell numbers, motility, morphology, or DNA. Some of themore subtle measures, such as changes in chromatin packaging, specific changes in the DNAstructure, imprinting errors, altered methylation patterns, altered template function, or transcriptionin the early embryo may affect progeny outcome and may serve as biomarkers for the assessmentof sperm quality or even toxicant exposure. The recent establishment of primordial germcell–derived permanent female and male murine embryonic cell lines presents an interesting avenueto explore the establishment of an in vitro method to assess germ cell damage and quality. 103Nevertheless, the use of an “isolated germ cell” model system to assess male-mediated developmentaltoxicity would presuppose that the male germ cell is the exclusive target. It assumes thatthe surrounding cellular organization and environment (germ cells at different stages of differentiation,and Sertoli, myoid, and Leydig cells) and the role of the oocyte are unimportant in translatingany male germ cell “hit” into developmental toxicity.1. Sperm Chromatin Packaging and FunctionExposure to a toxicant may affect sperm chromatin packaging and function; such an effect maybe detected as a change in gene expression profile, in the accessibility of chromatin to specificdyes, in the ability to decondense, or in pronuclear formation or template function. Interestingly,cyclophosphamide exposure affected male germ cell gene expression in a cell-, stage-, and treatment-specificmanner. Exposure to a single high dose most affected round spermatids, especiallytheir expression of heat shock proteins, their cochaperones, and genes associated with DNA repair.Chronic low dose treatment resulted in an overall decrease in gene expression in both pachytenespermatocytes and round spermatids (Figure 5.3). 104,105The ability of the sperm nucleus to decondense and serve as a template is another parameterthat could be very useful in determining the site of action of a toxicant on the germ cell. Decondensationof the chromatin of the mammalian spermatozoon takes place after fertilization, restoringthe paternal genome to an active conformation. 106 In vivo, the complete decondensation process ischaracterized by reduction of the disulfide bonds of the protamines, followed by degradation ofthese nuclear proteins and their replacement with histones. 107,108 Spermatozoa can be decondensedin vitro by incubating them with a reducing agent, such as dithiothreitol, and a protease, such asproteinase K. 102Although there is little in the literature on the effects of most drugs or toxicants on the abilityof spermatozoal nuclei to decondense, exposure to cyclophosphamide has been shown to alter invitro decondensation. 102 Spermatozoa from rats treated with cyclophosphamide for 1 week showedthe same decondensation pattern as did those from the control group. Conversely, while thedecondensation pattern of spermatozoa from rats exposed to cyclophosphamide for 6 weeks wassimilar to that of control for the first 60 min, marked chromatin dispersion was noted in the next30 min. The cell area, perimeter, curvature, length, and straight length were all significantly lessthan those of control spermatozoa. We speculate that other drugs may also alter the decondensation© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 135Paternal Cyclophosphamide TreatmentEffects on Germ CellsDNA damageAltered template functionAltered gene expressionEffects in EmbryosDNA damageDecreased DNA synthesisDysregulation of zygoticgene activationDecreased cell proliferationAbnormal Progeny OutcomeFigure 5.3Diagrammatic representation of the effects of paternal cyclophosphamide exposure on male germcells and embryos.pattern of spermatozoa, perhaps by affecting the cross-linking of protamines, and that this mayhave consequences on the ability of the male genome to be activated in the early embryo.Spermatozoal DNA template function is a measure of sperm nuclear decondensation. Byincubating rat sperm in vitro with cytoplasmic extracts of Xenopus laevis eggs and assessingchromatin decondensation and DNA synthesis, Sawyer and colleagues investigated how chemicaldamage affected nuclear activation. 109 Whereas exposure to a cross-linking agent blocked decondensation,treatment with a DNA base modifier, hydroxylamine, enhanced decondensation, inducedgross chromatin abnormalities, and increased [ 3 H]TTP incorporation into activated sperm nuclei. 109The availability of spermatozoal DNA for template function was not affected by 1 week of treatmentwith cyclophosphamide, but was markedly affected after 6 weeks of treatment with this drug. 110Pronucleus formation by human sperm has been studied extensively in denuded hamster eggsas an endpoint in fertility assessment. Using this approach, we demonstrated that the decondensationof spermatozoa from cyclophosphamide-treated males was more rapid than that of control spermatozoa.111 In addition, male pronucleus formation was early in rat oocytes sired by drug-treatedmales. One possibility is that cyclophosphamide-induced chromatin damage prevents “normal”condensation during spermiogenesis.A disturbance in male germ cell chromatin condensation, remodeling, and pronucleus formationmay result in dysregulation of zygotic gene activation, leading to adverse effects on embryonicdevelopment (Figure 5.3). In fact, transcription is turned on earlier in the male pronucleus than inthe female pronucleus. 112 The assessment of RNA synthesis in embryos sired by control andcyclophosphamide-treated males revealed that total RNA synthesis ([ 32 P]UTP incorporation) wasconstant in one to eight cell embryos sired by drug-treated fathers, while in control embryos RNAsynthesis increased fourfold, to peak at the four-cell stage. 111 Moreover, BrUTP incorporation intoRNA and Sp1 transcription factor immunostaining were increased and spread over both the cytoplasmicand nuclear compartments in two-cell embryos sired by cyclophosphamide-treated males. 111In contrast, both BrUTP incorporation and Sp1 immunostaining were in the nuclear compartmentonly in embryos fertilized by control spermatozoa.The profile of expression of specific genes was altered in embryos sired by drug-treated males,even as early as the one-cell stage. 111,113,114 In the one- and two-cell stage embryo, the relativeabundance of transcripts for candidate DNA repair, imprinted, growth factor, and cell adhesion© 2006 by Taylor & Francis Group, LLC

136 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONgenes was elevated significantly above control in embryos sired by cyclophosphamide-treatedmales; a peak in the expression of many of these genes was not observed until the eight-cell stagein control embryos. 111,113,114 Thus, paternal drug exposure temporally and spatially dysregulated ratzygotic gene activation, altering the developmental clock.2. Sperm Genetic IntegrityGenetic damage of the male germ cell is of major concern; such damage may be transmitted tothe offspring and lead to abnormal progeny outcome, not only in the F 1 generation but also insubsequent generations. Even fairly severely DNA-damaged sperm may be capable of fertilization.Moreover, some germ line alterations appear to cause phenotypic malformations. Measurementsof genetic damage to the male germ cell include dominant lethal and specific locus mutation testsas well as cytogenetic, fluorescent in situ hybridization (FISH), and polymerase chain reaction(PCR) based methods for the detection of DNA damage or mutations.Nonmammalian test systems have been used extensively in screening banks of drugs andchemicals for mutagenicity. These tests have been invaluable in selecting chemicals for further invivo genetic mutagenicity testing. One example of such a system involves the Japanese medaka.Use of arbitrarily primed PCR and fingerprinting in fish treated with γ-irradiation allows changesin the genomic DNA of individual progeny to be detected as bands lost or gained. 115 By takingadvantage of suitable reporter genes, one can use male germ cells of Drosophila to detect a spectrumof genetic damage, ranging from recessive lethal (or visible) mutations, deletions, reciprocaltranslocations, chromosome loss, or dominant lethal mutations to aneuploidy. 116A number of in vivo animal tests rely heavily on dominant characteristics in inbred mice.Dominant lethal and specific locus mutation tests are examples of in vivo animal tests that havebeen used extensively to identify the chemicals that are capable of mutating germ cells. 65,117–119 Inthese tests, male rodents are treated either acutely or chronically with the chemical to be investigatedand mated to control females. For the dominant lethal test, the outcome measured is embryolethality,usually postimplantation. Thus, the spermatozoa are capable of fertilizing an egg, but the conceptusfails to develop normally, either at the time of implantation or shortly thereafter. The specific locusmutation test uses mice with mutations in a number of loci coding for “visible” features, such ascoat color, and evaluates the ability of the chemical in question to cause a mutation in the malegerm cell at these loci. This approach has been very valuable in detecting a number of mammaliangerm cell mutagens. Both of these approaches use progeny outcome to measure the effects of thedrug on the male germ cell. More recently, transgenic mice have been used as a model with whichto study gene mutations during different phases of spermatogenesis. 120,121 The results of specific-locusmutation studies have suggested that exposure of spermatogonia to chemicals or radiation yields fewlarge lesions, while large lesions are common after exposure of postspermatogonial germ cells.At the chromosome level, it may be possible to detect bulky deletions, aneuploidy, or chromosomalduplications by use of cytogenetic approaches. 122 The inherent difficulty in cytogeneticanalysis of spermatozoa has been the lack of mitosis and hence of chromosomal structure. It wasonly by fertilizing denuded hamster eggs with the sperm in question that chromosomal structuresin the male pronucleus could be analyzed. This approach was used to identify the effects of age,x-irradiation, and drugs on the chromosomal banding pattern of human sperm. 123 More recently,FISH was developed for the analysis of aneuploidy in the sperm genome. 124 This method involvesthe staining of specific chromosomal regions with fluorescent-labeled complementary DNAsequences, allowing the identification of sperm with chromosomal abnormalities, such as trisomiesor aneuploidies. FISH has been used extensively to study human sperm and was adapted for therat by Wyrobek’s group to apply this powerful tool to toxicology. 125Other tests that have detected DNA damage in male germ cells include the unscheduled DNAsynthesis assay, the DNA alkaline elution assay, the single cell gel electrophoresis (SCGE or comet)assay, and the sperm chromatin structure assay (SCSA TM ). In the unscheduled DNA synthesis assay,© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 137the repair of chemically induced DNA damage is assessed in postmitotic male germ cells bymeasuring the incorporation of radiolabeled thymidine into their DNA. 126 Because of their activeDNA replication, unscheduled DNA synthesis cannot be determined in spermatogonia or preleptoteneprimary spermatocytes. Unscheduled DNA synthesis does not occur in later stage spermatidsor in spermatozoa because of the loss of the enzymes involved in DNA repair; both of these phasesof spermatogenesis are thus susceptible to damage leading to developmental toxicity. Therefore,this test is selectively useful for assessing DNA repair in late spermatocytes and early spermatids.In the alkaline elution assay, germ cells that have been exposed to a test drug or chemical arelysed and decondensed on a filter prior to the addition of a highly alkaline buffer. 110,127,128 Underalkaline conditions, the DNA unwinds and is eluted through the filter at a rate reflecting the extentto which the test chemical has resulted in DNA single-stand breaks or cross-links. One week oftreatment with cyclophosphamide caused DNA single-strand breaks that could be detected only inthe presence of proteinase K in the lysis solution, but no DNA cross-links were observed. 110 Incontrast, 6 weeks of treatment with cyclophosphamide induced a significant increase in both DNAsingle-strand breaks and cross-links in spermatozoal nuclei; the cross-links were due primarily toDNA–DNA linkages. 110 In general, there was a close correlation between the DNA damageresponses of human and rat testicular cells as assessed with alkaline elution. 129 The detection andlocalization of drug adducts within the genome of the male germ cell would add additionalinformation about the target of such DNA damaging agents.Although alkaline elution provides a powerful test of the interaction of a drug with chromatin ina large population of cells, it cannot be used on an individual cell basis. In the comet assay, individualcells, lysed and decondensed, are electrophoresed in agar after treatment with alkali to assess doublestrandbreaks or under neutral conditions to identify single-strand breaks. 130 The electric field causesthe migration of fragments of DNA. The smaller the fragment, the greater is the migration or “comet.”The DNA can be visualized with a fluorescent dye, and the relative amount of DNA that migrated, aswell as the distance it migrated, provides quantitative data on DNA integrity after drug exposure. Thereis a highly significant correlation between DNA fragmentation as detected by the comet assay inejaculated spermatozoa and infertility. 130 When the alkaline comet assay was used to assess DNAdamage in one-cell embryos sired by cyclophosphamide-treated males, a significantly higher percentage(68%) of the embryos fertilized by drug-exposed spermatozoa displayed a comet indicative of DNAdamage compared with embryos sired by control males (18%). 114The sperm chromatin structure assay is an indirect indicator of DNA damage because it measuresthe amount of single-stranded DNA after treatments that normally do not denature sperm DNA(heat or acid pH). 131,132 The test employs the unique metachromatic and equilibrium stainingproperties of acridine orange, a dye that fluoresces green when intercalated into double-stranded“native” DNA and red when bound to single-stranded “denatured” DNA (or RNA). Alternateapproaches to assess sperm genetic integrity include the terminal deoxynucleotidyl transferase(TDT)-mediated d UTP nick-end labeling (TUNEL) and in situ nick translation (NT) assays. 133 Inthese assays, sperm are labeled with fluorescently tagged DNA precursors at sites of single- and/ordouble-stranded breaks in DNA. Endogenous DNA breaks in human spermatozoa have beendemonstrated with the NT assay. 99There is evidence in most genetic mutation test systems for “hot spots” or loci that are moresusceptible to mutations. This specificity has also been observed for the specific locus mutationtest. 117 The importance of “specific” genes or chromosomes as targets in mediating the adverseeffects of chemicals on male germ cells is not known. Dubrova and colleagues have suggested theuse of hypervariable tandem repeat loci to evaluate germ line mutation induction in mice andhumans. 134,135 These loci are capable of detecting changes in mutation rates in samples fromrelatively small populations because they have a very high spontaneous mutation rate in bothhumans and mice. 136–139The position of specific genes on DNA loops attached to the nuclear matrix is constant. 95 Hence,we can speculate that specific gene loci may be targeted even by “nonspecific” exposures. Any© 2006 by Taylor & Francis Group, LLC

138 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONselectivity of the effect of male-mediated developmental toxicants on sperm genetic integrity maybe based on the DNA sequence and/or conformation in chromatin, as well as on epigeneticmodifications.3. Epigenetic ChangesThere is evidence that the male and female genomes are not equivalent during development. 140,141The presence of both the maternal and paternal genome is essential for normal development in therodent. Surani and his co-workers have suggested that chromosomes are imprinted in a germ-linespecific manner during gametogenesis. 142,143 Thus, the absence of the male genome in a zygoteleads to embryos with poor development of the extraembryonic tissues, while the absence of thematernal genome results in embryos having markedly reduced embryonic tissues. The exposure ofmale germ cells to a developmental toxicant could be embryolethal if the toxicant targeted genesthat are paternally imprinted. A number of studies have implicated the methylation status of variousregions of the gene in imprinting. 144 Maternally imprinted genes, such as Snrpn, Mest, and Peg3,are unmethylated in sperm and 100% methylated in mature oocytes, whereas the 5′ region of H19,a paternally imprinted gene, is completely methylated in sperm and unmethylated in oocytes. 145The treatment of male rats with 5-azacytidine, a drug that blocks DNA methylation, resulted in anincrease in preimplantation loss when germ cells were first exposed as spermatogonia or spermatocytes.146,147The male genome is required for normal development of the trophectoderm. In embryos siredby cyclophosphamide-treated male rats, cell death occurred selectively in those tissues derived fromthe inner cell mass, while the trophoblast-derived trophectoderm cells appeared morphologicallynormal. 148 Thus, exposure of the male rat to cyclophosphamide may affect paternal genes essentialfor the development of inner cell mass–derived tissues in the embryo, sparing those genes requiredfor normal trophectoderm development. This is not what would have been predicted from thenuclear transplantation experiments cited above. However, this is consistent with the tissue specificityof the effects of radiation on early embryos. 149 Transgenic mouse experiments have shownthat inner cell mass–derived tissues are eliminated in mice deficient in fibroblast growth factor-4. 150 Thus, inner cell mass cells have different growth factor requirements than the trophectoderm.Moreover, the male genome is essential for the development of the inner cell mass as well astrophectoderm-derived cells.IV. SUMMARY AND CONCLUSIONSOver the past few decades it has become clear that drugs given to the father may affect his progeny’soutcome. The range of effects that can occur encompasses infertility and reduced fertility, as wellas malformations, growth retardation, and behavioral alterations in the progeny. Furthermore, it isapparent that the germ cell line of the progeny may be affected. Technological advances in themethods available to study changes in DNA at the molecular level help to elucidate the molecularmechanisms mediating these consequences of paternal drug exposure for the progeny. Althoughfew of these observations have been extended to the human, it seems essential that carefully designedclinical and epidemiologic studies be undertaken to establish the extent to which such effects maycontribute to infertility and/or abnormal progeny outcome in humans.ACKNOWLEDGMENTSThe studies from our laboratories were done with the support of grants from the Canadian Institutesof Health Research.© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 139References1. Robaire, B. and Hales, B.F., Eds., Advances in Male-Mediated Developmental Toxicity. Advances inExperimental Medicine and Biology, Vol. 518, Kluwer Academic/Plenum Press, New York, 2003.2. Shepard, T.H., Ed., Catalog of Teratogenic Agents, 7th ed., Johns Hopkins University Press, Baltimore,MD, 1992.3. Sawyer, D.E. and Aitken, R.J., Male mediated developmental defects and childhood disease, Reprod.Med. Rev., 8, 107, 2001.4. McDonald, A.D., McDonald, J.C., Armstrong, B., Cherry, N.M., Nolin, A.D., and Robert, D., Fathers’occupation and pregnancy outcome, Br. J. Ind. Med., 46, 329, 1989.5. Brender, J.D. and Suarez, L., Paternal occupation and anencephaly, Am. J. Epidemiol., 131, 517, 1990.6. Olshan, A.F., Teschke, K., and Baird, P.A., Birth defects among offspring of firemen, Am. J. Epidemiol.,131, 312, 1990.7. Olshan, A.F., Teschke, K., and Baird, P.A., Paternal occupation and congenital anomalies, Am. J. Ind.Med., 20, 447, 1991.8. Savitz, D.A., Whelan, E.A., and Kleckner, R.C., Effect of parents’ occupational exposures on risk ofstillbirth, preterm delivery, and small-for-gestational-age infants, Am. J. Epidemiol. 129, 1201, 1989.9. Savitz, D.A., Sonnenfeld, N.L., and Olshan, A.F., Review of epidemiologic studies of paternal occupationalexposure and spontaneous abortion, Am. J. Ind. Med., 25, 361, 1994.10. Friedman J.M., Does Agent Orange cause birth defects? Teratology, 29, 193, 1984.11. Field, B. and Kerr, C., Reproductive behaviour and consistent patterns of abnormality in offspring ofVietnam veterans, J. Med. Genet., 25, 819, 1988.12. Kaye, C.I., Rao, S., Simpson, S.J., Rosenthal, F.S., and Cohen, M.M., Evaluation of the chromosomedamage in males exposed to Agent Orange and their families, J. Craniofac. Genet. Dev. Biol. Suppl.,1, 259, 1985.13. Stellman, S.D., Stellman, J.M., and Sommer, J.F. Jr., Health and reproductive outcomes amongAmerican legionnaires in relation to combat and herbicide exposure in Vietnam, Environ. Res., 47,150, 1988.14. Arbuckle, T.E., Lin, Z., and Merv, L.S., An exploratory analysis of the effect of pesticide exposureon the risk of spontaneous abortion in an Ontario farm population. Environ. Health Perspect., 109,851, 2001.15. Bonde, J.P., Joffe, M., Apostoli, P., Dale, A., Kiss, P., Spano, M., Caruso, F., Giwercman, A., Bisanti,L., Porru, S., Vanhoorne, M., Comhaire, F., and Zschiesche, W., Sperm count and chromatin structurein men exposed to inorganic lead: lowest adverse effect levels, Occup. Environ. Med., 59, 234, 2002.16. Little, J. and Vainio, H., Mutagenic lifestyles? A review of evidence of associations between germcellmutations in humans and smoking, alcohol consumption and use of ‘recreational’ drugs, Mutat.Res., 313, 131, 1994.17. Savitz, D.A., Schwingl, P.J., and Keels, M.A., Influence of paternal age, smoking, and alcoholconsumption on congenital anomalies, Teratology, 44, 429,1991.18. Davis, D.L., Paternal smoking and fetal health, Lancet, i, 123, 1991.19. Robbins, W.A., FISH (fluorescence in situ hybridization) to detect effects of smoking, caffeine, andalcohol on human sperm chromosomes, in Advances in Male-Mediated Developmental Toxicity.Advances in Experimental Medicine and Biology, Vol. 518, Robaire, B. and Hales, B.F., Eds., KluwerAcademic/Plenum Press, New York, 2003, chap 6.20. Wong, W.Y., Thomas, C.M.G., Merkus, H.M.W.M., Zielhuis, G.A, Doesburg, W.H., and Steegers-Theunissen, R.P.M., Cigarette smoking and the risk of male factor subfertility: minor associationbetween cotinine in seminal plasma and semen morphology, Fertil. Steril., 74, 930, 2000.21. Vogt, H.J., Heller, W.D., and Borelli, S., Sperm quality of healthy smokers, ex-smokers, and neversmokers,Fertil. Steril., 45,106, 1986.22. Zavos, P.M., Correa, J.R., Karagounis, C.S., Ahparaki, A., Phoroglou, C., Hicks, C.L., and Zarmakoupis-Zavos,P.N., An electron microscope study of the axonemal ultrastructure in human spermatozoafrom male smokers and nonsmokers, Fertil. Steril., 69, 430, 1998.23. Zenzes, M.T., Bielecki, R., and Reed, T.E., Detection of benzo(a)pyrene diol epoxide-DNA adductsin sperm of men exposed to cigarette smoke, Fertil. Steril., 72, 330, 1999.© 2006 by Taylor & Francis Group, LLC

140 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION24. Dellarco, V.L., Mavournin, K.H., and Waters, M.D., Aneuploidy data review committee: Summarycompilation of chemical data base and evaluation of test methodology, Mutat. Res., 167, 149, 1986.25. Senturia, Y.D., Peckham C.S., and Peckham, M.J., Children fathered by men treated for testicularcancer, Lancet, ii, 766, 1985.26. Fried, P., Steinfeld, R., Casileth, B., and Steinfeld, A., Incidence of developmental handicaps amongthe offspring of men treated for testicular seminoma, Int. J. Androl., 10, 385, 1987.27. Mulvihill, H.J., Reproductive outcomes among men treated for cancer, in Male-Mediated DevelopmentalToxicity, Olshan, A.F. and Mattison, D.R., Eds., Plenum Press, New York, 1994, p. 197.28. Olshan, A.F., Breslow, N.E., Daling, J.R., Falleta, J.M., Grufferman, S., Robison, L. L., et al. Wilm’stumor and paternal occupation, Cancer Res, 50, 3212, 1990.29. Eliasson, R. and Dornbusch, K., Secretion of metronidazole into the human semen, Int. J. Androl., 3,236, 1980.30. Malmborg, A.S., Antimicrobial drugs in human seminal plasma, J. Antimicrobe Chemother., 6, 483, 1978.31. Mann, T. and Lutwak-Mann, C., Passage of chemicals into human and animal semen: mechanismsand significance, CRC Crit. Rev. Toxicol., 11, 1, 1982.32. Forrest, J.B., Turner, T.T., and Howards, S.S., Cyclophosphamide, vincristine and the blood testisbarrier, Invest. Urol., 18, 443, 1981.33. Robaire, B. and Hales, B.F., Post-testicular mechanisms of male-mediated developmental toxicity, inMale-Mediated Developmental Toxicity, Mattison, D.R. and Olshan, A.F., Eds., Plenum Press, NewYork, 1994, p. 93.34. Soyka, L.F., Peterson, J.M., and Joffe, J.M., Lethal and sublethal effects on the progeny of male ratstreated with methadone, Toxicol. Appl. Pharmacol., 45, 797, 1978.35. Lutwak-Mann, C., Observations on the progeny of thalidomide-treated male rabbits, Br. Med. J., 1,1090, 1964.36. Hales, B.F., Smith, S., and Robaire, B., Cyclophosphamide in the seminal fluid of treated males:Transmission to females by mating and effects on progeny outcome, Toxicol. Appl. Pharmacol., 84,423, 1986.37. Hales, B.F. and Robaire, B., The male-mediated developmental toxicity of cyclophosphamide, inMale-Mediated Developmental Toxicity, Mattison, D.R. and Olshan, A.F., Eds., Plenum Press, NewYork, 1994, p. 105.38. Ratcliffe, J.M., Paternal exposures and embryonic or fetal loss: the toxicologic and epidemiologicevidence, in Male-Mediated Developmental Toxicity, Mattison, D.R. and Olshan, A.F., Eds., PlenumPress, New York, 1994, p. 185.39. Winder, C., Reproductive and chromosomal effects of occupational exposure to lead in the male,Reprod. Toxicol., 3, 221, 1989.40. Alexander, B.H., Checkoway, H., van Netten, C., Kaufman, J.D., Vaughan, T.L., Mueller, B.A., andFaustman, E.M., Paternal occupational lead exposure and pregnancy outcome. Int. J. Occup. Environ.Health, 2, 280, 1996.41. Foster, W.G., McMahon, A., and Rice, D.C., Sperm chromatin structure is altered in cynomolgusmonkeys with environmentally relevant blood lead levels, Toxicol. Ind. Health, 12, 723, 1996.42. Johansson, L. and Wide, M., Long-term exposure of the male mouse to lead: effects on fertility,Environ. Res., 41, 481, 1986.43. Stowe, V.M. and Goyer, R., Reproductive ability and progeny of F1 lead-toxic rats. Fertil. Steril., 22,755, 1971.44. Brady, K., Herrera, Y., and Zenick, H., Influence of paternal lead exposure on subsequent learningability of offspring, Pharmacol. Biochem. Behav., 3, 561, 1975.45. Slutsky, M., Levin, J.L., and Levy, B.S., Azoospermia and oligospermia among a large cohort ofDBCP applicators in 12 countries, Int. J. Occup. Environ. Health, 5, 116, 1999.46. Leone, M., Costa, M., Capitanio, G.L., Palmero, S., Prati, M., and Leone, M.M., Dibromochloropropane(DBCP) effects on the reproductive function of the adult male rat, Acta Eur. Fertil.,19, 99,1988.47. Whorton, M.D. and Foliart, D.E., Mutagenicity, carcinogenicity and reproductive effects of dibromochloropropane(DBCP), Mutat. Res.,123, 13, 1983.48. Thornton, S.R., Schroeder, R.E., Robison, R.L., Rodwell, D.E., Penney, D.A., Nitschke, K.D., andSherman, W.K., Embryo-fetal developmental and reproductive toxicology of vinyl chloride in rats,Toxicol. Sci., 68, 207, 2002.© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 14149. Anderson, D., Genetic and reproductive toxicity of butadiene and isoprene, Chem. Biol. Interact.,135–136, 65, 2001.50. Pacchierotti, F., Adler, I.D., Anderson, D., Brinkworth, M., Demopoulos, N.A., Lahdetie, J., Osterman-Golkar, S., Peltonen, K., Russo, A., Tates, A., and Waters, R., Genetic effects of 1,3-butadiene andassociated risk for heritable damage, Mutat. Res., 397, 93, 1998.51. Shukla, Y. and Taneja, P., Mutagenic potential of cypermethrin in mouse dominant lethal assay,Environ. Pathol. Toxicol. Oncol., 21, 259, 2002.52. Jha, A.M. and Bharti, M.K., Mutagenic profiles of carbazole in the male germ cells of Swiss albinomice, Mutat. Res., 500, 97, 2002.53. Tyla, R.W., Friedman, M.A., Losco, P.E., Fisher, L.C., Johnson, K.A., Strother, D.E., and Wolf, C.H.,Rat two-generation reproduction and dominant lethal study of acrylamide in drinking water, Reprod.Toxicol., 14, 385, 2000.54. Cai, L., Hales, B.F., and Robaire, B., Induction of apoptosis in the germ cells of adult male rats afterexposure to cyclophosphamide, Biol. Reprod., 56, 1490, 1997.55. Adler, I.D., Kliesch, U., Jentsch, I., and Speicher, M.R., Induction of chromosomal aberrations bydacarbazine in somatic and germinal cells of mice, Mutagenesis, 17, 383, 2002.56. Witt, K.L. and Bishop, J.B., Mutagenicity of anticancer drugs in mammalian germ cells, Mutat. Res.,355, 209, 1996.57. Hooser, S.B., van Dijk-Knijnenburg, W.C., Waalkens-Berendsen, I.D., Smits-van Prooije, A.E., Snoeij,N.J., Baan, R.A., and Fichtinger-Schepman, A.M., Cisplatin-DNA adduct formation in rat spermatozoaand its effect on fetal development, Cancer Lett., 151, 71, 2000.58. Seethalakshmi, L., Flores, C., Kinkead, T., Carboni, A.A., Malhotra, R.K., and Menon, M., Effectsof subchronic treatment with cis-platinum on testicular function, fertility, pregnancy outcome, andprogeny. J. Androl., 13, 65,1992.59. Russell, L.B. and Rinchik, E.M., Structural differences between specific-locus mutations induced bydifferent exposure regimes in mouse spermatogonial stem cells, Mutat. Res., 288, 187, 1993.60. Nagao, T. and Fukikawa, K., Genotoxic potency in mouse spermatogonial stem cells of triethylenemelamine,mitomycin C, ethylnitrosourea, procarbazine, and propyl methanesulfonate as measuredby F 1 congenital defects, Mutat. Res., 229, 123, 1990.61. Kato, M., Makino, S., Kimura, H., Ota, T., Furuhashi,T., and Nagamura, Y., Sperm motion analysisin rats treated with adriamycin and its applicability to male reproductive toxicity studies, J. Toxicol.Sci., 26, 51, 2001.62. Russell, L.B., Hunsicker, P.R., Kerley, M.K., Johnson, D.K., and Shelby, M.D., Bleomycin, unlikeother male-mouse mutagens, is most effective in spermatogonia, inducing primarily deletions, Mutat.Res., 469, 95, 2000.63. Parvinen, L.M., Soderstrom, K.O., and Parvinen, M., Early effects of vinblastine and vincristine onthe rat spermatogenesis: analyses by a new transillumination-phase contrast microscopic method, Exp.Pathol. (Jena), 15, 85, 1978.64. Nomura, T., Male-mediated teratogenesis: ionizing radiation/ethylnitrosourea studies, in Male-MediatedDevelopmental Toxicity, Mattison, D.R. and Olshan, A.F., Eds., Plenum Press, New York, 1994,p. 117.65. Russell, L.B., Effects of spermatogenic cell type on quantity and quality of mutations, in Male-Mediated Developmental Toxicity, Mattison, D.R. and Olshan, A.F., Eds., Plenum Press, New York,1994, p. 37.66. Trasler, J.M., Hales, B.F., and Robaire, B., Paternal cyclophosphamide treatment of rats causes fetalloss and malformations without affecting male fertility, Nature, 316, 144, 1985.67. Trasler, J.M., Hales, B.F., and Robaire, B., Chronic low dose cyclophosphamide treatment of adultmale rats: effect on fertility, pregnancy outcome and progeny, Biol. Reprod., 34, 275, 1986.68. Trasler, J.M., Hales, B.F., and Robaire, B., A time course study of chronic paternal cyclophosphamidetreatment in rats: effects on pregnancy outcome and the male reproductive and hematologic systems,Biol. Reprod., 37, 317, 1987.69. Qiu, J., Hales, B.F., and Robaire, B., Adverse effects of cyclophosphamide on progeny outcome canbe mediated through post-testicular mechanisms in the rat, Biol. Reprod., 46, 926, 1992.70. Anderson, D., Bishop, J.B., Garner, R.C., Ostrosky-Wegman, P., and Selby, P.B., Cyclophosphamide:review of its mutagenicity for an assessment of potential germ cell risks, Mutat. Res., 330, 115, 1995.© 2006 by Taylor & Francis Group, LLC

142 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION71. Hales, B.F. and Robaire, B., Reversibility of the effects of chronic paternal exposure to cyclophosphamideon pregnancy outcome in rats, Mutat. Res., 229, 129, 1990.72. Fabricant, J.D., Legator, M.S., and Adams, P.M., Post-meiotic cell mediation of behavior in progenyof male rats treated with cyclophosphamide, Mutat. Res., 119, 185, 1983.73. Auroux, M., Dulioust, E., Selva, J., and Rince, P., Cyclophosphamide in the F 0 male rat: physical andbehavioral changes in three successive adult generations, Mutat. Res., 229, 189, 1990.74. Generoso, W.M., Cattanach, B., and Malashenko, A.M., Mutagenicity of selected chemicals in mammals;the heritable translocation test, in Comparative Chemical Mutagenesis, de Serres, F.J. andShelby, M.D., Eds., Plenum Press, New York, 1981, p. 681.75. Backer, L.C., Gibson, M.J., Moses, M.J., and Allen, J.W., Synaptonemal complex damage in relationto meiotic chromosome aberration after exposure of male mice to cyclophosphamide, Mutat. Res.,203, 317, 1988.76. Sotomayor, R.E. and Cumming, R.B., Induction of translocations by cyclophosphamide in differentgerm cell stages of male mice: cytological characterization and transmission, Mutat. Res., 27, 375,1975.77. Orgebin-Crist, M-C. and Olsen, G.E., Epididymal sperm maturation, in The Male in Farm AnimalReproduction, Courot, M., Ed., Martinus Nijhoff, Amsterdam, 1984, p. 80.78. Chellman, G.J., Bus, J.S., and Working, P.K., Role of epididymal inflammation in the induction ofdominant lethal mutations in Fischer 344 rat sperm by methyl chloride, Proc. Natl. Acad. Sci. USA,83, 8087, 1986.79. Chellman, G.J., Morgan, K.T., Bus, J.S., and Working, P.G., Inhibition of methyl chloride toxicity inmale F-344 rats by the anti-inflammatory agent BW755C, Toxicol. Appl. Pharmacol., 85, 367, 1986.80. Qiu, J., Hales, B.F., and Robaire, B., Adverse effects of cyclophosphamide on progeny outcome canbe mediated through post-testicular mechanisms in the rat, Biol. Reprod., 46, 926, 1992.81. Hales, B.F., Crosman, K., and Robaire, B., Increased postimplantation loss and malformations amongthe F 2 progeny of male rats chronically treated with cyclophosphamide, Teratology, 45, 671, 1992.82. Wyrobek, A.J., Methods and concepts in detecting abnormal reproductive outcomes of paternal origin,Reproductive Toxicol., 7, 3, 1993.83. Robaire, B., Smith, S., and Hales, B., Suppression of spermatogenesis by testosterone in adult malerats: effect on fertility, progeny outcome and pregnancy, Biol. Reprod., 31, 221,1984.84. Handelsman, D.J., Farley, T.M., Pregoudor, A., and Waites, G.M., Factors in non-uniform inductionof azoospermia by testosterone enanthate in normal men. World Health Organization Task Force forMethods for the Regulation of Male Fertility. Fertil. Steril., 63, 125, 1995.85. Robaire, B., Pryor, J., and Trasler, J.M., Eds., Handbook of Andrology, American Society of Andrology,Allen Press, Inc., Lawrence, KS, 1995.86. Slott, V.L., Suarez, J.D., Poss, P.M., Linder, R.E., Strader, L.F., and Perreault, S.D., Optimization ofthe Hamilton-Thorn computerized sperm motility analysis system for use with rat spermatozoa intoxicological studies, Fund. Appl. Toxicol., 21, 298, 1993.87. Linder, R.E., Strader, L.F., Slott, V.L., and Suarez, J.D., Endpoints of spermatotoxicity in the rat aftershort duration exposures to fourteen reproductive toxicants, Reprod. Toxicol., 6, 491, 1992.88. Higuchi, H., Nakaoka, M., Kawamura, S., Kamita, Y., Kohda, A., and Seki, T., Application ofcomputer-assisted sperm analysis system to elucidate lack of effects of cyclophosphamide on ratepididymal sperm motion, J. Toxicol. Sci., 26, 75, 2001.89. Xu, L.C., Wang, S.Y., Yang, X.F., and Wand, X.R., Effects of cadmium on rat sperm motility evaluatedwith computer assisted sperm analysis, Biomed. Environ. Sci., 14, 312, 2001.90. Bleil, J.D. and Wasserman, P.W., Identification of a ZP-3 binding protein on acrosome-intact mousesperm by photoaffinity crosslinking, Proc. Natl. Acad. Sci. USA, 87, 5563, 1990.91. Akutsu, H., Tres, L.L., Tateno, H., Yanagimachi, R., and Kierszenbaum, A.L., Offspring from normalmouse oocytes injected with sperm heads from the azh/azh mouse display more severe sperm tailabnormalities than the original mutant, Biol. Reprod., 64, 249, 2001.92. Hansen, M., Kurinczuk, J.J., Bower, C., and Webb, S., The risk of major birth defects after intracytoplasmicsperm injection and in vitro fertilization, N. Engl. J. Med., 346, 725, 2002.93. Robaire, B. and Hermo, L., Efferent ducts, epididymis, and vas deferens: structure, functions andtheir regulation, in The Physiology of Reproduction, Vol. 1, Knobil, E. and Neill, J.D., Eds. RavenPress, New York, 1988, p. 999.© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 14394. Robaire, B. and Hinton, B.T., Eds. The Epididymis: From Molecules to Clinical Practice — AComprehensive Survey of the Efferent Ducts, the Epididymis and the Vas Deferens. Kluwer Academic/PlenumPublishers, New York, 2002.95. Ward, W.S. and Coffey, D.S., Specific organization of genes in relation to the sperm nuclear matrix,Biochem. Biophys. Res. Commun., 173, 20, 1990.96. Carrell, D.T. and Liu, L., Altered protamine 2 expression is uncommon in donors of known fertility,but common among men with poor fertilizing capacity, and may reflect other abnormalities ofspermiogenesis, J. Androl., 22, 604, 2001.97. Sega, G.A., Alcota, R.P.V., Tancongco, C.P., and Brimer, P.A., Acrylamide binding to the DNA andprotamine of spermiogenic stages in the mouse and its relationship to genetic damage, Mutat. Res.,216, 221, 1986.98. Dearfield, K.L., Douglas, G.R., Ehling U.H., Moore, M.M., Sega, G.A., and Brusick, D.J., Acrylamide:a review of its genotoxicity and an assessment of heritable genetic risk, Mutat. Res., 330, 71, 1995.99. Manicardi, G.C., Bianchi, P.G., Pantano, S., Azzoni, P., Bizzaro, D., Bianchi, U., and Sakkas, D.,Presence of endogenous nicks in DNA of ejaculated human spermatozoa and its relationship tochromomycin A3 accessibility, Biol. Reprod., 52, 864, 1995.100. Bizzaro, D., Manicardi, G.C., Bianchi, P.G., Bianchi, U., Mariethoz, E., and Sakkas, D., In-situcompetition between protamine and fluorochromes for sperm DNA, Mol. Hum. Reprod., 4, 127, 1998.101. Lolis, D., Georgiou, I., Syrrou, M., Zikopoulos, K., Konstantelli, M., and Messinis, I., ChromomycinA3-staining as an indicator of protamine deficiency and fertilization, Int. J. Androl., 19, 23, 1996102. Qiu, J.P., Hales, B.F., and Robaire, B., Effects of chronic low dose cyclophosphamide exposure onthe nuclei of rat spermatozoa, Biol. Reprod., 52, 33, 1995.103. Klemm, M., Genschow, E., Pohl, I., Barrabas, C., Liebsch, M., and Spielmann, H., Permanentembryonic germ cell lines of BALB/cJ mice — an in vitro alternative for in vivo germ cell mutagenicitytests, Toxicol. In Vitro, 15, 447, 2001.104. Aguilar-Mahecha, A., Hales, B.F., and Robaire, B., Acute cyclophosphamide exposure has germ cellspecific effects on the expression of stress response genes during rat spermatogenesis, Mol. Repro.Dev., 60, 302, 2001.105. Aguilar-Mahecha, A., Hales, B.F., and Robaire, B., Chronic cyclophosphamide treatment alters theexpression of stress response genes in rat male germ cells, Biol. Reprod., 66, 1024, 2002.106. Naish, S.J., Perreault, S.D., and Zirkin, B.R., DNA synthesis in the fertilizing hamster sperm nucleus:sperm template availability and egg cytoplasmic control, Biol. Reprod., 36, 245, 1987.107. Perreault, S.D., Naish, S.J., and Zirkin, B.R., The timing of hamster sperm nuclear decondensationand male pronucleus formation is related to sperm nuclear disulfide bond content, Biol. Reprod., 36,239, 1987.108. Perreault, S.D. and Zirkin, B.R., Sperm nuclear decondensation in mammals: role of sperm-associatedproteinase in vivo, J. Exp. Zool., 224, 253, 1982.109. Sawyer, D.E., Hillman, G.R., Uchida, T., and Brown, D.B., Altered nuclear activation parameters ofrat sperm treated in vitro with chromatin-damaging agents, Toxicol. Sci., 44, 52, 1998.110. Qiu, J., Hales, B.F., and Robaire, B., Damage to rat spermatozoal DNA after chronic cyclophosphamideexposure, Biol. Reprod., 53, 1465, 1995.111. Harrouk, W., Khatabaksh, S., Robaire, B., and Hales, B. F., Paternal exposure to cyclophosphamidedysregulates the gene activation program in rat pre-implantation embryos, Mol. Reprod. Dev., 57, 214,2000.112. Worrad, D.M., Ram, P.T., and Schultz, R.M., Regulation of gene expression in the mouse oocyte andearly preimplantation embryo: developmental changes in Sp1 and TATA box-binding protein, TBP,Development, 120, 2347, 1994.113. Harrouk, W., Robaire, B. and Hales, B.F., Paternal exposure to cyclophosphamide alters cell-cellcontacts and activation of embryonic transcription in the pre-implantation rat embryo, Biol. Reprod.,63, 74, 2000.114. Harrouk, W., Codrington, A., Vinson, R., Robaire, B., and Hales, B.F., Paternal exposure to cyclophosphamideinduces DNA damage and alters the expression of DNA repair genes in the rat preimplantationembryo, Mutat. Res. — DNA Repair, 461, 229, 2000.115. Kubota, Y., Shimada, A., and Shima, A., DNA alterations detected in the progeny of paternallyirradiated Japanese medaka fish (Oryzias latipes), Proc. Natl. Acad. Sci. USA, 92, 330, 1995.© 2006 by Taylor & Francis Group, LLC

144 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION116. Vogel, E.W. and Nivard, M.J.M., Model systems for studying germ cell mutagens: from flies tomammals, in Advances in Male-Mediated Developmental Toxicity, Advances in Experimental Medicineand Biology, Vol. 518, Robaire, B. and Hales, B.F., Eds., Kluwer Academic/Plenum Press, New York,2003, p. 99.117. Favor, J., Specific-locus mutation tests in germ cells of the mouse: an assessment of the screeningprocedures and the mutational events detected, in Male-Mediated Developmental Toxicity, Mattison,D.R. and Olshan, A.F., Eds., Plenum Press, New York, 1994, p. 23.118. Ehling, U.H., Dominant mutations in mice, in Male-Mediated Developmental Toxicity, Mattison, D.R.and Olshan, A.F., Eds., Plenum Press, New York, 1994, p. 49.119. Hales, B.F. and Cyr D.G., Study designs for the assessment of male-mediated developmental toxicity,in Advances in Male-Mediated Developmental Toxicity, Robaire, B. and Hales, B.F. Eds., KluwerAcademic/Plenum Press, New York, 2003, p. 271.120. van Delft, J.H., Bergmans, A., and Baan, R.A., Germ-cell mutagenesis in lambda lacZ transgenicmice treated with ethylating and methylating agents: Comparison with specific-locus test, Mutat. Res.,388, 165, 1997.121. Douglas, G.R., Gingerich, J.D., Soper, L.M., and Jiao, J., Toward an understanding of the use oftransgenic mice for the detection of gene mutations in germ cells, Mutat. Res., 388, 197, 1997.122. Allen, J.W., Collins, B.W., Cannon, R.E., McGregor, P.W., Afshari, A., and Fuscoe, J.C., Aneuploidytests: cytogenetic analysis of mammalian male germ cells, in Male-Mediated Developmental Toxicity,Mattison, D.R. and Olshan, A.F., Eds., Plenum Press, New York, 1994, p. 59.123. Martin, R.H., Rademaker, A., Hildebrand, K., Barnes, M., Arthur, K., Ringrose, T., Brown, I.S., andDouglas, G., A comparison of chromosomal aberrations induced by in vivo radiotherapy in humansperm and lymphocytes, Mutat. Res., 226, 21, 1989.124. Wyrobeck, A.J., Weier, H.-U., Robbins, W., Mehraein, Y., and Pinkel, D., Detection of sex-chromosomalaneuploidies in human sperm using two-color fluorescence in situ hybridization, Environ. Molec.Mutagen., 19, 72, 1992.125. Lowe, X.R., de Stoppelaar, J.M., Bishop, J., Cassel, M., Hoebee, B., Moore, D. 2nd, and Wyrobek,A.J., Epididymal sperm aneuploidies in three strains of rats detected by multicolor fluorescence insitu hybridization, Environ. Mol. Mutagen., 31, 125, 1998.126. Bentley, K.S., Sarrif, A.M., Cimino, M.C., and Auletta, A.E., Assessing the risk of heritable genemutation in mammals: Drosophila sex-linked recessive lethal test and tests measuring DNA damageand repair in mammalian germ cells, Environ. Molec. Mutagen., 23, 3, 1994.127. Kohn, K.W., Erickson, L.C., Ewig, R.A.G., and Friedman, C.A., Fraction of DNA from mammaliancells by alkaline elution, Biochemistry, 15, 4629, 1976.128. Sega, G.A., Sluder, A.E., McCoy, L.S., Owens, J.G., and Generoso, E.E., The use of alkaline elutionprocedures to measure DNA damage in spermiogenesis stages of mice exposed to methyl methanesulfonate,Mutat. Res., 159, 55, 1986.129. Bjorge, C., Brunborg, G., Wiger, R., Holme, J.A., Scholz, T., Dybing, E., and Soderlund, E.J., Acomparative study of chemically induced DNA damage in isolated human and rat testicular cells,Reprod. Toxicol., 10, 509, 1996.130. Irvine, D.S., Twigg, J.P., Gordon, E.L., Fulton, N., Milne, P.A., and Aitken, R.J., DNA integrity inhuman spermatozoa: relationships with semen quality, J. Androl., 21, 33, 2000.131. Evenson, D.P., Alterations and damage of sperm chromatin structure and early embryonic failure, inTowards Reproductive Certainty: Fertility and Genetics Beyond 1999, Jannsen, R. and Mortimer, D.,Eds., Parthenon Publishing Group Ltd, New York, 1999, p. 313.132. Evenson, D.P., Larson, K., and Jost, L.K., The sperm chromatin structure assay (SCSA TM ): clinicaluse for detecting sperm DNA fragmentation related to male infertility and comparisons with othertechniques. Andrology Lab Corner. J. Androl., 23, 25, 2002.133. Perreault, S.D., Aitken, R.J., Baker, H.W.G., Evenson, D.P., Huszar, D., Irvine, S., Morris, I.O., Morris,R.A., Robbins, W.A., Sakkas, D., Spano, M., and Wyrobek, A.J., Integrating new tests of sperm geneticintegrity into semen analysis: breakout group discussion, in Advances in Male-Mediated DevelopmentalToxicity, Advances in Experimental Medicine and Biology, Vol. 518, Robaire, B. and Hales,B.F., Eds., Kluwer Academic/Plenum Press, New York, 2003, p. 253.134. Dubrova, Y.E., Jeffreys, A.J., and Malashenko, A.M., Mouse minisatellite mutations induced byionizing radiation, Nature Genet. 5, 92, 1993.© 2006 by Taylor & Francis Group, LLC

PATERNALLY MEDIATED EFFECTS ON DEVELOPMENT 145135. Dubrova, Y.E., Nesterov, V.N., Krouchinsky, N.G., Ostapenko, V.A., Neumann, R., Neil, D.L., andJeffreys, A.J., Human minisatellite mutation rate after the Chernobyl accident, Nature, 380, 683, 1996.136. Jeffreys, A.J., Royle, N.J., Wilson, V., and Wong, Z., Spontaneous mutation rate to new length allelesat tandem-repeat hypervariable loci in human DNA, Nature, 332, 278, 1988.137. Vergnaud, G. and Denoeud, F., Minisatellites: mutability and genome architecture, Genome Res., 10,899, 2000.138. Kelly, R., Bulfield, G., Collick, A., Gibbs, M., and Jeffreys, A.J., Characterization of a highly unstablemouse minisatellite locus: evidence for somatic mutation during early development, Genomics, 5,844, 1989.139. Gibbs, M., Collick, A., Kelly, R., and Jeffreys, A.J., A tetranucleotide repeat mouse minisatellitedisplaying substantial somatic instability during early preimplantation development, Genomics,17,121, 1993.140. Barton, S.C., Surani, M.A.H., and Norris, M.L., Role of paternal and maternal genomes in mousedevelopment, Nature, 311, 374, 1984.141. McGrath, J. and Solter, D., Completion of mouse embryogenesis requires both the maternal andpaternal genomes, Cell, 37, 170, 1984.142. Surani, M.A.H., Barton, S.C, and Norris, M.L., Development of reconstituted mouse eggs suggestsimprinting of the genome during gametogenesis, Nature, 308, 548, 1984.143. Surani, M.A.H., Evidences and consequences of differences between maternal and paternal genomesduring embryogenesis in the mouse, in Experimental Approaches to Mammalian Embryonic Development,Rossant, J. and Pedersen, R. A., Eds., Cambridge University Press, New York, 1986, p. 401.144. Bartolomei, M.S., Webber, A.L., Brunkow, M.E., and Tilghman, S.M., Epigenetic mechanisms underlyingthe imprinting of the mouse H19 gene, Genes Dev., 7, 1663, 1993.145. Lucifero, D.L., Mertineit, C., Clarke, H.J., Bestor, T.H., and Trasler, J.M., Methylation dynamics ofimprinted genes in murine germ cells, Genomics, 79, 530, 2002.146. Doerksen, T. and Trasler, J.M., Developmental exposure of male germ cells to 5-azacytidine resultsin abnormal preimplantation development in rats, Biol. Reprod., 55,1155, 1996.147. Doerksen, T., Benoit, G., and Trasler, J.M., DNA hypomethylation of male germ cells by mitotic andmeiotic exposure to 5-azacytidine results in altered testicular histology, Endocrinology, 141, 3235,2000.148. Kelly, S.M., Robaire, B., and Hales, B.F., Paternal cyclophosphamide treatment causes postimplantationloss via inner cell mass-specific cell death, Teratology, 45, 313, 1992.149. Goldstein, L.S., Spindle, A.L., and Pedersen, R.A., X-ray sensitivity of the preimplantation mouseembryo in vitro, Radiat. Res., 62, 276, 1975.150. Feldman, B., Poueymirou, W., Papaioannou, V.E., DeChiara, T.M., and Goldfarb, M., Requirementof FGF-4 for postimplantation mouse development, Science, 267, 246, 1995.© 2006 by Taylor & Francis Group, LLC

CHAPTER 6Comparative Features of Vertebrate EmbryologyJohn M. DeSessoCONTENTSI. Introduction ........................................................................................................................147II. Comparative Placental Characteristics ..............................................................................148III. Embryological Processes ...................................................................................................154IV. Comparative Embryological Milestones............................................................................159Acknowledgments ..........................................................................................................................162References ......................................................................................................................................189References for Tables 3 to 12........................................................................................................191I. INTRODUCTIONMulticellular animals have limited life spans. Consequently, for a species to survive, a mechanismmust exist for the successive production of new generations. The solution to this problem lies inthe process of reproduction. This process typically involves the presence of two sexes, the productionby each sex of specialized cells called gametes, and a complicated series of events resulting in thejoining of two gametes to form a new individual. Gametes are referred to as haploid cells becausethey contain one-half the number of chromosomes found in somatic cells of the particular species.Male gametes (spermatozoa) are generated in the testes and are small, motile cells, millions ofwhich are produced daily. In contrast, female gametes (ova) are large, nonmotile cells that developin the ovaries. Relatively few ova are produced, and only a few hundred mature during thereproductive lifetime of female mammals.Fertilization is the union of a single spermatozoon and an ovum. It occurs in the femalereproductive tracts of birds and mammals and produces a new single celled organism, the zygote.Fertilization restores the diploid number of chromosomes, so that the zygote has the same amountof genetic material as did the somatic cells of its parents. Fertilization in mammals and birds alsodetermines the sex of the zygote and initiates the process of cleavage. Cleavage is a rapid seriesof mitotic divisions that allows the relatively large amount of cytoplasm contributed by the ovumto be divided into progressively smaller cells.In mammals, fertilization occurs in the uterine tube (oviduct). Cleavage divisions occur as thezygote progresses to the uterus, where it will become attached to the maternal uterine wall. Duringthis time, the zygote is surrounded by the zona pellucida, an acellular mucopolysaccharide layerthat prevents the zygote from implanting prematurely. When the zygote reaches the uterus, it is a147© 2006 by Taylor & Francis Group, LLC

148 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 6.1Gestational milestones for mammalsSpeciesGestational Milestone aA b B C D EPrimitive EarlyPartialUsualImplantation Streak Differentiation Closure c ParturitionRat 5–6 8.5 10 15 21–22Mouse 5 6.5 9 15 19–20Rabbit 7.5 7.25 9 18 30–32Hamster 4.5–5 7 8 13 16Guinea pig 6 12 14.5 ~29 67–68Monkey 9 17 21 ~44–45 166Human 6–7 13 21 ~50–56 266aIn gestational days; day of confirmed mating is gestational day 0.bLetters refer to positions on Figure 6.4 (conceptual roadmap of embryonic development).cMarks the end of major organogenesis for most organ systems.cluster of small cells surrounded by the zona pellucida; this cluster is called a morula. Subsequently,the zona pellucida thins, ruptures, and eventually disappears, while the morula cavitates to becomea sphere of cells surrounding a fluid-filled cavity. At this stage, the zygote is termed a blastocyst.In most mammals that are used in experimental studies, the blastocyst arises between days 5and 8 of gestation and attaches to the uterine wall during this time (Table 6.1). Two populationsof cells are recognized in a blastocyst. They include the outer sphere of cells, called the trophoblastthat gives rise to the placenta and fetal membranes, and a small cluster of cells on the inside, theinner cell mass that gives rise to the embryo proper.II. COMPARATIVE PLACENTAL CHARACTERISTICSOne of the earliest tasks of the blastocyst is the establishment of a mechanism for maintenance ofnutrient supply and disposal of metabolic wastes. This is accomplished through the developmentof a placenta from the trophoblast. The placenta and fetal (extraembryonic) membranes are temporaryorgans that form early in development and exist for a brief period compared to the life spanof the organism. Because of their importance to embryonic development, however, they will bedescribed in some detail before we return to further discussion of the embryo proper.The extraembryonic membranes provide nutrition, respiration, metabolic waste elimination, andprotection to the embryo and fetus, in addition to assisting in the establishment of embryonicvascularity. The four fetal membranes of vertebrates are the amnion, chorion, allantois, and yolksac. Not all vertebrate species exhibit all four membranes; for instance, animals that lay eggs inwater (anamniota) do not possess an amnion.A placenta is an organ composed of fetal and parental tissues that are intimately apposed forthe purpose of physiological exchange. 1 The fetal tissues of the placenta include one or more ofthe extraembryonic membranes (e.g., yolk sac, allantois), whereas the parental tissue is usuallypart of the uterus. The types of placentas can be described by the fetal membranes that participatein the apposition of fetal to maternal tissues. In general, the definitive placentas of eutherianmammals are formed from the outermost membrane of the embryonic vesicle (the avascularchorion), which is augmented by, and receives vascularization from, the allantois. This type ofplacenta is a chorioallantoic placenta. Most marsupial species develop placentas from the chorion,which is vascularized by the yolk sac. This type of placenta is called a choriovitelline (or yolk sac)placenta. (See discussion of the rodent “inverted” yolk sac placenta below.)The placenta and fetal membranes are tissues with diverse structures and functions. In addition,these tissues are dynamic and modify both their structures and functions during gestation. Consequently,when assessing the role of the placenta in developmental toxicity, one must be aware not© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 149Diffuse(Placenta Diffusa)Multiplex(Placenta Cotyledonaria)Banded(Placenta Zonaria)Discoid(Placenta Discoidalis)Figure 6.1Types of placentas, classified by shape. The images depict the outer surface of the chorion, thefetal membrane that is apposed to the maternal reproductive tract. The white territories representthe smooth portions (chorion laeve); the gray regions represent the part of the chorion (chorionfrondosum) that is modified to increase the surface area between embryo and mother. Note thatbanded placentas may exist as complete bands (illustrated) or as partial belts. Note also thatdiscoid placentas may exist as single placentas, as in humans, or as paired structures, as in rhesusmonkeys (illustrated).only of interspecies differences in placental structure and function, but also of differences in thestructure and function of the same placenta at different times in gestation.Placentas are classified according to their gross appearance, their mode of implantation, theirtype of modification of the chorionic surface to increase surface area, and the intimacy of embryonicinvasion into maternal tissues. 2,3 Because these differences can influence the efficiency or rate oftransfer of materials between mother and embryo, they will be described briefly.The outermost fetal membrane is the chorion. To the naked eye, it appears either as a smoothmembrane (chorion laeve) or as a roughened or fuzzy membrane (chorion frondosum). The distributionof the villous areas of the chorion may take on one of four shapes (see Figure 6.1).1. Diffuse (placenta diffusa): Villi are maintained over the entire chorion (e.g., pigs,* horses, humans[early in gestation], lemurs).2. Multiplex (placenta cotyledonaria): Villi are grouped in discrete rosettes (cotyledons) that areseparated by regions of smooth chorion (e.g., cattle, sheep, deer, ruminants).3. Banded (placenta zonaria): Villi assume a girdle-like configuration around the middle of thechorionic sac (e.g., carnivores — dogs, cats [complete band], bears [less than half]).4. Discoid (placenta discoidalis): Villi are grouped into one or two disk-shaped regions (e.g., insectivores,bats, rodents, nonhuman primates, definitive human placenta).* Villi in the pig are actually plicate elevations.© 2006 by Taylor & Francis Group, LLC

150 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONLumen of UterusCentral(Superficial)YolkSacChorionLumen of UterusEccentricYolkSacChorionInterstitialChorionFigure 6.2Types of placentas, classified by mode of implantation. The depth of implantation into the uterinewall increases from central, in which the conceptus essentially lies in the uterine lumen, to interstitial,in which the conceptus resides completely within the uterine wall and the uterine lumen is obliterated.The relationship of the chorionic sac to the uterine wall and lumen can be described in termsof the extent or the depth of embryonic implantation into the uterine wall, and three general typesof implantation can be distinguished 4 (see Figure 6.2).1. Central (superficial): The chorionic sac remains in contact with the main uterine lumen (e.g.,ungulates, carnivores, monkeys).2. Eccentric: The chorionic sac lies in a pocket or fold that is partially separated from the uterinelumen (e.g., rodents — early in gestation).3. Interstitial: The chorionic sac penetrates the uterine mucosa and loses contact with the uterinelumen (e.g., guinea pigs, human beings, rodents — late in gestation).In rodents, the uterine lining of a pregnant female assumes a characteristic topography whileawaiting the arrival of the blastocysts. The uterine mucosa appears scalloped, with evenly spacedindentations (or implantation chambers) along the long axis of each uterine horn. One blastocystwill come to occupy each implantation chamber in such a manner as to make the relationshipbetween the chorion and the uterine lining eccentric. With further development, the rodent embryowill completely embed itself into the uterine wall, making the relationship interstitial.The modifications of the chorionic surface to increase the area of contact between the chorionfrondosum of the embryo and the maternal reproductive tract also demonstrate species differences. 3,5Plicate: The surface of the chorion exhibits elevated ridges or folds (e.g., pigs).Villous: The chorionic surface exhibits fingerlike projections of embryonic tissue that project intomaternal blood. The maternal circulatory pattern is described as entering lacunae, or pools, in whichthe villi are bathed (e.g., primates).© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 151Labyrinthine: The chorionic surface exhibits anastomosing cords or trabeculae of embryonic tissuethrough which maternal blood flows. The maternal circulatory pattern is described as labyrinthine(e.g., insectivores, rodents, bats).Great species differences also exist with respect to the layers of embryonic and maternal tissuesthat are interposed between their respective circulations. The invasiveness of the trophoblast canbe gauged by the amount of maternal tissue that is eroded. 2,3,6,7 The three most common placentaltypes, as classified by extent of invasiveness, are described below (see Figure 6.3).1. Epitheliochorial: The least invasive type of placenta. No maternal tissue is destroyed. The sixlayers separating the maternal bloodstream from the embryonic bloodstream are maternal capillaryendothelium, maternal uterine connective tissue, uterine epithelium, trophoblast, embryonic connectivetissue, and embryonic capillary endothelium (e.g., pigs, horses).2. Endotheliochorial: The trophoblast invades the endometrium and connective tissue, allowing thetrophoblast to approach the maternal capillaries. The four layers interposed between maternal andembryonic circulations are maternal capillary endothelium, trophoblast, embryonic connectivetissue, and embryonic capillary endothelium (e.g., dogs, cats).3. Hemochorial: The trophoblast eliminates all maternal tissue, allowing the trophoblast to come intodirect contact with the maternal blood. The three layers that separate the maternal from theembryonic circulation are trophoblast, embryonic connective tissue, and embryonic capillary endothelium(e.g., lagomorphs, bats, rodents, primates).The gestational periods of rodents and lagomorphs (rabbits and guinea pigs) are brief (16 to68 days). Consequently, development in these species occurs rapidly. Because the definitive chorioallantoicplacenta is not established until a competent embryonic circulatory system is operative(at about the 20 somite stage), these species develop an early placenta that uses the membranes oftheir rather large yolk sacs. This early placenta is frequently termed the “inverted” yolk sac placentabecause portions of the outer yolk sac membranes attenuate and become discontinuous at the planeof apposition to the uterine wall, leaving the epithelium of the inner yolk sac membrane in virtualcontact with the uterine lumen and epithelium. The inverted yolk sac placenta ferries nutritivesubstances to the embryo by a histiotrophic process that entails the pinocytosis by yolk sac epithelialcells of maternally derived macromolecules found in uterine secretions and the subsequent breakdownof those macromolecules within lysosomal vacuoles, followed by diffusion into the embryo.In contrast to the inverted yolk sac placenta, the chorioallantoic placenta accomplishes the exchangeof nutrients, gases, and metabolic wastes between mother and embryo by means of a hemotrophicinterchange of solutes between the respective circulations. Such a direct exchange can movematerials between mother and embryo more efficiently and rapidly in the chorioallantoic placentathan can the multistep, lysosome-dependent process of the inverted yolk sac placenta.While the importance of the rodent inverted yolk sac placenta is greatly diminished after theestablishment of the chorioallantoic placenta, it does remain functional throughout gestation. Forpurposes of temporal comparison, the rat inverted yolk sac placenta develops about gestational day6.5 to 7, whereas the rat chorioallantoic placenta is established at approximately gestational day11 to 11.5. Inverted yolk sac placentas do not develop in humans or other primates. Table 6.2summarizes the placental, uterine, and other gestational characteristics of humans and six commonlyused experimental mammals.Since the placenta is the interface between the embryo and the maternal environment, it is thesite of absorption, transfer, and metabolism of nutrients and foreign compounds. In the not sodistant past, the placenta was believed to be a barrier that prevented the movement of all unwantedxenobiotic (foreign) compounds into the embryo. The thalidomide tragedy of the late 1950s andearly 1960s dispelled that idea. The placenta was reconceptualized as a sieve that retarded oreliminated the transfer of molecules that weighed greater than around 1000 Da or were highlycharged, highly polar, or strongly bound to (serum) proteins. Currently, however, it is recognized© 2006 by Taylor & Francis Group, LLC

EmbryoMother(Endometrium)Figure 6.3EmbryonicConnective TissueEmbryonic CapillaryCytotrophoblastUterineEpitheliumMaternalConnective TissueMaternal CapillaryMaternal Blood CellEmbryonicConnective TissueEmbryonic CapillaryCytotrophoblastSyncytiotrophoblastMaternal CapillaryMaternal ConnectiveTissueMaternal Blood CellEpitheliochorial Endotheliochorial Hemochorial(early)EmbryonicConnective TissueEmbryonic CapillaryCytotrophoblastSyncytiotrophoblastTrophoblastic LacunaMaternal ConnectiveTissueMaternal Blood CellTypes of placentas, classified by extent of invasiveness. The three most common placental types are depicted in a series that illustrates the progressiveloss of tissue layers between the maternal and embryonic vascular systems. Six layers are interposed between the two circulations in theepitheliochorial placenta, four layers in the endotheliochorial placenta, and three layers in the hemochorial placenta.152 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Table 6.2Comparative reproductive and placental features in selected experimental mammals and humansFeature Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey HumanEstrus cycle (days) 4–6 3–9 None 4 16 28(menstrual cycle)28(menstrual cycle)Ovulation stimulus Spontaneous Spontaneous Coitus Spontaneous Spontaneous Spontaneous SpontaneousUterus Bicornuate Bicornuate Duplex Bicornuate Bicornuate Simplex SimplexUsual no. of offspring 6–14 8–16 6–9 5–10 3–4 1 1Gestation (days) 22 19 30–32 15–16 67–68 166 266Implantation type Eccentric —earlyInterstitial — lateEccentric —earlyInterstitial — lateSuperficial Interstitial Interstitial Superficial InterstitialClassification by fetal Early Early Early Early Earlymembranes that Inverted yolk sac Inverted yolk sac Inverted yolk sac Inverted yolk sac Inverted yolk sac Chorioallantoic Chorioallantoiccontribute to placenta Definitive Definitive Definitive Definitive DefinitiveChorioallantoic Chorioallantoic Chorioallantoic Chorioallantoic ChorioallantoicPlacental shape Discoid Discoid Discoid Discoid Discoid Bidiscoid DiscoidInternal placental Labyrinthine Labyrinthine Labyrinthine Labyrinthine Labyrinthine Villous VillousstructurePlacental relation tomaternal tissuesHemotrichorial Hemotrichorial Hemodichorial Hemotrichorial Hemomonochorial Hemomonochorial HemomonochorialCOMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 153© 2006 by Taylor & Francis Group, LLC

154 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthat there exists a broad diversity of mechanisms for transporting molecules through the placenta.The transport mechanisms 8–10 include both simple diffusion for most molecules (e.g., urea, oxygen,carbon dioxide) and carrier-mediated transport. The carrier-mediated mechanisms include activetransport (e.g., for sodium/potassium, calcium, amino acids), facilitated diffusion (e.g., for D-glucose), and receptor-mediated endocytosis (e.g., for immunoglobulins, vitamin B-12). Thus, giventhe multiplicity of available transport mechanisms, when any substance is presented to the placenta,the question concerning entry into the embryo should not be whether placental transfer occurs, butrather by what mechanism and at what rate will transfer occur. The closest phenomenon to a barrierfunction is the expression in trophoblast cells of the multidrug resistance (mdr) gene family, whichencodes a p-glycoprotein on the surface of the trophoblast membrane of conceptuses 11–13 exposed tocertain xenobiotics. This phenomenon serves to limit the exposure of embryos to selected molecules.In addition to transferring nutritive molecules to the embryo, the placenta may metabolizesubstances, whether they are nutrients or xenobiotic compounds. 9,10,14 For example, in cattle andsheep, the placental trophoblast converts maternally delivered glucose to fructose, which is in turntransferred to the embryo. In those species, an intravenous dose of glucose to the pregnant femalecauses a dramatic rise in fetal plasma fructose concentrations, rather than a rise in fetal plasmaglucose. This illustrates the concept that placentas are not merely sieves but have the ability to altersome of the types of molecules that traverse them.Placentas also contain various enzymes that are capable of metabolizing xenobiotics. 15–17 Theseenzymes include reductases, epoxide hydrases, cytochrome P-450 monooxgenases, glucuronidases,and others. These enzymes are not present at all times during gestation but make their appearancesas the placenta (and embryo) mature. The presence (or absence) of these enzymes reflects thegenotype of the embryo rather than that of the mother. Placental enzymes can be induced byinducers of monooxygenases, such as phenobarbital, benzo(a)pyrene, and 3-methylcholanthrene.In addition, the formation of reactive intermediates from xenobiotic compounds by placental enzymepreparations has been demonstrated in vitro (e.g., see Chapter 12 of this volume).Placental toxicity, per se, is rarely cited as a primary mechanism for developmental toxicity.This does not mean that the importance of the placenta in development is not recognized, nor doesit mean that placental dysfunction can be discounted as playing a critical role in development.10,14,17–20 That the placenta plays a role in developmental toxicity is not in dispute; rather, ithas proved difficult to determine whether developmental toxicity arises as a result of direct placentaltoxicity or from combined effects on the materno-feto-placental unit. Examples of developmentaltoxicity that have been ascribed to some combination of mother, fetus, and placenta include reductionsin utero-placental blood flow subsequent to hydroxyurea, 21,22 altered transport of nutrients by azodyes, 23,24 immunotoxicants, 25,26 lectins, 71 and hemoglobin-based oxygen carriers, 72 as well as pathologicalchanges observed in the trophoblast after exposure to placental toxicants, such as cadmium. 27,28III. EMBRYOLOGICAL PROCESSESDevelopment from zygote to embryo to fetus to independent animal is a dynamic and carefullyorchestrated phenomenon that involves numerous simultaneous processes that occurs in specificsequences and at particular times during both gestation and the postnatal period. This is especiallytrue for rodents, wherein many of the organ systems of neonates have attained only the state ofmaturation found in late second or early third trimester human fetuses. 29 While it is imperative thatdevelopmental schedules be maintained, each embryo develops at its own rate, and there is someroom for adjustment to the schedules. That is, some developmental events may be delayed to acertain extent without adverse consequences. Thus, the gestational ages given for developmentalevents are merely averages of the observed events. Embryos within the same litter of polytocousspecies are frequently at different developmental stages, especially during early embryogenesis.This may have resulted from different times of fertilization as well as from differences in the rateat which each embryo progresses through its own developmental schedule.© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 155While many of the details concerning the development of embryos from various species (e.g.,length of gestation, size of fetuses, time at which developmental landmarks appear) differ, thesequence of developmental events and many other features and processes are remarkably consistentacross species. The following paragraphs will provide an overview of the consistencies and similaritiesof processes that take place in all embryos.The blastocyst is a small organism, the cells of which have relatively few distinguishingmorphological characteristics when observed with a microscope. There are two geographicallydistinct areas, the trophoblast and the inner cell mass. Not only does the blastocyst grow larger insize, but also the cells that comprise the blastocyst must become different in structure and in functionas development of the individual progresses. As mentioned previously, the trophoblast forms thefetal membranes, whereas the inner cell mass gives rise to the embryo proper.The cells of the inner cell mass quickly segregate into a two-layered disk that unequally transectsthe blastocyst cavity. One layer of cells (epiblast) is associated with the developing amniotic cavity;the other layer (hypoblast) is associated with the developing yolk sac cavity. The epiblast in turnrearranges by a process of cellular migration (variously called invagination, ingression, or gastrulation)30–33 into the three germ layers (ectoderm, mesoderm, and endoderm), as well as the notochord.The hypoblast gives rise to the epithelial lining of the yolk sac, but does not contribute tothe embryo proper.Specific tissues of the body are derived from each germ layer. The ectoderm will give rise tothe nervous system, skin, and adnexal dermal organs, including teeth, nails, hair, and both sweatand mammary glands. The derivatives of the mesoderm include cartilage, bone, muscle, tendons,connective tissue, kidneys, gonads, and blood. The endoderm gives rise to the linings of thealimentary, respiratory, and lower urinary tracts. The notochord serves as a primitive supportingtissue for the embryo and actively participates in the organization of the embryo. It degenerates,leaving no derivative except for the nucleus pulposus of each intervertebral disk.The primordia of the organ systems are formed from combinations of tissues derived from thegerm layers. To execute this process efficiently and accurately, many controls operate to maintainembryonic schedules and to control the fates of populations of cells, although there is some roomfor flexibility in these schedules and fates.To help understand how the development of the embryo proper unfolds in an orderly fashion,two important concepts will be explained. These relate to the potential fate of a given cell (embryoniccellular potency) and its state of differentiation. Briefly, embryonic cellular potency is the totalrange of developmental possibilities (i.e., all possible adult tissues) that an embryonic cell is capableof manifesting under any conditions. In contrast, differentiation is the process whereby an embryoniccell attains the intrinsic properties and functions that characterize a particular tissue. Differentiationis a progressive, continuous phenomenon that involves at least three steps: (1) determination, duringwhich stable biochemical changes occur within cells, but the changes are not apparent microscopically,(2) cytodifferentiation, during which those biochemical changes manifest themselves, resultingin the characteristic cytological and histological features that distinguish cell or tissue typesfrom one another, and (3) functional differentiation, during which the cell or tissue begins to actin a physiologically mature role (e.g., insulin synthesis and release by pancreatic islet cells).An embryonic cell’s potency and its state of differentiation are reciprocal characteristics. Cellsin the early stages of development, such as the blastomeres, the cells of the morula, or those ofthe inner cell mass of the blastocyst, are not differentiated; they are morphologically similar, andthey have the potential to become nearly any type of embryonic cell. The pluripotent cells of theinner cell mass constitute the population of embryonic stem cells, which hold the promise of curesfor many debilitating diseases. 73 As development proceeds, however, developmental decisions aremade concerning the fate of each cell. Thus, at later periods of gestation the cells have becomedifferent from one another. One cell may have become an endoderm cell lining the liver parenchyma,while another may be a mesoderm cell that is providing smooth muscle in the walls of a bloodvessel. The possible ultimate fates available to an endoderm cell are not the same as those of a© 2006 by Taylor & Francis Group, LLC

156 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmesoderm cell. Thus, the cells have restricted their potential. Also, they look different from oneanother because their state of differentiation has increased.Cells become increasingly differentiated with increasing gestational age, and their embryonicpotential decreases, as depicted in Figure 6.4. For both of these processes to occur in the propersequence to result in a well-formed, normal individual, mechanisms must be available to keeppopulations of cells on schedule. A primary means for accomplishing this is the process of tissueinteractions. As an example of tissue interactions, we discuss embryonic induction. 34Embryonic induction requires two populations of cells of developmentally dissimilar origin thatattain proximity to one another. Developmental information of a directive nature is released ortransferred from one population of cells (the inducer) for a finite period. The receiving populationof cells must be competent (i.e., able to react to the directive message) for a limited period. Thechange that is evoked in the competent tissue must be progressive, stable, and maturational. Oneimportant thing to note about this process is that the ability of one population of cells to send amessage and the second to receive and respond to the message is limited to a finite period (orwindow) that is intrinsic to each cell population. It is the transfer of developmental informationthrough these open “windows” that maintains the embryo on its schedule. Genetic control over thetiming and location of inducing and competent tissues is likely related to the sequence of expressionand spatial delimitation within the embryo of genes controlling the synthesis of transcription factorsand developmental control genes, such as the homeobox genes. 35–37The nature of the message substance or inducer has been investigated for many years. It is notknown what the medium of the message is in all cases. In some cases, the message appears torequire direct contact between the cells; in others it appears to be the release of a chemical substanceinto the extracellular space. In still other cases, a combination of the two appears to be required.For our purposes, however, the nature of the message is not as important as the fact that appropriatecommunication between the populations of cells has occurred in a timely fashion.It is important to recognize that, for a given cell, the information required to direct its differentiation(e.g., manufacture of cellular structural proteins, receptor molecules, and extracellularmatrix molecules) resides within the genetic material of its own nucleus, whereas the informationrequired to maintain developmental schedules usually comes from environmental stimuli (e.g.,inducer molecules as well as permissive and instructive signal molecules that are manufacturedand released by other embryonic cells). Successful development of an organism requires timelyinteractions between (normal) environmental stimuli and embryonic genes as they are expressedor repressed throughout development. 38 It should not be surprising, then, that abnormalities in eitheran embryo’s genetic material (i.e., mutations or chromosomal aberrations) or its environment canlead to developmental anomalies.The subcellular and molecular interactions that direct or contribute to the execution of thesedevelopmental processes (differentiation, induction, pattern formation), as well as to their controlby gene expression, are active areas of research that are beyond the scope of this brief overview.The reader is referred to more detailed texts and articles that capture this information. 32,33,39To respond to challenges external to the embryo or to untoward intrinsic influences on development,the embryo can possibly undergo internal rearrangements of its schedule or populationsof cells, thus maintaining normal, orderly development. This process has been termed embryonicregulation. Regulation is an important concept because it demonstrates that the process of embryonicdevelopment is a dynamic progression that is able to adjust to changing conditions.When an embryo is challenged by an environmental agent, many components contribute to theeventual outcome. Some of these are extraembryonic in nature, whereas others are embryologicalcomponents. Although the extraembryonic components are not the main thrust of this discussion,they will be addressed briefly because they can affect the rate and quality of embryonic development.First, the nature of the environmental agent itself must be considered. For instance, is it a physicalagent or a chemical agent? If it is a chemical agent, then its structure, polarity, and lipid solubilityare all important properties to be considered as they affect the amount of uptake of the chemical© 2006 by Taylor & Francis Group, LLC

Maturation andRelease ofGametesSpermSyngamy CleavageZygoteMorulaOvumFigure 6.4FertilizationGestational AgePre-ImplantationBlastocystA B C D EImplantationTrophoblastInner CellMassEmbryonic CellularPotencyEpiblastHypoblastGastrulationPrimitive StreakSyncytiotrophoblastCyttrophoblastChordamesodermGiant CellsGermLayersAmnioblastsExtra-EmbryonicMesenchymeNotochordMesodermEctodermEndodermDifferentiationIntermed.Lat. PlateParaxialNeuralPlatePeridermLining of YolkSac (extra-embroyonic)Diagrammatic representation of embryonic development. Fertilization is depicted on the left, and developmental maturation proceeds tothe right. The dashed arrow represents possible differentiative pathways; the series of arrowheads denotes the rapidly occurring celldivisions during cleavage. Diverging arrows represent developmental decisions made by tissues as they differentiate. With each succeedingdevelopmental decision, a cell’s developmental potential decreases, while its state of differentiation increases. The circled letters denotethe gestational milestones for the species listed in Table 6.1.Linigs ofMajorOrganogenesisGonadKidneySomaticSplachnicSomiteNeuralCrestFetalPeriodCentral NervousSystemSkinNailsAdnexa of SkinTeeth - EnamelDigestive TractLower GenitourinaryRespiratory TreeCellularDifferentiationBirthPigmentCellsPeripheralNervousSystenmOtherCOMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 157© 2006 by Taylor & Francis Group, LLC

158 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONinto the mother and the amount that will ultimately reach the embryo. In addition, each environmentalagent may be thought to act in a specific way on some aspect of embryonic metabolism,and this specificity may help to determine how the agent interferes with embryonic development.A second extraembryonic component is the dosage of the particular agent. The dosage is notas simple a component as one might assume; not all doses of proven teratogens cause birth defects.Typically, there is a lower dose range that allows most, or all, embryos to proceed through normaldevelopment, whereas higher doses may kill the embryo (and perhaps the mother as well). Inbetween those two doses, there is usually a rather narrow teratogenic range in which sufficientdamage is elicited in the embryo to disrupt developmental events without destroying it entirely. Inaddition, the dosage may be administered either acutely or chronically, and this also will affect thenature of any interference with embryonic development that may occur.A third extraembryonic component is the physiological state of the mother because she providesthe physical environment of the embryo. The state of the mother’s nutrition and her general stateof health are important, as is her ability to metabolize chemical agents and thereby change thenature of the compound to which the embryo may be exposed. 40 A fourth extraembryonic componentis the previously discussed efficiency of the maternal-fetal exchange through the placenta. Insummary then, the major nonembryonic considerations are the nature of the teratogenic agent, thedosage of the agent and timing of exposure, the maternal organism, and the effectiveness ofmaternal-embryonic exchange.There are also important embryological components that affect embryonic outcomes. 41–43 Thefirst of these components is the embryonic genotype and its expression, the theoretical basis ofwhich has been discussed elsewhere 38 and which is the topic of ongoing, in-depth research. 31,33,39In simplest terms, the embryonic genotype is an important embryonic consideration because itdetermines the inherent susceptibility of the embryo to exogenous agents at any given time duringdevelopment. Alterations in a cell’s DNA are the cause of mutations. Throughout most of the lifespan of mammals, DNA is replicated with great fidelity, and alterations to nonreplicating DNA,caused by environmental agents such as irradiation or chemicals, are rapidly repaired. There aretwo periods, however, when mammals are rather vulnerable to permanent changes in their DNA.One of these periods occurs during cleavage, when cell cycle times are shortest and extremelyrapid synthesis of DNA is required. The fidelity of DNA replication diminishes with the continuedrapidity of its synthesis. The other period is during the postmeiotic stage of gamete development.The greatest sensitivity occurs in males during spermiogenesis, when spermatozoa are maturing.The maturation of spermatozoa involves a process that drastically decreases the cytoplasm of thecells. In concert with the reduction of nonessential cytoplasm, the enzymes required for DNA repairare lost, leaving the maturing gametes unable to repair DNA damage. These topics have beendiscussed at length by others. 44,45A second important embryonic component is the stage of development of the embryo. In general,the time at which an agent acts on an embryo determines which tissues will be susceptible to theeffects of the agent. This means that susceptibility to a particular agent will vary greatly duringthe course of gestation. Agents that are applied, even at high doses, during the predifferentiationperiod (from the time of fertilization through formation of the blastocyst) typically produce noteratogenic response, although exposure of females to mutagens within a few hours of mating hasbeen reported to induce malformed offspring in some instances. 46 The reason why young embryosappear to be resistant to the effects of teratogens is not well understood; however, that resistanceis probably a result of either the lack of specialization by the cells of the zygote to form specificparts of the organism, thereby retaining their embryonic potency, or a large number of stem cells.As long as all or many cells of the zygote retain a high degree of potency, the destruction or damageof some of those cells can be tolerated because the embryo can still undergo sufficient regulationto allow normal development to proceed. Although it appears that the destruction of a small numberof undifferentiated cells in the embryo does not necessarily result in a structural malformation,there does appear to be a critical limit beyond which damaging even nonspecialized cells cannot© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 159be tolerated if the embryo is to live; if that critical limit is exceeded, the zygote will die. Further,nonlethal damage to the genome of pluripotent cells can result in a mosaic of tissues that exhibitincreased likelihood of disease or other pathology. 44During the period of early organogenesis (when the embryo begins to undergo differentiationand the establishment of the germ layers), the onset of greatest susceptibility to teratogenesis occurs.This is coincident with the processes of gastrulation or invagination. For mammals, this occursapproximately five days postconception in small rodents (e.g., hamsters and mice), and up to 10to 12 days postconception in primates. Not only is the onset of susceptibility to teratogenesissudden, but also the majority of teratogenic agents produce their highest incidences of malformationsat about this time. 41 Although there are no indications of the definitive organs in the embryoat that time, the cells of the germ layers have become determined (i.e., the morphologicallyundetectable aspect of differentiation has occurred) and have, therefore, lost some of their embryonicpotency. Thus, cells that have become determined are susceptible to teratogenic agents even thoughtheir ultimate morphology is not yet evident. For example, rat embryos that have been exposed tox-rays on gestational day 10 exhibit malformations of the kidney at term. 47 This is of interestbecause the definitive kidney of the rat develops from the metanephros, which does not appearuntil day 12 of gestation.This illustrates the concept that it is the stage of development at which an agent is effective,rather than the time at which it is administered, that determines the embryo’s susceptibility. 48 Thisconcept is important for those agents that might be stored in adipose tissues of the body. By wayof example, this has been used as a basis to allege that the vitamin A derivative, etretinate, mayhave caused malformations in the offspring of a woman who had terminated its use several monthsprior to conception. 49Not only do embryos themselves have a sudden onset of susceptibility to teratogenesis, but alsoeach organ of an embryo has a sensitive period for teratogenesis. 41,45 This sensitive (or critical)period is the time during which a small dose of a teratogen produces a great percentage of fetusesthat will exhibit malformations of the organ in question. The critical period coincides with the earlydevelopmental events and tissue interactions that occur within the organ. In general, the susceptibilityto teratogenesis decreases as differentiation and organogenesis proceed. This is because theproliferative and morphogenetic activities that characterize the early stages of the formation oftissues and organs become less prominent as the organ develops.As an embryo progresses through the period of organogenesis, and as differentiation continues,production of a given teratogenic effect requires increasingly higher doses of the teratogen. Thismeans that as organ systems and the embryo itself become progressively more differentiated, theybecome increasingly resistant to teratogenesis. Most of the organ systems have been laid down bythe period of late organogenesis and the early fetal period, and the critical events involved in theirformation have been completed. What remains to be accomplished during the remainder of prenataland postnatal development is the progressive growth and functional maturation of each organsystem. Strictly speaking, the majority of gross malformations become increasingly less problematic,although malformations of late-developing organs (e.g., kidneys, genitalia, and brain), alteredhistodifferentiation, growth retardation, and postnatal functional deficits (including neurobehavioralproblems) may still be caused.IV. COMPARATIVE EMBRYOLOGICAL MILESTONESBecause the primordia of the organ systems of an embryo are laid down in sequence, and notconcomitantly at any given time in gestation, each organ system is likely to be at a unique stageof differentiation. For this reason, agents given acutely during a particular period of gestation maycause malformations of one organ system but not of another, or they may cause different malformationsof the same organ system. Thus, the pattern of defects caused by any particular teratogen© 2006 by Taylor & Francis Group, LLC

160 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmay change if the time at which the agent is applied or if the time at which the agent is effectiveoccurs successively later in gestation. This has led to the construction of developmental schedulesfor embryos and, subsequently, both the use and misuse of those embryonic timetables. 50 It isimportant to realize that embryonic timetables can be used to determine at what developmentaltime a given organ is formed. From such a table, it is possible to ascertain the earliest and latestgestational times at which a particular organ system is likely to be grossly malformed by a noxiousagent. Such embryonic schedules are useful for determining whether an embryo was exposed duringor prior to the time of development for a given organ; they cannot identify the exact date at whichan embryo was exposed to a particular agent because, as mentioned previously, some agents havedelayed effects.The differences among species, especially with respect to the timing of prenatal developmentalevents, are the subjects of Tables 6.3 to 6.12. The tables present the times of appearance for eventsin the embryology of various organ systems for selected laboratory animal species. The timing ofsuch events is important if one wishes to investigate the normal development of a particular organsystem. Timing is also crucial to studies of the genesis of malformations of an organ, using ananimal model, or if one wishes to determine whether treatment with a specific agent is capable ofeliciting the malformation of a given organ system. In cases such as the latter, the investigator mustknow at what gestational time the organ system in question is undergoing organogenesis in theappropriate animal model. Thus, the reader is referred to Tables 6.3 to 6.12 for interspeciescomparisons of embryonic events related to development in general (Table 6.3); the circulatorysystem (Table 6.4); the digestive system (Table 6.5); selected endocrine glands (Table 6.6); therespiratory system (Table 6.7); the nervous system (Table 6.8); selected organs of special sense,i.e., eye, ear, and olfactory region (Table 6.9); the muscular, skeletal, and integumentary systems(Table 6.10); the excretory system (Table 6.11); and the reproductive system (Table 6.12). It isimportant to reemphasize that development does not end at birth and that it may be important tostudy events in postnatal animals. When the developmental phases of an organism’s life are scaledaccording to the appearance of developmental landmarks (rather strict chronology) so that thedevelopmental schedules of different species are congruent, they are being compared according tophysiologic time. This concept is explained in greater detail elsewhere. 29Although avian models are not considered to be relevant for the assessment of human developmentaltoxicity, data for the chicken are included because of its long-standing use in embryologicalstudies and because of its possible usefulness in assessment of the developmental toxicityof environmental pollutants toward wildlife. More complete texts and monographs should beconsulted for the detailed embryology of particular species (e.g., human, 32,33,51,52 rat, 53,54 mouse, 55–57hamster, 58 rabbit, 53,59 guinea pig, 55,60 rhesus monkey, 61,62 and chicken 63–65 ).Determination of the precise times in gestation for each species at which developmental eventstake place is difficult for a number of reasons. Even though the process of prenatal developmentproceeds sequentially, the rate at which it proceeds is neither standardized nor constant, even amongoffspring within the same litter. Development is based upon the expression of information containedwithin the genome of embryonic cells, and the timing of that expression is both triggered by andpermitted by signals in the environment of those cells. Thus, there can be substantial variation inthe time of appearance of rudimentary embryonic structures. This is especially true for those specieswith longer gestational periods.The timing of embryonic events is further complicated by the fact that the starting point fortiming (the instant of fertilization) is not known precisely. In most cases, the time of copulation isused as a surrogate for the time of fertilization. By convention, gestational age is measured fromthe time that mating is either observed (rabbits) or deduced from evidence of mating (such asobservation of a copulatory plug in mice or rats or finding sperm in a vaginal smear in rats, mice,or hamsters). When mating is deduced, the time of fertilization is usually considered to haveoccurred at 9:00 a.m. of the day that the observations were made. Thus, by embryological convention,9:00 a.m. of the day that the observations are made is set as day 0, hour 0 of gestation.© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 161In humans and other primates, gestational age is estimated by ovulation age (or, at times,menstrual age — which is about 14 days longer than ovulation age). This means that the actualtime of fertilization may be miscalculated by as much as 12 h in rodents and much longer inprimates. For avian embryos, including chickens, development is initiated approximately 24 h priorto laying.A final impediment to establishing the timing for embryonic events is caused by the samplesize (or number of examined specimens) from which the times have been derived. In some species,particularly humans and other primates, the number of available specimens is quite small, leadingto variations in the timing of events reported by the source documents. Thus, to group or classifyembryos by their stage of development rather than by the time after fertilization, some investigatorsreport other embryonic characteristics, such as crown-rump length, number of somites, or externalfeatures, as surrogates for gestational age. 52,53,63,66–70The aforementioned challenges to determining timing have led to the inclusion of several entriesfor many developmental events in Tables 6.3 to 6.12. These tables present the timing of developmentalevents of seven mammalian species and the chicken, organized by organ system. The entriesinclude the estimated time during gestation and (where appropriate and available) the surrogatedescriptors somite number or crown-rump length. Where source data have diverged, the entries aregiven as ranges. It should be emphasized, however, that even though the timing for the developmentalevents may be somewhat imprecise for certain events, the order of developmental eventswithin a given organ system rarely changes.ACKNOWLEDGMENTSThe author is grateful to Ms. Sue Walter for her diligence and patience in extracting data andpreparing multiple revisions of the tables for this manuscript. The author also wishes to thank Ms.Judith Pals, Mrs. Pauline Kapoor, and Mr. Michael Yang for their assistance with the figures.© 2006 by Taylor & Francis Group, LLC

Table 6.3DescriptionOne cell(in oviduct)Two cells(in oviduct)Four cells(in oviduct)Eight–twelvecellsMorula(in uterus)Free blastocyst(in uterus)Comparative early developmental milestonesAge a(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenSize Somites(mm) Ref.1 0.07 1–4,6,9,122 0.08 1–4,6,9,123 0.08 1–4,6,9,123.25 0.08 1–4,6,9,123.5 0.08 1–4,6,9,125 0.12 1–4,6,9,12Age SomitesAge Som-Age Som-Age Som-Age Som-Age Size Som-(d)Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) (mm) b ites Ref.1 5,7–9 1 42 1 33 1 0.13 61,621 5,7–9,120.33 12,282.25 12 0.46 12,28Age(d/h)1–1.5 42 0.96 49, 1 12 2 0.12 62,63 3h c 12561.67 12,42 1.25 50 1.5 12 1.5–2 0.12 62,63 3.25h c 122 5,7–9 3.5 56 3 0.1 633 5,7–9 3 42 4 177 4 414 5,7–9,12Implantation 5.5–6 0.28 10–12 4.5–5 10,34 7–7.5 10,12Shellmembraneformed inoviduct (chick)Shell of eggformed inuterine portionof oviduct(chick)Hypoblastformed6 1–4,6,9Primitive streak 8.5–9 1 1–4,6,9,10,12Neural folds 9 1 1–4,6,9,11FirstmyocardialcontractionsYolk sac;exocolemHead process/notochord9.5 1.5 1–4 1,2,9,13,14,18,24,549.5 1.5 1–4 1–4,6,9,124.5 5,7–9,125,7–9,10,357.5 5,7–9 7.75–8.257.25 10,578 24 8.5 9 24,573.5–4 42 4.75 56 5–8 33 5 0.1 634.5–6 10,12,42,436–6.5 10,51–53,569 10,12,46,606–7.5 10,12,64~5 42 6–6.5 51 9 46 7–8 0.5 656.5–7 10,42 12–131–4 27 7.75 5 44 14–14.58–8.25 12–1342,43,4416–16.57 36 7.5 44 11–1810,15 15–17 10,33 13.5–170.3–1.29,10,35,6615,55 20–21 3 33 18–21 1.5–2 1–4 36,37,41,67,68,70NASomitesRef.3.5– 864.5h c4.5– 8624h c7–19h 10,8622–26h 1 86,8723 56 21–24 2–3.5 4–12 59,71 1.5d 30,867 42 12 46 11–13 0.15 9,72 2d 8755 16–18 33 18 36 19–22h 86162 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Amniotic cavity 9.5 1.5 1–4 1–4,6,9,12Start of somitephase9.5–101.5 1–4 11,12 7.75–8Allantois arises 10 2 1–4 1–4, 6,9, 12Oral membraneperforates10 2 5–12 1,2,9,12,18,22–267.75 5,7–9 6.5–7.5 12,42,447.25–7.759–9d2h5,7–9,34,37 7.75–8.255,7–9,3519± 5,7,26,389 51,58 10 46,178 8 41 2d 871–4 27 7.5 42 14.5 59 20–21 46,60 19–21 1.5–2 1–4 37,64,67,70,738–10 99 7.75 5 42,44 11.75–1315,55,5610 10 8.5 44 27–28 10,33 26–30 3.3–4 17 12,26,39,74,75,76,77,7816.5–192d 1–4 8735,41 2–3d 30–362.2–3d 29–32Ten somites 10.5 10 10 8.5 10 10 8.5 10 10 8 10 10,42 15 10 10 23 10 10 25 10 10 1.5d 10 10,12Fusion ofneural folds(early)AnteriorneuroporeclosedBothneuroporesclosedDorsal flexuredisappears;embryo curvesventrallyAnterior limbbud appears10.5–10.7510.5–10.7510.5–11.510.5–11.5Tail bud 11–11.5Hind limb bud 11.5–appears d 12Hand (forepaw)raysabcd2.4 17 1,2,4,9,10,11,18,23,272.4 13–2011 3.3 21–2513.5–143.3 21–253.8 26–2811,27 8.5–9 10–149d 1h 18± 5,7,8,3910,11 9–9.5 10 9.5–10.51–4,6,9,12,271–4,6,9,12 9.5–9.751–4,6,9,121–4,6,9,10,123,4,10–12,27,30–328.5–9 5,7–9,3923± 5,7–9,3927,994 9–9.5 10,2710.5–1110,9927,57,998.25 12–138.5 17–2042,44 14–158.5–9 10 15.25–15.58.5–8.7517–208.75 17–2015,55 21–23 33 22–24 2–3.5 4–12 66–68,71,73,74,79–8210,44 15.25 10 25 10 24–26 2.5–4.913–2010 28–31 10,33 25–28 1042,44 25–27 33 26± 3942,44,46,4716.5 23 56 25–28 33,46 26 3–5 21–2910,39,71,73,8386,8726,8826–29h 3–4 282.3d 1039,73 51–56h 26–289.5 5,7–9 9.5 12 8–8.5 7–9 12,44 26 12 29 3.8 12 ca. 50–52h10–10.312.3 7,10,29,405,7–10 11–1210,27,57,9914.5 10,27,299 44 17.5–18.510.25–1110,42,44,4822–23.7529 10,56 28–30 10,33,6028–32 4–6 30–329,10,60,71,73,84,8520–212.2–3d 29–3210,56 34–35 10,46 35–37 8–11 10 4.75d 30,86Age is measured in hours and days from the time of evidence of intromission. For rats, mice, hamsters, and guinea pigs age is counted from 9:00 am on the morning of discovery of either sperm inthe vaginal lavage or a copulatory plug. In rabbits, it is measured from time of observation of mating. For primates, age is measured from the midpoint of the cycle (14 d after onset of last menses). Inchickens age is generally given as “incubation age” or time after laying. The actual age of the chicken embryo is approximately 24 to 25 h older than the incubation age.Crown-rump length.Preincubation age.Hindlimb bud forms earlier in rodents than in primates.30,868610,86COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 163© 2006 by Taylor & Francis Group, LLC

Table 6.4DescriptionVascularizationof yolk sacBilateral heartprimordia inventral wall ofcoelom; twodorsal aortaeFusing hearttubesFirstmyocardialcontractionsDorsalmesocardiumdisappearsAortic archarteriesformingComparative gestational milestones in circulatory system developmentAge a(d)7.25–8.5Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.0.6 1,2,9,13–229–9.5 1 6 1,2,7–10,12,13–22,249.5 1.5 1–4 1,2,8,9,13,14,24,5410–12Age(d)ca 8d8hRef.4 5,7,8,907 5,7,8,108 5,7,8,248d21h10,24 8.5–11Aortic arch I 10 2 5–12 2,8,9,13,14,18,24Sinus venosus;umbilicalvessels;cardinal veins;endocardiumS–shapedheartAnteriorcardinalsAortic archesI & IIDorsal aortaefuse10.5 2.4 16–2010 2 10–1210.5 2.5 16–2010.5–112–2.4 11–202,8,9,13,14,182,8,9,13,14,18,242,8,9,13,14,182,8,9,13,14,18,898d13hAge(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)7.5 44 13.5 56 16–18 33 19–20 1.5 74,84,96–1038.25 10 7.75 5 42,44 14.5 15,56 2–6 908.5–9 12 10,248–8.25 12–138.5 9 24 8–8.25 12–1315 90 8.75–9.2510,24 9–11 10,249 7,81,248.5 10,24 9.5 21 10,248d14h10,42–448–9.5 10 15.5–21.5Ref.Age(d/h)Size Somites(mm) aSomitesRef.20–29h 4 12,8615 10,15 22 10,33 21 2 10,24 1.2d 1042– 44 16.5 23 15,56 21–24 2–3.5 4–12 59,71 1.5d 30,8644 16–1710 22–30 10 22–32 2–4 10,12,249.25 24 8 44 21–23 33 22 2 7 7,12,81,248.5 44 17.5 29 56 24–26 13–208.5 10,44 16 10 25 10,33,4433 24–27 2.5–4.513–209012,7125–27 3.3 10,12,241.5–4.5d10,2433–38h 9–10 8848–54h 24–279–10 91 8 44 15.5 13 56 14 91 40h 13 129d 4h 20± 92,24 9.5 24 8.5 44 15.5 13 56 21–23 33 30 4 28 12,92,2450–55h 20–269d 4h 20± 92,24 9 44 16.5 23 56 27 3.3 20 92,24 56h 24–2710,2488,2424164 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Posteriorcardinal veinsestablishedTen-somitestageFirst circulation 10–10.510.5 2.5 16–202,9,13,14,1810.5 10 10 8.5–8.752.4 16–20Duct of Cuvier(commoncardinal vein)Aortic arch III 11 3.3 21–25Posteriorcardinal;branches ofanteriorcardinal onmesencephalonSeptationoccurringFour aorticarches;I regressing,IV still smallAtrioventricularcanalEndocardialcushionsappearAortic archesIII, IV, & VIBeginninginterventricularseptumVitelline veinsanastomosewith liver plexus2,9,13,14,18,1769d 4h 20± 7,39 8 44 16.5 23 56 26± 7,397,10 10 8.5 10 10 8 10 10 15 10 10 23 10 10 23–25 7,10 10 1.5d 10 10,128.5 5,7,24 26 3 71 2d 18 2412.5 6.2 12 8.5 44 16.5 23 56 24 14 83 ca. 45h 15 882,9,13,14,1810 27± 7,93 9.75 24,2713.5 8.5 12 9 44 88,2417.511.5–12.511.75–134.2 29–3112 5.1 34–3512– 5.1–6 35–12.375 4010,12,242,9,13,14,18,242,9,13,14,182,9,13,14,18,2412.125 5.2 36 2,9,13,14,1812.125–135.2 36 2,9,13,14,18,2410.5–11.510,40,2413 34 10,249 44 16.5 23 56 27–28 33 28–31 4 7,24,93,50–55h 26 88,2429 56 24 14 83 ca 45h 15 889.25 10 19.5 10 28 10 28–37 3.5–6 10,24,74,90,99,104–11110 24 11.75 24 9.25 44 17.5 29 56 28–29 33 31–34 4–4.6 12,24,839d18h2.3d 26 10,243d 36 8825 16 83 3–4d ca.30–4426± 39,24 11 24 26–28 39,71 3–4d ca30–3410.5 7,24,939–10.511 24 9.5 44 19 35 56 29–30 33 32–35 6 24,93 4d 88,2476,24 12 24 8.5 29 44 29–35 6 24,76 4d 88,2412.5 6.2 12 10.5 39 9.25 44 26 398888,24COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 165© 2006 by Taylor & Francis Group, LLC

Table 6.4DescriptionIntersegmentalarterysupplyinganterior limbbudPrimordium ofatrioventricularvalveSeptumprimumPulmonary veinenters leftatriumEndocardialcushions fusedSubcardinalveins formedInitiation ofaortic –pulmonaryseptumOstiumsecundumSeptumsecundumForamen ovalepresentSubcardinalanastomosisInferior venacava entersheartComparative gestational milestones in circulatory system development (continued)Age a(d)12–12.5Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken5.1 34–3512.375 6 39–4012.5 6.2 41–42Ref.2,9,13,14,18,242,9,13,14,182,9,13,14,18Age(d)11–11.5Ref.7,24,90Age(d)Ref.Size Somites(mm)SomitesSomitesAge SomitesAge Som-Age Som-Age Size Som-Age Som-(d)Ref. (d) ites Ref. (d) ites Ref. (d) (mm) a ites Ref. (d/h) ites Ref.9.25 44 19 35 56 5 83 4d 8823.75 56 31–35 4.3–5.474,84,95,104–11134 24 9.25 44 18.5 31 56 28–37 3.5–6 7,24,71,74,84,90,95,104–11150–55h 26 88,2411.5 7,90 20.75 39 56 35–40 8–10 7,74, 84,90,95,104–111ca 50h 20 8811 77,24 14 24 35–40 6–8 24,74, 5.5–6d 2477,84,96–98,100–10311.5 94 9.5 44 28 4 94 70h 30– 883611.5 77,24 10 44 32–35 7–8 24,71, 4d 247713.25 24 11 24 12.75 24 40 8–10 24 5d 2414 24 13 24 13 44 40 8–10 2413–14 8 46–482,9,13,14,18,2411–1210,24 14 10,2410–13 10,44 19.25–2138 10,56 34 10 41–44 8–10 10,24 5d 10,2412.5 7,94 11.5 44 11 7,9412.5 7,24,949.75 44 45–51 15–20 7,24,946–6.5d 24166 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

PosteriorcardinaldegenerationRight dorsalaorta betweenarches III andIV disappearsInterventricularseptumcompleteBursa ofFabricius;posterodorsalwall of cloaca(chick only)TruncalseptationcompleteInitiation ofaortic andpulmonarysemilunarvalues15 24 13 7,24,9315 c 12 2,9,13,14,18,2415.5 10 13–14Fetalcirculatorysystem isestablished b15.5 14.2 64 2,9,13,14,18ab12.5 94 11 44 15 9413 7,24,9514 24 11 44 41–42 12–14 7,24,9316.5 24 11 44 43–46 13–17 7,24,71,9510,77 16.5 10 15 10 22 10 36 10 35–46 11–14 10,71,777–7.5d 248d 245–6d 885–7d 10,8813.5 77,24 35–38 77,24 5.5–6d 24Crown-rump length.Differs from that in humans mainly by persistence of a capacious vitelline circuit in addition to allantoic (umbilical) circuit.cCompletion of rat membranous interventricular septum may occur as late as postnatal day 7 (179, 180).COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 167© 2006 by Taylor & Francis Group, LLC

Table 6.5DescriptionForegut andoral platePharyngealpouchesappear;stomodeumOral membraneperforatesLiverprimordiumComparative gestational milestones in digestive system developmentAge a(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.9.5 1.5 1–4 1,2,9,10,18,22–2610 2 5–12 2,9,18,25,2610 2 5–12 1,2,9,18,22–2610.5–112.4 13–20Hindgut 11 3.3 21–25Secondpharyngealpouch11 3.3 21–251,2,9,10,18,22–261,2,9,18,22–262,9,18,25Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)7.8 0 10,26 8.5 10 7.75–8 5 10,44 14–14.58.3–8.89–9d2h8.8 14–15Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)10,56 20.5 10 20.5–22Ref.2.1 2 10,26,71,1154 26 8 44 24–26 33 24–27 3.3 7 26,39,76,77,81,96,11619± 26 10 10 8.5 44 27–28 21–2910,26 9.5 10 8.5 10,44 16 10 24–26 13–2010,33 26–30 3.3–4 17 12,26,39,76–78,96,11610,33 21–27 2–3.3 10,26,39,78,116–1188–8.5 6–7 10,26 9 10 8 10,44 15.5 10,56 21 33 21.5 7 10,26,818d19hGallbladder b NA 9.625–9.7Pancreas,ventralPancreas,dorsalVitelline ductcloses11 26 9.7–1114 26 9.5 57 8.5 44 15.5 56 25–26 33 14 9125± 10,26 11.5 10 8.7 10 19 10 28–29 10,33 26–30 3.3–4 10,26,39,71,787,26 29–30 33 31–35 4.3–7.526,39,71,96,117,119,120Age(d/h)23h–1.1dSize Somites(mm) aSomitesRef.10,26,8836–39h 12 26,882.2–3d 29–3226,8850–56h 22 10,12,26,8850–53h 21 10,12,26,882.8–3.5d3d 2611 26 9.7 26 28–29 33 26 3.5 26,71 4 35 26,8811 3.3 21–25Cloaca 11 3.3 21–25Liver epithelialcords1–4,6,9,18,22–251,2,9,18,22–2511.5 10,26 9.5–9.6259.5 5,7–927 3.3 39,76,77,96,11625± 10,26 10.5 10 8.75–9 10,44 16.5 10 26 25 10,26,393 1010,26,88168 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

StomachappearsThirdpharyngealpouch;laryngotrachealgrooveFirst and thirdpharyngealpouches touchectoderm,secondruptured intovisceral grooveUmbilicalhernia beginsUrorectalseptumappearsTracheaseparates fromesophagusPrimordium ofbile ductAnal plateposterior togenital tubercleTongueprimordium;tuberculumimparFusion ofdorsal andventralpancreasTip of tonguefreeDental lamina;upper andlower incisorbuds forming11.5 10,26 11.5 10,26 10.5 10 8.5 10 16.5 10 28–29 10,33 28–32 3.5–4 10,26,39,76,77,84,96,11611.5 3.8 26–282,9,18,2512.125 5.2 36 1,2,9,18,22–2512.25–1412.375–176 39–409 11–12.32,9,10,26,112,11312.5 1,2,9,18,22–2610,26 12.5–14.58.75 44 19 35 57 27–28 21–243d 10,2633 28 4.5 71 50–55h 23–248.75 44 25 14 8310 9.3–11 10 22–2310 33–35 10 36–45 8 10,26,84,12115 10 9.5 10,44 28–48 4.3–6 10,39,76,77,84,96,114,11611 26 9.75 44 29–31 26,71 4–4.5d 1212.5 6.2 12 10 44 20.75 40 5712.5 6.2 41–4213 8 46–4814.5 10.5 56–6014.5 10.5 26–601,2,9,18,22–251,2,9,18,22–251,2,9,18,22–251,2,9,18,22–254.75–6d9.5 44 31–37 4.3–6 96,122 68h 35 889.75 44 20.75 40 57 36–40 6–10 39,76,77,84,12311.5 7,26 11.5 44 35–36 33 35–44 8–14 7,26,71,77,96,12410.5 44 23.75 5711.5 44 20.75 40 57 40–48 8–15.6125–12812104 886 d 88COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 169© 2006 by Taylor & Francis Group, LLC

Table 6.5DescriptionFusion ofmandibular andlower jawelementscompletedMandibularglands;mucosa nearmandibularsymphysis tobase of tongueAnalmembraneperforatesComparative gestational milestones in digestive system development (continued)Age a(d)14.5–15Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken10.5–1256–63Ref.1,2,9,18,22–2515 12 10,12,26Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)Ref.19.75 38 57 37–38 33 38 8 39,76,77,84,12340–50 8–17 96,116,12414 10 10 10 13 10 39–40 33 45–50 16.5 10,26,39,76,77,84,96,116Maximal size of 15.5 14.2 64 9 14.5 5,7–9 12.5 44 9wk 167umbilicalherniaPalatal folds 17 10,11 15 10 12 10 56–63 stage of union) c 167uniting (not allat the same19.5 10 26 10 45–46 10,33 10,12,83,116,Umbilicalhernia reducedabcd17–18.516–20Crown-rump lengthRats do not have gallbladders.In the chick palatal folds do not fuse.Transient structure; teeth do not form.1–4,6,9,10–12,2716–16.55,7–10 20 10 13 10 45–48 10,33 8.5–10wk26–45 9,10,39,66,76,77,84,96,116,121Age(d/h)Size Somites(mm) aSomites8 88Ref.6 10,26N/A18 10170 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Table 6.6DescriptionComparative gestational milestones in endocrine system developmentAge(d)Pharyngeal PouchesTwo pharyngealpouchesThreepharyngealpouchesRat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken11 3.3 21–2511.5 3.8 26–28Ref.2,9,18,252,9,18,25Four to six pharyngealpouches(ultimobranchialcomplex)12.125 5.2 36 2,9,18,25Epiphysis — Pineal GlandPineal;epiphysealevaginationAdrenal GlandAdrenal gland,corticalcomponent;coelomicepitheliumAdrenal gland,medullarycomponent;migratory cellsof neural crestandsympatheticgangliaPancreasPancreas —dorsal14–14.59.5–10.552–601,2,4,9,18,23,25,129–134Age(d)8.5–8.89d15hRef.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)6–14 91,115 24–26 33 27 3.3 12– 1491,11525± 135 27–28 21– 33 28 3.5 22 13529Ref.Age(d/h)11.5 44 28–29 33 13.5 83 68h 12,8811.5 77 11 44 32–38 60 33–48 13–17 60,77 52–64h 30–3512.5 10 11 10 18 10 10 10 23 10 34 10 3.25d 37–4013.5 8.5 49–5111–12.1252,9,18,25,129–1345.2 36 2,9,10,18,25,129–13411–12.5Size Somites(mm) aSomites44 44 16 83 4–7d 889.5 10 10 10 9.5 10 17.5 10 28–29 10,33 28 10 3d 35 10,88Ref.8810,88COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 171© 2006 by Taylor & Francis Group, LLC

Table 6.6DescriptionPancreas —ventralIslets ofLangerhanswithinpancreaticdiverticulaComparative gestational milestones in endocrine system development (continued)Age(d)11.5–12.125Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.5.2 36 2,9,10,18,25,129–13413 8 46–48Pancreas fused 13 8 46–48ThymusThymus 12.375–12.5Thymus andparathyroiddetaching from3rd pouchThyroidThyroid 10.5–12Thyroid showsopendiverticulumfrom floor ofmouthVesicularultimobranchialbody (lateralthyroid)detachingUltimobranchialvesiclesdetached frompharynx6 39–4013 8 46–482.4 13–2011.75 4.2 29–3113 8 46–4813.5 8.5 49–512,9,18,25,129–1342,9,10,18,25,129–1342,9,10,18,25,129–1342,9,18,25,129–1342,9,10,18,25,129–1342,9,18,25,129–1342,9,18,25,129–1342,9,18,25,129–134Size Somites(mm)Age SomitesAge Som-Age Som-Age Som-Age Som-Age Size Som-Age Som-(d)Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) (mm) a ites Ref. (d/h) ites Ref.9.7 10 11.5 10 9.5 10 29–30 10,33 31–32 10 4d 10,888–9wk40–50 96,117,1378–9d 8811.5 10 14 10 11.5 10 35–36 10 40–44 10 6d 1012 10 12.5 10 8.75–9 10,44 * 30–40 10 6–8d 10,8811.5 44 23 838.5 10 9.5 10 8.5 10 16.5 23 10,57 28–29 10,33 24–27 10 40h 12 10,8817.5 29 579.5 24± 136 12 44 20 136172 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

ParathyroidParathyroids 12.5 6.2 41–42Parathyroidsattached to leftand right wingsof thyroid17–18 16–20Hypophysis — Pituitary GlandRathke’s pouchappearsNeuralhypophysealevaginationRathke’s pouchclosed off,connected tooral ectodermStalk ofRathke’s pouchdetached fromstomodealepitheliumPars intermediathin-walled;pars distalis —trabecular andsecondaryvesiclesa2,9,10,18,25,1292,9,18,25,12911 5,7,10 8.75–9 10,44 35–38 10 6–8 10,885,7,13210.5 10 8.5–9 23+ 10,39 9.5 10 8.5 10 15.5 13 10,57 28–32 10,60 28–34 10,39,6011.5–11.7513.5–14Crown-rump length.4.2 10,12,608.5 49–512,9,11,18,25,12915.5 14.3 64 2,9,11,18,25,11.5 10,77 12 10 10 10 18.5 10 30–34 10,60 30–42 8–11 10,60,7712 60 9 44 19.75 38 57 14 8312.5 138 12 44 36–42 60 19 13814 60 40–44 60 53–54 22–24 6050–52h 20 10,883 10COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 173© 2006 by Taylor & Francis Group, LLC

Table 6.7DescriptionLaryngotracheal groove2nd pharyngealpouchPrimary lungdiverticulumPrimarybronchiTracheaseparated fromesophagusSecondarybronchiAsymmetriclung buds;3 bronchialareas in rightlung budBucconasalmembranerupturedMajor bronchialdivisionsPalatal shelvesuniting (Not allat the samestage of union)DevelopingalveoliabcComparative gestational milestones in respiratory systemAge(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken11 3.3 21–2511 3.3 21–2511.5–1211.75–12.53.8 26–284.2 29–3112.75 7 43–4512.75–137 43–45Ref.1,2,9,18,22–252,9,18,252,9,10,18,252,9,18,252,9,18,252,9,10,18,25Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)18–18.5Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomites8.75 7 8.5 44 24–26 339.5–9.7525± 7,10,76Ref.Age(d)Ref.Age(d/h)56 25 83 ca.50–54hSize Somites(mm) aSomitesRef.23 8810.5 10 9 10 16 10 27 10 26–28 3.3 10,71, 76,84,140–1423d 109 44 18.5 56 29–30 33 29 4.5 71 96h 8811 76 9 44 16.5 56 29–32 6 71,76 96h 8812 7,77 18.5 10 35–38 9 71,7710.5–11.510,76 12 10 9.5–10 10,44 18.5–21.513 60 15–15.2515.5 10 13 10 15 10 10.5–1117 10 15 10 19.5 10 12–12.5Crown-rump lengthPattern of avian lung development is different from that of mammals.In the chick, palatal shelves do not fuse.10,56 29 10 32 10 NA b 1060 36–42 60 48–51 16–18 6010,44 21.5 10,56 36 10 46 10 NA b 1010,44 26 10 45–46 10,33 57 10 NA c 1016.5 139 85 139174 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Table 6.8DescriptionComparative gestational milestones in nervous system developmentAge(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)Primitive streak 9 10 8 10 7.25 10 7 10 13 10 17 10 17 10 12h 10Neural plate 9–9.5 1 1,2,4, 7 5,7–10 8 10 7.5 10 13.5 10 20 10 18–19 1–1.5 10,39 1d 109,10,18,23Elevated brainplate, neuralfolds9.5 1.5 1–4 1,2,4,9,11,18,237.75–8.251–4 27 14.5 2 56 ca19–201–2 115 23–26h 4 88Neural crest forganglia of IXand X; spinalflexuresometimespresentTrigeminal,neural crestcomponent;neural crest atlevel ofmetencephalonGanglia of VIIand VIII; neuralcrest andposterodorsalepibranchialplacode of firstpharyngealgrooveFusion ofneural folds(early)AnteriorneuroporecompletelyclosedOtic pits; opticvesicles, andauditory pitsappearBothneuroporesclosed —anterior first10 2 5–12 1,2,4,9,18,2310 2 5–12 1,2,4,9,18,2310 2 5–12 1,2,4,9,18,2310.5—10.7510.5–10.7510.5–10.7510.5–11.52.4 17 1,2,4,10,11,18,23,2711,27 8.5–9 10–149d1h11,27 8d22h18± 5,7–9,399–9.5 10,2716 39 8.5–9 10–1410,11 9–9.5 10 9.5–10.58.5 17–208.5 17–2027 8.25 12–138.5 17–2027 8.25 12–13Ref.Age(d/h)24±1 3–5 92,144 ca45–49hSize Somites(mm) aSomites44 ca.40–45h 13–1417 8844 ca 45h 15 8844 14–1510 8.5–9 10 15.25–15.515,55 22–24 2–3.5 4–12 66–68,71,73,74,79–8210,44 15.25 10 25 10 24–26 2.5–4.913–2044 20 3910 25–31 10 25–28 1010,39,71,73,83Ref.8826–28h 3–4 282.3d 10COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 175© 2006 by Taylor & Francis Group, LLC

Table 6.8DescriptionOptic vesicle;optic pitsNeural tubedifferentiatesinto the threeprimary brainvesicles;anteriorneuroporeclosedFive brainvesiclesOtic cyst andotic pit closed;endolymphaticappendage;deep cervicalflexurePosteriorneuroporeclosingThickened lensdisc; lensplacodeEndolymphaticevaginationShallowolfactory pitsComparative gestational milestones in nervous system development (continued)Age(d)10.5–1110.5–12Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.Age(d)Ref.Age(d)2 4–13 27,89 8.5–9 10–142.4 13–2011.5 3.8 26–2811.5–12Posterior 11.5–neuropore 12closes bCerebralhemispheres(early)1,2,4,9,10,18,231,2,4,9,18,2327 9.1–108 5,7–10 8.25–8.5Ref.27,99Age(d)8.5 17–2027 8.5 17–20AgeRef. (d)44 15–1610,44 15–15.3Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)56 21–23 33 24± 2.5–5 13–2010,15 25–30 10,60 26±1 3.8–4.9Ref.71,73,92,1448.75 44 17.5 29 56 30–33 60 33± 7–11 77,144,14519± 39,60 10 99 16.75 56 28–30 60 28–32 4–6 39,60,14611 11 9.5 24± 76 9+ 99 27 60 26± 7612–12.1255.2 36 1,2,4,9,10,18,23Pontine flexure 13.5 8.5 49–51Vomeronasalorgan9.625 25± 76 11 99 28–29 33 28–35 5.4 38 76,84,8510 60 9 44 19 56 28–30 6010 60 11–1227,999 44 28–30 33,6011,27 10.5 60 10 99 9 44 28–30 60 26–27 3–6 21–291,2,4,9,18,2310 27± 5,7–9,7710.5 5,7–9,6011 10,99Age(d/h)Size Somites(mm) aSomitesRef.29–33 7 8610,39 33–38 10 10,8871,73,769 10 17 10 29 10 29–33 5.5–9 10,73,76,77,14410 99 30–33 60 35–38 7–9 6011.5 77 37± 773d 10,88176 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Formation ofchoroid plexusof lateral andfourthventricles13.5 8.5 49–511,2,4,9,18,2312.5–135,7–9,60,14311 44 36–42 60 48–51 16–18 60,143Cerebellum 14 10 12 10 15 10 11 10 19 10 30–36 10,60 37 10 4.5d 10Lens cavity12 60 34–36 60 42–44 11–14 60crescentic;primitive lensfibers present;primordialsemicircularducts;primordialcochlear duct;nasal finSuperior andinferior colliculiseparate12.5 77 37± 77Earliest reflexresponsesobservedNeostriatum;telencephaliccortexCrown-rump length.Posterior neuropore closes earlier in primates than in rodents.ab41±1 18–23 144,147,148~17 83 8–12d 88COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 177© 2006 by Taylor & Francis Group, LLC

Table 6.9DescriptionEyeComparative gestational milestones in the development of sense organsAge(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)Optic sulci 10 2 2 8.54 23± 39 14 39Optic vesicle 10.5 10 10 9.5 20± 10,39 9 10 8 10 10 15.5 13 10,56 23–25 10 10,33 21–24 2–2.8 10,74 26–33h 10,88formingOptic bulbs 10.5 2.4 2 23± 39,60Lens placode 11.5 3.8 2 10 24± 39,60 9.5 44 28–30 33,60 27–35 3.3– 5.4 84,149,15120– 21176Invagination oflens placodesbeginsExtrinsicpremusclemasses appearLens vesicleforming;choroid fissure11.75 4.2 2 10–10.560,76 9.25 44 17.5 29 56 28–32 60 27–32 4–7 30+ 71,73,7631–44 4.3–14Ref.152Age(d/h)Size Somites(mm) aSomites48h 8812.125 5.2–6 2 10 44 16.5 56 35–38 60 48h 88–12.375Lens separated 12.5 6.2 2,10 11.5 10 11.5 10 10 10,44 18 10 32 10 35–37 6 10,84,149,151,153,154Primordium ofhyaloid arteryLens vesicleclosedDefinite retinalpigmentationDifferentiationof retinaChoroid fissurefusing; hyaloidcanal12.5 6.2 2 10 44 35–37 6 84,149,151,153,15413–14 6.2 2,11,2711–11.511–1260 19.75 56 32–34 60 34.5 7–9 7125± 39,60 14–1513.5 8.5 2 12 149 15–15.2527,5730–36 60 33–44 7–14 25 39,60,73,77,8442–44 12–14 84,149,15160 34–36 60 42–44 14.5 60,1492.5d 103–3.5d 40–43100h 88Ref.86,88178 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Optic nervefibers presentDifferentiationof corneaAnteriorchamberdifferentiatingCiliary bodydifferentiation;primordium ofchoroideascleraEar14 9.5 2,10 13 10 15 10 11 10 21.5 10 39 10 46–48 14.6–15.613 60 15–15.2510,84,149,15160 36–42 60 48–51 16–18 60,84,149,1514d 1015 12 2 14 9 40–44 9 53–54 22–24 60 6d 8856–70 26–45 84,149,151Otic placodes 10 2 2 8.54 23± 39 27 14 39Otic cups 10.5 2.4 2,11 23± 39Otic vesicleformingOtocyst(closure)Otic vesicleswith shortendolymphaticductEndolymphaticsac appearspinched offfrom oticvesicleThickened,hollowepithelialprimodia ofsemicircularcanalsCochlear andvestibularregionsSeparation ofutricular andsaccularregionsCochleaappearing11.5 10 8.5–8.7511–11.53.3–3.813 10,39 9 10 8.5 10 15.5 13 10,56 25 10 21–29 2–6 10,73,1468d 882.3d 102,11 24± 39 9 44 29 33 30 4 17 39,146 11d 12 8811.75 4.2 2 10.5 76 29–32 5–7 29± 73,7612–13 11,27 11 77 30 33 33–37 6.5 77,146,15512–13 6.2 2,27 11–11.511.5 77 38–44 8–13 35± 77,155,15660 30–34 60 35–42 7–9 6012.5 6.2 2 14.5 150 33 150 6–7d 8813.5 10 12 10 13 10 10 10 20.5 10 37 10 44 10 7d 10COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 179© 2006 by Taylor & Francis Group, LLC

Table 6.9DescriptionOne or moresemicircularcanals formedCondensationsof ocularmusclesOtic capsulecartilaginousOcular musclesinnervatedOssification ofanteriorprocess ofmalleusOlfactoryOlfactoryplacodesComparative gestational milestones in the development of sense organs (continued)Age(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.Age(d)Ref.Age(d)Ref.Age(d)14 9.5 2 12 77 34–36 60 44–48 13.5–1514 9.5 2Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)Ref.37± 77,84,146,149,151,155,156Age(d/h)Size Somites(mm) aSomitesRef.5–6d 8815 10 14.5 10 20 10 11 10 22 10 42 10 56 10 8d 1015 12 2 6d 8811–11.5Olfactory pit 12.125 5.2–5.62–3.8 4–13 2,89 10 24–272 10.5–1156–70 28 160,16139,76 8.75 44 28–29 33 28–31 7–9 17–2539,76,77,157,15860,76 18.5 56 30–33 60 33–38 7–8 29± 73,76,84,141,15952–64h 29–32Olfactory 12.5 6.2 2 27± 39 4–6d 88nerve; olfactoryepitheliumOlfactory bulbs 13.5 10 11 10 14 10 11 10 23 10 38 10 37 10 7d 10Partlycartilaginousnasal septumand capsule15 12 2 6d 88Crown-rump length.a28 8888180 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Table 6.10DescriptionMuscular SystemComparative gestational milestones in muscular, skeletal, and integumentary system developmentAge(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)Primitive streak 9 10 8 10 7.25 10 7 10 13 10 17 10 17 10 0.5d 10First pharyngeal 10 2 5–12 1–4,6, 8–8.5 5,7–9 9.5 27 23 10 83arch9Ten-somite stage10.5 10 10 8.5 10 8.5 10 10 8 10 10 15 10 10 23 10 10 25 10 10 1.5d 10 10Pharyngealarches I and II,clefts I and IIAnterior limbbudThreepharyngealarchesHind limb bud b 11.5–12Appearance of4th pharyngealarchesMaxillaryprocesses meetnasolateral andmedialprocessesSubdivision offorelimb budCervical sinusobliteratedDigital rays(forepaw)Pleuroperitonealcanal closed;completediaphragm11 3.3 21–2511.5 3.8 26–283.8 26–2811.75 4.2 29–3112–13 7 43–4512.75–1413–13.513.5–147 43–458–8.5 46–5115 12 61–631–4,6,91–4,6,91–4,6,9,101–4,6,91–4,6,9,271–4,6,9,271–4,6,9,18,253,4,10,27,29–321–4,6,99.5–9.7523± 5,7–9,3910 5,7–9,6010–10.35,7–9,109.5–9.7510.5–1127 8.5 17–2027,578.75 17–2044 24–27 3.3 13–2042,44,46,4716.5 23 56 25–26 46 26 3–5 21–299.75 27 8.75 44 16.5 23 56 28–30 60 28–32 4–6 21–2411–1210.25 5,7–9 9.75–10.510.5 60 12–1312 5,7–9 13–1412.3 7,10,29,409 12–1310,27,5714.5 10,27,299 44 17.5–18.529 10,56 28–30 10,60 28–32 4–6 30–3227 9 44 16.5 23 56 28–29 33 32 4.6–5 33–34Ref.Age(d/h)Size Somites(mm) aSomitesRef.9,71, 83–85,912–3d 22 86,8739,73 51–56h 26– 30,862860,84,92,1669,10,60,71,73,84,8553–55h 24–272.2–3d 29–3283,84 3.5d 43–4427 9.25 44 37–38 33 42–45 12–13 83 23–26h 4–5 8827 10.5 44 20.75 39 56 30–32 60 31–35 5–7 6027 10 44 23.75 56 33–34 33 40–44 8–15 83,8410.25–1110,42,44,4822–23.7586,8710,8610,56 34–35 10,46 35–37 8–11 10 4.75d 30,8626 56 20 8386COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 181© 2006 by Taylor & Francis Group, LLC

Table 6.10DescriptionSkeletal SystemSubdivision offorelimb budSubdivision ofhindlimb budMesenchymalcondensationfor ribsAuditoryossicles;mesodermabove dorsalextremity oftubotympaniccavityDigital rays(forepaw)Anlagen ofcentra andneural archesFirst sign ofelbowDistinct fingerrays; rim ofhand platecrenated;primitivepalatineprocessesChondrificationcenters in ribsPrimordialMeckel’scartilagesInterdigitalnotches in handplateFirst sign ofwristsComparative gestational milestones in muscular, skeletal, and integumentary system development (continued)Age(d)12.75–1413–14.5Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken7 43–458.5 19–5114 9.5 52–5515–15.5Ref.1–4,6,9,271–4,6,9,10,11,271,2,9,13–15,17–22Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)10.5 60 12– 27 10.5 44 20.75 39 56 30–32 60 31–35 5–7 601311 60 30–33 60 37 8–11 7312 162 10.5 44 31–33 162 5d 8812–12.55,7–9,10,6014.5 10 11–14 10,44,60Ref.Age(d/h)Size Somites(mm) aSomites11 44 5–6d 8822 10 34–38 10,60 37–48 11–17 10,60,73, 84,1219.5 44 18.5 31 56 28 1674.75d 1013 60 36–42 60 48–51 16–18 60 4.5–5d 8613 60 14 60 35–38 60 44–48 13–17 6013 162 11 44 23.75 56 15 16213 60 14 60 35–38 60 44–48 13–17 6027 13.5 77 16–1727 36–42 60 37± 7714 60 40–44 60 53–54 22–24 60Ref.182 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Distalphalanges offingersseparatedFirst seven ribschondrified andin contact withsternumDigitalseparation ofhindpawsCompleteseparation ofdigits offorelimbPalatal foldsuniting (not allat the samestage of union)Primaryossificationcenter inhumerus withtrabeculaeOssificationcenters presentin all ribs16.5–1715.5–16.516.5–17Integumentary SystemMilk lineappearsNasolacrimalgroovePeridermpresentSkindifferentiatedinto stratumgerminativumand stratumintermedium17–18 1–4,6,9,10,1112.125–135.2–8 8–36 1–4,6,9,13–15,17–2211 14 60 40–44 60 53–54 22–24 6014.5 162 30–35 16227 40–44 60 45 16–18 6011,27 17–17.515–16.55,7–9,1027 46–50 60 56–60 27–31 6019.5 10 12 10 26 10 44–48 10,60 54–57 23–28 10,60 NA c15.5 123 43–44 33 50 12315.5 163 68 16312 5,7–9 14–1527 10 44 21.75 56 7 8313–14 27 11.5 60 30–34 60 37–42 8–11 6012 164 16 16413.5 165 22 165COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 183© 2006 by Taylor & Francis Group, LLC

Table 6.10DescriptionStratumgranulosumVibrissarypapillae appearon maxillaryprocessHair follicleprimordiaappearingDistinctauricularhillocksEyelids ––smallectodermalfoldsFeather germs(chick)First trunk hairpapillae appearabcComparative gestational milestones in muscular, skeletal, and integumentary system development (continued)Age(d)14–14.514–14.5Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken9.5 52–559.5 52–5515 12 61–6315.5–16.5Ref.1,2,9,11,13–15,17–22,271–4,6,9,271,2,9,13–15,17–22Size Somites(mm)Age SomitesAge Som-Age Som-Age Som-Age Som-Age Size Som-(d)Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) (mm) a ites Ref.15.5 164 60 16414–1511.5 60 14–1513 5,7–9,60,15411,27 17.5–18.75Crown-rump length.Hindlimb bud forms earlier in the rodents than in primates.In the chick, palatal folds do not fuse.5,7,8,2710.5 44 23.75 56 N/A12 4427 14 60 32–34 33,60 37–48 8–17 60,73 N/A15–15.2560 36–42 60 48–51 16–18 60,15427 12.5 44 NAAge(d/h)Somites6.5–7d 86Ref.184 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Table 6.11DescriptionIntermediatemesodermthickensnephrogeniccordPronephrosappearsTen-somitestageNephrogeniccord withmesonephrictubules andductMesonephrosappearsComparative gestational milestones in excretory system development (continued)Age a(d)Kidney: 11.75–mesonephric 12duct entersurogenitalsinus or cloacaPrimitivemesonephrictubules, mostlysolid; WolffianductdiscontinuousGerminalepithelium(testis)appearingKidney: uretericbudUrorectalseptumdividing cloacaUreteric budwithmetanephric“cap”Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.10 2 5–12 2,9,112,113Size Somites(mm)Age SomitesRef. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) (mm) a ites Ref.Age Som-Age Som-Age Som-Age Som-Age Size Som-(d)8.75 7 7.75 44 15.5 13 56 22 2.1 7110 10,26 22 10,26 1.4–1.5d10.5 10 10 8.5 10 10 8.5 10 10 8 10 10 15 10 10 23 10 10 25 10 10 1.5d 10 1011 3.3 21–252,9,112,1138d21h–9.515–245,7–9,38,79Age(d/h)8.75 44 16.5 56 25–28 3.5 10 71,79 1.75d 30,8611.5 10 10,26 9.5 10,26 24–25 3.5 10 26,38 2.3–3d 10 10,2611.75–12.125 4.2–5.24.2 29–3129–362,9,10,26,112,1132,9,112,11312 10,26 11.5–12.512.3 26 11–11.512.375–176–6.2 39–421,9,10,18,22–25,112,11312.5 2,9,10,26,112,11311 5,7–9,10,26,7611.5 10 9–9.5 10,44 17.5–1929–3517.5–21.75 29–3510,56 28–29 10,33 28± 4.5 10,26,71,76Somites3d 1056 28 3.5 71 1.75d 19–3210,26 13 10 9 10 33–34 10 38–40 10,26 4d 10,2676,114 11.5 10 9.25–9.344 19–2035 56 29–30 33 28–29 26,76 4d 11415 10 9.5 10 28–48 1013 10 9.5 44 21 10 31–32 10 32 6 10,26,71Ref.10,26885d 10COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 185© 2006 by Taylor & Francis Group, LLC

Table 6.11DescriptionKidney:metanephrosParamesonephric ductappearsTesteshistologicallydifferentiatedParamesonephric ductreaches cloacaRectum andurogenitalsinuscompletelyseparatedaComparative gestational milestones in excretory system development (continued)Age a(d)12.5–12.8Crown-rump length.Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken13.5 8.5 49–51Size Somites(mm)Age SomitesAge Som-Age Som-Age Som-Age Som-Age Size Som-Age Som-Ref. (d)Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) ites Ref. (d) (mm) a ites Ref. (d/h) ites Ref.10,26 11 10 23 10 10 10 23 10 38–39 10 35–37 10,26 6d 10,262,9,10,26,112,11313.5 10,26,11410 5,7–9,10,2612 10,26 16.5 10,2615 10 11 10,44 23.75 56 35–36 10 42–44 10,26 4d 10,2612 10,26 26 10,2636–39 10,26 46–48 10,26 13d 10,2615.5 10,26 14 10 20 10 13 10 26 10 37–38 10 49–56 10,26 7d 1017 16 2,9,112,11315 44 43 59 N/A186 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Table 6.12DescriptionComparative gestational milestones in reproductive system developmentAge a(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human ChickenRef.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)Primitive streak 9 10 8 10 7.25 10 7 10 13 10 17 10 17 10 0.5d 10Germ cells inyolk–sacepitheliumPronephrosappearsTen-somitestageGerm cells inmesenteryMesonephrosappearsGerm cellmigrationreachesborders ofmesonephricridgesGerminalepitheliumMesonephricduct entersurogenitalsinusIndifferentgonadal folds;rapid increasein number ofgerm cellsGerm cells ingenital ridges,end ofmigration9.5–10 1.5–2 1–12 1–4,6,9,18,112,131,134,168–1718–8.5 5,7–9 27 3.3 13–2010 114 22 114 1.4– 10,261.5d10.5 10 8.5 10 10 8.5 10 10 8 10 10 15 10 10 23 10 10 25 10 1.5d 10 1010.5–11.52.4–3.813–281–4,6,9,18,112,131,134,168–1719.5 5,7–9 29 3.8 25–2711.5 114 9.5 114 24 114 3d 11411.75 4.2 29–312,9,18,112,131,134,168–17112 10,11412 10,11412.125 5.2 36 2,9,112,11312.75 7 43–452,9,18,112,131,134,168–17111.5–12.520.75 39 56 31 4.3 30–3210,114 13 10 9 10 33–34 10 38–40 8 3–4d 38 10,88,11411 76,114 11.5 10 9.25–9.3Ref.9,84,85,9184,85,1169,84,8510,44 19 35 10,56 28–29 10,33 28 4.5 10,26,71,76,11433–34 3335–37 6 9,84,121,174Age(d/h)Size Somites(mm) aSomitesRef.COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 187© 2006 by Taylor & Francis Group, LLC

Table 6.12DescriptionParamesonephric ductappearsGonads beginsexualdifferentiationHistologicdifferentiationof testesGonad, retecords; instromabetweengenitalprimordiumandmesonephrosOogonia; germcells insecondarysexual cords ofovarian cortexParamesonephric ducts reachurogenitalsinusIndifferentexternalgenitaliaDifferentiationof male andfemale externalgenitaliaaComparative gestational milestones in reproductive system development (continued)Age a(d)Rat Mouse Rabbit Hamster Guinea Pig Rhesus Monkey Human Chicken13.5 8.5 49–5113.5 8.5 49–5113.5–14.5Crown-rump length.10.5 56–6014.5 10.5 56–60Ref.2,9–10,112–1142,9,18,25,129–1342,9,18,112,114,131,134,168–1712,9,18,112,131,134,168–171Age(d)Ref.10 5,7–9,10,11412.5 5,7–9,172Age(d)Ref.Age(d)Ref.Age(d)Ref.Age(d)Size Somites(mm)SomitesSomitesSomitesSomitesSomitesRef.Age(d)15 10 11 10,44 23.75 56 35–36 10,33 42–44 10,114 4d 10,88,11417 8312 11,173 38–39 33 46–48 14–16 85,114,16,175Ref.Age(d/h)Size Somites(mm) aSomitesRef.5.5d 11412.5 172 21.75 56 17 83 5d 8817 83 7–8d 8815.5 14.2 64 2,9,18,112,114,131,134,168–17149–56 114 7d 8819 10 37 1056–70 26–45 9,66,84,121188 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 189References1. Mossman, H.W., Comparative morphogenesis of the foetal membranes and accessory uterine structures,Contrib. Embryol. Carneg. Inst., 26, 129, 1937.2. Beck, F., Comparative placental pathology and function, Environ. Health Perspect., 18, 5, 1976.3. Ramsey, E.M., The Placenta: Human and Animal, Praeger, New York, 1982.4. Enders, A.C., Mechanisms of implantation of the blastocyst, in Biology of Reproduction: Basic andClinical Studies, Velardo, J.T. and Kasprow, B A., Eds., III Pan American Congress of Anatomy, NewOrleans, LA, 1972, p. 313.5. Reynolds, S.R.M., On growth and form in the hemochorial placenta: an essay on the physical forcesthat shape the chorionic trophoblast, in Biology of Reproduction: Basic and Clinical Studies, Velardo,J.T. and Kasprow, B.A., Eds., III Pan American Congress of Anatomy, New Orleans, LA, 1972, p. 7.6. Freese, U.E., The maternal-fetal vascular relationships in the human and rhesus monkey, in Biologyof Reproduction: Basic and Clinical Studies, Velardo, J.T. and Kasprow, B.A., Eds., III Pan AmericanCongress of Anatomy, New Orleans, LA, 1972, p. 335.7. Harris, J.S.W. and Ramsey, E.M., The morphology of human uteroplacental vasculature, Contrib.Embryol., 38, 43, 1966.8. Wild, A.E., Endocytic mechanisms of protein transfer across the placenta, Placenta, 1, 165, 1981.9. Miller, R.K., Koszalka, T.R., and Brent, R.L., Transport mechanisms for molecules across placentalmembranes, in Cell Surface Reviews, Poste, G. and Nicholson, G., Eds., Elsevier/North-Holland,Amsterdam, 1976, p. 145.10. Miller, R.K., Placental transfer and function: The interface for drugs and chemicals in the conceptus,in Drug and Chemical Action in Pregnancy: Pharmacologic and Toxicologic Principles, Fabro, S.and Scialli, A.R., Eds., Marcel Dekker, New York, 1986, p. 123.11. Arceci, R.J., Baas, F., Raponi, R., Horwitz, S.B., Housman, D., and Croop, J.M., Multidrug resistancegene expression is controlled by steroid hormones in the secretory epithelium of the uterus, Mol.Reprod. Dev., 25, 101, 1990.12. Nakamura, Y., Ikeda, S., Furukawa, T., Sumizawa, T., Tani, A., Akiyama, S., and Nagata, Y., Functionof P-glycoprotein expressed in placenta and mole, Biochem. Biophys. Res. Commun., 27, 235, 1997.13. Smit, J.W., Huisman, M.T., van Tellingen, O., Wiltshire, H.R., and Schinkel, A.H., Absence orpharmacological blocking of placental P-glycoprotein profoundly increases fetal drug exposure, J.Clin. Invest., 104, 1441, 1999.14. Miller, R.K., Levin, A.A., and Ng, W.W., The placenta: relevance to toxicology, in Reproductive andDevelopmental Toxicity of Metals, Clarkson, T., Nordberg, G., and Sager, P., Eds., Plenum Press, NewYork, 1983, p. 569.15. Juchau, M.R., Mechanisms of drug biotransformation reactions in the placenta, Fed. Proc., 31, 48, 1972.16. Juchau, M.R., Drug biotransformation in the placenta, Pharmacolog. Therapeut., 8, 501, 1980.17. Juchau, M.R. and Rettie, A.E., The metabolic role of the placenta, in Drug and Chemical Action inPregnancy: Pharmacologic and Toxicologic Principles, Fabro, S. and Scialli, A. R., Eds., MarcelDekker, New York, 1986, p. 153.18. Kelman, B.J., Effects of toxic agents on movements of material across the placenta, Fed. Proc., 38,2246, 1979.19. Goodman, D.R., James, R.C., and Harbison, R.D., Placental toxicology, Food Chem. Toxicol., 20,123, 1982.20. Juchau, M.R., The role of the placenta in developmental toxicology, in Developmental Toxicology,Snell, K., Ed., Praeger, New York, 1982, p. 187.21. Millicovsky, G. and DeSesso, J.M., Uterine versus umbilical vascular clamping: differential effectson the developing embryo, Teratology, 22, 335, 1980.22. Millicovsky, G., DeSesso, J.M., Clark, K.E., and Kleinman, L. I., Effects of hydroxyurea on maternalhemodynamics during pregnancy: a maternally mediated mechanism of embryotoxicity, Am. J. Obstet.Gynecol., 140, 747, 1981.23. Beck, F., Lloyd, J.B., and Griffiths, A., Lysosomal enzyme inhibition by trypan blue: a theory ofteratogenesis, Science, 157, 1180, 1967.24. Williams, K.E., Roberts, G., Kidston, M.E., Beck, F., and Lloyd, J.B., Inhibition of pinocytosis in ratyolk-sac by trypan blue, Teratology, 14, 343, 1976.© 2006 by Taylor & Francis Group, LLC

190 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION25. New, D.A.T. and Brent, R.L., Effect of yolk-sac antibody on rat embryos grown in culture, J. Embryol.Exp. Morphol., 27, 543, 1972.26. Brent, R.L., Beckman, D.A., Jensen, M., Koszalka, T.R., and Damjanov, I., The embryopathologiceffects of teratogenic yolk sac antiserum, Trophoblast Res., 1, 335, 1983.27. Parizek, J., Vascular changes at sites of oestrogen biosynthesis produced by parenteral injection ofcadmium salts: the destruction of placenta by cadmium salts, J. Reprod. Fertil., 7, 263, 1964.28. Dencker, L., Possible mechanisms of cadmium fetotoxicity in golden hamsters and mice: uptake bythe embryo, placenta and ovary, J. Reprod. Fertil., 44, 461, 1975.29. Morford, L.L., Henck, J.W., Breslin, W.J., and DeSesso, J.M., Hazard identification and predictabilityof children’s health risks from animal data, Environ. Health. Perspect., 112, 266, 2004.30. Browder, L.W., Erickson, C.A., and Jeffrey, W. R., Developmental Biology, 3rd ed., W. B. Saunders,Philadelphia, 1991.31. Gilbert, S.N., Developmental Biology, 5th ed., Sinauer Associates, Sunderland, MA, 1997.32. Larsen, W.J., Human Embryology, 3rd ed, Churchill Livingstone, New York and London, 2001.33. Carlson, B.M., Human Embryology and Developmental Biology, 3rd ed., C.V. Mosby, St. Louis, 2004.34. Hay, E. D., Embryonic induction and tissue interaction during morphogenesis, in Birth Defects,Littlefield, J.W. and De Grouchy, J., Eds., Excerpta Medica, Amsterdam, 1978, p. 126.35. Holland, P.W.H. and Hogan, B.L.M., Expression of homeobox genes during mouse development: areview, Genes Dev., 2, 773, 1998.36. Kessel, M. and Gruss, P., Murine developmental control genes, Science, 249, 374, 1990.37. DeRoberts, E.M., Oliver, G., and Wright, C.V.E., Homeobox genes and the vertebrate body plan, Sci.Am., 7, 46, 1990.38. Edelman, G., Topobiology: An Introduction to Molecular Embryology, Basic Books, New York, 1988.39. National Research Council, Scientific Frontiers in Developmental Toxicology and Risk Assessment,National Academy Press, Washington, DC, 2000.40. DeSesso, J.M., Maternal factors in developmental toxicity, Teratogenesis, Carcinog. Mutagen., 7, 225,1987.41. Wilson, J.G., Environment and Birth Defects, Academic Press, New York, 1973.42. Beckman, D.A. and Brent, R.L., Basic principles of teratology, in Medicine of the Fetus & Mother,Reece, E.A., Hobbins, J.C., Mahoney, M.J., and Petrie, R.H., Eds., J. B. Lippincott, Philadelphia,1992, p. 293.43. DeSesso, J.M. and Harris, S.B., Principles underlying developmental toxicity, in Toxicology and RiskAssessment, Fan, A.M. and Chang, P., Eds., Marcel Dekker, New York, 1996, p. 37.44. Mohrenweiser, H. and Zingg, B., Mosaicism: the embryo as a target for induction of mutations leadingto cancer and genetic disease, Environ. Mol. Mutagen., 25(Suppl. 26), 21, 1995.45. Drost, J.B. and Lee, W.R., Biological basis of germline mutation: comparisons of spontaneous germlinemutation rates among Drosophila, mouse, and human, Environ. Mol. Mutagen., 25(Suppl. 26), 48,1995.46. Generoso, W.M., Rutledge, J.C., Cain, K.T., Hughes, L.A., and Downing, D.J., Mutagen-induced fetalanomalies and death following treatment of females within hours after mating, Mutat. Res., 199, 175,1988.47. Wilson, J.G., Embryological considerations in teratology, in Teratology: Principles and Techniques,Wilson, J.G. and Warkany, J., Eds., University of Chicago Press, Chicago, 1965.48. Ritter, E.J., Scott, W.J., and Wilson, J.G., Relationship of temporal patterns of cell death and developmentto malformations in the rat limb: possible mechanisms of teratogenesis with DNA synthesisinhibitors, Teratology, 7, 219, 1973.49. Lammer, E.J., A phenocopy of the retinoic acid embryopathy following maternal use of etretinate thatended one year before conception, Teratology, 37, 472, 1988.50. Warkany, J., Sensitive or critical periods in teratogenesis: uses and abuses of embryologic timetables,Congenital Malformations, Year Book Medical Publishers, Chicago, 1971, p. 49.51. Hamilton, W.J. and Mossman, H.W., Hamilton, Boyd and Mossman’s Human Embryology, 4th ed.,Williams and Wilkins, Baltimore, 1972.52. O’Rahilly, R. and Muller, F., Developmental Stages in Human Embryos, Publication No. 637, CarnegieInstitute, Washington, DC, 1987.© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 19153. Edwards, J.A., The external development of the rabbit and rat embryo, in Advances in Teratology,Woolam, D.H., Ed., Academic Press, New York, 1968, p. 239.54. Hebel, R. and Stromberg, M.W., Anatomy and Embryology of the Rat, BioMed Verlag, Worthsee,Germany, 1986.55. Nelsen, O.E., Comparative Embryology of the Vertebrates, McGraw-Hill, New York, 1953.56. Snell, G.D. and Stevens, L.C., The early embryology of the mouse, in Biology of the LaboratoryMouse, Green, E.L., Ed., Blakiston, Philadelphia, 1966, p. 205.57. Rugh, R., The Mouse: Its Reproduction and Development, Burgess, Minneapolis, 1968.58. Boyer, C.C., Chronology of development of the golden hamster, J. Morphol., 92, 1, 1953.59. Waterman, A.J., Studies on the normal development of the New Zealand White strain of rabbit, Am.J. Anat., 72, 473, 1943.60. Scott, J.P., The embryology of the guinea pig. I. Table of normal development, Am. J. Anat., 60, 397,1937.61. Heuser, C.H. and Streeter, G.L., Development of the macaque embryo, Contrib. Embryol., 29, 15,1941.62. Hendrickx, A.G. and Sawyer, R.H., Embryology of the rhesus monkey, in The Rhesus Monkey, Vol.2, Bourne, G.H., Ed., Academic Press, New York, 1975, p. 141.63. Hamburger, V. and Hamilton, H.L., A series of normal stages in the development of the chick embryo,J. Morphol., 88, 49, 1951.64. Patten, B.M., Early Embryology of the Chick, 4th ed., McGraw-Hill, New York, 1951.65. Arey, L.B., Developmental Anatomy, 5th ed., W.B. Saunders, Philadelphia, 1965.66. Streeter, G.L., Developmental horizons in human embryos (XI-XII), Contrib. Embryol., 30, 211, 1942.67. Streeter, G.L., Developmental horizons in human embryos (XIII-XIV), Contrib. Embryol., 31, 27,1945.68. Streeter, G.L., Developmental horizons in human embryos (XV-XVIII), Contrib. Embryol., 32, 133,1948.69. Wanek, N., Muneoka, K., Holler, D.G., Burton, R., and Bryant, S.V., A staging system for mouselimb development, J. Exp. Zool., 249, 41, 1989.70. Witschi, E., Development of Vertebrates, W.B. Saunders Company, Philadelphia, 1956.71. DeSesso, J.M., Niewenhuis, R.J., and Goeringer, G.C., Lectin teratogenesis II. Demonstration ofincreased binding of concanaualih A to limb buds of rabbit embryos during the sensitive period,Teratology, 39, 395, 1989.72. Holson, J.F., Stump, D.G., Pearce, L.B., Watson, R.E., and DeSesso, J.M., Mode of action: yolk sacpoisoning and impeded histiotrophic nutrition. HBOC-related congenital malformations, Crit. Rev.Toxicol., in press, 2005.73. DeSesso, J.M. and Lavin, A.L., Stem cell research. Sigma, Fall edition, 2002, p. 11. Available onlineat http://www.mitretek.org/pubs/sigma/fall2002/chap3.pdf.References for Tables 3 to 121. Butcher, E.O., The development of the somites in the white rat (Mus norvegicus albinus) and the fateof the myotomes, neural tube, and gut in the tail, Am. J. Anat., 44, 381, 1929.2. Henneberg, B., Normentafel zur Entwicklungsgeschichte der Wanderratte (Rattus norvegicus Erxleben),G. Fischer, Jena, Germany, 1937.3. Huber, G.C., The development of the albino rat, Mus norvegicus albinus, J. Morphol., 26, 1, 1915.4. Long, J.A. and Burlingame, M.L., The development of the external form of the rat with someobservations on the origin of the extraembryonic coelom and fetal membranes, Univ. Calif., (Berkeley)Mem. Zool., 43, 143, 1938.5. MacDowell, E.C., Allen, E., and MacDowell, C.G., The prenatal growth of the mouse, J. Gen. Physiol.,11, 57, 1927.6. Nicholas, J.S. and Rudnick, D., Development of rat embryos of egg-cylinder to head-fold stages inplasma cultures, J. Exptl. Zool., 78, 205, 1938.© 2006 by Taylor & Francis Group, LLC

192 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION7. Otis, E.M. and Brent, R., Equivalent ages in mouse and human embryos, Anat. Record, 120, 33, 1954.8. Snell, G.D., The Early Embryology of the Mouse, Blakiston, Philadelphia, 1941.9. Witschi, E., The Development of Vertebrates, W.B. Saunders, Philadelphia, 1956.10. Hoar, R.M. and Monie, I.W., Comparative development of specific organ systems, in DevelopmentalToxicology, Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York 1981, p. 13.11. Christie, G., Developmental stages in somite and post-somite rat embryo, based on external appearance,and including some features of the macroscopic development of oral cavity, J. Morphol., 114,263, 1964.12. Altman, P., and Dittmer, D.S., Eds., Growth Including Reproduction and Morphological Development,Federation of American Societies for Experimental Biology, Washington, D.C., 1962, 304.13. Burlingame, P.L. and Long, J.A., The development of the external form of the rat, with someobservations on the origin of the extraembryonic coelom and foetal membranes, Univ. Calif. (Berkeley)Publs. Zool., 43, 249, 1939.14. Goss, C.M., Development of the median coordinated ventricle from the lateral hearts in rat embryoswith thirteen somites, Anat. Record, 112, 761, 1952.15. Harman, M.T. and Prickett, M., The development of the external form of the guinea-pig (Cavia Cobaya)between the ages of eleven days and twenty days of gestation, Am. J. Anat., 49, 351, 1932.16. Krehbiel, R.H., Cytological studies of the decidual reaction in the rat during early pregnancy and theproduction of deciduomata, Physiol. Zool., 10, 212, 1937.17. Marit, C., Les premiers stades de la formation du myocarde chez le rat, Bull. Acad. Roy. Med. Belg.,24, 451, 1959.18. Witschi, E., unpublished, Embryology of the Rat, State Univ. Iowa, Iowa City, 1961.19. Aoto, T., J. Fac. Sci. Hokkaido Univ., 13, 364, 1957.20. Herrmann, J., Nicholas, J.S., and Vosgian, M.E., Proc. Soc. Exptl. Biol. Med., 72, 454, 1949.21. Johnson, M.L., Time and order of appearance of ossification centers in albino mouse, Am. J. Anat.,52, 241, 1933.22. Schour, I. and Massler, M., The teeth, in The Rat in Laboratory Investigation, Griffith, J.Q. and Farris,E.J., Eds., J.B. Lippincott, Philadelphia, 1942.23. Adelmann, H.B., The development of the neural folds and cranial ganglia of the rat, J. Comp. Neurol.,39, 19, 1925.24. Sissman, N.J., Developmental landmarks in cardiac morphogenesis: comparative chronology, Am. J.Cardiol., 25, 141, 1970.25. Rogers, W.M., Development of pharynx and pharyngeal derivatives in white rat (Mus norvegicusalbinus), Am. J. Anat., 44, 283, 1929.26. Hoar, R.M., Comparative developmental aspects of selected organ systems. II. Gastrointestinal andurogenital systems, Environ. Health Perspect., 1, 61, 1976.27. Edwards, J.A., The external development of the rabbit and rat embryo, in Advances in Teratology,Vol. 3, Woollam, D.H.M., Ed., Academic Press, New York, 1968, p. 239.28. Nelsen, O.E., Comparative Embryology of the Vertebrates, Blakiston, New York, 1953.29. Davies, J. and Hesseldahyl, H., Comparative embryology of mammalian blastocysts, in Biology ofthe Blastocyst, Blandau, R.J., Ed., University of Chicago Press, Chicago, 1971, p. 27.30. Monie, I.W., Comparative development of rat, chick and human embryos, in Teratologic WorkshopManual (supplement), Pharmaceutical Manufacturers Associations, Berkeley, 1965, p. 146.31. Nicholas, J.S., Experimental methods and rat embryos, in The Rat in Laboratory Investigation, Farris,E.J. and Griffith, J.Q., Eds., Hafner Publishing, New York, 1962, p. 51.32. Witschi, E, Development of the rat, in Growth Including Reproduction and Morphological Development,Altman, P. and Dittmer, D.S., Eds., Federation of American Societies for Experimental Biology,Washington, D.C., 1962, p. 304.33. Hendrickx, A.G. and Sawyer, R.H., Embryology of the rhesus monkey, The Rhesus Monkey, Volume II:Management, Reproduction and Pathology, Bourne, G.H., Ed., Academic Press, New York, 1975, p. 141.34. Theiler, K., The House Mouse. Development and Normal Stages from Fertilization to 4 Weeks of Age,Springer, Berlin, Heidelberg, New York, 1972.35. Heuser, C.H., Rock, J., and Hertig, A.T., Two human embryos showing early stages of the definitiveyolk sac, Contribs. Embryol. Carnegie Inst. Wash., 31, 85, 1945.© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 19336. Heuser, C.H., A presomite human embryo with a definite chorda canal, Contribs. Embryol. CarnegieInst. Wash., 23, 251, 1932.37. Ludwig, E., Uber einen operativ gewonnenen menschlichen Embryo mit einem Ursegmente (EmbryoDal), Morphol. Jahrb, 59, 41, 1928.38. West, C.M., A human embryo of twenty-five somites, J. Anat., 71, 169, 1937.39. Streeter, G.L., Developmental horizons in human embryos. Description of age group XI, 13 to 20somites, and age group XII, 21 to 29 somites, Contribs. Embryol. Carnegie Inst. Wash., 30, 211, 1942.40. Gruneberg, H., The development of some external features in mouse embryos, J. Hered., 34, 88, 1943.41. Tuchmann-Duplessis, H., David, G., and Haegel, P., Illustrated Human Embryology: Embryogenesis,Vol. 1, Hurley, L.S. Translators, Masson & Co., Paris, 1972.42. Graves, A.P., Development of the golden hamster Cricetus auratus waterhouse, during the first ninedays, Amer. J. Anat., 77, 219, 1945.43. Boyer, C.C., Development of the Golden Hamster Cricetus Auratus with Special Reference to theMajor Circulatory Channels, doctoral dissertation, Duke University, Durham, NC, 1948.44. Boyer, C.C., Chronology of the development for the golden hamster, J. Morph., 92, 1, 1953.45. Tanimura, T., Nelson, T., Hollingsworth, R.R., and Shepard, T.H., Weight standards for organs fromearly human fetuses, Anat. Rec. 171, 227, 1971.46. Heuser, C.H. and Streeter, G.L., Development of the macaque embryo, Contrib. Embryol., 29, 15,1941.47. Iffy, L., Shepard, T.H., Jakobovits, A., Lemire, R.J., and Kerner, P., The rate of growth in youngembryos of Streeter’s horizons XIII to XXIII, Acta Anat. (Basel), 66, 178, 1967.48. Shenefelt, R.E., Morphogenesis of malformations in hamsters caused by retinoic acid: relation to doseand stage of treatment, Teratology, 5, 103, 1972.49. Lams, H., Etude de l’oeuf de cobaye aux premiers stades de l’embryogenese, Arch. de Biol., T. 28,229, 1913.50. Squire, R.R., The living egg and early stages of its development in the guinea pig, Carnegie InstitutionContributions to Embryology, 137, 225, 1932.51. Selenka, E., Studien uber Entwicklungsgeschichte der Thiere, Bd. 3, Wiesbaden, 1884.52. Maclaren, N., Development of Cavia: implantation., Trans. Roy. Soc. Edinburgh, 55, 115, 1926.53. Sansom, G.S. and Hill, J.P., Observations on the structure and mode of implantation of the blastocystof Cavia, Trans. Zool. Soc. London, 21, 295, 1931.54. Hall, E.K., Further experiments on the intrinsic contractility of the embryonic rat heart, Anat. Record,118, 175, 1954.55. Huber, G.C., On the anlage and morphogenesis of the chorda dorsalis in mammalia, in particular theguinea pig (Cavia cobaya), Anat. Rec., 14, 217, 1918.56. Scott, J.P., The embryology of the guinea pig. I. Table of normal development, Am. J. Anat., 60, 397, 1937.57. Waterman, A.J., Studies of normal development of the New Zealand White strain of rabbit, Am. J.Anat., 72, 473, 1943.58. Hensen, V., Beobachtungen uber die Befruchtung und Entwicklung des Kaninchens und Meerschweinchens,Zeitsch. f. Anat. u. Entwick., Bd.1, S.213, 353, 1876.59. Shepard, T.H., Growth and Development of the Human Embryo and Fetus. Endocrine and GeneticDiseases of Childhood, Gardner, L.I., Ed., W.B. Saunders, Philadelphia, 1969.60. Gribnau, A.A.M. and Geijsberts, L.G.M., Developmental stages in the Rhesus monkey (Macacamulatto), Adv. Anat. Embryol. Cell Biol., 68, 1, 1981.61. Menkin, M.F. and Rock, J., In vitro fertilization and cleavage of human ovarian eggs, Am. J. Obstet.Gynecol, 55, 440, 1948.62. Shettles, L.B., Observations on human follicular and tubal ova, Am. J. Obstet. Gynecol., 66, 235, 1953.63. Hertig, A.T., Adams, E.C., and Mulligan, W.J., On the preimplantation stages of the human ovum: adescription of four normal and four abnormal specimens ranging from the second to the fifth day ofdevelopment, Contribs. Embryol. Carnegie Inst. Wash., 35, 199, 1954.64. O’Rahilly, R., Early human development and the chief sources of information on staged humanembryos, Eur. J. Obstet Gynecol. Reprod. Biol., 9/4, 273, 1979.65. Hertig, A.T. and Rock, J., Two human ova of the pre-villous stage, having a developmental age ofabout seven and nine days respectively, Contribs. Embryol. Carnegie Inst. Wash., 31, 65, 1945.© 2006 by Taylor & Francis Group, LLC

194 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION66. Keibel, F., in Manual of Human Embryology, Vol. 1, Keibel, F. and Mall, F.P., Eds., Lippincott,Philadelphia, 1910, p. 59.67. Heuser, C.H. and Corner, G.W., Developmental horizons in human embryos. Description of age groupX, 4 to 12 somites, Contribs. Embryol. Carnegie Inst. Wash., 36, 29, 1957.68. Politzer, G. and Hann, F., Z. Uber die Entwicklung der branchiogenen Organe beim Menschen, Anat.Entwicklungsgeschichte, 104, 670, 1935.69. Sensenig, E.C., The development of the occipital and cervical segments and their associated structuresin human embryos, Contribs. Embryol. Carnegie Inst. Wash., 36, 141, 1957.70. Sternberg, H., Z., Beschreibung eines menschlichen Embryos mit vier Ursegmentpaaren, nebstBemerkungen uber die Anlage und fruheste Entwicklung einiger Organe beim Menschen, Anat.Entwicklungsgeschichte, 82, 747, 1927.71. Lemire, R.J., Loeser, J.D., Leech, R.W., and Alvord, E.W., Eds., Normal and Abnormal Developmentof the Human Nervous System, Harper and Row, Hagerstown, MD, 1975.72. Hertig, A.T. and Rock, J., Two human ova of the pre-villous stage, having an ovulation age of abouteleven and twelve days respectively, Contribs. Embryol. Carnegie Inst. Wash., 29, 127, 1941.73. O’Rahilly, R. and Muller, F., Human Embryology & Teratology, Wiley-Liss, New York, 1992.74. Bartelmez, G.W. and Evans, H.M., Development of the human embryo during the period of somiteformation including embryos with 2 to 16 pairs of somites, Contribs. Embryol. Carnegie Inst. Wash.,17, 1, 1926.75. Hochstetter, F., Beitrage zur Entwicklungsgeschichte des menschilichen Gehirns, F. Deuticke, Leipzig,1929.76. Streeter, G.L., Developmental horizons in human embryos. Description of age group XIII, embryosabout 4 or 5 millimeters long, and age group XIV, period of indentation of the lens vesicle, Contribs.Embryol. Carnegie Inst. Wash., 31, 27, 1945.77. Streeter, G.L., Developmental horizons in human embryos. Description of age groups XV, XVI, XVII,and XVIII, being the third issue of a survey of the Carnegie collection, Contribs. Embryol. CarnegieInst. Wash., 32, 133, 1948.78. Atwell, W., A human embryo with seventeen pairs of somites, Contribs. Embryol. Carnegie Inst.Wash., 21, 1, 1930.79. Corner, G.W., A well-preserved human embryo of 10 somites, Contribs. Embryol. Carnegie Inst.Wash., 20, 81, 1929.80. Ludwig, E., Embryon humain avec dix paires de somites mesoblastiques, C. R. Ass. Anat., 24, 580,1929.81. Payne, F., General description of a 7-somite human embryo, Contribs. Embryol. Carnegie Inst. Wash.,16, 115, 1925.82. Bartelmez, G.W., The subdivisions of the neural folds in man, J. Comp. Neurol., 26, 79, 1923.83. Hamilton, Boyd, & Mossman’s Human Embryology, 4th ed., Hamilton, W.J. and Mossman, H.W.,Eds., The Williams & Wilkins Co., Baltimore, 1972.84. Streeter, G.L., Developmental horizons in human embryos. Age groups XI-XXIII, Contribs. Embryol.Carnegie Inst. Wash., Embryol. reprint vol. 2, 1951.85. Witschi, E., Migration of the germ cells of human embryos from the yolk sac to the primitive gonadalfolds, Contribs. Embryol. Carnegie Inst. Wash., 32, 67, 1948.86. Hamburger, V. and Hamilton, H.L., A series of normal stages in the development of the chick embryo,J. Morphol., 88, 49, 1951.87. Patten, B.M., Early Embryology of the Chick, McGraw-Hill, New York, 1951.88. Hamilton, H.L., Lillie’s Development of the Chick. H. Holt, New York, 1952.89. Hebel, R. and Stromberg, M.W., Anatomy and Embryology of the Laboratory Rat, BioMed Verlag,Fed. Rep. Germany, 1986.90. Wilson, J., Embryology of the Human Heart, Ward’s Natural Science Est., Inc., Rochester, New York, 1945.91. Heuser, C.H., A human embryo with 14 pairs of somites, Contribs. Embryol. Carnegie Inst. Wash.,22, 135, 1930.92. Davis, C.L., Description of a human embryo having twenty paired somites, Contribs. Embryol.Carnegie Inst. Wash., 15, 1, 1923.93. Congdon, E.D., Transformation of the aortic-arch system during the development of the humanembryo, Contribs. Embryol. Carnegie Inst. Wash., 14, 47, 1922.© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 19594. McClure, C.F.W. and Butler, E.G., The development of the vena cava inferior in man, Am. J. Anat.,35, 331, 1925.95. Kramer, T.C., The partitioning of the truncus and conus and the formation of the membranous portionof the interventricular septum in the human heart, Am. J. Anat., 71, 342, 1942.96. Arey, L.B., Developmental Anatomy, 5th ed, W.B. Saunders, Philadelphia, 1946.97. Bloom, W. and Bartelmez, G.W., Hematopoiesis in young human embryos, Am. J. Anat., 67, 21, 1940.98. Gilmour, J.R., Normal haemopoiesis in intra-uterine and neonatal life, J. Pathol. Bacteriol., 52, 25,1941.99. Pitt, J.A. and Carney, E.W., Development of a morphologically-based scoring system for postimplantationNew Zealand White rabbit embryos, Teratology, 59, 88, 1999.100. Jageroos, B.H., On early development of vascular system; development of blood and blood-vesselsin chorion of man, Acta Soc. Med. Duodecim, B, 19, 1, 1934.101. Kampmeier, O.F., Hemopoietic foci in wall of thoracic duct, and cellular constituents of its lymphstream in human fetus, Am. J. Anat., 42, 181, 1928.102. Kling, C.A., Studien uber die Entwicklung der Lymphdrusen beim Menschen, Arch. Mikroskop. Anat.u. Entwicklungsmech., 63, 575, 1904.103. Sabin, F.R., in Manual of Human Embryology, Vol. 2, Edibel, F. and Mall, F.P., Eds., J.B. Lippincott,Philadelphia, 1912, p. 709.104. Davis, C.L., Development of human heart from its first appearance to stage found in embryos of 20paired somites, Contribs. Embryol. Carnegie Inst. Wash., 19, 245, 1927.105. Ebert, J., An analysis of the synthesis and distribution of the contractile protein, myosin, in thedevelopment of the heart, Proc. Natl. Acad. Sci. U. S., 39, 333, 1953.106. Goss, C.M., The physiology of the mammalian heart before circulation, Am. J. Physiol., 13, 146, 1942.107. Marcell, M.P. and Exchaquet, J.P., L’electrocardiogramme du foetus humain avec un cas de doublerhythme auriculaire verifie, Arch. maladies coeur et vasseaux, 31, 504, 1938.108. Odgers, P.N.B., The development of the pars membranacea septi in the human heart, J. Anat., 72,247, 1938.109. Odgers, P.N.B., The development of the atrio-ventricular valves in man, J. Anat., 73, 643, 1939.110. Robb, J.S., Kaylor, C.T., and Trumers, W.G., A study of specialized heart tissue at various stages ofdevelopment of the human fetal heart, Am. J. Med., 5, 324, 1948.111. Walls, E.W., The development of the specialized conducting tissue of the human heart, J. Anat., 81,93, 1947.112. Hall, K., The structure and development of the urethral sinus in the rat, J. Anat., 70, 413, 1936.113. Ludwig, E., Uber die Anlage der Nach-niere beider Maus, Acta Anat., 4, 193, 1947.114. Hoar, R.M., Comparative female reproductive tract development and morphology, Environ. HealthPerspect., 24, 1, 1978.115. Ingalls, N.W., A human embryo at the beginning of segmentation, with special reference to the vascularsystem, Contribs. Embryol. Carnegie Inst. Wash., 11, 61, 1920.116. Patten, B.M., Human Embryology, 2nd Ed., Blakiston, New York, 1953.117. Hamilton, W.J., Boyd, J.D., and Mossman, H.W., Human Embryology, W. Heffer, Cambridge, 1952.118. Sulik, K.K. and Sadler, T.W., Postulated mechanisms underlying the development of neural tubedefects. Insights from in vitro and in vivo studies, Ann N Y Acad. Sci., 678, 8, 1993.119. Bremer, J.L., Description of a 4 mm. human embryo, Am. J. Anat., 5, 456, 1906.120. Simkins, C.S., Textbook of Human Embryology, F. A. Davis, Philadelphia, 1931.121. His, W., Anatomie menschlicher Embryonen. Atlas, Vogel, Leipzig, 1880-1885.122. Nelson, W., et al., Congenital absence of gall bladder, Surgery, 25, 916, 1949.123. Streeter, G.L., Developmental horizons in human embryos (fourth issue): a review of the histogenesisof cartilage and bone, Contribs. Embryol. Carnegie Inst. Wash., 33, 149, 1949.124. Odgers, P.N.B., Some observations on the development of the ventral pancreas in man, J. Anat., 65,1, 1930.125. Ahrens, H., Die Entwicklung der menshlichen zähne, Anat. Hefte, 48, 167, 1913.126. Chase, S.W., unpublished, Western Reserve Univ., Cleveland, OH, 1955.127. Churchill, H.R., Histology and Histogenesis of the Human Teeth, J.B. Lippincott, Philadelphia, 1935.128. Orban, B., Oral Histology and Embryology, C.V. Mosby, St. Louis, 1953.129. Harland, M., Early histogenesis of thymus in white rat, Anat. Record, 77, 247, 1940.© 2006 by Taylor & Francis Group, LLC

196 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION130. Jost, A., The role of fetal hormones in prenatal development, Harvey Lectures, 55, 201, 1961.131. Price, D., Normal development of prostate and seminal vesicles of rat with study of experimentalpostnasal modifications, Am. J. Anat., 60, 79, 1936.132. Rogers, W.M., Fate of ultimobranchial body in white rat (Mus norvegicus albinus), Am. J. Anat., 38,349, 1927.133. Sehe, C.T., Steriod hormones at early developmental stages of vertebrates, Ph.D. Thesis, State Univ.Iowa, Iowa City, 1957, p. 1.134. Witschi, E. and Dale, E., Gen. Comp. Endocrinol, Suppl. 1, 356, 1962.135. Girgis, A., Description of a human embryo of twenty-two paired somites, J. Anat., 60, 382, 1926.136. Weller, G.L., Development of the thyroid, parathyroid and thymus glands in man, Contribs. Embryol.Carnegie Inst. Wash., 24, 93, 1933.137. Frazer, J.E., A Manual of Embryology, Bailliere, Tindall, and Cox, London, 1940.138. Atwell, W., The development of the hypophysis cerebri in man, with special reference to the parstuberalis, Am. J. Anat., 37, 159, 1926.139. Bender, K., Uber die entwicklung der lungen, Z. f. Anat. Entwicklungsgeschichte, 75, 639, 1925.140. Heiss, R., Zur Entwicklung und Anatomie der menschlichen Lunge, Arch. Anat. Physiol. Anat. Abt.,1919, p. 1.141. Keibel, F. and Elze, C., Normentafeln zur Entwicklungsgeschichte des Menschen, G. Fischer, Jena,Germany, 1908.142. Puiggros-Sala, J.Z., Uber die Entwicklung der Lungenanlage des Menschen, Anat. Entwicklungsgeschichte,106, 209, 1937.143. Weed, L.W., The development of the cerebro-spinal spaces in pig and in man, Carnegie Contrib.Embryol., 5, 3, 1917.144. Bartelmez, G.W. and Dekaban, A.S., The early development of the human brain, Contribs. Embryol.Carnegie Inst. Wash., 35, 12, 1962.145. Hines, M., Studies in the growth and differentiation of the telencephalon in man. The fissura hippocampi,J. Comp. Neurol., 34, 73, 1922.146. Anson, B.J., The early development of the membranous labyrinth in mammalian embryos, with specialreference to the endolymphatic duct and the utriculo-endolymphatic duct, Anat. Record, 59, 15, 1934.147. Fitzgerald, J.E. and Windle, W.F., Some observations on early human fetal movements, J. Comp.Neurol., 76, 159, 1941.148. Hooker, D., Fetal reflexes and instinctual processes, Psychosomat. Med., 4, 199, 1942.149. Bach, L. and Seefelder, R., Atlas zur Entwicklungsgeschichte des menschlichen Auges, W. Englemann,Leipzig, 1911–1912.150. Streeter, G.L., On the development of the membranous labyrinth and the acoustic and facial nervesin the human embryo, Am. J. Anat., 6, 139, 1906.151. Scammon, R.E. and Armstrong, E.L., On the growth of the human eyeball and optic nerve, J. Comp.Neurol., 38, 165, 1924 or 1925.152. Gilbert, P., The origin and development of the head cavities in the human embryo, J. Morphol., 90,149, 1952.153. Keibel, F., in Manual of Human Embryology, Vol. 2, Keibel, F. and Mall, F.P., Eds., J.B. Lippincott,Philadelphia, 1912, p. 218.154. Mann, I., The Development of the Human Eye, Cambridge University Press, England, 1928.155. Bast, T.H. and Anson, B.J., The Temporal Bone and the Ear, C.C. Thomas, Springfield, IL, 1949.156. Bast, T.H., Anson, B.J., and Gardner, W.D., The developmental course of the human auditory vesicle,Anat. Record, 99, 60, 1947.157. Gilbert, M.S., The early development of the human diencephalon, J. Comp. Neurol., 62, 81, 1935.158. His, W., Die Entwicklung des menschlichen Gehirns wahrend der ersten Monate, S. Hirzel, Leipzig,1904.159. Schaeffer, J.P., The lateral wall of the cavum nasi in man with special reference to the variousdevelopmental stages, J. Morphol., 21, 613, 1910.160. Richany, S.F., Anson, B.J., and Bast, T.H., The development and adult structure of the malleus, incus,and stapes, Quart. Bull. Northwestern Univ. Med. School, 28, 29, 1954.161. Richany, S.F., Anson, B.J., and Bast, T.H., The development and adult structure of the malleus, incus,and stapes, Ann. Otol. Rhinol. Laryngol., 63, 399, 1954.© 2006 by Taylor & Francis Group, LLC

COMPARATIVE FEATURES OF VERTEBRATE EMBRYOLOGY 197162. Sensenig, E., The early development of the human vertebral column, Carnegie Contrib. Embryol.,33, 21, 1949.163. Noback, C. and Robertson, G., Sequences of appearance of ossification centers in the human skeletonduring the first five prenatal months, Am. J. Anat., 89, 1, 1951.164. Patten, B., Human Embryology, Blakiston, Philadelphia, 1946.165. Hogg, I., Sensory nerves and associated structures in the skin of human fetuses of 8 to 14 weeks ofmenstrual age correlated with functional capability, J. Comp. Neurology., 75, 371, 1941.166. Johnson, F.P., A human embryo of twenty-four pairs of somites, Contribs. Embryol. Carnegie Inst.Wash., 6, 125, 1917.167. Arey, L.B., Developmental Anatomy, 7th ed., W.B. Saunders Company, Philadelphia, 1965.168. Mintz, B., Embryological development of primordial germ cells in the mouse: influence of newmutation WJ 1 , J. Embryol. Exptl. Morphol., 5, 396, 1957.169. Mintz, B., Continuity of the female germ line from embryo to adult, Arch. Anat. Microscop. (Paris),48, 155, 1959.170. Raynaud, A., Recherches embryologiques et histologiques sur la différenciation sexuelle normale dela souris, Bull. Biol. Fr. Belg., 29, 1, 1942.171. Thomson, J.D., Proc. Iowa Acad. Sci., 49, 475, 1942.172. Gillman, J., The development of the gonads in man, with a consideration of the role of fetal endocrinesand the histogenesis of ovarian tumors, Carnegie Contrib. Embryol., 32, 81, 1948.173. Ortiz, E., The embryological development of the Wolffian and Mullerian ducts and the accessoryreproductive organs of the golden hamster (Cricetus auratus), Anat. Rec., 92, 371, 1945.174. Hamilton, W.J., Boyd, J.D., and Mossman, H.W., Human Embryology, 2nd ed., Williams and Wilkins,Baltimore, 1952, p. 87.175. Witschi, E., in The Ovary, Grady, H.G. and Smith, D.E., Eds., Williams and Wilkins, Baltimore, 1962.176. Holson, J.F., Relative Transport Roles of Chorioallantoic and Yolk Sac Placentae on the 12th and 13thDays of Gestation, Doctoral dissertation, University of Cincinnati, 1973.177. Lewis, W.H. and Hartman, C.G., Early cleavage stages of the egg of the monkey (macacus rhesus),Carnegie Contrib. Embryol., 24,187, 1933.178. Wislocki, G.B. and Streeter, G.L., On the placentation of the macaque (Macaca mulatta) from thetime of implantation until the formation of the definitive placenta, Carnegie Contrib. Embryol., 27,1, 1938.179. Solomon, H.W., Wier, P.J., Fish, C.J., Hart, T.K., Johnson, C.M., Prosobiec, L.M., Gowan, C.C.,Maleef, B.E., and Kerns, W.D., Spontaneous and induced alterations in the cardiac membranousventricular septum of fetal, weaning, and adult rats, Teratology, 55, 185–194, 1997.180. Fleeman, T.L., Cappon, G.D., and Hurtt, M.E., Postnatal closure of membranous ventricular septaldefects in Sprague-Dawley rat pups after maternal exposure with trimethadione, Birth DefectsResearch: Part B: Developmental and Reproductive Toxicology, 71, 185–190, 2004.© 2006 by Taylor & Francis Group, LLC

PART IIHazard and Risk Assessment andRegulatory Guidance© 2006 by Taylor & Francis Group, LLC

CHAPTER 7Developmental Toxicity Testing — MethodologyRochelle W. Tyl and Melissa C. MarrCONTENTSI. Introduction and Objective ................................................................................................207II. Materials and Methods.......................................................................................................208A. Test Substance ...........................................................................................................2081. Chemical Safety and Handling............................................................................2092. Dose Formulation and Analyses..........................................................................209B. Test Animals ..............................................................................................................2101. Species and Supplier ...........................................................................................2102. Live Animals and Species Justification...............................................................2113. Total Number, Age, and Weight..........................................................................211C. Animal Husbandry.....................................................................................................2131. Housing, Food, and Water...................................................................................2132. Environmental Conditions ...................................................................................2143. Identification ........................................................................................................2154. Limitation of Discomfort.....................................................................................2155. Mating..................................................................................................................216III. Experimental Design..........................................................................................................219A. Study Design..............................................................................................................219B. Dose Selection ...........................................................................................................220C. Allocation and Treatment of Mated Females ...........................................................2201. Allocation.............................................................................................................2202. Treatment of Mated Females...............................................................................221D. Observation of Mated Females .................................................................................2281. Clinical Observations...........................................................................................2282. Maternal Body Weights .......................................................................................2293. Maternal Feed Consumption ...............................................................................229E. Necropsy and Postmortem Examination...................................................................2341. Maternal ...............................................................................................................2342. Fetal......................................................................................................................236F. Statistics .....................................................................................................................252G. Data Collection ..........................................................................................................252IV. Storage of Records.............................................................................................................253201© 2006 by Taylor & Francis Group, LLC

202 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONV. Compliance with Appropriate Governmental and AAALAC International Regulations......254VI. Reports...............................................................................................................................254A. Status Reports............................................................................................................254B. Final Report...............................................................................................................254VII. Personnel............................................................................................................................255VIII. Study Records to Be Maintained.......................................................................................255Acknowledgments..........................................................................................................................256References......................................................................................................................................256Soon after the horror of the thalidomide disaster in the late 1950s and early 1960s, resulting inover 8000 malformed babies in 28 countries, the U.S. Food and Drug Administration (FDA)assumed regulatory responsibilities for requiring specific testing paradigms “for the appraisal ofsafety of new drugs for use during pregnancy and in women of childbearing potential.” 1 A letterwas sent from the Chief of the Drug Review Branch, FDA, to all corporate medical directors, 1establishing what became known as the Guidelines for Reproductive Studies for Safety Evaluationof Drugs for Human Use. These guidelines (Figure 7.1) encompassed three test intervals.1. Phase I (Segment I): Prebreeding and mating exposures for both sexes and exposure to pregnantdams until implantation on gestational day (GD) 6, to provide information on possible effects onbreeding, fertility, preimplantation, and embryonic development to midgestation (GD 13 to 15 inthe absence of maternal exposure), and implantation (Figure 7.1A).2. Phase II (Segment II): Exposures during major organogenesis, to provide information on possibleeffects on in utero survival and morphological growth and development, including teratogenesis(Figure 7.1B).3. Phase III (Segment III): Exposures from the onset of the fetal period through weaning of theoffspring, to provide information on maternal parturition and lactation, on F 1 offspring late intrauterineand postnatal growth and development to reproductive maturity, and production of F2offspring (Figure 7.1C).The procedures for Segment II studies have essentially been followed by the U.S. EnvironmentalProtection Agency (EPA), 2–4 Japan, 5 Canada, 6 Great Britain, 7 and the Organization for EconomicCooperation and Development (OECD). 8Recently, attempts have been made to make testing guidelines for reproductive and developmentaltoxicity studies consistent among major nations. The International Conference on Harmonisation(ICH), representing the FDA, the European Community (EC), and Japan, has promulgatedtesting guidelines for registering pharmaceuticals within the three regions. 9 These guidelines arepresented graphically in Figure 7.2. They are similar to the original FDA guidelines (Figure 7.1)and assess exposure during prebreeding, mating, and gestation, until implantation on GD 6 (Study4.1.1), exposures from implantation to weaning (Study 4.1.2), exposures only during organogenesis(Study 4.1.3), and combined single- and two-study designs. ICH Study 4.1.3 is, in fact, identicalto the original FDA Phase II study, and ICH Study 4.1.2 is similar to the FDA Phase III study,except that exposures start at the beginning of organogenesis rather then at the end. Final testingguidelines from the EPA Office of Prevention, Pesticides and Toxic Substances (OPPTS) 10 andfrom the EPA Toxic Substances Control Act (TSCA) 11 for developmental toxicity studies have alsobeen promulgated recently. The FDA has also recently revised its developmental toxicity testingguideline 12 as has the OECD. 13All previous governmental developmental toxicity testing guidelines specified exposure asbeginning after implantation is complete and continuing until the completion of major organogenesis(closure of the secondary palate). This corresponds to GD 6 through 15 for rodents and GD 6through 18 (FDA and TSCA) or GD 7 through 19 (Federal Insecticide, Fungicide, and RodenticideAct; FIFRA) for rabbits, if the day of impregnation is designated GD 0. The 1994 ICH guideline 9© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 203A. Phase I: Fertility and General Reproductive Performance StudyF0malesQ10-week PBEF0 Qfemales2-week PBEF1MGGD6GD13–15F0 femalesNF0 malesNlittersInformation on: breeding, fertility, nidation, embryonic developmentB. Phase II: Developmental Toxicity StudyGD 0Q MGGD6 GD 15N GD 17–18 (mice) or 20–21 (rats)rodent femalesGD 0Q MGGD6 GD 18–19 N GD 29–30rabbit femalesInformation on: embryotoxicity, fetotoxicity, teratogenicityC. Phase III: Perinatal and Postnatal StudyGLF0 ratfemalesGD 0Q MF1GD 15 pnd 0 pnd 21N F0pnd 0 Wbirth NPWRMNF1 malespnd 0GNF1 femalesand F2 litters(pnd 4)Information on: fetal development, parturition, lactation, peri-, neo-, and postnatal effects,adult offspring structures and functionsQ = QuarantineDirect exposure of adultsM = MatingG = GestationPossible indirect exposure from transplacental L = Lactationand/or translactational transferW = WeanDirect exposure of offspring if test material is N = Necropsyadministered via feed or waterGD = Gestational daypnd = Postnatal dayNo exposurePWR = Postwean retention ofselected F1 offspringFigure 7.1FDA study designs.© 2006 by Taylor & Francis Group, LLC

204 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. Study of Fertility and Early Embryonic Development (4.1.1), Rodent (see Phase I)PBEQGF0 males4 weeksM GD 0 GD 6 GD 15 GD 20QF0 females3 weeksORNNNF0 femalesF0 femalesF0 malesAssess: Maturation of gametes, mating behavior, fertility, preimplantation, implantationB. Study for Effects on Prenatal and Postnatal Development, Including Maternal Function(4.1.2), Rodent (see Phase III)pndQMGD0GD6Gparturitionpnd 0LWselected F1 pupspnd 21NF0 femalesM GNF1 males4LNF1 femalesand F2 littersAssess: Toxicity relative to nonpregnant females, prenatal and postnatal development ofoffspring, growth and development of offspring, functional deficits (behavior, maturation,reproduction)C. Study for Effects on Embryo-Fetal Development (4.1.3), Rodent and Nonrodent (see Phase II)QGD 0MGN F0 females and F1 litters on GD 20GD GD6 15Assess: Toxicity relative to nonpregnant females, embryo/fetal death, altered growth ofoffspring, and structural changes of offspring in uteroFigure 7.2International Conference on Harmonization (ICH) study designs.has retained this duration of exposure only during major organogenesis. In a departure from theprevious guidelines, the recently finalized developmental toxicity testing guidelines by U.S. EPA(OPPTS), 14 FDA, 9 and OECD 8 specify exposure during the entire gestational period (Figure 7.3),from GD 0 through scheduled sacrifice at term or from GD 6 to term (see Section C, Allocationand Treatment of Mated Females, 2b. Duration of Administration, for a discussion of the rationalefor new start and end times of administration). There are other differences as well (see Table 7.1).The Phase II study, the major topic for this chapter, had traditionally been termed a “teratologystudy,” since the initial focus was on structural malformations (terata). It is currently more appropriatelytermed a “developmental toxicity study,” as it evaluates (and the term “developmentaltoxicity” encompasses) a spectrum of possible in utero outcomes for the conceptus, including death,malformations, functional deficits, and developmental delays. 16–19 There has been discussion onwhether this test, as structured, does assay developmental toxicity, 20 or whether it is even necessary© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 205D. Single Study Design (4.2), Rodents (combine 4.1.1 and 4.1.2)F0 malesQPBEF0 femalesQM3 weeksGD 0 G pnd 21NF0 malesGD 20N1/2 F0 femalesand F1 littersL W pnd 21selected F1 pupsN1/2 F0 femalesand F1 littersM GNF1 malesNF1 femalesand F2 littersE. Two-Study Design (4.3), Rodents4.1.1 with 4.1.2: 1/2 F0 females and F1 litters necropsied on GD 201/2 F0 females and F1 litters necropsied on pnd 21(retained selected F1 pups followed through mating and gestation of F2 litters)Figure 7.2 (continued)Q = QuarantinePBE = Prebreed Exposure PeriodM = MatingG = GestationL = LactationW = WeanN = NecropsyGD = Gestational Daypnd = Postnatal DayDirect exposure of adultsfor assessing developmental risk. 21 However, the consensus is that this study protocol, scientificallydesigned and performed, provides useful, critical information for human risk assessment of potentialdevelopmental toxicants. 14,15,22It is important to note that a Phase II Study evaluates only structural growth, development, andsurvival of offspring only during in utero development. The conceptuses are evaluated at term. Theparameters evaluated are in utero demise (resorption, fetal death), fetal body weights, and the sizeand morphology of external, visceral, and skeletal structures. For example, if the organs are theright size, shape, and color and in the correct location, they are judged to be normal. There is noassessment for microscopic integrity or function and no way to assess structural and/or functionaleffects that might have occurred (or become evident) during postnatal life if the fetuses had beenborn. 23,24Because there was (and is growing) concern about postnatal sequelae to in utero structuraland/or functional insult, as well as a recognition that exposure of a developing system may resultin qualitatively or quantitatively different effects than exposure of an adult system, the Phase IIIstudy was designed to investigate postnatal consequences to late in utero exposures (Figure 7.1C). 1In brief, the Phase III study consists of exposure of pregnant rats to the test agent from the end oforganogenesis (GD 15), through histogenesis (during the fetal stage), through parturition (birth),and through lactation until the offspring are weaned (postnatal day [PND] 21). The offspring are“exposed” only via possible transplacental and/or lactational (via the milk) routes. There are usuallythree test material groups and a vehicle control group, with at least 20 litters per group; exposureof the dam is usually by gavage (to minimize disruption of the mother and her litter and to controlthe internal dose). During gestation, the dam is weighed periodically and feed consumption ismeasured. Dams and pups are weighed, sexed, and examined externally, with food consumptionmeasured at birth (PND 0) and repeatedly during the lactation period (e.g., on PND 0, 4, 7, 14,© 2006 by Taylor & Francis Group, LLC

206 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONGQGD 0 GD 6MNGD 17–18 (mice)GD 20–21 (rats)GD 29–30 (rabbits)N fetusesat termORGQGD 0 GD 6MNGD 17–18 (mice)GD 20–21 (rats)GD 29–30 (rabbits)N fetusesat termKey:Q = QuarantineM = MatingG = GestationN = NecropsyGD = Gestational dayDirect exposure of dams/doesPossible indirect exposure of conceptuses viatransplacental transfer of parent compound and/or metabolitesNo exposureFigure 7.3New EPA, OECD, and FDA prenatal developmental toxicology (“Phase II”) study exposure durations.and 21). Litters are culled to eight pups on PND 4. The time of acquisition of developmentallandmarks, such as surface righting reflex, pinna (external ear) detachment, incisor eruption, eyeopening (pups are born blind with eyes shut), auditory startle (pups are born deaf with the externalauditory meatus [ear canal] closed), and midair righting reflex, is recorded. The age at testis descentmay also be recorded, occurring in male rats late in lactation, typically on PND 16 to 20 in theCD ® (SD) rat. If the pups are maintained after weaning, then vaginal patency (opening of thevaginal canal) and/or preputial separation are monitored, along with motor activity (initial exploratorybehavior as well as habituated behavior); learning and memory may also be assessed. Thistest provides information on the last “trimester” of pregnancy, delivery, maternal-pup interactionsand behaviors (such as pup retrieval, nursing, grooming, nest building, etc.), and pup postnatalgrowth and development.At weaning, the dam is sacrificed and the number of uterine implantation scars counted toobtain information on prenatal (postimplantation) loss; pups can be necropsied at weaning or later,with target tissues examined histologically. The pups may also be raised to adulthood and matedto ascertain any effects of early indirect exposure on reproductive competence.This chapter is designed to provide the methodology to perform a Phase II study according tocurrent U.S. governmental testing guidelines and in compliance with Federal Good Laboratory© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 207Table 7.1Differences between old and new developmental toxicity test guidelinesEvent/Parameter Previous Requirements a Current Requirements bMaternal EvaluationsAssignment to dose group Not specified Assignment by a body weightdependent random procedureDefinition of high dose level Should induce some overt maternaltoxicity, but not more than 10%maternal deathsShould induce developmentaland/or maternal toxicity, but notmore than 10% maternal deathsTest substance administration:period of dosingTest substance administration:dose adjustmentNumber of pregnant animals attermination (presumed pregnantanimals assigned to study)Maternal postmortem data: ovariancorpora lutea countsFetal EvaluationsRodents: Assignment of fetuses forevaluationDuring the major period oforganogenesis: days 6–15 inrodents and 6–18 (or 7–19 FIFRA)in rabbitsDosage based upon the bodyweight at the start of testsubstance administration oradjusted periodically by bodyweightRodents: 20 per groupRabbits: 12 per groupData required for all species exceptmice (TSCA only)One-third to one-half of each litterassigned for skeletal evaluation,the remainder for visceralevaluationDose from implantation throughtermination (days 6–20 or 21 inrats, 6–17 or 18 in mice, and 6–29or 30 in rabbits): option to beginon GD 0; ICH retains originaldosing periodDosage adjusted periodicallythroughout the period ofadministration by body weightRodents and rabbits: 20 litters pergroup (females with implantationsites at termination)Data required for all species(including mice)One-half of each litter assigned forskeletal evaluation, the remainderfor visceral evaluationRabbits: Coronal sectioning Not required Required (50% serial sections,50% coronal sections)Ossified and cartilaginous skeletalevaluationOnly ossified specified (alizarin redS stain)Both ossified and cartilaginousskeletal examination required(unspecified method of staining;usually alizarin red S for ossifiedbone and alcian blue forcartilaginous bone and otherstructures), all speciesaRequirement under FIFRA, TSCA, and FDA.bOPPTS (EPA TSCA and FIFRA) Draft Guidelines, Public Draft, February 1996, 10 and U.S. EnvironmentalProtection Agency, OPPTS, Health Effects Test Guidelines, OPPTS 870.3700, Prenatal Developmental ToxicityStudy, Public Draft, U.S. Government Printing Office, Washington, D.C., February, 1996 and U.S. EnvironmentalProtection Agency, Fed. Regist. 62(158), 43832, August 15, 1997, and U.S. Environmental Protection Agency,Fed. Regist., 56(234), 63798, 1991.Practice (GLP) Regulations. 25–27 The rest of this chapter is therefore organized according to theheadings (I through X) for a typical GLP-compliant study protocol, as currently followed in theauthors’ laboratory (Table 7.2). An additional useful reference for techniques and methods for andproblems encountered in the performance of a Phase II study is a small book by Pamela Taylor. 28I. INTRODUCTION AND OBJECTIVEThe protocol introduction should indicate the reason for the study and the objective of assessingdevelopmental toxicity (including teratogenicity) in the test animal species after in utero exposurefrom implantation to term.© 2006 by Taylor & Francis Group, LLC

208 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 7.2Typical protocol contentsIntroduction and ObjectiveMaterials and MethodsTest substance:CharacterizationIdentification (CAS No., lot/batch number, supplier)Chemical safety and handlingDosage formulation and analysesAnimals:Species and supplierJustification for live animals and speciesTotal number, age, and weightQuarantineAnimal Husbandry:Housing, food, and waterEnvironmental conditionsAnimal IdentificationLimitation of DiscomfortMatingExperimental DesignStudy designDose selectionAllocation and treatment of mated femalesObservation of mated femalesMaternal clinical observationsMaternal body weightsMaternal food consumptionNecropsy and postmortem evaluation:MaternalFetalStatisticsStorage of RecordsCompliance with Governmental and AAALAC RegulationsReportsStatus reportsFinal reportPersonnelStudy Records to be MaintainedReferencesATTACHMENT I – Certificate of Analysis (of the specific wwwlot/batch number of test materialATTACHMENT II — Material Safety Data SheetATTACHMENT III — etc.A. Test SubstanceII. MATERIALS AND METHODSThe test material should be characterized as to sponsor designation, chemical name, CAS Registrynumber, molecular formula, molecular weight, supplier, lot (or batch) number, chemical purity,appearance, solubility, and storage conditions. All of these parameters should be supplied by thestudy sponsor or by the performing laboratory. Information on the vehicle selected and the amountof test article required should also be included.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 2091. Chemical Safety and HandlingIf any relevant published toxicity information is available (e.g., from sponsor, RTECS, Toxline,Medline, etc.), it should be accessed and extracted. Any chemical-specific information pertainingto the toxic properties of the test material (i.e., eye irritation, skin irritation, sensitization, anticholinesteraseactivity, etc.) should be detailed in this section. The same is true of any chemical-specifichandling information, such as “hygroscopic,” “light sensitive,” “temperature sensitive” (e.g., storefrozen at 20 ± 5°C or –80 ± 5°C; refrigerated at 5 to 10°C; room temperature). A Certificate ofAnalysis and a Material Safety Data Sheet (MSDS), or Experimental Safety Data Sheet (ESDS),or other formal written safety information should be incorporated into the protocol (e.g., asattachments), and/or read and understood by all participating staff (with appropriate documentation).2. Dose Formulation and AnalysesPrior to the start of the study, representative formulations of the test material, in vehicle atconcentrations encompassing the range of dose levels to be employed in the study, must be assayedfor homogeneity and stability. Samples for homogeneity testing should be obtained from representativelocations (e.g., the top, middle, and bottom of a container of solution or suspension or fromthe left, right, and center of a V-shell diet blender). Stability of formulations should be ascertainedunder storage conditions (e.g., refrigerator or freezer) and at room temperature. The duration ofstorage stability assessments should allow for formulation and dose level verification analyses priorto use and time to reformulate and/or reanalyze, if necessary, before administration to the animals.The duration of room temperature stability assessments depends on the route of administrationselected and should allow for bolus dose administration (gavage) in the animal room (usually 1 to4 h/d maximum), for cutaneous application (usually 6 to 8 h/d), for dose administration in feed orwater (usually 9 d, to allow for 7 d of presentation and a “safety net” if the next formulation is notappropriate for administration), etc. Dose level verification should be performed on all doses foreach formulation if the formulation interval is reasonable (e.g., weekly or every 2 weeks) or onfirst, middle, and last formulation if formulations are frequent (based on stability data). Forgeneration of exposure concentrations of materials such as gases, aerosols, or dusts, uniformity ofconcentration level within the chamber and actual (analytical) concentrations in the breathing zoneof the test animals must also be established prior to the study’s start.To prepare oral or cutaneous doses, the following equation is useful:Concentration (mg/ml) =dose level (mg/kg)dose volume (ml/kg)To prepare feed or water dosage formulations, the dose level may be expressed in ppm,percentage (weight/weight), or a constant level of intake (in milligrams per kilogram per day). Forfeed or water dosing, the actual intake in milligrams per kilogram per day can be calculated basedon the amount of feed or water consumed per interval (converted to grams per day), the animal’saverage body weight over the feeding interval, and the percentage of test material in the diet orwater. For example, for a 0.5% dietary dose to a rat weighing an average of 300 g (0.3 kg) overthe interval, and eating 20 g feed/d, the actual intake is computed as follows:20 g/d × 0.0050.3 kg= 0.10.3= 0.333 g/kg/ d = 333 mg/kg/d© 2006 by Taylor & Francis Group, LLC

210 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONWhen you want to provide a fixed intake in milligrams of the test agent per kilogram of theanimal’s body weight per day (mg/kg/d), the concentration in the diet or water must be adjusted,usually weekly, based on food or water intake and the projected test animal body weight for thenext week (if possible by dose group, or based on historical control data). For example, if thetargeted intake is 500 mg/kg/d, the projected daily feed consumption is 30 g/d, and the projectedbody weight at the midpoint of the next interval is 400 g, the calculation for the dietary concentrationis:Conc (g/g) ==intake (g/kg/d) × body weight (kg)feed consumption (g/d)0.500 × 0.430= 0.0067 g/g equiv. to 0.67% in dietFor oral doses (gavage), since the dosing volume is usually kept constant (in milliliters perkilogram), the concentration will vary by dose level; the dosing volume is usually adjusted basedon each animal’s most recent body weight. For example, if the animal weighed 350 g and thedosing volume was 5 ml/kg, then the dosing volume on that particular dosing day would be:5ml x= =1.75 ml1000 g 350 gIf the animal weighed 375 g at the next weigh day, the dosing volume would be 1.875 ml, etc.For cutaneous application, the volume may vary by dose if the chemical is administered “neat”(undiluted).B. Test Animals1. Species and SupplierMice, rats, and rabbits are the most commonly used species for developmental toxicity studies.Using mice or rats as well as rabbits satisfies the rodent/nonrodent FDA and EPA testing requirements.These three species also have the most extensive historical databases available. For rodents,both inbred and outbred strains are available; many strains are more or less sensitive to specific orgeneral chemical insult during gestation, and this strain diversity in sensitivity is also seen withregard to the target organs that may be affected. Both types of strains are capable of changing overtime as a result of genetic drift, founder effect, selection, and/or new mutations. Each performinglaboratory must have a historical control database for each test species/strain used to submitguideline studies to governmental regulatory agencies. Recent control values from the authors’laboratory are listed in Table 7.3 of this chapter, and additional historical control databases arelisted in the Appendix of this volume.It is imperative that the animals come from a reputable supplier. Most reputable commercialbreeders now have extensive health quality control programs as well as genetic monitoring (toensure the genetic integrity of their animal strains). For a laboratory to maintain a reliable historicaldatabase, it is best if all animals come from the same supplier at the same location. In addition, asdiscussed later in this chapter, the laboratory should maintain a regular program of testing animalsfor common laboratory animal diseases, regardless of the supplier. There is no absolute certaintythat the animals will be disease free as received or that they will remain so in one’s own facility.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 211Table 7.3Summary of historical control maternal endpoints aGestational Days of DosingParameter 6–15 6–19Number of dams 178 136Maternal body weight (GD 0) (g) 247.56 ± 1.27 244.00 ± 1.29Maternal body weight (GD 20 at sacrifice) (g) c 410.24 ± 2.54 377.71 ± 2.03Maternal body weight change (gestation) (g) c 162.72 ± 1.87 133.71 ± 1.60Maternal body weight change (corrected) (g) c 73.85 ± 1.27 52.95 ± 1.32Gravid uterine weight (g) b,c 88.87 ± 1.22 80.76 ± 1.11Maternal liver weight (g) b,c 17.97 ± 0.17 16.33 ± 0.13Relative maternal liver weight (g) a,d 4.42 ± 0.03 4.32 ± 0.03aIncludes all pregnant control dams until terminal sacrifice on GD 20. Reportedas the mean ± S.E.M.; GD = gestational daybWeight change during gestation minus gravid uterine weightcp ≤ 0.05 significant difference between the two groupsdp ≤ 0.001, significant difference between the two groupsSource: Marr, M.C., Myers, C.B., Price, C.J., Tyl, R.W., and Jahnke, G.D., Teratology,59, 413, 1999 (tables provided by the authors).2. Live Animals and Species JustificationThe use of live vertebrate animals must be justified for federal GLPs, and the protocol must be submittedto each organization’s Institutional Animal Care and Use Committee (IACUC). The justification usuallypresented is that the sponsor requested the animals, and that this test with live animals is required bythe applicable governmental testing guidelines, e.g., preclinical testing for new drug development(FDA), for pesticide registration or Data Call-In (FIFRA), for a premanufacturing notice (PMN) of acommercial chemical under TSCA, for a mandated Test Rule or negotiated test agreement (TSCA), orfor a significant new use registration (SNUR) for a commercial chemical (TSCA). It should be statedthat alternative test systems are not available and/or currently accepted by the scientific community forthe assessment of chemical effects on prenatal mammalian growth and development. This is usuallythe section where historical control data available in the performing laboratory are noted for thespecies/strain on test along with any published historical control databases.3. Total Number, Age, and WeightRat and mouse females are sexually mature at approximately 50 days of age. 29 Male rodents donot have sperm in the cauda epididymis until 60 (mice) or 70 (rats) days of age. Most developmentaltoxicity studies use female rodents 8 to 10 weeks of age and males 10 to 12 weeks of age (ifbreeding is to be done in-house). An 8-week-old female CD ® rat weighs approximately 200 g, an8-week-old female CD-1 ® mouse weighs approximately 20 g (based on Charles River Laboratories’growth charts), and female rabbits should be between 4 and 6 months old (2.5 to 5 kg). Doesmature earlier than bucks, which do not attain adult sperm levels until 6 to 7 months of age. 29 Aprime consideration is to use the most reproductively sound age for any animal, but use animalsas young as possible to avoid the costs of purchasing older animals.Numbers are based on guideline requirements. The new guidelines require a minimum of 20pregnant animals per dosage group for rodents and rabbits. The previous guidelines required 12pregnant does per group for rabbits. Inseminated rats and mice typically have at least a 90%pregnancy rate. Therefore, putting 25 sperm- or copulation plug-positive females per dose groupon study should be sufficient, taking into consideration the historical pregnancy rate in the laboratoryfor the species and strain to be used. Ordering 30 to 50% extra female rodents, if doing one-ononein-house mating to obtain 100 sperm-/plug-positive rodents over a 3- to 4-d period, is a goodguideline. A 10% increase over the number of pregnant rabbits desired (i.e., 22 mated/inseminated© 2006 by Taylor & Francis Group, LLC

212 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONdoes per group), to obtain 20 litters at term per group, is also a reasonable guideline, as rabbits dooccasionally spontaneously abort or deliver early without regard to treatment. In the authors’laboratory, pregnancy rates for naturally bred rabbits are typically over 95%.a. Physical ExaminationUpon arrival, while the animals are being uncrated, a well-trained laboratory animal technicianshould inspect each animal for external alterations of the head, trunk, appendages (limbs and tail),and orifices (mouth, anus, genitourinary tract) and for congenital defects, such as microphthalmia.The condition of the coat, eyes, ears, and teeth should also be evaluated. Any abnormal clinicalobservation should be brought to the attention of the investigator. Rabbits are commonly fed arationed amount upon arrival and during quarantine to alleviate the onset of the mucoid enteritis(enteropathy) occasionally brought on by shipping stress. Covance Research Products has suggestedno feed for the first 24 h (water ad libitum) and an increase of 25 g/d, up to a 125-g ration. Forrabbits received timed-pregnant, a half ration is recommended for the first 24 h (65 to 70 g) andthen full ration (120 to 150 g) or ad libitum feeding.b. Pathogen Antibody ScreenRodents and rabbits may be purchased certified pathogen antibody free (with documentation fromthe supplier), but additional quality control evaluations are commonly done by many laboratories.In fact, many testing protocols require in-house health quality control. Each shipment of animalsshould be quarantined on arrival, and quality control evaluation should be initiated within one dayafter receipt. On the day after receipt, five rats per sex should be randomly chosen from the shipmentof animals. They should be sacrificed and their blood collected for assessment of viral antibodystatus. Commercial testing laboratories generally offer rat and mouse viral screening assays. Atypical viral screen for rats, available from BioReliance Corporation (Rockville, MD), consists ofevaluation for the presence of antibodies against the following: Toolan H-1 virus (H-1), Sendaivirus, pneumonia virus of mice (PVM), rat coronavirus/sialoacryoadenitis (RCV/SDA), Kilham ratvirus (KRV), CAR bacillus, Mycoplasma pulmonis (M. Pul.), and parvo. Thus, health status isassured from two independent sources other than the toxicology laboratory. Survival quality control(QC) is possible with rabbits because they may be bled for serum from the central ear artery.c. Flotation and/or Tape Test for Intestinal ParasitesMost endoparasites can be detected by fecal examination. This may be accomplished by directsmear, in which a small amount of feces is mixed with a drop of saline on a slide and then coverslipped. The preparation is then examined for the presence of parasites or parasitic ova. Protozoanparasites will be motile. For fecal flotation, a larger specimen is mixed in a solution of zinc sulfate,sodium nitrate, or supersaturated sugar, with a specific gravity of 1.2 to 1.8. The mixture iscentrifuged or simply allowed to settle. The tube is filled to the top with the flotation solution, anda clean coverslip is placed on top of the tube so that it touches the liquid. This should stand for15 to 30 min. The cover slip is then removed and placed on a microscope slide. The slide shouldthen be examined for the presence of parasite ova and coccidial oocysts. To simplify the process,a commercial kit may be used. The method of cellophane tape may be used by pressing a strip oftape to the animal’s perianal area. The tape is then placed on a microscope slide and examined. 30d. HistopathologyOccasionally, extra animals are ordered and sacrificed upon arrival, and slides of likely target organsfor disease are prepared for histological examination. These may include representative sections© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 213Table 7.4Recommended parameters for caging and environment for common laboratory animalsBody Cage Floor Area Cage Height Relative Dry-Bulb TemperatureAnimal Weight in. 2 cm 2 in. cm Humidity (%)°C °FMouse 15–25 g 12.0 77.42 5 12.70 30–70 18–26 64–79>25 g >15.0 96.78 5 12.70Rat 500 g >70.0 451.64 7 17.78ft 2 m 2Rabbit < 2 kg 1.5 0.14 14 35.56 30–70 16–21 61–722–4 kg 3.0 0.28 14 35.564–5.4 kg 4.0 0.37 14 35.56>5.4 kg >5.0 0.46 14 35.56Source: Data from National Research Council, Guide for the Care and Use of Laboratory Animals, Institute ofLaboratory Animal Resources, Commission on Life Sciences, National Research Council. National AcademyPress, revised 1996, Tables 2.1, 2.2, 2.3, and 2.4.of the liver, spleen, kidneys, gastrointestinal tract (esophagus, stomach, duodenum, jejunum, ileum,cecum, colon, and rectum), lungs, lymph nodes (submaxillary and mesenteric), and reproductiveorgans (testes, epididymides, prostate, seminal vesicles, coagulating gland, vagina, corpus andcervix uteri, oviducts, and ovaries). This adds increased costs to studies and is generally unnecessaryfor the relatively short-term developmental toxicity studies if a reputable vendor is used. (For longerterm studies, such as multigeneration, Phase I or III studies, this procedure is recommended.)C. Animal Husbandry1. Housing, Food, and WaterAnimals should be singly housed (except during quarantine or mating for rodents) so that foodconsumption can be determined (as required by testing guidelines) and so that there will be noconfounding factors from group dynamics. For example, a dominant female rat or mouse will eatmore than others, a dominant female mouse may overgroom (to the point of “barbering”) othercohabited mice, and stress levels of variously ranked females may vary, with consequences unrelatedto chemical exposure.Rodents can be housed in solid-bottom polycarbonate or polyethylene cages with stainless-steelwire lids (e.g., from Laboratory Products, Rochelle Park, NJ), using hardwood or other wellcharacterizedbedding (see below). Alternatively, they can be maintained in stainless-steel, wiremeshcages mounted in steel racks, with Deotized ® paperboard (e.g., from Shepherd SpecialtyPapers, Inc., Kalamazoo, MI) placed under each row of cages to collect solid and liquid excreta(copulation plugs for rats will also be detectable on the paperboard if the animals are individuallyhoused in hanging cages). Rabbits are housed in stainless-steel cages with mesh flooring (e.g., fromHoeltge, Inc., Cincinnati, OH), with a pan lined with paperboard beneath each cage to collectexcreta. The dimensions of the cages as required by the NRC Guide 31 (update of the NIH Guide 32 )are presented in Table 7.4, by species and body weight range.Feed must be certified and analyzed (usually by the supplier) and is usually available ad libitumthroughout the study (after any initial adjustments in food presentation for rabbits; see previoussection on quarantine). For rodents, feed can be pelleted and available on the cage lid or in feeders© 2006 by Taylor & Francis Group, LLC

214 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONwithin the cage, or it can be ground and available in feeders within the cage. For rabbits, pelletedfeed is presented in feeders attached to the cage, and it must be changed frequently. For rodents,a good feed source is Purina (PMI ® Nutrition International) Certified Rat and Mouse Chow, No.5002 (ground or pelleted); for rabbits, we have used Purina Certified Rabbit Chow (e.g., No. 5322or high fiber diet No. 5325, for animals on restricted feed). A high-fiber diet that stimulates hindgutmotility, reduces enteritis, and protects against fur chewing and formation of trichobezoars (hairballs) in the stomach is necessary for rabbits. 33 Additional analyses for possible contaminants and/ornutrient levels can be provided by commercial sources. Feeds should be stored at or below 15°Cto 21°C (60°F to 70°F) and should not be used more than 6 months past the milling date.Drinking water must meet EPA standards for potable water and must be analyzed for contaminants.If the water is taken directly from a municipal water supply, the supplier can provide analysesshowing levels of contaminants to be within acceptable limits for human consumption. If theincoming water is further treated (e.g., acidified, deionized, filtered, deionized/filtered, distilled),then these procedures must be documented and posttreatment analyses recorded. The water, regardlessof treatment, may also be analyzed by an independent testing laboratory (e.g., Balazs Laboratory,Inc., Sunnyvale, CA. The water can be presented in plastic (polypropylene, polycarbonate)or glass bottles with stainless-steel sipper tubes and rubber stoppers, or via an automatic wateringsystem (e.g., from Edstrom Industries, Inc., Waterford, WI).2. Environmental Conditionsa. Light CyclesAccording to the NRC Guide (1996), light levels of 30 foot candles are adequate for most routinecare procedures. Illumination of excessive intensity and duration may cause retinal lesions in albinorats and mice. 32,34 The color balance of light should approach that of sunlight to allow the mostaccurate observations of the conditions of the animals’ eyes and other body parts, for which coloris an important factor. 30 Light cycles of 12 h light/12 h dark seem to be adequate to promotebreeding of rodents. Ovulation in mice generally occurs during the midpoint of the dark cycle.Continuous light will depress cycling; therefore, documented quality control of computer- or timercontrolledroom light timing is essential. Light cycles for rabbits can vary from 8 to 10 h of lightfor males and 14 to 16 h of light for females. An intermediate compromise (i.e., 12:12) is adequatefor breeding. Shortening of the lighted segment of the light cycle may bring on autumnal sexualdepression.b. Temperature and Relative HumidityTemperature and relative humidity are two of the most important factors in an animal’s environmentbecause of their effects on metabolism and behavior. Consequently, they may have an effect on theanimal’s biological reactions to various test agents (Table 7.4). The range of temperatures suggestedin the NRC Guide 31 is slightly lower than each species’ thermoneutral zone, but allows optimalcomfort, reactivity, and adaptability. 35 Hyperthermia is of concern in pregnant animals, but the coretemperature of pregnant Sprague-Dawley (SD) rats is less affected by heat stress than that ofnonpregnant rats. 36 Of more concern is the effect of elevated temperature on male fertility.Typically, sustained temperatures above 85°F (29.4°C) will result in temporary infertility. Therefore,careful monitoring of temperature deviations and immediate response to any equipmentfailure is essential in a facility engaged in reproductive and developmental toxicity studies ifbreeding is done in-house. Relative humidity levels, as suggested by the NRC Guide, 31 are 30%to 70% for rodents and rabbits (Table 7.4). Excessively low humidity in rodent rooms can cause“ring-tail” in neonates.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 215c. VentilationThe suggested rate of ventilation in the NRC Guide 31 (10 to 15 air changes per hour) is designedto help create an odor-free environment. This, along with an adequate frequency of bedding changes,ensures an acceptably low level of ammonia (< 20 ppm) in the microenvironment of the cage.d. NoiseThe NRC Guide’s 31 recommendations for limitation and reduction of noise inherent in the day-todayoperations of an animal facility include separating the animal rooms from the procedures thatinvolve the most noise (such as cage washing and refuse disposal). Noisy animals, such as dogsand primates, should be separated from rodents and rabbits. Continuous exposure to acousticallevels above 85 dB can cause reduced fertility in rodents. 37e. ChemicalsAromatic hydrocarbons from cedar shavings and pine bedding material can induce the biosynthesisof hepatic microsomal enzymes, so softwood shavings should be avoided for developmental toxicitystudies. Excellent hardwood bedding materials include Ab-Sorb-Dri ® hardwood chips (LaboratoryProducts, Inc., Garfield, NJ), Sani-Chip ® cage litter (P.J. Murphy Forest Products Corporation,Montville, NJ), and Alpha-Dri ® purified α-cellulose (Shepherd Specialty Papers, Inc., Kalamazoo,MI). Alpha-Dri may be more expensive but is more absorbent; its use may require fewer cagechanges and therefore save time and money.Bedding can be sterilized and certified, and should be changed as often as is necessary to keepanimals dry and clean. Litter should be emptied from cages in areas other than the animal roomsto minimize exposure to aerosolized waste. Although rodents will successfully mate followingrecent cage changes, it is best to allow the male time to mark “his” cage with urine prior tointroduction of the female for mating; male rabbits generally mate better if their bedding has notbeen changed immediately prior to mating. It is unwise to change a littering rodent’s cage within24 h of delivery; therefore, the last suggested cage change is usually on GD 20 (day of vaginalsperm or copulation plug is GD 0).3. IdentificationRodents may be identified by ear tag, tail tattoo, ear notch, toe notch, or implant. An inexpensiveand highly reliable method, if done properly, is the tail tattoo. This is accomplished with a tattoogun. The tail is cleaned with soap and water or alcohol. The tattoo needle is dipped into the ink,and the individual ID is slowly “written” onto the tail. The animal is typically restrained in a rodentrestrainer for this process. Ear tags are inexpensive, but if not applied properly, they may poseirritation problems. Multiply housed animals may also lose tags through grooming or playing.Some commercial rabbit suppliers ear tag their kits or subcutaneously implant an identificationchip (e.g., AVID microchips, AVID Microchip I.D. Systems, Folsom, LA) at weaning. This allowseasy identification for particular clients, following genetic traits (such as excessive aggression, hairgrowth pattern, and litter size) and tracking individuals for mating (to preclude sibling matings).Rabbits’ ears may also be tattooed, but this requires sedation. Toe clipping (once very common formice) is not currently considered appropriate for humane reasons and for ease in “reading” numbers.Ear notching (usually in association with toe clips) is also not currently done; multiply-housedanimals create new “notches” as they interact, and notches can be difficult to read.4. Limitation of DiscomfortU.S. Department of Agriculture regulations require annual documentation of animals (currentlylarge mammals, including dogs, cats, rabbits, nonhuman primates, etc., but it is anticipated that© 2006 by Taylor & Francis Group, LLC

216 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrats, mice, and birds may soon be included) that are subjected to more than momentary or slightpain or distress without the benefit of anesthetic, analgesic, or tranquilizing drugs. As part of thestudy protocol, it is necessary and appropriate to state that some adult toxicity may be caused byexposure to the high and/or mid doses. It should be anticipated that those oral doses employed (orthe cutaneous doses applied) will not result in irritation or corrosion to the gastrointestinal tract(or the skin) of the test animals. Discomfort or injury to animals should be limited, so if any animalbecomes severely debilitated or moribund, it should be immediately and humanely terminated afterappropriate anesthesia. All necropsies should be performed after terminal anesthesia. If blood is tobe collected, it should be collected after appropriate anesthesia. Animals should not be subjectedto undue or more than slight or momentary pain or distress.5. Matinga. RodentsRats are continually polyestrous, with a 4- to 5-d estrous cycle (estrus lasts approximately 12 h).Breeding considerations for developmental toxicity studies include the number of pregnant femalesthat can be processed in one day at sacrifice, the range of mating days, and the need to limit thenumber of females mated to any given male in a dosage group. Females are mated at between 8and 10 weeks of age. Males may be mated at approximately 12 weeks of age and appear to breedsuccessfully for 6 to 8 months after this. Monogamous (one female with one male) or polygamous(harem) mating may be employed. Disadvantages of the monogamous scheme include the need tohouse a larger population of males. Polygamous mating pairs two to three (rarely as high as six)females with one male. If the male is housed with multiple females and mates with more than oneon the same night, the distinct possibility exists that females impregnated later will have reducedpregnancy rates and/or litter size (due to reductions in the number of ejaculated sperm in latermatings). This can be disadvantageous since the number of females employed per male per dosegroup must be limited to reduce the likelihood that heritable malformations from a male will affectstudy outcome.Females may be checked by means of vaginal lavage with 0.9% saline for determination oftheir estrous cycle stage (i.e., proestrus, estrus, metestrus, or diestrus) using fixed cells and Toluidineblue stain, if necessary. Only the females in estrus or late proestrus would be used for mating.Charles River Laboratories can now provide guaranteed rats with 4-d cycles from their Portage,MI, facility. One caveat is that the stress of shipment may shift the cycle (usually by one day). Thefemales’ estrous stage should be confirmed by the performing laboratory upon arrival. Conversely,females may be paired one-to-one with a male, regardless of estrous stage. Each morning followingpairing, females are checked by vaginal lavage for the presence of sperm. If sperm negative, femalesremain with the male and are checked daily until found sperm positive. A medicine droppercontaining approximately 0.1 ml of saline is inserted in the vaginal opening, and the fluid is expelledinto the vaginal canal and then drawn back into the eye dropper. The contents are expelled onto amicroscope slide and then examined for the presence of sperm using a light microscope (400×magnification). The presence of a copulation plug in the vagina (for mice) or a dropped copulationplug (for rats) is also a positive sign of mating. In solid-cage systems, vaginal lavage or manualchecking for the presence of a copulatory plug is necessary. In a hanging cage system, the presenceof a dropped plug on the shelf below the cage may be used as evidence of mating in monogamouslypaired rats (the plug dries, shrinks, and falls out within 6 to 8 h postcoitus). Once the female isfound sperm- or plug-positive, she is typically weighed and removed to her own cage. This dateis normally designated as GD 0 (rarely, GD 1).One-to-one breeding of rats, without regard to the female’s estrous cycle, will usually generate25% to 30% sperm-positive females per day (i.e., in a 4- to 5-d cycle, theoretically 20% to 25%of the animals are in the appropriate stage for mating on any given day).© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 217Mouse estrous cycles are also 4 to 5 d, with an evening estrous period of 12 h. Females maybe primed prior to mating. 38 Group-caged female mice may enter a phase of anestrus, which isterminated by the odor or presence of a male (from a pheromone excreted in male urine). A malein a wire-mesh cage can be placed in a larger cage containing a group of females 3 d prior to thedesired day of mating (or females can be housed in the presence of bedding taken from a male’scage). This synchronizes the estrous cycle to typically yield a 50% plug-positive rate when thefemales are mated.Female mice are introduced into the male’s cage (not vice versa and not into a clean cage; themale will more likely be successful if he has previously marked his territory) in the late afternoon,and the following morning the females are checked for the presence of copulation plugs in thevagina. Plugs are normally readily apparent, but if not, a pair of closed, rounded-point forceps aregently inserted into the vaginal opening and then opened to expose any plug present.Timed-pregnant rodents are now available from most suppliers. This can be advantageousbecause of space limitations and the high cost of maintaining a male breeding colony, but theanimals are more expensive, and there is no time for a quarantine period prior to GD 0 followingreceipt in the research laboratory. Also, shipping stress can result in implantation failure in miceand rats.b. RabbitsRabbits are induced ovulators, releasing ova 9 to 13 h after copulation; ovulation may also beinduced by an injection of luteinizing hormone (LH). The number of females to be mated eachday is dependent, as with rodents, on the number that can be successfully processed at sacrifice.1) Natural Mating — A receptive doe (characterized by a congested, moist, purple vulva) willraise her hindquarters (flagging) to allow copulation when placed in a buck’s cage. Once copulationhas occurred, the female may be introduced into a second male’s cage to increase the fertility rate.A high success rate is obtained even with randomly mated females (i.e., a group of females chosento breed on a given day regardless of receptivity). If a female shows no interest, another male maybe tried. If the male shows no interest, the female’s hindquarters may be placed over his nose tomake sure he notices her. Some males may attempt to mate in inappropriate positions, but thetechnician can manually place the female in a better position.Occasionally, a female will refuse to mate. This will have to be considered an unsuccessfulmating, and an attempt can be made the next day. Only rarely will a doe fail to mate during thetwo or three days of mating for a study.2) Artificial Insemination — Ovulation is induced by intravenous injection (marginal ear vein)of 20 to 25 IU of human chorionic gonadotropin (hCG). This may be done immediately prior tomating or up to 4 to 5 h prior to mating. Sperm is collected from bucks by use of artificial vaginas(AV), which have been previously described. 39,40 An AV can be assembled easily by use of 1-in.rubber latex tubing, two bored neoprene stoppers, K-Y Jelly, and glass centrifuge tubes (seeFigure 7.4A). The latex can be replaced as it becomes brittle with use and washing. Glass inseminationtubes (see Figure 7.4B) have become fairly expensive to have custom made but will lastfor years with careful handling. The AVs are heated in an incubation oven to approximately 40°Cto 45°C prior to use. Semen is collected by using a tubally ligated female rabbit (or one with acontraceptive implant). The technician holds the AV, the opening of which has been coated withK-Y Jelly, between the teaser’s hind legs. The male will normally mate and ejaculate within 1min. (The teaser should be sterilized in case the male is faster than the technician and “misses”the AV.) Alternatively, a sleeve made from a female rabbit skin is placed on the arm of the technician.The buck mounts the sleeve and ejaculates in the AV held in the technician’s hand. This eliminatesthe need for live, sterilized does.© 2006 by Taylor & Francis Group, LLC

218 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. Artificial Vagina and Collecting Tube25 mm25 mm#5 Neoprene stopper with 18 mmhole bored through the stopper#6 1/2 Neoprene stopper with18 mm hole boredCollecting tubeArtificial vagina6" length of 1" diameter Penrose drain tubing is sufficient for 1 A.V.The two stoppers are held together with an adhesive by 3M- Scotch Grip RubberAdhesive #1300.B. Artificial Insemination TubeFigure 7.4Artificial insemination equipment.The advantage of artificial insemination is that the semen sample can be evaluated for spermcount, viability, and motility before insemination to assure that an optimal sample is introducedinto the female. This can be accomplished manually by using a blood diluting pipette and ahemocytometer. The semen is drawn up to the 0.5 mark on the pipette, followed by 0.9% salineto the 1.01 mark. The sample is gently mixed, and a drop is released into the groove of thehemocytometer. The number of sperm in five 0.2 × 0.2 mm squares is counted at 400× magnification.If a count of greater than 10 per five squares is obtained (i.e., 10 million sperm), then the sampleis acceptable for use. An additional advantage is that up to five females may be inseminated withone sample, thus reducing the number of bucks needed in the breeding colony. Precise recordsmust be kept of the male’s performance, including fetal outcomes; any males producing malformedcontrol kits must be culled from the colony.By use of a 1-cc syringe, a 0.25-ml sample of undiluted semen is drawn into the inseminationtube. The female to be inseminated is placed upside down with her head between a seated technician’slegs. The hind legs are grasped by the technician and spread. A second technician insertsthe glass insemination tube until the pelvic brim is felt and then rotates the tube 180˚ and continuesto insert it up to a depth of 7 to 10 cm. The semen is then injected and the tube withdrawn. If thedoe urinates during insemination, another semen sample is injected 15 to 30 min later.The day of successful natural mating or the day of artificial insemination is usually designatedGD 0. One major consequence of hormonally priming the female (necessary for artificial insemination,less commonly used for natural breeding) is the possibility of “superovulation.” A largenumber of eggs are ovulated, but many are not implanted, resulting in a large percentage ofpreimplantation loss.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 219Table 7.5Summary of developmental endpoints in control littersGestational Days of DosingParameter 6–15 6–19All littersNo. litters examined 178 136No. corpora lutea per dam 16.93 ± 0.23 15.45 ± 0.18No. implantation sites/litter a,b 15.79 ± 0.22 14.69 ± 0.20Percent preimplantation loss/litter a 6.66 ± 1.04 5.47 ± 0.79No. resorptions/litter (percent postimplantation loss/litter) a 2.99 ± 0.41 3.61 ± 0.67No. (percent) litters with resorptions 56 (31.5) 52 (38.2)No. late fetal deaths/litter a 0.0 0.0No. adversely affected implants/litter c,d,e 1.10 ± 0.12 0.67 ± 0.08No. (percent) litters with adversely affected implants d 94 (52.81) 63 (46.32)Live litters e 4.42 ± 0.03 4.32 ± 0.03No. live fetuses/litter b,c 15.30 ± 0.22 14.21 ± 0.21Average fetal body weight (g)/litter (sexes combined) a,e 3.671 ± 0.026 3.566 ± 0.024aReported as the mean ± S.E.M.bp ≤ 0.05, significant difference between the two groupscIncludes all dams pregnant at terminal sacrifice on GD 20; litter size = No. implantation sitesper damdAdversely affected = nonlive plus malformedep ≤ 0.01; significant difference between the two groupsfIncludes only dams with live fetuses; litter size = no. live fetuses per damSource: Marr, M.C., Myers, C.B., Price, C.J., Tyl, R.W., and Jahnke, G.D., Teratology, 59, 413,1999 (tables provided by the authors).No. corpora lutea – No. implantation sites× 100No. corpora luteaFor example, the value for percent preimplantation loss in the authors’ laboratory for artificiallyinseminated does with hormonal priming is 30.74 ± 2.90% (88 does) versus the range of valuesfor naturally bred does (not primed) of 5.82%–17.48% (224 does) (Table 7.6).3) Vendor Supplied Timed-Pregnant Rabbits — Vendors (e.g., Covance Research Products,Denver, PA) are now offering timed-mated rabbits and have done extensive studies with numerouslaboratories that show the fertility rate based on gestational day of shipment. Unless the toxicologyfacility doing the study is in close proximity to the vendor, it is not possible to measure foodconsumption from GD 0. The advantage of purchasing timed-mated rabbits is eliminating theupkeep of a large number of bucks and the technical time expended for mating procedures andrecordkeeping, but there is no on-site quarantine period prior to GD 0 with these animals. In theauthors’ laboratory, we waive the prebreed quarantine. We monitor the purchased timed-mated doesfrom GD 1, 2, or 3 (depending on when they arrive at our facility) until the start of dosing on GD6 for body weight, weight changes, feed consumption, clinical signs, etc., so this time serves as apredosing “quarantine” period in lieu of a premating quarantine.A. Study DesignThis section should document the following:III. EXPERIMENTAL DESIGN© 2006 by Taylor & Francis Group, LLC

220 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION• The number of mated females per group• The number of days (and gestational days) of treatment• The dose levels in mg/kg/day (as assigned), mg/ml (as formulated), ppm, dietary percentage, andin mg/m 3 , etc.• The dosing volume in ml/kg, if appropriate• Projected in-life study performance dates should also be documented as follows:– Animals arrive at performing laboratory– Dates of GD 0 (for rodents, the end date is tentative since it will depend on performance ofthe breeding pairs)– Inclusive dates of treatment (the end date will depend on GD 0 dates)– Dates of scheduled sacrifice (the end date will depend on GD 0 dates)The date of anticipated submission of the draft report may also be included in this section (or underthe section on reports).B. Dose SelectionAll current governmental testing guidelines call for at least three dose or concentration levels plusa concurrent vehicle control (four groups total). The top dose level should be chosen to result indemonstrable maternal (and possibly developmental) toxicity. Maternal indicators can includereductions in body weight or weight gain, treatment-related clinical signs of toxicity, sustainedreductions in food and/or water consumption, and maternal mortality (not to exceed 10%). 41,42 Themiddle dose should result in minimal maternal (and possible developmental) toxicity. The low doseshould be a NOAEL (No Observable Adverse Effect Level) for both maternal and developmentaltoxicity. Optimally, the selection of doses should be based on results of range-finding studies inpregnant animals. A distant second choice on which to base doses would be a 14-day repeated dosestudy in the same species and strain by the same route in the same sex (although the animals werenot pregnant, and pregnancy changes many physiologic and toxicologic parameters). Informationon absorption, distribution, metabolism, and excretion in the test species (and sex) would be veryuseful, but it is rarely available for commercial chemicals and pesticides (most likely available fornew drug preclinical work) and is almost never available for pregnant females. The spacing of thedoses may be arithmetic or logarithmic; the low dose should be selected, if possible, to allowapplication of safety factors during risk assessment and still be above any expected human exposurelevels. The doses selected and their justification should be documented in this section.C. Allocation and Treatment of Mated Females1. AllocationThe number of females assigned to dose groups is defined by the protocol. Animals can be assignedto dose groups totally randomly (by computer, by a table of random numbers, by numbered cardsdealt randomly, etc.), or they can be assigned on the basis of body weight (i.e., stratified by bodyweight but randomly within body weight classes).For mated females to be assigned to study by stratified randomization, GD 0 weights arearranged in ascending order, from the lightest to the heaviest on the first GD 0 date, and assignedin the same order or reverse order (i.e., descending order from heaviest to lightest) on subsequentdays. Beginning with the lightest weight, animals are assigned to groups stratified by body weight(number of animals per group equals number of treatment groups). Within each stratified group,one animal will be randomly assigned to each treatment group. In the event that the total numberof animals inseminated on a given day is not an even multiple of the number of treatment groups,the stratified group with the heaviest body weights (the last group filled by stratified randomization)will be assigned randomly (one animal per treatment group) until all animals have been assigned.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 221When consecutive breeding days are required, then female assignment to dose groups on eachsubsequent GD 0 will place females in sequence beginning with the dose group following the lastdose group assigned on the previous day. In other words, the mated females will be assigned tocomplete the last incomplete stratification group (if necessary) and then assigned to the nextstratification group until all GD 0 females for that day are assigned.The objective at the end of the mating period is to have all groups with very similar mean bodyweights (not statistically significantly different) with very similar indicators of variance (e.g.,standard deviation, standard error) and representatives from each mating date in each group.Obviously, if some of these females turn out not to be pregnant at scheduled sacrifice (and so arenot included in summarized and analyzed study data, such as body weights, weight changes), thesummarized GD 0 weights for reporting purposes (based on pregnant animals) will not be identicalto the summarized weights initially run on GD 0 to guarantee homogeneity of study groups. Thispossibility should be explained in an Standard Operating Procedure (SOP) or memorandum to thestudy records. The values should still be very similar.2. Treatment of Mated Femalesa. Routes of AdministrationVehicles used for gavage dosing should be innocuous. Suspensions may be administered orally,but injections (i.v., i.p., or i.m.) require the compound be soluble in a physiologically compatible,innocuous vehicle such as water, dextrose solution, or physiological saline (0.9% NaCl). The vehiclemay affect the outcome, e.g., corn oil may result in slower absorption than water, or in a differentcontrol embryofetal profile. 431) Gavage — Commonly used vehicles include water, corn (or other) oil, and aqueous (up to0.5% to 1.0%) methylcellulose, with or without polysorbate 800. The compound may be solublein the vehicle, or a suspension may be created. The only limiting factor is the physical nature ofthe compound to be administered in the vehicle. The formulation should be easily drawn up intothe dosing needle used (based on inner diameter of the needle) and be amenable to being expelledfrom the syringe without loss of homogeneity.Rodents and rabbits may be gavaged with blunt-ended, stainless-steel feeding needles (e.g.,from Popper and Sons, Inc., New Hyde Park, NY), 18 gauge and 1.5 in. long for mice and 16 or18 gauge and 2 in. long for rats. A 13-gauge, 6-in. stainless-steel feeding needle is normally usedfor rabbits. 44 These needles are attached to appropriately sized syringes to deliver the entire doseto the animal while allowing accurate measurement of the dosing volume. Alternatively, a flexible,rubber catheter may be used for oral gavage dosing in rabbits. This involves using a Nelaton ®French catheter from 8 to 12 in. long attached to a syringe adapter. The rabbit is put in a restrainer,and a plastic bit is inserted into the rabbit’s mouth; the catheter is eased into the hole in the bitand down into the stomach. Rats may also be dosed by catheter. Dosing volumes are typically 5to 10 ml/kg for mice; 2.5, 5, or 10 ml/kg for rats; and 1 to 5 ml/kg for rabbits (lower volumesshould be employed for corn oil solutions or suspensions). These volumes allow administration ofthe compound into a full stomach. For rodents, the compound is drawn into an appropriately sizedsyringe with the stainless-steel feeding needle attached. Since the stainless-steel feeding tubes goonly halfway down the esophagus to the stomach, care must be taken while inserting the needle toavoid inadvertently inserting the needle into the trachea. A well-trained technician who is alert to signsof a distressed animal can avoid such a misinsertion of the needle and a resultant “lung shot” (i.e.,introduction of the dosing agent into the trachea and/or lungs). In rodents, a lung shot results inimmediate demise; in rabbits, sequelae may not be detected until necropsy. A disadvantage of the steelfeeding tube is efflux of the compound should it be injected too quickly, or as more often happens, theesophagus constricts, causing backwash. This will often result in aspiration pneumonia.© 2006 by Taylor & Francis Group, LLC

222 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONWhen catheters are used, the dosing compound is drawn directly into the syringe. Then theadapter with the catheter is attached to the syringe, and the catheter is fed through the bit completelyinto the stomach. If the tube is guided into the trachea, the rabbit will twitch its ears and cough,signaling the technician to withdraw the tube and try again. Steady movement of the nostrils shouldbe apparent while the compound is being injected by use of a rubber catheter. With this method,it is necessary to “wash” the catheter with approximately 1 ml of vehicle after dosing, to administerthe last milliliter of compound remaining in the catheter.2) Dosing via the Diet — The test compound may be introduced directly into ground rodent orrabbit feed, depending on the characteristics of the compound, including its palatability and solubility.The test compound may be mixed with a solvent prior to mixing with the feed, or it may bemicroencapsulated to limit release to a particular part of the intestines or to ensure that unpalatablecompounds will be eaten.Ground feed for rodents is made available in the home cage in glass feed jars that are chemicallyresistant, easily sanitized, and transparent (allowing cage-side observation of the feeder). The jarsare typically fitted with a stainless-steel lid. Mouse feed jars are also fitted with a wire cylinderthat allows access to the feed but prevents nesting in the feed. Rats will generally use their pawsto scoop out the feed and then lick their paws. The feed jar is weighed at intervals based on thesize of the jar, the amount consumed, and the stability of the compound in the feed. This allowsthe calculation of grams of feed eaten per kilogram of animal body weight per time interval (usuallydaily), and therefore quantitates (in milligrams or gram per kilogram body weight per day) theactual amount of test article consumed. Most feed consumption occurs during the dark cycle, sothe rodent is being “dosed” periodically at night. This differs from gavage, where the test materialis generally given as a once-daily bolus dose in the morning.Rabbits will not eat ground food; 33,45 therefore, any ground, feed-based dose formulation mustbe pelleted prior to administration to rabbits. This may be done in-house or by commerciallaboratories. Because most performing laboratories do not have pelleting capabilities, many compoundsthat may be administered to rodents in the feed are given to rabbits by gavage.Before glass feed jars are used for another study, they should be washed. Then, several jarsthat had been used for the high-dose feed should be rinsed, employing a solvent appropriate forthe test compound they had contained. The jar rinse is then analyzed for the presence of the testcompound. If the results of the analysis are negative, the jars are considered suitable for use inanother study. If not, additional washing may be done, or the jars may be soaked with solvent andreanalyzed.3) Dosing via Drinking Water — The major consideration for use of drinking water as a vehiclefor compound administration is the palatability of the test compound in water. If the animals won’tdrink their water because of its smell or taste, you cannot successfully dose them by this route.Water bottles are used for the dosing solution. These may be polypropylene or polycarbonate ifthe test compound is nonreactive with plastic. Glass bottles may be used (either clear or amber),depending on the reactivity of the compound to light. In our experience, plastic bottles are unsuitablefor rabbit studies because rabbits tend to play with the bottles to the extent that measurement ofwater consumption is almost impossible. Glass bottles in sturdy holders seem to be more reliable.Water consumption is determined by weighing the bottle at intervals determined by the stabilityof the compound and the volume of water consumed daily by the animal. Bottle washes shouldalso be done prior to use for another study to verify absence of test compound.4) Injection — Intravenous — The blood vessels used for intravenous (i.v.) injection in therodent are normally the dorsal and ventral tail veins. Rodents are dosed i.v. at between 2.5 and 10ml per kilogram of body weight. To dilate the vein, the rodent is placed in a warming box undera heat lamp for a few seconds, or the tail is placed in a warm water bath at 37°C. The rodent is© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 223placed in a Plexiglas box that has a narrow V opening in one side. The rodent’s body remains in thebox with the tail extending to the outside through the V opening. Alternatively, the rat can be suspendedin a cone-shaped container (attached to a ring stand) with the tail hanging down in warm water.To ensure that a length of vein will remain patent on subsequent days of dosing, the site of thefirst injection should be as near the tip of the tail as possible. Subsequent injection sites would belocated more proximal to the body. Holding the tip of the tail, the technician inserts the needle(normally 26 gauge attached to a 1-cc syringe) into the vein at a slight angle and slowly injectsthe test article. A localized swelling at the injection site indicates the dose was not delivered i.v.Following withdrawal of the needle, gauze is placed over the injection site, and pressure is applieduntil bleeding stops.Rabbits are dosed via the marginal ear vein, usually at 1 ml/kg of body weight. The rabbit isplaced in a restrainer that allows full access to the ears. The hair from the ear to be dosed is pluckedfrom the tip to the base of the ear at the ear margin to expose the vein. A 25- or 26-gauge needleis used with the appropriate size syringe. The vein is compressed proximal to the injection site.The needle is inserted into the vein (pointing toward the head), the pressure on the vein is released,and the dosing solution is injected slowly. The needle is then removed, and gauze is held firmly overthe injection site until the bleeding stops. The ear used is alternated each day, and the injection sitesshould be selected beginning at the tip of the ear and subsequently moving toward the head. Syringesshould be changed after each dosage group, and needles should be changed after each animal.Subcutaneous — In rodents, subcutaneous (s.c.) injections are normally made in the interscapulararea on the back of the animal. The volume used is generally up to 10 ml/kg of body weight,with needle size dependent on the size of the animal and the viscosity of the dosing formulation.The loose skin behind the neck is grasped between the thumb and forefinger. The needle is insertedthrough the fold of skin but not into the underlying muscle. Loose skin behind the neck or in thehip area of the rabbit is used for s.c. injections. Syringes should be changed between dosage groups,and needles should be changed after each animal is dosed.Intraperitoneal — A rat is restrained by holding its head and thorax. A mouse is grasped bythe scruff of the neck, and the tail is twined around a finger to control the hindquarters. The needleshould be introduced rapidly into a point slightly left or right of the midline and halfway betweenthe pubic symphysis and xiphisternum on the ventral abdominal surface. A 25- or 26-gauge needleis used, and the typical injection volume is 2.5 to 10 ml/kg of body weight.Rabbits are held by the scruff of the neck by a seated technician, with the rabbit’s hips restingin the technician’s lap and the rabbit facing another technician who performs the dosing. The needleis inserted through the skin of the abdomen just lateral to the midline and just posterior to the areaof the umbilicus, pointed toward the spine. The syringe is changed between dose groups, and theneedle is replaced between injections. Intraperitoneal injections in late pregnancy run the risk ofinjection directly into the gravid uterus, since a great deal of the space in the peritoneal cavity istaken up by the uterus.Intramuscular — The site of intramuscular dosing is usually the haunch or upper hind leg(“thigh” region), with the needle inserted into the muscle; repeated dosing should alternate dosingsites (right, left, and repeat).5) Inhalation Exposures — Mated females can be exposed in their home cages (if they arewire-mesh hanging cages), with food and water sources removed (to prevent inadvertent exposurevia ingestion of absorbed or dissolved test material), or the animals can be transferred to exposurecages. Automatic watering systems can be employed within the exposure chambers to providedrinking water ad libitum during the exposure periods. The amount of water exposed to the testatmosphere in the tip of the “nipple” is very small. Therefore, the amount of dissolved test materialin the water available for drinking is also very small. Animals are usually exposed 6 to 8 h/d. Thisinterval is determined from the time the desired concentration is reached at the start of exposure,using a calculated t 99 (time required to attain 99% of the target concentration) until the exposure© 2006 by Taylor & Francis Group, LLC

224 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONsystem is shut down, with subsequent exhaust until the calculated t 01 (time to 1% of the targetconcentration) is achieved. Individual animal clinical observations should begin as soon as possibleafter the chamber is opened. General observations on animals viewed through the chamber duringexposures should also be made (although not all animals will be visible, and the identity of a givenanimal may not be ascertainable) since the postexposure observations may not coincide with thetime of maximal clinical response.6) Nose- or Head-Only Exposures — Nose-only or head-only exposures are usually reservedfor exposure to aerosols, dusts, or mists (which the animal can groom from the fur and ingestand/or absorb through the intact skin), or for radiolabeled, difficult to obtain, or expensive testmaterials (to reduce the amount of material needed for exposures). These exposures require restraint(and therefore cause stress); the confinement equipment must minimize compression of the rapidlyexpanding abdominal contents and prevent additional heat stress. Commercially available equipmentfor nose-only or head-only exposures can be purchased for rabbits, rats, and mice. 46 It maybe appropriate to adapt the animal to the exposure apparatus for one to two days prior to the onsetof exposures to minimize stress.7) Subcutaneous Insertion — Subcutaneous insertion of test material, an implant, or anosmotic minipump (for continuous infusion) is done under anesthesia. An incision is made on theinterscapular dorsum, and one or more subcutaneous pockets is created with blunt forceps in theconnective tissue space above the muscular layer. The implant material is inserted, and the incisionsite is stitched or closed with wound clips. The implant material is usually inserted at the start ofthe dosing period, and pre- and postexposure weights of the implanted material can be used toascertain actual “dose.” Alternatively, the implanted material can be retrieved at the scheduledterminal gestational sacrifice. Concurrent control group animals undergo the same procedures(anesthesia, pocket formation, implantation of an implant, wound closure, etc.), with the implantempty or filled with the vehicle.8) Cutaneous Application — The EPA has examined the acceptability and interpretation ofcutaneous developmental toxicity studies. 47–49 The procedures for this type of administration are asfollows. The dosing site on the interscapular dorsum and up to 10% of the total body surface areshaved or clipped prior to initial application, with repeated shaving or clipping as needed duringthe dosing period. For a mouse, the application site is 2.5 × 2.5 cm, and 0.1 ml of the test articleis applied (i.e., 3 ml/kg for a 30-g animal); for a rat, the site is 5 × 5 cm or 7.5 × 7.5 cm, with 0.5to 0.6 ml applied (1.5 to 1.8 ml/kg for a 300-g animal); for a rabbit, the site is 7.6 × 7.6 or 10 ×10 cm, with 1 to 2 ml applied (0.3 to 0.6 ml/kg for a 3-kg animal).For a nonoccluded application, the animal is manually restrained, and the test material is usuallyapplied by syringe over the prepared site. A rat or rabbit may be provided with a collar (see Figure7.5A) to preclude access to the dosing site, but the collar should not prevent access to food or water.Collars sized for mice (approximately 2-in. outer diameter and ¾-in. inner diameter) are now availablecommercially (by special order from Lomir Biomedical, Inc., Quebec J7V7M4, Canada).Occlusion (covering the dosing site) reduces any volatilization of test material, so more testmaterial remains on the site, and the test animal does not inadvertently get exposed by inhalationof the volatilized material. Occlusion also prevents access to the dosing site during the animal’sgrooming, so more test material remains on the site and the test animal does not inadvertently getexposed by ingestion of test material contaminating the hair. It further maximizes and/or acceleratesabsorption through the skin by increasing both temperature and humidity at the dosing site. 49Methods of occlusion (Figure 7.5) include application of sterile gauze and Vetwrap (3MAnimal Care Products, St. Paul, MN) over the dosing site, usually accompanied by use of an“Elizabethan collar” (Lomir Biomedical, Inc., Quebec J7V7M4, Canada). This technique is commonlyused on rabbits and rats (Figure 7.5A). 49–51 A second method useful in rabbits and rats© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 225A85–1234CollarTapeGauzeApplication siteShaved areaBVelcroGauzeShaved areaFore limb holeCScreen platformScreen compartmentGauzeShaved areaTapsHead clip attached to neck springNeck spring catchChin restNeck spring pivotFigure 7.5Methods of topical occlusion for pregnant animals. (From Tyl, R.W., York, R.G., and Schardein,J.L., Reproductive and developmental toxicity studies by cutaneous administration. In Health RiskAssessment: Dermal and Inhalation Exposure and Absorption of Toxicants, Wang, R.G.M., Knaak,J.B., and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 1992. With permission.)involves use of a Spandex ® elastic jacket, with forelimb holes, Velcro ® closures on the back, anda plastic sheet on the dorsal underside to overlay the dosing site; these are available commercially(e.g., Lomir Biomedical, New York, NY). The jacket is placed on a flat surface, the test animal isplaced with its forelegs in the holes, and the test article is applied. Sterile gauze is placed over thesite (if appropriate), and the jacket flaps are brought up over the site and snugly attached by theVelcro strips. In most cases, a collar is not needed (Figure 7.5B). 49–52© 2006 by Taylor & Francis Group, LLC

226 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA method developed for occluded cutaneous application in mice (Figure 7.5C) 49,53 utilizes astainless-steel mesh tray, compartmentalized by stainless-steel fences between each animal slot.Each compartment has an opening at one end for the animal’s head, with a chin rest and a neckspring. Once the mouse is restrained, the dorsal dosing site is prepared, the test article is applied,and sterile gauze is placed over the site. Adhesive tape is secured across the dosing sites of allanimals in the tray by clips on each fence.In each case, the site is examined, and the test article and occlusion are applied at the start ofeach daily dosing period. At the end of each dosing period, the occlusion is removed, the gauze isdiscarded, and the site is gently washed and examined. A useful system of scoring the applicationsite is provided by Draize et al. 54 for edema (swelling), erythema (redness), and any eschar (necrotictissue) formation.b. Duration of AdministrationThe recent governmental testing guidelines specify exposure beginning after implantation is complete(or on the day of insemination; see below) and continuing until the day before scheduledsacrifice at term. This corresponds to GD 6 (or 0) through 19 (or 20) for rats, GD 6 (or 0) through17 (or 18) in mice, and GD 6 (or 0) through 28 (or 29) for rabbits, if the day sperm or a copulationplug is found (rodent) or mating is observed (rabbit) is designated GD 0. In the previous protocols,the guidelines specified exposure only from GD 6 to the end of major organogenesis (closure ofthe secondary palate), GD 6 through 15 for rodents, GD 6 through 18, 55 or GD 7 through 19 56 forrabbits.The rationale for starting exposures after implantation is complete is based on two possibleconfounding scenarios:• If the initial (parent) test material is teratogenic and the metabolite(s) is not, and if metabolism isinduced by exposure to the parent compound, then exposure beginning earlier than implantation(with concomitant induction of enzymes and enhanced metabolism) will result in the conceptusesbeing exposed to less of the teratogenic moiety, and the study may be falsely negative.• If the test chemical and/or metabolites interferes with implantation, then exposure prior to implantationwill result in few or no conceptuses available for examination.However, there are situations when initiation of exposure should begin prior to completion ofimplantation. These include:• For exposure regimens that are anticipated to result in slow systemic absorption (e.g., cutaneousapplication, subcutaneous insertion, or dosing via feed or water, steady-state), maximal bloodlevels may not be attained until the very end of (or beyond) organogenesis if exposures begin onGD 6 or 7.• For materials that are known to have cumulative toxicity (due to buildup of chemical and/or insult)after repeated exposures, exposure should begin on GD 0 (or earlier) so that the conceptuses aredeveloping in a fully affected dam.• For materials that are known to deplete essential components (such as vitamins, minerals, cofactors,etc.), exposures should begin early enough so that the dam is in a depleted state by the start oforganogenesis (or by GD 0).• For materials that are innocuous as the parent chemical but that are metabolized to teratogenicforms, exposures should begin early enough so that postimplantation conceptuses are exposed tomaximal levels of the teratogenic metabolites.• For test materials that are known not to interfere with implantation, exposure encompasses theentire gestational period and offspring are available for examination.In the case of exposures that do not involve bolus dosing (i.e., gavage), such as dosing via thefeed or water, and subcutaneous implants, the duration of each daily exposure is essentially© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 227continuous for the entire dosing period ad libitum. For cutaneous application and inhalationexposures, the daily duration is usually 6 to 8 h (corresponding to a human workday), althoughsome studies have employed exposures of 22 h/d, with 2 h/d for housekeeping and allowing theanimals to eat and drink unencumbered.The termination of exposures was previously specified as the end of major organogenesis. It issignaled by the closure of the secondary palate and the change in designation of the conceptusfrom embryo to fetus. The cessation of exposure prior to term allowed for a postexposure recoveryperiod for both the dam and the fetuses, and an assessment could be made regarding whether theobserved maternal effects (body weights, clinical observations) are transient or permanent. However,fetal evaluations take place at term, and there is no commonly employed way to detect early adverseeffects on the conceptus that resolve (are repaired, compensated for, or result in in utero demise)earlier in gestation. What was observed at term was the net result of the original insult and anyrepair or compensation that occurred subsequently in the conceptus, as well as any observableeffects on the dam after a postexposure period. Thus, no detailed information on the dam wasobtained during the exposure period, and there was no way to distinguish effects during exposurefrom those occurring afterward. The new developmental toxicity testing guidelines require exposureuntil term, which includes the fetogenesis period.The consequences to exposures continuing until term are as follows:• There is no postdosing recovery period so the outcome at term is not confounded by insult duringdosing and compensation in the postdosing period. Example: There is no increase in maternal feedconsumption or weight gain in postdosing period and no obscuring of effects on fetal body weight(since most fetal body weight gain is in the “last trimester”). There may be increased in utero fetaldeath due to continued direct or indirect toxic insult.• The NOAEL may allow a lower dose (than if the dosing ended on GD 15), which may be a moreappropriate value.• The incidence of fetal malformation may be underestimated because more malformed fetuses maydie with prolonged maternal dosing.• Dams will require more test material as they gain weight in the “last trimester” (to maintainconstant dose in milligrams per kilogram per day), so maternal and/or fetal toxicity may be greaterat a given dose (i.e., the maternal liver will have to metabolize more test material).• Fetal malformation rates will probably not increase because they are usually induced during theperiod of major organogenesis (GD 6 through 15), common to both dosing regimens.• Maternal assessments at scheduled necropsy on GD 20 (e.g., toxicokinetic, biochemical, physiologicalendpoints) reflect effects of continued dosing (not after a postdosing recovery period). Asatellite group of dams to be sacrificed at the end of dosing under the shorter dosing period forsuch assessments is, therefore, not necessary (this saves animals, time, and money).• More test material is necessary (more dosing days and dams are heavier in the “last trimester,” somore test material is administered to maintain a constant dose in milligrams per kilogram per day).• The costs to perform a Phase II study with prolonged dosing are increased because more testchemical is required and there is increased labor to dose the dams on the additional days.The rationale for continuing exposure until term includes:• Maternal exposure until term is a better model for human exposure than exposure only during aportion of gestation, with abrupt cessation at the end of embryogenesis.• Continued maternal responses at term (e.g., changes in organ weights, hematology, clinical chemistry,histopathology) can be better interpreted in terms of causality; there is no maternal postexposureperiod for compensatory changes to occur.• Many systems continue to develop in the fetal period, both in terms of increases in cell size andthe number and differentiation of specialized cells, tissues, and organs (e.g., central nervous,pulmonary, renal, gastrointestinal systems, etc.). The effects on these processes occurring in thepresence of continual maternal exposure will be manifested at term (for most of the systems) andwill not be confounded by compensatory processes that may occur in a postexposure period.© 2006 by Taylor & Francis Group, LLC

228 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION• The male reproductive system is established and differentiates internally and externally in utero,beginning on GD 13 to 14 in rodents, so effects from possible endocrine-active or reproductivelytoxic compounds with other mechanisms may be detected. However, at term, only the testes andepididymides can be reasonably assessed, and most effects are not discernible until weaning,puberty, or adulthood (such as morphological and/or functional effects on accessory sex organs,adult testicular and epididymal spermatogenesis, and sperm transit).Recently in the authors’ laboratory, a comparison was made of parameters of maternal anddevelopmental effects in control CD ® (SD) rats dosed from GD 6 through 15 versus from GD 6through 19. 57 Although the GD 6 through 15 dosing was employed for earlier studies (1992 to1997) and the GD 6 through 19 dosing was employed for more recent studies (1996 to 1998), therewas an overlap in time between studies employing the different dosing durations.As anticipated, control maternal body weights and weight gain, gravid uterine weight, absoluteand relative liver weight, and maternal body weight change (GD 20 – GD 0, minus weight of thegravid uterus) were all significantly lower with the extended dosing period (the control damsreceived vehicle only, with no exposure to test chemical; Table 7.6). Litter size and fetal bodyweights were also reduced with the extended maternal dosing period (in the absence of any testchemical; Table 7.6). Interestingly and unexpectedly, uterine implantation sites per litter were alsoreduced as were the number of adversely affected fetuses per litter (Table 7.6). The same was trueof percent fetuses per litter with visceral malformations, total (any) malformations, visceral variations,and total (any) variations (Table 7.7 and Table 7.8).The authors concluded that the longer dosing regimen with no recovery period resulted insignificant depression of maternal body weight and weight gain endpoints, as well as a reductionin fetal body weights. Presumably, the stress of handling and dosing is responsible for thesedifferences. The three vehicles used (methylcellulose, corn oil, and water) were equally representedin both data sets. The concern is whether dosing with a potentially toxic test material will resultin even further reductions due to a synergistic effect of the longer dosing period and the toxicityof the test material. The decrease in the number of implants and live fetuses may be due to thedifferences in times of performance of the two groups of studies. In the early 1990s, Charles RiverLaboratories selected all offspring from larger litters (i.e., rather than a set number per litter,regardless of litter size) as breeders, so that the average litter size rose. In the latter 1990s, a morebalanced selection program was instituted (the CD ® [SD] “international gold standard” IGS) to haltand reverse the increasing litter sizes. The decreased incidence of hydronephrosis (a commonmalformation), as well as enlarged lateral ventricles of the brain and rudimentary ribs on lumbarI (both variations), may represent genetic drift in this strain over the years evaluated. The relativedevelopmental delay of the fetal skeleton, shown by the increased incidences of dumbbell cartilageand bipartite ossification centers in the thoracic centra, is likely due to decreased fetal body weightsat term in the litters under the longer dosing regimen (i.e., the fetuses are delayed in late gestationaldevelopment but are appropriate for their size). 58These considerations are the basis for the requirement in the new OPPTS, 10 FDA, 12 and OECD 13final developmental toxicity testing guidelines that exposures continue until terminal sacrifice. 16D. Observation of Mated FemalesThe better and more complete the profile on maternal effects, the better the interpretation ofembryofetal findings in the context of maternal toxicity, if any. 58–611. Clinical ObservationsClinical observations of all animals should be made once daily during the pre- and posttreatmentperiods, at least once daily at dosing, and at least once after dosing for gavage studies (with timingof observation based on initial findings during the dosing period) throughout the dosing period.Observations should be made for (but not limited to) the potential findings listed in Table 7.9.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 229In the authors’ laboratory, mated females are observed closely on the first day of dosing at anumber of postdosing time points (e.g., 15 and 30 min and 1, 2, and 4 h). This is done to observeany possible treatment-related clinical signs, to identify the times at which such signs are maximallyexpressed (for determination of times of observation for subsequent dosing days), and to documentwhether the signs are transient (and the time frame if they are).2. Maternal Body WeightsMated females should be weighed on GD 0 and at least once more on GD 3 prior to the exposureperiod (if exposures are started on GD 6). Females should be weighed frequently during the exposureperiod, at least every three days (e.g., GD 6, 9, 12, 15, 17 [mice], 18, and 20 [rats]; GD 6, 9, 12, 15,18, 21, 24, 27, and 29 or 30 [depending on the day of necropsy] for rabbits). 42 Assessment of maternalweight is very useful for detection of early consequences of dosing (e.g., taste aversion from gavage,which may become more profound over time, or palatability problems from treated food or water,which may resolve over time) and stress for other routes (e.g., inhalation, cutaneous application),compounded by any effects from the chemical. Maternal weights can also be used to ascertain whetherearly effects are transient and if there is complete litter loss prior to term. For example, CD ® rats willusually gain 40 to 60 g from GD 0 to 6 if they are pregnant (F344 rats will gain at least 20 g, andCD-1 ® mice will gain 2 to 4 g). Subsequent weight loss to GD 0 levels may indicate total litter loss.Maternal body weight changes should be calculated for the preexposure period if employed(GD 0 to 6 or 7), for time increments during exposure, and for the total exposure period (GD 6 to9, 9 to 12, 12 to 15, 15 to 18, and 18 to 20, and 6 to 20 for rats; GD 6 to 9, 9 to 12, 12 to 15, 15to 18, 18 to 21, 21 to 24, 24 to 27, 27 to 30, and 6 to 30 for rabbits). Weight changes (calculatedper animal and summarized for each dosage group) are more sensitive indicators of effects thanare body weights per se. Such measures can detect early, transient weight reductions during thedosing period and compensatory increases late in the dosing period.Total gestational weight change (GD 0 to 17 or 18 for mice, GD 0 to 20 or 21 for rats, andGD 0 to 29 or 30 for rabbits) is also calculated. Gestational weight change corrected for the weightof the gravid uterus (weight change during gestation minus the weight of the gravid uterus) andcorrected terminal body weight allow measurement of a weight change or terminal weight of thedam, with no contribution from effects on the conceptuses. This corrected parameter will provideinformation on maternal toxicity per se, unconfounded by reduced numbers and/or body weightsof the litter. This correction can only be made using the weight of the term gravid uterus. Theimpact on maternal body weights of effects on the conceptuses (e.g., reduced number of live fetuses,reduced embryo/fetal body weights) during gestation cannot be ascertained unless extra femalesare included in the various treatment groups and are necropsied during the exposure period toprovide input on body weight, gravid uterine weight, and status of the offspring.3. Maternal Feed ConsumptionFeed consumption of singly housed animals should be measured throughout gestation on the samegestational days on which the maternal animals are weighed. This is done by measuring the fullfeeder (or the feed alone) at the start of each interval and the “empty” feeder (or the remainingfeed alone) at the end of each interval. Pretreatment, treatment, posttreatment, and gestational feedconsumption should also be calculated.Feed consumption should be presented as grams per animal per day and grams per kilogramper day, with the latter obviously factoring in the weight of the female during the consumptioninterval. For example, if a rat that weighed 300 g at the start and 350 g 2 d later ate 25 g/d duringthe interval, then the average food consumption is as follows:25 g/d(0.300 kg + 0.350 kg)2=25 g/d0.325 kg= 76.92 g/kg/d© 2006 by Taylor & Francis Group, LLC

Table 7.6Recent historical control data for developmental toxicity studies in common laboratory animalsParameter CD ® (SD) Rats CD-1 ® Mice NZW Rabbits a Rabbits bNZWNo. pregnant 672 71 227 88No. live litters 672 70 224 83No. fetuses 10,033 841 1800 558No. corpora lutea/litter 14.91–18.74 12.43–13.86 8.31–10.38 10.48 ± 0.26No. implants/litter 13.71–16.75 11.57–13.35 7.58–9.06 7.19 ± 0.31Percent preimplantation loss/litter 0.55–14.76 5.12–9.25 5.82–17.48 30.74 ± 2.90No. (%) litters with nonlive implants 5 (13.04)–17 (70.83) d 7 (30.43)–17 (68.00) 3 (14.29)–17 (42.11) 37 (42.0)No. nonlive implants/litter 0.22–1.21 0.61–1.00 0.14–0.74 0.85 ± 0.15Percent nonlive implants/litter 1.18–6.82 4.85–11.50 1.85–16.15 14.46 ± 2.80No. live fetuses/litter 13.30–16.24 10.96–12.70 7.00–8.50 6.72 ± 0.31Male 6.21–8.72 5.42–6.87 3.50–4.74 3.39 ± 0.23Female 6.32–8.65 5.04–6.96 3.50–4.64 3.33 ± 0.21Percent male fetuses/litter 41.83–57.06 44.19–54.87 43.04–51.51 49.68 ± 2.85Fetal body weight/litter (g) c 3.430–3.866 1.007–1.042 48.09–53.30 50.23 ± 0.96Males (g) 3.519–4.063 1.007–1.065 48.14–54.16 49.65 ± 1.00Females (g) 3.424–3.791 0.999–1.021 46.96–52.79 48.86 ± 0.88No. (%) fetuses with malformations 0 (0.00)–48 (10.84) 3 (1.01)–31 (12.30) 1 (0.64)–16 (10.53) 32 (5.7)No. (%) litters with malformed fetuses 0 (0.00)–13 (56.52) 3 (12.50)–14 (60.87) 1 (5.26)–9 (42.86) 21 (25.3)No. malformed fetuses/litter 0.00–1.66 0.13–1.35 0.05–0.80 0.39 ± 0.10Male 0.00–0.97 0.08–1.00 0.00–0.40 0.25 ± 0.08230 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Female 0.00–0.69 0.004–0.39 0.00–0.40 0.16 ± 0.05Percent malformed fetuses/litter 0.00–10.60 1.14–12.18 0.62–10.57 5.54 ± 1.48Male 0.00–10.36 1.67–18.44 0.00–9.50 5.21 ± 1.61Female 0.00–12.64 0.60–5.41 0.00–9.25 5.63 ± 2.00No. (%) fetuses with variations 16 (4.95)–202 (45.60) 38 (13.01)–161 (54.21) 51 (42.02)–158 (65.60) 343 (61.5)No. (%) litters with variant fetuses 10 (43.48)–30 (100.00) 17 (73.91)–23 (95.83) 11 (85.00)–45 (100.00) 77 (92.8)No. variant fetuses/litter 0.70–6.96 1.65–6.71 3.22–5.96 4.13 ± 0.31Male 0.39–3.72 1.04–2.88 1.30–3.22 2.24 ± 0.22Female 0.30–3.24 0.61–3.83 1.83–2.89 2.19 ± 0.19Percent variant fetuses/litter 5.22–46.55 12.87–54.70 42.53–62.70 60.74 ± 3.62Male 6.27–42.60 16.11–55.52 38.40–59.45 57.66 ± 4.46Female 3.82–49.57 8.86–54.63 47.74–82.36 63.24 ± 4.11Note:For data involving ranges as No. (%)–No. (%), the specific percent is not necessarily based on the specific number. The range ofnumbers and the range of percentages are not necessarily the same since different studies had different numbers of litters/fetuses.The values are the lowest to the highest numbers and the lowest to the high percentages.aNaturally bred, without hormonal priming.bArtificially inseminated, with hormonal priming; values are presented as grand mean of study control group means ± SEM.cSexes combined.Source: From the Reproductive and Developmental Toxicology Laboratory of the authors at RTI; data are presented as range of studycontrol group means. For CD ® rats, 28 studies; for CD-1 ® mice, 3 studies; and for naturally bred rabbits, 11 studies.DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 231© 2006 by Taylor & Francis Group, LLC

232 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 7.7Summary and statistical analysis of control rat fetal malformationsand variationsGestational Days of DosingParameter 6–15 6–19No. fetuses examined a 2723 1932No. litters examined b 178 136MalformationsNo. fetuses with external malformations/litter c,d 0.02 ± 0.01 0.02 ± 0.01Percent fetuses with external malformations/litter c,d 0.16 ± 0.08 0.13 ± 0.08No. (percent) fetuses with external malformations d 4 (0.1) 3 (0.2)No. (percent) litters with external malformations e 4 (2.2) 3 (2.2)No. fetuses with visceral malformations/litter c,d 0.55 ± 0.10 0.11 ± 0.04Percent fetuses with visceral malformations/ litter c,d,f 6.92 ± 1.36 1.41 ± 0.49No. (percent) fetuses with visceral malformations d 98 (6.4) 15 (1.6)No. (percent) litters with visceral malformations e 43 (24.2) 9 (6.6)No. fetuses with skeletal malformations/litter c,d 0.04 ± 0.02 0.05 ± 0.02Percent fetuses with skeletal malformations/litter c,d 0.89 ± 0.42 0.77 ± 0.03No. (percent) fetuses with skeletal malformations d 8 (0.4) 7 (0.7)No. (percent) litters with skeletal malformations e 7 (3.9) 7 (5.1)No. fetuses with any malformations/litter c,d 0.61 ± 0.10 0.18 ± 0.05Percent fetuses with malformations/litter c,d,f 4.73 ± 0.87 1.22 ± 0.30No. (percent) fetuses with malformations d 109 (4.0) 25 (1.3)No. (percent) litters with malformations e 53 (29.8) 17 (12.5)VariationsNo. fetuses with external variations/litter c,d 0.02 ± 0.01 0.0 ± 0.0Percent fetuses with external variations/litter c,d 0.10 ± 0.07 0.0 ± 0.0No. (percent) fetuses with external variations d 3 (0.01) 0 (0.0)No. (percent) litters with external variations e 2 (1.1) 0 (0.0)No. fetuses with visceral variations/litter c,d 3.62 ± 0.23 1.15 ± 0.11Percent fetuses with visceral variations/litter c,d,f 44.9 ± 2.89 16.04 ± 1.54No. (percent) fetuses with visceral variations d 645 (41.9) 157 (16.3)No. (percent) litters with visceral variations e 140 (78.7) 79 (58.1)No. fetuses with skeletal variations/litter c,d 0.97 ± 0.10 1.00 ± 0.11Percent fetuses with skeletal variations/litter c,d 10.87 ± 1.21 14.40 ± 1.58No. (percent) fetuses with skeletal variations d 172 (9.4) 136 (14.1)No. (percent) litters with skeletal variations e 94 (52.8) 74 (54.4)No. fetuses with variations/litter c,d 4.46 ± 0.24 2.15 ± 0.015Percent fetuses with variations/litter c,d,f 29.09 ± 1.50 15.26 ± 1.05No. (percent) fetuses with variations d 793 (29.1) 293 (15.2)No. (percent) litters with variations e 160 (89.9) 113 (83.1)aOnly live fetuses were examined for malformations and variationsbIncludes only litters with live fetusescReported as the mean ± S.E.M.dFetuses with one or more malformations or variationseLitters with one or more fetuses with malformations or variationsfp ≤ 0.001; significant difference between the two groupsSource: Marr, M.C., Myers, C.B., Price, C.J., Tyl, R.W., and Jahnke, G.D., Teratology,59, 413, 1999 (tables provided by the authors).Technicians should make notations regarding spilled feeders, feed in bottom of cage, etc., to accountfor “outlying” values not appropriate for inclusion in summarization and analysis (statistical testsfor outliers can also be used to identify suspect values). Water consumption should be assessed inexactly the same way as feed consumption. If the animals are dosed via feed or water, the actualintake of test material in grams or milligrams per kilogram of body weight per day can be easilycalculated. This is done by using the value for consumption in grams per kilogram per day multipliedby the percentage of test material in the diet or water.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 233Table 7.8Summary of morphological abnormalities in control CD ® rat fetuses:listing by defect type aGestational Days of DosingParameter 6–15 6–19Fetal Malformations and VariationsTotal no. fetuses examined externally b 2723 1932Total no. fetuses examined viscerally b 1538 d 966Total no. fetuses examined skeletally b 1838 d 965 eTotal no. litters examined c 178 136External MalformationsAnasarca 1 (1)Cleft palate 1 (1)Anal atresia 1 (1) 2 (2)Agenesis of the tail 1 (1)Short, thread-like tail 2 (2)Short tail 1 (1)Umbilical hernia 1 (1)Visceral MalformationsAbnormal development of the cerebral hemispheres 1 (1)Abnormally shaped heart 1 (1)Hydronephrosis:Bilateral 2 (2) 6 (4)Left 3 (2) 3 (2)Right 1 (1)Hydroureter:Bilateral 54 (28) 4 (3)Left 36 (19) 2 (2)Right 3 (3) 3 (3)Skeletal MalformationsUnossified frontals, parietals, and interparietals 1 (1)Discontinuous rib 1 (1)Discontinuous rib cartilage 1 (1)Fused rib cartilage 1 (1)Fused cartilage: thoracic centrum 1 (1)Bipartite cartilage, bipartite ossification center: thoracic centrum 6 (5) 6 (6)External VariationsHematoma:Neck 2 (1)Forelimb 1 (1)Visceral VariationsEnlarged lateral ventricle of brain (full):Bilateral 100 (43) 18 (15)Left 20 (13) 8 (7)Right 19 (14) 4 (4)Enlarged lateral ventricle of brain (half):Bilateral 129 (52) 48 (28)Left 12 (9) 21 (15)Right 11 (10) 7 (5)Enlarged lateral ventricle of brain (partial):Bilateral 283 (63)Left 14 (11)Right 14 (13)© 2006 by Taylor & Francis Group, LLC

234 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 7.8Summary of morphological abnormalities in control CD ® rat fetuses:listing by defect type a (continued)Gestational Days of DosingParameter 6–15 6–19Enlarged nasal sinus 1 (1) 2 (2)Agenesis of the innominate artery 1 (1) 5 (5)Pulmonary artery and aorta arise side by side from heart 1 (1)Extra piece of liver tissue on median liver lobe 1 (1)Very soft tissue: liver 1 (1)Blood under kidney capsule: Right 1 (1)Small papilla: Right 1 (1)Distended ureter:Bilateral 37 (21) 27 (18)Left 20 (17) 15 (12)Right 12 (9) 11 (10)Skeletal VariationsMisaligned sternebrae 2 (2)Unossified sternebra (I, II, III, and/or IV only) 5 (4)Rib on lumbar I:Bilateral full 3 (2)Left full 4 (4)Right full 1 (1) 1 (1)Bilateral rudimentary 31 (20) 6 (4)Left rudimentary 40 (29) 7 (6)Right rudimentary 15 (12) 3 (3)Short rib: XIII 5 (3) 2 (2)Wavy rib 1 (1)Incomplete ossification, cartilage present: thoracic centrum 1 (1)Dumbbell cartilage, normal ossification center: thoracic centrum 1 (1)Dumbbell cartilage, dumbbell ossification center: thoracic centrum 33 (19) 3 (3)Dumbbell cartilage, bipartite ossification center: thoracic centrum 10 (10) 40 (23)Normal cartilage, bipartite ossification center: thoracic centrum 36 (22) 70 (32)Normal cartilage, no ossification center: thoracic centrum 1 (1)Normal cartilage, unossified ossification center: thoracic centrum (VI-XIII only) 1 (1)aA single fetus may be represented more than once in listing individual defects. Data are presented as numberof fetuses (number of litters). See Table 7.7 for summary and statistical analysis of fetal malformation/variationincidence.bOnly live fetuses were examined.cIncludes only litters with live fetuses.dFor many studies, a single fetus was examined externally, viscerally, and skeletally.eOne fetus was lost during processing prior to skeletal examination.Source: Marr, M.C., Myers, C.B., Price, C.J., Tyl, R.W., and Jahnke, G.D., Teratology, 59, 413, 1999 (tablesprovided by the authors).E. Necropsy and Postmortem Examination1. MaternalMated females that die during the course of the study should be necropsied in an attempt todetermine the cause of death, with target tissues saved for possible histopathology. Females thatappear moribund should be humanely euthanized (by CO 2 asphyxiation for rodents and i.v. euthanasiasolution for rabbits) and necropsied to attempt to determine the cause of the morbidity, withtarget tissues saved for optional histopathology. Females showing signs of abortion or prematuredelivery should also be sacrificed, as described above, as soon as the event is detected. They shouldbe subjected to a gross necropsy, and any products of conception should be saved in neutral buffered© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 235Table 7.9Some typical maternal clinical observationsAlteration of:Body positionActivityCoordinationGaitUnusual behavior, such as:Head flickingCompulsive biting or lickingCirclingRooting in bedding aPresence of:HemorrhagesPetechiaeConvulsions or tremorsIncreased salivationaIncreased lacrimation or other ocular dischargeRed-colored tears (chromodacryorrhea)Nasal dischargeIncreased or decreased urination or defecation (including diarrhea)PiloerectionMydriasis or miosis (enlarged or constricted pupils)Unusual respiration (fast, slow, gasping, or retching)VocalizationaSalivation pre- or postdosing and/or rooting in bedding postdosing mayindicate taste aversion (from gavage dosing). Palatability problemsfrom dosed feed or water are most likely to be detectable from reducedfeed or water intake initially, with “recovery” back to normal intakevalues over time.10% formalin. Approximately 1.0 to 1.5 d before expected parturition (GD 17 or 18 for mice, 20or 21 for rats, and 29 or 30 for rabbits), all surviving dams/does should be killed by CO 2 asphyxiation(rodents) or i.v. injection into the marginal ear vein (rabbits). They should be laparotomized, theirthoracic and abdominal cavities and organs examined, and their pregnancy status confirmed byuterine examination. Uteruses that present no visible implantation sites or that contain one-hornpregnancies should be stained with ammonium sulfide (10%) to visualize any implantation sitesthat may have undergone very early resorption. 62 Immersion of the uterus in ammonium (or sodium)sulfide renders it useless for fixation and histopathologic examination. When it is to be subsequentlyfixed and examined, the authors stain the fixed uterus with potassium ferricyanide by a methodthey developed to detect resorption sites. The stain is then washed out of the uterus, and the uterusis replaced in fixative for subsequent histopathologic examination (the staining does not interferewith subsequent evaluations).At sacrifice, the body, liver, any identified target organs, and the uterus of each mated femaleshould be removed and weighed. Any maternal lesions should be retained in appropriate fixativefor possible subsequent histologic examination. The ovarian corpora lutea should be counted todetermine the number of eggs ovulated (to allow calculation of preimplantation loss and to put thenumber of implants in perspective). For rabbits, each corpus luteum is a round, convex body witha central nipplelike projection; for rats, each corpus luteum is a round, slightly pink, discreteswelling (counted with the naked eye or under a dissecting microscope). For mice, the ovariancapsule must be removed (peeled off) and the corpora lutea counted under a dissecting microscope;each appears as a rounded, slightly pink, discrete swelling. For partially resorbed litters, the numberof corpora lutea may be less than the number of total implantation sites since corpora lutea involute(to become small corpora albicantia) when they no longer sustain live implants (note that theconcordance between number of resorbed implants and the loss of corpora lutea is not exact).Corpora lutea can be discounted for summarization when the count is less than the number of© 2006 by Taylor & Francis Group, LLC

236 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONimplants, or the number of corpora lutea can be set to correspond to the number of implants tocorrect for involuting corpora lutea, typically in partially resorbed litters. The uterine contents (i.e.,number of implantation sites, resorptions, dead fetuses, and live fetuses) should be recorded (seebelow). Dead fetuses should be weighed, examined externally, and saved in neutral buffered 10%formalin.2. Fetala. Anesthesia and EuthanasiaOnce fetuses are removed from the uterus, several options for anesthesia or euthanasia are acceptable,but one may be preferred over the other, depending on the constraints of each institution’sAnimal Care and Use Committee. Rodent and rabbit fetuses may be anesthetized or euthanized byi.p. (or sublingual) injection of a sodium pentobarbital solution. Alternatively, the USDA considershypothermia an acceptable form of anesthesia for rodent fetuses. The fetus is placed on a wettedpaper towel over ice, which induces anesthesia by lowering the core body temperature below25°C. 63,64 This method does not interfere with the internal examination of organ systems, assometimes occurs with injected anesthetics. Rabbit fetuses are routinely euthanized by intraperitonealinjection of 0.25 ml of sodium pentobarbital solution. Anesthesia and/or euthanasia must beachieved prior to any further examinations.b. Implantation Detection, Designation, and RecordingOnce the uterus is removed, weighed, and opened, the status of implantation sites can be evaluated.Beginning at a designated landmark specified by the individual laboratory’s SOPs (e.g., starting atthe left ovarian end and moving, in order, to the cervix and up to the right ovary, or alternativelybeginning at the left ovary and moving to the cervix, then starting at the right ovary and movingto the cervix), the status of implantation sites is determined and the location of the cervix specified(since uterine position affects the reproductive outcome in unexposed and exposed conceptuses). 65Implantation sites can be characterized as follows:Live: Fetus exhibits spontaneous or elicited movement.Dead: Nonlive fetus with discernible digits in any or all paws, body weight at or greater than 0.8 gfor rats, 0.3 g for mice, or 10.0 g for rabbits.Full: Nonlive fetuses with discernible digits below the weights listed above for fetuses designated as “dead.”Late: Some embryonic or fetal tissue from fetal remains to a full fetus with discernible limb buds butno discernible digits; fetuses may show signs of autolysis or maceration and may appear pale orwhite.Early: Evidence of implantation sites visualized only after staining fresh uterine preparation withammonium sulfide; metrial glands and decidual reaction only; maternal placenta; maternal and fetalplacenta.Once implantation status is recorded, the umbilical cord of each fetus is gently pinched withforceps to occlude blood flow, and the cord is cut as close to the body as is practical. The placentasare examined and discarded (some laboratories weigh and/or retain placentas in fixative for possiblesubsequent histopathology). The ICH testing guideline 9 requires gross examination of the placentas.The OPPTS draft guidelines 10 called for closer examination of placentas (e.g., weight and retentionin fixative), but this suggestion did not make it to the final guideline. 66Each live fetus is given a fetal number, while dead fetuses and early or late resorptions aredesignated by a letter (e.g., D for dead, L for late resorption, E for early resorption) since no furthertracking will be done. To induce hypothermia, rodent fetuses may be placed in order (by uterinehorn) on a wetted paper towel over ice, or they may be placed in a compartmentalized box over a© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 237bag of crushed ice. Alternatively, the fetuses may be euthanized with sodium pentobarbital andplaced in a compartmentalized box. Each fetus may be individually identified by a plastic or papertag (e.g., jewelry “price tag”) on an alkali-resistant string (e.g., dental floss) secured around thefetus’ middle, below the rib cage, and above the pelvis.Rabbit fetuses may be numbered on the top of the head with a permanent marker once theyare removed from the amniotic sac, or they may be individually identified by a tag as describedabove. This must be done quickly, as rabbit fetuses are active, and their position in the uterus canbe uncertain if they are not marked immediately. The head is gently blotted dry, and a permanentmarker is used to gently record the fetal number on the top of the head. Excessive pressure maycause intracranial swelling, thus creating artifacts observed during the craniofacial examination.c. External ExaminationAll live fetuses are weighed, normally to a hundredth of a gram (thousandth of a gram for mice),in numerical order. Dead fetuses are also weighed. After weighing, each fetus is held under amagnifying lens and its sex (for rodents) is determined by gauging the anogenital distance (AGD;the distance between the anus and the genital tubercle [papilla]). In male CD ® (SD) rat fetuses, thehistorical control range for AGD is approximately 1.8 mm; in females, the distance is about onehalfthat, approximately 0.85 mm). In CD-1 ® mouse term fetuses, the male AGD is approximately1.2 mm, and the female AGD is approximately 0.6 mm. AGD can be accurately and preciselymeasured with an eyepiece diopter (accurate to 0.01 mm), attached to a stereo microscope, and amicroscope stage grid. Alternatively, vernier calipers or a micrometer eyepiece on a dissectingmicroscope, calibrated with a micrometer stage, can be used. Rabbit fetuses are normally not sexedexternally, as they have no distinct sex difference in AGD.The fetuses are carefully examined externally. 67 The examination should proceed from head totail in an orderly manner (with the aid of a magnifying lens for rats and/or a dissecting microscopefor mice, if necessary). The contour of the cranium should be noted in profile and full-face view.The bilateral presence of the eye bulges should be noticed; they should be symmetrical and ofnormal size and position. The eyelids should be closed. The pinnae (external ear flaps) should notbe detached from the head at this developmental stage (rodents only), and they should be checkedfor symmetry, size, shape, and location. The alignment of the upper and lower jaw is examined inprofile. A clean angle should be formed with the upper jaw and nose protruding slightly furtherthan the lower jaw. The upper jaw is checked face-on for notches, furrows, or distortions. Thetongue and palate are checked by depressing the tongue while opening a pair of closed forcepsthat have been inserted between the upper and lower jaws. The status of the teeth is also checkedin rabbit fetuses.The skin covering the head and the rest of the body is checked carefully for continuity, andany abnormalities in color, texture, or tone are noted. Irregular swellings, depressions, bumps, orecchymoses (subdermal hematomas) are recorded as well. The overall posture of the fetus shouldbe observed at this time. The fore- and hindlimbs are checked carefully for normal size, proportions,and position. The number and disposition of the digits is noted (four digits plus a dewclaw on theforelimbs, five digits on the hind limbs), and the depth of the interdigital furrows is explored bygently pressing against the paws with the forceps to spread the digits.Abnormalities of the general body form, failure of dorsal or ventral midline closure, and defectsinvolving the umbilicus are typically quite obvious but must be examined carefully for accuratedescription. The anus, external genitalia, and tail are examined next. The anal opening must bepresent and in the proper location. The external genitalia are checked for general shape and size.The tail is also examined for normal length and diameter (to detect thread-like tail), as well asabnormal kinking or curling and localized enlargements or constrictions. Dead fetuses may besaved for visceral and skeletal examination or may be preserved in neutral buffered 10% formalinfor possible subsequent examination.© 2006 by Taylor & Francis Group, LLC

238 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. RODENTCrown-rump lengthB. RABBITCrown rump lengthFigure 7.6Measurement of crown-rump length in term (GD 20) fetuses.Other evaluations, such as crown-rump length measurement, may be done prior to visceralexamination. For measurement of crown-rump length, each fetus is placed on its side in a naturalfetal position on a metric rule calibrated in millimeters (rodents) or centimeters (rabbits), or ismeasured with a vernier caliper or micrometer eyepiece. Length is measured from the crown tip(dorsal aspect of the head, corresponding to the hindbrain) to rump base (excluding the tail) andrecorded (see Figure 7.6). For example, the crown-rump length for control CD ® rat fetuses rangesfrom 30 to 40 mm (3.0 to 4.0 cm) and that for control New Zealand White (NZW) rabbit fetusesranges from 8 to 12 cm.d. Visceral Examination—Staples’s Technique 681) Selection — Current guidelines specify that 50% of the litter for rodents and 100% for rabbitsmust be evaluated for soft tissue alterations (malformations or variations). A random or arbitraryprocess of selection should be employed, such as evaluating every other fetus. In the case of 50%visceral evaluations, alternating even- or all odd-numbered fetuses from sequential litters could beselected (or one could select even-numbered fetuses from a dam with an even study number andodd-numbered fetuses from a dam with an odd study number, etc.). The designated rodent fetusesfor soft tissue evaluation will also have their heads removed, with appropriate unique identification© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 239(e.g., individual, labeled scintillation vials for rodents; a large jar for each litter with individuallynumbered heads for rabbits) for subsequent fixation, decalcification, and soft tissue craniofacialexamination. Since all rabbit fetuses will be examined viscerally, selection for which ones willhave their heads removed and evaluated can be made, as above, for the selection of rodent fetusesfor visceral examination.The tools needed for rodents include microdissecting scissors, forceps, and ultra microdissectingor iridectomy scissors (for heart cuts in rodents). All visceral examinations should be performedunder a dissecting microscope. Each fetus designated for visceral examination is placed in thesupine position (on its back) on a wax block with a number corresponding to the fetal number, andthe fetal limbs are secured by rubber bands or dissecting pins. A ventral midline incision is madefrom the umbilicus, cutting caudal to the genital tubercle and cutting cranially to the neck. The cutshould first be made through the skin so that the sternebrae are visible, and then through themuscular abdominal wall and the ribcage, slightly lateral to and on the right of the sternebrae. Oncethe incision is completed, the ventral attachment of the diaphragm is located. Working to eitherside from this point, the technician checks the diaphragm for herniations or other abnormalities(i.e., thinness) and carefully clips it away from the rib cage. Once the diaphragm has been clippedaway, the rib cage may be gently opened on either side and secured to the block with pins.2) Thoracic Viscera — The fetal viscera are examined sequentially, generally beginning withthe thoracic organs and moving caudally. The bilobed thymus is first examined for normal size,shape, and coloration, and removed. The lungs should be pink and frothy in appearance, withclearly visible alveoli. For rodents, there should be three lung lobes on the right side of the fetus,one on the left, and a small intermediate lobe that crosses the midline and lies slightly over to theleft side. Rabbits have three lobes on the right side and two mediastinal lobes on the left side. Thetrachea and esophagus are checked for normal alignment and their relationship to the vessels ofthe heart, i.e., the aortic and pulmonary arches, should be ventral to (in front of) the trachea andesophagus.The heart and great vessels should be examined carefully. Each vessel entering and leaving theheart should be the proper size and should course normally as far as it can be traced. The atria ofthe heart may be heavily engorged (this is not abnormal), but the heart should have a smooth,rounded appearance and a well-defined apex slightly left of center. A light indentation is usuallyvisible over the area of the interventricular septum. The pericardium, a thin transparent membranethat surrounds and cushions the heart, should be carefully stripped away. The first heart cut (withultra microdissecting scissors for rodents and sharp microdissecting scissors in rabbits) begins tothe right of the ventral midline surface at the apex. It extends anteriorly and ventrally into thepulmonary artery. When opened, this incision permits examination of the tricuspid valve, papillarymuscles, chordae tendineae, right side of the interventricular septum, atrioventricular septum, andsemilunar valve of the pulmonary artery. The second heart cut originates slightly to the left of theventral midline surface of the apex and extends anteriorly and ventrally into the aorta. Because thepulmonary artery lies ventral to the aorta, it is cut as this incision extends into the aorta. Followingthe second cut, these structures are visible: bicuspid valve, papillary muscles, chordae tendineae,left side of the interventricular septum, atrioventricular septum, and semilunar valve of the aorta.3) Abdominal Viscera — The liver lobes (four including the median, right lateral, left lateral,and caudate) are counted and checked for fusion, texture, and normal coloration. Rats do not havea gallbladder. Rabbits’ gallbladders vary more in shape and size than those of other species. Thestomach plus the lower esophagus, spleen, pancreas, and small and large intestines are examinednext. The stomach and intestines should be filled with a viscous green or yellowish fluid (bile plusswallowed amniotic fluid). The kidneys should then be examined; the left kidney should be locatedslightly more caudal than the right. The ureters are checked for normal size and location and forcontinuity from the renal hilus to the urinary bladder. Then, each kidney is cut transversely (or the© 2006 by Taylor & Francis Group, LLC

240 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONleft longitudinally and the right transversely) so that the renal papilla and renal pelvis may beexamined. Any abnormal dilation of the ureters or the renal pelvis (more common in rats) is noted.The urinary bladder should be checked at this time.The sex of each fetus is then determined by carefully inspecting the gonads. Rodent testes aredefined by their round to bean shape and the tortuous spermatic artery that runs along the side ofeach testis, with associated epididymides (caput [head], corpus [body], and cauda [tail]) on thelateral side of the testes. The testes should be located low in the pelvis on either side of the urinarybladder. Additional reproductive structures can be seen in the term rat fetus with a magnifying lensor dissecting microscope. For males, these include the white, threadlike efferent ducts (vasaefferentia) from the testes to the caput epididymis, the single, thin ductus deferens (vas deferens,Wolffian duct) from the cauda epididymis to the urethra and the thick gubernaculums, which extendscaudally from the cauda epididymis to the inguinal ring. In rabbits, the testes are in the lowerquadrant; they are slightly oval, and their superficial blood vessels can be seen.Ovaries are small and elongated. Their location in rodents is high in the pelvis, just inferior tothe kidneys, and they are cupped by the funnel-shaped ends of the oviducts (uterine tubes, fallopiantubes). The bicornuate (“two horned”) uterus is continuous with the proximal ends of the oviducts.For females, in addition to the ovaries, oviducts, and uterus, the cranial suspensory ligament canbe seen (with magnification) extending from each ovary anteriorly to the diaphragm along thedorsal body wall. Rabbit ovaries are pinkish and elongated but, like the testes, are lower in theabdomen than is the case for rodents. Once the visceral examination is concluded, the viscera areremoved in toto, and the eviscerated carcass is prepared for skeletal staining.e. Visceral Examination—Wilson’s TechniqueAn alternative method for visceral examination of rodent fetuses involves fixation and decalcificationin Bouin’s solution of fetuses selected for visceral examination. 17, 69–74 Each fetus is then seriallycross-sectioned through the head (see head examination, below) and trunk regions by freehandcutting with a razor or scalpel blade. The advantages to this technique include examination of thefetuses at the convenience of the technical staff (since the fetuses are fixed) and retention of sectionsfor documentation or subsequent histologic confirmation of a lesion. The disadvantages include:(1) inability to examine the same fetus for visceral and skeletal alterations (since Bouin’s fixativedecalcifies the skeleton and the serial sections preclude such examination), 75 (2) inability to usecolor changes in fresh specimens and/or blood flow through the heart and great vessels as aids indetection and diagnosis of circulatory alterations, (3) difficulty in sectioning each fetus in a litterand every litter in precisely the same way (e.g., through the heart) to assure comparability ofsections, and (4) difficulty in visualizing morphological alteration from serial cross-sections. Acomparison of Staples’s versus Wilson’s technique in rat fetuses indicated that Staples’s techniqueidentified more heart and great vessel malformations than Wilson’s technique did. 76 An alternativemicrodissection technique after fixation has been presented by Barrow and Taylor. 77Surface staining of Wilson’s sections can enhance detection of visceral alterations. 78 SinceBouin’s fixative contains picric acid, which is reactive, unstable, explosive, and also a strong irritantand allergen (and therefore a safety and health hazard), and Bouin’s fixative can result in fragilefriable soft tissues, alternative fixatives that do not contain picric acid have been suggested. Theuse of a modified Davidson’s fixative for fetuses has been reported 79 (and for fixation of testes andeyes 80 ), with subsequent serial sectioning of fetuses by Wilson’s technique. The authors 79 reportthat side-by-side comparisons of results with the two fixatives indicate Davidson’s fixative producessuperior contrast and definition of organs and vessels in the cranial, thoracic, and abdominal regions,with the tissues remaining moist for a longer time. The procedure reported is as follows. Fetusesare immersed in modified Davidson’s fixative (14 ml ethanol, 6.25 ml glacial acetic acid, 37.5 mlsaturated commercial grade 37% formaldehyde, and 42.25 ml distilled water for 100 ml of fixative)© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 241for 1 week, rinsed twice with tap water, and stored in 70% isopropyl alcohol. Another decalcifyingfixative that does not contain picric acid is Harrison’s, with similar benefits to Davidson’s. Mostlaboratories utilize the Staples’s technique 68 or modifications of the Staples’s technique 77,81 forvisceral examination of rodent fetuses. Wilson’s technique cannot be used for rabbit fetuses sincecurrent guidelines require both visceral and skeletal examination of all fetuses of that species.f. Craniofacial ExaminationPrior to or after visceral examination (for each fetus scheduled for craniofacial examination), thehead is removed and placed in Bouin’s solution or another decalcifying fixative (see above) forfixation and decalcification. Rodent heads may be put individually in a scintillation vial approximatelythree-quarters full of fixative. Because rabbits have identification numbers on top of theirheads, the heads from an entire litter may be put in a large container for fixation. (Some laboratoriesperform only a single midcoronal section on rabbit heads at sacrifice; this provides information oneye structure and on the status of the lateral ventricles of the cerebrum [e.g., is there hydrocephaly?]but does not provide information on other areas of the fore-, mid-, and hindbrain.) Rodent headsshould remain in the fixative for approximately 72 h prior to examination, and rabbit heads shouldremain for about a week. The following recipe may be used to prepare 6 L of Bouin’s fixative:Saturated picric acid (57.13 g in 4200 ml distilled water)37% Formaldehyde (1428 ml)Glacial acetic acid (286 ml)The equipment necessary for craniofacial evaluation includes a dissecting microscope, scalpelsor razor blades, and forceps. The head is removed from the bottle and blotted dry. Up to seven cutsmay be made using a sharp, clean blade. They should always be made in the sequence listed, butadditional cuts may be added if a structure appears abnormal, is missed, or must be verified asmissing. The cuts should be smooth and perpendicular to the cutting surface. The following descriptionsand the head cuts are modified from Wilson 17,69 and van Julsingha and Bennett 82 (Figure 7.7).Before the heads are cut, they are examined for any grossly apparent abnormalities that shouldbe more carefully explored as the cutting proceeds. The first cut is made with the head turned noseupward (use a pair of blunt forceps to grip the head) and exposes the tongue, palate, upper lips,and lower jaw. It is a ventrodorsal section (i.e., horizontal section) beginning at the mouth andcoursing immediately inferior to the ears. The tongue should be lifted from the palate after the cut ismade, and the palate and upper lips examined for incomplete closure (clefting). The pattern of therugae should be examined for correct closure of the palate (the rugae should not be misaligned wherethey meet in the midline). The nasopharyngeal opening, the cochlea of the ears, and the base of thebrainstem (medulla oblongata) may also be visualized. The flat surface produced by this cut simplifiesthe remaining slices by stabilizing the head, which should now be turned over (anterodorsal side up).The second cut is made about half way between the tip of the nose and the foremost corner ofthe eye slits. The nasal passages, nasal conchae, nasal septum, palate, and insertions for the vibrissaeshould be visible on either side of the cut.The third cut is made through the eyes. Tooth primordia and the Harderian glands may bevisible, in addition to the nasal septum, nasal passages, nasal conchae, and palate. Both eyes,including cornea, lens, and retina, are visible in cross-section. The cerebral hemispheres and theanterior-most portions of the two lateral ventricles are readily visualized, as are the mandibularrami on either side. Occasionally, the interventricular foramen may be bisected, and a small portionof the thalamus and third ventricle will be visible. Other structures that should be located in thissection include the optic nerves and/or the optic chiasma, the nasopharynx, the soft palate, andcranial nerves V and VII.© 2006 by Taylor & Francis Group, LLC

242 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION432Brain olfactory labesVitreous chamberwith lensCuts and faces ofsections shown below1ComeaPalate3Nasal sinusRetina1CerebralhemispheresLateralventriclesNasal sinusesThirdventricleNasal seplum42PalateFigure 7.7Diagram of five sections typically observed during examination of a normal term (GD 20) fetalrodent head.The fourth cut is made through the largest vertical diameter of the cranium, well in front ofthe ear flaps. The thalamus, surrounding the slitlike third ventricle, will occupy a large central area,and the left and right cerebral hemispheres should surround it dorsolaterally. The lateral ventriclesand cerebellum can be visualized, as can the pons and medulla oblongata (brainstem). After thiscut, the final posterior section of the brain should be lifted from the cranial vault for a superficialobservation of the cerebellum and surrounding structures.After examination, the fetal head sections (rodents) may be stored in labeled cassettes (e.g.,Tissue-Tek Uni-Cassette, Miles, Inc., Diagnostics Division, Elkhart, IN) and sealed in plastic bags(one per litter), with 70% ethanol as a preservative. Rabbit head sections from each fetus may besealed in an individual, plastic, heat-sealable bag with 70% ethanol. All the bags from a litter canbe stored in a single larger bag.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 243g. Skeletal ExaminationCurrently, the remaining 50% of the litter for rodents (intact, not decapitated) and 100% of thelitter for rabbits are processed and examined for skeletal alterations. Fetal carcasses are evisceratedand prepared for staining with alizarin red S (for ossified bone) 83 or “double stained” with alizarinred S and alcian blue (for ossified and cartilaginous bone plus other permanent cartilaginousstructures, such as the tip of the nose, external pinnae, etc.). 84,85 New guidelines require evaluationof both cartilaginous and ossified skeletal components but do not specify how (double staining isthe preferred method and strongly recommended).1) Single Staining for Ossified Bone — Two and one-half days is the minimum time requiredto hand stain rodent fetuses for skeletal evaluation. However, commercially available automaticstainers (e.g., the Sakura Teratology Processor from Sakura Finetek, USA, Inc., Torrance, CA)allow very rapid single or double staining of fetuses, with appropriate documentation for GLPcompliance.Once fetuses are eviscerated, they are immersed in 70% ethanol at least overnight in preparationfor staining. 86,87 The ethanol is drained off the next morning and replaced with 1% potassiumhydroxide (KOH) solution containing alizarin red S (25 mg/liter of solution) for 24 h. After 24 h,this is poured off and replaced with a fresh 1% KOH solution (without alizarin). 88 Six hours later,the 1% KOH is replaced with 2:2:1 solution (2 parts 70% ethanol to 2 parts glycerin to 1 partbenzyl alcohol). The next morning the 2:2:1 solution is decanted and replaced with a 1:1 solution(1 part 70% ethanol to 1 part glycerin). The skeletons can remain in this solution indefinitely forevaluation and storage.Seven to eight days is the minimum time required to process a rabbit fetus for skeletalexamination. Once the fetus has been eviscerated, it is skinned, placed in a partitioned plastic box,and either air dried overnight or placed in 70% ethanol. The following morning, the ethanol isdrained off and replaced with KOH solution containing alizarin red S as described above. Fortyeighthours later, this staining solution is decanted and replaced with a fresh 1% KOH solution.After up to 4 d, depending on the size of the fetus, the 1% KOH is replaced with the abovementioned2:2:1 solution. The next morning the 2:2:1 solution is decanted and replaced with 1:170% ethanol to glycerin for evaluation and storage.2) Double Staining for Ossified Bone and Cartilaginous Tissue — Rodents — Fetusesshould be skinned (or the skin should be separated from the underlying tissue at least) afterevisceration to allow the alcian blue to penetrate. The fetus is immersed in a 70°C water bath forapproximately 7 s (the fetus will appear to “unfold”). The epidermis is then peeled off the body,paws, and tail. This is best done by a gentle rubbing between the thumb and fingers, as the skeletonat this point is extremely fragile. The skeleton is placed in 95% ethanol overnight. The followingmorning, the ethanol is drained and replaced with alcian blue stain (15 mg alcian blue to 80 ml95% ethanol to 20 ml glacial acetic acid). This is decanted after 24 h and replaced with 95% ethanolfor 24 h. The following morning, the 95% ethanol is decanted, and the skeletons are placed inalizarin red S staining solution (25 mg/l of 1% KOH) for 24 to 48 h. The stain is drained andreplaced with 0.5% KOH for 24 h; this step may not be necessary for small specimens. After the0.5% KOH solution is decanted, it is replaced with 2:2:1 solution (70% ethanol to glycerin tobenzyl alcohol). The following morning the fetal skeletons are placed in 1:1 70% ethanol to glycerinfor evaluation and storage. 89,90Rabbits — The process is the same as for staining with alizarin red S, with the following alterations.After skinning, the fetal skeletons are placed in 95% ethanol. The following morning the ethanolis decanted and replaced with the alcian blue stain (see rodent staining for formulation). Thespecimens remain in the alcian blue for approximately 24 h and are then placed in 95% ethanol© 2006 by Taylor & Francis Group, LLC

244 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONfor 24 h. The following morning, the fetal skeletons are placed in the alizarin red S solution, andthe remainder of the process is as outlined above for single staining of rabbit fetuses.h. Skeletal EvaluationThe following description of a fetal skeletal exam is based on using a double-stained preparation.In a double-stained skeleton, the ossified bone will be stained red to purple and the cartilage blue.The staining of the cartilage allows the examiner to ascertain whether the underlying structure ispresent (i.e., is the bone merely not yet ossified, although the cartilage anlage is present, or is theunderlying cartilage actually missing?). Similarly, a cleft in the cartilage plate of the sternumexplains bipartite ossification sites of the sternebrae, and a misalignment of the sternebrae may beseen by the abnormal fusion of the cartilage plate. The normalcy of the cervical vertebrae can onlybe determined in a double-stained specimen, as these are not normally ossified in the term fetalskeletal preparation.It is critical that the examiner be familiar with the normal appearance of bone structures in thefetus and with the degree of ossification that should be evident on the sacrifice date. Accurateskeletal descriptions enable the investigator to pinpoint accelerated or retarded ossification. 90–94The skeletal system consists of axial and appendicular sections. The axial skeleton includes theskull, vertebral column, sternum, and ribs. The pelvic and pectoral girdles and the appendagescomprise the appendicular skeleton.The paired bones of the skull (Figure 7.8), which must be identified during the skeletalexamination, are described as follows. The exoccipitals and supraoccipital form the posterior wallof the cranial cavity. The supraoccipital is fused in rats on GD 19 and on GD 28 in rabbits. However,this bone has paired aspects in GD 18 mice. The interparietal forms the posterior portion of thecranial roof, and the more anterior parietals form part of the roof and sides of the cranial cavity.Frontals and nasals cover the anterior portion of the brain. The premaxillae, maxillae, zygomatics,and squamosals are the bones of the face and the upper jaw. The zygomatic bone, squamosal bone,and zygomatic process of the maxilla combine to form the zygomatic arch. The mandibles (unfusedmedially at this stage of development) form the lower jaw. An incisor originates from each mandiblein the rabbit skeleton.Usually, six ossified sternebrae are present (Figure 7.9). In rodents, sternebra 5 ossifies last,number 6 second to last, and number 2 (and/or 4) third to last. The sternebrae fuse to form thesternum or breastbone. The sternum consists of a cartilage plate, with the distal cartilage portionof the first seven ribs fused to it. The sternum is formed from two symmetrical halves that fuseduring development. Incomplete fusion will result in a sternal cleft or perforation of the cartilage,as well as misalignment of the ossified sternebrae.Vertebrae are the basic structural units of the vertebral column (Figure 7.10 and Figure 7.11).Each vertebra consists of a ventral centrum and paired lateral arches. The vertebral column articulatesanteriorly with the exoccipital bones of the skull. The vertebrae are divided into the followinggroups: cervical (7 vertebrae), thoracic (typically 12 vertebrae in rabbits, 13 in rodents), lumbar (7vertebrae in rabbits, 6 in rodents), sacral (4 vertebrae), and caudal (number of vertebrae varies withtail length).The centrum consists of an ossified center surrounded by cartilage. The caudal centra are entirelycartilage. The examiner should note unossified centra (in a normal GD 20 rat, thoracic centra I toVII are normally unossified; all lumbar and sacral centra should be ossified) and misaligned,bipartite, dumbbell, or unilateral ossification centers. If the cartilage anlage is affected, it shouldbe recorded along with any findings for the ossified portion. The ossified portion of the centra mayor may not be bipartite or dumbbell when the cartilage portion is bipartite or dumbbell. The centraare also examined for fusion of the cartilage between centra. The vertebral arch is almost completelyossified except for a cartilage tip, which may or may not be present on the transverse process,© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 245Figure 7.8Dorsal and ventral views of the skull from a normal term (GD 20) rat fetus.depending on the relative maturity of the fetus. The spinous process is entirely cartilaginous andcan be examined only from the dorsal side. The cartilage of the transverse processes of sacral centraI, II, and III is normally fused, providing extra support for the pelvis. The two most anterior cervicalvertebrae are the atlas and the axis. Each thoracic vertebra articulates dorsally with a pair of ribs.Rabbits usually have 12 pairs of ribs, whereas rodents usually have 13 pairs (Figure 7.12). Eachspecies often has a full rib, rudimentary rib, or an ossification site at lumbar I. In rodents, this mayonly be unilateral, but in rabbits, it is usually bilateral. Additional rib structures are typicallyclassified as extra, full, or supernumerary ribs if they are at least one-half or greater than the lengthof rib I or XIII; as rudimentary ribs if they are less than one-half the length of rib I or XIII; or asossification sites if they are small, round dots. Each rib consists of a cartilage tip proximal to thevertebral arch, an ossified middle portion, and a distal cartilage portion. The distal cartilages of© 2006 by Taylor & Francis Group, LLC

246 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSternebra ISternebra IISternebra IIISternebra IVSternebra VSternebra VIOssified areas are open, cartilage is shaded.Figure 7.9Sternum of a normal term (GD 20) rat fetus.ribs I to VII are attached to the sternum, and ribs VIII to XI curve upward, with their tips in closeproximity. Ribs XII and XIII are free distally. Extra ribs may also be found on cervical arch VIIand in rats are sometimes found only as cartilage; this extra cartilage is sometimes found to befused to rib I of the same side.Paired dorsal scapulae and ventral clavicles comprise the pectoral girdle (Figure 7.10). Thescapulae are flat, trapezoidal bones with anterior-dorsal projections; the clavicles are elongated,© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 247Figure 7.10View of axial and appendicular skeleton of a normal term (GD 20) rat fetus (decapitated).curving, slender bones that articulate laterally with the scapulae and medially with the sternum(postnatally). The forelimb skeleton consists of the humerus in the upper foreleg, the radius andulna in the lower foreleg, and carpals, metacarpals, and phalanges in the forefoot (Figure 7.13). Inrodents, the only bones of the forefoot that have ossified by the time of sacrifice on GD 20 are themetacarpals, located between the carpals and phalanges (by GD 21, phalanges are visible). In mice,usually only metacarpals II to IV are ossified at term; in rats, metacarpals II to V are normallyossified at term.Three pairs of bones are the basic units of the fetal pelvic girdle (Figure 7.14). Individuallythey are the ilium, ischium, and pubis. The femur, patella, tibia, and fibula are the bones of thethigh and hindleg, and the tarsals, metatarsals, and phalanges make up the skeletal support of thehindfoot. The five metatarsals ossify first. The hindfoot of the rabbit has only four digits.If data are being collected by hand, rather than by use of computer software, there is the optionof a checklist of bones that must be checked off as normal or any deviations recorded. The otheroption is recording any deviations from normal — any and all developmental delays that are usually© 2006 by Taylor & Francis Group, LLC

248 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFRONT Centrum BACKTransverseprocessVertebralarchFRONTSpinous processSacralIIIIlliumIIIOssified areas are open, cartilage is shaded.Thoracic centraCaudal centraFigure 7.11Enlarged views of selected elements of the axial skeleton of a term (GD 20) rat fetus.manifested as lack of ossification of sternebrae (V, VI, and/or II in rats, VI in mice) and as bipartiteor dumbbell-shaped cartilage or unossified centra (in rats). There is less variability in the developmentof the skeleton if the sacrifice day is GD 21 for rats and GD 18 for mice, but the risk ofdelivery prior to scheduled sacrifice is increased.During the fetal examinations, all unusual (notable) observations are recorded or described bythe examining technician. This is done for each individual fetus on the appropriate data collectionsheet or in the appropriate locations on the computer terminal display. The observations willsubsequently be coded as malformations (M) or variations (V) by a designated staff member (withinput from the study director, laboratory supervisor, veterinarian, pathologist, etc., as appropriate),based on previously established criteria and designations (usually defined by the study director andlaboratory supervisor). The determination of M or V status for each observation is made in thecontext of three salient factors:• In the test animal and human literature on malformations, there is no case to date where ateratogenic agent causes a new, “never before seen” malformation. What is detected is an increasein the incidence of the malformation(s) above those seen in the general population. The current© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 249Figure 7.12Detailed view of ribs from a term (GD 20) rat fetus.Figure 7.13Detailed view of forepaw and hindpaw skeletal elements from a term (GD 20) rat fetus.view is that teratogenic agents act on susceptible genetic loci or on susceptible developmentalevents. Therefore, the response seen is influenced by the genetic background and will vary byspecies, strain, race (in the case of humans), and individual (the last is more relevant to outbredstrains and/or to genetically heterogeneous populations than to inbred strains).• There is genetic predisposition to certain malformations that characterizes specific species, strains,races, and individuals. Historical control data are indispensable (along with concurrent controls)for determining the designation and occurrence of the present finding(s) in the context of the© 2006 by Taylor & Francis Group, LLC

250 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIlliumFemurPubisTibiaFibulaMetatarsalsIshiumFigure 7.14Detailed view of the pelvic and hind limb skeletal elements from a term (GD 20) rat fetus.background “noise” of the population on test or at risk. For comparison purposes, the most recentMARTA/MTA compiled control data from several strains of rats, mice, and rabbits can be accessedat the following URL: http://hcd.org/search/abnormality.asp. Additional historical control databasesor teratogenicity studies with control data exist for a number of commonly employed test animals,for example: CD ® rat; 95–98 CD-1 ® mouse; 96,99,100 F-344 rat; 101–116 B6C3F1 mouse; 104,117 and NewZealand White rabbit. 118–124• The general considerations for designation of a finding as an M or a V are imprecise, may varyfrom study to study and teratologist to teratologist, may be relatively arbitrary, and are notnecessarily generally accepted. 125The following sections present general classification criteria employed in the authors’ laboratoryto aid in the appropriate categorization of experimental findings.i. Malformations• Incompatible with or severely detrimental to postnatal survival, e.g., ventricular septal defect(VSD), diaphragmatic hernia, exencephaly, anencephaly, spina bifida, cleft palate• Irreversible without intervention, e.g., hydroureter, hydronephrosis, VSD• Involve replication, reduction (if extreme), or absence of essential structure(s) or organs, e.g.,limbs, organs, major blood vessels, brain, heart chambers or valves, skeletal components unossifiedin an abnormal pattern• Result from partial or complete failure to migrate, close, or fuse, e.g., cleft palate, cleft lip, ectopicorgans, facial clefts, renal agenesis, lung agenesis, open neural tube• May include syndromes of otherwise minor anomalies• Exhibit a concentration- or dose-dependent increased incidence across dose groups, with a quantitativeand/or qualitative change across dose groups, e.g., meningocoele → meningomyelocoele→ meningoencephalocoele → exencephaly; foreshortened face → facial cleft → facial atresia;short tail → no tail→anal atresia; short rib → missing rib; brachydactyly (short digits) → oligodactyly(absence of some digits) → adactyly (absence of all digits); missing distal limb bones(hemimelia) → missing distal and some proximal limb bones → amelia (missing limb)© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 251j. Transitional FindingsThese may be upgraded to “malformation” or downgraded to “variation” status, depending onseverity and/or frequency of occurrence.• Nonlethal and generally not detrimental to postnatal survival• Generally irreversible• Frequently may involve reduction or absence of nonessential structures, e.g., innominate artery• Frequently may involve reduction in number or size (if extreme) of nonessential structures or mayinvolve their absence• Exhibit a dose-dependent increased incidencek. Variations• Nonlethal and not detrimental to postnatal survival• Generally reversible or transitory, such as wavy rib or the reduced ossification in a cephalocaudalsequence frequently seen associated with immaturity or delayed development as result of toxicity,e.g., reduced ossification in fore- and hindpaws, caudal vertebrae, pubis (but usually not ilium orischium), skull plates, sternebrae (especially 5, 6, 2, or 4, in that order), cervical centra (especially1) 126• May occur with a high frequency and/or not exhibit a dose-related increased incidence, e.g., reducedrenal papilla and/or distended ureter in CD ® rat fetuses at term, 127 dilated lateral cerebral ventricles(in rodents), extra ribs (on lumbar vertebra I) in rat and rabbit fetuses• Detectable change (if not extreme) in size of specific structures (subjective), e.g., renal papillaone-half to one-quarter normal width; spleen greater than 1.5 times normal; kidney less than onehalfnormal size, kidney two times normal size; liver lobe less than one-half or up to 2 timesnormal size.A profile of recent historical control data from the authors’ laboratory of developmental toxicitystudies on CD ® rats, CD-1 ® mice, and New Zealand White rabbits is presented in Table 7.6. Thesetest animal species exhibit characteristics that maximize their usefulness for such studies, such as:• High pregnancy rate with very low levels of spontaneous full litter resorptions• Large litters (total implants) with low preimplantation loss• Large live litters with low postimplantation loss• Approximately equal sex ratios in fetuses• Consistent fetal body weights• Low malformation rates• Reasonable variation rates (high enough to provide sensitivity to test agents that increase theincidence of variations and/or exacerbate the variations to malformations; low enough to allowdetection of treatment-related increases)The interpretation of fetal findings requires a “weight of the evidence” approach (examplesfollow). If the incidence of fetal malformations is increased at the middle dose but not at the highdose, but the in utero mortality is highest at the high dose, then it is likely that the most affectedconceptuses at the high dose died, especially with the prolonged dosing period specified in the newguidelines. In this case, the lack of a dose-response pattern is spurious, and an umbrella parameterlike affected implants (nonlive plus malformed) will indicate the real dose-response pattern.In the presence of reduced fetal body weights, it is typical to observe increased incidences ofreduced ossification, especially in the skeletal areas that ossify last in utero, such as sternebrae V andVI of the prenatal sternum, bipartite or dumbbell ossification sites in the thoracic centra, and anteriorand posterior phalangeal segments of the paws. 58 Reduced fetal body weights are commonly associatedwith dilated renal pelves (a delay in growth of the renal papilla; 98 and/or dilated (enlarged) lateralventricles of the cerebrum without compression of the cerebral walls (again a delay in growth). 57© 2006 by Taylor & Francis Group, LLC

252 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONConversely, an increase in the incidence of an ossification site, partial or full rib on lumbar 1(i.e., a 14th rib in rodents or a 13th rib in rabbits), considered a variation in the authors’ laboratory,may indicate an alteration in developmental patterning (e.g., Hox genes), genetic drift over time,and/or may predict possible greater effects at higher doses. 125 In the authors’ laboratory, cervicalribs (ossification site, partial, or full ribs) are considered malformations because of their rarity inhistorical controls and the possibility that they may have postnatal consequences, including alterationof blood flow to the head. Biological plausibility must also be considered. Kimmel andWilson 125 conclude that “the best basis for rational interpretation of such anatomical variants iscumulative (historical) data on untreated controls, vehicle-treated controls, and various experimentalgroups in the strain or stock of animals used in a given teratogenicity study.”F. StatisticsAs part of protocol development, the choice of statistical analyses should be made a priori, althoughspecific additional analyses may be appropriate once the data are collected. The unit of comparisonis the pregnant female or the litter and not the individual fetus, as only the dams are independentlyand randomly sorted into dose groups. 128 The fetus is not an independent unit and cannot berandomly distributed among groups. Intralitter interactions are common for a number of parameters(e.g., fetal weight or malformation incidence). Two types of data are collected: (1) ordinal/discretedata, which are essentially present or absent (yes or no), such as incidences of maternal deaths,abortions, early deliveries, or clinical signs, and incidences of fetal malformations or variations;and (2) continuous data, such as maternal body weights, weight changes, food and/or waterconsumption, organ weights (absolute or relative to body or brain weight), and fetal body weightsper litter. For both kinds of data, three types of statistical analyses are performed. Tests for trendsare available and appropriate to identify treatment-related changes in the direction of the data(increases or decreases), overall tests are performed for detecting significant differences amonggroups, and specific pairwise comparisons (when the overall test is significant) are made to theconcurrent vehicle control group values. Pairwise comparisons are critical to identify statisticallysignificant effects at a given dose relative to the concurrent vehicle control group. Continuous dataare designated parametric (requiring “normally distributed” data) or nonparametric (“distributionfree”), with different specific tests employed for the three types of statistical analyses, dependingon whether the data are parametric or nonparametric. (See Chapter 17 for a more completediscussion of statistical approaches and appropriate tests.)G. Data CollectionHistorically, data have been collected by hand and subsequently entered into various softwaresystems for statistical analyses and summarization (the latter was also typically done by hand insome laboratories). Summary tables were then constructed by hand or word processing. Currently,most laboratories are using (or acquiring) automated online, real-time data collection systems.These systems are either constructed in-house or purchased as a turnkey package available from anumber of vendors (e.g., Provantis ® (Instem), ISIS ® BioComp, PathTox ® , WIL ® , TopCat, Teros ® ,etc.). In-house systems can be designed to meet the specific needs and procedures of the performinglaboratory; purchased systems are standardized (with modifications available at additional cost).Regardless of the data collection method and analysis, the requirements under GLPs must besatisfied. Also, various regulatory agencies have provided guidance on how to purchase, validate,and acceptance test the automated systems (see Chapter 18 for details on GLPs). The advantageswith automated systems include: more efficient use of technical staff, no need for back-enteringdata, less errors on data collection, elimination of paperwork, ease of QA auditing, and thepossibility of electronically submitting the finished report to the appropriate cognizant federal orinternational agency (see Section V of this chapter for a discussion of electronic record keeping© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 253and reporting). One note: an audit trail of changes to a dataset is, if anything, even more importantfor electronic data capture than for hand-collected data. The audit trail (required, permanent, andaccessible by appropriate staff) must contain the “who, what, when, and why” of each change.IV. STORAGE OF RECORDSAll original study records, including all original data sheets, should be bound and stored in thesecure archives of the performing laboratory, under the control of the quality assurance officer. Anybiological samples collected during the course of the study (e.g., slides, blocks, wet tissues, fetalskeletons, sectioned fetal heads, dead fetuses, maternal gross lesions, etc.) should be placed insecure storage in the archives. Work sheets and computer printouts generated in the statisticalanalysis of data should also be stored in the archives. Copies of the final study report should befiled with the contract laboratory as well as with the sponsor. All study records, data, and reportsshould be maintained in storage in accordance with the appropriate federal guidelines, e.g., for theregistration lifetime of the pesticide (FIFRA), storage is for 10 years after submission of the finalreport (FDA, EPA, TSCA), and in accordance with the appropriate governmental GLPs, e.g., EPA(FIFRA), 26 EPA (TSCA), 27 or FDA. 25 These records and samples may be released to the sponsorupon written request. See Table 7.10 for a list of typical study records to be maintained, and forfurther details regarding GLP compliance, refer to Chapter 18.Table 7.10Study records to be maintainedProtocol and any amendmentsList of standard operating proceduresAnimal requisition and receipt recordsQuarantine recordsTemperature and humidity records for the treatment room(s)Animal research facility room log(s)Deionized water analysis (if appropriate)Feed type, source, lot number, dates used, certification, contaminantsDose code records containing rx code (if used), color code (if used), and concentrationsMating/insemination recordsRandomization recordsAssignment to study recordsBulk chemical receipt, storage, and use recordsDose formulation recordsAnalytical chemistry report(s)Disposal of archival dose formulation samplesPoststudy shipment of bulk chemical to supplierDosing recordsClinical observationsMaternal body weightsMaternal food and/or water consumptionNecropsy sheets (in the event of maternal mortality)Teratology sacrifice records: results of external, visceral, and skeletal examination sheetsFetal carcasses and head sectionsWet tissues, blocks, and slides (if appropriate)Computer printoutsPhotographs (if taken)CorrespondenceAny deviations from the protocolDraft reportFinal report© 2006 by Taylor & Francis Group, LLC

254 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONV. COMPLIANCE WITH APPROPRIATE GOVERNMENTALAND AAALAC INTERNATIONAL REGULATIONSThe performing laboratory must be operated in compliance with the appropriate governmental GLPs(see above). The animal research facility should be accredited by AAALAC International (Associationfor Assessment and Accreditation of Laboratory Animal Care International). Thus, the study should beconducted in compliance with appropriate GLPs and in compliance with the appropriate testing guidelines(e.g., FIFRA; 3,4,56 TSCA, 2,10 FDA, 55,129 ) and AAALAC International accreditation standards. Itmay be useful if the performing laboratory is also approved by the Ministry of Agriculture, Forestryand Fisheries (MAFF), Japan, for toxicology studies on agricultural chemicals, so that work performedunder EPA (FIFRA) GLPs and OPPTS or OECD testing guidelines will be acceptable to the Japaneseregulatory agencies. The quality assurance unit of the performing laboratory must review the protocoland any amendments; inspect all in-life phases, necropsy, and fetal evaluations; audit the raw andsummarized data and the final report; and provide a compliance statement to that effect for the finalreport. For further details regarding GLP compliance, refer to Chapter 18.With the increasing use of automated data collection systems, the possibility and usefulness ofsubmitting the final report in toto electronically to the appropriate federal or international agencyhas been recognized. To date, the FDA 130 has promulgated regulations for electronic records andelectronic signatures (known affectionately as CFR, Part II), and the EPA has published 131 aproposed rule on establishment of electronic reporting and electronic records (known asCROMERRR, Cross Media Electronic Reporting and Recordkeeping Rule).The objective of both the FDA and EPA initiatives is to “set forth the criteria under which theagency considers electronic records, electronic signatures, and handwritten signatures executed toelectronic records to be trustworthy, reliable, and generally equivalent to paper records as handwrittensignatures executed on paper.” 130 The EPA proposal rule 131 would “allow electronic reportingto EPA by permitting the use of electronic document recovery systems to receive electronicdocuments in satisfaction of certain document submission requirements in EPA’s regulations.” TheEPA acknowledges that “the electronic records criteria in [their] rule are not as detailed as thatcontained in FDA’s 21 CFR Part II.” 131 Laboratories providing studies for submission to the EPAor FDA should read the referenced documents in their entirety. One concern the authors of thischapter have is whether the electronic submissions would be limited to “read only” or would haveto be sent in a format that allowed manipulation by regulators (as is apparently one objective ofthe FDA code). The integrity of the text and data submitted by the performing laboratory must bemaintained. The study director will have no control over any alterations or manipulations to thedata (e.g., deletions of an animal’s data, a different interpretation of results, etc.) postsubmission.A. Status ReportsVI. REPORTSStatus reports should be provided to the sponsor at a frequency specified by the sponsor. The contentand format of these reports should remain flexible so as to accommodate unforeseen situations.B. Final ReportThe final report should include an abstract, objectives, materials and methods, results (narrativeand summary tables), discussion, conclusions, references, any deviations from the protocol, a GLPcompliance statement, a QA statement, and appendixes. The appendixes should include analyticalchemistry reports (if appropriate), data from individual mated females, individual embryo and fetaldata by litter and uterine location, histopathology and clinical chemistry reports (if appropriate),© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 255and the study protocol with any amendments. If the test material is a pesticide, the report mustinclude the Statement of (No) Data Confidentiality (page 2 of report), the GLP compliance statement(page 3), and the FIFRA Flagging Statement (page 4), as well as the QA statement.Data in the final report should be summarized in tabular form, showing for each dose group:numbers of animals at the start of test, pregnant animals, ovarian corpora lutea, and total uterineimplantations; numbers and percentages of live fetuses; pre- and postimplantation loss; numbersof fetuses and litters with any external, visceral, or skeletal abnormalities; percentage of fetusesper litter with malformations and variations, by sex per litter (if possible). The findings should beevaluated in terms of the observed effects and the exposure levels (doses) producing effects. Inaddition, specific information on toxic responses should be reported by dose, species and strainemployed, date of onset and duration of each abnormal sign and its subsequent course, foodconsumption, body weight and weight changes, uterine weight data, and fetal findings (number oflive/dead, resorptions, fetal body weight, sex, and external, visceral, and skeletal alterations [malformations,variations]). Historical developmental toxicity control data, both published and fromthe performing laboratory, should also be considered, if appropriate. It is anticipated that a NOAELwould be identified for both maternal and developmental toxicities.The final report for this study, in draft form, should be submitted to the sponsor within a certaininterval (e.g., 3 months of the last sacrifice date). Within a specified interval (e.g., 30 days) ofreceipt of any comments from the sponsor’s representative, at least one copy of the fully signedfinal report should be submitted to the sponsor. If the test material is evaluated under FIFRA testingguidelines, the final report must be in FIFRA format, according to EPA PR Notice 86-5.VII. PERSONNELAll senior staff involved in the study (“key personnel”) should be identified by name and responsibility.Technical support staff are listed by name and responsibility or title in the authors’laboratory, but that is not required. The list should include:• Study director (required)• Sponsor’s representative or study monitor (if appropriate)• Animal research facility veterinarian and staff supervisor/manager• Developmental toxicology laboratory staff supervisor/manager• Data analyst• Chemical handling supervisor (chemical receipt, disbursement, return to sponsor)• Dosing formulation supervisor• Supervisor for analytical chemistry (dosing formulations)• Supervisor for clinical chemistry (if appropriate)• Histology supervisor (if appropriate)• Veterinary pathologist (if appropriate)• Supervisor/manager of quality assurance unitIt is always useful to add a statement to the study protocol (prior to its initial submission tothe sponsor when the study is initially commissioned) such as “Additional team members, if any,will be documented in the study records and listed in the final report.” Otherwise, the protocol willhave to be amended if new staff are added (it will also have to be amended if identified staff leaveor do not work on the study).VIII. STUDY RECORDS TO BE MAINTAINEDA typical list of study records to be maintained is presented in Table 7.10. Most of the entries areself-explanatory. The listing is there to provide a listing of all SOPs used by title and effective date© 2006 by Taylor & Francis Group, LLC

256 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(at least) or an actual copy of each SOP. Since SOPs evolve (new ones are written, old ones areretired, current ones are revised or amended), it is necessary to provide the specific SOPs that werein use at the time the study was performed.ACKNOWLEDGMENTSThe authors wish to acknowledge with profound appreciation the dedication, expertise, and efforts ofDr. Tyl’s staff at the University of Connecticut, Chemical Industry Institute of Toxicology, Bushy RunResearch Center, and especially at RTI. Special thanks go to C.B. Myers, data specialist at RTI, forsummarizing historical control data for Table 7.3 and to C.A. Winkie at RTI for her expert and patienttyping (and retyping) of this chapter. The authors have learned a great deal and thoroughly enjoyedtheir long association with the science and art, as well as with the practitioners of developmental toxicitytesting and research as the discipline has grown and evolved. 11,114 We cannot imagine a more exciting,rewarding, or relevant way to serve science and improve the human condition.References1. Goldenthal, E.I., Guidelines for Reproduction Studies for Safety Evaluation of Drugs for Human Use,U.S. Food and Drug Administration, Washington, D.C., 1966.2. U.S. Environmental Protection Agency, Toxic Substances Control Act (TSCA) test guidelines: finalrule, Fed. Regist., 50, 39412, 1987.3. U.S. Environmental Protection Agency, Pesticides Assessment Guidelines (FIFRA), Subdivision F.Hazard Evaluation: Human and Domestic Animals (Final Rule), NTIS (PB86-108958), Springfield,VA, 1984.4. U.S. Environmental Protection Agency, Pesticide Assessment Guidelines, Subdivision F Hazard Evaluation:Humans and Domestic Animals, Series 83-3, Rat or Rabbit Developmental Toxicity Study,June, 1986 (NTIS PB86-248184), as amended in Fed. Regist., 53(86), Sect. 158.340, May 4, 1988.5. Ministry of Health and Welfare, Japanese Guidelines of Toxicity Studies, Notification No. 118 of thePharmaceutical Affairs Bureau, 2. Studies of the Effects of Drugs on Reproduction, Yakagyo Jiho Co.,Ltd., Tokyo, Japan, 1984.6. Canada Ministry of Health and Welfare, Health Protection Branch, The Testing of Chemicals forCarcinogenicity, Mutagenicity and Teratogenicity, Ministry of Ottawa, Ottawa, 1973.7. Department of Health and Social Security, United Kingdom, Committee on Safety of Medicines: Notesfor Guidance on Reproduction Studies, Department of Health and Social Security, Great Britain, 1974.8. Organization for Economic Cooperation and Development (OECD), Guideline for Testing of Chemicals:Teratogenicity, Director of Information, OECD, Paris, France, 1981.9. U.S. Food and Drug Administration, International Conference on Harmonisation (ICH), Guidelinefor detection of toxicity to reproduction for medicinal products, Section 4.1.3, study for effects onembryo-fetal development, Fed. Regist., 59(183), 48749, September 22, 1994.10. U.S. Environmental Protection Agency, OPPTS, Health Effects Test Guidelines, OPPTS 870.3700,Prenatal Developmental Toxicity Study, Public Draft, U.S. Government Printing Office, Washington,D.C., February, 1996.11. U.S. Environmental Protection Agency, TSCA, Toxic Substances Control Act Test Guidelines, FinalRule 40 CFR Part 799.9370, TSCA Prenatal Developmental Toxicity, Fed. Regist., 62(158), 43832,August 15, 1997.12. U.S. Food and Drug Administration, Guidelines for Developmental Toxicity Studies, Center for FoodSafety and Applied Nutrition Redbook 2000, IV.C.9.b, July 20, 2000.13. Organization for Economic Cooperation and Development, OECD Guideline for the Testing of Chemicals;Proposal for Updating Guideline 414: Prenatal Developmental Toxicity Study, pp. 1–11, adoptedJanuary 22, 2001.14. U.S. Environmental Protection Agency, Guidelines for the health assessment of suspect developmentaltoxicants, Fed. Regist., 51, 34028, 1986.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 25715. U.S. Environmental Protection Agency, Part V, Guidelines for developmental toxicity risk assessment;notice, Fed. Regist., 56(234), 63798, 1991.16. Wilson, J.G., The evolution of teratology testing, Teratology, 20, 205, 1979.17. Wilson, J.G., Principles of teratology, in Environment and Birth Defects, Academic Press, New York,1973, p. 11.18. Johnson, E.M., Screening for teratogenic potential: Are we asking the proper questions? Teratology,21, 259, 1980.19. Johnson, E.M., Screening for teratogenic hazards: nature of the problems, Ann. Rev. Pharmacol.Toxicology, 21, 417, 1981.20. Johnson, E.M. and Christian, M.S., When is a teratology study not an evaluation of teratogenicity?J. Am. Coll. Toxicol., 3, 431, 1984.21. Marks, T.A., A retrospective appraisal of the ability of animal tests to predict reproductive anddevelopmental toxicity in humans, J. Am. Coll. Toxicol., 10(5), 569, 1991.22. U.S. Environmental Protection Agency, Proposed amendments to the guidelines for the health assessmentof suspect developmental toxicants; request for comments; notice, Fed. Regist., 54, 9386, 1989.23. Daston, G.P., Rehnberg, B F., Carver, B., Rogers, E.H., and Kavlock, R.J., Functional teratogens ofthe rat kidney. I. Colchicine, dinoseb and methyl salicylate, Fundam. Appl. Toxicology, 11, 381, 1988.24. Daston, G.P., Rehnberg, B.F., Carver, B., and Kavlock, R.J., Functional teratogens of the rat kidney.II. Nitrofen and ethylenethiourea, Fundam. Appl. Toxicology, 11, 401, 1988.25. U.S. Food and Drug Administration, Good Laboratory Practice Regulations for Nonclinical LaboratoryStudies, Code of Federal Regulations (CFR), 229, April 1, 1988.26. U.S. Environmental Protection Agency, Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA)good laboratory practice standards; final rule, Fed. Regist., 54, 34051 (40-CFR-792) August 17, 1989.27. U.S. Environmental Protection Agency, Toxic Substances Control Act (TSCA), good laboratorypractice standards; final rule, Fed. Regist., 54, 34033, August 17, 1989.28. Taylor, P., Practical Teratology, Academic Press, New York, 1986.29. Harkness, J.E. and Wagner, J.E., The Biology and Medicine of Rabbits and Rodents, 3rd ed., Lea andFebiger, Philadelphia, 1989.30. American Association for Laboratory Animal Science, Training Manual Series, Vol. 3, for LaboratoryAnimal Technologist, Stark, D.M. and Ostrow, M.E., Eds., AALAS, Cordova, TN, 1991.31. National Research Council, Guide for the Care and Use of Laboratory Animals, Institute of LaboratoryAnimal Resources, Commission on Life Sciences, National Research Council, National AcademyPress, revised 1996.32. National Institutes of Health, Guide for the Care and Use of Laboratory Animals, PHS NIH Pub. No.86-23, revised, U.S. Department of Health and Human Services, Bethesda, MD, 198533. Cheeke, P.R., Nutrition and nutritional diseases, in The Biology of the Laboratory Rabbit, 2nd ed.,Manning, P.J., Ringler, D.H., and Newcomer, C.E., Eds., Academic Press, New York, 1995, p. 321.34. Bellhorn, R.W., Lighting in the animal environment, Lab. Anim. Sci., 30, 440, 1980.35. Besch, E.L., Environmental quality within animal facilities, Lab. Anim. Sci., 30, 385, 1980.36. Wilson, N.E. and Stricker, E.M., Thermal homeostasis in pregnant rats during heat stress, J. Comp.Physiol. Psych., 93, 585, 1979.37. Peterson, E.A., Noise and laboratory animals, Lab. Anim. Sci., 30, 422, 1980.38. Whitten, W.K., Modification of the oestrus cycle of the mouse by external stimuli associated with themale, J. Endocrinol., 13, 399, 1956.39. Hafez, E.S.E., Ed., Reproduction and Breeding Techniques for Laboratory Animals, Lea and Febiger,Philadelphia, 1970.40. Bredderman, P.J., Foote, R.H., and Yassen, A. M., An improved artificial vagina for collecting rabbitsemen, J. Reprod. Fertil., 7, 401, 1974.41. Chernoff, N., Rogers, J.M., and Kavlock, R.J., Review paper. An overview of maternal toxicity andprenatal development: considerations for developmental toxicity hazard assessments. Toxicology, 59,111, 1989.42. Schwetz, B.A. and Tyl, R.W., Consensus workshop on the evaluation of maternal and developmentaltoxicity workgroup iii report: low dose extrapolation and other considerations for risk assessment —models and applications. Teratogen., Carcinogen., Mutagen., 7(3), 321, 1987.43. Kimmel, C.A., Price, C.J., Sadler, B.M., Tyl, R.W., and Gerling, F.S., Comparison of distilled water (DW)and corn oil (CO) vehicle controls from historical teratology study data, Toxicologist, 5, 185, 1985.© 2006 by Taylor & Francis Group, LLC

258 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION44. Feldman, D.B., Simplified gastric intubation in the rabbit, Lab. Anim. Sci. 27, 1037, 1977.45. American Association for Laboratory Animal Science, Training Manual Series, Vol. 1, for AssistantLaboratory Animal Technician, Stark, D.M. and Ostrow, M.E., Eds., AALAS, Cordova, TN, 1991.46. Tyl, R.W., Ballantyne, B., Fisher, L.L., Fait, D.L., Dodd, D.E., Klonne, D.R., Pritts, I.M., and Losco,P.E., Evaluation of the developmental toxicity of ethylene glycol aerosol in CD-1 ® mice by nose-onlyexposure, Fundam. Appl. Toxicol., 27(1), 49, 1995.47. Francis, E.Z. and Kimmel, C.A., Proceedings of the workshop on the acceptability and interpretationof dermal developmental toxicity studies, Teratology, 39, 453, 1989.48. Kimmel, C.A. and Francis, E.Z., Proceedings of the workshop on the acceptability and interpretationof dermal developmental toxicity studies, Fundam. Appl. Toxicol., 14, 386, 1990.49. Tyl, RW., York, R.G., and Schardein, J.L., Reproductive and developmental toxicity studies bycutaneous administration, in Health Risk Assessment through Dermal and Inhalation Exposure andAbsorption of Toxicants, Wang, R.G., Knaak, J.B., and Maibach, H.I., Eds., CRC Press, Boca Raton,FL, 1993, p. 229.50. Tyl, R.W., Fisher, L.C., France, K.A., Garman, R.H., and Ballantyne, B., Evaluation of the teratogenicityof bis(2-dimethylaminoethyl) ether after dermal application in New Zealand white rabbits, J.Toxicol. Cutaneous Ocular Toxicol., 5, 263, 1986.51. Tyl, R.W., Fisher, L.C., Kubena, M.F., Vrbanic, M.A., Gingell, R., Guest, D., Hodgson, J.A., Murphy,S.R., Tyler, T.R., and Astill, B.D., The developmental toxicity of 2-ethylhexanol (2-EH) applieddermally to pregnant Fischer 344 rats, Fundam. Appl. Toxicol., 19, 176, 1992.52. Fisher, L.C., Tyl, R.W., and Kubena, M.F., Cutaneous developmental toxicity study of 2-ethylhexanol(2-EH) in Fischer 344 rats, Teratology, 39(5), 452, 1989.53. Tyl, R.W., Fisher, L.C., Kubena, M.F., Vrbanic, M.A., and Losco, P.E., Assessment of the developmentaltoxicity of ethylene glycol applied cutaneously to CD-1 ® mice, Fundam. Appl. Toxicol., 27, 155, 1995.54. Draize, J.H., Woodard, G., and Calvery, H.O., Methods for the study of irritation and toxicity ofsubstances applied to the skin and mucous membranes, J. Pharmacol. Exp. Therap., 82, 377, 1944.55. U.S. Food and Drug Administration, Toxicological Principles for the Safety Assessment of Direct FoodAdditives and Color Additives Used in Food (“Redbook”), Appendix II, Guidelines for teratogenicitytesting in rat, hamster, mouse, and rabbit, 108, 1982.56. U.S. Environmental Protection Agency, OPP Guideline 83-3, Teratogenicity Study, Pesticide AssessmentGuidelines, Subdivision F, Hazard Evaluation, Human and Domestic Animals, Office of Pesticidesand Toxic Substances, Washington, D.C., 1982.57. Marr, M.C., Myers, C.B., Price, C.J., Tyl, R.W., and Jahnke, G.D., Comparison of maternal anddevelopmental endpoints for control CD ® rats dosed from implantation through organogenesis orthrough the end of gestation, Teratology, 59, 413, 1999 (tables provided by the authors).58. Kimmel, G.L., Kimmel, C.A., and Francis, E.Z., Evaluation of maternal and developmental toxicity,Teratogen. Carcinogen. Mutagen., 7, 203, 1987.59. Khera, K.S., Maternal toxicity — a possible factor in fetal malformations in mice, Teratology, 29,411, 1984.60. Khera, K.S., Maternal toxicity — a possible etiological factor in embryo-fetal deaths and fetalmalformations of rodent-rabbit species, Teratology, 31, 129, 1985.61. Clark, R.L., Robertson, R. T., Minsker, D. H., Cohen, S.M., Tocco, D.J., Allen, H.L., James, M.L.,and Bokelman, D.L., Diflunisal-induced maternal anemia as a cause of teratogenicity in rabbits,Teratology, 30(3), 319, 1984.62. Salewski, E., Färbemethode zum makroskopischen nachweis von implantationsstellen am uterus derratte, Naunyn-Schmiedebergs Arch. Exp. Pathol. Pharmakol., 247, 367, 1964.63. Lumb, W.V. and Jones, E.W., Veterinary Anesthesia, Lea and Febiger, Philadelphia, 1973, p. 452.64. Blair, E., Hypothermia, in Textbook of Veterinary Anesthesia, Soma, L., Ed., Williams and Wilkins,Baltimore, MD, 1979.65. Woollam, D.H.M. and Millen, J.W., Influence of uterine position on the response of the mouse embryoto the teratogenic effects of hypervitaminosis A, Nature, 190, 184, 1961.66. U.S. Environmental Protection Agency, Office of Prevention, Pesticides and Toxic Substances, HealthEffects Test Guidelines OPPTS 870.3700, Prenatal Developmental Toxicity Study; final guideline,August 1998.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 25967. Edwards, J.A., The external development of the rabbit and rat embryo, in Advances in Teratology,Vol. 3, Woolam, D.H.M., Ed., Academic Press, New York, 1968.68. Staples, R.E., Detection of visceral alterations in mammalian fetuses, Teratology, 9, 37, 1974.69. Wilson, J.G., Embryological considerations in teratology, in Teratology: Principles and Techniques,Wilson, J.G. and Warkany, J., Eds., University of Chicago Press, Chicago, 1965, p. 251.70. Wilson, J.G., Methods for administering agents and detecting malformations in experimental animals,in Teratology: Principles and Techniques, University of Chicago Press, Chicago, 1965, p. 262.71. Wilson, J.G., Environmental and Birth Defects, Academic Press, New York, 1973.72. Wilson, J.G., Reproduction and teratogenesis: Current methods and suggested improvements, J. Assoc.Off. Anal. Chem., 58, 657, 1975.73. Wilson, J.G., Critique of current methods for teratogenicity testing and suggestions for their improvement,in Methods for Detection of Environmental Agents That Produce Congenital Defects, Shepard,T., Miller, J.R., and Marois, M., Eds., American Elsevier, New York, 1975, p. 29.74. Wilson, J.G. and Fraser, F.C., Eds., Handbook of Teratology, Plenum Press, New York, 1977.75. Monie, I.W., Dissection Procedures for Rat Fetuses Permitting Alizarin Red Staining of Skeleton andHistological Study of Viscera, Supplement to Teratology Workshop Manual, Berkeley, CA, 1965, p. 163.76. Fox, M.H. and Goss, C.M., Experimentally produced malformations of the heart and great vessels inrat fetuses, Amer. J. Anat., 102, 65, 1958.77. Barrow, M.V. and Taylor, W.J., A rapid method for detecting malformations in rat fetuses, J. Morphol.,127, 291, 1969.78. Faherty, J.F., Jackson, B.A., and Greene, M.F., Surface staining of 1 mm (Wilson) slices of fetusesfor internal visceral examination, Stain Tech., 47, 53, 1972.79. Creasy, D.M., Jonassen, H., Lucacel, C., Agajanov, T., Gondek, H., and Schoenfeld, H., ModifiedDavidson’s fixative as an alternative to Bouin’s for evaluation of fetal soft tissue anomalies using theWilson’s serial sectioning technique, Society of Toxicologic Pathology, 21st Annual Symposium,meeting program abstract no. P35, 73, Denver, CO, 2002.80. Latendress, J.R., Warbritton, A.R., Jonassen, H., and Creasy, D.M., Fixation of testes and eyes usinga modified Davidson’s fluid: Comparison with Bouin’s fluid and conventional Davidson’s fluid,Toxicol. Pathol. 30, 524, 2002.81. Stuckhardt, J.L. and Poppe, S.M., Fresh visceral examination of rat and rabbit fetuses used interatogenicity testing, Teratogen. Carcinogen. Mutagen., 4, 181, 1984.82. van Julsingha, E.B. and Bennett, C.G., A dissecting procedure for the detection of anomalies in therabbit foetal head, in Methods in Prenatal Toxicology, Neubert, D., Merker, H J., and Kwasigrouch,T.E., Eds., George Thieme Publishers, Stuttgart, Germany, 1977, p. 126.83. Dawson, A.B., A note on the staining of the skeleton of cleared specimens with alizarin red S, StainTech., 1, 123, 1926.84. Inouye, M., Differential staining of cartilage and bone in fetal mouse skeleton by alcian blue andalizarin red S, Congenital Anomalies, 16, 171, 1976.85. Kimmel, C.A. and Trammel, C., A rapid procedure for routine double staining of cartilage and bonein fetal and adult animals, Stain Tech., 56, 271, 1981.86. Peltzer, M.A. and Schardein, J.L., A convenient method for processing fetuses for skeletal staining,Stain Tech., 41, 300, 1966.87. Staples, R.E. and Schnell, V.L., Refinements in rapid clearing technic in the KOH-alizarin red Smethod for fetal bones, Stain Tech., 39, 62, 1964.88. Crary, D.D., Modified benzyl alcohol clearing of Alizarin-stained specimens without loss of flexibility,Stain Tech., 37, 124, 1962.89. Marr, M.C., Myers, C.B., George, J.D., and Price, C.J., Comparison of single and double staining forevaluation of skeletal development: The effects of ethylene glycol (Eg) in CD ® rats, Teratology, 37,476, 1988.90. Marr, M.C., Price, C.J., Myers, C.B., and Morrissey, R.E., Developmental stages of the CD ® (Sprague-Dawley) rat skeleton after maternal exposure to ethylene glycol, Teratology, 46, 169, 1992.91. Fritz, H., Prenatal ossification in rabbits as indicative of fetal maturity, Teratology, 11, 313, 1974.92. Spark, C. and Dawson, A.B., The order and time of appearance of centers of ossification in the foreandhindlimbs of the albino rat, with special reference to the possible influence of the sex factor, Am.J. Anat., 41, 411, 1928.© 2006 by Taylor & Francis Group, LLC

260 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION93. Strong, R.M., The order, time and rate of ossification of the albino rat (Mus norvegicus albinus)skeleton, Am. J. Anat., 36, 313, 1928.94. Walker, D.G. and Wirtschafter, Z.T., The Genesis of the Rat Skeleton, Charles C. Thomas, Springfield,IL, 1957.95. Banerjee, B.N. and Durloo, R.S., Incidence of teratological anomalies in control Charles River CD ®strain rats, Toxicology, 1, 151, 1973.96. Perraud, J., Levels of spontaneous malformations in the CD ® rat and the CD-1 ® mouse, Lab. Anim.Sci., 26, 293, 1976.97. Charles River Laboratories, Embryo and Fetal Developmental Toxicity (Teratology) Control Data inthe Charles River Crl:CD ® BR Rat, Charles River Laboratories, Inc., Wilmington, MA, 1988.98. Woo, D.C. and Hoar, R.M., Reproductive performance and spontaneous malformations in controlCharles River CD ® rats: a joint study by MARTA, Teratology, 19, 54A, 1979.99. Fritz, H., Grauwiler, J., and Himmler, H., Collection of control data from teratological experimentsin mice, rat and rabbits, Arzneim-Forsch/Drug Res., 28, 1410, 1978.100. Palmer, A.K., Sporadic malformations in laboratory animals and their influence in drug testing, DrugsFetal Dev. Adv. Biol. Med., 27, 45, 1972.101. DePass, L.R. and Weaver, E. V., Comparison of teratogenic effects of aspirin and hydroxyurea in theFischer 344 and Wistar strains, J. Toxicol. Environ. Health, 15, 297, 1982.102. Wolkowski-Tyl, R., Jones-Price, C., Ledoux, T.A., Marks, T.A., and Langhoff-Paschke, L., Teratogenicityevaluation of technical grade dinitrotoluene in the Fischer 344 rat, Teratology, 23, 70A, 1981.103. Wolkowski-Tyl, R., Jones-Price, C., Ledoux, T.A., Marks, T.A., and Langhoff-Paschke, L., Teratogenicityevaluation of aniline HCl in the Fischer 344 rat, Teratology, 23, 70A, 1981.104. Wolkowski-Tyl, R., Phelps, M., and Davis, J.K., Structural teratogenicity evaluation of methyl chloridein rats and mice after inhalation exposure, Teratology, 27, 18, 1983.105. Wolkowski-Tyl, R., Jones-Price, C., Marr, M.C., and Kimmel, C.A., Teratologic evaluation of diethylhexylphthalate (DEHP) in Fischer 344 rats, Teratology, 27, 85A, 1983.106. Jones-Price, C., Wolkowski-Tyl, R., Marks, A., Langhoff-Paschke, L., Ledoux, T. A., and Reel, J. R.,Teratologic and postnatal evaluation of aniline hydrochloride in the Fischer 344 rat, Toxicology Appl.Pharmacol., 77, 465, 1985.107. Hayes, W.C., Cobel-Geard, S.R., Hanley, T.R., Murray, J.S., Freshour, N.L., Rao, K.S., and John,J.A., Teratogenic effects of vitamin A palmitate in Fischer 344 rats, Drug Chem. Toxicology, 4, 283,1981.108. Peters, M.A. and Hudson, P.M., Effects of totigestational exposure to ethchlorvynol on developmentand behavior, Ecotoxicology Environ. Safety, 5, 494, 1981.109. Peters, M.A., Hudson, P.M., and Dixon, R.L., Effect of totigestational exposure to methyl n-butylketone on postnatal development and behavior, Ecotoxicology Environ. Safety, 5, 291, 1981.110. John, J.A., Hayes, W.C., Hanley, T.R., Cobel-Geard, S.R., Quast, J.F., and Rao, K.S., Behavioral andteratogenic potential of vitamin A in the Fischer 344 rat, Teratology, 24, 55A, 1981.111. Keller, W.C., Inman, R.C., and Back, K.C., Evaluation of the embryotoxicity of JP10 in the rat,Toxicologist, 1, 30, 1981.112. Snellings, W.M., Pringle, J.L., Dorko, J.D., and Kintigh, W.J., Teratology and reproduction studieswith rats exposed to 10, 33, or 100 ppm of ethylene oxide (EO), Toxicol. Appl. Pharmacol., 48, A84,1979.113. Sunderman, F., Shen, S.K., Mitchell, J.M., Allpass, P.R., and Damjanov, I., Embryotoxicity and fetaltoxicity of nickel in rats, Toxicology Appl. Pharmacol., 43, 381, 1978.114. Sunderman, F., Allpass, P.R., Mitchell, J.M., Baselt, R.C., and Albert D.M., Eye malformations inrats: induction by prenatal exposure to nickel carbonyl, Science, 203, 550, 1979.115. Bus, J.S., White, E.L., Tyl, R.W., and Barrow, C.S., Perinatal toxicity and metabolism of n-hexane inFischer 344 rats after inhalation exposure during gestation, Toxicol. Appl. Pharmacol., 51, 295, 1979.116. Olson, C.T. and Back, K.C., Methods for teratogenic screening of air force chemicals, US NTIS ADReport, AD-A052002, 1978.117. Wolkowski-Tyl, R., Lawton, A.D., Phelps, M., and Hamm, Jr., T.E., Evaluation of heart malformationsin B6C3F1 mouse fetuses induced by in utero exposure to methyl chloride, Teratology, 27, 197, 1983.118. Cozens, D.D., Abnormalities of the external form and of the skeleton in the New Zealand Whiterabbit, Food. Cosmet. Toxicology, 3, 695, 1965.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL TOXICITY TESTING — METHODOLOGY 261119. DeSesso, J.M. and Jordan, R.L., Drug-induced limb dysplasias in fetal rabbits, Teratology, 15, 199,1977.120. Hartman, H.A., The fetus in experimental teratology, in The Biology of the Laboratory Rabbit,Weisbroth, S.H., Flatt, R.E., and Kraus, A.L., Eds., Academic Press, New York, 1974, p. 92.121. McClain, R.M. and Langhoff, L., Teratogenicity of diphenylhydantoin in the New Zealand whiterabbit, Teratology, 21, 371, 1980.122. Palmer, A.K., The design of subprimate animal studies, in Handbook of Teratology, Vol. 4, Wilson,J.G. and Fraser, F.C., Eds., Plenum Press, New York, 1978, p. 215.123. Stadler, J., Kessedjian, M.-J., and Perraud, J., Use of the New Zealand white rabbit in teratology:incidence of spontaneous and drug-induced malformations, Food Chem. Toxicol., 21, 631, 1983.124. Woo, D.C. and Hoar, R.M., Reproductive performance and spontaneous malformations in controlNew Zealand white rabbits: a joint study by MARTA, Teratology, 25, 82A, 1982.125. Kimmel, C.A. and Wilson, J.G., Skeletal deviations in rats: malformations or variations? Teratology,8, 309, 1973.126. Aliverti, V., Bonanomi, L., Giavini, E., Leone, V.G., and Mariani, L., The extent of fetal ossificationas an index of delayed development in teratogenic studies on the rat, Teratology, 20, 237, 1979.127. Woo, D.C. and Hoar, R.M., Apparent hydronephrosis as a normal aspect of renal development in lategestation of rats: the effect of methyl salicylate, Teratology, 6, 191, 1972.128. Weil, C.S., Selection of the valid number of sampling units and a consideration of their combinationin toxicological studies involving reproduction, teratogenesis, or carcinogenesis, Fd. Cosmet. Toxicol.,8, 177, 1970.129. U.S. Food and Drug Administration, Guidelines for Reproduction Studies for Safety Evaluation ofDrugs for Human Use, U.S. Food and Drug Administration, Washington, D.C., 1966.130. U.S. Food and Drug Administration, Federal Food, Drug, and Cosmetic Act, Commissioner of Foodand Drugs, Title 21, Chapter I of Code of Federal Regulations (FR), amended by addition of Part II:“Part II — Electronic Records, Electronic Signatures.” Fed. Regist., 62(54), 13464–13466 (March 20,1997)131. U.S. Environmental Protection Agency, Part V, 40 CFR parts 3, 51, et al., Establishment of ElectronicReporting: Electronic Records; Proposal Rule. Fed. Regist., 66(170), 46162–46195, 2001.© 2006 by Taylor & Francis Group, LLC

CHAPTER 8Nonclinical Juvenile Toxicity TestingMelissa J. Beck, Eric L. Padgett, Christopher J. Bowman,Daniel T. Wilson, Lewis E. Kaufman, Bennett J. Varsho,Donald G. Stump, Mark D. Nemec, and Joseph F. HolsonCONTENTSI. Introduction ........................................................................................................................264A. Objectives...................................................................................................................265B. Background................................................................................................................2651. Historical Perspective of Pediatric Therapeutics ................................................2662. Adverse Events ....................................................................................................2673. Agency Intervention/Regulatory Guidelines.......................................................267II. Importance of Juvenile Animal Studies.............................................................................272A. Differences in Drug Toxicity Profiles between Developing andDeveloped Systems....................................................................................................272B. Utility of Studies in Juvenile Animals......................................................................2731. Clinical Considerations........................................................................................273III.2. Predictive Value ...................................................................................................274Design Considerations for Nonclinical Juvenile Toxicity Studies....................................274A. Intended or Likely Use of Drug and Target Population...........................................274B. Timing of Exposure in Relation to Phases of Growth and Developmentof Target Population ..................................................................................................275C. Potential Differences in Pharmacological and Toxicological Profilesbetween Mature and Immature Systems...................................................................276D. Use of Extant Data ....................................................................................................2771. Pharmacokinetic (PK) and Toxicokinetic (TK) Data in Adult Animals ............2772. Extant Adult Nonclinical Animal Toxicity Data.................................................2803. Stand-Alone Juvenile PK and TK Studies..........................................................280IV. Juvenile Toxicity Study Design .........................................................................................280A. Types of Studies ........................................................................................................2801. General Toxicity Screen ......................................................................................2812. Adapted Adult Toxicity Study.............................................................................2813. Mode of Action Study .........................................................................................283263© 2006 by Taylor & Francis Group, LLC

264 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONB. Model Selection.........................................................................................................2841. Species .................................................................................................................2842. Sample Size .........................................................................................................284C. Exposure ....................................................................................................................2861. Route of Administration ......................................................................................2862. Frequency and Duration of Exposure .................................................................2903. Dose Selection .....................................................................................................290D. Organization of Test Groups .....................................................................................2921. Within-Litter Design............................................................................................2922. Between-Litter Design.........................................................................................2933. Fostering Design..................................................................................................2944. One Pup per Sex per Litter Design.....................................................................295E. Potential Parameters for Evaluation..........................................................................2951. Growth and Development....................................................................................2952. Food Consumption...............................................................................................2993. Serum Chemistry and Hematology .....................................................................2994. Macroscopic and Microscopic Evaluations.........................................................3015. Physical and Sexual Developmental Landmarks and Behavioral Assessments .....3016. Reproductive Assessment ....................................................................................318F. Statistical Design and Considerations.......................................................................318V. Conclusions ........................................................................................................................319Acknowledgments ..........................................................................................................................319References ......................................................................................................................................319I. INTRODUCTIONThe use of animals to evaluate the potential adverse effects of xenobiotics (environmental chemicalsor therapeutic agents) is required as an essential component of safety and risk assessment to ensurehuman health. However, many of the traditional study designs use direct dosing in adult animalsto provide safety data to extrapolate for human risk assessment, a practice that provides less thanadequate safety data for the pediatric population. The predictive value of the study designs forpurposes of risk assessment following direct dosing of xenobiotics to juvenile animals has laggedfar behind that of risk assessment based on direct dosing in adult animals or prenatal exposure.Although a considerable volume of material has been written on juvenile or pediatric pharmacokinetics,1 less attention has been given to juvenile or pediatric toxicity studies that employ directadministration of xenobiotics during critical periods of development. Many nonclinical safetyevaluations of therapeutic agents are conducted in adult animals, with the result that many drugsapproved for use in the adult human population are not adequately labeled for use in the pediatricpopulation. Physicians prescribe therapeutic agents off-label for children to treat ailments such asdepression, epilepsy, severe pain, gastrointestinal problems, allergies, high blood pressure, andattention deficit hyperactivity disorder (ADHD). Even though many of the drugs prescribed forthese indications have been extensively evaluated in adults, very little, if any, safety or efficacydata have been obtained from controlled studies in juvenile animals or clinical trials in children.Historically, physicians have determined the dosage of therapeutics for use in children basedupon professional experience and the child’s body weight. The problems with this approach arethat children are not just “little adults,” and the differences in sensitivity between children andadults can be chemical specific. Developing organs or organ systems and metabolic pathways inchildren and fully developed adult systems can react very differently to drugs. Because of theirsmall stature, infants and/or children are often assumed to be more susceptible to the toxic effectsof drugs or chemicals in the environment. However, their susceptibility depends on the substance© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 265Table 8.1Proposed pediatric age classificationPreterm newborn infantsTerm newborn infantsInfants and toddlersChildrenAdolescentsBorn prior to 38 weeks of gestation0 to 27 days of age28 days to 23 months of age2 to 11 years of age12 to 16-18 years of ageSource: Modified from Guidance for Industry: E11 Clinical Investigationof Medicinal Products in the Pediatric Population (2000). 6and the exposure situation (e.g., timing of exposure during critical developmental periods). In somecases, there may be no difference in response between children and adults. In other cases, differentphysiological and metabolic factors, pharmacokinetics, and behavioral patterns can render childrenmore or less sensitive than adults.Results of studies in juvenile animal models indicate that exposure to certain environmentalchemicals, 2 drugs, 3 and ionizing radiation 4 can lead to potential developmental and/or functionaldeficits. These deficits may range from immunomodulation (e.g., suppression of vital componentsof the immune system) to altered or poorly regulated processes that may be debilitating (e.g.,behavioral impairment). Concern regarding potential toxic effects in children has increased in recentyears with the recognition that children are a unique target population. However, the inherent risks,ethical issues, and costs associated with extensive human pediatric testing can be significant.One way to address the concerns that children may be more sensitive than adults to drug orchemical exposures is to conduct safety evaluations designed to target critical periods of developmentfollowing direct exposure in juvenile animals. Results obtained from testing at the appropriatedevelopmental stage in animal models may be used to characterize and extrapolate both efficacyand safety in the pediatric population. More importantly, information from juvenile animal studiesmay support reducing the number of children required for pediatric clinical trials.A. ObjectivesThe purpose of this chapter is to provide a brief introduction to the regulatory history of pediatricsafety assessment and to present information for the design, conduct, and interpretation of nonclinicaljuvenile safety studies for extrapolation to the pediatric population. Examples of possiblevariations in study designs are presented. In addition, selection of the appropriate juvenile animalmodel, advantages and disadvantages of various animal species, and some mistakes to avoid in thepractical aspects of conducting the studies are presented. Where appropriate, examples are givenregarding the differences in organ system maturation between animal species and their relationshipto human development. Other topics of importance for study design, including dose selection;organization of test groups (e.g., litter design); route, frequency, and timing of exposure; andendpoints to measure are discussed.B. BackgroundNewborns (preterm or term), infants, toddlers, children, and adolescents are recognized as differentfrom adults in many ways with respect to behavioral, developmental, and medical requirements.Equally important is the recognition that developmental differences exist within the pediatricpopulation as well. Grouping the pediatric population into age categories is to some extent subjective,but classifications such as the one in Table 8.1 provide a basis for considering age groupswithin the pediatric population in which developmental differences may exist. Indeed, more thana century ago, Dr. Abraham Jacobi, father of American pediatrics, recognized the age differencesbetween his pediatric patients, as well as the need for age-related drug therapy, when he wrote,© 2006 by Taylor & Francis Group, LLC

266 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION“Pediatrics does not deal with miniature men and women, with reduced doses and the same classof disease in smaller bodies, but…its own independent range and horizon.” 5Traditionally, physicians have recognized that tissues and organs in children mature and functiondifferently depending on the developmental period of life, and that anomalies in these earlydevelopmental processes could be associated with disease progression found specifically in thepediatric population. 6 Despite the known differences in the physiological ontogeny from early childdevelopment to adulthood, there was very little difference in the way in which drugs were prescribedbetween the pediatric and adult populations. 7Prior to the increased concern of off-label use of drugs in children, as well the considerationfor using developmental pharmacology in the therapeutic evaluation process, there was a lack ofappropriate drug-dosing guidelines to help in the determination of pediatric drug dose levels to beused in the clinic. 8 Typically, adjustments for drug dose in children were based upon rudimentaryformulary calculations that used age or relative body size (e.g., Young’s Rule, Cowling’s Rule, orClark’s Rule), as well as labeling information derived from both nonclinical and clinical adultstudies. 9,10 These practices were based upon default assumptions that there are no developmentaldifferences for a drug’s pharmacokinetic or pharmacodynamic characteristics between children andadults, and that children and adults have similar disease progression. However, pediatric growthand development are not one-dimensional processes, and age-associated changes in body compositionand organ development and function are variables that can have an impact on drug toxicityand/or efficacy, as well as on disease progression. 9 Therefore, these prescribing practices were notconsidered sufficient for individualizing drug doses across the entire course of pediatric development.8 As a result, the use of dose level equations is being replaced by normalization of the drugdose for either body weight or body-surface area, in combination with understanding the safetyand pharmacokinetic data obtained from testing drugs in pediatric clinical trials and/or from juvenileanimal studies. 8,11–131. Historical Perspective of Pediatric TherapeuticsMany prescription drugs and biological therapeutics are marketed with little or no dosage informationfor administration to pediatric patients. This information, if available for use in the pediatric population,is provided to physicians in the product label (package insert). Therefore, the drugs without adequatelabeling information which are given to children are unlicensed or prescribed off-label. The off-labelprescription of a drug therapy approved by the U.S. Food and Drug Administration (FDA) to a patientoutside the specification of the product license involves drugs being administered by an unapprovedroute, formulation, or dosage, or outside an indicated age range. 14 According to a report in the FDAConsumer Magazine, some classes of drugs and biologics, such as vaccines and antibiotics, generallyhave adequate labeling information for pediatric use. 15 However, pediatric labeling for other classes ofdrugs has been deficient. Examples include steroids to treat chronic lung disease in preterm neonates, 16agents to treat gastrointestinal disorders, 17,18 prescription pain medications, 19 antihypertensives, 20 andantidepressants. 21 In addition, some age groups have less labeling information available to them thanothers. For example, children under 2 years of age have virtually no pediatric use information on drugproducts in several drug class categories. 22Despite regulatory efforts to increase the rate of labeling of pharmaceuticals for pediatricindications, little has changed in the way that these products are prescribed for off-label use in thepast two decades. According to some literature, more than half of the drugs approved every yearthat are likely to be used in children are not adequately tested or labeled for treating patients inthe pediatric population. 15 A survey of the 1973 Physicians’ Desk Reference (PDR) showed that78% of drugs listed contained either a disclaimer or lacked sufficient dose information for pediatricuse. 23 A later survey of the 1991 PDR indicated that 81% of the listed medications contained informationdisclaiming use in children or limiting use to specific age groups in the pediatric population. 24 Another© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 267survey of new molecular entities (NMEs; an NME contains an active substance that has not previouslybeen approved for marketing in any form in the United States) approved by the FDA from 1984through 1989 showed that 80% of the listed drugs were approved without label information for thepediatric population. 25 In addition, the FDA identified the 10 drugs most commonly prescribed tochildren in 1994 that lacked sufficient pediatric labeling information. 15 Together, these drugs wereprescribed more than five million times in the pediatric population. Between 1991 and 2001, theFDA approved 341 NMEs. 26 Of these NMEs, only 69 (20%) were labeled for use in children atthe time they were initially approved. Although the percentages varied from year to year, there wasno apparent trend toward an increase over time for NMEs with pediatric labeling. In addition,between 1991 and 1997, the FDA also gathered statistics on the number of NMEs that werepotentially useful in the pediatric population. Of 140 such drugs, only 53 (38%) were labeled foruse in children when they were initially approved. This relatively low percentage of NMEs developedand/or labeled for pediatric use is somewhat surprising, considering that some published dataindicate that more children than adults in the United States are taking prescription medications. 27,28The lack of drug approval for pediatric use does not imply that a drug is contraindicated. Itsimply means that insufficient data are available to grant approval status and that the risks or benefitsof using a drug for a particular indication have not been examined. This leads to the followingbasic issues concerning the off-label use of drugs in children: the lack of age-related dosingguidelines, the lack of age-related adverse-effect profiles, and the unique pediatric clinical problemsthat must be considered when prescribing off-label. 7 These issues create an ethical dilemma forthe physician in that a decision must be made to either deprive a child of the potential therapeuticbenefits of using a drug or to use it despite disclaimers or the lack of sufficient safety and efficacydata. The latter choice has a potential consequence of underdosing with a resulting lack of efficacyor overdosing with a resulting increase in toxicity.2. Adverse EventsThe history of pediatric medicine is well documented with progressive improvements in children’shealth care. However, interspersed throughout the medical successes have been many therapeutictragedies of adverse drug reactions in the pediatric population (refer to Figure 8.1 for examples).Some of the tragedies include kernicterus in newborns from administration of sulfa drugs, 29 seizuresand cardiac arrest from the local anesthetic bupivacaine, 30 neurological effects from the unexpectedabsorption through the skin of infants bathed with hexachlorophene, 31 fatal reactions in neonatesexposed to benzyl alcohol used as a preservative in dosing formulations, 32 fatal hepatotoxicity tochildren 2 years or younger administered valproic acid as part of a multiple anticonvulsant therapy, 33cardiotoxic effects of anthracyclines in the treatment of childhood cancer, 34 and cardiovascularcollapse (i.e., gray-baby syndrome) in neonates administered chloramphenicol as a treatment forsevere bacterial infections. 35All of these events have the common underlying theme of introducing treatments into thepediatric population without having adequately investigated their potential harm. It is these typesof misfortunes that have been primarily responsible for the major changes in laws and regulationsthat guide the testing and marketing of new drugs.3. Agency Intervention/Regulatory GuidelinesMany drugs marketed in the United States and used in the pediatric population lack safety andefficacy information derived from this population. In many cases, safety data from clinical trialsin adults, supported by nonclinical studies in adult animals, have supported the use of a therapeuticagent in the pediatric population. In the case of a potential therapeutic agent or new environmentalchemical, the question with respect to early safety studies is usually, “Has the compound been© 2006 by Taylor & Francis Group, LLC

Figure 8.11775Pott DescribesScrotal Cancer inChimneySweeps2004CERHRPublishesFluoxetineReportPaint DustIdentified asMajor Sourceof Lead inChildren19781904CPSC BansUse of LeadPaints inHousing2003PediatricResearchEquity Act2002BPCA19342000Children’sHealth Act1951Phenylketonuria Retinopathy ofDiscovered PrematurityLinked to OxygenTherapy1973EPA Begins LeadPhase-Out fromGasoline1979NCTRCollaborativeBehavioralStudy1980Reye’sSyndromeLinked toAspirinFDAPediatricRuleIssuedInfluencing events and key times in guideline development and their changes for juvenile toxicology.19981956Sulfa DrugsAssociated withChloramphenicol-KernicterusGray Baby1971 Syndrome1972HexachloropheneHuman EffectsHexachloropheneEffects on Rat BrainsReported1982BenzylAlcoholInduces“GaspingSyndrome”1997FDAMA1996FQPA19571966Wilson’sPrinciples1991First FIFRADNTRequirement1990CAAAmendmentBanned Lead& LeadAdditives inGasoline268 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 269evaluated in adult animals?” Better questions would have been, “Has the substance been evaluatedin developing animals?” or “Have the effects in adult animals been studied following direct exposureto developing offspring?” Despite the lack of pediatric and nonclinical juvenile toxicity studies,many of the therapeutic drugs on the market today have been used safely in children. For example,acetaminophen has been prescribed for many decades to children. 36 Although there has been someformal pharmacological testing of acetaminophen following direct administration to children, agreat deal of information about its safety and efficacy has been acquired over time from anecdotalreports. An extrapolation from the clinical experience in adults was used to provide initial dosinginformation for children. In general, the risk of adverse events from acetaminophen administrationappears to be lower in pediatric patients than adults. 37 However, regardless of the very low incidenceof adverse effects, acetaminophen toxicity still remains a concern because it is widely used throughoutthe pediatric population. 38 In general, many drugs prescribed to the pediatric population haveinsufficient information, either anecdotally and/or from controlled studies, on which to base theappropriate dosage regimen.The lack of appropriate safety or efficacy information for the use of therapeutic agents inchildren may have been due in part to the trepidation that physicians had about testing drugs inchildren. Many in the health community argued that it was unethical to put children at risk,especially since children could not give legal consent. 39 Despite these concerns, children were stillprescribed drugs that the FDA had approved for use in adults only. Even though it was known thatjuveniles, especially neonates and infants, metabolized some drugs differently than adults, thetherapeutic benefits were assumed to outweigh the unknown risks. In addition to the ethical concernsfor testing drugs in children, pediatric trials and/or nonclinical juvenile studies would also addmillions to the cost of drug development. Physicians can legally prescribe adult products to childrenoff-label. Thus, the drug industry had little incentive to increase the number of drugs targeting thepediatric population or to evaluate existing drugs currently on the market that could be or werebeing administered to children. Instead, many companies opted to add a disclaimer on the labelindicating that safety and effectiveness have not been established in the pediatric population. Eventhough off-label usage may have been supported by much professional experience, it was recognizedby the lay, medical, scientific, and governmental communities that this could not continue to bethe type of evidence on which to base safe and effective drug therapies for use in the pediatricpopulation. Indeed, as illustrated in Figure 8.1, there already existed scientific evidence of pediatrictoxicities that were not readily predicted from adult use of the same drugs.With the increasing concern that pediatric patients should be recognized as a sensitive subpopulationin risk assessment, and with drug development targeting conditions specific to the pediatricpopulation, new regulations were enacted to increase the extent to which drugs were tested priorto use in children (refer to Figure 8.1 for a brief history of pediatric drug labeling). Two of themost prominent U.S. regulations used to advance pediatric risk assessment for drug developmentinclude the Pediatric Rule 40 and the Best Pharmaceuticals for Children Act (BPCA). 41The Pediatric Rule established the presumption that all candidate drugs and biologics must beevaluated in the pediatric population. Under the Rule, pediatric safety and effectiveness data mustbe included for New Drug Applications (NDAs), Biologics License Applications (BLAs), supplementalapplications for new active ingredients, and new indications, dosage forms, dosing regimens,or routes of administration. For currently marketed drugs and biologics, the Rule required pediatricstudies for products that were used for a labeled indication in 50,000 or more pediatric patients,where the absence of adequate labeling could pose a significant risk to these patients. In addition,the Rule required pediatric studies for products that would provide a meaningful therapeutic benefitover existing treatments for pediatric patients for one or more of the claimed indications where theabsence of adequate labeling information could pose significant risk in these patients. However,under the Rule, the FDA had no authority to require a manufacturer to conduct pediatric studiesfor products used for off-label indications.© 2006 by Taylor & Francis Group, LLC

270 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIn December of 2000, the Pediatric Rule was challenged in federal court. 42 The lawsuit statedthat the Rule was not valid, arguing that the FDA had no statutory authority to promulgate theRule. The court decided in favor of the lawsuit, struck down the Rule, and enjoined the FDA fromits enforcement. However, despite the ruling, legislation was eventually introduced in Congress towrite the Pediatric Rule into law. As a result, the Pediatric Research Equity Act of 2003 amendedthe Federal Food, Drug and Cosmetic Act to provide the FDA with the statutory authority to requirethe drug industry to submit assessments regarding the use of drugs and biologics in pediatric patientsin certain defined circumstances. 43 In fact, the authority granted in the new legislation follows manyof the same key elements of the former Pediatric Rule. 44The BPCA 41 reauthorized until 2007 the 6 months of additional drug patent protection originallyenacted in the Food and Drug Administration Modernization Act (FDAMA) of 1997, whichincorporated an incentive-based program for encouraging the drug industry to provide pediatricdrug labeling information. 45 In addition, the BPCA made available alternative mechanisms, suchas studies by the National Institutes of Health (NIH), to obtain information about generic or patenteddrugs that a manufacturer does not study in children. The BPCA authorized the FDA to requestpediatric studies of marketed drugs upon determination that the information relating to their usein the pediatric population provides additional health benefits or because their use predisposeschildren to greater risk. The BPCA required the NIH (in consultation with the FDA) to develop,prioritize, and publish an annual list of approved drugs for which additional studies are needed toassess safety and effectiveness in the pediatric population. The FDA can issue written requests forthe conduct of pediatric or nonclinical juvenile studies to manufacturers of approved NDAs (drugswith or without patent exclusivity). If there is no response from the manufacturers, the NIH mustpublish a request for contract proposals to conduct the studies. Following their completion, reportsof the studies, including all data generated, must be submitted to the NIH and FDA. Followingsubmission, the FDA has a 180-day period to negotiate label changes with the manufacturer andpublish in the Federal Register a summary of the report and a copy of the requested label changes.If a manufacturer does not agree to the recommended labeling changes, then the FDA refers therequest to the Pediatric Advisory Subcommittee for review. After the subcommittee makes itsrecommendations to the FDA, the manufacturer is requested to make the labeling changes. If themanufacturer fails to agree, the FDA may deem the drug misbranded.The combination of mandatory (Pediatric Rule, now law under the Pediatric Research EquityAct of 2003) and incentive-based (FDAMA, now reauthorized under BPCA) legislative policiesappears to have contributed to an increased number of drug studies involving children, whichresulted in pediatric labeling changes 46 and an increase in the number of drugs being developedspecifically for diseases in the pediatric population. 47Although these legislative policies advocate the collection of pediatric safety and effectivenessdata for marketed drugs (Pediatric Rule and BPCA) and biological products (Pediatric Rule only),very little, if any, guidance was provided to indicate specifically what to do to obtain adequatesafety information for pediatric labeling. There are, however, several guidance documents thatprovide outlines of critical issues in pediatric drug development and approaches to the safe, efficient,and ethical study of drug products in the pediatric population. 6,48 These guidance documentsacknowledge that assessing the effects (safety and effectiveness) of drugs in pediatric patients isthe primary goal of any pediatric clinical trial. However, pediatric clinical trials should be ethicallyconducted without compromising the well-being of pediatric patients or predisposing them to morethan minimal risk in cases where there may be only slight, uncertain, or no beneficial returns. 49Therefore, previously collected, relevant safety data from adult studies, as well as data fromnonclinical repeated-dose toxicity, reproductive toxicity, and genotoxicity studies, should be consideredin an effort to anticipate and reduce potential hazards in pediatric subjects before a clinicaltrial is initiated. Given that there are developmental differences between pediatric and adult patientsthat can affect risk assessment and that acceptable concordance has been shown for the predictability© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 271of drug toxicity between adult animals and humans, 50 nonclinical juvenile studies should be usefulfor predicting potentially adverse events prior to the initiation of pediatric clinical trials.Recognizing the important benefits of juvenile toxicity studies, in February 2003 the FDAissued a draft guidance document on nonclinical safety evaluation of pediatric drug products. 51 Thisdocument provides general information on the nonclinical safety evaluation of drugs intended forthe pediatric population. In addition, the guidance presents some conditions under which nonclinicaljuvenile studies are considered meaningful predictors of pediatric toxicity and recommends pointsto consider for the design of nonclinical juvenile studies. Because the document attempts toencompass most pediatric products (e.g., NMEs, drugs used off-label, or pediatric-only indications),there is no one conventionally accepted nonclinical juvenile study design that is recommended toaddress the need for, timing of, or the most appropriate parameters to include for assessment. Thishas led to the current practice for determining the need for nonclinical juvenile studies on a caseby-casebasis. This approach should take into consideration the following issues for determiningthe need for conducting a nonclinical juvenile study: (1) the intended likely use of the drug in thepediatric population, (2) the timing and duration of dosing in relation to the growth and developmentin children, (3) the potential differences in pharmacological or toxicological profiles betweenchildren and adults, and (4) the evaluation of whether available adult data from nonclinical orclinical studies support reasonable safety for drug administration in pediatric patients. For specificstudy design considerations for a nonclinical juvenile study, the guidance document discussesappropriate selection of animal species, age at initiation of dosing, sex, sample size, route, frequency,and duration of drug administration, dose selection, toxicological end points, and timing of monitoring.The content of this chapter spans some of the same considerations recommended in theguidance document and attempts to further elaborate on specific examples of study design andimplementation.In regard to industrial and environmental chemicals, the U.S. Environmental Protection Agency(EPA), along with many in the medical community, recognize the necessity of hazard identificationstudies for pesticides, toxic substances, and environmental pollutants to address the potential foradverse effects to pediatric development. 52,53 Hazard and dose-response information are used in therisk assessment process and are typically derived from laboratory animal studies. These includethe prenatal developmental toxicity study in two species, 54 the two-generation reproduction studyin rodents, 55 and for some chemicals, the developmental neurotoxicity study in rats. 56 The reproductionand developmental neurotoxicity studies include juvenile animals that have been potentiallyexposed to a test chemical through placental transfer during in utero development and until weaningfrom lactational exposure. The juvenile animals may also be directly exposed through treated-dietconsumption during the latter part of the preweaning period. The data used for risk assessment inthe juvenile population from these types of animal studies commonly assume the following: (1)offspring exposure to the chemical has in fact occurred, (2) in the absence of kinetic data, levelsof chemical exposure in the offspring are similar to those in adults, and (3) exposures of youngrodents (prenatal and/or weanling animals) are similar to potential exposures in downstream developmentalperiods (e.g., neonates, infants, children, or adolescents). In recognition that these assumptionsmay not provide the most accurate data for risk assessment in infants, children, or adolescents(the presumed sensitive populations), the EPA recommends that consideration be given to directdosing in immature animals to characterize potential adverse effects during critical developmentalperiods. 52 These studies can be conducted on a chemical-specific basis, in which modifications toexisting guideline protocols may be used to specifically address the developing animal. Thisapproach in study design has been pursued by the EPA Office of Pesticide Programs since 1999to evaluate potential developmental neurotoxic effects and age-related sensitivity to cholinesteraseinhibition by organophosphorous pesticides. 57 The EPA has begun advocating the expanded use ofbiomarkers and toxicokinetic data to help characterize direct exposure and/or aid in the design oftargeted animal studies for risk assessment for the pediatric population.© 2006 by Taylor & Francis Group, LLC

272 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONII. IMPORTANCE OF JUVENILE ANIMAL STUDIESA. Differences in Drug Toxicity Profiles between Developing andDeveloped SystemsSome drugs or environmental chemicals in the pediatric population show different toxicity profilescompared with those of adults. Intrinsic differences between developed and immature systems mayresult in an increase or decrease in drug toxicity in the pediatric population. From a physiologicaland anatomical perspective, many factors can contribute to these differences in toxicity profiles.Some of these factors include rapid cell proliferation during early pediatric development, agerelateddifferences in body composition (e.g., fat, muscle, and bone composition, and water contentand distribution), level of hepatic enzyme development, age-related differences in protein synthesisand function in the pancreas and gastrointestinal tract, composition and amount of circulatingplasma proteins that bind to free drug, maturation of renal function, degree of immune systemfunctional maturity, ontogeny of receptor expression, intestinal motility and gastric emptying, andmaturation of the blood–brain barrier.As stated previously, neonates (preterm and term), infants, children, and adolescents should notbe considered simply as “little adults” when toxicological risk is evaluated. Instead, pediatricpatients are a unique population for assessing risk associated with drug therapies or chemicalexposures. From the viewpoint of pediatric medicine, there is a large potential for children to bemore sensitive to drug effects during growth and development, which encompasses the periodbetween birth and adulthood. The structural and functional characteristics of many cell types,tissues, and organ systems differ significantly between children and adults because of rapid growthand development. These developmental differences can lead to different outcomes for drug absorption,distribution, metabolism, and elimination, which can have an impact on toxicity and/or efficacy.For example, as a result of reductions in both basal acid output and the total volume of gastricsecretions during the neonatal period, the gastric pH is relatively higher (greater than 4) than thegastric pH in older children and in adults. 58,59 Subsequently, oral administration of acid-labilecompounds to neonates can lead to greater drug bioavailability than in older children or adults. 60In contrast, drugs that are weak acids may require larger oral doses during the early postnatal periodto achieve the therapeutic levels observed in older children or adults. 61 In addition, age-dependentchanges in body composition can alter the physiological compartments in which a drug may bedistributed. 9 For example, compared to adults, neonates and infants generally have relatively largerextracellular and total-body water spaces associated with adipose stores, resulting in a higher ratioof water to lipid composition. 62 This could potentially lead to changes in the volume of drugdistribution, affecting free plasma levels of the drug.From a functional perspective, a fully mature immune system is lacking in the pediatricpopulation, where adult response antibody levels for immunoglobulin G (IgG) and immunoglobulinA (IgA) are not achieved until about 5 and 12 years of age, respectively. 63 Experimental animalstudies, and to a lesser extent human studies, describing adverse immunological effects in neonatespresumed to be exposed to toxic agents during the prenatal or early postnatal period have beenpublished. 64–66 Of particular concern is that aberrant effects on the immune system often appearmore severe and/or persistent when the drug or chemical exposure occurred peri- or postnatally,compared with exposure in adults.In regard to normal cognitive and motor function, several neurodevelopmental processesundergo relatively rapid postnatal change from birth to 3 years of age when compared with thoseof adults. These processes include migration of neurons in the cerebellum, synapse formation inthe association cortex, closure of the blood brain barrier, and myelination in the somatosensory,visual, and auditory areas of the brain. 67 When administered to preterm infants for the treatmentof bronchopulmonary dysplasia, inhaled steroids can affect postnatal development during the period© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 273in which these critical neurological processes occur. 68 The resulting toxic effects have been implicatedas the cause of an increased association with the occurrence of cerebral palsy. 69Children metabolize some drugs in an age-related manner, and adverse events may not bepredicted from adult experiences. For example, children may be more sensitive to adverse eventsdespite exposures being equal to or less than those in adults because maturational differences inenzyme activities and/or concentrations may preclude a simple scale down from adult to pediatricdoses. Administration of the antibiotic chloramphenicol in neonates (less than 1 month of age)leads to a toxic condition described as “gray baby syndrome.” 35 This condition can be fatal via aform of circulatory collapse associated with excessive and sustained serum concentrations of theunconjugated drug. Neonates are susceptible because of the absence of Phase II enzymes responsiblefor the conjugation reaction converting the active drug to a biologically inactive, water-solublemonoglucuronide. 70 Because this is a dose-dependent toxicity, chloramphenicol must be administeredto neonates at a lower dosage than is administered from infancy through adulthood.In contrast, hepatotoxicity from acetaminophen-induced depletion of glutathione levels is lesssevere in neonates and young children than in adults. 71 The major pathway for the metabolism oftherapeutic doses of acetaminophen is through conjugation reactions with sulfate and glucuronide.The minor metabolism pathway for the drug is oxidation by the mixed-function oxidase P450system (mainly CYP2E1 and CYP3A4) to a toxic, electrophilic metabolite, N-acetyl-p-benzoquinoneimine(NAPQI). 72 Under conditions of excessive NAPQI formation or reduced glutathione stores(e.g., acetaminophen overdose), the toxic metabolite covalently binds to proteins and the lipidbilayer, resulting in hepatocellular death and subsequent liver necrosis. 36 Because young childrenhave lower enzyme activity levels for the minor pathway and higher glutathione stores and rate ofglutathione turnover, there is less toxic intermediate formed.B. Utility of Studies in Juvenile Animals1. Clinical ConsiderationsTraditionally, drugs have not been sufficiently evaluated in the pediatric population. The reason ismultifactorial and relates to ethical issues, technical constraints, financial and practical concerns,and the potential for long-term adverse effects, as well as deficiencies in previous laws andregulations to require more labeling information. 73 It is considered unethical for a child to beprescribed medications that have not been adequately evaluated; however, studies involving proceduressubmitting children to more than minimal risk with only slight, uncertain, or no benefit arealso regarded as unethical. 49 Under federal regulations, children are considered a vulnerable groupand require additional protection as research subjects. 74 In addition, obtaining the assent of a childand the permission of a parent or guardian is not the same as obtaining informed consent from acompetent adult. 39 Furthermore, the parameters utilized to monitor children’s health and safetyduring clinical trials are of major technical and ethical concern.The financial and practical challenges in drug development involving children include the smallnumber of patients with certain age-related medical conditions, the expense of studying childrenin various stages of development, difficulties in developing formulations appropriate for children,difficulty in recruitment of pediatric patients, and the possibility that an approved drug will be usedby only a small number of pediatric patients. 26 In addition, previous FDA regulations allowed theeffectiveness of a drug treatment to be extrapolated from studies in adults to children if theprogression of the medical condition being treated and the effects of the drug were sufficientlysimilar. 75 The adult data were usually augmented with pharmacokinetic studies obtained in children.However, the safety of a drug prescribed in children cannot always be extrapolated from dataobtained in adults; therefore, the drug and course of treatment could potentially be more or lesstoxic in children. Because of these issues, governmental institutions and professional organizations© 2006 by Taylor & Francis Group, LLC

274 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONhave started to address the need to evaluate drugs for children in a similar manner as they are foradults.2. Predictive ValueThe traditional model for prediction of potential hazards in pediatric patients is to consider studydata (safety, effectiveness, and exposure) from previous adult human studies, as well as data fromnonclinical repeated-dose toxicity, reproductive toxicity, and genotoxicity studies. 48 However, thismodel inherently assumes (correctly or incorrectly) that a similar disease progression occursbetween children and adults, that there will be a comparable response to the drug for a specificindication, and that there are no differences between adults and developing children that may affectthe safety profile. Therefore, given that there are known differences between pediatric and adultpatients that can affect risk-benefit analysis, and that acceptable concordance has been shown forthe predictability of drug toxicity between adult animals and humans, nonclinical juvenile studiesshould be used to predict potentially adverse events prior to the initiation of pediatric clinicaltrials. 50,76In general, some juvenile animals (e.g., rodents, dogs, minipigs, nonhuman primates) exhibitage-related and developmental characteristics similar to those in the human pediatric population,thus making them suitable for toxicity testing. 77 Because of these similarities, nonclinical juvenilestudies have been shown to be useful for identifying, evaluating, or predicting age-related toxicitiesin children. For example, the adverse effects of phenobarbital on cognitive performance in childrenare predicted by administration of the drug to the developing rodent during periods of criticalneurodevelopment. 78,79 Increased sensitivity of human infants to hexachlorophene neurotoxicity wasreplicated and evaluated in juvenile rats of comparable developmental age. 80 Proconvulsant effectsobserved in developing rodents treated with theophylline could be predictive of risk to similareffects in children. 81 In addition, a juvenile rat model has been used to understand the toxic effectson craniofacial growth and development observed in young children administered prophylactictreatments (irradiation with or without chemotherapeutic drugs) to reduce the recurrence of childhoodacute lymphoblastic leukemia. 4 Unfortunately, most of these examples of the predictive utilityof nonclinical juvenile studies for identifying adverse events occurred after the adverse events wereidentified in children.III. DESIGN CONSIDERATIONSFOR NONCLINICAL JUVENILE TOXICITY STUDIESWhen designing nonclinical juvenile toxicity studies, various factors must be considered to obtaindata that allow the assessment of the toxicological profile of a compound in young animals andprovide information pertaining to specific endpoints deemed important based on existing clinicaland/or nonclinical data. For example, the intended use of a drug and the intended target populationmust be addressed, along with the developmental period of exposure in the target population. Theduration of clinical use of a drug must also be considered when determining the proper exposureperiod in animal studies. The test article exposure regimen in an animal model should correspondto the intended exposure period in the human target population. Finally, to the extent possible,inter- and intraspecies physiological, pharmacological, and toxicological profiles should be understood.A. Intended or Likely Use of Drug and Target PopulationJuvenile toxicity studies are often conducted in the rat, which is a species that reaches adult statusmuch more rapidly than humans (as illustrated in Figure 8.2). This compression of physiological© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 275100Rat% Adult StatusHuman00 5101520Age (years)Figure 8.2Time to develop adult characteristics. Rats, because of their shorter life span compared to humans,reach adult status much more rapidly. This compression of physiological time must be consideredwhen determining the appropriate exposure periods in juvenile toxicity studies.time in the rat and other nonhuman species relative to humans must be considered when determiningthe appropriate time of initiation and duration of exposure in the nonclinical juvenile study. Forexample, if a compound is administered for a number of years in children, the critical period ofexposure in the rat may be merely weeks. Additional thought must be given to the ontogeny ofspecific organ systems in the animal versus the human. Maturation of several organ systems (e.g.,the renal system) occurs in utero in humans but occurs in the early postnatal period in rodents.Therefore, the window of exposure in nonclinical juvenile studies must take into account thephysiological age correlation between humans and the animal model, as well as differences intiming of development of specific organ systems of interest. As an example, Figure 8.3 illustratesthe comparative age categories based on CNS and reproductive system development in severalspecies. 82B. Timing of Exposure in Relation to Phases of Growth and Developmentof Target PopulationThere are multiple examples that illustrate the idea of physiological time and differences indevelopment between humans and animals. Well-documented examples include the differencesbetween humans and rats in the timing of formation of the functional units of the lung (alveoli)and those of the kidney (nephrons). Detailed reviews of the development of the kidneys, lungs, andother organ systems can be found in Appendix C of this text. In the human, alveolar proliferationis initiated during gestational week 36 (approximately). It is completed by 2 years of age, and theexpansion stage is completed by 8 years of age. In the rat, however, alveolar development occursstrictly during the postnatal period, with proliferation occurring between postnatal days (PND) 4and 14, and the expansion stage is completed by PND 28. 66 Based on this information, it wouldbe appropriate to expose young rats during the first 4 weeks of postnatal life in order to study theeffects of a compound on alveolar development. The rat, therefore, is a widely accepted model forstudying the effects of compounds on lung development, while other species, such as the rabbit,sheep, pig, and monkey are not acceptable for a postnatal assessment of lung development becauseof the advanced stage of lung development seen in these species at birth. 83Nephrogenesis also occurs prenatally in the human but postnatally in the rodent. 84–86 Nephrogenesisis completed by gestational week 34 to 35 in humans, 85 while it has been reported that thisprocess is completed by PND 11 in the rat, with further kidney maturation extending until postnatalweek 4 to 6. 87 Along with considering the relative timing of morphological development of thekidney between humans and test species, functional development should also be considered. The© 2006 by Taylor & Francis Group, LLC

The underlying physiology of juveniles and adults leading to the differences in metabolic clearanceand/or activation of compounds should be identified, if possible, prior to initiating juvenile toxicitytesting. Table 8.2 lists several factors responsible for the different responses to toxicant exposurein adults versus children. As mentioned previously in this chapter, the timing of exposure to a drug276 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONRatB≤910214590DaysMinipigB241426WeeksDog B 0.5 3 6 20 28 WeeksNonhumanPrimateB0.563648MonthsHumanB2821216Days/YearsPre-TermNeonateTermNeonateInfant/Toddler Child AdolescentOntogenyBBirthFigure 8.3Comparative age categories based on overall CNS and reproductive development. (From Buelke-Sam, J., Comparative schedules of development in rats and humans: implications for developmentalneurotoxicity testing, presented at the 2003 Annual Meeting of the Society of Toxicology, Salt LakeCity, UT, 2003.) When designing juvenile toxicity studies, specific organ system development mustbe considered in conjunction with overall developmental age categories.rat model is appropriate in this respect when investigating potential nephrotoxicants because severalaspects of kidney function, including glomerular filtration rate, concentrating ability, and acid-baseequilibrium, mature postnatally in both humans and rats. 87 Therefore, postnatal exposure in the ratwould be appropriate when examining drug- or chemical-induced effects on nephrogenesis andfunctional kidney development.Extensive investigations have also been performed to elucidate the ontogenies of immune systemfunction, bone growth, and the reproductive and central nervous systems in various species, includinghumans (cf. Appendix C). As an example, the ontogeny of cholinesterase was examined in fivecompartments in the rat, including the plasma, red blood cells, brain, heart, and diaphragm. 88 Ona whole-tissue basis, cholinesterase in the heart and diaphragm increased with age from gestationday 20 through PND 21 and then remained relatively constant through adulthood, while in thebrain, cholinesterase increased until approximately 6 weeks of age. However, when normalized tothe weight of the tissue, cholinesterase activity in the heart and diaphragm increased until PND 11and decreased thereafter until adulthood, while it continued to increase with age and growth of thebrain.C. Potential Differences in Pharmacological and Toxicological Profilesbetween Mature and Immature Systems© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 277or chemical is an important factor in the resulting effects seen in exposed individuals. During theneonatal and postnatal periods, a multitude of developmental processes occur, including CNSdevelopment, cardiovascular system development, and reproductive system maturation. Becauseyoung children are growing and maturing, they may express unique susceptibilities to toxicantexposure. For example, while a chemical or drug that delays sexual maturation may not have aneffect on a sexually mature adult, profound adverse effects may occur in a child exposed prior toor during puberty. Similarly, if a drug or chemical affects long bone growth by disruption of theepiphyseal plates and subsequent endochondral bone formation, the resulting toxicity would bemanifested to a greater degree in an exposed child compared with an adult. Since the growth plateshave already fused in the adult, manifestations of toxicity due to growth plate disruption would beunlikely. This idea of critical exposure periods is one that resonates throughout this chapter. It iscrucial to understand the developmental processes occurring in a test system and the temporalrelationship to human development.If a chemical causes alterations because of a long-term cascade of events, it is more likely thata child exposed to the material will develop symptoms while an adult who is exposed may not.Ultraviolet irradiation and the subsequent development of skin cancer is an example of a delayedresponse, which would be more likely to occur the earlier the exposure took place. Other examplesof toxicants in which toxicity is manifested in increasing severity with age include methylmercuryand triethyltin. 67,89–91 By disrupting neurodevelopmental processes in the neonate, these compoundscause life-long aberrations in motor and sensory function that worsen with age.D. Use of Extant DataTable 8.2Factors responsible for differencesin risk assessment between childrenand adultsGrowth and developmentTime between exposure and manifestationDiet and physical environmentParameters of toxicity assessmentBiochemical and physiological responsesDrug and chemical dispositionExposure/behavior patternSource: Modified from Roberts, R.J., Similarities &Differences Between Children & Adults: Implicationsfor Risk Assessment, Guzelian, P.S., Henry, C.J., andOlin, S.S., Eds., ILSI Press, Washington, 1992, p. 11.1. Pharmacokinetic (PK) and Toxicokinetic (TK) Data in Adult AnimalsBecause of physiological differences between adult and juvenile systems, the bioavailability andbiotransformation of xenobiotics may be difficult to predict accurately in juveniles when only adultdata are available. 92 For example, young children and animals have the potential to be exposed tomuch higher levels of chemicals administered dermally (on a milligram per square meter) basisbecause of their large surface-area-to-body-weight ratio. 93Gastric pH in juvenile animals is generally higher than in adults, thereby increasing theabsorption of basic molecules and decreasing absorption of acidic molecules. Differences in gastrointestinal(GI) absorption and motility in the juvenile model may also affect the toxicologicalprofile of drugs and chemicals. Gastrointestinal motility in infants and neonatal children is low,whereas in older children GI motility is higher. For example, the average GI absorption rate of leadin infants is approximately fivefold higher than in adults. 94 The difference in absorption of leadbetween infants and adults, coupled with the rapid neurodevelopmental processes occurring during© 2006 by Taylor & Francis Group, LLC

278 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION30PigHuman25= Birth% Fat in Body201510Rat5RabbitGP0 10 30 100 300Days after ConceptionFigure 8.4Comparative ontogeny of fat. The human and guinea pig deposit fat prior to birth. (After Adolph,E.F. and Heggeness, F.W., Growth, 35, 55, 1971.)infancy (neural migration, brain topographical mapping), exacerbates the devastating results forthose who are exposed during the early postnatal period.Distribution of a drug throughout the body may also differ between adults and infants becauseof differences in plasma protein binding. Generally, newborns display less protein binding thanadults because of a combination of factors. First, albumin from infants has less affinity for certaindrugs, but adult binding capacity is developed during the first year of life. In addition, basiccompounds may not be bound to a high degree because of low levels of alpha-1-acid glycoproteinin the neonate. 9 Lastly, neonatal serum can contain high levels of free fatty acids that may causedissociation of drugs from albumin. 93Another factor that influences distribution of chemicals throughout the body is fat composition(see Figure 8.4 for comparative fat ontogeny). 95 Since many drugs are lipophilic, the body fatcontent (which changes during development) can influence the sequestration and availability offree drug. Total body water, extracellular fluid, and intracellular fluid levels also change duringmaturation and may lead to differences in drug distribution between developing and fully developedorganisms. 94 See Figure 8.5 and Figure 8.6 for comparative perinatal water content. In general,species with longer gestational periods have lower water contents at birth. 95Drug receptor expression and binding may also vary between mature and immature organisms.It has been postulated that the differences in response to phenobarbital and methylphenidate inchildren compared with adults result from differences in drug-receptor interactions. Phenobarbitalincreases the effectiveness of GABA (an inhibitory neurotransmitter) and inhibits the release ofglutamate (an excitatory neurotransmitter). Interestingly, phenobarbital is a sedative in adults butproduces hyperactivity in children. 76 Conversely, methylphenidate, a dopamine reuptake inhibitorthat also affects norepinephrine levels, is a sedative in children but a stimulant in adults. 76Metabolism of chemicals varies with age. Generally, metabolism in the human infant is somewhatslower than in the adult. 96 Although dependent on the specific enzyme, Phase I metabolismis fully developed in humans by 3 years of age, while the Phase II enzymes continue to developlater in life. The development of metabolic systems in laboratory animals follows patterns similarto those of humans, with lower enzyme activity at birth and a rapid increase in metabolic activitywithin the first few postnatal months. 94 Many enzymes have unique developmental expression© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 27995% Water in a Fat-Free Body908580RatMouseRabbitDogGuineaPigPigHuman7510 30100Days after Conception300Figure 8.5Comparative water content at birth. Longer gestations develop drier (“denser”) animals. (AfterAdolph, E.F. and Heggeness, F.W., Growth, 35, 55, 1971.)% Water in a Fat-Free Body95908580RatRabbitGPPig= BirthHuman751030100Days after Conception300Figure 8.6Comparative perinatal water content. Water fraction decreases with age in all species. (After Adolph,E.F. and Heggeness, F.W., Growth, 35, 55, 1971.)patterns. Therefore, the toxicological or pharmacological profile of a xenobiotic often differsdepending on age.Lastly, differences in excretion rate exist between infants and adults. 97 Generally, the glomerularfiltration rate is lower in young children, 87 which may increase the half-life of a compound that isreadily excretable by the kidneys. Biliary excretion follows a similar time course of developmentas renal clearance, with biliary excretion of compounds being low in the neonate and increasing© 2006 by Taylor & Francis Group, LLC

280 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONwith age. 97 To illustrate this point, biliary excretion of methylmercury is approximately 10-foldlower in neonatal rats than in adults, which may contribute to the longer half-life and increasedsensitivity observed in young rats. 98 This difference in excretion of methylmercury is a result oflower excretion of glutathione into the bile in neonatal animals. Since methylmercury is carriedinto the bile as a glutathione conjugate, the lack of glutathione excretion in turn causes lowerexcretion rates of methylmercury.2. Extant Adult Nonclinical Animal Toxicity DataPediatric administration of pharmaceuticals has been primarily based on adult toxicity data, anddoses have been adjusted for body weight or surface area without regard to differential sensitivitybetween children and adults. As previously discussed, one example of the differing responsivenessbetween adults and children to drugs is the lower toxicity of acetaminophen observed in children.Alternatively, children may be more sensitive than adults to certain compounds because of developmentalprocesses occurring in the growing child. For example, corticosteroids cause a decreasein growth velocity in children but do not have a similar effect in adults, as their bone growth hasalready ceased. 99 Other examples of juvenile toxicants that are less toxic to adults include phenobarbital,hexachlorophene, and valproic acid. It is becoming increasingly obvious that there aremany factors that must be considered when using adult data to evaluate the pharmacological andtoxicological profiles of drugs administered to young children.3. Stand-Alone Juvenile PK and TK StudiesAs mentioned earlier in the chapter, the pharmacokinetic profile of a drug can differ dramaticallybetween juveniles and adults and between animals and humans. Thus, the existing pharmacokineticdata of the drug from clinical and nonclinical adult studies may not be of value when predictingpediatric exposure to the drug, and the manufacturer may wish to conduct a stand-alone juvenilepharmacokinetic and toxicokinetic study prior to administration of the drug in the clinic. Thesestand-alone studies provide several advantages to the manufacturer, including the ability to comparepharmacokinetic data between the juvenile and adult animal to examine potential differences inexposure across ages that may require selection of different dosages for the different aged populationsand for the pediatric population in the clinic. Additionally, by evaluating differences inexposure profiles across the species, such studies can aid the manufacturer in determining whetherthe animal model chosen is an appropriate species for use in the definitive nonclinical juvenilestudy. Finally, the studies can provide limited toxicity data and may guide the investigator to selectadditional endpoints for evaluation in the definitive study.A. Types of StudiesIV. JUVENILE TOXICITY STUDY DESIGNJuvenile toxicity studies are generally conducted using one of three designs: (1) adapted adulttoxicity studies (usually reproductive study designs that are modified to address specific concernsin the juvenile), (2) general toxicity screens, and (3) focused or mode of action studies. Each ofthese study designs serves an important role in the nonclinical development of a drug that will beadministered to pediatric patients and for those chemicals for which human adolescent exposuremay be unavoidable. One or more of these study designs may be required to evaluate the juvenilesafety profile of the test agent for risk assessment in the pediatric population. It cannot be stressedenough that researchers need to stay in contact with the agency to which the study will be submittedand receive their input prior to conducting the study to ensure the acceptability of the finished study.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 2811. General Toxicity ScreenThe term “general toxicity screen” suggests that basic toxicity endpoints, such as clinical observations,body weight, food consumption, clinical pathology, and histopathology will be evaluated;however, these studies may encompass additional endpoints. While growth and pathology endpointsplay a key role in the general toxicity screening study, they may be accompanied by more specializedevaluations, including assessments of reproductive performance, behavior, and immune competence.These study designs often initiate dose administration prior to or at the time of weaning,and measure growth and development into young adulthood. Typically, these studies are conductedin rodents and require many animals to address relevant concerns. However, it is not uncommonfor these studies to be conducted in large animal models. A large animal model provides someadvantages over the rodent, including the capability of longitudinal assessments of various physiologicalendpoints. The general toxicity screen also commonly includes a posttreatment (or recovery)period, in which persistent and/or latent effects of the test article can be examined in comparisonto changes observed during treatment. An example of one such study design (the within-litter studydesign) is illustrated in Figure 8.7.The general toxicity screen serves an important purpose in the nonclinical juvenile study batterybecause it may reveal target organs not identified in adult toxicity studies. In addition, the generaltoxicity screen may serve to highlight areas that require further research to define the mode ofaction of the test article.2. Adapted Adult Toxicity StudyFor the purposes of this chapter, the authors define an adapted adult toxicity study to be areproductive toxicity study in which the design is modified to include a satellite juvenile phase(e.g., immunotoxicity, neurotoxicity, endocrine modulation screen, comet assay, or brain morphometricassessment). For some chemicals, the adapted adult design may be sufficient to addresspediatric concerns in lieu of conducting a general toxicity screen. These modified studies have theadvantage of a transgenerational exposure regimen, wherein the parental population is first exposed,followed by exposure of the juvenile population (either with or without concurrent parental exposure);this exposure model may more closely parallel the real-world situation for some chemicals.For example, the design for a two-generation reproductive toxicity study required by the EPA andthe OECD encompasses a 10-week premating exposure period that begins, for the F 1 generation,immediately following weaning. 55,100 During this period, growth and sexual developmental landmarksare evaluated. If there is concern that the infant and/or toddler may be directly exposed tothe test article, appropriate adjustments to the onset of dose administration for the F 1 generationmay be considered. The adjustments may consist of oral intubation of the offspring (in a studywhere administration is by gavage), modified feeding canisters (see Figure 8.8 for a diagrammaticrepresentation of a Lexan ® rat creep feeder) that would allow measurement of whole-litter offspringfood consumption (in a study where administration is via the diet), or specially designed inhalationchambers that allow litters to be exposed concurrently with dams (Figure 8.9). In the same way,for some pre- and postnatal development studies, concerns over the bioavailability of the test articlein the neonate dictate that a modified approach for dose administration must be used. For thesesituations, pharmaceutical companies have developed a strategy wherein maternal animals areadministered the test article during gestation, followed by a combination of maternal and offspringdose administration during the postnatal period.In cases where inadequate offspring exposure has been demonstrated following delivered dosesto the dam only, both the EPA and OECD have requested modifications to the developmentalneurotoxicity study design to incorporate direct administration of the test article to the F 1 offspringprior to weaning. 52,167 These requests have been made to the registrants for certain classes ofcompounds and reflect heightened public concern for special populations at risk.© 2006 by Taylor & Francis Group, LLC

282 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTotal of30/sex/groupSubset of10/sex/groupSubset of20/sex/groupAdditional6/sex/groupPND 4PND 7PND 21Cull to 4/sexBegin dosingWeanCull to 4/sexTK phlebotomy3/sex/groupPND 25Begin VP examPND 35Begin BPS examPND 62PND 63Motor activityStartle responseLearning and memoryEnd DosingClinical pathologyNecropsyHistologyMotor activityStartle responseLearning and memoryTK phlebotomy3/sex/groupPND 85LD 0LD 7PND 130Begin mating trialDams deliverNecropsy dams and pupsNecropsy paternal animalsAdministration PeriodFigure 8.7Within-litter study design. A robust within-litter design requires approximately 36 litters obtainedvia in-house mating or time-mated animals. VP, vaginal patency; BPS, balanopreputial separation.A specific type of adapted adult toxicity study design that has been proposed as an option forevaluating postnatal toxicity is the transgenerational protocol. This design has been proposed toevaluate effects of chemicals on the development of the reproductive system that may not otherwisebe detected by standard testing paradigms. This alternative design has been suggested because someendocrine-active chemicals were negative in standard multigeneration studies (likely because ofpostweaning selection of only one pup per sex per litter). 101 The dosing regimen for this transgenerationaldesign is in utero and lactational exposure (with the option of direct dosing of theoffspring), followed by necropsy of all offspring in adulthood. 102 The salient points to this designare that more offspring per litter are used (though fewer litters per group) and more endpointsrelated to endocrine-active compounds are evaluated. An important feature of this design is thatmultiple offspring per litter are examined following sexual maturity (since earlier examination may© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 28310 cm20 cmTop View19 cmBedding Material2 cmSide ViewFront ofshoebox6.5cm19 cm3.5 cmRear ViewFigure 8.8Construction of a Lexan ® creep feeder shoebox caging insert for rats. This system allows rat pupsaccess to treated feed but prevents access by the dam. Untreated feed for consumption by thedam must be offered in a separate cage to prevent preferential consumption of untreated diet bythe pups. During the second half of the lactational period, dams can be separated from their pupsfor at least 8 h with minimal impact on either.30 cmD C 14.1 cm C ABBFigure 8.9A 4.7-liter glass and anodized aluminum inhalation chamber. (A) Inlet pipe that carries air/testatmosphere at 2.5 to 3.0 l/min. (B) End cap with O-ring that seals chamber. (C) Dispersal platethat distributes atmosphere throughout chamber and prevents aerosols from affecting animalsdirectly. (D) Outlet pipe that attaches to manifold.not fully characterize effects on reproductive organs that are not yet fully developed), decreasingthe potential for false negatives of low-incidence responses. An example of this type of study(although without direct dosing of pups) is the one-generation extension study sponsored by theEPA that evaluated vinclozolin and di(n-butyl) phthalate. That investigation demonstrated enhancedsensitivity in detecting low-incidence responses as statistically and biologically significant effects(thus reducing the chance of false negatives). 1033. Mode of Action StudyMode of action studies are usually conducted when disruption of a specific organ system or endpointhas been demonstrated to be of concern in the juvenile and/or adult. Alternatively, mode of action© 2006 by Taylor & Francis Group, LLC

284 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONstudies may be undertaken when the registrant believes that the toxicity observed in a generaltoxicity screening study is species specific and not relevant to humans. These studies are designedto address specific issues and/or class effects, and do not necessarily involve large numbers ofanimals or long exposure periods. They do not generally include many endpoints; however, thoseparameters that are evaluated are typically much more closely examined than would be the casein a general toxicity screen. For example, in a study designed to examine the toxic potential of theangiotensin converting enzyme inhibitor, ramipril, 16 rat pups per sex per group were administeredthe test article on one of two discrete days during development (PND 14 or 21). Clinical chemistrieswere assessed together with macroscopic and microscopic examination of several organs 3 or 14days later. 92 In addition, the pharmacokinetic profile of the parent and metabolite were evaluatedon the days of dose administration. Although only limited numbers of animals were used (8 persex per group per age at necropsy), statistically significant changes in renal structure and functionwere noted, and dramatic differences were observed in the exposure of the juvenile rats to boththe parent molecule and metabolite when ramipril was administered on PND 14 or 21.B. Model SelectionSelection of the animal model is based on the study objective, the available data for the test articlein the model, the availability of historical control data, and the experience of the researcher andtesting laboratory with the model chosen. Several key factors must be considered when selectingthe animal model to be used in a juvenile toxicity study. These factors include the species, age,sex, sample size, and relevance to human development. 511. SpeciesA majority of nonclinical juvenile studies utilize the rodent as the animal model of choice for anumber of reasons. For adapted adult toxicity studies, the animal model is dictated by basic guidelinerequirements for the standard study, which in most cases tends to be the rodent. Moreover, ratsand mice are well characterized with regard to growth and development, given the large numberof reproductive, developmental, and general toxicity studies that have been conducted with thesespecies. However, very little historical control data, such as the ontogenic profile of clinicalchemistry parameters (refer to Table 8.3 for one example in rats) or the microscopic changes thatoccur in many organ systems during development, are available for more traditional toxicityendpoints in developing animals.Despite the paucity of historical control data for traditional toxicity endpoints in the developingrodent, there is a greater breadth of literature regarding the developmental changes that occur inthe rodent than in other species selected for juvenile toxicity testing (e.g., dog, swine, and nonhumanprimate). Furthermore, if the intent is to administer the test article during the period of adolescentdevelopment, larger animal species will typically require much longer exposure periods than therodent, given the longer developmental spans of larger species. Dog, swine, or nonhuman primatestudy designs also require more test article, housing space, technical involvement, and socializationthan a comparable rodent study. However, large animal models afford some key advantages in anonclinical juvenile study. For example, the larger circulating blood volume in these speciesprovides the opportunity to evaluate changes in clinical chemistry parameters very early in developmentand repeatedly over time in the same animal; longitudinal assessments of standard clinicalchemistry panels are extremely difficult, if not impossible, to conduct in rodents, because of thesmall circulating blood volume.2. Sample SizeThe number of animals required for a nonclinical juvenile toxicity study entails a balance betweenachieving the statistical power required to detect potentially adverse outcomes and the logistical© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 285Table 8.3Ontogeny of serum chemistry parameters in the Crl:CD ® (SD)IGS BR ratParameter PND 17 PND 24 PND 28 PND 35Albumin (g/dl)M 3.2 ± 0.2 3.8 ± 0.1 3.9 ± 0.2 4.1 ± 0.2F 3.4 ± 0.2 3.8 ± 0.2 3.9 ± 0.2 4.3 ± 0.2Total protein (g/dl)M 4.7 ± 0.2 4.9 ± 0.1 5.1 ± 0.3 5.5 ± 0.3F 4.9 ± 0.1 4.9 ± 0.2 5.1 ± 0.3 5.6 ± 0.4Total bilirubin (mg/dl)M 0.5 ± 0.3 0.2 ± 0.1 0.2 ± 0.1 0.1 ± 0.1F 0.4 ± 0.1 0.2 ± 0.1 0.1 ± 0.1 0.1 ± 0.1Urea nitrogen (mg/dl)M 18.7 ± 3.9 11.7 ± 2.9 11.9 ± 2.6 12.5 ± 2.1F 18.1 ± 3.4 12.8 ± 2.7 11.9 ± 2.4 12.8 ± 2.6Alkaline phosphatase (U/l)M 304 ± 27 333 ± 74 332 ± 65 368 ± 74F 302 ± 31 373 ± 92 342 ± 55 295 ± 89Alanine aminotransferase (U/l)M 21 ± 9 59 ± 9 58 ± 11 73 ± 16F 19 ± 6 56 ± 8 56 ± 13 62 ± 13Glucose (mg/dl)M 277 ± 12 264 ± 26 215 ± 28 259 ± 51F 284 ± 34 265 ± 42 235 ± 27 217 ± 30Cholesterol (mg/dl)M 183 ± 13 74 ± 9 78 ± 10 71 ± 9F 205 ± 22 76 ± 13 83 ± 14 76 ± 10Phosphorus (mg/dl)M 13.2 ± 0.4 13.3 ± 1.2 13.9 ± 1.1 13.4 ± 0.5F 13.6 ± 0.5 13.0 ± 0.7 13.1 ± 1.2 12.0 ± 1.1Potassium (mEq/l)M 8.65 ± 1.03 7.77 ± 0.91 7.33 ± 0.85 7.73 ± 0.75F 8.67 ± 1.17 7.57 ± 0.93 7.08 ± 0.55 6.9 ± 0.68Sodium (mEq/l)M 136 ± 2 143 ± 2 144 ± 4 144 ± 1F 137 ± 2 143 ± 2 145 ± 4 144 ± 2Aspartate aminotransferase (U/l)M 104 ± 24 130 ± 24 111 ± 18 116 ± 18F 109 ± 15 126 ± 15 119 ± 20 101 ± 20Creatinine (mg/dl)M 0.1 ± 0.05 0.1 ± 0.05 0.1 ± 0.00 0.2 ± 0.08F 0.1 ± 0.05 0.1 ± 0.05 0.1 ± 0.05 0.2 ± 0.05Glutamyltransferase (U/l)M A 0.9 ± 0.6 0.8 ± 0.5 0.8 ± 0.4F 0.9 ± 0.6 0.5 ± 0.2 0.9 ± 0.3 0.7 ± 0.5Globulin (g/dl)M 1.5 ± 0.1 1.1 ± 0.1 1.2 ± 0.2 1.4 ± 0.2F 1.4 ± 0.1 1.1 ± 0.1 1.2 ± 0.1 1.3 ± 0.2Data represent the means ± standard deviations of eight rats/sex/age, except for glutamyltransferase, which was often belowinstrument range for individual animals. (Collected at WIL Research Laboratories, LLC.)A = Below instrument range; M = males; F = females.difficulty involved in conducting the study. If the purpose of the study is to evaluate neurobehavioraldeficits that may result from juvenile rodent exposure to the test article, it may be of benefit to usea sample size of 20 animals per sex per group to provide enough power to detect statisticallysignificant changes in performance. 104 However, if the species of choice for the same study is thebeagle, animal costs, test article requirements, and the feasibility of conducting quantitative behavioralassessments (e.g., locomotor activity) may limit the sample size, increasing the likelihood ofType II errors.© 2006 by Taylor & Francis Group, LLC

286 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe availability of sufficient litters to produce the number of animals required for the study maybe a limiting factor, and overselection from a small number of litters to yield the desired number ofanimals per group may compromise the study (a discussion of litter-based selection will follow). Inmore focused study designs, smaller numbers of animals may be adequate to accurately assess thetoxic potential of the test article, as demonstrated in the ramipril study described previously. 92 Inreproductive and developmental toxicity studies adapted to address specific concerns in the juvenile,the availability of an adequate sample size is of less concern because these studies are typicallyconducted in the rodent, with approximately 20 to 30 litters per group available for analysis. Selectionof a subset of animals from these litters for juvenile testing is quite feasible in the adapted adult studydesigns, and the researcher is still capable of fulfilling the basic requirements of the standard study.A critical issue that must be addressed when determining the appropriate number of animalsfor use in a study is the effect of the litter on the response to test article exposure. It was demonstratedin the Collaborative Behavioral Teratology Study (CBTS) that offspring from the same litter respondmore similarly to a test article than do nonlittermates. 105 This response is known as the litter effect.Therefore, guidelines for developmental toxicity studies require that the litter be the experimentalunit for statistical and biological analysis. 54,106–110 Unfortunately, many researchers fail to carry thisrequirement into the postweaning period. Instead, littermates are treated as though they are notsiblings. Determination of the appropriate number of animals for use for a study should take intoaccount the number of litters available for selection. Attempts must be made to select animals fromas many litters as possible in order to create the required sample size for each group. This can beaccomplished by conducting within-litter selections, fostering, or selecting only one animal per sexper litter (between-litter selection) for analysis. Between-litter designs should only be consideredwhen culling is planned, when data from siblings are used to obtain a litter mean, or when siblingsare selected for different endpoints of analysis. The issue of litter-based selection is addressed inmore detail later in the chapter.C. ExposureWhen designing a juvenile toxicity study, the proper exposure paradigm must be determined. Thisinvolves selection of an acceptable route of administration, consideration of the frequency andduration of exposure and the timing of dose administration, and appropriate dose level selection.When feasible, the route of exposure should match the expected exposure scenario for the pediatricpopulation.1. Route of AdministrationChoosing an acceptable route of administration may seem an easy task for most test articles, inthat the most likely route of clinical administration or exposure to the human population wouldalready have been determined. However, there are many more challenges when the route ofadministration is selected for a juvenile toxicity study. The first issue that must be considered isthe feasibility by which the test article may be administered via a particular route. Although thecommon route of administration in preweaning animals is by gavage, oral intubation can betechnically challenging at some ages (Figure 8.10 and Figure 8.11). For example, in a PND 4mouse pup, the technical challenges involved in administering the test article can be extreme. Whenthe test article must be administered as a viscous suspension through a small gauge cannula, thesechallenges may be impossible to overcome. While a rat pup is several times larger than a mousepup of the same age (at PND 1, rats weigh approximately 6 g whereas mice weigh approximately2 g), the size of the animal and the susceptibility of the esophagus to perforation early in postnatallife cannot be ignored. The staff involved in administering doses to these young animals must bethoroughly trained; never assume that an individual who is capable of dosing an adult animal willbe able to dose a neonate of the same species. At the authors’ laboratory, staff members are required© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 2879 9:45 AM8 11:03 AMFigure 8.10Example of gavage technique for PND 4 to 10 rat pups, using 24-gauge stainless-steel dosingcannula.to intubate 30 pups once daily from PND 3 through PND 28, with no more than one death duringthe dosing period, to be qualified to gavage juvenile rodents on study. Based upon this trainingregimen, the authors have found that the success rate for oral intubation of juvenile rats is nearly100% (Table 8.4).In the authors’ experience, a dosage volume of 5 to 10 ml/kg has been found to be ideal. Experimentationmay be required to determine dosage volumes that are technically feasible for the vehicle.Finally, the time required to administer a dose should be considered when choosing the route.This includes predosing activities, as well as the actual dosing of the animal. Since developinganimals have difficulty maintaining core body temperature early in life and time away from thedam can have a negative impact on development, the time it takes to identify and segregatelittermates prior to dosing should be minimized. If prolonged separation from the dam is unavoidable,the environment supporting the pups should be warmed appropriately.Other routes of administration may also be considered, including dietary, dermal, inhalation,and intravenous. Each of these routes poses different technical challenges to the researcher, dependingon the species and age of animal involved. For example, dietary administration may seem© 2006 by Taylor & Francis Group, LLC

288 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION13 11:45 AM13 11:45 AMFigure 8.11Example of gavage technique for PND 21 to 35 rat pups, using 21-gauge Teflon ® dosing cannula.appropriate if it is the likely route of human exposure, but the age at which pups start to consumesolid food must be considered. If dietary exposure commences prior to weaning, specialized feedingapparatuses may be required to prevent maternal consumption of the test diet, while allowingoffspring consumption (refer to Figure 8.8). In this scenario, untreated diet must be provided tothe dam in a separate cage, and the dam must be removed from the litter each day for up to eighthours in order to prevent consumption of the untreated maternal diet by the offspring. Because ratpups do not consume appreciable quantities of diet prior to PND 14, the dietary route is notappropriate when exposure is required early in postnatal life. Furthermore, a within-litter testingscheme (with all groups represented in the litter) employing the dietary route of administration isnot possible prior to weaning.Inhalation exposure may be the most appropriate choice in some circumstances, but issues ofanimal welfare and inhalation capacity of the test species may make this choice difficult toimplement. For example, the inhalation route has been used for modified developmental neurotoxicitystudies, which typically use the rat model. 111 In this exposure paradigm, both the dam and© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 289Table 8.4Flexible cannula intubation error rates in juvenile Crl:CD ® (SD)IGS BR ratsAge atIntubationNumberofAnimalsNumber ofDosesAdministeredConfirmedIntubationErrors aSuspectedIntubationErrors bIntubationError Rate,%PND 4-31 256 7,168 4 0 0.056PND 7-13 3681 25,767 0 5 0.019PND 7-34 256 7,936 0 0 0.000PND 14 or 21 376 376 0 0 0.000PND 14-41 1800 50,400 0 0 0.000PND 19-21 212 636 2 0 0.314PND 21-34 520 14,560 0 0 0.000PND 22-41 285 5,700 0 0 0.000All ages combined 7386 112,543 6 5 0.010aIntubation errors were confirmed at necropsy by observation of a perforated esophagus.bSuspected intubation errors resulted in animal death, with reddened esophagus and/or foamy contentsin trachea and/or lungs noted at necropsy.Source: Data were collected at WIL Research Laboratories, LLC (9/22/00 – 6/10/04).litter are exposed simultaneously to the test article during early postnatal life of the pups untilweaning. Therefore, specialized whole-body inhalation chambers must be used for the animalsduring exposure (refer to Figure 8.9). If the pups are exposed without the dam, the temperature ofthe inhalation chamber must be maintained appropriately since neonatal rats are inefficient atregulating core body temperature, although this is not recommended early in lactation. The inhalationroute of administration for juvenile rodents must be whole-body until the animal reachessufficient size to make head- or nose-only inhalation possible. The whole-body route of exposureintroduces the possibility that the juvenile and/or the maternal animal (if concurrently exposed)may also be exposed orally (during grooming) and dermally. Finally, neonatal animals in thesemodified inhalation chambers may be exposed to lower than desired concentrations of the testarticle as a result of filtration when nursing. 77 In contrast, exposure by the inhalation route presentsfewer technical difficulties when using larger species because they may be exposed via head- ornose-only methodology, with some acclimatization.Less common routes of administration, such as intravenous, present technical challenges thatare related to animal size. It is technically more feasible to administer a test article intravenouslyto a puppy or a piglet than to a rodent pup. There are more sites into which the dose may beadministered in a larger animal (e.g., jugular, saphenous, marginal ear, or femoral vein) than in arodent. If intravenous administration is required and the animal of choice is the rodent, theinvestigator may have to compromise by using older animals at the onset of dose administration.Continuous infusion would not be possible in rodent juvenile studies that initiate prior to weaning.Moreover, the logistical challenges involved in continuous infusion in any species may be dauntinggiven the potential need to provide new catheters at multiple intervals throughout the study tocompensate for growth of the animal.The dermal route is another option for administration of the test article to the juvenile, althoughInternational Life Sciences Institute (ILSI) does not recommend this route for juvenile rodentsbecause of architectural differences between rodent and human skin. 77 The use of this route shouldbe limited to the postweaning period to prevent unintended oral administration to the dam duringgrooming and/or maternal rejection of the offspring because of changes in olfactory cues causedby application of the material. Moreover, if siblings are gang-housed after weaning, a period ofseparation may be required for those studies requiring dermal applications. Finally, to preventingestion of the test article, Elizabethan collars (see Chapter 7, Figure 7.5A) may be required, andthe size of these collars should be monitored closely and appropriately adjusted to accommodatethe growth of the animal during the dose administration period.© 2006 by Taylor & Francis Group, LLC

290 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONOther routes of administration that are less commonly employed in nonclinical juvenile studiesare subcutaneous, intramuscular, and intraperitoneal injection. These routes are adaptable to thejuvenile animal, depending upon the age and species of the model.2. Frequency and Duration of ExposureThe frequency of dose administration prior to weaning will most often be once daily, with dosagesbased on daily body weight determination. Previous studies in adult animals, which may includekinetic information, will aid in developing the most appropriate regimen for the test article. However,the researcher is cautioned not to overlook differences in absorption, distribution, metabolism, andelimination (ADME) profiles between juvenile and adult animals, such as differences in the concentrationand/or activity of key metabolic pathways. Carefully designed pharmacokinetic studiesin juvenile rodents have revealed marked differences in clearance rates, peak plasma concentrations,and area under the curve (AUC) values for test articles when administered at different ages in thesame model. 92 These differences clearly demonstrate that developmental changes occurring inmetabolic pathways may result in significant quantitative differences in systemic exposure andinfluence the selection of exposure frequency.In addition, the treatment period must be considered when juvenile toxicity studies are designed.Consideration should be given to covering the entire period of growth, from the neonate to thesexually mature animal. In rodent models, test article administration from weaning until or beyondthe age when mating would occur may be required. Alternatively, administration may begin atapproximately PND 4 to 7, which is roughly equivalent to a preterm human infant, and continueuntil just after the males enter puberty, at approximately 6 to 7 weeks of age. At a minimum, thistreatment period, coupled with the adult general and reproductive toxicity studies that are alreadyrequired, would encompass all periods of the animal’s life span. If there is concern that a test articlemay reveal oncogenic potential as a result of juvenile (or even prenatal) exposure, a subset of animalsfrom a pre- and postnatal development study may be selected to become part of a carcinogenicity study,with dose administration beginning immediately following weaning. Dose administration throughoutgrowth may be impractical in large animal species (longer duration and higher cost), for which theentire growth and development period may span several months to years. At a minimum, in situationswhere a large animal model is the species of choice, the period of dose administration for a generalscreening study should cover the period of target organ system development (if known).A recovery (or posttreatment) period may be included in the design of the study. For some testarticles, changes in functional parameters are expected because of the pharmacological action.Therefore, there may be less concern regarding acute effects of the test article (e.g., locomotoractivity changes immediately following administration of a sedative) and more concern aboutwithdrawal, persistent, and/or latent effects. In addition, when the posttreatment period containsthe same assessments as those in the treatment period, comparisons can be made regarding theseverity of the changes, persistence of these changes, and their relationship to the expected pharmacologyof the test article. If functional assessments are not included in the treatment period,interpretation of findings during the posttreatment period becomes much more difficult. In thisscenario, responses in the juvenile study must be compared with those that may have been observedin studies with adult animals.3. Dose SelectionFor the general toxicity screen, the draft FDA guidance document recommends that the high doseproduce frank toxicity and that a no-observed-adverse-effect level be identified, if possible. 51However, when enzymatic pathways involved in metabolism of the test article are immature,different sensitivities may be observed between the adult and the juvenile. Moreover, functionalchanges in developing organ systems can create periods of greater sensitivity in the juvenile, with© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 291resulting toxicities that would not be evident in the adult. Therefore, dose levels used for adultanimals may not be applicable to juvenile animals. As an example, the benzodiazepines produceparadoxical responses in opposition to their anxiolytic properties, including convulsions. 112 Thoseconvulsions were characterized in preweaning rats and were postulated to result from differencesin the Type 1 and Type 2 benzodiazepine receptor sites expressed in immature and mature rats.Another example of apparent differential sensitivity between adults and juveniles relating toenzyme system competence is that of the pyrethroids. 113 As an example, the LD 50 for cypermethrin(a type II pyrethroid) is far lower in juvenile rats than in adults of the same strain. However, whentri-o-tolyl phosphate was used to inhibit drug metabolism prior to treatment in the adult to mimicthe inherently lower esterase level in the juvenile rat, the observed LD 50 at both developmentalstages became similar. 114 An interesting age-related manifestation of toxicity of a type II pyrethroidevaluated in the authors’ laboratory in the juvenile rat is a gait abnormality that was termed“carangiform ataxia.” This abnormality is a fish-tailing gait wherein the forepaws pedal in a forwarddirection while the posterior third of the body (behind the rib cage) rapidly oscillates in a fishlikemotion with maximal lateral abdominal flexion. Carangiform ataxia is temporally limited andgenerally corresponds to the time of locomotor function development in the rat during whichpivoting is observed (approximately PND 6 to 11). 115 This age-specific manifestation of pyrethroidtoxicity is speculated to result from the limited neural development in the hindquarters of the ratat this age; there is no known adult correlate. This gait abnormality was observed in juvenile ratsat a dose level approximately 10-fold lower than the dose level that produced maternal toxicity,emphasizing both the apparent sensitivity of juveniles and the possibility of unique manifestationsof toxicity based on physiological development. For current thinking and future directions in thestudy of developmental neurotoxicity of pyrethroid insecticides, the reader is referred to a recentreview article by Shafer et al. 116Dose selection in a general sense requires careful consideration of adult nonclinical data, datafrom analogs, existing human exposure data, and knowledge of the organ systems that are developingduring the proposed period of dose administration. In addition, well-designed range-findingstudies are essential for the selection of acceptable dose levels. In these studies, gross changes inbody weight, food consumption, and clinical chemistry, and macroscopic changes in organ structureand weight can be evaluated. Although less commonly included in dose range-finding studies inadult animals, pharmacokinetic profiling of the test article can be extremely valuable in the selectionof dose levels for the definitive juvenile study. The pharmacokinetic profile (e.g., C max , AUC) inthe juvenile after direct administration can be compared with previously developed profiles in adultanimals. Any potential differences that are identified can provide guidance as to the modificationsin the dosage levels, dosage regimen, vehicle, etc., that may be required. Moreover, to date manynonclinical juvenile studies are being conducted after some initial data have been developed in thehuman. In these circumstances, inclusion of a pharmacokinetic phase in the dose range-findingstudy will guide the researcher to select dose levels that will more accurately reflect the humanexposure scenario or that may even indicate the appropriateness of the species as a model for thehuman situation.As stated previously, the FDA draft guidance document recommends that the high dose shouldproduce frank toxicity. 51 One way to ensure that toxicity is observed is by reaching the maximumtolerated dose (MTD). Since clinical studies in the pediatric population will only characterize theefficacy of the drug, and perhaps some side effects, reaching an MTD in the nonclinical juvenilestudy will allow for an examination of the range of potential adverse outcomes following exposure.However, several challenges may be imposed by the use of the MTD. For example, reaching anMTD in an adult study will result in toxicity that may be limited to a single organ system.Conversely, administration of the test article to a juvenile animal population at a dose level at orabove the MTD may not only result in an adverse outcome in a particular organ or organ systembut may also result in confounding growth retardation. The authors have experience using doselevels that were far above the MTD for the juvenile. One example was a developmental neurotoxicity© 2006 by Taylor & Francis Group, LLC

292 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONstudy in rats of methimazole, a drug used to treat hyperthyroidism. 117 Although not strictly a juveniletoxicity study, in this design the test article was administered in the drinking water from gestation day6 through lactation day 21. Body weight gain was decreased early in the lactational period when pupsdid not consume water; this effect was exacerbated later in the lactational period when pups began toindependently consume water containing the test article. Growth retardation can have a profoundinfluence on the development of many organ systems and on the ability to interpret direct effects ofthe test article compared with secondary changes resulting from the growth retardation.D. Organization of Test GroupsSeveral organizational models are available, each with advantages and disadvantages. These modelsare based on the principle of the litter as the experimental unit of testing and analysis and includethe within-litter, between-litter, fostering, and one pup per sex per litter designs.1. Within-Litter DesignThe within-litter design is based on the idea that all dosage groups should be equally representedin each litter (refer to Figure 8.7 for a schematic representation of a study design using the withinlitterapproach). Depending on the number of dose levels in a particular study design or the sizeof the litter, a true within-litter design may be difficult to achieve. In these cases, a modifiedapproach, the split-litter design, may be employed. In such designs, no two same-sex siblings areassigned to the same dosage group, and multiple litters are required to obtain a sample size of oneper sex for each group. In either case, the litter effect is distributed across all dosage groups. Referto Figure 8.12 for a depiction of the distribution of pups in within-litter and split-litter designs.AEach DamGroup 1 male pup Group 2 male pup Group 3 male pup Group 4 male pupGroup 1 female pup Group 2 female pup Group 3 female pup Group 4 female pupBDam1Dam 2Group 1male pupGroup 2male pupGroup 3male pupGroup 4male pupGroup 5male pupGroup 6male pupGroup 1male pupGroup 2male pupGroup 1female pupGroup 2female pupGroup 3female pupGroup 4female pupGroup 5female pupGroup 6female pupGroup 1female pupGroup 2female pupCEach DamGroup 1 male pup Group 1 male pup Group 1 male pup Group 1 male pupGroup 1 female pup Group 1 female pup Group 1 female pup Group 1 female pupFigure 8.12Distribution of pups in various study designs. (a) For within-litter designs, each treatment groupis represented in each litter by one male and one female pup. (b) For split-litter designs, eachtreatment group is represented by no more than one male and one female pup per litter, but allgroups may not be represented in each litter. (c) For between-litter designs, all pups in the litterare assigned to the same treatment group.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 293Considering that the response to a test article is more alike in siblings than in offspring fromdifferent litters, 105 the within-litter approach provides the advantage of exposing genotypicallysimilar offspring to different levels of the test article. Another advantage of the within-litter designis that litter-specific maternal and environmental factors are appropriately distributed across dosagegroups. This design also provides ethical and practical advantages, in that it reduces the numberof animals (dams with litters) required for the study and minimizes the time required for doseadministration and data collection.However, the within-litter design has some key disadvantages, including the possibility of crosscontaminationbecause of exposure to the test article or metabolite when administration beginsprior to weaning. For example, siblings assigned to the control or lower dosage groups may consumefeces from siblings assigned to higher dose groups, or they may be dermally exposed to the testarticle through contact with contaminated bedding. Also, the dam may be inadvertently exposedto the test article while grooming the neonate (or if she has consumed one of her young), and theoffspring may then be exposed to the test article while nursing.2. Between-Litter DesignRather than representing all dosage groups in the litter, the between-litter design assigns the entirelitter to the same dosage group (refer to Figure 8.12). Compared with the within-litter design, thisis a very straightforward design in principle. Between-litter designs have the advantage of decreasedchances of cross-contamination of dosage levels and simplification of dose administration procedures.Figure 8.13 presents a schematic representation of a study design using the between-litterapproach.For the between-litter design, it is expected that the parental genetics and maternal care wouldbe similar for all offspring in the litter. A disadvantage of this design is that it requires a largenumber of animals. In a rodent juvenile toxicity screen with a sample size of 25 per group, thenumber of progeny being dosed each day can exceed 1500, depending on whether standardizationof litters is performed. This alone can create a logistically challenging study design. When theendpoints of analysis are also considered, these studies may become too demanding for manylaboratories to conduct properly.Between-litter designs are incorrectly implemented when the litter effect is ignored (i.e., thepup is considered the statistical unit instead of the litter). For a detailed exposition of the littereffect in statistical analysis, refer to Chapter 9 of this text. For example, when incorrectly applied,each sibling is assigned to the same dosage group as a separate unit for statistical analysis. Ratherthan decreasing the effect of the litter, incorrect use of the between-litter design can enhance theeffect of the litter on the observed response. For this reason, between-litter designs should only beused when the researcher plans to cull the litter with increasing age, when data from littermatesare used to obtain a litter mean at those times when the entire litter has been examined for aparticular endpoint, or when littermates are selected for different endpoints.Few studies have been designed to specifically evaluate the same endpoints by use of both thewithin-litter and between-litter designs. In one such study, pups treated with 9 mg triethyltin/kgfrom the within-litter design weighed less than those in the between-litter design from PND 14 to20. In addition, the hyperactivity induced in young adult animals by triethyltin in the within-litterdesign was higher than that in the between-litter design. 118 Conversely, in another study, 6-hydroxydopamineinduced hyperactivity was less in a modified within-litter design (half of pups treated)than it was in a between-litter design (all pups treated). In addition, 6-hydroxydopamine-treatedpups in the within-litter design demonstrated better avoidance than those in the between-litterdesign. 119 Although these results do not conclusively demonstrate that the within-litter design ismore scientifically relevant, they do illustrate that different results may be obtained depending onthe organization of the test groups. No definitive regulatory guidance has been provided to indicatewhich approach should be used.© 2006 by Taylor & Francis Group, LLC

294 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTotal of25/litters/groupSubset of25/sex/groupSubset of25/sex/groupSubset of 25–50/sex/groupPND 4Cull to 4/sexPND 7Begin dosingTK phlebotomy3/sex/groupPND 21WeanPND 25Begin VP examPND 35Begin BPS examMotor activityStartle responseLearning and memoryPND 62End DosingTK phlebotomy3/sex/groupPND 63Clinical pathologyNecropsyHistology1/2 for CNS perfusion1/2 for general tissuesMotor activityStartle responseLearning and memoryPND 85LD 0LD 7PND 130Begin mating trialDams deliverNecropsy dams and pupsNecropsy paternal animalsAdministration PeriodFigure 8.13Between-litter study design. A robust between-litter design requires approximately 100 littersobtained via in-house mating or via time-mated animals.3. Fostering DesignThe fostering approach has been employed by some researchers in an effort to minimize the effectof the litter and genetic bias on the study. In this design, offspring are fostered to new mothersimmediately after birth. Although on the surface this approach may seem relatively simple, it canbecome difficult to track and assess the effects of genetics or gestational influences on the observedresponses to a test article. A true fostering study does not merely replace the litter of one femalewith that of another. Instead, it replaces the original litter of one female with one offspring fromeach of several litters of other females (Figure 8.14). Fostering also requires that a sufficient numberof litters be born on the same day. If insufficient litters are available for fostering on a given day,© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 295A: Prior to Fostering B: After FosteringDam ADam BDam CDam ADam BDam CA1 A2 A3 A4A5 A6 A7 A8B1 B2 B3 B4B5 B6 B7 B8C1 C2 C3 C4C5 C6 C7 C8B1 C2 D3 E4F5 G6 H7 I8C1 D2 E3 F4G5 H6 I7 A8D1 E2 F3 G4H5 I6 A7 B8Dam DDam EDam FDam DDam EDam FD1 D2 D3 D4D5 D6 D7 D8E1 E2 E3 E4E5 E6 E7 E8F1 F2 F3 F4F5 F6 F7 F8E1 F2 G3 H4I5 A6 B7 C8F1 G2 H3 I4A5 B6 C7 D8G1 H2 I3 A4B5 C6 D7 E8Dam GDam HDam IDam GDam HDam IG1 G2 G3 G4G5 G6 G7 G8H1 H2 H3 H4H5 H6 H7 H8I1 I2 I3 I4I5 I6 I7 I8H1 I2 A3 B4C5 D6 E7 F8I1 A2 B3 C4D5 E6 F7 G8A1 B2 C3 D4E5 F6 G7 H8Figure 8.14Fostering procedure. (A) Prior to fostering; a total of at least nine litters of eight pups on one dayare required for a fostering procedure that would create an entirely new litter. B. After fostering;no pups remain with original dams and no siblings are present in the fostered litter.then all the litters born on that day cannot be fostered. Therefore, careful planning must be usedto ensure efficient use of litters. Finally, fostering does not remove the effect of the litter entirely.Rather, the care provided by the foster dam is as important to the litter effect as that provided bythe birth dam. Thus, using the fostering approach does not entirely alleviate concerns of using thebetween-litter design, although genetic factors should be evenly distributed. The authors haveconducted several studies using the fostering design in rats and have observed that an environmentallitter effect was imposed on the fostered litter such that pups in the fostered litter gained weightat a similar rate. If the within-litter design is to be used, the fostering approach is not recommendedbecause it offers very little improvement in control of bias but increases the complexity of thestudy. Moreover, for logistical reasons, the fostering approach is not recommended for studies withlarge numbers of litters.4. One Pup per Sex per Litter DesignThe final organizational design that is used in juvenile toxicity studies is the selection of one pupper sex per litter. This design is the preferred approach for selecting a subset of offspring from apre- and postnatal development study. This approach eliminates the chance for cross-contamination(if the animals are group-housed) and avoids the large number of animals that would be dosed ina between-litter design.E. Potential Parameters for Evaluation1. Growth and DevelopmentThe FDA’s current draft guidance states that nonclinical juvenile studies should include appropriategrowth measurements, such as body weight, growth velocity, crown-rump length, tibia length, andorgan weights. 51 The utility of body weight and body weight change as sensitive measures oftoxicity has been previously described in other texts and therefore is not discussed in detail here.© 2006 by Taylor & Francis Group, LLC

296 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION450400Body Weight (g)350300250200150100500FemalesMales10 11 12 13 14 15 16 17 21 28 35 42 49 56 63 70 72Postnatal DayFigure 8.15Body weight development of the Crl:CD ® (SD)IGS BR rat. Each body weight interval representsthe mean weight (± S.E.M.) of litters by sex (n = 20 to 25 litters per interval) collected at WILResearch Laboratories, LLC (2003–2004).Body Weight (kg)1211109876543210MalesFemales1 6 12 18 24 30 36 42 56 70 84 98 112 126 140 154 168 182Age (days)Figure 8.16Body weight development of the beagle dog. Each body weight interval represents the meanweight (±S.E.M.) of litters by sex (n = 11 to 16 litters per interval) collected at WIL ResearchLaboratories, LLC (2001–2002).An important factor regarding body weight change in juvenile study designs is the relatively rapidgrowth during the early postnatal period (Figure 8.15 and Figure 8.16 illustrate growth curves forrats and dogs, respectively). Body weight in the rat pup increases by approximately twofold betweenPND 1 and 7 and approximately threefold between PND 7 and 21. In the postweaning period, malerat body weight increases twofold between approximately PND 28 and 42 and again between PND42 and 70, after which body weight gain is not so dramatic. Female rat postweaning body weightdoubles between approximately PND 28 and 49, but increases only approximately 30 to 40%between PND 49 and 70. Consequently, body weight should be measured frequently prior to 10weeks of age (at least twice weekly) to evaluate subtle changes that may not be detected by useof longer intervals. Furthermore, frequent body weight measurements will allow more accuratedosage calculations during rapid periods of growth.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 297PND 3 PND 9 PND 42 PND 70 PND 139Figure 8.17Examples of full-body DEXA scans of anesthetized beagle dogs from each postpartum developmentalage category (WIL Research Laboratories, LLC, 2001–2002).One example illustrating differential toxicity on growth and development with age and gendercomes from a study of methylphenidate. When administered to female rats from PND 5 through24 via subcutaneous injection (s.c.) once daily (35 mg/kg/day), the reductions in mean body weight,femur length, and pituitary weight were approximately 9%, 5%, and 16%, respectively, and werestatistically significant compared with controls on PND 25. 120 After a 30-day recovery period, meanbody weight, femur length, and pituitary weight in the methylphenidate-treated females were similarto those of control animals. These data in females were consistent with data from previous studiesin male rats. 121,122 However, no such effects were observed in PND 55 female rats treated fromPND 35 through 54. 120 When methylphenidate was administered similarly (35 mg/kg/day, s.c.)from PND 35 through 54 to male rats, a statistically significant decrease in mean body weight ofapproximately 10% by PND 55 was seen (no effect was observed on mean femur length or pituitaryweight).In addition, evaluation of the overall growth and/or function of organ systems that developpostnatally (e.g., skeletal, renal, pulmonary, neurological, immunological, and reproductive systems)is recommended by the FDA guidance document. 51 As suggested above, additional directgrowth measurements, such as crown-rump length or long bone length (tibia or femur), can beassessed longitudinally. Bone growth can be evaluated by noninvasive techniques, such as dualenergy x-ray absorptiometry (DEXA) or computerized tomography (refer to Figure 8.17 for anexample of DEXA beagle full-body scans). A recent review comparing important developmentalmilestones demonstrated that patterns of postnatal skeletal growth and development betweenhumans and some animal models are similar. 123 Additional measures of growth, including craniofacialdimensions, femur length, and body length following exposure to radiation and/or chemotherapeuticcompounds have been described in an animal model. 4Changes in body composition may be considered in addition to any changes in body weight injuvenile animals (see Table 8.5 for beagle reference data). Total body composition (including bonemineral density and content, lean mass, and total body fat) can also be evaluated longitudinallythrough imaging techniques such as DEXA. Water content is very high during early prenataldevelopment, decreasing perinatally and with advancing age (refer to Figure 8.5 and Figure 8.6). 95Consequently, preweaning animals have higher water content and less fat than mature animals. Thiscan result in a larger volume of distribution for water soluble compounds than in older animals. 1 Withrespect to fat content, only the guinea pig and humans begin depositing fat prenatally (Figure 8.4). 95© 2006 by Taylor & Francis Group, LLC

Table 8.5Total body composition by DEXA for beagle dogsPostnatal Day aParameter b 3 9 21 30 42 70 115 139BMD (g/cm 2)M 0.23 ± 0.014 (9) 0.32 ± 0.016 (15) 0.38 ± 0.033 (14) 0.41 ± 0.036 (14) 0.47 ± 0.031 (14) 0.56 ± 0.033 (14) 0.67 ± 0.027 (14) 0.7 ± 0.028 (10)F 0.24 ± 0.011 (8) 0.33 ± 0.019 (16) 0.39 ± 0.025 (14) 0.42 ± 0.029 (14) 0.47 ± 0.023 (14) 0.55 ± 0.02 (14) 0.64 ± 0.018 (14) 0.66 ± 0.049 (10)BMC (g)M 3.2 ± 1.07 (9) 5.2 ± 2.37 (15) 15.6 ± 6.67 (14) 25.6 ± 9.13 (14) 44.9 ± 11.89 (14) 112 ± 25.96 (14) 197.2 ± 32.49 (14) 244.3 ± 35.17 (10)F 3.3 ± 0.87 (8) 5.2 ± 1.55 (16) 15.6 ± 5 (14) 24.6 ± 6.9 (14) 43.6 ± 8.56 (14) 104.1 ± 15.49 (14) 171.2 ± 24.76 (14) 213.9 ± 29.49 (10)Soft tissue (g)M 410 ± 65.7 (9) 658 ± 141 (15) 1066 ± 237.5 (14) 1374 ± 305.6 (14) 2123 ± 397.9 (14) 3881 ± 694.2 (14) 6227 ± 1059.6 (14) 6952 ± 991.8 (10)F 406 ± 49.7 (8) 654 ± 102.6 (16) 1077 ± 210.4 (14) 1319 ± 236.4 (14) 2027 ± 320.1 (14) 3599 ± 439.8 (14) 5512 ± 849.9 (14) 6227 ± 856.6 (10)Lean (g)M 376 ± 56.3 (9) 602 ± 120.8 (15) 965 ± 190.9 (14) 1258 ± 257.7 (14) 1881 ± 306.9 (14) 3215 ± 536.6 (14) 5607 ± 786.9 (14) 6263 ± 739.3 (10)F 374 ± 41.5 (8) 600 ± 94.7 (16) 966 ± 170.8 (14) 1172 ± 260.2 (14) 2511 ± 2831 (14) 3056 ± 348 (14) 5090 ± 685.1 (14) 5739 ± 698.6 (10)Fat (g)M 35 ± 17.4 (9) 57 ± 31.7 (15) 101 ± 65 (14) 117 ± 62.3 (14) 242 ± 128 (14) 686 ± 177.2 (14) 621 ± 323.2 (14) 689 ± 313.9 (10)F 32 ± 14 (8) 54 ± 22.5 (16) 111 ± 62.4 (14) 118 ± 68.2 (14) 231 ± 117.5 (14) 542 ± 128.8 (14) 422 ± 194.5 (14) 488 ± 210.7 (10)Fat (%)M 8.3 ± 3.3 (9) 8.2 ± 3.7 (15) 8.9 ± 4.06 (14) 8.1 ± 2.93 (14) 10.9 ± 4.37 (14) 17.4 ± 2.15 (14) 9.5 ± 3.38 (14) 9.5 ± 3.21 (10)F 7.9 ± 2.79 (8) 8.1 ± 3.29 (16) 9.9 ± 4.07 (14) 8.5 ± 3.52 (14) 11.1 ± 4.67 (14) 15 ± 2.27 (14) 7.3 ± 2.33 (14) 7.5 ± 2.54 (10)DEXA BW (g)M 414 ± 66.5 (9) 663 ± 142.9 (15) 1081 ± 243.5 (14) 1400 ± 313.9 (14) 2168 ± 408.5 (14) 3993 ± 718.8 (14) 6425 ± 1088.8 (14) 7196 ± 1024 (10)F 410 ± 50.4 (8) 659 ± 103.5 (16) 1093 ± 214.8 (14) 1343 ± 242.2 (14) 2071 ± 328.2 (14) 3703 ± 453.8 (14) 5683 ± 872.7 (14) 6441 ± 882.3 (10)Scale BW (g)M 351 ± 56.7 (9) 572 ± 118.2 (15) 990 ± 223.9 (14) 1232 ± 271.9 (14) 1951 ± 353.5 (14) 3595 ± 677.6 (14) 6492 ± 1190.8 (14) 7363 ± 1099.8 (10)F 347 ± 50.7 (8) 570 ± 87.2 (16) 993 ± 194.8 (14) 1215 ± 224.2 (14) 1817 ± 241 (14) 3265 ± 409.7 (14) 5782 ± 965 (14) 6641 ± 921.9 (10)BW diff. (%)M 15.2 ± 2.27 (9) 13.4 ± 5.68 (15) 8.4 ± 1.4 (14) 11.4 ± 4.78 (14) 9.2 ± 7.35 (14) 10.1 ± 3.93 (14) –0.9 ± 2.79 (14) –2.2 ± 1.75 (10)F 15.4 ± 5.3 (8) 13.3 ± 5.01 (16) 9.2 ± 1.38 (14) 9.6 ± 1.9 (14) 11.8 ± 4.7 (14) 11.8 ± 2.45 (14) –1.6 ± 2.8 (14) –3.2 ± 2.38 (10)aValues are expressed as: mean ± SD (N = number of litters).bBMD, bone mineral density; BMC, bone mineral content; Soft tissue, soft tissue [lean (g) + fat (g)]; DEXA BW, body weight calculated from the DEXA total body scan [soft tissue(g) + BMC (g)]; Scale BW, body weight recorded from a scale just prior to scanning; BW Diff., percent difference between the DEXA calculated body weight and the scale bodyweight [DEXA BW (g) – Scale BW (g)]×100, M, male; F, female.Source: Data were collected at WIL Research Laboratories, LLC (2001-2002).298 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 2992. Food ConsumptionMonitoring food consumption in juvenile studies is a general measure of animal health and apotential indicator of toxicity, just as it is in other bioassays. During the preweaning period, it isnot typical to evaluate food consumption for individual offspring. If animals are single-housed afterweaning, food consumption can be reliably evaluated. As shown in Figure 8.18, food consumption,starting on PND 28, increases throughout the growth phase in the rat, if calculated on a gram peranimal per day basis (although females appear to plateau around PND 39). Interestingly, foodconsumption in the rat actually decreases over time if compared relative to body weight (refer toFigure 8.18), with remarkable similarity between males and females. This decrease in food consumption,when calculated on a gram per kilogram body weight basis, has also been observed inhumans. 94Species-specific considerations should be made in the design and interpretation of food consumptiondata. One consideration is that feeding patterns in rodents are linked to circadian rhythms,thus most of the food intake occurs during the dark cycle. In contrast, humans, minipigs, nonhumanprimates, and dogs are not influenced in this manner.3. Serum Chemistry and HematologyThe current FDA draft guidance for juvenile studies states that clinical pathology can be useful, 51but these evaluations may be limited by the technical feasibility of obtaining adequate samplevolumes for analysis, particularly in the case of rodents. To evaluate a standard clinical chemistrypanel prior to PND 28 in the rat, researchers must use terminal blood collections, and samples mayhave to be pooled to obtain adequate sample sizes in rats younger than PND 17. A review of ratand dog standard serum chemistry panels from juvenile animals reveals specific patterns of changethroughout postnatal development (Table 8.3 and Table 8.6). 92,124 For example, in both the rat andthe dog, cholesterol, total bilirubin, and serum urea nitrogen levels decreased markedly with age.However, alanine aminotransferase (ALT) increased by approximately 50% in beagles during thefirst 20 weeks of life, while in the rat, ALT increased by approximately 300% from 2 to 5 weeksof age.Prior to investigating effects of a test article on clinical pathology parameters, as much informationas possible about the ontogeny of these parameters should be gathered. For example, if achemical inducing α-2 urinary globulin accumulation were under investigation in a juvenile malerat, the researcher should be cognizant that α-2 urinary globulin is not produced by the liver insubstantial amounts until the time of puberty. 125–126 In addition to evaluation of standard enzymepanels, specific enzymes associated with a known mode of action may be evaluated. One exampleis cholinesterase activity, which increases dramatically from birth until PND 42 in the rat brain(Figure 8.19). 88 An understanding of the ontogeny of this type of enzyme system can be critical inthe design and interpretation of nonclinical juvenile toxicity studies to evaluate compounds thattarget specific enzyme systems.Reference hematology values at defined developmental intervals should also be consideredwhen designing a juvenile study, especially if the class of compounds has known effects onhematopoeisis or the developing immune system. There is a lack of hematology reference valuesat different developing time points for small laboratory animals (e.g., mice and rats) because ofthe technical challenges associated with collecting enough terminal blood from individual animalsto evaluate a full panel of hematology parameters for reasonable interpretation. For larger animals(e.g., dogs, nonhuman primates, and pigs) the technical challenges of obtaining the blood volumesrequired for a full hematology panel are not so great. A review of the standard hematology panelfrom juvenile dogs reveals some specific developmental differences from birth to adolescence (referto Table 8.7). 124 At birth, reticulocytes are larger and more numerous than those of adults, asindicated by the higher mean reticulocyte counts and mean corpuscular volumes (MCV) during© 2006 by Taylor & Francis Group, LLC

300 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION180.0160.0140.0Food Consumption (g/kg/day)120.0100.080.060.040.020.0Males Females0.028–32 32–36 36–39 39–43 43–46 46–50 50–53 53–57 57–60 60–64Postnatal Interval in Days35.030.0Food Consumption (g/animal/day)25.020.015.010.05.0MalesFemales0.028–32 32–36 36–39 39–43 43–46 46–50 50–53 53–57 57–60 60–64Postnatal Interval in DaysFigure 8.18Food consumption in juvenile Crl:CD ® (SD)IGS BR rats during development (PND 28 throughPND 64).© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 30155000500004500040000Activity (U/brain)Activity (U/g tissue)403530Activity (U/brain)3500030000250002000015000100005000252015105Activity (U/g tissue)0FetusPND 4PND 11AgePND 21PND 42Dam0Figure 8.19Ontogeny of female Crl:CD ® (SD)IGS BR rat brain cholinesterase (fetal tissue pooled regardlessof sex). Data were collected at WIL Research Laboratories, LLC.the first month following birth. As neonatal dogs develop from birth to 3 months of age, thesevalues decline to normal adult ranges. In contrast, many of the other hematology values (neutrophil,lymphocyte, and platelet counts) have similar values from birth through adolescent development.4. Macroscopic and Microscopic EvaluationsMacroscopic and microscopic evaluations are also recommended by the FDA draft guidancedocument for nonclinical juvenile studies. 51 The importance of gross examinations is similar to thatfor standard toxicity studies, although with less emphasis on chronic lesions. Macroscopic changesconsistent with hypoplasia may correlate with changes in growth in juvenile studies. One exampleis a change in the relative size of organs or structures such as the kidneys, liver, testes, or bones.If there are known target organs, based on data from adult animals or humans, these organs shouldbe examined in the juvenile. However, histopathology varies with stage of organ maturity (e.g.,reproductive organs, such as the testes), and these differences between developing and adult animalsmust be considered. As an example, basophilic tubules observed in regenerating tubular epitheliumin an adult kidney are considered evidence of ongoing degeneration of the renal tubule and areconsidered an adverse finding. 127 However, in a juvenile study of the angiotensin converting enzyme(ACE) inhibitor, ramipril, the finding of basophilic tubules on PND 17 and 28 was noted for bothwell-developed tubules and for clusters of relatively undeveloped tubules in both control and treatedanimals. 92 In that study, the adverse change caused by ramipril was the increased incidence ofbasophilic tubules in the treated animals, potentially suggesting a delay in development of thekidneys of these animals. Another ACE inhibitor was also shown to cause arrested maturation ofthe developing kidney (immature glomeruli, distorted and dilated tubules, and relatively few andshort, thick arterioles) when administered to newborn rats. 1285. Physical and Sexual Developmental Landmarks and Behavioral AssessmentsThe assessment of growth can involve evaluation of a variety of different parameters, includingbody weights and landmarks of physical and sexual development. Physical developmental landmarksare commonly included as endpoints in reproductive and developmental toxicity studies© 2006 by Taylor & Francis Group, LLC

Table 8.6Serum chemistry values for beagle dogsPostnatal Day aParameter b 7 14 21 28 42 60 85 108 135ALB (g/dl)M 2.3 ± 0.22 (8) 2.3 ± 0.13 (8) 2.5 ± 0.17 (9) 2.9 ± 0.17 (7) 3 ± 0.13 (7) 2.9 ± 0.13 (7) 3 ± 0.12 (7) 3 ± 0.12 (7) 3.1 ± 0.13 (7)F 2.3 ± 0.1 (9) 2.3 ± 0.1 (9) 2.5 ± 0.16 (10) 2.9 ± 0.17 (8) 3 ± 0.15 (8) 3 ± 0.14 (8) 3.1 ± 0.09 (8) 3.1 ± 0.1 (8) 3.2 ± 0.13 (8)ALP (U/l)M 223 ± 50.3 (8) 199 ± 38.2 (9) 168 ± 7.5 (7) 178 ± 12.3 (7) 232 ± 34.1 (7) 246 ± 55.3 (7) 253 ± 29 (7) 209 ± 27.4 (7) 167 ± 19.6 (7)F 214 ± 45 (9) 183 ± 53.3 (10) 168 ± 16.7 (8) 167 ± 12 (8) 217 ± 41.8 (8) 247 ± 51.6 (8) 253 ± 22.5 (8) 213 ± 26.6 (8) 168 ± 25.9 (8)ALAT (U/l)M 28 ± 17.2 (8) 15 ± 3.1 (9) 17 ± 2.6 (7) 24 ± 5.3 (7) 36 ± 11.9 (7) 33 ± 5.7 (7) 37 ± 6.7 (7) 39 ± 7.5 (7) 40 ± 3.2 (7)F 24 ± 11.5 (9) 17 ± 6.4 (10) 17 ± 3.5 (8) 21 ± 7.5 (8) 32 ± 13.7 (8) 31 ± 6.3 (8) 38 ± 5.5 (8) 39 ± 7.4 (8) 41 ± 2.9 (8)ASAT (U/l)M 53 ± 25 (8) 34 ± 11.7 (9) 29 ± 6.2 (7) 28 ± 6.4 (7) 30 ± 1.8 (7) 34 ± 8.3 (7) 35 ± 12.1 (7) 36 ± 7 (7) 30 ± 3.3 (7)F 49 ± 26.6 (9) 31 ± 10.6 (10) 28 ± 8 (8) 23 ± 6.5 (8) 29 ± 4.6 (8) 32 ± 6 (8) 31 ± 7.9 (8) 34 ± 3 (8) 31 ± 4 (8)TOT. BIL (mg/dl)M 0.7 ± 0.22 (8) 0.5 ± 0.28 (8) 0.4 ± 0.11 (9) 0.4 ± 0.13 (7) 0.2 ± 0.07 (7) 0.2 ± 0.06 (7) 0.2 ± 0.03 (7) 0.1 ± 0.05 (7) 0.1 ± 0.03 (7)F 0.9 ± 0.49 (9) 0.5 ± 0.24 (9) 0.4 ± 0.06 (10) 0.3 ± 0.1 (8) 0.3 ± 0.11 (8) 0.2 ± 0.06 (8) 0.2 ± 0.08 (8) 0.2 ± 0.04 (8) 0.2 ± 0.03 (8)UREA N (mg/dl)M 30 ± 17 (8) 21 ± 5.5 (9) 16 ± 1.6 (7) 15 ± 2.6 (7) 13 ± 3.3 (7) 11 ± 2.4 (7) 15 ± 4.7 (7) 15 ± 5.1 (7) 11 ± 2 (7)F 26 ± 3.7 (9) 21 ± 4.6 (10) 16 ± 3.2 (8) 14 ± 1.4 (8) 13 ± 2.9 (8) 11 ± 3 (8) 13 ± 1.8 (8) 14 ± 3.5 (8) 12 ± 2.9 (8)CHOL (mg/dl)M 285 ± 128.8 (8) 218 ± 46.1 (9) 233 ± 41.9 (7) 239 ± 46 (7) 156 ± 36.3 (7) 127 ± 14 (7) 143 ± 10.8 (7) 166 ± 10.5 (7) 183 ± 10.7 (7)F 249 ± 39.4 (9) 241 ± 39.7 (10) 263 ± 49 (8) 259 ± 42.1 (8) 171 ± 37.5 (8) 126 ± 11.7 (8) 143 ± 19.4 (8) 166 ± 16.5 (8) 174 ± 14.7 (8)CREAT (mg/dl)M 0.1 ± 0.061 (8) 0.06 ± 0.046 (9) 0.03 ± 0.039 (7) 0.04 ± 0.045 (7) 0.06 ± 0.059 (7) 0.14 ± 0.057 (7) 0.14 ± 0.073 (7) 0.19 ± 0.099 (7) 0.29 ± 0.078 (7)F 0.08 ± 0.067 (9) 0.06 ± 0.066 (10) 0.05 ± 0.039 (8) 0.04 ± 0.048 (8) 0.07 ± 0.051 (8) 0.16 ± 0.044 (8) 0.13 ± 0.057 (8) 0.21 ± 0.065 (8) 0.32 ± 0.079 (8)302 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

GLT (U/L)M 77.4 ± 44.5 (8) 16.7 ± 8 (8) 5.4 ± 0.91 (9) 3.3 ± 0.73 (7) 2.8 ± 0.73 (7) 3 ± 0.6 (7) 3.8 ± 0.35 (7) 3.2 ± 0.41 (7) 3.7 ± 0.58 (7)F 58.8 ± 33.66 (9) 18.2 ± 7.68 (9) 6.1 ± 1.68 (10) 3.5 ± 1.05 (8) 2.6 ± 0.57 (8) 3.2 ± 0.76 (8) 7.4 ± 10.15 (8) 3.4 ± 0.5 (8) 3.5 ± 0.68 (8)GLOB (g/dl)M 1.6 ± 0.08 (8) 1.4 ± 0.08 (8) 1.4 ± 0.09 (9) 1.3 ± 0.13 (7) 1.4 ± 0.1 (7) 1.9 ± 0.25 (7) 2.2 ± 0.13 (7) 2.1 ± 0.13 (7) 2 ± 0.16 (7)F 1.5 ± 0.13 (9) 1.4 ± 0.14 (9) 1.4 ± 0.11 (10) 1.4 ± 0.13 (8) 1.4 ± 0.13 (8) 1.9 ± 0.26 (8) 2.1 ± 0.17 (8) 2.1 ± 0.16 (8) 2 ± 0.06 (8)GLUC (mg/dl)M 120 ± 11.8 (8) 121 ± 7.5 (9) 128 ± 12 (7) 135 ± 13 (7) 131 ± 8.2 (7) 125 ± 10.1 (7) 113 ± 4.8 (7) 116 ± 7.1 (7) 109 ± 5.7 (7)F 122 ± 13.9 (9) 123 ± 10.7 (10) 129 ± 11.9 (8) 136 ± 10.3 (8) 135 ± 10.5 (8) 126 ± 8.7 (8) 117 ± 5.5 (8) 109 ± 5.9 (8) 106 ± 6.3 (8)K + (mEq/l)M 5.7 ± 0.19 (8) 5.7 ± 0.44 (8) 5.8 ± 0.44 (9) 5.9 ± 0.49 (7) 6 ± 0.49 (7) 5.7 ± 0.52 (7) 5.4 ± 0.38 (7) 5.4 ± 0.3 (7) 5 ± 0.14 (7)F 5.8 ± 0.48 (9) 5.7 ± 0.48 (9) 5.8 ± 0.42 (10) 5.9 ± 0.55 (8) 6 ± 0.53 (8) 5.5 ± 0.67 (8) 5.3 ± 0.34 (8) 5.2 ± 0.3 (8) 5.1 ± 0.3 (8)Na + (mEq/l)M 144 ± 5.7 (8) 144 ± 2.5 (9) 144 ± 1.5 (7) 146 ± 1.8 (7) 146 ± 0.8 (7) 149 ± 2 (7) 149 ± 1 (7) 148 ± 1.4 (7) 149 ± 1.2 (7)F 142 ± 3.8 (9) 144 ± 3.1 (10) 145 ± 1.8 (8) 147 ± 1.6 (8) 147 ± 1.3 (8) 149 ± 2.4 (8) 149 ± 0.8 (8) 149 ± 2 (8) 150 ± 1.8 (8)I. PHOS (mg/dl)M 9.8 ± 1.13 (8) 9.5 ± 0.5 (8) 9.8 ± 1.1 (9) 9.5 ± 0.72 (7) 9.6 ± 0.29 (7) 9.3 ± 0.37 (7) 9.2 ± 0.39 (7) 8.5 ± 0.34 (7) 7.9 ± 0.28 (7)F 10.6 ± 1.23 (9) 10 ± 0.74 (9) 9.6 ± 0.57 (10) 9.5 ± 0.57 (8) 9.4 ± 0.32 (8) 9.1 ± 0.56 (8) 9.2 ± 0.36 (8) 8.6 ± 0.28 (8) 7.8 ± 0.41 (8)PROT (g/dl)M 3.9 ± 0.26 (8) 3.6 ± 0.17 (8) 3.9 ± 0.2 (9) 4.2 ± 0.22 (7) 4.4 ± 0.11 (7) 4.8 ± 0.29 (7) 5.2 ± 0.19 (7) 5.1 ± 0.17 (7) 5.1 ± 0.21 (7)F 3.7 ± 0.16 (9) 3.7 ± 0.2 (9) 3.9 ± 0.19 (10) 4.2 ± 0.26 (8) 4.4 ± 0.2 (8) 4.8 ± 0.23 (8) 5.2 ± 0.17 (8) 5.2 ± 0.14 (8) 5.2 ± 0.16 (8)aValues are expressed as: Mean ± SD (N = number of litters).bSerum Chemistry Key: P, parameter; ALB, albumin; ALP, alkaline phosphatase; ALAT, alanine aminotransferase; ASAT, aspartate aminotransferase; TOT. BIL, total bilirubin; UREAN, urea nitrogen; CHOL, cholesterol; CREAT, creatinine; GLT, glutamyltransferase; GLUC, glucose; I.PHOS; inorganic phosphorus; PROT, total protein; M, males; F, females.Source: Data were collected at WIL Research Laboratories, LLC (2001–2002).NONCLINICAL JUVENILE TOXICITY TESTING 303© 2006 by Taylor & Francis Group, LLC

Table 8.7Parameter bHematology values for beagle dogsPostnatal Day a7 14 21 28 42 60 85 108 135WBC (thousands/µl)M 12.7 ± 2.12 (8) 10.4 ± 1.84 (8) 12.9 ± 3.43 (9) 12.6 ± 1.46 (7) 13.2 ± 2.56 (7) 13.3 ± 3.6 (7) 15.2 ± 3.33 (7) 12.2 ± 2.56 (7) 10.2 ± 1.88 (7)F 13.7 ± 3.11 (8) 10.1 ± 2.95 (8) 10.5 ± 2.98 (10) 11.7 ± 1.13 (8) 12.5 ± 2.39 (7) 13 ± 3.09 (8) 13.8 ± 2.82 (8) 11.4 ± 3.22 (8) 10.1 ± 2.53 (8)RBC (millions/µl)M 4 ± 0.41 (8) 3.6 ± 0.27 (8) 3.6 ± 0.22 (9) 3.6 ± 0.24 (7) 3.7 ± 0.36 (7) 4.5 ± 0.34 (7) 5.2 ± 0.39 (7) 5.7 ± 0.31 (7) 5.9 ± 0.49 (7)F 4 ± 0.33 (8) 3.6 ± 0.45 (8) 3.6 ± 0.27 (10) 3.7 ± 0.23 (8) 3.9 ± 0.29 (7) 4.6 ± 0.47 (8) 5.2 ± 0.42 (8) 5.7 ± 0.34 (8) 6 ± 0.47 (8)HB (g/dl)M 11.8 ± 1.18 (8) 9.8 ± 0.75 (8) 9.4 ± 0.42 (9) 8.8 ± 0.53 (7) 8.6 ± 0.86 (7) 10.2 ± 0.76 (7) 11.7 ± 0.65 (7) 12.9 ± 0.55 (7) 13.5 ± 0.98 (7)F 11.5 ± 1.44 (8) 9.9 ± 1.23 (8) 9.5 ± 0.6 (10) 9.4 ± 0.59 (8) 9.1 ± 0.8 (7) 10.6 ± 0.92 (8) 11.9 ± 0.71 (8) 13.2 ± 0.68 (8) 13.9 ± 1.03 (8)HCT, %M 32 ± 3.1 (8) 27 ± 1.7 (9) 26 ± 1.2 (7) 25 ± 1.6 (7) 24 ± 2.4 (7) 29 ± 1.8 (7) 33 ± 2 (7) 36 ± 1.7 (7) 37 ± 2.6 (7)F 31 ± 2.8 (8) 27 ± 3.2 (10) 26 ± 1.6 (8) 26 ± 1.6 (8) 26 ± 2 (8) 30 ± 2.7 (8) 33 ± 2.1 (8) 37 ± 1.9 (8) 38 ± 3 (8)MCV (fl)M 79 ± 2.5 (8) 75 ± 2.7 (9) 72 ± 1.6 (7) 69 ± 2.3 (7) 65 ± 1.6 (7) 64 ± 2.1 (7) 63 ± 2 (7) 63 ± 1.4 (7) 63 ± 1.4 (7)F 76 ± 3.3 (8) 75 ± 2.5 (10) 73 ± 2.1 (8) 70 ± 2.1 (8) 66 ± 1.3 (8) 65 ± 1.7 (8) 64 ± 2 (8) 64 ± 1.7 (8) 63 ± 1.6 (8)MCH (µµg)M 29 ± 1.1 (8) 27 ± 1.2 (9) 26 ± 0.5 (7) 25 ± 0.4 (7) 23 ± 0.9 (7) 23 ± 1.2 (7) 23 ± 0.7 (7) 23 ± 0.6 (7) 23 ± 0.6 (7)F 29 ± 2.2 (8) 28 ± 1.2 (10) 26 ± 0.7 (8) 25 ± 0.5 (8) 23 ± 0.8 (8) 23 ± 0.6 (8) 23 ± 0.7 (8) 23 ± 0.6 (8) 23 ± 0.4 (8)MCHC (g/dl)M 37 ± 1.1 (8) 37 ± 1.2 (9) 36 ± 0.4 (7) 36 ± 0.8 (7) 35 ± 0.7 (7) 36 ± 0.8 (7) 36 ± 0.9 (7) 36 ± 0.2 (7) 37 ± 0.4 (7)F 38 ± 1.9 (8) 37 ± 1 (10) 36 ± 0.6 (8) 36 ± 0.6 (8) 35 ± 0.8 (8) 36 ± 0.6 (8) 36 ± 0.5 (8) 36 ± 0.3 (8) 37 ± 0.5 (8)PLAT (thousands/µl)M 431 ± 102.2 (8) 361 ± 110.7 (9) 392 ± 128.8 (7) 394 ± 101.9 (7) 360 ± 86.3 (7) 331 ± 107.7 (7) 440 ± 77.3 (7) 442 ± 47.9 (7) 401 ± 90.6 (7)F 491 ± 126.5 (8) 412 ± 72.7 (10) 442 ± 107.7 (8) 427 ± 91.4 (8) 413 ± 90.6 (8) 330 ± 103.9 (8) 409 ± 94.8 (8) 422 ± 42.1 (8) 386 ± 81.3 (8)RETIC (%)M 6.7 ± 7.54 (7) 4.5 ± 5.77 (7) 1.5 ± 0.29 (7) 1.4 ± 0.31 (7) 2.1 ± 0.69 (7) 1.7 ± 0.66 (7) 1.7 ± 0.63 (7) 0.9 ± 0.18 (7) 0.7 ± 0.15 (7)F 6.4 ± 7.92 (8) 4.1 ± 3.68 (8) 1.5 ± 0.24 (9) 1.4 ± 0.36 (8) 2 ± 0.52 (8) 1.4 ± 0.54 (8) 1.4 ± 0.28 (8) 0.9 ± 0.17 (8) 0.7 ± 0.21 (8)RETIC AB (millions/µl)M 0.31 ± 0.363 (7) 0.16 ± 0.2 (7) 0.05 ± 0.011 (7) 0.05 ± 0.011 (7) 0.08 ± 0.027 (7) 0.08 ± 0.032 (7) 0.09 ± 0.035 (7) 0.05 ± 0.011 (7) 0.04 ± 0.011 (7)F 0.25 ± 0.288 (8) 0.14 ± 0.126 (9) 0.05 ± 0.01 (8) 0.05 ± 0.014 (8) 0.07 ± 0.018 (8) 0.06 ± 0.027 (8) 0.07 ± 0.017 (8) 0.05 ± 0.012 (8) 0.04 ± 0.014 (8)© 2006 by Taylor & Francis Group, LLC304 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION

N (thousands/µl)M 7 ± 1.94 (8) 6.2 ± 1.27 (8) 7.7 ± 2.55 (9) 7.3 ± 0.66 (7) 7.6 ± 1.44 (7) 9.2 ± 3.47 (7) 10.3 ± 2.6 (7) 7.4 ± 1.53 (7) 5.8 ± 1.56 (7)F 8.1 ± 2.34 (8) 5.6 ± 2.22 (8) 5.6 ± 2.29 (10) 6.5 ± 0.92 (8) 6.8 ± 1.42 (7) 8.7 ± 2.65 (8) 9.2 ± 1.88 (8) 6.4 ± 2.36 (8) 5.7 ± 1.29 (8)L (thousands/µl)M 4.4 ± 0.66 (8) 3.2 ± 0.95 (8) 4 ± 1.06 (9) 3.9 ± 0.47 (7) 4.3 ± 1 (7) 3 ± 0.92 (7) 3.5 ± 0.79 (7) 4 ± 0.98 (7) 3.7 ± 1.04 (7)F 4.2 ± 0.96 (8) 3.5 ± 0.97 (8) 4.1 ± 1.09 (10) 4 ± 0.51 (8) 4.4 ± 0.95 (7) 3.5 ± 1.31 (8) 3.3 ± 1.07 (8) 4.1 ± 1.42 (8) 3.8 ± 1.41 (8)MO (thousands/µl)M 1.2 ± 0.41 (8) 1 ± 0.29 (8) 1 ± 0.7 (9) 1.3 ± 0.72 (7) 1.1 ± 0.77 (7) 0.9 ± 0.51 (7) 1.3 ± 0.85 (7) 0.8 ± 0.44 (7) 0.5 ± 0.24 (7)F 1.3 ± 0.73 (8) 0.9 ± 0.6 (8) 0.8 ± 0.33 (10) 1 ± 0.54 (8) 1.2 ± 0.79 (7) 0.7 ± 0.5 (8) 1.2 ± 0.9 (8) 0.8 ± 0.47 (8) 0.6 ± 0.47 (8)E (thousands/µl)M 0.15 ± 0.21 (8) 0.03 ± 0.1 (9) 0.06 ± 0.059 (7) 0.02 ± 0.06 (7) 0.09 ± 0.182 (7) 0.09 ± 0.114 (7) 0.09 ± 0.134 (7) 0.07 ± 0.09 (7) 0.16 ± 0.161 (7)F 0.06 ± 0.177 (8) 0.07 ± 0.127 (10) 0.11 ± 0.159 (8) 0.04 ± 0.082 (8) 0.03 ± 0.041 (8) 0.13 ± 0.233 (8) 0.06 ± 0.119 (8) 0.05 ± 0.085 (8) 0.08 ± 0.113 (8)B (thousands/µl)M 0.04 ± 0.052 (8) 0.03 ± 0.044 (9) 0.03 ± 0.055 (7) 0.12 ± 0.243 (7) 0.12 ± 0.104 (7) 0.04 ± 0.057 (7) 0.07 ± 0.093 (7) 0.03 ± 0.036 (7) 0.02 ± 0.026 (7)F 0.04 ± 0.088 (8) 0.01 ± 0.016 (10) 0.01 ± 0.015 (8) 0.06 ± 0.051 (8) 0.11 ± 0.151 (8) 0.03 ± 0.069 (8) 0.04 ± 0.046 (8) 0.03 ± 0.054 (8) 0.05 ± 0.077 (8)N (%)M 54 ± 8.3 (8) 59 ± 6.1 (9) 60 ± 6.3 (7) 58 ± 2.3 (7) 58 ± 5.9 (7) 68 ± 8.6 (7) 67 ± 7.6 (7) 60 ± 4.3 (7) 56 ± 7.4 (7)F 59 ± 8.1 (8) 54 ± 10.5 (10) 51 ± 9.9 (8) 56 ± 4.6 (8) 54 ± 7.5 (8) 66 ± 8.1 (8) 67 ± 8.4 (8) 56 ± 9.3 (8) 56 ± 9.9 (8)L (%)M 35 ± 6.2 (8) 31 ± 5.9 (9) 32 ± 6.1 (7) 31 ± 4.3 (7) 33 ± 4.7 (7) 23 ± 7.8 (7) 24 ± 5.7 (7) 33 ± 2.7 (7) 38 ± 5.2 (7)F 31 ± 7.1 (8) 36 ± 8 (10) 39 ± 7.3 (8) 34 ± 5.3 (8) 35 ± 3.9 (8) 28 ± 9.1 (8) 25 ± 6.1 (8) 37 ± 10.4 (8) 37 ± 7.5 (8)MO (%)M 9.4 ± 4.34 (8) 9.7 ± 3.44 (8) 7.7 ± 4.22 (9) 10.1 ± 4.98 (7) 7.9 ± 4.74 (7) 7.2 ± 4.28 (7) 8.1 ± 4.37 (7) 6.7 ± 3.48 (7) 5.1 ± 1.9 (7)F 9.1 ± 4.13 (8) 9.1 ± 4.94 (8) 7.9 ± 3.62 (10) 8.6 ± 4.44 (8) 9.2 ± 5.09 (7) 5 ± 2.74 (8) 7.9 ± 4.77 (8) 6.6 ± 3.28 (8) 5.1 ± 4.38 (8)E (%)M 1.2 ± 1.62 (8) 0.3 ± 1 (8) 0.6 ± 0.74 (9) 0.2 ± 0.53 (7) 0.6 ± 1.08 (7) 0.6 ± 0.63 (7) 0.7 ± 0.94 (7) 0.5 ± 0.62 (7) 1.4 ± 1.41 (7)F 0.6 ± 1.77 (8) 0.7 ± 1.27 (8) 1.5 ± 2.41 (10) 0.4 ± 0.78 (8) 0.3 ± 0.31 (7) 1.1 ± 1.64 (8) 0.4 ± 0.79 (8) 0.6 ± 0.96 (8) 0.9 ± 1.38 (8)B (%)M 0.27 ± 0.398 (8) 0.28 ± 0.441 (9) 0.23 ± 0.369 (7) 1.05 ± 1.932 (7) 0.82 ± 0.7 (7) 0.4 ± 0.572 (7) 0.39 ± 0.456 (7) 0.25 ± 0.348 (7) 0.18 ± 0.263 (7)F 0.25 ± 0.535 (8) 0.05 ± 0.158 (10) 0.14 ± 0.147 (8) 0.55 ± 0.473 (8) 0.94 ± 1.054 (8) 0.22 ± 0.364 (8) 0.26 ± 0.355 (8) 0.28 ± 0.452 (8) 0.44 ± 0.729 (8)aValues are expressed as: Mean ± SD (N = number of litters).bHematology Key: P, parameter; WBC, white blood cells; RBC, erythrocytes; HB, hemoglobin; HCT; hematocrit; MCV, mean cell volume; MCH, mean cell hemoglobin; MCHC,mean cell hemoglobin concentration; PLAT, platelets; RETIC, reticulocytes; AB, absolute; N, neutrophils; L, lymphocytes; MO, monocytes; E, eosinophils; B, basophils; M, males;F, females.Source: Data were collected at WIL Research Laboratories, LLC (2001-2002).NONCLINICAL JUVENILE TOXICITY TESTING 305© 2006 by Taylor & Francis Group, LLC

306 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 8.8Developmental landmarks for beagle dogsconducted for both environmental chemicals and drugs and are required endpoints in the OECDdraft guideline for developmental neurotoxicity studies. 129 Recently, studies evaluating juveniletoxicity have included these parameters, as they provide key indicators of perturbations in growthand function, and because the dose administration period for these studies often encompasses agesat which many of these landmarks are acquired.In addition to physical landmarks of development, the study of immediate onset and latentbehavioral expression of CNS damage associated with exposures during prenatal and early postnataldevelopment has been accorded a higher level of concern in recent years. Tests to evaluate theseaspects of development are routinely performed during the International Conference on Harmonisation(ICH) pre- and postnatal development study, 106 as well as under the more robust and focusedEPA and OECD developmental neurotoxicity protocols. 56,129 A plethora of standard testing paradigmshave been proposed, validated, and/or used in the process of identifying and characterizingagents that may represent a threat to the developing nervous system. The authors have chosenrepresentative examples of these tests, and the following discussion should not be construed to bea comprehensive treatment of the subject. The reader is referred to the indicated references foradditional exposition.a. Physical Landmarks of DevelopmentEyelidSeparationIncisorEruptionVaginalPatencyTestesDescentBalanopreputialSeparationM F M F F M MDay of acquisition (PND)Mean 15.5 15.6 20.2 20.3 21.3 22.4 50.7S.D. 1.62 1.81 2.09 3.30 4.36 4.72 7.34N (litters) 16 17 16 17 16 16 14Body weight (g)Mean 792 774 927 877 915 1057 2572S.D. 123.7 169.5 157.2 162.1 247.9 338.6 607.6N (litters) 16 17 16 17 16 14 14M = male, F = femaleSource: Data were collected at WIL Research Laboratories, LLC (2001–2002).There are several accepted measures of growth and functional development of the juvenile animalthat have been employed in a variety of study designs, including the multigeneration studiessubmitted to the EPA and the pre- and postnatal development studies submitted to the FDA.Although the most robust historical control data exist for the rat, landmarks can and have beenevaluated in a variety of species. An example data set for these parameters collected in the beagledog is presented in Table 8.8. Inclusion of these landmarks of development may provide insightinto potential developmental delays occurring as a result of direct administration of a test article.However, the use of these landmarks is always contingent on the age at onset of exposure and theduration of the study in general. It is also important to note that while these landmarks are stillevaluated in tests for development, many are positively correlated with body weight, and developmentaldelays or accelerations are typically a function of neonatal weight. 130,131 Therefore, merelycollecting body weights at key developmental stages may be as effective at elucidating changes ingrowth and development without requiring the addition of labor-intensive neonatal evaluations.Moreover, the OECD draft guideline for developmental neurotoxicity studies has suggested thatthese endpoints are “recommended only when there is prior evidence that these endpoints willprovide additional information.” 129 In addition, the ICH guideline for reproductive and developmentaltoxicity studies states that “the best indicator of physical development is body weight.” 131© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 307Table 8.9 Example of developmental delay in surface righting reflexSex Control T1 T2 T3A. Total Number of Pups Available for AssessmentMales 5.1 ± 0.10 5.0 ± 0.08 5.1 ± 0.13 5.1 ± 0.19Females 5.1 ± 0.15 5.1 ± 0.17 5.1 ± 0.19 5.2 ± 0.27Combined 5.1 ± 0.10 5.1 ± 0.11 5.1 ± 0.14 5.2 ± 0.18B. % Positively Responding (Number of Pups)MalesTotal pups 95 119 100 108PND 5 95% (90) 98% (117) 94% (94) 85% (92)PND 6 100% (95) 100% (119) 100% (100) 100% (108)FemalesTotal Pups 108 112 110 102PND 5 86% (93) 89% (100) 88% (97) 78% (80)PND 6 100% (108) 99% (111) 100% (110) 98% (100)PND 7 100% (112) 100% (102)CombinedTotal Pups 203 231 210 210PND 5 90% (183) 94% (217) 91% (191) 82% (172)PND 6 100% (203) 100% (230) 100% (210) 99% (208)PND 7 100% (231) 100% (210)Notes: Surface righting response evaluated in Crl:CD ® (SD)IGS BR rat offspring,beginning on PND 5, with assessment continuing until positiveevidence of righting was observed (WIL Research Laboratories,LLC). The same data are presented in A and B. Section A presentsthe mean ± S.D. day of acquisition of surface righting for males,females, and combined sexes. Section B presents the percentage(total number) of rat offspring acquiring surface righting on eachday of assessment.The ICH guideline specifically mentions several preweaning developmental landmarks, includingpinna detachment, hair growth, and incisor eruption, and notes that these measures are highlycorrelated with body weight and postcoital age, especially “when significant differences in gestationlength occur.” Moreover, functional developmental landmarks, such as surface and air righting,were also noted in the ICH guideline to be dependent on physical development. While a completediscussion of the many developmental landmarks is beyond the scope of this chapter, the authorshave chosen to address a few key evaluations, and we encourage the reader to review the extensiveliterature on this topic for more detail on specific evaluations.One commonly evaluated landmark of functional development is surface righting. The surfacerightingreflex evaluates the ability of the animal to regain a normal upright stance when placed ina supine position on a flat surface. 132 This reflex is present soon after birth in rats, although thetime required to attain an upright stance decreases as the animal ages. 133 At the authors’ facility,this test is performed beginning on PND 5 (one day after litter standardization) for the rat, with a15-s window for test completion. Neonates with negative responses (no observed righting) areevaluated on sequential days until a positive response is noted. Under these testing conditions, theapproximate average age at which righting occurs is PND 5. Because the age at onset of testingequals the age at which a positive response is observed, a delay in acquisition of this landmark ismost easily noted as an increase in the number of animals (on a litter basis) with a negative responseon the first day of testing. An example of such a delay is provided in Table 8.9. Although not© 2006 by Taylor & Francis Group, LLC

308 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONevident in this example, it is important to note that a statistical difference from control in the meanage at acquisition for this test is a key indicator of a test article–related change and should not beignored.Hair growth, incisor eruption, and eyelid separation are well accepted landmarks of physicaldevelopment, and both incisor eruption and eyelid separation were required endpoints of analysisin the CBTS. 105 For each of these tests, the neonate is evaluated from the first day of testing untila positive response is noted. Of course, evaluation of the acquisition of some landmarks is notpossible in species with such long gestational lengths that they are typically born with hair (dog,pig) or with separated eyelids (pig).Habituation is another functional reflex that is often assessed in a variety of study designs.Habituation is the progressive decrease in response to repeated stimulation that cannot be attributedto receptor fatigue or adaptation. 134 Glass and Singer characterized habituation as the most importantmechanism by which humans adapt and survive in modern society. 135 It is ubiquitous in nature andmost multicellular organisms have been shown to exhibit some form of habituation. 136 The regulatoryagencies consider development of the habituation response to be a key endpoint for evaluation.56,129 Habituation can also be used as a measure of basic learning, and in the CBTS, it wasdetermined to be a consistent method for detecting toxicity caused by methyl mercuric chloride. 104This reflex can be evaluated with a number of automated functional tests, such as the auditorystartle habituation test and the locomotor activity test. More detailed information on this reflex isprovided below as part of the discussion of each behavior.As Lochry pointed out in her review, landmarks of physical and functional development areoften highly correlated with body weight. 130 Therefore, when test article–related changes in landmarksof physical and functional development are observed, correlative changes in behavioralassessments may be noted as well. For example, delays in the acquisition of physical landmarkscan indicate a general developmental delay that may influence the normal ontogeny of locomotoractivity or the habituation to a startle stimulus. It is also critical to note that preweaning developmentallandmarks may be affected by maternal behavior. If the mother becomes incapable ofadequately rearing her young, growth of the offspring will be negatively impacted. This detrimentalmaternal influence can profoundly affect the offspring, should be carefully monitored during thecourse of the study, and must be considered during evaluation of the data.b. Landmarks of Sexual DevelopmentTwo additional standard landmarks of sexual development that are included in both juvenile andadult reproduction studies are balanopreputial separation and vaginal patency (see Chapters 9 and10 for descriptions). These landmarks are acquired by males or females, respectively, prior to theonset of puberty and are triggered by the presence of reproductive hormones. 137Evaluation of balanopreputial separation is typically conducted daily until there is evidencethat the prepuce has separated from the glans penis. 138 There is some thought that preputialseparation can be accelerated by repeated examinations; however, to ensure that the day of acquisitionis not overlooked, many laboratories will begin the examinations between PND 35 to 40 inthe rat. 137 In the authors’ laboratory, the mean age at acquisition of balanopreputial separation inthe Crl:CD ® (SD)IGS BR rat is approximately PND 45. The authors have also developed limiteddata in the New Zealand White rabbit, which indicate that balanopreputial separation is achievedon approximately PND 72. In addition, data developed at the authors’ laboratory in the beagle dogreveal that the mean age at acquisition of balanopreputial separation is approximately PND 51 (seeTable 8.8). As with the other landmarks of development, delays and accelerations in the acquisitionof balanopreputial separation can occur. Acquisition is correlated with body weight, 139,140 but otherdetermining factors, such as the presence of reproductive hormone stimulation or endocrinemodulatingchemicals, including 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene, can also disruptnormal development. 141 Testes descent is another landmark of sexual development in the male that© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 309can be assessed in a number of species and is preferred by some researchers when evaluating sexualmaturation in mice. Recently published data from three investigations in Swiss albino mice suggestthat testes descent is very consistent over time and across laboratories, and occurs at approximatelyday 23 or 24 of postnatal life. 142–144In females, the acquisition of vaginal patency is detected by examining the vagina to determineif the septum covering the vagina has ruptured. 145 Typically, the initial age at testing is betweenPND 25 to 30 in the rat. In the authors’ laboratory, the mean age at acquisition of vaginal patencyin the Crl:CD ® (SD)IGS BR rat is approximately PND 33. In addition, in a limited data set developedin the authors’ laboratory, the mean age at acquisition for vaginal patency in the New ZealandWhite rabbit was shown to be approximately PND 29. In comparison, data developed at the authors’laboratory for the beagle dog indicate that the mean age at acquisition of vaginal patency isapproximately PND 21 (see Table 8.8). Swiss albino mice acquire vaginal patency at approximatelyday 23 or 24 of postnatal life. 142–144Vaginal patency is positively correlated with body weight, and substantial changes in growthcan alter the day of acquisition. 139 Aside from effects on growth, estrogenic and antiestrogeniccompounds can dramatically influence the day on which vaginal opening is first observed. 137 Onepotential confounder in the observation of vaginal patency is the somewhat rare occurrence of athin strand of tissue remaining over the vaginal canal. The presence of the thin strand of tissuedoes not interfere with estrous cyclicity or mating, and it can be removed by the observer or duringthe mating process. 137 However, there seems to be some disagreement as to whether the rat shouldbe considered as having acquired vaginal patency if this tissue strand is still present. Gray andOstby noted that when animals with the thin strand of tissue were eliminated from the analysis,there was no effect of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on the mean age at acquisition.146 As with many laboratories, the authors consider a rat to have acquired vaginal patency ifonly the thin strand of tissue spans the lumen.Given the potential for acceleration in the acquisition of these sexual developmental landmarksby potent endocrine modulators, it is recommended that the age of initial evaluation should be wellin advance of any normal age of acquisition. In the authors’ laboratory, male and female rats areinitially evaluated on PND 35 and 25, respectively. That timing is used because diethylstilbestrol,a potent estrogen, has been shown to accelerate vaginal opening in a female pubertal assay to 29.7and 24.1 days of age at dose levels of 1 and 5 µg/kg, respectively. 147 The mean age at acquisitionof vaginal patency in that study was prior to the initial evaluation age used by most laboratories,but PND 25 should be early enough for most cases, unless the test article is a suspected endocrinemodulating compound.Another key point to consider when evaluating vaginal patency is the variability of the test ina given test species. The mean age at acquisition of vaginal patency in mice has been shown to beaffected by cohabitation with adult males and by the presence of bedding from male cages. 137 Thispotential confounding influence should be taken into consideration when designing a juvenile studyin mice.When evaluating effects on balanopreputial separation or vaginal patency, it is important toconsider changes in other growth parameters to determine whether there is a pattern of developmentaldelay or acceleration. Changes in the mean age at acquisition of either of these endpointsin the absence of effects on either body weight or the mean age at acquisition of other physicallandmarks can be a key indicator of endocrine modulation by a test article. However, when themean age at acquisition of balanopreputial separation or vaginal patency is altered in the presenceof clear changes in body weight, these effects are often considered secondary to overall changesin growth. In these cases, it is imperative to evaluate the entire data set for other similar patternsof effect.Ashby and Lefevre reported a relationship between the age at acquisition of balanopreputialseparation and group mean body weight. 140 They showed that in the absence of changes in bodyweight, changes in the age at acquisition of balanopreputial separation in Alpk:APfSD rats of more© 2006 by Taylor & Francis Group, LLC

310 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthan two days would likely be test article–related. Similarly, Merck performed calculations usinghistorical data to determine the power of a standard two-tailed test to detect changes in the age atacquisition of both balanopreputial separation and vaginal patency. 137 As an example, their calculationsdemonstrated that with a sample size of 20 Sprague-Dawley rats/group, a 2.5-day changein the mean age at acquisition of balanopreputial separation would have a probability of 0.87 ofbeing a true change from control. Likewise, the data indicated that a difference in the mean age atacquisition of vaginal patency of 2.0 days would have a probability of 0.87 of being a true changefrom control. 137c. Behavioral AssessmentsThe decision to conduct behavioral testing as part of the juvenile study can be a difficult onebecause assessment of behavior may be required for other agents in addition to neuroactivecompounds. Factors influencing the decision include the age at initiation of clinical dose administration,the duration of the dose administration period, whether the central or peripheral nervoussystems are still developing in the species, and whether the test article has been shown to causeadverse changes in organ systems that are key to internal homeostasis in adult animals. As mentionedearlier, there are critical windows during which organ systems may be susceptible to test articleadministration. For example, the development of the human nervous system extends well beyondbirth. 148 If the pediatric indication includes the period of nervous system development, regulatoryagencies may recommend an evaluation of behavioral endpoints. In addition, functional deficits inkey organ systems can produce detrimental effects on behavior that are separate from directneurotoxicity. However, these potential secondary changes in behavior should not be ignoredbecause they can provide important clues to the clinician regarding the types of side effects thatpediatric patients may experience. The next several sections describe basic behavioral testingparadigms that may be employed to evaluate changes in motor function and coordination, sensoryperception, and cognition. The list of endpoints discussed is not meant to cover every possibleparameter, and the reader is invited to review the extensive literature on these topics for more detail.1) Locomotor Activity — Locomotor activity can be generally quantified as total, fine, andambulatory activities. There are several styles of test chambers for rats, and the authors’ facilityemploys an open-field environment surrounded by four-sided black plastic enclosures to decreasethe potential for distraction by extraneous environmental stimuli. Activity is measured electronicallywith a series of infrared photobeams surrounding the clear plastic rectangular open field environment.The animal’s activity can be expressed as total and ambulatory activity (for rodents), althoughthis is not a universally applied approach. Total activity is generally the sum of both gross and finemotor movements (any photobeam break during the testing paradigm). Ambulatory activity is ameasure of only gross movements (two or more consecutive photobeam breaks, i.e., the animal ismoving from one location to another). Testing is usually conducted over a 60- to 90-min session,although some laboratories conduct locomotor activity test sessions over a 23-h period. The lengthof the test session is determined by the desire to demonstrate habituation during a particular test.Many laboratories have based the session length on EPA guideline recommendations that “the testsession should be long enough for motor activity to approach asymptotic levels by the last 20percent of the session for nontreated control animals.” 56 Appropriate validation and positive controlstudies must be conducted prior to the conduct of any animal study to determine the session lengthnecessary to evaluate habituation, using the equipment available for testing.For the rat, total activity will normally be relatively low at PND 13, will increase to a maximumpreweaning level by PND 17, and will decrease by PND 21. This ontogenic profile of activity hasbeen well characterized. 149 For rats and mice, locomotor activity is generally assessed longitudinallyduring early postnatal life (PND 13, 17, and 21 for rats) and again at young adulthood (PND 60to 61) in standard developmental neurotoxicity studies. 56 More limited assessments are generally© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 311Total Counts27502500225020001750150012501000750Male500 Female250023 41 63 79 95 111 127 148Postnatal Day of Locomotor Activity AssessmentFigure 8.20Locomotor activity development of the beagle dog. Mean number of total counts for litters by sexat each day of locomotor activity assessment (n = 7 to 8 litters per time point). Data were collectedat WIL Research Laboratories, LLC.conducted for the pre- and postnatal development studies, 131 although these assessments are generallyconducted in a longitudinal fashion as well.The first three time points listed above for rats examine the ontogeny of habituation, thedevelopment of motor activity coordination, and the presence or absence of the characteristic patternof early rodent activity. 150–151 In more traditional reproduction and developmental neurotoxicitystudies, the assessment at approximately PND 60 examines potential latent alterations in motoractivity and the persistence of changes observed prior to weaning. Other ages may be substitutedor added as appropriate for the age at which dose administration occurs and to assess potentialrecovery from or latency of motor activity decrements after juvenile exposure.The use of different species, such as the dog, will require adjustments in the timing for evaluatingchanges in the ontogenic profile of activity. Figure 8.20 illustrates the development of locomotoractivity in beagle puppies over a period of approximately 18 weeks. 124 If no data are availableregarding the ontogeny of motor activity in the species selected for the juvenile study and if it isconsidered necessary to understand the potential effects of the test article on this profile, somespecies characterization will be required prior to study initiation or else selection of a differentanimal model should be considered. Many nonrodent species have substantially longer periods ofmotor development than those of rodents (refer to Figure 8.20 for ontogeny of motor function inthe dog), requiring longer study durations to assess changes in the ontogeny of motor activity. 115When evaluating locomotor activity, the pattern of habituation should be examined within thecontrol group at each age of assessment. In the rat on PND 13, there should be very little changein total activity across each test interval. A test interval is defined in this text as a period duringwhich activity counts are collected and then presented. A test interval is typically between 10 and15 min and is determined through validation of the equipment. Between PND 17 and 21, a graphof cross-interval activity should begin to show a near-hyperbolic shape. The first test interval shouldcontain the most movement, with each subsequent interval showing an approximately 30 to 50%decrease in movement (when the total activity session is divided into four 15-min intervals) untilthe final test interval, when very little change should occur. In testing paradigms in which six ornine intervals (usually 10 min each) are used for data analysis, the magnitude of interintervaldecreases will likely differ from that described above. Interval data from motor activity sessionsat adulthood should demonstrate a well-defined habituation curve. There should be no consistentsex-related differences among young animals, although gender-specific differences may developwith age.© 2006 by Taylor & Francis Group, LLC

312 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPotential test article–related effects observed in motor activity should be compared with thoseobserved in other behavioral and/or physiological endpoints. For example, if a delay is suspectedin the normal pattern of motor activity development, similar changes indicative of delay or decreasedperformance in age- or weight-sensitive parameters in a functional observational battery, such asgrip strength, rotarod performance, or gait, may also be expected. In comparison with a change inmultiple related behavioral endpoints, a change in only one motor activity parameter in the absenceof any other change in behavior may carry less weight relative to test article exposure.Some locomotor activity test systems evaluate activity in both a vertical plane and along ahorizontal surface. These systems contain two levels of photobeams. The lower level detectsinterruptions in photobeams resulting from gross movements and grooming, while the upper leveldetects interruptions caused by rearing. Rearing is typically noted in aroused adult rats and providesan indication of exploratory behavior. 133 When examined in an open field environment, it has beendemonstrated that rearing is not well developed in rats until PND 18 and continues to increase infrequency in response to a novel environment beyond weaning. Rearing decreases over time in anovel environment. An interpretation of this parameter in the locomotor activity assessment shouldfirst consider whether rearing should be present in the animal at the age tested. Also, if a reductionin rearing counts is not observed over the testing interval, this should be considered an importanttest article–related change.2) Auditory Startle Response — The auditory startle test used most often at the authors’ laboratoryis based on the procedures used in the CBTS. 145 In this testing paradigm, the animal isplaced in an enclosure atop a force transducer in a sound-attenuated chamber. Background noiseis used to dampen external environmental stimuli and enhance the response to the startle stimulus.152–153 The force applied to the transducer is measured for some period after a high-pitched noiseof approximately 115 dB(A) is produced. Examples of the resultant data include the maximumresponse to the stimulus and the time to maximum response. The data are typically presented asblocks of trials for a given assessment. Other paradigms include prepulse inhibition, which providesa stimulus (auditory or tactile) prior to the startle burst to detect an attenuated response. 152 Morecomplex startle testing paradigms, involving a combination of habituation and prepulse inhibition,have been successfully used at some testing facilities. 154As with the locomotor activity assessment, the auditory startle response paradigm is typicallyconducted longitudinally. The standard ages of assessment in most reproductive and developmentalneurotoxicity studies are PND 20 and 60 for the rodent. 56 Within a particular assessment interval,the pattern of habituation should be examined. During preweaning assessments in the rodent at theauthors’ laboratory, there is typically an attenuated change in peak response across each trial block.At young adulthood, peak response to the startle stimulus should exhibit a characteristic habituationcurve, with the greatest response observed during the first trial block. Each subsequent trial blockshould demonstrate a reduction in force, until the final intervals, when very little change is expected.Latency to peak response does not change remarkably over a given test session. Typically youngercontrol animals will show a slight increase in latency to peak response over the five blocks of trials.No gender-specific differences in peak response should be observed for preweaning animals,although gender-related differences become apparent with increasing age and body weight (becausemales generally weigh more than females).It is important to compare potential test article–related effects observed in the auditory startleresponse testing paradigm with those for other behavioral and/or physiological endpoints. Forexample, if deficits in peak response related to neuromotor fatigue or disruption are observed, theinvestigator might also expect to observe adverse effects on motor activity and grip strength. Severedeficits in body weight can confound interpretation of the effects of the test article on the auditorystartle response and may mask subtle changes in neuromotor function in this testing paradigm.For nonrodent species, alternative methods for evaluating auditory startle habituation may berequired. The authors are aware of only two systems capable of evaluating auditory startle response© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 313Table 8.10Biel maze learning and memory paradigmDay Number of Trials Path Data Collected Parameter Evaluated1 4 Straight channel Time to escape Swimming2 2 (Trials 1 and 2) Path A (forward) Time to escape and number of errors Learning3 2 (Trials 3 and 4) Path A (forward) Time to escape and number of errors Learning4 2 (Trials 5 and 6) Path B (reverse) Time to escape and number of errors Learning5 2 (Trials 7 and 8) Path B (reverse) Time to escape and number of errors Learning6 2 (Trials 9 and 10) Path B (reverse) Time to escape and number of errors Learning7 2 (Trials 11 and 12) Path A (forward) Time to escape and number of errors Memoryin large animals, although basic startle habituation can be evaluated in some forms of cognitivetesting, such as the conditioned eyeblink response in the rabbit. Instead, many testing facilitiesevaluate sensory perception in large animal models by use of standard assessments in a functionalobservational battery examination. These measures of sensory perception provide an indication ofhearing and responsiveness but do not easily afford the opportunity for quantitative evaluation ofthe response magnitude or evaluation of habituation.3) Learning and Memory — Although there are several forms of learning and memory testingparadigms in use, four accepted models will be discussed in this section, including the Biel watermaze, the Morris water maze, active avoidance, and passive avoidance. Each model has its ownadvantages and limitations that have been presented in various forums and are not expounded uponin this chapter.The model of learning and memory used at the authors’ laboratory is the complex eight-unitT-maze (Biel water maze), 155 initially applied to neurotoxicity testing by Vorhees and colleagues. 156Each animal is assessed for acquisition of learning and memory at a single stage of development;for standard reproduction and developmental neurotoxicity studies, the most common ages of onsetof testing are approximately PND 22 or 62. According to the current EPA developmental neurotoxicitytesting guideline, 56 it is not appropriate to test an animal at more than one stage by use ofthe same learning and memory paradigm because it may confound the results of both the learningand memory components of the test. However, if two different measures of learning are used inthe same study, the same animals may be used at multiple ages because there would be no expectedconfounding influences of one testing period on the other. This would provide the additional benefitof a longitudinal assessment of learning and memory. Regardless, for the Biel maze procedure usedat the authors’ testing facility (refer to Table 8.10), the first day of testing is used to measure theanimal’s ability to navigate a straight channel, followed by 5 days of testing during which theanimal is required to learn first the forward path and then the reverse path. On the final day oftesting, animals are probed for the ability to remember the forward path. The first four trials(conducted on testing days 2 to 3) are designed to measure learning in the forward path and shortertermmemory of that path. The second six trials (conducted on testing days 4 to 6) are designedto measure learning and shorter-term memory in the reverse path. The last two trials (conductedon testing day 7) are designed to measure long-term memory of the forward path.An alternative model of learning and memory used in many laboratories is a maze developedby Morris. 157 Like the Biel maze, the Morris maze involves immersion of the rodent in water,although the maze in this case is a circular tank divided into quadrants. The animal is placed inthe tank and must learn the location of the submerged platform through a series of trials. After asufficient number of trials, the platform is moved to a different quadrant, and the animal must learnthe new location. For all trials, the time spent in each quadrant of the tank and the number of errors(entries into the wrong quadrant) are recorded. 158 Learning can be observed as an increase in thetime spent in the correct quadrant, as well as by a decrease in the number of errors. Memory isassessed in one of two ways. If the animal is not tested for a period after learning the location ofthe platform, memory can be observed by relatively few errors and relatively long periods spent© 2006 by Taylor & Francis Group, LLC

314 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONin the correct quadrant. Alternatively, if the platform is placed in a new location, memory is notedas a dramatic decrease in the time spent in the correct quadrant and a large increase in the numberof errors, while the animal learns the new location.One of the most popular tests of learning and memory is passive avoidance. This assessmentinvolves the use of an aversive stimulus to shape the animal’s behavior. One typical design involvesplacement of the rodent in a lighted chamber, with access to a dark chamber. 159 As rats prefer adarkened environment, the rat will cross into the darkened chamber. Upon entry, an electric shockis administered to the animal through the floor plate, ending the trial. The latency for the rat tocross into the darkened chamber is recorded. After one or more training trials, the rat is placed inthe lighted chamber and the latency to cross into the darkened chamber is again recorded. Anincrease in the latency to cross is considered evidence of learning. To determine the retention oftraining (i.e., memory), the animal is returned to the testing apparatus after a period of days, andthe latency to cross is recorded. 154In contrast to the passive avoidance paradigm, active avoidance requires that the animal performa task to avoid the aversive stimulus. 159 In this evaluation, an animal is placed into a chamber andallowed access to another chamber. A conditional stimulus, such as a tone or light, is presented,whereupon the animal must cross into the opposite chamber within a given time frame. If theanimal does not cross within the allotted time, an aversive stimulus (typically an electric foot shock)is applied until the animal moves to the other chamber, and an escape event is recorded. If theanimal crosses within the allotted time, an avoidance is recorded. The typical measures of learningare the numbers of escapes and the numbers of avoidances over a series of trials.A common assessment paradigm for active avoidance is the two-way shock avoidance, whichalters the placement of the electric shock between the two chambers. In this paradigm, the rat mustlearn that the conditional stimulus is applied prior to the presentation of the shock and that to avoidor escape the aversive stimulus it must reenter a chamber in which it was previously exposed tothe shock. 159 Another paradigm of active avoidance is the Y-maze. 160–164 In this assessment, threearms are used in place of the two chambers for the two-way shock avoidance. The rat is placed inan arm of the maze, the conditional stimulus is applied, and the rat must move to the appropriatearm to avoid the aversive stimulus. If the rat crosses to the correct arm within the allotted time, anavoidance is recorded. If the rat crosses to the incorrect arm within the allotted time, an error isrecorded, and the aversive stimulus is applied. If the rat does not cross within the allotted time, anescape event is recorded when the rat crosses after being exposed to the aversive stimulus.For water-based assessments of learning and memory, when comparing treated groups to thecontrol group, one must first determine whether the test article–treated groups demonstrate difficultiesin swimming ability. If there is a significant change in this parameter, all other endpointsin the assessment may be affected. Developmental delays can impair swimming ability. This usuallylengthens the latency to escape within the first few days of testing during the early postweaningperiod, without concomitant effects on the number of errors committed. Increased time to escapethroughout early postweaning testing and/or during later testing may indicate clear neuromotorimpairment. Changes in learning in the Biel maze will be detected by alterations in the slopes ofthe curves (response versus time; Figure 8.21 and Figure 8.22) for the forward and reverse paths.In the authors’ experience, the slope should appear somewhat similar to a habituation curve,although asymptotic levels are rarely achieved in either learning phase. Therefore, a potential changemay be represented by a flattened curve (animal never learns or cannot remember) or a curve thatreaches asymptotic levels (animal is a fast learner). Changes in memory may be detected in severalways. First, test article–related changes in shorter-term memory may be detected by alterations inthe slope of the line between the two trials on the same day. Effects on shorter-term memory mayalso present as changes in the slope of the line between two sequential sessions on different days.Test article–related effects on long-term memory may be detected as significantly longer latenciesto escape or as a greater mean number of errors in trial 11 than in trial 1, with relatively littlechange between trials 11 and 12.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 315Time (sec)180160140120100806040200Swim 1 2 3 6 7 8 9 10 11 124 5Trials0 ppm10 ppm30 ppm100 ppmFigure 8.21Times to escape the Biel maze for PND 22 male rats (WIL Research Laboratories, LLC).Errors (counts)40353025201510501 2 3 4 5 6 7 8 9 10 11 12Trials0 ppm 10 ppm 30 ppm 100 ppmFigure 8.22Number of errors committed by PND 22 male rats during Biel maze testing (WIL ResearchLaboratories, LLC).When evaluating passive avoidance, the change in latency over time becomes the measure ofboth learning and memory. Since the aversive stimulus typically is only applied during the trainingsession(s), learning can be observed as an increase in the latency to cross to the darkened chamber(when multiple training sessions are used). Upon completion of the training session(s), the aversivestimulus is no longer applied and memory can be examined by the latency to cross to the darkenedchamber. An additional assessment of learning can be observed as a decrease in the latency to crosswith repeated tests, as the animal learns that the aversive stimulus is no longer being applied tothe floor plate in the darkened chamber.Evaluation of the data for active avoidance depends on the paradigm being used. With two-wayshock avoidance, learning can be detected as a decrease in the number of escapes and an increasein the number of avoidances for a series of trials. In the conduct of the Y-maze avoidance test,learning is observed as a decrease in the number of escapes and in the number of errors, while thenumber of avoidances increases with learning. Again, memory can be observed as a relatively smalldifference in the performance of the animal between the training and memory sessions.© 2006 by Taylor & Francis Group, LLC

316 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIn the overall assessment of the response, it is important to evaluate interrelated endpoints inother behavioral and/or physiological evaluations before drawing conclusions. For example, anincrease in the mean latency to escape from a straight channel during Biel maze swimmingevaluation might also be correlated with decreased motor activity, decreased peak response to anauditory stimulus, or reduced body weight. The Biel maze test evaluates a unique form of habituation.Immersion in water provides an aversive stimulus to rats. Therefore, rats must habituate tothe test conditions while also learning and remembering the escape route from the maze. It is likelythat a significant test article–related effect on habituation in locomotor activity or the auditorystartle response may also manifest itself in the early phases of the learning component of the Bielmaze test. Finally, changes in neuromotor function can negatively affect the ability to interpret allthe described assessments of learning and memory. If an animal is not able to locomote or hasimpaired motor coordination, it may be unable to perform the required tasks adequately. Thus,these data should be examined in combination with data from other measures of neuromotorfunction, such as the locomotor activity assessment or grip strength, rotarod performance, andlanding foot splay, in the functional observational battery assessment.4) Functional Observational Battery Assessments — The current FDA draft guidance forjuvenile toxicity studies 51 states that well-established methods should be used to monitor keyfunctional domains of the central nervous system and these should include assessments of reflexontogeny, sensorimotor function, locomotor activity, reactivity, and learning and memory. Many ofthese domains can be evaluated in the functional observational battery (FOB) assessment. The FOB,in combination with an automated measure of locomotor activity, was designed to quickly screenfor neurobehavioral deficits in Tier 1 adult neurotoxicity studies. 165 It has been incorporated into avariety of study designs, including acute and subchronic neurotoxicity studies, developmentalneurotoxicity studies, and more recently, nonclinical juvenile studies. 56,129,166,167 The FOB consistsof a series of endpoints that assess six functional domains, as first defined by Moser. 165 At theauthors’ testing facility, the FOB includes the parameters listed in Table 8.11.The FOB can be adapted to nonrodent species with some modifications. At the authors’laboratory, FOB testing paradigms have been developed for use in the dog and nonhuman primate,and the authors are aware of FOB tests designed for use in the rabbit. 168 Although evaluationscannot necessarily be conducted in an open-field arena for some large animal models (e.g., nonhumanprimates and rabbits), deficits in neuromotor function that may affect gait can be characterizedby modifying standard assessments used in rodents.Regardless of the animal model chosen and the specific endpoints evaluated in the FOB, it isimperative that appropriate controls (e.g., training, interobserver reliability, validation) be implementedat the testing facility. First, the investigator must ensure that observers are appropriatelytrained in the assessments, and the observer should understand the types of abnormal neurobehavioraloutcomes that may be encountered. Of equal importance, the observer must be able torecognize the normal behavior of the test species. In the authors’ laboratory, new personnel are notpermitted to be trained in the rat FOB until after a year of employment. It is our position that ittakes at least a year to develop a good understanding of normal rat behaviors, and even then, FOBtraining requires that the individual evaluate 60 control animals in addition to approximately 20animals that have been treated with various positive control agents.Another important control in the conduct of the FOB is the use of blinded examinations. Sincemany of the endpoints included in the FOB are subjective in nature, it is possible for the observerto unintentionally bias the results by applying slightly harsher expectations to animals in higherdose groups than to control animals. Therefore, neurotoxicity testing guidelines require FOBassessments to be conducted without knowledge of the animal’s dosage group assignment. 166,167The testing laboratory must also ensure that observer variability be minimized. This can beaccomplished by using only one observer for the entire study, which may not always be possible,or by conducting specialized positive control studies designed to minimize interobserver variability.© 2006 by Taylor & Francis Group, LLC

Table 8.11 Functional domains as described by Moser and associated battery testsHome CageObservationsSensoryObservationsHandlingObservationsNeuromuscularObservationsPhysiologicalObservationsOpen FieldObservationsBiting - E Air righting reflex - N Ease of handling animal in hand - E Grip strength - fore- and hindlimb - N Body temperature - P Arousal - EConvulsions or tremors - E Approach response - S Ease of removal from cage - E Hindlimb extensor strength - N Body weight - P Backing - CFeces consistency - A Eyeblink response - S Eye prominence - A Hindlimb foot splay - N Catalepsy - P Bizarre or stereotypic behavior - EPalpebral closure - A,C Forelimb response - S Fur appearance - A Rotarod performance - N Convulsions or tremors - EPosture - N Hindlimb extension - S Lacrimation or chromodacryorrhea - A Gait - NOlfactory orientation - S Mucous membranes, eye, or skin color - A Gait score - NPupil response - A Muscle tone - N Grooming - NStartle response - S Palpebral closure - A,C Mobility - NTail pinch response - S Piloerection - P Rearing - CTouch response - S Red or crusty deposits - A Time to first step (seconds) - NRespiratory rate or character - PUrination or defecation - ASalivation - ANote:Endpoints observed during the functional observational battery assessment conducted at the authors’ testing facility. Functional domains, as defined by Moser (Moser,V.C., J. Am. Coll. Toxicol., 10, 661, 1991) examined by each endpoint are as follows: A = autonomic, N = neuromuscular, S = sensorimotor, E = CNS excitability, C =CNS activity, P = physiologicalNONCLINICAL JUVENILE TOXICITY TESTING 317© 2006 by Taylor & Francis Group, LLC

318 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThese interobserver reliability studies should be conducted periodically to prevent drift in theobservations by a single observer or across multiple observers. 166,169Finally, FOB assessments in large animal species should be conducted only after the observerhas developed a rapport with the subject. For example, prior to conducting the FOB in nonhumanprimates, the observer should enter the testing room and sit quietly for several minutes to allowthe animals to become acclimated to the observer’s presence.In adult neurotoxicity studies conducted with the rat, pretest FOB assessments are performedto provide baseline data on the behavior and neuromotor function of the animal. Pretest evaluationsare particularly important when conducting studies with dogs and nonhuman primates becausethese species have distinct personalities and behaviors that are intrinsic to the individual animal.It is, therefore, especially important to understand what the normal behavior for these animals isprior to administration of a drug that may influence that behavior. Unfortunately, pretest evaluationsusually are not possible for nonclinical juvenile studies because animal behavior and neuromotorfunction change with age, and pretest evaluations of neonatal animals are unlikely to predict thepattern of responses that these animals will display as adults. For example, in a juvenile rat studyin which dose administration is initiated on PND 7, gait cannot be evaluated during the predosingperiod because rats at this age are incapable of coordinated locomotion. For the same reason,posttreatment FOB assessments may not be comparable to assessments conducted during thetreatment period because neurological development may have progressed to a point beyond whichjuvenile toxicities may be observed.6. Reproductive AssessmentAlthough reproductive performance assessment is required in the standard battery of tests conductedfor submission to the EPA and FDA, 55,100,131,170 it has also been included in many juvenile studydesigns for a number of reasons. As with other endpoints, differential sensitivity to the test articlefollowing juvenile or adult exposure may lead to different effects on reproduction. In addition, theperiod of exposure often encompasses sexual maturation of the test system. Therefore, reproductiveperformance may be affected after juvenile exposure, and such assessments should be consideredas part of a juvenile toxicity program. Regardless of the reason for inclusion in the study, specificdetails of the design, conduct, and interpretation of these assessments are discussed in Chapters 9and 10, and the reader is referred to those chapters for extensive overviews.F. Statistical Design and ConsiderationsOne of the most important elements in the design of any study is the statistical methodology. Thestatistical procedures employed in a study are critically dependent upon the other elements of thedesign, including the test species, sample size, animal assignment, experimental methodology, anddata collected. Each of these elements is discussed in detail in earlier sections of the chapter.Therefore, this section addresses specific issues with regard to statistical approaches. It cannot beemphasized enough that the effect of the litter cannot be ignored in the statistical analysis of juveniledata. Regardless of the age of the animal, the litter has been shown to play a key role in the responseof the animal to toxic insult, although it has been demonstrated that the effect of the litter decreaseswith increasing age. 105Many endpoints evaluated in a juvenile study are collected longitudinally. These endpoints,including body weights, food consumption, clinical pathology, and many of the behavioral tests,exhibit patterns of change over time. Often, laboratories statistically evaluate mean values at discreteintervals from test article–treated groups compared with the control group, while eyeballing alterationsin the patterns over the assessment period. One method by which patterns are evaluated isthrough graphical presentation of the data. While it is clear that one can visually identify grosschanges in body weight or food consumption by inspecting a graph, subtle changes in the pattern© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 319of growth may be overlooked. Moreover, altered patterns may not be identified by simple one-wayanalyses of variance (ANOVA), which are designed to evaluate snapshots in time. That is becausethe ANOVA proceeds by parsing the data into arbitrarily discrete blocks, which may have little todo with the actual pattern of change. However, one statistical approach that lends itself well toboth a biological and statistical evaluation of the data is the repeated measures analysis of variance(RANOVA).For a more complete discussion of the conduct and interpretation of the RANOVA, the readeris referred to the multitude of existing statistical texts. Briefly, however, the RANOVA is a meansof detecting a main effect of treatment, a main effect of time, and an interaction effect of treatmentby time. 171 Often, a main effect of treatment is a change in the cumulative mean of a particulardata set in a treated group compared with the control group. Main effects of treatment, whenpresented graphically, would be evident by a shift in the overall data, without a change in the shapeof the response curve. For many endpoints, main effects of time are not typically examined becausethe investigator expects to observe a change in the data with time (e.g., body weights changedramatically with age during the growth spurt of a juvenile animal). Interaction effects of treatmentby time are observed when the difference from control changes over time, as evident by a differencein the slopes of the response curves for control and treated animals. More complex repeatedmeasures analyses can also be conducted, using such factors in the analyses as sex and behavioralexperience. An example of one such analysis was presented as part of the CBTS, and the readeris referred to these texts for excellent reviews of the analyses. 105,171V. CONCLUSIONSChildren are a unique population for health risk assessment; they require special considerationswhen exposed to environmental toxicants or when prescribed drugs with insufficient labelinginformation. Numerous tragedies affecting the pediatric population have shaped current scientific,medical, and public opinion, resulting in legislation that has encompassed both incentive-basedvoluntary participation and mandatory requirements for assessing the effects of xenobiotics inchildren. Although pediatric clinical trials for risk assessment are more common today, nonclinicaljuvenile animal studies are increasingly recommended or required for early hazard identification.If nonclinical juvenile safety studies are undertaken, challenges inherent to their design includeunderstanding the complexity of physiological and anatomic differences that exist between immatureand mature organisms. The concept of physiological time is the basis for identifying the criticalwindows for exposure, and for selection of an appropriate animal model and the endpoints evaluated.Although guidance documents have been published, current nonclinical juvenile study designs aregenerally developed on a case-by-case basis. Factors influencing study design include the age ofthe target population, availability of safety and exposure data, relevancy of endpoints evaluated,and practicality of execution.Regulatory guidelines and study designs will continue to evolve as more is learned about organsystem maturation in various animal models, and that will lead to better predictive value ofnonclinical juvenile studies to humans. Increased focus on safety assessment for pediatric populationsshould minimize the likelihood of future tragedies while improving the overall protection ofchildren’s health.ACKNOWLEDGMENTSThe authors would like to acknowledge with deep appreciation the efforts of various individualswho contributed in tangible ways to the completion of this project. Special thanks go to Dr. GeraldSchaefer (WIL Research Laboratories, LLC) and Ms. Judy Buelke-Sam (Toxicology Services) for© 2006 by Taylor & Francis Group, LLC

320 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONtheir thoughtful comments and suggestions regarding our discussion of behavioral assessments.Jon Hurley, Kim Rhodes, Robert Wally, and Matthew Coffee compiled and verified much of theWIL historical control data referenced in this chapter. The authors would also like to recognize thenumerous staff members at WIL, too many to mention individually, who through their dedication,hard work, and meticulous data collection over the past 10 to 20 years have laid the foundationfor this chapter.References1. Ginsberg, G., Hattis, D., Sonawane, B., Russ, A., Banati, P., Kozlak, M., Smolenski, S., and Goble,R., Evaluation of child/adult pharmacokinetic differences from a database derived from the therapeuticdrug literature, Toxicol. Sci., 66, 185, 2002.2. Rice, D.C., Behavioral impairment produced by low-level postnatal PCB exposure in monkeys,Environ. Res., 80, S113, 1999.3. Rofael, H.Z., Turkall, R.M., and Abdel-Rahman, M.S., Immunomodulation by cocaine and ketaminein postnatal rats, Toxicology, 188, 101, 2003.4. Schunior, A., Zengel, A.E., Mullenix, P.J., Tarbell, N.J., Howes, A., and Tassinari, M.S., An animalmodel to study toxicity of central nervous system therapy for childhood acute lymphoblastic leukemia:effects on growth and craniofacial proportion, Cancer Res., 50, 6455, 1990.5. Halpern, S.A., American Pediatrics: The Social Dynamics of Professionalism, 1880-1980, Universityof California Press, Berkeley, CA, 1988, p. 52.6. U.S. Food and Drug Administration, Guidance for Industry: E11 Clinical Investigation of MedicinalProducts in the Pediatric Population, U.S. Department of Health and Human Services, Food and DrugAdministration, Rockville, Maryland, December 2000.7. Cuzzolin, L., Zaccaron, A., and Fanos, V., Unlicensed and off-label uses of drugs in paediatrics: areview of the literature, Fund. Clin. Pharmacol., 17, 125, 2003.8. Kearns, G.L., Impact of developmental pharmacology on pediatric study design: overcoming thechallenges, J. Allergy Clin. Immunol., 196, S128, 2000.9. Kearns, G.L., Abdel-Rahman, S.M., Alander, S.W., Blowey, D.L., Leeder, J.S., and Kauffman, R.E.,Developmental pharmacology: drug disposition, action, and therapy in infants and children, N. Engl.J. Med., 349, 1157, 2003.10. Blumer, J.L., Off-label uses of drugs in children, Pediatrics, 104, 598, 1999.11. Leeder, J.S. and Kearns, G.L., Pharmacogenetics in pediatrics: implications for practice, Pediatr. Clin.North Am., 44, 55, 1997.12. Brent, R.L., Utilization of animal studies to determine the effects and human risks of environmentaltoxicants (drugs, chemicals, and physical agents), Pediatrics, 113, 984, 2004.13. Preclinical data may support single study for pediatric cancer drugs, The Pink Sheet, 66, 35, 2004.14. Sutcliffe, A.G., Prescribing medicines for children: major problems exist, but there are some promisingdevelopments, BMJ, 319, 70, 1999.15. Nordenberg, T., Pediatric drug studies: protecting pint-sized patients, FDA Consumer Magazine, 33(3),May–June 1999. Accessed April 21, 2004 at http://www.fda.gov/fdac/features/1999/399_kids.html.16. Weiler, H.A., Paes, B., Shah, J.K., and Atkinson, S.A., Longitudinal assessment of growth and bonemineral accretion in prematurely born infants treated for chronic lung disease with dexamethasone,Early Hum. Dev., 47, 271, 1997.17. Evans, J.S. and Huffman, S., Update on medications used to treat gastrointestinal disease in children,Curr. Opin. Pediatr., 11, 396, 1999.18. Martin, R.M., Wilton, L.V., Mann, R.D., Steventon, P., and Hilton, S.R., Unlicensed and off labeldrug use for paediatric patients, BMJ, 317, 204, 1998.19. Conroy, S. and Peden, V., Unlicensed and off label analgesic use in paediatric pain management,Paediatr. Anaesth., 11, 431, 2001.20. Shacter, E. and DeSantis, P.L., Labeling of drug and biologic products for pediatric use, Drug Inf. J.,32, 299, 1998.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 32121. Antidepressant prescriptions by pediatricians have doubled: FDA analysis, The Pink Sheet, 66, 37,2004.22. Pina, L.M., Center ids top 10 drugs used off-label in out-patient setting, News Along the Pike, 3(1),6–7, 17 January 1997. Accessed April 20, 2004 at http://www.fda.gov/cder/pike/jan97.pdf.23. Wilson, J.T., Pragmatic assessment of medicines available for young children and pregnant or breastfeedingwomen, in Basic and Therapeutic Aspects of Perinatal Pharmacology, Morselli, P.L., Garattini,S., and Sereni, F., Eds., Raven Press, New York, 1975, p. 411.24. Gilman, J.T. and Gal, P., Pharmacokinetic and pharmacodynamic data collection in children andneonates, Clin. Pharmacokinet., 23, 1, 1992.25. U.S. Food and Drug Administration, Center for Drug Evaluation and Research, Office of DrugEvaluation Statistical Report, U.S. Department of Health and Human Services publication 89–233530,Rockville, MD, 1989.26. Steinbrook, R., Testing medications in children, N. Engl. J. Med., 347, 1462, 2002.27. Blassingame, K.M., Spending on childrens’[sic] Rx rises rapidly, BenefitNews.com, January 2003.Accessed May 11, 2004 at http://www.benefitnews.com/pfv.cfm?id=3882.28. Express Scripts Drug Trend Report, Express Scripts ® , 2001. Accessed May 10, 2004 http://www.express-scripts.com/other/news_views/outcomes2001/dt_report.pdf.29. Andersen, D.H., Blanc, W.A., Crozier, D.N., and Silverman, W.A., A difference in mortality rate andincidence of kernicterus among premature infants allotted to two prophylactic antibacterial regimens,Pediatrics, 18, 614, 1956.30. Gunter, J.B., Benefit and risks of local anesthetics in infants and children, Paediatr. Drugs, 4, 649,2002.31. Anderson, J.M., Cockburn, F., Forfar, J.O., Harkness, R.A., Kelly, R.W., and Kilshaw, B., Neonatalspongioform myelinopathy after restricted application of hexachlorophene skin disinfectant, J. Clin.Pathol., 34, 25, 1981.32. Lovejoy, F.H., Fatal benzyl alcohol poisoning in neonatal intensive care units, Am. J. Dis. Child., 136,974, 1982.33. Dreifuss, F.E., Santilli, N., Langer, D.H., Sweeney, K.P., Moline, K.A., and Menander, K.B., Valproicacid hepatic fatalities: a retrospective review, Neurology, 37, 379, 1987.34. Kremer, L.C.M. and Caron, H.N., Anthracycline cardiotoxicity in children, N. Engl. J. Med., 351,120, 2004.35. Weiss, C.F., Glazko, A.J., and Weston, J.K., Chloramphenicol in the newborn infant: a physiologicalexplanation of its toxicity when given in excessive doses, N. Engl. J. Med., 262, 787, 1960.36. American Academy of Pediatrics Committee on Drugs, Acetaminophen toxicity in children, Pediatrics,108, 1020, 2001.37. Lesko, S.M. and Mitchell, A.A., The safety of acetaminophen and ibuprofen among children youngerthan two years old, Pediatrics, 104, 1, 1999.38. Kearns, G.L., Leeder, J.S., and Wasserman, G.S., Acetaminophen overdose with therapeutic intent, J.Pediatr., 132, 5, 1998.39. American Academy of Pediatrics Committee on Bioethics, Informed consent, parental permission,and assent in pediatric practice, Pediatrics, 95, 314, 1995.40. U.S. Food and Drug Administration, Regulations requiring manufacturers to assess the safety andeffectiveness of new drugs and biological products in pediatric patients: final rule, Fed. Regist., 63,66632, 1998 (codified at 21 CFR §231).41. Best Pharmaceuticals for Children Act, Pub. L. 107–109, 107th U.S. Congress (January 4, 2002).Accessed March 17, 2004 http://www.fda.gov/opacom/laws/pharmkids/contents.html.42. History of pediatric studies, rule, legislation and litigation, Biotechnology Industry Organization,Washington, DC, January 8, 2002. Accessed October 29, 2004 http://www.bio.org/reg/action/pedhist.asp.43. Pediatric Research Equity Act of 2003, Pub. L. 108–155, 108th U.S. Congress (December 3, 2003).Accessed March 15, 2004 at http://www.fda.gov/opacom/laws/prea.html.44. Congress gives FDA the power to require pediatric studies. Arnold & Porter Update, December 2003.Accessed April 4, 2004 at http://www.arnoldporter.com/pubs/files/Advisory_FDA_Pediatric_Studies.PDF.45. Food and Drug Administration Modernization Act of 1997, Pub. L. 105–115, 105th U.S. Congress(November 21, 1997). Accessed April 10, 2004 at http://www.fda.gov/cder/guidance/105-115.htm.© 2006 by Taylor & Francis Group, LLC

322 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION46. U.S. Food and Drug Administration, Center for Drug Evaluation and Research, Pediatric ExclusivityLabeling Changes as of May 7, 2004. Accessed May 12, 2004 at http://www.fda.gov/cder/pediatric/labelchange.htm.47. Drug companies are developing nearly 200 medicines for children, in New Medicines in Developmentfor Children, The Pharmaceutical Research and Manufacturers of America (PhRMA), May 29, 2002.Accessed April 20, 2004 at http://www.phrma.org/newmedicines/resources/2002-05-29.57.pdf.48. ICH Expert Working Group, ICH Harmonised Tripartite Guideline, Maintenance of the ICH Guidelineon Non-Clinical Safety Studies for the Conduct of Human Clinical Trials for Pharmaceuticals, M3(M),International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticalsfor Human Use (ICH), Geneva, Switzerland, Finalized July, 1997.49. American Academy of Pediatrics Committee on Drugs, Guidelines for the ethical conduct of studiesto evaluate drugs in pediatric populations, Pediatrics, 95, 286, 1995.50. Olson, H., Betton, G., Robinson, D., Thomas, K., Monro, A., Kolaja, G., Lilly, P., Sanders, J., Sipes,G., Bracken, W., Dorato, M., Van Deun, K., Smith, P., Berger, B., and Heller, A., Concordance of thetoxicity of pharmaceuticals in humans and in animals, Regul. Toxicol. Pharmacol., 32, 56, 2000.51. U.S. Food and Drug Administration, Center for Drug Evaluation and Research, Guidance for Industry:Nonclinical Safety Evaluation of Pediatric Drug Products, U.S. Department of Health and HumanServices, Rockville, MD, 2003 (Draft).52. U.S. Environmental Protection Agency, Toxicology data requirements for assessing risks of pesticideexposure to children’s health: Report of the Toxicology Working Group of the 10X Task Force, April28, 1999 (Draft). Accessed October 13, 2004 http://www.epa.gov/scipoly/sap/1999/may/10xtx428.pdf.53. Schettler, T., Stein, J., Reich, F., Valenti, M., and Wallinga, D., In Harm’s Way: Toxic Threats to ChildDevelopment, Greater Boston Physicians for Social Responsibility (GBPSR), Cambridge, MA, 2000.Accessed October 29, 2004 at http://psr.igc.org/ihwrept/ihwcomplete.pdf.54. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Prenatal Developmental ToxicityStudy, Office of Prevention, Pesticides and Toxic Substances (OPPTS), 870.3700, 1998.55. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Reproduction and FertilityEffects, Office of Prevention, Pesticides and Toxic Substances (OPPTS), 870.3800, 1998.56. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Developmental NeurotoxicityStudy, Office of Prevention, Pesticides and Toxic Substances (OPPTS), 870.6300, August 1998.57. Kimmel, C.A. and Makris, S.L., Recent developments in regulatory requirements for developmentaltoxicology, Toxicol. Letters, 120, 73, 2001.58. Agunod, M., Yamaguchi, N., Lopez, R., Luhby, A.L., and Glass, G.B., Correlative study of hydrochloricacid, pepsin and intrinsic factor secretion in newborns and infants, Am. J. Digest. Dis., 14,400, 1969.59. Rødbro, P., Krasilnikoff, P.A., and Christiansen, P.M., Parietal cell secretory function in early childhood,Scand. J. Gastroenterol., 2, 209, 1967.60. Huang, N.N. and High, R.H., Comparison of serum levels following the administration of oral andparenteral preparations of penicillin to infants and children of various age groups, J. Pediatr., 42, 657,1953.61. Morselli, P.L., Antiepileptic drugs, in Drug Disposition during Development, Morselli, P.L., Ed.,Spectrum, New York, 1977, p. 60.62. Siber, G.R., Echeverria, P., Smith, A.L., Paisley, J.W., and Smith, D.H., Pharmacokinetics of gentamicinin children and adults, J. Infect. Dis., 132, 637, 1975.63. Miyawaki, T., Moriya, N., Nagaoki, T., and Taniguchi, N., Maturation of B-cell differentiation abilityand T-cell regulatory function in infancy and childhood, Immunol. Rev., 57, 61, 1981.64. Barnett, J.B., Developmental immunotoxicology, in Experimental Immunotoxicology, Smialowicz,R.J. and Holsapple, M.P., Eds., CRC Press, Boca Raton, FL, 1996, p. 47.65. Holliday, S.D. and Smialowicz, R.J., Development of the murine and human immune system: differentialeffects of immunotoxicants depend on time of exposure, Environ. Health Perspect., 108, 463,2000.66. Dietert, R.R., Etzel, R.A., Chen, D., Halonen, M., Holladay, S.D., Jarabek, A.M., Landreth, K., Peden,D.B., Pinkerton, K., Smialowicz, R.J., and Zoetis, T., Workshop to identify critical windows ofexposure for children’s health: immune and respiratory systems work group summary, Environ. HealthPerspect., 108, 483, 2000.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 32367. Rice, D. and Barone, S., Jr., Critical periods of vulnerability for the developing nervous system:evidence from human and animal models, Environ. Health Perspect., 108, 511, 2000.68. Bos, A.F., Dibiasi, J., Tiessen, A.H., and Bergman, K.A., Treating preterm infants at risk for chroniclung disease with dexamethasone leads to an impaired quality of general movements, Biol. Neonate,82, 155, 2002.69. Barrington, K.J., The adverse neuro-developmental effects of postnatal steroids in the preterm infant:a systemic review of RCTs., BMC Pediatr., 1, 1, 2001.70. Gupta, A. and Waldhauser, L.K., Adverse drug reactions from birth to early childhood, Pediatr. Clin.N. Am., 44, 79, 1997.71. Miller, R.P., Roberts, R.J., and Fischer, L.J., Acetaminophen elimination kinetics in neonates, children,and adults, Clin. Pharmacol. Thera., 19, 284, 1976.72. Corcoran, G.B., Mitchell, J.R., Vaishnav, Y.N., and Horning, E.C., Evidence that acetaminophen andN-hydroxyacetaminophen form a common arylating intermediate, N-acetyl-p-benzoquinoneimine,Mol. Pharmacol., 18, 536, 1980.73. Coté, C.J., Kauffman, R.E., Troendle, G.J., and Lambert, G.H., Is the “therapeutic orphan” about tobe adopted, Pediatrics, 98, 118, 1996.74. U.S. Food and Drug Administration, Federal policy for the protection of human subjects: additionalprotections for children involved as subjects in research (technical amendment), Fed. Regist., 56,28032, 1991 (codified at 45 CFR §46.404-406).75. U.S. Food and Drug Administration, Specific requirements on content and format labeling for humanprescription drugs: revision of the “pediatric use” subsection in the labeling, Fed. Regist., 59, 64240,1994 (codified at 21 CFR §201.57).76. Roberts, R.J., Overview of similarities and differences between children and adults: implications forrisk assessment, in Similarities & Differences Between Children & Adults: Implications for RiskAssessment, Guzelian, P.S., Henry, C.J., and Olin, S.S., Eds., ILSI Press, Washington, D.C., 1992, p. 11.77. ILSI Risk Science Institute Expert Working Group on Direct Dosing of Pre-weaning Mammals inToxicity Testing, Principles and Practices for Direct Dosing of Pre-weaning Mammals in ToxicityTesting and Research, Zoetis, T. and Walls, I., Eds., ILSI Press, Washington, DC, 2003.78. Diaz, J., Schain, R.J., and Bailey, B.G., Phenobarbital-induced growth retardation in artificially rearedrat pups, Biol. Neonate, 32, 77, 1977.79. Fonseca, N.M., Sell, A.B., and Carlini, E.A., Differential behavioral responses of male and femaleadult rats treated with five psychotropic drugs in the neonatal stage, Psychopharmacologia, 46, 253,1976.80. Steinschneider, M., Hexachlorophene, in Experimental and Clinical Neurotoxicity, Spencer, P.S. andSchaumburg, H.H., Eds., Oxford University Press, Oxford, 2000, p. 630.81. Yokoyama, H., Onodera, K., Yagi, T. and Iinuma, K., Therapeutic doses of theophylline exert proconvulsanteffects in developing mice, Brain Dev., 19, 403, 1997.82. Buelke-Sam, J., Comparative schedules of development in rats and humans: implications for developmentalneurotoxicity testing, presented at the 2003 Annual Meeting of the Society of Toxicology,Salt Lake City, UT, 2003.83. Zoetis, T. and Hurtt, M.E., Species comparison of lung development, Birth Defects Res. Part B Dev.Reprod. Toxicol., 68, 121, 2003.84. Fouser, L. and Avner, E.D., Normal and abnormal nephrogenesis, Am. J. Kidney Dis., 21, 64, 1993.85. Gomez, R.A. and Norwood, V.F., Recent advances in renal development, Curr. Opin. Pediatr., 11,135, 1999.86. Kleinman, L.I., Developmental renal physiology, Physiologist, 25, 104, 1982.87. Zoetis, T. and Hurtt, M.E., Species comparison of anatomical and functional renal development, BirthDefects Res. Part B Dev. Reprod. Toxicol., 68, 111, 2003.88. Beck, M.J., Nemec, M.D., Schaefer, G.J., Stump, D.G., and Buelke-Sam, J., Validation of peripheraltissue cholinesterase activity assessment in rats administered chlorpyrifos by gavage [Abstract],Teratology, 61, P5, 2002.89. Rice, D.C., Age-related increase in auditory impairment in monkeys exposed in utero plus postnatallyto methylmercury, Toxicol. Sci., 44, 191, 1998.90. Rice, D.C., Evidence for delayed neurotoxicity produced by methylmercury, Neurotoxicology, 17,583, 1996.© 2006 by Taylor & Francis Group, LLC

324 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION91. Barone, S., Jr., Stanton, M.E., and Mundy, W.R., Neurotoxic effects of neonatal triethyltin (TET)exposure are exacerbated with aging, Neurobiol. Aging, 16, 723, 1995.92. Beck, M.J., Ching, S.V., Leung, E., Moorman, A.R., Lai, A.A., Radovsky, A., and Buelke-Sam, J.,An acute toxicity and pharmacokinetic study of the ace inhibitor, ramipril (R), in juvenile rats[Abstract], Birth Defects Res. Part A Clin. Molecular Teratol., 70, 294, 2004.93. Snodgrass, W.R., Physiological and biochemical differences between children and adults as determinantsof toxic response to environmental pollutants, in Similarities and Differences Between Childrenand Adults: Implications for Risk Assessment, Guzelian, P.S., Henry, C.J., and Olin, S.S., Eds., ILSIPress, Washington, DC, 1992, p. 35.94. Plunkett, L.M., Turnbull, D., and Rodricks, J.V., Differences between adults and children affectingexposure assessment, in Similarities and Differences Between Children and Adults: Implications forRisk Assessment, Guzelian, P.S., Henry, C.J. and Olin, S.S, Eds., ILSI Press, Washington DC, 1992,p. 79.95. Adolph, E.F. and Heggeness, F.W., Age changes in body water and fat in fetal and infant mammals,Growth, 35, 55, 1971.96. Kacew, S., General principles in pharmacology and toxicology applicable to children, in Similaritiesand Differences Between Children and Adults: Implications for Risk Assessment, Guzelian, P.S., Henry,C.J. and Olin, S.S, Eds., ILSI Press, Washington DC, 1992, p. 24.97. Radde, I.C., Mechanisms of drug absorption and their development, in Textbook of Pediatric ClinicalPharmacology, Macleod, S.M. and Radde, I.C., Eds., PSG Publishing Co., Littleton, MA, 1985, p. 17.98. Ballatori N. and Clarkson T.W., Developmental changes in the biliary excretion of methylmercuryand glutathione, Science, 216, 61, 1982.99. Salvatoni, A., Piantanida, E., Nosetti, L., and Nespoli, L., Inhaled corticosteroids in childhood asthma:long-term effects on growth and adrenocortical function, Paediatr. Drugs, 5, 351, 2003.100. Organization for Economic Cooperation and Development (OECD), Guideline for Testing of Chemicals,Guideline 416, Two-Generation Reproduction Study, Paris, France, 2001.101. Gray, L.E., Jr. and Foster, P.M.D., Significance of experimental studies for assessing adverse effectsof endocrine-disrupting chemicals, Pure Appl. Chem., 75, 2125, 2003.102. Gray, L.E. Jr., Wilson, V., Noriega, N., Lambright, C., Furr, J., Stoker, T.E., Laws, S.C., Goldman, J.,Cooper, R.L., and Foster, P.M., Use of the laboratory rat as a model in endocrine disruptor screeningand testing, ILAR J., 45, 425, 2004.103. George, J.D., Tyl, R.W., Hamby, B.T., Myers, C.B., and Marr, M.C., One-Generation Extension Studyof Vinclozolin and Di-n-butyl Phthalate Administered by Gavage on Gestational Day 6 to PostnatalDay 20 in Cd ® (Sprague-Dawley) Rats, presented at the Endocrine Disruptor Methods ValidationSubcommittee (EDMVS) Seventh Plenary Meeting, June 5–6, 2003. Accessed October 29, 2004 athttp://www.epa.gov/scipoly/oscpendo/docs/edmvs/onegenextensionfinalreport5803.pdf.104. National Academy of Sciences, Sample size determination, in Guidelines for the Care and Use ofMammals in Neuroscience and Behavioral Research, National Academies Press, Washington, DC,2003, p. 176, Appendix A.105. Buelke-Sam, J., Kimmel, C.A., Adams, J., Nelson, C.J., Vorhees, C.V., Wright, D.C., Omer, V.St.,Korol, B.A., Butcher, R.E., Geyer, M.A., Holson, J.F., Kutscher, C.L., and Wayner, M.J., Collaborativebehavioral teratology study: results, Neurotoxicol. Teratol., 7, 591, 1985.106. U.S. Food and Drug Administration, ICH guideline on detection of toxicity to reproduction formedicinal products, Fed. Regist., 59, 183, 1994.107. Organization for Economic Cooperation and Development, Guidelines for Testing of Chemicals.Section 4, No. 414, Teratogenicity, adopted January 22, 2001.108. Office of Food Additive Safety, IV.C.9.b., Guidelines for developmental toxicity studies, Redbook2000—Toxicological Principles for the Safety Assessment of Food Ingredients, Office of PremarketApproval, U.S. FDA, CFSAN, College Park, MD, 2000.109. U.S. Food and Drug Administration, Center for Biologics Evaluation and Research, Guidance forIndustry—Considerations for Reproductive Toxicity Studies for Preventive Vaccines for InfectiousDisease Indications, U.S. Department of Health and Human Services, Rockville, MD, 2000 (Draft).110. Japanese Ministry of Agriculture, Forestry and Fisheries (MAFF), Developmental Toxicity TestingGuidelines, 1985.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 325111. Vitarella, D., James, R.A., Miller, K.L., Struve, M.F., Wong, B.A., and Dorman, D.C., Developmentof an inhalation system for the simultaneous exposure of rat dams and pups during developmentalneurotoxicity studies, Inhalation Toxicol., 10, 1095, 1998.112. Barr, G.A. and Lithgow, T., Effect of age on benzodiazepine-induced behavioural convulsions in rats,Nature, 302, 431, 1983.113. Sheets, L.P., A consideration of age-dependent differences in susceptibility to organophosphorous andpyrethroid insecticides, Neurotoxicology, 21, 57, 2000.114. Cantalamessa, F., Acute toxicity of two pyrethroids, permethrin, and cypermethrin in neonatal andadult rats, Arch. Toxicol., 67, 510, 1993.115. Kallman, M.J., Assessment of motoric effects, in Developmental Neurotoxicology, Harry, G.J., Ed.,CRC Press, Boca Raton, 1994, pp. 103–122.116. Shafer, T.J., Meyer, D.A., and Crofton, K.M., Developmental neurotoxicity of pyrethroid insecticides:critical review and future research needs, Environ. Health Perspect. (in press, accessed October 29,2004 at http://ehp.niehs.nih.gov/members/2004/7254/7254.pdf).117. Schaefer, G.J., Nemec, M.D., Herberth, M.T., Pitt, J.A., Radovsky, A.E., and Knapp, J.F., Validationof developmental neurotoxicity endpoints in rats administered methimazole in drinking water[Abstract], The Toxicologist, 66, 234, 2002.118. Ruppert, P.H., Dean, K.F., and Reiter, L.W., Comparative developmental toxicity of triethyltin usingsplit-litter and whole-litter dosing, J. Toxicol. Environ. Health, 12, 73, 1983.119. Pearson, D.E., Teicher, M.H., Shaywitz, B.A., Cohen, D.J., Young, J.G., and Anderson, G.M., Environmentalinfluences on body weight and behavior in developing rats after neonatal 6-hydroxydopamine,Science, 209, 715, 1980.120. Pizzi, W.J., Rode, E.C., and Barnhart, J.E., Differential effects of methylphenidate on the growth ofneonatal and adolescent rats, Neurotoxicol. Teratol., 9, 107, 1987.121. Pizzi, W.J., Rode, E.C., and Barnhart, J.E., Methylphenidate and growth: demonstration of a growthimpairment and a growth-rebound phenomenon, Dev. Pharmacol. Ther., 9, 361, 1986.122. Greely, G.H. and Kizer, J.S., The effects of chronic methylphenidate treatment on growth and endocrinefunction in the developing rat, J. Pharmacol. Exp. Ther., 215, 545, 1980.123. Zoetis, T., Tassinari, M.S., Bagi, C., Walthall, K., and Hurtt, M.E., Species comparison of postnatalbone growth and development, Birth Defects Res. Part B Dev. Reprod. Toxicol., 68, 86, 2003.124. Padgett, E.L., Chengelis, C.P., Rhodes, K.K., Holbrook, D.L., Haas, M.C., Lehman, J.L., Bell, C.R.,and Nemec, M.D., Postpartum developmental reference data from beagle dogs over four definedpediatric stages, The Toxicologist, 66, 236, 2002.125. Roy, A.K., Schiop, M.J., and Dowbenko, D.J., Regulation of the hepatic synthesis of alpha-2 urinaryglobulin and its corresponding messenger RNA in maturing male rats, FEBS Letters, 70, 137, 1976.126. Neuhaus, O.W. and Flory, W., Age-dependent changes in the excretion of urinary proteins by the rat,Nephron., 22, 570, 1978.127. Gopinath, C., Prentice, D.E. and Lewis, D.J., Atlas of Experimental Toxicological Pathology, KluwerAcademic Publishers, Boston, 1987, p. 78.128. Gomez R.A., Sequeira Lopez, M.L., Fernandez, L., Chernavvksy, D.R., and Norwood, V.F., Thematuring kidney: development and susceptibility, Ren. Fail., 21, 283, 1999.129. Organization for Economic Cooperation and Development (OECD), Guideline for the testing ofchemicals, Developmental Neurotoxicity Study, Proposal for a New Guideline 426, September 2003(Draft).130. Lochry, E.A., Concurrent use of behavioral/functional testing in existing reproductive and developmentaltoxicity screens, J. Am. Coll. Toxicol., 6, 433, 1987.131. ICH Expert Working Group, ICH Harmonised Tripartite Guideline, Detection of Toxicity to Reproductionfor Medicinal Products, S5A, International Conference on Harmonisation of TechnicalRequirements for Registration of Pharmaceuticals for Human Use (ICH), Geneva, Switzerland, FinalizedJune 1993.132. Adams, J., Methods in behavioral teratology, in Handbook of Behavioral Teratology, Riley, E.P. andVorhees, C.V., Eds, Plenum Press, New York, 1986, p. 67.133. Altman, J. and Sudarshan, K., Postnatal development of locomotion in the laboratory rat, Anim. Behav.,23, 896, 1975.© 2006 by Taylor & Francis Group, LLC

326 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION134. Rinaldi, P.C. and Thompson, R.R., Age, sex and strain comparison of habituation of the startle responsein the rat, Physiol. Behav., 35, 9, 1985.135. Glass, D.C. and Singer, J.E., Urban Stress: Experiments on Noise and Social Stressors, AcademicPress, New York, 1972, p. 10.136. Thompson, R.F. and Spencer, W.A., Habituation: a model phenomenon for the study of neuronalsubstrates of behavior, Psychol. Rev., 73, 16, 1966.137. Clark, R.L., Endpoints of reproductive system development, in An Evaluation and Interpretation ofReproductive Endpoints for Human Health Risk Assessment, Daston, G. and Kimmel, C.A., Eds.,ILSI Press, Washington DC, 1998, 10.138. Korenbrot, C.C., Huhtaniemi, I.T., and Weiner, R.I., Preputial separation as an external sign of pubertaldevelopment in the male rat, Biol. Reprod., 17, 298, 1977.139. Cameron, J.L., Metabolic cues for the onset of puberty, Horm. Res., 36, 97, 1991.140. Ashby, J. and Lefevre, P.A., The peripubertal male rat assay as an alternative to the Hershbergercastrated male rat assay for the detection of anti-androgens, oestrogens and metabolic modulators, J.Appl. Toxicol., 20, 35, 2000.141. Kelce, W.R., Stone, C.R., Laws, S.C., Gray, L.E., Kemppainen, J.A., and Wilson, E.M., PersistentDDT metabolite p,p’-DDE is a potent androgen receptor antagonist, Nature, 375, 581, 1995.142. Hossain, M., Devi, P.U., and Bisht, K.S., Effect of prenatal gamma irradiation during the late fetalperiod on the postnatal development of the mouse, Teratology, 59, 133, 1999.143. Devi, P.U. and Hossain, M., Effect of early fetal irradiation on the postnatal development of mouse,Teratology, 64, 45, 2001.144. Torrente, M., Colomina, M.T., and Domingo, J.L., Effects of prenatal exposure to manganese onpostnatal development and behavior in mice: influence of maternal restraint, Neurotoxicol. Teratol.,24, 219, 2002.145. Adams, J., Buelke-Sam, J., Kimmel, C.A., Nelson, C.J., Reiter, L.W., Sobotka, T.J., Tilson, H.A., andNelson, B.K., Collaborative behavioral teratology study: protocol design and testing procedures,Neurobehav. Toxicol. Teratol., 7, 579, 1985.146. Gray, L.E. Jr. and Ostby, J.S., In utero 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters reproductivemorphology and function in female rat offspring, Toxicol. Appl. Pharmacol., 133, 285, 1995.147. Kim, H.S., Shin, J.H., Moon, H.J., Kang, I.H., Kim, T.S., Kim, I.Y., Seok, J.H., Pyo, M.Y., and Han,S.Y., Comparative estrogenic effects of p-nonylphenol by 3-day uterotrophic assay and female pubertalonset assay, Reprod. Toxicol., 16, 259, 2002.148. Wood, S.L., Beyer, B.K., and Cappon, G.D., Species comparison of postnatal CNS development:functional measures, Birth Defects Res. Part B Dev. Reprod. Toxicol., 68, 391, 2003.149. Campbell, B.A., Lytle, L.D., and Fibiger, H.C., Ontogeny of adrenergic arousal and cholinergicinhibitory mechanisms in the rat, Science, 166, 635, 1969150. Buelke-Sam, J., Kimmel, G.L., Webb, P.J., Slikker, W. Jr., Newport, G.D., Nelson, C.J., and Kimmel,C.A., Postnatal toxicity following prenatal reserpine exposure in rats: effects of dose and dosingschedule, Fundam. Appl. Toxicol., 4, 983, 1984.151. Buelke-Sam, J., Ali, S.F., Kimmel, G.L., Slikker, W. Jr., Newport, G.D., and Harmon, J.R., Postnatalfunction following prenatal reserpine exposure in rats: neurobehavioral toxicity, Neurotoxicol. Teratol.,11, 515, 1989.152. Hoffman, H.S. and Fleshler, M., Startle reaction: modification by background acoustic stimulation,Science, 141, 928, 1963.153. Ison, J.R. and Hammond, G.R., Modification of the startle reflex in the rat by changes in the auditoryand visual environments, J. Comp. Physiol. Psychol., 75, 435, 1971.154. Vorhees, C.V., Acuff-Smith, K.D., Schilling, M.A., Fisher, J.E., Moran, M.S., and Buelke-Sam, J., Adevelopmental neurotoxicity evaluation of the effects of prenatal exposure to fluoxetine in rats,Fundam.. Appl. Toxicol., 23, 194, 1994.155. Biel, W.C., Early age differences in maze performance in the albino rat, J. Gen. Psych., 56, 439, 1940.156. Vorhees, C.V., Butcher, R.E., Brunner, R.L., and Sobotka, T.J., A developmental test battery forneurobehavioral toxicity in rats: a preliminary analysis using monosodium glutamate, calcium carrageenanand hydroxyurea, Toxicol. Appl. Pharmacol., 50, 267, 1979.157. Morris, R.G.M., Spatial localization does not require the presence of local cues, Learn. Motivat., 12,239, 1981.© 2006 by Taylor & Francis Group, LLC

NONCLINICAL JUVENILE TOXICITY TESTING 327158. Olton, D.S. and Wenk, G.L., Dementia: animal models of the cognitive impairments produced bydegeneration of the basal forebrain cholinergic system, in Psychopharmacology: The Third Generationof Progress, Meltzer, H.Y., Ed., Raven Press, New York, 1987, p. 941.159. Tilson, H.A. and Harry, G.J., Neurobehavioral toxicology, in Neurotoxicology, Abou-Donia, M.B.,Ed., CRC Press, Boca Raton, FL, 1999, p. 527.160. Barrett, R.J., Leith, N.J., and Ray, O.S., Permanent facilitation of avoidance behavior by d-amphetamineand scopolamine, Psychopharmacologia, 25, 321, 1972.161. Barrett, R.J., Leith, N.J., and Ray, O.S., A behavioral and pharmacological analysis of variablesmediating active-avoidance in rats, J. Comp. Physiol. Psychol., 82, 489, 1973.162. Barrett, R.J., Leith, N.J., and Ray, O.S., An analysis of the facilitation of avoidance acquisitionproduced by d-amphetamine and scopolamine, Behav. Biol., 11, 189, 1974.163. Barrett, R.J. and Steranka, L.R., An analysis of d-amphetamine produced facilitation of avoidanceacquisition in rats and performance subsequent to drug termination, Life Sci., 14, 163, 1974.164. Caul, W.F. and Barrett, R.J., Shuttle-box vs. Y-maze avoidance: The value of multiple responsemeasures in interpreting active avoidance performance, J. Comp. Physiol. Psychol., 84, 572, 1973.165. Moser, V.C., Applications of a neurobehavioral screening battery, J. Am. Coll. Toxicol., 10, 661, 1991.166. U.S. Environmental Protection Agency, Health Effects Test Guideline, Neurotoxicity Screening Battery,Office of Prevention, Pesticides and Toxic Substances (OPPTS), 870.6200, August 1998.167. Organization for Economic Cooperation and Development (OECD), Guideline for Testing of Chemicals,Guideline 424, Neurotoxicity Study in Rodents, Paris, France, 1997.168. Hurley, J.M., Losco, P.E., Hermansky, S.J., and Gill, M.W., Functional observational battery (FOB)and neuropathology in New Zealand White rabbits [Abstract], The Toxicologist, 15, 247, 1995.169. Moser, V.C., Screening approaches to neurotoxicity: a functional observational battery, J. Am. Coll.Toxicol., 8, 85, 1989.170. Office of Food Additive Safety, IV.C.9.a., Guidelines for reproduction studies, Redbook 2000—ToxicologicalPrinciples for the Safety Assessment of Food Ingredients, Office of Premarket Approval,U.S. FDA, CFSAN, College Park, MD, 2000.171. Nelson, C.J., Felton, R.P., Kimmel, C.A., Buelke-Sam, J., and Adams, J., Collaborative behavioralteratology study: statistical approach, Neurobehav. Toxicol. Teratol., 7, 587, 1985.© 2006 by Taylor & Francis Group, LLC

CHAPTER 9Significance, Reliability, and Interpretationof Developmental andReproductive Toxicity Study FindingsJoseph F. Holson, Mark D. Nemec, Donald G. Stump, Lewis E. Kaufman,Pia Lindström, and Bennett J. VarshoCONTENTSI. Introduction ........................................................................................................................330A. Basis of Chapter ........................................................................................................330B. Overview of Chapter .................................................................................................331II. Significance of Developmental and Reproductive Toxicology .........................................333A. Historical Background on Predictive Power of Animal Studieswith Regard to Human Developmental and Reproductive Outcomes......................333B. Animal-to-Human Concordance for Reproductive Toxicity.....................................337III. Study Design Considerations.............................................................................................342A. History of Study Design............................................................................................3441. Pharmaceuticals ...................................................................................................3442. Agricultural and Industrial Chemicals ................................................................347B. Species .......................................................................................................................348C. Characterization of Dose-Response Curve ...............................................................350D. Dose Selection and Maximum Tolerated Dose.........................................................352E. General Statistical Considerations.............................................................................3531. Statistical Power ..................................................................................................3532. The Litter Effect ..................................................................................................353IV. Dose Range–Characterization Studies vs. Screening Studies...........................................354A. Dose Range–Characterization Studies ......................................................................354B. Developmental and Reproductive Toxicity Screening Studies.................................358C. Converging Designs...................................................................................................361V. Developmental Toxicity Studies ........................................................................................361A. Guideline Requirements ............................................................................................3621. Animal Models ....................................................................................................3622. Timing and Duration of Treatment .....................................................................362B. Interpretation of Developmental Toxicity Study Endpoints .....................................3641. Maternal Toxicity and Its Interrelationship with Developmental Toxicity ........365329© 2006 by Taylor & Francis Group, LLC

330 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION2. Endpoints of Maternal Toxicity...........................................................................3663. Intrauterine Growth and Survival........................................................................3684. Fetal Morphology ................................................................................................371VI. Reproductive Toxicity Studies ...........................................................................................382A. Guideline Requirements ............................................................................................3821. Animal Models ....................................................................................................3822. Timing and Duration of Treatment .....................................................................385B. Interpretation of Reproductive Toxicity Study Endpoints ........................................3851. Viable Litter Size/Live Birth Index.....................................................................3862. Neonatal Growth..................................................................................................3873. Neonatal Survival Indices....................................................................................3874. Prenatal Mortality ................................................................................................3885. Assessment of Sperm Quality .............................................................................3886. Weight and Morphology (Macroscopic and Microscopic) of ReproductiveOrgans ..................................................................................................................3907. Estrus Cyclicity and Precoital Interval................................................................3918. Mating and Fertility Indices and Reproductive Outcome ..................................3929. Duration of Gestation and Process of Parturition...............................................39410. Endpoints Assessed during Lactation, Nesting, and Nursing...........................39511. Sexual Behavior.................................................................................................39812. Landmarks of Sexual Maturity..........................................................................39813. Sex Ratio in Progeny.........................................................................................40014. Functional Toxicities and CNS Maturation ......................................................40115. Oocyte Quantitation...........................................................................................401VII. Evaluation of Rare (Low Incidence) Effects....................................................................401A. Introduction................................................................................................................401B. Scope of the Problem ................................................................................................402C. Criticality of Accurate and Consistent Historical Control Data...............................404D. Case Studies...............................................................................................................4041. Dystocia ...............................................................................................................4042. ACE Fetopathy ....................................................................................................4053. Retroesophageal Aortic Arch...............................................................................4064. Lobular Agenesis of the Lung.............................................................................407E. Approach to Evaluating Rare Findings.....................................................................408VIII. The Role of Expert Judgment ..........................................................................................409A. Integrity of Database .................................................................................................410B. Biologic Dynamics and Dimensions.........................................................................412C. Relevancy and Risk Analysis ....................................................................................413IX. Conclusion ........................................................................................................................414Acknowledgements ........................................................................................................................416References ......................................................................................................................................418A. Basis of ChapterI. INTRODUCTIONThe impetus for this chapter was the editor’s request for, and the present authors’ goal to produce,the first in-depth publication on the interpretation of guideline developmental and reproductivetoxicity study findings. In recent years, increasing numbers of individuals without training in either© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 331of these disciplines have begun working in areas requiring the oversight, execution, and interpretationof these types of toxicity studies. For this reason, the authors have supplemented the datainterpretation theme with information concerning the significance and reliability of the studies.While recognizing that developmental toxicology is a subdiscipline of reproductive toxicology,the authors have separated them for reasons of convenience when examining these complex topics.However, regulatory protocol designs may entail exposure during, and evaluation of, both reproductionand development. Therefore, these terms occasionally have been used interchangeably inthis chapter.The “significance” aspect of this chapter is presented initially to demonstrate the scientific basisfor viewing these studies as strong signals of potential human hazard. The “reliability” componentsof this chapter are interspersed, as they are associated with the particular measures or proceduresof import. The sections on interpretation constitute the remainder of the chapter.The guidance provided in this chapter is in part based on the extensive database at WIL ResearchLaboratories, Inc. (WIL) and is augmented by selected reports from the open literature. For 15years, WIL has been one of a small number of contract research organizations (CROs) conductinglarge numbers of reproductive and developmental toxicity studies for regulatory submission. Theauthors have been involved with the oversight, review, and/or interpretation of more than 1500 suchstudies over a composite of 80 person-experience years. The experience gained at our laboratoryis based on consistent management, staffing, and methodologies, with studies conducted primarilyat a single facility maintaining best husbandry practices, training, and progressive approaches. Inaddition, many additional nonguideline studies have been designed and conducted to verify and/orclarify findings or elucidate modes of action for a wide array of products and chemistries, thusproviding additional information concerning variability and validity of certain endpoints.Central themes of this chapter include the following: (1) validity and concordance of experimentalanimal studies to human outcome scenarios for both developmental and reproductive toxicityendpoints, (2) analyses of selected aspects of the guidelines and their effects on reliability of thedata, (3) appropriate statistical paradigms, in simple language, with computational examples, (4)critique and guidance relative to experimental techniques, (5) new (including previously unpublished)data for selected anatomic variants and other key measures common to these studies, (6)guidance for interpretation of substatistical findings (including rare events), and (7) the role ofkinetic data in the interpretation of study outcomes.It is not possible to cover all contingencies or combinations of findings in the limited space ofthis chapter. For example, in a single developmental toxicity study, the possible number of permutationsof the variables may exceed tens of thousands. The numerous values and “trigger levels”of various endpoints cited herein probably extrapolate to other laboratory venues, databases, andexperimental scenarios. However, exceptions may occur because of differences in levels of controlof husbandry, training, and consistency of applying state-of-the-science standard operating procedures.In the final analysis, however, recognition of the interrelatedness of various endpoints is thebasis for study interpretation.B. Overview of ChapterThis chapter begins with a brief review and discussion of animal-to-human concordance of developmentaland reproductive toxicity. A historical perspective is presented, which includes earlierwork conducted in the late 1970s and early 1980s. This work forms the basis for many of today’sregulatory and scientific practices in reproductive toxicology. Much of that early work has not beenpublished in the open literature; therefore, parts of it are presented here. Of equal importance isthe significant role of this work in validating animal models as reliable predictors of potentialhazards to human development and reproduction. The latter fact elevates the importance of bestpossible practices in the design, conduct, interpretation, and extrapolation (use) of these data. Thoseissues are of great importance because animal studies are, and will remain for the foreseeable© 2006 by Taylor & Francis Group, LLC

332 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.1Sources of regulatory agency guidance for interpretation of study dataTitle Agency Date Type Status Ref.Guidelines for Developmental Toxicity RiskAssessmentStandard Evaluation Procedure:Developmental Toxicity StudiesStandard Evaluation Procedure: ReproductiveToxicity StudiesGuidelines for Reproductive Toxicity RiskAssessmentReviewer Guidance: Interpretation of StudyResults to Assess Concerns about HumanReproductive and Developmental ToxicitiesEPA 1991 Developmental toxicity Older a 39EPA 1992 Developmental toxicity Older a 147EPA 1993 Reproductive toxicity Older a 147EPA 1996 Reproductive toxicity Older a 40FDA 2001 Reproductive anddevelopmental toxicityDraft 5a“Older” refers to documents written prior to significant regulatory guideline–driven changes in study design,methodology, and/or endpoints required.future, the linchpin between hazard identification and protection of human development and reproduction.This chapter also presents a brief comparison of salient differences between the various regulatoryguidelines and requirements under which developmental and reproductive toxicity studiesare conducted. This is followed by a practical discussion on interpretation of the various dataendpoints derived from these studies. Such interpretation has not been addressed consistently orsufficiently in-depth, although various regulatory agencies have produced guidance documents onthe subject (see Table 9.1).In addition to the regulatory agency documents listed in Table 9.1, DeSesso et al. 1 and Palmerand Ulrich 2 have produced general overviews of the subject. To date, a committee of the InternationalLife Sciences Institute (ILSI) has compiled the most comprehensive treatment of the interpretationof reproductive toxicology data. 3 However, this report is oriented toward risk assessment and doesnot directly address the interpretive aspects of the relevant endpoints that would typically be requiredfor people making direct use of these data sets. That effort had been initiated, in part, out of concernexpressed by some segments of industry because the newly promulgated U.S. EnvironmentalProtection Agency (EPA) guidelines contain requirements for many new endpoints. The most recentpublication of guidance for the interpretation of reproductive toxicity data was produced in 2002under the aegis of the European Centre for Ecotoxicology and Toxicology of Chemicals (ECE-TOC). 4 This document provides guidance in the form of a structured approach to evaluation ofreproductive toxicity data, taking into account the possible role of maternal toxicity and presentingexamples from several fertility and developmental toxicity studies.The above-mentioned publications present varying structures in which developmental andreproductive toxicity data may be evaluated. For example, the reviewer guidance document producedby the U.S. Food and Drug Administration (FDA) 5 broadly classifies the various indicatorsof developmental toxicity (mortality, dysmorphogenesis, alterations to growth, and functionaltoxicities) and reproductive toxicity (effects on fertility, parturition, and lactation). It utilizes anumerical scheme to assign levels of concern to the various findings. Regulators use these levelsof concern in the overall assessment of risk to human reproduction and development. This classificationscheme, popularly known as “the Wedge” (due to its pyramidal depiction), is primarilyused to judge signal strength and is useful as an overall framework in which to evaluate the data.No system, however, no matter how well intentioned, can replace the breadth of scientific experience(including technical competency) and judgment garnered from the conduct and reporting of manydevelopmental and reproductive toxicity studies in the same laboratory. Experience remains thecrux of this level of application of the scientific method. In fact, scientific interpretation of the datadrives the application of such classification schemes. Hence, providing guidance to interpretationof these data, as contained in this chapter, is a crucial exercise.© 2006 by Taylor & Francis Group, LLC

Prior to the early 1980s, numerous peer-reviewed articles were published comparing effects inlaboratory animals and human beings. 6–9 Although animal studies to predict adverse human findingswere federally mandated, there appeared to be a number of different views regarding the reliabilityof these animal models. The view at that time conceptualized the relationship between noxiousinsults to development as having four possible outcomes (malformation, death, functional impairment,and growth retardation), rather than constituting a continuum of response (see Figure 9.1).Experimental evidence to the contrary was found in work by Snow and Tam, 10 who utilized alongitudinal study design to demonstrate that the relationship between a toxic stimulus and adversedevelopmental outcomes is complex and exhibits a temporal dimension in both detection andexpression. In their study, summarized in Table 9.2, they showed that a number of manifestationsresulted from a single treatment of mice with mitomycin-C on gestational day (GD) 6 and thatthese manifestations varied dramatically, depending on the stage of development examined.Further consideration of a number of known human teratogens led one of the present authors(J.F. Holson) to conclude that human manifestations of teratogenicity across exposure levels weremost commonly multiple outcomes (see Table 9.3). Indeed, in these human cases, multiple manifestationsof developmental toxicity (then referred to as teratogenicity) commonly resulted fromexposure to the same external dose during the same period of development.DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 333Toxic stimulusMalformationsDeathFunctional impairmentsGrowth retardationToxic stimulusGrowthretardationMalformationFunctionalimpairmentDeathFigure 9.1Relationship between toxic stimulus and adverse developmental outcome.Table 9.27½ days8½ days10½ days13½ daysBirthPostnatalMaturityDevelopmental toxicity following a single dose ofmitomycin-C given to pregnant mice on gestational day 6Embryos smaller but morphologically normalEmbryos smaller but morphologically retardedEmbryos smaller but morphologically almost normalEmbryos smaller, defects in one litter, reduced litter sizeSize normal at birth, stillborn pupsDeath (64% at 21 days), runts, motor defects, tremorsReduced fertilitySource: Snow, M.H.L. and Tam, P.P.L., Nature, 279, 555, 1979.The appendix to this chapter contains the abbreviations used in this chapter.II. SIGNIFICANCE OF DEVELOPMENTALAND REPRODUCTIVE TOXICOLOGYA. Historical Background on Predictive Power of Animal Studieswith Regard to Human Developmental and Reproductive Outcomes© 2006 by Taylor & Francis Group, LLC

334 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.3Ionizing radiationsDiethylstilbestrolAlcoholMultiple manifestations of known human embryotoxinsIUGR aFunctional impairmentsChildhood cancerMalformationsDeathMalformationsClear-cell adenocarcinoma of cervixFunctional changesMalformationsFunctional impairmentIUGRLethalityaIntrauterine growth retardation.Source: From U.S. Food and Drug Administration, National Center ForToxicological Research, Reliability of Experimental Studies for PredictingHazards to Human Development (Kimmel, C.A., Holson, J.F., Hogue,C.I., and Carlo, G.L., Eds.), NCTR Technical Report for Experiment No.6015, Jefferson, AR, 1984.The above issues were recognized by J.F. Holson to be crucial for the use of animals in developmentaland reproductive risk assessment, and became the basis for a concerted effort led by J.F. Holson,funded by the FDA and carried out at the National Center for Toxicological Research (NCTR) toevaluate the concordance (predictive reliability) of experimental animal findings to known humansituations of xenobiotic-induced maldevelopment. In 1984, a final report was completed for this project,entitled Reliability of Experimental Studies for Predicting Hazards to Human Development. 11 Thisreport was never published in the open literature. Selected highlights are presented here because theyhave direct bearing on the subject of significance of findings from regulatory reproductive toxicitystudies. Also, the findings of that study were used to reinforce the value of experimental studies asanimal models in EPA data assessment guidelines for developmental toxicology.In Table 9.4, the five original assumptions that were made for assessing reliability of animalstudies are presented. Unfortunately, in none of the other publications on this topic at the time ofthis research, nor since, were these assumptions considered in evaluating concordance.In Table 9.5, a list of agents known to be human prenatal toxicants at the time of this study ispresented. The list has not been updated from the original report. Thus, it reflects the originalfindings of this study of concordance and provides a historical perspective. This list was dictatedby Assumption 1 in Table 9.4 that concordance between animals and humans could not be assessedin any other manner than by using established effects in the human. For each agent, the first yearfor which there was a report is presented in the second column. The species for which adverseeffects on prenatal development were discovered are presented in the third column, with the firstspecies presented in parentheses.A closer evaluation of these data leads to several inferences. First, agents known to disrupt humandevelopment affect multiple species. Second, with the exception of thalidomide and polychlorinatedbiphenyls, all agents disrupting human development were first reported in a laboratory animal study.Third, in general, scientists, physicians, and regulators did not heed these initial reports of laboratoryanimal studies as adequate signals of hazard. Were one to update this list with current knowledge,similar relationships would be found, with the exception that much more credence is now given tofindings from well-designed and interpreted animal developmental toxicity studies. That credenceemanates from this NCTR study and experience gained during the intervening period.The fifth assumption made in the NCTR study was that sensitivity of the model was based oncomparability of “effect levels.” These comparisons are presented in Table 9.6. At the time of thestudy, little pharmacokinetic (PK) or toxicokinetic (TK) data were available so only the “administereddose” could be used. However, with the exception of thalidomide, there was remarkable© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 335Table 9.4Original assumptions for assessing reliability of animal tests1. Only agents with an established effect in humans and adequate informationfor both humans and animals could be evaluated for concordance of effects.• Compounds for which no effect was indicated may actually have beennegative or may have been a false negative due to the inability to detecteffects because of inadequate power of the studies.2. Statistical power of the study designs had to be considered in order to evaluateadequacy of the data and apparent “species differences” in response.• Situations in which large animal studies may have been matched to a fewcase reports and a conclusion drawn as to the poor predictability of theanimal studies were noted and reevaluated.3. The multiplicity of endpoints in developmental toxicity comprise a continuumof response (i.e., dysmorphogenesis, prenatal death, intrauterine growthretardation, and functional impairment represent different degrees of adevelopmental toxicity response).• Although this assumption would be debated by some, the weight ofexperimental and epidemiological evidence tended to support rather thanrefute the assumption.• The examples of fetal alcohol syndrome, diethylstilbestrol, andmethylmercury were discussed in support of this assumption.4. Manifestations of prenatal toxicity were not presumed to be invariable amongspecies (i.e., animal models were not expected to exactly mimic humanresponse).• Also, the human population has exhibited an array of responses that weredetermined by the magnitude of the exposure, timing of the exposure,interindividual differences in sensitivities due to genotype, interaction withother types of exposure, and interaction among all of these factors.• Just as the human and rat are not the same, all human subjects are notidentically responsive to exogenous influences.5. Sensitivity was based on comparability of the “effect levels” among species.• This was necessary because for most established human developmentaltoxicants there was still not adequate dose-response information availableto compare sensitivities among species.Source: Summarized from U.S. Food and Drug Administration, National Centerfor Toxicological Research, Reliability of Experimental Studies for Predicting Hazardsto Human Development (Kimmel, C.A., Holson, J.F., Hogue, C.I., and Carlo,G.L., Eds.), NCTR Technical Report for Experiment No. 6015, Jefferson, AR,1984.Table 9.5Agents that cause developmental toxicityAgent Year Species a Ref.Alcohol(ism) 1919 (rat), gp, ch, hu, mo 148Aminopterin 1950 (mo & rat), ch, hu 149Cigarette smoking 1941 (rab), hu, rat 150Diethylstilbestrol 1939 (mo & rat), hu, mi 151Heroin/morphine 1969 (rat), ha, hu, rab 152Ionizing radiation 1950 (mo), ha, hu, rat, rab 153Methylmercury 1965 (rat), ca, hu, mo 154Polychlorinated biphenyls 1969 (hu), rat 155Steroidal hormones 1943 (monk), ha, hu, mo, rat, rab 156Thalidomide 1961 (hu), mo, monk, rab 157,158aca — cat, ch — chicken, ha — hamster, gp — guinea pig, hu — human, mi —mink, mo — mouse, monk — monkey, rat — rat, rab — rabbit.Source: Adapted from U.S. Food and Drug Administration, National CenterFor Toxicological Research, Reliability of Experimental Studies for PredictingHazards to Human Development (Kimmel, C.A., Holson, J.F., Hogue, C.I., andCarlo, G.L., Eds.), NCTR Technical Report for Experiment No. 6015, Jefferson,AR, 1984.© 2006 by Taylor & Francis Group, LLC

336 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.6Effect levels for selected developmental toxicants in humans and test speciesAgent Species Dose ResponseAminopterin Human 0.1 mg/kg/day Death and/or malformationsRat 0.1 mg/kg/day Death and/or malformationsDiethylstilbestrol Human 0.8–1.0 mg/kg Genital tract abnormalities and/or deathMouse 1 mg/kg Genital tract abnormalities and/or deathIonizing radiation Human 20 rd/day MalformationsRat and mouse 10–20 rd/day MalformationsCigarette smoking Human 20 cigarettes/day Growth retardationRat >20 cigarettes/day Growth retardationThalidomide Human 0.8–1.7 mg/kg MalformationsMonkey 1.25–20 mg/kg MalformationsRabbit 150 mg/kg MalformationsSource: Adapted from U.S. Food and Drug Administration, National Center For Toxicological Research,Reliability of Experimental Studies for Predicting Hazards to Human Development (Kimmel, C.A., Holson,J.F., Hogue, C.I., and Carlo, G.L., Eds.), NCTR Technical Report for Experiment No. 6015, Jefferson,AR, 1984.similarity in effect levels among animals and people for the examples chosen. With the dramaticincrease in analytical capabilities and the more generalized use of metabolic, and PK and pharmacodynamic(PD) studies, comparison of adverse effects is now more routinely done based on thearea under the curve (AUC) and the peak concentration (C max ) of the agent. Hence, differences ininternal exposure among species are often demonstrated, and these reveal that most historicalinstances that were initially believed to be “species differences” in response actually resulted fromdifferences in bioavailability or metabolism. Specific examples of true “biologically based differences”in embryonal response have not been demonstrated. This is understandable given theconserved nature of embryology and development and its great similarity among mammalian andsubmammalian species. For further discussion of this topic, the reader is referred to an excellentreport prepared by the Committee on Developmental Toxicology of the National Research Council. 12Examples of compounds that are developmentally toxic in humans are presented in Table 9.7.As indicated in this table, the embryo or fetal toxicity of a substance may manifest at exposurelevels that do not significantly affect the mother during pregnancy, at levels that cause maternalstress or alter the physiological state of the mother, or at or near levels that produce overt maternaltoxicity. Animal developmental toxicity studies have revealed a similar range of responses. Thisstriking similarity in the range of relationships between maternal toxicity (or absence thereof) anddevelopmental outcome in humans is mirrored in laboratory animal studies. This point is importantbecause it further confirms the similarities among species in regard to maternal toxicity anddevelopmental outcome. It also indicates that the presence of maternal toxicity cannot a priori beassumed as the causative agent of maldevelopment in any particular instance.The database from the epidemiology and animal studies in the NCTR study is sufficient interms of relative power. Along with identification of dose-response relationships, it permits thededuction that, overall, toxic insults to the maternal and embryonal or fetal organisms are similarenough for all agents to presume concordance between animal models and humans, even whenonly external dose is considered. Inclusion of PK and TK studies has refined (and will continue torefine) the quantitative aspects of extrapolation of experimental findings to human risk assessment.Table 9.8 presents a list of publications and reports that address the issue of animal-to-humanconcordance. These are presented along with key attributes for each because they represent excellentsource material. Except for the first two reports, none contains critical analyses of the primaryliterature or use preestablished evaluation criteria.As the findings from the NCTR study indicate, further bolstered by the other studies andpublications listed in Table 9.8, the significance of experimental indications of developmentaltoxicity should garner great concern. Many scientists not working in the reproductive toxicology© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 337Table 9.7 Relationship between maternal and fetal toxicityfor established human developmental toxicantsEmbryo and fetal toxicity produced at or near exposure levels that elicitovert maternal toxicity:• Aminopterin• Methylmercury• Polychlorinated biphenylsEmbryo and fetal toxicity produced at exposure levels that cause maternalphysiologic changes or stress:• Cigarette smoking• Steroidal hormones• Ethanol consumptionEmbryo and fetal toxicity produced at exposure levels that do not producesignificant maternal effects during pregnancy:• Ionizing radiation• Diethylstilbestrol• ThalidomideSource: Adapted from Holson, J. F., Estimation and extrapolation of teratologicrisk, in Proceedings of NATO Conference on in vitro Toxicity Testingof Environmental Agents: Current and Future Possibilities, NATO, MonteCarlo, Monaco, 1980.area, corporate representatives, animal rights activists, and lay people are often confused or misguidedby inconsistencies in technical reports or lay publications that allege differences betweenhuman experience and animal studies. Such allegations present a problem for the future of thediscipline because they have contributed to past catastrophes. In Table 9.9, an attempt has beenmade to summarize the reasons for the high frequency of such misrepresentations. These reasonsare self-explanatory, with the exception of the last entry, which the authors have included out ofnecessity because none can predict all possibilities for the future. No established or generallyaccepted example of an agent that has adversely affected human development but has not inducedadverse developmental effects in an animal model currently exists. The reader is referred to theoriginal assumptions made in the NCTR study to better reconcile the reasons for apparent discrepancieslisted in Table 9.9.Developmental toxicity studies in laboratory animals provide the only controlled and ethicalmeans of identifying the potential risks of a large array of drugs, chemicals, and other agents tothe developing human embryo and fetus. However, they cannot replace direct human evidence orexperience. Often, the findings from the animal models will be supplemented by confirmation inpregnancy registries. These registries involve collection of information from human pregnanciesthrough use of questionnaires and may be conducted after product approval and/or marketing.However, pregnancy registries are variably sensitive, and rarely serve as early predictors of developmentalhazard. Therefore, the design and interpretation of results from animal developmentaltoxicity studies are critical to the hazard identification phase of risk assessment, and in the case ofequivocal findings, such studies will guide the use of postmarketing registry analyses.B. Animal-to-Human Concordance for Reproductive ToxicityReproductive toxicology is also a complex field of science, as many authors of textbooks and reviewarticles have declared in their opening paragraphs. The reason for this complexity is that sexualreproduction involves the union of gametes from two genders, resulting in progeny that in turnexperience development and maturation of the myriad processes that enable continuation of the“life cycle.” That cycle comprises several life stages, each having a unique biology and temporallyentrained acquisition of features. Reproductive physiology (and its dynamic anatomy) includes© 2006 by Taylor & Francis Group, LLC

338 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.8 Studies and publications concerning animal-to-human concordance for developmentaland reproductive toxicityAuthors Ref. AttributesHolson, 1980(Proceedings of NATOConference)Holson et al., 1981(Proceedings of ToxicologyForum)NCTR Report No. 6015, 1984Nisbet and Karch, 1983(Report for the Council onEnvironmental Quality)Brown and Fabro, 1983(journal article)Hemminki and Vineis, 1985(journal article)Buelke-Sam and Mactutus,1990 (journal article)Newman et al., 1993 (journalarticle)11, 95,159Interdisciplinary team(epidemiologists and developmental toxicologists)Critical analysis of primary literatureApplied criteria for acceptance of data/conclusions in reportsand included power considerationsEstablished and applied concept of multiple developmental toxicityendpoints as representing signals of concordanceQualitative outcomes and external dose comparisons madeNo measures of internal doseCritical analysis of primary literature129 Many chemicals and agents addressedLimited review of primary literatureNot a critical analysis of primary literatureRelied on authors’ conclusionsNo power analysesLimited use of internal dose information160 Not a critical analysis of primary literatureMade use of findings from other reviewsExcellent review and presentation of overall concordance issues130 Interspecies inhalation doses adjustedRelied on authors’ conclusionsTwenty three occupational chemicals and mixturesNo measures of internal dose161 Small number of agents coveredCritical review of the qualitative and quantitative comparability ofhuman and animal developmental neurotoxicityLimited use of internal dose measures132 Provided detailed informationOnly four drugs evaluatedEmphasis on morphologyFocus on NOAELs aNo measures of internal doseShepard, 1995 (book, 8th ed.) 162 Computer-based annotated bibliographyCatalog of teratogenic agentsNot a critical analysisLimited comments regarding animal-to-human concordance for alimited number of agentsNo use of internal dose measuresSchardein, 2000 (book, 3rd ed.) 163 Extensive compilation of open literatureNot a critical analysisVariably relied on authors’ conclusionsNo measures of internal dose nor criteria for inclusion or exclusionof studiesOnly partially devoted to concordance issuesaNo-observed-adverse-effect level.many of biology’s most complicated endocrinologic feedback and control mechanisms. It furtherencompasses the transient development (appearance and disappearance) of structures and functions,the roles of which are crucially timed to comprise the reproductive cycle. To our knowledge, andbased on a recent search of the literature, there is no published rigorous and critical analysis–basedstudy of concordance between humans and laboratory animal models for reproductive toxicityoutcomes. Several publications in the early 1980s (refer to Table 9.10) attempted to address theissue, but were limited in scope because of data gaps (e.g., lack of exposure quantification, lackof animal and/or human data for key reproductive endpoints). Admittedly, because of the numerousendpoints typically evaluated in reproduction studies, in contrast to the more focused issues of© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 339Table 9.9Reasons for apparent discrepancies between animal and human effectsWhen it is reported or found that an agent causes developmental effects in test animals, but not in humans, thismay be due to the following factors:• The animal study is flawed or improperly interpreted.• The chemical does not reach the marketplace so no human exposure occurs.• Exposure occurs but is not high enough to cause an effect.• Large enough populations with documented exposures are difficult to identify, and the effects may beconfounded by other factors.• Effects may occur but are below the level of detection for epidemiological studies.• The agent actually does not affect development in humans.Table 9.10Publications concerning animal-to-human concordance for reproductive toxicityAuthors Ref. AttributesBarlow and Sullivan, 1982 (book) 164 Many industrial chemicals addressedCritical analysis of primary literatureDid not apply criteria for inclusion of chemicals reviewed based onextent of available dataNo power analysisLimited number of endpoints comparedLimited internal dose information and human external exposure datalackingU.S. Congress, Office ofTechnology Assessment, 1985(chapter in book)165 Eleven agents coveredNot a critical analysis of primary literatureLimited to one reproductive toxicity concordance table (createdentirely from previously referenced Barlow and Sullivan (1982) andNisbet and Karch (1983) reviews) and a very brief generaldiscussionOnly addressed concordance of effect (no assessment ofconcordance of dose)developmental toxicity studies, development of a comprehensive review would be a daunting task.Thus a crucial need remains for an organized and critical analysis of the primary literature inreproductive toxicology to evaluate the concordance of regulatory reproductive toxicity studies tohuman reproductive outcomes. There have been attempts to compare effects in laboratory animalswith those in humans for specific endpoints or by gender. One such study, Ulbrich and Palmer, 13examined 117 substances or substance classes and proposed that histopathology and organ-weightanalysis were the best general-purpose means for detecting substances that affect male fertility.These authors likewise concluded that sperm analysis was a realistic alternative to histopathologyand organ-weight analysis when these proved to be impractical. Further, it has been reported thatpredictability may be observed in results from well-validated models, 14 and several publicationshave reviewed factors that are important in assessing risk to the male reproductive system. 14–17Certainly, however, the most critical feature of an animal model, as with the use of models in allareas of toxicology, should be the use of TK data, especially biotransformation data.In mammalian systems, perturbation of the reproductive organs or disturbances in the integratedcontrol of the process of reproduction can originate from multiple sources. For example, cadmiumdisrupts testicular function secondarily to its main effect on vascular integrity to the vesselssupplying the testes. The nematocide dibromochloropropane (DBCP) has several sites of action,including the testis and epididymis. Through conversion to α-chlorohydrin, DBCP causes epididymalvascular damage (progressing to sperm granulomas and spermatoceles) and decreases epididymalrespiratory activity and motility, providing the basis of its putative action in the epididymis. 18Both cadmium and DBCP ultimately induce infertility but act via divergent pathways. Standardhazard evaluation studies would not reveal the mode of action of either agent, and follow-up studieswould be required to identify the ultimate target site.© 2006 by Taylor & Francis Group, LLC

340 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION30002500No. of Animals/Study2000150010005000Acute OralToxicitySubchronicToxicity2-Year CancerBioassayDevelopmentalToxicity2-GenerationReproductionFigure 9.2Comparison of the number of animals on study in various experimental designs.Changes in a specific reproductive process in which the mode of action is well understood andconserved among species are considered a clear threat to human health. Command of the issueswill afford the most enlightened use (or will prevent misuse) of the data for the intended purpose.Reproductive and developmental toxicity studies are large in scale and complex (Figure 9.2). Nomatter what systematic approach is applied to the data to formulate conclusions, the final interpretationmust consider interrelated endpoints collectively if the risk assessment process is to proceedefficiently and effectively.One potential problem related to extrapolation of reproductive hazard from animals to humansis embedded in the phenotypic diversity of humanity. Expansion of the human population overmany years has given rise to a heterogeneous gene pool. Animals that are purpose-bred forexperimental applications are genotypically more homogeneous than, and less predictive of, theoutlier responders in the diverse human population. The logical assumption is that humans are morelikely to be predisposed to idiosyncratic toxicologic responses. A second potential problem is thatwith reproductive toxicity, as compared to developmental toxicity, there are many more structuresand processes in the maternal and paternal animals that may be affected or manifest toxicity. Thisis in contrast to prenatal development, with its highly conserved embryo and fetus, which has amuch more limited phenotypic variety among species. While some phenotypic diversity may existin embryos and fetuses among species, it is certainly far less than in the adult life stage, in whichnumerous structures reside in either or both of the parental sexes. Hence, caution must be exercisedwhen extrapolating effects from animal models to humans for many aspects of reproductive toxicity.Nevertheless, the finding of reproductive toxicity in a well-conducted guideline study should beconsidered to constitute a strong signal, which currently could be negated only by appropriateinformation on mode of action or the existence of an adequate study in humans. The use ofepidemiology has become increasingly important in establishing cause and effect relationships. 19It is an essential scientific need that an organized and critical analysis of the primary literaturein reproductive toxicology be conducted to evaluate the concordance of regulatory reproductivetoxicity studies to human reproductive outcomes. Such a study would need to be modeled after theFDA-NCTR study discussed in the beginning of this chapter. Assumptions for inclusion andexclusion of papers, consideration of power, establishment of exposure parameters, and criteria forconcordance would need to be developed and used.Figure 9.3 is an idealized depiction of the relationship between progression of time in developmentand the degree of phenotypic variability, both on an ontogenic and species basis. The earlier© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 341Extent of differentiationEmbryonicperiodFetalperiodPostnatalperiodDegree ofphenotypicvariabilityBirthTime in development (age)Figure 9.3Relationship between progression of time in development and the degree of phenotypic variability.in development, the more similar or conserved are the fundamental processes of development andtheir similarity among species. As development progresses, expression of individual and “speciesspecific”characteristics increases. In middle and late postnatal development, it would be expectedthat species-specific attributes that have evolved for adaptive success will be expressed and mayrepresent vulnerable phenomena, not present in humans or having less importance to human health.Thus, on a theoretical basis, it might be expected that the later in development a model is used,the greater the number of species differences that might be encountered among models. It wouldbe expected that for mammalian species these differences would be few and not so numerous asto thwart the rational use of laboratory models. With the advent of regulations for juvenile (pediatric)studies, experience over the next decade will be key to answering this question. The timing of onsetof functions may be problematic, as the time of parturition in different species becomes a confoundingfactor for conducting the experimental studies. An example is maturation of the renin-angiotensinsystem, which occurs periparturitionally and postnatally in the rat and prenatally in the human.DBCP is an insecticide that has been widely used on banana plantations and as a fumigant forgrain and soil. Reported effects among workers at DBCP production sites in California and in Israelincluded prolonged azoospermia and oligospermia, with a strong correlation between the durationof employment and testicular function. Potashnik and Porath published data indicating that agriculturalworkers exposed to DBCP in Israel had impaired fertility or were sterile. 20 These investigatorsreported that while follicle stimulating hormone (FSH) and luteinizing hormone (LH) weresignificantly increased, testosterone levels were not significantly decreased in men who wereseverely affected by the DBCP exposure. 20 Similarly, increased gonadotropin levels were correlatedwith decreased sperm counts in agricultural applicators in California. 21 Meistrich et al. reportedprolonged oligospermia in LBNF 1 rats injected with DBCP. 22 Within 6 to 20 weeks post-treatment,only 20% of the seminiferous tubules contained differentiating germ cells. A majority of the tubules(70%) had germinal epithelium and Sertoli cells but no differentiating germ cells. Morphologicalteration of the Sertoli cells was noted in the remaining 10% of the tubules. In contrast to thehuman hormone levels, both intratesticular testosterone and gonadotropin levels were increased inthe DBCP-treated rats. The authors concluded that DBCP exerted its effects by inducing Type Aspermatogonia to undergo apoptosis rather than differentiation. These data in humans and rodentsdemonstrate adequate concordance for hazard identification purposes.Rats and humans exhibit other differences in the reproductive cycle, such as the intrinsic featureof rat reproduction in which the LH surge is synchronized with the onset of heat to ensure fertility.Agents that block norepinephrine binding to alpha-2 receptors, inhibitors of dopamine beta-hydroxylaseactivity, gamma-amino butyric acid (GABAnergic) receptor agonists, delta-9-tetrahydrocannabinol,© 2006 by Taylor & Francis Group, LLC

342 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONand methanol attenuate or ablate the amplitude and alter the timing of the LH surge in the rat,thereby delaying or preventing ovulation. 23 Alteration of the LH surge in rats appears to be acommon site of vulnerability for a wide array of agents and chemistries. Most often the results ofsuch disruptions were initially observed as changes in fertility as well as changes in indices suchas fertility, ovulation, implantation efficiency, and litter size in two-generation reproductive performancetoxicity studies. Administration of a single dose of LH-disrupting agents on the day ofproestrus can produce multiple adverse reproductive outcomes. By interfering with the hypothalamicregulation of the preovulatory surge of LH, these agents accelerate the normal loss of ovariancycling in rats by inducing early onset of the normal age-associated impairment of central nervoussystem (CNS)–pituitary control of ovulation. The hormonal milieu present in the aging female ratis one of persistent estrogen secretion (from persistently developing follicles) and no secretion ofprogesterone. This milieu is the precise endocrine environment known to facilitate the developmentof mammary gland tumors in rats reported in a number of studies. 24–26 However, it is unlikely thatthis response bears relevance to humans because menopause in women results from depletion ofprimordial ovarian follicles and the subsequent decline in estrogen levels. Although inhibition ofovulation in the rat by these agents is relevant to human reproduction, effects on human fertilityare generally considered threshold-driven responses. Therefore, the benefits of drugs inducing thesechanges can outweigh the risks if a sufficiently large margin of safety can be demonstrated. Mostregulatory attention, however, focuses on the basis of carcinogenesis because it is viewed as obeyinga linear dose-response, even when the mode of action is on a reproductive process.From the previous discussion, the need for a well-conceived and conducted concordance studybecomes evident. Likewise, the complexities of such an endeavor are also obvious.III. STUDY DESIGN CONSIDERATIONSDevelopmental toxicity studies are designed to assess effects on prenatal morphogenesis, growth,and survival. These studies are generally required for all regulated agents (e.g., drugs, biologics,pesticides, industrial chemicals, food additives, veterinary drugs, and vaccines). While the studydesigns are similar for the various compounds that are tested, customized guideline requirementsare available to address unique properties and characteristics of each of the above categories. Forexample, the basic developmental toxicity study design for drugs suggests initiation of exposureeither at implantation or immediately following mating. However, administration of vaccines takesplace prior to mating to ensure that an immune response will occur during gestation. A pivotalconsideration in the evaluation of biologics is the frequent lack of demonstrable maternal internaldose. In these cases, a key pharmacologic biomarker may serve as a proxy measure of exposure.The key features to all developmental toxicity designs are: (1) that maternal exposure is sustainedthroughout major organogenesis and (2) that the dam is necropsied 1 to 2 days prior to expectedparturition so that all products of conception can be evaluated.In contrast to developmental toxicity studies, reproductive toxicity studies are designed to assessfertility and reproductive outcome. Since the advent of the 1966 FDA guidelines, 27 the testing ofmedicinal products has used a segmented approach to reflect human exposure. Because mostpharmaceuticals are administered for discrete courses of therapy, possess short half-lives, andminimally (if at all) bioaccumulate, the segmented approach using administration to a singlegeneration is appropriate. Alternatively, reproductive toxicity testing of industrial chemicals, pesticides,and food additives is intended to assess effects of low-level exposure (intentional and/orunintentional) over a significant portion of the life span that may result in bioaccumulation. Thetest animals are dosed via routes of administration that mimic human exposure. Later sections ofthis chapter provide more in-depth discussion of specific aspects of the guideline developmentaland reproductive toxicity study designs relating to data interpretation. Figure 9.4 presents a schematiccomparison of the various developmental and reproductive toxicity study designs.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 343A B C D EFPremating toConceptionConception toImplantationFertility and EarlyEmbryonic Development10W 4W 2W ICH 4.1.1(Segment I)Estrous Cyclicity MatingCorpora Lutea FertilityImplantation SitesPre-Implantation LossSpermatogenesisOne-, Two-, or MultigenerationStudyEstrous CyclicityMatingFertilityCorpora LuteaImplantation SitesPre-Implantation LossSpermatogenesisImplantationto Closure ofHard PalateEmbryo/FetalDevelopmentPI LossViable FetusesMalformationsDev. VariationsFetal WeightScreening Studies4W 2W OECD 421, OPPTS 870.3550Hard-PalateClosure to End ofPregnancyBirth toWeaningDenotes Dosing PeriodC MAX ICH 4.1.3OECD 414OPPTSAUC 870.3600870.3700(Segment II)Postimplant. LossViable FetusesMalformationsDev. VariationsFetal WeightC MAXPre- and Postnatal DevelopmentF 0ICH 4.1.2AUC(Segment III) F 1 ????????ParturitionLitter SizeLandmarks of Sexual DevelopmentNeurobehavioral AssessmentAcoustic Startle ResponseMotor ActivityLearning & MemorySingle/Multigeneration10WOECD 415, 416, OPPTS 870.3800, FDA Redbook I, NTP RACBSatellite PhaseF 1 ????????F 2 ????????ParturitionLitter SizeLandmarks of Sexual DevelopmentNeurobehavioral AssessmentAcoustic Startle ResponseMotor ActivityLearning & MemoryHistopathologyWeaning toSexualMaturityGestation LengthPup ViabilityPup WeightOrgan WeightsF 1 Mating andFertilityGestation LengthPup ViabilityPup WeightOrgan WeightsF 1 Mating andFertility HormonalAnalyses OvarianQuantificationEstrous CyclicityImplantation SitesOECD 478OPPTS870.5450Dominant Lethal AssayZygote/EmbryolethalityMatingFertilityLimited:MalformationsDev. VariationsOPPTS870.3500Chernoff-KavlockAssayLimited:MalformationsVariationsAssess recoverythrough multiplemating trialsParturitionGestation LengthLitter SizeHistopathologyParturitionGestation LengthLitter SizePup ViabilityPup WeightPup ViabilityPup WeightOrgan WeightsUterotrophic AssayEstrogenicityAnti-EstrogenicityHershberger AssayAndrogenicityAnti-AndrogenicityFigure 9.4Schematic comparison of the design of various developmental and reproductive toxicity studies.© 2006 by Taylor & Francis Group, LLC

344 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. History of Study DesignFigure 9.5 depicts the major milestones in the evolution of developmental and reproductive toxicology(DART), including examples of various human tragedies that provided the impetus forimprovement of study designs. In this section, pertinent events surrounding the historical developmentof DART study designs are addressed initially, followed by a similar treatment of the studydesigns relevant to agricultural and industrial chemicals.1. PharmaceuticalsRegulations governing the development, and particularly the safety assessment, of new drugs havechanged dramatically since the establishment of the FDA. A variety of factors have driven thedynamics. These include human tragedies arising from the absence of (or deficiencies in) appropriatedata, sociologic pressures, political pressures, governmental initiatives (including harmonization oftesting guidelines to promote consistency among nations and to reduce redundant studies), economicpressures, and consumer advocacy. The net result has been continual advancement of the complexityand breadth of safety assessment protocols. Figure 9.5 highlights selected events in chronologicalorder.Prior to formation of the FDA in 1930, Congress passed the original Food and Drugs Act in1906. 28 This action restricted the interstate commerce of misbranded and adulterated foods, drinks,and drugs. The Food, Drug and Cosmetic Act (FDC Act) of 1938 29 strengthened the provisions ofthe 1906 act. The most important provisions of this legislation were the introduction of the landmarkconcept of mandatory evaluation of drug safety in animal studies before marketing and the establishmentof safe tolerances for unavoidable poisons. The FDC Act laid the foundations for the firstFDA laboratory animal studies in the early 1940s. Just prior to passage of this law, diethylstilbestrol(DES) was synthesized as the first active estrogen mimic and was intended to be a therapeutic formaintenance of pregnancy in humans. Green et al. published the initial reports of adverse reproductiveeffects (intersexuality in rodents) in 1939, 30,31 but clinicians and scientists widely andmistakenly discounted the significance of the animal data. Approximately 35 years later, Herbst etal. established the causal and latent relationship of DES exposure in utero to vaginal clear-celladenocarcinoma and reproductive tract dysplasia in progeny at about the onset of puberty. 32,33Procedures for the Appraisal of the Toxicity of Chemicals in Foods, published in 1949, 34 wasamong the earliest FDA guidance documents for the evaluation of specific effects of drugs onreproduction in rats. The default assumptions of risk focused on the following concepts: (1) thatdrugs were intended to be administered for short periods of time such that insults to the body wouldbe limited and (2) that with properly designed toxicity studies, risk-benefit analysis could be usedas a justification for introduction of the drug in humans. Further, the document suggested placementof more emphasis on the toxicologic responses resulting from acute and subacute exposures coveringat least 10% of the life span of the species. Results of chronic bioassays were required duringevaluation of chemical food additives because potential exposure represented a greater portion ofthe life span, human exposure was not volitional, and most additives were without direct benefitto health.In the 1950s, reproductive toxicity tests gained more importance, especially when deleteriousand selective responses were expected in sex organs. The studies were conducted concurrently withthe chronic toxicity study. The experimental approach included the exposure of 16 males and 8females per group for 100 days, followed by two mating trials (F 1a and F 1b litters). Selected offspringfrom the F 1b litters were exposed continuously after weaning until sexual maturity, and the assessmentof functional reproductive outcome was repeated as for the first generation. The studies© 2006 by Taylor & Francis Group, LLC

Figure 9.51906Food andDrugs Act20001973DBCPFDARedbook20001938Food, DrugandCosmetic Act1975Red DyeNumber 21998OPPTS1950First US FDALaboratoryAnimal SafetyStudies1972FIFRA1997FDAMAOngoingmethylmercury andDES exposuresDES(Herbst & Scully)19961971FQPASDWA1979TSCA1994ICHGuidelines1962ThalidomideEpidemic1967–68AgentOrange/2,4,5-T& TCDD1993“ACE-Fetopathy”CoinedFDARedbook IIMajor milestones in the evolution of developmental and reproductive toxicology.1963Conference onPrenatal DrugEffectsNCTRCollaborativeBehavioralStudy19921979InternationalConcern onDecreasingFertility1985NCTRCollaborativeStudy ReportedNCTRExtrapolationModels inTeratogenesisStudy19801982FDARedbook IIsotretinoinApprovedWilson’sPrinciples1966GoldenthalGuidelinesFirst ACE-FetopathyCase Report1981NCTRConcordanceStudy(Teratology vs.DevelopmentalToxicology)OECDGuidelinesDEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 345© 2006 by Taylor & Francis Group, LLC

346 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONcontinued through a third generation to produce F 3a and F 3b progeny. Offspring from the secondmating of the F 2 generation were euthanized at weaning for histologic evaluation. Specific endpointssuggested by the guidance document were limited and included tracking success of mating andfertility, number of offspring in the litters at postnatal day (PND) 1, 5, and 21, and weight of thelitter at 21 days postpartum. Paired-feeding studies were encouraged when issues of palatabilityarose or when large doses prevented normal nutritional intake. The Association of Food & DrugOfficials of the United States endorsed these concepts of reproductive toxicity study design. 35In September 1960, the FDA received a New Drug Application (NDA) for the sedative-hypnoticthalidomide (under the brand name Kevadon ® ). The FDA refused to approve the NDA, largelybecause of the lingering concerns that there were insufficient data to support safe use in humans.By 1962, the causal relationship between thalidomide exposure during human pregnancy and limbreduction defects emerged, initially in Europe and Australia. At that time, testing for prenataltoxicity was not a mature field of science. Because of the thalidomide tragedy, the PharmaceuticalManufacturer’s Association formed the Commission on Drug Safety in 1962 and sponsored workshopswith the FDA to advance the understanding of test methods for the discipline. 36 Hence, thediscipline of teratology arose. At approximately the same time, Congress enacted the Kefauver-Harris Amendments (modifying the FDC Act of 1938) requiring drug manufacturers to provide theFDA with data that drugs were both safe and effective prior to marketing. Using the informationgleaned from the workshops, the FDA launched the segmented three-phase testing paradigm, 27which remains the essence of the modern safety assessment guidelines for developmental andreproductive toxicity hazard assessment.The “fertility and general reproductive performance study,” formerly termed the Segment Istudy, was devised to identify effects following the pairing of treated males and females. Theexposure regime consists of a 10-week premating period for the males (encompassing an entirespermatogenic cycle) and a 2-week period (encompassing two to three estrous cycles) for females.The resultant offspring are evaluated for effects postnatally, through weaning. The study was alsointended to screen for male-mediated “dominant lethal” effects. The “teratology study” (SegmentII) was designed to reveal effects on morphogenesis, intrauterine growth, and intrauterine survivalcaused by prenatal exposures commencing after implantation and continuing through the period ofmajor organogenesis. The “perinatal and postnatal study” (Segment III), also known as the “preandpostnatal [PIP] study,” originally employed exposure from late gestation through weaning, Itwas intended to characterize effects on late fetal development, the process of parturition, alterationsin maternal behavior during lactation (nesting and nursing), and growth and salubrity of offspringduring postnatal development. Recent emphasis on juvenile toxicity has increased awareness thatadverse effects can be elicited during the late phase of gestation and early postnatal life despitetreatment periods that are short relative to most treatment regimens for reproductive toxicity studies.A major innovation to the 1966 guidelines included the addition of endpoints addressingfunctional toxicity. Wilson had described the four major manifestations of developmental toxicityas dysmorphogenesis, prenatal mortality, intrauterine growth retardation (IUGR), and functionaldeficit. 6 Although studies conducted under the 1966 guidelines provided experimental data for thefirst three aforementioned outcomes, functional toxicity to the CNS and reproductive impairmentfollowing in utero and postnatal exposures were not widely studied. The event responsible forchanging this thinking stemmed from latent effects of the widespread methylmercury poisoning atMinamata Bay in Japan during the late 1950s and early 1960s. The resultant horrific and devastatingneurobehavioral effects of this tragedy led to decades-long study of neurobehavioral teratogenicity,culminating in the execution and reporting of the NCTR Collaborative Behavioral Teratology Study(CBTS). 37 The CBTS, published in 1985, proved to be instrumental in formalizing neurobehavioralassessments in the United States and directly influenced changes in the European Union (EU) andJapanese Ministry of Health, Labour and Welfare (MHLW) guidelines.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 347Considerable country-to-country variation in the principles of study design plagued productdevelopment efforts globally during the 1980s and early 1990s. Pharmaceutical development programsutilized the 1966 guidelines to permit marketing in the United States. However, additionalstudies containing design modifications addressing the concern regarding functional toxicity weregenerally required prior to marketing a drug internationally.Under the auspices of the International Conference on Harmonisation (ICH) in 1993, a memorandumof understanding standardized requirements across a broader range of regulatory agencies.Many of the concepts of modern study design emerged from this effort. Major innovations includedthe encouragement of internal dose determination to better quantify exposures across species,studies targeting effects on fertility and early embryonic development, and expansion of the treatmentregimen in the study of pre- and postnatal development to better address functional toxicity.In addition to providing guidance regarding study designs, the FDA has recently published adraft guidance document for integrated assessment of developmental and reproductive toxicity studyresults. 5 This document describes a suggested process for estimating the human developmental andreproductive risks resulting from drug exposure when only animal data are available.2. Agricultural and Industrial ChemicalsShortly after the formation of the EPA in 1970 and in response to the growing number of chemicalsentering commerce, the Office of Pesticide Programs proposed guidelines for registering pesticidesin the United States. 38 These guidelines were codified (CFR Vol. 43, No. 163) and includedprovisions for teratogenicity (163.83-3) and reproduction (163.83-4) studies.Under the 1978 guidelines, EPA’s pesticide evaluation was a modification of the FDA’s historicalthree-generation approach to pesticides and food colorants. Representing a significant departurefrom classic FDA paradigms, the design was reduced to an assessment of two generations producingsingle litters in each generation. The EPA reasoned that the litters produced from the first matingcycle in each generation yielded highly variable results because of the relatively immature statusof the mothers. To mitigate this confounding variable, the agency adopted the single litter pergeneration approach and delayed the onset of the mating cycle until the maternal animals were ofsufficient age to produce uniform and stable litters. A rationale for eliminating the third generationwas the introduction of the first requirements for mutagenicity studies in the 1978 proposedguidelines. The assumption was that a battery of mutagenicity studies was collectively moresensitive than a three-generation study for identifying genetic damage to germ cells. Another reasonfor reducing the number of generations was the agency’s experience that for persistent chemicals,cumulative toxicity was rarely expressed for the first time in the third generation.Requirements for the reproductive toxicity test were fundamentally unchanged when the guidelineswere fully enacted in 1982. However, significant changes were introduced when the Officeof Prevention, Pesticides and Toxic Substances (OPPTS) 870.3800 guidelines were promulgatedin 1998. Differences in the 1998 guidelines arose from concerns over endocrine-active compounds.These concerns led to addition of new endpoints, such as landmarks of sexual development,semenology, estrous cyclicity, ovarian follicle quantification, and weanling organ weights, toimprove overall sensitivity.For studies designed to evaluate the effects of prenatal exposure on fetal outcome (teratogenicitystudies and prenatal developmental toxicity studies), the concepts of study design have beenremarkably constant since the advent of the 1978 guidelines proposed by the Federal Insecticide,Fungicide and Rodenticide Act (FIFRA). The major developments in content of those guidelinesinvolve extending maternal exposure during the late fetal period, strengthening statistical power(group size) in the second species (rabbit), and requiring a more balanced evaluation of the productsof conception for visceral and skeletal abnormalities in the rodent. The requirement for exposure© 2006 by Taylor & Francis Group, LLC

348 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONduring the late fetal period appeared initially in the 1978 guideline, was later deemphasized in the1982 FIFRA and 1984 Toxic Substances Control Act (TSCA) guidelines, and was reintroducedunder the OPPTS 870.3800 guidelines in 1998.In addition to the guidelines for conduct of developmental and reproductive toxicity studies,the EPA has provided guidance documents to aid in interpretation and risk assessment of the dataderived from these studies. 39,40 These risk assessment guidelines represent the best effort thus farat codifying and documenting a thorough approach to risk assessment for the discipline.B. SpeciesIdeally, the principal factor in selecting the test species for developmental and reproductive toxicitystudies would be that they respond to toxicity in the same manner as humans. That is not possible,so historically, animal models were not selected for any reasons other than size, availability,economics, fecundity, and (probably) not appearing too anthropomorphic. The rat, mouse, andrabbit have fortunately proven to be acceptable surrogates, and the advent of higher-quality PKand TK studies has improved the utility of these models. On a scientific basis, and assuming theaforementioned, several key attributes of these models include the following: (1) basic physiologyand anatomy that are known or well-studied, (2) similar PK and TK profiles with humans, (3)comparable PD, and (4) in the absence of such information, apparently sufficient similarities inanatomical structure and reproductive physiology to permit comparison with humans. The prevailingassumption is that mammalian systems are most appropriate. The ideal test system would be easilymaintained in a laboratory environment, possess significant fecundity, display stable reproductiveindices, be polytocous, have a relatively low historical incidence of spontaneous structural malformations,and have other characteristics suitable for evaluating significant numbers of chemicalentities cost effectively. The rat is often preferred because the endocrinology and reproductivephysiology of this species have been thoroughly studied. In addition, general pharmacology modelshave been fully validated in the rat. Typically, animal models should be nulliparous becauseconfirmation of pregnancy in previously gravid females is frequently problematic (because ofimplantation scars). However, in the case of the rabbit and dog, use of proven breeders can lend stabilityto the reproductive indices. For most product development efforts, the rat or mouse is selected initially,but another relevant and presumably susceptible species is typically required. Alternatives include therabbit, guinea pig, hamster, dog, and nonhuman primate. Perceived advantages and disadvantages ofeach model are listed in Table 9.11 (enhanced from the ICH testing guidelines).In assessment of developmental toxicity data, the investigator must be cognizant of a keydifference in reproductive physiology among rodents (particularly rats), rabbits, and humans.Prolactin mediates maintenance of early pregnancy in the rat, unlike the rabbit and human. In thecase of the rabbit, progesterone secreted from the corpus luteum sustains maintenance of pregnancy.The fundamental control in the human and rabbit changes with advancing pregnancy to control viathe fetal-placental unit. This feature of endocrinology may predispose the rabbit to a higher riskof postimplantation loss and abortion when only a few implantations and limited hormonal signalingexist to maintain luteal function. This concept is illustrated by the data presented in Table 9.12. Of60 control rabbits evaluated at laparohysterectomy that had a single implantation site, 26 (43.3%)had completely resorbed litters while 6 additional rabbits (10.0%) terminated pregnancy prematurelyvia abortion. The remaining 28 animals had normal reproductive outcomes. Animals with twoimplants also were at high risk for abnormal reproductive outcomes. There was less associationbetween number of implantation sites and abnormal reproductive outcome in cases where theintrauterine contents consisted of three to five implants. Rabbits with greater than five implantsusually have no difficulty maintaining pregnancy to term. Therefore, when interpreting the significanceof abortion rates and total litter loss in compound-treated rabbits, the data must be evaluatedin light of the number of embryos that were implanted in the uterus following fertilization.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 349Table 9.11Comparison of attributes of species used in developmental and reproductivetoxicity studiesModel Perceived Advantages Perceived DisadvantagesRatStudies of kinetics possibleLarge and well-characterized historicaldatabaseUsed extensively in repeated-dose toxicitystudiesPolytocousHigh fertilityProcesses of biotransformation, hormonalcascades, and organ system ontogeny wellunderstoodShort gestation periodCost-effectiveSusceptible to dopamine agonists (dependenceon prolactin for maintenance of early pregnancy)Prone to premature reproductive senescencefollowing treatment with GABAnergic and otherCNS-active agentsIncreased susceptibility to Leydig cell tumorsIncreased susceptibility to mammary tumorsInverted yolk sac placentation aLimited fetal periodMouseHamsterGuinea pigRabbitStudies of kinetics possibleRequire less test articlePolytocousProcesses of biotransformation, hormonalcascades, and organ system ontogeny wellunderstoodHigh fertilityInbred, knock-out and transgenic modelsavailableCost-effectiveAlternative rodent modelPolytocousRequire less test articleHigh fertilityVery short gestation periodCost-effectiveCommonly used in biologic researchAlternate rodent modelCost-effectiveTrue fetal phaseStudies of kinetics possiblePolytocousSignificant fetal periodLarge fetus is ideal for visceralmicrodissection procedureSemen can be obtained for longitudinalassessmentCost-effectiveHigh basal metabolic rateVisceral microdissection procedure hampered bysmall size of fetusCompressed developmental windows; prone tomalformation clustersLimited blood sampling sitesInverted yolk sac placentation aLimited fetal periodNot routinely used in repeated-dose toxicitystudiesVisceral microdissection procedure hampered bysmall size of fetusCheek pouches (can conceal dosage)Inverted yolk sac placentation for all of gestationCannibalismFetal period occurs extrauterinally in partPoor depth of historical databaseLimited blood sampling sitesSmall litter sizeLonger gestation periodSensitive to local gastrointestinal disturbances(e.g., antibiotics)Limited blood sampling sitesNot routinely used in repeated-dose toxicitystudiesPoor depth of historical databaseLower fertilityConsume diet inconsistentlyProne to abortion and toxemiaInduced ovulatorSensitive to local gastrointestinal disturbances(e.g., antibiotics)Not routinely used in repeated-dose toxicitystudiesProne to resorption when few implantations arepresentInverted yolk sac placentation aFerret Nonrodent modelSignificant fetal periodNot routinely used in repeated-dose toxicitystudiesPoor depth of historical databaseSeasonal breedersExpensive© 2006 by Taylor & Francis Group, LLC

350 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.11Comparison of attributes of species used in developmental and reproductivetoxicity studies (continued)Model Perceived Advantages Perceived DisadvantagesDogStudies of kinetics possibleNonrodent modelOffspring number sufficientRoutinely used in repeated-dose toxicitystudiesSemen can be obtained for longitudinalassessmentRadiographs possible to study fetalmorphologyPlacentation more like humans(no inverted visceral yolk sac)Significant fetal periodSeasonal breedersPoor depth of historical databaseHigh inbreeding coefficientExpense may lead to small group sizesLimited history of use with some validation studiesNonhumanprimateSize and multiple sampling sitesStudies of kinetics possibleNonrodent modelTrue fetal phaseRoutinely used in repeated-dose toxicitystudiesPhylogenetically close to humansReproductive physiology more similar to thatof humansExcellent model to assess biologicsSemen can be obtained for longitudinalassessmentRadiographs possible to study fetalmorphologyPlacentation similar to humans(no inverted visceral yolk sac)Kinetics often different than in humansGenerally single offspring per pregnancyLong gestation periodProne to abortionPoor depth of historical databaseStudies of fertility and functional reproductiveoutcome problematicInterpretation dependent on adequate historicalcontrol data due to cost factors usually leadingto small group sizesExpense and limited availability may lead to smallgroup sizesSome species are seasonal breeders(e.g., Rhesus)aListed as a negative attribute because it is anatomically different than in higher orders of mammals; however,probably should only be considered a negative, that is, producing nonconcordant effects, for large, proteinaceousmolecules or other agents that affect proteolysis and/or pinocytosis.Source: Holson, J.F., Pearce, L.B., and Stump, D.G., Birth Defects Res. (Part B), 68, 249, 2003.C. Characterization of Dose-Response CurveBroadly defined, the dose response of a particular chemical is the dynamic relationship betweenexposure and biological response(s) in the test system. In toxicity studies, plotting the effect(abscissae) versus a series of doses (ordinates) describes the dose-response relationship of anadverse finding for a population of animals. This model assumes that all members of the populationare either positive responders or nonresponders, and the relationship is defined as a quantal doseresponserelationship. 41 Dose-response curves are typically classified into one of three generalmodels: (1) threshold model, (2) linear model, or (3) U-shaped model. The threshold model foradverse responses assumes that there is a dose at which no effect is produced. The linear modelassumes that some effect occurs at any dose. With increasing dose, the shape of the dose-responsecurve will change until either a maximal response or maternal mortality occurs. For cancer endpoints,the linear model is typically assumed and applied. For noncancer endpoints, includingdevelopmental and reproductive toxicity, the threshold model is typically used. U-shaped modelshave been described for important nutritional substances required for homeostasis and for certainhormones. 42 For example, at low doses associated with dietary deficiency, adverse effects maymanifest. However, as the dose increases through the essential range, the adverse response isdiminished or ablated. As the dose is increased further, an adverse response manifests and increaseswith increasing dose similar to that described for the threshold and linear models.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 351Table 9.12Modal distribution of spontaneous litter complications in control Hra:(NZW)SPF rabbitsDistributionofImplantation SitesFemales withCompleteLitter ResorptionFemalesthatAbortedTotal AbnormalReproductiveOutcomesNo. Sites No. Animals No. % No. % No. %1 60 26 43.3 6 10.0 32 53.32 60 12 20.0 2 3.3 14 23.33 85 1 1.2 3 3.5 4 4.74 118 2 1.7 0 0.0 2 1.75 130 1 0.8 1 0.8 2 1.56 188 0 0.0 0 0.0 0 0.07 237 0 0.0 2 0.8 2 0.88 246 1 0.4 1 0.4 2 0.89 238 1 0.4 2 0.8 3 1.310 134 0 0.0 1 0.7 1 0.711 87 0 0.0 2 2.3 2 2.312 48 0 0.0 1 2.1 1 2.113 9 0 0.0 1 11.1 1 11.114 6 0 0.0 0 0.0 0 0.015 0 0 0.0 0 0.0 0 0.016 1 0 0.0 0 0.0 0 0.0Data tabulated from 84 definitive developmental toxicity studies (1647 females) conducted at WIL ResearchLaboratories, Inc. (1992 to 2003). Average historical litter size = 6.5 viable fetuses.In application, investigators typically attempt to use at least three graded dose levels to characterizea range of doses in toxicity studies that attempt to identify a no-effect level (NOEL), athreshold dose, and the maximum tolerated dose (MTD). Because the slope of the dose-responsecurve is often steeper in developmental toxicity studies than in other toxicity studies, large incrementsbetween dose levels should be avoided after achievement of the threshold. When a deleteriousresponse increases in frequency with ascending dose, the dose-response relationship may be thecritical factor in discriminating between effects that are treatment related and those that arise frombiological variability. Because the dose-response curve characterizes the potency of an agent, itprovides a common means to compare and contrast the attributes of various compounds relativeto margin of safety, therapeutic index, potency, and efficacy.One example of research aimed at investigating this important concept (i.e., presence or absenceof threshold in developmental toxicity) is the work of Shuey et al. with the antineoplastic agent,5-fluorouracil (5-FU). 43 Although this research effort was thorough and well conceived, it may notbe directly applicable to all developmental toxicity scenarios. The agent, 5-FU, was selected forstudy because it is a teratogen in humans and animals, and the mode of action was known (inhibitionof the enzyme thymidylate synthase [TS] that blocks DNA synthesis and cell proliferation). Theprocess is saturable, and a threshold can be demonstrated by blocking the active sites for thisenzyme. The method for analysis of TS in embryonic tissues was sufficiently sensitive to detectchanges well below those required to induce IUGR and structural malformations. The investigatorsdemonstrated that inhibition of TS in the fetus was measurable in the absence of associated changesin morphology or growth. Hence, a threshold model for 5-FU developmental toxicity was purported.However, the extent to which this phenomenon applies broadly to xenobiotics is not known.Additionally, whether the absence of morphologic effects in the presence of inhibition of TSconstitutes reserve or repair remains unclear.Many scientists believe mammalian embryos are capable of a high rate of repair. However, thisremains largely unquantified and speculative. The subject has not been studied to the extent thatany quantifiable correlation to a toxic insult can be made. Additionally, belief in the direct role ofmaternal homeostatic mechanisms in embryonal protection is based on knowledge of a limitednumber of endogenous elements and compounds that are preferentially taken up by the developing© 2006 by Taylor & Francis Group, LLC

352 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONconceptus at the mother’s expense (e.g., calcium). However, the role of these homeostatic mechanismsin protecting the conceptus at the time of insult by a xenobiotic is neither well studied norunderstood, largely because of the limited assessment of maternal biochemistry and physiology indevelopmental and reproductive toxicity studies. Furthermore, given that the timing of developmentis so critical (e.g., movement of palatal shelves in concert with head growth), it would seem thatrepair mechanisms would have to be exquisite and rapid to protect against xenobiotic disruptionof key temporal relationships.Many homeostatic mechanisms, including active, facilitated, and pinocytotic transport processes,assure supply of essential nutrients to developing tissues at the expense of maternal levels.Although such protective mechanisms exist, there is ample evidence to indicate that severe perturbationsin maternal homeostasis may have adverse effects on development. Nevertheless, it is oftensurprising that extensive effects on maternal physiology may occur in the absence of adversedevelopmental effects. Refer to Chapter 4 in this text and Section V.B.1 of this chapter for furtherdiscussion of the role of maternal toxicity in developmental outcome.D. Dose Selection and Maximum Tolerated DoseA fundamental tenet of well-designed toxicity studies is the characterization of toxicity at an MTD.Regulatory agencies, in general, require the highest dose in a hazard identification study to characterizetoxicologic responses at the MTD. In practice, however, there is little agreement regardingthe working definition of an MTD and its value and toxicologic relevance. In some instances, adose that results in decrements in maternal or parental body weight gain of 5% to 10% (withoutinducing mortality) is of sufficient magnitude to define toxic effects at the upper range of the doseresponsecurve. Additional information such as TK data, 44,45 manifestations of developmentaltoxicity, and/or indications of reproductive dysfunction or failure may be used as evidence of anachieved MTD. In a typical dose-response study, a minimum of two additional treatment groupsutilizing appropriate fractions of the MTD are evaluated, along with a control group. However, amore rigorous approach using more than three dose groups has been suggested, 46 and appropriatelyconducted dose–range-finding studies are essential in characterizing the lower segment of the doseresponsecurve.The approach of using an MTD as a high-dose level in nonclinical safety studies is notuniversally accepted. The opposing viewpoint is founded principally on the premise that for humanexposure, whether direct (as in the case of pharmaceutical entities or food additives) or indirect(through consumption of residues, as in the case of pesticides), evaluation of effects at the MTDgrossly overpredicts risk. For example, inorganic arsenic injected intraperitoneally at high dosesresults in prenatal mortality and an array of structural fetal malformations in rodents, presumablythrough saturation of the methylation pathway in the liver. 47,48 Interestingly, doses administeredorally at levels far exceeding relevant environmental exposures posed insignificant risk to thedeveloping embryo, even when significant maternal toxicity occurred. 49,50 This example illustratesthat for trace exposures, parenteral routes of delivery may create an exaggerated concern, eventhough it is physically not possible to deliver enough inorganic arsenic transplacentally under theseconditions to elicit dysmorphogenesis. In this example, use of the MTD resulted in exaggerateddoses that overwhelmed absorptive and metabolic pathways that have evolved to provide detoxificationfor semicontinuous environmental exposure. Barring situations of acute intoxication, thereis little benefit to evaluating and committing resources to such an exposure scenario, especiallycompared with other routes of exposure more appropriate to humans. At the point when maternaltoxicity manifests, pathways of elimination and metabolism may change dramatically, and saturationof plasma protein binding may occur, resulting in exaggerated exposure and less precise extrapolationbetween animal models and the human. Therefore, knowledge gained from PK and PDstudies is essential in bridging this gap.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 353Prediction of hazard in large human populations from small-scale animal studies is difficultbecause of the limited statistical power of the latter. Regulatory decisions derived from animalstudies may be based on the results from 20 to 25 pregnancies per dose level, whereas humanexposure may involve thousands or millions. Studying toxicity at the MTD presumes maximizationof response, which ameliorates the difference in power between the two situations. Thus, bioassaysutilizing conventional numbers of sampling units and doses near the therapeutic dose may notindicate what would be the important toxicologic events in larger human populations. Likewise,for chemicals, use of the MTD in the absence of kinetic data indirectly, though crudely, establishesexposure and maximizes the ability to detect sensitive responders. Advocates of the MTD conceptalso contend that studying toxicity approaching the mortality portion of the dose-response curveis essential in the routine evaluation of pharmaceuticals to provide clinicians the types of adverseevents that they may encounter at relevant therapeutic doses in clinical trial patients. Clearly, anegative data set in a preclinical study diminishes the ability of the clinical investigator to rapidlyidentify and manage dose-limiting toxicities. Data generated at the MTD may assist in identificationof the symptomatology that is critical for assessment of human toxicity or overdose and for selectingantidotes or appropriate courses of therapy.E. General Statistical ConsiderationsStatistical considerations play an especially important role in the evaluation of developmental andreproductive toxicity data. Both continuous (e.g., fetal body weight) and binary (e.g., postimplantationloss) measures are produced by these studies, and proper interpretation requires considerationof the statistical power of the study and the litter effect.1. Statistical PowerStatistical power is the probability that a true effect will be detected if it occurs. It is formallydefined as 1β, where β is the probability of committing a Type II error (false negative). 51 Power isdependent on the sample size, background incidence, and variability of the endpoint in question,and the significance (α) level of the analysis method. For example, the sample size (number oflitters) needed to detect a 5% or 10% change in an endpoint is dramatically lower for a continuousmeasure with low variability, such as fetal body weight, than for a binary response with highvariability, such as embryolethality (resorptions). 52 Consequently, significant changes in fetal weightare often detected at dose levels lower than those at which effects on embryo or fetal survival areobserved. However, despite an adequate sample size, the large number of endpoints evaluated in adevelopmental or reproductive toxicity study dictates that several spurious statistically significantdifferences will likely occur because given a significance level of 0.05, a Type I error (false positive)will occur 5% of the time. For example, because a standard developmental toxicity study withANOVA 53 /Dunnett’s 54 and Kruskal-Wallis/Mann-Whitney 55 statistical analyses performed on allparametric and nonparametric data, respectively, may involve as many as 100 to 300 individualstatistical hypothesis tests, the possibility exists for numerous spurious statistical findings.A replicated or unbalanced study design may also augment the statistical power of a study. Forexample, a large-scale, replicated dose-response study of the herbicide 2,4,5-trichlorophenoxyaceticacid demonstrated that such a study design may assist in resolution of problems of interspeciesvariability determination, high- to low-dose response extrapolation, and reproducibility of low-leveleffects. 522. The Litter EffectThe litter must be considered the experimental unit in developmental and reproductive toxicitystudies because the litter is the unit that is randomized, and individual fetuses or pups within litters© 2006 by Taylor & Francis Group, LLC

354 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONdo not respond completely independently. 41,56 The propensity for fetuses of a given litter to exhibitsimilar responses to toxic insult, thereby artificially inflating the apparent group response, has beendesignated the “litter effect.” Mathematically, using the litter as the experimental unit to accountfor this influence requires a determination of the percentage of embryos or fetuses within eachlitter that are affected. A grand mean is then calculated from the individual litter means. Using thisapproach, the variance among litters (standard deviation and standard error) can also be calculated.While the litter proportion calculation is generally applied to embryo and fetal survival data,many investigators fail to use this approach to analyze fetal malformation data. These investigatorssimply determine a percentage of litters with at least one malformed fetus, failing to apply correctlitter-based statistics. With this approach, the number of malformed fetuses within each litter is nottaken into account (e.g., a litter with 1 of 12 fetuses malformed is given the same weighting as alitter with 6 of 12 fetuses malformed). Therefore, variance among litters cannot be determined.Initially determining the percentage of malformed fetuses within a litter, followed by calculatinga litter grand mean, is the most appropriate way to analyze fetal malformation data.Table 9.13 presents five comparative examples of calculations using the litter as the experimentalunit. These examples include the following endpoints: numbers of resorptions (prenatal death),malformations, and fetal weight. For each example, an explanation of the derivation of the pertinentendpoint is provided. These examples contrast the incorrect and correct litter-based statistics andclearly demonstrate the different means of computation. Surprisingly, even though considerableresearch has been conducted and literature published on this topic in the 1960s and 1970s, fewlaboratories, and even fewer commercial software programs, properly calculate these values.Depending on a particular data set, the correct and incorrect means may differ little, but variance(sum-of-mean-square-derived) may not be obtained at all using the improper calculations. Whenthe incorrect statistic is used, the N (number of fetuses) is exaggerated, increasing false positiveresults, and in the case of fetal weight, “within-litter variance components” are not determined orincorporated. The latter omission often fails to identify the within-litter response dimension (i.e.,increase in number of affected fetuses per litter), which becomes lost in the overall among-littervalue. Because of the incorrect statistics, litters with large or small numbers of fetuses will bedisproportionately weighted in the analysis. In addition, the analysis will not be able to assessclustering of effect in limited numbers of litters. The precise calculations, which obey the litterunit, were first applied in the late 1970s at the NCTR. 56,57 Use of the appropriate method ofcalculation can make significant differences in interpretations. This becomes of paramount importanceto postnatal data evaluations because litter influences persist well beyond weaning. 37IV. DOSE RANGE–CHARACTERIZATION STUDIESVS. SCREENING STUDIESDose range–characterization studies provide information to design a definitive study (one used forhazard identification in risk assessment). Under most circumstances, dose range–characterizationstudies do not eliminate a test article from development, although unexpected results can lead tothat decision. Conversely, toxicologic screening studies may be used expressly for selecting productcandidates. The following sections discuss the differences between dose range–characterization andscreening studies, and where the two types of studies may converge.A. Dose Range–Characterization StudiesDose range–characterization studies (also referred to as dose range–finding, preliminary, pilot, ordose-finding studies) are an essential component of a valid research program. These studies provideinvestigators with information necessary to properly select dose levels for definitive developmentaland reproductive toxicity studies. They represent a different level of interpretation and are pivotal© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 355Table 9.13Litter proportion calculation examplesExample 1: Resorptions (Prenatal Deaths)Resorptions Implantation Sites Postimplantation LossLitter (No.) (No.) (%)1 0 15 0%2 0 14 0%3 5 10 50%4 0 13 0%5 1 15 7%Incorrect StatisticsNumerator = 6 = Total no. of resorptionsDenominator = 67 = Total no. of implantationsIncidence = 9.0% = Total no. of resorptions/total no. of implantationsAmong-litter variability = — = IncalculableIn Example 1, the numbers of resorptions are summed and divided by the total number of implantation sites.This calculation [(5 + 1)/67*100 = 9.0%] provides a simple incidence, without regard to weighting of effectsby litter, as with correct litter-based statistics.Correct Litter-Based StatisticsNumerator = 57% = Sum of postimplantation loss (%) per litterDenominator = 5 = Total no. of littersTrue %PL = 11.3% = Sum of postimplantation loss (%) per litter/total no. of littersAmong-litter variability = 21.8% = Standard deviation of postimplantation lossUsing Example 1, correct litter-based statistics are calculated by summation of the percent per litterpostimplantation loss and division by the number of litters, the randomized, true experimental unit[(50% + 7%)/5 = 11.3%].Example 2: Resorptions (Prenatal Deaths)Resorptions Implantation Sites Postimplantation LossLitter (No.) (No.) (%)1 1 15 7%2 1 14 7%3 2 10 20%4 1 13 8%5 1 15 7%Incorrect StatisticsNumerator = 6 = Total no. of resorptionsDenominator = 67 = Total no. of implantationsIncidence = 9.0% = Total no. of resorptions/total no. of implantationsAmong-litter variability = — = IncalculableIn Example 2, the numbers of resorptions are summed and divided by the total number of implantation sites,as in Example 1. Note that while the distribution of resorptions has changed dramatically from Example 1, theoverall number of resorptions is the same, and the incorrect statistics do not change [(1 + 1 + 2 + 1 + 1)/67*100 =9.0%].Correct Litter-Based StatisticsNumerator = 48% = Sum of postimplantation loss (%) per litterDenominator = 5 = Total no. of littersTrue %PL = 9.6% = Sum of postimplantation loss (%) per litter/total no. of littersAmong-litter variability = 5.8% = Standard deviation of postimplantation loss© 2006 by Taylor & Francis Group, LLC

356 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.13Litter proportion calculation examples (continued)In Example 2, correct litter-based statistics are again calculated by summation of the percent per litterpostimplantation loss and division by the true experimental unit [(7% + 7% + 20% + 8% + 7%)/5 = 9.6%]. Notethat by use of the correct litter-based statistics, the change (from Example 1) in resorption distribution throughoutthe litters is reflected in both the percent per litter of resorptions and the among-litter variability (expressed hereas standard deviation).Example 3: MalformationsLitterNo. of FetusesMalformedTotal No.of FetusesPercentMalformed1 0 16 0%2 0 14 0%3 3 10 30%4 0 18 0%5 0 15 0%Incorrect Statistics 1Numerator = 3 = Total no. of fetuses malformedDenominator = 73 = Total no. of fetusesIncidence = 4.1% = Total no. of fetuses malformed/total No. of fetusesAmong-litter variability = — = IncalculableWith incorrect statistics 1 in Example 3, the numbers of malformed fetuses are summed and divided by the totalnumber of fetuses. This calculation [(3)/73*100 = 4.1%] provides a simple incidence, without regard to weightingof effects by litter, as with correct litter-based statistics.Incorrect Statistics 2Numerator = 1 = Total No. of litters with at least 1 malformed fetusDenominator = 5 = Total No. of littersFalse % = 20% = Total No. of litters with at least 1 malformed fetus/total No. of littersAmong-litter variability = — = IncalculableBy applying incorrect statistics 2 to Example 3, the number of litters containing at least one malformed fetus isdivided by the total number of litters [1/5*100 = 20%]. This approach may have some utility when comparingamong studies where one is evaluating dispersion among litters for a specific-type finding (i.e., resorptions, atype of malformation, or a type of variant). This should only be used as a supplement to true litter-based statistics.Correct Litter-Based StatisticsNumerator = 30% = Sum of percent malformed per litterDenominator = 5 = Total No. of littersTrue %PL = 6.0% = Sum of percent malformed per litter/total No. of littersAmong-litter variability = 13.4% = Standard deviation of percent malformed per litterUsing Example 3, correct litter-based statistics are calculated by summation of the percent malformed fetusesper litter and division by the number of litters, the randomized, true experimental unit [(30%/5 = 6.0%].Example 4: MalformationsLitterNo. of FetusesMalformedTotal No.of FetusesPercentMalformed1 0 16 0%2 0 14 0%3 1 10 10%4 0 18 0%5 2 15 13%© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 357Table 9.13Litter proportion calculation examples (continued)Incorrect Statistics 1Numerator = 3 = Total no. of fetuses malformedDenominator = 73 = Total no. of fetusesIncidence = 4.1% = Total no. of fetuses malformed/total no. of fetusesAmong-litter variability = — = IncalculableIn Example 4, the numbers of malformed fetuses are summed and divided by the total number of fetuses, asin Example 3. Note that while the distribution of malformed fetuses has changed from Example 3, the overallnumber of malformed fetuses is the same, and the incorrect statistics 1 do not change [(3)/73*100 = 4.1%].Incorrect Statistics 2Numerator = 2 = Total no. of litters with at least 1 malformed fetusDenominator = 5 = Total no. of littersFalse % = 40% = Total no. of litters with at least 1 malformed fetus/total no. of littersAmong-litter variability = —- = IncalculableIn Example 4, incorrect statistics 2 are again calculated by dividing the number of litters containing at least 1malformed fetus by the total number of litters [2/5*100 = 40%]. Note that by using incorrect statistics 2, thechange (from Example 3) in malformation distribution is exaggerated; it ignores all of the unaffected fetuses inevery litter, and it obligatorily produces a high incidence due to just spontaneous occurrences, potentially maskingreal treatment-related effects.Correct Litter-Based StatisticsNumerator = 23% = Sum of percent malformed per litterDenominator = 5 = Total No. of littersTrue %PL = 4.7% = Sum of percent malformed per litter/total No. of littersAmong-litter variability = 6.5% = Standard deviation of percent malformed per litterIn Example 4, correct litter-based statistics are again calculated by summation of the percent malformed fetusesper litter and division by the true experimental unit [(10% + 13%%/5 = 4.7%]. Note that by use of the correctlitter-based statistics, the change (from Example 3) in malformation distribution throughout the litters is reflectedin both the percent per litter of malformed fetuses and the among-litter variability (expressed here as standarddeviation).Example 5: Fetal weightLitterUterine Position a and Weight (g)Litter mean 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 181 3.6 3.8 3.6 3.5 3.7 3.4 3.8 3.6 3.5 3.7 3.4 3.8 3.52 2.8 2.9 2.7 3.0 3.0 2.9 2.4 3.1 2.8 3.0 2.7 2.4 2.6 2.5 2.9 3.0 2.8 2.5 2.43 3.6 3.8 3.7 3.4 3.8 3.6 3.5 3.7 3.4 3.3 3.7 3.7 3.5 3.7 3.44 3.6 3.8 3.8 3.7 3.6 3.5 3.7 3.3 3.7 3.4 3.8 3.8 3.6 3.75 4.2 4.1 4.3aUterine position begins at distal left uterine horn continuing through cervix and to distal right uterine horn.Incorrect StatisticsNumerator = 198.9 = Sum of all fetal weightsDenominator = 59 = Total no. of fetusesIncorrect group mean fetal weight (g) = 3.4 = Sum of all fetal weights/total no. of fetusesVariability = 0.46 = Standard deviation of individual fetal weightsIn Example 5, a continuous variable, fetal weight (g), is used to illustrate further the use of incorrect statistics.All individual fetal weights are summed and divided by the total number of fetuses in the group. By use of theseincorrect statistics, litters with higher numbers of fetuses will be inappropriately weighted in the calculation. Inaddition, this approach further confounds hypothesis testing by artificially reducing the variance around thecontinuous data endpoint through inflation of the degrees of freedom.© 2006 by Taylor & Francis Group, LLC

358 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.13Litter proportion calculation examples (continued)Correct Litter-Based StatisticsNumerator = 17.8 = Sum litter mean fetal weightsDenominator = 5.0 = Total no. of littersTrue group mean fetal weight (g) = 3.6 = Sum litter mean fetal weights/total no. of littersAmong-litter variability = 0.52 = Standard deviation of litter mean fetal weightsBy use of correct litter-based statistics [(3.6 + 2.8 + 3.6 + 3.6 + 4.2)/5 = 3.6 g], each litter is weighted similarly,and among-litter variability can be calculated. Example 5 demonstrates a 0.2-g difference in group mean fetalweight between the correct and incorrect statistics. Because a similar difference between a treated group anda concurrent control group of this relative magnitude in a continuous variable is often test article–related, theuse of incorrect statistics can confound data interpretation.to conducting an overall developmental and reproductive toxicity program. Interpretation of datacollected from dose range–finding studies will not differ appreciably from interpretation of thedefinitive study results, with two important exceptions. First, the number of dams used in pilotstudies is substantially smaller (usually 5 to 10 animals, versus 22 to 25 animals per group in thedefinitive study). Lower statistical power, because of the use of fewer animals, will potentiallyconfound data interpretation, especially when pregnancy rates are less than 100%. Standard statisticalanalyses may be performed for a pilot data set; however, they must be interpreted with greatercaution because of the reduced statistical power afforded by the low numbers of animals per group.The lower statistical power in these studies can increase both Type I and Type II errors. In addition,historical control data derived from definitive studies must be applied to dose range–characterizationstudies with caution. With diminished statistical power, the distribution of values around the centraltendency will be much larger for the preliminary study. Furthermore, not all endpoints are evaluatedin a range-finding study, making it unlikely to elicit strong conclusions from the data. Nevertheless,these studies are critical to conducting a successful definitive study. Specifically, they allow theinvestigator to develop a preliminary estimate of the threshold of response and the MTD, and theymay provide early evidence of developmental or reproductive toxicity.B. Developmental and Reproductive Toxicity Screening StudiesToxicity screens are simplified studies or models designed to identify agents having a certain setof characteristics that will either exclude the agents from further investigation (e.g., drug candidates)or cause them to be assigned for further, more rigorous investigations (e.g., industrial and agriculturalchemicals).With respect to drug evaluation, screening studies can provide a number of advantages overperforming guideline studies. These advantages include, but are not limited to, fewer animals andless test article required, more rapid execution, earlier attrition of candidate products, improvedpotency and selectivity evaluation, and economic savings. In the case of chemical testing, advantagesof employing a screening regimen may include resource conservation, quantitative structure-activityrelationships (QSARs) database creation and expansion, and increased ability to evaluate additionalchemicals.Pharmaceutical companies have historically used a number of developmental and reproductivetoxicity screens for determining whether to halt or continue test article development (e.g., Hershbergerand uterotrophic assays); however, the FDA has not published guidelines for screeningstudies. Conversely, the EPA and the Organisation for Economic Cooperation and Development(OECD) have promulgated a variety of in vitro and in vivo screening studies for evaluation ofdevelopmental and reproductive toxicity potential. Only the in vivo screens are discussed in thischapter; refer to Chapter 16 of this text and Brown et al. 58 for discussion of in vitro screens. The© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 359Table 9.14Guideline-driven developmental and reproductive toxicity screens, by agencyGuideline NumberType of Screen EPA/OPPTS OECDPreliminary developmental toxicity screen (Chernoff-Kavlock assay) 870.3500 —Reproduction and developmental toxicity screening test 870.3550 421Combined repeated dose toxicity study with the reproduction and developmentaltoxicity screening test870.3650 422Table 9.15Selected features of developmental and reproductive toxicity screening studiesOPPTS 870.3550/ OPPTS 870.3650/OPPTS 870.3500 OECD 421 OECD 422Species Rat or mouse Rat or mouse Rat or mouseGroup size 15 females 10 per sex 10 per sexTreatment regimen GD 7 to 11 or toclosure of hardpalateMale: 28 daysFemale: 14 daysPremating through LD a 3Male: 28 daysFemale: 14 daysPremating through LD 3Toxicokinetics None None NoneMating period Not applicable Two weeksTwo weeks(bred prior to startof dosing)Behavioral testing None None FOB b motor activityOrgan weights None Testes, epididymides Testes, epididymides, liver, kidneys,adrenals, thymus, spleen, brain,heartHistopathology None Testes, epididymides,accessory sex organs,ovariesTestes, epididymides, accessory sexorgans, ovaries, uterus,gastrointestinal tract, urinary tract,lungs, central & peripheral nervoussystem, liver, kidneys, adrenals,thymus, spleen, heart, bone marrowPup necropsy Dead pups only External exam only External exam onlyNote: Additional subchronic endpoints: hematology, blood biochemistryaLactational day.bFunctional observational battery.in vivo screening guidelines for chemical evaluation and their respective requirements are presentedin Table 9.14 and Table 9.15. Another useful screen is the Reproductive Assessment by ContinuousBreeding (RACB) screen used routinely by the National Toxicology Program. 59While screening studies will identify potent toxicants if appropriate endpoints are includedand/or are correlated, Type I and Type II statistical errors can also result from the reduced samplesizes employed. The specificity of a screen refers to its capacity to produce false positives (TypeI error), while the sensitivity of a screen speaks to its ability to produce false negatives (Type IIerrors). The authors’ experiences with the Reproduction/Developmental Toxicity Screening Test(OECD 421/OPPTS 870.3550) and Combined Repeated Dose Toxicity Study with the Reproduction/DevelopmentalToxicity Screening Test (OECD 422/OPPTS 870.3650) confirm that theseerrors are not infrequent. In particular, these OECD and OPPTS screens are useful if employed astrue screens. However, they are not useful if they are considered apical studies for hazard identificationwhen there is significant potential for human exposure. An expert panel of the GermanChemical Society’s Advisory Committee on Existing Chemicals has concluded that OECD 421and 422 screening tests are neither alternatives to definitive studies (i.e., those conducted underOECD guidelines 414, 415, and 416) nor replacements for these studies, particularly because theplanned validation of these screening tests has not been completed. 60© 2006 by Taylor & Francis Group, LLC

360 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.16Screens used as an antecedent to a two-generation reproductive toxicity studyScreen (S) Two-Generation (2-G) Concordance (%)Adult Endpoints F 0 F 0 F 1 2-G F 0 vs. S F 0 2-G F 1 vs. S F 0Mortality 2/9 1/9 3/8 89 75Body weight 5/9 8/9 7/8 67 63Food consumption 5/9 8/9 6/8 67 50Clinical observations 4/9 3/9 2/8 67 63Fertility index 0/9 0/9 0/8 100 100Mating index 0/9 0/9 0/8 100 100Implantation index 3/9 2/9 1/8 89 88Postimplantation loss 3/8 1/9 1/8 89 88Live birth index 3/8 1/8 1/8 63 63Organ weights 3/6 4/8 3/8 67 80Gestation length 0/8 0/8 0/8 100 100Dystocia 0/7 0/8 0/8 100 100Neonatal endpoints (F 1 ) (F 1 ) (F 2 ) 2-G F 1 vs. S F 1 2-G F 2 vs. S F 1Survival 2/8 1/8 2/8 63 75Body weight 3/8 5/8 5/8 75 75Note: For endpoint-finding enumeration, the denominator represents the number of test articles for which bothscreening and two-generation studies were performed. The numerator reflects the number of studies forwhich an effect was noted. For this table, five of the nine studies used extrapolated dose levels, that is;dose levels in these studies that were close enough to have considered them comparable for the observedeffects. Concordance reflects cumulative agreement (both positive and negative) between the screeningresults and the definitive results from each generation of the two-generation reproductive toxicity study,expressed as a percentage.Source: Data from studies conducted at WIL Research Laboratories, Inc.Table 9.17Examples of failures in sensitivity of screensScreen (S) Two-Generation (2-G) Concordance (%)Adult Endpoints F 0 F 0 F 1 2-G F 0 vs. S F 0 2-G F 1 vs. S F 0Fertility index 0/3 0/3 3/3 100 0Mating index 0/3 0/3 3/3 100 0Live birth index 1/3 1/3 2/3 100 67Organ weights 0/1 2/3 2/2 (33) (0)Gestation length 0/3 0/3 1/3 100 67Dystocia 0/3 1/3 1/3 67 67Neonatal endpoints (F 1 ) (F 1 ) (F 2 ) 2-G F 1 vs. S F 1 2-G F 2 vs. S F 1Sex ratio 0/3 1/3 2/3 67 33Hypospadias Not measured 1/3 Not measured (0) (0)Note: The denominator represents the number of test articles for which both screening and full two-generationstudies were performed. The numerator reflects the number of studies for which an effect was noted.For this table, dose levels were identical between the screening studies and the definitive two-generationreproductive toxicity studies. Concordance reflects cumulative agreement (both positive and negative)between the screening results and the definitive results from each generation of the two-generationreproductive toxicity study, expressed as a percentage.Employing the guideline-minimum number of animals per group in these two general designscan often result in effects that appear to be nondose-responsive and/or inconclusive when facedwith rare (low incidence) events. Table 9.16 presents selected endpoint-by-endpoint results forscreens performed at the authors’ laboratory when used as an antecedent to guideline two-generationstudies (OECD 416/OPPTS 870.3800). Table 9.17 presents data from three additional case studiesin which multiple failures of sensitivity (false negatives) occurred.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 361From the data presented in Table 9.16 and Table 9.17, the sensitivity and specificity limitationsintrinsic to the referenced screens are apparent. Strong conclusions regarding developmental andreproductive hazards can be drawn only from definitive studies containing adequate sample sizes,multiple endpoints, appropriate treatment duration, and ideally, some assessment of internal exposure(TK).The screening studies examined above are short-term experiments used to select and/or sort aseries of molecules on the basis of developmental and reproductive toxicity potential. These studiesmay be used to predict potential hazard, but the results generally cannot be used for risk assessment.Heuristically, two axioms have arisen: (1) if the primary intent of the screen is to reduce the numberof animals used, careful consideration must be given to the possible necessitation of subsequentstudies because of poor characterization of the dose-response curve, and (2) if the intent of thescreen is to reduce resource consumption, an analogy to the first point exists with the exception ofthe application to agents not developed for biologic activity, with limited human exposure andeconomic significance.C. Converging DesignsDose range-characterization studies can converge with screening studies for selected purposes. Thisconvergence often occurs in development of drugs in particular pharmacologic classes. An examplein the authors’ laboratory is a hybridized dose range–characterization and definitive embryo andfetal development study that was employed to screen retinoic acid analogs for teratogenic potency.The hybridized design utilized more animals than are typically employed for dose range-characterizationbut fewer than would be necessary for a definitive study. Generally, several candidateanalogs were evaluated concomitantly at high dose levels with one concurrent control group andone comparator group treated with a known potent analog (all-trans-retinoic acid). Teratogenicpotency was determined through a postmortem examination of greater scope than is regularlyemployed for a dose range-characterization study, as it consisted of both visceral and skeletalcomponents. Those analogs with the least expression of the classic terata were identified as potentialcandidates for further development. This type of specialized screening study can be adapted forother pharmacologic classes possessing known teratogenic potential. This, however, is only possiblewhen the investigator is able to correlate the selected study endpoints with a known overall patternof developmental effects. Furthermore, since teratogenic responses to certain classes of compoundsare well known and expected, an abbreviated treatment regimen could be employed, targeting acritical, susceptible period of organogenesis.V. DEVELOPMENTAL TOXICITY STUDIESThis section begins with a presentation of the key differences between various regulatory guidelinerequirements for developmental toxicity studies, including a brief discussion on the importance ofanimal models as well as the timing and duration of treatment. The interrelatedness of endpointsis discussed relative to both development (or delay thereof) and to the complexities of unravelingany interaction between developmental and maternal toxicity. A discussion of specific endpointsfrom these studies, within the context of the various interrelationships that may occur, is presentedwith historical control data to aid in determining toxicologic relevance. An in-depth evaluation ofthe challenges inherent in interpretation of fetal morphology data, including an exposition ofexperimental considerations relative to these critical examinations, is then presented. The sectionconcludes with case studies from the authors’ laboratory, illustrating the difficulties inherent ininterpreting the significance of several fetal developmental variations.© 2006 by Taylor & Francis Group, LLC

362 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. Guideline RequirementsTable 9.18 presents selected differences in study design requirements between various agencyguidelines. The OPPTS of the EPA has produced prenatal developmental toxicity testing guidelinesfor use in the testing of pesticides and toxic substances. This guidance document harmonized thepreviously separate testing requirements under TSCA and FIFRA. The OECD and the JapaneseMinistry for Agriculture, Forestry and Fisheries (MAFF) have developed guidelines very similarto those of the EPA. For medicinal products, the ICH developed technical requirements for theconduct of developmental toxicity studies in support of pharmaceutical products registration forthe European Union, Japan, and the United States. For food ingredients, the Center for Food Safety& Applied Nutrition (CFSAN) within the FDA has developed guidelines for developmental toxicitystudies. Finally, the Maternal Immunization Working Group of the Center for Biologics Evaluationand Research (CBER) at the FDA has developed a draft guidance document for developmentaltoxicity studies of vaccines.1. Animal ModelsRelevance of the mode of action of the test agent, when known, is the most important factor in thechoice of animal models for developmental toxicity testing, regardless of guideline requirements.Developmental toxicity studies are typically performed in two species, one rodent and one nonrodent.Historically, the rat and rabbit have been the preferred rodent and nonrodent species, respectively,for these types of studies. However, alternatives to these species exist (refer to Section III.Bof this chapter for further details).A recent publication has proposed the use of a tiered approach to developmental toxicity testingfor veterinary pharmaceutical products for food-producing animals. 61 Using this approach, if teratogenicitywere to be observed in the rodent, no testing in a second species would be required.However, if a negative or equivocal result for teratogenicity were to be observed in the rodent, thentesting in a second species, preferably the rabbit, would be conducted.In special cases, investigators may use the dog or nonhuman primate model in lieu of the rabbit.However, the cost of these models and the need for specialized techniques and experience to conducta successful study necessitate a well-articulated, scientific rationale justifying their use.An example of effective use of the canine model in developmental toxicology was a series ofstudies performed with a hemoglobin-based oxygen carrier (HBOC) intended to be an oxygenbridge in lieu of human blood transfusion. A preliminary developmental toxicity study in the ratrevealed significant embryo toxicity (lethality, growth retardation, and structural malformations inmultiple organ systems) following infusion of the test article on GD 9. 62 These effects coincidedwith the prechorioallantoic placental period of development when histiotrophic nutrition via theinverted visceral yolk sac placenta (InvYSP) is essential to development in rodents. The results ofthe preliminary study led to the hypothesis that the embryo toxicity in the rat was unique to thepresence of the InvYSP during the time the malformations were occurring, and that these findingsrepresented a false positive signal for human developmental toxicity. 63 Subsequent rat whole-embryoculture studies confirmed a pronounced effect on endocytotic and proteolytic functions in thevisceral yolk sac placenta. Because the canine model and humans do not possess an InvYSP, aseries of phased-exposure studies in the dog was conducted. The results of the canine studiesrevealed no evidence of developmental toxicity, substantiating the hypothesis. 622. Timing and Duration of TreatmentOf particular note in the comparison of developmental toxicity study guidelines is the differencebetween EPA and ICH requirements for the timing and duration of treatment. The EPA guidelinesrequire dose administration throughout the period of prenatal organogenesis (beginning at implantation© 2006 by Taylor & Francis Group, LLC

Table 9.18Selected comparison points between various regulatory agency guideline requirements for developmental toxicity studiesFDA/Redbook 2000 FDA/Vaccines a Japan/MAFFEPA (1998) ICH (1994) OECD (2001) (2000 Draft) (2000 Draft) (1985)Ref. 166 167 168 169 170 171Choice of animal model Rat and rabbit Rat and rabbit Rat and rabbit Rat and rabbit Vaccine should elicit Rat and rabbitimmune responseSingle or multiple Multiple Multiple Multiple Multiple Single MultiplespeciesRandomization Yes Yes Yes Yes Yes Not mentionedGroup SizeRodentsApprox. 20 females withimplants at necropsy16–20 litters Approx. 20 females withimplants at necropsyApprox. 20 females withimplants at necropsy16–20 litters per phase At least 20 pregnantfemalesNonrodentsApprox. 20 females withimplants at necropsy16–20 litters Approx. 20 females withimplants at necropsyApprox. 20 females withimplants at necropsy16–20 litters per phase At least 12 pregnantfemalesTiming and duration oftreatment/ExposurePeriod of majororganogenesis bImplantation to closure ofhard palatePrior to mating andimplantation to birthPeriod of majororganogenesisMorphologic examinationof fetuses cRodentsNonrodentsApprox. 1/2 of each litterfor skeletal alterations(preferably bone andcartilage)Remaining fetuses fromeach litter for soft tissuealterations (acceptableto examine all fetusesviscerally followed byskeletal examination)All fetuses for both softtissue and skeletalanomalies (capita of 1/2removed and evaluatedinternally, thusprecluding skeletalexamination of thosecapita)1/2 of each litter forskeletal alterationsRemaining fetuses fromeach litter for soft tissuealterationsEach fetus for soft tissue(by dissection) andskeletal anomaliesImplantation to day priorto scheduledlaparohysterectomy bApprox. 1/2 of each litterfor skeletal alterationsRemaining fetuses fromeach litter for soft tissuealterationsAll fetuses for both softtissue and skeletalanomalies (capita of 1/2removed and evaluatedinternally, thusprecluding skeletalexamination of thosecapita)Implantation to day priorto scheduledlaparohysterectomy bApprox. 1/2 of each litterfor skeletal alterationsRemaining fetuses fromeach litter for soft tissuealterations (acceptableto examine all fetusesviscerally followed byskeletal examination)All fetuses for both softtissue and skeletalanomalies (capita of 1/2removed and evaluatedinternally, thusprecluding skeletalexamination of thosecapita)1/2 of each litter forskeletal alterationsRemaining fetuses fromeach litter for soft tissuealterationsEach fetus for soft tissue(by dissection) andskeletal anomalies1/3 to 1/2 of each litter forskeletal alterationsRemaining fetuses fromeach litter for soft tissuealterationsEach fetus for soft tissue(by dissection) andskeletal anomaliesaThe FDA guidelines for developmental toxicity studies of vaccines also include a requirement for a follow-up phase from birth to weaning for evaluation of effects on preweaningdevelopment and growth, survival, developmental landmarks, and functional maturation.bIf preliminary studies do not indicate a high potential for preimplantation loss, treatment in the de nitive study may include the entire period of gestation (typically gestational days 0–20for rats and 0–29 for rabbits).cICH guidelines do not require examination of the low- and middose fetuses for soft tissue and skeletal alterations where evaluation of the control and high-dose animals did not revealany relevant ndings .DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 363© 2006 by Taylor & Francis Group, LLC

364 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONor alternatively at fertilization), whereas ICH guidelines require administration only during theperiod of major organogenesis (implantation through closure of the hard palate). This may appearto be a deficiency in the ICH guidelines, although the design was likely predicated on the presumptionthat pre- and postnatal studies would also be conducted, establishing exposure during the lategestational period. However, under regulatory regimens that do not include pre- and postnatalstudies, as in the case of biologics, the absence of late gestational treatment may result in theinability to detect adverse effects on functional maturational changes and osteogenesis that occurduring this period.An understanding of the toxicokinetics of the test agent may also require broadening the durationof treatment to include the period prior to implantation. In studying the effects of agents that requireprotracted administration to achieve steady state in the maternal organism, it is advisable to begindosing in advance of the processes under evaluation. An example of this understanding was reflectedin a series of developmental toxicity studies of inorganic arsenic, in which freely circulating arseniclevels were maximized by administration prior to fertilization to compensate for sequestration ofthe test agent in erythrocytes. In one of these studies, females received the test substance for 14days prior to mating, throughout mating, and continuing until GD 19, so that circulating levels ofarsenic in the dams were high enough to ensure transplacental delivery to the conceptuses. 49 Thetiming and approach to these issues is compound specific and based on kinetics, pharmacodynamics,or other exposure information.Alternatively, factors such as induction of detoxification enzymes may make beginning doseadministration at implantation, rather than during the preimplantation phase, more appropriate todetect significant effects on early differentiation.B. Interpretation of Developmental Toxicity Study EndpointsTable 9.19 presents the endpoints typically evaluated in developmental toxicity studies, listed firstin approximate chronological order of collection, and then ranked by approximate sensitivity ofthe endpoints. In this ranking, endpoints considered most sensitive are those that are most oftenaffected in general or by lower doses of xenobiotics.These endpoints may be either continuous or binary measures. Fetal body weight, one of thefew endpoints that is continuous, is also the most sensitive because it is gravimetrically determined.Given good laboratory techniques, fetal body weight should exhibit the least variance because ofthe lack of subjectivity inherent in its collection. Because sensitivity of an endpoint may bemethodologically or biologically dependent, the ranking in Table 9.19 is relatively arbitrary andshould not be considered a definitive categorization. Mode of action may determine which measureis more sensitive, although in most cases of developmental toxicity established in humans andlaboratory models, fetal weight is the most sensitive and is often associated with arrays of developmentaldisruption. Here, the critical importance of examining patterns of effect on multipleendpoints and reconciliation of the biological plausibility of the overall effect cannot be overemphasized.Patterns of effect on multiple endpoints, arising often from depression of fetal body weight,are the key to understanding the arrays of developmental disruption associated with toxic insult toin utero progeny. Because events such as body cavity development (e.g., gut rotation and retraction)and midline closures (neural tube, hard palate, spine, thorax, and abdomen) proceed so systematicallyand harmoniously, retardation of growth may simply shift the timing of these discretedevelopmental events by a matter of hours or days in most species. For example, an insult thatmanifests as omphalocele or cleft palate may result from developmental delay due to intrauterinegrowth retardation rather than a direct effect on morphogenesis of those structures. This distinctionmay be important in risk assessment. Fraser 64 and Holson et al. 57 have presented examples involvingpalatal closure in mice. Furthermore, growth retardation may be associated with functional impairmentand/or death. Therefore, all components of a developmental toxicity data set must be evaluated© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 365Table 9.19Endpoints of developmental toxicity studiesApproximate ChronologicalOrder of CollectionMaternal survivalMaternal clinical observationsMaternal body weightMaternal body weight changeMaternal gravid uterine weightMaternal net body weightMaternal net body weight changeMaternal food consumptionMaternal necropsy findingsMaternal clinical pathology aMaternal organ weights aFetal viabilityCorpora luteaImplantation sitesPreimplantation lossPostimplantation lossEarly resorptionsLate resorptionsFetal weightsMaleFemaleCombinedFetal malformationsFetal developmental variationsRanked by Sensitivity(Most Sensitive to Least Sensitive)Fetal weightsMaleFemaleCombinedPostimplantation lossEarly resorptionsLate resorptionsFetal viabilityFetal malformationsFetal developmental variationsMaternal body weightMaternal body weight changeMaternal gravid uterine weightMaternal net body weightMaternal net body weight changeMaternal food consumptionMaternal survivalMaternal clinical observationsMaternal necropsy findingsMaternal clinical pathology aMaternal organ weights aCorpora luteaImplantation sitesPreimplantation lossaMaternal clinical pathology and organ weight data can provide extremelyuseful information for resolving issues of maternal toxicity, but these endpointsare not required by regulatory agencies for developmental toxicitystudies and therefore, unfortunately, typically are not collected.holistically and under the premise that selective effects on discrete endpoints, although they mayexist, are the exception to the rule of patterns of effect. Figure 9.6 illustrates the intrinsic interrelationshipsamong developmental toxicity endpoints.1. Maternal Toxicity and Its Interrelationship with Developmental ToxicityCharacterization of maternal toxicity is essential because of the nature of the developing organismwithin the maternal milieu. Progeny in utero are dependent upon the maternal animal for theirphysical environment, nutrients, oxygen, and metabolic waste disposal. The relationship of maternaltoxicity to embryo and fetal toxicity is subject to various interpretations; therefore, due diligencemust be given to the effects of the test agent on the dam, especially to determine whether thepregnant female is more sensitive to a given agent than is the conceptus. Embryo and fetal effectsthat occur in the presence of frank maternal toxicity sometimes, although perhaps not alwayslogically, elicit less concern from a human risk assessment perspective than those effects on theconceptus that occur at maternally nontoxic doses.In a developmental toxicity study, the dam and the products of conception coapt and areinterdependent in many ways. Thus, it is biologically plausible that maternal toxicity may affectdevelopment of the progeny in utero. However, if toxic effects manifest in both the maternal andfetal organisms, there is no way to determine the relationship of the findings without, at minimum,measurement of concomitant plasma levels of the test agent in both organisms. In an acute dosingregimen, the dam would be expected have a higher C max so fetal exposure, as judged from plasmalevels of an agent, will lag behind that of the mother in time, and often in magnitude. Therefore,fetal effects that occur in the presence of maternal toxicity cannot necessarily be attributed to the© 2006 by Taylor & Francis Group, LLC

366 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONClinicalfindingsStressDAMStressBody weight andfood consumptionUterusPlacentaCONCEPTUSGrowthDevelopmentaldelayMidline Closuredefects (e.g.,omphalocele)Cleft PalateNTDMorphogenesisSyndromes/spectrummultiple malformationsEarlyresorptionsDead termfetusesSurvivalLateresorptionsCatch-upgrowth andrepairDecreasedossificationExquisite effects(Single Organ System)Incompatiblewith lifePrenataldeathDevelopmental variationsFigure 9.6Depiction of intrinsic interrelationships among developmental toxicity endpoints.effects on the mother. It has long been known that additional information on clinical pathology,organ weights, and measures of pharmacodynamics would further clarify this relationship. Unfortunately,the regulatory impetus to gather such information has not been forthcoming. Because thisinvestment has not been made, the debate and uncertainty in this regard will continue.Study of the ontogeny of physiologic regulation has revealed that mammalian maternal organismsare exquisitely adapted to protect their developing organisms through homeostatic protectivemechanisms. 65–68 An understanding of this phenomenon will help prevent underestimation of developmentalsensitivity that can occur when fetal effects are evaluated only in relation to maternaltoxicity. In the final analysis, the debate over the relationship between maternal toxicity anddevelopmental toxicity becomes more an issue of risk than one of causality. After all, the mother’shealth is still important, regardless of whether the progeny are affected. However, from a riskmanagement perspective, maternal effects are not necessarily equivalent to adverse developmentaloutcomes. The reader is referred to Chapter 4 of this text for a more detailed discussion of maternallymediated developmental toxicity.2. Endpoints of Maternal ToxicityData interpretation for a developmental toxicity study begins with an evaluation of the basicmaternal data collected (survival, clinical observations, body weights, food consumption, andnecropsy findings). Robust assessment of maternal toxicity includes measurements of organ weights© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 367Table 9.20Maternal gestation body weight gain ranges, by species/strainMouse Rat RabbitEndpoint Crl:CD ® -1(ICR)BR Crl:CD ® (SD)IGS BR Hra:(NZW)SPFMean total body weight gain during gestation a (g) 24–29 122–171 434–820Mean net body weight gain during gestation a,b (g) 4–8 41–76 71–427Mean percentage of total body weight gain attributedto the products of conception75–82 55–67 47–93Note: The day of observation of evidence of mating (rodents) or artificial insemination (rabbits) was designatedgestational day 0. Approximate ages at GD 0 were 80-100 days, 12-15 weeks, and 5-7 months for mice, rats,and rabbits, respectively. Laparohysterectomies were conducted on GD 18 for mice, GD 20 for rats, and GD 29for rabbits.aGestational days 0–18 for mice, GD 0–20 for rats, and GD 0–29 for rabbits.bLean body mass of dam/doe.Source: Data tabulated from studies conducted at WIL Research Laboratories, Inc., including 10 mouse studies(1991 to 2001), 25 rat studies (1999 to 2003), and 25 rabbit studies (1998 to 2003).and clinical pathology evaluations. The onset, relationship to dose, and timing of clinical findingsof toxicity may be important factors to note, depending upon the effects observed in the developingembryos or fetuses. For example, an adverse effect on the gestating female that manifests clinicallyduring a critical developmental window could be implicated in the subsequent morphologic alterationof the offspring.In the absence of anatomic and clinical pathology evaluations and organ weight measurements,maternal body weight data (usually accompanied by food and/or water consumption data) providethe clearest measure of maternal toxicity in these studies. Thorough evaluation of maternal growthshould include assessment of body weight and body weight gain at minimum intervals of 3 to 4days throughout the treatment period, gravid uterine weight (weight of the uterus and contents),net body weight (the terminal body weight exclusive of the weight of the uterus and contents), andnet body weight change (the overall body weight change during gestation exclusive of the weightof the uterus and contents). Body weight deficits of 5% or greater that are sustained over a periodof several days are generally considered to be a signal of an adverse effect on maternal growth.Table 9.20 presents reference data from the authors’ laboratory for mean total and net body weightgain during gestation for the most commonly used species, as well as the mean percentages of totalweight gain during gestation that are due to the products of conception.Maternal food consumption should be evaluated and presented for the corresponding intervalsof body weight gain. Food consumption measurements in these studies are critical for monitoringof maternal homeostasis. Many studies have revealed an association between dietary restriction inmice, rats, and rabbits and adverse outcomes on the progeny during both gestation and lactation.Studies of dietary restriction have demonstrated that reductions in food consumption of as little as10% of the normal dietary intake (approximately 7 to 8, 20 to 25, and 150 to 200 g/d for mice,rats, and rabbits, respectively) may be associated with increased prenatal death, dysmorphogenesis,and/or growth retardation. 69–75 Therefore, reduced maternal food intake of 10% or greater in adevelopmental toxicity study may be an indication not only of maternal toxicity but also ofsecondary insult to the developing progeny. However, the extent to which dietary restriction mimicsmaternal inanition, failed weight gain, or weight loss due to compound-related toxicity is not known.Regulatory developmental toxicity studies are generally not designed to separate maternal anorexiceffects from other potential insults to the developing progeny.Despite the general correlation in various species between restricted maternal food consumptionor fasting and adverse developmental outcome, exceptions to this rule do exist. In a recent studyconducted at the authors’ laboratory, a proprietary compound that produced excessive maternaltoxicity when tested in Crl:CD ® (SD)IGS BR rats (extreme decrements in maternal food consumptionand body weight gain) at the high-dose level had no resultant (or concomitant) effect on thedeveloping progeny. At the high-dose level, mean food consumption over the treatment period (GD© 2006 by Taylor & Francis Group, LLC

368 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION6 to 18) was 14 g/animal/d, compared with the expected mean of 22 g/animal/d in the controlgroup. Mean net (corrected) body weight in the high-dose group was 20% lower than the controlgroup value. However, no adverse effects on fetal survival, weight, or morphology manifested atany dose level tested. This example illustrates the evolutionary principle of conservation of progeny,even at the expense of the mother, and also shows that frank maternal toxicity will not alwayscause adverse fetal outcome.3. Intrauterine Growth and SurvivalBasic gestational data collected for a developmental toxicity study at laparohysterectomy (mistakenlyreferred to as cesarean section) include parameters such as numbers of pregnant and nonpregnantfemales, corpora lutea, implantations, early and late resorptions, alive and dead fetuses,abortions, fetal body weight, and fetal sex ratio. From these data, the indices of pre- and postimplantationloss may be calculated on a proportional litter basis (refer to Section III.E.2 for a detaileddiscussion of the litter effect). Intrauterine parameters may provide important information withwhich to assess the relative concern of fetal morphologic findings. These endpoints are evaluatedwithin the context of the health of the dam because the well-being of the gestating mother mayinfluence fetal outcome.a. Preimplantation LossPreimplantation loss is determined by comparing the number of corpora lutea produced with thenumber of successful implantations, as indicated below (presented on a proportional litter basis).Summation per group (%) =Σ preimplantion loss/litter (%)Number litters/groupwherePreimplantation loss/litter (%) =(Number corpora lutea – Number implantation sites)/litterNumber corpora lutea/litter× 100In a study where females are treated prior to implantation, an increase in preimplantation lossmay indicate an adverse effect on gamete transport, fertilization, the zygote or blastula, and/or theprocess of implantation itself. If treatment occurs at or after actual nidation, an increase in preimplantationloss probably reflects typical biological variability. However, in either case if the datareflect an apparent dose-response relationship and/or are remarkably lower than the historical controlrange, further studies may be necessary to elucidate the mode of action and extent of the effect.Table 9.21 presents, by species, typical ranges for numbers of corpora lutea and implantation sites,as well as mean litter proportions of preimplantation loss.b. Postimplantation LossEffects on survival of the embryo/fetus are manifested by an increase in postimplantation loss.Postimplantation loss should be calculated on a proportional litter basis as indicated below.Summation per group (%) =Σ postimplantion loss/litter (%)Number of litters/group© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 369Table 9.21Historically observed means (ranges) for preimplantation parameters, by species and strainMouse Rat Rat RabbitEndpoint Crl:CD ® -1(ICR)BR Crl:CD ® (SD)BR Crl:CD ® (SD)IGS BR Hra:(NZW)SPFCorpora lutea (no./litter) 15.0 (13.1–17.3) 16.8 (14.4–20.5) 17.7 (16.0–19.4) 10.4 (7.9–13.6)Implantations (no./litter) 13.0 (11.2–14.3) 15.0 (11.5–18.3) 15.9 (14.6–17.5) 7.0 (5.1–9.3)Preimplantation loss (%/litter) 12.7 (5.7–19.2) 10.5 (2.8–25.4) 9.5 (4.0–15.7) 30.7 (6.6–52.0)The day of observation of evidence of mating (rodents) or artificial insemination (rabbits) was designatedgestational day 0. Laparohysterectomies were conducted on GD 18 for mice, GD 20 for rats, and GD 29 for rabbits.Source: Data tabulated from studies conducted at WIL Research Laboratories, Inc., including 20 Crl:CD ® -1(ICR)BR mouse studies (1984-2000), 158 Crl:CD ® (SD)BR rat studies (1982-1997), 61 Crl:CD ® (SD)IGS BR ratstudies (1998-2003), and 81 Hra:(NZW)SPF rabbit studies (1992-2003). These four databases contain informationderived from 460, 3585, 1452, and 1608 pregnant control dams/does, respectively.Table 9.22EndpointHistorically observed means (ranges) for postimplantation parameters by species and strainMouseCrl:CD ® -1(ICR)BRRatCrl:CD ® (SD)BRRatCrl:CD ® (SD)IGS BRRabbitHra:(NZW)SPFTotal postimplantation loss 7.4 (3.2–12.0) 5.7 (2.2–13.5) 4.4 (2.0–8.6) 9.1 (0.6–23.4)(%/litter)Early resorptions (%/litter) 5.7 (3.2–10.4) 5.6 (1.8–13.5) 4.4 (1.5–8.6) 7.9 (0.6–22.7)Late resorptions (%/litter) 1.4 (0.0–3.2) 0.1 (0.0–3.2) 0.1 (0.0–0.8) 1.2 (0.0–6.2)Dead fetuses 0.3 (0.0–1.2) 0.0 (0.0–1.4) 0.0 (0.0–0.0) 0.0 (0.0–0.6)Viable fetuses (%/litter) 92.6 (88.0–96.8) 94.3 (86.5–97.8) 95.6 (91.4–98.0) 90.9 (76.6–99.4)The day of observation of evidence of mating (rodents) or artificial insemination (rabbits) was designatedgestational day 0. Laparohysterectomies were conducted on GD 18 for mice, GD 20 for rats, and GD 29 for rabbits.Source: Data tabulated from studies conducted at WIL Research Laboratories, Inc., including 20 Crl:CD ® -1(ICR)BR mouse studies (1984 to 2000), 158 Crl:CD ® (SD)BR rat studies (1982 to 1997), 61 Crl:CD ® (SD)IGSBR rat studies (1998 to 2003), and 81 Hra:(NZW)SPF rabbit studies (1992 to 2003).wherePostimplantation loss/litter (%) =Number dead fetuses + resorptions (early and late)/litterNumber corpora lutea/litter×100In rodents and rabbits, postimplantation loss may manifest as early resorptions, late resorptions,or dead fetuses. In both rodents and rabbits, the dead conceptus undergoes gradual degradation,followed by maternal reabsorption, and is referred to as a resorption; in rabbits, it may be abortedinstead of being reabsorbed. Early or late resorptions are identified by the absence (early) or presence(late) of distinguishable features, such as the head or limbs. In guideline developmental toxicitystudies, it is not possible to determine from evaluation of the products of conception whetherintrauterine deaths were spontaneous in origin, the result of malformations, or the result of a directtoxic insult to the conceptus.Table 9.22 provides historical ranges for postimplantation loss in mice, rats, and rabbits. Asindicated in this table, there is typically an inverse relationship between mean litter proportions oftotal postimplantation loss and viable fetuses, except in the case of significant differences in thenumbers of successful implantations between control and treatment groups. Because of the variabilitythat is typically observed in these data, increased postimplantation loss in a test substance–treatedgroup is typically considered an adverse effect only when it reaches a level that isat least double that observed in the concurrent controls. More confidence may be placed in thedecision if a dose-response relationship is present and the concurrent control mean is within thehistorical control range.© 2006 by Taylor & Francis Group, LLC

370 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAn additional factor that must be considered for rat developmental toxicity studies is the numberof females per treatment group with an increased number of resorptions. Table 9.23 presents themodal distribution of the total numbers of females with resorptions in the authors’ Crl:CD ® (SD)IGSBR rat historical control database. These data indicate that an increase in the number of femalesthat have three or more resorptions is a signal of developmental toxicity.c. Prenatal GrowthThe most sensitive and reliable indicator of an alteration to intrauterine growth is a reduction infetal body weight. Fetal body weight collection in developmental toxicity studies occurs on theday of laparohysterectomy, generally GD 18, 20, and 29 for mice, rats, and rabbits, respectively.These time points represent the day prior to the expected day of delivery for each species.The consistency of fetal body weights, particularly for the rat model, has enabled investigatorsto censor those data that resulted from erroneous determination of evidence of mating. Incorrectassessment of the timing of mating in the Crl:CD ® (SD)IGS BR rat, even if only displaced by 24hours, may result in mean litter weights greater than 5.0 g (compared with the normal weight of3.6 g on GD 20). When evaluated against the considerable historical control database compiled inthe authors’ laboratory, the conclusion is that the fetuses are actually 21 or 22 days old, as opposedto the intended age of 20 days. In these presumably rare cases, the heavier litters should not beincluded in the group mean.A mature historical control database provides the best means of gauging the reasonableness ofa group mean fetal body weight (presented on a litter basis). It has been the authors’ observationsthat in mature historical control databases, the range of variation for control rat fetal body weightsis only 0.3 to 0.6 g, depending upon the sex and strain evaluated (see Table 9.24). In rabbits thisvariation is greater, but it is still small enough to enable consistent interpretation of study results.Furthermore, male fetal body weights are typically greater than female fetal body weights for themost commonly used species.d. Fetal Sex RatioTable 9.23Resorptions(No.)Modal distribution ofresorption rates inCrl:CD ® (SD)IGS BR ratsFemales(No.)0 7881 4472 1613 354 95 26 27 38 09 110 111 1Total 1450Source: Data tabulated from 61 Crl:CD ®(SD)IGS BR rat studies (1998–2003) conductedat WIL Research Laboratories, Inc.The sex ratio per litter, evaluated in conjunction with mean fetal body weights for males andfemales, may reveal whether or not the test agent preferentially affects survival of a particular© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 371Table 9.24Historically observed litter means (ranges) for fetal body weight by species and strainMouse Rat Rat RabbitEndpoint Crl:CD ® -1(ICR)BR (g) Crl:CD ® (SD)BR (g) Crl:CD ® (SD)IGS BR (g) Hra:(NZW)SPF (g)Combined sexes 1.31 (1.21–1.36) 3.5 (3.3–3.9) 3.6 (3.4–3.8) 46.9 (39.2–51.8)Males NA NA 3.7 (3.5–3.9) 47.3 (39.9–51.7)Females NA NA 3.5 (3.4–3.7) 45.8 (38.2–49.9)Note:NA = Data not available. The day of observation of evidence of mating (rodents) or artificial insemination(rabbits) was designated gestational day 0. Laparohysterectomies were conducted on GD 18 for mice,GD 20 for rats, and GD 29 for rabbits.Source: Data tabulated from studies conducted at WIL Research Laboratories, Inc., including 20 Crl:CD ® -1(ICR)BR mouse studies (1984–2000), 158 Crl:CD ® (SD)BR rat studies (1982-1997), 61 Crl:CD ® (SD)IGS BR ratstudies (1998–2003), and 81 Hra:(NZW)SPF rabbit studies (1992–2003). These four databases contain informationderived from 460, 3585, 1452, and 1608 pregnant control dams/does, respectively. From these litters,5552, 50,858, 22,047, and 10,278 viable fetuses, respectively, were available for evaluation.gender. It has been hypothesized that the hormone levels of both parents around the time ofconception may partially control the offspring sex ratio and manifestation of sex-biased malformations.76,77 However, such effects are rarely observed at the embryo or fetal stage of development inhazard identification studies. Mild variations from an equal male to female sex distribution ratiofrequently occur, and little significance is attached to values that fall within normally expectedranges. The historically observed mean percentage of males per litter for Crl:CD ® -1(ICR)BR mice(1984 to 2000) and Crl:CD ® (SD)IGS BR rats (1998 to 2003) is typically tighter (44.0% to 58.0%per litter) than the observed range for Hra:(NZW)SPF rabbit litters (37.9% to 74.1% per litter, 1982to 2002) in the authors’ laboratory. This difference is likely due to mathematical variability resultingfrom the smaller sample sizes for rabbits (average of 6.5 fetuses/litter) when compared with mice(12.1 fetuses/litter) or rats (15.2 fetuses/litter). For further discussion of this topic, the reader isreferred to Section VI.B.13.4. Fetal Morphologya. Context and Interrelatedness of FindingsMeasures of normalcy in morphogenesis are the most important endpoints in developmental toxicitystudies. Beyond good study design and conduct, it is important to remember that assessment ofeffects on prenatal development occurs in a single snapshot (at the time of laparohysterectomy).Not all fetuses in these litters are at the same point in their developmental schedules. The discipline’scurrent methods of assessment allow us to examine, at least macroscopically, every structure inthese fetuses. Normal morphology may be confounded by disruptions in the temporal pattern ofdevelopment. These insults may retard development by a direct effect on the conceptus (retardationof its growth), so that by the time the products of conception are evaluated, the natural timing andpatterns of development are affected, including increasing cortical masses with decreasing cavityvolumes, degree of cavitation of the brain ventricles and kidneys, closure of the neural tube, anddevelopment of the thorax and abdominal walls. More severe retardation of fetal growth mayproduce cardiac valve defects and altered histogenesis.Actual diagnosis or designation of a defect may be confused by gradations of tissue types andappearances in affected structures, because microscopic evaluation procedures are not routinelyemployed in these studies. Examination of skeletal and cartilaginous structures may be confoundedby imprecise staining procedures that do not enable quantification of the structures that are stained.Attempts have been made to semiquantify these procedures, but these methods have not beenvalidated or generally applied. 78,79 Skeletal examination is further confounded because for the mouseand rat, the time of laparohysterectomy (usually 24 h prior to expected delivery) coincides withthe period of peak osteogenesis. Ossification of rodent fetal bones occurs rapidly during the last© 2006 by Taylor & Francis Group, LLC

372 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION48 h of gestation. Because of these factors, the skeletal system in rodent fetuses is immature whenit is evaluated.Additional critical factors should be kept in mind when interpreting fetal morphology data. Forexample, the terminology and experimental basis for teratologic evaluation originated from thehuman medical field and human anatomy, for which no comparable detailed knowledge base exists(e.g., regarding human sternal and spinal cord development). Therefore, the certainty of many ofthese observations is not well founded relative to the human experience. Because of the incompletenessof the human database for newborn morphology (or lack of publication of these data),the interspecies differences and similarities for many of these body regions and structures are notwell understood. Also problematic in these guideline studies is that species with short in uterodevelopmental periods have been selected for use. In these species, the temporal spacing of complexdevelopmental processes may make the model especially vulnerable to subtle disruptions in growthand timing. Finally, in species with short gestation periods, examination of specimens one day priorto birth fails to address a critical period in morphogenesis and histogenesis that would naturallyoccur in utero in human development.Because of the numerous potential permutations of effects in developmental toxicity studies,the authors have chosen to summarize a critical approach to evaluation of the data in the formpresented in Table 9.25. It should be remembered that each data set is unique, and therefore thesequestions are intended to serve as a basic framework (as opposed to a “cookbook”) within whichto evaluate developmental toxicity studies. Evaluation of data from a developmental toxicity studyoccurs at two levels. First, all data must be examined to differentiate between effects that manifestat statistically significant levels and those that manifest substatistically. This superficial review ofthe data is necessary for thorough reporting and regulatory Good Laboratory Practice (GLP)completeness. Most important is interpretation of the toxicologic significance of the affectedendpoints. Table 9.26 presents selected examples of factors affecting interpretation of the data.b. Experimental Considerations1) Training of Prosectors — Macroscopic examination of fetal specimens requires extensivetraining and an understanding of developmental morphology. The potential impact to productdevelopment and, hence, human health may be substantial if any effects are unrecognized orerroneously created. This situation parallels that of a histopathologist’s importance for the outcomeof a carcinogenicity study, with one major difference. Professionals with veterinary degrees obtainboard certification and learn both standardized nomenclature and a process for classification ofhistopathology. In contrast, there are no formally recognized training programs or certificationboards for professionals who carry out reproductive and developmental toxicity studies. Investigatorsmust rely upon on-the-job training to conduct morphologic evaluations of progeny, where oneor two findings may be critical. In addition (or perhaps as a consequence), a universally acceptednomenclature and classification system for developmental morphologic effects has yet to be developedand adopted. With regard to standardized nomenclature for malformations in common laboratoryanimals, an FDA-sponsored effort in the mid-1990s (also supported by the InternationalFederation of Teratology Societies) resulted in the publication “Terminology of DevelopmentalAbnormalities in Common Laboratory Mammals (Version 1).” 80 Currently, the Terminology Committeeof the Teratology Society is in the process of updating this glossary.The following example illustrates the importance of a standardized nomenclature and classificationscheme. Muscular ventricular septal defects have never been detected in among over 21,000control Crl:CD ® (SD)IGS BR rat fetuses evaluated at WIL. However, other laboratories have reportedup to 200 ventricular septal defects (membranous and muscular) in similar databases. Realizingone practical aspect (i.e., degree of technical competence of personnel) of this discrepancy betweenlaboratories is important. Are technicians not recognizing these defects because they lack training?© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 373Table 9.25QuestionsKey considerations during assessment of developmental toxicity dataAre minor test article-treatedgroup changes inappropriatelyweighted by uncommonlylow/high control group values?Is the pregnancy rate significantlyreduced (either relative to theconcurrent control group or in allgroups)?Are clinical signs of toxicitypresent?Is there an increased incidenceof abortion (in rabbits)?Does decreased maternal bodyweight gain correlate with similarchanges in food consumption?Do body weight deficits occur ina dose-related manner?Does maternal body weight gain“rebound” after cessation oftreatment?Do maternal body weight gainand food consumption decreasefollowing cessation oftreatment?Is maternal net body weightaffected?Is food consumption reduced?Is maternal body weight gainreduced during a specific periodof gestation?Is gravid uterine weight affected?Is the percentage of affectedlitters increased relative tocontrols?Is the percentage of affectedfetuses per litter increasedrelative to controls?Is viable litter size decreased?Implications and CommentsAlthough the data from test article-treated groups initially should becompared with the concurrent control group, use of the laboratory’shistorical control data will enable identification of atypical concurrent controlvalues.Reduced pregnancy rates caused by the test article are rare, consideringthe onset of treatment in most developmental toxicity studies (typically afterimplantation). The power of the study (ability to detect dose-responserelationship) may be impaired.Severe clinical signs of toxicity that disrupt maternal homeostasis andnutritional status may result in subsequent insults to the products ofconception. Conversely, frank increases in terata in the absence ofsignificant maternal toxicity probably signal an exquisite effect onmorphogenesis and suggest that the embryo or fetus is more sensitive.Related maternal data must be examined carefully (females may havestopped eating), including the historical control rate of abortion andtemporal distribution of events (abortions during GD 18 to 23 are rare,compared to those occurring later in gestation).Concomitant reductions in body weight and food consumption usuallyindicate systemic toxicity, and in some cases signal a maternal CNS effect.Dose-related effects on body weight generally indicate a compound-relatedeffect. If maternal mortality is present at the high-dose level, effects onbody weight gain may not manifest because of elimination of sensitivemembers of the group. In this case, body weight effects occurring only atlower doses may represent an adverse effect.A rebound in body weight gain following treatment may indicate recoveryfrom the effects of the compound.This may indicate a withdrawal effect during the fetal growth phase becauseof maternal CNS dependency (e.g., opioid compounds). However, it isdifficult to correlate such an effect with fetal outcome.A decrease in maternal net body weight is most likely an indicator ofsystemic toxicity. However, reduced maternal body weight in the absenceof an effect on maternal net body weight indicates an effect on intrauterinegrowth and/or survival, rather than maternal systemic toxicity.Changes in food consumption generally signal either palatability issues orsystemic toxicity caused by the test substance. Reduced food consumptionwithout a concomitant effect on body weight gain is likely a transient effectand probably not adverse. If food utilization is unaffected when food intakeis reduced, the test article is probably affecting caloric intake (i.e., a testarticle or vehicle with high caloric content may cause an animal to consumeless food with minimal or no net effect on body weight gain).Decreased maternal body weight gain during the late gestational periodmay be associated with reduced fetal skeletal mineralization. In all cases,risk to the developing embryo or fetus is greater the longer the reductionin maternal body weight gain is sustained.Reduced gravid uterine weights are usually due to reduced viable litter sizeand/or reduced fetal weights. In rare cases, effects on the placenta or theuterus may be initially recognized by nonfetus-related changes in graviduterine weight.This may indicate a maternal dimension if the proportion of fetuses in thoselitters is not similarly affected.This probably indicates a proximate fetal effect.Decreased viable litter size generally represents either increased resorptionrate (usually attributable to the test agent) or decreased numbers of corporalutea and/or implantation sites (not likely to be a test substance–relatedeffect unless treatment began at or prior to fertilization).© 2006 by Taylor & Francis Group, LLC

374 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.25QuestionsKey considerations during assessment of developmental toxicity data (continued)Is the resorption rate increased?Is fetal weight subtly decreasedin the high-dose group (e.g.,mean rat fetal weights of 3.6,3.6, 3.6, and 3.4 g in control,low-, mid-, and high-dosegroups, respectively)?Does a correlation exist betweenlow-weight-for-age fetuses anddysmorphogenesis?Is a syndrome or spectrum ofeffects present?Is the total malformation rate pergroup affected?Are there qualitative differencesin the relative functional utility ofaffected structures or systems?Is fetal ossification generallyretarded (evidenced by AlizarinRed a and/or Alcian Blue buptake)?Is a dose-related responseobserved when the endpoint of“affected implants” is evaluated?Are there similar effects in otherstudies?aDawson, A. B., Stain Technol., 1, 123, 1926.bInouye, M., Congen. Anom., 16, 171, 1976.Implications and CommentsComplete litter resorptions are rare (3 in 1452 Crl:CD ® (SD)IGS BR ratlitters). Therefore, even one completely resorbed rat litter in the high-dosegroup merits attention. Furthermore, an increased incidence of controlfemale rats with more than three resorptions usually constitutes a signalof developmental toxicity (refer to Table 9.23). Refer to Section III.B. for adiscussion of the relationship between number of implantation sites andincreased resorption and/or abortion rate in rabbits.If other signals of developmental toxicity are present, the decrease in fetalweight is probably adverse. If no other signals of developmental toxicityexist, and the low fetal weight is within the laboratory’s historical controlrange, it is probably not an adverse effect. A robust, highly consistenthistorical control database in this instance would indicate that fetusesweighing 3.4 g are able to survive and thrive, and therefore the weightreduction is likely to be temporary. Such a conclusion may be corroboratedwith data from postnatal assessment studies.If malformations are limited to low-weight-for-age fetuses, the malformationsmay be due to generalized growth retardation (e.g., omphalocele in rabbitsor cleft palate in rats).A syndrome of effects indicates multiple insults on a specific organ systemduring development (e.g., tetralogy of Fallot, or a cascade of events duringheart development, beginning with ventricular septal defects and includingpulmonary stenosis and valvular defects). A spectrum of dysmorphogeniceffects implies a less targeted, more generalized response (e.g., vertebralagenesis occurring with ocular field defects).Several organ systems may have slight malformation rate increases thatmay not be outside the historical control range when evaluated individually.However, in summation they may signal an increased generalizeddysmorphogenic effect.Bent long bones (e.g., femur) are of greater concern than bent ribs.Unossified centra or sternebrae are not of great import if the underlyingcartilaginous structures are present and properly articulated. Malformedcaudal vertebrae or facial papillae in rats would not merit the concern thatsimilar alterations to analogous human organs or structures would elicit.Delayed ossification may be due to developmental delay resulting fromintrauterine growth retardation or may stem from properties of the testagent expected to cause delayed mineralization of skeletal structures (e.g.,calcium chelation, altering of blood urea nitrogen, alkaline phosphataseinhibition, or changes in parathyroid hormones or vitamin D).For this endpoint, each implantation is counted once (tabulated as“affected”), whether the outcome is an early or late resorption, malformedfetus, or dead fetus. As the rate of malformation increases, the rate of lateresorption declines. Likewise, as the number of early resorptions rises, thenumbers of late resorptions and malformed fetuses decline. Refer to Figure9.7 for an idealized graphical depiction of this interrelationship betweenmost common fetal outcomes.Comparison to effects in a second species developmental toxicity study,malformations manifested as anatomical or functional changes in the preandpostnatal development study, and maternal toxicity relative to systemictoxicity observed in subchronic toxicity studies may indicate that thepregnant animal is more susceptible to the test agent than the nonpregnantanimal.Alternatively, is the laboratory confident (because of the highly trained technicians) that there areno such occurrences? A highly trained and stable group of professionals would indicate the latter.Such professionals would have been trained to recognize “normalcy” when judging fetal morphology,the most important criterion in these kinds of evaluations. In addition, they would be mostlikely to have developed and be using a consistent approach to nomenclature and definitions ofmalformations and variations.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 375Table 9.26Factors affecting interpretation of developmental toxicity dataProcedural ArtifactsDiagnosisFailure to recognize rare findings when presentFixationGeneral tissue shrinkageOcular opacities due to formaldehyde fixationMishandling of specimensDeformed bone and cartilage (e.g., apparent arthrogryposis caused by uterine compression becausefetuses were not removed from the uterus in a timely manner)Hematomas or petechial hemorrhagesImprecise heart cross-sectionsEndpoint Interrelatedness (Refer to Table 9.25 for Further Details)Absence of dose-related response in malformation rate because of dose-related increase in postimplantationlossFetal weight inversely related to litter size (intralitter competition)Late gestation maternal body weight gain relative to intrauterine survivalGeneralized intrauterine growth retardation vs. specific structural malformations (e.g., omphalocele, cleftpalate, and increased renal cavitation)Temporal PatternsContinuum of responses (effect of lumping versus splitting of findings)Overrepresenting an effect by tabulating both malformations and variations in the same body region,potentially confounding delineation of the NOAELInsult to specific anatomic region manifested as variations at lower dose levels and as malformations athigher dose levelsShifts in the timing of intrauterine death (early to late resorption) relative to dose-response curve (referto Figure 9.8 for an idealized graphical depiction)Lack of a dose-response relationship because of saturation of absorption, binding, and/or receptor capacitySymmetrical expression of the abnormalityIdeally, to minimize the bias introduced by subjectivity, the same individuals should performthe fetal evaluations across all dosage groups in a single study. To minimize unnecessary variabilitybetween outcomes, the same personnel should conduct the entire reproductive and developmentalprogram for a particular test agent, if possible.2) Methodology — Prenatal morphogenesis may be evaluated via one of two primary methods:(1) the older Wilson 81 method, involving serial freehand razor sections of decalcified specimens,or (2) microdissection of nonfixed fetuses (commonly referred to as fresh dissection 82 or the Staplesmethod 83 ). Though combinations of these two methods have been published (e.g., the Barrow andTaylor technique 84 ), the Wilson and fresh dissection methods are the most commonly employedapproaches. Following is a discussion of advantages and disadvantages of each method.The primary advantage of the fresh dissection method is that all fetuses can be examined forboth soft tissue and skeletal changes (see Figure 9.7), thus maintaining the greatest possiblestatistical power of a study. In addition, artifactual changes as a result of fixation can occur withthe Wilson method (refer to Table 9.26), and visualization of spatial relationships in two dimensions(e.g., following structures that transcend multiple sections) is much more difficult to master thanwhen they are visualized in situ (refer to Table 9.27 for additional discussion of the numerouspractical and scientific advantages of fresh dissection of fetuses).Another reason that the Wilson sectioning method, which allows the processed specimens tobe stored and examined later, is not the best approach is that this method necessitates a priori“subsetting” of the fetuses into those that will be examined viscerally and those that will beexamined skeletally. Such a practice is usually followed simply out of convenience; it is undesirablestatistically and may diminish the value of the study. For example, from the external perspective,a fetus may manifest multiple malformations. Complete characterization of underlying internalmorphologic defects would require both thorough visceral and skeletal examinations. However, the© 2006 by Taylor & Francis Group, LLC

376 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION4 Groups (of 25 Dams) × 350 Fetuses= 1400 FetusesFreehandSectionWhole-BodyMicrodissection1 /2 Skeletal 1 /2 Visceral 1 /2 Control andHigh GroupSkeletal andVisceral100% Skeletaland VisceralFigure 9.7Comparison of the number of fetuses available for visceral and skeletal examination using thefreehand section and whole-body microdissection techniques.MalformationsEarlyresorptions%AffectedLateresorptionsIncreasing doseFigure 9.8Idealized depiction of the relationship of dose and adverse developmental outcome.Wilson sectioning method requires the investigator to commit to either a visceral or skeletalexamination, but not both.Some of the rationale behind a priori subsetting of the fetuses may be defensible if theinvestigator is using a high dose of a compound that will affect at least 15% to 20% of the fetuses.However, this rationale presumes that each compound tested will evoke such a response and thatthis would be known prior to the study, which is not usually the case. Therefore, adverse effectson fetal morphology are much more likely to be overlooked if a large number of compounds aretested by the subsetting approach.Further compounding the visceral-skeletal subsetting problem is the ICH guideline–drivenoption of only evaluating control and high-dose group fetuses when no apparent effects are observedat the high-dose level. This approach assumes that any treatment-related findings that occur will© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 377Table 9.27Advantages of fresh dissection of fetusesProsectors are more easily trained.Good visualization with less chance of missing a cleft, breach, or absence of structures (e.g., valve in heart),because of thickness and perpendicularity of cross-section of the sample.Direct visualization of structures, as one would learn in anatomy, instead of having to rotate images mentally inthree dimensions.Coloration (organs, vessels, tissues).Direct ascertainment of vessel patency by tactile manipulation of vessels to determine patency and if blood is moving.Evaluation of vascular tissue perfusion by coloration.All fetuses can be examined for both soft tissue and skeletal changes.Avoids the difficulty of sectioning every fetus in exactly the same way as is needed to ensure comparability ofviews among fetuses.Decreased probability of creating artifacts due to fixative tissue shrinkagebe discernible in the high-dose group (refer to Section VII for a further discussion of the dangersinherent in this assumption). The statistical power of these studies is already low relative toextrapolation to the human population, and when the dosage group has already been subsetted priorto evaluation, the probability of detecting treatment-related effects declines dramatically. Therefore,the authors conclude that morphologic evaluation of all fetuses by use of the fresh dissection methodis the most powerful approach for developmental toxicity studies, yielding the most conclusiveresults for hazard identification.c. Relative Severity of Findings (Malformations vs. Variations)A critical first step in fetal morphologic evaluation is determining the relative severity of thealteration. This determination is complicated by the potential continuum of responses betweennormal and extremely deviant fetal morphology. Fetal dysmorphogenic findings have been reportedas developmental deviations, structural changes, malformations, anomalies, congenital defects,anatomic alterations, terata, structural alterations, deformations, abnormalities, anatomic variants,or developmental variations, with little consistency across laboratories in definition of the terms.The current authors have chosen to report, and accordingly define, these external, visceral, andskeletal findings as either developmental variations or malformations. Variations are defined asalterations in anatomic structure that are considered to have no significant biological effect onanimal health or body conformity and/or occur at high incidence, representing slight deviationsfrom normal. Malformations are defined as those structural anomalies that alter general bodyconformity, disrupt or interfere with normal body function, or may be incompatible with life.d. Presentation of FindingsFetal morphologic findings should be summarized by: (1) presenting the incidence of a given findingboth as the number of fetuses and the number of litters available for examination in the group and(2) considering the litter as the basic unit for comparison and calculating the number of affectedfetuses in a litter on a proportional basis as follows:whereSummation per group (%) =Σ viable fetuses affected per litter (%)Number of litters per groupViable fetuses affected per litter (%) =Number of viable fetuses affected per litterNumber of viable fetuses per litter× 100© 2006 by Taylor & Francis Group, LLC

378 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.28Most commonly occurring developmental variations in control Crl:CD ® (SD)IGS BR ratsTotal Number Examined (1998–2003) Fetuses Litters % per LitterExternal 22,047 1447 —None observed in control ratsVisceral 21,853 1447 —Major blood vessel variation a 15 15 0.0–0.4Renal papilla(e) not developed and/or distended ureter(s) 5 4 0.0–0.8Hemorrhagic ring around the iris 3 3 0.0–0.3Hemorrhagic iris 1 1 0.0–0.3Skeletal 21,843 1446 ––Cervical centrum no. 1 ossified 3,919 1065 6.6–32.1Sternebra(e) no. 5 and/or no. 6 unossified 1,920 632 0.3–23.114th rudimentary rib(s) 1,328 565 1.4–15.1Hyoid unossified 325 208 0.0–3.47th cervical rib(s) 121 91 0.0–2.7Reduced ossification of the 13th rib(s) 119 74 0.0–3.0Sternebra(e) no. 1, 2, 3, and/or 4 unossified 50 44 0.0–1.3Sternebra(e) malaligned (slight or moderate) 32 31 0.0–0.825 presacral vertebrae 30 20 0.0–2.027 presacral vertebrae 29 21 0.0–1.87th sternebra 19 5 0.0–2.514th full rib(s) 18 14 0.0–0.9Bent rib(s) 18 18 0.0–4.0aThe most commonly occurring manifestations of this finding are: (1) right carotid and right subclavian arteriesarising independently from the aortic arch (no brachiocephalic trunk), (2) left carotid artery arising from thebrachiocephalic trunk, and (3) retroesophageal right subclavian artery.Source: Data collected at WIL Research Laboratories, Inc.e. Significance of Developmental VariantsAn important issue in interpreting developmental toxicity study data is the developmental (anatomic)variant. Table 9.28 and Table 9.29 present the most commonly occurring developmentalvariations observed in control Crl:CD ® (SD)IGS BR rats and Hra:(NZW)SPF rabbits, respectively,in the authors’ laboratory. The total numbers of descriptors presented in these tables (17 for ratsand 23 for rabbits) represent approximately 42% of the total number of descriptors for each specieswhen findings from both control and test substance-exposed animals are tabulated. At least 41unique developmental variations have been observed over the past 6 years of investigations usingthe Crl:CD ® (SD)IGS BR rat in the authors’ laboratory (61 studies, 1447 litters with viable fetuses).In the case of the Hra:(NZW)SPF rabbit, at least 53 different developmental variations have beenobserved in 1529 litters (81 studies) from 1992 through 2003.Developmental variants are of somewhat lesser concern to the investigator and regulatoryreviewer than malformations; however, they often pose a great dilemma if not interpreted properly.In assessment of whether a fetal morphologic deviation represents a malformation or a variation,the factors listed in Table 9.30 must be considered.Findings that are classified as developmental variants must then be assessed for their toxicologicsignificance and potential impact on hazard identification decisions. In the latter case, the EPAhistorically has been much more mindful of developmental variants, often differentiating the NOELand NOAEL based upon the context of these variants, than has the FDA. 39 The toxicologicsignificance (human versus animal) of developmental variants generally will be dependent uponthe combined weight of their impact on salubrity and their fate (the extent to which they are© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 379Table 9.29Most commonly occurring developmental variations in control Hra:(NZW)SPF rabbitsTotal Number Examined (1992–2003) Fetuses Litters % per LitterExternal 10,278 1529 —Twinning 1 1 0.0–0.8Visceral 10,278 1529 —Accessory spleen(s) 1,198 681 4.8–33.2Major blood vessel variation a 565 329 0.0–17.5Gallbladder absent or small 150 115 0.0–7.8Retrocaval ureter 142 110 0.0–5.4Hemorrhagic ring around the iris 46 33 0.0–3.6Spleen — small 6 6 0.0–1.0Hemorrhagic iris 4 4 0.0–0.8Liver — pale 2 2 0.0–0.6Eye(s) — opacity 2 1 0.0–1.0Accessory adrenal(s) 1 1 0.0–0.7Renal papilla(e) not developed and/or distended ureter(s) 1 1 0.0–1.2Skeletal 10,278 1529 —13th full rib(s) 4,082 1240 19.4–59.113th rudimentary rib(s) 1,982 1042 8.1–32.527 presacral vertebrae 1,724 766 4.5–32.1Hyoid arch(es) bent 504 357 0.0–22.2Sternebra(e) no. 5 and/or 6 unossified 448 274 0.0–11.4Sternebra(e) with threadlike attachment 146 121 0.0–9.1Sternebra(e) malaligned (slight or moderate) 117 108 0.0–5.0Extra site of ossification anterior to sternebra no. 1 106 84 0.0–7.4Accessory skull bone(s) 80 69 0.0–5.07th cervical rib(s) 73 59 0.0–7.725 presacral vertebrae 35 31 0.0–7.4aThe most commonly occurring manifestations of this finding are: (1) right carotid and right subclavian arteriesarising independently from the aortic arch (no brachiocephalic trunk), (2) left carotid artery arising from thebrachiocephalic trunk, and (3) retroesophageal right subclavian artery.Source: Data collected at WIL Research Laboratories, Inc.Table 9.30Observational determinantsof anatomic or functionaldeviationsDegree of deviation from averageMagnitude of morphologic changeNature of functional impact, if anyIncidence (prevalence) aImpact on salubrityCosmetic significanceaIn humans, the convention has been touse an incidence of less than 4% and nomedical or surgical significance. In experimentalstudies, no convention has beenestablished.reversible). The difficulty is that for animal models the impact on salubrity is generally unknown(unless adult bone scans are conducted to follow the fate of skeletal variations). However, if a cleardose-related response manifests, additional factors, including statistical power of the study andpotential for temporal changes (due to breeding, penetrance, environmental factors, etc.), must beaddressed.© 2006 by Taylor & Francis Group, LLC

380 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.31Determination of the significance of developmental variationsLearned scientific debatePerforming confirmatory studies if necessaryDetermining if an effect is reversible with maturation or if significant remodeling is likelyComparing the anatomy of mothers and progeny (e.g., vascular variations)Assessing the prevalence of paternal contributions (e.g., gallbladder variations in the rabbit)Taking advantage of newer technologies when possible (e.g., assessing potential expression of the variant inthe adult phenotype via computed topography [CT] scan or bone densitometry)A suggested approach to distinguishing the significance of developmental variants is outlinedin Table 9.31. A combination of the steps may determine such significance.Further discussion of related examples — the rabbit gallbladder issue and expression of 27presacral vertebrae and supernumerary ribs in rats and rabbits — is presented in the followingsections.1) Gallbladder Variations in Rabbits — A fetal observation in the rabbit that has generatedconsiderable regulatory discussion and discord is the “absent” or “small” gallbladder. The classicwork by Sawin and Crary clearly demonstrated that the spontaneous incidence of “absent” gallbladderwas significant and variable in several stocks of rabbits. 85 Sawin and Crary adroitly pointedout that most of the literature referenced the “presence” or “absence” of the gallbladder as a speciescharacteristic, rather than an individual characteristic. Sawin and Crary studied genetic influenceson the gallbladder in 25 unrelated stocks, several of which had been inbred for as many as 22generations. Of particular note to the present day use of the Hra:(NZW)SPF stock rabbit were thepresent authors’ studies demonstrating that some animals repeatedly produce offspring lacking agallbladder, even when the authors’ investigation of the phenomenon was extended to include otherstocks. The “absence” of the gallbladder was part of a graded series of variations in its size andshape.In the authors’ experience, it is relatively rare to see a true and confirmed (by an additionalstudy) dose-related increase in gallbladder absence. More often than not, these scenarios appeardose related but are not reproducible. Of course, there may be exceptions.Over the course of hundreds of studies, the authors compiled a historical database on thisobservation and analyzed paternal contributions to size and/or presence of the gallbladder. Thesedata indicate that there is a highly variable and sometimes dramatic trend that certain stock buckssire litters with many more fetuses having either a “small” or “absent” gallbladder (refer to Table 9.32for representative examples). Continued monitoring of the colony of breeders and maintainingappropriate records for use in interpreting related fetal findings is critical. However, purchasingdate-mated females has become commonplace in some laboratories engaged in reproductive toxicology.Without sire records, these paternally influenced fetal observations and their relationshipto treatment cannot be ascertained or appropriately analyzed.2) Supernumerary Ribs and Presacral Vertebrae — Historically, the authors’ laboratory hasclassified 27 presacral vertebrae (PV) and supernumerary ribs (SR) as developmental variations.These variants are routinely quantified and occur in control populations at well-characterizedfrequencies. It is the authors’ opinion that 27 PV and 13th full/rudimentary rib, in the absence ofother fetal effects, should not automatically be considered a finding of toxicologic concern orteratogenicity in developmental toxicity studies with rabbits. Furthermore, caution is warrantedwhen considering whether an increased occurrence of 27 PV or 13th full or rudimentary rib isrelevant when extrapolating to human development.The thoracolumbar border that gives rise to 27 PV and SR is highly labile in rabbits, makingits significance in developmental toxicity problematic. Results of 10,037 fetal skeletal evaluations© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 381Table 9.32Stock Male No.Paternally influenced prevalence of absent or reducedgallbladder in New Zealand white rabbit fetusesHra:(NZW)SPF Rabbits dPercent FetusesAffected per Litter aNo. ofLitters bNo. with AbsentGallbladder/Litter c3508 20.2 109 8/72876 20.0 40 5/48457 14.6 48 0/02877 11.1 166 5/42871 3.9 152 7/6Lsf:NZW Rabbits e481 58.3 36 20/13252 27.0 37 10/7445 24.1 29 8/7251 4.8 21 1/1aPercent of fetuses per litter with absent or reduced gallbladder.bTotal litters sired.cNo. fetuses/No. litters.dA limited number of bucks (5) is presented to demonstrate the range of values,representing more than 100 animals in the entire Hra:(NZW)SPF rabbit databaseat WIL Research.eFour bucks were used to demonstrate the range of values in the Lsf:NZWrabbit database at WIL Research.from 79 rabbit studies recorded in the WIL historical control database showed that 13th full rib,13th rudimentary rib, and 27 PV are the first, second, and third most frequently noted skeletalvariations, respectively, in control rabbits. The mean litter percent of fetal 27 PV was 17.4% (rangeof 5% to 32%; Q1–Q3 interquartile range of 13% to 21%). One-half of the control females carriedat least one fetus with 27 PV, and over 1% of control females exhibited a 100% occurrence of fetal27 PV. The mean litter percentages of fetuses with 13th full or 13th rudimentary rib were 39%(range of 19% to 59%; Q1–Q3 interquartile range of 35% to 44%) and 20% (range of 8% to 32%;Q1–Q3 interquartile range of 17% to 23%), respectively. There was also a strong positive correlationbetween SR and 27 PV among control populations. The presacral region is more labile in rabbitsthan in rats and mice, which exhibit mean litter percentages of 27 PV of 0.15% and 0.13%,respectively. The mean litter percent of supernumerary or rudimentary ribs is also lower than inrabbits, i.e., 6.1% in rats and 12.9% in mice.A recent study examined the prevalence of 27 PV and SR in 62 adult male and 100 adult femalerabbits by radiography. 86 The incidence of 27 PV among adult control rabbits was 22%. Thisincidence agreed well (especially given the difference in N) with the mean litter percent of 27 PVamong fetal rabbits (17%), suggesting that there is no detriment or decrease in survival of progenywith this variant. These findings also provide evidence that the presence of 27 PV is withoutsignificant effect on salubrity and suggest little or no remodeling of these structures in the rabbit,as there is in rodents. This radiographic study further showed that 54.6% of adult control rabbitshave some type of SR (e.g., full, rudimentary, bilateral, unilateral or in combination). Publishedreports have suggested that both chemically induced and control fetal SR can resolve duringmaturation in rats 87–90 and that both 12 and 13 pairs of ribs are wild-type phenotypes in adultrabbits. 91,92 The radiographic study found that 80% of adult rabbits with 27 PV also exhibited 13thfull rib. 86 In other words, the occurrence of 13th full rib (considered wild type) in the rabbit occurspredominantly with additional vertebrae in the axial skeleton. Separate laboratories have reported© 2006 by Taylor & Francis Group, LLC

382 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONsimilar correlations between PV and SR in adult rabbits. 88,90,93 It is also worth noting that approximately5% of humans exhibit supernumerary (3%) or missing (2%) presacral vertebrae. 94In the absence of any other fetal effects (i.e., intrauterine growth retardation, major or minormalformations, fetal death, or functional impairment), an increased occurrence of 27 PV and/or13th full ribs is not sufficient evidence of developmental toxicity. This is especially true in therabbit model, where there is a high background incidence of these skeletal variations in controls,where 12 or 13 pairs of ribs is considered the wild-type phenotype, and where 13th full rib iscommonly associated with 27 PV.The usefulness of endpoints such as 27 PV and 13th full rib in predicting risk for the humanpopulation is complex and has not been well studied or validated. Reviews of known teratogenicagents (e.g., thalidomide, alcohol, ionizing radiation, steroidal hormones, heroin, or morphine) havebeen published, taking into account the endpoints evaluated, the relative power of the studies, thedose-response patterns, and overall toxicity to the maternal and fetal systems. 11,95–97 Based on thelimited number of agents for which adequate published data were found, it was determined thatthe most reliable endpoints for predicting risk of developmental toxicity in humans were fetalgrowth retardation, malformations, functional impairment, and spontaneous abortion. There iscurrently inadequate information regarding skeletal developmental variations to make judgmentsconcerning concordance or nonconcordance between human and animal data. The lack of concordanceacross animal species (e.g., rabbits versus rodents) for 27 PV and 13th full rib suggests thatthese findings might simply reflect species-specific anatomic variation, without detrimental effectson salubrity.VI. REPRODUCTIVE TOXICITY STUDIESThis section begins with a discussion of the various regulatory guideline requirements for reproductivetoxicity studies. Specific endpoints measured in these studies are then addressed, andguidance on their interpretation is provided, based on the authors’ collective experience andhistorical data.A. Guideline RequirementsTable 9.33 presents the salient differences between the various regulatory guidelines for reproductivetoxicity study designs that may affect interpretation of the data. In the United States, OPPTS (withinthe EPA) has developed reproductive toxicity testing guidelines for use in the testing of pesticidesand toxic substances. These guidelines harmonized the previously separate testing requirementsunder TSCA and FIFRA, and the OECD in Europe has developed guidelines very similar to thoseof the EPA. For medicinal products, ICH developed technical requirements for the conduct ofreproductive toxicity studies for supporting the registration of pharmaceutical products for the EU,Japan, and the United States. Finally, for food ingredients, CFSAN (within the FDA) has developedguidelines for reproductive toxicity studies.1. Animal ModelsThe rat is the recommended species for both the EPA and OECD multigeneration study and theICH fertility and pre- and postnatal development studies. There are several advantages to using therat, such as a high fertility rate, a large historical control database, and cost effectiveness. However,the rat is not always the most appropriate model for these studies. Based on mode of action, kinetics,nature of the compound (e.g., recombinant protein) or sensitivity, other species, such as the rabbit,dog, or nonhuman primate, have at times been used instead of the rat. However, because the duration© 2006 by Taylor & Francis Group, LLC

Table 9.33Selected comparison points between various regulatory agency guideline requirements for fertility and reproductive toxicity studiesComparison Point FDA/Redbook 2000 (2000 Draft) OECD (2001) EPA (1998) ICH (1994)Reference 172 173 174 175,176Species Preferably rat Preferably rat Preferably rat Preferably ratGroup size30/sex/group (F 0 ) andYield preferably not less than 20 Yield approx. 20 pregnant 16–20 litters25/sex/group (maximum of2/sex/litter) for F 1 , to yieldapprox. 20 pregnantfemales/generationpregnant females/generation females/generationAge at start of treatment 5–9 weeks 5–9 weeks 5–9 weeks Young, mature adults at time ofmatingNumber of generations directly 2, optional 3 2 2 1exposedDuration of treatment pergenerationMales: 10 weeks prior to andthroughout the mating periodMales: 10 weeks prior to andthroughout the mating periodMales: 10 weeks prior to andthroughout the mating periodMales: 4 weeks prior to andthroughout the mating periodFemales: 10 weeks prior tomating, throughout mating andpregnancy and up to weaningof the offspringEstrous cycle determination Minimum of 3 weeks prior tocohabitation and duringcohabitationMating period 1:1 until copulation occurs or 2to 3 weeks have elapsedSpermatogenesis assessmentIf there is evidence of malemediatedeffects on developingoffspring, all (at least of thecontrol and high dose) malesper generation assessed forsperm motility and morphologyand enumeration ofhomogenization-resistantspermatids and caudaepididymal spermFemales: 10 weeks prior tomating, throughout mating andpregnancy and up to weaningof the offspringPrior to cohabitation andoptionally during cohabitation1:1 until copulation occurs or 2weeks have elapsed. In casepairing is unsuccessful,remating of females with provenmale from same group could beconsideredAt least 10 males per generationassessed for sperm motility andmorphology and enumerationof homogenization-resistantspermatids and caudaepididymal spermFemales: 10 weeks prior tomating, throughout matingperiod and pregnancy and upto weaning of the offspringMinimum of 3 weeks prior tocohabitation and throughoutcohabitation1:1 until copulation occurs oreither 3 estrous periods or 2weeks have elapsed. If matinghas not occurred, animalsshould be separated withoutfurther opportunity for matingAll (at least of the control andhigh dose) males pergeneration assessed for spermmotility and morphology andenumeration ofhomogenization-resistantspermatids and caudaepididymal spermFemales: 2 weeks prior tomating, throughout matingperiod and through at leastimplantationAt least during the mating period1:1; most laboratories would usea mating period of between 2and 3 weeksSperm count in epididymides ortestes, as well as sperm viability(later changed to optional)DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 383© 2006 by Taylor & Francis Group, LLC

Table 9.33Organ weights — adultsAdult histopathologyExamination of offspringUterus, ovaries, testes,epididymides (total and cauda),seminal vesicles, prostate,brain, liver, kidneys, adrenals,spleen, pituitary, known targetorgans10 control and high dose animalsselected for matingFrom as soon as possible afterbirth until weaningUterus, ovaries, testes,epididymides (total and cauda),seminal vesicles, prostate,brain, liver, kidneys, adrenals,spleen, thyroid, pituitary, knowntarget organsAll control and high dose animalsselected for matingFrom as soon as possible afterbirth until weaningUterus, ovaries, testes,epididymides (total and cauda),seminal vesicles, prostate,brain, liver, kidneys, adrenals,spleen, pituitary, known targetorgans10 control and high dose animalsselected for matingFrom as soon as possible afterbirth until weaningNot requiredNot requiredAt any point after mid-pregnancyStandardization of litters Optional Optional Optional Optional as part of a pre- andpostnatal development studyDevelopmental landmarks Age of vaginal opening andpreputial separationAge of vaginal opening andpreputial separationAge of vaginal opening andpreputial separationAssessment recommended aspart of a pre- and postnataldevelopment studyOffspring functional testsOrgan weights — pupsSelected comparison points between various regulatory agency guideline requirements for fertility and reproductive toxicity studies (continued)Comparison Point FDA/Redbook 2000 (2000 Draft) OECD (2001) EPA (1998) ICH (1994)Optional, excellent vehicle toscreen for potentialdevelopmental neurotoxicantsBrain, spleen and thymus from atleast two pups/sex/litter atweaningRecommended when separatestudies on neurodevelopmentare not availableBrain, spleen and thymus fromone pup/sex/litter at weaningNot requiredBrain, spleen and thymus fromone pup/sex/litter at weaningTeratology phase Optional as either F 2b or F 3b litter Separate study Separate study OptionalImmunotoxicity screening Optional for F 0 , F 1 , and F 2 Not required Separate study Not requiredgenerationsAssessment recommended aspart of a pre- and postnataldevelopment studyNot required384 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 385and cost of the study are typically greatly increased, reproductive toxicity studies using suchalternative species are only rarely performed.2. Timing and Duration of TreatmentThe EPA, OECD, and FDA food additives requirements for reproductive toxicity studies aredesigned to assess the effects of long-term exposure, while the treatment periods for drug studiesare much shorter and less comprehensive. The rationale behind the long-term exposure regimensemployed by the EPA and OECD is to mimic the ambient, low-level, long-term exposures toindustrial chemicals in the workplace and to low-level residues of pesticides in the diet. Similarly,exposure to food additives may occur over a lifetime. In contrast, treatment during defined stages(phases) of reproduction in the ICH guidelines generally better reflects human exposure to medicinalproducts and allows more specific identification of stages at risk. In the EPA, OECD, and FDAmultigeneration studies, both generations of parental animals are exposed prior to mating andthroughout the mating, gestation, and lactation periods in the rat. The ICH studies are divided intothree segments: (1) fertility assessment, (2) embryo-fetal development assessment, and (3) pre- andpostnatal development assessment. Within each of the ICH segments (phases), the direct exposureperiod is also more limited than is required in the EPA and OECD study designs. For example, a10-week premating period is recommended for the EPA, OECD, and FDA multigeneration studies,while the premating period in the ICH fertility study is generally only 4 and 2 weeks for the malesand females, respectively. In a collaborative study of 16 compounds, the shorter premating maleexposure period for ICH studies was found to be appropriate for detection of drug effects on malefertility. 98 In that collaborative study, the 4-week premating period was found to be as effective asa 9-week premating period for identification of male reproductive toxicants, albeit in a limited setof agents. In addition, exposure ends at the time of implantation in the ICH fertility study butcontinues throughout the entire gestation and lactation periods in the EPA, OECD, and FDAmultigeneration study. Unlike the EPA, OECD, and food additive multigeneration studies, exposureof offspring following weaning cannot be assumed in the ICH pre- and postnatal study.Because of the short premating treatment regimen in the ICH fertility study, compound administrationdoes not begin until the rats are 8 to 9 weeks of age, which is after puberty has occurred.This is in contrast to the EPA, OECD, and food additive design in which direct F 1 exposure beginsat weaning (prior to sexual maturation) and F 0 exposure begins typically at 5 to 9 weeks of age,during the process of sexual maturation in the females and prior to maturation in the males.Another difference among the guidelines is that only the OECD requires adult thyroid weightevaluation. This is a surprising omission from the food additive and EPA guidelines because of therole of the thyroid in development and sexual maturation, recent welling concerns for endocrinemodulation, and questions and concerns regarding the thyroid toxicity of ammonium perchlorate.99,100B. Interpretation of Reproductive Toxicity Study EndpointsIn this section, the endpoints listed in Table 9.34 (male-female or coupled) for evaluating reproductivetoxicity are discussed. These endpoints are listed first in approximate chronological orderof collection and then ranked by approximate sensitivity of the endpoints (those endpoints that arebolded may be assessed in and/or appropriate for evaluation in humans). In this ranking, endpointsconsidered most sensitive are those that are most often affected in general or those most oftenaffected by lower doses of xenobiotics. In the following sections, discussion of Crl:CD ® SD(IGS)BR rat historical control data for reproductive toxicity endpoints is based on the database compiledin the authors’ laboratory. Unless specifically noted otherwise, these data originate from 55 studies(92 separate matings) conducted during 1996 to 2002.© 2006 by Taylor & Francis Group, LLC

386 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.34Endpoints of reproductive toxicity studiesApproximate Chronological Order of CollectionEstrous cyclicityPrecoital intervalSexual behaviorMating/fertility indices and reproductive outcomeDuration of gestationParturitionNeonatal survival indicesPrenatal mortalityNesting and nursing behaviorAssessment of sperm qualityWeight and morphology (macroscopic andmicroscopic) of reproductive organsOocyte quantificationQualitative and quantitative physiologic endpointsrevealing unique toxicities of pregnancy andlactationViable litter size/live birth indexSex ratio in progenyNeonatal growthLandmarks of sexual maturityFunctional toxicities and CNS maturationLearning and memoryRanked by SensitivityViable litter size/live birth indexNeonatal growth aNeonatal survival indicesPrenatal mortalityAssessment of sperm qualityWeight and morphology (macroscopic and microscopic)of reproductive organsEstrous cyclicityPrecoital intervalMating and fertility indices and reproductiveoutcomeDuration of gestationParturitionLandmarks of sexual maturityFunctional toxicities and CNS maturationLearning and memoryQualitative and quantitative physiologic endpointsrevealing unique toxicities of pregnancy and lactationNesting and nursing behaviorSexual behaviorSex ratio in progenyOocyte quantificationaBolded endpoints may be assessed and/or appropriate for evaluation in humans.Because sensitivity of an endpoint may be methodologically or biologically dependent, theranking in Table 9.34 is relatively arbitrary and should not be considered a definitive ranking ofendpoints. Mode of action may determine which measure is more sensitive.1. Viable Litter Size/Live Birth IndexThe calculation of live litter size and live birth index is as follows:Live litter size =Live birth index =Total viable pups on day 0Number litters with viable pups on day 0Number of live offspringNumber live offspring delivered × 100As an apical measure, viable litter size is frequently the most sensitive indicator of reproductivetoxicity and is a very stable index. Decreased numbers in this endpoint relative to the concurrentcontrol group can result from a reduction in the ovulatory rate and timing, from shifts in the timingof tubal transport, and from reductions in implantation rate, postimplantation survival, or spermparameters (motility or concentration). In assessing the number of viable pups, it is important toexamine the litters as soon as possible following birth to ascertain the number of live pups versusthe number of stillborn pups. In addition, cannibalism of the progeny by the dam may result in aninaccurate determination of the number of pups born. However, this concern has probably beenoverstated as a technical problem except in instances where offspring are born dead or malformed,or there is a specific effect on CNS function by the test agent. Problems with cannibalism havedecreased because of improvements in husbandry and laboratory practices, coupled with betterdefinedstocks of animals. Optimal offspring quantification can be accomplished by examining thedam at least twice each day (in the morning and afternoon) during the period of expected parturition(beginning on GD 21). These frequent examinations are necessary to detect dystocia. Within the© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 387authors’ laboratory, the historical control mean live litter size for Crl:CD ® (SD)IGS BR rats is 14.1pups per litter. When sufficient numbers of animals are included in the study to produce approximately20 litters per group, compound-treated group mean reductions of as little as one pup perlitter may be an indicator of an adverse effect. Decrements of 1.5 pups per litter and greater arestrong signals of reproductive toxicity. This is especially true if differences from the control groupoccur in a dose-related manner.2. Neonatal GrowthA change in neonatal growth compared to the concurrent control is another endpoint that oftenindicates toxicity in a reproductive toxicity study. Adverse effects can be manifested by reducedbirth weights caused by growth retardation in utero and/or by reduced body weight gain followingbirth. Reductions in body weight gain following birth may be a result of indirect exposure to thetest article through nursing or may be attributable to decreased milk production by the dam. Groupmean male pup birth weights are generally approximately 0.5 g greater than group mean femalebirth weights. It is of critical importance to evaluate the neonatal growth curve in conjunction withlitter size to control for the confounding effects of within-litter competition. This can be accomplishedby using a nested analysis of variance, in which the covariate is live litter size. 56 Meanabsolute pup body weight differences of 10% or more will typically result in statistical significance.In studies that use the dietary route of exposure, reductions in offspring body weight that areobserved only during the week prior to weaning (postnatal days [PND] 14 to 21) are most likelya result of direct consumption of the feed by the pups. In a study using a fixed concentration ofthe test article in the diet, the milligram per kilogram exposure of these pups during the week priorto weaning and immediately following weaning is often much greater than that of the dam. Thisis because of the higher proportional consumption of food by the pups. It is also important toconsider route of exposure when comparing pup body weights with historical control data. Forexample, in the typical EPA multigeneration study design via whole-body inhalation, the dams areremoved from the pups and are exposed to the test material for 6 h/d. Because of this separationfrom the litter, pup body weights at weaning on PND 21 are generally approximately 20% lessthan the weights of pups from a study employing an oral (dietary, gavage, or drinking water) routeof exposure.3. Neonatal Survival IndicesNeonatal survival indices, in conjunction with offspring body weights, are used to gauge disturbancesin postnatal salubrity, growth, and development. When evaluating postnatal survival, it isimportant to maintain the litter as the experimental unit. However, it is also useful to inspect theabsolute numbers of deaths in each group. Standardization of litter size is an option in all guidelines.The United States convention is to standardize litters on PND 4, while litter standardization is oftennot performed in Europe. Neither approach is clearly right or wrong. The main advantage of littersize standardization is reduction of the variance in offspring weight due to litter size differences.This approach will improve sensitivity and make the historical control data range less variable.However, using this approach can decrease the likelihood of identifying outlier pup growth, becausesome pups are randomly euthanized. Assuming that litters are standardized on PND 4 (e.g., to 8or 10 pups), the method for calculating the mean litter proportion of postnatal survival is as follows:Postnatal survival between birth andPND 0 or PND 4 (prior to selection)(% per litter)=Σ(Viable pups per litter on PND 0or PND 4/No. pups born per litter)Number litters per group× 100© 2006 by Taylor & Francis Group, LLC

388 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPostnatal survival for allother intervals (% per litter)=Σ(Viable pups per litter at the end of interval N/viable pups per litter at the start of interval N)Number of litters per group× 100where N= PND 0 to 1, 1 to 4 (prior to selection), 4 (post-selection) to 7, 7 to 14, 14 to 21, or 4(postselection) to 21.Adverse effects on pup survival occur most frequently during the period prior to litter-sizestandardization (culling) on PND 4. Control group pup survival is generally high prior to culling,with group mean values greater than 90%. Following culling, survival is usually greater than 95%.Total litter losses occur infrequently in the rat. In the authors’ laboratory, approximately 1% ofdams that deliver have total litter loss. Therefore, a treatment group with just two total litter losseswould constitute at least a fourfold increase over historical control. While total litter loss isoccasionally observed in single litters in a control group, it is very rare to see two or more totallitter losses in a control group in a guideline study (N of approximately 25 to 30 females per group).Moreover, relative to the authors’ historical database, more than one total litter loss or mean litterpostnatal survival values below 90% in compound-treated groups are typically indicators of toxicityin a guideline study.4. Prenatal MortalityPrenatal mortality is assessed by comparing the number of former implantation sites in the dam(determined at necropsy) with the number of pups born to that dam. This endpoint is not so preciseas the viable litter size because of the uncertainty regarding the total number of pups born andbecause of the length of time passing from the implantation process to determination of the numberof implantation sites (3 weeks following parturition). Prenatal mortality remains a useful parameter;however, in the authors’ experience, there may be up to a 15% error in this parameter. Therefore,this endpoint must ultimately be evaluated relative to other reproductive parameters. A standardmultigeneration study includes approximately 100 litters per generation, and it is not feasible toobserve active parturition of each dam. As stated previously, parturition observations are generallyperformed two or three times per day for a study. Cannibalism can occur prior to identification ofthe total number of pups born within a litter. Because of the potential for cannibalism, the accuracyof prenatal mortality is not so great as the accuracy of the number of former implantation sites(determined by direct examination of the uterus of the dam following weaning). The term “unaccounted-forsites” should be used instead of postimplantation loss in reproduction studies. Thenumber of unaccounted-for sites for a dam is calculated as follows:Unaccounted-for sites =Number of former implantation sites – number of pups bornNumber of former implantation sitesThe number of unaccounted-for sites provides a good estimate of the prenatal mortality contributingto the perinatal mortality measured postnatally. Prenatal mortality generally occurs at alow incidence in the rat. Mean numbers of unaccounted-for sites greater than 1.5 per litter areusually an indicator of test article–induced prenatal mortality.5. Assessment of Sperm QualitySperm quality is assessed by determination of motility, concentration, and morphology. For assessmentof sperm motility (when collected from the epididymis or vas deferens, more aptly termed© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 389spermatozoon motility), the use of computer-assisted sperm analysis (CASA) systems has becomethe industry standard technology, although it is not required. The original methodology of manuallycounting cells by use of a hemocytometer and motility determination through video techniqueswill suffice, although the CASA system appears to be more accurate, precise, and cost-effectiveover the course of multiple studies. Based on user-defined criteria, the CASA system analyzessperm motion and determines total motility (any sperm motion) and progressive motility (netforward sperm motion). The CASA system uses stroboscopic illumination at 662 nm (a series oflight flashes of 1-msec duration at a rate of 60 Hz) to obtain precise optical images that are thenconverted into a digital format. The digital images are analyzed by processing algorithms todetermine the properties of sperm motion.A recent effort was undertaken to compare different techniques between multiple laboratories. 101There are no standards for sperm motility assessment; therefore, procedures and criteria vary amonglaboratories. It is thus incumbent on each laboratory to validate its CASA system under conditionsof use. The media used for dilution of the sperm sample and the incubation conditions must bevalidated to show that they do not affect sperm motility. In addition, gate, size, and intensityparameters on a CASA system must be adjusted to distinguish between motile and nonmotile spermand between sperm and debris. Determination of the percentage of progressively motile spermrelies on user-defined thresholds for average path velocity and straightness or linear index.Rat sperm can be obtained from the cauda epididymis or from the vas deferens. Mean controlgroup values for total rat sperm motility are usually 80% to 90%. In the authors’ laboratory, controlrat progressive motility is typically 10% lower than total motility. When assessing rodent spermmotility, one must be cognizant that the sperm being analyzed are not fully mature. Sperm collectedfrom the epididymis and vas deferens are not accompanied by the secretory fluids from the prostate,seminal vesicles, and bulbourethral gland that are added prior to ejaculation. Therefore, spermmotility and viability determined as part of the standard rodent evaluation may not precisely reflectthe actual sperm motility and viability of an ejaculated semen sample. However, ejaculated semencan be obtained and analyzed from the rabbit, dog, and primate. The advantages of using thesethree species instead of the rodent for sperm assessment are several: (1) longitudinal assessmentof sperm quality may be performed, (2) assessment of reversibility of adverse effects on spermquality may be conducted in the same animal, (3) baseline (pretest) data may be obtained, and (4)an absolute number of ejaculated motile sperm may be determined. Sperm assessment in speciesother than rodents is not conducted frequently. However, following guidance from regulatoryagencies, the authors’ laboratory has measured sperm motility in nonrodent species by use of aCASA system.Concentration of spermatozoa in the rodent is measured by homogenizing a testis or epididymisand then counting the number of homogenization-resistant sperm. The concentration values arenormalized to the weight of the tissue. Separate enumeration of spermatozoa in the testis andepididymis is necessary to determine the site of insult. If a test material adversely affects spermatogenesis,then both testicular spermatid and epididymal sperm numbers will be reduced. However,if sperm maturation is affected, testicular spermatid counts will be normal, while epididymalsperm counts will be reduced. Frequently, reductions in sperm concentration are accompanied byreductions in sperm motility and reproductive organ weights. If large reductions are observed inany of these parameters, male fertility rates may be reduced.Sperm morphology is determined qualitatively by light microscopy. Quantitative CASA measurementsof sperm morphology (primarily the head) can be made. However, because there are nopublished sources of the range of normal values for these measurements and the significance ofslight alterations in these measurements on fertility is not known, the majority of laboratories donot use them. The most common approach for morphologic assessment is to prepare a wet-mountslide and assess any abnormalities in the head, midpiece, and tail of the sperm. 102 Alterations includedetachment of the head from the tail, dicephalic heads, irregularly shaped heads, and coiled tails.A certain degree of subjectivity is involved in the morphologic assessment. By use of the criteria© 2006 by Taylor & Francis Group, LLC

390 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.35Historically observed mean spermatogenesis data by species and strain aRat Mouse RabbitEndpoint [Crl:CD ® (SD)IGS BR] [Crl:CD ® -1(ICR)BR] [Hra:(NZW)SPF]Sperm motility (%) 84.3 81.4 75.8Sperm production rate (sperm/g testis/day) 14.3 × 10 6 28.9 × 10 6 16.2 × 10 6Epididymal sperm concentration (sperm/g 470 × 10 6 616 × 10 6 638 × 10 6epididymis)Morphologically abnormal sperm (%) 1.1 1.8 4.1Source: Data from WIL Research Laboratories, Inc. historical control database.described by Linder et al., 102 morphologically abnormal control sperm are generally observed at alow frequency (

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 391spermatogonia is more likely to produce a permanent deficit in fertility than a compound that causesspermatid retention. Interpretation of testicular histopathologic data should be conducted in conjunctionwith assessment of sperm parameters, testicular weights, fertility data, and macroscopicdata. The testicular histopathologic examination may be the most sensitive indicator of toxicity,particularly when a lesion is observed to have been dose responsive. Presence of debris and sloughedcells in the epididymides can also provide indicators of adverse reproductive effects. Rarely areadverse reproductive effects manifested primarily in the prostate or seminal vesicles.Female reproductive organs that are usually weighed include the uterus and the ovaries; nonreproductiveorgans include the brain, pituitary, and thyroid gland. When evaluating uterine weight,it is essential to determine the stage of estrus of the animal at the time of necropsy. This is bestaccomplished by histopathologic examination of the vagina and uterus. A lavage of vaginal contentsmay also be taken prior to necropsy and examined by light microscopy. However, the stagedetermined from the vaginal lavage is approximately one-half day behind the stage determined byhistopathologic examination of the uterus and vagina. Therefore, results of histopathologic examinationsmay not be consistent with vaginal lavage data, particularly if the vaginal lavage isperformed several hours prior to the necropsy. Fluid retention and cellular proliferation vary greatlyin female rats, depending on the stage of estrus. Weights are highest during the estrus stage andlowest during the diestrus stage. Unless uterine weight is correlated to stage of estrus, false positiveand false negative interpretations may result.Reproductive organs subjected to histopathologic examination in the female include the ovaries,uterus, vagina, and mammary glands. Examination of these organs in combination with vaginalcytology data provides a determination as to whether the female is cycling normally, has enteredan anovulatory stage, or has entered premature reproductive senescence.7. Estrus Cyclicity and Precoital IntervalThe normal rat estrous cycle is four or five days in duration. For the original and extensive treatmentof this topic, refer to Long and Evans. 105 The techniques and principles for determining the estrouscycle have remained the same as in this original reference. However, the application and understandingof parameters of the estrous cycle have greatly improved in the last two decades. In theauthors’ laboratory, the estrous cycle is divided into the following four stages: estrus (E), metestrus(M), diestrus (D), and proestrus (P). During E, the majority of the cells present are cornifiedepithelial cells. These cells are irregularly shaped and may or may not have remnants of a nucleus.Some nucleated epithelial cells may be present; however, no leukocytes are present. Estrus usuallylasts 10 to 15 h. During M, nucleated epithelial cells are surrounded by increasing numbers ofleukocytes. Metestrus is the shortest stage, lasting approximately 6 h. At the beginning of D, celltypes are equally distributed among leukocytes, cornified epithelial cells, and nucleated epithelialcells. During the later part of this stage, the nucleated epithelial cells become more spherical inappearance. Diestrus is the longest stage, lasting two to three days. During P, the majority of cellsare nucleated epithelial cells, with some cornified epithelial cells that become more prevalent asthe E stage approaches. The nucleated epithelial cells appear spherical in shape and are usuallygrouped in clusters. Proestrus usually lasts 8 to 12 h.To determine the estrous cycle pattern in rats, vaginal lavages are performed once per day.Female rats will only be receptive for mating on the evening of the P stage. Therefore, animalsthat do not proceed through a cycle will not mate. It is not unusual for animals to occasionallyhave an extended cycle (six or more days); however, repeated extended cycles may be an indicatorof reproductive toxicity.A variety of methods are used to evaluate the estrous cycle. One approach is to determine thepercentage of days that an animal is in each stage of the cycle (i.e., the percentage of days in E,M, D, or P). Another approach is to determine the number of animals that have one or more irregularcycles (i.e., a cycle of six or more days or a cycle of three or fewer days). Lavages are usually© 2006 by Taylor & Francis Group, LLC

392 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONperformed only once per day. Therefore, one or more of the shorter stages (P, E, and M) may bemissed in a normal cycle, and this may confound evaluation of mean estrous cycle length. In theauthors’ laboratory, the average cycle length is calculated for complete estrous cycles (i.e., the totalnumber of returns to M or D from E or P). Estrus cycle length is determined by counting thenumber of days from the first M or D in a cycle to the first M or D in a subsequent cycle. In thisapproach, it is assumed that if an animal has a cycle pattern of P followed by an M or a D, theanimal actually went through E, but the estrus stage was missed because vaginal cytology wasexamined only once per day.The composition of the cell types and the movement through the estrous cycle is hormonallycontrolled. During D, estradiol levels are low and gradually rise because of increased secretion bythe growing follicles. This results in vaginal epithelial cell proliferation and the transition of thevaginal cytology into P. The elevated levels of estradiol trigger a surge in prolactin and luteinizinghormone. The surge in LH then triggers ovulation early in the morning of E. Serum estradiol levelsdecline dramatically on the evening of P as the follicle begins to luteinize and vaginal epithelialcell proliferation declines. Estrus cycle length is also dependent on progesterone secretion by theovarian follicles and corpora lutea. In the rat, the hypothalamic regulation of the LH surge inresponse to elevated levels of estradiol is lost when an aging female enters reproductive senescence.As a result, ovulation does not occur, and the levels of estradiol remain persistently elevated.Evaluation of vaginal cytology in this condition reveals a persistent E state. The elevated estradiollevels stimulate prolactin release, producing mammary duct ectasia, and ultimately, galactoceles.If this normal aging process is accelerated in test article-treated groups, fertility would be affected,and this would be an indication of reproductive toxicity. An array of chemicals and drugs has beenreported to cause this state of premature senescence through similar mechanisms. In a state ofpersistent D, estradiol levels are low, and follicular development and ovulation do not occur, againresulting in reduced fertility.The precoital interval is defined as the number of days from the initiation of cohabitation of abreeding pair until evidence of copulation is observed. As previously stated, female rats will onlybe receptive to the male on the evening of P. Therefore, precoital interval data should be evaluatedin conjunction with the estrous cycle data. For example, an increase or decrease in the precoitalinterval for an exposure group compared with the control group is not an adverse effect if theanimals in the exposure group are cycling normally and mate on the evening of the first P followingthe start of the breeding period. However, an increase in the precoital interval can provide weightof-evidencesupport for identifying slight effects on estrus cyclicity. Investigators also need to becognizant that precoital intervals cannot be calculated for animals that do not cycle or mate.Therefore, this endpoint cannot be assessed in isolation without also evaluating estrous cycling andthe mating index. Improved sensitivity of detecting effects on the precoital interval and conduct ofstudies to evaluate further such effects may be made more cost-effective by using the “shortenedmating period” methodology. 56 In this method, the mating period may be reduced to a 2-h periodby the use of perineal manipulation and its elicitation of the ear-quivering response.8. Mating and Fertility Indices and Reproductive OutcomeTraditionally, assessment of functional aspects of reproduction, such as calculation of mating andfertility indices, has been an important cornerstone of well-designed studies and has served a keyrole in the identification and characterization of injury to the reproductive system. 40 Table 9.36presents the historically observed range of values for male and female mating and fertility indicescalculated from studies conducted in the authors’ laboratory.The mating index is a mathematical expression derived from post facto observations confirmingthat coitus has occurred. The index is generally considered a reliable measure of normal sexualbehavior and libido. It also provides indirect information relative to the function and status of thehypothalamic-pituitary-gonadal axis. In other words, changes in this index may arise from disturbance© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 393Table 9.36Historically observed functional reproductive indicesfor [Crl:CD ® (SD)IGS BR rats] aMean(%)25thQuartile (%)75thQuartile (%)Male Mating IndexNo. of males with evidence of matingTotal no. of males used for matingFemale Mating IndexNo. of females with evidenceof mating (or confirmed pregnancy)Total no. of females used for mating× 100 96.1 95.0 100.0× 100 97.3 95.9 100.0Male Fertility IndexNo. of males siring a litterTotal no. of males used for mating× 100 89.6 86.7 95.0Female Fertility IndexNo. of femaleswith confirmed pregnancyTotal no. of females used for mating× 100 90.9 88.0 96.0Source: Data from WIL Research Laboratories, Inc. historical control database.in the CNS or from imbalances in hormonal cascades. Indirect measures of mating performanceare obtained by observing either the presence of sperm from a vaginal lavage (microscopically) ora copulatory plug in the vagina or below the mating cage. The latter observation technique mayalso be used for the mouse. However, results are more variable, and the endpoint becomes lesssensitive because retention of the vaginal plug is longer and mistiming of pregnancy may be aconsequence. The process of confirming that coitus has occurred is confounded because, evenamong the most technically skilled, the rate of proficiency of timing the event by vaginal lavageis only 90% to 93%, in the authors’ experience. Thus, up to 10% of the data for a given breedingcycle may be unavailable for the final interpretation. For this reason, differences in the matingindex that represent one or two fewer successful matings in a group should not cause an increasedlevel of concern. However, decrements in this endpoint may also signal that abnormalities havebeen introduced. Such abnormalities may consist of changes in the duration and/or the orderlysequence of stages of the rodent estrous cycle, or they can also arise from deficits in sperm quality.These decrements may occur in the presence of disturbances in the elapsed time between the startof cohabitation and observed evidence of mating (referred to as the precoital interval). The precoitalinterval is usually reported as a fraction of the number of days required for mating. Most rodentswould be expected to show evidence of mating within the first estrous cycle (4 or 5 days). It isnoteworthy that normal precoital intervals may occur without successful impregnation if effects onintromission are observed. In addition, the females are rarely synchronized at the onset of thebreeding period in terms of stage of estrus, and statistical transformations may be necessary toaccount for these variations. In essence, to determine biological significance of differences in groupmean data of less than two days between treated groups and the control group, one must includea detailed assessment of the individual data.On rare occasions, an unexpectedly high number of females may show evidence of copulationon the first morning after pairing. This is thought to be a result of forced copulation of females© 2006 by Taylor & Francis Group, LLC

394 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthat are not in the receptive phase of the cycle (proestrus). A reduction in the number of successfulpregnancies can be expected in this instance. 106The fertility index is the number of animals inducing pregnancy or becoming pregnant dividedby the number of mating sets, and it is the primary measure describing the outcome of a matingtrial. In the rodent, several intromissions must occur prior to successful conception. These intromissionstrigger uterine contraction, enhanced sperm transport, and hormonal changes that prepare theuterus to accept the blastulae. When the fertility index is evaluated in isolation from other endpoints,it is generally considered a poor predictor of perturbations in the reproductive system. Therefore,it is essential to evaluate other interrelated endpoints (e.g., organ weights, macroscopic and microscopictissue changes, estrous cycle data, and oocyte quantification) to provide initial insight intothe processes of the reproductive system that have been compromised. Some reproductive toxicantsare so selective in their actions that a full characterization of the response is only possible by meansof specially designed studies in which “pulse doses” are administered acutely over discrete phasesof the reproductive cycle. At the other extreme are reproductive toxicants for which a decline inthe fertility index is the most sensitive endpoint. The best example of the latter can be found in areport produced by the Center for the Evaluation of Risks to Human Reproduction (CERHR) onthe reproductive and developmental toxicity of 1-bromopropane (1-BP). 107 A study of the reproductivetoxicity of 1-BP was conducted via the inhalation route at the authors’ laboratory underthe OPPTS 870.3800 guideline. Reproductive performance was highly compromised in this studyover a broad range of exposure concentrations (100, 250, 500, and 750 ppm). In the first parentalgeneration (F 0 ), the blockade of pregnancy was complete at 750 ppm (0% fertility index). At thehigher exposure concentrations (over 250 ppm), numerous reproductive endpoints were affected(generally statistically significantly), and they included the following: increases in estrous cyclelength and ovarian follicular cysts; and decreased ovarian weight, numbers of corpora lutea, fertilityindices, and litter size; and/or changes in sperm quality. At the lowest exposure concentration, 100ppm, the only alteration was a decrease in the male and female fertility indices (greater than 20%).The decreased fertility indices were interpreted to constitute the most sensitive indicator of reproductivetoxicity for 1-BP, judged by the dose response expressed. As an apical measure, the fertilityindex may reflect cumulative, immeasurable decrements in many subordinate processes and unmeasuredendpoints (e.g., sperm capacitation and tubal transport). The CERHR Expert Panel did notconsider the diminution in this key functional endpoint to be an effect. Presumably, the finalinterpretation of the lowest-observed-adverse-effect level was arrived at merely on the basis of lackof statistical significance, rather than from comparisons to normally expected ranges (the laboratory’shistorical control data).9. Duration of Gestation and Process of ParturitionThe length of gestation is a derived value as determined through these regulatory study designs,which compares the elapsed time between fertilization and initiation of parturition. In theCrl:CD ® (SD)IGS BR rat, the average length of gestation is 22 days, but it ranges from 21 to 23days postcoitus. Near the period of expected parturition, the dams are observed several times dailyto establish, as closely as possible, the onset of parturition. The mean data are generally reportedas a fractional value, with units expressed in days. The duration of gestation is highly predictablein control populations and has been well characterized in a variety of laboratory animal modelsand stocks. Exposures that result in premature birth yield pups that have lower body mass becauseof the decreased gestation length. These low-weight litters are underdeveloped and at increasedrisk for mortality. The greater the prematurity, the more pronounced are the effects on developmentand survival. Conversely, prolongation of the gestation period may induce other adverse sequelae,including maternal death, increased pup mortality, and other confounded measures. In the mostsevere instances of delayed parturition, mortality rates among the dams may be excessive becauseof possible septicemia associated with the degenerating products of conception and other related© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 395major events such as circulatory collapse. The process of parturition is usually completed within1 to 2 h in rats, with longer times expected for larger litters. Dystocia may be a manifestation ofmaternal toxicity resulting from exposures that cause prolonged labor, difficult labor, or terminationof the contractions necessary to evacuate the uterus. Pharmaceutical agents, such as those antihypertensivesthat evoke blockade of calcium channels, may also directly cause dystocia. 108 Dystociais a relatively rare event in the control population in standard laboratory strains; in the authors’laboratory, dystocia has been observed in less than 0.5% of Crl:CD ® (SD)IGS BR rat litters (7occurrences in 1905 litters from 1997 through 2003). Therefore, the appearance of dystocia inexposed animals should always trigger a heightened level of concern.10. Endpoints Assessed during Lactation, Nesting, and Nursinga) Endocrine EndpointsNear the end of gestation, rapid changes in maternal endocrine levels occur to prepare the mammarygland for the functional demands of nurturing and supporting the offspring. The milk ejection reflexin the rodent is triggered by the suckling stimulus and involves the release of oxytocin from storesin the pituitary gland. This action, in turn, induces proliferative changes in the mammary tissueand stimulates milk production and release. Agents, or their metabolites, that cause disturbancesin these pathways can also cause adverse changes in milk composition and/or milk quantity.Likewise, some chemical agents or their metabolites that enter the milk from the maternal circulationmay suppress postnatal development and maturation. Adverse effects on the growth of nursing pupsmay also be caused by factors unrelated to the dam. For example, agents that disrupt the cascadeof developmental events necessary for palatal closure can result in poor suckling and retardedgrowth or mortality in offspring. The support of lactation and nursing behavior may be dependenton audible cues emitted from the pups. 109,110 Agents that interfere with these cues may increasematernal neglect and subsequent postnatal wastage.b) Blood and Tissue DosimetryDirect measurements of milk production, milk composition, and dispersion of parent drugs and/ormetabolites are rarely conducted in standard guideline studies. Figure 9.9 amplifies key considerationsin the design and interpretation of studies that include bioanalytical assays of variouscomponents of the blood and tissues following maternal exposure during the reproductive cycle.Blood and tissue dosimetry in the postnatal animal may also establish and clarify exposure of theprogeny to the test compound. Ensuring exposure of the progeny to the test compound during thelactational period has received increasing interest and support, and is more frequently included insafety assessment studies. This approach also provides the strongest basis for designing studieswith improved relevance to protection of children’s health.Figure 9.9 is a schematic depiction of the comparative modes of translocation of materialsbetween the maternal systemic circulation, placenta, and embryo or fetus vs. the maternal systemiccirculation, mammary glands, and neonate. This comparison demonstrates bidirectional fluxes ofmaterials (endogenous and xenobiotic) between the maternal circulation, the placenta, and theembryo or fetus. On the other hand, unidirectional translocation of like materials occurs betweenthe maternal circulation, the mammae, and the neonate. In the former case, flux between the embryoor fetus, placenta, and maternal circulation is intravenous (IV) in nature, with the only unidirectionalabsorptive activity occurring in the placental membranes, either chorioallantoic or yolk sac. In thelatter case of maternal circulation to the mammae and neonate, translocation is secretive and entailsgastrointestinal absorption by the neonate. Gastrointestinal absorption involves dynamic acquisitionof specific absorptive functions and maturation of the overall gastrointestinal physiology over theperiod of postnatal development. Figure 9.9 illustrates that for estimation of prenatal exposure of© 2006 by Taylor & Francis Group, LLC

396 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPrenatalTreatmentPostnatalEmbryo/Fetus Placenta Mother Mammae NeonatePrenatalPostnatalDrug transfer to offspring Nearly all transferred Apparent selectivity (“barrier”)Drug levels in offspring C max and AUC measured Not routinely measuredMaternal blood versusoffspring levelsExposure route tooffspringMaternal level often a surrogateModulated IV exposure via placentaMaternal levels probably NOT a goodpredictorOral via immature gastrointestinal tractCommentary Timing of exposure is critical Extent of transfer to milk and neonatalbioavailability is key to differentiatingindirect (maternal) effectsFigure 9.9Comparative modes of translocation of materials between the mother and the offspring.MaternalToxicresponseDevelopmentalLog of doseFigure 9.10 Hypothetical relationship of the log of dose and toxic response for maternal and developmentaltoxicity.the products of conception, sampling of maternal blood and/or plasma concentrations of a xenobioticis usually a good dosimeter. Hence, sampling the maternal blood for estimation of exposure of theembryo or fetus prenatally suffices as a surrogate for direct sampling of the embryo or fetus, whichmay not be feasible because of limitations of sample sizes and analytical sensitivity. Conversely,measuring levels of a xenobiotic in maternal milk is a poor dosimeter of exposure in postnatalprogeny because of the secretory nature of milk production and the immaturity of the developinggastrointestinal tract. Therefore, the best estimate of postnatal exposure is made by sampling directlythe blood and/or plasma of the progeny. In both instances, prenatal and postnatal, these toxicokineticdeterminations are crucial to examining relative sensitivities of the adult vs. the developing offspring.The latter determination is crucial to differentiating relative biologic susceptibility fromdifferences in exposure of the two life stages.Without the toxicokinetic information, plotting and comparison of the dose-response curves fordams vs. progeny (prenatal or postnatal) are based solely on the less sensitive, more subjective,and less direct toxicologic observations routinely recorded in these reproductive toxicity studies.Figure 9.10 further depicts this very important relationship by displaying two hypothetical logdose-responsecurves. The placement and slopes of the curves for maternal versus developmental© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 397(prenatal or postnatal) responses might suggest a biologically based or mediated difference insensitivity of the two organisms. However, the difference between the curves may be merely dueto differences in exposure. This is an important dimension to the overall issue of interpretation ofindividual parameters in these bioassay studies, as well as to the overall processes of risk assessmentand risk management. Measurement of the levels of a xenobiotic in maternal milk is better thanno information; however, it is not the best indicator of postnatal exposure of the progeny. For earlyembryonic development in which embryonic weight is in the microgram wet-weight range, maternalblood levels of a xenobiotic remain the most practical and reasonable estimate of prenatal exposure.c) Milk CompositionGenerally, the presence or absence of nursing is evaluated qualitatively by observing milk in thestomach of the pups. This is possible in early postnatal life because the abdominal tissues overlyingthe stomach are translucent at that time, and the absence of fur allows for visualization. Observationsof decreased milk content in pups that die early in postnatal development may signal an adverseeffect in nesting and nursing behavior. However, this observation provides little insight relative tothe mode of action of the agent to which the pups are exposed. While quantitative evaluation ofmilk composition, milk quality, and xenobiotic kinetics by means of hormonal induction and manualexpression techniques may be accomplished in guideline studies, these tests are not routinely carriedout.Unlike the breadth of knowledge that has been developed describing changes in the physiologicstatus of the prenatal rat, there is considerably less data characterizing the progressive changes inmilk composition that occur from birth until weaning. 111 Therefore, quantitative evaluation of thefollowing endpoints may be important considerations for further study: potassium, lactose, sodium,serum albumin, transferrin, and casein. Consistent with the human, increases in the concentrationof potassium and lactose and decreases in sodium concentration begin from 12 h before birth to48 h postpartum in the rat. These events correspond to the transition from the production ofcolostrum to that of milk. Serum albumin in the rat increases by a factor of two within a few daysafter delivery, but it does not increase in the human. Levels of transferrin in rat milk are low fromparturition until lactational day (LD) 10, when marked elevations occur that are sustained throughweaning. Casein is present in milk at term, and the concentration increases steadily until weaning.In response to the termination of nursing at the end of the postnatal period (LD 20), the levels ofboth lactose and potassium decline dramatically, while sodium content rises.When conducting a reproduction study, it is important to recognize and observe innate behavioralcharacteristics of rodents as they care for their progeny because alterations in these behaviorsmay represent an overt change in the CNS status of the mother. Rodents, particularly mice, areexpert nest builders and constantly strive to maintain an orderly arrangement of the pups to enhancesuckling. Suckling occurs frequently over both the light and dark cycles. Rodents will also immediatelyretrieve pups removed from the nest.d) Maternal Physiological EndpointsIt is important to compare physiologic endpoints in maternal animals, such as body weight duringgestation and lactation, to those obtained during the premating phase. This can help to determinewhether unique toxicities have occurred under the conditions of pregnancy and/or lactation. Foodconsumption will be elevated during gestation; this is due to elevations in circulating levels ofprogesterone. On the other hand, the rapidly changing physiologic status of the dams and the fluxin hormones from the ovary and thyroid during pregnancy, coupled with rapid weight gain, tendto confound the utility of organ weight relative to body weight comparisons. Likewise, otherconfounders in reproductive toxicity studies can involve flaws in the features of experimental designrelative to administered dose. For example, the traditional approach of correcting individual doses© 2006 by Taylor & Francis Group, LLC

398 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONfor gavage delivery based on current body weights should be refined for advancing pregnancy. Ifnot, excessive overdosing of the maternal animals can induce profound effects during the periodof expected parturition, when the mass of the products of conception (but not necessarily the leanbody mass of the maternal animal) are maximal.11. Sexual BehaviorWhile it is clear that normal sexual behavioral function is dependent on the complex integrationof input from several organ systems, little direct evaluation is devoted to this important measurein animal studies. This void in the data is often created by practicality at the operational levelbecause of the nocturnal breeding habits of rodents. In females, the most direct measure of sexualbehavior involves the observation of lordosis to confirm receptivity, and this behavior is tied to thetiming of late proestrus. Indirect measures of this endpoint are collected and evaluated as describedin previous sections.In males, periods of mounting and intromission can be quantified and may be included asexperimental endpoints if class effects are anticipated. The timing and number of intromissionsand release of ejaculatory plugs in the rat model can be used to assess potential effects on sexualbehavior. The extent to which these endpoints relate to human sexual behavior is unclear, however,because the authors are unaware of any studies of the concordance or validity of such measures.From a basic hazard identification and/or screening perspective, these measures may have valueand as such are commonly published in the neurobehavioral literature. Nevertheless, sexual behaviorendpoints have been given little attention in the guidelines for reproductive toxicity studies. Indeed,the presence of ejaculatory plugs in cage pans may be a useful observation, but this parameter isnot commonly reported in the toxicologic literature.Rats have a specialized anatomical structure, termed the coagulating gland, which rapidlyproduces an ejaculatory plug after coitus. The function of this structure is to provide a physicalbarrier to prevent leakage of the seminal plasma from the vagina. Classically, little attention hasbeen directed at quantifying changes in the frequency of these emissions in guideline reproductivetoxicity studies. In the authors’ laboratory, several guideline two-generation studies have revealedqualitative and dose-related increases in the number of ejaculatory plugs per animal in comparisonwith values from concurrent control animals. Interestingly, the response was expressed severalweeks before the onset of the first generation mating cycle. This effect was initially encounteredin inhalation studies using whole-body inhalation exposure regimens. More recently, it has beenobserved occasionally in studies utilizing more common routes of exposure. Several authors studyingsexual behavior in rats have reported that rats have daily spontaneous ejaculations. 112–114 It hasalso been reported that rats normally lick their genitals at the time of ejaculation and consume theplugs. 115 It is conceivable that toxicants could be eliminated from the body in these structures, andpoor palatability might preclude consumption. Thus, qualitative differences in the number ofejaculatory plugs observed in these studies may have no origin in neuroendocrine dysfunction.Nevertheless, the increased production of the ejaculatory plugs in the aforementioned studies hadno impact on the fertility of the males, and the relevance of these findings to human risk assessmentis unknown.12. Landmarks of Sexual MaturityThe primary landmarks of sexual development are anogenital distance, balanopreputial separation(BPS) in males, and vaginal patency (VP) in females. In rodents, anogenital distance is greater formales than for females. Anogenital distance (AGD) is typically measured at birth, and reflects thelinear distance between the genital tubercle and the anus. Mean values for males and females arenot presented from the authors’ laboratory because techniques for measuring AGD vary from onelaboratory to another, based on anatomical reference points. It is critical that consistent procedures© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 399Rear ViewAnusMeasurementAnusGenital TubercleGenital TubercleTowardsTailMaleAnusSide ViewGenitalTubercleFemaleTowardsHeadMeasurementMeasurement is made from the caudal margin of the anus to the caudal margin ofthe genital tubercle.Figure 9.11 Measurement landmarks of anogenital distance.be used within a single laboratory for this measurement so that observer bias is controlled. Adiagram of the technique employed at the authors’ laboratory is presented in Figure 9.11.The skin should be stretched to facilitate consistency of the measurement. In addition, thetechnician should examine equal numbers of pups within each treatment group. The absolute AGDis directly correlated with body weight (i.e., pups with greater weight have larger AGDs). Whenstatistically analyzed, the AGD should be normalized relative to the cube root of the body weight. 116© 2006 by Taylor & Francis Group, LLC

400 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe most frequently observed adverse effect on the AGD is a reduced distance in males in responseto antiandrogenic agents or 5α-reductase inhibitors. When mean litter anogenital distances ofapproximately 20 litters are evaluated, differences of 5% or greater are generally indicators ofreproductive toxicity.BPS is evaluated by expressing the penis to determine if the prepuce can be retracted from theglans penis. An animal is not considered positive for this endpoint unless complete retraction ofthe prepuce can be performed. BPS is mediated by androgens, and males that do not achieve BPSwill be infertile. 117 In control Crl:CD ® (SD)IGS BR male rats in the authors’ laboratory, the meanday of acquisition of BPS is PND 44.8 (the range is 41.6 to 49.0). Because body weight caninfluence the timing of BPS, it is important to record the weight of the animal on the day BPS isobserved. When growth retardation (as measured by reduced body weight) is observed, the day ofacquisition for BPS is likely to be delayed. 117,118 Preputial separation in males occurring in parallelwith delays in vaginal patency in females indicates a generalized growth delay, as opposed to aselective endocrine-mediated mechanism. In the absence of an effect on body weight, a change inthe day of acquisition of BPS of 2 days or greater is typically an indicator of toxicity.VP is evaluated by examining the vagina to determine if the septum covering the vagina hasruptured. VP generally occurs around the time of first ovulation and occurs in response to anincrease in serum estradiol levels as females enter puberty. A female is not considered positive forthis endpoint unless the septum is completely absent. Occasionally, a thin strand of tissue willremain over the vagina. This threadlike strand would not preclude an animal from mating; as such,the animal would still be considered to have achieved VP. In control Crl:CD ® (SD)IGS BR femalerats in the authors’ laboratory, the mean day of acquisition of VP is PND 33.5 (range is 31.7 to38.8). As described for BPS, body weight can be a confounder in interpreting the timing of VP. Inthe absence of an effect on body weight, a change in the day of acquisition of VP of 2 days orgreater is typically an indicator of toxicity.13. Sex Ratio in ProgenySeveral publications have reported sex-ratio effects of organochlorines (e.g., 2,3,7,8-tetrachlorodibenzo-p-dioxin[TCDD]) on human populations, although cross-study results have not been consistent.Lowered offspring sex ratio has been associated with paternal exposure to TCDD followingan accidental exposure at Seveso in northern Italy in 1976 (0.38 male to female ratio), 119 and fortwo cohorts of pregnancies fathered by Russian workers at a pesticide-producing factory in the cityof Ufa, Bashkortostan, Russia, from 1961 through 1988 (0.40 male to female ratio). 120 The Sevesodata indicated a greater effect in men that were exposed before the age of 19; however, no similarcorrelation could be made for the Ufa cohorts. In contrast to these studies, conflicting results werefound in a study of paternal occupational exposure to TCDD in factories producing Agent Orangein the United States. In this study, serum samples collected in 1987 revealed no differences inoffspring sex ratio between pregnancies fathered by workers with high levels of TCDD exposureand those of nonexposed neighborhood referents. 121In contrast to the above-mentioned putative human examples, selective effects on survival of aparticular gender are rarely observed in animal hazard identification studies. In the authors’ experiencewith approximately 1500 studies, a confirmed test article–related shift in sex ratio has neverbeen observed. Conceivably, alterations could arise from direct effects on the male germ cellpopulations that would favor the production or destruction of Y or X chromosome-bearing sperm.An alternate hypothesis would involve direct insults on the male or female conceptuses; however,the plausibility of this phenomenon is not fully established. Mild shifts in the male to female sexdistribution ratio often occur, and little significance is focused on values that fall within normallyexpected ranges (44% to 56%, based on the authors’ historical control data for Crl:CD ® (SD)IGSBR rats).© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 40114. Functional Toxicities and CNS MaturationThe study of immediate onset and latent behavioral expression of CNS damage associated withexposures during prenatal and early postnatal development has been accorded a higher level ofconcern in recent years. Tests to evaluate these aspects of development are routinely performedduring the ICH pre- and postnatal study of development, 122 as well as under the more robust andfocused EPA and OECD developmental neurotoxicity protocols. 123,124 A plethora of standard testingparadigms have been proposed, validated, and/or used in the process of identifying and characterizingagents that may represent a threat to the developing nervous system. The reader is referredto Chapter 8 of this text for additional exposition.15. Oocyte QuantitationIn cases where female fertility is reduced, oocyte quantitation may provide information regardingthe mode of action of an agent. Oocyte quantitation is used to assess whether the follicle maturationprocess in the ovary is occurring normally. 125,126 Three major types of follicles may be quantified:small or primordial, growing, and antral follicles. Small or primordial follicles contain isolatedoocytes or oocytes with a single surrounding layer of granulosa cells. Growing follicles have anoocyte surrounded by multiple layers of granulosa cells. Antral follicles have a central oocyte withina fluid-filled space bordered by hundreds of layered granulosa cells. For quantifying the folliclenumbers, serial sections of the ovary are usually cut approximately 100 µm apart. By cutting theserial sections 100 µm apart, the same primordial and growth follicles will not be counted multipletimes. Many laboratories combine the primordial and growing follicles when reporting oocytenumbers. Counting antral follicles is more problematic. Because these follicles have cross sectionsof 100 to 550 µm, the antral follicles will be counted multiple times when sections taken 100 µmapart are examined. Therefore, quantification of antral follicles is typically not performed. Thereis another potential confounder in quantifying the small and growing follicles. For animals that areanovulatory (having fewer numbers of corpora lutea and antral follicles), increased numbers ofprimordial and growing follicles will often be recorded because the larger follicles are not presentto obscure some of the small and growing follicles. In the authors’ experience, changes in thenumbers of small and growing follicles are rarely observed.A. IntroductionVII. EVALUATION OF RARE (LOW INCIDENCE) EFFECTSThis section addresses the most critical, long existing, and often overlooked or misunderstood issueof rare event (low incidence) developmental and reproductive findings. Rare events may be causedby a xenobiotic or may be spontaneous. The difficulties in determining whether a small increasein response is related to treatment or is a chance occurrence are not straightforward but arecommonly unappreciated. To the best of our knowledge, this is the first published attempt to fullycharacterize the attendant issues and recommend a paradigm for resolution.The daunting task for most scientists working in this discipline is how to distinguish betweenthose findings occurring spontaneously and those produced by a xenobiotic. The limited statisticalpower of the bioassays, relative to the large potential number of human beings exposed, dictatesthis to be a nonstatistical exercise and one that must rely on probabilistic theory, intuition fordevelopmental processes and patterns, and arrays and timing of effects.The material in this section was initially developed for a presentation at the 2002 WinterToxicology Forum. 127 At that meeting of largely FDA and private-sector scientists, rare (lowincidence) events and the difficulties they present for risk assessment, product licensure, and labeling© 2006 by Taylor & Francis Group, LLC

402 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONwere the topics of interest because of recent experience with the pharmaceutical class effects ofinhibitors of the cyclooxygenase enzymes (COX inhibitors). Although there have been manyprevious examples of rare-effect issues, most have gone unnoticed or unpublicized, and not untilthis major class effect was identified was the topic given any focused regulatory attention. Becausethis issue is so important and intrinsic to developmental and reproductive toxicity bioassays, thissection of the chapter was modified from the Toxicology Forum presentation in which both thetopics of developmental and reproductive effects, and cancer were addressed. It is important tounderstand that in developmental and reproductive toxicity studies, most responses are likely tomanifest at substatistical rates. For a review of the COX-2 rare events evaluation, see Cook et al. 128In this section, examples from the authors’ historical control database illustrate the importanceof accruing and managing consistent control data (over time and based on good animal husbandryand techniques) to assist in the interpretation of results. The advantages of having a highly skilledand stable work force are also discussed. A few case studies are included to demonstrate how datawere interpreted in selected controversial study outcomes. Finally, a stepwise approach to evaluatingrare findings and low incidence occurrences is presented.B. Scope of the ProblemThe goal for regulators is to gain knowledge of whether a condition, procedure, or chemical ordrug exerts adverse effects on reproduction or development. Their mandate is to evaluate risk tohumans with a sufficient degree of confidence to enable sound decision-making. This is accomplishedby critically reviewing results of animal safety studies and then extrapolating those resultsto humans. A high degree of concordance has been shown between human adverse effect scenariosand what can be demonstrated in rodents (as well as in the rabbit). 11,129–132 However, when theincidence of a finding is too low for statistical identification or confirmation, it becomes moredifficult to separate an adverse effect from normal biological variability. For example, what doesa ventricular septal defect in one high-dose group fetus (out of 150 to 325 fetuses evaluated) signifyin terms of potential human effects? Likewise, what is the human hazard of two to three rodentswith total litter loss during the lactation period?The FDA has defined rare incidence as “…an endpoint that occurs in less than 1% of the controlanimals in a study and in historical control animals.” 5 This definition, from a statistical point ofview, has an inherent problem. Animal safety studies are not (per guideline) designed with such adegree of statistical power (number of animals) that the statistical significance of 1 in 100 occurrencescan be detected. Furthermore, upon evaluation of historical control data, responses for almostall of the critical developmental defects are seen to occur at a frequency of less than 1 in 100.Table 9.37 provides a comparison of overall spontaneous malformation prevalence in differentspecies. 133 The data demonstrate the challenge of detecting rare events in safety testing. In additionto the rare event–related statistical issues, incidences of malformations are much lower in laboratoryanimals than in humans, further complicating extrapolation. The background incidence of spontaneousmalformations is particularly low in the rat. Because of this, the rat is a sensitive model fordetecting rare events.Concern for two to three malformations per group in animal studies is appropriate, even thoughthere may not be statistically significant differences from the concurrent control group for theincidences of individual or total malformations. The challenge is how to differentiate the effect ofa test agent from mere biological variation.An example of the challenge inherent in evaluating low incidence findings occurred in a recentstudy in the authors’ laboratory in which athyrosis was noted upon macroscopic examination in ahigh-dose group Crl:CD ® (SD)IGS BR rat fetus (this malformation had not been seen previouslyin approximately 80,000 control and test fetuses of this rat strain). Several other fetuses distributedin multiple litters in the high-dose group had enlarged or rudimentary thyroids, and cleft palatewas observed at 4.5% per litter in that same dosage group. Although athyrosis occurred in only© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 403Table 9.37Spontaneous malformation rates invarious species aSpecies Mean (%) Range (%) Fetuses (N)Rat 0.33 0–1.6 9643Mouse 1.2 0–3 5207Rabbit 3.2 0–10 4708Dog 5.5 5.3–5.7 167Human 4 3–9 Multiple surveysSource: Data for rat, mouse, rabbit, and dog malformationrates originate from the historical control databaseat WIL Research Laboratories, Inc. Human data wereobtained from the March of Dimes (2003) and personalcommunication from Ken Jones (2003).Table 9.38Selected malformation rates in control fetuses(means of study means)Rat Incidence Rabbit IncidenceMalformation (% per Litter) a (% per Litter) bVentricular septal defect 0.00 0.02Cleft lip or palate 0.02 0.04Abdominal wall defect (including gastroschisis) 0.04 0.06Hydrocephaly 0.03 0.20Spina bifida 0.00 0.17Renal agenesis 0.01 0.02Diaphragmatic hernia 0.00 0.04Retroesophageal aortic arch 0.01 0.01a61 Crl:CD ® (SD)IGS BR rat developmental toxicity studies conducted at WIL (1998 to 2003)b81 Hra:(NZW)SPF rabbit developmental toxicity studies conducted at WIL (1992 to 2003)one fetus, this finding was considered an adverse effect in view of the following: (1) the rare natureof the finding, (2) other clear signals of structural malformation (cleft palate) at the high dose level,and (3) other alterations in fetal thyroid structure, revealing a pattern of insult. The relationship totreatment of a single case of athyrosis occurring independently would have been much less clear.To further illustrate the challenge of discerning the significance of rare event findings, a fewexamples of malformations that have occurred as rare events or low incidence findings in numerousscenarios are listed in Table 9.38. Among the 21,000+ control Crl:CD ® (SD)IGS BR rat fetuses inthe authors’ historical control database, ventricular septal defects, spina bifida, and diaphragmatichernia have never been detected. However, spina bifida and diaphragmatic hernia have beenobserved in treated fetuses. Because any occurrence of a malformation is such a rare event, therelationship between the agent tested and the occurrence of a malformation is difficult to discern.On the other hand, the low background incidence of malformations makes it easier to distinguishsubtle teratogenic events in those cases where the agent causes only a few malformations, givenappropriately rigid and consistent historical control values. In the absence of such standards, thehistorical control database will be of limited use for identification of teratogens, and interpretationof the findings becomes even more confounded and without a reliable basis for comparison.Further complicating matters is the interdependence of endpoints. The developing embryo orfetus, for example, is entirely reliant upon the maternal milieu for proper growth and development.Maternal stress has been shown to cause cleft palate in mice. 70 Antibiotics may produce embryoor fetal lethality and/or cause abortion in rabbits by adversely affecting the maternal gut flora, 75,134and the rat embryo and fetus has been shown to depend upon maternal zinc for proper development.135 Therefore, determining whether a fetal malformation is caused by the maternal condition© 2006 by Taylor & Francis Group, LLC

404 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONor is a direct effect of the test agent presents its own challenges and is discussed in detail in Chapter4 of this text.C. Criticality of Accurate and Consistent Historical Control DataA comparison with the laboratory’s historical control data should be a first step toward determiningwhether a small increase or decrease (not statistically significant) in an endpoint constitutes atreatment-related effect. To illustrate this point, control mean values and ranges from over 1300litters for selected reproductive endpoints are shown in Table 9.39. These data originate from theCrl:CD ® (SD)IGS BR rat historical control database at the authors’ laboratory. Since inception ofcollecting these data for this animal model (1996), endpoints such as mean viable litter size, survivalbefore or on PND 4, total litter loss, and newborn pup weights have been found to be very consistentwhen coupled with good animal husbandry. Occurrences above or below means and ranges of theseendpoints may, therefore, indicate the threshold of a positive signal for a treatment-related effect.Live litter size and early pup viability (before PND 4) are sensitive endpoints. Neonatal deathsmay be small as a mean percentage, and a death rate exceeding approximately 5% per group shouldbe closely evaluated for a potential treatment-related effect. Likewise, a decrease in PND 1 pupweight below a mean of 6.5 g is most likely past the threshold of a positive signal. Finally, intoday’s modern facilities with good animal husbandry and highly skilled technicians, total litterloss for one or two dams in the same group probably constitutes a treatment-related effect, evenin the absence of statistical significance. These examples demonstrate the importance of creatinga historical control database and using it consistently when interpreting overall results from individualstudies.D. Case StudiesThe following discussion examines scenarios involving three different compounds that demonstratethe difficulties with rare event data interpretation and consequences that may occur if the signalsare missed in safety testing.1. DystociaTable 9.39Historical control data and values indicating a positive signalEndpoint Historical Control Positive Signal ThresholdViable litter size Mean = 14.1 ± 0.94 Decrease of ≥1Survival before/on PND 4 Mean = 95.9% ≤91%Range = 91.3%–99.3%Total litter loss Mean = 1.21% 1 is equivocal(N = 1905 litters) 2 is a stronger signalPND 1 pup weights Mean = 7.1 g ± 0.25 ≤6.5 g is a strong signalRange = 6.5–7.6 g(N = 1100 litters)Source: Data from WIL Research Laboratories, Inc. Crl:CD ® (SD)IGS BR rat reproductivehistorical control database (55 studies conducted during 1996–2002).The example shown in Table 9.40 refers to a two-generation reproductive toxicity and developmentalneurotoxicity study with a second mating phase for the F 1 generation. 136 The agent tested, octamethylcyclotetrasiloxane(D4), was used industrially as well as in over-the-counter products andwas administered via whole-body vapor inhalation. Five groups of 30 rats per sex received target© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 405Table 9.40Rates of dystocia in a two-generationreproductive toxicity study ofoctamethylcyclotetrasiloxaneExposure (ppm) 0 70 300 500 700No./sex 30 30 30 30 30F 0 0 0 0 2/24 3/26F 1 -1st 0 0 0 0 1/17F 1 -2nd 0 0 1/21 1/18 0/12exposures of 0, 70, 300, 500, or 700 ppm of D4. Exposure of the F 0 and F 1 males were conductedfor 10 weeks before mating and through the mating period. Exposures of the F 0 and F 1 femaleswere conducted for 10 weeks before mating, throughout mating and gestation (suspended from GD21 through LD 4), and from LD 5 through weaning of the offspring.One of the endpoints affected by D4 was parturition. To determine whether dystocia in thisstudy was treatment related, comparisons were made to the laboratory’s historical control rate of0.6%. In the F 0 generation, there were five instances of dystocia in a total of 150 litters (3.33%);two in 24 litters (8.3%) at 500 ppm and three in 26 litters (11.5%) at 700 ppm. In the F 1 generationfirst mating, the single incidence (1/17) at 500 ppm was 5.9%. In the F 1 generation second mating,there were two occurrences, 1/21 (4.8%) at 300 ppm and 1/18 (5.6%) at 500 ppm. Because of thelack of evidence of dystocia in the concurrent control groups for either generation and the verylow incidence of dystocia in the historical control data, the staff at the conducting laboratory vieweddystocia as a treatment-related effect.This conclusion could be challenged because no statistically significant incidence of dystociahad occurred in any generation, there was no consistent pattern across generations and/or matings,and there was no apparent dose-response pattern. There were, however, plausible explanations fordystocia being a treatment-related effect in this study. Statistical significance was not (and wouldnever have been) detected because the effect occurred at such a low incidence. In addition, theoffspring of those animals exhibiting dystocia in the F 0 generation were not represented in the F 1generation because of death. Therefore, an apparently decreased response in the first F 1 mating wasthought to be the result of loss of the more sensitive animals from the second generation. Whentwo more instances of dystocia were observed in the F 1 generation second mating (including oneat a lower exposure level than previously observed), the laboratory’s conclusion was strengthened,thereby confirming the generational effect.This effect was initially not acknowledged by many toxicologists because of the low incidenceat which it occurred. This case study demonstrates how a rare event finding may be misinterpretedif one ignores any comparison with historical data and only relies upon finding a dose-responserelationship, consistency across generations, and statistical significance.2. ACE FetopathyThe fetal effects of angiotensin converting enzyme (ACE) inhibitors that cross the placenta werefirst recognized in the human. Subsequent reevaluations of animal data revealed that a positivesignal (increased postnatal mortality) had been present in the original studies. However, the low(and not statistically significant) incidence above concurrent controls of this finding was deemednot treatment related. Current awareness of the real risks for birth defects when using ACE inhibitorsthat cross the placenta allows practitioners to minimize risk and gain therapeutic benefits for treatingpregnancy-related hypertension. When ACE inhibitors that cross the placenta are used in the thirdtrimester, these antihypertensive agents cause fetal hypotension. Additional fetal effects are illustratedin Figure 9.12.Major organogenesis (defined as the time from implantation to closure of the hard palate) isunaffected by ACE inhibitors; however, other effects on the developing organism are severe,© 2006 by Taylor & Francis Group, LLC

406 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONACE inFetalhypotensionRenalCompromise(anuria)OligohydramniosCalvarial hypoplasiaNeonatal anuriaIUGRDeathFigure 9.12 Spectrum of fetal effects following maternal administration of ACE inhibitors.including renal compromise resulting in anuria. This in turn leads to a reduced volume of amnioticfluid (oligohydramnios), hypoxia, and calvarial hypoplasia. Humans born with this syndrome arestunted because of the anuria, and many die at an early age.What happened in the risk assessment of these ACE inhibitors? The original animal safetytesting was extensive and of good quality. 137 In spite of this, the small increases in postnatalmortality, evident in the initial reproduction studies, were not identified as positive treatment-relatedeffects, probably because these increases were marginally statistically significant.An elegant series of postnatal studies (reported by Spence et al.), in which ACE inhibitors weredirectly administered to pups, revealed increased mortality (greater than 30%), growth retardation,and renal alterations (anatomical and functional). 138,139 There is a temporal difference in developmentalschedules between humans and rodents. In humans, renal glomerular function matures inutero, while in the rat, it matures postnatally on day 17 or 18. Thus, initial lack of recognition ofthis difference led to an incorrect outcome in the extrapolation of results to humans. It was notuntil additional confirmatory studies were conducted that the complete picture emerged, and theACE inhibitors became the first examples of non-CNS functional developmental toxicants.3. Retroesophageal Aortic ArchThis case, involving a rat developmental toxicity study conducted in the authors’ laboratory withan antibiotic to be used via the oral mucosal route, again demonstrates the criticality of maintaininga robust and up-to-date historical control database. Single occurrences of retroesophageal aorticarch in the mid- and high-dose groups of the study presented an interpretation dilemma. Thesemalformations were initially considered to be a treatment-related rare event, but upon closerevaluation of the findings within the context of the historical control data were interpreted torepresent increased penetrance of this trait in Crl:CD ® (SD)IGS BR strain rats. However, reassessmentof the data after two additional years of historical control data compilation pointed back tothe original interpretation, that the findings were related to treatment with the test article.The drug was tested for developmental effects in both the rat and rabbit during July 2000. Inrats, it was administered orally by gavage as a single daily dose from GD 6 through 17 at a doseof 0, 5, 25, or 100 mg/kg/d. No maternal toxicity or effects on intrauterine growth and survivalwere observed. However, retroesophageal aortic arches were present in one fetus each at 25 and100 mg/kg/d. Retroesophageal aortic arch is a malformation in which the great arch of the aortafrom the heart develops behind the esophagus. It requires a major change in the early developmentof cardiac outflow, and it is a rarely occurring malformation. In spite of this malformation, bothfetuses were apparently thriving based on their body weights, which were similar to their respectivemean litter (3.6 and 3.5 g) and group mean litter (3.6 and 3.5 g) weights. Viscerally, these fetuseshad no evidence of impaired peripheral perfusion. Furthermore, in the concurrently performedembryo and fetal development study in Hra:(NZW)SPF rabbits, no retroesophageal arches wereobserved in the 427 fetuses (63 litters) from does administered the test article orally at doses upto 60 mg/kg/d.The appearance of retroesophageal aortic arch in two test article–treated groups was deemedunlikely to be a spontaneous event, considering that this finding was observed for only 2 of 9642fetuses evaluated viscerally from 635 litters in 27 studies (1998 to 2001) in the WIL© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 407Table 9.41Dose (mg/kg/d)Crl:CD ® (SD)IGS BR rat historical control database (refer to Table 9.41). While such a distribution(in the case study) arising merely from chance is statistically possible, the authors instead positedthe explanation described below, based on the temporal placement of these findings in the historicalcontrol database at the time. Because the only instances of retroesophageal aortic arch in thehistorical control database at the time of the study were clustered in the year prior to conduct ofthe case study (1999), the occurrence of these malformations in this study was interpreted to reflectincreased penetrance of this trait. Further evidence of increased penetrance at the time was derivedfrom single occurrences of these findings in test article–treated groups (unrelated to test articletreatment) of two studies that had been recently conducted for other sponsors.The occurrence of retroesophageal aortic arches in the case study, while unusual, was notascribed to treatment because the incidences of this finding in the mid- and high-dose groups werewithin the WIL historical control data range, the WIL historical control database suggested anincreased penetrance of this trait, and no concordance was observed in the concurrently performedembryo and fetal development study in rabbits.Further maturation of the WIL historical control database (addition of 2 years and 12,211 controlfetuses; refer to Table 9.41) revealed only a single additional finding of retroesophageal aortic arch(in 2002). Since the time of the case study, another unrelated test article demonstrated a nearlysingular effect on major blood vessels, with the predominant malformation being retroesophagealaortic arch (100% of the high-dose group litters had fetuses with this finding). A xenobiotic hassubsequently been shown to exquisitely affect major blood vessel development, while no sustainedincrease in the historical control incidence of retroesophageal aortic arch has been observed.Therefore, the authors now consider the original two occurrences of this finding in the case studyto likely have been associated with treatment.4. Lobular Agenesis of the LungRetroesophageal aortic arch in Crl:CD ® (SD)IGS BR rat developmentaltoxicity studies aCase Study aGroup IncidencesIncidences in HistoricalControl Databases0 5 25 100 1 b 2 cFetuses affected/total 0/386 0/368 1/353 1/351 2/9642 3/21,853 dfetuses evaluatedTotal litters evaluated 24 24 23 22 635 1447Mean %/litter affected 0.0 0.0 0.3 0.3 0.0–0.3 0.0–0.3aThe case study was conducted during June 2000.bDatabase 1 is the database existing at the time of initial interpretation of the case studydata (27 studies conducted from April 1998 through March 2001).cDatabase 2 is the current database (61 studies conducted from April 1998 through June2003); this database is a continuation of Database 1.dThe initial two findings of retroesophageal aortic arch were from studies conducted in1999, and the third instance of this finding occurred in a study conducted in 2002.Source: All studies referenced were conducted at WIL Research Laboratories, Inc.Although ultimately discounted in the aforementioned case, penetrance does remain a confoundingfactor in interpretation of rare event manifestation. As an example, Table 9.42 presents the temporaldistribution of another malformation, lobular agenesis of the lung (right accessory lobe), inHra:(NZW)SPF rabbits in the authors’ historical control database.During a 2-year period (1999 and 2000), lobular agenesis of the lung was observed in 13 of2951 fetuses evaluated (average yearly incidence of 0.45%). In the previous and subsequent 3-yearperiods, only 2 of 3002 and 4 of 2763 fetuses, respectively, were observed with this finding (averageyearly incidences of 0.05% and 0.17%, respectively). The increased incidence of lobular agenesis© 2006 by Taylor & Francis Group, LLC

408 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.42Temporal distribution of lobular agenesis of the lung in control Hra:(NZW)SPFrabbit fetuses aYear 1996 1997 1998 1999 2000 2001 2002 2003Total fetuses evaluated 1088 519 1395 1659 1292 1177 707 879Lobular agenesis of the lungFetuses affected (no.) 1 0 1 6 7 1 3 0Fetuses affected (%) 0.09 0.00 0.07 0.36 0.54 0.08 0.42 0.00aData presented are from 67 Hra:(NZW)SPF rabbit developmental toxicity studies (1996 to 2003)in the historical control database at WIL Research Laboratories, Inc.Table 9.43may indicate a heritable trait and, as such, may represent localized, time-bounded penetrance. Thisincrease in the incidence of such a rare finding during a 2-year period may or may not be consideredremarkable. However, given the temporal boundaries of the increase, the apparent penetrance oflobular agenesis may significantly affect interpretation of a study conducted during that time frame.Because the finding was still a rare event during 1999 to 2000, a spontaneously occurring distributionof findings solely in treated groups was certainly possible. Without systematic maintenance ofhistorical control data and recognition of penetrance as it occurs, the hypothetical distribution ofa few occurrences of lobular agenesis in the treated groups would likely be interpreted as a testarticle–related finding. Such an interpretation could possibly reduce the margin of safety of acandidate pharmaceutical, potentially keeping a needed therapy from patients.E. Approach to Evaluating Rare FindingsTypical reaction to rare events andsubsequent scenariosDisbelief; rely on lack of statistical significanceComparison to concurrent controlComparison to laboratory’s historical control databaseComparison to other historical control databasesAsk experience and opinions of othersConstruct explanation to negateAgency rejectsRepeat study or label appropriatelyTable 9.43 presents the typical reaction to the manifestation of low incidence findings and thesubsequent scenario. The recommended paradigm for interpretation of rare events includes a numberof steps that should be considered to determine if a finding is biologically relevant (even thoughit is not statistically significant) when compared with the concurrent control. These steps are listedin Table 9.44; however, their order may vary depending on the situation.When evaluating and interpreting rare events, the best practice for resolution of the relationshipbetween the agent tested and the findings observed is use of a large historical control database forcomparison. Such a database should be developed at the laboratory performing the studies, usingconsistent methodology, staff, and animal husbandry, as well as appropriately designed confirmatorystudies. The design of confirmatory studies may involve increasing the number of animals overall,as well as in the control group, to counteract the 3:1 probability of a rare event occurring in thetreatment groups versus the control group (see Figure 9.13).Evaluation of whether a signal is real should also include comparisons of results from otherreproductive and developmental studies, as well as observations in other species.With the current designs of animal safety testing protocols, it is not possible to detect rareevents or small changes in endpoints by use of the guideline-recommended statistical methodologies.Other approaches (such as the most current version of the Monte-Carlo analysis) will haveto be used.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 409Table 9.44Stepwise ApproachDose response(including TK a )Historical controlcomparisonAdditional statistical testsApproach to determining biological relevance of rare event findingsSecond species comparison(including TK data)Other studies comparisonConfirmatory studyMechanistic studiesConsideration and CautionPossibility of any rare event occurring is 3:1 in favor of test article–treated groupsMean and range values (laboratory’s own data, then others)Values near or outside upper or lower limits of historical control range provide astrong signal that rare event is “real”Monte-Carlo analysis to examine sampling error and predict population estimatesrelative to sampling valuesEstablishment of similar internal dose (C max and/or AUC) criticalConcordant outcome from embryo, fetal, and pre- and postnatal studies (e.g.,slight decrease in fetal BW in developmental study, along with slight decreasein PND 1 BW in reproductive study) is considered to be “real” eventIncreased N per group; increased N in control group; increased dose; limitedexposure regime (if developmental timing of organ or system affected is known,only administer doses to achieve same AUC on limited regime, which couldilluminate whether the underlying embryology and the outcome make sense)Evaluate pharmacological action relative to ontogeny of receptors, signalingpathways, and other possible modes of actionaN = number, BW = body weight, TK = toxicokinetics, PND = postnatal day, AUC = area under the curve.1ControlHigh3LowMidFigure 9.13 Probability of a rare event occurring in the treatment groups relative to the control group.Because concurrent control groups are limited in size (N), they represent, at best, the currentrate of high-incidence findings and/or continuous variable values. They are poor estimators of thetrue population statistic. To estimate the incidence of a developmental defect in the population ofa given species for occurrences of less than 1 in 100 requires approximately 100 control data setsfor reliability. Therefore, concurrent controls have little value in discerning treatment-related versusspontaneous occurrences for the majority of developmental toxicity endpoints (other than fetalweight and embryolethality). The best probabilistic approach to matching current findings tooccurrences in a historical control population is a Monte-Carlo analysis. 140VIII. THE ROLE OF EXPERT JUDGMENTThis section will address additional factors that must be considered when interpreting developmentaland reproductive toxicity data.© 2006 by Taylor & Francis Group, LLC

410 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 9.45Integrity of databaseFrequently asked key regulatory questions following data package submissionWas an established, validated model used?Was a NOAEL (NOEL) demonstrated?Does increasing dose increase severity or incidence?Biologic dynamics anddimensionsRelevancy and riskanalysisWhen was the effect exerted?Is the effect reversible?Do indications of sensitization (generational effects or imprinting) exist?Is the effect gender specific?Were appropriate TK comparisons made between experimental studies andestimated human PK or exposure scenarios?Are there significant differences in the pattern, timing, or magnitude of exposurebetween guideline studies and human scenarios?Is concordance of effects among species evident?Is the mode of action known or deducible?Is the mechanism of action known?The final report for a developmental or reproductive toxicity study is most often submitted tothe appropriate regulatory agency. Following submission, careful regulators not only will questionthe study director’s final interpretation but will also examine issues pertaining to the integrity ofthe database, key biological dynamics and dimensions, and the relevance and risk to the humanexposure scenario. Table 9.45 presents 12 key questions frequently asked by representatives ofregulatory agencies concerning the aforementioned issues.The expert toxicologist must anticipate the regulators’ concerns and ask these questions priorto data interpretation and regulatory submission, as well as considering them prior to study design.The following subsections will address these common questions pertaining to the integrity of thedatabase, key biologic dynamics and dimensions, and the relevancy and risk related to the humanexposure scenario.A. Integrity of DatabaseQuestions surrounding the integrity of the data generated from any toxicity study should beparamount in the mind of an expert toxicologist. These questions involve the ability of the studydesign to detect potential toxicologic changes under study, the training of personnel conducting astudy, and the possible presence of dose-response relationships, among others. Without theseunderpinnings, the utility of toxicologic data, and any interpretations derived therefrom, is diminished.1. Was an established, validated model used?While the rat and rabbit are the most commonly used species in studies of toxicity to development,and the rat is the primary test system for reproductive toxicity determinations, examples of theuse of the dog and primate have been discussed in this chapter. Probably the single most importantcriterion of an appropriate model is metabolism that is similar to that of the human, especiallywhen a metabolite is the developmentally toxic agent. Typically, this level of understanding ofmetabolism is not determined a priori except in certain pharmaceutical development situations.2. Was a NOAEL (or NOEL) demonstrated?From numerous examples in this chapter, the potential problems that arise from relying solely onstatistical significance and/or the presence of a dose-response relationship are evident. The absenceof a dose response does not fully negate the findings, especially in the case of rare events orthreshold responses. Regulators often will rely heavily on concurrent control data for comparison.Slight changes in response across treatment groups relative to the concurrent control, in the absenceof statistical significance or a clear dose-response relationship, may be an indication of toxicity.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 411Table 9.46How maldevelopment differs from tumorigenesisNot more frequent with timeMyriad possible underlying mechanismsAmniotic banding, oligohydramnios-skull dysgenesisInterference with signaling pathways (e.g., TGF-β)MutationMaternal influences possibleMultiple endpoints interrelatedWeight alterations causing cleft palate and neural tube defectsOccurs early in life and hence greater economic and social impactTable 9.47Selected differences in developmental toxicity vs. oncogenicity endpoint ascertainmentSmaller group sizes (25 litters vs. minimum 50/sex/group)Macroscopic examination (histopathology very rare)Physical limitations based on sizeLess standardized nomenclatureNo certification of personnel, controls over training, etc.This has potentially more of an affect because of the earlier inclusion of women in clinical trialsInvolves coapt organisms (dam and fetuses)Always potential for maternal influence, but fetal effects may affect the dam as wellDynamic morphology and functionACE fetopathy exampleAnimals evaluated in the midst of changing morphologyNo two points in development are the sameExposure hourly and daily key to outcomeAn important aspect of human studiesIn these instances, the use of historical control data may be the more appropriate comparison fordetermining relationship to treatment. However, some data sets will always appear uninterpretable.In these cases, additional studies with increased dosage levels, perhaps with an unbalanced design,can be useful to confirm or refute the initial findings.3. Does increasing dose increase severity and/or incidence?Despite the concerns raised above, the presence of a dose-response relationship can provide reasonsfor enhanced concern. Slightly increased responses (relative to the control group) at the lowestexposure level and greater responses with increasing exposure provide confidence that the effectsobserved at the lowest exposure level are likely to have been related to treatment, even in theabsence of statistical significance. The proper spacing of dose levels is critical to discernment ofdose-response relationships. Because of animal-to-animal variability, differences between doselevels of less than twofold may result in overlapping of responses with increasing exposure.Conversely, dose levels too widely spread (generally greater than fivefold) will hinder characterizationof the dose-response relationship. A likely outcome of this scenario is identification of ano-effect dose and one response level; however, it is likely that no information regarding the slopeof the dose-response curve will be obtained. In either case, improper spacing of dose levels canmake identification of the threshold response level problematic.It is interesting to contrast the dose-response relationships of developmental toxicity andcarcinogenicity. The cancer endpoint (nonepigenetic) is considered to exhibit linear dose-responsebehavior, whereas most regulators consider reproductive or developmental toxicity to be a thresholdphenomenon. Both developmental defects and cancer are generally irreversible endpoints, andcertain reproductive lesions may also be irreversible. The basis for distinction between carcinogenicityand developmental toxicity is not totally rational, nor is it entirely consistent with ourunderstanding of the possible mechanisms of carcinogenicity or with the dose-response outcomesof the two. Many differences exist between these two toxicologic endpoints (see Table 9.46 andTable 9.47) that contribute variably to this distinction (linear versus threshold). Nevertheless, it is© 2006 by Taylor & Francis Group, LLC

412 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmore probable that both obey similar dose-response behavior at corresponding exposure levels,with the exception that developmental toxicity can be caused by nonstoichiometric phenomena.B. Biologic Dynamics and DimensionsA full command of the biologic dynamics and dimensions surrounding the possible manifestationsof developmental and reproductive toxicity is essential for proper data interpretation. Questionsunder this umbrella address the discernment of the timing and reversibility of effects, the recognitionof those effects that appear in a sensitization pattern, and the resolution of gender-specific effects.4. When was the effect exerted?Identification of the timing of a developmental or reproductive insult is important for extrapolationof animal data to effects in humans for subsequent risk assessment. Intimate knowledge ofdevelopmental and reproductive processes will guide the investigator in designing subsequentstudies to determine whether single or multiple exposures were necessary for the insult, to determinewhether central or peripheral sites of action were affected, and to elucidate mode of action.Understanding the mode of action of the test substance in animal models is essential to determiningrelevance of the insult to humans, particularly for effects on fertility and other reproductiveprocesses. There are numerous examples of agents that require only a limited window of exposureto elicit an adverse response. For example, agents that block the LH surge in female rats canadversely affect fertility following a single dose on the day of proestrus. These agents would beregulated quite differently than those agents requiring extended exposure to attain sufficientbioaccumulation to elicit an adverse response.5. Is the effect reversible?Reversibility of developmental and reproductive insults may greatly influence regulation of thetest agent. Assessment of reversibility for fertility toxicants may be accomplished through use ofa recovery (nonexposure) period. Doses at or near the MTD may induce irreversible changes;however, those same effects may be reversible at lower dosage levels, indicating a threshold forreversibility. Interrelatedness of the endpoints in developmental toxicity studies makes it difficultto assess reversibility. Disruption of the schedule of in utero development may have consequencespostnatally. In the case of intrauterine growth retardation, even though postnatal growth maynormalize to the level of concurrent controls, developmental schedules may be disturbed, resultingin increased postnatal loss. Postweaning assessments beyond measurement of bodyweight, suchas evaluation of neurobehavioral endpoints, may be important in revealing latent effects that signallack of reversibility.6. Do indications of sensitization (generational effects or imprinting) exist?Generational effects and imprinting represent some of the most insidious risks to human health.The multigeneration study is designed to assess the potential for increased risk in subsequentgenerations, relative to the initially exposed (F 0 ) generation. Specious decreases in adverse reproductiveresponses in subsequent generations must be recognized as such, as adverse effects canbe masked by the loss of sensitive individuals from the F 0 /F 1 parental animal populations (selectionphenomenon). See Section VII.D.1 for a description of the D4 example from the authors’ laboratory.Sensitizing phenomena can be separated from bioaccumulative effects by the use of toxicokineticdata to compare internal exposures in offspring with those of the parental generation.7. Is the effect gender-specific?In determining if the effect observed is gender specific, one can identify the population at riskfrom the exposure. With agents not producing overt toxicity, discerning a gender basis may notalways be straightforward. However, gender-specific effects can often be deduced from organhistopathology or other endpoints (e.g., reproductive hormone levels). In classic fertility and twogenerationreproductive toxicity studies, in which both males and females are given the test article,additional studies generally will be required to determine if one sex is preferentially affected.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 413Alternatively, additional breeding phases can be introduced to the extant study designs, using naivemales and/or females in an effort to minimize the number of animals required. When identified,gender-specific effects can give valuable guidance toward identification of possible modes of action.C. Relevancy and Risk AnalysisDespite adequate integrity of the database and resolution of particular biologic dynamics anddimensions, consideration of the relevancy of the data and subsequent risk analysis must beaddressed. These issues include interspecies and temporal comparisons of measures of dosimetryand knowledge of mode and mechanism of action.8. Were appropriate toxicokinetic comparisons made between experimental studies and estimatedhuman pharmacokinetics or exposure scenarios?The appropriate toxicokinetic assessments are key to standardizing internal dose, establishingbioavailability, and determining possible dose-dependent kinetics and/or potential changes inmetabolism over the course of exposure. Adverse developmental and reproductive effects may bethe result of sufficiently high C max or AUC exposures or a combination thereof. Because examplesof both cases exist (e.g., valproic acid for C max and cyclophosphamide for AUC), the observedeffects must be considered in terms of the proper internal exposure measure. In the case of at leastone chemical (2-methoxyethanol), dysmorphogenesis can be attributed to both the AUC and C max ,depending on the critical period of development. In mice, 2-methoxyethanol exposure on GD 8caused exencephaly, with a high correlation to the maternal plasma C max , independent of thematernal AUC. 141 However, 2-methoxyethanol exposure of mice on GD 11 elicited paw malformationswith a better correlation to the AUC than to the C max . 142 Therefore, prenatal response tointernal dose (as measured by C max or AUC) at one specific point in development does notnecessarily apply to all points in development.Further use of toxicokinetic data may help determine whether effects observed in neonatesoccur because of exposure during gestation or lactation. Elaborate schemes involving milk production,collection, consumption, and bioavailability may be constructed to address neonatalexposure questions. However, the use of simple fostering studies (mutual exchange of littersbetween treated and untreated dams) with direct toxicokinetic measurements from offspring bloodmay provide a more reliable and straightforward means of answering those questions.9. Are there significant differences in the pattern, timing, or magnitude of the exposure betweenguideline studies and human scenarios?Standard experimental animal models employed in hazard identification have greatly compressedperiods of organogenesis relative to the timing of human development. Single daily administrationsof drugs with short half-lives may result in very low systemic exposures during critical developmentalwindows and may not be adequate for assessing developmental toxicity over the entiresusceptible period. These cases may require multiple administrations each day or alternate exposuremethods (e.g., continuous infusion or dietary administration) to maximize exposure duration.Multiple administrations per day have become more commonplace in the last 5 years, with theadvent of more extensive use of toxicokinetics.10. Is concordance of effects among species evident?Effects manifesting in multiple species signal enhanced concern for human risk, and the concernincreases further when the number of species increases. However, the absence of such multiplicitydoes not necessarily diminish concern. 5 Response variability between species most often appearsto indicate a mode of action difference that can be critical to appropriate animal model selectionfor human hazard identification. “Apparent” lack of concordance among species can arise viaseveral different scenarios. A difference in sensitivity, implying a biological basis (i.e., not amaternally controlled or metabolic basis), is often cited as a primary reason for discordance betweenspecies. However, this claim has historically been made without comparing appropriate measuresof internal dose (exposure) among the species in question. Examples where an apparent lack of© 2006 by Taylor & Francis Group, LLC

414 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONconcordance based on sensitivity issues was resolved through exhaustive measures of AUC andC max are the cases of isotretinoin and valproic acid. For isotretinoin, comparisons of maternal andembryonic blood levels of 13-cis-retinoic acid (isotretinoin) and all-trans-retinoic acid, as well astheir predominant metabolites, revealed significant differences between the more sensitive (primatesand lagomorphs) and less sensitive (rodents) species in elimination, glucuronidation, andplacental transport. 143–146Another common source of false discordance is lack of appropriate statistical power in acomparison species. Because of the prohibitive cost and animal welfare concerns of large groupsizes, misleading discordance frequently occurs when the dog or nonhuman primate is used as thetest species.11. Is the mode of action known or deducible?Insight into the mode of action of an agent under study is highly desirable when the toxicologistis faced with the interpretation of possible equivocal developmental and reproductive effects.Knowledge concerning the pharmacologic action or class effect may aid the investigator in targetingspecialized study designs and obtaining a better assessment of the degree to which the findingsrepresent a hazard to human reproduction.12. Is the mechanism of action known?If the mechanism of action is defined as the effect precipitating the initiation of pathogenesis, thenthe precise action of no known human developmental toxicant has been elucidated. However,increasing knowledge of mode and mechanism of actions is being gained through research and isprobably most advanced in the case of the retinoids. The value of knowledge of mechanism ofaction is commonly extolled and obvious.The adequacy of the current ICH, EPA OPPTS, and OECD developmental and reproductivetoxicity guideline studies to answer these 12 questions is presented in Table 9.48.IX. CONCLUSIONThe goal of nonclinical reproductive toxicity studies is identification of potential human hazards.This chapter is meant to be a practical guide to the significance, reliability, and interpretation ofreproductive and developmental toxicity study findings. When reviewing these data, one must bemindful that the validity and reliability of the study are based on many assumptions. Theseassumptions include the following:1. Staff for study execution is well-trained and competent.2. Animal husbandry follows best practices.3. Animal model selected is appropriate.4. Sample size is sufficient for statistical power.5. Dosage levels are adequately spaced and selected.6. Kinetics are well defined.In addition, knowledge of the mode of action of test substances is becoming more availableand will continue to be useful in the evaluation of reproductive and developmental toxicity studyfindings.Once the investigator is confident that these assumptions have been met, the data set then canbe presumed valid and reliable for extrapolation to the human situation. In order to draw a conclusionregarding reproductive toxicity, the investigator must understand the interrelatedness of the endpointsunder study. Often, subtle reproductive insults can be recognized by identification of a patternof effect across multiple endpoints. Historical control data developed at the performing laboratoryunder similar conditions is essential in assessment of the significance of rare event findings.© 2006 by Taylor & Francis Group, LLC

Table 9.48What do regulatorswant to know?Extent to which guideline studies answer key regulatory questionsICH EPA OPPTS (8700.) OECDFert. P/P DT DT 2-G DNT DT 1-G 2-G Screen DNT4.1.1 4.1.2 4.1.3 3700 3800 6300 414 415 416 421 4261 Validated model Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes2 NOAEL determined Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes3 Dose response No Yes Yes Yes Yes No Yes Some Yes Some No4 Insult timing elucidated No No No No No No No No No No No5 Reversibility No No No No No No No No No No No6 Imprinting phenomenon No No No No Yes No No No Yes No No7 Gender basis No Yes Yes Yes No Yes Yes No No No Yes8 TK pro led Yes Yes Yes No No No No No No No No9 Exposure mimicked Some Some Some ? ? ? ? ? ? ? ?10 Interspecies concordance No No Yes Yes No No Yes No No No No11 Mode of action No No No No No No No No No No No12 Mechanism of action No No No No No No No No No No NoNote: Fert = fertility; P/P = pre and postnatal; DT = developmental toxicity; 2-G = two-generation; DNT = developmental neurotoxicity;1-G = one-generation.DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 415© 2006 by Taylor & Francis Group, LLC

416 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONDespite confirmation of the aforementioned assumptions and the presence of a rich and accuratehistorical control database, the investigator may not be able to draw definitive conclusions becauseof equivocal findings. In this scenario, the only recourse is to perform additional studies, perhapswith increased sample size and/or increased dosage levels.Important decisions must be made based on these data sets. In the case of drugs, many of theendpoints assessed in nonclinical reproductive safety studies may not be principal endpoints inclinical trials (e.g., functional assessments, hormone analyses, and intrauterine growth and development).For environmental chemicals, human exposure is often unknown and unmonitored, andcould involve multiple compounds. As a result, determination of a proximate human reproductivetoxicant is often impossible. Therefore, reproductive hazard identification in animal studies isparamount in protecting human health.Human tragedies have led to the refinement and enhancement of regulatory guideline requirements.As a result, there exists a general confidence in the ability to detect reproductive hazards.This is especially true when effects are manifested across multiple well-validated animal modelsand/or studies with overlapping treatment regimes or endpoints. As our knowledge of the basicbiology of animal models and humans advances, study designs and the strength of assumptionsused for animal-to-human extrapolation will continue to improve. The challenge facing investigatorsis application of emerging scientific principles to hazard identification and ultimately risk assessment.Faulty study design and interpretation can keep a needed treatment from an affected humanpopulation or conversely expose those populations to unnecessary risk.ACKNOWLEDGEMENTSThe authors would like to acknowledge with deep appreciation the efforts of various individualswho contributed in tangible ways to the completion of this project. Special thanks to KerinClevidence, Michelle Edwards, and Jeanette Howell, who labored diligently to compile and verifymuch of the WIL historical control data referenced in this chapter. The authors would also like torecognize the numerous staff members at WIL, too many to mention individually, who throughtheir dedication, hard work, and meticulous data collection over the past 10 to 20 years have laidthe foundation for this chapter.REFERENCES1. DeSesso, J.M., Harris, S.B., and Swain, S.M., The design, evaluation, and interpretation of developmentaltoxicity tests, in Toxicology and Risk Assessment: Principles, Methods, and Applications, Fan,A.M. and Chang, L.W., Eds., Marcel Dekker, New York, 1996, p. 227.2. Palmer, A.K. and Ulbrich, B., Data presentation in developmental toxicity studies: an aid to evaluation?in Handbook of Developmental Toxicology, Hood. R.D., Ed., CRC Press, Boca Raton, FL, 1997, chap. 8.3. Daston, G. and Kimmel, C.A., Eds., An Evaluation and Interpretation of Reproductive Endpoints forHuman Health Risk Assessment, International Life Sciences Institute, Washington, DC,1998.4. European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC), Guidance on Evaluationof Reproductive Toxicity Data, Monograph No. 31, 2002.5. U.S. Food and Drug Administration, Reviewer Guidance: Integration of Study Results to AssessConcerns about Human Reproductive and Developmental Toxicities (Draft), Center for Drug Evaluationand Research, 2001.6. Wilson, J.G., Environment and Birth Defects, Academic Press, New York, 1973.7. Brent, R.L., Drug testing for teratogenicity: its implications, limitations and application to man, inDrugs and Fetal Development, Klingberg, M.A., Abramovici, A., and Cheneke, J., Eds., Plenum Press,New York, 1972, p. 31.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 4178. Nishimura, H. and Tanimura, T, Chemical Aspects of the Teratogenicity of Drugs, Elsevier, New York,1976.9. Fraser, F.C., Relation of animal studies to the problem in man, in Handbook of Teratology, Vol. 1,Wilson, J.G. and Fraser, F.C., Eds., Plenum Press, New York, 1979, p. 75.10. Snow, M.H.L. and Tam, P.P.L., Is compensatory growth a complicating factor in mouse teratology?,Nature, 279, 555, 1979.11. U.S. Food and Drug Administration, National Center for Toxicological Research, Reliability of ExperimentalStudies for Predicting Hazards to Human Development (Kimmel, C.A., Holson, J.F., Hogue,C.I., and Carlo, G.L., Eds.), NCTR Technical Report for Experiment No. 6015, Jefferson, AR, 1984.12. National Research Council, Scientific Frontiers in Developmental Toxicology and Risk Assessment,National Academy Press, Washington, DC, 2000.13. Ulbrich, B. and Palmer, A.K., Detection of effects on male reproduction — a literature survey, J. Am.Coll. Toxicol., 14, 293, 1995.14. Clegg, J.A., Adverse drug reactions, Toxicol. Rev., 13, 235, 1994.15. Buiatti, E, Barchielli, A., Geddes, M., Nastasi, L., Kriebel, D., Franchini, M., and Scarselli, G., Riskfactors in male infertility, Arch. Environ. Health, 39, 266, 1984.16. Paul, M. and Himmelstein J., Reproductive hazards in the workplace: what the practitioner needs toknow about chemical exposures, Obstet. Gynecol., 71, 921, 1988.17. Thomas, M.J. and Thomas, J.A., Toxic responses of the reproductive system, in Casarett & Doull’sToxicology, the Basic Science of Poisons, Klassen, C.D., Ed., McGraw-Hill, New York, 2001, p. 673.18. Waller, D.P., Killinger, J.M., Zaneveld, L.J.D., Physiology and toxicology of the male reproductivetract, in Endocrine Toxicology, Thomas, J.A., Korach, K.S., and McLachlan, J.A., Eds., Raven, NewYork, 1985, p. 269.19. Scialli, A.R. and Lemasters, G.K., Epidemiologic aspects of reproductive toxicology, in ReproductiveToxicology, 2nd ed., Witorsch, R.J., Ed., Raven Press, 1995, p. 241.20. Potashnik, G. and Porath, A., Dibromochloropropane (DBCP): a 17-year reassessment of testicularfunction and reproductive performance, J. Occup. Environ. Med., 37, 1287, 1995.21. Goldsmith, J.R., Dibromochloropropane: epidemiological findings and current questions, Ann. N. Y.Acad. Sci., 26, 837, 1997.22. Meistrich, M.L., Wilson, G., Shuttlesworth, G.A., and Porter, K.L., Dibromochloropropane inhibitsspermatogonial development in rats, Reprod. Toxicol., 17, 263, 2003.23. Cooper, R.L., Neuroendocrine control of female reproduction, in Reproductive and Endocrine Toxicology(Comprehensive Toxicology, Vol. 10), Boekelheide, K., Chapin, R.E., Hoyer, P.B., and Harris,C., Eds., Elsevier, New York, 1997, p. 273.24. Eldridge, J.C., Wetzel, L.T., Stevens, J.T., and Simpkins, J.W., The mammary tumor response intriazine-treated female rats: a threshold-mediated interaction with strain and species-specific reproductivesenescence, Steroids, 64, 672, 1999.25. Cooper, R.L., Goldman, J.M., and Rehnberg, G.L., Neuroendocrine control of reproductive functionin the aging female rodent, J. Am. Geriatric Sci., 34, 735, 1986.26. Cooper, R.L., Stoker, T.E., Goldman, J.M., Parrish, M.B., and Tyrey, L., Effect of atrazine on ovarianfunction in the rat, Reprod. Toxicol., 10(4), 257, 1996.27. U.S. Food and Drug Administration, Guidelines for Reproduction Studies for Safety Evaluation ofDrugs for Human Use, 1966.28. U.S. Federal Food and Drugs Act of 1906, Public Law No. 59-384, 34 Stat. 768, 1906.29. U.S. Federal Food, Drug and Cosmetic Act, 21 U.S.C. 301, 1938.30. Green, R., Burrill, M., and Ivy, A., Experimental intersexuality: the paradoxical effects of estrogenson the sexual development of the female rat, Anat. Rec., 74, 429, 1939.31. Green, R., Burrill, M., and Ivy, A., Experimental intersexuality: modification of sexual developmentof the white rat with synthetic estrogen, Proc. Soc. Exp. Biol. Med., 41, 169, 1939.32. Herbst, A.L., Ulfelder, H., and Poskanzer, D.C., Adenocarcinoma of the vagina: association of maternalstilbestrol therapy with tumor appearance in young women, N. Eng. J. Med., 284, 878, 1971.33. Herbst, A.L., Cole, P., Colton, T., Robboy, S.J., and Scully, R.E., Age-incidence and risk of diethylstilbestrol-relatedclear cell adenocarcinoma of the vagina and cervix, Am. J. Obstet. Gynecol., 128,43, 1977.© 2006 by Taylor & Francis Group, LLC

418 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION34. Lehman, A.J., Laug, E.P., Woodard, G., Draize, J.H., Fitzhugh, O.G., and Nelson, A.A., Procedures forthe appraisal of the toxicity of chemicals in foods, Food Drug Cosmetic Law Quarterly, September, 1949.35. Mauer, I.,Ed., Appraisal of the Safety of Chemicals in Foods, Drugs and Cosmetics, The Associationof Food and Drug Officials of the United States, Washington, D.C., 1959.36. Hendrickx, A.G. and Binkerd, P.E., Nonhuman primates and teratological research, J. Med. Primatol.,19, 81, 1990.37. Buelke-Sam, J., Kimmel, C.A., Adams, J., Nelson, C.J., Vorhees, C.V., Wright, D.C., Omer, V.St.,Korol, B.A., Butcher, R.E., Geyer, M.A., Holson, J.F., Kutscher, C.L., and Wayner, M.J., Collaborativebehavioral teratology study: results, Neurotoxicol. Teratol., 7, 591, 1985.38. U.S. Environmental Protection Agency, Office of Pesticide Programs. Proposed guidelines for registeringpesticides in the United States; Hazard evaluation: Humans and domestic animals, Fed. Regist.,43, 37336, 1978.39. U.S. Environmental Protection Agency, Guidelines for developmental toxicity risk assessment; notice,Fed Regist., 56, 63798, 1991.40. U.S. Environmental Protection Agency, Reproductive toxicity risk assessment guidelines; notice, Fed.Regist., 61, 56274, 1996.41. Eaton, D.L. and Klaassen, C.D., Principles of toxicology, in Casarett & Doull’s Toxicology: The BasicScience of Poisons, 6th ed., Klaassen, C.D., Ed., McGraw-Hill, New York, 2001, p. 11.42. vom Saal, F.S., Timms, B.G., Montano, M.M., Palanza, P., Thayer, K.A., Nagel, S.C., Dhar, M.D.,Ganjam, V.K., Parmigiani, S., and Welshons, W.V., Prostate enlargement in mice due to fetal exposureto low doses of estradiol or diethylstilbestrol and opposite effects at high doses, Proc. Natl. Acad.Sci. USA, 94, 2056, 1997.43. Shuey, D.L., Lau, C., Logsdon, T.R., Zucker, R.M., Elstein, K.H., Narotsky, M.G., Setxer, R.W.,Kavlock, R.J., and Rogers, J.M., Biologically-based dose-response modeling in developmental toxicology:biochemical and cellular sequelae of 5-fluorouracil exposure in the developing rat, Toxicol.Appl Pharmacol., 126, 129, 1994.44. Young, J.F. and Holson, J.F., Utility of pharmacokinetics in designing toxicological protocols andimproving interspecies extrapolation, J. Environ. Pathol. Toxicol., 2, 169, 1978.45. Scott, W.J. Jr., Collins, M.D., and Nau, H., Pharmacokinetic determinants of embryotoxicity in ratsassociated with organic acids, Environ. Health Perspect., 102(Suppl. 11), 97, 1994.46. Faustman, E.M., Allen, B.C., Kavlock, R.J., and Kimmel, C.A., Dose-response assessment for developmentaltoxicity. I. Characterization of database and determination of no observed adverse effectlevels, Fundam. Appl. Toxicol., 23, 478, 1994.47. Hood, R.D., Effects of sodium arsenite on fetal development, Bull. Environ. Contam. Toxicol., 7, 216,1972.48. DeSesso, J.M., Jacobson, C.F., Scialli, A.R., Farr, C.H., and Holson, J.F., An assessment of thedevelopmental toxicity of inorganic arsenic, Reprod. Toxicol., 12, 385, 1998.49. Holson, J.F., Stump, D.G., Clevidence, K.J., Knapp, J.F., and Farr, C.H., Evaluation of the prenataldevelopmental toxicity of orally administered arsenic trioxide in rats, Food Chem. Toxicol., 38, 459,2000.50. Stump, D.G., Holson, J.F., Fleeman, T.L., Nemec, M.D., and Farr, C.H., Comparative effects of singleintraperitoneal or oral doses of sodium arsenate or arsenic trioxide during in utero development,Teratology, 60, 283, 1999.51. National Academy of Sciences, Sample size determination, in Guidelines for the Care and Use ofMammals in Neuroscience and Behavioral Research, National Academies Press, Washington, DC,2003, p. 176, Appendix A.52. Nelson, C.J. and Holson, J.F., Statistical analysis of teratologic data: problems and advancements, J.Environ. Pathol. Toxicol., 2, 187, 1978.53. Steel, R.G.D. and Torrie, J.H., Principles and Procedures of Statistics, A Biometrical Approach, 2nded., McGraw, New York, 1980, pp. 504-506.54. Dunnett, C.W., New tables for multiple comparisons with a control, Biometrics, 20, 482, 1964.55. Kruskal, W.H. and Wallis, W.A., Use of ranks in one-criterion variance analysis, J. Am. StatisticalAssoc., 47, 583, 1952.56. Holson, J.F., Scott, W.J., Gaylor, D.W., and Wilson, J.G., Reduced interlitter variability in rats resultingfrom a restricted mating period, and reassessment of the “litter effect”, Teratology, 14, 135, 1976.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 41957. Holson, J.F., Gaines, T.B., Nelson, C.J., LaBorde, J.B., Gaylor, D.W., Sheehan, D.M., and Young,J.F., Developmental toxicity of 2,4,5-trichlorophenoxyacetic acid (2,4,5-T). I. Multireplicated doseresponsestudies in four inbred strains and one outbred stock of mice, Fundam. Appl. Toxicol., 19,286, 1992.58. Brown, N.A., Spielmann, H., Bechter, R., Flint, O.P., Freeman, S.J., Jelinek, R.J., Koch, E., Nau, H.,Newall, D.R., Palmer, A.K., Renault, J.Y., Repetto, M.F., Vogel, R., and Wiger, R., Screening chemicalsfor reproductive toxicity: the current alternatives; the report and recommendations of an ECVAM/ETSworkshop (ECVAM workshop 12), Altern. Lab. Anim., 23, 868, 1995.59. Chapin, R.E. and Sloane, R.A., Reproductive assessment by continuous breeding: evolving studydesign and summaries of ninety studies, Environ. Health Persp., 105(Suppl. 1), 199, 1997.60. Gesellschaft Deutscher Chemiker (GDCh — German Chemical Society), Evaluation of OECD ScreeningTests 421 (Reproduction/Developmental Toxicity Screening Test) and 422 (Combined RepeatedDose Toxicity Study with the Reproduction/Developmental Toxicity Screening Test), edited by theGDCh Advisory Committee on Existing Chemicals (BUA), BUA Report 229, S. Hirzel WissenschaftlicheVerlagsgesellschaft, Stuttgart, 2002.61 Hurtt, M.E., Cappon, G.D., and Browning, A., Proposal for a tiered approach to developmental toxicitytesting for veterinary pharmaceutical products for food-producing animals, Food Chem. Toxicol., 41,611, 2003.62. Stump, D.G., Holson, J.F., Pearce, L.B., Rentko, V.T., and Gawryl, M.S., Finding of developmentaltoxicity studies of HBOC-201 in rodent and canine models [abstract], Birth Defects Res. (Part B),68, 250, 2003.63. Holson, J.F., Pearce, L.B., and Stump, D.G., A probable false positive finding of prenatal toxicity inthe rodent model with a high molecular weight protein oxygen therapeutic: evidence and implications[abstract], Birth Defects Res. (Part B), 68, 249, 2003.64. Fraser, F.C., The multifactorial/threshold concept — use and misuses, Teratology, 14, 267, 1976.65. Adolph, E.F., Ontogeny of physiological regulations in the rat, Quart. Rev. Biol., 32, 89, 1957.66. Gibson, K.J. and Lumbers, E.R., Mechanisms by which the pregnant ewe can sustain increased saltand water supply to the fetus, J. Physiol., 445, 569, 1992.67. Gibson, K.J. and Lumbers, E.R., The roles of arginine vasopressin in fetal sodium balance and as amediator of the effects of fetal “stress,” J. Develop. Physiol., 19, 125, 1993.68. Torchinsky, A., Fein, A., and Toder, V., Modulation of embryo sensitivity to teratogen by nonspecificintrauterine immunopotentiation, Toxicol. Methods, 5, 131, 1995.69. Szabo, K.T. and Brent, R.L., Reduction of drug-induced cleft palate in mice, Lancet, 1, 1296, 1975.70. Hemm, R.D., Arslanoglou, L., and Pollock, J.J., Cleft palate following prenatal food restriction inmice: association with elevated maternal corticosteroids, Teratology, 15, 243, 1977.71. Ellington, S., In-vivo and in-vitro studies on the effects of maternal fasting during embryonic organogenesisin the rat, J. Repro. Fertil., 60, 383, 1980.72. Ikemi, N., Imada, J., Goto, T., Shimazu, H., and Yasuda, M., Effects of food restriction on the fetaldevelopment during major organogenesis in rats, Cong. Anom., 33, 363, 1993.73 Petrere, J.A., Rohn, W.R., Grantham II, L.E., and Anderson, J.A., Food restriction during organogenesisin rabbits: effects on reproduction and the offspring, Fundam. Appl. Toxicol., 21, 517, 1993.74. Matsuzawa, T., Nakata, M., Goto, I., and Tsushima, M., Dietary deprivation induces fetal loss andabortions in rabbits, Toxicology, 22, 255, 1981.75. Clark, R.L., Robertson, R.T., Peter, C.P., Bland, J.A., Nolan, T.E., Oppenheimer, L., and Bokelman,D.L., Association between adverse maternal and embryo/fetal effects in norfloxacin-treated and fooddeprivedrabbits, Fundam. Appl. Toxicol., 7, 272, 1986.76. James, W.H., Evidence that mammalian sex ratios at birth are partially controlled by parental hormonelevels at the time of conception, J. Theor. Biol., 180, 271, 1996.77. James, W.H., Exposure to chemicals, offspring sex ratios, and their relevance to teratology [letter],Teratology, 62, 75, 2000.78. Menegola, E., Broccia, M.L., Di Renzo, F., and Giavini, E., Comparative study of sodium valproateinducedskeletal malformations using single or double staining methods, Reprod. Toxicol., 16, 815,2002.79. Sucheston, M.E., Hayes, T.G., and Eluma, F.O., Relationship between ossification and body weight ofthe CD-1 mouse fetus exposed in utero to anticonvulsant drugs, Terat. Carcinog. Mutagen., 6, 537, 1986.© 2006 by Taylor & Francis Group, LLC

420 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION80. Wise, L.D., Beck, S.L., Beltrame, D., Beyer, B.K., Chahoud, I., Clark, R.L., Clark, R., Druga, A.M.,Feuston, M.H., Guittin, P., Henwood, S.M., Kimmel, C.A., Lindstrom, P., Palmer, A.K., Petrere, J.A.,Solomon, H.M., Yasuda, M., and York, R.G., Terminology of developmental abnormalities in commonlaboratory animals, Teratology, 55, 249, 1997.81. Wilson, J.G., Embryological considerations in teratology. In Teratology: Principles and Techniques;Wilson, J.G. and Warkany, J., Eds., The University of Chicago Press, Chicago, 1965, p. 251.82. Stuckhardt, J.L. and Poppe, S.M., Fresh visceral examination of rat and rabbit fetuses used interatogenicity testing, Terat., Carcin. Mutagen., 4, 181, 1984.83. Staples, R.E., Detection of visceral alterations in mammalian species [abstract], Teratology, 9, 37,1974.84. Barrow, M.V. and Taylor, W.J., A rapid method for detecting malformations in rat fetuses, J. Morphol.,127, 291, 1969.85. Sawin, P. and Crary, D., Morphogenetic studies of the rabbit. X. Racial variations in the gall bladder,Anat. Rec., 110, 573, 1951.86. Pitt, J.A., Snyder, L.T., Smith, B.L., Varsho, B.J., Nemec, M.D., and Holson, J.F., An x-ray study ofthe incidence of supernumary ribs and presacral vertebrae in adults rats, mice and rabbits, Teratology,65, 328, 2002.87. Chernoff, N., Rogers, J.M., Turner, C.I., and Francis, B.M., Significance of supernumerary ribs inrodent developmental toxicity studies: postnatal persistence in rats and mice, Fundam. Appl. Toxicol.,17, 448, 1991.88. Foulon, O., Girard, H., Pallen, C., Urtizberea, M., Repetto-Larsay, M., and Blacker, A.M., Inductionof supernumerary ribs with sodium salicylate, Reprod. Toxicol., 13, 369, 1999.89. Wickramaratne, G.A. de S., The post-natal fate of supernumerary ribs in rat teratogenicity studies, J.Appl. Toxicol., 8, 91, 1988.90. Foulon, O., Jaussely, C., Repetto, M., Urtizberea, M., and Blacker, A.M., Postnatal evolution ofsupernumerary ribs in rats after a single administration of sodium salicylate, J. Appl. Toxicol., 20,205, 2000.91. Green, E.L., The inheritance of a rib variation in the rabbit, Anat. Rec., 74, 47, 1939.92. Greenaway, J.G., Partlow, G.D., Gonsholt, N.L., and Fisher, R.S., Anatomy of the lumbosacral spinalcord in rabbits, J. Amer. Animal Hosp. Assoc., 37, 27, 2001.93. Beck, S.L., Assessment of adult skeletons to detect prenatal exposure to trypan blue in mice, Teratology,28, 271, 1983.94. Moore, K. and Persaud, T., Eds., The Developing Human: Clinically Oriented Human Embryology,6th ed., W.B. Saunders, Philadelphia, 1998, p. 414.95. Holson, J.F., Kimmel, C.A., Hogue, C.I., and Carlo, G.L., Suitability of experimental studies forpredicting hazards to human development, in Proceedings of the 1981 Annual Winter Meeting of theToxicology Forum, Arlington, VA, 1981.96. Holson, J.F., DeSesso, J., Jacobson, C., and Farr, C., Appropriate use of animal models in theassessment of risk during prenatal development: an illustration using inorganic arsenic, Teratology,62, 51, 2000.97 National Academy of Sciences, Evaluating Chemical and Other Agent Exposures for Reproductiveand Developmental Toxicity, National Academy Press, Washington, DC, 2001.98. Takayama, S., Akaike, M., Kawashima, K., Takahashi, M., and Kurokawa, Y., A collaborative studyin Japan on optimal treatment period and parameters for detection of male fertility disorders inducedby drugs in rats, J. Am. Coll. Toxicol., 14, 266, 1995.99. Barton, H.A. and Andersen, M.E., Endocrine active compounds: from biology to dose responseassessment, Crit. Rev. Toxicol., 28, 363, 1998.100. Sterner, T.R. and Mattie D.R., Perchlorate literature review and summary: developmental effects, metabolism,receptor kinetics and pharmacological uses, NTIS Technical Report (NTIS/ADA367421), 1998.101. Kato, M., Fukunishi, K., Ikegawa, S., Higuchi, H., Sato, M., Horimoto, M., and Ito, S., Overview ofstudies on rat sperm motion analysis using a Hamilton-Thorne Sperm Analyzer — collaborativeworking study, J. Toxicol. Sci., 26, 285, 2001.102. Linder, R.E., Strader, L.F., Slott, V.L., and Suarez, J.D., Endpoints of spermatotoxicity in the rat aftershort duration exposures to fourteen reproductive toxicants, Repro. Toxicol., 6, 491, 1992.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 421103. O’Connor, J.C., Davis, L.G., Frame, S.R., and Cook, J.C., Evaluation of a Tier I screening batteryfor detecting endocrine-active compounds (EACs) using the positive controls testosterone, coumestrol,progesterone, and RU486, Toxicol. Sci., 54, 338, 2000.104. Russell, L.D., Ettlin, R., Sinha Hikim, A.P., and Clegg, E.D., Histological and HistopathologicalEvaluation of the Testis, Cache River Press, Clearwater, FL, 1990.105. Long, J.A. and Evans, H.M., The oestrous cycle in the rat and its associated phenomena, Mem. Univ.Calif., 6, 1, 1922.106. Foster, P.M.D., Reproductive performance, in An Evaluation and Interpretation of ReproductiveEndpoints for Human Health Risk Assessment, Daston, G. and Kimmel, C., Eds., International LifeSciences Institute, Washington, DC, 1998, p. 5.107. NTP-CERHR Expert Panel Report on the Reproductive and Developmental Toxicity of 1-Bromopropane,National Toxicology Program (NTP), Center for the Evaluation of Risks to Human Reproduction(CERHR), 2002.108. Horimoto, M., Takeuchi, K., Iijima, M., and Tachibana, M., Reproductive and developmental toxicitystudies with amlodipine in rats and rabbits, Oyo Yakuri, 42, 167, 1991.109. Zippelius, H.M. and Schleidt, W.M., Ultraschallante bei jungen Mausen [Ultrasounds in young mice],Naturwissenschaften, 43, 502, 1956.110. Noirot, E. and Pye, D., Sound analysis of ultrasonic distress calls of mouse pups as a function oftheir age, Animal Behavior, 17, 340, 1969.111. Nicholas, K.R. and Hartmann, P.E., The foetoplacental unit and the initiation of lactation in the rat,Aust. J. Biol. Sci., 34, 455, 1981.112. Orbach, J., Spontaneous ejaculation in rat, Science, 134, 1072, 1961.113. Kihlström, J.E., Diurnal variation in the spontaneous ejaculations of the male albino rat, Nature, 209,513, 1966.114. Beach, F.A., Variables affecting “spontaneous” seminal emission in rats, Physiol. Behav., 15(1), 91, 1975.115. Orbach, J., Miller, M., Billimoria, A., and Solhkhah, N., Spontaneous seminal ejaculation and genitalgrooming in rats, Brain Res., 5, 520, 1967.116. Gallavan, Jr., R.H., Holson, J.F., Stump, D.G., Knapp, J.F., and Reynolds, V.L., Interpreting thetoxicologic significance of alterations in anogenital distance: potential for confounding effects ofprogeny body weights, Repro. Toxicol., 13, 383, 1999.117. Clark, R.L., Endpoints of reproductive system development, in An Evaluation and Interpretation ofReproductive Endpoints for Human Health Risk Assessment, Daston, G. and Kimmel, C.A., Eds.,International Life Sciences Institute, Washington, DC, 1998, p. 10.118. Ashby, J. and Lefevre, P.A., The peripubertal male rat assay as an alternative to the Hershbergercastrated male rat assay for the detection of anti-androgens, oestrogens and metabolic modulators, J.Appl. Toxicol., 20, 35, 2000.119. Mocarelli, P., Gerthoux, P.M., Ferrari, E., Patterson, Jr., D.G., Kieszak, S.M., Brambilla, P., Vincoli,N., Signorini, S., Tramacere, P., Carreri, V., Sampson, E.J., Turner, W.E., and Needham, L.L., Paternalconcentrations of dioxin and sex ratio of offspring, Lancet, 355(9218), 1838, 2000.120. Ryan, J.J., Amirova, Z., and Carrier, G., Sex ratios of Russian pesticide producers exposed to dioxin,Environ. Health Persp., 110, A 699, 2002.121. Schnorr, T.M., Lawson, C.C., Whelan, E.A., Dankovic, D.A., Deddens, J.A., Piacitelli, L.A., Reefhuis,J., Sweeney, M.H., Conally, L.B., and Fingerhut, M.A., Spontaneous abortion, sex ratio, and paternaloccupational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin, Environ. Health Perspec., 109, 1127,2001.122. International Conference on Harmonisation (ICH), Tripartite Guideline on Detection of Toxicity toReproduction for Medicinal Products, Fed. Regist., Section 4.1.2, September 22, 1994.123. U.S. Environmental Protection Agency, Health Effects Test Guideline, Developmental NeurotoxicityStudy, OPPTS 870.6300, August 1998.124. Organization for Economic Cooperation and Development, Guideline for the testing of chemicals,Developmental Neurotoxicity Study, Proposal for a New Guideline 426, September 2003 Draft.125. Bolon, B., Bucci, J.J., Warbritton, A.R., Chen, J.J., Mattison, P.R., and Heindel, J.J., Differentialfollicle counts as a screen for chemically-induced ovarian toxicity in mice: results from continuousbreeding bioassays, Fundam. Appl. Toxicol., 39, 1, 1997.© 2006 by Taylor & Francis Group, LLC

422 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION126. Bucci, T.J., Bolon, B., Warbritton, A.R., Chen, J.J., and Heindel, J.J., Influence of sampling on thereproducibility of ovarian follicle counts in mouse toxicity studies, Reprod. Toxicol., 11, 689, 1997.127. Holson, J.F., Interpretation of low-incidence findings in developmental and reproductive toxicitystudies, presented at the 2002 Annual Winter Meeting of the Toxicology Forum, Washington, DC, 2002.128. Cook, J.C., Jacobson, C.F., Gao, F., Tassinari, M.S., Hurtt, M.E., and DeSesso, J.M., Analysis of thenonsteroidal anti-inflammatory drug literature for potential developmental toxicity in rats and rabbits,Birth Defects Res. Part B Dev. Reprod. Toxicol., 68, 5, 2003.129. Nisbet, I.C.T. and Karch, N.J., Chemical Hazards to Human Reproduction, Noyes Data Corp., ParkRidge, IL, 1983.130. Hemminki, K. and Vineis, P., Extrapolation of the evidence on teratogenicity of chemicals between humansand experimental animals: chemicals other than drugs, Teratog. Carcinog. Mutagen., 5, 251, 1985.131. Kimmel, C.A., Rees, D.C., and Francis, E.Z., Eds., Proceedings of the Workshop on the Qualitativeand Quantitative Comparability of Human and Animal Developmental Neurotoxicity, Neurotoxicol.Teratol., 12, 173, 1990.132. Newman, L.M., Johnson, E.M., and Staples, R.E., Assessment of the effectiveness of animal developmentaltoxicity testing for human safety, Reprod. Toxicol., 7, 359, 1993.133. March of Dimes Birth Defects Foundation, 1996, Birth Defects and Infant Mortality: A National andRegional Profile, March of Dimes Birth Defects Foundation, New York.134. Frohberg, H., Gleich, J., and Unkelbach, H.D., Reproduction toxicological studies on cefazedone,Arzneim. Forsch., 29, 419, 1979.135. Duffy, J.Y., Overmann, G.J., Keen, C.L., and Daston, G.P., Cardiac abnormalities induced by zincdeficiency are associated with alterations in the expression of genes regulated by zinc-finger transcriptionfactors, Teratology, 59, 382, 1999.136. Stump, D.G., Holson, J.F., Kirkpatrick, D.T., Reynolds, V.L., Siddiqui, W.H., and Meeks, R.G.,Evaluation of octamethylcyclotetrasiloxane (D4) in a 2-generation reproductive toxicity study in rats,The Toxicologist 54, 370, 2000.137. Youreneff, M.A., Singh, A.R., Hazelette, J.R., Yau, E.T., and Traina, V.M., Teratogenic evaluation ofbenazepril hydrochloride in mice, rats, and rabbits, J. Am. Coll. Toxicol., 9, 647, 1990.138. Spence, S.G., Allen, H.L., Cukierski, M.A., Manson, J.M., Robertson, R.T., and Eydelloth, R.S.,Defining the susceptible period of developmental toxicity for the the AT 1 -selective angiotensin IIreceptor antagonist Losartan in rats, Teratology, 51, 367, 1995.139. Spence, S.G., Cukierski, M.A., Manson, J.M., Robertson, R.T., and Eydelloth, R.S., Evaluation ofthe reproductive and developmental toxicity of the AT 1 -selective angiotensin II receptor antagonistLosartan in rats, Teratology, 51, 383, 1995.140. Sweeney, L.M., Tyler, T.R., Kirman, C.R., Corley, R.A., Reitz, R.H., Paustenbach, D.J., Holson, J.F.,Whorton, M.D., Thompson, K.M., and Gargas, M.L., Proposed occupational exposure limits for selectethylene glycol ethers using PBPK models and Monte Carlo simulations, Toxicol. Sci., 62, 124, 2001.141. Terry, K.K., Elswick, B.A., Stedman, D.B., and Welsch, F., Developmental phase alters dosimetryteratogenicityrelationship for 2-methoxyethanol in CD-1 Mice, Teratology, 49, 218, 1994.142. Clarke, D.O., Elswick, B.A., Welsch, F., and Conolly, R.R., Pharmacokinetics of 2-methoxyethanoland 2-methoxyacetic acid in the pregnant mouse: a physiologically based mathematical model, Toxicol.Appl. Pharm., 121, 239, 1993.143. Nau, H., Teratogenicity of isotretinoin revisited: species variation and the role of all-trans-retinoicacid, J. Am. Acad. Dermatol., 45, S183, 2001.144. Tzimas, G., Thiel, R., Chahoud, I., and Nau, H., The area under the concentration-time curve of alltrans-retinoicacid is the most suitable pharmacokinetic correlate to the embryotoxicity of this retinoidin the rat, Toxicol. Appl. Pharm., 143, 436, 1997.145. Collins, M.D., Tzimas, G., Hummler, H., Bürgin, H., and Nau, H., Comparative teratology andtransplacental pharmacokinetics of all-trans-retinoic acid, 13-cis-retinoic acid, and retinyl palmitatefollowing daily administration in rats, Toxicol. Appl. Pharm., 127, 132, 1994.146. Nau, H., Embryotoxicity and teratogenicity of topical retinoic acid, Skin Pharmacol., 6(Suppl.1), 35,1993.147. U.S. Environmental Protection Agency, Standard Evaluation Procedure, Developmental Toxicity Studies,Health Effects Division, Office of Pesticides Programs, Washington, DC, 1993.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY STUDY FINDINGS 423148. Arlitt, A.H., The effect of alcohol on the intelligent behavior of the white rat and its progeny, Psychol.Monogr., 26, 1, 1919.149. Thiersch, J.B. and Phillips, F.S., Effect of 4-amino-pteroylglutamic acid (aminopterin) on earlypregnancy, Proc. Soc. Exp. Biol. Med., 74, 204, 1950.150. Shoeneck, F.J., Cigarette smoking in pregnant women, N.Y. State J. Med., 41, 1945, 1941.151. Greene, R.R., Burrill, M.W., and Ivy, A.C., Experimental intersexuality. The effects of estrogens onthe antenatal sexual development of the rat, Am. J. Anat., 67, 305, 1940.152. Geber, W.F. and Schramm, L.C., Comparative teratogenicity of morphine, heroin and methadone inthe hamster, Pharmacologist, 11, 248, 1969.153. Russell, L.B, X-ray induced developmental abnormalities in the mouse and their use in the analysisof embryological patterns, J. Exp. Zool., 114, 545, 1950.154. Matsumoto, H., Koya, G., and Takeuchi, T., Fetal Minamata disease: A neuropathological study oftwo cases of intrauterine intoxication by methyl mercury compound, J. Neuropathol. Exp. Neurol.,24, 563, 1965.155. Taki, I., Hisanaga, S., and Amagase, Y., Report on Yusno (chlorobiphenyls poisoning). Pregnant womenand their fetuses, Fukuoka Acta Med., 60, 471, 1969.156. van Wagenen, G. and Hamilton, J.B., The experimental production of pseudohermaphroditism in themonkey, Essays Biol., 583, 1943.157. Lenz, W., Kindliche Missbildungen nach Medikament wahrend der Draviditat?, Dtsch. Med. Wochenschr.,86, 2555, 1961.158. McBride, W.G., Thalidomide and congenital abnormalities, Lancet, 2, 1358, 1961.159. Holson, J.F., Estimation and extrapolation of teratologic risk, in Proceedings of NATO Conference onin vitro Toxicity Testing of Environmental Agents: Current and Future Possibilities, NATO, MonteCarlo, Monaco, 1980.160. Brown, N.A. and Fabro, S.F., The value of animal teratogenicity testing for predicting human risk,Clin. Obstet. Gynecol., 26(2), 467, 1983.161. Buelke-Sam, J. and Mactutus, C., Workshop on the qualitative and quantitative comparability of humanand animal developmental neurotoxicity: testing methods in developmental neurotoxicity for use inhuman risk assessment, Neurotoxicol. Teratol., 12, 269, 1990.162. Shepard, T.H., Catalog of Teratogenic Agents, 8th ed., The Johns Hopkins University Press, Baltimore,1995.163. Schardein, J.L., Chemically Induced Birth Defects, 3rd ed., Marcel Dekker, New York, 2000.164. Barlow, S.M. and Sullivan, F.M., Reproductive Hazards of Industrial Chemicals, Academic Press,London, 1982.165. U.S. Congress, Office of Technology Assessment, Reproductive Health Hazards in the Workplace,OTA-BA-266, U.S. Government Printing Office, Washington, DC, 1985, p. 161.166. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Prenatal Developmental ToxicityStudy, Office of Prevention, Pesticides and Toxic Substances (OPPTS) 870.3700, 1998.167. U.S. Food and Drug Administration, International Conference on Harmonisation; Guideline on detectionof toxicity to reproduction for medicinal products. Fed. Regist., 59(No. 183), 1994.168. Organization for Economic Cooperation and Development, Guidelines for Testing of Chemicals.Section 4, No. 414: Teratogenicity, adopted 22 January 2001.169. U.S. Food and Drug Administration, Office of Food Additive Safety, IV.C.9.b., Guidelines for developmentaltoxicity studies, Redbook 2000—Toxicological Principles for the Safety Assessment of FoodIngredients, College Park, MD, 2000.170. U.S. Food and Drug Administration, Guidance for Industry—Considerations for Reproductive ToxicityStudies for Preventive Vaccines for Infectious Disease Indications, Rockville, MD, 2000 (Draft).171. Japanese Ministry of Agriculture, Forestry and Fisheries (MAFF), Developmental Toxicity TestingGuidelines, 1985.172. U.S. Food and Drug Administration, Office of Food Additive Safety, IV.C.9.a., Guidelines for reproductionstudies, Redbook 2000—Toxicological Principles for the Safety Assessment of Food Ingredients,College Park, MD, 2000.173. OECD, OECD Guideline for Testing of Chemicals, Proposal for Updating Guideline 416, Two-Generation Reproduction Study, OECD, Paris, France, 2001.© 2006 by Taylor & Francis Group, LLC

424 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION174. U.S. Environmental Protection Agency, Health Effects Test Guidelines; Reproduction and FertilityEffects, Office of Prevention, Pesticides and Toxic Substances (OPPTS) 870.3800, 1998b.175. ICH Expert Working Group, ICH Harmonised Tripartite Guideline, Detection of Toxicity to Reproductionfor Medicinal Products, S5A, International Conference on Harmonisation, Geneva, Switzerland,1994.176. ICH Expert Working Group, Maintenance of the ICH Guidelines on Toxicity to Male Fertility, anAddendum to the ICH Harmonised Tripartite guideline, Detection of Toxicity to Reproduction forMedicinal Products, S5B(M), International Conference on Harmonisation, Geneva, Switzerland, 1994.© 2006 by Taylor & Francis Group, LLC

CHAPTER 10Testing for Reproductive Toxicity*Robert M. ParkerCONTENTSI. Introduction ........................................................................................................................428A. Brief Overview of DART Study Guidelines.............................................................4281. Fertility Studies....................................................................................................4292. Developmental Toxicity Studies..........................................................................4303. Reproduction Studies...........................................................................................4304. Reproductive Assessment by Continuous Breeding ...........................................4315. Developmental Neurotoxicity Study ...................................................................4326. Combined Repeated Dose Toxicity Study with the Reproduction andDevelopmental Toxicity Screening Test..............................................................4337. Dominant Lethal Study .......................................................................................435B. Endpoints for Reproductive Studies..........................................................................4361. Couple-Mediated Endpoints ................................................................................4362. Male-Specific Endpoints and Paternally Mediated Effects ................................4363. Female-Specific Endpoints and Maternally Mediated Effects ...........................436II. Reproductive Performance.................................................................................................437A. Overview....................................................................................................................437B. Treatment Duration for Males in Fertility and Reproduction Studies .....................437C. Cohabitation and Evidence of Mating ......................................................................4381. Vaginal Plugs .......................................................................................................4382. Vaginal Smears ....................................................................................................438D. Reproductive Indices .................................................................................................4391. Male Mating and Fertility Index .........................................................................4392. Female Fertility Index .........................................................................................4403. Mean Gestational Length ....................................................................................4424. Litter Size ............................................................................................................4425. Live-Birth Index ..................................................................................................4436. Survival Index......................................................................................................4437. Sex Ratio..............................................................................................................444III. Male Reproductive Toxicology..........................................................................................444A. Biology.......................................................................................................................444B. Observations ..............................................................................................................445C. Male Body Weight and Food Consumption..............................................................445D. Male Necropsy...........................................................................................................445* The views expressed in this chapter are those of the author and do not necessarily reflect the views of Hoffmann-LaRoche, Inc.425© 2006 by Taylor & Francis Group, LLC

426 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONE. Male Reproductive Organ Weight.............................................................................446F. Histopathological Evaluation of the Male Reproductive Organs .............................4461. Testis ....................................................................................................................4472. Epididymides, Accessory Sex Glands and Pituitary...........................................448G. Sperm Evaluations .....................................................................................................4491. Sperm Number or Sperm Concentration.............................................................4492. Computer-Assisted Sperm Analysis....................................................................4503. Cauda Epididymal Sperm Motility .....................................................................4544. Sperm Morphology..............................................................................................456H. Seminiferous Epithelium Staging..............................................................................4571. Spermatogenesis ..................................................................................................4572. The Staging of Spermatogenesis.........................................................................458IV. Female Reproductive Toxicology ......................................................................................458A. Biology.......................................................................................................................458B. Observations ..............................................................................................................459C. Female Body Weight .................................................................................................459D. Female Necropsy .......................................................................................................4591. Gestation Day 13 Uterine Examination (Segment I Studies).............................4592. Gestation Day 20 or 21 Term Cesarean Section (Segment II Studies)..............4593. End of Lactation (Segment III Studies) ..............................................................460E. Female Reproductive Organ Weight .........................................................................4601. Ovary....................................................................................................................4602. Uterus, Oviducts, Vagina, and Pituitary Gland...................................................460F. Histopathological Evaluation of the Female Reproductive Organs..........................4611. Ovaries .................................................................................................................4612. Uterus...................................................................................................................4613. Oviducts ...............................................................................................................4614. Vagina and External Genitalia.............................................................................4615. Pituitary................................................................................................................4626. Mammary Gland and Lactation ..........................................................................4627. Reproductive Senescence ....................................................................................4638. Developmental and Pubertal Alterations.............................................................463G. Vaginal Cytology .......................................................................................................4631. Description...........................................................................................................4632. Vaginal Lavage Methodology..............................................................................4643. Estrous Cycle Staging .........................................................................................4654. Estrous Cycle Evaluation ....................................................................................465H. Corpora Lutea Counts ...............................................................................................466I. Uterine Evaluation .....................................................................................................4671. Assessment of Resorptions..................................................................................4672. Assessment of Implantation in “Apparently Nonpregnant” Dams.....................467J. Oocyte Quantitation...................................................................................................467V. Parturition and Litter Evaluations......................................................................................468A. Parturition ..................................................................................................................468B. Milk Collection..........................................................................................................469C. Litter Evaluations.......................................................................................................4691. Litter Observations ..............................................................................................4692. Gender Determination .........................................................................................4703. Viability................................................................................................................4704. External Alterations .............................................................................................4705. Missing Pups........................................................................................................470© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 427VI.6. Pups Found Dead.................................................................................................4707. Dam Dies before Scheduled Sacrifice.................................................................4708. All Pups in a Litter Die.......................................................................................4719. Gender Determination Errors ..............................................................................471Pup Evaluations..................................................................................................................471A. Pup Body Weights .....................................................................................................4711. Pup Birth Weight .................................................................................................4712. Postnatal Pup Body Weights ...............................................................................4713. Crown-Rump Length ...........................................................................................4714. Anogenital Distance ............................................................................................472B. Developmental Landmarks........................................................................................4721. Balanopreputial Separation..................................................................................4722. Vaginal Patency ...................................................................................................4733. Day of First Estrus ..............................................................................................4734. Pinna Unfolding...................................................................................................4735. Hair Growth .........................................................................................................4736. Incisor Eruption ...................................................................................................4747. Eye Opening ........................................................................................................4748. Nipple Retention..................................................................................................4749. Testes Descent .....................................................................................................474C. Behavioral or Reflex Testing.....................................................................................4741. Surface-Righting Reflex ......................................................................................4742. Cliff Avoidance ....................................................................................................4743. Forelimb Placing..................................................................................................4754. Negative Geotaxis Test ........................................................................................4755. Pinna Reflex.........................................................................................................4756. Auditory Startle Reflex........................................................................................4757. Hind Limb Placing ..............................................................................................4758. Air Righting Reflex .............................................................................................4759. Forelimb Grip Test (Categorical) ........................................................................47610. Pupil Constriction Reflex ....................................................................................476D. Examination of Rodents for Hypospadias ................................................................4761. In-life Procedure for Male Rodents ....................................................................4762. Postmortem Procedure for Male Rodents...........................................................4773. Postmortem Procedure for Female Rodents .......................................................477VII. Functional Observation Battery .........................................................................................477A. Adult FOB .................................................................................................................4771. In-Cage Observations ..........................................................................................4772. Removing from Cage Observations ....................................................................4783. Standard Open Field Arena Observations...........................................................4784. Manipulations in the Standard Open Field Arena ..............................................4785. Grip Strength Manipulations ...............................................................................4786. Foot Splay............................................................................................................4787. Pupillary Response ..............................................................................................4798. Body Temperature................................................................................................479B. FOB in Preweanling Rats..........................................................................................4791. In Cage Observations ..........................................................................................4792. Removing from Cage Observations ....................................................................4793. Standard Open Field Arena Observations...........................................................4794. Manipulations in the Standard Open Field Arena ..............................................480References ......................................................................................................................................480© 2006 by Taylor & Francis Group, LLC

428 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONI. INTRODUCTIONThe objective of this chapter is to provide the methodologies for evaluations used during theperformance of reproductive toxicity studies. Traditional developmental and reproductive toxicity(DART) studies in rodents are a major source of data on the effects of potential developmental andreproductive toxicants. Reproductive toxicities include structural and functional alterations that mayaffect reproductive competence (fertility, parturition, and lactation). Evaluations of fertility, pregnancy,lactation, and maternal and paternal behaviors provide measures of the consequences ofreproductive injury. These evaluations provide information concerning gonadal function, estruscyclicity, mating behavior, conception, parturition, lactation, weaning, and the growth and developmentof the offspring. Developmental toxicities are generally those that affect the F 1 or F 2generations. The four manifestations of developmental toxicity as described by Wilson 1 are mortality,dysmorphogenesis (structural alterations), alterations to growth, and functional alterations.Mortality due to developmental toxicity may occur at any time from conception to adulthood.Dysmorphogenic effects are generally seen as malformations or variations to the skeleton or visceraof the offspring. Alterations to growth are generally seen as growth retardation, although excessivegrowth or early maturation may also be seen. Functional alterations could include any persistentchange in normal physiologic or biochemical function, but typically only reproductive function anddevelopmental neurobehavioral effects are measured in these studies.A. Brief Overview of DART Study GuidelinesThe purposes of DART studies are to determine whether a test substance has the potential to causeadverse effects on the male and female reproductive system or the developing conceptus, and todetermine a developmental and/or reproductive no observable adverse effect level (NOAEL) forthe test substance. The developmental or reproductive NOAEL is the highest treatment level testedthat shows no developmental or reproductive effects, respectively. It should be noted that DARTstudy designs were not developed to cover every male and female reproductive parameter, developmentalparameter, specific target organ, or mechanism of toxicity. For example, traditional DARTstudies do not provide information on maternal endocrine status, the levels of an agent or itsmetabolites in the milk and the fetus or pup, placental pathology, or reproductive senescence.The reproductive cycle has historically been divided into three segments for DART studies.Segment I study (also known as reproduction and fertility studies) covers the period of premating,cohabitation and mating, and early pregnancy through implantation (gestation day [GD] 6 in therat). Segment II study (also known as teratology or developmental toxicity studies) covers pregnancyfrom implantation through major organogenesis (closure of the hard palate; GD 15 in the rat) upto the day before delivery. Segment III study (perinatal and postnatal studies) covers late pregnancyand postnatal development (usually until weaning at postnatal day [PND] 21). Multigenerationalstudies (studies covering two to three generations) are required for chemicals that pose a low-levelchronic exposure risk in humans. Combination studies such as Segment I and II and Segment IIand III studies are acceptable as long as all required parameters and minimum study requirementsfor each segment are met. Another example of an acceptable combinatory study is the 90-DayRepeated Dose Subchronic Study with a One-Generation Reproduction Study.DART study guidelines have been promulgated by the U.S. Food and Drug Administration(FDA), 2–7 the U.S. Environmental Protection Agency (EPA), 8–11 the Japan Ministry of Agriculture,Forestry and Fisheries (MAFF), 12,13 Canada, 14,15 Great Britain, 16 World Health Organization, 17 andthe Organization for Economic Cooperation and Development (OECD). 18–22 The ICH (InternationalConference on Harmonization of the Technical Requirements for Registration of Pharmaceuticalsfor Human Use) 6,7,23 has been concerned with unifying the study guideline requirements for pharmaceuticalsafety assessment for the European Union, Japan, and the FDA. Guideline compliantreproductive and fertility studies must be performed in accordance with associated Good Laboratory© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 429Practice Regulations, 24–26 and biosafety 27 and animal welfare guidelines. 28,29 A number of theseguidelines and additional references for reproduction studies can be found in the appendixes forChapter 19.1. Fertility StudiesFertility studies 4–7,10 (also known as Segment I studies) test for toxic effects resulting from treatmentwith the test substance before mating (males and/or females), during cohabitation and mating, anduntil implantation (GD 6). For females this study design detects effects on the libido, estrous cycle,ovulation, mating behavior, oviductal transport, development of preimplantation stages of theembryo, and implantation. For males this study design permits detection of functional effects onlibido, mating, and sperm quality (motility, count, and morphology) that may not be detected byhistological examinations of the male reproductive organs alone. Males are treated with the testsubstance 2, 4, or 10 weeks (depending upon guideline requirements) prior to cohabitation andmating, and until termination. Females are treated 2 weeks prior to cohabitation, through cohabitationand mating, and until implantation (usually GD 6). The dam is killed on either GD13 or 21and necropsied, and the fetuses are evaluated for viability (see Figure 10.1). Reproductive endpointsevaluated include mating, pregnancy rate, implantation sites, conceptus viability, corpora lutea, preandpostimplantation losses, and litter size. The Combined Male and Female Fertility Study (FDASegment I and ICH s5 and s5a) guidelines are described in detail in Appendix 1 of Chapter 19. Inmale or female fertility studies, the test substance is administered only to males or females,respectively.F 0 PrematingMatingF 0 GestationTreatment Period0 2Weeks46GD 21F1 Fetal ExamStudy TaskNontreatment periodF 0 PrematingMatingF 0 GestationEstrous smearsOffspring Development &Behavior Testing0 24Weeks6GD 7GD 21F1 Fetal ExamF 0 Gestation orF 0 Premating Matingpostmating0 2 4 Weeks 6902Weeks4GD 7GD 13F1 Fetal ExamFigure 10.1Fertility study schematics (FDA Segment I and ICH s5 and s5a).© 2006 by Taylor & Francis Group, LLC

430 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONGD 0 GD 15F 0 GestationGD 0 F 0 GestationGD 20GD 6GD 21F1 Fetal ExamGD 6 GD 21F1 Fetal ExamFigure 10.2Developmental toxicity study schematics (FDA Segment II and EPA, OPPTS 870.3700 andOECD 414).2. Developmental Toxicity StudiesDevelopmental toxicity 9,14,17 or teratology studies (also known as Segment II studies) are used todetect adverse effects of treatment on the pregnant female and development of the embryo andfetus from implantation (GD 6) to closure of the hard palate (GD 15); however, more recently thetreatment period has been extended to GD 20 (see Figure 10.2), with necropsy and fetal evaluationon GD 21. Rabbit teratology studies dose on GD 6 through 18 with necropsy and fetal evaluationsperformed on GD 29. The endpoints assessed include maternal toxicity (body weight, clinical andnecropsy observations, feed consumption), embryo-fetal death (early and late resorptions), growthretardation, and structural alterations (malformations and variations). Functional deficits cannot bedetermined using this paradigm. See Tyl and Marr 30 for detailed discussion of the developmentaltoxicity study. This study is also described in Appendixes 1 and 2 of Chapter 19.3. Reproduction StudiesThe perinatal and postnatal study 4–7 (also known as the Segment III study) design detects adverseeffects on the pregnant and lactating female and on development of the conceptus and the offspringfollowing treatment of the female from implantation (GD 6) or the end of organogenesis (GD 15)through lactation and weaning (usually PND 21). This study is described in detail in Appendix 1and 2 of Chapter 19.The one-generation reproduction testing guideline (OECD 415 18 ) requires treatment of the P 1males at least 10 weeks and females at least 2 weeks prior to cohabitation, during cohabitation andmating, until termination after the mating period for the males, and through gestation and lactationuntil weaning at PND 21 for the females. F 1 pups are killed on PND 21 and evaluated (see Figure10.3). This study is designed to provide information concerning the effects of a test substance onmale and female reproductive capacity, including gonadal function, the estrous cycle, matingbehavior, conception, gestation, parturition, lactation, and weaning, and on the growth and developmentof the offspring. The study may also provide information about the effects of the testsubstance on neonatal morbidity, mortality, and target organs in the offspring (F 1 generation), andpreliminary data on prenatal and postnatal developmental toxicity. It may even serve as a guide forsubsequent tests.The two-generation reproduction testing guideline (OPPTS 870.3800 10 and OECD 416 22 ) isdesigned to provide information concerning the effects of a test substance on the integrity andperformance of the male and female reproductive systems, as described above for the one-generationstudy, while including an F 2 generation. The study design is similar to the one-generation studyfor the P 1 and F 1 generations; however, the test substance is administered postweaning to selectedF 1 offspring (usually one male and one female per litter) during their growth into adulthood, duringcohabitation and mating, and during gestation, parturition, and birth, continuing until the F 2 generationis weaned (PND 21) (see Figure 10.4). For further description of this study, refer to theOPPTS 870.3800 guideline presented in Appendix 2 of Chapter 19.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 431F 0F 0 LactationGD 0 GD 15DeliveryUterine ExamLD 21F 1 LactationF 1 PostweaningF 1 MatingF 1 Gestation03WeeksGD 13F 0F 0 LactationGD 0 GD 6DeliveryLD 21F 1 LactationF 1 PostweaningF 1 MatingF 1 Gestation03WeeksGD 21F 2 Fetal ExamTreatment PeriodStudy TaskNontreatment periodEstrous smearsOffspring Development &Behavior TestingFigure 10.3 Reproduction study schematics (OPPTS 870.3800 and OECD 415).4. Reproductive Assessment by Continuous BreedingData from reproductive assessment by continuous breeding (RACB) studies 31,32 identify hazards toreproduction, help to characterize their toxic effects, and indicate dose-response relationships. EachRACB study is separated into four tasks (see Figure 10.5).Task 1 is a modified 4- to 5-week range-finding study consisting of a 1-week treatment periodfor five treatment levels and concurrent control group, followed by a 1-week cohabitation treatmentperiod, a 3-week gestation treatment period, and treatment continuing until delivery.Task 2 is the main study. A control plus three treatment levels are treated for 1 week prior tocohabitation, and for 18 weeks while the rats are cohoused as breeding pairs. Normally, four littersare delivered per adult pair during the 18-week cohabitation period. The last litter is weaned atPND 21, with one pup per gender per litter selected to continue in Task 4. If Task 2 results arenegative, then Task 3 is not conducted and Task 4 is conducted with control and high treatmentgroups only. If Task 2 shows toxic effects, Task 3 is performed, and Task 4 requires all four groups(the control and three treatment groups).© 2006 by Taylor & Francis Group, LLC

432 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONF 0 PrematingMating Gestation Lactation0 DeliveryF 1 Postweaning Mating GestationLactationWeaningDeliveryF 0 PrematingMating GestationLactationF 2 Lactation0 DeliveryF 1 Postweaning Mating Gestation LactationTreatment PeriodStudy TaskNontreatment periodWeaningDeliveryEstrous smearsOffspring Development &Behavior TestingF 2b LactationFigure 10.4 Two-generation reproduction study schematics (OPPTS 870.3800 and OECD 416).Task 3, if required, is the crossover mating trial (no treatment required) that is performed todetermine whether the effects are couple-mediated or single-sex mediated.Task 4 is the F 1 generation test. Treatment with the test substance starts at weaning (PND 21),with each pup receiving the same treatment that the P 1 generation received. When 80 days of age,F 1 breeding pairs are cohabited for 1 week and separated when mating is confirmed. The F 1 femalecontinues on treatment and delivers its litter (F 2 ), and the adult F 1 and neonatal F 2 generationanimals are killed and necropsied.5. Developmental Neurotoxicity StudyThe developmental neurotoxicity study (OPPTS 870.6300 33 and OECD 426 34 ) design providesinformation for use in evaluating the potential for neurotoxic effects in offspring after exposure toa test substance in utero and/or via maternal milk or by direct dosing of the pup during the lactationperiod. Female rats are administered the test substance once daily beginning on GD 6 and continuingthrough PND 21. If there is no lactational transfer of the test substance or active metabolites, thendirect dosing of the pups beginning as early as PND 4 may be required. Female rats will be evaluatedfor adverse clinical signs observed during parturition, and for the duration of gestation, for littersize, live litter size, and pup viability at birth. F 1 generation males and females are tested as pupsand as adults for spontaneous locomotor activity, learning and memory, reflex development, andtime to sexual maturation. The functional observational battery (FOB) for pups and adults is© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 433Control*DL******Low***Medium******DL*High*Task 1Dose Range-FindingTask 2Continuous BreedingHolding RearingPeriod**Task 4Second Generation−6 0 15Weeks18 21 29 32Figure 10.5Reproductive assessment by continuous breeding schematic. See text for discussion of eachtask. The dark bars indicate treatment while the empty bars indicate no treatment. The twoseparate bars indicate single housing while the thicker bar indicates cohousing. The descendingarrows indicate the birth of a litter. DL, dominant lethal; *, limited necropsy; ** full necropsy.(Adapted from Chapin and Sloan, Environ. Health Perspect., 105(Suppl 1), 199, 1997.)performed at specified intervals. Neurohistopathological examination and morphometry are performedon selected brain regions (see Figure 10.6).6. Combined Repeated Dose Toxicity Study with the Reproductionand Developmental Toxicity Screening TestThe combined repeated dose toxicity study with the reproduction and developmental toxicityscreening test (OECD 422 20 and OPPTS 870.3650 35 ) is being required in the testing program ofhigh production volume (HPV) chemicals. HPV chemicals are those chemicals that are producedin or imported to the United States in amounts over 1 million pounds per year. The Chemical Rightto-KnowInitiative (1998) is the response to an EPA study that found that very little basic toxicityinformation is publicly available on most of the HPV commercial chemicals made and used in theUnited States. For further information concerning this program, go to http://www.epa.gov/opptintr/chemrtk/sidsappb.htm. This study design provides limited information on systemic toxicity andneurotoxicity following repeated treatments. It provides initial information on possible effects onmale and female reproductive performance, such as gonadal function, mating behavior, conception,development of the conceptus, and parturition. It is not an alternative to, nor does it replace, theexisting guidelines in OPPTS 870.3700, OPPTS 870.3800, OPPTS 870.6200, and OPPTS 870.7800,or the equivalent OECD studies. This study cannot be used as a definitive study for the setting ofa developmental or reproductive NOAEL.Males are treated for a minimum of 4 weeks (2 weeks precohabitation, during the mating period,and routinely 2 weeks postmating). Females are treated for 2 weeks precohabitation, during the© 2006 by Taylor & Francis Group, LLC

434 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAcclimation6 daysF 0 Female RatsArrive at TestingFacilityCohabitation1 weekDG 6Start of ExposureExposure Period DG 6 - DL 211st PossibleDelivery(DG 21)Study SchematicLast PossibleDelivery(DG 25)PARTURITIONLactation Day 1End of Exposure(DL 21)Day of Sacrifice(DL 22)SUBSET 1One Pup/Sex/Litter/Dosage GroupF 1GenerationSUBSET 5One Pup/Sex/Litter/Dosage GroupBody WeightsSUBSET 2One Pup/Sex/Litter/Dosage GroupSUBSET 3One Pup/Sex/Litter/Dosage GroupSUBSET 4One Pup/Sex/Litter/Dosage GroupBody WeightsDP 22: SACRIFICE/BRAIN WEIGHTSTen Pups/Sex/Dosage GroupClinical Signs, Body & FeedWeightsClinical Signs, Body & FeedWeightsClinical Signs, Body & FeedWeightsDay 22 SacrificeNEUROPATHOLOGICAL EXAMTen Pups/Sex/Dosage Group(Control & High DosageGroups Read First)Sexual Maturation:DP 28: Vaginal PatencyDP 39: Preputial SeparationSexual Maturation:DP 28: Vaginal PatencyDP 39: Preputial SeparationSexual Maturation:DP 28: Vaginal PatencyDP 39: Preputial SeparationDPs 23 to 25 (retest one week later):PASSIVE AVOIDANCEDPs 14, 18, 22, & 58 to 62 days:MOTOR ACTIVITYDP 80 (approximately):SACRIFICE/BRAIN WEIGHTSTen Pups/Sex/Dosage GroupAbbreviations:DG = Day of (Presumed) GestationDP = Day PostpartumDL = Day of LactationM = MaleF = FemaleDPs 58 to 62 (retest one week later):WATERMAZEDPs 23 & 58 to 62 days:AUDITORY STARTLEHABITUATIONNEUROPATHOLOGICAL EXAMTen Pups/Sex/Dosage Group(Control & High DosageGroups Read First)SACRIFICESACRIFICEFigure 10.6Developmental neurotoxicity study schematic (OPPTS 870.6300 and OECD 426 [draft]).F 0 PrematingMating F 0 Postmatingor Gestation0 2 4 6902Weeks4DeliveryDL 4Figure 10.7Combined repeated dose toxicity study with the reproduction and developmental toxicity screeningtest schematic (OPPTS 870.3650 and OECD 422).mating period, during pregnancy, and at least 4 d after delivery (day before the scheduled kill). AFOB and a motor activity assessment are conducted pretest and after the fourth week of treatment.Clinical pathology data are collected in the fourth week. Tissues are collected at necropsy for acomplete histopathological evaluation. Pups are sacrificed on day 4 of lactation (DL 4) (see Figure 10.7).© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 435MGMGGD 13Uterine ExaminationMGMGMGMGM = Mating PeriodG = Gestation PeriodMGMGMGMGFigure 10.8 Dominant lethal study schematic (OPPTS 870.5450 and OECD 478).7. Dominant Lethal StudyA dominant lethal mutation is a mutation occurring in a germ cell that does not cause dysfunctionof the gamete but is lethal to the fertilized egg or developing embryo. Dominant lethal effects aregenerally thought to be the result of chromosomal damage (structural and numerical anomalies),but gene mutations and toxic effects cannot be excluded. In the dominant lethal study (OECD 478 36and OPPTS 870.5450 37 ), adult males receive a single treatment with the test substance (see Figure10.8). Individual males are mated sequentially (usually at weekly intervals) to one or two virginfemales per week. Females are cohoused with the males for at least the duration of one estrouscycle or until mating has been confirmed. The number of males (usually 25) in each treatmentgroup should be sufficient to provide between 30 and 50 pregnant females per interval. A concurrentpositive control (e.g., triethylenemelamine, cyclophosphamide, or ethyl methanesulfonate), a negative(vehicle) control, and three treatment levels should be used. The females are sacrificed in thesecond half of the pregnancy (usually around GD 13), and the uterine contents are examined forthe numbers of implantation sites and numbers of live or dead conceptuses. The numbers of corporalutea in the ovaries are also counted, and pre-and postimplantation losses are calculated. The matingindex and the fertility index are also calculated (see below). The dominant lethal effect is basedon a statistically significant decrease in the number of live embryos conceived with treated spermcompared with controls. A test substance that produces a statistically significant dose-relateddecrease in the number of live embryos at any one of the test points is considered mutagenic inthe test system.The number of mating periods should ensure that spermatogenesis is adequately covered. Theduration of spermatogenesis is approximately 63 d in the rat; therefore the intervals need to extend© 2006 by Taylor & Francis Group, LLC

436 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONover a total period of approximately 10 weeks. The weekly intervals (from the time of test substanceadministration to mating) permit sampling of germ cells exposed as mature sperm (effects seen infirst week), spermatids (effects seen in the second through fourth weeks), spermatocytes (effectsseen in the fifth through seventh weeks), and spermatogonial and stem cell stages of maturation(effects seen in the eight through tenth weeks).B. Endpoints for Reproductive StudiesThe endpoints of reproductive toxicity in test species can be divided into three categories: couplemediated,male-specific, and female-specific endpoints. 381. Couple-Mediated EndpointsCouple-mediated endpoints are reproductive endpoints where both parents may have a contributoryrole. Couple-mediated endpoints from reproduction studies may include mating behaviors (matingrate, time to mating), pregnancy rate, pre- and postimplantation loss, number of implantation sites,gestation length, litter size (total and live), number of live and dead offspring, sex ratio, birth weight,external and internal malformations and variations, postnatal weight, postnatal structural and functionaldevelopment, offspring survival, and offspring reproduction. Most regulatory reproductionand fertility guidelines specify cohabitation of treated male rats with treated female rats, complicatingthe resolution of gender-specific influences.2. Male-Specific Endpoints and Paternally Mediated EffectsData on the potential male reproductive toxicity of a test substance may be collected from manyguideline studies including acute, subchronic, chronic, reproduction, and fertility studies. Malespecificendpoints that are determined in these studies include, but are not limited to: body weight;mating behavior (libido, mounting, erection, ejaculation); reproductive organ weights; histopathologyof the testes, epididymides, seminal vesicles, prostate, and pituitary; sperm quality (count,viability, morphology, motility); sperm transport, maturation, and storage in the epididymis; productionof seminal fluid; and production and secretion of hormones (LH, FSH, testosterone,estrogen, prolactin, etc.) in the pituitary-hypothalamus-gonadal axis.Male-only treatment with a variety of agents has been shown to produce adverse effects inoffspring including pre-and postimplantation loss, growth and behavioral deficits, and structuralmalformations. 39–41 Many of the chemicals reported to cause paternally mediated effects displaygenotoxic activity and may cause transmissible genetic alterations. However, other mechanisms ofinduction of male-mediated effects are also possible, 42 including nongenetic (e.g., presence of drugin seminal fluid) or epigenetic (e.g., an effect on imprinted gene expression from the paternal alleles,resulting in loss of gene function). Additional studies will be needed to clarify the mechanisms ofpaternal exposure associated with adverse effects on offspring. If a test substance is identified ascausing a paternally mediated adverse effect on offspring in the test species, the effect should beconsidered an adverse reproductive effect.3. Female-Specific Endpoints and Maternally Mediated EffectsThe reproductive life cycle of the female may be divided into several phases (embryo-fetal,prepubertal, cycling adult, pregnant, lactating, and reproductively senescent). Detailed descriptionsof these phases are beyond the scope of this chapter but are readily available. 43 Detailed descriptionsof maternally mediated effects on development are beyond the scope of this chapter but are describedby Hood and Miller. 44 Studies should be conducted to detect adverse effects occurring during anyof these phases. Regulatory developmental and reproductive studies primarily detect adverse effects© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 437on the female’s ability to become pregnant, on pregnancy outcome, and on offspring survival anddevelopment. Adverse alterations in the nonpregnant female reproductive system have been reportedat treatment levels below those that result in reduced fertility or produce adverse effects onpregnancy or pregnancy outcomes. 45–47Female-specific endpoints of reproductive toxicity include, but are not limited to: organ weights,histopathology, estrous cycle onset, length, and other characteristics, ovarian follicular development,ovulation, mating behavior, production and secretion of hormones in the pituitary-hypothalamusgonadalaxis, fertility, gestation length, number of corpora lutea, pre- and postimplantation losses,parturition, lactation, nursing behavior, uterine decidualization and implantation, placentation, andsenescence.Unlike that of the male, the status of the adult female reproductive system fluctuates. Innonpregnant rats, the ovarian and uterine structures (and other reproductive organs) change throughoutthe estrous cycle. Although not cyclic, other changes normally occur during pregnancy, lactation,and return to cyclicity after lactation. These normal fluctuations may affect or confound theevaluation of female reproductive endpoints. It is important to be aware of the reproductive statusof the female at necropsy, including estrous cycle stage. This facilitates interpretation of effectswith endpoints, such as uterine weight and histopathology of the ovary, uterus, and vagina. 38A. OverviewII. REPRODUCTIVE PERFORMANCEAlthough methods are available to quantify aspects of sexual performance in most test species,routinely no direct evaluation of sexual behavior is performed in most regulatory reproductivetoxicology studies. Sexual behavior requires complex interactions between the neural, endocrine,and reproductive components; therefore, it is susceptible to disruption by a variety of toxic substancesand pathologic conditions. Treatment-related alterations of sexual behavior in either sex oftest species may represent potentially significant reproductive concern for humans.B. Treatment Duration for Males in Fertility and Reproduction StudiesThe fertility and reproduction study designs assume that data from previously conducted toxicitystudies (e.g., histopathology and weights of reproductive organs from the acute and subchronicstudies) will be used to select the appropriate treatment levels. The duration of treatment selectedshould be adequate for determining effects on spermatogenesis. Because of the duration of spermatogenesis,most fertility and reproductive studies require treatment of the males for approximately10 weeks prior to mating.Provided that no effects were found in a repeated dose toxicity study (of at least 4 weeksduration), a premating treatment interval of 2 weeks for females and 2 weeks for males should beused in ICH fertility studies. 6,7 The reduction of the premating treatment period by the ICH wasbased upon a reevaluation 48,49 of the basic research on the effects of male reproductive toxicantson the process of spermatogenesis. The ICH concluded: (1) that compounds that induce selectiveeffects on male reproduction are rare; (2) that mating is an insensitive means of detecting effectson spermatogenesis in rodents; (3) that good pathological and histopathological examination of themale reproductive organs provides a more sensitive means of detecting effects on spermatogenesis;(4) that compounds that affect spermatogenesis almost invariably affect postmeiotic stages; and (5)that there are no conclusive examples of a male reproductive toxicant that could only be detectedafter treating the males for 9 to 10 weeks and then mating them. Studies conducted in Japan 50indicate that a 2-week premating treatment period is adequate to detect toxicity in male reproductiveorgans. Because the 2-week repeated dose study was validated as being as effective as a 4-week© 2006 by Taylor & Francis Group, LLC

438 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrepeated-dose study for determining male reproductive organ toxicity, treatment of the males for2 weeks before mating is considered acceptable in ICH fertility studies. Justification for the lengthof the premating treatment period should be stated in the protocol. In fertility studies, treatmentshould continue throughout the cohabitation period until termination for males and at least throughimplantation for females. Multigenerational studies require a premating treatment period of at least10 weeks for both P 1 and F 1 males and females.C. Cohabitation and Evidence of MatingFemales are introduced into the male’s cage in the afternoon, and they are checked for evidenceof mating the following morning. When a receptive sexually competent female rat is placed witha sexually competent male rat, the male will attempt to mount within a relatively short time. Thetime from female placement into the cage until the first mount is defined as the latency to firstmount. Prior to ejaculation, the male will attempt several (six to nine) mounts with intromission.In standard reproductive toxicology studies, the evidence of sexual receptivity and successful matingis based on the presence of copulatory plugs or sperm-positive vaginal lavage. However, these posthoc findings do not demonstrate that male sexual activity resulted in adequate sexual stimulationof the female rat. 51 Failure to adequately stimulate the female has been shown to impair spermtransport in the genital tract of female rats, thereby reducing the probability of successful fertilization.52 The reduced fertility (as calculated in the fertility index) might be erroneously attributedto an effect on the spermatogenic process in the male or on fertility of the female, rather than onaltered mating behavior. Effects on sexual behavior should be considered as adverse reproductiveeffects. However, impairment to sexual behavior that is secondary to more generalized systemiceffects (e.g., severe acute intoxication, impaired motor activity, or general lethargy) should not beconsidered an adverse reproductive effect.1. Vaginal PlugsWhen the male rat intromits and ejaculates, the semen quickly coagulates, forming a copulatoryplug. Secretions from the male accessory sex glands are responsible for copulatory plug formation,without which rodent fertility can be impaired. Copulatory plugs can remain in situ (in the vagina),or they can be dislodged from the vagina and fall into the cage pan. In male-treated fertility studies,fewer plugs found in the treated groups when compared with the control group indicates that malemating behavior may have been altered.Mating (copulation) is confirmed by the presence of a copulatory plug or by the observationof spermatozoa in the vaginal smear (“sperm positive”). While in cohabitation, the females andtheir cages should be checked daily for the presence of copulatory plugs. (Note: The recordingmethodology presented here is an example only.) If a copulatory plug is observed in situ (in thevagina), a V is recorded on the cohabitation sheet. Those animals observed to have a copulatoryplug in situ do not require a vaginal lavage. A V may also be recorded on the sample slide next tothe animal number to show that no vaginal lavage sample was required for that animal. The presenceof copulatory plug in the cage pan is recorded by writing a C on the cohabitation sheet. (Note:Copulatory plugs in the cage pan are not sufficient evidence that mating has occurred.) If only acage pan plug is observed or when no evidence for mating is observed, then a vaginal lavage shouldbe performed to detect the presence of sperm.2. Vaginal SmearsVaginal lavage for the presence of sperm can be performed at any time of day; however, morningis recommended. The samples should be collected at approximately the same time each day overthe course of the study. Before any vaginal lavage samples are taken, the aliquot of fresh saline© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 439should be verified as contaminant-free (e.g., free of dirt, dust, spermatozoa, epithelial cells). Thefemale rat is removed from the cage and its identification is verified. One or two drops (approximately0.25 ml) of physiological (0.9%) saline are withdrawn from the aliquot into a new, cleandropping pipette. The tip of the dropping pipette is gently inserted into the vaginal canal. Thepipette bulb is firmly but gently depressed to expel the saline from the pipette into the vagina. Thesaline is gently drawn back into the dropping pipette, and the pipette is removed from the vaginalcanal. If the flush appears very clear, it is recommended that the lavage be repeated. Care must betaken not to insert the tip too far into the vaginal canal, because excessive stimulation of the cervixcan result in pseudopregnancy (prolonged diestrus). The female is returned to her cage. The contentsof the pipette are placed onto a clean glass slide. Each slide is numbered, a clear top and bottomof each slide is identified, and a data sheet is marked to indicate the location of each vaginal smear.All slides should be handled carefully to avoid mixing vaginal lavage samples from differentanimals. If samples become mixed, they should be discarded and the sampling procedure repeatedfor those animals. After vaginal lavage sampling has been completed for the day, the saline shouldbe reevaluated for contamination that may have occurred during sample collection.The vaginal lavage slides are allowed to air dry and are examined microscopically. When spermare present in the vaginal smear, a positive (+) is recorded. When no sperm are present in thevaginal smear, a negative (–) is recorded. A second vaginal lavage sample may be required to moreconclusively determine mating status (e.g., when the cage pan plugs are present but the vaginallavage sample is not sperm positive or only a few sperm were present). When an additional vaginallavage sample is collected and examined, an “a” for additional sample is noted, and the results ofthe second lavage are recorded on the cohabitation sheet.D. Reproductive Indices1. Male Mating and Fertility Indexa. Male Mating IndexThe male mating index is reported in studies where the male is the test system. It is calculatedusing the following formula:Male mating index =number of males with confirmed matingnumber of males cohabitated× 100Historical control values for the male mating index are approximately 94% with a range of 80%to 100%. In fertility or reproduction studies where both sexes are treated, this index can be usedas an indicator of reproductive toxicity, but not of specific male or female reproductive toxicity.Confirmation of male mating is usually based on the presence of a copulatory plug or sperm in thevaginal smear. Mating and pregnancy can also be confirmed by the unexpected delivery of a litteror by the discovery of resorbed uterine implantation sites in the postmortem uterine evaluation ofpresumed nonpregnant females (no confirmed mating and considered not pregnant). The malemating index can be influenced by many factors, including physical impairment, acute intoxication,alterations to the neuroendocrine-gonadal axis affecting libido, hormonal imbalance, etc. Otherfactors that might affect the male mating index include estrous cycle disruption (prolonged cyclesor persistent estrus or diestrus) and lack of female receptivity.b. Male Fertility IndexThe male fertility index is calculated using the following formula:© 2006 by Taylor & Francis Group, LLC

440 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONMale fertility index =number of males impregnating a femalenumber of males cohabitated× 100This index measures the male’s ability to produce sperm that are capable of impregnating a female.Because mating does not imply impregnation, this index supplies additional information. The malefertility index can be influenced by many of the same factors as those described above for the malemating index, as well as many other factors, such as sperm capacitation and transport, fertilization,oviductal milieu, development of the conceptus, and nidation. Historical control values for the malefertility index are approximately 85%, with a range between 70% and 100%. Again, in fertility orreproduction studies where both sexes are treated, the male mating index can be used as an indicatorof reproductive toxicity but not of specific male or female reproductive toxicity.2. Female Fertility Indexa. Fertilization RateFertilization rate (the proportion of ova that were fertilized divided by the total number of ovaavailable) is seldom reported in reproduction studies because the data collection must be done veryearly in gestation. Multiple factors are involved in the timing and integrity of gamete and zygotetransport. These factors are quite susceptible to chemical perturbation and may affect the survivalof the gametes, fertilization, and development of the conceptus.b. Female Mating IndexThe female mating index is calculated using the following formula:This index measures the female’s ability to mate. Mating implies that the female was in a periodof behavioral heat (estrus), was sexually attractive to the male (emitting chemicals that attract themale), and performed mating behaviors (solicitational behaviors [such as ear wiggling, hopping,and darting] and receptive behaviors [lordosis response]). 51 In routine reproductive toxicity testing,the mating is not actually observed, so unusual mating behaviors are not identified. Historicalcontrol values for the female mating index are approximately 94%, with a range between 80% and100%. The female mating index can be influenced by many factors, such as physical impairment,acute intoxication, alteration of the neuroendocrine-gonadal axis, diminished libido, hormonalimbalance, and estrous cycle disruptions. When both sexes are treated, the female mating indexcan be used as an indicator of reproductive toxicity but not of specific male or female reproductivetoxicity.c. Female Fertility IndexFemale mating index =number of females mating× 100number of cohabitated femalesThe Female Fertility Index is calculated using the following formula:Female fertility index =number of pregnant femalesnumber of females cohabitated × 100© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 441This index measures the female’s ability to achieve pregnancy. The female fertility index canbe influenced by many of the same factors described above for the male fertility and female matingindices. Historical control values for the female fertility index are approximately 85%, with aminimum of 72% and a maximum of 100%. Again, when both sexes are treated, the female fertilityindex can be used as a general indicator of reproductive toxicity.d. Fecundity or Pregnancy IndexThe fecundity or pregnancy index is calculated using the following formula:Fecundity index =number of pregnant femalesnumber of females with confirmed mating × 100This index measures the female’s ability to achieve pregnancy. Historically, the fecundity orpregnancy index for Sprague-Dawley rats averages 93.0% (SD ±9.88%). 53 The fecundity or pregnancyindex can be influenced by many of the same factors described above in the male fertilityand female mating indices. Again, the fecundity or pregnancy index can only be used as a generalindicator of reproductive toxicity when both sexes are treated.e. Implantation Index and Pre- and Postimplantation LossesThe implantation index is calculated by dividing the number of implantation sites by the numberof corpora lutea. Historical control values for the implantation index are approximately 90%, witha range between 85% and 100%.Implantation index =number of implantsnumber of corpora lutea × 100Disruption of early developmental processes may contribute to a reduction in fertilization rate andincreased early embryonic death prior to implantation. This can be calculated as the preimplantationloss, which is the number of corpora lutea minus the number of implantation sites, divided by thenumber of corpora lutea.Preimplantation loss =number of corpora lutea – number of implantsnumber of corpora lutea× 100Embryonic deaths after implantation plus fetal deaths are calculated as postimplantation loss, which,in fertility studies, is the total number of implantation sites minus the number of live conceptuses,divided by the number of implantation sites.Postimplantation loss =number of implants – number of viable fetusesnumber of implants× 100Postimplantation loss can be determined in reproduction studies following delivery of a litter. Theuterus is examined (usually at necropsy after weaning of the litter), and the number of implantationsites is determined. Postimplantation loss in reproduction studies is the total number of implantationsites minus the number of full-term pups, divided by the number of implantation sites. The historical© 2006 by Taylor & Francis Group, LLC

442 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONaverage preimplantation loss (in developmental toxicity studies) is 7.6%, while the average postimplantationloss in developmental toxicity and natural delivery studies is 6.0 and 8.8%, respectively. 53To identify the cause of increased preimplantation loss, additional studies using direct assessmentsof fertilized ova and early embryos would be necessary. Pre- and postimplantation loss occursin untreated as well as treated rodents and contributes to the normal variation in litter size.f. Gestation IndexThe gestation index is calculated using the following formula:Gestationindex=number of females with live bornnumber of females with evidence of pregnancy × 100This index measures the female’s ability to maintain pregnancy, based on having delivered at leastone live pup. Historically, the gestation index for Sprague-Dawley rats averages 97.3% (SD ±10.26%). 53 In RACB studies, this index is also known as the pregnancy rate (the proportion ofpairs with confirmed mating that have produced at least one pregnancy within a fixed period). Thegestation index is not a sensitive indicator of reproductive performance because all litters are treatedas having equal biological significance regardless of their size.g. Precoital Interval (Mean Time to Mating)The precoital interval or mean time to mating (the average number of days after the initiation ofcohabitation required for each pair to mate) is a useful indicator of altered reproductive function.Precohabitational estrous cycle data and cohabitational vaginal smears usually provide enoughevidence to determine when mating has occurred. The normal precoital interval for rodents is 1 to4 d, depending on the estrous cycle stage at the time of initiation of cohabitation. Historical controlvalues for the precoital interval are approximately 3.3 ± 0.4 d, with a range between 2.8 and 3.8d. If the stage of the estrous cycle at the time of cohabitation is known, this initial variance can becorrected during the data analysis. An increased precoital interval suggests impaired sexual behaviorin one or both partners or abnormal estrous cyclicity.3. Mean Gestational LengthMean gestational length (duration of pregnancy) is calculated as the duration from GD 0 (day ofpositive evidence of mating) to the day of parturition. This information is easily derived from thebreeding records and maternal records. Historically, the mean gestational length for Sprague-Dawley rats averages 22.3 d (SD = 0.44). 53 Decreased gestational length is associated with decreasedpup weight and viability, while increased gestational length is associated with higher birth weights(due to the longer growth period in the uterus). However, increased gestational length may also beassociated with dystocia (difficult labor and/or delivery), which may result in death or the physicalimpairment of the dam and/or the pups. Changes in gestational length can be due to several factors,including alteration in hormone levels and fetal growth retardation.4. Litter SizeLitter size is the number of offspring delivered and should be determined at or as soon after birthas possible. Litter size is an important endpoint in the overall evaluation of reproductive performancein rodents and is frequently more sensitive than the other reproductive indices. For most reproductionstudies, the mean litter size is calculated at standardized time points (PND 0, 4, 7, 14, and 21).© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 443The day of birth is defined as PND 0. The litter size on PND 0 is calculated using the followingformula:Litter size PND0 =total number of pups delivered (live and stillborn)number of dams that deliveredLitter size should include dead as well as live offspring; therefore, numbers of live and deadoffspring should be determined. Historically, the litter size PND 0 for Sprague-Dawley rats averages14.75 pups (14.46 liveborn and 0.29 stillborn). 53 Dams that die or that resorb their entire litter areexcluded from the litter size calculation. This value can be affected by maternal cannibalism, acommon occurrence with stillborn or malformed pups. While cannibalism of healthy pups byuntreated dams does occur, excessive cannibalism by treated dams should be considered an adversetreatment-related maternal behavior. The ovulation rate (number of ova available for fertilization),fertilization rate, implantation rate, and postimplantation survival rate affect litter size. A decreasedlitter size may indicate an adverse reproductive effect. Again, in fertility or reproduction studieswhere both sexes are treated, the litter size can be used as a nonspecific indicator of reproductivetoxicity.5. Live-Birth IndexThe live birth index is calculated using the following formula:Live birth index =number of pups born alivetotal number of pups born × 100Normal delivery usually requires about 1 hour, depending on the number of pups in the litter. Litterviability checks are best done immediately after delivery; however, this usually is not possiblebecause most litters are delivered late at night, when the technical staff is not present. All newbornpups should be examined for viability and for external abnormalities. It is difficult to determinevisually whether the pup was stillborn (died in utero) or died shortly after birth. One way todistinguish the stillborn pups from the pups that died shortly after birth is by floating the lungs insaline (see below). Historically, the live birth index for Sprague-Dawley rats averages 98.0%. 536. Survival IndexSurvival indices are primary endpoints in a reproduction study, reflecting the offspring’s ability tosurvive postnatally to weaning. Survival and viability indices are usually measured on PND 4, 7,14, and 21. Survival indices are calculated using the following formula:Survival index =total number of live pups (at designated timepoint)number of pups born× 100If litter standardization or culling of the litter is performed on PND 4, then the PND 4 survivalindex should be based on the preculling number of offspring. Historically, the survival index forSprague-Dawley rats at PND 4 averages 97.24% (derived from the percentage of dead pups betweenPND 0 and 4). 53 All subsequent survivability index calculations will use the number of offspringremaining after standardization as the denominator in the equation. If litter standardization wasperformed, the PND 21 calculation is known as the lactation index. If litter standardization wasnot performed, the PND 21 calculation is known as the weaning index.© 2006 by Taylor & Francis Group, LLC

444 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPup survival endpoints can be affected by the toxicity of the test substance, either by directeffects on the offspring (i.e., structural or functional developmental defects, low birth weight,impaired suckling ability), indirectly through effects on the dam (i.e., lactational ability of the dam,maternal neglect, acute intoxication during the treatment period), or by a combination of maternaland offspring effects. Pup survival may also be impaired by nontreatment-related factors, such aslitter size.7. Sex RatioOffspring gender in mammals is determined at fertilization of an ovum by sperm with an X or aY chromosome. Altered sex ratios may be related to several factors, including selective loss ofmale or female offspring, sex-linked lethality (genetic germ cell abnormalities), abnormal productionof X or Y chromosome–bearing sperm, or hormonal alterations that result in intersex conditions(masculinized females or feminized males). Determination of gender is generally done shortly afterparturition, and the data are usually presented as a ratio, using the following format:Sex ratio =number of male offspringnumber of female offspringThese data are often presented as a percentage of male offspring, using the following formula:Percentage of male offspring =number of male offspringtotal number of offspring × 100Historically, the percentage of male offspring for Sprague-Dawley rats averages 50.0% (SD±3.39%). 53 The calculations should include both live and dead pups because sex can easily bedetermined (except in the case of cannibalized pups or in some sexually ambiguous offspring).Treatment effects on sex ratio can indicate an adverse reproductive effect, especially if embryonicor fetal loss is observed. Because of the altered anogenital distance, visual determination of sex isoften incorrect when the pups have an intersex condition. The correct diagnosis is usually determinedlater during lactation, when the condition becomes obvious, or at the postmortem evaluation,when the examination reveals male or female reproductive organs.A. BiologyIII. MALE REPRODUCTIVE TOXICOLOGYEvaluation of the rodent male reproductive system usually consists of a gross evaluation (includingorgan weights) and histological evaluation of the testes, epididymides, vasa deferentia, seminalvesicles, coagulating gland, and prostate, and evaluation of the sperm. There are primarily foursperm measures required in the reproductive testing guidelines: epididymal sperm count, motility,and morphology, and testicular spermatid head count.The male rat reproductive system consists of the testis (with scrotum), epididymis, vas deferens,ampullary gland, seminal vesicle, coagulating gland, bulbourethral gland, prostate, preputial gland,and penis. While the male genital protuberance is generally larger than the females’, it is not areliable method for determining gender of a pup. Determination of the sex of a newborn rat pupis based on the visual examination of the relative distance between the genital tubercle and the© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 445anus. This distance is larger in the male than female (approximately 3.5 mm versus 1.4 mm). 54 Thetestes are situated in the abdomen of the newborn male and do not descend into the scrotum untilthe rat is 4 to 6 weeks old. When viewed externally, the large testicles make the scrotal sacs protrude.The inguinal canal, unlike that of the human, remains open in the adult rat, so the testes may bewithdrawn into the abdomen. Also visible externally is the aperture of the prepuce. The prepuceis a loose sheath that protects the penis. The penis and the external orifice of the urethra can beexposed for examination by gentle backward pressure at the sides of the prepuce.B. ObservationsThere are three basic types of observations performed by trained technicians: a viability check, aclinical observation, and an extended or detailed clinical observation. A viability check is a dailycage-side observation performed once (at the start of the workday) or twice (at the start and endof the workday) to determine mortality, morbidity, early delivery, or abortion. During the viabilitycheck, food and water availability should also be monitored. The clinical observation is a moredetailed evaluation of the physical or behavioral characteristics of the rat. The rat is removed fromthe cage and observed for any abnormalities, physical or behavioral. This observation is usuallyperformed prior to treatment and often at a selected interval after treatment. The extended or detailedclinical observation is performed in an open field arena by a technician trained to administer theFOB. The extended or detailed clinical observation is essentially a FOB without including forelimband hind limb grip strength, hind limb foot splay, and body temperature measurements (see below).Extended or detailed clinical observations are usually performed weekly or monthly dependingupon the protocol.C. Male Body Weight and Food ConsumptionBody weights are recorded daily, twice weekly, or weekly, depending upon the guideline requirements.Food consumption is usually recorded at the same interval as body weight. Body weightand food consumption should be recorded at approximately the same time of day during the courseof the study. Recording male body weight during treatment provides an indicator of the generalhealth status that may be relevant in the interpretation of male reproductive effects. Body weightloss or reduction in body weight gain can be caused by several factors, including systemic toxicity,stress (such as restraint), treatment-induced anorexia, or decreased feed or water consumption dueto altered palatability. Caloric restriction studies that caused small reductions in body weight haveshown little effect on the male reproductive organs or function. 55 If a direct link between bodyweight change and the male reproductive effect cannot be established, then male reproductivesystem alterations should be considered adverse reproductive effects and not secondary to theoccurrence of systemic toxicity. However, in the presence of severe body weight change or othersignificant systemic toxic effects, it should be stated that an adverse effect on a male reproductiveendpoint occurred, but the effect may have resulted from a systemic toxic effect.D. Male NecropsyThe male is usually terminated after the cohabitation period; however, it may be advisable tocontinue treatment and delay termination until after pregnancy outcome is known. In the event ofan equivocal effect on fertility, the treated males can be mated with untreated females to determineif the effect was male mediated. The male is euthanized and subjected to a complete necropsy. Thereproductive organs, and other tissues specified in the protocol, are weighed and saved for possiblehistopathological evaluation. At necropsy, sperm quality may also be assessed (see below).© 2006 by Taylor & Francis Group, LLC

446 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONE. Male Reproductive Organ WeightMale reproductive organ weights are collected for the testes, epididymides, seminal vesicles (withcoagulating glands), and prostate. The pituitary gland is also routinely harvested and weighed. Theorgan weight values should be reported as both absolute weights and relative weights (organ weightto body weight ratio). Organ weight to brain weight data is often reported because the weight ofthe brain in the adult rat usually remains quite stable. 56 Mean normal testis weight varies onlyslightly within a given experimental species, 57,58 suggesting that absolute testis weight should be agood indicator of testicular injury. However, reduction in testis weight is often detected only attreatment levels greater than those required to produce significant effects on other gonadal parameters.59–61There can be a temporal delay between testicular cell death and testicular weight change. Whilecell death would be expected to reduce testicular weight, testicular edema and inflammation, cellularinfiltration, and/or Leydig cell hyperplasia may actually increase testicular weight. Sloughed deadgerminal epithelium cells can cause a blockage of the efferent ducts, causing increased testis weightas a result of fluid accumulation. 62,63 While testis weight may not be a good indicator for certainadverse testicular effects, a significant alteration in testicular weight should be considered an adversemale reproductive effect.Epididymal and cauda epididymal weights are often required in reproductive studies, especiallywhen sperm parameters are evaluated. Because sperm are within the epididymis and thereforecontributes to its weight, reductions in sperm production can alter the epididymal weight. Thepresence of lesions such as sperm granulomas or leukocytic infiltration (inflammation) in theepididymal epithelium can also affect the epididymal weight.Because the seminal vesicle and prostate are androgen-dependent organs, weight changes inthese organs may reflect alterations in endocrine status or testicular function. The seminal vesiclesand prostate can respond differently to a test substance; therefore, information will be improved ifthe weights were collected separately (with and/or without their secretory fluids). Differential lossof secretory fluids prior to weighing can cause artifactual organ weights or increased variation inorgan weight values.Pituitary weight can provide information on the reproductive status of an animal. However, thepituitary contains many other cell types that are responsible for the regulation of numerous physiologicfunctions unrelated to reproduction. Therefore, alterations in pituitary weight do not necessarilyreflect altered reproductive function. Selective histological stains may be used to confirmalterations in regions of the pituitary associated with reproductive functions. If the pituitary weightchange can be associated with changes in reproductive function of the pituitary, then the pituitaryweight change should be considered an adverse reproductive effect.Significant changes in male reproductive organ weights (absolute and/or relative) should beconsidered an adverse male reproductive effect. However, lack of reproductive organ weight effectsshould not be used to negate significant changes in other reproductive endpoints.F. Histopathological Evaluation of the Male Reproductive OrgansHistopathological evaluations of the male reproductive organs have a prominent role in toxicityassessment. Detailed descriptions of the histology and pathology of the male reproductive systemare beyond the scope of this chapter; however, brief discussions of the organs comprising the malereproductive system are presented. Routinely evaluated male reproductive organs include the testes,epididymides, prostate, seminal vesicles, coagulating glands, and pituitary. Histopathological evaluationscan be especially useful by (1) providing location (including target cells) and severity ofthe lesion, (2) providing comparative data from a variety of reproductive study designs, (3) providinga relatively sensitive indicator of damage in short-term dosing studies, and (4) describing recoveryafter the cessation of treatment. When similar targets or mechanisms exist in humans, the basis for© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 447interspecies extrapolation is strengthened. Significant, biologically meaningful histopathologicalalterations of male reproductive organs should be considered adverse reproductive effects.1. TestisNumerous chapters 64–66 and books 67–69 have been written concerning the testis. While detaileddescriptions of testicular histology and spermatogenesis are beyond the scope of this chapter, areview is presented.The testis is covered by the tunica albuginea, a tough fibrous capsule. The rat testis containsvery little intertubular connective tissue, in contrast to the human testis, which is partitioned byconnective tissue septa. However, the rat testis is similar to the human testis in having two majorcompartments: the interstitial compartment and the seminiferous tubule compartment. The interstitialcompartment contains the connective tissue, interstitial (Leydig) cells, nerves, blood vessels,and lymphatic vessels. Unlike human testicular capillaries, rat testicular capillaries are not fenestrated.The appearance of Leydig cells in the developing testes coincides with the production oftestosterone. Macrophages are commonly observed in the interstitium and may account for up to25% of the interstitial cells.A limiting membrane or boundary tissue formed by lymphatic endothelial cells, myoid cells,and acellular components (basal lamina) bounds the seminiferous tubule compartment. The myoidcells are contractile and provide the major motive force for the movement of sperm and fluidthrough the seminiferous tubule. The basal lamina and the myoid cells provide the structural baseon which the Sertoli cells and the basal compartment cells of the seminiferous epithelium rest.Sertoli cells extend around the periphery of the tubule through to the lumen, with invaginationssurrounding germ cells at various stages of differentiation. Specialized junctions between the Sertolicells form the blood-tubule barrier in the rat at approximately 18 days of age. After that, the Sertolicells do not divide again. Sertoli cells maintain metabolic support for the seminiferous epithelium,maintain hormonal control of spermatogenesis and testicular function, and have a phagocyticfunction.The seminiferous tubules are highly convoluted loops that have their two ends connected toshort segments of straight tubules (tubuli recti). Although highly convoluted, the tubules straightenalong the long axis of the testis. A transverse histological section through the long axis of the testisallows examination of cross-sectioned seminiferous tubules. The tubuli recti connect to the retetestis, which is connected to the cephalic portion of the epididymis by 10 to 20 long and tortuousefferent ducts (ductuli efferentes).The most sensitive endpoint for the determination of testicular toxicity is histopathology.Testicular histology and sperm parameters are linked, 70 and alterations in testicular structure areusually accompanied by alterations in testicular function. However, some functional changes mayoccur in the absence of detectable structural changes in the testis. Alterations in the testicularstructure and function are also linked to alterations in epididymal structure and function, but changesin epididymal function may occur in the absence of testicular or epididymal structural changes.Proper fixation (using Bouin’s, Davidson’s, or a comparable fixative for at least 48 h) andcareful trimming, embedding (paraffin), sectioning (5 to 6 µm), and staining (hematoxylin andeosin or periodic acid-Schiff’s-hematoxylin stain) of testicular tissue 71,72 improves the quality ofthe histological evaluation of spermatogenesis. Care must be taken to minimize the compressionof the testes during tissue dissection, as it can result in artifactual damage to the testes (such assloughing of germ cells from the seminiferous epithelium).Either a longitudinal or transverse section of the testis can be taken. The advantages of alongitudinal section are that there is more tissue to examine and that longitudinal sections of theseminiferous tubules allow determination of consecutive stages of the spermatogenic cycle. Theadvantage of a transverse section is a consistent cross-section of seminiferous tubules, allowingspermatogenic staging, quantification of cell numbers, and tubular measurements (such as seminiferous© 2006 by Taylor & Francis Group, LLC

448 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONtubule diameters). Most technicians take a midline transverse section. However, the testis hasasymmetrical structures (e.g., rete testis) that may be missed by a transverse section.Histopathological evaluations of the testis can determine whether sloughed cells are present inthe tubule lumen, whether the germinal epithelium is degenerating or severely depleted, or whethermultinucleated giant cells are present. Evaluation of the seminiferous tubules should include thepachytene spermatocyte. This cell type has the longest life of the primary spermatocytes with themost RNA and protein synthesis. The pachytene spermatocyte appears to be uniquely susceptibleto toxicants. 73 The rete testis should be examined for dilation due to obstruction or disturbances influid dynamics and for the presence of proliferative lesions and rete testis tumors. Both spontaneousand chemically induced lesions in the rete testis often appear as germ cell degeneration anddepletion. More subtle lesions, such as missing germ cell types or retained spermatids, can reducethe number of sperm being released into the tubule lumen. Such effects may not be detected whenless than adequate methods of tissue preparation have been used. Also, knowledge of the detailedtesticular morphology and spermatogenesis assists in the identification of lesions that may accompanylower treatment levels or result from short-term exposure. 68Several methodologies for qualitative or quantitative assessment of testicular tissue are availablethat can assist in the identification of testicular lesions, including the use of “spermatogenic cyclestaging.” 68,74 To provide high resolution cellular morphologic detail required for spermatogeniccycle staging, semithin (2 µm) sections will have to be produced from testes embedded in glycolmethacrylate resin. A detailed description of spermatogenesis and spermiogenesis required forstaging of spermatogenesis is beyond the scope of this chapter; however, detailed morphologicaldescriptions with light and electron microscopic photographs of these processes are available in anexcellent book by Russell et al. 68Rodent males produce sperm in numbers that greatly exceed the minimum requirements forfertility, particularly as evaluated in reproductive studies that allow multiple matings. 75,76 In somestrains of rats and mice, sperm production can be drastically reduced (by up to 90% or more)without affecting fertility. 77–79 Human male sperm production appears to be much closer to theinfertility threshold; therefore, less severe sperm count reductions may cause human infertility.Negative results in rodent studies that are limited to only fertility and pregnancy outcomes provideinsufficient information to conclude that the test substance poses no reproductive hazard in humans.2. Epididymides, Accessory Sex Glands and PituitaryThe basic morphology of other male reproductive organs (epididymides, accessory sex glands, andpituitary) has been described 80,81 as well as the histopathological alterations that may accompanycertain disease states. 82,83 While detailed descriptions of histology and pathology of these organsare beyond the scope of this chapter, a discussion is presented.The epididymis, a single, long, highly tortuous tube surrounded by connective tissue, has threeportions: the caput (head) located on the anterior pole of the testis, a narrow corpus (body) lyingalong the dorsomedial aspect of the testis, and the cauda (tail) located on the posterior pole of thetestis. The cephalic portion of the epididymis is almost entirely encapsulated in fat. The ductus(vas) deferens is the continuation of the epididymis. The lumen of the vas deferens becomes lesstortuous leading to the prostatic urethra. The secretions of the seminal vesicles, coagulating glands,ampullary glands, dorsal and ventral prostate glands, and bulbourethral glands are added to thesperm in the prostatic and penile urethra to produce semen. Preputial glands do not contribute tothe semen, but rather secrete musklike compounds.A longitudinal section of the epididymis that includes all epididymal segments should beincluded for histopathological examination. During dissection of the epididymis from the testis,care must be taken not to cut the testicular capsule and cause extrusion of the seminiferous tubules,as such damage may disrupt the testicular architecture. Histopathological evaluation of the epididymisshould include information about the caput, corpus, and cauda segments. Presence of debris© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 449and/or sloughed cells in the epididymal lumen is an indicator of damage to the germinal epitheliumor to the excurrent ducts of the epididymis. The presence of lesions such as leukocytic infiltration,sperm granulomas, or absence or paucity of clear cells in the cauda epididymal epithelium shouldbe recorded.The density of sperm and the presence of germ cells in the lumena of the three epididymalsegments reflect time-dependent events in the testis. Sperm in the caput were released from thetestis only a few days earlier, while the sperm in the cauda were released approximately 14 daysearlier. A careful examination of the accessory sex glands is indicated when decreased copulatoryplugs are observed during cohabitation.Histopathologic evaluation of the pituitary should include cellular morphology of gonadotropinproducingcells.G. Sperm EvaluationsThe sperm quality parameters are sperm number, sperm motility, and sperm morphology. Althougheffects on sperm production can be reflected in other measures, such as testicular spermatid countor cauda epididymis weight, no other measures are adequate to reflect effects on sperm morphologyor motility. The application of digital video technology to sperm motility analysis allows a detailedevaluation of sperm motion, including precise information on individual sperm tracks. Computerassistedsperm analysis (CASA) also provides permanent storage of the actual sperm track images,which can be reanalyzed as necessary, either manually or computer assisted. With CASA technology,sperm velocity (rectilinear and curvilinear) and track amplitude and frequency can be obtainedon large numbers of sperm. Chemically induced alterations in sperm motion have been detectedby use of CASA technology, 84–86 and often these changes are correlated with the fertility of theexposed animals. 87–89 Studies indicate that significant reductions in sperm velocity are associatedwith infertility, even when the percentage of motile sperm is not affected. It is also important tobe able to distinguish between the sperm with progressive motility and those showing other patternsof movement. 90While changes in spermatogenesis and sperm maturation endpoints have been related to fertilityin several test species, the ability to predict infertility from these data (in the absence of actualfertility data) may not be reliable. Fertility is dependent not only on having adequate sperm count,but also on having normal sperm. When sperm quality is good, then a significant reduction in spermcount is required to affect fertility. Rat sperm counts have to be reduced substantially before fertilityis affected. 91 In fertile and infertile men, the distribution of sperm count overlaps, with the meansperm count for fertile men being greater than for infertile men. 92 However human fertility is usuallyreduced when sperm counts decrease below 20 million per milliliter. 93 If the sperm counts arenormal in rodents, sperm motility or velocity must be greatly reduced before fertility is affected. 87Reduction in sperm count or quality may not cause infertility in rats, but the reductions may bepredictive of infertility.1. Sperm Number or Sperm ConcentrationSperm number or sperm concentration for test species, as well as sperm motility, morphology, andviability, may be determined from ejaculated, epididymal, or testicular samples. 90 Rat spermevaluations routinely use sperm from the cauda epididymis. Sperm count is reported as an absolutesperm count and relative to the cauda epididymal weight. However, because the number of spermcontributes to the weight of the epididymis, expression of the data as a ratio may actually mask areduction in sperm count. As one would expect, epididymal sperm counts can be directly influencedby level of sexual activity. 75,94 Sperm production data may be derived by counting the number ofhomogenization-resistant testicular spermatids. 75,78,95,96 These counts are a measure of sperm productionfrom the stem cells and their development through spermatocytogenesis and spermiogenesis. 78,97 The© 2006 by Taylor & Francis Group, LLC

450 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONspermatid count may serve as a substitute for quantitative histological analysis of sperm production 68if the duration of treatment is sufficient for a lesion to be fully expressed. However, spermatidcounts may be misleading when the duration of treatment is too short.2. Computer-Assisted Sperm AnalysisTechnological advances in computer-assisted sperm analysis have made objective evaluation of themotile characteristics of rat sperm possible. CASA systems utilize a collection of components(VCR, microscope, multiple microscope objectives, optical disk drive, computer hardware, andCASA software) to determine sperm concentration, sperm motility and other motion parameters,and sperm morphology. CASA systems usually have a microprocessor-controlled stepping motorthat can move a temperature-controlled microscope stage to exact positions. Light (of a selectedwavelength) is provided by a strobe source that illuminates the image field with a series of shortflashes (7 to 60 flashes/sec, each of 1- to 3-msec duration). The CASA system then collects a seriesof phase contrast or fluorescent images, with up to 20 image fields analyzed per specimen. Motileand stationary objects from each field of view are located and identified. Moving objects (spermheads) are identified by use of background subtraction techniques, and their mean size and brightnessare calculated. In the case of rat sperm, only the head is detected as motile.After the motile cells are identified, the static sperm are then discriminated from detritus(typically leukocytes, erythrocytes, epithelial cells, spermatogonia, and cellular debris). The staticsperm must be properly counted to allow percent motility to be calculated. The static rat spermhead and its connected midpiece can be identified and combined by the computer to form a singleelongated object. The extreme elongation allows static sperm to be distinguished from other staticcells and debris. Three modes of discrimination are employed to identify the static sperm: (1) sizeand intensity, (2) elongation (ratio of head width to head length), and (3) use of a DNA-specificfluorescent stain that stains the highly condensed DNA in the sperm head. Most of the other staticcell types are basically round; therefore, elongation of the head and midpiece is a useful separator.Detritus does not routinely have significant DNA, and thus it does not stain or fluoresce.Motile sperm are tracked by comparing successive frames by means of a nearest neighboralgorithm that selects the closest object from the next frame. The individual sperm track is calculatedby joining the successive sperm positions with straight lines. Typically rat sperm move at a rate of200 to 400 µm/sec or less, so a spermatozoon will usually move less than 3 to 6 µm betweenframes collected at 60 frames/sec.After the motile and nonmotile sperm have been identified, counted, and tracked, the cellconcentrations are calculated. Concentration calculations require the number of motile and nonmotilesperm, the area of the field of view, the dilution ratio, and the depth of the specimen chamber.The dilution factor is a variable and needs to be precisely controlled. The CASA system calculatessperm count and concentration, motility, progressive velocity, track speed, path velocity, straightness,linearity, cross beat frequency, and amplitude of lateral head displacement (see Figure 10.9).The following list describes the various sperm motility endpoints:ParameterTotal sperm countedTotal concentration (millions/ml)ConcentrationNumber motileNumber progressivePath velocity (VAP)DefinitionNumber of static and motile sperm counted in all elds analyz edCalculated using concentration (M/ml) multiplied by the ejaculate volumeThe number of sperm per ml of sample determined by dividing the totalsperm count by the total volume of the elds analyz edPopulation of sperm that are moving at or above a minimum de ned speedThe number of cells moving with both the VAP greater than the mediumaverage path velocity and STR greater than the threshold straightnessTotal distance along the smoothed average position of the sperm dividedby the time elapsed© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 451ParameterProgressive velocity (VSL orrectilinear velocity)Track speed (VCL or curvilinearvelocity)Amplitude of lateral headdisplacement (ALH)Straightness (STR)Linearity (LIN)MotilitySperm per gramDefinitionStraight-line distance from the beginning to the end of the sperm track,divided by the time elapsedTotal distance between the actual point-to-points for a given sperm duringacquisition, divided by the time elapsedMean width of the head oscillation as the sperm swimsAverage value of the ratio VSL/VAP (STR measures the departure of thesperm path from a straight line)Average value of the ratio VSL/VCL (LIN measures the departure of thesperm track from a straight line)The ratio of number of motile sperm to the total number of spermNumber of sperm, in millions per gram of tissueSperm MotionTrack Speed(VCL)Path Velocity(VAP)ProgressiveVelocity (VSL)Figure 10.9Illustration of sperm velocities and motion parameters.a. Epididymal Sperm CountsSamples should be processed and analyzed sequentially across treatment groups (i.e., first animalin each group followed by the second animal in each group). (Note: The Hamilton-Thorne IntegratedVisual Optical System [IVOS] is used only as an example of a CASA system. Cryo Resourcesmakes the CellSoft Computer Assisted Sperm Analyzer, the other CASA system available for ratsperm analysis at this time. No endorsement of either system is intended or implied.)Routinely, the first printout of the summary report for each animal should be maintained in thestudy records and defined as raw data. An optical image of each field analyzed should be storedon an optical disk that can be maintained for future analysis. The following description of the useof a CASA system assumes that the system has been properly installed, the sperm parameters havebeen defined, and the system has been validated for use in a study. Select the study setup (e.g.,Analysis Setup No. 7 [Rat Ident]). A unique filename for each study should be entered for printingASCII files. The setup report for each day should be printed and maintained in the study records.Check each day of use to determine that the CASA system is correctly identifying sperm byselecting the “playback” option. Approximately 20 sperm per field should be counted.© 2006 by Taylor & Francis Group, LLC

452 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1) Epididymal Sample Preparation — When the rat is euthanized, the epididymides areremoved and weighed. If the selected epididymis cannot be evaluated immediately, it can be storedfrozen until evaluation. To freeze the epididymis, place it in a Bitran ® bag, flash freeze it in liquidnitrogen, and store it in an ultralow freezer (–70°C) until ready for analysis. When ready to evaluatethe epididymis, thaw the specimen. For fresh tissue or after thawing the frozen tissue, excise thecauda epididymis from the caput and corpus. Trim off the fat, and cut the vas deferens close to thecauda. Weigh and record the cauda weight. (Note: If the cauda weight is less than expected, thevolumes must be adjusted.) Place the cauda epididymis in a glass vial containing 15 ml of diluent,consisting of Triton X-100 [0.05% (v/v)] in 0.9% (w/v) NaCl (i.e., 1.8 g NaCl and 100 µl TritonX-100 in 200 ml deionized water). Homogenize (e.g., with a Polytron ® homogenizer at setting 5)on ice for approximately 1 min. Bring the homogenate up to a final volume of 40 ml and vortexon high until the sample is mixed. Transfer 2 ml of homogenate to a scintillation vial. Sonicate(e.g., with a Sonifier ® at setting 3) on ice for approximately 30 sec. Store vials on ice until analysis.2) Epididymal Sperm Count Analysis — Place 200 µl of well vortexed homogenate into theIDENT stain tube (for DNA staining) and vortex to mix thoroughly. Incubate at 37° ± 2°C for 3to 5 minutes. Using a standard pipette tip, transfer approximately 6 µl of the suspension to a clean2X-CEL 20-µm slide with coverslip. (Note: The 2X-CEL 20-µm slides from the vendor may notbe clean. Clean before use, if necessary.) Load the slide into the IVOS, and let the temperatureequilibrate for approximately 20 sec. Move to the fifth or sixth field of cells using the Jog In/JogOut button. Select the Ident Motile option and focus the sample on the screen. Visually examinethe sample to determine if it is acceptable. If the image field is unacceptable (e.g., clumping isevident, too many or too few sperm are present, excessive debris or bubbles in the field), repeatthe steps until an acceptable image field is found. Use Manual Select to select the first field to beanalyzed. Move to the next field and repeat the selection process until a total of 10 fields have beenselected. Run Start Scan. (Note: If a field is out of focus, select Start Scan, refocus, and repeatsteps until 10 fields have been counted.) The summary page printout should be maintained in thestudy records.b. Testicular Homogenization-Resistant Spermatid Counts1) Testicular Sample Preparation — When the rat is euthanized, the testes are removed andweighed. (Note: If the testis weight is less than expected, adjusted volumes must be used.) If thetestis cannot be evaluated immediately, it can be stored frozen until evaluation (as described abovefor the epididymis). For fresh or thawed testis, remove and weigh the tunica albuginea. Place theparenchyma in a glass vial containing 15 ml of diluent (see above for diluent composition).Homogenize on ice for approximately 30 sec. Bring the homogenate up to a final volume of 20ml and vortex on high until the sample is mixed. Transfer 2 ml of homogenate to a scintillationvial. Sonicate on ice for approximately 45 sec. Store vials on ice (approximately 4°C) until analysis.This process varies in the duration of homogenization and sonication from the methods publishedby Meistrich and van Beek. 972) Testicular Homogenization-Resistant Spermatid Count Analysis Using CASA — Usethe same methodology described for the epididymal sperm count using the CASA system (above).(Note: If no stained spermatids are present, check that the stain was properly mixed. If the stainwas not properly mixed, invert the stain kit vial two or three times, tapping it each time, and thenrevortex with homogenate and allow time to stain.)c. Manual Epididymal and Testicular Sperm Counts1) Epididymal Sample Preparation — Use the same methodology described for the epididymalsperm count (above), except homogenize the epididymis for 30 sec to 3 min to adequately© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 453disrupt the tissue. Bring the homogenate to the appropriate final volume (generally between 50 and100 ml), using deionized water. Record the final volume in the study records.2) Testicular Sample Preparation — For fresh tissue or after thawing the frozen tissue, weighand record the testis weight. Remove and weigh the tunica albuginea. Place the parenchyma in aglass vial containing 30 ml of diluent. Homogenize the parenchyma for 30 sec to 3 min to adequatelydisrupt the tissue. Bring the homogenate up to appropriate final volume (generally between 50 and100 ml), using deionized water. Record the final volume in the study records.3) Manual Sperm and Spermatid CountsLoading the Hemocytometer — Manual concentration analysis is conducted by producing ahomogenate of standard volume from a cauda epididymis or testis sample of known weight andcounting the number of sperm heads in a known volume of homogenate by use of a hemocytometer.Place the coverslip on the hemocytometer. Transfer 10 µl of the suspension into both chambers ofthe hemocytometer. Touch the center edge of the coverslip with the pipette tip and allow eachchamber to fill by capillary action. To facilitate counting, place the loaded hemocytometer in ahumid area (e.g., a Petri dish with moist gauze) for approximately 5 min so the sperm heads cansettle to a common focal plane.Counting using the Hemocytometer 96 — Count the four corner squares and the middle squarein each chamber. Include cells touching the middle line at the top and the left side. Do not countcells touching the middle line at the bottom and on the right side. Obtain two sets of hemocytometercounts (total of four chambers counted). If there are 15 sperm or less per counting chamber, recordall counts. If there are greater than 15 sperm, determine if a recount is necessary within each setof hemocytometer counts. Generally, if there is less than a 20% difference between hemocytometercounts, the sperm head counts are recorded and accepted. If there is a 20% to 50% differencebetween counts, the counts are recorded, the hemocytometer is cleaned and reloaded, and the newsample is counted and recorded. If greater than 50% difference exists between counts, the countis recorded but not used for tabulation. Then the hemocytometer is cleaned and reloaded, and anew sample is counted and recorded. (Note: Remember to always clean the hemocytometer andcover slip immediately after completion of counting by rinsing with deionized water and air dryingor drying with lens tissue.)4) Epididymal Calculations for Manual Sperm and Spermatid CountsSperm per Cauda EpididymisSperm per cauda epididymis =mean count multiplied by dilution volumevolume of counting chamberFor example, with a mean chamber count of 20, a dilution volume of 100 ml, and a chamber volumeof 0.0001 ml, the calculation would be (20 × 100)/0.0001 = 2 × 10 7 sperm per cauda epididymis.Sperm per Gram Cauda EpididymisSperm per gram cauda epididymis =sperm count per caudaweight of cauda (grams)For example, with a cauda sperm count of 2 × 10 7 and a cauda weight of 0.25 g, the calculationwould be 2 × 10 7 /0.25 = 8 × 10 7 sperm per gram epididymis.© 2006 by Taylor & Francis Group, LLC

454 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION5) Calculations for Manual Testicular Sperm and Spermatid CountsSpermatids per TestisSpermatids per testis =mean count multiplied by dilution volumevolume of counting chamberFor example, with a mean chamber count of 20, the dilution volume of 100 ml, and a chambervolume (hemocytometer factor) of 0.0001 ml, the calculation would be (20 × 100)/0.0001 = 2 ×10 7 spermatids per testis.Spermatids per Gram TestisSpermatids per gram testis =For example, with a testicular sperm count of 2 × 10 7 and a parenchyma weight of 0.25 g, thecalculation would be 2 × 10 7 /0.25 = 8 × 10 7 spermatids per gram testis.Spermatid Production — Spermatid production can also be expressed as number of spermatidsproduced per day per gram of tissue. This calculation uses the amount of time (in days) duringspermatogenesis while these cells are resistant to homogenization (4.84 d for mice and 6.10 d forrats).Daily sperm production =For example, with a value for spermatids per gram testis of 8 × 10 7 and a daily production factorof 6.10 (rat), the calculation would be 8 × 10 7 spermatids per gram testis/6.10 d = 1.3 × 10 7spermatids per day per gram testis.3. Cauda Epididymal Sperm Motilityspermatid count per testisweight of parenchyma (in grams)spermatid count per gram tissuedays resistant to homogenizationSperm motility estimates may be obtained from sperm samples collected from the cauda epididymisor vas deferens, or from ejaculated sperm samples. Sperm motility can be recorded and evaluatedimmediately, or the sperm images can be recorded, stored, and analyzed later, either by CASA ormanually. 84,98–100 Sperm motility is a reproductive endpoint and an integral part of some reproductivetoxicity tests. 101–103 Standardized procedures within the laboratory are essential. Sperm motility canbe affected by many factors, including (1) method of sample collection and handling, (2) elapsedtime between sampling and observation, (3) dilution medium characteristics (e.g., pH and concentrationof energy sources), (4) temperature (of dilution medium, slides, and room), (5) spermdilution, and (6) the microscopic chamber employed for the observations. 87,90,104–106The procedures described below can be used to collect, prepare, and analyze sperm motility insamples from the rat epididymis by means of a CASA system (e.g., IVOS). These procedures aregenerally chronological, but it is not absolutely necessary that the order presented be rigidlyfollowed. The sample collection methods generally follow the procedures outlined in Toth et al.. 107An image of each field analyzed should be stored on an optical disk. These images can be maintainedfor future analysis. A printout of the summary report is routinely maintained in the study recordsand is often defined as the raw data.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 455a. Procedures1) Set-up of the IVOS for Determining Sperm Motility — Turn the optical disk drive on firstand off last. Check the magnification with a 100-µm grid prior to the start of the study or when changingfrom analyzing counts to motility. Document magnification in the study records. Calibrate the phaseannulus daily before using the IVOS and document in the study records. Check the illumination dailybefore using the IVOS and document the low photometer reading (reading must be between 95 and105) in the study records. Select the appropriate analysis setup. Assign a unique filename to each studyfor printing ASCII files. Print the Setup Report daily for each study and maintain in the study records.2) Sample Collection — Prepare the appropriate volume of 10 mg/ml bovine serum albumin/phosphatebuffered saline (BSA/PBS) solution daily. (Note: Use an unopened bottle of PBS foreach day of solution preparation. Weigh 1.0 g BSA for every 100 ml of PBS required and mix untildissolved.) Place 10 ml of the BSA/PBS solution into Petri dishes for each sample and incubate atapproximately 36°C until use. Document the incubator temperature in the study records each dayof usage. Prepare test tubes with appropriate amounts of BSA/PBS diluent to be used for 1:9 dilution(i.e., 100 µl sample to 900 µl diluent), cover them with Parafilm ® , and incubate them at approximately36°C until use. Euthanize the male rat. (Note: Tissue harvesting should be done immediately. Donot allow the carcass to become chilled.) Open the abdominal cavity and excise the epididymis.Place the epididymis in a tared weigh boat, weigh it, and record the weight. Isolate the caudal portionof the epididymis with a scalpel blade. Place the cauda epididymis in a tared weigh boat, weigh it,and record the weight. Transfer the cauda epididymis to a prelabeled Petri dish containing 10 ml ofBSA/PBS diluent prewarmed to approximately 36°C. Pierce the epididymis four to six times, usingan 18-gauge needle. Cover the Petri dish and incubate it at approximately 36°C to allow for spermswimout. Five minutes is the optimal amount of time for swimout; 5 to 8 min is an acceptable time.It is important that swimout times be kept uniform. Record incubation time in the study records.Upon removal of the dish from the incubator, gently swirl the Petri dish and its contents to dispersethe sample. Place the appropriate amount of sperm solution into the dilution tube (e.g., 100 µl in a1:9 dilution tube), re-cover it with Parafilm, and gently mix it by inverting the tube. Transferapproximately 15 µl of the diluted suspension to each chamber of an 80-µm 2X CEL slide.3) Sample Analysis — Load the chamber into the IVOS. Perform a preliminary check to determinethe suitability of the dilution used. Twenty to forty sperm per analysis field is considered asuitable dilution. If the sample appears too dilute, redilute (1:4) from the original sample. (Note: TheJog In/Jog Out button is used to move through the various sample fields.) Enter the appropriate diluentratio (e.g., 1:9, 1:4, etc.), animal number, initials, and date in the information fields. Select the firstsample field to be analyzed and press Manual Scan. (Note: The technician must check that eachsample field is acceptable for analysis, i.e., no large debris or air bubbles, etc.) If the sample field isnot acceptable, select the next field or prepare a new slide for analysis. Press the Jog In/Jog Outbutton to move to the next field, and press Manual Scan to select the next field to be analyzed. PressStart Scan after the desired sample fields have been selected. Review the count summary that appearson the screen to ensure that at least 200 sperm have been analyzed. If less than 200 sperm have beenanalyzed, move to a new field and select by pressing Add Scan. Keep adding new sample fields untilat least 200 cells have been analyzed. Print the summary and maintain printouts in the study records.4) TroubleshootingSperm Are Present but Not Motile — If sperm are present but not motile, check the pH of themedium (should be 7.2), the temperature of the medium (approximately 36°C), the time passedsince sample collection (approximately 15 min or less), the latency from death to harvesting tissue,and the necropsy observations concerning the left epididymis, the left testis, or the seminal vesicles.pH Is out of Range — If the pH is discovered to be less than 7.1, note the deviation in the commentssection of the information screen and ensure that the correct pH is listed in the pH field. For© 2006 by Taylor & Francis Group, LLC

456 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONsubsequent samples, place the medium into the Petri dish earlier, giving it more time to shift towardthe alkaline side. If placing the medium into the dish earlier does not allow the pH to shift backinto range, then try adjusting the medium with 1 N NaOH.If the pH is discovered to be greater than 7.3, note the deviation in the comments section of theinformation screen and ensure that the correct pH is listed in the pH field. For subsequent samplesadjust the medium with 1N HCl.Temperature Is out of Range — Record the slide warmer temperature in the comments section of theinformation screen with the added comment that the medium temperature was out of range. Adjust theslide warmer temperature and wait until it is within range before preparing another sample.Too Much Time Has Passed Since the Sample Collection — Record that an excessive amount oftime has passed in the comment section of the information screen. Reduce the time for samplecollection by ensuring that the epididymis is collected, trimmed, and weighed as quickly as possible.Complete sample collection prior to all other steps of the necropsy, or assign the remaining necropsysteps to another technician.Increased Latency from Death to Harvesting Tissue — Ensure that no excess time passes betweenthe time of death and the time of sample collection.Check Necropsy Observations — If there are necropsy findings for the epididymis, testis, or seminalvesicles, note relevant observations in the comments section of the information screen.No Sperm Are Present — First, prepare another slide. If no sperm are present, record the finding.Check to see if there are necropsy observations concerning the epididymis, the testis, or the seminalvesicles that might explain the lack of sperm.4. Sperm MorphologySperm morphology refers to structural evaluation of the sperm. Sperm morphology can be evaluatedfrom vas deferens or ejaculated samples, but it is routinely evaluated using cauda epididymalsamples. Sperm morphology profiles are relatively stable in a normal individual (and a strain withina species) over time, which may enhance their use in the detection of spermatotoxic events. 108 Thetraditional sperm morphology methodology characterizes defects of the sperm head, midpiece, andflagellum in either stained preparations (by brightfield microscopy) 109 or fixed unstained preparations(by phase contrast microscopy). 90,110 Data should be collected that categorize the spermabnormality by type and determine the frequency of occurrence. However, because rodents have ahigh proportion of normal sperm with little morphological variability, it has been recommendedthat the sperm should be classified as normal or abnormal for statistical analysis so that changesin the proportion of total abnormal sperm can detected. 90 Objective, quantitative approaches (e.g.,automated sperm morphology analysis) are now available and should result in a higher level ofconfidence than more subjective measures. 111a. Sperm Morphology (from Samples Not Previously Used for Motility Analysis)Prepare 1% Eosin Y solution in deionized water. Euthanize the male rat using CO 2 anesthesia andexsanguination. Make a midline abdominal incision, grasp the fat surrounding the caput epididymis,and pull the testis into the abdominal cavity. Excise the testis by cutting the vas deferens about 0.5cm from the cauda epididymis, dissect any remaining tissues, and place the testis plus epididymisin a weighing pan. Carefully trim away surrounding fat and remove the intact epididymis from thetestis. Weigh testis and epididymis (to the nearest 0.1 mg). The cauda region is separated from thecorpus at the point just below where the vas deferens exits the epididymis. Isolate the caudaepididymis with a scalpel blade, and weigh it to the nearest 0.1 mg.Transfer the cauda epididymis to a Petri dish containing 10 ml of Dulbecco’s phosphate-bufferedsaline prewarmed to 37°C. (Note: As an alternate buffered saline, use 1% Medium 199 modified[M199] in Hank’s Salts with 20 mM HEPES buffer and 5% BSA added, but without sodiumbicarbonate and L-glutamine.) Pierce the epididymis 4 to 6 times, using an 18-gauge needle, and© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 457incubate it approximately 15 min at 37°C. Periodically gently swirl the Petri dish and its contentsto facilitate release of sperm from the cauda. Pipette the suspension gently approximately 20 times,using a 200-µl pipette with wide-bore tip to disperse sperm aggregates while being careful not tocreate air bubbles in the suspension. Prepare a 1:1 dilution of sperm suspension in 1% Eosin Ystain (i.e., 200 µl sample to 200 µl Eosin Y stain). Gently tap the bottom of the test tube to mix.Incubate at room temperature for approximately 30 to 60 min to allow for staining. Gently remixthe suspension with the 200 µl pipette. Prepare a smear (using a procedure similar to making ablood smear) on a labeled microscope slide. A Miniprep ® Blood Smearing Instrument (GeometricData, Wayne, PA) can be used to prepare uniform smears. Prepare two to four additional slidesfrom each sample to use for confirmation or for additional evaluation. Air dry slides overnight.Samples may be fixed (using methanol or formalin) and a cover slip applied with Permount orother suitable medium for long-term storage.b. Sperm Morphology (from Samples Previously Used for Motility Analysis)Prepare 1% Eosin Y solution in deionized water. Incubate cauda epididymis sample at 37°C forapproximately 15 min. Follow the same procedures described above. Remove the cauda epididymisfrom the dish and save it in an appropriate fixative for possible histological examination.c. Evaluation of Sperm MorphologyA minimum of 200 to 1000 sperm per animal should be morphologically examined at 400× to1000× magnification. Slides should be quickly scanned to determine the quality of the preparation.Common artifacts are clumping, slides that contain too many sperm, and improperly prepared spermsmears. Most sperm smears will contain at least 500 sperm. Sperm on the periphery of the slideare routinely excluded from examination because of the greater likelihood of preparation artifacts(broken sperm, clumping, etc.).The assessment and classification of rat sperm morphology is a subjective technique. Abnormalitiesmay affect the head, midpiece, or tail, 109 and some sperm may have multiple abnormalities.Values for normal sperm vary between laboratories, ranging from 94% to 99.7%. 70 In normal humanmales, up to 50% of the sperm may be misshapen. There are no standard classifications for ratsperm, 112 but many investigators use the sperm morphology classifications proposed by Linder etal. 110 Spermatozoa were classified as normal, normally shaped head separated from the flagellum,misshapen head separated from the flagellum, misshapen head with a normal flagellum, misshapenhead with a normal flagellum, and normal head with degenerative flagellum. Other descriptionsinclude banana-head, pinhead, two-head, two-tailed, bent tail, and blunt hook.H. Seminiferous Epithelium Staging1. SpermatogenesisWhile no quantitative assessment of the spermatogenic cycle (e.g., frequency distribution of tubulesin each stage) is required by current reproductive regulatory guidelines, qualitative knowledge ofstaging is needed to evaluate the normality of the cellular composition of the seminiferous tubule(e.g., which cells are missing or are inappropriately present). A detailed description of spermatogenesisand spermiogenesis is beyond the scope of this chapter; however, detailed morphologicaldescriptions with light and electron microscopic photographs of these processes are available in abook by Russell et al. 68 The process of formation and development of the spermatozoon is knownas spermatogenesis. Spermatogenesis is routinely divided into three phases: the proliferative, meiotic,and spermiogenic phases.© 2006 by Taylor & Francis Group, LLC

458 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFlagellar development is a continuous process, continuing until sperm release. Sperm areessentially immotile when released into the lumina of the seminiferous tubules. Sperm motility isdeveloped in the epididymis, but vigorous motility occurs only after deposition into the femalereproductive tract.2. The Staging of SpermatogenesisWhile no guideline studies require staging of spermatogenesis, the FDA Redbook 2000 states“testicular tissue should be examined with a knowledge of testis structure, the process of spermatogenesis,and the classification of spermatogenesis.” This statement implies that at least the pathologistshould be aware of spermatogenesis staging. Other regulatory reproductive study guidelineshave similar statements.The classification of spermatogenesis is based on the examination of cross-sectioned seminiferoustubules, and the organization of their cells has been extensively studied in the rat. 113,114 Astage is “a defined grouping of germ cell types at particular phases of development” in crosssectionedtubules. Each cell type of the cell association is morphologically integrated with othersin its developmental processes, 115 and each stage has a defined germ cell composition. The ratspermatogenic cycle has been divided into 14 stages (Roman numerals I to XIV).Single cell associations develop and remain in one region of the tubule, called a segment.Spermatids will develop from Stage I through Stage XIV. Although the cells do not move withinthe tubule, the cell associations are distinctly arranged along the length of a tubule. Starting at therete testis and proceeding into both ends of the seminiferous tubule, segments will show a linkedorder of stages decreasing as they go further into the tubule. For example, if the initial segmentnext to the rete testis were a Stage IX, the next segment further away from the rete testis wouldbe at Stage VIII, then Stage VII, VI, and so forth. The linking of the adjacent consecutive cellassociations is called the continuity of segmental order. The pattern of decreasing segmental orderin a direction away from the rete testis is called the descent in segmental order. The descent insegmental order continues from both ends of the seminiferous tubule (away from the rete testis)until the two patterns of descent meet at a point within the tubule. This point is called the site ofreversal. The site of reversal is not equidistant from the rete testis. This decreasing segmentalpattern is not perfect, as some variations, called modulations, do occur (e.g., Stage X, IX, VIII, IX,VIII, VII). A complete series of stages along a tubule (including modulations) is called a wave ofthe seminiferous epithelium. Many waves (usually about 24) are usually present in a seminiferoustubule. In rats, a wave averages 2.6 cm in length, but waves are not uniform in length (rangingfrom 1 to 5 cm). Spermatogenesis (from spermatogonia to spermatozoa) requires about 63 d in theSprague-Dawley rat. The cell cycle duration is the time that it takes to go from Stage I to StageXIV. The cell cycle duration is 12.9 to 13.3 d in the Sprague-Dawley rat. 116A. BiologyIV. FEMALE REPRODUCTIVE TOXICOLOGYEvaluation of the rodent female reproductive system usually consists of a gross evaluation (includingorgan weights) plus a histological evaluation of the ovary, uterus, cervix, vagina, and mammary glands.Vaginal cytology is assessed for determination of the estrous cycle. Female reproductive function isregulated through complicated interactions involving the central nervous system (particularly the hypothalamus),pituitary, ovaries, the reproductive tract, and the secondary sexual organs. In addition,nongonadotrophic components of the endocrine system may also affect the female reproductive system.To understand the significance of effects on female reproductive endpoints, the relationships betweenthe reproductive hormones and the reproductive organs must be understood.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 459B. ObservationsThere are three basic types of observations — a viability check, a clinical observation, and anextended or detailed clinical observation — which are performed as described above for males.C. Female Body WeightBody weights are recorded daily, twice weekly, or weekly, depending upon the guideline requirementsand reproductive phase (premating, gestation, or lactation). During cohabitation and gestation,body weights are recorded daily or at protocol-specified intervals (e.g., GD 0, 6, 9, 12, 15,18, and 20). Maternal and pup body weights in the lactation or postnatal period are usually recordedweekly (PND 0, 7, 14, 21). Food consumption during gestation is usually recorded at the sameintervals as body weights. Body weights and food consumption should be recorded at approximatelythe same time of day during the course of the study. Monitoring female body weight during treatmentprovides an index of the general health status of the animal that may be relevant in the interpretationof reproductive effects. Female body weight fluctuates normally because estrogen and progesteroneinfluence food intake, energy expenditure, 117 water retention, and fat deposition rates. 118,119 Reductionin food intake and body weight in the female rat is one of the most sensitive noninvasiveindicators of an estrogenic compound. In addition to endocrine-related factors, body weight lossor reduction in body weight gain can be caused by other factors, including systemic toxicity, stress(such as restraint), treatment-induced anorexia, or decreased feed or water consumption due toaltered palatability caused by the test substance.Unless a direct causal link between the observed female reproductive effect and decreased bodyweight or body weight gain can be established, female reproductive system alterations should beconsidered an adverse effect and not secondary to the occurrence of systemic toxicity. Therefore,in the presence of mild to moderate body weight changes, alteration in a female reproductivemeasure should be considered an adverse female reproductive effect. In the presence of severebody weight change, it can be stated that an adverse effect on a female reproductive endpointoccurred, but the effect may have resulted from a more generalized toxic effect.D. Female Necropsy1. Gestation Day 13 Uterine Examination (Segment I Studies)The dams are routinely necropsied around GD 13 (early in the second half of gestation). The damis euthanized, and the uterine contents are examined (as described below). Uterine position andviability (live or dead) of each implant is recorded, as are the corpora lutea counts from each ovary.The dam is subjected to a complete necropsy, and the reproductive organs and other tissues specifiedin the protocol are weighed and saved for possible histopathological evaluation. Certain guidelinesrequire termination on GD 20 or 21 (term Cesarean section).2. Gestation Day 20 or 21 Term Cesarean Section (Segment II Studies)The dam is euthanized, the uterus is removed and weighed (recorded as gravid uterine weight),and the uterine contents are exposed. The number and uterine position of each resorption (early orlate) and/or fetus (viable or dead) are recorded. The number of corpora lutea from each ovary iscounted and recorded. The individual term fetuses are removed from their membranes, and theumbilical cord is clamped and cut. Each placenta is examined and weighed, if required, or discarded.The empty uterus is weighed and the value recorded. Tag, box position, or other suitable means ofidentification individually identifies each fetus. The fetus is examined externally for any abnormalities,sexed, and weighed. Abnormalities are classified as either malformations or variations. A© 2006 by Taylor & Francis Group, LLC

460 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmalformation is “a permanent structural change that may adversely affect survival, developmentor function” while a variation is a “divergence beyond the usual range of structural constitutionthat may not adversely affect survival or health.” 11 In rat studies, half of the fetuses are subjectedto visceral evaluation while either the remaining half or all of the fetuses are processed for skeletalevaluations. For a detailed discussion of necropsy, uterine evaluation, and the fetal gross, visceral,and skeletal evaluations used in the developmental toxicity study, see Tyl and Marr. 303. End of Lactation (Segment III Studies)In Segment III and multigenerational studies, the dam is killed after the weaning of the offspringon PND 21. The dam is euthanized, subjected to a complete necropsy, and the reproductive organsand other tissues specified in the protocol are weighed and saved for possible histopathologicalevaluation.E. Female Reproductive Organ Weight1. OvaryOvarian weight in the normal rat does not show significant fluctuations throughout the estrouscycle. However, persistent polycystic ovaries, oocyte depletion, decreased corpus luteum formation,luteal cysts, altered pituitary-hypothalamic function, and reproductive senescence may be associatedwith ovarian weight changes. Ovarian gross morphology and histology should be examined todetect alterations associated with ovarian weight changes.2. Uterus, Oviducts, Vagina, and Pituitary GlandUterine weight fluctuates throughout the estrous cycle, peaking at proestrus, when the uterus isdistended with watery fluid (sometimes misdiagnosed as hydrometra) in response to increasedestrogen secretion. Compounds that inhibit steroidogenesis and cyclicity can cause the uterus tobecome small and atrophic, thereby decreasing the uterine weight. Uterine weight change has beenused as a basis for comparing relative potency of estrogenic compounds in bioassays. 120The oviducts are not routinely evaluated or weighed in reproductive toxicity studies. Theoviductal weight fluctuates depending upon the stage of the estrous cycle. The oviductal lumencontains a variable amount of fluid, which is greatest at metestrus, the period of ovulation. 121 Grossevaluations can be of value in detecting morphologic anomalies, such as agenesis and segmentalaplasia.Vaginal weight changes usually parallel uterine weight changes during the estrous cycle,although the magnitude is less. The vaginal and cervical epithelia show cyclic changes in morphologyand thickness associated with stages of the estrous cycle.Pituitary gland weight can provide information into the reproductive status of the female.Pituitary weight normally increases with age and also increases during pregnancy and lactation.Increased pituitary weight often precedes cancerous lesions induced by estrogenic compounds, andconversely, reduced pituitary weight may result from decreased estrogenic stimulation. 122 However,the pituitary contains other cell types that are responsible for the regulation of a variety ofphysiologic functions unrelated to reproduction. Therefore, changes in pituitary weight do notnecessarily reflect reproductive impairment. Selective histological stains should be used to confirmchanges in regions of the pituitary associated with reproductive functions. If the pituitary weightchanges can be associated with changes in reproductive functions of the pituitary, then such pituitaryweight changes should be considered as adverse female reproductive effects.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 461F. Histopathological Evaluation of the Female Reproductive Organs1. OvariesNumerous articles, 123 chapters, 124–127 reviews, 128,129,131 and books 130 have been written concerningthe ovary and ovulation. A detailed description of ovarian histology and ovulation is beyond thescope of this chapter. The ovary serves a number of functions that are critical to reproductiveactivity, including ovulation of oocytes and production of hormones. The developing ovarianfollicles produce estrogen, while the corpora lutea formed after ovulation produce progesterone.Histopathological evaluation of the three major compartments of the ovary (i.e., follicular, luteal,and interstitial) plus the capsule and stroma should be performed to reveal possible toxicologicalconditions. 132–134 Methods described below can be used to quantify ovarian follicular development.Delayed ovulation can alter oocyte viability and cause trisomy and polyploidy in the conceptus.135–137 Altered follicular development, ovulation failure, or altered corpus luteum formation orfunction can result in disruption of cyclicity and reduced fertility. Therefore, significant increasesin follicular atresia, evidence of oocyte toxicity, interference with ovulation, or altered corpusluteum formation or function should be considered adverse female reproductive effects. In addition,significant changes in folliculogenesis or luteinization (e.g., increased cystic follicles, luteinizedfollicles), altered puberty, and/or premature reproductive senescence should be considered adversefemale reproductive effects.2. UterusThe histological appearance of the normal uterus fluctuates according to the stage of the estrous cycleand pregnancy. The endometrium consists of simple columnar epithelium with simple tubular glandswith little or no branching. During proestrus, the epithelium becomes more cuboidal, possibly a resultof compression and stretching of the dilated uterus. During proestrus and estrus, inflammatory cells(mostly neutrophils) infiltrate the myometrium and endometrial stroma. Inhibition of ovarian activitywith the resultant reduction in steroid secretion can result in endometrial hypoplasia and atrophy, aswell as altered vaginal cytology. Uterine hypertrophy and hyperplasia can also be caused by prolongedestrogen or progesterone treatment, respectively. Significant dose-related increases in endometrialhyperplasia, hypoplasia, or aplasia should be considered adverse female reproductive effects.3. OviductsThe oviducts are not routinely examined histologically in reproductive toxicity studies, althoughgross and histological evaluations of the oviduct for anomalies induced during development (agenesis,segmental aplasia, and hypoplasia) can be performed. Oviductal hypoplasia or hyperplasiacan occur because of reduced estrogen stimulation or prolonged estrogenic stimulation, respectively.Significant dose-related increases in oviductal hyperplasia or hypoplasia should be consideredadverse female reproductive effects.4. Vagina and External GenitaliaDevelopmental abnormalities of the vagina (agenesis, hypoplasia, dysgenesis) can be due to geneticfactors, prenatal toxicant exposure, or may have an unknown etiology. In utero exposure of femalesto agents that disrupt reproductive tract development can cause development of ambiguous externalgenitalia (e.g., masculinization of female rodents with increased anogenital distance). 138 Errors ingender determination can be made when observing newborn pups with ambiguous external genitalia,© 2006 by Taylor & Francis Group, LLC

462 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONresulting in erroneous at-birth sex ratio observations. Offspring from sex steroid–treated rodentfemales commonly have malpositioned vaginal and urethral ducts. Agents with estrogenic activity(such as DES) can accelerate the occurrence of vaginal patency while other agents (such asphytoestrogens and TCDD) can delay vaginal opening. Infantile or malformed vagina or vulva(including masculinized vulva or increased anogenital distance) and/or altered timing of vaginalopening should be considered adverse developmental effects.The estrous cycle can be monitored in rodents by observing changes in the vaginal smearcytology. 139,140 Repetitive occurrence of the estrous cycle stages at regular normal intervals suggeststhat neuroendocrine control of the cycle and ovarian responses to that control are normal. A morecomplete description of the vaginal cytology associated with the estrous cycle is presented below.In reproduction and fertility studies, vaginal smear cytology should be examined daily for atleast three normal estrous cycles (approximately 2 weeks) prior to treatment, after onset of treatment,and before necropsy. 141 In addition to the determination of estrous cycle length and the occurrenceor persistence of estrus or diestrus, daily vaginal smear data provide the incidence of spontaneouspseudopregnancy, an altered endocrine state reflected by persistent diestrus. The daily vaginal smeardata can be used to distinguish pseudopregnancy from pregnancy based on the number of days thesmear remains leukocytic. The evaluation of vaginal cytology also can detect onset of reproductivesenescence in rodents. 142 Generally, the estrous cycle will be lengthened or the females will becomeacyclic. Lengthening of the cycle may be a result of increased duration of estrus or diestrus. Asignificant alteration in the interval between occurrences of estrus for a treatment group whencompared with the control group should be considered an adverse female reproductive effect.5. PituitaryIn histological evaluations of the rodent anterior pituitary, the relative sizes of the acidophils andbasophils have been reported to vary during the reproductive cycle and pregnancy. 143 Significanthistopathological damage involving cells in the anterior pituitary that control gonadotropin orprolactin production should be considered an adverse reproductive effect.6. Mammary Gland and LactationThe development and growth, histology, and pathology of the rat mammary glands have been welldocumented. 144 Mammary tissue is highly endocrine dependent for development and function. 145–147Development of the six pairs of mammary glands in rats occurs between birth and puberty.Toxic agents can adversely affect mammary gland size, histological appearance, and milkproduction and release, and many exogenous chemicals and drugs or their metabolites can betransferred into milk. 46,148,149 Milk may contain lipophilic agents (mobilized from the adipose tissue)at concentrations greater than those present in the blood or organs of the dam, thereby exposingthe suckling offspring to elevated levels of such agents.During lactation, the mammary glands can be dissected, weighed, and processed for histologicalexamination. Techniques have been developed for measuring mammary tissue development, milkproduction, and milk composition in rodents. 147 Milk production may be estimated by weighingmilk-deprived litters before and after nursing or by measuring the milk collected from the stomachsof pups at necropsy. The DNA, RNA, and lipid content of the mammary gland and the compositionof the milk can be measured as indicators of toxicity.Reduced pup growth may be caused by altered milk palatability, reduced milk availability, orby ingestion of a toxic agent secreted into the milk. However, other factors unrelated to lactationcan cause reduced pup growth (e.g., altered maternal behavior, stress, litter size, or poor suckingability in the pup). Significant milk production decrements or alterations in milk quality, whethermeasured directly or reflected in altered development of the pup, should be considered adversereproductive effects. 38© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 4637. Reproductive SenescenceThere is a loss of the regular ovarian cycles and associated normal cyclical changes in the uterineand vaginal epithelium with advancing age in the female rat. Although the mechanisms causingreproductive senescence are not thoroughly understood, there are age-dependent alterations thatoccur within the hypothalamic-pituitary control of ovulation. 150,151 Cumulative estrogen exposuremay play a role in the loss of ovarian function in rats since adulthood estrogen treatment canaccelerate the loss of function. 152 In contrast, depletion of oocytes is the principal cause of the lossof ovarian cycling in humans. 153 Early onset of reproductive senescence in females (e.g., asevidenced by cessation of normal cycling, ovarian pathology, or an endocrine profile that isconsistent with reproductive senescence) should be considered an adverse female reproductiveeffect.8. Developmental and Pubertal AlterationsThe Guidelines for Developmental Toxicity Risk Assessment 11 and the Guidelines for ReproductiveToxicity Risk Assessment 38 are available as more detailed references to determine whether effectsinduced or observed during the pre- or perinatal period should be considered adverse. Significanteffects on sexual development (e.g., malformations of the internal or external genitalia, alteredanogenital distance) or age at puberty, whether early or delayed, should be considered adverse.Other adverse effects for females include alterations in age for various developmental landmarks(nipple development, vaginal opening) or alterations in vaginal cyclicity.G. Vaginal Cytology1. DescriptionEvaluation of the estrous cycle is used to detect alterations in the normal female reproductivesystem. Beginning at puberty (approximately 45 days of age in the rat), distinctive cyclic changesin uterine cytology occur in response to the ovarian hormones, estradiol and progesterone. Femalerats are polyestrous, with normal cycles ranging between 4 and 5 days. Estrous cycles continuethroughout the reproductive lifespan of the female, except during pregnancy and lactation orpseudopregnancy. Vaginal lavage is a simple, noninvasive method of collecting vaginal cells fordetermination of the estrous cycle. Vaginal lavage can also be used to confirm mating (presence ofsperm). Changes in vaginal cytology reflect changes in circulating ovarian hormone levels (estradioland progesterone). Monitoring of the estrous cycle in rats over several weeks provides informationabout the length of the cycle and about alterations in cyclicity, such as the induction of prolongedor persistent estrus or diestrus (pseudopregnancy).The estrous cycle is separated into three or four distinct phases: metestrus (also called DiestrusI), diestrus (also called Diestrus II), proestrus, and estrus. 154 The metestrus or diestrus (Diestrus Iand II) duration is usually 2 to 3 d, the proestrus duration is usually 1 d, and the estrus durationis also 1 d. Metestrus (Diestrus I) vaginal smears contain leukocytes, few polynucleated epithelialcells, and a variable number of cornified epithelial cells (some of these cells are beginning toroll). 140 Diestrus (Diestrus II) vaginal smears contains a mixture of cell types, primarily leukocytes,a variable number of polynucleated epithelial cells, and few to no cornified epithelial cells.During this period, vaginal epithelial cells are proliferating and sloughing off into the vaginallumen. The proestrus vaginal smear contains numerous, round, polynucleated epithelial cells (inclumps or strands), a variable number of leukocytes, and few to no cornified epithelial cells.Serum estradiol falls rapidly on the evening of proestrus as the ovarian follicles leutinize.Ovulation occurs early in the morning of vaginal estrus. Proliferation of the vaginal epitheliumdecreases, and cornified epithelial cells predominate. The estrus vaginal smear contains no© 2006 by Taylor & Francis Group, LLC

464 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONDiestrusFull DiestrusLeukocytesLeukocytesProestrusProestrusCornifiedepithelial cellsLeukocyteNucleatedepithelial cellsClumped nucleatedepithelial cellsEstrusCornified epithelialcellsIrregularly shapedcellsProlonged EstrusFigure 10.10Estrous cycle evaluation.leukocytes, few or no polynucleated epithelial cells, and many flat cornified cells. The changes invaginal cytology are shown in Figure 10.10.2. Vaginal Lavage MethodologyVaginal lavage for the determination of the stage of estrous can be performed at any time of theday; however, it is important that the samples be collected at approximately the same time eachday over the course of a study. A pre- and postcheck of the saline should be made. (Note: If glasspipettes are used, the rubber bulbs can be saved and washed for future use, while the glass stemsshould be discarded after each sampling. If plastic pipettes are used, the pipette should be discardedfollowing each sampling. Dropping pipettes should be held with the tip pointing down whenever© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 465they contain a vaginal lavage or saline sample, to avoid contaminating the bulb. Contaminated bulbsshould be discarded.)The female rat is removed from the cage, and her identification is verified. One or two drops(approximately 0.25 ml) of physiological saline are drawn into a new clean dropping pipette, andthe tip of the pipette is gently inserted into the vaginal canal. The pipette bulb is firmly but gentlydepressed to expel the saline into the vagina. The saline is gently drawn back into the droppingpipette, and the pipette is removed from the vaginal canal. If the flush is very clear, the lavageshould be repeated. Care must be taken not to insert the tip too far into the vaginal canal becausestimulation of the cervix can result in pseudopregnancy (prolonged diestrus). The animal is returnedto its cage. The contents of the pipette are then delivered onto either (1) a clean glass slide or (2)a ring slide containing designated areas for the placement of each sample (vaginal smear). Eachslide should be numbered, the top and bottom of each slide clearly identified, and a correspondingrecord made indicating the animal number and location for each vaginal smear. Technicians shouldrecord the date, the start and end times of sample collection, and their initials on appropriate forms.All slides should be handled carefully to avoid mixing vaginal lavage samples from differentanimals. If samples become mixed, they are discarded, and the sampling procedure is repeated forthose animals.3. Estrous Cycle StagingEach vaginal lavage sample (fresh or stained) is examined microscopically (low power, 100×). Thetime and the technician’s initials should be recorded on the estrous cycle evaluation data sheetwhen sample evaluations begin. Cellular characteristics of vaginal smears reflect structural changesof vaginal epithelium. The stage of the estrous cycle is determined by recognition of the predominantcell type present in the smear. Diestrus smears contain predominantly leukocytes, proestrus smearscontain predominantly round, polynucleated epithelial cells (in clumps or strands), and estrus smearscontain predominantly flat cornified cells. The stage of the estrous cycle is recorded for each animalon the appropriate data sheet. The end time and the technician’s initials should be recorded on theestrous cycle evaluation data sheet.4. Estrous Cycle EvaluationMeasurement of estrous cycle length is done by selecting a particular stage and counting until therecurrence of the same stage (usually day of first estrus, day of first diestrus, or day of proestrus).The estrous cycle lengths should be calculated for each female. It is often easier to visualize theestrous cycle by plotting the data (Figure 10.11). The group mean estrous cycle length should becalculated, and the number of females with a cycle length of 6 d or greater should be calculated.There is a general consensus that a single cycle with a diestrus period of 4 d or longer or an estrusperiod of 3 d or longer is aberrant. Cycles that have two or more days of estrus during most of thecycles are classified as showing persistent estrus. Cycles that have four or more days of diestrusduring most of the cycles are classified as showing persistent or prolonged diestrus. Pseudopregnancyis a persistent diestrus (approximately 14 days in duration). Constant estrus and diestrus, ifobserved, should be reported. Persistent diestrus indicates at least temporary and possibly permanentcessation of follicular development and ovulation, and thus at least temporary infertility. Persistentdiestrus, or anestrus, may be indicative of toxicants that interfere with follicular development, thatdeplete the pool of primordial follicles, or that perturb gonadotropin support of the ovary. Theovaries of anestrous females are atrophic, with few primary follicles and an unstimulated uterus,and as expected, serum estradiol and progesterone are abnormally low. The presence of regularestrous cycles after treatment does not necessarily indicate that ovulation occurred because lutealtissue may form in follicles that have not ruptured. However, that effect should be reflected inreduced fertility. Altered estrous cyclicity or complete cessation of vaginal cycling in response to© 2006 by Taylor & Francis Group, LLC

466 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONDays PretreatmentDay of Treatment12345678910111213141234567891011121314A P E d d P E d d P E d d P E d d P E d d P E d d P E d dB P E d d P E d d P E d d P E d d d d d d d d d d d d d dC P E d d P E d d P E d d P E d d P E E E E E E E E E E ED P E d d P E d d P E d d P E d d P E E d d P E E E d d PE P E d d P E d d P E d d P E d d P E d d d d d P E d d dA P E d d P E d d P E d d P E d d P E d d P E d d P E d dB P E d d P E d d P E d d P E d d d d d d d d d d d d d dC P E d d P E d d P E d d P E d d P E E E E E E E E E E ED P E d d P E d d P E d d P E d d P E E d d P E E E d d PE P E d d P E d d P E d d P E d d P E d d d d d P E d d dABCDEFigure 10.11Graphical display of vaginal smear data. Three methods of depicting the estrous cycle data arepresented. The rst method presents the data on a spreadsheet with the estrus in bold. Thesecond method presents the estrus and proestrus shaded. The third method uses stacked bars(estrus, 3 bars; proestrus, 2 bars; and diestrus, 1 bar). Cycle A represents a normal cycle. CycleB shows the pattern in constant diestrus or pseudopregnancy. Cycle C shows constant estrus.Cycle D shows persistent or prolonged estrus. Cycle E shows persistent or prolonged diestrus.(Adapted from Cooper and Goldman in An Evaluation and Interpretation of Reproductive Endpointsfor Human Risk Assessment, 1999.)toxicants should be considered an adverse female reproductive effect. Subtle changes of cyclicitycan occur at doses below those that alter fertility. However, subtle changes in cyclicity withoutassociated changes in reproductive or hormonal endpoints would not be considered adverse.H. Corpora Lutea CountsThe corpus luteum is formed from the thecal and granulosa cells of the postovulatory follicle andis a transitory endocrine organ. 155 Progesterone is the major hormone produced by the corpus luteumand is necessary for implantation and the maintenance of the pregnancy. Each corpus luteum appearsas a round, slightly pink, discrete swelling on the surface of the ovary. Corpora lutea can be countedwith the naked eye or under a dissecting microscope to determine the number of eggs ovulated.The number of corpora lutea should be recorded for each ovary. In a litter with no obvious signsof resorption, the number of corpora lutea should be equal to or exceed the number of totalimplantation sites. For partially resorbed litters, the number of corpora lutea may actually be less© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 467than the total number of implantations. When some of the corpora lutea are no longer required tosupport the remaining conceptuses, some of the corpora lutea involute to become corpora albicantia.However, the concordance between the number of corpora lutea that involute and the number ofresorbed implants is not precise.Corpora lutea can be either discounted when the count is less than the number of implants orthe number of corpora lutea can be set to correspond to the number of implants to correct forinvoluting corpora lutea in partially resorbed litters. 30 The corpora lutea count is used in determiningpre- and postimplantation losses. A reduction in the number of corpora lutea or an increase in thepre- and/or postimplantation loss should be considered an adverse reproductive effect.I. Uterine Evaluation1. Assessment of ResorptionsAn implanted conceptus that is undergoing autolysis is termed a resorption. An early resorption isdefined as a conceptus that has implanted but has no recognizable embryonic characteristics evidentupon examination. An early resorption is usually visualized by ammonium sulfate staining or byobservation through pressed glass plates. A late resorption is defined as fetal remains or tissuesthat have recognizable fetal characteristics (such as limb buds with no discernable digits present)and are undergoing autolysis. In some fertility studies where the necropsy is performed near term,fetuses are classified as live (fetus exhibits normal appearance, spontaneous movement, heart beat,etc.) or dead (nonliving full size fetus with discernable digits, with a body weight greater than 0.8g for rats). In fertility studies where the pregnant female rodents are necropsied around GD 13 to15, the conceptuses are categorized as live (normal appearance, heart beat, etc.) or dead (discolored,undergoing autolysis). The dead classification in these studies is analogous to resorption. Thenumber and uterine positions of any early and late resorptions and viable and dead fetuses arerecorded. Segment III and multigeneration studies routinely evaluate the reproductive organs of thedam at necropsy, postweaning.2. Assessment of Implantation in “Apparently Nonpregnant” DamsWhen dams appear not to be pregnant at necropsy, a thorough examination of the uterus must beperformed. If no implantation sites are observed, the opened uterus is stained with a 10% aqueoussolution of ammonium sulfide. 156 Record (under maternal gross findings) that the uterus was placedin stain. If implantation sites are found after staining, the pregnancy status must be updated.J. Oocyte QuantitationThe U.S. EPA OPPTS 807.3800 Reproductive and Fertility Effects Guideline 10 requires a quantitativeevaluation of the primordial follicle population of the ovary. The postlactational ovary shouldcontain primordial and growing follicles, as well as the large corpora lutea of lactation. Ovarianhistology, oocyte quantitation, and differential follicle count methodology have been described. 157Ovarian follicles have been classified into 10 categories. 158 These 10 categories have been reducedinto three classes for the purposes of quantification: primordial follicles, growing follicles, andantral follicles. The earliest primordial follicle is an oocyte with no surrounding granulosa cells(type 1), while the latest primordial follicle has a complete single layer of granulosa cells surroundingthe oocyte. The earliest growing follicle has an oocyte that is starting to grow, and thesurrounding cells have formed two layers (granulosa and theca), while the latest growing folliclehas multiple layers of granulosa and thecal cells but no evidence of antrum formation. The earliestantral follicle has a large oocyte, multiple cell layers, and a developing antrum, while the latestantral (Graafian) follicle has mature multiple layers of granulosa and thecal cells and a large antrum.© 2006 by Taylor & Francis Group, LLC

468 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONHistopathological examination should detect qualitative depletion of the primordial folliclepopulation; however, reproductive study guidelines require a quantitative evaluation of primordialfollicles of the F 1 females. The number of animals, ovarian section selection method, and sectionsample size should be statistically appropriate for the evaluation procedure used. The guidelinescite ovarian follicular counting methodologies 159–161 that use a differential count of the primordialfollicles, growing follicles, and antral follicles. The current EPA Multigeneration ProtocolModification 10 requires ovaries from 10 randomly selected females from highest treatment andcontrol groups. Five 5-µm ovarian sections are made from the inner third of each ovary, at least100 µm apart. The number of primordial follicles (and small growing follicles) is counted, and thepresence or absence of growing follicles and corpora lutea is confirmed for comparison of highesttreatment group and control group ovaries. If statistically significant differences occur, the nextlower treatment group will be evaluated.Differential follicular counts are associated with age of the female, number of animals evaluated,number and location of the ovarian sections, the number of sections per ovary, section thickness,differences between evaluators, and the criteria for defining primordial follicles. Ovarian folliclecount appears to be more sensitive endpoint than ovarian weight. In 18 RACB studies using Sprague-Dawley rats, 161 the number of small follicles counted had a wide variation, ranging from 147 ± 57to 556 ± 144, and another investigator reported 301 ± 13 small follicles. 131 A decrease in ovarianfollicle count is usually considered a biomarker of an adverse reproductive effect because norecovery is possible. A recent review of follicle count historical control data in rats 162 suggests that:(1) with a large number of animals, there is relatively good replication within a generation (whenevaluated by the same person); (2) there is great variability within individual ovaries and sections;(3) random selection of ovarian sections has a high probability of biasing effects; (4) ovarian folliclecounts do not appear to be correlated with either ovarian weight or body weight at termination;and (5) ovarian follicle counts should not be the sole endpoint used in risk assessment. However,even with the great variability between reported values, a detectable decrease in follicle count ina treated group when compared with the control group is considered an adverse effect.A. ParturitionV. PARTURITION AND LITTER EVALUATIONSFemale rats should be housed to litter in boxes with bedding material no later than day 20 ofpresumed gestation (2 d prior to the day of expected delivery; earlier if required by protocol). Thedams are observed periodically throughout the day for stretching, visible uterine contractions,vaginal bleeding, and/or placentas in the nesting box. These signs are strong indicators of parturitiononset, and indicate that dams should be monitored closely for delivery of the first pup.As soon as a pup is found, parturition is considered initiated. If required by protocol, the timeshould be recorded. A marker (e.g., a tape “flag”) may be placed on the outside of the nesting box,indicating a delivery in progress. Once delivery has begun, the dams are checked periodically duringthe day for evidence of difficulty in labor or delivery (e.g., a pup only partially delivered, a damcold to the touch and pale). Any apparent difficulty is recorded, and a supervisor or veterinarianshould be notified if the difficulty seems extreme or unusual.The calendar date on which delivery of a litter appears to be completed is defined as lactationday 0 (LD 0), postpartum day 0 (PND 0), and day 0 of age. The following maternal behaviorsoffer evidence that parturition is complete: (1) removal of amniotic sacs, placentas, and umbilicalcords, and grooming of the pups by the dam; (2) self-grooming by the dam; (3) nesting behaviorby the dam; and (4) nursing. However, parturition may be complete without all of these behaviorspresent. Gentle palpation of the dam may also be used. If the time of completion of parturition is© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 469required, that information is recorded. Individual pup observations, such as appearance and viability,should not be recorded until delivery is completed. Apparently dead pups should be removed fromthe nesting box during parturition to preclude cannibalization by the dam. The pups should beretained in a container such as a weigh boat, labeled with the dam number, until pup statuses arerecorded following completion of parturition. Pups delivered but not present at the end of parturitionbecause of cannibalization should be recorded as “undetermined sex, undetermined viability.”In order to preclude disruption of delivery and/or maternal care, dams in the process of deliveringpups should not be administered the test substance. If a dam has completed parturition by the timethe last additional animal in the normal sequence has been dosed, this dam can be administeredthe test substance also. If the dam continues to actively deliver after daily test substance administrationis completed, the dam’s daily administration can be skipped. A delayed or missed daily test substanceadministration and the reason for its occurrence should be recorded on the appropriate form.Occasionally, additional pups are born to a litter that has already been weighed and observedon PND 0. These pups are handled as follows:If delivery of a litter has been completed and it is within the same day, the additional pup(s) and damare weighed on “LD 0/PND 0” (for example, a Monday).If an additional pup(s) is found on PND 1 (for example, Tuesday), when this litter is counted, thefollowing comment (or a comment that conveys the same information) should be entered on thelitter observation form for PND 1. “One additional (sex) pup present that was delivered aftercompletion of weighing and recording of litter observations for all pups delivered on PND 0.” Theadditional pup(s) is not weighed until PND 4; the next scheduled weighing interval for that litter(for example, Friday). The pup(s) is tattooed if required by protocol.Edits will be required to change the number of pups in the litter. The date of the dam’s DL 0remains the same (it is not changed to the date that the additional pup was found). Pups are weighedon the days specified in the protocol. On those days when pup body weights are recorded, bedding isusually changed and each litter is evaluated for maternal and pup nesting behavior. Each litter shouldbe checked for viability and counted every day, with the number and status of the pups recorded.B. Milk CollectionPrior to milk collection, the pups are withheld from the dam for approximately 4 h. At least 5 minprior to milk collection, 0.05 ml of oxytocin (20 units/ml) is administered intravenously to the dam.The time that the oxytocin is administered should be documented on the appropriate form. Thedam is then held in a position that facilitates access to the teats. Milk is expressed from one teatat a time by applying gentle digital pressure. Initially the pressure is applied at the base of the teatand then is directed distally. The milk is collected into a microtainer or other suitable container(labeled with the protocol and animal number, date, identification of sample, and test system, i.e.,rat milk). The time of collection and volume of milk collected is documented on the appropriateform. If more milk is required, then another teat is selected and milked. When the milk collectionis completed, the dam is returned to her litter.C. Litter Evaluations1. Litter ObservationsThe entire litter is observed for general appearance and maternal and pup nesting behavior. Thefollowing items should be checked: (1) that the pups are warm and clean, (2) that there is evidenceof a nest, (3) that the pups are grouped together, and (4) that the pups are nursing or have milkpresent in their stomachs (milk spot present).© 2006 by Taylor & Francis Group, LLC

470 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION2. Gender DeterminationThe nesting box is removed from the rack, the dam’s identification is verified, the dam is placedinto the holding cage, the pups are removed from the nesting box, and the gender of each pup isdetermined. Anogenital distance (see below) is used to determine gender (the male anogenitaldistance is greater than the female anogenital distance). Pups that have been cannibalized to theextent that their gender cannot be determined are recorded as “sex unable to be determined due todegree of cannibalization.” If a malformation precludes sex determination, this is recorded (forexample, “tail absent, no apparent anus, sex could not be determined”). If the malformed pup isdead, the sex is determined by internal examination.3. ViabilityThe viability of each pup is determined. Live pups are recorded as alive. Dead pups are necropsied,the trachea is tied off with string, and the lungs are removed. The lungs are then placed in acontainer of water. If the dead pup’s lungs float, the pup is assumed to have taken a breath, andthe viability is recorded as “born alive, but found dead.” If the lungs sink, the pup is assumed tohave never breathed, and the viability is recorded as “stillborn.” Dead pups can be cannibalized tothe point where viability cannot be determined. The sex and viability status are recorded for eachpup (including dead, cannibalized, and malformed pups) in all litters.4. External AlterationsEach pup in the litter is evaluated for external alterations. The following items are evaluated foreach pup: the general shape of the head; presence and completeness of features; trunk continuity;absence of bruises, lesions, or depressions; length and shape of limbs (toes are counted); lengthand diameter of the tail; presence of anus; milk in stomach; and whether partial cannibalizationhas occurred. All live pups are weighed. If a live, partially cannibalized pup is weighed, its weightis footnoted in the raw data.5. Missing PupsIf a pup cannot be found during the daily litter count, the pups are sorted by gender to determinethe sex of the missing pup(s) by process of elimination. The missing pup is recorded as “missing,presumed cannibalized.” If the pups have been tattooed or otherwise identified, the actual numberof the missing pup should be recorded.6. Pups Found DeadAll found dead pups are necropsied. Stomach contents (i.e., “stomach contains or does not containmilk”) of all dead pups should be recorded. Unless otherwise dictated by the protocol, dead pupsfound on LD 0 to 4 with gross lesions are fixed in Bouin’s fixative. Gross lesions from pups founddead on LD 5 through weaning are fixed in 10% neutral buffered formalin. Because of postmortemautolysis, dead pups are not routinely weighed.7. Dam Dies before Scheduled SacrificeIf a dam with a litter dies before her scheduled sacrifice, the pups are sacrificed and necropsied.The sacrifice and necropsy observations are recorded.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 4718. All Pups in a Litter DieIf all the pups in a litter die, the dam may remain in the study until her scheduled sacrifice date,or the dam may be sacrificed at the discretion of the study director.9. Gender Determination ErrorsIf it is discovered that an error was made in the determination of pup gender, the appropriate editshould be written and the raw data (manual form or computer printout) corrected for that day andeach preceding day. For example, if a sexing error is found at necropsy of a culled pup on LD 7,an edit is written to correct the gender, and the data from the previous lactation days are corrected.After the litter observations, the bedding in the nesting box can be replaced with clean beddingor a new box with clean bedding can be obtained. The dam is weighed and observed for clinicalsigns, and the feed jar is weighed (if appropriate). The dam and pups are placed into the clean box,and the dam’s cage tag and feed jar are transferred to the new box. The nesting box is returned tothe rack.A. Pup Body Weights1. Pup Birth WeightVI. PUP EVALUATIONSBirth weight should always be measured for each pup on the day of parturition. Data from individualpups are reported as the entire litter weight and as weight by sex per litter. Growth rate in utero isinfluenced by the normality of the fetus, uterine milieu, uterine position, and fetal gender, withfemales tending to be smaller than males. 163 Maternal nutritional status, intrauterine growth rates,litter size, and gestation length influence birth weights. Individual pups in small litters tend to belarger than pups in larger litters. Therefore, reduced birth weights that can be related to large littersize or increased birth weights that can be related to small litter size should not be considered anadverse effect unless the altered litter size is considered treatment related and survivability and/ordevelopment of the offspring have been compromised.2. Postnatal Pup Body WeightsPostnatal weights are dependent on litter size and on pup gender, birth weight, suckling ability,and normality, as well as on maternal milk production. With large litters, small or weak offspringmay not thrive and may show further impairments in growth or development. Because one cannotdetermine whether growth retardation or decreased survival rate was due solely to the increasedlitter size, these effects are usually considered adverse developmental effects. Conversely, pupweights in very small litters may appear comparable to or greater than control weights and thereforemay mask decreased postnatal weights in other litters.3. Crown-Rump LengthCrown-rump lengths for fetuses are measured with a caliper. Fetuses may be euthanized beforemeasuring the crown-rump length. Place each fetus on its side on a flat surface, and place thecaliper with one prong touching the top of the head and the opposite prong touching the base ofthe tail. Do not stretch or compress the fetus. Record the measurements in millimeters.© 2006 by Taylor & Francis Group, LLC

472 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION4. Anogenital DistanceAnogenital distance is the length between the anus and the genital tubercle. In rodents and primates,the anogenital distance is greater for males than females. 54 For rodents, anogenital distances aregreater for males than females beginning on GD 17 and continuing through PND 21. Meananogenital distances at birth for males and females are approximately 3.5 and 1.4 mm, respectively.The increased growth of this region occurs in response to testosterone. The genders of rabbit fetusesor neonates cannot be reliably determined using anogenital distance but can be determined byexamining the gonads.Anogenital distances for rat fetuses (GD 20 or 21) or pups on PND 0 to 3 are measured witha micrometer and a stereomicroscope. Measurements taken on pups on PND 4 or later should bedone using a caliper. The fetus or PND 0 to 3 pup is held in the technician’s hand, and the tail israised with the other hand to an 80° to 90° angle from the horizontal, exposing the anus. (Note:Be careful not to pull the tail, as this stretches the anogenital distance.) The anogenital area isbrought into focus with a calibrated stereomicroscope. The anogenital distance is measured fromthe cranial (or anterior) edge of the anus, which comes to a point, to the base (or posterior edge)of the genital tubercle. The base of the tubercle is not clearly differentiated as it slopes into theanogenital area. A baseline between the distinct edges of the genital tubercle is visually estimated.It is very important that the anus and the base of the genital tubercle be kept in the same focalplane. The length from the base of the genital tubercle to the cranial edge of the anus is recorded.Hold the PND 4 or older pup by its tail, keeping the tail at an 80° to 90° angle from thehorizontal. The arms of the caliper should be aligned as follows: for males, the anogenital distanceis measured from the cranial (or anterior) edge of the anus to the base (or posterior edge) of theanogenital aperture; and for females, the anogenital distance is measured from the cranial edge ofthe anus to the base of the urinary aperture (not the base of the vulva). The anogenital distance isrecorded in millimeters. (For further description, see Chapter 9, Figures 9.9 through 9.11)B. Developmental LandmarksThe onset of various developmental landmarks can be used to assess postnatal development. Sexualmaturation landmarks (balanopreputial separation, vaginal patency) are required or recommendedin perinatal-postnatal, multigenerational studies, and developmental neurotoxicity studies, wherethe offspring are raised to adulthood. Additional developmental landmarks can be assessed includingpinna detachment, hair growth (pilation), incisor eruption, eye opening (pups are born with eyelidsclosed), nipple development, and testes descent. The development of the following reflexes canalso be evaluated: surface righting reflex, cliff avoidance, forelimb placing, negative geotaxis, pinnareflex, auditory startle reflex (pups are born with the external auditory canal closed), hind limbplacing, air righting reflex, forelimb grip test, and pupillary constriction reflex. The examinationsmust begin prior to the landmark’ historical day of onset and continue daily until each animal inthe litter meets the criterion. 164 Data for each day’s testing should be expressed as the number ofpups that have achieved the criterion for each developmental landmark, divided by the total numberof pups tested in the litter. Forelimb grip test and pupillary constriction tests are conducted on PND21 only. Data for the forelimb grip test and pupillary constriction tests should be expressed as thenumber of pups that have achieved the criterion, divided by the total number of pups tested in thelitter. The postnatal days listed below for each developmental landmark were compiled from severalsources and are subject to variability between laboratories and subtle differences in assessment ofthe criteria. 164–1661. Balanopreputial SeparationIn rats, balanopreputial separation is considered to result from balanopreputial membrane cornification,which leads to the detachment of the prepuce from the glans penis in the rat. Preputial© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 473separation occurs dorsolaterally and then ventrally on the penis and down the shaft of the penis. 54The prepuce remains attached to the glans penis on its ventral surface by the frenulum. (Note: Theprocess of development of the prepuce in humans is different from that of the rat.) Male rodentsare examined for balanopreputial separation beginning on PND 22 (mouse), PND 27 (hamster), orPND 35 to 40 (rat). Published preputial separation age ranges from PND 32 to 42 (Japanesestudies) 167–169 or PND 41 to 46 (American studies), 54,170 depending on the observation criterion. TheJapanese preputial separation criterion is based on the age that the prepuce separates from the glanspenis, while the American criterion is met when the prepuce completely retracts from the glanspenis and the foreskin is cuffed around the base of the penis. Therefore, it is important to definethe criterion in the protocol.Each male rodent is removed from its cage and held in a supine position. Because manipulationof the prepuce can accelerate the process of preputial separation, the males must be examinedgently. Gentle digital pressure is applied to the sides of the prepuce, and the criterion is met whenthe prepuce completely retracts from the head of the penis (see above for alternate criterion). Theforeskin can be attached along the shaft of the penis, but it cannot be attached to the opening ofthe urethra. Each male rodent is examined daily until acquisition or until PND 55, whichever isearlier. Body weight should be recorded on the day the criterion is met.2. Vaginal PatencyAfter the canalization of the vagina occurs in rats, the vaginal opening remains covered by a septumor membrane. The age when the septum is broken or no longer evident is described as the age ofvaginal patency, and vaginal patency is the most readily determined marker for puberty in rats. 54Published Sprague-Dawley rat vaginal patency values range from PND 30.8 to 38.4. 169,171 Femalerats are examined beginning a few days prior to expected age of maturation (e.g., PND 25 to 28),and the examinations are continued until the criterion for patency has been achieved or until PND43, whichever comes first. The female is removed from the cage and held in a supine position,exposing the genital area. Pressure is gently applied to the side of the vaginal opening to see if theseptum or membrane remains. When the membrane is present, the area has a slight “puckered”appearance; however, when the membrane has broken, the vaginal opening is about the size of apinhead. The criterion has been met when the vagina is completely open. Body weight should berecorded on the day the criterion is met.3. Day of First EstrusAfter vaginal opening has been determined, vaginal smears can be taken daily to determine theday of first estrus (defined as the first day on which only cornified epithelial cells are observed ona vaginal smear). First estrus normally occurs within a few days after vaginal opening (approximatelyPNDs 40 to 45) in the Sprague-Dawley rat. 1724. Pinna UnfoldingBeginning on PND 1, each pup in a litter is examined individually until all pups in the litter meet thecriterion for pinna unfolding (i.e., the first day that the point of a pinna [earflap] is detached from arodent pup’s head). Each pup is examined closely for detachment of either the right or left pinna fromthe side of the head, and criterion is met when the point of either pinna is detached. This usually occurswithin PND 1 164 to 7 165 in Sprague-Dawley rats, with the average acquisition around PND 2 or 3.5. Hair GrowthBeginning on PND 1, rodent pups are examined for the appearance of hair. All pups in each litterare examined daily until bristles appear on the dorsal surface of all pups in the litter. Criterion is© 2006 by Taylor & Francis Group, LLC

474 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmet when there are bristles on the dorsal surface. This usually occurs within PND 1 to 5 164 inSprague-Dawley rats.6. Incisor EruptionRodent pups are examined for the appearance of incisor eruption beginning on PND 7. Each pupin every litter is carefully examined for eruption through the gum of either an upper or lower incisor.Criterion is met when an incisor is visible. This usually occurs within PND 8 to 16 164,165 in Sprague-Dawley rats, with the average acquisition around PND 11. 1737. Eye OpeningBeginning on PND 10, all pups in each litter are examined daily until all pups in the litter meetthe criterion for eye opening. Each pup is carefully examined for any break in the membraneconnecting the upper and lower eyelids, and criterion is met when there is a break in the membraneof either eye. This usually occurs within PND 11 164 to 18 165 in Sprague-Dawley rats, with theaverage acquisition around PND 13 or 14.8. Nipple RetentionOn PND 11 through 13, all males in each litter should be examined for the presence of areolaand/or nipples, which should no longer be present by PND 12 or 13. Each pup is carefully examinedfor the presence of nipples and/or areola (usually only in the thoracic region) by brushing the haircoat against the nap. The data are usually presented as percent of males with areolae and percentof males with nipples.9. Testes DescentMale rodents are examined for testis descent beginning on PND 19. Each male rodent is removedfrom its cage and examined for the presence of one or both testes in the scrotum. This usuallyoccurs within PND 20 to 29 164 in Sprague-Dawley rats, with the published average acquisitionbetween PND 18 165,169 and 25. 166C. Behavioral or Reflex Testing1. Surface-Righting ReflexThe surface-righting reflex is evaluated beginning on PND 1 for rat pups. Surface righting isregaining the normal position after the pup is placed on its back. This is a complex coordinatedaction requiring many different muscles in the neck, trunk, and limbs. Pups are placed on theirbacks on a flat surface and quickly released. The criterion is met when the pup regains its normalposition on four paws on the floor within 5 sec. This usually occurs between PND 1 and 9 164,165 inSprague-Dawley rats, with the average acquisition around PND 6 169 or 7. 1662. Cliff AvoidanceCliff avoidance is evaluated beginning on PND 1 for all pups. Cliff avoidance is the behavior ofcrawling away from an edge of a flat surface edge (cliff). Pups are placed on a table or platform,with their front paws over the edge. The criterion is met when the pup attempts to crawl away fromthe edge within a 10-sec period. This usually occurs between PND 2 164 and 12 165 in Sprague-Dawleyrats, with the average acquisition around PND 8 or 9. 166© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 4753. Forelimb PlacingBeginning on PND 7, rodent pups are tested for forelimb placing or lifting in response to tactilestimulus on the dorsal surface of the foot. The pup is suspended by holding the scruff of the neckso that one forelimb is in contact with a stainless-steel plate (e.g., cage tag holder) held horizontalto the working surface by a clamp and ring stand. The dorsum of the suspended foot is gentlytouched with a thin (approximately 2 to 3 mm in diameter) metal rod. The pup must immediatelyraise and place the suspended foot on the rod to meet the criterion. This usually occurs aroundPND 9 to 10. 1654. Negative Geotaxis TestBeginning on PND 7, rodent pups are tested for the ability to change from a downward to anupward orientation on an inclined plane of approximately 30°. Each pup is placed facing “downhill”on a platform tilted at a 30° angle. The pup must turn 180° to face “uphill” within a 60-sec intervalto meet the criterion (a pup with its body positioned sideways and its head facing “uphill” doesnot meet the criterion). This usually occurs between PND 7 and 14 165 in Sprague-Dawley rats, withthe average acquisition around PND 7 169 or 8. 1665. Pinna ReflexThe pinna reflex tests the somatomotor component of the 7th cranial nerve in rats. The presenceof the pinna reflex is evaluated daily, beginning on PND 13. The inner surface (near the concha)of the pinna is lightly touched with a filament or the tip of an artist’s brush. The criterion is metwith the presence of any movement of the pinna (sudden twitch or flattening of the ear) made inresponse to the applied stimulus. If the first ear tested does not respond to the stimulus, the oppositeear is tested. This reflex usually occurs around PND 14 165 in Sprague-Dawley rats.6. Auditory Startle ReflexBeginning on PND 10, all pups are examined daily for the auditory startle reflex. The auditorystartle reflex is noted as a sudden flinch or cessation of ongoing movement following the auditorystimulus. Each nesting box is removed from the study room and placed in a quiet room. Littermatesremain outside the testing room in order to mitigate habituation to the auditory stimulus. The pupis placed into a container, such as a beaker, with approximately 600 ml of bedding material andtaken into the testing room. A clicker is held directly above the beaker, but not touching it, and theclicker stimulus is delivered. Any observable whole body response (e.g., flinching, jumping, andfreezing of activity) meets the criterion for the startle response. This usually occurs between PNDs12 and 13 164,165 in Sprague-Dawley rats.7. Hind Limb PlacingBeginning on PND 14, rodent pups are tested for hind limb placing in response to a tactile stimuluson the dorsal surface of the foot. Each pup is held so that one hind limb is in contact with themetal plate (as in IV.C.3. above). The dorsum of the suspended foot is gently touched with a thinrod. The pup must immediately raise and place the suspended foot on the rod to meet the criterion.This usually occurs around PND 16. 1658. Air Righting ReflexThe air righting reflex is the ability of pups to land on all four paws when dropped from an invertedposition. All pups in each litter are tested once each day beginning on PND 14 until all pups inthe litter demonstrate the reflex. The pup is held in a supine position above a well-padded surface© 2006 by Taylor & Francis Group, LLC

476 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONand then released. Rat and hamster pups are held approximately 38 cm above the surface; mousepups are held 17 to 20 cm above the surface. The pup must land on all four limbs to meet thecriterion. This usually occurs between PND 8 164,165 and 18 in Sprague-Dawley rats, with the averageacquisition around PND 16 169 or 17. 166 The use of videotape to record the response will allow moreinformation to be obtained, such as response speed, character, and progression. 174,1759. Forelimb Grip Test (Categorical)Each pup is tested for forelimb grip on PND 21. A thin rod (approximately 2 to 3 mm in diameter)is supported by a ring stand suspended horizontally above the padded surface. The pup is held soit can grasp the thin rod with its forepaws and is then released. It must remain suspended for atleast 1 sec to meet the criterion. The number of pups that met the criterion, divided by the totalnumber of pups tested, is recorded.10. Pupil Constriction ReflexAll pups in each litter are examined once on PND 21 for direct and consensual pupillary constrictionof both eyes in response to the beam from a penlight. This test evaluates the autonomic componentof cranial nerve reflexes. Each nesting box is transferred from the study room to a quiet isolatedtesting room, sufficiently dim to dilate the pupils of the eyes of the pups. Each pup is testedindividually, with the following requirements to meet the criterion:The pup is removed from the nesting box and the penlight is directed into one eye. Immediateconstriction of the pupil of the eye being tested is the initial requirement.Without turning the light off, it is immediately directed into the contralateral eye, the pupil of whichshould already be constricted.The light is turned off for a minimum of 5 sec. This sequence is then repeated for the other eye.Responses should be identical to meet the criterion.D. Examination of Rodents for HypospadiasIn males, hypospadias is defined as an opening of the urethra at an abnormal location on the ventralsurface of the penis. In normal males, the urethral opening is located at the tip of the penis. Inaffected males, the urethral opening extends a variable distance from the base of the penis. Otherphysical characteristics that may be seen in the affected males are inability to fully extend thepenis, distended preputial glands, undescended testes, and decreased anogenital distance. In females,hypospadias is defined as an opening of the urethra into the vagina rather that at the urinary apertureadjacent to the vaginal opening.1. In-life Procedure for Male RodentsMale rodents are examined beginning on PND 39 (rat) or PND 22 (mouse). Each male is removedfrom its cage and held in a supine position. Gentle digital pressure is applied to the sides of therodent’s prepuce. The ventral surface of the penis is examined for the location of the urethralopening. The presence of hypospadias is documented as follows: If the prepuce has not separatedand the location of the urethral opening is not apparent, the observation is documented as “hypospadiasnot determined” (this observation is continued daily until it is determined whether hypospadiasis present); if the urethral opening is located at the tip of the penis, the observation is recordedas “hypospadias absent”; or if the urethral opening is located ventrally anywhere along the shaft of thepenis, the observation is recorded as “hypospadias present.” Severity scores, if required, are assigned.If the penis can be extended and the urethral opening is located approximately one-half to two-thirdsthe distance from the base of the penis, then the severity is recorded as “slight.” If the penis cannot befully extended and the urethral opening is located approximately one-quarter to one-half the distance© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 477from the base of the penis, the severity is recorded as “moderate.” Moderate hypospadias may also beassociated with distended preputial glands and undescended testes. If the penis cannot be extended andthe urethral opening is located at the base of the penis, the severity is recorded as “extreme.” Extremehypospadias is usually associated with distended preputial glands, undescended testes, and visualpresence of the os penis as the unattached flaps of penile tissue spread laterally.2. Postmortem Procedure for Male RodentsThe male rodent is euthanized and then held in a supine position while gentle digital pressure isapplied to the sides of its prepuce. Forceps are used to extend the penis, and the ventral surface ofthe penis is examined for the location of the urethral opening. The presence of hypospadias andseverity scores are documented as described above.3. Postmortem Procedure for Female RodentsThe female rodent is euthanized, its abdomen is cut open, and a dye (e.g., gentian violet) is injectedinto its bladder. Gentle pressure is applied to the distended bladder, and the location of the urethralopening is determined by watching for exiting dye. The presence of hypospadias is documentedas follows: (1) if the dye exits from the urinary aperture adjacent to the vaginal opening, theobservation is documented as “hypospadias absent,” or (2) if the dye exits from the vagina, theobservation is documented as “hypospadias present.”VII. FUNCTIONAL OBSERVATION BATTERYNeurobehavioral screening methods, such as the FOB, have been routinely used to identify potentialneurotoxicity of new and existing chemicals, primarily in adult rats. The U.S. EPA publishedguidelines for neurotoxicity screening tests, including the FOB and motor activity, in 1985 176 andsubsequently revised them in 1991 177 and 1998. 178 These methods have been validated and largedatabases exist for effects of a wide range of chemicals. The FOB consists of approximately 20 to30 endpoints that were selected to evaluate aspects of nervous system function. Most of the datacollected is subjective. Therefore, the quality of the data depends primarily on the observer’s abilityto detect and describe changes in the animals’ behaviors. The former American Industrial HealthCouncil has developed a video and manual to train observers and to achieve consistency in theFOB across laboratories. 179Training of the technicians must be a continuous process, and variability between techniciansshould be minimized. The same observer should perform the FOB on all of the animals at a timepoint and, ideally, at all time points. When performing the FOB, the observer should be unawareof the animal’s treatment group. Stable environmental conditions (e.g., light level, temperature,background noise) must be maintained or data may be difficult to interpret. An assessment of thevarious tests used in the FOB is presented in Ross. 175 A detailed or extended clinical observation,a modified FOB (without grip strength, foot splay, or body temperature measurements), is currentlybeing used by many laboratories. FOB trained technicians have greater training in neurologicalassessment, and therefore, the quality of the information collected in the detailed or extendedclinical observations using FOB trained technicians is increased. 180A. Adult FOB1. In-Cage ObservationsThe adult rat is observed while it is in its home cage for the following characteristics: posture (e.g.,sitting, lying on one side, limbs spread), palpebral (eyelid) closure (e.g., open, closed, drooping),© 2006 by Taylor & Francis Group, LLC

478 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONwrithing, circling, biting, and gait and coordination abnormalities. The observations and/or scoresare recorded on the appropriate form.2. Removing from Cage ObservationsWhile the adult rat is being removed from its cage, it is observed for the following characteristics:ease of removal, ease of handling, vocalizations, muscle tone (e.g., limp, rigid), bite marks on tailand or paws, palpebral closure, piloerection, fur appearance (includes observation of skin, mucousmembranes, and eyes), lacrimation, salivation, and exophthalmos. The observations and/or scoresare recorded on the appropriate form.3. Standard Open Field Arena ObservationsThe adult rat is placed in the standard open field arena for approximately 2 min. The techniciansobserves and assesses the following characteristics: presence of righting reflex, posture, respirationease, rate of respiration, convulsions, tremors, muscle fasiculations, muscle spasm, grooming, gaitand coordination abnormalities (e.g., unbalanced, ataxic, unable to move), level of arousal (e.g.,stupor, excited), palpebral closure, diarrhea, polyuria, rearing (number of times both forepaws arelifted above cageboard), and vocalizations. The observations and/or scores are recorded on theappropriate form.4. Manipulations in the Standard Open Field ArenaWhile the adult rat is in the open field arena, the technician observes and assesses the followingcharacteristics: approach and touch response, auditory stimulus, and tail pinch. For the approachand touch response, the rat is rapidly approached toward the front of its head with a blunt probe(e.g., a finger) and then is touched between the eyes. The observer records the presence or absenceof the approach and touch response. For auditory stimulus, an appropriate stimulus (e.g., fingercricket or snap) is presented directly above the rat’s head and observer records the presence orabsence of the flinch reflex or other responses. For the tail pinch, the observer grasps the middleportion of the tail firmly between the thumb and index finger and pinches the tail. The observerrecords the presence or absence of a tail pinch response. Additional observations are noted whenpresent (e.g., sniffing, head weaving, Straub tail, syncope [loss of consciousness], stereotypicbehavior).5. Grip Strength ManipulationsThe technician selects the appropriate setting for the force gauge (for example, a Chatillon DigitalForce gauge) and tares it. The adult rat is grasped to provide support for the weight of the animal.The paws are brought into contact with the appropriate grip surface, ensuring that only forelimbpaws come into contact with the forelimb grip surface and only the hind limb paws come intocontact with the hind limb grip surface. The rat is allowed to grasp the surface with both pawsbeing tested. The technician then pulls the animal with a fluid motion away from the gauge untilthe paws release. The technician performs at least three forelimb and three hind limb trials for eachanimal. The gauge should be tared between trials, and the gauge readouts are documented on theappropriate form.6. Foot SplayThe technician applies ink to both heels of the hind paws of the adult rat. The rat is grasped bythe base of the tail with one hand and around the thorax with the other hand and held approximately© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 47932 cm above a sheet of paper. When the rat is still, the technician releases it, and it drops onto thepaper. The technician marks the position of the first footprints made and measures and records thedistance (in cm) between the centers of the backs of the heel prints.The rationale for the use of the width of the foot splay measurement, as an indicator of neurologicalfunction, is not clear. Increased foot splay width is considered an adverse neurological effect. Theanatomical basis for this reflex is unknown; see Ross 175 (p. 470) for a discussion of this parameter.7. Pupillary ResponseThis test evaluates the autonomic component of cranial nerve reflexes. Light from a penlight isshined into one eye of the adult rat while the other eye is shielded from the light. The normalresponse to the light is pupillary constriction. This procedure is repeated for the other eye, and theresponses are recorded on the appropriate form.8. Body TemperatureThe adult rat’s rectal temperature is determined with a clean, calibrated temperature probe. Alubricant (such as K-Y Jelly ® ) is placed on the tip of probe. The animal is removed from its cageand gently restrained to expose the anus, while also limiting its movement. The tip of the lubricatedprobe is gently inserted into the rectum until the bulb end is no longer visible (approximately 0.6to 0.95 cm). (Note: Handle only the cable and not the probe after insertion of the tip of the probeinto the rectum because touching the probe may alter temperature readings.) The temperature isrecorded when the digital display has stabilized (approximately 10 to 20 sec).B. FOB in Preweanling RatsThe Developmental Neurotoxicity Study (OPPTS 870.6300) requires the assessment of the centralnervous system of 10 male and 10 female offspring per dose group on PND 4, 11, 21, 35, 45, and60. The standard FOB is appropriate for rats 21 days or older, but some of the assessments areinappropriate for preweanling rats because they cannot be performed with young rats (PND 4 or11) or the data values obtained are too variable. The following points were recommended 181 to beexcluded from the pup evaluation: pupillary response, ataxia, foot splay, home cage activity, easeof removal from the home cage, approach response, touch response, body temperature, and gripstrength.1. In Cage ObservationsThe preweanling rat is observed while it is in its home cage for the following characteristics:posture, palpebral closure, writhing, circling, biting, and gait and coordination abnormalities.2. Removing from Cage ObservationsWhile the rat pup is being removing from its cage, it is observed for the following characteristics:ease of removal, ease of handling, vocalizations, bite marks on tail and or paws, palpebral closure,piloerection, fur appearance (includes observation of skin and mucous membranes), lacrimation,salivation, and exophthalmos.3. Standard Open Field Arena ObservationsThe preweanling rat is placed in the standard open field arena for approximately 2 min. Thetechnicians observes and assesses the following characteristics: surface righting reflex, posture,© 2006 by Taylor & Francis Group, LLC

480 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrespiration ease, rate of respiration, convulsions, tremors, grooming, gait and coordination abnormalities,arousal, palpebral closure, diarrhea, urination, and vocalizations.4. Manipulations in the Standard Open Field ArenaWhile the preweanling rat is in the open field arena, the technician observes and assesses theauditory stimulus response. An auditory stimulus (e.g., finger cricket or snap) is presented directlyabove the preweanling rat’s head and the observer records the presence or absence of the flinchreflex or other responses.References1. Wilson, J.G., Environment and Birth Defects, Academic Press, New York, 1973.2. U.S. Food and Drug Administration, Bureau of Foods. Guidelines for Reproduction Studies for SafetyEvaluation of Drugs for Human Use, Washington, DC, 1966.3. U.S. Food and Drug Administration, Bureau of Foods, Toxicological Principles for the Safety Assessmentof Direct Food Additives and Color Additives Used in Foods, Washington, DC, 1982.4. U.S. Food and Drug Administration, Center for Food Safety and Applied Nutrition, ToxicologicalPrinciples for the Safety Assessment of Direct Food Additives and Color Additives Used in Foods,“Redbook II” (draft), Washington, DC, 1993.5. U.S. Food and Drug Administration, International Conference on Harmonisation; Guideline on detectionof toxicity to reproduction for medicinal products, Fed. Regist., Part IX, 59 (183), 48746–48752, 1994.6. U.S. Food and Drug Administration, International Conference on Harmonisation; Guideline on detectionof toxicity to reproduction for medicinal products; Addendum on toxicity to male fertility, Fed.Regist., April 5, 1996, Vol. 61, No. 67, 1996.7. U.S. Food and Drug Administration, International Conference on Harmonisation; Maintenance ofthe ICH guideline on toxicity to male fertility: An addendum to the ICH tripartite guideline on detectionof toxicity to reproduction for medicinal products, Amended November 9, 2000.8. U.S. Environmental Protection Agency, Reproductive and Fertility Effects. Pesticide AssessmentGuidelines, Subdivision F. Hazard Evaluation: Human and Domestic Animals, Office of Pesticidesand Toxic Substances, Washington, DC, EPA 540/9-82-025, 1982.9. U.S. Environmental Protection Agency, Health Effects Test Guidelines OPPTS 870.3700, PrenatalDevelopmental Toxicity Study, Washington, DC, EPA 712-C-96-207, 1996, p. 1.10. U.S. Environmental Protection Agency, Health Effects Test Guidelines OPPTS 870.3800, Reproductionand Fertility Effects, Washington, DC, EPA 712-C-96-208, 1996, p. 1.11. U.S. Environmental Protection Agency, Guidelines for developmental toxicity risk assessment, Fed.Regist. 56(234): 63798–63826, 1991.12. Japan Ministry of Agriculture, Forestry and Fisheries, Guidance on Toxicology Study Data for Applicationof Agricultural Chemical Registration, 59 NohSan No. 4200, 1985, p. 45.13. Japan Ministry of Agriculture, Forestry and Fisheries, Guidelines for Screening Toxicity Testing ofChemicals, 59 NohSan No. 4200, 1985, p. 209.14. Canada, Health Protection Branch, The Testing of Chemicals for Carcinogenicity, Mutagenicity andTeratogenicity, Ministry of Health and Welfare, Canada, 1977.15. Drug Directorate Guidelines, Toxicological Evaluation 2.4, Reproductive Studies, Ministry of NationalHealth and Welfare, Health Protection Branch, Health and Welfare, Canada, 1990.16. Committee on the Safety of Medicines (CSM), Notes for Guidance on Reproduction Studies, Departmentof Health and Social Security, Great Britain, 1974.17. World Health Organization (WHO), Principles for the testing of drugs for teratogenicity, WHO Tech.Rept. Series, No. 364, Geneva, 1967.18. Organization for Economic Cooperation and Development, OECD 415: One-Generation ReproductionToxicity Study (Original Guideline, adopted 26th May 1983), OECD Paris, 1983.19. Organization for Economic Cooperation and Development, OECD 421: Reproduction/DevelopmentalToxicity Screening Test (Original Guideline, adopted 27th July 1995), OECD Paris, 1995.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 48120. Organization for Economic Cooperation and Development, OECD 422: Combined Repeated DoseToxicity Study with the Reproduction/Developmental Toxicity Screening Test (Original Guideline,adopted 22nd March 1996), OECD Paris, 1996.21. Organization for Economic Cooperation and Development, OECD 414: Prenatal DevelopmentalToxicity Study (Updated Guideline, adopted 22nd January 2001), OECD Paris, 2001.22. Organization for Economic Cooperation and Development. OECD 416: Two-Generation ReproductionToxicity Study (Updated Guideline, adopted 22nd January 2001), OECD Paris, 2001.23. International Conference on Harmonization of Technical Requirements of Pharmaceuticals for HumanUse, Detection of toxicity to reproduction of medicinal products, Fed. Regist., 59(1831), 48746–48752,1994.24. U.S. Food and Drug Administration, Good Laboratory Practice Regulations; Final Rule, 21 CFR Part58, September 4, 1987.25. U.S. Environmental Protection Agency, Toxic Substance Control Act (TSCA), Good Laboratory PracticeStandards, Final Rule, 40 CFR Part 792, August 17, 1989.26. U.S. Environmental Protection Agency. Federal Insecticide, Fungicide and Rodenticide Act (FIFRA),Good Laboratory Practice Standards, Final Rule, 40 CFR Part 160, August 17, 1989.27. U.S. Department of Health and Human Services, Biosafety in Microbiological and Biomedical Laboratories,4th ed., Richmond, J.Y. and McKinney, R.W., Eds., Public Health Service Centers forDisease Control and Prevention, and National Institutes of Health, Washington, D.C., 1999.28. Institute for Laboratory Animal Research, National Research Council, Guide for the Care and Useof Laboratory Animals, Washington, DC, 1996.29. National Institute of Mental Health, Methods and Welfare Considerations in Behavioral Research withAnimals: Report of the National Institute of Mental Health Workshop, Morrison, A.R., Evans, H.L.,Ator, N.A. and Nakamura, R.K., Eds., NIH Publication No. 02-5083, U.S. Government Printing Office,Washington, DC, 2002.30. Tyl, R.W. and Marr, M.C., Developmental toxicity testing — methodology, in Handbook of DevelopmentalToxicology, Hood R.D., Ed., CRC Press, Boca Raton, FL, 1997, p. 175.31. Gulati, D.K., Hope, E., Teague, J., and Chapin, R.E., Reproductive toxicity assessment by continuousbreeding in Sprague-Dawley rats: a comparison of two study designs, Fundam. Appl. Toxicol., 17,270, 1991.32. Chapin, R.E. and Sloane, R.A., Reproductive assessment by continuous breeding: evolving studydesign and summaries of ninety studies, Environ. Health Perspect., 105(Suppl 1), 199, 1977.33. U.S. Environmental Protection Agency, Health Effects Test Guidelines; OPPTS 870.6300, DevelopmentalNeurotoxicity Study, Office of Prevention, Pesticides and Toxic Substances, August 1998.34. Organization for Economic Cooperation and Development, OECD 426: Developmental NeurotoxicityStudy (Revised Draft Guideline, October 1999), OECD Paris, 1999.35. U.S. Environmental Protection Agency, Health Effects Test Guidelines; OPPTS 870.3650, CombinedRepeated Dose Toxicity Study with the Reproduction/Developmental Toxicity Screening Test, Officeof Prevention, Pesticides and Toxic Substances, July 2000.36. Organization for Economic Cooperation and Development. OECD 478: Genetic Toxicology: RodentDominant Lethal Test (Updated Guideline, adopted 4 April 1984), OECD Paris, 1984.37. U.S. Environmental Protection Agency. Health Effects Test Guidelines; OPPTS 870.5450, RodentDominant Lethal Assay, Office of Prevention, Pesticides and Toxic Substances, August 1998.38. Risk Assessment Forum, U.S. Environmental Protection Agency. Guidelines for reproductive toxicityrisk assessment, Fed. Regist., 61(212), 56274–56322, October 31, 1996.39. Davis, D.L., Friedler, G., Mattison, D., and Morris, R., Male-mediated teratogenesis and other reproductiveeffects: biologic and epidemiologic findings and a plea for clinical research, Reprod. Toxicol.,6, 289, 1992.40. Colie, C.F., Male mediated teratogenesis, Reprod. Toxicol., 7, 3, 1993.41. Savitz, D.A., Sonnenfeld, N.L., and Olshan, A.F., Review of epidemiologic studies of paternal occupationalexposure and spontaneous abortion, Am. J. Ind. Med., 25, 361, 1994.42. Trasler, J.M. and Doerksen, T., Teratogen update: paternal exposures — reproductive risks, Teratology,60, 161, 1999.43. Knobil, E., Neill, J.D., Greenwald, G.S., Markert, C.L., and Pfaff, D.W., The Physiology of Reproduction,Raven Press, New York, 1994.© 2006 by Taylor & Francis Group, LLC

482 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION44. Hood, R.D. and Miller, D.B., Maternally mediated effects on development, in Handbook of DevelopmentalToxicology, Hood R.D., Ed., CRC Press, Boca Raton, FL, 1997, p. 61.45. Le Vier, R.R. and Jankowiak, M.E. The hormonal and antifertility activity of 2,6-cis-diphenylhexamethylcyclotetrasiloxanein the female rat, Biol. Reprod., 7, 260, 1972.46. Sonawane, B.R. and Yaffe, S.J., Delayed effects of drug exposure during pregnancy: reproductivefunction, Biol. Res. Pregnancy, 4, 48, 1983.47. Cummings, A.M. and Gray, L.E., Methoxychlor affects the decidual cell response of the uterus butnot other progestational parameters in female rats, Toxicol. Appl. Pharmacol., 90, 330, 1987.48. Ulbrich, B. and Palmer, A.K., Detection of effects on male reproduction — a literature survey, J. Am.Coll. Toxicol.,14, 293, 1995.49. Takayama, S., Akaike, M., Kawashima, K., Takahashi, M., and Kurokawa, Y., Collaborative study inJapan on optimal treatment period and parameters for detection of male fertility disorders in ratsinduced by medicinal drugs, J. Am. Coll. Toxicol.,14, 66, 1995.50. Sakai, T., Takahashi, M., Mitsumori, K., Yasuhara, K., Kawashima, K., Mayahare, H., and Ohno, Y.,Collaborative work to evaluate toxicity on male reproductive organs by 2-week repeated dose toxicitystudies in rats. Overview of the studies, J. Toxicol. Sci., 25, 1, 2000.51. Cooper, R.L., Goldman, J.M., and Stoker, T.E., Measuring sexual behavior in the female rat, in Methodsin Toxicology, Volume 3, Part B. Female Reproductive Toxicology, Heindel, J.J. and Chapin, R.E.,Eds., Academic Press, San Diego, CA, 1993, p. 42.52. Adler, N.T. and Toner, J.P., The effect of copulatory behavior on sperm transport and fertility in rats,in Reproduction: Behavioral and Neuroendocrine Perspective, Komisaruk, B.R., Siegel, H.I., Chang,M.F., and Feder, H.H., Eds., New York Academy of Science, New York, 1986, p. 21.53. MARTA (Middle Atlantic Reproduction and Teratology Association) and MTA (Midwest TeratologyAssociation), Historical Control Data (1992–1994) for Developmental and Reproductive ToxicityStudies using the Crl:CD ® (SD)BR Rat, Charles Rivers Laboratories, March 1996. (Available on-lineat www.crivers.com.)54. Clark, R.L., Endpoints of reproductive system development, in An Evaluation and Interpretation ofReproductive Endpoints for Human Risk Assessment, Daston, G. and Kimmel, C., Eds., ILSI Press,Washington, DC, 1999, p. 28.55. Chapin, R.E., Gulati, D.K., Barnes, L.H., and Teague, J.L., The effects of feed restriction on reproductivefunction in Sprague-Dawley rats, Fundam. Appl. Toxicol., 20, 23, 1993.56. Stevens, K.R. and Gallo, M.A., Practical considerations in the conduct of chronic toxicity studies, inPrinciples and Methods of Toxicology, Hayes, A.W., Ed., Raven Press, New York, 1989, p. 237.57. Schwetz, B.A., Rao, K.S., and Park, C.N., Insensitivity of tests for reproductive problems, J. Environ.Pathol. Toxicol., 3, 81, 1980.58. Blazak, W.F., Ernst, T.L., and Stewart, B.E., Potential indicators of reproductive toxicity, testicularsperm production and epididymal sperm number, transit time and motility in Fischer 344 rats, Fundam.Appl. Toxicol., 5, 1097, 1985.59. Berndtson, W.E., Methods for quantifying mammalian spermatogenesis: a review, J. Anim. Sci., 44,818, 1977.60. Foote, R.H. and Berndtson, W.E., The germinal cells, in Reversibility in Testicular Toxicity Assessment,Scialli, A.R. and Clegg, E.D., Eds., CRC Press, Boca Raton, FL, 1992, p. 1.61. Ku, W.W., Chapin, R.E., Wine, R.N., and Gladen, B.C., Testicular toxicity of boric acid (BA):relationship of dose to lesion development and recovery in the F344 rat, Reprod. Toxicol., 7, 305, 1993.62. Hess, R.A., Moore, B.J., Forrer, J., Linder, R.E., and Abuel-Atta, A.A., The fungicide Benomyl (methyl1-(butylcarbamoyl)-2-benzimidazolecarbamate) causes testicular dysfunction by inducing the sloughingof germ cells and occlusion of efferent ductules, Fundam. Appl. Toxicol., 17, 733, 1991.63. Nakai, M., Moore, B.J., and Hess, R.A., Epithelial reorganization and irregular growth followingcarbendazim induced injury of the efferent ductules of the rat testis, Anat. Rec., 235, 51, 1993.64. Boorman, G.A., Chapin, R.E., and Mitsumori, K., Testis and epididymis, in Pathology of the FischerRat, Boorman, G.A., Eustis, S.L., Elwell, M.R., Montgomery, C.A., Jr., and MacKenzie, W.F., Eds.,Academic Press, San Diego, 1990, p. 405.65. Russell, L.D., Normal testicular structure and methods of evaluation under experimental and disruptiveconditions, in Reproductive and Developmental Toxicity of Metals, Clarkson, T.W., Nordberg, G.F.,and Sager, P.R., Eds., Plenum Publishing Co., New York, 1983, p. 227.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 48366. De Kretser, D.M. and Kerr, J.B., The cytology of the testis, in The Physiology of Reproduction, Knobil,E. and Neill, J., Eds., Raven Press, New York, 1988, p. 837.67. Johnson, A.D. and Gomes, W.R., The testis, in Advances in Physiology, Biochemistry and Function,Vol. IV, Academic Press, New York, 1977.68. Russell, L.D., Ettlin, R., Sinha-Hikim, A.P., Clegg, E.D., Eds., Histological and HistopathologicalEvaluation of the Testis, Cache River Press, Clearwater, FL, 1990.69. Scialli, A.R. and Clegg, E.D., Reversibility in Testicular Toxicity Assessment, CRC Press, Boca Raton,FL, 1992.70. Chapin, R.E. and Conner, M.W., Testicular histology and sperm parameters, in An Evaluation andInterpretation of Reproductive Endpoints for Human Risk Assessment, Daston, G. and Kimmel, C.,Eds., ILSI Press, Washington, DC, 1999, p. 28.71. Chapin, R.E., Morphologic evaluation of seminiferous epithelium of the testis, in Physiology andToxicology of Male Reproduction, Lamb, J.C. and Foster, P.M.D., Eds., Academic Press, New York,1988, p. 155.72. Hess, R.A. and Moore, B.J., Histological methods for evaluation of the testis, in Methods in Toxicology:Male Reproductive Toxicology, Chapin, R.E. and Heindel, J.J., Eds., Academic Press, San Diego,1993, p. 52.73. Mangelsdorf, I. and Buschman, J., Extrapolation from Animal Studies to Humans for the Endpoint ofMale Fertility, Federal Institute for Occupational Safety and Health, Dortman/Berlin/Dresden, Germany,2002.74. Hess, R.A., Quantitative and qualitative characteristics of the stages and transitions in the cycle ofthe rat seminiferous epithelium: light microscopic observations of perfusion-fixed and plastic-embeddedtestes, Biol. Reprod., 43, 525, 1990.75. Amann, R.P., A critical review of methods for evaluation of spermatogenesis from seminal characteristics,J. Androl., 2, 37, 1981.76. Working, P.K., Male reproductive toxicity: comparison of the human to animal models, Environ.Health, 77, 37, 1988.77. Aafjes, J.H., Vels, J.M., and Schenck, E., Fertility of rats with artificial oligozoospermia, J. Reprod.Fertil., 58, 345, 1980.78. Meistrich, M.L., Quantitative correlation between testicular stem cell survival, sperm production, andfertility in the mouse after treatment with different cytotoxic agents, J. Androl. 3, 58, 1982.79. Robaire, B., Smith, S., and Hales, B.F., Suppression of spermatogenesis by testosterone in adult malerats: effect on fertility, pregnancy outcome and progeny, Biol. Reprod., 31, 221, 1984.80. Fawcett, D.W., Bloom and Fawcett: A Textbook of Histology, W.B. Saunders, Philadelphia, PA, 1986.81. Boorman, G.A., Elwell, M.R., and Mitsumori, K., Male accessory sex glands, penis and scrotum, inPathology of the Fischer Rat, Boorman, G.A., Eustis, S.L., Elwell, M.R., Montgomery, C.A., Jr., andMacKenzie, W.F., Eds., Academic Press, San Diego, 1990, p. 419.82. Jones, T.C., Mohr, U., and Hunt, R.D., Genital System, Springer-Verlag, New York, 1987.83. Haschek, W.M. and Rousseaux, C.G., Handbook of Toxicologic Pathology, Academic Press, NewYork, 1991.84. Toth, G.P., Stober, J.A., Read, E.J., Zenick, H., and Smith, M.K, The automated analysis of rat spermmotility following subchronic epichlorohydrin administration: methodologic and statistical considerations,J. Androl.,10, 401, 1989.85. Slott, V.L., Suarez, J.D., Simmons, J.E., and Perreault, S.D., Acute inhalation exposure to epichlorohydrintransiently decreases rat sperm velocity, Fundam. Appl. Toxicol., 15, 597, 1990.86. Klinefelter, G.R., Laskey, J.W., Kelce, W.R., Ferrell, J., Roberts, N.L., Suarez, J.D., and Slott, V.,Chloroethylmethanesulfonate-induced effects on the epididymis seem unrelated to altered Leydig cellfunction, Biol. Reprod., 51, 82, 1994.87. Toth, G.P., Stober, J.A., George, E.L., Read, E.J., and Smith, M.K., Sources of variation in thecomputer-assisted motion analysis of rat epididymal sperm, Reprod. Toxicol., 5, 487, 1991.88. Oberlander, G., Yeung, C.H., and Cooper, T.G., Induction of reversible infertility in male rats by oralornidazole and its effects on sperm motility and epididymal secretions, J. Reprod. Fertil., 100, 551, 1994.89. Slott, V.L., Jeffay, S.C., Suarez, J.D., Barbee, R.R., and Perreault, S.D., Synchronous assessment ofsperm motility and fertilizing ability in the hamster following treatment with alpha-chlorohydrin, J.Androl., 16, 523, 1995.© 2006 by Taylor & Francis Group, LLC

484 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION90. Seed, J., Chapin, R.E., Clegg, E.D., Darney, S.P., Dostal, L., Foote, R.H., Hurtt, M.E., Klinefelter,G.R., Makris, S.L., Schrader, S., Seyler, D., Sprando, R., Treinen, K.A., Veeranachaneni, R., and Wise,L.D., Methods for assessing sperm motility, morphology, and counts in the rat, rabbit and dog: aconsensus report, Reprod. Toxicol., 10, 237,1996.91. Klinefelter, G.R., Laskey, J.W., Perreault, S.D., Ferrell, J., Jeffay, S., Suarez, J., and Roberts, N., Theethane dimethanesulfonate-induced decrease in the fertilizing ability of cauda epididymal sperm isindependent of the testis, J. Androl. 15, 318, 1994.92. Meistrich, M.L. and Brown, C.C., Estimation of the increased risk of human infertility from alterationsin semen characteristics, Fertil. Steril., 40, 220, 1983.93. World Health Organization, WHO Laboratory Manual for the Examination of Human Semen andSperm-Cervical Mucus Interaction, 3rd Ed., Cambridge University Press, Cambridge, 1992.94. Hurtt, M.E. and Zenick, H., Decreasing epididymal sperm reserves enhances the detection of ethoxyethanolinduced spermatotoxicity, Fundam. Appl. Toxicol., 7, 348, 1986.95. Cassidy, S.L., Dix, K.M., and Jenkins, T., Evaluation of a testicular sperm head counting technique usingrats exposed to dimethoxyethyl phthalate (DMEP), glycerol alpha-monochlorohydrin (GMCH), epichlorohydrin(ECH), formaldehyde (FA), or methyl methanesulphonate (MMS), Arch. Toxicol. 53, 71, 1983.96. Blazak, W.F., Treinen, K.A., and Juniewicz, P.E., Application of testicular sperm head counts in theassessment of male reproductive toxicity, in Methods in Toxicology: Male Reproductive Toxicology,Chapin, R.E. and Heindel, J.J., Eds., Academic Press, San Diego, 1993, p. 86.97. Meistrich, M.L. and van Beek, M.E.A.B., Spermatogonial stem cells: assessing their survival andability to produce differentiated cells, in Methods in Toxicology: Male Reproductive Toxicology,Chapin, R.E. and Heindel, J.J., Eds., Academic Press, San Diego, 1993, p. 106.98. Boyers, S.P., Davis, R.O., and Katz, D.F., Automated semen analysis, Curr. Probl. Obstet. Gynecol.Fertil., 12, 173, 1989.99. Yeung, C.H., Oberlander, G., and Cooper, T.G., Characterization of the motility of maturing ratspermatozoa by computer-aided objective measurement, J. Reprod. Fertil., 96, 427, 1992.100. Slott, V.L. and Perreault, S.D., Computer-assisted sperm analysis of rodent epididymal sperm motilityusing the Hamilton-Thorne motility analyzer, in Methods in Toxicology: Male Reproductive Toxicology,Chapin, R.E. and Heindel, J.J., Eds., Academic Press, San Diego, 1993, p. 319.101. Morrissey, R.E., Schwetz, B.A., Lamb, J.C., Ross, M.D., Teague, J.L., and Morris, R.W., Evaluationof rodent sperm, vaginal cytology, and reproductive organ weight data from National ToxicologyProgram 13-week studies, Fundam. Appl. Toxicol., 11, 343, 1988.102. Toth, G.P., Stober, J.A., Zenick, H., Read, E.J., Christ, S.A., and Smith, M.K., Correlation of spermmotion parameters with fertility in rats treated subchronically with epichlorohydrin, J. Androl., 12,54, 1991.103. Morrissey, R.E., Lamb, J.C., Morris, R.W., Chapin, R.E., Gulati, D.K., and Heindel, J.J., Results andevaluations of 48 continuous breeding reproduction studies conducted in mice, Fundam. Appl. Toxicol.,13, 747, 1989.104. Slott, V.L., Suarez, J.D., and Perreault, S.D., Rat sperm motility analysis: methodologic considerations,Reprod. Toxicol., 5, 449, 1991.105. Chapin, R.E., Filler, R.S., Gulati, D., Heindel, J.J., Katz, D.F., Mebus, C.A., Obasaju, F., Perreault,S.D., Russell, S.R., Schrader, S., Slott, V., Sokol, R.Z., and Toth, G., Methods for assessing rat spermmotility, Reprod. Toxicol., 6, 267, 1992.106. Schrader, S.M., Chapin, R.E., Clegg, E.D., Davis, R.O., Fourcroy, J.L., Katz, D.F., Rothmann, S.A.,Toth, G., Turner, T.W., and Zinaman, M., Laboratory methods for assessing human semen in epidemiologicstudies: a consensus report, Reprod. Toxicol., 6, 275, 1992.107. Toth, G.P., Read, E.J., and Smith, M.K., Utilizing Cryo Resources CellSoft computer-assisted spermanalysis system for rat sperm motility studies, in Methods in Toxicology: Male Reproductive Toxicology,Chapin, R.E. and Heindel, J.J., Eds., Academic Press, San Diego, 1993, p. 303.108. Zenick, H., Clegg, E.D., Perreault, S.D., Klinefelter, G.R., and Gray, L.E., Assessment of malereproductive toxicity: a risk assessment approach, in Principles and Methods of Toxicology, Hayes,A.W., Ed., Raven Press, New York, 1994, p. 937.109. Filler, R., Methods for evaluation of rat epididymal sperm morphology, in Methods in Toxicology:Male Reproductive Toxicology, Chapin, R.E. and Heindel, J.J., Eds., Academic Press, San Diego,1993, p. 334.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 485110. Linder, R.E., Strader, L.F., Slott, V.L., and Suarez, J.D., Endpoints of spermatotoxicity in the rat aftershort duration exposures to fourteen reproductive toxicants, Reprod. Toxicol., 6, 491, 1992.111. Davis, R.O., Gravance, C.G., Thal, D.M., and Miller, M.G., Automated analysis of toxicant-inducedchanges in rat sperm head morphometry, Reprod. Toxicol., 8(6), 521, 1994.112. UK Industrial Reproductive Toxicology Discussion Group (IRDG), Computer-Assisted Sperm Analysis(CASA) Group, Rat Sperm Morphological Assessment, Guideline Document, 1st ed., October2000. (http://www.irdg.co.uk/IRDG_spermmorphology.doc)113. Clermont, Y. and Perey, B., The stages of the cycle of the seminiferous epithelium of the rat: practicaldefinitions in PA-Schiff-hematoxylin-eosin stained sections, Rev. Canad. Biol., 16:451, 1957.114. Leblond, C.P. and Clermont, Y., Definition of the stages of the cycle of the seminiferous epitheliumin the rat, Ann. N.Y. Acad. Sci. 55:548, 1952.115. Russell, L.D., Ettlin, R., Sinha Hikim, A.P., and Clegg, E.D., The classification and timing of spermatogenesis,in Histological and Histopathological Evaluation of the Testis, Russell, L.D., Ettlin, R.,Sinha Hikim, A.P., and Clegg, E.D., Eds., Cache River Press, Clearwater, FL, 1990, chap. 2.116. Keller, K.A., Developmental and reproductive toxicology, in Toxicology Testing Handbook, Principles,Applications, and Data Interpretation, Jacobson-Kram, D. and Keller, K.A., Eds., Marcel Dekker,Inc., New York, 2001, p. 209.117. Wade, G.N., Gonadal hormones and behavioral regulation of body weight, Physiol. Behav., 8, 523,1972.118. Galletti, F. and Klopper, A., The effect of progesterone on the quantity and distribution of body fatin the female rat, Acta Endocrinol., 46, 379, 1964.119. Hervey, E. and Hervey, G.R., The effects of progesterone on body weight and composition in the rat,J. Endocrinol., 37, 361, 1967.120. Kupfer, D., Critical evaluation of methods for detection and assessment of estrogenic compounds inmammals: strengths and limitations for application to risk assessment, Reprod. Toxicol., 2, 147, 1987.121. Leininger, J.R. and Jokinen, M.P., Oviduct, uterus and vagina, in Pathology of the Fischer Rat,Boorman, G.A., Eustis, S.L., Elwell, M.R., Montgomery, C.A., Jr., and MacKenzie, W.F., Eds.,Academic Press, San Diego, 1990, p. 443.122. Cooper, R.L., Chadwick, R.W., Rehnberg, G.L., Goldman, J.M., Booth, K.C., Hein, J.F., and McElroy,W.K., Effect of lindane on hormonal control of reproductive function in the female rat, Toxicol. Appl.Pharmacol., 99, 384, 1989.123. Hirschfield, A.N., Histological assessment of follicular development and its applicability to riskassessment, Reprod. Toxicol., 1, 71, 1987.124. Baker, T., Oogenesis and ovarian development, in Reproductive Biology, Balin, H. and Glasser, S.,Eds., Exerpta Medica, Amsterdam, 1972.125. Alison, R.H., Morgan, K.T., and Montgomery, C.A., Jr., Ovary, in Pathology of the Fischer Rat,Boorman, G.A., Eustis, S.L., Elwell, M.R., Montgomery, Jr., C.A., and MacKenzie, W.F., Eds.,Academic Press, San Diego, 1990, p. 429.126. Eppig, J., Mammalian oocyte development in vivo and in vitro, in Elements of Mammalian Fertilization,Vol. 1, Wassarman, P.M., Ed., CRC Press, Boca Raton, FL, 1991, p. 57.127. Wassarman, P.M., The mammalian ovum, in Physiology of Reproduction, Vol. 1, Knobil, E. and Neill,J.D., Eds., Raven Press, New York, 1988, p. 69.128. Crisp, T.M., Organization of the ovarian follicle and events in its biology: oogenesis, ovulation oratresia, Mutat. Res., 296, 89, 1992.129. Eppig, J.J., Oocyte control of ovarian follicular development and function in mammals, Reproduction,122, 829, 2001.130. Peters, H. and McNatty, K.P., The Ovary, Paul Elek, New York, 1980.131. Plowchalk, D.R., Smith, B.J., and Mattison, D.R., Assessment of toxicity to the ovary using folliclequantitation and morphometrics, in Methods in Toxicology: Female Reproductive Toxicology, Heindel,J.J. and Chapin, R.E., Eds., Academic Press, San Diego, 1993, p. 57.132. Kurman, R. and Norris, H.J., Germ cell tumors of the ovary, Pathol. Annu., 13, 291, 1978.133. Kwa, S.L. and Fine, L.J., The association between parental occupation and childhood malignancy, J.Occup. Med., 22, 792, 1980.134. Langley, F.A. and Fox, H., Ovarian tumors: classification, histogenesis, etiology, in Haines and Taylor’sObstetrical and Gynaecologic Pathology, Fox, H., Ed., Churchill Livingstone, Edinburgh, 1987, p. 542.© 2006 by Taylor & Francis Group, LLC

486 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION135. Fugo, N.W. and Butcher, R.L., Overripeness and the mammalian ova. I. Overripeness and earlyembryonic development, Fertil. Steril., 17, 804, 1966.136. Butcher, R.L. and Fugo, N.W., Overripeness and the mammalian ova. II. Delayed ovulation andchromosome anomalies, Fertil. Steril., 18, 297, 1967.137. Butcher, R.L., Blue, J.D., and Fugo, N.W., Overripeness and the mammalian ova. III. Fetal developmentat midgestation and at term, Fertil. Steril., 20, 223, 1969.138. Gray, L.E. and Ostby, J.S., In utero 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters reproductivemorphology and function in female rat offspring, Toxicol. Appl. Pharmacol., 133, 285, 1995.139. Long, J.A. and Evans, H.M., The oestrous cycle in the rat and its associated phenomena, Mem. Univ.Calif., 6, 1, 1922.140. Cooper, R.L., Goldman, J.M., and Vandenbergh, J.G., Monitoring of the estrous cycle in the laboratoryrodent by vaginal lavage, in Methods in Toxicology: Female Reproductive Toxicology, Heindel, J.J.and Chapin, R.E., Eds., Academic Press, San Diego, 1993, p. 45.141. Kimmel, G.L., Clegg, E.D., and Crisp, T.M. Reproductive toxicity testing: a risk assessment perspective,in Reproductive Toxicology, Witorsch, R.J., Ed., Raven Press, New York, 1995, p. 75.142. LeFevre, J. and McClintock, M.K., Reproductive senescence in female rats: a longitudinal study ofindividual differences in estrous cycles and behavior, Biol. Reprod., 38, 780, 1988.143. Holmes, R.L. and Ball, J.N., The Pituitary Gland: A Comparative Account, Cambridge UniversityPress, Cambridge, 1974.144. Boorman, G.A., Wilson, J.T., van Zwieten, M.J., and Eustis, S.L., Mammary gland, in Pathology ofthe Fischer Rat, Boorman, G.A., Eustis, S.L., Elwell, M.R., Montgomery, C.A., Jr., and MacKenzie,W.F., Eds., Academic Press, San Diego, 1990, p. 295.145. Wolff, M.S., Lactation, in Occupational and Environmental Reproductive Hazards, Paul, M., Ed.,Williams and Wilkins, Baltimore, 1993, p. 60.146. Imagawa, W., Yang, J., Guzman, R., and Nandi, S., Control of mammary gland development, in ThePhysiology of Reproduction, Knobil, E. and O’Neill, J.D., Eds., Raven Press, New York, 1994, p. 1033.147. Tucker, H.A., Lactation and its hormonal control, in The Physiology of Reproduction, Knobil, E. andO’Neill, J.D., Eds., Raven Press, New York, 1994, p. 1065.148. American Academy of Pediatrics Committee on Drugs, The transfer of drugs and other chemicalsinto human milk, Pediatrics, 93, 137, 1994.149. Oskarsson, A., Hallen, I.P., and Sundberg, J., Exposure to toxic elements via breast milk, Analyst,120, 765, 1995.150. Cooper, R.L., Conn, P.M., and Walker, R.F., Characterization of the LH surge in middle-aged femalerats, Biol. Reprod., 23, 611, 1980.151. Finch, C.E., Felicio, L.S., and Mobbs, C.V., Ovarian and steroidal influences on neuroendocrine agingprocesses in female rodents, Endocrinol. Rev., 5, 467, 1984.152. Brawer, J.R. and Finch, C.E., Normal and experimentally altered aging processes in the rodenthypothalamus and pituitary, in Experimental and Clinical Interventions in Aging, Walker, R.F. andCooper, R.L., Eds., Marcel Dekker, New York, 1983, p. 45.153. Mattison, D.R., Clinical manifestations of ovarian toxicity, in Reproductive Toxicology, Dixon, R.L.,Ed., Raven Press, New York, 1985, p. 109.154. Cooper, R.L. and Goldman, J.M., Vaginal cytology, in An Evaluation and Interpretation of ReproductiveEndpoints for Human Risk Assessment, Daston, G. and Kimmel, C., Eds., ILSI Press, Washington,DC, 1999, p. 42.155. Nelson, S. and Gibori, G., Dispersion, separation, and culture of the different cell populations of therat corpus luteum, in Methods in Toxicology: Female Reproductive Toxicology, Heindel, J.J. andChapin, R.E., Eds., Academic Press, San Diego, 1993, p. 340.156. Salewski, E., Farbemethode zum makroskopischen Nachweis von Implantationsstellan am Uterus derRatte. Naunyn Schmiedebergs, Arch. Exp. Pathol. Pharmakol., 247, 367, 1964.157. Heindel, J.J., Oocyte quantitation and ovarian histology, in An Evaluation and Interpretation ofReproductive Endpoints for Human Risk Assessment, Daston, G. and Kimmel, C., Eds., ILSI Press,Washington, DC, 1999, p. 57.158. Pedersen, T. and Peters, H.J., Proposal for the classification of oocytes and follicles in the mouseovary, Reprod. Fert., 17, 555, 1968.© 2006 by Taylor & Francis Group, LLC

TESTING FOR REPRODUCTIVE TOXICITY 487159. Bolon, B., Bucci, T.J., Warbritton, A.R., Chen, J.J., Mattison, D.R., and Heindel, J.J., Differentialfollicle counts as a screen for chemically induced ovarian toxicity in mice: results from continuousbreeding bioassays, Fundam. Appl.Toxicol., 39, 1, 1997.160. Bucci, T.J., Bolon, B., Warbritton, A.R., Chen, J.J., and Heindel, J.J., Influence of sampling on thereproducibility of ovarian follicle counts in mouse toxicity studies, Reprod. Toxicol., 11(5), 689, 1997.161. Heindel, J.J., Thomford, P.J., and Mattison, D.R., Histological assessment of ovarian follicle numberin mice as a screen of ovarian toxicity, in Growth Factors and the Ovary, Hirshfield, A.N., Ed., Plenum,New York, 1989, p. 421.162. Christian, M.S. and Brown, W.R., Control primordial follicle counts in multigeneration studies inSprague-Dawley (“gold standard”) rats, The Toxicologist, 66(1-S), March 2002.163. Tyl, R.W., Developmental toxicity in toxicologic research and testing, in Perspectives in Basic andApplied Toxicology, Ballantyne, B., Ed., John Wright, Bristol, 1987, p. 203.164. Bates, H.K., Cunny, H.C., and Kebede, G.A., Developmental neurotoxicity testing methodology, inHandbook of Developmental Toxicology, Hood, R.D., Ed., CRC Press, Boca Raton, FL, 1997, p. 291.165. Henck, J.W., Developmental neurotoxicology: testing and interpretation, in Handbook of Neurotoxicology,Volume II, Massaro, E.J., Ed., Humana Press, Totowa, NJ, 2002, p. 461.166. Iezhitsa, I.N., Spasov, A.A., and Bugaeva, L.I., Effects of bromantan on offspring maturation anddevelopment of reflexes, Neurol. Toxiclo., 23, 213, 2001.167. Kai, S. and Kohmura, H., Comparison of developmental indices between Crj:CD(SD)IGS rats andCrj:CD(SD) rats, in Biological Reference Data on CD(SD)IGS-1999, Matsuzawa, T. and Inoue, H.,Eds., CD(SD)IGS Study Group, Yokohama, 1999, p. 157.168. Ohkubo, Y., Furuya, K., Yamada, M., Akamatsu, H., Moritomo, A., Ishida, S., Watanabe, K., andHayashi, Y., Reproductive and developmental data on Crj:CD(SD)IGS rats, in Biological ReferenceData on CD(SD)IGS-1999, Matsuzawa, T. and Inoue, H., Eds., CD(SD)IGS Study Group, Yokohama,1999, p. 153.169. Ema, M., Fujii, S., Furukawa, M., Kiguchi, M., Ikka, T., and Harazano, A., Rat two-generationreproductive toxicity study of bisphenol A, Repro. Toxicol., 15, 505, 2001.170. Lewis, E.M., Christian, M.S., Barnett, J., Jr., and Hoberman, A.M., Control preputial separation datafor F1 generation CRL Sprague-Dawley (“gold standard”) rats in EPA developmental neurotoxicology,EPA multigeneration and FDA peri-postnatal studies, The Toxicologist, 66(1), 1149, 2002.171. Hoberman, A.M., Christian, M.S., Lewis, E.M., and Barnett, J., Jr., Control vaginal opening data forF1 generation CRL Sprague-Dawley (“gold standard”) rats in EPA developmental neurotoxicology,EPA multigeneration and FDA peri-postnatal studies, The Toxicologist, 66(1), 1137, 2002.172. Tinwell, H., Haseman, J., Lefevre, P.A., Wallis, N., and Ashby, J., Normal sexual development of twostrains of rat exposed in utero to low doses of Bisphenol A, Toxicol. Sci., 68, 339, 2002.173. Weisenburger, W.P., Neurotoxicology, in Toxicology Testing Handbook, Jacobson-Kram, D. and Keller,K.A., Eds., Marcel Dekker, New York, 2002, p. 255.174. Vorhees, C.V., Acuff-Smith, K.D., Moran, M.S., and Minck, D.R., A new method for evaluating airrightingreflex ontogeny in rats using prenatal exposure to phenytoin to demonstrate delayed development,Neurotoxicol. Teratol., 16(6), 563, 1994.175. Ross, J.F., Tier 1 neurological assessment in regulated animal safety studies, in Handbook of Neurotoxicology,Volume II, Massaro, E.J., Ed., Humana Press, Totowa, NJ, 2002, p. 461.176. U.S. Environmental Protection Agency, Toxic Substances Control Act Testing Guidelines, 40 CFRpart 798 subpart G section 798.6050, Fed. Regist., 50, 39458–39460, 1985.177. U.S. Environmental Protection Agency, Neurotoxicity Guidelines, Addendum 10, Pesticide AssessmentsGuidelines Subdivision F, Publication number PB 91-154617, National Technical InformationServices, Springfield, VA, 1991.178. U.S. Environmental Protection Agency, Health Effects Test Guidelines, OPPTS 870.6200, NeurotoxicityScreening Battery, EPA 712-C-98-238, 1998.179. Moser, V.C. and Ross, J.F., Eds., Training Video and Reference Manual for a Functional ObservationalBattery, American Industrial Health Council, Washington, DC, 1996.180. Ross, J.F., Mattsson, J.L., and Fix, A.S., Expanded clinical observations in toxicity studies: historicalperspectives and contemporary issues, Regul. Toxicol. Pharmacol., 28, 17, 1998.181. Moser, V.C., The functional observational battery in adult and developing rats, Neurotoxicology, 21(6),989, 2000.© 2006 by Taylor & Francis Group, LLC

CHAPTER 11The U.S. EPA Endocrine DisruptorScreening Program: In Vitro and In VivoMammalian Tier I Screening AssaysSusan C. Laws, Tammy E. Stoker, Jerome M. Goldman, Vickie Wilson, L. Earl Gray, Jr.,and Ralph L. CooperCONTENTSI. Introduction ........................................................................................................................489II. In Vitro Assays for the Detection of Endocrine Disrupting Chemicals............................490A. Estrogen and Androgen Receptor Competitive Binding Assays ..............................490B. Androgen and Estrogen Receptor-Dependent Gene Transcription Assays ..............493C. Steroidogenesis ..........................................................................................................496D. Aromatase Assay .......................................................................................................500III. In Vivo Assays for the Detection of Endocrine Disrupting Chemicals ............................502A. Uterotrophic Assay ....................................................................................................502B. Hershberger Assay .....................................................................................................505C. Male and Female Pubertal Protocols ........................................................................507D. In Utero-Lactational Exposure ..................................................................................510IV. Considerations for Evaluation of Data from the TIS Battery...........................................514V. Summary ............................................................................................................................515Acknowledgments ..........................................................................................................................516References ......................................................................................................................................516I. INTRODUCTIONIn response to emerging concerns that environmental chemicals may have adverse effects on humanhealth by altering the function of the endocrine system, 1 the Food Quality Protection Act mandatedthat the U.S. Environmental Protection Agency (U.S. EPA) develop and implement an endocrinedisruptor screening program (EDSP). Working toward this goal, the U.S. EPA is currently implementinga proposed EDSP that is designed to detect chemicals that alter the estrogen, androgen,and thyroid systems in humans, fishes, and wildlife. 2 This program, based largely upon recommendationsmade in 1998 by the Endocrine Disruptor Screening and Testing Advisory Committee(EDSTAC), 3 will allow for: (1) the initial sorting and prioritization of the 80,000 plus chemicals489© 2006 by Taylor & Francis Group, LLC

490 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONunder the purview of the EPA, (2) the identification of chemicals for further testing, using a TierI screening (TIS) battery that includes both in vitro and in vivo mammalian and ecotoxicologicalassays, and (3) the characterization of adverse effects and establishment of dose-response relationshipsfor hazard assessment, using a Tier II testing battery. In addition to developing a prioritysetting approach for chemical selection, the U.S. EPA is currently engaged in a process of optimizationand validation of those Tier I and Tier II assays that have been proposed for inclusion inthe final screening and testing program. To guide this process forward, in 2001 the U.S. EPAestablished the Endocrine Disruptor Methods Validation Subcommittee (EDMVS) made up ofgovernment and nongovernment scientists, along with stakeholder representatives from variousinterest groups, to advise and review ongoing and newly initiated methodological studies. Theagency is also working closely with the Interagency Coordinating Committee for the Validation ofAlternative Methods (ICCVAM) 4 and the Organisation for Economic Cooperation and Development(OECD), 5 who have agreed to assist with the validation of a number of the assays. Specifically,each assay that will be included in the final TIS or Tier II testing batteries will initially undergo aprevalidation process that will optimize the protocol and assure that the assay can be conductedby a qualified laboratory. The validation process will include a thorough evaluation to demonstratethe biological relevance of each assay, as well as the assay’s ability to be replicated within alaboratory and among multiple laboratories. Performance criteria for each assay will be determinedduring the validation process and will include estimates of intra- and interlaboratory variance.Here we review the mammalian in vitro and in vivo assays that will likely be included ascomponents of the U.S. EPA’s final TIS battery. An overview of each assay and the mode of actionthat each reflects are shown in Table 11.1. Each assay is discussed with respect to its purpose,technical issues, strengths, limitations, and interpretation of the data. Criteria for evaluating datafrom the TIS battery, as well as avenues for identifying new assays that will enhance and/or refinethe first-generation screening and testing batteries, are discussed.II. IN VITRO ASSAYS FOR THE DETECTIONOF ENDOCRINE DISRUPTING CHEMICALSA. Estrogen and Androgen Receptor Competitive Binding AssaysIn vitro, cell-free or whole-cell mammalian androgen receptor (AR) and estrogen receptor (ER)competitive binding assays were proposed for the TIS battery to detect compounds that bind tothese receptors and potentially affect normal hormone action. These assays can be used to determinethe relative binding affinities of environmental chemicals for the receptor, as compared with thoseof a high affinity radiolabeled hormone, such as 17-β-estradiol for the ER and R1881 or dihydrotestosterone(DHT) for the AR. The ability of a steroid hormone to bind to its respective receptorplays a central role in steroid hormone action. After the receptor binds to its ligand in an intactcell, the protein undergoes a conformational change that facilitates the formation of receptor-ligandcomplexes. These complexes can then bind to specific sequences on DNA and initiate the transcriptionof target genes 6 (Figure 11.1). Environmental chemicals that compete with endogenousligands for AR or ER binding have the potential to either induce hormone-dependent transcriptionalactivity on their own (agonist) or to block normal hormone function by preventing the endogenoushormone from binding to the receptor (antagonist). Although there are slight differences betweenthe ER and AR competitive binding assays, the basic steps for each are the same. The assay measuresthe ability of a radiolabeled ligand to bind with its respective receptor in the presence of increasingconcentrations of a test chemical (e.g., competitor). A test chemical with a high affinity for thereceptor will compete with the radiolabeled ligand for binding to the receptor at a lower concentrationthan a chemical that does not have a high affinity for the receptor. Specifically, the steps ofthe assay are: (1) the receptor and radiolabeled ligand are incubated alone or in the presence of© 2006 by Taylor & Francis Group, LLC

Table 11.1Mammalian assays under consideration for the TIS battery of the U.S. EPA’s EDSPThyroidMode of ActionEstrogen Androgen SteroidAssay Agonist Antagonist Agonist Antagonist Synthesis HPG g HomeostasisIn Vitro AssaysEstrogen ReceptorCompetitive binding a X XGene transcription b X XAndrogen ReceptorCompetitive binding a X XGene transcription b X XSteroidogenesis aAromatase a,cIn Vivo AssaysUterotrophic e X XHershberger e X XPubertal developmentFemale a X X X X XMale a,c X X X X XIn utero–lactational f X X X X X X XaAssay is currently undergoing validation by the EPA.bMethods have been reviewed by an ICCVAM Expert Panel, but no specific assays have been selected for validation.cAlternative assay recommended by EDSTAC.dAssay detects inhibition of aromatase only.eThe OECD is directing the validation of this assay in collaboration with the EPA.fAlternative assay under development by the EPA to evaluate the effects of pre- and perinatal exposures.gHypothalmic-pituitary-gonadal axis.XX dTHE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 491© 2006 by Taylor & Francis Group, LLC

492 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSHBGSSBloodSHBGS3SS2SGonad1C3SS4Target CellCytoplasm5SSSRHSP6HSPSR7S SR RHREtf8GENEAmRNADNANucleus9BiologicalResponseProtein Ar10mRNAtRNAaamRNAFigure 11.1The steroid receptor pathway. Estrogen (S) and testosterone are synthesized in the gonads fromcholesterol and enter the blood where they are transported by steroid hormone binding globulins(SHBG). S binds to its receptor, causing a transformational change in the macromolecule [e.g.,loss of heat shock proteins (HSP)]. This activates the steroid-receptor complex and allows thereceptors to bind in dimer formation to the S hormone response element (HRE). Gene transcriptionis initiated, and new mRNA is produced that results in the synthesis of a new protein.increasing concentrations of a nonradioactive competitor; (2) once the reaction has reached a steadystatecondition, the ligand-receptor complexes (bound) are separated from the unbound ligand (free);(3) the amount of radioactivity bound is quantitated; and (4) the results are analyzed and evaluated. 7Historically, the source of the receptor for ER and AR binding assays has been cytosolic extractsprepared from tissues (usually uterus for ER and prostate or epididymides for AR), and their useis well documented in the scientific literature. 8–14 However, other options are available, includingwhole-cell binding assays and the use of recombinant receptors. 15–19 Steroid-receptor binding assaysare performed at various concentrations of test compound (usually over six to seven orders ofmagnitude) in competition with a fixed concentration of radiolabeled high-affinity reference ligand(such as [ 3 H]-17-β-estradiol for ER and [ 3 H]-R1881 or [ 3 H]-DHT for AR) to generate a doseresponsecurve. The test chemical is added to a tissue preparation or to a recombinant receptorpreparation, and the amount of displacement of the radiolabeled ligand by the test chemical isdetermined. Assay tubes are also included to assess both total (no competitor present) and nonspecificbinding (low-affinity binding not reflective of interaction with the ER or AR). Nonspecificbinding is detected by measuring the amount of radioligand bound in the presence of a saturatingconcentration of unlabeled reference ligand (e.g., 100× the concentration of radioligand). Thenonspecific binding is subtracted from the total binding for each concentration of test chemical tocalculate the specific binding at that concentration. The IC 50 (inhibitory concentration 50%) is theconcentration of test chemical that inhibits the specific binding of the radioactive ligand by 50%.Each type of in vitro binding assay has its strengths and limitations. These assays only assessthe ability of the chemical to bind the receptor and are unable to differentiate agonist from antagonistresponses. Binding assays using ER or AR preparations from tissues have been used extensively,are relatively simple to conduct, and produce consistent results with practice and proper handling.In addition, tissues from multiple species are amenable for use in these assays. Whenever tissuepreparations are used, one must keep in mind that other steroid receptors are likely to be presentas well as the receptor of interest. This is a particular concern when conducting assays for the© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 493androgen receptor because ligands for the androgen receptor may also interact with the glucocorticoidreceptor (GR) or progesterone receptor (PR). In this case, a compound such as triamcinoloneacetonide, which binds to and blocks PR and GR but not AR, can be added to the buffer to ensurespecific AR binding. 20,21 Limitations of this type of assay include the use of animals to acquire thetissue, use of radioactive ligand, and the fact that they are not easily amenable to a high-throughputsystem.The use of cultured cells and/or recombinant receptors eliminates the need to isolate receptorsfrom animal tissues, can be performed with human cells or receptors, and has the possibility ofadaptation for high throughput screening. With the use of recombinant receptors, the researcheralso has more control over the receptors that are present in the preparation. For example, transientexpression of AR in cells that do not express PR or GR eliminates the need for additional blockingcomponents in the buffer. 16,18 Assays can also be conducted with purified recombinant receptor. 17,19In this case, additional factors, such as heat shock proteins, may need to be added to help stabilizethe receptor in its proper conformation for binding. Cell-based assays also require cell cultureequipment, thorough training in cell culture techniques, and the use of radioactivity. Althoughpurified recombinant receptor techniques have not been used as extensively as receptor preparationsfrom tissues, they have the potential to provide an alternative method for the TIS battery.Data interpretation is an important component of steroid-receptor binding studies. Typically,the data from each competitive binding assay are plotted on a semilog graph so that a sigmoid plotis obtained. The data can be plotted as counts per minute bound or as percent of total specificbinding on the Y axis versus the concentration of the test chemical on the X axis. If the test chemicaland the radiolabeled ligand compete for a single common binding site, the competitive bindingcurve will have a shape (Figure 11.2A) determined by the law of mass action. 22 Specifically, thecurve should descend from 90% to 10% specific binding over approximately an 81-fold increasein the concentration of the test chemical (e.g., this portion of the curve will cover approximatelytwo log units, Figure 11.2B). A binding curve that drops dramatically (i.e., from 70% to 0%) overone order of magnitude should be questioned, as should one that is U-shaped (i.e., percent boundat first decreases with increasing concentration of competitor but then begins to increase, Figure 11.2C).In the both cases, something has happened to the dynamics of the binding assay, and the reactionis no longer following the law of mass action (e.g., the test chemical may be binding to multiplebinding sites or the test chemical may not be soluble at higher concentrations in the assay bufferand is precipitating from the solution). In these cases, an additional K i experiment may be requiredto determine if the test chemical is truly a competitive inhibitor. While it is possible to calculate aK i value from the IC 50 by use of the Cheng-Prusoff equation, 23 the use of this equation is only validif the competitive binding curve reflects a pure competition for a single binding site. 24 An experimentallyderived K i requires adding increasing concentrations of radiolabeled ligand in the presenceof fixed concentrations of the test chemical and then plotting the data on a double reciprocal plot. 25,26A pattern of lines that intersect on the ordinate axis is characteristic of competitive inhibition(Figure 11.3). Slopes obtained from each of the double reciprocal plots are then replotted as afunction of the inhibitor concentration. The slope and intercept of the replot can be used to calculatea K i value for the test chemical. 26Performance criteria for binding assays are being developed by the Office of Science, Coordinationand Policy (OSCP) of the U.S. EPA through an agreement with the ICCVAM. Even thoughsteroid receptor binding assays have been used for years, they have not been rigorously standardizedand validated as required by the EDSP. Currently, protocols for ER and AR competitive bindingassays are undergoing the standardization and validation process, under the direction of the OSCP.B. Androgen and Estrogen Receptor-Dependent Gene Transcription AssaysCell-based gene transcriptional activation assays were also proposed by EDSTAC, either to complementor as an alternative to the ER and AR competitive binding assays. These assays can be© 2006 by Taylor & Francis Group, LLC

494 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPercent Specific BindingPercent Specific BindingPercent Specific Binding1251007550250-121251007550251251000-12755025(A)(B)IC50-11 -10(C)0-12 -1190%81 fold10%-11 -10 -9 -8 -7 -6Competitor (M)-10-9 -8 -7 -6Competitor (M)-9 -8 -7 -6Competitor (M)-5-5-5-4-4-4-3EstradiolEstradiolCompetitive binder-3EstradiolChemical 1Chemical 2-3Figure 11.2Examples of data from a steroid competitive binding assays. (A) Typical competitive binding curve,demonstrating that inert estradiol and [ 3 H]-E 2 compete for a single common binding site (ER),following the law of mass action. (B) Comparison of competitive binding curves for inert estradioland a competitive inhibitor. (C) Examples of data that do not follow the law of mass action. TheU-shaped curve produced by Chemical 1 is indicative of a solubility problem. The dramatic dropfrom 100% to 12% binding over a one order of magnitude increase in Chemical 2 indicates aproblem with the dynamics of the binding assay.used to identify compounds that bind to the ER or AR and, unlike standard receptor binding assays,can also discriminate agonists from antagonists because they assess the ability of a chemical toaffect hormone-dependent gene transcription. A vector carrying a hormone response element forthe receptor of interest attached to a reporter gene is introduced into the cells. In some cases, asecond vector that expresses the receptor (if not already endogenous to the cells) is also cotransfectedinto the cells. When the ligand-receptor complex binds to the hormone response element in dimerformation, the transcription of the reporter gene is initiated. The level of expression of the reportercan then be measured. The assays are conducted by diluting a range of concentrations of the testchemical in the cell culture medium. The test chemical is examined alone to assess its ability toactivate hormone-dependent gene transcription (agonism) or in the presence of a high-affinity ligand(such as 17-β-estradiol for ER or DHT for AR) to assess its ability to interfere with or blockhormone-dependent gene transcription (antagonism).© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 4951/Bound (nM)2015105(A)050 µM100 µM200 µM-1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.51/Total (nM)9(B)7Slope531-100 0 100 200Competitor (µM)Figure 11.3Experiment demonstrating competitive inhibition of steroid receptor binding by a test chemical.(A) Double-reciprocal plots of inhibition of [ 3 H]-ligand specific binding. (B) Slope-replot analysisfor the determination of the inhibition constant (x intercept = K i ). These data represent a testchemical that is a competitive inhibitor of estrogen receptor binding with a K i = 89 µM.There are benefits as well as limitations to the use of mammalian versus yeast cells for theseassays. Yeast-based assays are simple, sensitive, and generally less expensive to conduct than assaysusing mammalian cells. However, with yeasts there are rank-order differences in potencies ofendocrine active test chemicals, 27–29 as well as differences in permeability of compounds throughthe yeast cell wall relative to the plasmalemma of mammalian cells. In this respect, assays utilizingmammalian cells are thought to be more representative of risk to humans.A number of in vitro transcriptional activation assays using mammalian cells have been developedthat assess either estrogen- or androgen-dependent gene transcription. Transient transfectionassays in which both the receptor and a hormone-responsive reporter are cotransformed intocells 18,30–34 have been widely used. However, in transient transfection assays, target DNA sequencesare over expressed and therefore do not reflect physiological levels of receptor. In addition, transfectionefficiencies (i.e., percentage of cells expressing the receptor) may vary greatly from assayto assay. Unfortunately, the responsiveness is limited in time, as the transgenes are usually lostwithin 72 h, so transfections must be repeated with each new set of assays.Adenoviral transduction, which uses a replication-defective virus to deliver biologically competentgenes, has also been used to develop androgen receptor transcriptional activation assays. 35Since the virus is replication defective, it represents no hazard of infection. Both the AR and areporter gene are introduced into either CV-1 or MDA-MB-453 cells. As with transient transfectionassays, transduction requires that the procedure be repeated for each new set of assays. However,viral infection appears to be much more efficient than transfection because induction is consistentlyhigh, with much lower variability. Transduction and transfection techniques require similar facilities,and both can be easily accomplished with basic cell culture supplies and equipment (culture hoodand cell culture incubator). Stably transformed cells eliminate the need for repeated transient© 2006 by Taylor & Francis Group, LLC

496 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONtransfections and reduce the variability associated with this technique. However, stably transformedcell lines utilizing endogenous or transformed androgen receptor combined with a stably transformedhormone-responsive luciferase (or other) tagged reporter were not readily available untilrecently. These are genetically modified cell lines that are derived by transfection followed byantibiotic selection and clonal expansion. Ideally, this methodology produces a stable populationof cells that will respond uniformly to stimuli. 36–41Transcriptional activation assays have several strengths. They utilize a well-defined mechanismof action to distinguish agonists from antagonists and can be performed using mammalian cells.Thus, the results are likely applicable to humans. Transient transformation and viral-transductionassays are highly responsive (i.e., give high levels of induction in comparison with vehicle controls)and provide more control over the specificity of response. For example, androgen receptors orglucocorticoid receptors can be transfected into cells separately, and the responses can be assessedindependently. Both adenoviral-transduced and stable cell lines are amenable to a 96-well plateformat and give a consistent response with low variability, thus making them attractive for adaptationto high throughput systems (HTS). Generally, in vitro transcriptional activation assays are costeffective and, with proper training and equipment, are rapid and easy to perform. Cell lines derivedfrom human tissue can be used to develop these assays, so it is human receptors that are used inthe assessments. In addition, only a small amount of test chemical is needed to run multiple assays.Transcriptional activation assays also have some limitations. Compared with traditional receptorbinding assays that use cytosolic preparations, there is a requirement for cell culture equipment,techniques, and training. The presence of more than one receptor in the cells (especially if usingcells with endogenous receptors) that can activate the hormone response element of the reportermay be viewed as a limitation. This requires additional assays using selective competitors todifferentiate individual receptor activity. Transient transfection assays require separate transformationsfor each set of assays conducted, can have high interassay variability, and are generally notsuited for high throughput systems. Stable cell lines have many advantages, as discussed above,but require a considerable investment of time and resources to produce each cell line. Anotherconcern is the possibility of false negatives and false positives. These systems assess the testchemical, but if a particular test chemical requires metabolic activation to produce its hormonallyactive metabolite, it could produce a false negative in the assay. It is also necessary to assess theissue of cytotoxicity when working with cell-based assays. In tests for antagonist activity, decreasesin reporter activity could be due to either the test chemical or general cellular toxicity from thecompound. The incorporation of cell toxicity tests into routine chemical testing is an importantextension of the method. Otherwise, a test chemical could be falsely identified as an antagonist.C. SteroidogenesisThe EDSTAC final report recommended two assays for the evaluation of effects of environmentalchemicals on steroidogenesis. The first, a sliced testis protocol, was designed to measure testosteroneproduction in vitro from testicular parenchymal fragments. The second, an aromatase assay designedto assess the potential effect of toxicant exposure on the conversion of testosterone to estradiol,was recommended as an alternate approach. Thus, the focus on steroidogenesis was on two majorbiologically active hormones, testosterone and estradiol, both generated in the later portion of thegonadal steroidogenic pathway (Figure 11.4).The rationale for recommending the sliced testis assay was based on a number of previousstudies that demonstrated the utility of the method for identifying toxicant-induced endocrinealterations. 42,43 In general, this method measures testosterone release in vitro under both nonstimulatedand human chorionic gonadotropin (hCG)-stimulated conditions over a period of hoursfollowing testis decapsulation. The assay may be conducted using testes from animals followingin vivo exposure, or it may evaluate in vitro testosterone production by using testes from animalswith no prior exposure to the test chemical. Since a number of variations in the procedure have© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 497CholesterolSex Steroid SynthesisHOP450sccPregnenoloneHOStAR/PBRCH 3C = O3β-HSDCH 3C = O17α-hydroxylase/17,20 lyaseHO17α-hydroxylase/17, 20 lyase17 hydroxyPregnenoloneCH 3C = OOH3β-HSDCH3C = OOH∆5 Pathway17α-hydroxylase/17,20 lyaseHO17α-hydroxylase/17,20 lyaseODihydroepiandrosterone3β-HSDOOProgesterone∆4 PathwayO17 hydroxyProgesteroneStAR - Steroidogenic Acute Regulatory ProteinPBR - Peripheral benzodiazepine receptorP450scc - CYP11A (side chain cleavage)3β-HSD - 3β-hydroxysteroid dehydrogenase/∆5-4-isomerase17α - hydroxylase/17,20 lyase - CYP17P450arom - CYP19 (aromatase)5α-DHT - 5α-dihydrotestosteroneAndrostenedioneOP450arom 17β-HSDOHO5α-reductase5α-DHTTestosteroneOHOEstroneP450 arom17β-HSDOHHOEstradiolFigure 11.4The sex steroid pathway. Chemical structures and enzymes for the synthesis of steroids.been employed over the years, optimization and standardization of the protocol has been a prerequisiteto using the method as a screen. Consequently, as part of the prevalidation process of theassay for possible use in the TIS battery, a number of experiments are being conducted to optimizethe type of incubation medium, gaseous environment, temperature, parenchymal fragment size,time to achieve initial baseline testosterone secretion, appropriate hCG concentration, and measuresof tissue viability. Moreover, the optimal age range of the animals from which the tissues areobtained is being established.The basic sliced testis protocol is depicted in Figure 11.5. Following an initial replacement ofmedium to eliminate any hormone released by trauma, sampling from incubated fragments isperformed at 1-h intervals. This establishes a baseline hormonal secretion and gauges the magnitudeof the response to a stimulating concentration of hCG. However, some modifications in the protocol(Table 11.2) will likely emerge from the optimization process currently being conducted under theoversight of the U.S. EPA.For use in characterization of the effects of environmental toxicants on steroid synthesis, thesliced testis approach can be compared with the evaluation of circulating steroid concentrations inresponse to in vivo exposures and to the assessment of steroid production in vitro by use of enrichedor purified gonadal cell preparations. All have strengths and limitations when considered in thecontext of a screening application for endocrine disruptors (Table 11.3). Assessments of serumsteroid concentrations are easily performed, whether done under terminal conditions or as part ofa periodic sampling during an in vivo study. The serum samples can be obtained without extensivetraining, and characteristics of xenobiotic absorption, distribution, metabolism, and excretion(ADME) can be evaluated with an in vivo exposure paradigm. However, the specificity of the effecton steroidogenesis would be questionable. Any alteration in testosterone secretion observed following© 2006 by Taylor & Francis Group, LLC

498 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSlowly Shake andIncubate34°C, 95% O 2 , 5% CO 2MediumReplaced0.5 hrFreeze Freeze FreezeFreezehCG ChallengeMeidum 1991.5 ml(0.75 ml/50 mg tissue)Sample 100 µlthen hCGchallenge in1.5 ml mediaSample100 µlSample100 µlSample100 µlWeigh0 hr 1.5 hr +2.5 hr +3.5 hrBaselinehCGSample+4.5 hrWeighMediumReplaced0.5 hrFreeze Freeze Freeze FreezeNo ChallengeMedium 1991.5 ml(0.75 ml/50 mg tissue)Sample 100 µlthen mediareplacedwithout hCGSample100 µlSample100 µlSample100 µl0 hr 1.5 hr+2.5 hr +3.5 hr +4.5 hrBaselineSampleFigure 11.5Example of the sliced testes protocol. Medium replacement after an initial 30-min incubationminimizes the influence of traumatic testosterone release on baseline secretion. Aliquots ofmedium are then sampled on an hourly basis under hCG-stimulated or nonstimulated conditions.the in vivo exposure, for example, could reflect either a direct effect on steroidogenesis or a responsesecondary to an indirect gonadotropic stimulation. An alternative approach would be to administerthe test chemical systemically and assess the gonadal secretory response in vitro, using testiculartissue taken at necropsy. In this case, unlike the case of in vitro only exposures, ADME conditionswould still be maintained. Also, temporal controls over steroid sampling can be more strictlyregulated under culture conditions where hormonal stimulation is artificially imposed. This typeof stimulation, often using luteinizing hormone (LH) or hCG, is employed to evaluate the effectsof a xenobiotic on the responsiveness of the gonadal preparation to gonadotropic stimulation inculture. Costs tend to be higher than for strictly in vitro exposures because the animal usage isgreater and larger amounts of the compound under study are needed. A third alternative is toevaluate gonadal preparations exposed only in vitro. In this case, any measured alteration in steroidsecretion would reflect a direct effect on the process of steroidogenesis.The sliced testis protocol recommended by EDSTAC employs fragments of parenchymal tissue,which are relatively straightforward to prepare. Moreover, animal usage is reduced if the invitro–only exposure paradigm is used. Multiple parenchymal tissue fragments from a single testisor testis pair can be randomly assigned to different treatment conditions, minimizing the influenceof animal-to-animal variability. Nevertheless, one of the issues raised concerning the utility of asliced (or minced) testis approach to the assessment of testosterone secretion is the question ofvariability of the obtained data in comparison to other types of incubated testicular preparations.For example, isolated Leydig cells have been used for the same purpose and have the advantageof increased cellular uniformity and sensitivity. Both crude enriched 44–47 and purified 48–50 Leydigcell preparations have been employed to measure testosterone secretion, with the latter (85% to90% purity) exhibiting less variability in controls, as reflected by coefficients of variation (COV)calculated from reported control data (Table 11.4). On the other hand, crude Leydig cell preparations© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 499Table 11.2Potential modifications to sliced testes protocol currently under considerationFragment size: The protocol included in the EDSTAC final report suggested using a parenchymal fragment sizeof 50 mg. This would allow a significant number of samples from a single testis to be distributed across multiplechemical treatments for both nonstimulated and hCG-stimulated conditions, thereby reducing the variabilityinherent in using individual animals for exposures to different xenobiotics. At the same time, use of such smallerfragment sizes would minimize the number of animals required for testing. Ultimately, the determination of anappropriate amount of tissue will depend on the variability in hormonal secretion observed.Incubation medium: Incubation times for the sliced testis protocol are short term and typically extend no longerthan 5 h. While sterile-filtered Medium 199 has frequently been used for gonadal incubations, a variety of nutrientmedia (e.g., Krebs-Ringers bicarbonate, Dulbecco’s modified Eagle’s medium) should be acceptable.Nevertheless, modest disparities in secretion with different media may exist. There has been a trend away fromusing phenol red as a pH indicator in these media, since a growing number of publications have shown that itpossesses estrogenic properties a–c and may have an adverse effect on Leydig cell testosterone production. dGaseous environment: Incubations of testicular fragments and enriched Leydig cell preparations havecommonly been carried out under 95% air (approximately 19% oxygen) and 5% CO 2 . Oxygenated media havealso been employed, e,f and it is not likely that over the short term the increase in oxygen content in the incubationmedium would have an adverse effect on tissue viability.Tissue viability: For any type of in vitro system, it is essential that an assessment of cell and tissue viabilitybe performed because reductions in steroid secretion could be attributable to a general toxicant-induced celldeath and not a targeted alteration in steroidogenesis. The use of techniques relying on perturbations in cellmembrane integrity (e.g., trypan blue, propidium iodide, lactic dehydrogenase) are later stage indices ofcytotoxicity and may not identify earlier events in cell death. Alternatively, measures of mitochondrial integrity(i.e., tetrazolium-based MTT, MTS assays) are more likely to reflect a cytotoxic insult taking place within the 4to 5 hour time frame of the sliced testis procedure. An estimation of impairments in mitochondrial membranepotential using specific dye probes appears to be a very sensitive index of early apoptotic changes, althoughsuch procedures are more appropriate for cell-based assays and their utility for assessments of tissue cytotoxicitywould have to be determined.aHofland, L.J., van Koetsveld, P., Koper, J.W., den Holder, A., and Lamberts, S.W., Weak estrogenic activity ofphenol red in the culture medium: its role in the study of the regulation of prolactin release in vitro, Mol. Cell.Endocrinol., 54, 43, 1987.bWelshons, W.V., Wolf, M.F., Murphy, C.S., and Jordan, V.C., Estrogenic activity of phenol red, Mol. Cell.Endocrinol., 57, 169, 1988.cDumesic, D.A., Renk, M., and Kamel, F., Estrogenic effects of phenol red on rat pituitary cell responsivenessto gonadotropin-releasing hormone, Life Sci., 44, 397, 1989.dAbayasekara, D.R., Kurlak, L.O., Band, A.M., Sullivan, M.H., and Cooke, B.A., Effect of cell purity, cellconcentration, and incubation conditions on rat testis Leydig cell steroidogenesis, In Vitro Cell Dev. Biol., 27A,253, 1991.eWilker, C E., Welsh, T.H. Jr, Safe, S.H., Narasimhan, T.R., and Johnson, L., Human chorionic gonadotropinprotects Leydig cell function against 2,3,7,8-tetrachlorodibenzo-p-dioxin in adult rats: role of Leydig cell cytoplasmicvolume, Toxicology, 95, 93, 1995.fHarris, G.C. and Nicholson, H.D., Characterization of the biological effects of neurohypophysial peptides onseminiferous tubules, J. Endocrinol., 156, 35, 1998.(approximately 12% to 15% purity) show COVs that are comparable with those obtained usingsliced testicular tissue. As a screening approach, the increase in the technical demands of theisolation procedure, the costs, and the time necessary to complete the Leydig cell enrichmentprocess are factors militating against this option.One additional option for consideration is the use of stabilized steroidogenic cells. A growingnumber of studies in steroid toxicology have employed cells derived from mouse or rat Leydig celltumors. 51–53 Depending on which portion of the sex steroid pathway is of concern, different celllines can be used. The mouse MA-10 Leydig cell tumor line has been shown to be particularlyuseful for detection of alterations in steroidogenic acute regulatory (StAR) protein, cytochromeP450 side chain cleavage enzyme, and 3-hydroxysteroid dehydrogenase activity, all of which affectprogesterone synthesis early in the steroidogenic pathway. 54–56 In contrast, the KGN human granulosatumor cell line has been shown to be particularly high in P450 aromatase activity, making itpotentially valuable for assessment of toxicant-induced insult to estradiol production. 57 However,this cell line shows low 17-hydroxylase and 17,20-lyase activity, which would make it unsuitablefor detection of effects on androstenedione synthesis. One cell line that appears to hold promise© 2006 by Taylor & Francis Group, LLC

500 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 11.3Strengths and limitations of different assessments of testosterone productionType of Assessment Strength LimitationIn vivo exposure and invivo samplingIn vivo exposure and invitro sampling (slicedtestis protocol)In vitro exposure and invitro sampling (slicedtestis protocol)In vitro exposure and invitro sampling (enrichedand purified Leydig cells)• Relatively easy to perform,extensive training not necessary• Can account for xenobioticabsorption, distribution,metabolism and excretion (ADME)• Not difficult to perform• ADME parameters maintained• Temporal conditions of stimulationmore strictly controlled.• Not difficult to perform• Testicular specificity of response• Animal usage and costs reduced• Multiple tissue fragments fromsingle testis or testis pair can berandomly assigned to treatmentgroups, minimizing animal-toanimalvariability• Increased uniformity in cellpopulation• Increased sensitivity of response• Purified Leydig cell preparationsshow reduced variability inresponse• Testicular specificity of responsequestionable• Animal usage and costs higher than forin vitro exposure conditions• Testicular specificity of responsequestionable• Animal usage and costs higher than forin vitro exposure conditions.• Variability of data tends to be higher thanfor purified Leydig cell preparations• Sensitivity of response less than for morepurified cell preparations• ADME factors absent• Technically more demanding, particularlyfor purified Leydig cells• Increase in cost over sliced testisapproach• ADME factors absentTable 11.4Control Group Coefficients of Variation: Testosterone SecretionPreparation Non-stim. LH*/hCG-stim. ReferenceSliced Testis 1023282350Crude Leydig Cell Prep(~12–15%)Purified Leydig Cell Prep(80–95%)for evaluations of xenobiotic alterations in steroid production is the H295R line, which is derivedfrom human adrenocortical tumor cells. This cell line appears to contain all of the enzymes of thesteroidogenic pathway 58–60 and has already been used to a limited extent to evaluate effects onaromatase activity. 61,62 Since the use of a cell line would eliminate the necessity for animals andthe line is available through the American Type Culture Collection (ATCC), this approach offersdistinct advantages for studying the entire sex steroid pathway. For these reasons, its utility iscurrently under study by the U.S. EPA for possible substitution in the TIS battery.D. Aromatase Assay452640151196—1327—2922—302725*1223—81212Lau & Saksena (1981) Int. J. Androl 4:291.Gray et al. (1995) TAP 130:248.Powlin et al. (1998) Tox. Sci. 46:61Wilker et al. (1995) Toxicology 95:93.Banczerowski et al. (2001) Br.Res. 906:25.Chambon et al. (1985) Andrologia 17:172.Kan et al. (1985) J. Steroid Biochem. 23:1023.Raji & Bolarinwa (1997) Life Sci. 61:1067.Laskey & Phelps (1991) TAP 108:296.Ronco et al. (2001) Toxicology 159:99.Nagata et al. (1999) FEBS Lett. 444:160.Romanelli et al. (1997) Life Sci. 61:557.Klinefelter et al. (1991) TAP 107:460.Guillou et al. (1985) FEBS Lett. 184:6.Incubation parameters: 10 5 –10 6 cells/well; 3–4h collection period; 100mlU hCG or *50 ng/ml oLH stimulationAromatase (CYP19) is a cytochrome P450 enzyme complex that converts androgens to estrogens(Figure 11.4). The enzyme is present in multiple tissues, such as the ovary, placenta, uterus, testis,© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 501brain, and adipose tissue, and is essential for normal reproductive development and function inboth males and females. As indicated in the previous section, an in vitro aromatase assay is currentlyincluded as an alternative assay in the TIS battery. This assay may be used if the final screeningbattery does not include the female pubertal assay, a protocol that can also detect alterations in thesynthesis of estrogen. The need for an assay to screen specifically for changes in aromatase activityis justified by numerous reports that environmental contaminants can affect the activity of thisenzyme. Several flavonoids and related phytoestrogens, 63–65 as well as certain pesticides, 66 havebeen shown to inhibit aromatase activity in vitro. Studies to evaluate the effects on reproductivefunction in mammalian and nonmammalian species following exposure to several pesticides,including fenarimol, imazalil, prochloraz, and triadimefon, have demonstrated reduced fertility asa result of decreased serum estrogens and altered gonadotropin concentrations. 67–69In vitro methods may be readily used to identify chemicals that inhibit aromatase activity byacting as suicide substrates. Generally, homogenates or microsomal preparations of aromatasecontainingtissues, such as ovaries, testes, brain, or human placenta, are incubated for a designatedperiod with the test chemical, radiolabeled androstenedione or testosterone, and essential cofactors.The estrogen product, estrone or estradiol, that forms during the enzymatic reaction can be isolatedby use of thin-layer chromatography or high-pressure liquid chromatography (HPLC) and quantifiedby liquid scintillation spectroscopy. Alternatively, the tritiated water assay is a method that measuresthe amount of 3 H 2 O released during incubation with [1β- 3 H]-androstenedione. The basis of thismethod is that 1 Mol of 3 H 2 O is released during the aromatization of 1 Mol of the A-ring of[1β- 3 H]-androstenedione to estrone. 70 The advantage of the 3 H 2 O method is that it does not requireany chromatography to separate steroid substrate. Rather, the amount of 3 H 2 O released is measuredin the aqueous phase following extractions with organic solvents and dextran-coated charcoal.Results from the product isolation method and the 3 H 2 O method have been shown to be comparablefor human placental microsomes, rodent ovarian microsomes and tissues, and cell culture systems.Aromatase activity is typically reported as enzyme velocity or rate (nanomole product formed permilligram of protein per minute). The source of tissue used for these two methods can vary,depending upon the level of aromatase activity required for a particular study. Human placentascollected at term contain very high levels of aromatase and provide a good source for a microsomalpreparation. Rodent ovarian microsomes and recombinant-human aromatase can also be usedsuccessfully with these methods although the level of aromatase activity is lower than that observedusing human placental microsomes.A number of in vitro cell culture systems have also been used to evaluate the effects ofenvironmental chemicals on aromatase activity. These cell culture systems have the advantage ofpotentially detecting both inhibition and induction of enzyme activity. Additionally, some cell linesmay provide metabolic activation of some test chemicals as well as providing information on theability of the chemical to cross the cell membrane. Cell lines most commonly used include thehuman JEG-3 and JAR choriocarcinoma cells, 71 R2C rat Leydig carcinoma cells, 72 MCF-7 humanmammary carcinoma cells, 73 and H295R human adrenocortical carcinoma cells. 61,62,74 The JEG-3and JAR cell lines display a fairly high basal level of aromatase activity, a characteristic that isadvantageous when evaluating chemicals that might have an inhibitory effect. In contrast, the H295Rcell line has been used to detect induction of aromatase activity by a number of chlorotriazineherbicides (e.g., atrazine, simazine, and propazine) and the fungicide vinclozolin. 61,62,74 Elevatedlevels of CYP19 mRNA detected using RT-PCR indicate that the induction of enzyme activity inthese cells is regulated at the level of gene transcription rather than by a change in the activationof the enzyme itself. Although induction of aromatase activity following in vivo exposure to thesechemicals has not been currently documented, Stoker et al. have reported increased serum estroneand estradiol concentrations in male rats exposed to atrazine, 75 suggesting that aromatase activitymay be induced. Another potential advantage of the H295R cell line is that it has been reported topossess multiple promoters that are responsible for the tissue-specific regulation of aromatasesynthesis in mammals. 61 Thus, this cell line may provide an opportunity to determine whether a© 2006 by Taylor & Francis Group, LLC

502 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONparticular test chemical has an effect on aromatase activity through one of a number of promotersites, as opposed to other cell lines that do not possess multiple promoters.The approach employed to evaluate whether a particular chemical can disrupt the regulation ofaromatase activity depends upon a number of factors. Chemicals that inhibit aromatase activity byacting as suicide substrates can be identified with fairly simple in vitro techniques. Since inhibitionof aromatase activity has been demonstrated to be a common mechanism for a variety of environmentalchemicals, in vitro assays using microsomal preparations and the production of 3 H 2 O toevaluate aromatase activity provide a relatively easy and inexpensive screening method. However,chemicals that might induce aromatase activity by altering the regulation of enzyme synthesis orstability require the use of in vitro whole cell systems and/or in vivo studies that evaluate multiplearomatase-containing tissues. Finally, the design of in vivo studies is critical for understanding (1)whether a chemical has a direct effect on aromatase activity versus an indirect effect mediatedthrough a change in the regulation of the hypothalamic-pituitary-gonadal axis and (2) whether anyeffect on aromatase is tissue dependent. For example, circulating estrogen concentrations areinfluenced by the levels and cyclicity of pituitary gonadotropins. 76 Thus, the evaluation of aromataseactivity after longer exposures to a given chemical may not reflect its primary mode of action. Additionally,since aromatase expression is under the control of tissue-specific promoters regulated bydifferent cohorts of transcription factors, 77 it is important to evaluate multiple tissues for aromataseactivity in the in vivo studies. In some cases there could be local alterations in estrogen biosynthesisin a tissue that would not be reflected by an alteration of overall circulating hormone levels. 78A. Uterotrophic AssayIII. IN VIVO ASSAYS FOR THE DETECTIONOF ENDOCRINE DISRUPTING CHEMICALSThe rodent uterotrophic assay, recommended by EDSTAC as one of the core in vivo assays for theTIS, is generally accepted as a robust, technically simple, and reliable method for detectingestrogenic activity. The assay was first developed during the mid-1930s for use in both mice andrats, 79–81 and continues to be routinely used today as the gold standard to detect both potent andweak estrogen agonists 82 as well as antagonists. 83,84 Specifically, the assay measures the ability ofa test chemical to significantly increase the weight of the rodent uterus after three consecutive daysof exposure, or in the case of an antagonist, to prevent an estrogen-dependent increase in uterineweight. The assay is based upon an estrogen mode of action that includes estrogen binding to itsreceptor, initiation of gene transcription, and induction of uterine growth. 85,86 A number of studieshave correlated in vitro ER binding affinity with a uterotrophic response following in vivo exposureto a broad range of pharmaceuticals and environmental chemicals. 87–89In collaboration with the U.S. EPA, the OECD has agreed to direct the optimization andvalidation of the rodent uterotrophic assay for the TIS battery. The OECD has concentrated itsinternational validation program on evaluating the performance of two versions of the uterotrophicassay, one using sexually immature female rats and the other using adult ovariectomized (OVX)rats. Both of these versions provide an animal model that is devoid of endogenous estrogen, therebyensuring that any observed estrogenic response will be a direct influence of the test chemical.Owens and Ashby 82 have published an overview of the OECD protocol for the uterotrophic assay,along with an extensive review of the biological basis and use of the assay throughout the past 60years. In brief, sexually immature female (less than 21 days of age) or adult OVX rats (approximately60 days of age at ovariectomy with a minimum of 14 days’ recovery) are dosed with a testchemical by oral gavage or subcutaneous injection for three consecutive days, at approximately24-h intervals. Body weights are monitored daily, and the absolute uterine weights of both the wetuterus (includes imbibed fluid) and blotted uterus are recorded at necropsy, 24 h following the final© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 503dose. Uterine weights are evaluated by analysis of variance, with body weight at necropsy as acovariable.Although there are numerous issues that must be considered during the optimization andvalidation process for the uterotrophic assay, the OECD has developed a strategy and begun tomethodically move through the process. 90 Initially, three major variables that required considerationincluded species selection (e.g., mouse versus rat), age of the test animal (sexually immature versusadult OVX), and route of administration (oral gavage versus subcutaneous injection [s.c.]). Additionalfactors, such as animal strain, animal husbandry (influence of diet or vehicle), and length oftime between last dose and necropsy (6 versus 24 h), that could possibly confound the results alsorequired special consideration.Drawing upon the vast amount of historical data and published literature on the uterotrophicassay, the OECD has based its approach to the validation process on scientific rationale and practicalexperience. The group chose to standardize the protocol with the laboratory rat as opposed to themouse, as the former is the preferred rodent model in toxicology testing paradigms for regulatorypurposes. It was also decided that studies would be designed to address questions regarding dosingroute, strain variation, and animal husbandry during the multiple laboratory evaluation. Phase I ofthe OECD validation process was designed to test, refine, and standardize the sexually immatureand the adult OVX rat uterotrophic assays using a high-potency estrogen, ethinyl estradiol, and theestrogen receptor antagonist, ZM 189,154. This phase was also designed to provide an initialdetermination of intra- and interlaboratory variability. Phase 2 was designed to demonstrate thecapability of the standardized uterotrophic protocols to identify estrogenicity in a set of testchemicals composed of weak estrogens and a known negative, as well as to continue to documentintra- and interlaboratory variations. The potential effects of low levels of isoflavones in laboratoryanimal chow were also evaluated to determine if the animal chows containing higher concentrationsof phytoestrogens would hamper the detection of estrogenic activity during tests of weak agonists.The results from Phases 1 and 2 of the validation process for the uterotrophic assay have recentlybeen published. Phase I studies were conducted by 19 laboratories from Japan, Denmark, France,Germany, Korea, the Netherlands, the United Kingdom, and the United States. 90 The study evaluatedfour protocols: (1) sexually immature, female rat model, with oral dosing for 3 days, (2) sexuallyimmature, female rat model, with dosing by s.c. injection for 3 days, (3) adult, OVX rat model,with dosing by s.c. injection for 3 days, and (4) adult, OVX rat model with extended dosing bys.c. injection for 7 days. Animals were exposed to ethinyl estradiol (0.03 to 3.0 µg/kg/d), or ZM189,154 (0.1 and 1.0 µg/kg/d) coadministered with ethinyl estradiol (0.3 µg/kg/d). All animals werekilled 24 h after the last dose. Other experimental conditions, such as animal strain, vehicle, diet,and bedding, varied between laboratories. Data from these studies demonstrated that all laboratoriessuccessfully conducted each of the four protocols, and that each protocol detected a significantincrease in uterine growth following exposure to ethinyl estradiol. For each protocol there was goodagreement among laboratories with respect to the lowest effective dose (LOEL) of ethinyl estradiol.Although the shape of the dose-response curve for ethinyl estradiol did vary between laboratories, therewas general agreement in dose response when the 10% and 50% effective doses (ED 10 and ED 50 ) werecalculated. Given the differences in animal strain and husbandry, the robust nature of each of theprotocols for detecting increases in uterine weight was demonstrated. In addition, preliminary evidencethat the protocols are a feasible method for detecting antagonists was also provided. The recommendationsresulting from these studies included: (1) moving forward with the validation of both thesexually immature and adult, OVX animal models, (2) refining the protocols to address minor technicalissues, and (3) continuing to test the protocols, using several weak estrogen agonists.Studies conducted under Phase 2 of the OECD validation process are reported by Kanno etal. 91,92 and Owens et al. 93 One study evaluated the same four protocols as employed in Phase I,using ethinyl estradiol, several weak estrogens (genistein, methoxychlor, nonylphenol, bisphenolA, and o,p′-DDT), and a negative control, dibutylphthalate. Twenty-one laboratories from ninenations participated in the study, although not all conducted each protocol or tested all of the© 2006 by Taylor & Francis Group, LLC

504 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONchemicals. Overall, the uterotrophic assay provided reproducible results within the same laboratoryand across the participating laboratories for all the test chemicals. 91 With the exception that thelowest effect levels (LOELs) differed for the test chemicals by dosing routes, no major differenceswere noted between the protocols. As a result, the sexually immature and adult, OVX animal modelswere judged to be qualitatively equivalent to one another, and both protocols will continue to moveforward through the validation process. There were some false positives for dibutylphthalate inwhich 5 of 36 experiments resulted in a statistical increase in uterine weight as compared with thecontrol (vehicle). Of these, three sets of data for the dibutylphthalate had uterine weights significantlyhigher than the vehicle control, yielding a false positive rate of 8% for this test chemical. 92These data reinforce the importance of using a weight-of-the-evidence approach for data interpretation,using all the assays within the TIS battery. Finally, a comparison of the phytoestrogen contentin the diets used in the participating laboratories showed that levels lower than 325 to 350 µg/gtotal genistein equivalents did not impair the responsiveness of the uterotrophic assay when testingthe weak estrogen-receptor agonists, nonylphenol and bisphenol A. 93 However, recommendationsto monitor dietary phytoestrogen content will most likely be included in the final OECD guidelinesfor the uterotrophic assay as well as recommendations to monitor control uterine weights to becertain they fall within a range consistent with historical data.The advantages of using the uterotrophic assay as one of the in vivo components of the TISbattery are exemplified by its extensive use as an indicator of an estrogen receptor–mediatedresponse. Historical data, as well as data recently collected during the OECD validation studies,have proven the assay to be both biologically relevant (e.g., it detected chemicals known to bindto the estrogen receptor) and reliable (in that results can be replicated within a laboratory andbetween laboratories). The robust nature of the assay has also been proven through the ability ofmultiple laboratories to replicate results, even with variations in experimental conditions, such asstrain, dosing route, animal husbandry, and animal model (e.g., sexually immature versus adult,OVX female rats). Within the TIS battery, the uterotrophic assay is one of the most technicallysimple. It requires less time to complete and fewer animals per treatment group than the femalepubertal protocol, and in some cases it appears to be more sensitive to weak environmentalestrogens. For example, bisphenol A is known to possess a weak binding affinity for the estrogenreceptor 89,94 and to increase uterine weight following oral or s.c. exposures. 89,95 However, thischemical did not alter the age of vaginal opening when administered by oral gavage from 21 to 35d of age at doses two to three times higher than the dose required to stimulate uterine growth. 89Ashby and Tinwell 95 also reported no change in the age of vaginal opening following a 3-d exposureto bisphenol A (600 and 800 mg/kg) by oral gavage, although uterine weights were significantlyincreased in the same animals. In the same study when bisphenol A was administered by s.c.injection, 4 of 7 (600 mg/kg) and 3 of 8 (800 mg/kg) animals displayed early vaginal opening.Thus, although uterotrophic responses were similar following oral or s.c. exposures to bisphenolA, only the s.c. route altered the age at vaginal opening.While the uterotrophic assay has many strengths, the method is limited in that it can only detectestrogenic or antiestrogenic activity. The female pubertal protocol, for example, is capable ofdetecting a broader range of mechanisms of action. In addition, there are some technical caveatsthat may avert false negatives: Guidelines for the dose selection for the uterotrophic assay must bedeveloped to ensure that test chemicals with short-term estrogenic activity will not be missed ifthe endpoints are measured 24 h after the last dose. Several reports have demonstrated that thelength of the uterotrophic response following exposure to some chemicals is much shorter than forothers. Thus, the dosing route and timing following exposure may dictate whether a significantchange in uterine weight may still be detected at necropsy. 79,85,96,97 A comparison of the uterotrophicresponses 6 and 24 h following oral exposures to 4-tert-octylphenol or bisphenol A showedsignificant increases in uterine weights at the 6 h but not at the 24 h time point. 89 Yamasaki et al. 98also reported a higher increase in uterine weight at 6 h as compared with the 24 h time point© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 505following the last of three daily s.c. injections of bisphenol A. However, in the later study theincreased uterine weight at 24 h was still significantly different from the control. The higher dosesof bisphenol A used in the OECD validation studies (e.g., 600 mg/kg by oral gavage and 300 mg/kgfor s.c. injection) did result in significant increases in uterine weight 24 h following the last dose. 91Thus, it is imperative that the final protocol for the uterotrophic assay include dose selection criteriato allow a sufficient dose of the test chemical for each treatment route that will enable the detectionof estrogenic activity at the time of necropsy.B. Hershberger AssayThe Hershberger assay is an in vivo short-term test similar in concept to the rodent uterotrophicassay discussed above. Both assays measure changes in specific tissues that normally respond toendogenous steroid hormones. The assay conditions are designed to achieve low endogenoushormone levels and examine target tissues that are highly responsive to administration of exogenoushormones. The Hershberger assay provides a method to determine whether a compound possessesandrogenic or antiandrogenic activity. It involves weighing tissues in castrated, sexually immaturemale rats following exposure to a test chemical with or without androgen supplementation. 99 Thebiological relevance of the assay is based upon the fact that accessory sex tissues and glands dependon androgen stimulation to gain and maintain weight during and after puberty. 100The Hershberger assay may also present the opportunity to detect inhibitors of 5-α-reductaseby taking advantage of a differential response among the accessory sex glands and tissues. Testosteroneis converted by 5-α-reductase to DHT, which has significantly higher binding affinity thantestosterone for the androgen receptor. The levator ani and bulbocavernosus (LABC) muscles lackthe enzyme 5-α-reductase, while in other tissues the enzyme is present (e.g., ventral prostate,seminal vesicles, and coagulating glands). It is plausible then that tissues with 5-α-reductase activityare more responsive to compounds that are metabolized to DHT than tissues without the enzyme.Acknowledging the substantial amount of historical data and the long standing use of theHershberger assay for the detection of androgens and antiandrogens, 99–104 the EDSTAC recommendedthat the Hershberger be a component of the TIS battery. As with the uterotrophic assay,the OECD agreed to assist the OSCP, U.S. EPA, by directing the validation process for theHershberger assay. Using an approach similar to that used for the validation of the uterotrophicassay, the OECD designed a series of studies that will be conducted in multiple laboratoriesthroughout the world. Phase I was designed to evaluate the standardized protocol, using a referenceandrogen and an antiandrogen. Phase II of the process will evaluate the ability of the assay to detectother androgens, several weak antiandrogens, and a 5-α-reductase inhibitor. Estimates of intra- andinterlaboratory variance will be determined throughout the validation process.The protocol being evaluated currently by the OECD uses castrate, sexually immature malerats. The animals are castrated after peripuberty (generally occurs after postnatal day [PND] 42)by removing both the testes and epididymides. In most rat strains, such as the Sprague-Dawley,Long-Evans, or Wistar, peripuberty is expected to take place at approximately 6 weeks of age,within an average age range of 5 to 7 weeks. Peripuberty is marked by prepuce separation fromthe glans penis (GP). Experimentally, preputial separation is necessary so that the GP can beproperly dissected and accurately weighed. At the peripubertal stage of sexual development, theGP and other androgen-dependent sex accessory tissues are sensitive to androgens, having bothandrogen receptors and the appropriate steroidogenic enzymes. The advantage of using a rodentof this age is that the sex accessory tissues have a high sensitivity and small relative weight, whichminimizes variation in responses between individual animals. In addition, castration removes theinhibitory feedback of testosterone on the hypothalamic-pituitary axis, resulting in a characteristicincrease in the concentration of LH. Depending upon the action of the test substance in thehypothalamus, the pattern of LH secretion with increasing doses of the test substance may be altered.© 2006 by Taylor & Francis Group, LLC

506 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFollowing a minimum of 7 days after castration (but no later than PND 60 because the age atnecropsy should not be greater than PND 70), the males are assigned to treatment groups (n = 6),housed 2 to 3 per cage, and dosed by oral gavage for 10 consecutive days with one of severaldosages of the test chemical. For detection of androgenic activity, only the test chemical isadministered. To detect antiandrogenic activity, the test chemical is administered with and withoutthe reference androgen agonist, testosterone propionate (TP) 0.1 mg/rat/d or 0.4 mg/kg/d, s.c. Themales are killed approximately 24 hours later on the 11th day. The five sex accessory tissues, theventral prostrate (VP), seminal vesicles with coagulating glands (SV), LABC muscles, Cowper’sgland (or bulbourethral gland, COW), and GP are carefully dissected, trimmed of excess adheringtissue and fat, and weighed to the nearest 0.1 mg. Each tissue is handled with particular care toavoid loss of fluids or desiccation. In addition to the sex accessory tissues, it is useful to includeweights of the liver, adrenals, and kidney, as they may provide supplementary information aboutthe systemic toxicity, target organs, and other effects of the test substance. For example, possibleinduction of testosterone metabolizing enzymes in the liver or an impact on adrenal steroidogenesiscould occur with exposure to some test chemicals. Optional endpoints, such as the measurementof serum testosterone and LH concentrations, may also provide possible insight into metabolismchanges and hypothalamic responses. Considerations of the assessment of thyroid function (serumtetraiodothyronine [T 4 ], tri-iodothyronine [T 3 ], and thyroid-stimulating hormone [TSH]; and thyroidweight and histology) and adrenal steroidogenesis also seem reasonable as optional endpoints.If the evaluation of each chemical requires necropsy of more animals than is reasonable for asingle day, necropsy may be continued over two consecutive days. In this case, the work could bedivided so that necropsy of three animals per treatment per day (one cage) takes place on the firstday. Statistical comparisons in individual laboratories will be made for the different sex accessorytissues by analysis of variance. For an androgen agonist, the mean of each variable at each doseof the test chemical will be compared with the vehicle control. A statistically significant increasein tissue weight will be considered a positive androgen agonist result. For an androgen antagonist,the means from the test chemical coadministered with TP will be compared with the TP-onlycontrol. A statistically significant decrease in tissue weight will be considered a positive antagonistresult. If more than one set of comparisons is required, all comparisons will be conducted separatelyfor each test group against its control.Several issues should be considered to adequately interpret the results from the Hershbergerassay. Differences in body weight may be a source of variability in the weight of tissues of interest(especially the liver) within and among groups of animals. This variability could potentially reducethe assay sensitivity and possibly lead to false positives or false negatives. Therefore, variations inbody weight should be both experimentally and statistically controlled. The statistical analysisshould be conducted both with and without body weight as a covariate. Because toxicity may alsoaffect body weight, the body weight on the first day of dosing, as opposed to the body weight atnecropsy, should be used as the covariate in these cases.Experimental control of body weight is accomplished in two steps. The first step involvesselection of animals with relatively small variation in body weight for the study cohort from alarger population of animals. Use of unusually small or large animals should be avoided. Areasonable level of body weight variation within the study cohort can be tolerated (i.e., 20% of themean body weight for the cohort population, e.g., 175 ± 35 g). The second step involves theassignment of animals to different treatment groups (n = 6) by a randomized complete blockapproach. Under this approach, animals are randomly assigned to treatment groups so that eachgroup’s body weights at the beginning of the study have the same mean and standard deviation.Phase 1 of the OECD validation process has recently been completed and summarized. 105 Inthese studies, the androgenic response to TP and the antiandrogenic effects of flutamide (FLU)were evaluated in 17 laboratories. Data from these studies showed that the assay was robust andreproducible across several laboratories, even if the participating laboratory used a variation of the© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 507recommended protocol (e.g., strain of rat, diet, modest differences in the age of castration). Alllaboratories and all protocols were successful in detecting increases in the weights of the accessorysex organs and tissues in response to TP and in detecting the antiandrogenic effects of FLU. Therewas also good agreement among laboratories with regard to the dose responses obtained.Phase II of the validation process, currently in progress, will evaluate the performance of thelaboratories in using the Hershberger protocol to detect two additional agonists (methyl testosteroneand trenbolone), weak AR antagonists (vinclozolin, procymidone, linuron, p,p´-DDE), and a5-α-reductase inhibitor (finasteride). Each chemical will be examined in three to five laboratoriesto determine reproducibility and to provide estimates of intra- and interlaboratory variation. TheOECD will use data from these studies to set performance criteria and to evaluate whether it isplausible to use the assay as a component of the TIS battery.Finally, the strengths and limitations of the Hershberger assay should be addressed whileevaluating the method for use in the TIS battery. As was noted for the uterotrophic assay, theHershberger assay is supported by a wealth of historical data. The biological relevance is clearlydocumented and supported by the physiological dependence of the sex accessory glands and tissueson androgens during puberty. Phase I of the OECD validation process has demonstrated the robustnature of the assay in the ability of multiple laboratories to replicate the results, even with variationsin animal strain. One disadvantage of the Hershberger assay is that it requires surgical castration,a potentially stressful procedure that requires technical skill. Another potential limitation of theassay is that although the androgens play a predominant role in the growth and maintenance ofmale reproductive structures, several other factors can influence organ weights. These factors includethe thyroid and growth hormones, prolactin, estrogens, and EGF, 106 and could possibly lead to falsepositives. The fact that the Hershberger assay allows for the evaluation of optional adjunct measuresthat can facilitate the identification of the mechanism of action may help to alleviate this problem.C. Male and Female Pubertal ProtocolsTwo in vivo assays currently undergoing validation for the TIS battery are the male and femalepubertal protocols. These protocols provide methods for evaluating the effects of chemicals onpubertal development and thyroid function in the rat through alterations in the estrogen, androgen,or thyroid hormone systems. Additionally, the protocols are able to detect steroid-associated toxicinsult and allow for an evaluation of additional apical effects on reproductive function, such ashypothalamic alterations, that may not be associated with an alteration in steroidogenesis or asteroid-receptor mechanism. Thus, while the pubertal protocols are less mechanistically targetedthan the uterotrophic and Hershberger assays, they do allow for a broader casting of the endocrinedisruptor net in the TIS battery.The 21-day dosing period used in the female pubertal protocol encompasses the period ofdevelopment in the rat during which the female brain begins to respond to the positive feedbackof estrogen, resulting in the occurrence of the first estrous cycle and ovulation. The age of vaginalpatency (vaginal opening) is correlated with the first estrus and provides a noninvasive endpoint tomonitor the onset of puberty. Since pubertal development is dependent upon the presence ofestrogen, this assay is capable of detecting chemicals with estrogenic or antiestrogenic (estrogenreceptor or steroid-enzyme mediated) activity, as well as chemicals that alter pubertal developmentvia changes in LH, follicle stimulating hormone (FSH), prolactin, or growth hormone secretion,or that alter hypothalamic neurotransmitter function. Finally, endpoints are included in the protocolto detect alterations in thyroid hormone homeostasis. Goldman et al. 107 have previously describedthis protocol and reviewed in detail the historical data that support the biological relevance of theendpoints (Figure 11.6). In brief, the assay is conducted using weanling Sprague-Dawley or Long-Evans female rats (culled to 8 to 10 per litter on PND 3). Females are gavaged with the test chemicalin corn oil (2.5 to 5.0 ml/kg) between 0700 and 0900 (lights 14:10, on 0500 h) from PND 22 to© 2006 by Taylor & Francis Group, LLC

508 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION3 or 4CullpupsAge (Postnatal Day)21 25 30 35 40 45VOPPSFemales: Dose, Examine for VO, Daily BWWeanAssign to Trt.Males: Dose, Examine for PPS, Daily BWFemaleNecropsyTissue wts.T4, TSHHistology50 53MaleNecropsyTissue wts.T4, TSHHistologyFigure 11.6Overview of the male and female protocols for the assessment of pubertal development andthyroid function. Rats are weaned on PND 21, ranked by body weight and litter, and randomlyassigned to treatment groups such that the mean body weights and variances are approximatelyequal for all groups. Females are dosed from PND 22 to 42, while males are dosed from PND23 to 53. Endpoints of pubertal development are evaluated throughout the dosing period. Animalsare killed on PND 42 (females) or 53 (males). At necropsy, tissues are weighed and serum isused for hormone assay.42, with 15 females per treatment group. Required endpoints include growth, age at vaginal openingwith vaginal cytology until necropsy, serum T4 and TSH, liver, kidney, pituitary, adrenal, uterine(with and without fluid) and ovarian weights, and thyroid histology. Optional endpoints includeserum T 3 , estradiol, prolactin, thyroid weight, systemic and vaginal tissue histology, ex vivo ovarianand pituitary hormone production, hypothalamic neurotransmitter concentrations, and extension ofestrous cycle length.The male pubertal protocol, an assay recommended as an alternate method by EDSTAC, isconducted over a 31-d period, during which a number of pubertal indices are measured. This assayis capable of detecting chemicals with antithyroid, estrogenic, androgenic, or antiandrogenic (ARor steroidogenic enzyme mediated) activity and agents that alter pubertal development via changesin FSH, LH, prolactin, growth hormone, or hypothalamic function. Weanling Sprague-Dawley orLong-Evans rats, housed two to three per cage, are dosed by oral gavage from PND 23 to 53. 108Required endpoints include growth; age at preputial separation; serum T4 and TSH; seminal vesicleplus coagulating gland (with and without fluid), ventral prostate, and LABC weights; epididymaland testis weights and histology; and thyroid histology. Optional endpoints include serum T 3 ,testosterone, estradiol, LH, prolactin, and liver, kidney, adrenal, and pituitary weights, ex vivo testisand pituitary hormone production, and hypothalamic neurotransmitter concentrations.From a historical perspective, the male and female pubertal protocols use a wealth of robustand sensitive endpoints capable of detecting a broad variety of endocrine disrupting chemicals. 107,108However, prior to use in the U.S EPA’s EDSP, the protocols will undergo additional testing todocument the biological relevance of the endpoints, as well as the ability to produce data that areconsistent within a single laboratory and across different laboratories for a broad range of chemicals.In addition, several technical issues are currently being addressed to facilitate the successful useof either of the protocols, to reduce intralaboratory variability, and to interpret the data correctly.For example, there have been reports that some laboratory rodent dietary formulations containlevels of phytoestrogens that are sufficient to induce alterations in uterine weight and histology. 109,110As a result, there is some debate about the influence of the choice of diet on the pubertal endpoints,and whether a low phytoestrogen content in commercially prepared dietary formulations maypresent a confounding factor in the interpretation of test data. There are studies demonstrating noneonatal or pubertal alterations in the offspring of dams maintained on a phytoestrogen containingdiet, 111 and the validation studies for the uterotrophic assay suggest that phytoestrogen contents ofless than 325 to 350 µg/g total genistein equivalents did not impair assay responsiveness. 93 However,© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 509some consideration should be given to the selection of a diet with minimal or undetectable levelsof estrogenic activity.Another technical concern that has generated debate is whether a reduction in body weight gainduring the study might have a confounding effect on the endpoints associated with pubertaldevelopment. While it is known that drastic reductions (greater than 25% to 50% as compared withcontrols) 112–115 can influence pubertal development, body weight decrements of 10% or less do notappear to cause significant changes in vaginal opening (VO) or preputial separation (PPS). 116–118Thus, the dose selection criteria for the pubertal studies will remain at or just below the maximumtolerated dose (MTD), which is typically defined as a 10% lower body weight (as compared withthe control) by the end of the dosing period. While it is generally recognized that doses of a testchemical at or below this level will have no adverse systemic toxicity that would interfere with theendpoint measures used in these protocols, this will be thoroughly evaluated during the validationprocess. Additionally, the fact that estrogenic compounds can decrease appetite and food consumptionshould be noted, and a drop in body weight following dosing with an estrogenic chemicalshould not automatically be considered indicative of general systemic toxicity. Differences in themagnitude of body weight decrement may differ by animal strain, a factor that should be consideredwhen using an MTD that was determined in another strain.The final pubertal protocols will likely contain recommendations that will help to minimizevariability in the hormone and tissue weight endpoints. For example, variability in hormoneconcentrations can be reduced by carefully monitoring the time of day and the method by whichthe animals are euthanized. Carbon dioxide can be used as the method of euthanasia if only thethyroid hormones and TSH are to be measured, as a stress-related change in thyroid hormone doesnot generally occur in a short amount of time. 119 However, if additional endocrine endpoints, suchas levels of the stress-responsive hormone, prolactin, are to be measured, decapitation would berecommended. In all cases, efforts should be made to prevent any stressors that might occur priorto euthanasia (animal transfer, cage changes, excessive noise) to avoid unwanted variability in thehormone data. In addition, because of the circadian rhythms in serum pituitary and steroid hormonesecretions, necropsy should be performed within a narrow window of time following the last dose,although no time of day is specified for the necropsy of either females or males. For example, thereis a circadian rhythm in the secretion of TSH in the rat, with a zenith at midday 120–122 and minimumsecretion at the start of the dark period. Thus, a restricted period for the necropsy during a periodof basal secretion would assure more consistent results. Finally, although the female pubertalprotocol does not currently recommend necropsy on a specific day of the estrous cycle, it isimportant to understand the endocrine status of the animal at necropsy because serum hormoneconcentrations and the weights of the reproductive organs will vary throughout the estrous cycle.Either synchronizing the necropsy on a specific day of the estrous cycle or at least identifying theday of the cycle on which the females are necropsied will greatly facilitate proper interpretationof the hormone and tissue weight data. For example, uterine histology can be used to determinechanges in uterine epithelium that correlate to changes in serum steroid and gonadotropin concentrations,123 as is the uterine weight for detection of estrogenic compounds. 124,125 However, withoutconsidering the hormonal profile and the day of the estrous cycle, the use of these data is questionable.This is also true for ovarian histological assessments and the measurement of any pituitarypeptide hormones, such as prolactin or luteinizing hormone.Although both the male and female pubertal protocols are undergoing the validation processfor the U.S. EPA’s EDSP, only one of the protocols will likely be included in the final TIS battery.While both protocols are technically simple and provide numerous sensitive endpoints that havebeen used for decades to detect endocrine alterations, each protocol possesses its own unique setof strengths and weaknesses when used as part of the TIS battery. The endpoints in the femalepubertal protocol are quite sensitive to estrogens and antiestrogens 126,127 but are not as useful for© 2006 by Taylor & Francis Group, LLC

510 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthe detection of androgenic or antiandrogenic activity as compared with the endpoints of the malepubertal protocol. For example, dietary administration of nonylphenol affects VO but not PPS, 128and methoxychlor accelerates VO at dosage levels nearly fivefold lower than those needed to delayPPS in the male. 128,129 The female pubertal protocol would also provide a better screen for chemicalsthat inhibit aromatase activity. Marty et al. 117 detected a delay in the age of VO with ketoconazole(100 mg/kg) and the aromatase inhibitor fadrozole (0.6 and 6 mg/kg), using the female pubertalprotocol. Replication of results between laboratories has also been well documented for the pubertalendpoints. 118,130,131 Conversely, the male pubertal protocol is preferable for detecting androgenicand antiandrogenic activity. Although the Hershberger assay is also a sensitive assay for the detectionof chemicals that alter androgen receptor function, the male pubertal assay more comprehensivelyassesses the endocrine system. The male pubertal protocol has been successfully used to identifychemicals that affect the hypothalamic-pituitary axis, 132–134 inhibit steroid hormone synthesis, 135alter androgen receptor function, 136 or alter thyroid hormone homeostasis. 137–139 The effects ofseveral toxicants on the timing of PPS have been replicated successfully by several laboratories.For example, the ability of vinclozolin to delay PPS at 100 mg/kg/d has been demonstrated in twolaboratories by different investigators. 136,140 In addition, several laboratories using methoxychlortreatment have also delayed PPS at similar dosage levels. 129,141 A limitation of the male protocolis that it requires a longer dosing period than either the Hershberger assay or female pubertalprotocol. Also, additional in vivo and in vitro tests would be required to identify the specificmechanism of action following a delay in PPS and altered sex accessory gland weights to determinewhether the effect is mediated through an antiandrogen or a disruption of hypothalamic-pituitaryfunction. 108,142D. In Utero-Lactational ExposureThe EDSTAC believed that the recommended T1S battery, if validated, would have the necessarybreadth and depth to detect any currently known disruptors of estrogen, androgen, and thyroidhormone homeostasis. There was concern, however, that chemical substances or mixtures couldproduce effects from prenatal or prehatch exposure that would not be detected from pubertal oradult exposure. 3 Furthermore, there were differing views in the EDSTAC about whether there isscientific evidence of known endocrine disruptors or reproductive toxicants that can affect theprenatal stage of development without affecting the adult or prematuration stages, and whethereffective doses and affected endpoints may differ among the three life stages.The EDSTAC therefore recommended that the U.S. EPA take affirmative steps, in collaborationwith industry and other interested parties, to develop a protocol for a full life cycle (i.e., withembryonic exposure and evaluation of the adult offspring) developmental exposure screening assaythat can be subjected to standardization and validation. The EDSTAC believed such an assay mustinvolve prenatal exposure and retention of offspring through puberty to adulthood and providestructural, functional, and reproductive assessment. Furthermore, the EDSTAC recommended thatif such an assay was identified, standardized, and validated, the decision to include it in the finalTIS battery would involve an evaluation of its potential to replace one or more of the originallyrecommended TIS assays and the overall impact on the cost effectiveness of the TIS battery. Theproposed protocol has been identified by the EPA as the in utero/lactational exposure testing protocol(Figure 11.7) and has been assigned for development under the EDSP. 143 The objective of thisbioassay is to detect effects mediated by alterations in the estrogen, androgen, and thyroid-signalingpathways as a consequence of exposure during pre- and postnatal development in the laboratoryrat. The treatment paradigm allows for an evaluation of effects on organogenesis, sexual differentiation,and puberty. The U.S. EPA considered at least four variations of the in utero/lactationalexposure protocol (Table 11.5). After deliberation, it was agreed that protocol D would be examinedfurther for standardization. In this protocol, dosing of the dam begins on GD 6 and treatment is© 2006 by Taylor & Francis Group, LLC

F0 femalesFigure 11.7Mgd0gd6GPoral gavagepnd0AGDALOverview of the in utero/lactational exposure protocol.123WN F0 femalespnd21F1; uterotrophicCohort; 1 female/litterF1; female pubertalCohort; 4 females/litterF1; male pubertalCohort; 4 males/littersc injectionoral gavageoral gavageF0 dams - serum T4, TSH on pnd 21 (aorta)pnd 21–24; NVPpnd 21– 42;VP, FEAGD on pnd 21pnd 21–70;PPSAGD on pnd 21pnd 4standardize littersto 5 females and4 males/littergross pathologyon cullsKey:M = matingRN = examination for retained nipples in F1 males and femalesG = gestationW = wean (pnd 21)gd = gestational dayN = necropsyP = parturitionVP = acquisition of vaginal patency (females)pnd = postnatal dayPPS = acquisition of preputial separation (males)AGD = anogenital distanceFE = first estrusL = lactationUVD = urethral vaginal distanceA = examination in for areolae on pnd 12–13direct exposure to F0 dams and postweaning F1 offspring6 hrs. past last dose on pnd 24; body,ovary, and uterine (wet) weights; serumE2, T4, and TSH; uterine histology(begin pnd 22), vaginal smear and N onpnd 42; RN, organ weights, grosspathology, T4, T3, TSH, AGD, UVD,thyroid and ovarian histology(begin on pnd 23), N pnd 70, RN, organweights, serum T4, TSH, histology:testes, epididymides, thyroid, AGD,spermatid head count, epididymalspermTHE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 511© 2006 by Taylor & Francis Group, LLC

512 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 11.5Comparison of the parameters included in protocols A, B, C, and DProtocolParameter A B C DMaternal Data (F 0 )Dosing Period GD 6–PND 21 GD 6–PND 21 GD 6–PND 21 GD 6–PND 21Body weight X X X XFood consumption X X X XGestational length X X X XMean litter size X X X XMating index X — — —Fertility index X — — —Preimplantation loss X X X XClinical signs and behavior X X X XGross necropsy (PND 21) X X X XOrgan weights X X X XHistopathology X X X XHormone concentrations — X X XOffspring Data (F 1 )Dosing Period (preweaning) GD 6-PND 21 GD 6-PND 21 GD 6-PND 18 GD 6-PND 21Dosing Period (postweaning) — — PND 18-21 (F) PND 21-24 (F)PND 18-42 (F) PND 21-42 (F)PND 18-52 (M) PND 21-60 (M)Litter size X X X XViability X X X XClinical signs X X X XBody weight X X X XSex ratio X X X XGestational index X — — —Life birth index X — — —Survival index (preweaning) X X X XLactation index X X X XAnogenital distance — X X XPND-4 necropsy — X X XRetained nipples — X X (males only) X (male and female)Developmental LandmarksPinna detachment X — — —Eye opening X — — —Vaginal patency X X X XPreputial separation X X X XFunctional AssessmentEye examination X — — —Auditory function X — — —Motor activity X — — —Learning and memory X — — —Uterotrophic assay — — X (PND 21) X (PND 24)Male weanling necropsy — X X —Female pubertal evaluation — X (undosed) X (direct dose) X (direct dose)Male pubertal evaluation — X (undosed) X (direct dose) X (direct dose)F 1 gross necropsyX (16 weeksof age)X (at puberty) X (PND 21or puberty)X (PND 24or puberty)Organ weights X X X XHistopathology X X X XHormone concentrations — X X XF 1 mating X — — —© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 513Table 11.5Comparison of the parameters included in protocols A, B, C, and D (continued)ProtocolParameter A B C DF 2 evaluation (PND 4) X — —Study duration a 21 weeks 12 weeks 12 weeks 12 weeksCost $350,000 b $300,000 c $350,000 c $350,000 cNote: GD = gestation day, PND = postnatal day, X = parameter is included in the protocol, — = parameter isnot included in the protocol, M = male, F = female.aStudy duration includes quarantine, and assumes in-house bred animals.bStudy cost does not include hormone assays, but includes chemistry and histopathology costs.cStudy cost includes all chemistry, histopathology, and hormone assay costs.continued until PND 21. The dam is monitored for clinical signs daily throughout the treatmentperiod. Body weight and food consumption are monitored during this time. Beginning on PND 20,the female is also observed twice daily for evidence of littering. On PND 21 (day of weaning), thedam is killed and blood is collected for subsequent T 4 and TSH determinations. The dam is thenexamined for gross morphological changes, and the uterus is examined for implantation scars,which are counted and recorded.For the F 1 progeny, the in utero/lactational protocol requires the following: At birth, the pupsare counted, sexed, and examined externally. Each live pup is then weighed individually and itsanogenital distance is measured. On PND 4, the size of each litter is adjusted to nine pups (5females and 4 males). Gross pathology is performed on the culls. There should be a maximum of10 litters per dosage group. Pups are then weighed and examined daily. The presence or absenceof nipples and areolae on the ventrum is recorded for all F 1 male and female offspring on PND11-13.On PND21, the F 1 offspring are then divided into three different treatment cohorts, whichinclude:• The uterotrophic cohort, in which one female from each litter is selected and dosed (s.c.) fromPND 22 to 24. On PND 24, the female is necropsied 6 h after the last dose and the body weight,ovarian weight, and uterine weight (wet) are recorded. Also, blood is collected for subsequentserum estradiol, T 4 , and TSH determinations.• The F 1 female pubertal cohort, comprising two dosed and two nondosed females from each litter.Dosing (oral gavage) begins on PND 21 and continues until PND 42 or 43. During this time thefemale is examined for vaginal patency, and her body weight is recorded. Clinical observationsare made twice daily. Beginning on the day of VO, daily vaginal smears are obtained to determinethe first estrous cycle and whether vaginal cycles are present prior to necropsy.• The F 1 male pubertal cohort, composed of two dosed and two nondosed males from each litter.Dosing (oral gavage) begins on PND 21 and continues until PND 70. Body weights are recordeddaily, and clinical observations are made twice daily. Beginning on PND 23, each F 1 study maleis examined daily for PPS.The in utero/lactational protocol lists a number of specific endpoints that must be carefullyexamined at necropsy for each cohort (Table 11.5). In using a developing system as the basis forthe test, it is understood that modes of action other than those of the estrogen, androgen, andthyroid-signaling pathways may be involved in the induction of toxicity. As such, any observedeffects must be interpreted in light of the overall weight of the evidence as to whether they areendocrine dependent. It does not appear that this assay would be suitable as an initial screen forendocrine disrupting chemicals because of the number of animals required and the length of theexposure. However, the assay might be useful as a bridge between the TIS battery and the Tier IItesting assays (e.g., Tier 1.5), or as an alternative to the standard multigenerational tests in somelimited cases.© 2006 by Taylor & Francis Group, LLC

514 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIV. CONSIDERATIONS FOR EVALUATIONOF DATA FROM THE TIS BATTERYWithin the conceptional framework of the EDSTAC’s recommendations for the EDSP, the assaysin the TIS battery were intended to identify chemicals capable of altering estrogen, androgen, andthyroid hormone systems. Criteria for selection of the assays to be incorporated in the TIS batteryincluded the use of assays that would (1) maximize sensitivity to minimize false negatives, (2)include a range of organisms representing differences in metabolism, (3) detect a range of modesof action and of taxonomic groups, and (4) include enough diversity among the endpoints to permitweight-of-evidence conclusions. Thus, when using the TIS battery as a whole, the EDSTAC believedthat the screening battery would possess the necessary breadth and depth to detect all currentlyknown modes of action for endocrine disrupting chemicals. Those chemicals producing significantchanges in the endpoints in both in vitro and in vivo TIS assays would be further evaluated by TierII testing assays to characterize the dose response, relevance of the exposure route, sensitive lifestages, and hazard potential.Following the completion of the validation studies for the TIS assays, the data will be reviewedby a FIFRA Science Advisory Panel (SAP), which will provide recommendations for the selectionof the TIS assays that will be used to screen the first set of chemicals. 144 It is likely that the SAPwill also recommend an approach for evaluating the data from the TIS battery, and the scientificbasis for moving forward with Tier II testing. In cases where there may be equivocal results, itmay be prudent to replicate certain TIS assays before moving forward to the Tier II testing. Grayet al. 145 have suggested that replication of some TIS assays in such cases may almost entirelyeliminate statistical false positives and minimize animal use by not moving directly to a multigenerationalstudy. Alternatively, the in utero/lactational protocol might be employed as a Tier 1.5study to provide additional data prior to engaging in a more extensive multigenerational study.Finally, the concept of tailoring the Tier II tests using results of the TIS battery will be considered.For example, if a test chemical produces androgenic or antiandrogenic effects in the TIS, thenendpoints such as anogenital distance at birth and nipple and areola retention in infant rats shouldbe included in the mammalian multigenerational test because these are sensitive, permanent biomarkersthat are highly correlated with malformations and reproductive organ weight changes laterin life. 145 Moreover, alterations in estrogen receptor–related assays or altered steroid hormonesynthesis might lead one to include different endpoints in the Tier II tests.Although the specific assays under consideration for the Tier II testing battery are not formallyreviewed in this chapter, the EDSTAC used the following criteria for assay selection: assays that(1) include the most sensitive developmental life stage, (2) identify the specific hazard caused byexposure to the chemical and the dose response, and (3) cover a range of taxa. As such, the EDSTACrecommended a two-generation mammalian reproductive toxicity study, an avian reproduction test,fish and mysid life cycle tests, and an amphibian development and reproduction test as the initialcandidates for the Tier II testing assays. While these assays were designed to work as a whole andto provide dose response data for any observed adverse effects, any mechanistic data obtained fromthe TIS battery will likely also be used to support the potential relevance to humans and wildlifeduring hazard identification and risk assessment.The EPA updated its standard mammalian two-generation test in 1998 and included newendocrine sensitive endpoints that would be of value for EDSP Tier II testing. 146 More recently theagency has begun evaluating the utility of extending the observation period of the F 1 generationthrough puberty and adulthood, as compared with the standard protocol, to determine whetheradditional information could be obtained that would better detect endocrine disrupting chemicals.This study will use two known antiandrogens, dibutylphthalate and vinclozolin, and involve observationof male offspring survival, growth, and development through lactation, weaning, puberty,and adulthood following maternal exposure from GD 6 to PND 20. Secondary hypotheses to be© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 515tested include whether some effects might be detected more easily in the adult as compared withthe weanling, and if increasing the number of males observed at either age would be beneficial.Data from these studies may ultimately increase the sensitivity of the mammalian multigenerationtest for the detection of endocrine disrupting chemicals.Potential alternatives to the standard mammalian two-generation test include several transgenerationalprotocols that have been developed recently. Many of these protocols typically use fewerlitters (7 to 10 per dosage group) but examine all animals in the litter. 147–150 Even though theprotocols employ fewer animals, they provide more statistical power to detect reproductive effectsin the F 1 generation. In addition, the measurement of anogenital distance (AGD) at 1 to 2 days ofage and areolae and nipples at 12 to 13 days of age in the F 1 generation are included. Theseendpoints are not only extremely useful “biomarkers” of an effect on androgen function, but alsothe changes in AGD and nipples can be permanent. Moreover, these endpoints are highly correlatedwith malformations and other alterations in androgen-dependent tissues. 148 Unfortunately neitherof these markers is measured in the F 1 generation under the EPA standard two-generation testguidelines. In this regard, the proposal that the Tier II testing assays should be based upon theresults of TIS remains an important consideration for obtaining the most relative data on thesuspected mode of action. Another transgenerational protocol, described as alternative reproductiontest (ART) 140,151 or alternative mammalian reproductive test (ARMT), 3 has been used to studyxenoestrogens, such as the phthalates, the herbicide, linuron, and the fungicide, fenarimol. Thisprotocol initiates oral exposure by daily gavage of P 0 male and female rats at weaning and continuesthrough puberty, mating, gestation, and lactation. The F 1 are monitored throughout life but notdosed. While this protocol is longer than the in utero/lactational protocol, it may be appropriatefor xenoestrogens and inhibitors of steroidogenesis that affect the P 0 as well as the F 1 generationin the lower dosage groups.V. SUMMARYIn this chapter we have briefly discussed the in vitro and in vivo assays that are currently underconsideration for use in the TIS battery of the EPA’s EDSP. The current TIS battery represents theproduct of considerable scientific evaluation and deliberations by the EDSTAC, EDMVS, OECD,ICCVAM, and the EPA’s EDSP. The strengths and limitations of these assays, as well as their usefulnessfor detecting known modes of action for the disruption of the mammalian endocrine system, have beendescribed. The assays are currently in various stages of the validation process that will document thebiological relevance and quantify intra- and interlaboratory variance as measures of assay reliability.The ultimate goal of the EPA is to provide a set of assays that will be capable of detecting chemicalswith endocrine disrupting activity, while minimizing the occurrence of false negatives.The EPA issued a notice of its approach for selecting the chemicals for the initial round ofscreening. 152 Of the approximately 87,000 chemicals under its regulatory purview, the agency isproposing to select and screen 50 to 100 chemicals drawn from pesticide active ingredients andhigh production volume chemicals, with some having pesticide inert uses. As previously recommendedby the SAP subcommittee, data resulting from the screening process will be evaluated byan independent external panel of experts to determine whether the program needs to be improvedor further optimized. In addition, as the agency moves forward with its EDSP, new methods andstate-of-the-art approaches will continue to be evaluated for possible substitution in the TIS andTier II testing batteries. New approaches that will reduce animal use, such as using QSAR modelsor HTS technology, will continued to be developed and evaluated to identify chemicals likely todisrupt the endocrine system. Thus, the assays used in the EDSP will continue to be refined andupdated as the program evolves toward full implementation.© 2006 by Taylor & Francis Group, LLC

We would like to acknowledge the efforts of our colleagues Dr. Joseph Merenda, Mr. Gary Timm,Mr. Jim Kariya, Mr. Greg Schweer, and the staff in OSCP, EPA, that have the direct responsibilityfor coordinating the implementation of the EDSP; Battelle Northwest Pacific Laboratories forproviding the ER competitive binding curves shown in Figure 11.2; Janet Ferrell for assistancewith the references; and Dr. Audrey Cummings and Dr. Ron Hood for their reviews and helpfulcomments on earlier drafts of the manuscript.This manuscript has been reviewed in accordance with the policy of the National Health andEnvironmental Effects Research Laboratory, U.S. Environmental Protection Agency, and approvedfor publication. Approval does not signify that the contents necessarily reflect the views and policyof the Agency, nor does mention of trade names or commercial products constitute endorsementor recommendation for use.References1. The International Programme on Chemical Safety, http://www.who.int/pcs/.2. U.S. EPA, Endocrine Disruptor Screening Program Web Site, http://www.epa.gov/scipoly/oscpendo/index.htm.3. U.S. EPA, Office of Prevention, Pesticides, and Toxic Substances, EDSTAC Final Report, http://www.epa.gov/scipoly/oscpendo/edspoverview/finalrpt.htm.4. ICCVAM, NICEATM, http://iccvam.niehs.nih.gov/methods/endocrine.htm.5. OECD Endocrine Testing and Assessment. http://www.oecd.org/document/62/0,2340,en_2649_34377_2348606_1_1_1_1,00.html.6. Tsai, M. and O’Malley, B.W., Molecular mechanisms of action of steroid/thyroid receptor superfamilymembers, Ann. Rev. Biochem., 63, 451, 1994.7. Bylund, D.B and Toews, M.L., Radioligand binding methods: practical guide and tips, Am. J. Physiol.,265(5 pt 1), L421, 1993.8. Evans, R.M., The steroid and thyroid hormone receptor superfamily, Science, 240, 889, 1988.9. Nonneman, D.J., Ganjam, V.K., Welshons, W.V., and Vom Saal, F.S., Intrauterine position effects onsteroid metabolism and steroid receptors of reproductive organs in male mice, Biol. Reprod., 47, 723, 1992.10. Tekpetey, F.R. and Amann, R.P., Regional and seasonal differences in concentrations of androgen andestrogen receptors in ram epididymal tissue, Biol. Reprod., 38, 1051, 1988.11. Traish, A.M., Muller, R.E., and Wotiz, H.H., Binding of 7 alpha, 17 alpha-dimethyl-19-nortestosterone(mibolerone) to androgen and progesterone receptors in human and animal tissues, Endocrinology,118, 1327, 198612. Zava, D.T., Landrum, B., Horwitz, K.B., and McGuire, W.L., Androgen receptor assay with [3H]methyltrienolone(R1881) in the presence of progesterone receptors, Endocrinology, 104, 1007, 1979.13. NIEHS, Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM),NTP Interagency Center for the Evaluation of Alternative Toxicological Methods, ICCVAM Evaluationof In Vitro Test Methods for Detecting Potential Endocrine Disruptors: Estrogen Receptor and AndrogenReceptor Binding and Transcriptional Activation Assays. http://iccvam.niehs.nih.gov/methods/endodocs/edfinrpt/edfinrpt.pdf.14. NIEHS, Interagency Center for the Evaluation of Alternative Toxicological Methods (NICETM).Current Status of Test Methods for Detecting Endocrine Disruptors: In Vitro Androgen ReceptorBinding Assays. http://iccvam.niehs.nih.gov/methods/endodocs/final/arbndbrd/arbndall.pdf15. Andersen, H.R., Andersson, A., Arnold, S.F., Autrup, H., Barfoed, M., Beresford, N.A., Bjerregaard,P., Christensen, L.B., Gissel, B., Hummel, R., Jorgensen, E.B., Korsgaard, B., Le Guevel, R., Leffers,H., McLachlan, J., Moller, A., Nielsen, J.B., Olea, N., Oles-Karasko, A., Pakdel, F., Pedersen, K.,Perez, P., Skakkeboek, N.E., Sonnenschein,C., Sots, A.M., Sumpter, J.P., Thorpe, S.M., and Grandjean,P., Comparison of short-term estrogenicity tests for identification of hormone-disrupting chemicals,Environ. Health Perspect., 107, 89, 1999.516 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONACKNOWLEDGMENTS© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 51716. Wong, C., Kelce, W.R., Sar, M., and Wilson, E.M., Androgen receptor antagonist versus agonistactivities of the fungicide vinclozolin relative to hydroxyflutamide, J. Biol. Chem., 270, 1998, 1995.17. Bauer, E.R.S., Daxenberger, A., Petri, T., Sauerwein, H., and Meyer, H.H.D., Characterization of theaffinity of different anabolics and synthetic hormones to the human androgen receptor, human sexhormone binding globulin and to the bovine progestin receptor, APMIS, 108, 838, 2001.18. Lambright, C.L., Ostby, J., Bobseine, K., Wilson, V.S., Hotchkiss, A.K., Mann, P.C., and Gray, L.E.,Jr., Cellular and molecular mechanisms of action of linuron: an antiandrogenic herbicide that producesreproductive malformations in male rats, Toxicol. Sci., 56, 389, 2000.19. Kuiper, G.G., Lemmen, J.G., Carlsson, B., Corton, J.C., Safe, S.H., van der Saag, P.T., van der Burg,B., and Gustafsson, J., Interaction of estrogenic chemicals and phytoestrogens with estrogen receptorb, Endocrinology, 139, 4252, 1998.20. Kyakumoto, S., Sato, N., Nemoto, T., Ohara-Nemoto, Y., and Ota, M., Binding of [ 3 H]methyltrienoloneto androgen receptor in rat liver, Biochim. Biophys. Acta., 800, 214, 1984.21. Murthy, L.R., Johnson, M.P., Rowley, D.R., Young, C.Y., Scardino, P.T., and Tindall, D.J., Characterizationof steroid receptors in human prostate using mibolerone, Prostate, 8, 241, 1986.22. GraphPad Software, Inc., Introduction to Radioligand Binding, http://www.graphpad.com/curvefit/introduction9e.htm23. Cheng, Y. and Prusoff, W.H., Relationship between the inhibition constant (Ki) and the concentrationof inhibitor which causes 50 percent inhibition (I 50 ) of an enzymatic reaction, Biochem. Pharmacol.,22, 3099, 1973.24. Mailman, R.B. and Lawler, C.P., Toxicant-receptor interactions: fundamental principles, in Introductionto Biochemical Toxicology, Hodgson, E. and Levi, P., Eds., Appleton and Lange, Norwalk, CT, 1994, p. 31925. Kelce, W.R., Monosson, E., Gamcsik, M.P., Laws, S.C., and Gray, L.E., Jr., Environmental hormonedisruptors: evidence that vinclozolin developmental toxicity is mediated by antiandrogenic metabolites,Toxicol. Appl. Pharmacol., 126, 276, 1994.26. Segel, I.H., Enzyme Kinetics, John Wiley & Sons, New York, pp. 100–159.27. Gaido, K.W., Leonard, L.S., Lovell, S., Gould, J.C., Babai, D., Portier, C.J., and McDonnell, D.P.,Evaluation of chemicals with endocrine modulating activity in a yeast-based steroid hormone receptorgene transcription assay, Toxicol. Appl. Pharmacol., 143, 205, 1997.28. Lyttle, C.R., Damian-Matsumura, P., et al., Human estrogen receptor regulation in a yeast modelsystem and studies on receptor agonists and antagonists, J. Steroid Biochem. Mol. Biol., 42, 677, 1992.29. Kohno, H., and Gandini, O., et al., Anti-estrogen activity in the yeast transcription system: Estrogenreceptor mediated agonist response, Steroids, 59, 572, 1994.30. Gaido, K.W., Leonard, L.S., et al., Differential interaction of the methoxychlor metabolite 2,2-bis-(p-hydroxyphenyl)-1,1,1-trichloroethane with estrogen receptors alpha and beta, Endocrinology,140(12), 5746, 1999.31. Kemppainen, J.A., Langley E., et al., Distinguishing androgen receptor agonists and antagonists:distinct mechanisms of activation by medroxyprogesterone acetate and dihydrotestosterone, Mol.Endocrinol., 13, 440, 1999.32. Lobaccaro, J.M., Poujol, N., Terouanne, B., Georget, V., Fabre, S., Lumbroso, S., and Sultan, C.,Transcriptional interferences between normal or mutant androgen receptors and the activator protein1 — dissection of the androgen receptor functional domains, Endocrinology, 140, 350, 1999.33. Tremblay, A., Tremblay G.B., et al., EM-800, a novel antiestrogen, acts as a pure antagonist of thetranscriptional functions of estrogen receptors alpha and beta, Endocrinology, 139, 111, 1998.34. Vinggaard, A.M., Joergensen E.C., et al., Rapid and sensitive reporter gene assays for detection ofantiandrogenic and estrogenic effects of environmental chemicals, Toxicol. Appl. Pharmacol., 155,150, 1999.35. Hartig, P.C., Bobseine, K.L., et al., Development of two androgen receptor assays using adenoviraltransduction of MMTV-luc reporter and/or hAR for endocrine screening, Toxicol. Sci., 66, 82, 2002.36. Blankvoort, B.M., de Groene, E.M., et al., Development of an androgen reporter gene assay (AR-LUX)utilizing a human cell line with an endogenously regulated androgen receptor, Annal. Biochem., 298,93, 2001.37. Legler, J., van den Brink, C.E., et al., Development of a stably transfected estrogen receptor-mediatedluciferase reporter gene assay in the human T47D breast cancer cell line, Toxicol. Sci., 48, 55, 1999.© 2006 by Taylor & Francis Group, LLC

518 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION38. Rogers, J.M. and Denison, M.S., Recombinant cell bioassays for endocrine disruptors: developmentof a stably transfected human ovarian cell line for the detection of estrogenic and antiestrogenicchemicals, In Vitro Mol. Toxicol., 13, 67, 2000.39. Terouanne, B., Tahiri, B., et al., A stable prostatic bioluminescent cell line to investigate androgenand antiandrogen affects, Mol. Cell. Endocrinol., 160, 39, 2000.40. Wilson, V.S., Bobseine, K., et al., A novel cell line, MDA,kb2, that stably expresses an androgenandglucocorticoid-responsive reporter for the detection of hormone receptor agonist and antagonist,Toxicol. Sci., 66, 69, 2002.41. Zacharefwski, T.R., Meek, M.D., et al., Examination of the in vitro and in vivo estrogenic activitiesof eight commercial phthalate esters, Toxicol. Sci., 46, 282, 1998.42. Rehnberg, G.L., Linder, R.E., Goldman, J.M., Hein, J.F., McElroy, W.K., and Cooper, R.L., Changesin testicular and serum hormone concentrations in the male rat following treatment with m-dinitrobenzene,Toxicol. Appl. Pharmacol., 95, 255, 1988.43. Gray, L.E., Klinefelter, G., Kelce, W., Laskey, J., Ostby, J., and Ewing, L., Hamster Leydig cells areless sensitive to ethane dimethanesulfonate when compared to rat Leydig cells both in vivo and invitro, Toxicol. Appl. Pharmacol., 130, 248, 1995.44. Wilker, C.E., Welsh, T.H., Jr., Safe, S.H., Narasimha, T.R., and Johnson, L., Human chorionicgonadotropin protects Leydig cell function against 2,3,7,8-tetrachlorodibenzo-p-dioxin in adult rats:role of Leydig cell cytoplasmic volume, Toxicology, 95, 93, 1995.45. Chambon, M., Grizard, G., and Boucher, D., Bromocryptine, a dopamine agonist, directly inhibitstestosterone production by rat Leydig cells, Andrologia, 17, 172, 1985.46. Kan, P.B., Hirst, M.A., and Feldman, D., Inhibition of steroidogenic cytochrome P-450 enzymes inrat testis by ketoconazole and related imidazole anti-fungal drugs, J. Steroid Biochem., 23, 1023, 1985.47. Laskey, J.W. and Phelps, P.V., Effect of cadmium and other metal cations on in vitro Leydig celltestosterone production, Toxicol. Appl. Pharmacol., 108, 296, 1991.48. Guillou, F., Martinat, N., and Combarnous, Y., Rapid in vitro desensitization of the testosteroneresponse in rat Leydig cells by sub-active concentrations of porcine luteinizing hormone, FEBS Lett.,184, 6, 1985.49. Klinefelter, G.R., Laskey, J.W., and Roberts, N.L., In vitro/in vivo effects of ethane dimethanesulfonateon Leydig cells of adult rats, Toxicol. Appl. Pharmacol., 107, 460, 1991.50. Romanelli, F., Fillo, S., Isidori, A., and Conte, D., Stimulatory action of endothelin-1 on rat Leydigcells: involvement of endothelin-A subtype receptor and phospholipase A2-arachidonate metabolismsystem, Life Sci., 61, 557, 1997.51. Mack, S.O., Lorence, M.C., Andersson, S., and Mason, J.I., Expression of cytochrome P-450(17)alpha, 3 beta-hydroxysteroid dehydrogenase/delta 5,4-isomerase, and steroid 5 alpha-reductase in ratH540 Leydig tumor cells, Mol. Cell. Endocrinol., 74R, 11, 1990.52. Stocco, D.M, King, S., and Clark, B.J., Differential effects of dimethylsulfoxide on steroidogenesisin mouse MA-10 and rat R2C Leydig tumor cells, Endocrinology,136, 2993, 1995.53. Nikula, H., Talonpoika, T., Kaleva, M., and Toppari, J., Inhibition of hCG-stimulated steroidogenesisin cultured mouse Leydig tumor cells by bisphenol A and octylphenols, Toxicol. Appl. Pharmacol.,157, 166, 1999.54. Walsh, L.P., Webster, D.R., and Stocco, D.M., Dimethoate inhibits steroidogenesis by disruptingtranscription of the steroidogenic acute regulatory (StAR) gene, J. Endocrinol., 167, 253, 2000.55. Dees, J.H., Gazouli, M., and Papadopoulos, V., Effect of mono-ethylhexyl phthalate on MA-10 Leydigtumor cells, Reprod. Toxicol., 15, 171, 2001.56. Chen, L.Y., Huang, Y.L., Liu, M.Y., Leu, S.F., and Huang, B.M., Effects of amphetamine on steroidogenesisin MA-10 mouse Leydig tumor cells, Life Sci., 72, 1983, 2003.57. Nishi, Y., Yanase, T., Mu, Y., Oba, K., Ichino, I., Saito, M., Nomura, M., Mukasa, C., Okabe, T., Goto,K., Takayanagi, R., Kashimura, Y., Haji, M., and Nawata, H., Establishment and characterization ofa steroidogenic human granulose-like tumor cell line, KGN, that expresses functional follicle-stimulatinghormone receptor, Endocrinology, 142, 437, 2001.58. Logié, A., Boulle, N., Gaston, V., Perin, L., Boudou, P., Le Bouc, Y., and Gicquel, C., Autocrine roleof IGF-II in proliferation of human adrenocortical carcinoma NCI H295R cell line, J. Mol. Endocrinol.,23, 23, 1991.© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 51959. Ohno, S., Shinoda, S., Toyoshima, S., Nakazawa, H., Makino, T., and Nakajin, S., Effects of flavonoidphytochemicals on cortisol production and on activities of steroidogenic enzymes in human adrenocorticalH295R cells, J. Steroid Biochem. Mol. Biol., 80, 355, 2002.60. Harvey, P.W. and Everett, D.J., The adrenal cortex and steroidogenesis as cellular and moleculartargets for toxicity: critical omissions from regulatory endocrine disruptor screening strategies forhuman health?, J. Appl. Toxicol., 23, 81, 2003.61. Sanderson, J.T., Boerma, J., Lansbergen, G.W., and van den Berg, M., Induction and inhibition ofaromatase (CYP19) activity by various classes of pesticides in H295R human adrenocortical carcinomacells, Toxicol. Appl. Pharmacol., 182, 44, 2002.62. Sanderson, J.T., Letcher, R.J., Heneweer, M., Giesy, J.P., and van den Berg, M., Effects of chloro-striazineherbicides and metabolites on aromatase activity in various human cell lines and on vitellogeninproduction in male carp hepatocytes, Environ. Health Perspect., 109, 1027, 2001.63. Kellis, J.T., Jr. and Vickery, L.E., Inhibition of human estrogen synthase (aromatase) by flavones,Science, 225, 1032, 1984.64. Kellis, J.T., Jr., Nesnow, S., and Vickery, L.E., Inhibition of aromatase cytochrome P450 (estrogensynthase) by derivative of alpha-napthoflavone, Biochem. Pharmacol., 35, 2887, 1986.65. Ibrahim, A.R. and Abul-Hajj, Y.J., Aromatase inhibition by flavonoids, J. Steroid Biochem. Mol. Biol.,37, 257, 1990.66. Vinggaard, A.M., Hnida, C., Breinholt, V., and Larsen, J.C., Screening of selected pesticides forinhibition of CYP19 aromatase activity in vitro, Toxicol. In Vitro, 12, 227, 2000.67. Hirsh, K.S., Weaver, D.E., Black, L.J., Falcone, J.F., and MacLusky, N.J., Inhibition of central nervoussystem aromatase activity: a mechanism for fenarimol-induced infertility in the male rat, Toxicol.Appl. Pharm., 91, 235, 1987.68. Kragie, L., Aromatase in primate pregnancy: a review, Endocr. Res., 28, 121, 2002.69. Vinggaard, A.M., Nellemann, C., Dalgaard, M., Jorgensen, E.B., and Andersen, H.R., Antiandrogeniceffects in vitro and in vivo of the fungicide prochloraz, Toxicol. Sci., 69, 344, 2002.70. Lephart, E.D. and Simpson, E.R., Assay of aromatase activity, Methods Enzymol., 206, 477, 1991.71. Letcher, R.J., Van, H.I., Drenth, H.J., Norstrom, R.J., Bergman, A., Safe, S., Pieters, R.M., and vanDen, B.M., Cytotoxicity and aromatase (CYP19) activity modulation by organochlorines in humanplacental JEG-3 and JAR choriocarcinoma cells, Toxicol. Appl. Pharmacol., 160, 10, 1999.72. Doody, K.M., Murry, B.A., and Mason, J.I., The use of rat Leydig tumor (R2C) and human hepatoma(HEPG2) cells to evaluate potential inhibitors of rat and human steroid aromatase, J. Enzyme Inhib.,4, 153, 1990.73. Catalano, S., Marsico, S., Giordano, C., Mauro, L., Rizza P, Panno, M.L., and Ando, S., Leptinenhances, via AP-1, expression of aromatase in MCF-7 cell line, J. Biol. Chem., 279(19), 19908, 2004.74. Sanderson, J.T., Seinen, W., Giesy, J.P., and van den Berg, M., 2-Chloro-s-triazine herbicides inducearomatase (CYP 19) activity in H295R human adrenocortical carcinoma cells; A novel mechanismfor estrogenicity, Toxicol. Sci., 54, 121, 2000.75. Stoker, T.E., Laws, S.C., Guidici, D.L., and Cooper, R.L., The effect of atrazine on puberty in maleWistar rats: an evaluation in the protocol for the assessment of pubertal development and thyroidfunction, Toxicol. Sci., 58, 50, 2000.76. Ojeda, S.R. and Urbanski, H.F., Puberty in the rat, in The Physiology of Reproduction, 2nd Ed.,Knobil, E. and Neill, J.D., Eds., Raven Press, New York, 1999.77. Simpson, E.R., Clyne, C., Rubin, G., Boon, W.C., Robertson, K., Britt, K., Speed, C., and Jones, M.,Aromatase — a brief overview, Annual Rev. Physiol., 64, 93, 2002.78. Miller, W.R. and O’Neill, J., The importance of local synthesis of estrogen within the breast, Steroids,50, 537, 1987.79. Levin, L. and Tyndale, H.H., The quantitative assay of follicle stimulating substances, Endocrinology,21, 619, 1937.80. Dorfman, R.I., Gallagher, T.F., and Koch, F.C., The nature of the estrogenic substance in human maleurine and bull testis, Endocrinology, 19, 33, 1935.81. Bulbring, E. and Burn, J.H., The estimation of oestrin and of male hormone in oily solution, J. Physiol.,85, 320, 1935.© 2006 by Taylor & Francis Group, LLC

520 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION82. Owens, J.W. and Ashby, J., Critical review and evaluation of the uterotrophic bioassay for theidentification of possible estrogen agonists and antagonists: in support of the validation of the OECDuterotrophic protocols for the laboratory rodent, Critical Rev. Tox., 32, 445, 2002.83. Wakeling, A.E. and Bowler, J., Novel antioestrogens without partial agonist activity, J. Steroid.Biochem., 31, 645, 1988.84. MacGregor J.I. and Jordan, V.C., Basic guide to the mechanisms of antiestrogen action, Pharmacol.Rev., 50, 151, 1998.85. Anderson J.N., Clark J.H., and Peck, E.J., Jr., The relationship between nuclear receptor-estrogeninding and uterotrophic responses, Biochem. Biophys. Res. Comm., 48, 1460, 1972.86. Clark, J.H. and Peck, E.J., Jr., Female Sex Steroids, Receptors and Function, Springer-Verlag, Berlin,1979.87. Zacharewski, T.R., Meek, M.D., Clemons, J.H., Wu, Z.F., Fielden, M.R., and Matthews, J.B., Examinationof the in vitro and in vivo estrogenic activities of eight commercial phthalate esters, Toxicol.Sci., 46, 282, 1998.88. Ashby J., Tinwell, J., Pennie, W., Brooks, A., Lefevre, P.A., Beresford, N., and Sumpter, J.P., Partialand weak oestrogenicity of the red wine constituent resveratrol, consideration of its superagonistactivity in MCF-7 cells and its suggested cardiovascular protective effects, J. Appl. Toxicol., 19, 39,1999.89. Laws, S.C., Carey, S.A., Ferrell, J.M., Bodman, G.J., and Cooper, R.L., Estrogenic activity of octylphenol,nonylphenol, bisphenol A and methoxychlor in rats, Toxiocol. Sci., 54, 154, 2000.90. Kanno, J., Onyon, L., Haseman, J., Fenner-Crisp, P., Ashby, J., and Owens, J.W., The OECD programto validate the rat uterotrophic bioassay to screen compounds for in vivo estrogenic responses, Phase1, Environ. Health Perspect., 109, 785, 2001.91. Kanno, J., Onyon, L., Peddada, S., Ashby, J., Jacob, E., and Owens, W., The OECD program tovalidate the rat uterotrophic bioassay, phase two: dose response studies, Environ. Health Perspect.,111(12), 1530, 2003.92. Kanno, J., Onyon L., Peddada, S., Ashby, J., Jacob, E., and Owens, W., The OECD program to validatethe rat uterotrophic bioassay, phase two — coded single dose studies, Environ. Health Perspect.,111(12), 1550, 2003.93. Owens, W., Ashby, J., Odum, J., and Onyon, L., The OECD program to validate the rat uterotrophicbioassay, phase two: dietary phytoestrogen analyses, Environ. Health Perspect., 111(12), 1559, 2003.94. Gould, J.C., Leonard, L.S., Maness, S.C., Wagner, B.L., Conner, K., Zacharewski, T., Safe, S.,McDonnell, D.P., and Gaido, K.W., Bisphenol A interacts with the estrogen receptor alpha in a distinctmanner from estradiol, Mol. Cell Endocrinol., 142, 203, 1998.95. Ashby, J. and Tinwell, J., Uterotrophic activity of bisphenol A in the immature rat, Environ. HealthPerspect. 106, 719, 1998.96. Katzenellenbogen, B.S., Biology and receptor interactions of estriol and estriol derivatives in vitroand in vivo, J. Steroid Biochem., 20, 1033, 1984.97. Jordan, V.C., Mittal, S., Gosden, B., Koch, R., and Lieberman, M.E., Structure-activity relationshipsof estrogens, Environ. Health Perspect., 61, 97, 1985.98. Yamasaki, K., Sawaki, M., Noda, S., and Takatuki, M., Effects of olive, corn, sesame or peanut oilon the body weights and reproductive organ weights of immature male and female rats, Exp. Animals,.50, 173, 2001.99. Hershberger, L., Shipley, E., and Meyer, R., Myotrophic activity of 19-nortestosterone and othersteroids determined by modified levator ani muscle method, Proc. Soc. Exp. Biol. Med., 83, 175, 1953.100. Ojeda, S.R., Andrews, W.W., Advis, J.P., and White, S.S., Recent advances in the endocrinology ofpuberty, Endocr. Rev., 1, 228, 1980.101. Wainman, P. and Shipounoff, G.C., The effects of castration and testosterone propionate on the striatedperineal musculature in the rat, Endocrinology, 29, 975, 1941.102. Eisenberg, E., Gordan, G.S., and Elliott, E.W., Testosterone and tissue respiration of the castrate malerat with a possible test for myotrophic activity, Endocrinology, 45, 113, 1949.103. Korenchevsky, V., The assay of testicular hormone preparations, Biochem. J., 26, 413, 1932.104. Korenchevsky, V. and Dennison, M., The assay of transdehydroandrosterone and its effects on maleand female gonadectomized rats, Biochem. J., 29, 1514, 1935.© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 521105. OECD, Final OECD report of the work towards the validation of the rat Hershberger assay: Phase-1, androgenic response to testosterone propionate and anti-androgenic effects of flutamide.ENV/JM/TG/EDTA(2002)1/REV2, October 2002, Paris.106. Luke, M.C. and Coffey, D.S., Human androgen receptor binding to the androgen response elementof prostate specific antigen, J. Androl., 15, 41, 1994.107. Goldman, J.M., Laws, S.C., Balchak, S.K., Cooper, R.L., and Kavlock, R.J., Endocrine-disruptingchemicals, prepubertal exposures and effects on sexual maturation and thyroid activity in the femalerat. A focus on the EDSTAC recommendations, Crit. Rev. Toxicol., 30, 135, 2000.108. Stoker, T.E., Parks, L.G., Gray, L.E., and Cooper, R.L., Endocrine-disrupting chemicals, prepubertalexposures and effects on sexual maturation and thyroid function in the male rat. A focus on theEDSTAC recommendations, Crit. Rev. Toxicol., 30, 197, 2000.109. Thigpen, J.E., Setchell, K.D., Ahlmark, K.B., Locklear, J., Spahr, T., Caviness, G.F., Goelz, M.F.,Haseman, J.K., Newbold, R.R., and Forsythe, D.B., Phytoestrogen content of purified, open- andclosed-formula laboratory animal diets, Lab. Anim. Sci., 49, 530, 1999.110. Boettger-Tong, H., Murthy, L., Chiappetta, C., Kirkland, J.L., Goodwin, B., Adlercreutz, H., Stancel,G.M., and Makela, S., A case of a laboratory animal feed with high estrogenic activity and its impacton in vivo responses to exogenously administered estrogens, Environ. Health Perspect., 106, 369, 1998.111. Casanova, M., You, L., Gaido, K.W., Archibeque-Engle, S., Janszen, D.B., and Heck, H.A., Developmentaleffects of dietary phytoestrogens in Sprague-Dawley rats and interactions of genistein anddaidzein with rat estrogen receptors and in vitro, Toxicol. Sci., 51, 236, 1999.112. Engelbregt, M.J.T., vanWeissenbruch, M.M., Pop-Snijders, D., Lips, P., and Delemarr-van de Waal,H.A., Body mass index, body composition, and leptin at onset of puberty in male and female ratsafter intrauterine growth retardation, and after early postnatal food restriction, Ped. Res., 50, 474, 2001.113. O’Connor, J.C., Davis, L.G., Frame, S.R., and Cook, J.C., Evaluation of a Tier I screening batteryfor detecting endocrine-active compounds (EACs) using the positive controls testosterone, coumestrol,progesterone, and RU486, Toxicol. Sci., 54, 338, 2000.114. Merry, B.J. and Holehan, A.M., Onset of puberty and duration of fertility in rats fed a restricted diet,J. Reprod. Fert., 57, 253, 1979.115. Kennedy, G.C. and Mitra, J., Body weight and food intake as initiating factors for puberty in the rat,J. Physiol., 166, 408, 1979.116. Cooper, R.L., Stoker, T.E., Ferrell, J., Leffler, K., Bremser, K., and Laws, S.C., Impact of body weightchange on the EDSTAC Tier I male and female pubertal protocols, Toxicologist, 72(S-1), 129, 2003.117. Marty, M.S., Crissman, J.W., and Carney, E.W., Evaluation of the EDSTAC female pubertal assay inCD rats using 17-beta-estradiol, steroid biosynthesis inhibitors, and a thyroid inhibitor, Toxicol. Sci.,52, 269, 1999.118. Laws, S.C., Ferrell, J.M., Stoker, T.E., Schmid, J., and Cooper, R.L., The effect of atrazine on pubertyin female Wistar rats: an evaluation in the protocol for assessing pubertal development and thyroidfunction, Toxicol. Sci., 58, 366, 2000.119. Armario, A., Lopez-Calderon, A., Jolin, T., and Balasch, J., Response of anterior pituitary hormonesto chronic stress. The specificity of adaptation, Neurosci. Biobehav. Rev., 10, 245, 1986.120. Jordan, D., Veisseire, M., Borson-Chazot, F., and Mornex, R., Postnatal development of TRH andTSH rhythms in the rat, Hormone Res., 27, 216, 1987.121. Selgas, L., Arce, A., Esquifino, A.I., and Cardinali, D.P., Twenty-four-hour rhythms of serum ACTH,prolactin, growth hormone, and thyroid-stimulating hormone, and of median-eminence norepinephrine,dopamine, and serotonin, in rats injected with Freund’s adjuvant, Chronobiol. Int., 14, 253, 1997.122. Pallardo, L.F., Pericas, I., and Jolin, T., Thyroid iodine uptake, thyroid iodine secretion and plasmaTSH levels in male rats during the day and night, Acta Endocrinol. (Copenhagen), 83, 517, 1976.123. Spornitz, U.M., Socin, C.D., and Dravid, A.A., Estrus stage determination in rats by means of scanningelectron microscopic images of uterine surface epithelium, Anat. Rec., 254, 116, 1999.124. Bulger, W.H., Feil, V.J., and Kupfer, D., Role of hepatic monooxygenases in generating estrogenicmetabolites from methoxychlor and from its identified contaminants, Mol. Pharmacol., 27, 115, 1985.125. Grunert, G., Porcia, M., and Tchernitchin, A.N., Differential potency of oestradiol-17 beta anddiethylstilboestrol on separate groups of responses in the rat uterus. J. Endocrinol., 110, 103, 1986.© 2006 by Taylor & Francis Group, LLC

522 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION126. Kim, H.S., Shin, J.H., Moon, H.J., Kim, T.S., Kang, I.H., Seok, J.H., Kim, I.Y., Park, K.L., and Han,S.Y., Evaluation of the 20-day pubertal female assay in Sprague-Dawley rats treated with DES,tamoxifen, testosterone, and flutamide, Toxicol. Sci., 67, 52, 2002.127. Kim, H.S., Shin, J.H., Moon, H.J., Kang, I.H., Kim, T.S., Kim, I.Y., Seok, J.H., Pyo, M.Y., and Han,S.Y., Comparative estrogenic effects of p-nonylphenol by 3-day uterotrophic assay and female pubertalonset assay, Reprod. Toxicol,, 16, 259, 2002.128. Chapin, R.E., Delaney, J., Wang, Y., Lanning, L., Davis, B., Collins, B., Mintz, N., and Wolfe, G.,The effects of 4-nonylphenol in rats, a multigeneration reproductive study, Toxicol. Sci., 52, 80, 1999.129. Chapin, R.E., Harris, M.W., Davis, B.J., Ward, S.M., Wilson, R.E., Mauney, M.A., Lockhart, A.C.,Smialowicz, R.J., Moser, V.C., Burka, L.T., and Collins, B.J., Fundam. Appl. Toxicol. 40, 138, 1997.130. Ashby, J., Tinwell, H., Stevens, J., Pastoor, T., and Breckenridge, C.B., The effects of atrazine on thesexual maturation of female rats, Regul. Toxicol. Pharmacol., 35, 468, 2002.131. Gray, L.E., Jr, Ostby, J., Ferrell, J., Rehnberg, G., Liner, R., Cooper, R.L., Goldman, J.M., Slott, V.,and Laskey, J., A dose-response analysis of methoxychlor-induced alterations in reproductive developmentand function in the rat, Fundam. Appl. Toxicol., 12, 92, 1989.132. Hostetter, M.W. and Piacsek, B.E., The effect of prolactin deficiency during sexual maturation in themale rat, Biol. Reprod., 17, 574, 1977.133. Ramaley, J.A. and Phares, C.K., Delay of puberty onset in males due to suppression of growthhormone, Neuroendocrinology, 36, 32, 1983.134. Stoker, T.E., Laws, S.C., Guidici, D.L., and Cooper, R.L., The effect of atrazine on puberty in maleWistar rats, an evaluation in the protocol for the assessment of pubertal development and thyroidfunction, Toxicol. Sci., 58, 50, 2000.135. Marty, M.S., Crissman, J.W., and Carney, E.W., Evaluation of the male pubertal onset assay to detecttestosterone and steroid biosynthesis inhibitors in CD rats, Toxicol. Sci., 60, 285, 2001.136. Monosson, E., Kelce, W.R., Lambright, C., Ostby, J., and Gray, L.E., Jr., Peripubertal exposure to theantiandrogenic fungicide, vinclozolin, delays puberty, inhibits the development of androgen-dependenttissues, and alters androgen receptor function the male rat, Toxicol. Ind. Health, 15, 65, 1999.137. Stoker, T.E., Ferrell, J., Hedge, J.M., Crofton, K.M., Cooper, R.L., and Laws, S.C., Assessment ofDE-71, a commercial polybrominated diphenyl ether (PBDE) mixture, in the EDSP male pubertalprotocol, Toxicologist, 72(S-1), 135, 2003.138. Marty, M.S., Crissman, J.W., and Carney, E.W., Evaluation of the male pubertal assay’s ability todetect thyroid inhibitors and dopaminergic agents, Toxicol. Sci., 60, 63, 2001.139. Yamasaki, K., Tago, Y, Nagai, K., Sawaki, M., Noda, S., and Takatsuki, M., Comparison of toxicitystudies based on the draft protocol for the ‘Enhanced OECD Test Guideline No. 407’ and the researchprotocol of ‘Pubertal Development and Thyroid Function in Immature Male Rats’ with 6-n-propyl-2-thiouracil, Arch. Toxicol., 76, 495, 2002.140. Anderson, S.A., Pearce, S.W., Fail, P.A., McTaggart, B.T., Tyle, R.W., and Gray, L.E. Jr., Validationof the Alternative Reproductive Test Protocol (ART) to assess toxicity of methoxychlor in rats,Toxicologist, 15, 1995.141. Gray, L.E., Jr., Ostby, J., Sigmon, R., Ferrell, J., Rehnberg, G., Linder, R., Cooper, R., Goldman, J.,and Laskey J., The development of a protocol to assess reproductive effect of toxicants in the rat,Reprod. Toxicol., 2, 281, 1988.142. Huhtaniemi, I., Nikula, H., Rannikko, S., and Clayton, R., Regulation of testicular steroidogenesis bygonadotrophin-releasing hormone agonists and antagonists, J. Steroid Biochem., 24, 169, 1986.143. In Utero/Lactational Exposure Testing Protocol, http://www.epa.gov/scipoly/oscpendo/meet-ings/2002/march/inuterolactationprotocol.pdf.144. U.S. EPA, About the Science Advisory Panel, http://www.epa.gov/oscpmont/sap/about.htm.145 Gray, L.E., Jr., Ostby, J., Wilson, V., Lambright, C., Bobseine, K., Hartig, P., Hotchkiss, A., Wolf, C.,Furr, J., Price, M., Parks, L., Cooper, R.L., Stoker, T.E., Laws, S.C., Degitz, S.J., Jensen, K.M., Kahl,M.D., Kahl, J.J., Korte, E.A., Makynen, E.A., Tietge, J.E., and Ankley, G.T., Xenoendocrine disruptorstieredscreening and testing: filling the data gaps, Toxicology, 181, 371, 2002.146. U.S. Environmental Protection Agency, Prevention, Pesticides and Toxic Substances. Health EffectsTest Guidelines, OPPTS 870.3700, Prenatal Developmental Toxicity Study, EPA 712-C-98-207,August 1998.© 2006 by Taylor & Francis Group, LLC

THE U.S. EPA ENDOCRINE DISRUPTOR SCREENING PROGRAM 523147. McIntyre, B.S., Barlow, N.J., Wallace, D.G., Maness, S.C., Gaido, K.W., and Foster, P.M., Effects ofin utero exposure to linuron on androgen-dependent reproductive development in the male Crl,CD(SD)BR rat, Toxicol. Appl. Pharmacol., 167, 87, 2000.148. McIntyre, B.S., Barlow, N.J., and Foster, P.M., Male rats exposed to linuron in utero exhibit permanentchanges in anogenital distance, nipple retention, and epididymal malformations that result in subsequenttesticular atrophy, Toxicol. Sci., 6, 62, 2002.149. McIntyre, B.S., Barlow, N.J., and Foster, P.M., Androgen-mediated development in male rat offspringexposed to flutamide in utero, permanence and correlation of early postnatal changes in anogenitaldistance and nipple retention with malformations in androgen-dependent tissues, Toxicol. Sci., 62,236, 2001.150. Gray, L.E., Wolf, C., Lambright, C., Mann, P., Price, M., Cooper, R.L., and Ostby, J., Administrationof potentially antiandrogenic pesticides (procymidone, linuron, iprodione, chlozolinate, p,p’-DDE,and ketoconazole) and toxic substances (dibutyl- and diethylhexyl phthalate, PCB 169, and ethanedimethane sulphonate) during sexual differentiation produces diverse profiles of reproductive malformationsin the male rat, Toxicol. Ind. Health, 15, 94, 1999.151. Anderson, S.A., Fail, P.A., McTaggart, B.J., Tyl, R.W., and Gray, L.E., Jr., Reproductive toxicity ofmethoxychlor in corn oil to male and female Long-Evans hooded rats using the alternative reproductiontest protocol (ART). Biol. Reprod., 50(Suppl. 1), 101, 1994.152. U.S. EPA, Endocrine Disruptor Screening Program, Proposed Chemical Selection Approach for InitialRound of Screening (EPA Fact Sheet). http://www.epa.gov/scipoly/oscpendo/factsheet.htm.© 2006 by Taylor & Francis Group, LLC

CHAPTER 12Role of Xenobiotic Metabolism in Developmentaland Reproductive ToxicityArun P. KulkarniCONTENTSI. Introduction ........................................................................................................................527II. Overview of Xenobiotic Metabolism ................................................................................527A. Phase I Reactions.......................................................................................................5281. Cytochrome P450 ................................................................................................5282. Flavin-Containing Monooxygenase.....................................................................5293. Peroxidases and Prostaglandin Synthase ............................................................5294. Lipoxygenase .......................................................................................................5295. Lipid Peroxidation ...............................................................................................530B. Phase II Reactions .....................................................................................................5301. UDP-Glucuronosyl Transferase...........................................................................5302. Sulfotransferase....................................................................................................5313. Glutathione Transferase.......................................................................................531III. Assay of Xenobiotic Metabolizing Enzymes ....................................................................531A. General Comments ....................................................................................................531B. Assay Methods ..........................................................................................................5321. Cytochrome P450 Estimation..............................................................................5322. Cytochrome P450–Mediated Xenobiotic Oxidation...........................................5323. Flavin-Containing Monooxygenase.....................................................................5334. Lipoxygenase .......................................................................................................5335. Prostaglandin Synthase........................................................................................5346. Peroxidase............................................................................................................5357. Epoxide Hydrase..................................................................................................5368. Glutathione Transferase.......................................................................................5369. UDP-Glucuronosyl Transferase...........................................................................53710. Sulfotransferase....................................................................................................537IV. Metabolism of Developmental Toxicants ..........................................................................538A. Maternal Liver ...........................................................................................................5381. Cytochrome P450 ................................................................................................5382. Flavin-Containing Monooxygenase.....................................................................5403. Epoxide Hydrase..................................................................................................540525© 2006 by Taylor & Francis Group, LLC

526 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION4. UDP-Glucuronosyl Transferase...........................................................................5405. Glutathione Transferase.......................................................................................540B. Uterus and Decidua ...................................................................................................5401. Cytochrome P450 ................................................................................................5402. Peroxidase............................................................................................................5413. Lipid Peroxidation ...............................................................................................5414. Lipoxygenase and Prostaglandin Synthase .........................................................5415. Flavin-Containing Monooxygenase.....................................................................542C. Intrauterine Conceptual Tissues ................................................................................5421. Animal Species ....................................................................................................5422. Humans ................................................................................................................542D. Embryo.......................................................................................................................5431. Cytochrome P450 ................................................................................................5432. Prostaglandin Synthase........................................................................................5443. Lipoxygenase .......................................................................................................5454. Peroxidase............................................................................................................545E. Fetus...........................................................................................................................5461. Cytochrome P450 ................................................................................................5462. Flavin-Containing Monooxygenase.....................................................................5473. Prostaglandin Synthase........................................................................................5474. Peroxidase............................................................................................................5475. Lipoxygenase .......................................................................................................5476. Epoxide Hydrase..................................................................................................5487. Glutathione Transferase.......................................................................................5488. UDP-Glucuronosyl Transferase...........................................................................5489. Sulfotransferase....................................................................................................548F. Fetal Membranes .......................................................................................................5491. Cytochrome P450 ................................................................................................5492. Peroxidative Enzymes..........................................................................................549G. Placenta......................................................................................................................5491. Cytochrome P450 ................................................................................................5492. Flavin-Containing Monooxygenase.....................................................................5503. Prostaglandin Synthase........................................................................................5504. Peroxidase............................................................................................................5505. Lipoxygenase .......................................................................................................5516. Lipid Peroxidation-Coupled Cooxidation of Xenobiotics ..................................5537. Other Enzymes of Xenobiotic Oxidation............................................................5548. Epoxide Hydrase..................................................................................................5559. UDP-Glucuronosyl Transferase...........................................................................55510. Sulfotransferase....................................................................................................55511. Glutathione Transferase.......................................................................................555V. Role of Xenobiotic Metabolism in Reproductive Toxicology ..........................................556A. General.......................................................................................................................556B. Male Reproductive System........................................................................................5561. Testicular Enzymes of Xenobiotic Metabolism..................................................5562. Male Reproductive Toxicants ..............................................................................556C. Female Reproductive System ....................................................................................5571. Ovarian Enzymes of Xenobiotic Metabolism.....................................................5572. Female Reproductive Toxicants ..........................................................................558VI. Conclusions and Future Research Needs ..........................................................................559References ......................................................................................................................................560© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 527I. INTRODUCTIONIf it is a must to ingest food, drink water, and breathe, it is equally critical for survival to have theability to defend ourselves from pernicious chemicals entering the body. Human exposure topotentially harmful natural or man-made xenobiotics is inevitable, and it is only a matter of extent.Evolutionary forces have endowed humans and animals with the necessary armor in the form ofbiochemical machinery composed of enzymes, each with unique ability to transform and neutralizelipophilic xenobiotics into hydrophilic metabolites suitable for exit from the body via feces andurine. Unfortunately, biochemical accidents do sometimes occur, leading to the generation ofpowerful ultimate toxicants. That process, instead of providing protection, predisposes body tissuesand cells to demise or destruction. Xenobiotics interact with enzymes as substrates, inhibitors,activators, or inducers.Knowledge of xenobiotic metabolism is vital to understanding the developmental and reproductivetoxicity of chemical agents that require biotransformation. The bulk of the availableinformation originates from studies conducted with adult rodent liver. However, extrahepatic tissuespossess the ability to metabolize xenobiotics, and these tissues display major qualitative andquantitative differences in their metabolic capacity. Unfortunately, human data are very limited forconceptual tissues. Similarly, the evidence linking metabolism with developmental or reproductivetoxicity of chemicals is not always transparent. Several excellent reviews covering earlier progressin developmental toxicology are available, providing lists of original publications. The reader isencouraged to consult these reviews covering xenobiotic metabolism during pregnancy 1,2 and inembryonic, 3–6 fetal, 6–8 and placental 6,8–10 tissues. This chapter mainly focuses on the data publishedin recent years. Methodology employed in the study of major enzymes involved in xenobioticmetabolism is also briefly described.Since interpretation of in vivo xenobiotic metabolism data in terms of enzymes involved in aparticular reaction is often difficult, many scientists focus their efforts on well-designed in vitrostudies. This experimental approach permits investigators studying xenobiotic metabolism to: (1)establish the role of a specific enzyme, (2) examine the enzyme of interest under fully definedassay conditions, (3) make qualitative and quantitative estimates of the metabolites of a toxicant,(4) reveal the modulatory effects of endobiotics and other factors, and (5) study the effects of invivo exposure to chemicals on different pathways. In light of this, only select publications primarilydealing with in vitro xenobiotic metabolism are considered here. No attempt is made to present acomprehensive review of all aspects.II. OVERVIEW OF XENOBIOTIC METABOLISMThe process of transforming lipophilic xenobiotics to polar entities in the body is widely perceivedto occur in two distinct metabolic phases. During Phase I, functionalization, an atom of oxygen isincorporated in the chemical, and functional groups, such as –OH or –COOH, are either introducedor exposed. In other cases, xenobiotic bioconversion involves reduction or hydrolysis. Althoughmost reactions reduce chemical toxicity (detoxication), some Phase I metabolites are more reactivethan their parent compounds (bioactivation). They interact with critical macromolecules, such asprotein, RNA, and DNA, and initiate toxic sequelae. Phase II reactions consist of synthetic reactionsin which functional or polar groups are conjugated with endobiotics and include sulfation, glucuronidation,acetylation, methylation, and conjugation with reduced glutathione (GSH) or aminoacids. These reactions increase xenobiotic hydrophilicity and greatly promote excretion. In mostcases the conjugates are biologically nonreactive. Bioactivation of xenobiotics during Phase IIreactions can also occur, but it is less common.The increased polarity induced by conjugation greatly retards the ability of compounds to diffuseacross the lipid bilayer of the cell membrane. The exit of conjugates from cells is facilitated by© 2006 by Taylor & Francis Group, LLC

528 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmembers of the ATP-binding cassette superfamily of active transport proteins (e.g., multidrugresistance associated protein [MRP], multispecific organic anion transporter [MOAT], G-glycoproteins,and others). Active cellular efflux of xenobiotics by these transporters is often referred to asPhase III detoxication and elimination, because Phase I and/or Phase II reactions often precede orare required for Phase III processes. 11 A few investigators use the term “Phase III metabolism” torefer to a process that involves further metabolism of products emanating from Phase II reactions. 12This is exemplified by enzymatic catalysis of GSH conjugates leading to mercapturic acid synthesis.Another example is bioactivation of cysteine conjugate after methylation or by a pyridoxal phosphate–dependentβ-lyase. 13 A brief description of major enzymes involved in Phase I and Phase IIis given below.A. Phase I ReactionsOxidation represents the Phase I reaction most, if not all, xenobiotics encounter inside the body.It is mediated by microsomal cytochrome P450, flavin-containing monooxygenase, peroxidase,prostaglandin synthase, lipoxygenase, and processes occurring during lipid peroxidation. Dehydrogenasesand amine oxidases also contribute to oxidation of certain xenobiotics. Reduction ofxenobiotics is mediated by azoreductases and nitroreductases, while epoxide hydrolases, esterases,and amidases affect hydrolytic cleavage of specific linkages in xenobiotics.The liver was believed to be the chief site for the generation of ultimate developmental andreproductive toxicants, but that is not true for those chemicals in which the ultimate toxicant is areactive species or a radical with a relatively short biological half-life. Also P450, an enzyme knownfor its broad substrate specificity and dominant role in xenobiotic metabolism in the liver, waspresumed to be the sole catalyst responsible for the bioactivation of most developmental toxicantsin conceptual tissues, especially placenta. The research efforts devoted to this enzyme system forthe past several years resulted in unjustifiable neglect of alternate pathways for xenobiotic metabolismand hampered their exploration. Recent reports (discussed later) documenting the role ofother enzymes capable of xenobiotic oxidation in conceptal tissues have changed the direction ofresearch, and as a result, a new picture is emerging.Of the various Phase I reactions, oxidation, which not only represents the first metabolic reactionfor most xenobiotics, but in many cases also results in bioactivation, has received the most attention.In view of this, an overview of the major enzymes involved in xenobiotic oxidation is presentedbelow.1. Cytochrome P450Of the enzymes involved in xenobiotic oxidation, cytochrome P450 (P450) is the most extensivelystudied. P450 is a heme thiolate protein (E.C. 1.14.14.1, nonspecific oxygenase). Multiple speciesof P450 enzymes (not isozymes, in a strict sense) are known to occur in different animal tissues.At least 154 gene products constitute this superfamily. The number of P450 genes is currentlyestimated to be 53 in humans alone. The total number of known P450 sequences exceeds 500. 14Human liver microsomes contain more than 15 different P450 enzymes capable of oxidation ofxenobiotics and endobiotics. The major human microsomal P450 enzymes involved in xenobioticoxidation include CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2E1, CYP1A2,CYP2D6, and CYP3A4. Similar information is available for laboratory animal species. The titerof each form varies interindividually within a species, depending upon tissue, gender, genetics,hormonal status, exposure to chemicals, and other factors. The obligatory components for all P450sto function as monooxygenases include NADPH, O 2 , and NADPH P450 reductase, while the needfor phospholipids, NADH, cytochrome B 5 , and NADH cytochrome B 5 reductase varies widelyamong different P450 species. The steady state concentration of NADPH in hepatocytes is estimatedto be about 100 µM.© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 529Xenobiotic catalysis by P450 involves multiple steps, some being rate limiting. The types of2e oxidation of xenobiotics catalyzed by P450 mainly include hydroxylation of aliphatic or aromaticcarbons, epoxidation of a C=C double bond, heteroatom (S– and N–) oxygenation and N-hydroxylation,and heteroatom (N–, S– and O–) dealkylation. A few examples of P450-mediated 1ereduction or xenobiotic oxidation are also known. Several excellent reviews on P450 cover mechanismsof catalysis, 15 purification and properties, 16 structure and function, 17 substrate specificity, 18–21induction, 22 inhibition and stimulation, 23 and species differences. 242. Flavin-Containing MonooxygenaseFlavin-containing monooxygenases (FMO; E.C. 1.14.13.8) are microsomal flavoproteins capableof one step 2e oxygenation of many heteroatom (N–, S–, and P–) containing chemicals, drugs, andpesticides. 25,26 FMO is primarily a detoxication enzyme, although a few examples of xenobioticbioactivation are known. At least six forms (FMO1 to FMO6) of mammalian FMOs are described,and some may present in multiple tissues of a given species. NADPH is the preferred cofactor,while NADH supports xenobiotic oxidation by FMO with about 40% efficiency. Relatively longlivedreduced hydroperoxyflavin is the oxidant responsible for the oxygenation of xenobiotics. Therelease of a water molecule and NADP + represent the rate limiting steps in the catalytic cycle ofFMO. 25,26 In general, FMO is more thermolabile than P450 and also exhibits a higher pH optimum(pH 8–10) in vitro. It is noteworthy that antibodies raised against FMO do not inhibit the enzyme.FMO is refractive to the inductive effects of many chemicals.3. Peroxidases and Prostaglandin SynthasePeroxidases, e.g., lactoperoxidase, myeloperoxidase, and tissue specific peroxidases, are hemeproteins. Among different peroxidases, protaglandin synthase (PGS; E.C. 1.14.99.1) is the mostextensively studied enzyme. 27,28 PGS is a bifunctional enzyme with distinct cyclooxygenase andperoxidase activities. Cyclooxygenase is responsible for bis-dioxygenation of free arachidonic acidto produce cyclic endoperoxide hydroperoxide (PGG 2 ). PGG 2 is reduced to PGH 2 by the peroxidaseactivity of the enzyme. In most tissues, free arachidonate occurs at a concentration of about 20µM. 27 Although PGS activity predominantly resides in endoplasmic reticulum, some activity canbe noted in the nuclear envelope. There are two structurally related isoforms of PGS, called COX-1 and COX-2, encoded by separate genes. COX-1 is the constitutive form that synthesizes prostaglandinsresponsible for “housekeeping” functions, while COX-2 is induced by proinflammatorycytokines and growth factors and is believed to play a role in inflammation and cell growth. Despitedifferences in sensitivity to inhibitors, substrate specificity, and expression, both forms of PGSdisplay a remarkable similarity in structure and catalytic function. Several mechanisms have beenproposed to explain xenobiotic oxidation catalyzed by PGS. 27,28 Stated briefly, many reducingcosubstrates are directly oxidized to free radical species by the peroxidase activity of PGS. Incertain cases, the free radical interacts with O 2 and generates a secondary oxidant capable ofxenobiotic oxidation. Lipid peroxyl radicals, formed as intermediates, can serve as potent oxidantsfor epoxidation and other xenobiotic bioactivation reactions. Both COX-1 and COX-2 occur inmany mammalian tissues, and several intrauterine tissues possess PGS activity. 6 Ram seminalvesicles express only COX-1 and represent a rich source of PGS activity. Because of its low contentin the liver, PGS is believed not to contribute significantly to hepatic xenobiotic metabolism. 27,284. LipoxygenaseLipoxygenases (LO) comprise a family of non–heme-iron containing cytosolic proteins that catalyzethe insertion of a single molecule of O 2 into polyunsaturated fatty acid (PUFA) at defined positions.Accordingly, 5-, 12-, and 15-LOs, and other LOs have been described in the literature. 29 The© 2006 by Taylor & Francis Group, LLC

530 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONstereospecific actions of these enzymes require a 1,4-cis-cis-pentadiene structure in the substratemolecule. This moiety is transformed into 1-hydroperoxy-2-trans, 4-cis-pentadiene. Subsequently,this relatively unstable hydroperoxy derivative of arachidonic acid undergoes catalysis, giving riseto a wide array of biologically active metabolites, such as leukotrienes, hydroxyeicosatetraenoicacids (HETEs), lipoxins, and hepoxilins. 29–31 The substrate requirements of LO are less stringentthan those of PGS. Thus, certain LOs not only use free PUFAs but can also peroxidize di- and triglycerides,lipoproteins, and even membrane-bound lipids. LOs are ubiquitous in plants and animals,occurring in several mammalian organs, including intrauterine tissues. 6,29–31 Various arachidonatemetabolites generated by LOs are essential for fertilization, implantation, normal growth anddevelopment of the conceptus, maintenance of pregnancy, and certain aspects of parturition. 6 Asingle LO protein not only exhibits dioxygenase activity toward PUFAs but also displays cooxidaseactivity toward a large number of xenobiotics. 29–31 The peroxidaselike (cooxidase) activity ofdifferent LOs is not only efficiently supported by PUFAs and their hydroperoxides but also bysubstitutes such as H 2 O 2 , tert-butyl hydroperoxide, and cumene hydroperoxide. 29–31 The principalmechanisms of xenobiotic oxidation include one-electron oxidation of reducing substances to freeradical species by LO. Concurrently, the active ferric LO is converted into inactive ferrous LO.Ferrous iron oxidation by hydroperoxide regenerates the active enzyme for the next catalytic cycle.The intermediate peroxyl radicals of PUFA generated during dioxygenation also serve as oxidantsand trigger epoxidation or sulfoxidation of xenobiotics. Additional mechanisms of LO-mediatedxenobiotic catalysis have been described. 29–315. Lipid PeroxidationLipid peroxidation is a normal biochemical process constantly occurring at a basal rate in essentiallyall cell types. It occurs in tissues as a metal-catalyzed nonenzymatic process, as well as processesmediated by LO and PGS. Lipid peroxidation products are essential for maintenance of pregnancyand for normal growth and development of the conceptus. They are involved in the induction oflabor and in the parturition process at term. Recent data show that increased lipid peroxidation dueto aberrant PGS and/or LO activities in reproductive tissues may be linked to various disorders,such as hypertension and preeclampsia in pregnancy. 32 Uncontrolled lipid peroxidation may alsobe responsible for the developmental or reproductive toxicities of some chemicals. The proposedmechanisms include the generation of reactive oxygen species (ROS), the toxic products of lipidperoxidation, and/or xenobiotic free radicals. 3,4 Lipid peroxidation-coupled cooxidation representsanother pathway for oxidation of certain xenobiotics. 1,6B. Phase II Reactions1. UDP-Glucuronosyl TransferaseGlucuronidation is a Phase II reaction catalyzed by two families of microsomal UDP-glucuronosyltransferases (UGTs: E.C. 2.4.17), with each family made up of several members. 33 The processinvolves transfer of glucuronic acid from UDP-glucuronic acid (UDPGA) to aglycone by UGT.The major site for glucuronidation is the liver. The concentration of UDPGA in the liver is estimatedat about 350 µM. UGT-mediated hepatic glucuronidation of xenobiotics is considered a “lowaffinity–high-capacity”process. UGT is inducible by xenobiotics. Operation of the enzyme in vivorequires at least two transporters. One is for the transport of UDPGA from cytoplasm to the lumenof endoplasmic reticulum, and the other exports glucuronide into the cytoplasm. The enzyme exhibitslatency because of its luminal location, and prior membrane disruption by, e.g., brief sonication ordetergent addition, is needed for full expression of activity. Many endobiotics and xenobiotics or theirmetabolites bearing –OH, –COOH, –NH 2 , and –SH groups undergo glucuronidation. Although most© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 531conjugates are hydrophilic, biologically inactive products, many examples of activation of drugs,carcinogens, and other chemicals are known. 332. SulfotransferaseSulfotransferases (SULTs: E.C. 2.8.2) are cytosolic enzymes that mediate the transfer of a sulfo(SO 3 ) moiety from 3′-phosphoadenosine-5′-phosphosulfate (PAPS) to a nucleophilic site of theacceptor molecule. Several excellent reviews provide detailed information on SULTs. 34 At least 10SULTs are described in humans alone. Because of the rate limiting availability of PAPS in liverand SULT inhibition at high substrate concentrations, sulfoconjugation is considered a “highaffinity–low-capacity”pathway of xenobiotic metabolism. Under normal conditions, PAPS contentis 30 to 70 nmol/g in the liver and 3 to 30 nmol/g in other tissues. Although many xenobiotics aredetoxified by sulfoconjugation, in a few cases, bioactivation does occur. 343. Glutathione TransferaseGlutathione transferases (GSTs; E.C. 2.5.1.18) are multifunctional proteins ubiquitously distributedin plants and animals. They effect thioether bond formation through the conjugation of GSH withchemicals containing an electrophilic center. GSH conjugation represents the first step in mercapturicacid synthesis. Additionally, certain GSTs isomerize unsaturated compounds, reduce hydroperoxides,serve as binding, storage, or transport proteins, and modulate signal transduction processes.35 GSTs belong to two distinct superfamilies. The first comprised at least 16 cytosolic proteinsin humans. The cytosolic GSTs have been assigned to eight families or classes, designated Alpha(α), Mu (µ), Pi (π), Sigma (σ), Theta (θ), Zeta (ζ), Omega (ω), and Kappa (κ). Four additionalmembers, Beta (β), Delta (δ), Phi (φ), and Tau (τ) occur in bacteria, insects, and plants. 35 Thesecond is composed of microsomal GSTs and contains at least six members in humans. 36 Thecofactor, GSH, is the most abundant intracellular nonprotein thiol. It occurs in rodent liver at aconcentration of about 10 mM.Electrophilic chemicals serve as substrates for GST. The process of GSH conjugation ofxenobiotics by GSTs primarily involves the displacement of an electron withdrawing group byGSH in a substrate containing a halide, sulfate, sulfonate, phosphate, or nitro group or the directaddition of GSH across a carbon-carbon double bond (Michael addition reaction). GST can opena strained epoxide ring during the GSH addition. In general, GSH conjugation leads to the loss ofbiological reactivity of the parent compound. However, GSH conjugation of geminal and vicinaldihaloalkanes generates mutagenic species. 13 With polyhalogenated alkenes, further metabolism ofthe initial GSH conjugates yields ultimate toxicants. 13,35A. General CommentsIII. ASSAY OF XENOBIOTIC METABOLIZING ENZYMESObservation of a specific xenobiotic metabolism reaction in a tissue may not necessarily identifythe catalyst involved. Multiple enzymatic pathways should be considered. For example, aldrinepoxidation, long believed to be an indication specific for P450, 37 is also mediated by PGS 20,21,27,28and LO. 29–31 To assess the importance of an enzyme, both qualitative and quantitative aspects shouldbe considered. The presence of a bioactivation enzyme signals the potential of that pathway forthe generation of an ultimate toxicant. However, confirmation of the pathway’s biological significancerequires more than mere demonstration of a barely detectable level of the enzyme in a tissue(e.g., P450 in nonsmoker’s placenta or in rodent embryos) by ultrasensitive methods. Although of© 2006 by Taylor & Francis Group, LLC

532 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONacademic interest, such a finding may not be meaningful from a practical point of view because itmay not provide an adequate explanation for the toxicity of the chemical. Since the quantity ofproduct formed is determined by the amount of enzyme, it is important that the bioactivationenzyme occurs in the tissue in biologically significant amounts. Furthermore, the catalytic efficiencyof the enzyme must be high enough to generate an amount of ultimate toxicant adequate to interactwith target molecules and induce toxicity. This implies that the generation of ultimate toxicantmust overwhelm the endogenous defense and repair mechanisms. That thresholds exist and themajority of babies born today are healthy and normal despite everyday low exposures of womento many chemicals throughout pregnancy attests to this contention. In this section, assay proceduresare given for estimating activities of major xenobiotic metabolizing enzymes.B. Assay Methods1. Cytochrome P450 EstimationThe most commonly used method is that of Omura and Sato, 38 which is based on the reducedcarbon monoxide (CO) difference spectrum.a. Reagents1. 0.1 M potassium phosphate buffer, pH 7.42. Sodium dithionite, reagent grade3. CO gas, reagent grade (use in fume hood)b. ProcedureWashed microsomes in the buffer (1 to 3 mg protein/ml) are placed in two matching cuvettes (1.0-cm light path). By use of a spectrophotometer equipped with a turbid sample accessory operatedin split beam mode, a baseline of equal absorption is recorded between 400 and 500 nm. Thecontents of the sample cuvette are saturated with CO (one bubble per second for about 30 sec). Afew crystals (about 1 mg) of sodium dithionite are added to both cuvettes. The cuvettes are coveredwith Parafilm ® , and the contents are gently mixed by inverting the cuvettes three or four times.Difference spectra are then recorded (400 to 500 nm) until absorbance at the 450-nm peak reachesmaximum. P450 content is determined using the following equation:[(A450 to 490 nm) observed (A450 to 490 nm) baseline]/0.091 = nmol P450/milliliterc. RemarkIn the case of significant hemoglobin (Hb) contamination, the procedure of Matsubara et al. 39 isrecommended.2. Cytochrome P450–Mediated Xenobiotic OxidationXenobiotic oxidation observed in a reconstituted system containing purified P450 enzyme and/orinhibition of the reaction by the inclusion of antibody to NADPH P450 reductase provide strongevidence for the involvement of P450. The results of experiments using preferred substrates incombination with selective inhibitor(s) and a specific antibody can provide useful data on thepossible contribution of individual P450 enzymes to xenobiotic oxidation in microsomes. P450 cancatalyze a wide spectrum of xenobiotic oxidation reactions. For each type of reaction, a choice ofsubstrate is available. Typical assay conditions employed in such experiments are presented below© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 533for aldrin epoxidation. With appropriate modification(s), these assay conditions can be adapted forthe study of other reactions.a. EpoxidationThe highly sensitive and reproducible procedure of Wolff et al., 37 which determines the rate ofaldrin epoxidation, is commonly used.b. Reagents1. 0.1 M potassium phosphate or Tris-HCl buffer, pH 7.42. 0.5 M MgCl 23. 1.0 mM NADPH or NADPH generating system (1 IU glucose-6-phosphate dehydrogenase/ml,0.5 mM NADP + , 10 mM glucose-6-phosphate)4. Aldrin solution in acetone or ethanol5. Dieldrin standard solution in hexanec. ProcedureAfter preincubation of microsomes in buffer with 0.25 mM aldrin at 37˚C for 3 to 5 min, thereaction is initiated by the addition of NADPH (1.0 ml final volume). The flasks are incubated for5 to 15 min at 37˚C with constant shaking. The reaction is terminated by the addition of 250 µlof 10% trichloroacetic acid (TCA), followed by chilling in an ice bath. The dieldrin produced inthe incubation medium is extracted by vortexing with 2.5 ml of hexane. The mixture is centrifugedat 2500 × g for 5 min to accelerate phase separation. Dieldrin is quantified by injecting 2 to 5 µlof hexane extract into a gas chromatograph equipped with FID or 63 Ni ECD. With ECD, levels ofdieldrin as low as 5 to 10 pg can be detected.3. Flavin-Containing MonooxygenaseThe activity of FMO can be readily measured by following methimazole-dependent oxygen uptakeor NADPH depletion. 40a. Reagents1. 50 mM Tris-glycine buffer, pH 8.52. 2.0 mM GSH3. 1.0 mM NADPH4. 0.2 mM methimazole5. 2.4 mM n-octylamineb. ProcedureMix microsomes with methimazole, GSH, n-octylamine, and the buffer in sample and referencecuvettes. Initiate the reaction by the addition of NADPH to the sample cuvette. A decrease in theabsorbance at 340 nm is followed for 3-5 min. An extinction coefficient of 6.22 mM –1 cm –1 is usedto calculate NADPH depletion rate, which equals to the rate of methimazole sulfoxidation.4. LipoxygenaseDioxygenase activity of tissue LO can be assayed either spectrophotometrically by estimating theconjugated dienes formed or by measuring oxygen uptake (Section III.B.5). The identity of 5-, 12-,© 2006 by Taylor & Francis Group, LLC

534 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONor 15-LO can be determined from reverse-phase HPLC of radioactive arachidonic acid metabolites.The spectrophotometric method 41 described below is simple, rapid, and sensitive for routine estimationof the rate of dioxygenase activity of LO. The protocol described in Section III.B.6 can be adaptedwith some modifications to measure the cooxidase (peroxidaselike) activity of LO. Hemoglobin (Hb)exhibits both pseudodioxygenase and pseudoperoxidase activities, 29–31 and so Hb contamination posesa serious problem for the estimation of tissue LO activity. The method devised in our laboratory forhuman term placental cytosolic LO (HTPLO) alleviates this problem. In brief, the protocol requiresthe treatment of cytosol with ZnSO 4 (0.5 mM). 42 The treatment selectively removes Hb (greater than97%) without affecting the dioxygenase and cooxidase activities of HTPLO. After centrifugation, thesupernatant is dialyzed thoroughly before assaying for dioxygenase or cooxidase activity.a. Reagents1. 50 mM Tris buffer, pH 9.0, ice cold, deaerated or argon bubbled2. Substrate solution is prepared fresh daily by dissolving 466 µl pure linoleic acid in 500 µl absoluteethanol in a test tube on ice. A drop of Tween-80 is added, the volume is adjusted with buffer to10 ml, and the mixture is vortexed briefly (5 sec). The substrate solution is protected from lightand stored on ice under argon.b. ProcedureA matched pair of quartz cuvettes is used. In the sample cuvette, a suitable amount (25 to 100 µl)of enzyme (ZnSO 4 treated, dialyzed 100,000 × g supernatant or purified LO) is mixed with buffer,while an equal volume of buffer is pipetted into the reference cuvette (no enzyme). After 2 min ofpreincubation at 37˚C, the reaction is initiated by the addition of linoleic acid solution (5 mM ordesired final concentration), first to the reference cuvette and then to the sample cuvette (1 to 3 mlfinal volume). The change in absorbance at 234 nm is recorded for 3 min (or longer if needed).The initial linear portion of the recording, reflecting conjugated diene formation, is used to computethe rate of reaction. Specific activity is calculated using a molar extinction coefficient of 25,000for linoleic acid hydroperoxide.c. RemarksThe optimum pH and substrate concentration vary, depending upon the tissue source of LO;therefore, it is essential to run pilot experiments if the optimum assay conditions are unknown.Arachidonic acid or another suitable PUFA can be substituted for linoleic acid. In case of a Hbcontamination problem, the use of (partially) purified enzyme is strongly recommended. Within anarrow range, dioxygenase activity is proportional to the amount of enzyme protein used in theassay. Specific activity decreases when a high enzyme concentration is used, due primarily to selfinactivationby fatty acid hydroperoxide accumulation in the medium. Native enzyme contains mostof its iron in the inactive (ferrous) state and commonly displays a lag phase, which may last up toseveral minutes. Especially in the case of crude enzyme preparations, the presence of endogenouschemicals (e.g., GSH, catecholamines) in the reaction medium also increases the lag period. 43 Priordialysis or addition of a low (micromolar) concentration of H 2 O 2 , 44,45 or other lipid hydroperoxidesis required to fully activate dioxygenase. Certain LOs, e.g., 5-LO, require calcium ions and/or ATPfor maximum expression of activity, and this aspect should be considered. 29,305. Prostaglandin SynthaseThe measurement of cyclooxygenase activity can be followed by O 2 uptake and is described below.The assay procedure outlined in Section III.B.6 can be adapted with necessary modifications forthe estimation of peroxidase activity of PGS.© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 535a. Reagents1. Arachidonic acid (substrate) solution can be prepared as described for linoleic acid in SectionIII.B.4. Arachidonate is relatively susceptible to autooxidation and therefore, arachidonate solutionsshould be stored under argon or nitrogen.2. 0.1 M potassium phosphate buffer, pH 7.4.b. ProcedureA microsomal suspension (0.5 to 2.0 mg protein) and the buffer are placed in a water-jacketedvessel maintained at 37˚C and containing a Clark-type oxygen electrode connected to an oxygenmonitor. The contents are allowed to equilibrate at 37˚C, and a stable baseline is established (2 to3 min). The reaction is started by adding the substrate solution (0.1 mM arachidonate) with amicrosyringe (final volume 2 ml) though the vessel’s outlet, and the oxygen uptake is recorded for3 to 5 min. The ambient oxygen concentration in the aerated buffer at 37˚C is approximately 220 µM.c. RemarksThis procedure can be adapted with suitable modifications for the measurement of the dioxygenaseactivity of LO. Similar to LO, the cyclooxygenase activity of PGS exhibits a linear response withina narrow range of enzyme protein concentration. Low or no activity is observed when high enzymeconcentration is used. When purified enzyme is reconstituted, the addition of heme (hematin),phenol, and GSH is required to observe maximal activity.6. PeroxidaseA number of compounds can be used to monitor tissue peroxidase activity. The most commonlyused method, described below for human placental peroxidase, is an adaptation of the spectrophotometricmethod reported for peroxidase-mediated tetraguaiacol formation from guaiacol in thepresence of H 2 O 2 . 46a. Reagents1. 50 mM Tris-HCl buffer, pH 7.2 (or appropriate buffer for the enzyme under study)2. 2.6 M guaiacol (O-methoxyphenol) solution in absolute ethanol3. 3.3 mM H 2 O 2 (dilute 30% hydrogen peroxide with the buffer)b. ProcedurePipette and mix a suitable amount of buffer, 5 to 50 µl of enzyme (absent in the reference cuvette),and guaiacol (13 mM; 5 µl/ml) in the sample and reference cuvettes, and preincubate for 2 to 3min at 37˚C. Initiate the reaction by the addition of H 2 O 2 solution (100 µl/ml or desirable volume)(1 or 3 ml final volume of reaction mixture). Record the increase in the absorbance at 470 nm for3 to 5 min. Specific activity is calculated using an extinction coefficient of 26.6 mM –1 cm –1 fortetraguaiacol. The initial linear portion of the response curve is used to compute the enzyme velocity.c. RemarksTetraguaiacol is the major product of the reaction. Small amounts of other metabolites are alsoformed. It is very important to consider the high peroxide specificity exhibited by different peroxidasesto support xenobiotic oxidation. Thus, H 2 O 2 serves as an efficient cofactor for xenobioticoxidation catalyzed by human placental peroxidase, 47 but it is a very poor cofactor for PGS. 27© 2006 by Taylor & Francis Group, LLC

536 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSimilarly, tert-butyl hydroperoxide or cumene hydroperoxide are less efficient in supporting peroxidaseactivity of PGS. The peroxidase activity of PGS requires the presence of PGG 2 or PUFAhydroperoxide in the medium. 27,28 Since these hydroperoxides are generated by the dioxygenaseactivity, cooxidation of xenobiotics by PGS and LO is routinely assayed using reaction mediasupplemented with PUFA alone. Peroxide specificity of LO is relatively broad. Not only H 2 O 248and PUFA hydroperoxides 29–31 but either tert-butyl hydroperoxide or cumene hydroperoxide 49 canalso efficiently support the cooxidase activity of LO.7. Epoxide HydraseThe procedure of Oesch et al., 50 which employs styrene oxide, is commonly used to assay epoxidehydrase (EH; E.C. 3.3.2.3) activity of tissue microsomes. The assay can also be performed using[ 3 H]BP-4,5-oxide.a. Reagents1. 7-[ 3 H]styrene oxide, 16 mM, in tetrahydrofuran containing 0.1% triethylamine2. 0.5 M tris-HCl buffer, pH 8.7b. ProcedureAdd substrate solution (20 µl) to the suitable amount of enzyme in buffer (300 µl final volume)and incubate for 5 to 10 min at 37˚C. The reaction is terminated by addition of 5 ml petroleumether (bp 30˚C to 60˚C) to the test tube. The mixture is vortexed and then centrifuged for 5 minat 1000 × g. The tube is then cooled by immersion in an ice-acetone bath. After the aqueous phaseis frozen, the upper organic phase is discarded. This extraction procedure is repeated twice. Finally,2 ml of ethyl acetate is added, and the contents are mixed and centrifuged as before. An aliquot(300 µl) of the ethyl acetate extract containing styrene glycol is transferred to a scintillation vialfor radioactivity counting. The incubation mixture containing boiled enzyme (100˚C for 10 min)or no enzyme is used as the control.c. RemarksEH activity has been observed in rodent liver nuclei, mitochondria, and microsomes (mEH) as wellas cytosol (sEH). Of these, the microsomal and cytosolic forms have received the most attention.According to some reports, selective substrates for mEH include phenanthrene oxide, BP-4,5-oxide,and cis-stilbene oxide while sEH hydrolyzes trans-stilbene oxide and trans-β-benzyl styrene oxidemore efficiently. Most substrates are potentially carcinogenic; therefore, routine precautions forcarcinogen and radioisotope handling should be taken.8. Glutathione TransferaseA number of assay procedures are available for assaying GST activity. This includes estimation ofdepletion of GSH or test substrate from the medium or quantitation of specific product formationfrom the test substrate, either by radiometry, spectrophotometry, or HPLC. The most commonlyused spectrophotometric method, described by Habig et al. 51 and modified by Radulovic andKulkarni 52 for human placental GST (GSTP1-1), is described below.a. Reagents1. 0.2 M 1-chloro-2,4-dinitrobenzene (CDNB) solution in acetone or ethanol2. 0.1 M potassium phosphate buffer, pH 6.53. 25 mM reduced glutathione solution in buffer4. Enzyme: purified or cytosol (100,000 × g supernatant)© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 537b. ProcedureThe rate limiting amount of GST (5 to 50 µl/ml; absent in the reference cuvette), GSH (2.5 mM;100 µl/ml), and buffer are pipetted into the sample and reference cuvettes and allowed to preincubatefor 2 to 3 min at 37˚C. The reaction is initiated by the addition of CDNB solution (1 mM, 5 µl/ml),and the increase in absorbance at 340 nm with time, reflecting the formation of the GSH adductof CDNB, is recorded. The specific activity is estimated by use of the initial linear portion of theresponse curve (1 to 3 min) and an extinction coefficient of 9.6 mM –1 cm –1 .c. RemarksThe inhibition of purified rat live and human placental GST by exogenous Hg is concentrationdependent, 53 and contamination of tissue cytosol with Hb is commonly noted. Thus, for accuratedetermination of GST activity, the method devised by Sajan and Kulkarni 53 to remove Hb frommammalian tissue cytosol prior to the assay is strongly recommended. GST occurs in multipleforms. 35,36 CDNB serves as a substrate for most, if not all, isozymes of GST. When unpurifiedenzyme is used, the use of preferred substrates, along with selective inhibitors and specific antibodies,can provide useful information regarding individual isozymes of GST.9. UDP-Glucuronosyl TransferaseSeveral spectrophotometric, radiometric, fluorometric, and HPLC methods are available in theliterature to assay different forms of UGT activity in tissue microsomes. One of the routinelyemployed spectrophotometric methods measuring p-nitrophenol glucuronidation is described below.a. Reagents1. 1.0 mM UDP-glucuronate (UDPGA), triammonium salt, adjusted to pH 7.4 with KOH2. 1.0 mM p-nitrophenol in 50 mM Tris-HCl buffer, pH 7.4, containing 10 mM MgCl 23. 50 mM Tris-HCl buffer, pH 7.44. 0.2 M glycine buffer containing 0.15 M NaCl, pH 10.45. Triton X-100 detergentb. ProcedureMix UDPGA (cofactor), p-nitrophenol (substrate), and Tris-HCl buffer (final volume 1.4 ml) in atube and preincubate for 3 min at 37˚C. Start the reaction by adding 0.25 to 1.0 mg Triton X-100activated microsomes (100 µl). Microsomes are activated prior to incubation by mixing 0.2 µlTriton X-100/mg microsomal protein. After incubation for 10 min at 37˚C, the reaction is terminatedby the addition of 5 ml of glycine buffer. Measure p-nitrophenol concentration spectrophotometricallyat 405 nm. Control incubation mixture is prepared by omitting UDPGA. The extinctioncoefficient of p-nitrophenol at pH greater than 10 is 18.1 mM –1 cm –1 .10. SulfotransferaseA variety of assay procedures are available to estimate SULT activity. 34 The general radiometricassay described below is versatile and convenient.a. Reagents1. 0.5 M sodium phosphate buffer, pH 6.52. 0.1 M 2-mercaptoethanol, prepared by adding 0.07 ml of the thiol to 10 ml H 2 O© 2006 by Taylor & Francis Group, LLC

538 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION3. 5 mM 2-naphthol, dissolved in acetone4.35S-labeled (0.1 to 1.0 µCi) plus cold PAPS5. 0.1 M barium acetate6. 0.1 M barium hydroxide7. 0.1 M zinc sulfateb. ProcedureMix buffer, mercaptoethanol, 2-naphthol (substrate), and PAPS (cofactor) and preincubate for 3min at 37˚C. Start the reaction by adding a suitable amount of enzyme (final volume 1.0 ml). After10 min incubation at 37˚C, terminate the reaction by addition of the phosphate buffer and 0.2 mlbarium acetate solution. Then add 0.2 ml barium hydroxide solution followed by 0.2 ml zinc sulfatesolution. Remove the precipitate by low speed centrifugation (2500 × g for 10 min) and repeat thebarium plus zinc step. The unreacted [ 35 S]PAPS remaining after the reaction is precipitated bybarium, but most sulfoconjugates are water soluble and are not precipitated. Therefore, the radioactivityremaining in the supernatant after centrifugation is used to quantify the sulfoconjugateformed. Controls are assayed as above, omitting only the substrate.c. RemarksThis procedure can be combined with other analytical procedures, such as thin layer chromatography,HPLC, or spectrophotometry using methylene blue reagent. If necessary, any other desirablesubstrate can substitute for the 2-naphthol. The optimum pH for the reaction varies by substrateand enzyme. This assay may not be suitable for some substrates if their sulfoconjugates are precipitatedto various degrees by the barium treatment (e.g., sulfoconjugates containing a carboxyl group).A. Maternal LiverIV. METABOLISM OF DEVELOPMENTAL TOXICANTSDuring pregnancy, several major shifts from the nonpregnant state occur in the maternal liver. Withsome xenobiotics these changes may not be of consequence to the conceptus if the chemical requiresprior metabolism and the ultimate toxicant is a reactive species with a relatively short biologicalhalf-life. In these cases, xenobiotic metabolism in the maternal liver determines the amount ofparent compound and its stable metabolite(s) reaching the conceptus. Logic dictates that thegeneration of ultimate toxicant must occur in the conceptus to exert toxicity on the embryo or fetus.1. Cytochrome P450a. GeneralThe literature is inconsistent and either no change or pregnancy-related decreases in P450 contenthave been reported. 54–56 Turcan et al. 57 observed a significant decrease in the proportion of P450in the high spin state in hepatic microsomes from pregnant rats. The spin state of the P450 isimportant in overall metabolism. A shift to a lower spin state may alter the redox potential of theNADPH cytochrome P450 reductase–hemoprotein coupling, causing a reduction in electron flowto the P450 and a decrease in the rate of substrate oxidation. Besides changes in the amount ofP450, Lambert et al. 58 reported synthesis of a new major pregnancy-specific form, P450 gest , in thehepatic microsomes of two mouse strains. The elevation of this P450 was maximal by gestationday (GD) 6. Another pregnancy-induced P450 has been identified in the rabbit lung. 59© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 539In animals, a significant increase in maternal liver weight during pregnancy is noted and is aresult of the proliferation of parenchymal cells. 54,60 This increases microsomal yield per liver. 54However, according to Combes et al., 61 one should keep in mind that the liver is not enlarged duringpregnancy in humans. Several reports describe changes in the rates of hepatic microsomal P450-mediated xenobiotic oxidation during pregnancy. 54,55,62–65 For example, in the mouse, the declinesin specific activity for N-demethylation of carbaryl and nicotine were 20% and 50%, respectively,on day 12 of gestation, and remained statistically lower on GD 18. 54 Similarly, the rates of 7-ethoxycoumarin O-dealkylase (ECOD) decreased 11% and 39% on GD 12 and 18, respectively. 54A report 64 suggested no pregnancy-related change in the N,N-dimethylaniline demethylase activityin mouse hepatic microsomes while aminopyrine N-demethylase showed a modest decline. Therates of dearylation of EPN, but not parathion, declined significantly by day 12 of pregnancy butreturned to nonpregnant levels by day 18. 54 Hepatic aldrin epoxidase activity continued to declinetoward term 54,65 but posted no change in total capacity. Parathion undergoes dearylation and desulfurationcatalyzed by hepatic P450. Although the dearylation of parathion exhibited no changein the rate, a 51% decline was noted in desulfurase activity in mouse liver during pregnancy. 66 Ourobservations suggest 54,64,65 that the total capacity of liver for dealkylation of nicotine, carbaryl,aminopyrine, N,N-dimethylaniline, and 7-ethoxycoumarin, and epoxidation of aldrin does notchange as result of pregnancy. Similar observations have been noted for parathion. 54,67b. InductionIn pregnant golden hamsters, ethanol treatment resulted in a 50% increase in CYP2E protein. 56Treatment of pregnant mice with polychlorinated biphenyls (PCBs) significantly increased thedimethylnitrosamine demethylase activity. 63 Pretreatment of pregnant rats with 2,3,7,8-tetrachlorodibenzo-p-dioxin(TCDD) on GD 10 has been reported to cause an approximate 15-fold inductionof maternal liver AHH on GD 20. 62 Cytochrome P450 IIIA1, which is normally undetectable inthe maternal liver, was significantly elevated at GD 15 to 21 following pregnenolone-16-carbonitrile(PCN) exposure of pregnant rats. 68Many reports have described the effect of chemical pretreatment of pregnant animals on theteratogenicity of test chemicals. Qualitative and quantitative changes in P450s were believed to beresponsible for the observed results. Cyclophosphamide teratogenicity depends on its bioconversionto phosphoramide mustard and/or acrolein. Rodent embryos are incapable of this activation andrequire supplementation with adult liver S 9 fraction as an activation system in the culture medium.More effective activation was noted with a liver S 9 fraction prepared from adult male rats pretreatedwith phenobarbital or a mixture of PCBs other than that from untreated or 3-methylcholanthrene(3-MC) pretreated rats. 69,70 In contrast to these results, an S 9 fraction prepared from adult male ratspretreated with either 3-MC, Aroclor 1254, or isosafrole, but not from phenobarbital or PCN, wasfound to be more effective in enhancing 2-acetylaminofluorene (AAF) dysmorphogenesis in ratembryos. 70,71 The required P450s are not induced in the maternal liver by pretreatment of rats with3-MC, Aroclor 1254, isosafrole, phenobarbital, or PCN. Transplacental exposure of GD 8 embryosto 3-MC resulted in a high percentage of malformed embryos following explantation on GD 10 ina medium containing AAF. 70,71For several reasons, the approach of using inducers in studies of chemical teratogenicity isseriously flawed, and considerable caution should be exercised in the interpretation of results ofsuch models. Under in vivo conditions, it is difficult to identify the enzyme responsible forbioactivation of the test chemical. For example, cyclophosphamide oxidation is not only mediatedby P450, but also by PGS 72 and LO, 73 and various chemicals also induce LOs. 29,30 The P450 inducers3-MC, BP, or phenobarbital are themselves teratogenic and thus may weaken and predisposeembryos to the teratogenicity of a test agent. Thus, the observed exaggerated response may not bedue to the test agent alone but may reflect a potentiative or synergistic interaction between inducerand test chemical. Furthermore, the residues of inducers and/or their toxic metabolites occur as© 2006 by Taylor & Francis Group, LLC

540 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONcontaminants in the liver microsomes or S 9 fractions of pretreated animals. Therefore, their use forin vitro embryo culture models may alter the response of embryos to test chemicals.2. Flavin-Containing MonooxygenaseRelatively little published information exists describing the dynamics of hepatic FMO-catalyzedxenobiotic metabolism during pregnancy. Although the hepatic dimethylaniline N-oxidase activitydid not change, lung FMO increased by 57% in 28-day pregnant rabbits. 74 In mice, no pregnancyrelatedchanges were noted in N-oxidase activity in liver, lung, or kidney microsomes 64 or for thehepatic S-oxidation of phorate. 543. Epoxide HydraseThe microsomal epoxide hydrase activity in guinea pigs toward styrene oxide declined 30% to49% in intestinal preparations at GD 54 and 67, 75 while a spurt of activity (a 20% to 67% rise)was noted in the lung, intestine, and kidney microsomes of 20-day pregnant rabbits. 75 No suchincrease in epoxide hydrolase activity was noted for the liver.4. UDP-Glucuronosyl TransferaseIn rats, pregnancy appears to decrease p-nitrophenol glucuronidation by 50% by GD 20, as comparedto the activity 8 days after parturition. 62 The maternal enzyme is inducible and increasedabout 15-fold at term when rats were treated with 5 µg of TCDD on the tenth day of pregnancy.5. Glutathione Transferasea. GeneralPregnancy did not alter maternal liver GST in mice when CDNB was used as the substrate. 76 Similarresults were reported for rat liver GST activity toward 3,4-dichloro-1-nitrobenzene (DCNB) atmidpregnancy. 77 However, the liver cytosol from 19- or 20-day pregnant rats exhibited a significantincrease in CDNB and 1,2-epoxy-3-(p-nitrophenoxy) propane conjugation and a decrease in DCNBconjugation activity. The authors also noted no change in GST activity toward p-nitrobenzylchloride.77 GSH conjugation of sulfobromophthalein in rats 78 and of styrene oxide in rabbits 75 isdepressed in rats because of pregnancy. Mitra et al. 79 demonstrated that in rats, the addition ofpurified maternal liver GST and GSH to the culture medium markedly enhances 1,2-dibromoethane(DBE) toxicity to embryos. In the absence of exogenous GST, the same concentrations of DBEand GSH caused only marginal toxicity, suggesting the absence of GST in rat embryos. Since thematernally generated toxic episulfonium ions of DBE do not escape the liver, in vivo bioactivationof DBE is expected to produce maternal liver toxicity but not embryotoxicity.b. InductionHepatic GST activity was noninducible by treatment of pregnant rats with phenobarbital. 80 However,induction of hepatic GST was reported in pregnant rats and mice exposed to 5,6-benzoflavone. 81Kulkarni et al. 76 observed a threefold induction of hepatic GST activity toward CDNB in 18-daypregnant mice exposed to trans-stilbene oxide, a potent inducer of the enzyme.B. Uterus and Decidua1. Cytochrome P450During GD 7 to 10, the weight of the rat embryo amounts to less than 1% of the whole implantationsite (embryo plus decidua). 82 No AHH activity could be detected in homogenates of whole implantation© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 541sites (GDs 7 through 10) or of the decidua (GD 11) of control rats. However, induction occurredwhen pregnant rats were pretreated with BP 24 h before sacrifice. 822. Peroxidasea. GeneralRat uterine peroxidase has been purified and characterized. 83 Nelson and Kulkarni 84 noted thatmembrane-bound uterine peroxidase activity was very high in nonpregnant mice but declinedrapidly with the onset of pregnancy, and only 0.2% remained by GD 18. An earlier study 85 reportedthat total (membrane-bound and luminal-soluble) uterine peroxidase activity is initially low butincreases on GD 7 in rats and mice, and this high level is maintained throughout pregnancy. Also,uterine peroxidase has been shown to oxidize estradiol to polymeric products capable of bindingto protein. 86 Byczkowski and Kulkarni 87 noted the generation of BP-7,8-dihydrodiol-9,10-epoxide,a postulated ultimate teratogen in animals, 88 from BP-7,8-dihydrodiol (BP-diol) by rat uterineperoxidase.b. InductionJellinck et al. 89 studied the characteristics of estrogen-induced peroxidase in mouse uterine luminalfluid. The peroxidase activity was stimulated by tyrosine and 2,4-dichlorophenol, and evidence waspresented establishing uterine peroxidase-catalyzed oxidation of diethylstilbestrol (DES). 90 A semiquinonemetabolite was presumed to be the primary intermediate produced during one-electronoxidation of DES by the peroxidase. In contrast to the widely known inducibility of peroxidase byestrogenic compounds, treatment of young female rats with 2,3,4,7,8-pentachlorodibenzofuran orTCDD was found to significantly lower the constitutive uterine peroxidase activity in a dosedependentmanner. 91 In cotreatment studies with 17-estradiol, TCDD antagonized the increase inuterine peroxidase activity.3. Lipid PeroxidationIn general, lipid peroxidation in the rat uterus is decreased during gestation. To eliminate anycontribution from conceptual tissues, investigators studied lipid peroxidation in the rat uterusfollowing deciduoma induction by intraluminal injection of arachis oil on GD 5. 92 The endogenouscontent of thiobarbiturate-reactive substances decreased significantly in uterine tissue duringdecidua induction. The lowest values were observed on GD 11 and 14, and the values increasedthereafter. Ascorbate-supported lipid peroxidation decreased during the progressive phase of inductionand increased later, during the regressive phase. Cumene hydroperoxide–triggered lipid peroxidationwas low, while that supported by NADPH did not change. Low levels of peroxidizedlipids were noted in human uterine microsomes in the absence of any exogenous cofactors. 93 Thepresence of this basal level of thiobarbiturate-reactive substances in the microsomes suggests invivo lipid peroxidation. An increase in malondialdehyde production was observed when increasingamounts of ascorbate, ferrous iron, and arachidonate were added to the incubation medium.4. Lipoxygenase and Prostaglandin SynthaseThe results of radioimmunoassays 94 indicated that leukotriene generation in rats remained unalteredduring days 1 to 3 but exhibited a marked increase on day 4, with a peak at noon. This was followedby a sharp decline on day 5, prior to implantation. Prostaglandin biosynthesis displayed an essentiallysimilar pattern. At midpregnancy in rats, myometrium incubated with the attached deciduareleased more prostanoids into the incubation medium than when incubated without decidua. 95 As© 2006 by Taylor & Francis Group, LLC

542 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONpregnancy progressed to GD 21, more release of prostanoids was noted. In animals, uterineprostaglandin release is pulsatile and can be stimulated by oxytocin or estradiol. 96 However, thishormonal stimulation is not sustained and diminishes even when the stimulus is applied continuously.This decline may be related to the liberation and availability of the substrate, arachidonicacid, from the cell membranes.In the nonpregnant human uterus, both endometrium and myometrium exhibit cyclooxygenaseas well as 5- and 12-LO activities toward arachidonate. 97 Prostanoid synthesis has been studied inhuman endometrium obtained at different phases of the ovarian cycle and in decidua in earlypregnancy. 98 Endometrial prostaglandin production was maximal during the midsecretory phase,while it was markedly suppressed in decidua. Essentially similar conclusions have been reachedby others. The production of arachidonic acid metabolites via cyclooxygenase and LO pathwaysoccurs in decidua vera at term. 99 Rees et al. 100 observed the release of different leukotrienes whendecidua samples obtained at term with different modes of delivery were examined by use of ashort-term incubation technique. Recently, Lei and Rao 101 noted the presence of immunoreactive15-LO in rough endoplasmic reticulum and found 15-HETE in myofilaments. Quite unexpectedly,both were also present in nuclear chromatin. Flatman et al. 102 purified and characterized 12-LOfrom human uterine cervix and found the highest specific activity to be associated with microsomes.5. Flavin-Containing MonooxygenaseSo far only one report has described the changes in microsomal FMO activity during pregnancy.Osimitz and Kulkarni 64 reported a decline in FMO activity in the mouse uterus during pregnancy.The activity decreased by 56% and 28% in 12- and 18-day pregnant mice, respectively, as comparedwith that of nonpregnant animals.C. Intrauterine Conceptual Tissues1. Animal SpeciesWhole rodent embryo culture is one of the most popular in vitro test systems employed to investigatethe developmental toxicity of xenobiotics. Even though the use of only the embryo proper is implied,in reality, preparations usually not only include embryo, but also ectoplacental cone and chorioallantoicplacenta, visceral yolk sac, and amniotic membranes. According to Schlede and Merker, 82decidua cannot be excised from 7- to 10-day-old rat embryos. Although a mixture of tissues iscommonly used, there appears to be no strong rationale behind the common practice of not includingother tissues of close proximity, such as the parietal yolk sac and Reichert’s membrane. In anycase, one should realize that the biological response observed following exposure to the testchemical is a net outcome of metabolic interactions of the xenobiotic with all the tissues involved.It is noteworthy that drug metabolizing enzymes, such as P450, are relatively more active andabundant in the extraembryonic tissues (especially visceral yolk sac) than in the embryo proper. 5 Asimilar argument can be made in the case of the chick embryo in ovo test system. Availability ofinformation on the collective ability of analogous preparations of HICT (human intrauterine conceptualtissues: embryo, together with amnion, chorion, placenta, decidua, and umbilical cord) to metabolizexenobiotics would be of great value in understanding the developmental toxicity of chemicals inhumans. Limited data collected in our laboratory are available on this aspect (detailed later).2. Humansa. PeroxidaseMembrane-bound peroxidase was isolated and purified from pooled samples of HICT of 4 to 6weeks’ gestation. 103 The multistep procedure involved CaCl 2 extraction of membranes and affinity© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 543chromatography on Concanavalin A sepharose 4B, followed by hydrophobic chromatography onphenylsepharose CL 4B. This yielded preparations of HICT peroxidase with relatively high specificactivity (15 µmol tetraguaiacol formed/min/mg protein), reflecting a 44% yield of activity and 95-fold purification. Besides guaiacol, the purified embryonic enzyme was able to oxidize all of thetested compounds (pyrogallol, o-dianisidine, p-phenylenediamine, N,N,N,N-tetramethyl phenylenediamine[TMPD], benzidine, 3,3′,5,5′-tetramethyl benzidine [TMBZ], epinephrine, bilirubin, andthiobenzamide) at a significant rate in the presence of H 2 O 2 . 103 Qualitatively similar results wereobserved with peroxidase isolated and purified from pooled samples of HICT of 10 weeks’ gestation.104 Further study 105 has shown that the teratogenic arylamine 2-aminofluorene (2-AF) is easilyoxidized by the purified HICT peroxidase. Kinetic data obtained under optimal assay conditionsyielded a K m value of 41 µM for 2-AF and a V max value of 1.2 µmol/min/mg protein. The reactiveintermediates generated during the oxidation of 2-AF were capable of binding to calf thymus DNAand protein (bovine serum albumin, BSA). Covalent binding of radioactivity (nmol of [ 3 H]2AFequivalent bound/min/mg enzyme protein/mg of DNA or BSA) occurred at the rate of approximately1.90 and 3.75 to DNA and to protein, respectively.b. LipoxygenaseLO activity has been partially purified from the cytosol of HICT at 4 106 or 8 to 10 104,107 weeks ofgestation by affinity chromatography. Besides model compounds, 104 the purified HICT LO wascapable of activation of teratogens in the presence of PUFAs. Thus, the enzyme easily epoxidizedaflatoxin B 1 (ATB 1 ) in the incubation media supplemented with linoleic acid. 107 It is interesting tonote that the rate of epoxidation was greater than that noted with purified term placental LO. Thepurified LO isolated from HICT of 4 weeks of gestation was also capable of oxidation of theproximal animal teratogen, BP-diol. 106 Both formation of the putative ultimate toxicant BP-diol-9,10-epoxide and significant covalent binding of radioactivity to protein were observed during thereaction supported by linoleic acid. 106c. Glutathione TransferaseDatta et al. 108 purified GST activity present in cytosol of HICT of 6 to 10 weeks of gestation usingaffinity chromatography and examined its ability to form GSH conjugates with xenobiotics. A 200-fold purification was achieved when GST activity toward CDNB was assayed. There was nosignificant difference in the specific activity toward CDNB among the HICT GST preparations of6-, 8-, and 10-weeks’ gestation. Besides CDNB, the enzyme was found to be active toward DCNB,ethacrynic acid, p-nitrobenzyl chloride, and 4-nitropyridine-oxide. Additionally, the enzyme alsowas capable of DBE activation. The rate of GSH conjugation of DBE was about twofold greaterwith HICT-GST of early gestation (6 weeks) as compared with those from later periods (8 and 10weeks). The covalent binding of [ 14 C]DBE to protein was approximately 10-fold higher when thereaction was mediated by HICT-GST from 6 weeks gestation compared with that from 10 weeksgestation. Taken together, these results strongly suggest that DBE may be a developmental toxicantin humans.D. Embryo1. Cytochrome P450a. GeneralFlint and Orton 109 evaluated the ability of 46 compounds to inhibit differentiation of rat midbrainand limb bud cells in culture. No increase in the extent of inhibition was observed with 24 out of© 2006 by Taylor & Francis Group, LLC

544 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION27 active compounds in the presence of rat liver S 9 fraction, suggesting that rat embryonic cellspossess enzymes capable of xenobiotic activation. Several attempts by other investigators to definitivelyidentify P450s in rodent, human, and chick embryos and extraembryonic tissues have metwith limited success. Recent reviews 2,5 of the published data (most of which are indirect, conflicting,and/or negative) on this subject have finally concluded that constitutive functional P450 enzymescapable of xenobiotic oxidation have not been definitely identified in rodent or human embryosduring organogenesis. It appears that, with a possible exception of P4503A7, the constitutive formsof P4501A1, P4501A2, and probably P4502E1 and P4502B are either absent or their signal is soweak that it can not be detected by ultrasensitive probes such as reverse transcriptase–polymerasechain reaction (RT-PCR). Thus, the marginally positive results claiming P450 involvement reportedby some investigators appear to be of questionable significance and relevance to the bioactivationof teratogens during human organogenesis.b. InductionAlthough a few studies 5 have reported the inductive effects of 3-MC, β-naphthoflavone, and PCNon rodent embryonic xenobiotic oxidation, published data are too scant for other inducers. P450sin rodent embryos are refractive to induction following exposure to phenobarbital. 70,71 However,those from MC-pretreated dams were capable of measurable oxidative biotransformation of AAF,suggesting P450 induction. 70,71 In preimplantation stage mice, activation of BP, possibly byCYP1A1, can be detected on GD 4. 110 Mouse embryos preexposed to 3-MC exhibit a higherincidence of BP-induced fetal abnormality, suggesting possible induction of CYP1A1. 111 Interestingly,the incidence of fetal anomalies was much lower in noninducible littermates. The signals forP4501A1 mRNA were detectable in conceptal tissues from 3-MC–treated rats only on GD 12. 112However, the intensity of the signal was lowest in tissues of the embryo proper. This casts doubtsregarding the presumed activation of teratogenic chemicals by embryonic P450s. In light of otherknown pathways, Wells and Winn 3 have recently questioned whether the elevated levels of embryonicP450 are sufficient to mediate teratologically relevant bioactivation of xenobiotics. Treatmentof eggs with 3,3′,4,4′-tetrachloro-biphenyl has been shown to result in a 2- and 14-fold increase inECOD and AHH activities, respectively, in livers from 5-day-old chick embryos. 1132. Prostaglandin SynthasePGS activity occurs in mouse 114 and rat 115 blastocysts before implantation. The liberation ofprostaglandin 6-keto-PGF 1 , PGE, and PGF from endogenous arachidonate was noted in GD 11 ratembryo homogenates. 116 As expected, nonsteroidal antiinflammatory drugs, such as aspirin,indomethacin, salicylate, meclofenamic acid, ibuprofen, 5,8,11-eicosatriynoic acid (ETI), and5,8,11,14-eicosatetraynoic acid (ETYA), caused dose-dependent inhibition of prostaglandin synthesis.Nevertheless, the authors opined that the inhibition of PGS is not likely to be the cause ofaspirin-induced limb defects in rats. Recent reports 117,118 have described the occurrence of PGSactivity in mouse embryos. Even though, the presence of PGS in human embryos is expected, itappears that relevant published data on this subject are not currently available.A series of investigations reported by Wells and coworkers provide compelling evidence implicating,in part, embryonic PGS as one of the major enzymes responsible for activation of phenytoinand other teratogenic compounds. The results supporting the hypothesis that phenytoin teratogenicityinvolves a free radical mechanism include oxidation of DNA and covalent binding toprotein, 118,119 oxidation of GSH to GSSG, 118,120 and partial protection from phenytoin-induced cleftpalate formation in mice when dams were pretreated with either EYTA, 120 caffeic acid (a dualinhibitor of PGS and LO), and aspirin (a cyclooxygenase inhibitor) or with α-phenyl-N-t-butylnitrone (PBN, a spin trap agent for free radicals). 121 Purified PGS in the presence of arachidonategenerates reactive metabolites from phenytoin that bind covalently to proteins and can be detected© 2006 by Taylor & Francis Group, LLC

y ESR spectrometry. 122 Further ESR studies 123 revealed that the putative unstable nitrogen-centeredfree radical and a stable carbon-centered radical are formed from phenytoin by PGS. A correlationwas noted between the amounts of free radical adducts formed by PGS and the teratogenicity ofseveral chemicals. Thus, highly teratogenic dimethadione produced substantially more free radicalsthan its minimally teratogenic parent compound, trimethadione. Phenytoin and its p-hydroxylatedmetabolite, 5-(p-hydroxyphenyl)-5-diphenylhydantoin, yielded similar amounts of free radicals.Both D- and L-isomers of mephenytoin and its N-demethylated metabolite generated differentamounts of free radicals. Among the chemicals that are less potent teratogens in humans, phenobarbitalyielded only minimal amounts of free radicals, while no free radical formation was detectablefrom carbamazepine. 123 Incubation of mouse embryos in culture with BP decreases anterior neuroporeclosure, turning, yolk sac diameter, and somite number, and similar to phenytoin, triggersDNA and protein oxidation. 118 Superoxide dismutase addition provides complete protection, suggestingthat BP embryotoxicity in mice may be mediated by ROS generated by free radicalsproduced during peroxidative metabolism of BP. Similarly, free radicals generated by peroxidasefrom thalidomide, the most potent known human teratogen, initiate teratogenic insult and DNAoxidation in rabbit embryos. These effects are abolished by the pretreatment with either PBN 124 oraspirin. 125Two laboratories have independently documented the existence of arachidonate LO activity in ratembryos. Earlier, Vanderhoek and Klein 126 presented evidence for the existence of LO activity inrat embryos. Our laboratory has not only confirmed the occurrence of LO in 9- and 10-day-old ratembryos but has also established the importance of this enzyme system in xenobiotic metabolism. 127The enzyme was found to oxidize benzidine, O-dianisidine, TMPD, and TMBZ in the presence oflinoleic acid. Although direct evidence for the presence of a LO pathway in the human embryoproper is still lacking, our earlier studies have shown an abundance of LO activity in HICT duringorganogenesis. The enzyme preparations were capable of oxidation of xenobiotics at biologicallysignificant rates (see Section IV.C.2.b for details).The hypothesis was tested that phenytoin teratogenicity in mouse embryos may, in part, resultfrom peroxidative bioactivation of phenytoin by LO to a toxic reactive free radical intermediate.3,120,128,129 Embryoprotective effects were observed when mouse embryos were cultured in vitroin the presence of ETYA, a dual inhibitor of the PGS and LO pathways. The combined treatmentresulted in a significant increase in yolk sac diameter, somite development, and crown-rump length,relative to phenytoin alone. These results suggest that embryonic PGS and/or LO play a major rolein phenytoin embryotoxicity. Direct evidence supporting phenytoin activation by LO is also available.128,129 It was shown that both arachidonate and linoleic acid support phenytoin oxidation bysoybean LO, leading to covalent binding of reactive intermediates to protein. As expected, themagnitude of covalent binding was diminished by the addition of LO inhibitors, such as NDGA,quercetin, BW755, or ETYA.ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 5453. Lipoxygenase4. PeroxidaseBy means of the diaminobenzidine staining reaction, Balakier 130 was able to detect endogenousperoxidase activity in early mouse embryos. The bulk of the activity was localized inside or closeto the numerous apical vacuoles in the endoderm. Ectoderm, mesoderm, ectoplacental cone, andtrophoblast cells did not contain peroxidase activity. Using guaiacol as a model substrate and H 2 O 2 ,Nelson and Kulkarni 84 observed easily measurable peroxidase activity in calcium extracts of mouseembryo membranes. The activity increased about sixfold between GD 12 and 14. Although publisheddata are not available for the human embryo proper, HICT isolates were found to possessconsiderable peroxidase activity at 4 to 6 weeks gestation. 103 Preparations of partially purified© 2006 by Taylor & Francis Group, LLC

546 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONenzyme, free of Hb contamination, were capable of oxidizing several xenobiotics, such as guaiacol,thiobenzamide, 3,3′-dimethoxybenzidine, TMPD, TMBZ, 2,2′-azinobis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), benzidine, pyrogallol, epinephrine, bilirubin, p-phenylenediamine, and 2-AF.E. FetusThe subject of xenobiotic metabolism in mammalian fetal tissues has been reviewed periodically.6–8,131–133 The following is a brief synopsis of the literature.1. Cytochrome P450a. GeneralIn rats, the total P450 content in fetal liver microsomes is approximately 10% of that in the maternalliver. 134 Several studies have suggested that the capacity of rodent fetal liver to mediate xenobioticoxidation is much lower (approximately 5%) than that observed in adults, and the activity becomesmore easily detectable toward term. However, information regarding the identity of P450s involvedin these reactions is still unclear. In the case of human fetal liver, early efforts were primarilyfocused on revealing the ability of various preparations to mediate xenobiotic oxidation. By use ofcell cultures, homogenates, postmitochondrial supernatant, or microsomes, several drug substrateswere reported to be oxidized by human fetal liver. 7,8,131,132,135 The total P450 content in the humanfetal liver varies from 20% to 70% of the level in adult liver. 131,132 Compared with rates in adult humanliver, the reported rates of aromatic and aliphatic hydroxylation and N-demethylation of differentxenobiotics in human fetal liver are generally very low. However, in the case of ethylmorphine demethylation,fetal liver microsomal activity may approach the levels noted in adult liver. 136,137Although the ability of human fetal liver microsomes to oxidize xenobiotics has been knownfor some time, very little was clear about the biochemical characteristics of P450s until purificationattempts by Kitada et al. 135 and Komari et al. 138 Now, the presence of several P450 enzymes isknown. Very low levels (about one-tenth the adult level) of CYP1A1 protein are expressed before20 weeks’ gestation. The enzyme is associated with ECOD and hydroxylation of AAF, and theactivities are inhibited by antibodies to CYP1A1 and 7,8-benzoflavone. 139,140 CYP1A2 and CYP1B1appear to be expressed at negligible levels in the human fetus. 141 As regards the expression of CYP2Cenzymes, CYP2C8 may be present as a major form, 139 while the concentration of 2C9, and 2C19may be extremely low or undetectable. 142 Although the expression of CYP2D6 is detectable at 11weeks of pregnancy by RT-PCR, 139 functional protein is found later after 20 weeks of gestation. 143CYP2E1 mRNA was detectable in fetal liver after 16 weeks gestation. 144 Both protein level andethanol oxidation rate are elevated in cultured fetal hepatocytes following exposure to ethanol orclofibrate. The results of RT-PCR probe and RNase protection assays suggest the presence ofCYP3A4 145 and possibly CYP3A5. 146 CYP3A7 is the major constituent that accounts for aboutone-third of human fetal liver P450. It has been characterized and purified. 147,148 This form is shownto be capable of metabolism of about a dozen chemicals and plays a major role in fetal xenobioticdealkylation or activation.b. InductionIn general, P450 responsible for xenobiotic oxidation in animal fetuses is not readily inducible bychemical exposure of the mother during pregnancy. This appears to be particularly true for phenobarbital-typeinducers, such as p,p-DDT. 149 Some elevation in certain monooxygenase activitiescan be detected if the maternal exposure to specific inducers (e.g., 3-MC) occurs near term.Simultaneous exposure to both types of inducers usually produces additive effects. A study reported© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 547an approximate fourfold increase in P450 content exclusively in fetal liver on GD 21 when 3-MCwas administered to pregnant rats on GD 18 and 19. 134 However, the reduced CO difference spectrumexhibited a peak at 450 nm, while the maternal liver microsomes displayed a peak at 448 nm, asexpected.AHH in fetal rat liver is induced following treatment of pregnant rats with TCDD. 62 Inductionwas evident 3 days prior to birth when pregnant rats were administered TCDD on GD 10 or 16but not on GD 5. TCDD treatment did not change the fetal liver microsomal P450 content.Dimethylnitrosamine demethylase activity in fetal mouse liver was low but readily detectable onGD 16. 63 When pregnant mice were treated with PCBs on GD 14, the fetal livers showed a significantincrease in enzyme activity on GD 16. Induction was not detectable in fetal liver when assayed onGD 19, after treatment of mothers on GD 15 or 17.Sherratt et al. 150 investigated induction of CYP1A1 in cultured rat fetal hepatocytes and foundthat exposure to 1,2-benzanthrecene alone resulted in a 60-fold increase. The inductive effect waspotentiated about threefold when dexamethasone was included in the culture medium. CYPIIIA1in fetal rat liver is inducible by PCN as early as GD 15. 68 Ethanol treatment of pregnant hamsterswas found to result in a twofold increase in the CYP2E protein in fetal livers. 56 Recently, Srivastavaet al. 151 reported a significant decrease in fetal hepatic microsomal P450 content and aminopyrineN-demethylase, aniline hydroxylase, and AHH activities on GD 20 following daily exposure ofpregnant rats to styrene.Negative findings about inducibility of P450 in the human fetus should not be construed asabsence of a mechanism of induction, considering the fact that in vivo exposures to chemicalsduring pregnancy are generally low. At midgestation, phenobarbital exposure of mothers or cigarettesmoking failed to induce fetal AHH or N–methyl aniline demethylase activities. 131 On the otherhand, induction of prazepam metabolism was indicated when cultured fetal liver explants wereexposed to phenobarbital. 152 AHH activity is induced when cells are exposed to BP, 3-MC, andbenz(a)anthracene. 131 In monolayer culture of human fetal hepatocytes, phenobarbital and α-naphthoflavoneexposures caused up to eightfold induction of AHH activity, while t-stilbene oxide orbenzanthracene treatment had no effect.2. Flavin-Containing MonooxygenaseIn human fetal liver, N-oxidation of N,N-dimethylaniline by FMO can be detected in microsomesas early as 13 weeks of gestation. 153 There appears to be no correlation between the rate of N-oxidation and the smoking habit of the mother.3. Prostaglandin SynthaseAlthough systematic studies on the role played by PGS in xenobiotic metabolism have not appearedin the literature, available data strongly suggest that this pathway is operative in human 154–156 andanimal 157 fetal tissues.4. PeroxidaseIn the mouse, peroxidase activity toward guaiacol in the presence of H 2 O 2 was found to increaselogarithmically between GD 12 and 18, resulting in a 50-fold increase. 84 A 400-fold purificationof peroxidase from whole human fetuses obtained at gestation weeks 8 to 10 has been reported. 1585. LipoxygenaseAnalyses of arachidonate metabolites suggested functional 5-, 12-, and 15-LO pathways in therabbit fetus. 157 Human fetal tissues exhibiting LO activity include lung, 154 adrenal, brain, heart,© 2006 by Taylor & Francis Group, LLC

548 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONkidney, and liver. 156 Partially purified LO from the cytosol of HICT at 8 to 10 weeks of gestation. 104,107oxidized model compounds, 104 and epoxidized ATB 1 in the presence of linoleic acid. 1076. Epoxide HydraseA study on prenatal changes in microsomal epoxide hydrase activity toward styrene oxide revealedthat the enzyme specific activity in rabbit fetal liver at GD 20 is less than 30% of the adult value. 75The activity was low in intestine, lung, and kidney on GD 28. In guinea pigs, fetal hepatic andintestinal epoxide hydrase activities were less than 10% of adult values, and levels of the enzymein lung and kidney were too low to measure. Phenobarbital caused dose-dependent induction (1.2to 3.1-fold) in human fetal hepatocytes obtained at 17 to 20 weeks’ gestation. 1597. Glutathione TransferaseHepatic GST titer is virtually undetectable in rodent embryos and early fetal liver, low near term,and reaches adult levels within a few weeks postpartum. 75,78,80,149,160 It appears that fetal GST activityis not induced when pregnant rats are treated with either DCNB or p,p-DDT. 149,160 On the otherhand, daily exposure of pregnant rats to styrene was found to be associated with a significant declinein GST activity in the fetal liver on GD 20. 151GST is abundant in human fetal liver. However, controversy surrounds the presence of acidic,basic, and neutral forms. 161–163 Mitra et al. 164 purified five GST isozymes from fetal liver (16 to 18weeks’ gestation) cytosol by affinity and ion-exchange HPLC. Two anionic forms with pI valuesof 5.5 (P-2) and 4.5 (P-3) and one basic form with the pI value of 8.7 (P-6) were clearly separated,and small amounts of two near neutral forms (P-4 and P-5) were also identified. Earlier, Radulovicet al. 165 studied the metabolism of methyl parathion by GST purified from human fetal livers at 14to 23 weeks of gestation. Methyl parathion was detoxified by O-dealkylation. HPLC, TLC, andradiometry indicated desmethyl parathion as the sole metabolite. Kulkarni et al. 166 reported theGSH conjugation of DBE by human fetal liver GST. Subsequently, 164 a novel model was devisedto assess the likelihood of developmental toxicity of DBE in humans. In the proposed in vitromodel, rat embryos (lacking GST) served as passive targets, and DBE bioactivation was mediatedby purified isozymes of human fetal liver GST. All of the tested human fetal liver GST isozymeswere found capable of bioactivating DBE. Covalent binding of radioactivity from DBE to DNAand protein suggested DBE bioactivation by GST, which was 144% and 212% higher, respectively,with the P-3 anionic form compared with the P-6 basic isoform. DBE bioactivation by the GST P-3 isoform resulted in developmental toxicity to cultured rat embryos. The results of this studysuggest that DBE may be a human developmental toxicant.8. UDP-Glucuronosyl TransferaseGlucuronide formation in the liver does not occur until after birth in guinea pigs, mice, and humans, 167,168and is barely measurable with o-aminophenol in rat fetuses near term. 160 In rats treated with TCDD onGD 5, fetal liver UGT activity toward p-nitrophenol and testosterone was first evident at GD 18 and22, respectively. 62 This activity increased rapidly, and a 20-fold increase in activity was observed byGD 22. In humans, UGT activity is measurable toward morphine, eostriol, and 2-naphthol in fetal livermicrosomes. 169 UGT activity in human fetal liver was low, about 0.1% of adult values with bilirubin, 170less than 5% with harmol, 171 and approximately 10% with 2-naphthol and morphine.9. SulfotransferaseCappiello et al. 172 conducted a comparative study of human SULT activity in the cytosolic fractionsof adult and fetal liver and placenta, using 2-naphthol as a substrate. Activity in fetal liver was© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 549approximately 15% of the value for adult liver. Activity of the enzyme, rather than the availabilityof PAPS, was suggested as the rate-limiting factor in the sulfation reaction. Both SULT and PAPSwere higher in fetal liver than in placenta. The rate of sulfation correlated with the rate of 2-naphtholglucuronidation; however, no relationship with gestational age was observed. 169F. Fetal Membranes1. Cytochrome P450In the rat, the fetal membranes mainly consist of the visceral yolk sac and a fine layer of amnion,and these are difficult to separate. 82 Homogenates of fetal membranes from untreated rats occasionallyexhibit detectable levels of AHH activity. Treatment of pregnant rats with BP 24 h priorto sacrifice resulted in a three- to eightfold higher induction of AHH activity in fetal membranesthan in placenta. 822. Peroxidative EnzymesWhen stimulated by calcium ions, rabbit amnion and splanchnopleure produced a mixture of 11-,12-, and 15-HETEs via the LO pathway. 173 Metabolite production was greater on GD 20 to 22 thanon GD 28 to 30. The cyclooxygenase products of arachidonate were more abundantly producedby these tissues in late pregnancy. Earlier studies suggested the operation of the cyclooxygenasepathway in human amnion 174 under the influence of estrogens. 175 Subsequent studies 99,100 showedthe existence of both the cyclooxygenase and LO pathways in human amnion and chorion laeve.The production of leukotrienes by amnion showed no difference when samples were collected afterlabor at term, after preterm labor, or after cesarean section at term. 100 In contrast to this, anotherstudy noted a threefold greater release of PGE 2 by amnion following spontaneous delivery comparedwith cesarean section at term. 176 The opposite was true for PGE 2 in decidua, while no differencein prostaglandin production was observed for chorion.Price et al. 177 studied immunohistochemical localization of PGS in human fetal membranes anddecidua. Decidualized stromal cells were stained most intensely and consisted of all cell types.Cytotrophoblast, as well as early placental villi and syncytiotrophoblast of all gestational ages,demonstrated lighter, more variable staining patterns. Regardless of gestational age, amnion stainedin a heterogeneous fashion, with some cells demonstrating intense staining and others having nostaining. A more recent study 178 indicated that tissues obtained after vaginal delivery releasedsignificantly higher quantities of leukotrienes than did tissues obtained following cesarean section.The leukotriene release was stimulated by oxytocin. Despite the presence of both PGS and LO,there is no published report on xenobiotic oxidation by these enzymes from human fetal membranes.G. Placenta1. Cytochrome P450a. Nonsmoker’s PlacentaNumerous studies on oxidative metabolism of xenobiotics in various preparations of placenta havebeen reported in the literature, and the information has been critically reviewed. 1–4,8–10,131,132 Basedon the results of over 25 years of research, Juchau 10 concluded that the microsomal P450 responsiblefor xenobiotic oxidation is either undetectable or occurs in extremely small amounts, in the placentasof nonsmoking women. Similar opinions have been echoed by other investigators. 1,3,6,8 Low butmeasurable deethylation of 7-ethoxycoumarin or 7-ethoxyresorufin, which is routinely observed inthe placentas of nonsmokers, may be an attribute of CYP19 responsible for steroid aromatization. 179© 2006 by Taylor & Francis Group, LLC

550 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe supportive evidence is weak and inconclusive for the suspected presence of CYP2E1, whilethe occurrence of other functional forms of P450 enzymes (CYP3A, detectable in first trimester,and CYP4B1) in human term placentas is debatable. 8b. InductionIn general, phenobarbital and other barbiturates, as well as ethanol, do not induce placentalxenobiotic enzymes at midgestation or at term. 131 The concomitant induction of P450 and xenobioticmonooxygenation activities in human placental microsomes by the components of cigarette smokehas been known for quite some time. 8–10 There is compelling evidence that only CYP1-1 is inducedas a result of maternal cigarette smoking. Although several enzyme activities are co-induced tosome extent by cigarette smoking, AHH and deethylation of ethoxycoumarin and ethoxyresorufinhave received the most attention. 8 Although a satisfactory explanation is still lacking, marked AHHinduction due to cigarette smoking during pregnancy is observed at term but not during the firstor second trimester. 8–10 Apparently, there is no report on the influence of cigarette smoking duringpregnancy on the terminal epoxidation of BP-diol by human placental microsomes. Wong et al. 180and Lucier et al. 181 studied inducibility of the P450 system in placentas of women exposed to PCBsand dibenzofurans. The results revealed a dramatic elevation (about 100-fold) of AHH activity.This increase in enzyme activity paralleled an increase (150-fold) in the placental microsomalprotein related to P450 form 6 (CYP1A1). No other P450 forms were detected. There was anexcellent correlation between induction of 7-ethoxyresorufin deethylase and AHH activities. Inductionof CYP1A1 and associated enzyme activities has also been noted in human chorionic carcinomacell line JEG-3 after treatment with 3-MC, BP, or TCDD. 82. Flavin-Containing MonooxygenaseBy use of N,N-dimethylaniline as a model substrate, the presence of FMO activity was observedin mouse placental microsomes. 64 The activity increased nearly fivefold when measured at GD 18,in comparison with GD 12. Subsequent experiments indicated that N,N-dimethylaniline N-oxidaseactivity is undetectable or absent in microsomes isolated from human term placentas of nonsmokers.Further evaluations with other substrates are needed before definitive conclusions can be reachedregarding the existence of this enzyme in human placenta.3. Prostaglandin SynthaseEarlier reports suggested the presence of PGS activity, albeit at very low levels, in human andrabbit placentas. 6 It was suggested that the low PGS activity in human placenta may, in part, berelated to circulating endogenous inhibitors during pregnancy. 182 Although the role played by thisenzyme in xenobiotic oxidation has never been rigorously tested in placental microsomes of anyanimal species, the results of a recent study indicate that neither benzidine oxidation nor the PGSantigen can be detected in microsomal preparations from term placentas. 1834. PeroxidaseAt least two studies have shown the presence of peroxidase activity in the placentas of mice andrats, measured with guaiacol as the model substrate. 84,85 The results of the mouse study indicatedthat peroxidase activity does not change during GD 7 through 20. 84Earlier peroxidase activity was described in crude human placental homogenates. 184 This activitywas detectable in human placenta as early as the sixth week of gestation and increased towardterm. Nelson and Kulkarni 47 confirmed the abundance of peroxidase activity in membranes of humanterm placentas from nonsmokers. They described a procedure for the isolation, solubilization, and© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 551partial purification of placental peroxidase free from contaminating heme proteins. The protocolwas refined further to obtain the enzyme with electrophoretic homogeneity. 185,186 The modifiedprocedure yields final enzyme preparations displaying over 110-fold purification compared to theinitial calcium extract of membranes, with an overall recovery of about 55% of enzyme activity.At present it is not known whether exposure to chemicals present in diet, cigarette smoke, etc.,modulates peroxidase activity.Several studies have now established peroxidase as one of the major pathways of xenobioticmetabolism in human placenta. Available evidence suggests that both membrane-bound and partiallypurified preparations of the enzyme display the ability to catalyze oxidation of several xenobioticsin the presence of H 2 O 2 . Hydrogen donors, such as guaiacol, pyrogallol, benzidine, TMBZ, 3,3′-dimethoxybenzidine, TMPD, ABTS, and endogenous chemicals (e.g., bilirubin, epinephrine) areeasily oxidized by the enzyme. 47,185 Thiobenzamide, a hepatotoxicant, appears to be one of the bestsubstrates, and its oxidation to a sulfoxide by purified peroxidase was found to proceed at the rateof about 18 µmol/min/mg protein. 186 Spectrophotometric and radiometric studies have yielded datasuggesting that 2-AF also serves as an excellent substrate and undergoes peroxidase-mediatedoxidation at an impressive rate. 187 It is reported that BP-diol, a proximate human developmentaltoxicant, is epoxidized to generate the putative ultimate toxicant BP-diol-9,10-epoxide by placentalperoxidase. 188,189 The presence of phenylbutazone or amino acids, e.g., tyrosine, was required toobserve a maximal rate of the reaction. 188,189A few studies 190–192 have reported a positive correlation between the use of phenothiazine drugsduring early pregnancy and induction of malformations, while a suggestive relationship of borderlinesignificance in terms of cardiovascular effects was noted in another study. 193 Sobel 194 opinedthat chlorpromazine induces abortion but causes no birth defects in the babies of mothers whoreceived the drug. It is widely accepted that prior oxidation is essential for the expression of thepharmacological and toxicological effects of phenothiazines. Yang and Kulkarni 195 studied in vitrometabolism of seven phenothiazines by purified placental peroxidase. All the phenothiazines testedwere found to undergo one-electron oxidation to their respective cation radicals. These relativelystable species can be detected spectrophotometrically. The highest rate of cation radical formationwas noted with promazine (approximately 350 nmol/min/mg protein), while the lowest rate wasobserved with trifluoperazine. More recently, eugenol, a naturally occurring plant compound widelyused as a food flavoring agent, an additive to fragrances, cosmetics, and tobacco products, and inmedicine, was found to be oxidized to a toxic quinone methide by purified placental peroxidase. 1965. LipoxygenaseThe concentration of free arachidonic acid in most tissues is about 20 µM. 27 Therefore, the availabilityof substrate, arachidonic acid or other PUFAs can be a rate limiting factor for tissue LOactivity. However, this may not be an issue for the placenta because free arachidonic acid alone isabundant (about 0.7 mM) in human term placenta. 197 Considering the questionable activities ofPGS and P450, it appears that arachidonic acid metabolism in human placenta occurs exclusivelyvia the LO pathway. The studies on leukotriene biosynthesis in human and rabbitplacentas 6,100,155,198,199 lend strong support to this contention. Human term placental LO (HTP-LO)has been purified and partially characterized. 199 At present, there is no published report detailingthe effect of dietary substances, ethanol consumption, or cigarette smoking during pregnancy onHTP-LO activity.Aberrant activities of LO and/or PGS in reproductive tissues have been implicated in variouspregnancy related hypertension disorders and preeclampsia. 6,32 The liberation of leukotrienes suggestsin vivo operation of the 5-LO pathway in human placenta. The production of LTB 4 is highin early pregnancy (7 to 12 weeks of gestation) but remains low during the third trimester, with asignificant increase during spontaneous labor at term. 198 The observations that plenty of LO activityoccurs in HICT at 4 weeks gestation 106 and midpregnancy 104 are in line with this report. No© 2006 by Taylor & Francis Group, LLC

552 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONdifference in leukotriene release was observed in placentas examined after labor at term, afterpreterm labor, or after cesarean section. 100 Leukotriene output is low in uncomplicated pretermlabor but markedly increases in chorioamnionitis-associated preterm labor. The fact that 5- and 12-HETE are generated when homogenates were incubated with 14 C-arachidonate signals that functional5- and 12-LO in human placenta is present at spontaneous vaginal delivery or after cesarean section.155,156 In subsequent studies, 100,200 the predominance of the 12-LO pathway was observed. Decreasedformation of 5- and 12-HETE from arachidonic acid was noted in preeclampsic placentas. 200Recently, cellular oxidative stress was proposed as one of the underlying mechanisms ofdevelopmental toxicity of many chemicals. 3,4 Concurrent depletion of cellular GSH plays a majorcontributory role in this process. GSH provides the first line of defense against excessive productionof xenobiotic free radicals within cells. While performing the sacrificial role as an intracellularnucleophilic trap, GSH is extensively consumed. In the absence of GSH participation, an interactionof xenobiotic free radicals with molecular oxygen leads to the liberation of ROS. 3,4 Other investigatorshave also stressed the role played by ROS in the developmental toxicity and teratogenicityof various chemicals. In this regard, it is noteworthy that although heme proteins, such as P450and various peroxidases, lack the ability to oxidize GSH, LO has been found to trigger its directoxidation in the absence of xenobiotics. This phenomenon, first noted with soybean LO in thepresence of linoleic acid, 201,202 is also displayed by HTP-LO in reaction media supplemented withdifferent PUFAs. 203 GSH oxidation by LO is accompanied by superoxide anion production. 201 Moreintense superoxide anion radical generation can be noted during LO-catalyzed linoleate-coupledcooxidation of NAD(P)H, 204 but not of ascorbate, 205 in the absence of oxidizable xenobiotics.The importance of these findings is further heightened by the demonstration that GS • generatedduring PUFA-supported GSH oxidation by LO can attack a C=C bond in xenobiotics, and theprocess ultimately leads to the formation of a GSH-xenobiotic conjugate. This novel mechanismof GSH conjugate formation from xenobiotics has been documented for ethacrynic acid withsoybean LO and HTP-LO in the presence of various PUFAs. 202,203 It is noteworthy that the rate ofsoybean LO-catalyzed ethacrynic acid conjugation with GSH was up to 1650-fold greater than therates reported in the literature for various preparations of purified GST. 202 This novel reaction withHTP-LO proceeds at a biologically significant rate (about 60% of the maximal velocity noted underoptimum assay conditions) under physiologically relevant conditions of pH and concentrations of GSHand arachidonic acid. 203 Further studies have shown that p-aminophenol, a developmental toxicant inanimals, also rapidly undergoes LO-mediated GSH conjugation to generate an ultimate toxic species. 206The study by Joseph et al. 199 has shown that purified HTP-LO cooxidizes pyrogallol, ABTS,benzidine, 3,3′-dimethoxybenzidine, TMPD, TMBZ, and guaiacol in the presence of linoleic acid.The oxidation rate was highest with TMPD and lowest with guaiacol, the most commonly usedperoxidase substrate. Hypervitaminosis-A has been linked with birth defects in animals and humans.Several studies have suggested that the parent molecule per se may not be teratogenic and bioactivationis essential. The study conducted by Datta and Kulkarni 207 suggested that all-trans retinolacetate is an excellent substrate for cooxidation by soybean LO, as well as HTP-LO, in the presenceof linoleic acid. The dioxygenase-generated peroxyl radical of linoleic acid was proposed to attackthe π-electrons of the C=C bond and bring about epoxidation of the all-trans retinol acetate. Earlier,HTP-LO-mediated epoxidation was also shown with the developmental toxicant BP-diol. 106 Significantcovalent binding of radiocarbon from BP-diol and a preponderance of trans-anti-7,8,9,10-tetrahydrotetrol suggested bioactivation of BP-diol by purified LO isolated from nonsmokers’placenta. These observations clearly explain the literature reports of detection of BP adducts inhuman term placentas from nonsmokers. 208 ATB 1 , a substituted coumarin, is a known animal andhuman teratogen. Many reports have documented biologically significant levels of ATB 1 in maternaland cord blood, 209,210 ATB 1 -DNA adducts in placentas and cord blood, 211 and transplacental carcinogenicity.212 It is widely accepted that ATB 1 requires prior epoxidation to cause toxicity. Thenotion that P450 is the sole catalyst for the bioactivation of this toxicant has been questioned. 6 Thepostulate that LO activates ATB 1 finds strong support in the demonstration that purified adult human© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 553liver LO 213 can efficiently perform this reaction. Datta and Kulkarni 107 observed that not only HTP-LO but also LO purified from HICT at early gestation can rapidly oxidize ATB 1 to an electrophilicspecies, the ATB 1 -8,9-epoxide, in the presence of PUFAs. Such findings explain, in part, thedevelopmental toxicity associated with in utero exposure to this fungal toxin.Occupational exposure of workers in the dye industry to arylamines, such as 4-aminobiphenyl(4-ABP) and benzidine, can result in bladder cancer. Currently, there is increasing concern overpossible transplacental carcinogenicity associated with gestational active or passive exposure to 4-ABP present in environmental tobacco smoke. Interestingly, the emission of 4-ABP is 30 timesgreater in sidestream smoke than in mainstream smoke. It crosses the placenta and binds to fetalHb, 214,215 and 4-ABP adducts have been detected in the fetuses of both smoking and nonsmokingmothers. In rats, the presence of detectable levels of DNA adducts has been noted in all fetal tissuesexamined following maternal dosing with 4-ABP. 216It is widely presumed that prior metabolism is obligatory to observation of 4-ABP toxicity. Theknown absence of P450, FMO, and PGS triggered interest in examining the possible metabolismof 4-ABP by HTP-LO. 217 Both soybean LO and HTP-LO oxidized 4-ABP to a stable metabolitethat was identified by gas chromatography–mass spectrometry (GC-MS) as 4,4′-azobis(biphenyl)when the reaction media were supplemented with either PUFAs or H 2 O 2 . HTP-LO was superiorto soybean LO as a catalyst of the reaction. Datta et al. 217 hypothesized that 4-ABP undergoes oneelectronoxidation by peroxy radicals generated by LO from PUFA to produce a free radical speciesthat dimerizes to the coupling azo product. Similar metabolic conversion of 4-ABP by LO from HICTis expected. Collectively, the results suggest that the cooxidation catalyzed by LO may be the underlyingbiochemical mechanism responsible for the transplacental toxicity of 4-ABP in humans.As with other xenobiotics, N-dealkylation of drugs is one of the major detoxication reactions.PUFAs support HTP-LO-mediated N-demethylation of phenothiazines, and thus far, the reactionhas been noted with promazine, chlorpromazine, promethazine, and trimeprazine. 218 Both soybeanLO and HTP-LO can mediate the reaction in the presence of H 2 O 2 . 219 It is noteworthy that besideslinoleic acid and H 2 O 2 , both cumene hydroperoxide and tert-butyl hydroperoxide efficiently coupleLO-mediated N-demethylation of phenothiazines. 49 In addition to studies of phenothiazines, animaldata are available for other drugs taken during pregnancy. Other drugs that undergo H 2 O 2 -dependentN-demethylation under the influence of soybean LO include aminopyrine, 220 imipramine, andrelated tricyclic antidepressants. 221 In addition, model compounds, such as, N-methyl aniline, N,Ndimethylaniline, N,N,N,N-TMBZ, N,N-dimethyl-p-phenylenediamine, N,N-dimethyl-3-nitroaniline,and N,N-dimethyl-p-toluidine are demethylated by HTP-LO in the presence of H 2 O 2 . 222Several organophosphate and carbamate pesticides have been reported to possess teratogenicproperties, and metabolism plays a major role in determining their developmental toxicity. 6 Currently,the information on this subject is scant for human conceptual tissues, especially placenta.Hu and Kulkarn 223 demonstrated that soybean LO detoxifies several pesticides by mediating N-demethylation in the presence of H 2 O 2 . In a subsequent study, purified HTP-LO was shown tosupport cooxidative N-demethylation of aminocarb, chlordimeform, dicrotofos, and zectran at abiologically significant level in the presence of linoleic acid. 218 It was proposed that the initial stepin xenobiotic N-dealkylation by LO involves one-electron oxidation of the substrate to produce acorresponding nitrogen-centered cation free radical. The radical undergoes a rapid decay to forman iminium cation, either by deprotonation or hydrogen atom abstraction. Subsequent hydrolysisof the iminium cation finally yields formaldehyde and a monodemethylated product.6. Lipid Peroxidation-Coupled Cooxidation of Xenobioticsa. Lipid PeroxidationMany studies have documented that pregnancy in human is associated with increased in vivo lipidperoxidation. Uotila et al. 224 found a 45% increase in serum levels of conjugated dienes when© 2006 by Taylor & Francis Group, LLC

554 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONpregnancy advanced from the first to the second trimester. An earlier study 225 reported conjugateddiene levels nearly doubled by 36 weeks of gestation, and the values were about 40% greater inwomen with preeclampsia. The evidence suggests that placenta (and/or other conceptual tissues)may be the major source of lipid peroxidation products during pregnancy, as the values forconjugated dienes return to normal within 24 h of delivery.In contrast to the compelling in vivo evidence, the reports on in vitro lipid peroxidation inplacental microsomes are inconsistent. Although the occurrence of ascorbate-supported nonenzymaticlipid peroxidation has been repeatedly demonstrated in human placenta, 1,6,93 anomalies existin the reports of NADPH-supported reactions. Thus Juchau and Zachariah 226 could not detect lipidperoxidation when human term placental microsomes were incubated with NADPH. Similar negativefindings were reported by Arvela et al. 227 with placental microsomes from rats, rabbits, andguinea pigs. Kulkarni and Kenel 228 pointed out that previous negative findings 226 may have beendue to suboptimal assay conditions employed to evaluate NADPH-dependent lipid peroxidation inhuman placental microsomes. The presence of chelated iron was found to be essential to the process.In fact, the results suggested a high susceptibility of human placental microsomes to lipid peroxidationbecause of an apparent lack of endogenous antioxidants in the microsomes. Several reportsfrom other labs 229,230 have now confirmed the occurrence of NADPH-dependent lipid peroxidationin human term placental microsomes and mitochondria.Many environmental pollutants initiate, promote, or undergo one-electron oxidation underaerobic conditions, and in the presence of lipids they promote lipid peroxidation. Using linoleicacid, the occurrence of the process has been demonstrated for copollutants such as reduced vanadium(vanadyl), hydrated sulfur dioxide (bisulfite), and asbestos fibers (Canadian chrysotile). 231–233Vanadium accumulates in human placenta at a concentration of about 3 ng/g. When added toplacental microsomes in the presence of NADPH, vanadium triggers intense lipid peroxidation. 234,235b. Effect of Redox Cycling AgentsIt has been proposed that redox cycling of certain nitroheterocycles, quinones, and pesticides, suchas nitrophen and paraquat, coupled with NADPH-cytochrome P450 reductase, may represent aserious embryotoxic or fetotoxic threat. 4,6 Mounting evidence points out that lipid peroxidationmay be one of the major contributing factors in the developmental toxicity of the agents listedabove. This contention finds support in the strong stimulatory effects exerted by paraquat onNADPH-supported lipid peroxidation in human placental microsomes or mitochondria. 230,236,237NADPH-dependent redox cycling of menadione in placental microsomes also leads to extensivelipid peroxidation. 237c. Cooxidation of XenobioticsEarlier reports 1,6 have shown that rat liver microsomes undergoing ascorbate- or NADPH-supportedlipid peroxidation are capable of cooxidation of xenobiotics. In light of these reports, the possibilityof lipid peroxidation-coupled cooxidation of xenobiotics in human term placenta was examined.NADPH- and iron-dependent lipid peroxidation of human term placental microsomes supportedcooxidation of both BP and BP-diol to several metabolites. 238 Cooxidation of BP-diol was alsoobserved during placental microsomal lipid peroxidation triggered by vanadium in the presence ofNADPH. 2357. Other Enzymes of Xenobiotic OxidationIt should be borne in mind that not all the enzymes responsible for xenobiotic bioactivation ofdevelopmental toxicants in human placenta are known. Similarly, the presumption that each teratogenrequiring oxidative bioactivation must serve as a substrate for one or more of the known© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 555enzymes described above may not always be true. The reported presence of indanol dehydrogenaseactivity in human term placenta of nonsmokers 239 represents a case in point. Earlier, the results ofan in vitro test suggested indan and 1-indanone to be teratogenic. 240 Thus, the noted rapid oxidationof 1-indanol to 1-indanone by human placental microsomes in the presence of either NAD + orNADP + represents one step in bioactivation. 2398. Epoxide HydraseIn rabbit and guinea pig placentas, microsomal epoxide hydrase activity toward styrene oxide isnear the limit of detection and exhibits a downward trend as pregnancy advances. 75 In humans,such activity is detectable in placenta as early as 8 weeks gestation. Apparently, the enzyme maturesearly during the first trimester since no gestational increase in epoxide hydrase activity was observedlater. 2419. UDP-Glucuronosyl TransferaseLucier et al. 62 measured placental UGT activity toward p-nitrophenol on GD 20 in rats. The activitywas induced threefold when pregnant rats were pretreated with TCDD on GD16, but not whenanimals were treated on GD 5 or 10. The ability of rabbit placenta to form oxazepam glucuronidein vitro was nearly doubled on GD 14 by pretreatment with phenobarbital but was not affected byoxazepam. 24210. SulfotransferaseCappiello et al. 172 measured SULT activity in human placentas at 11 to 27 weeks of gestation. Thevalues obtained represent about 6% of the activity observed in fetal liver. Placental PAPS contentwas estimated at 3.6 nmol/g tissue. The authors opined that the availability of SULT and not PAPSis the rate-limiting factor for sulfation of xenobiotics in human placenta.11. Glutathione Transferasea. GeneralThe occurrence of GST activity has been noted in the placentas of rabbits, guinea pigs, mice, rats,and monkeys. 1,6 In general, the activity measured with different substrates was found to declinegradually toward term. Earlier, conventional methods to purify human placental GST met withlimited success. Radulovic and Kulkarni 52 were the first to introduce a novel, rapid method combiningaffinity chromatography and HPLC for the purification of this GST. Their procedure resultedin high yields of electrophoretically homogeneous enzyme, with about 1500-fold purification. Theenzyme, now identified as GSTP1-1, can exist in two charge isomers. These forms can be separatedby HPLC and are interconvertible, depending upon the concentration of GSH in the surroundingmedium. 243 Since the tissue concentration of GSH was estimated as 0.11 mM, it was postulatedthat the enzyme predominantly exists as a high-affinity–low-activity conformer in vivo. Pretermplacental GST activity toward CDNB has been found to exhibit a twofold variation during weeks13 to 23 of gestation. 244 GST activity toward styrene oxide remains fairly constant between 8 and25 weeks of gestation, followed by a decline of about 50% at full term. 245 A significant inhibitionof the enzyme’s activity toward CDNB and other model substrates can be observed at relativelylow concentrations of Hb, 53 PUFAs, and their esters, 246 a variety of retinoids, 247 and nitrobenzenes. 248Besides being able to metabolize CDNB, DCNB, p-nitrobenzyl chloride, 4-nitropyridine-Noxide,and ethacrynic acid, 243–248 the purified GST1-1 can metabolize methyl parathion, an animalembryotoxicant and teratogen. 244 HPLC analyses revealed the presence of desmethyl parathion as© 2006 by Taylor & Francis Group, LLC

556 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthe sole metabolite from O-dealkylation. The metabolism of DBE by purified GST1-1 generatesreactive species that bind covalently to protein and DNA. 249 Rat embryos exhibit an inhibition ofgrowth and development when DBE is bioactivated in the culture medium by purified humanplacental GST.b. InductionIt appears that human placental GST activity is not altered following exposure of women to theconstituents of tobacco smoke. 245,250 Studies with laboratory animals have also indicated the refractivenature of placental GST to the effects of such inducers as trans-stilbene oxide, 76 phenobarbital,AAF, butylated hydroxyanisole, and 3-methyl-4-dimethylaminobenzene. 251V. ROLE OF XENOBIOTIC METABOLISM IN REPRODUCTIVE TOXICOLOGYA. GeneralConcerns regarding possible ill effects of chemicals present in the working and living environmenton human reproduction have markedly increased over the years. The pathways of hormone biosynthesis,gametogenesis, and secretory functions of accessory sex glands represent some of theobvious targets for reproductive toxicants. A multitude of mechanisms are expected to be responsiblefor reproductive failure. Unfortunately, in most cases, cause and effect relationships are not firmlyestablished. Although biotransformation of toxicants is fairly well understood, especially in animals,there is paucity of information on the metabolism of toxicants in the affected cell types and theidentity of the ultimate toxicants responsible for toxic insults.B. Male Reproductive System1. Testicular Enzymes of Xenobiotic MetabolismThe presence of P450, AHH activity, epoxide hydrase, and glutathione transferase activities in thetestes of laboratory animals has been documented. 252–254 These enzymes are induced followingexposure to TCDD. Seminal vesicle microsomes represent a rich source of PGS, 27 and LO activityhas been detected in testis, vesicular gland, prostate, Ledig cells, and spermatozoa. 62. Male Reproductive ToxicantsDrugs, pesticides, environmental chemicals and metals have been identified as potential testiculartoxicants. Affected men exhibit loss of libido, testicular atrophy, abnormal sperm morphology, lownumbers of competent spermatozoa, or dysfunctions related to erection, ejaculation, and secretionby accessory sex glands.Although the biological significance is currently unknown, low levels of residues of currentlyused organophosphorus and carbamate pesticides and those of now abandoned organochlorineinsecticides have been detected in food and water, human tissues, blood, and milk. 255 Organochlorines(e.g., DDT, dieldrin, heptachlor), carbamates (e.g., carbaryl), organophosphates (e.g., parathion),and other pesticides (e.g., DBE and 1,2-dibromo-3-chloropropane, DBCP) have beensuspected of causing testicular injury in animals, although supportive evidence for humans isavailable for only a few. The literature on P450-dependent oxidation of pesticides has beenreviewed. 19,21,26 The nematocide DBCP is probably the first example linking chemical exposureunder an occupational setting to testicular toxicity in workers. 256 Oligospermia, azoospermia, andincreased FSH and LH levels were noted in exposed men. Following discontinuation of exposure,© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 557reversal of DBCP toxicity was noted in oligospermic but not in azoospermic men. In rats, GSHconjugates are formed from DBCP in liver, kidney, and testis cytosol. 254 The identification of theultimate toxicant has been illusive, 257 but an Ames assay suggested that oxidative metabolism isneeded to generate genotoxic metabolite(s) from DBCP. 257Chlordecone (kepone) causes reversible testicular toxicity and low sperm count, decreasedsperm motility, abnormal morphology, and arrest in sperm maturation in exposed men. 258 In humans,as in animals, a small portion of kepone in the body is reduced to its alcohol or converted toglucuronides and/or GSH conjugates before fecal excretion. Whether kepone itself or one of itsmetabolites is the reproductive toxicant is not yet established. Among pesticides, linking exposuresto Agent Orange (a 1:1 mixture of 2,4-D and 2,4,5-T herbicides) sprayed in Vietnam to systemicillnesses in soldiers and increased incidences of miscarriages, stillbirths, and birth defects has beenfrustrating, illusive, and hotly debated for the past 30 years. Many conflicting reports have appearedin the literature. It has been suspected that TCDD contamination, which occurred during themanufacture of 2,4,5-T, may be a culprit.Testicular cancer due to exposure to soot in chimney sweeps was documented long ago. Sincethen, several polycyclic aromatic hydrocarbons (PAHs) are suspected to cause testicular toxicity.Many studies have documented that smoking adversely affects testicular function in men. Asignificant decrease in sperm count and motility and an increase in abnormal sperm morphologyhave been documented. Although over 3000 different chemicals occur in the cigarette smoke, PAHsand a few other compounds are the suspected toxicants. The blood-testis barrier does not preventexposure of the testis to PAHs, such as BP. P450-catalyzes epoxidation of BP to the proximate BPdiol.Further oxidation forms the ultimate 7,8-diol-9,10-epoxide, which is believed to exert toxicity.Evidence gathered in rats by use of perfusion of the testis and testis homogenates indicates thattesticular enzymes can activate BP as well as detoxify toxic reactive intermediates. 252,253 Theplasticizer diethylhexyl phthalate and its metabolite monoethylhexyl phthalate, resulting fromdealkylation, are testicular toxicants in rodents. 259 They cause sloughing of spermatids and spermatocytesand vacuolization of Sertoli cell cytoplasm. 2,5-Hexanedione, a metabolite of n-hexane,is a testicular toxicant, 260 but it is not known whether testicular P450 mediates this reaction. Ethyleneglycol monoethyl ether and its metabolite, 2-methoxy ethanol (2-ME), are testicular toxicants. 2612-Methoxyacetic acid, the oxidation product of 2-ME generated by sequential oxidation by alcoholdehydrogenase and acetaldehyde dehydrogenase, is believed to be the ultimate testicular toxicant.Among drugs, several agonists or antagonists of natural hormones, antineoplastic agents, anesthetics,and antibiotics are known to produce adverse effects on the male reproductive system.Cyclophosphamide is a therapeutic agent commonly used for the treatment of boys with nephroticsyndrome and Hodgkin’s disease. The results of a few studies have suggested that high drugexposure, especially later in the puberty, is associated with elevated risks of decreased spermatogenesis.Although the enzymes responsible for cyclophosphamide bioactivation in testis are notknown, published data suggest that P450, PGS, and LO are capable of its conversion to acroleinand phosphoramide mustard. 72,73 Among antipsychotic agents, chlorpromazine has been shown toincreases the rate of prolactin secretion that may be associated with breast engorgement in men.Chlorpromazine is metabolized by several heme containing proteins and LO. Therapeutic use ofimipramine and amitriptyline causes delayed orgasm in some men, while other antidepressantshave been shown to suppress spermatogenesis. Imipramine and related tricyclic compounds areknown to be metabolized by P450 and flavin-containing monooxygenase, as well as by LO. 219C. Female Reproductive System1. Ovarian Enzymes of Xenobiotic MetabolismThe evidence suggests that the ovary has the full complement of enzymes necessary for metabolismof xenobiotics. Thus, the demonstration of AHH activity and BP metabolism indirectly suggest the© 2006 by Taylor & Francis Group, LLC

558 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONpresence of P450 and epoxide hydrase activities in the murine ovary. 253 Various reports have shownactive prostaglandin synthesis in animal ovaries, while the LO pathway is operational in mammalianovary, preovulatory follicles, and oocytes. 6 Other metabolically competent reproductive tissues inclose proximity to the embryo, such as uterus, placenta, fetal membranes, etc., are also expectedto contribute to biotransformation of reproductive toxicants.2. Female Reproductive ToxicantsIn women, destruction of ova; interference with ovulation, fertilization, or implantation; early, late,or missed abortions, etc., are some of the lead indicators of unfavorable reproductive outcomes.Pesticide exposure is essentially unavoidable for women engaged in the agriculture industry.Epidemiological surveys have noted a positive correlation between the residues of organochlorinepesticides in maternal or cord blood, placenta, and fetal tissues and the incidence of spontaneousabortion, missed abortion, fetal prematurity, induction of early labor, premature delivery, andstillbirth. 262 Recent recognition that methoxychlor, dieldrin, o,p-DDT, toxaphene, and endosulfanmay act as endocrine disruptors has renewed interest in the reproductive toxicology of pesticides.Some data for other pesticides are available. 263 Many of these pesticides act as estrogen agonistsand/or androgen antagonists. Prochloraz reacts as both an estrogen and an androgen antagonist. 263Epoxidation of aldrin to dieldrin has long been employed as a model reaction catalyzed byP450 19–21 and has been noted with PGS. 21 The demonstration that soybean LO can mediate thisreaction suggests LO as an additional pathway for this metabolic step. 264 Heptachlor, isodrin, andchlordene also follow this path of metabolism. 19 Methoxychlor itself is not estrogenic, but itsmetabolites are. 265 Human liver microsomes catalyze primarily mono-O-demethylation of methoxychlor,and CYP2C19, CYP2C18, and CYP1A2 are more active in this catalysis. Further metabolismoccurs during longer incubation. The second O-demethylation yields bis-hydroxy methoxychlor,which is a more potent estrogen than mono-hydroxy methoxychlor. CYP2B6 and CYP1A2mediate ortho-hydroxylation, whereas CYP3A4 is most active in ortho-hydroxylation of monoandbis-demethylated methoxychlor. 265Chronic exposure to organophosphates causes abnormal menstruation, amenorrhea, and inductionof early menopause. 266 Animal studies have revealed reproductive toxicity and teratogenicityof several organophosphorous and carbamate insecticides. Metabolic catalysis of these pesticideshas been extensively studied in animals; however, data are very limited for human tissues. 19–21Desulfuration catalyzed by P450, FMO, 19–21,26 and LO 267 forms their oxons, which are more potentanticholinesterases. The pesticides lose their potency when metabolized by other pathways. Detoxicationby dealkylation occurs via P450 19–21 and LO. 218,223Many PAHs can cause ovarian toxicity. 253 Thus, BP, 3-MC, and DMBA all cause destructionof primordial oocytes in DBA/2N and C57BL/6N inbred mice. In rodents, the process is species,strain, age, dose, and metabolism dependent. However, oocyte destruction does not appear to belinked to the Ah locus or AHH activity. For example, conversion of BP to BP-diol epoxide isessential for toxicity. Although P450 was believed to be the catalyst responsible for this bioactivation,253 possible involvement of other enzymes cannot be ruled out at present because PGS, 27,28soybean LO, and animal tissue LOs 106,268,269 can perform this final activation. It is highly likely thatovarian PGS and/or LO also bioactivates BP and other PAHs.Several epidemiological studies have found cigarette smoking in women to decrease the ageof spontaneous menopause, suggesting oocyte destruction. BP has been the prime suspect; however,involvement of other chemicals is likely. Although specific details are sketchy, arylamines, such as4-ABP, benzidine, and 2-AF, may also contribute to female reproductive toxicity. These chemicalsare oxidized by P450, PGS, and LO. 27,28,217,270 Despite the obvious importance of this subject, studieson these toxicants with human ovarian tissue enzymes are still lacking.Acrylonitrile exposure in polymer industries seems to be associated with abnormal menses.Animal studies have revealed that acrylonitrile is a developmental toxicant and a teratogen. 271 They© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 559also point out that the ultimate toxicant may be the epoxide and/or cyanide released duringacrylonitrile oxidation. Earlier suggestions on the role played by P450 in acrylonitrile bioactivationhave been questioned. 272 The data point out the ability of LO to convert acrylonitrile to its epoxideand cyanide. 272 A few reports have similarly indicated reproductive toxicity for styrene and vinylchloride. Metabolic data for these chemicals are not currently available for human female reproductivetissues.DES is a synthetic estrogen that was used to prevent miscarriage. DES is carcinogenic inanimals, and its human use has been linked to the occurrence of vaginal and cervical clear celladenocarcinoma in young women. In utero exposure to DES has been associated with a highincidence of vaginal adenosis, cervical ectropion, and structural abnormalities of the uterus andvagina in girls, and with epididymal cysts, hypoplastic testes, varicoceles, and spermatozoal defectsin boys. 273 The results of several studies have led to the conclusion that estrogenic activity aloneis insufficient to explain DES toxicity. It is now established that DES oxidation to Z,Z-dienestrolis required. Horseradish peroxidase, uterine peroxidase, and PGS, but not P450, were shown to becapable of this reaction. 27,28,273 Recently, soybean LO has been found to activate DES by a oneelectronoxidation process. 274 It was proposed that two molecules of the initial DES oxidationproduct, DES-semiquinone, undergo dismutation to generate DES-quinone and DES. The quinonemetabolites can bind irreversibly to cellular macromolecules and may be the ultimate toxicants. Atpresent, there are no reports on human female reproductive tissue enzymes catalyzing DES activationto explain the toxicity reported in boys and girls.Although intake of phenothiazines during pregnancy appears to be relatively safe, conflictingreports on toxicities do exist in the literature. Sobel 194 suggested that chlorpromazine inducesabortions but causes no birth defects, while Slone et al. 193 noted a positive relationship betweenantenatal phenothiazine use and intelligence quotient score, birth defects, and perinatal mortalityrate. Phenothiazines are teratogenic in laboratory animals. Several oxidative enzymes can metabolizethese compounds. Recent studies with soybean LO have revealed another metabolic complexitywith possible unfavorable pharmacological or toxicological outcomes. 275,276 It was demonstratedthat the chlorpromazine cation radical, the product of one-electron oxidation of chlorpromazine byLO, can serve as a shuttle oxidant in a secondary reaction that stimulates oxidation of benzidineand other xenobiotics several-fold. The shuttle oxidant production was also noted with otherphenothiazines. 275,276 Antiepileptic drug therapy is associated with reproductive dysfunction. Forexample, the anticonvulsant drug phenytoin has been reported to cause hirsutism in young women.As pointed out earlier, phenytoin is metabolized by P450 and PGS, as well as by LO. 3 It remainsto be established whether phenytoin itself or a metabolite generated by human female reproductivetissues is responsible for the outcome.VI. CONCLUSIONS AND FUTURE RESEARCH NEEDSBecause of ethical and legal reasons, few controlled in vitro xenobiotic metabolism studies havebeen conducted with human reproductive tissues. Unfortunately, the paucity of human data dictatescontinued heavy reliance on information generated from laboratory animals. The thalidomidetragedy has taught us a lesson by clearly demonstrating that no animal species can be a completesurrogate for humans. It is therefore extremely important to consider both qualitative and quantitativespecies-specific differences in xenobiotic metabolizing enzymes between the laboratoryanimal species commonly employed in developmental and reproductive toxicity testing and humans.Notable among these are P450 and GST. The specific P450 content and the xenobiotic oxidizingability of adult human liver are significantly lower than is true of rodents. However, the reverse istrue when one considers fetal liver. Significant amounts of this hemoprotein occur in human fetalliver, while it is barely detectable in rodent fetal liver late in gestation. Further, the GST contentof human fetal liver equals or even exceeds adult levels, while this enzyme is barely detectable in© 2006 by Taylor & Francis Group, LLC

560 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrodent fetal liver, and the same is true for placental GST. Significant commonality resides in theexistence of peroxidative enzymes in the conceptal tissues of both humans and rodents. In theplacentas of nonsmokers, functional PGS and P450 are essentially absent. A relative abundance ofperoxidase and LO strongly suggest that they may be primarily responsible for oxidative xenobioticbiotransformation in human placenta. Although our understanding of the in vivo metabolism ofdevelopmental and reproductive toxicants provides some comfort, a wide gap remains with regardto the cause-and effect-relationships involved in toxicity, the chemical identities of the ultimatetoxicants, and the subcellular sites of their biosynthesis in human target tissues.References1. Kulkarni, A.P., Role of xenobiotic metabolism in developmental toxicity, in Handbook of DevelopmentalToxicology, Hood, R.D., Ed., CRC Press, Boca Raton, FL, 1997, chap.12.2. Miller, M.S., et al., Drug metabolic enzymes in developmental toxicology, Fundam. Appl. Toxicol.,34, 165, 1996.3. Wells, P.G. and Winn, L.M., Biochemical toxicology of chemical teratogenesis, Crit. Rev. Biochem.Mol. Biol., 31, 1, 1996.4. Wells, P.G., et al., Oxidative damage in chemical teratogenesis, Mut. Res., 396, 65, 1997.5. Juchau, M.R., Boutelet-Bochan, H., and Huang, Y., Cytochrome-P450-dependent biotransformationof xenobiotics in human and rodent embryonic tissues, Drug. Met. Rev., 30, 541, 1998.6. Kulkarni, A.P., Role of biotransformation in conceptal toxicity of drugs and other chemicals, Curr.Pharmaceut. Designs, 7, 833, 2001.7. Kitada, M. and Kamataki, T., Cytochrome P450 in human fetal liver: Significance and fetal-specificexpression, Drug Met. Rev., 26, 305, 1994.8. Hakkola, J., et al., Xenobiotic-metabolizing cytochrome P450 enzymes in the human feto-placentalunit: role in intrauterine toxicity, Crit. Rev. Toxicol., 28, 35, 1998.9. Pasanen, M. and Pelkonen, O., The expression and environmental regulation of P450 enzymes inhuman placenta, Crit. Rev. Toxicol., 24, 211, 1994.10. Juchau, M.R., Placental enzymes: cytochrome P450s and their significance, in Placental Toxicology,Rama Sastry, B.V., Ed., CRC Press, Boca Raton, FL, 1995, p. 197.11. LeBlanc, G.A. and Dauterman, W.C., Conjugation and elimination of toxicants, in Introduction toBiochemical Toxicology, 3rd Ed., Hodgson, E. and Smart, R.C., Eds., Wiley Interscience, New York,2001, chap. 6.12. Ioannides, C., Xenobiotic metabolism. An overview, in Enzyme Systems that Metabolise Drugs andOther Xenobiotics, Ioannides, C., Ed., Wiley, New York, 2002, chap.1.13. Anders, M.W. and Dekant, W., Glutathione-dependent bioactivation of haloalkenes, Annu. Rev. Pharmacol.Toxicol., 38, 501, 1998.14. Nelson, D.R., et al., P450 superfamily: update on new sequences, gene mapping, accession numbersand nomenclature, Pharmacogenetics, 6, 1, 1996.15. Guengerich, F.P., Comparisons of catalytic selectivity of cytochrome P450 subfamily members fromdifferent species, Chem.-Biol. Interact., 106, 161, 1997.16. Ryan, D.E. and Levin, W., Purification and characterization of hepatic microsomal cytochrome P-450, Pharmacol. Ther., 45, 153, 1990.17. Waterman, M.R., Cytochrome P450: cellular and structural considerations, Curr. Opinions Struct.Biol., 2, 384, 1992.18. Juchau, M.R., Substrate specificities and functions of the P540 cytochromes, Life Sci., 47, 2385, 1990.19. Kulkarni, A.P. and Hodgson, E., The metabolism of insecticides: the role of monooxygenase enzymes,Annu. Rev. Pharmacol. Toxicol., 24, 19, 1984.20. Kulkarni, A.P. and Hodgson, E., Metabolism of insecticides by mixed function oxidase systems, inDifferential Toxicities of Insecticides and Halogenated Aromatics, International Encylopedia of Pharmacologyand Therapeutics, Section 113, Matsumura, F., Ed., Pergamon Press, New York, 1984, chap. 2.21. Kulkarni, A.P., Naidu, A.K., and Dauterman, W.C., Oxidative metabolism of insecticides, in RecentAdvances in Insect Physiology and Toxicology, Gujar, G.T., Ed., Agricole Publishing Acad., NewDelhi, India, 1993, chap. 13.© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 56122. Batt, A.M., Enzyme induction by drugs and toxins, Clin. Chim. Acta, 209, 109, 1992.23. Halpert, J.R. and Guengerich, F.P., Enzyme inhibition and stimulation, in Biotransformation, Vol. 3,Comprehensive Toxicology, Guengerich, F.P., Ed., Elsevier Science, Oxford, 1997, chap.4.24. Smith, D.A., Species differences in metabolism and pharmacokinetics: Are we close to an understanding?Drug Met. Rev., 23, 355, 1991.25. Cashman, J.R., Human flavin-containing monooxygenase: substrate specificity and role in drug metabolism,Curr. Drug Met., 1, 1037, 2000.26. Hodgson, E., et al., Flavin-containing monooxygenase and cytochrome P450-mediated metabolismof pesticides: from mouse to human, Rev. Toxicol., 6, 231, 1998.27. Smith, W.L. and Marnett, L.J., Prostaglandin endoperoxide synthase: structure and catalysis, Biochim.Biophys. Acta, 1083, 1, 1991.28. Degan, G.H., Vogel, C., and Abel, J., Prostaglandin synthases, in Enzyme Systems that MetaboliseDrugs and Other Xenobiotics, Ioannides, C., Ed., John Wiley, Chichester, UK, 2002, chap.6.29. Kulkarni, A.P., Lipoxygenases, in Enzyme Systems that Metabolise Drugs and Other Xenobiotics,Ioannides, C., Ed., John Wiley, Chichester, UK, 2002, chap.7.30. Kulkarni, A.P., Lipoxygenase — a versatile biocatalyst for biotransformation of endobiotics andxenobiotics, Cell. Mol. Life Sci., 58, 1805, 2001.31. Kulkarni, A.P., Lipoxygenase-mediated metabolism of xenobiotics: relevance to toxicology and pharmacology,Recent Res. Dev. Drug Met. Disp., in press, 2002.32. Zahradnik, H.P., et al., Hypertensive disorders in pregnancy. The role of eicosanoids, Eicosanoids, 4,213, 1991.33. Bock, K.W., UDP-glucuronosyltransferases, in Enzyme Systems that Metabolise Drugs and OtherXenobiotics, Ioannides, C., Ed., John Wiley, Chichester, UK, 2002, chap.8.34. Glatt, H., Sulfotransferases, in Enzyme Systems that Metabolise Drugs and Other Xenobiotics, Ioannides,C., Ed., John Wiley, Chichester, UK, 2002, chap.10.35. Sherratt, P.J. and Hayes, J.D., Glutathione transferases, in Enzyme Systems that Metabolise Drugsand Other Xenobiotics, Ioannides, C., Ed., John Wiley, Chichester, UK, 2002, chap. 9.36. Jakobsson, P.J., et al., Membrane associated proteins in eicosanoid and glutathione metabolism(MAPEG). A widespread superfamily, Am. J. Resp. Crit. Care Med., 161, S20, 2000.37. Wolff, T., Erhard, D., and Wanders, H., Aldrin epoxidation, a highly sensitive indicator specific forcytochrome P450-dependent monooxygenase activities, Drug Met. Disp., 7, 301, 1979.38. Omura, T. and Sato, R., The carbon monoxide binding pigment of liver microsomes. I. Evidence forits hemoprotein nature, J. Biol. Chem., 239, 2370, 1964.39. Matsubara, T., et al., Quantitative determination of cytochrome P450 in rat liver homogenate, Anal.Biochem., 75, 596, 1976.40. Ziegler, D.M., Microsomal flavin-containing monooxygenase: oxygenation of nucleophilic nitrogenand sulfur compounds, in Enzymatic Basis of Detoxication, Vol. I, Jakoby, W.B., Ed., Academic Press,New York, chap. 9.41. Kulkarni, A.P. and Cook, D.C., Hydroperoxidase activity of lipoxygenase: a potential pathway forxenobiotic metabolism in the presence of linoleic acid. Res. Comm. Chem. Pathol. Pharmacol., 61,305, 1988.42. Hover, C. and Kulkarni A.P., A simple and efficient method for hemoglobin removal from mammaliantissue cytosol by zinc sulfate and its application to the study of lipoxygenase, Prost. Leuk. EssentialFatty Acids, 62, 97, 2000.43. Byczkowski, J.Z., Ramgoolie, J., and Kulkarni, A.P., Some aspects of activation and inhibition of ratbrain lipoxygenase, Int. J. Biochem., 24, 1691, 1992.44. Kulkarni, A.P., et al., Hydrogen peroxide: a potent activator of dioxygenase activity of soybeanlipoxygenase, Biochem. Biophys. Res. Commun., 166, 417, 1990.45. Kulkarni, A.P., et al., Dioxygenase and peroxidase activities of soybean lipoxygenase: Synergisticinteraction between linoleic acid and hydrogen peroxide, Res. Commun. Chem. Pathol. Pharmacol.,66, 287, 1989.46. Maehly, A.C. and Chance, B., Assay of catalases and peroxidases, Methods Biochem. Anal., 1, 385,1954.47. Nelson, J.L. and Kulkarni, A.P., Partial purification and characterization of a peroxidase activity fromhuman placenta, Biochem. J., 268, 739, 1990.© 2006 by Taylor & Francis Group, LLC

562 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION48. Kulkarni, A.P. and Cook, D.C., Hydroperoxidase activity of lipoxygenase: hydrogen peroxide-dependentoxidation of xenobiotics, Biochem. Biophys. Res. Commun., 155, 1075, 1988.49. Hover, C.G. and Kulkarni, A.P., Hydroperoxide specificity of plant and human tissue lipoxygenase: an invitro evaluation using N-demethylation of phenothiazines. Biochim. Biophys. Acta, 1475, 256, 2000.50. Oesch, F., Jerina, D.M., and Daly, J.A., A radiometric assay for hepatic epoxide hydrase activity with[7–3H]styrene oxide, Biochim. Biophys. Acta, 227, 685, 1971.51. Habig, W.H., Pabst, M.J., and Jakoby, W.B., Glutathione S-transferase, the first enzymatic step inmercapturic acid formation, J. Biol. Chem., 249, 7130, 1974.52. Radulovic, L.L. and Kulkarni, A.P., A rapid, novel high performance liquid chromatography methodfor the purification of glutathione S-transferase: An application to the human placental enzyme,Biochem. Biophys. Res. Commun., 128, 75, 1985.53. Sajan, M. and Kulkarni, A.P., A simple and rapid method for hemoglobin removal from mammaliantissue cytosol by zinc sulfate and its application to the study of glutathione transferase, Toxicol.Methods, 7, 55, 1997.54. Osimitz, T.G. and Kulkarni, A.P., Hepatic microsomal oxidative metabolism of pesticides and otherxenobiotics in pregnant CD1 mice, Pestic. Biochem. Physiol., 23, 328, 1985.55. Symons, A.M., Turcan, R.C., and Parke, D.V., Hepatic drug metabolism in pregnant rat, Xenobiotica,12, 365, 1982.56. Miller, M.S., et al., Expression of the cytochrome P-4502E and 2B gene families in the lungs andlivers of nonpregnant, pregnant and fetal hamsters, Biochem. Pharmacol., 44, 797, 1992.57. Turcan, R.G., et al., Drug metabolism, cytochrome P-450 spin state and phospholipid changes duringpregnancy in the rat, Biochem. Pharmacol., 30, 1223, 1981.58. Lambert, G.H., Lietz, H.W., and Kotake, A.N., Effect of pregnancy on the cytochrome P-450 systemin mice, Biochem. Pharmacol., 36, 1965,1987.59. Williams, D.E., et al., A prostaglandin w-hydroxylase cytochrome P450 (P-450 PG-w ) purified fromlungs of pregnant rabbits, J. Biol. Chem., 259, 14600, 1984.60. Cambell, R.M., Fell, B.F., and Mackie, W.S., Ornithine decarboxylase activity, nucleic acids, and cellturnover in the livers of pregnant rats, J. Physiol. (London), 241, 699, 1974.61. Combes, B., et al., Alterations in sulphobromophthalein sodium removal mechanism from blood duringnormal pregnancy, J. Clin. Invest., 42, 1431, 1963.62. Lucier, G.W., et al., Postnatal stimulation of hepatic microsomal enzymes following administrationof TCDD to pregnant rats, Chem.-Biol. Interact., 11, 15, 1975.63. Jannettl, R.A. and Anderson, L.M., Dimethylnitrosamine demethylase activity in fetal, suckling, andmaternal mouse liver and its transplacental and transmammary induction by polychlorinated biphenyls,J. Natl. Cancer Inst., 67, 461, 1981.64. Osimitz, T.G. and Kulkarni, A.P., Oxidative metabolism of xenobiotics during pregnancy: significanceof flavin-containing monooxygenase, Biochem. Biophys. Res. Commun., 109, 1164, 1982.65. Osimitz, T.G. and Kulkarni, A.P., Polyamine effects on cytochrome P-450 and flavin containingmonooxygenase-mediated oxidation of xenobiotics, Drug Met. Disp., 13, 197, 1985.66. Weitman, S.D., Vidicnik, M.J., and Lech, J.J., Role of pregnancy on the activation and detoxicationof parathion, Pharmacologist, 24, 93, 1982.67. Weitman, S.D. and Lech, J.J., Influence of pregnancy on the hepatic metabolism of parathion, Toxicologist,3, A78, 1983.68. Hulla, J.E. and Juchau, M.R., Occurrence and inducibility of cytochrome P450IIIA in maternal andfetal rats during prenatal development, Biochemistry, 28, 4871, 1989.69. Greenaway, J.C., et al., The in vitro teratogenicity of cyclophosphamide in rat embryos, Teratology,25, 335, 1982.70. Juchau, M.R., et al., Effect of 3-methylcholanthrene and phenobarbital on the capacity of embryosto bioactivate teratogens during organogenesis, Toxicol. Appl. Pharmacol., 80, 137, 1985.71. Juchau, M.R., et al., Generation of reactive dysmorphogenic intermediates by rat embryo in culture:effects of cytochrome P450 inducers, Toxicol. Appl. Pharmacol., 81, 533, 1985.72. Kanekal, S. and Kehrer, J.P., Metabolism of cyclophosphamide by lipoxygenases, Drug Met. Disp.,22, 74, 1994.73. Kanekal, S. and Kehrer, J.P., Evidence for peroxidase mediated metabolism of cyclophosphamide,Drug Met. Disp., 21, 37, 1993.© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 56374. Devereux, T.R. and Fouts, J.R., Effect of pregnancy or treatment with certain steroids on N,Ndimethylanilinedemethylation and N-oxidation by rabbit liver and lung microsomes, Drug Met. Disp.,3, 254, 1975.75. James, M.O., et al., The perinatal development of epoxide metabolizing enzymes activities in liverand extrahepatic organs of guinea pigs and rabbits, Drug Met. Disp., 5, 19, 1977.76. Kulkarni, A.P., et al., Glutathione S-transferase activity during pregnancy in the mouse: effects oftrans-stilbene oxide pretreatment, Biochem. Pharmacol., 33, 3301,1984.77. Polidoro, G., et al., Effect of pregnancy on hepatic glutathione S-transferase activities in the rat,Biochem. Pharmacol., 30, 1859, 1981.78. Combes, B. and Stakelum, G.S., Maturation of sulfobromophthalein sodium-glutathione conjugatingsystem in rat liver, J. Clin. Invest., 41, 750, 1962.79. Mitra, A.K., et al., Rat hepatic glutathione S-transferase mediated embryotoxic bioactivation ofethylenedibromide, Teratology, 46, 439, 1992.80. Bell, J.V., Hansell, M.M., and Ecobichon, D.J., The influence of phenobarbitone on maternal andperinatal hepatic drug metabolizing enzymes in the rat, Can. J. Physiol. Pharmacol., 53, 1147, 1975.81. Rouet, P., et al., Metabolism of benzo(a)pyrene by brain microsomes of fetal and adult rats and mice.Induction by 5,6-benzoflavone, comparison with liver microsomal activities, Carcinogenesis, 2, 919,1981.82. Schlede, E. and Merker, H.J., Benzo(a)pyrene hydroxylase activity in the whole implantation site,decidua, placenta and fetal membranes of the rat, Naunyn Schmiedebergs Arch. Pharmacol., 282, 59,1974.83. Keeping, H.S. and Lyttle, C.R., Monoclonal antibody to rat uterine peroxidase and its use in identificationof the peroxidase as being of eosinophil origin, Biochim. Biophys. Acta, 802, 399, 1984.84. Nelson, J.L. and Kulkarni, A.P., Peroxidase activity in the mouse uterus, placenta and fetus duringpregnancy, Biochem. Internat., 13, 131, 1986.85. Agrawal, P. and Laloraya, M.M., Ascorbate and peroxidase changes during pregnancy in albino ratand Swiss mouse, Am. J. Physiol., 236, E386, 1979.86. McNabb, T., Sproul, J., and Jellinck, P.H., Effects of phenols on the oxidation of estradiol by uterineperoxidase, Can. J. Biochem., 53, 855, 1975.87. Byczkowski, J.Z. and Kulkarni, A.P., Activation of benzo(a)pyrene-7,8-dihydrodiol in rat uterus: anin vitro study, J. Biochem. Toxicol., 5, 139, 1990.88. Shum, S., Jenson, N.M., and Nebert, D.W., The murine ah locus: in utero toxicity and teratogenesisassociated with genetic differences in benzo(a)pyrene metabolism, Teratology, 20, 365, 1979.89. Jellinck, P.H., Newbold, R.R., and McLachlan, J.A., Characteristics of estrogen-induced peroxidasein mouse uterine luminal fluid, Steroids, 56, 162, 1991.90. Metzer, M. and Jellinck, P.H., Peroxidase-mediated oxidation, a possible pathway for metabolicactivation of diethylstilbestrol, Biochem. Biophys. Res. Commun., 85, 874, 1978.91. Astroff, B. and Safe, S., 2,3,7,8-Tetrachlorodibenzo-p-dioxin as an antiestrogen: effect on rat uterineperoxidase activity, Biochem. Pharmacol., 39, 485, 1990.92. Devasagayam, T.P.A., Sivabalan, R., and Tarachand, U., Lipid peroxidation in the rat uterus duringdecidua induced cell proliferation, Biochem. Int., 21, 27, 1990.93. Falkay, G., Herczeg, J., and Sas, M., Microsomal lipid peroxidation in human uterus and placenta,Biochem. Biophys. Res. Commun.,79, 843, 1977.94. Malathy, P.V., Cheng, H.C., and Dey, S.K., Production of leukotrienes and prostaglandins in the ratuterus in endometrium during preimplantation period, Prostaglandins, 32, 605, 1986.95. Gu, W. and Rice, G.E., Arachidonic acid metabolites in pregnant rat uterus, Prost. Leukot. Essent.Fatty Acids, 42, 15, 1991.96. Poyser, N.L, A possible explanation for the refractiveness of uterine prostaglandin production, J.Reprod. Fertil., 91, 371, 1991.97. Demers, L.M., Rees, M.C.P., and Turnbull, A.C., Arachidonic acid metabolism by the non-pregnanthuman uterus, Prost. Leukotrienes Med., 14, 175, 1984.98. Ishihara, O., et al., Metabolism of arachidonic acid and synthesis of prostanoids in human endometriumand decidua, Prost. Leukotrienes Med., 24, 93, 1986.99. Mitchell, M.D. and Grzyboski, C.F., Arachidonic acid metabolism by lipoxygenase pathways inintrauterine tissues of women at term of pregnancy, Prost. Leukotrienes Med., 28, 303, 1987.© 2006 by Taylor & Francis Group, LLC

564 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION100. Rees, M.C.P., et al., Leukotriene release by human fetal membranes, placenta and decidua in relationto parturition, J. Endocrinol., 118, 497, 1988.101. Lei, Z.M. and Rao, C.V., The expression of 15-lipoxygenase gene and the presence of functionalenzyme in cytoplasm and nuclei of pregnant human myometria, Endocrinology, 130, 861, 1992.102. Flatman, S., et al., Biochemical studies on a 12-lipoxygenase in human uterine cervix, Biochim.Biophys. Acta, 883, 7, 1986.103. Joseph, P., Srinivasan, S., and Kulkarni, A.P., Xenobiotic oxidation during early pregnancy in human:peroxidase catalyzed chemical oxidation in conceptal tissues, Xenobiotica, 24, 583, 1994.104. Datta, K., et al., Peroxidative xenobiotic oxidation by partially purified peroxidase and lipoxygenasefrom human fetal tissues at 10 weeks of gestation, Gen. Pharmacol., 26, 107, 1995.105. Kulkarni, A.P. and Murthy, K.R., Xenobiotic metabolism in humans during early pregnancy: peroxidase-mediatedoxidation and bioactivation of 2-aminofluorene, Xenobiotica, 25, 799, 1995.106. Joseph, P., et al., Bioactivation of benzo(a)pyrene-7,8-dihydrodiol catalyzed by lipoxygenase purifiedfrom human term placenta and conceptal tissues, Reprod. Toxicol., 8, 307, 1994.107. Datta, K. and Kulkarni, A.P., Oxidative metabolism of aflatoxin B 1 by lipoxygenase purified fromhuman term placenta and intrauterine conceptal tissues, Teratology, 50, 311, 1994.108. Datta, K., et al., Glutathione S-transferase-mediated detoxification and bioactivation of xenobioticsduring early human pregnancy, Early Human Dev., 37, 167, 1994.109. Flint, O.P. and Orton, T.C., An in vitro assay for teratogens with cultures of rat embryo midbrain andlimb bud cells, Toxicol. Appl. Pharmacol., 76, 383, 1984.110. Filler, R. and Lew, K.J., Developmental onset of mixed function oxidase activity in preimplantationmouse embryos, Proc. Natl. Acad. Sci. USA., 78, 6991, 1981.111. Nebert, D.W., Genetic differences in drug metabolism: proposed relationship to human birth defects,in Handbook of Experimental Pharmacology, Vol. 65, Teratogenesis and Reproductive Toxicology,Johnson, E.M. and Kochhar, D.M., Eds., Springer-Verlag, New York, 1983, p. 49.112. Omiecinski, C.J., Redlich, C.A., and Costa, P., Induction and developmental expression of cytochromeP450iA1 messenger RNA in rat and human tissues: detection by the polymerase chain reaction, CancerRes., 50, 4315, 1990.113. Brunstrom, B., Activities in chick embryos of 7-ethoxycoumarin O-deethylase and aryl hydrocarbon(benzo[a]pyrene) hydroxylase and their induction by 3,3′,4,4′-tetrachlorobiphenyl in early embryos,Xenobiotics, 16, 865, 1986.114. Nimura, S. and Ishida, K.J., Immunohistochemical detection demonstration of prostaglandin E 2 inpreimplantation mouse embryos, J. Reprod. Fert., 80, 505, 1987.115. Parr, M.B., et al., Immunohistochemical localization of prostaglandin synthase in the rat uterus andembryo during the implantation period, Biol. Reprod., 38, 333, 1988.116. Klein, K.L., Scott, W.J., and Clark, K.E., Prostaglandin synthesis in rat embryo tissue: the effect ofnon-steroidal anti-inflammatory drugs in vivo and ex vivo, Prostaglandins, 27, 659, 1984.117. Liu, L. and Wells, P.G., Potential molecular targets mediating chemical teratogenesis: in vitro peroxidase-catalyzedphenytoin bioactivation and oxidative damage to proteins and lipids in murine maternaland embryonic tissues, Toxicol. Appl. Pharmacol., 134, 71, 1995.118. Winn, L.M. and Wells, P.G., Evidence for embryonic prostaglandin H synthase-catalyzed bioactivationand reactive oxygen species-mediated oxidation of cellular macromolecules in phenytoin andbenzo(a)pyrene teratogenesis, Free Rad, Biol. Med., 22, 607, 1997.119. Winn, L.M. and Wells, P.G., Phenytoin-initiated DNA oxidation in murine embryo culture, andembryoprotection by antioxidative enzymes superoxide dismutase and catalase: evidence for reactiveoxygen species mediated DNA oxidation in the molecular mechanism of phenytoin teratogenicity,Mol. Pharmacol., 48, 112, 1995.120. Miranda, A.F., Wiley, M.J., and Wells, P.G., Evidence for embryonic peroxidase-catalyzed bioactivationand glutathione-dependent cytoprotection in phenytoin teratogenicity: modulation by eicosatetraynoicacid and buthionine sulfoximine in murine embryo culture, Toxicol. Appl. Pharmacol., 124,230, 1994.121. Wells, P.G., et al., Modulation of phenytoin teratogenicity and embryonic covalent binding by acetylsalicylicacid, caffeic acid, and alpha-phenyl-N-t-butylnitrone: implications for bioactivation byprostaglandin synthetase, Toxicol. Appl. Pharmacol., 97, 192, 1989.© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 565122. Kubow, S.and Wells, P.G., In vitro bioactivation of phenytoin to a reactive free radical intermediateby prostaglandin synthase, horseradish peroxidase, and thyroid peroxidase, Mol. Pharmacol., 35, 504,1989.123. Parman, T., Chen, G., and Wells, P.G., Free radical intermediates of phenytoin and related teratogens.Prostaglandin H synthase-catalyzed bioactivation electron paramagnetic resonance spectrometry, andphotochemical product analysis, J. Biol. Chem., 273, 25079, 1998.124. Parman T., Wiley, M.J., and Wells, P.G., Free radical mediated oxidative DNA damage in the mechanismof thalidomide teratogenicity, Nature Med., 5, 582, 1999.125. Arlen, R.R. and Wells, P.G., Inhibition of thalidomide teratogenicity by acetylsalicylic acid: evidencefor prostaglandin H synthase-catalyzed bioactivation of thalidomide to a reactive intermediate, J.Pharmacol. Exp. Ther., 277, 1649, 1996.126. Vanderhoek, J.Y. and Klein, K.L., Lipoxygenases in rat embryo tissue, Proc. Soc. Exp. Biol. Med.,188, 370, 1988.127. Roy, S.K., et al., Lipoxygenase activity in rat embryo and its potential for xenobiotic oxidation, Biol.Neonate, 63, 297, 1993.128. Kubow, S. and Wells, P.G., In vitro evidence for lipoxygenase-catalyzed bioactivation of phenytoin,Pharmacologist, 30, A74, 1988.129. Yu, W.K. and Wells, P.G., Evidence for lipoxygenase-catalyzed bioactivation of phenytoin to ateratogenic reactive intermediate: in vitro studies using linoleic acid-dependent soybean lipoxygenase,and in vivo studies using pregnant CD-1 mice, Toxicol. Appl. Pharmacol., 131, 1, 1995.130. Balakier, H., Endogenous peroxidase in the visceral endoderm of early mouse embryos, J. Exp. Zool.,231, 243, 1984.131. Pelkonen, O., Environmental influences on human foetal and placental xenobiotic metabolism, Eur.J. Clin. Pharmacol., 18, 17, 1980.132. Pelkonen, O., Biotransformation of xenobiotics in the fetus, Pharmacol. Ther., 10, 261, 1980.133. Farrar, H.C. and Blumer, J.L., Fetal effects of maternal drug exposure, Ann. Rev. Pharmacol. Toxicol.,31, 525, 1991.134. Mizokami, K., et al., 3-Methylcholanthrene induces phenobarbital-induced cytochrome P450 hemoproteinin fetal liver and not cytochrome 448 hemoproteins induced in maternal liver of rats, Biochem.Biophys. Res. Commun., 107, 6, 1982.135. Kitada, M., et al., Four forms of cytochrome P450 in human fetal livers: purification and their capacityto activate promutagens, Jpn. J. Cancer Res., 82, 426, 1991.136. Rane, A. and Ackermann, E., Metabolism of ethylmorphine and aniline in human fetal liver, Clin.Pharmacol. Ther., 13, 663, 1972.137. Ladona, M.G., et al., Human fetal and adult liver metabolism of ethylmorphine, Biochem. Pharmacol.,38, 3147, 1989138. Komari, M., et al., Fetus specific expression of a form of cytochrome P450 in human livers, Biochemistry,29, 4430, 1990139. Hakkola, J., et al., Expression of xenobiotic metabolizing cytochrome P450 forms in human adultand fetal liver, Biochem. Pharmacol., 48, 59, 1994.140. Shimada, T., et al., Characterization of microsomal cytochrome P450 enzymes involved in oxidationof xenobiotic chemicals in human fetal livers and adult lungs, Drug Met. Disp., 24, 515, 1996.141. Hakkola, J., et al., Expression of CYP1B1 in human adult and fetal tissues and differential inducibilityof CYP1B1 and CYP1A1 by Ah receptor ligands in human placenta and cultured cells, Carcinogenesis,18, 391, 1997.142. Shimada, T., Misono, K.S., and Guengerich, F.P., Human liver microsomal cytochrome P450 mephenytoin4-hydroxylase, a prototype of genetic polymorphism in oxidative drug metabolism, J. Biol. Chem.,261, 909, 1986.143. Jacqz-Aigrain, E. and Cresteil, T., Cytochrome P450-dependent metabolism of dextromethorphan:fetal and adult studies, Dev. Pharmacol. Ther., 18, 161, 1992.144. Carpenter, S.J., Lasker, J.M., and Raucy, J.L., Expression, induction, and catalytic activity of theethanol-inducible cytochrome P450 (CYP2E1) in human fetal liver and hepatocytes, Mol. Pharmacol.,49, 260, 1996.145. Greuet, J., et al., The fetal specific gene CYP3A7 is inducible by rafampicin in adult hepatocytes inprimary culture, Biochem. Biophys. Res. Commun., 225, 689, 1996.© 2006 by Taylor & Francis Group, LLC

566 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION146. Schuetz, J.D., Beach, D.L., and Guzelian, P.S., Selective expression of cytochrome P4450 CYP3AmRNA in embryonic and adult human liver, Pharmacogenetics, 4, 11, 1994.147. Wrighton, S.A. and VandenBranden, M., Isolation and characterization of human fetal cytochromeP450 P450HLp2: a third member of the P450III gene family, Arch. Biochem. Biophys., 268, 144, 1989.148. Komari, M., et al., Molecular cloning and sequence analysis of cDNA containing the genetic encodingregion for human fetal liver cytochrome P450, J. Biochem., 105, 161,1989.149. Bell, J.U., Hansell, M.M., and Ecobichon, D.J., The influence of DDT on the ontogenesis of drugmetabolizing enzymes in the perinatal rats, Toxicol. Appl. Pharmacol., 35, 165, 1976.150. Sherratt, A.J., Banet, D.E., and Prough, R.A., Glucocorticoid regulation of polycyclic aromatic hydrocarboninduction of cytochrome P450IAi, glutathione S-transferase, and NAD(P)H:quinone oxidoreductasein cultured fetal rat hepatocytes, Mol. Pharmacol., 37, 198, 1990.151. Srivastava, S., Seth, P.K., and Srivastava, S.P., Altered activity of hepatic mixed function oxidase,cytochrome P450, and glutathione S-transferase by styrene in rat fetal liver, Drug Chem. Toxicol., 15,233, 1992.152. Nau, H., et al., Preparation, morphology and drug metabolism of isolated hepatocytes and liver organcultures from human fetus, Life Sci., 23, 2361, 1978.153. Rane, A., N-oxidation of a tertiary amine (N,N-dimethylaniline) by human fetal liver microsomes,Clin. Pharmacol. Ther., 15, 32, 1973.154. Saeed, S.A. and Mitchell, M.D., Arachidonate lipoxygenase activity in human fetal lung, Eur. J.Pharmacol., 78, 389, 1982.155. Saeed, S.A. and Mitchell, M.D., Formation of arachidonate lipoxygenase metabolites by human fetalmembranes, uterine decidua vera, and placenta, Prost. Leukotrienes Med., 8, 635, 1882.156. Saeed, S.A. and Mitchell, M.D., Conversion of arachidonic acid to lipoxygenase products by humanfetal tissues, Biochem. Med., 30, 322, 1983.157. Morykawa, M.J., et al. Arachidonic acid metabolites: effect on inflammation of fetal rabbit excisionalwounds, Inflammation, 16, 251, 1992.158. Dimitrijevic, L., et al., Purification and characterization of an estrogen-binding peroxidase from humanfetus, Biochimie, 61, 535, 1979.159. Peng, D., Pacifici, G.M., and Rane, A., Human fetal liver cultures: basal activities and inducibility ofepoxide hydrolases and aryl hydrocarbon hydroxylase, Biochem. Pharmacol., 33, 71, 1984.160. Stevens, L.A., A comparative study of enzymes in foetal, young and adult rat, Comp. Biochem.Physiol., 6, 129, 1962.161. Pacifici, G.M., Norlin, A., and Rane, A., Glutathione S-transferases in human fetal liver, Biochem.Pharmacol., 30, 3367, 1981.162. Guthenberg, C., et al., Two distinct forms of glutathione S-transferase from human fetal liver, Biochem.J., 235, 741, 1986.163. Kashiwada, M., et al., Purification and characterization of acidic form of glutathione S-transferase inhuman fetal livers: high similarity to placental form, J. Biochem., 110, 743, 1991.164. Mitra, A.K., et al., A novel model to assess developmental toxicity of dihaloalkanes in human:embryotoxic bioactivation of 1,2-dibromoethane by isozymes of human fetal liver glutathione S-transferase, Teratogen. Carcinog. Mutagen., 12, 113, 1992.165. Radulovic, L.L., Dauterman, W.C., and Kulkarni, A.P., Biotransformation of methyl parathion byhuman fetal liver glutathione S-transferases: an in vitro study, Xenobiotica, 17, 105, 1987.166. Kulkarni, A.P., Edwards, J., and Richards, I., Metabolism of 1,2-dibromoethane in the human fetalliver, Gen. Pharmacol., 23, 1, 1992.167. Dutton, G.J., Glucuronide synthesis in foetal liver and other tissues, Biochem. J., 71, 141, 1959.168. Burchell, B., UDP-glucuronyltransferase activity towards eostriol in fresh and cultured foetal tissuesfrom man and other species, J. Steroid Biochem., 5, 261, 1974.169. Pacifici, G.M., et al., Development of glucuronyltransferase and sulfotransferase toward 2-naphtholin human fetus, Dev. Pharmacol. Ther., 14, 108, 1990.170. Kawade, N. and Onishi, S., The prenatal and postnatal development of UDP-glucuronyltransferaseactivity towards bilirubin and the effect of premature birth on this activity in human liver, Biochem.J., 196, 257, 1981.171. Tan, T.M.C., Sit, K.H., and Wong, K.P., UDP-glucuronyltransferase activity toward harmol in humanliver and human fetal liver cells in culture, Anal. Biochem., 185, 44, 1990.© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 567172. Cappiello, M., et al., Sulfotransferase and its substrate: adenosine-3′-phosphate-5′-phosphosulfate inhuman fetal liver and placenta, Dev. Pharmacol. Ther., 14, 62, 1990.173. Elliott, W.J., et al., Arachidonic acid metabolism by rabbit fetal membranes of various gestationalages, Prostaglandins, 27, 27, 1984.174. Kinoshita, K. and Green, K., Bioconversion of arachidonic acid to prostaglandins and related compoundsin human amnion, Biochem. Med., 23, 185, 1980.175. Olson, D.M., Skinner, K, and Challis, J.R.G., Estradiol-17 and 2-hydroxyestradiol-17-induced differentialproduction of prostaglandins by cells dispersed from human intrauterine tissues at parturition,Prostaglandins, 25, 639, 1983.176. Cheung, P.Y.C. and Challis, J.R.G., Prostaglandin E 2 metabolism in the human fetal membranes, Am.J. Obstet. Gynecol., 161, 1580, 1989.177. Price, T.M., et al., Immunohistochemical localization of prostaglandin endoperoxide synthase inhuman fetal membranes and decidua, Biol. Reprod. 41, 701, 1989.178. Pasetto, N., et al., Influence of labor and oxytocin on in vitro leukotriene release by human fetalmembranes and uterine decidua at term gestation, Am. J. Obstet. Gynecol., 166, 1500, 1992.179. Meigs, R.A., The constitutive 7-ethoxycoumarin O-deethylase activity of human placentalmicrosomes: relationship to aromatase, Biochem. Biophys. Res. Commun., 145, 1012, 1987.180. Wong, T.K., et al., Correlation of placental microsomal activities with protein detected by antibodiesto rabbit cytochrome P450 isozyme 6 in preparations from humans exposed to polychlorinatedbiphenyls, quaterphenyls, and dibenzofurans, Cancer Res., 46, 999, 1986.181. Lucier, G.W., et al., Placental markers of human exposure to polychlorinated biphenyls and polychlorinateddibenzofurans, Environ. Hlth. Persp., 76, 79, 1987.182. Mortimer, G., Mackay, M.M., and Stimson, W.H., The distribution of pregnancy associated prostaglandinsynthase inhibitor in the human placenta, J. Pathol., 159, 239, 1989.183. Flammang, T.J., et al., Arachidonic acid dependent peroxidative activation of carcinogenic arylaminesby extrahepatic human tissue microsomes, Cancer Res., 49, 1977, 1989.184. Matkovics, B., Fodor, I., and Kovacs, K., Properties of enzymes III. A comparative study of peroxidedecomposingenzymes and superoxide dismutase of human placenta of different ages, Enzyme, 19,285, 1975.185. Joseph, P., et al., Peroxidase: a novel pathway for chemical oxidation in human term placenta, Placenta,13, 545, 1992.186. Joseph, P., Srinivasan, S., and Kulkarni, A.P., Placental peroxidase — further purification of the enzymeand oxidation of thiobenzamide, Placenta, 14, 309, 1993.187. Murthy, K.R., Joseph, P., and Kulkarni, A.P., 2-Aminofluorene bioactivation by human term placentalperoxidase, Terat. Carcinog. Mut., 15, 115, 1995.188. Kulkarni, A.P., Joseph, P., and Byczkowski, J.Z., Effects of phenylbutazone and amino acids onbenzo(a)pyrene-7,8-dihydrodiol co-oxygenation by partially purified placental peroxidase, Toxicologist,12, 197, 1992.189. Madhavan, N.D. and Naidu, K.A., Purification and partial characterization of peroxidase from humanplacenta of non-smokers: metabolism of benzo(a)pyrene-7,8-dihydrodiol, Placenta, 21, 501, 2000.190. Pumeau-Rouquette, C., Goujard, J., and Hual, G., Possible teratogenic effect of phenothiazines inhuman beings, Teratology, 15, 57, 1977.191. Edmund, M.J. and Craig, T.J., Antipsychotic drug use and birth defects: an epidemiological reassessment,Comp. Psychiatry, 25, 32, 1984.192. Altshuler, L.L., et al., Pharmacologic management of psychiatric illness during pregnancy: dilemmasand guidelines, Am. J. Psychiatry, 153, 592, 1996.193. Slone, D., et al., Antenatal exposure to phenothiazines in relation to congenital malformations perinatalmortality rate, birth weight and intelligence quotient score, Am. J. Obstet. Gynecol., 128, 486, 1977.194. Sobel, D.E., Fetal damage due to ECT, insulin coma, chlorpromazine, or reserpine, Arch. Gen.Psychiatry, 2, 606, 1960.195. Yang, X. and Kulkarni, A.P., Oxidation of phenothiazines by human term placental peroxidase in nonsmokers.Teratogen. Carcinog. Mutagen., 17, 139, 1997.196. Zhang, R. Kulkarni, K.A., and Kulkarni, A.P., Oxidation of eugenol by human term placental peroxidase,Placenta, 21, 234, 2000.© 2006 by Taylor & Francis Group, LLC

568 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION197. Ogburn, P.L., et al., Leukotriene B 4 , 6-keto-prostaglandin F 1 , thromboxane B 2 , and free arachidonicacid levels in human placental tissue, Ann. N. Y. Acad. Sci., 524, 434, 1988.198. Lopez Bernal, A., et al., Placental leukotriene B 4 release in early pregnancy and in term and pretermlabour, Early Hum. Dev., 23, 93, 1990.199. Joseph, P., Srinivasan, S.N., and Kulkarni, A.P., Purification and partial characterization of lipoxygenasewith dual catalytic activities from human term placenta, Biochem. J., 293, 83, 1993.200. Walsh, W.S. and Parisi, V.M., The role of arachidonic metabolites in preeclampsia, Semin. Perinatol.,10, 334, 1986.201. Roy, P., Sajan, M., and Kulkarni, A.P., Lipoxygenase mediated glutathione oxidation and superoxidegeneration, J. Biochem. Toxicol., 10, 111, 1995.202. Kulkarni, A.P. and Sajan, M., A novel mechanism of glutathione conjugation formation by lipoxygenase:a study with ethacrynic acid, Toxicol. Appl. Pharmacol., 143, 179, 1997.203. Kulkarni, A.P. and Sajan, M., Glutathione conjugation of ethacrynic acid by human term placentallipoxygenase, Arch. Biochem. Biophys., 371, 220, 1999.204. Roy, P., et al., Superoxide generation by lipoxygenase in the presence of NADH and NADPH, Biochim.Biophys. Acta, 1214, 171, 1994.205. Roy, P. and Kulkarni, A.P., Oxidation of ascorbic acid by lipoxygenase: effect of selected chemicals,Food Chem. Toxicol., 34, 563, 1996.206. Yang, X. and Kulkarni, A.P., Lipoxygenase-mediated biotransformation of p-aminophenol in thepresence of glutathione. Possible conjugate formation, Toxicol. Lett., 111, 253, 2000.207. Datta, K. and Kulkarni, A.P., Co-oxidation of t-retinol acetate by human term placental lipoxygenaseand soybean lipoxygenase, Repr. Toxicol., 10, 105, 1996.208. Manchester, D.K., et al., Detection of benzo(a)pyrene diol epoxide-DNA adducts in human placenta,Proc. Natl. Acad. Sci. USA., 85, 9243, 1988.209. Denning, D.W., et al., Transplacental transfer of aflatoxins in humans, Carcinogenesis, 11, 1033, 1990.210. Wild, C.P., et al., In-utero exposure to aflatoxin in West Africa, Lancet, 337, 1602, 1991.211. Heish, L.L. and Heish, T.T., Detection of aflatoxin B 1 -DNA adducts in human placenta and cordblood, Cancer Res., 53, 1278, 1993.212. Cameron, H.M. and Warwick, G.P., Primary cancer of the liver in Kenyan children, Brit. J. Cancer,36, 793, 1977.213. Roy, S. and Kulkarni, A.P., Aflatoxin B 1 epoxidation catalyzed by partially purified adult human liverlipoxygenase, Xenobiotica, 27, 231, 1996.214. Coghlin, P.H., et al., 4-aminobiphenyl-hemoglobin adducts in fetuses exposed to tobacco smokecarcinogen in utero, J. Natl. Cancer Inst., 83, 274, 1991.215. Hammond, S.K., et al., Relationship between environmental tobacco smoke exposure and carcinogenhemoglobinadduct levels in nonsmokers, J. Natl. Cancer Inst., 85, 474, 1993.216. Lu, L.J.W., et al., 32 P-postlabelling assay in mice of transplacental DNA damage induced by the environmentalcarcinogens safrole, 4-aminobiphenyl and benzo(a)pyrene, Cancer Res., 46, 3046, 1986.217. Datta, K., Sherblom, P.M., and Kulkarni, A.P., Co-oxidative metabolism of 4-aminobiphenyl bylipoxygenase from soybean and human term placenta, Drug Met. Disp., 25, 196, 1997.218. Hover, C.G. and Kulkarni, A.P., Human term placental lipoxygenase-mediated — demethylation ofphenothiazines and pesticides in the presence of polyunsaturated fatty acids, Placenta, 21, 647, 2000.219. Rajadhyaksha, A., et al., N-demethylation of phenothiazines by lipoxygenase from soybean and humanterm placenta in the presence of hydrogen peroxide, Teratogen. Carcinogen. Mutagen., 19, 211, 1999.220. Yang, X. and Kulkarni, A.P., N-dealkylation of aminopyrine catalyzed by soybean lipoxygenase inthe presence of hydrogen peroxide, J. Biochem. Toxicol., 12, 175, 1997.221. Hu, J., Sajan, M., and Kulkarni, A.P., Lipoxygenase-mediated N-demethylation of imipramine andrelated tricyclic antidepressants in the presence of hydrogen peroxide, Int. J. Toxicol., 18, 251, 1999.222. Hover, C. and Kulkarni, A.P., Lipoxygenase-mediated hydrogen peroxide-dependent N-demethylationof N,N′-dimethylaniline and related compounds, Chem.-Biol. Interact., 124, 191, 2000.223. Hu, J. and Kulkarni, A.P., Soybean lipoxygenase-catalyzed demethylation of pesticides, Pestic. Biochem.Physiol., 61, 145, 1998.224. Uotila, J., et al., Lipid peroxidation products, selenium-dependent glutathione peroxidase and vitaminE in normal pregnancy, Eur. J. Obstet. Gynecol. Reprod. Biol., 42, 95, 1991.225. Wickens, D., et al., Free radical oxidation (peroxidation) products in plasma in normal and abnormalpregnancy, Ann. Clin. Biochem., 18, 158, 1981.© 2006 by Taylor & Francis Group, LLC

ROLE OF XENOBIOTIC METABOLISM IN DEVELOPMENTAL AND REPRODUCTIVE TOXICITY 569226. Juchau, M.R. and Zachariah, P.K., Comparative studies on the oxidation and reduction of drugsubstances in human placenta versus rat hepatic microsomes, Biochem. Pharmacol., 24, 227, 1975.227. Arvela, P., Karki, N.T., and Pelkonen, R.O., Lipoperoxidation rates and drug oxidizing enzymeactivities in the liver and placenta of some animal species during the perinatal period, Experientia,32, 1311, 1976.228. Kulkarni, A.P. and Kenel, M.F., Human placental lipid peroxidation. I. Some characteristics of theNADPH-supported microsomal reaction, Gen. Pharmacol., 18, 491, 1987.229. Diamant, S., Kisselevitz, R., and Diamant, Y., Lipid peroxidation system in human placental tissue:general properties and the influence of age, Biol. Reprod., 23, 776, 1980.230. Klimek, J., Influence of lipid peroxidation on the progesterone biosynthesis in human placentalmitochondria, J. Steroid Biochem. Mol. Biol., 42, 729, 1992.231. Byczkowski, J.Z. and Kulkarni, A.P., Lipid peroxidation and benzo(a)pyrene derivative co-oxygenationby environmental pollutants, Bull. Environ. Contam. Toxicol., 45, 633, 1990.232. Kulkarni, A.P. and Byczkowski, J.Z., Effect of transition metals on biological oxidations, in EnvironmentalOxidants, Nriagu, J.O. and Simmons, M.S., Eds., John Wiley, New York, 1994, chap. 16.233. Byczkowski, J.Z. and Kulkarni, A.P., Oxidative stress and asbestos, in Environmental Oxidants,Nriagu, J.O. and Simmons, M.S., Eds., John Wiley, New York, 1994, chap.15.234. Byczkowski, J.Z., Wan, B., and Kulkarni, A.P., Vanadium-mediated lipid peroxidation in microsomesfrom human term placenta, Bull. Environ. Contam. Toxicol., 41, 696, 1988.235. Byczkowski, J.Z. and Kulkarni, A.P., Vanadium redox cycling and lipid peroxidation, Biochim. Biophys.Acta, 1125, 134, 1992.236. Kenel, M.F., Bestervelt, L.L., and Kulkarni, A.P., Human placental lipid peroxidation. II. NADPHand iron dependent stimulation of microsomal lipid peroxidation by paraquat, Gen. Pharmacol., 18,373, 1987.237. Byczkowski, J.Z. and Kulkarni, A.P., NADPH-dependent drug redox cycling and lipid peroxidationin microsomes from human term placenta, Int. J. Biochem., 21, 183, 1989.238. Byczkowski, J.Z. and Kulkarni, A.P., Lipid peroxidation-coupled co-oxygenation of benzo(a)pyreneand benzo(a)pyrene-7,8-diol in human term placental microsomes, Placenta, 11, 17, 1990.239. Kulkarni, A.P., Strohm, B., and Houser, W.H., Human placental indanol dehydrogenase: some propertiesof the microsomal enzyme, Xenobiotica, 15, 513, 1985.240. Braun, A.G. and Weinreb, S.L., Metabolism of thalidomide analogues inhibit attachment of cells toconcanavalin A coated surfaces, Teratology, 27, 33A, 1983.241. Pacifici, G.M. and Rane, A., Epoxide hydrase in human placenta at different stages of pregnancy,Dev. Pharmacol. Ther., 6, 83, 1983.242. Berte, F., et al., Ability of placenta to metabolize oxazepam and aminopyrine before and after drugstimulation, Arch. Int. Pharmacodyn., 182, 182, 1969.243. Radulovic, L.L. and Kulkarni, A.P., High performance liquid chromatographic separation and studyof the charge isomers of human placental glutathione S-transferase, Biochem. J., 239, 53, 1986.244. Radulovic, L.L., LaFerla, J.J., and Kulkarni, A.P., Human placental glutathione S-transferase mediatedmetabolism of methyl parathion, Biochem. Pharmacol., 35, 3473, 1986.245. Pacifici, G.M. and Rane, A., Glutathione S-transferase in the human placenta at different stages ofpregnancy, Drug Met. Disp., 9, 472, 1981.246. Mitra, A., et al., Inhibition of human term placental and fetal liver glutathione S-transferases by fattyacids and fatty acid esters, Toxicol. Lett., 60, 281, 1992.247. Kulkarni, A.A. and Kulkarni, A.P., Retinoids inhibit mammalian glutathione transferases, Cancer Lett.,91, 185, 1995.248. Sajan, M., Reddy, G., and Kulkarni, A.P., In vitro inhibition of mammalian glutathione transferasesby selected nitrobenzenes, Int. J. Toxicol., 19, 285, 2000.249. Mitra, A.K., et al., In vitro embryotoxicity of ethylene dibromide bioactivated by purified humanplacental glutathione S-transferase, Int. J. Occup. Med. Toxicol., 2, 101, 1993.250. Manchester, D.K. and Jacoby, E.H., Glutathione S-transferase activities in placentas from smokingand non-smoking women, Xenobiotica, 12, 543, 1982.251. Kitahara, A., et al., Changes in molecular forms of rat hepatic glutathione S-transferase duringchemical hepatocarcinogenesis, Cancer Res., 44, 2698, 1984.252. Dixon, R.L., Aspects of male reproductive toxicology, in Occupational Hazards and Reproduction,Hemminki, K, Sorsa, M., and Vainio, H., Eds., Hemisphere, New York, 1985, chap. 4.© 2006 by Taylor & Francis Group, LLC

570 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION253. Mattison, D.R., Shiromizu, K., and Nightingale, M.S., The role of metabolic activation in gonadaland gamete toxicity, in Occupational Hazards and Reproduction, Hemminki, K, Sorsa, M., and Vainio,H., Eds., Hemisphere, New York, 1985, chap.7.254. Miller, G.E. and Kulkarni, A.P., 1,2-dibromo-3-chloropropane as a substrate for hepatic, renal andtesticular glutathione S-transferases in the rat, Toxicologist, 3, 416A, 1983.255. Kulkarni, A.P. and Mitra, A., Pesticide contamination of food, in Food Contamination from EnvironmentalSources, Simons, M.S. and Nriagu, J., Eds., Adv. Environ. Sci. Technol., Vol. 23, John Wiley,New York, 1990, chap. 9.256. Babich, H., Davis, D.L., and Stotzky, G., Dibromochloropropane (DBCP): a review, Sci. Total Environ.,17, 207, 198.257. Miller, G.E., Brabec, M.J., and Kulkarni, A.P., Mutagenic activation of 1,2-dibromo-3-chloro-propaneby cytosolic glutathione S-transferases and microsomal enzymes, J. Toxicol. Environ. Hlth., 19, 503,1986.258. Guzelian, P.S., Comparative toxicology of chlordecone (kepone) in humans and experimental animals,Ann. Rev. Pharmacol. Toxicol., 22, 89, 1982.259. Gray, T.J.B. and Beamand, J.A., Effect of some phthalate esters and other testicular toxins on primarycultures of testicular cells, Food Cosmet. Toxicol., 22, 123, 1984.260. Boekelheide, K., 2,5-Hexanedione alters microtubule assembly: I. Testicular atrophy, not nervoussystem toxicity, correlates with enhanced tubulin polymerization, Toxic. Appl. Pharmacol., 88, 370,1987.261. Wang, W. and Chapin, R.E., Differential gene expression detected by suppression substractive hybridizationin ethylene glycol monomethyl ether-induced testicular lesion, Toxicol. Sci., 56, 165, 2000.262. Kulkarni, A.P., Treinen, K.A., and Bestervelt, L.L., Human placental Ca 2+ -ATPase: target for organochlorinepesticides? Trophoblast Res., 2, 329, 1987.263. Andersen, H.R., et al., Effects of currently used pesticides in assays for estrogenicity, adrogenicity,and aromatase activity in vitro, Toxicol. Appl. Pharmacol., 179, 1, 2002.264. Naidu, A.K., Naidu, A.K., and Kulkarni, A.P., Aldrin epoxidation: catalytic potential of lipoxygenasecoupled with linoleic acid oxidation, Drug Metab. Disp., 19, 758, 1991.265. Stresser, D.M. and Kupfer, D., Human cytochrome P450-catalyzed conversion of the proestrogenicpesticide methoxychlor into an estrogen. Role of CYP2C19 and CYP1A2 in O-demethylation, DrugMet. Disp., 26, 868, 1998.266. Nicoise, T., Chronic organophosphorus intoxication in women, J. Jpn. Assoc. Rural Med., 22, 756,1974.267. Naidu, A.K., Naidu, A.K., and Kulkarni, A.P., Role of lipoxygenase in xenobiotic oxidation: parathionmetabolism catalyzed by highly purified soybean lipoxygenase, Pestic. Biochem. Physiol., 41, 150,1991.268. Byczkowski, J.Z. and Kulkarni, A.P., Lipoxygenase-catalyzed epoxidation of benzo(a) pyrene-7,8-dihydrodiol, Biochem. Biophys. Res. Commun., 159, 1199, 1989.269. Byczkowski, J.Z. and Kulkarni, A.P., Linoleate-dependent co-oxygenation of benzo(a)pyrene-7,8-dihydrodiol by rat cytosolic lipoxygenase, Xenobiotica, 22, 609, 1992.270. Roy, S. and Kulkarni, A.P., Lipoxygenase: a new pathway for 2-aminofluorene bioactivation. CancerLett., 60, 33, 1991.271. Saillenfait, A.M., et al., Modulation of acrylonitrile-induced embryotoxicity in vitro by glutathionedepletion, Arch. Toxicol., 67, 164, 1993.272. Roy, P. and Kulkarni, A.P., Co-oxidation of acrylonitrile by soybean lipoxygenase and partially purifiedhuman lung lipoxygenase, Xenobiotica, 29, 511, 1999.273. Metzler, M., Biochemical toxicology of diethylstilbestrol, Rev. Biochem. Toxicol., 6, 191, 1984.274. Nunez-Delicado, E., Sanchez-Ferrer, A., and Garcia-Carmona, F., Hydroperoxidative oxidation ofdiethylstilbestrol by lipoxygenase, Arch. Biochem. Biophys., 348, 411, 1997.275. Hu, J. and Kulkarni, A.P., Metabolic fate of chemical mixtures. I. Shuttle oxidant effect of lipoxygenase-generatedradical of chlorpromazine and related phenothiazines on the oxidation of benzidineand other xenobiotics, Teratogen, Carcinogen. Mut., 20, 195, 2000.276. Persad, A.S., Stedeford, T.J., and Kulkarni, A.P., The hyperoxidation of L-dopa by chlorpromazine ina lipoxygenase catalyzed reaction, Toxicologist, 54, 1746A, 2000.© 2006 by Taylor & Francis Group, LLC

CHAPTER 13Use of Toxicokinetics in Developmental andReproductive ToxicologyPatrick J. WierCONTENTSI. Introduction ........................................................................................................................572II. Principles of Absorption, Distribution, Biotransformation, and Elimination ...................572A. Movement across Biological Membranes .................................................................572B. Plasma Protein Binding .............................................................................................573C. Volume of Distribution ..............................................................................................573D. Blood and Seminiferous Tubule Distribution ...........................................................574E. Placental Transfer ......................................................................................................574F. Biotransformation ......................................................................................................575G. Elimination.................................................................................................................576H. Physiological Changes in Pregnancy ........................................................................576I. Physiological Changes in Neonates ..........................................................................577III.Classical Pharmacokinetics................................................................................................578A. Simple Compartmental Models.................................................................................578B. Basic Pharmacokinetic Parameters............................................................................578IV. Physiologically Based Pharmacokinetics ..........................................................................582A. Physiologically Based Pharmacokinetic Models of Pregnancy................................582V. Sampling Strategies............................................................................................................583VI. Regulatory Expectations ....................................................................................................585A. ICH Reproduction Test Guideline.............................................................................585B. ICH Toxicokinetic Guideline.....................................................................................587VII.Toxicokinetics in Developmental and Reproductive Toxicology......................................587A. When Are Toxicokinetics Not Essential?..................................................................588B. Use of Toxicokinetics in Study Design.....................................................................588C. Use of Toxicokinetics in Study Interpretation ..........................................................588D. Use of Toxicokinetics in Risk Assessment ...............................................................590VIII. Conclusions ........................................................................................................................592References ......................................................................................................................................592571© 2006 by Taylor & Francis Group, LLC

572 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONI. INTRODUCTIONIt has long been recognized that access of a nontopical drug to its ultimate site of action requiresthe agent to enter and be distributed through the systemic circulation. Although a drug may bebound to plasma protein or blood cells, the aqueous phase of blood plasma is a vehicle fordistribution of drugs to tissues throughout the body. Tissues can sequester, metabolize, or eliminatethe drug from the body. Pharmacokinetics quantitatively defines the absorption, distribution, metabolism,and elimination of drugs in the context of understanding how these factors influence a drug’sefficacy. Quantitative expressions (pharmacokinetic parameters) are used to define the extent ofdrug absorption, the magnitude of drug binding to plasma proteins, and the rate of drug eliminationfrom blood plasma, to name a few examples.A logical extension of the principles of pharmacokinetics is their application toward understandinghow the absorption, distribution, metabolism, and elimination of a compound influenceits toxicity. If an orally ingested compound is poorly absorbed from the gastrointestinal tract,extensively bound to plasma proteins, rapidly metabolized to a nonreactive moiety, or quicklyeliminated by urinary excretion, the relative potential for toxicity is diminished. Conversely, greaterabsorption of a xenobiotic, formation of reactive metabolites, and/or prolonged retention of theagent in the body generally leads to a relatively higher potential for toxicity. Toxicokinetics is notsimply the collection of pharmacokinetic parameters in a toxicology study; it means quantitativelyestablishing how such parameters relate to the toxicity potential of the test article.Toxicokinetics is practically useful in reproductive and developmental toxicology for: (1)experimental study design, (2) interpretation of toxicology data, and (3) risk assessment, especiallyin extrapolation of findings between species or between different exposure conditions. Each ofthese is detailed by example in this chapter. To provide the necessary background, this chapterbegins with an overview of how chemicals are absorbed, distributed, biotransformed, and eliminatedin the body, followed by definition of classical pharmacokinetic parameters.II. PRINCIPLES OF ABSORPTION, DISTRIBUTION,BIOTRANSFORMATION, AND ELIMINATIONXenobiotics may enter the body by a variety of routes, classically divided into two broad categories,enteral (via the gastrointestinal tract) and parenteral (not via the gastrointestinal tract). The formerincludes sublingual, oral (by diet or intubation), and rectal administration. Parenteral routes includedermal, intranasal, ocular, inhalation, subcutaneous, intramuscular, and intravascular. In experimentalstudies, the route of administration can be based on a variety of considerations, for example,the route of administration in humans or physical characteristics of the test article. As subsequentlydescribed in this chapter, toxicokinetics can be an important part of the rationale for route ofadministration in experimental studies, for example, when striving to optimize the level of systemicexposure in animals.A. Movement across Biological MembranesBy any route of administration, fundamental processes govern the rate and extent of xenobiotictransit across biological membranes. These processes are important determinants of xenobioticabsorption and distribution. Many agents cross biological membranes by simple passive diffusion,which depends on the characteristics of the biological membrane, the concentration gradient ofunbound xenobiotic, and the molecular properties of the xenobiotic, particularly molecular radius,degree of ionization, and lipid-water parturition coefficient. Most xenobiotics with molecular weightless than 800 Da transfer across the placenta by passive diffusion. 1 In facilitated diffusion (e.g.,placental transfer of D-glucose or L-lactate), 2 the xenobiotic binds to a “carrier” molecule, forming© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 573a complex that diffuses across membranes, depending on the concentration gradient. Active transportis a cellular-energy requiring process for movement of molecules against a concentration gradient.Several essential nutrients (e.g., amino acids, calcium, vitamin B 12 ) cross the placenta by activetransport. 2 Xenobiotics may also be subject to carrier-mediated transport. The p-glycoproteins (alsoknown as multidrug-resistant [mdr] proteins), organic-anion transporters, and organic-cation transporterscan actively “pump” certain xenobiotics out of cells. Gastrointestinal absorption of severaldrugs is limited by mdr activity in intestinal enterocytes. 3 Placental mdr has been shown to modulatethe teratogenic potential of avermectin, as mdr null-mutant mice are more susceptible to theinduction of cleft palate by that drug. 4 Macromolecules can move across cellular membranes bycellular pinocytosis (engulfing of fluid) or endocytosis (receptor mediated internalization, e.g.,movement of iron-transferrin across hemochorial placentas). 2When movement of a substance (e.g., anesthetic gases such as nitrous oxide or halothane 1 )across cell membranes is very rapid relative to the rate of blood flow to the tissue, uptake into thattissue is termed “flow-limited.” Conversely, when uptake is rate limited by movement across cellmembranes (e.g., antibiotics such as penicillin or streptomycin 5 ), uptake is considered to be “membranelimited” (or “diffusion limited” when that movement occurs by passive diffusion).B. Plasma Protein BindingThe binding of a xenobiotic to blood plasma proteins can be an important determinant of itsdistribution and elimination. Many drugs noncovalently bind to albumins, globulins, or lipoproteins.5 The large molecular radius of the drug-protein complex can effectively retard passivediffusion across membranes and thereby limit drug distribution into tissues and limit glomerularfiltration, both of which can influence toxicity. The extent of plasma protein binding by a compoundcan be assessed by ultrafiltration or equilibrium dialysis and is commonly expressed as %fu, thefraction of drug unbound by plasma protein, i.e., the “free fraction.” It is not uncommon for thereto be species differences in %fu, and small differences can be important. For example, considerthe case of a drug with %fu = 0.5% (99.5% bound) in the rat and %fu = 1.5% (98.5% bound) inhumans. At the same total concentration in plasma, the “free” drug concentration, which is freelydiffusible across cellular membranes into target organs, is three times higher in humans than inrats. Such a difference could be associated with a greater potential for toxicity. At the same time,urinary excretion of the drug could be greater in humans and might lower the potential for toxicity.It should also be appreciated that %fu can vary according to total drug concentration, given a finitenumber of binding sites on plasma protein. 5Unexpected drug interactions can occur when there is competition for binding to plasmaproteins. If a drug with a low %fu, such as digitoxin (%fu = 2%), is only 10% displaced fromplasma protein binding, the “free” drug concentration increases sixfold (%fu increases from 2% to12%). 6 Such interactions can have adverse consequences. An example is kernicterus in infantsresulting from displacement of bilirubin from plasma proteins by sulfonamide drugs, which competefor the bilirubin binding site, leading to more “free” bilirubin diffusion into the brain. 6C. Volume of DistributionAn empirical indication of how extensively a xenobiotic is distributed throughout the body is givenby the apparent volume of distribution (V d ), the virtual fluid volume in which the total amount ofdrug in the body appears to be dissolved to account for the measured drug concentration. Conceptually,a drug restricted to blood plasma (and assuming that the drug is not bound, metabolized, oreliminated) has a V d equivalent to the total blood plasma volume (about 3 liters for a 70-kg human);a drug widely distributed throughout all intracellular and extracellular (interstitial) space has a V dequivalent to the total body water volume (about 41 liters). The V d can be determined experimentallyby giving the test article as a bolus intravenous dose, in which case:© 2006 by Taylor & Francis Group, LLC

574 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONV d (l) = intravenous dose (mg)/C (0) (mg/liter) (13.1)where C (0) is the unbound plasma drug concentration determined by extrapolation back to the timeof injection. A large V d can be due to extensive distribution throughout body water and/or extensivebinding to tissues. Xenobiotics that partition into body fat can have a V d many times greater thanthe total body water volume. In the absence of having tissue concentrations of xenobiotics tocompare, differences in V d can sometimes help explain differences in the toxicity potential amongdrugs with comparable absorption and elimination rates.D. Blood and Seminiferous Tubule DistributionDistribution of xenobiotics between blood plasma and the seminiferous tubule lumen is limited byseveral cells and surrounding layers, including capillary endothelium and basal lamina, lymphaticendothelium, peritubular myocytes, seminiferous tubule basal lamina, and Sertoli cell–to–Sertolicell junctions. 7 These layers can prevent the access of certain molecules (e.g., immunoglobulins,albumin, inulin, and certain lower molecular weight compounds) to the seminiferous tubule lumenand hence limit exposure of the germinal epithelium and its progeny. 8,9 Certain p-glycoproteinsthat can serve as efflux transporters and further control access of xenobiotics to germ cells havebeen detected in the testes. Testicular concentrations of verapamil are significantly higher in mdr1a(-/-) mice relative to mdr1a (+/+) mice given the same dose, 10 and the administration ofp-glycoprotein inhibitors has been shown to increases testicular penetration of several drugs. 10–12E. Placental TransferThere are multiple pathways by which the embryo or fetus can be exposed to xenobiotics duringdevelopment. Initially, the preimplantation blastocyst is freely exposed to solutes within the uterus.Embryonic exposure can be rapid (thiopental reaches the blastocyst within minutes of maternaladministration) and extensive (the blastocyst attains concentration of nicotine that are approximately10-fold greater than the maternal plasma concentration). 13,14 In rodents and rabbits, the predominantinterface between the maternal and embryonic circulations during very early organogenesis is theyolk-sac placenta, a saclike cavity of extraembryonic endoderm. Subsequently during gestation,the predominant interface in these species is the chorioallantoic placenta, beginning at the 20 somitestage of embryonic development (about day 11 and 10 postfertilization in the rat and rabbit,respectively). 15,16There are species differences in the histology of the chorioallantoic placenta, including differentnumbers of cell layers separating the maternal and fetal blood. Table 13.1 gives a simplifieddescription of the maternal-fetal interface in common species during the fetal period of development.Histological features of placentas vary during gestation. For example, the rabbit chorioallantoicplacenta is epitheliochorial on day 8 postcoitus (pc), endotheliochorial on day 9.5 pc, and hemochorialfrom day 10 pc. 16 Furthermore, placental histology is not entirely uniform. For example,human chorionic epithelium variably consists of one layer (syncytiotrophoblast) or two layers(cytotrophoblast and syncytiotrophoblast) over the villous surface.Most importantly, the placenta should not be regarded as a barrier nor as a sieve. It is bothactive and selective in the transport of nutrients and xenobiotics. While placental permeability tomost proteins is low, immunoglobulin G (IgG) is actively (via receptor-mediated endocytosis) andselectively (all subclasses of IgG, but not other immunoglobulins) transferred across the placenta. 2But also there are gaps in the intervening cell layers, so virtually any substance might to someextent cross the placenta; for example, even erythrocytes pass in both directions across the placenta.Therefore, the salient questions are the rate and extent of placental transfer, which can vary duringgestation as a result of the dynamic nature of placentation and can vary between species having© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 575Table 13.1Histological features of the chorioallantoic placenta in common speciesType of Placentation and Representative Animal SpeciesEpitheliochorial(Pig, Horse)Syndesmochorial(Sheep, Cow)Maternal BloodEndotheliochorial(Dog, Cat, Ferret)Hemochorial a(e.g., Human,Monkey, Rabbit,Rat, Mouse)Maternal capillaryendotheliumUterine connectivetissueUterine epitheliumChorionicepitheliumFetal connectivetissueFetal capillaryendothelium Fetal BloodaThe number (1 to 3) of trophoblast cell layers that form the chorionic epithelium varies by species and stageof gestation.Source: Amoroso, E.C., Placentation, in Marshall’s Physiology of Reproduction, 3rd ed., Parkes, A.S., Ed.,Longmans, London, 1952, Vol. II, chap. 15.different types of placentation. For example, placental transfer of digoxin is much greater in humansand rodents (hemochorial placenta) than in sheep (syndesmochorial placenta). 17F. BiotransformationThe term metabolism is properly used to describe the absorption, distribution, biotransformation,and elimination of a xenobiotic, but for the present discussion it can be understood to refer tochemical transformations catalyzed by enzymes. These transformations are broadly categorized asPhase I or Phase II reactions. 18 Phase I reactions include chemical oxidations, reductions, orhydrolyses that generally increase polarity (hydrophilicity) but may also result in a more reactivemetabolite. 19 Phase II reactions can further increase hydrophilicity (which generally enhanceselimination), typically by conjugation with polar moieties (e.g., via acetylation, sulfation, glucuronidation,and glutathione or amino acid conjugation). Biotransforming enzymes are widelydistributed throughout the body, including the gonads and placenta, 20–26 where “local” formationof reactive metabolites can feature in a compound’s reproductive or developmental toxicity. Thehuman fetal liver has significant capability to metabolize xenobiotics (much more so than those ofrodents) using cytochrome P450 enzymes, including CYP1A1, CYP2D6, and CYP3A7, whichaccount for up to half the total fetal liver P450 content. 27 There is active sulfation but limitedglucuronidation enzyme activity in human fetal liver. 27In some cases metabolism can be the rate limiting determinant of xenobiotic elimination, butmetabolism of a given agent may differ markedly between species or over time (e.g., because ofcytochrome P450 induction) within a given species. This will generally influence and be reflectedby changes in the pharmacokinetic parameters described below. However, attempts to correlatepharmacokinetic parameters with a xenobiotic’s toxicity potential are confounded if only the parentmolecule is considered and a metabolite mediates the toxicity (e.g., testicular toxicity of diethylhexyl phthalate, 28 ethylene glycol monomethyl ether, 29 and n-hexane 30 ; the teratogenicity of 2-methoxyethanol is discussed in later sections).© 2006 by Taylor & Francis Group, LLC

576 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONG. EliminationThe principal routes of elimination from the body are urinary excretion (typically for polar xenobioticsand metabolites), fecal excretion (for nonabsorbed material and via biliary excretion forproducts of hepatic metabolism), and exhalation (for gases and volatile compounds). As a resultof physiological changes, the rate of compound elimination can vary markedly during pregnancyand postnatal development, as discussed below.Xenobiotics can also be excreted in semen. Basic xenobiotics tend to achieve semen:bloodplasma concentration (S:P) ratios greater than unity (e.g., methadone 31,32 and naltrexone 33 attainS:P ratios as high as 14), neutral drugs (e.g., caffeine 34,35 acetaminophen 36 ) have a ratio near 1, andacidic drugs tend to have a low S:P ratio (e.g., acetylsalicylic acid, 0.12; valproic acid, 0.07). 37,38This apparent relationship between drug disposition in semen and drug pK a has been attributed tothe slightly acidic (pH 6.6) 39,40 prostatic fluid component of seminal plasma and the phenomenonof ion trapping. 37,41,42 Given equilibration of the nonionized form of a drug by passive diffusion, atransmembrane difference in total drug concentration depends on the pH gradient and drug pK a .Valproic acid (pK a = 4.95) is ionized to a lesser extent in prostatic fluid (pH 6.6) than in bloodplasma (pH 7.4), where ion trapping occurs. Although seminal vesicle fluid is slightly basic (pH7.4–8.0) and constitutes a major percentage of semen volume, partitioning into prostatic fluidappears to be a more important determinant (protein binding aside) of drug disposition in semen.On a mass balance basis, semen is not a major route of elimination (the average human ejaculatevolume is about 3 ml), but xenobiotics in semen can potentially alter spermatozoa and be transferredduring intercourse. This transfer can present toxicity directly (e.g., vincristine in semen may causevaginitis 43 ) or after absorption into the partner’s blood. Seminal transmission to a pregnant femalepresents a potential concern for a teratogenic agent, but the relatively small ejaculate volume (3ml) delimits the dose. For example, a drug with a maximum plasma concentration of 10 µg/ml thatequilibrates with semen gives a body weight adjusted dose less than 1 µg/kg to a 55-kg partner.Nonetheless, malformations were produced in naive female rabbits mated to male rabbits treatedwith thalidomide prior to mating, 44 so implications of seminal transmission for potent teratogensshould be considered.Xenobiotics may also be excreted in milk. The pH of milk is slightly acidic relative to that ofplasma, and hence weak acids, like phenytoin and valproic acid, appear in low concentration inmilk because of the phenomenon of ion trapping in plasma. 45 Milk contains 3% to 5% emulsifiedfat that favors excretion of lipophilic drugs, such as diazepam and chlorpromazine. 45 Drugs maytransfer from plasma to milk by passive diffusion or by carrier-mediated transport (e.g., cimetidineby the organic cation–transport system). 46 Most drugs have a milk to plasma ratio of 1.0 or less,and only about 15% of drugs studied achieve a milk to plasma ratio greater than twofold. 47Movement of drug between blood plasma and milk is bidirectional, and retrograde diffusion frommilk into plasma can occur as the plasma concentration declines. 45 Hence, excretion of drugs inmilk ultimately depends on the rate and extent of milk secretion and not simply the milk to bloodconcentration ratio. Consequently, the percentage of dose excreted in milk is generally less than1%. 48 The milk to blood plasma ratio per se also does not predict infant systemic exposure, whichis determined by the absolute concentration in milk, quantity of milk intake, oral bioavailability,and rate of drug clearance by the infant. 45,47,49 Furthermore, oral bioavailability is influenced by thecomposition of the milk. For example, calcium in milk seems to inhibit the absorption of tetracyclines.50H. Physiological Changes in PregnancyPhysiological changes in pregnancy can affect absorption, distribution, metabolism, and eliminationof xenobiotics. 51–53 Increased gastric emptying time prolongs residence time in the stomach and© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 577small intestine that could enhance oral bioavailability. Pulmonary tidal volume increases whileinterstitial changes in the lung decrease gas diffusion rates. There is increased blood flow to theskin that could enhance dermal absorption of chemicals. Plasma volume increases nearly 50%during human pregnancy, 54 along with an increase in total body water, which can result in a higherV d . At the same time, plasma albumin levels decrease as much as 38% while most lipid fractionsrise, 54 which can change %fu. There is no major sustained change in liver blood flow or xenobioticmetabolizing enzyme activities during pregnancy. 53 However, a xenobiotic with low hepatic extractiondue to high protein binding in nonpregnant females may be subject to enhanced hepaticelimination during pregnancy if %fu increases. 53 Human renal blood flow and glomerular filtrationrate increase up to 33% and 66%, respectively, 54 which promotes more rapid clearance of xenobioticseliminated predominantly unchanged by the kidneys. There can be significant species differencesin gestational physiology. For example, contrary to the situation in humans, there is noincrease in renal blood flow in pregnant rabbits. 55I. Physiological Changes in NeonatesReproductive toxicology studies can include indirect (via milk) or direct administration of testarticles to immature animals. Neonates, particularly rodents because of their relatively short periodof fetal development, and infants have physiological differences from adults that can result inqualitative or quantitative differences in compound absorption, distribution, biotransformation, orelimination including:• An immature gastrointestinal tract (e.g., fewer layers of intestinal epithelium, longer gastricemptying time, lower peptic activity), 56 which may result in enhanced absorption of substancesranging from lead 57 to insulin 58• Exponential development of lung alveoli from birth to 4 years of age in children 59• Greater total body water and extracellular fluid volume as a percentage of body weight that canaffect xenobiotic distribution 60• Generally lower plasma protein binding of drugs in newborns 61• Incompletely developed blood-brain barrier with greater hematoencephalic exchange 62• Minimal or low levels of xenobiotic-metabolizing enzymes 60,63–65• Reduced biliary excretion of xenobiotics 66,67• Low glomerular filtration rate (about one-third the adult rate) and renal tubular secretion rates 68,69which can result in a slower rate of renal excretion of xenobiotics 60These differences can sometimes account for the greater toxicity of certain compounds inneonates. The use of analgesics during labor and delivery can result in transplacental exposurefollowed by clinical toxicity in the neonate because of the slow rate of drug elimination. Forexample, meperidine readily equilibrates across the placenta, but its elimination half-life in theneonate is two- to sevenfold longer than in the adult. 70 Moreover the incomplete blood-brain barrierpotentially increases the risk of respiratory depression in the infant. 71 A second classical exampleis that of chloramphenicol, an anti-infective that produced cases of “gray-baby syndrome” (cyanosis)in infants because of hematotoxicity (blood dyscrasia). 72,73 About 90% of chloramphenicol isexcreted as a monoglucuronide adduct in adult humans. Premature and newborn infants have lowrates of both glucuronide conjugation and glomerular filtration that result in a longer half-life (10to 48+ h in newborns vs. 4 h in adults) 74 with elevated and prolonged plasma concentrations ofchloramphenicol. Greater sensitivity to toxicants is not always simply a matter of less drug metabolismin the infant. Theophylline is extensively metabolized in adults, with little free drug excretedin urine. The half-life of theophylline is extended in infants because of an undeveloped rate of N-demethylation, 75 but at the same time there is active methylation of theophylline to producecaffeine, 76 itself having a half-life nearly 20 times greater in infants (100 h) 77 vs. adults (4.9 h). 78© 2006 by Taylor & Francis Group, LLC

578 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONOne CompartmentModelTwo CompartmentModelk ak ak 121k el1k elk 212Figure 13.1One- and two-compartment pharmacokinetic models.III. CLASSICAL PHARMACOKINETICSIn classical pharmacokinetics, the body is represented by theoretical compartments of drug distribution.79–81 A complete model would require many compartments to account for the amount ofcompound distributed among intestinal lumen, blood, various tissues, bile, feces, urine, etc. Inpractice, data on chemical concentration in each “compartment” are rarely collected, but bloodplasma concentration data are readily obtained and generally provide a reliable indication ofsystemic exposure. However, in some cases (e.g., placental transfer studies), tissue samples mustbe collected to meet the objectives of the toxicokinetic study.A. Simple Compartmental ModelsIn basic modeling, the body is represented by one or two “compartments.” The simplest model isa one-compartment model in which blood plasma (or whole blood) is assumed to be in rapidequilibrium with all other compartments, including those where compound absorption (k a ) andelimination (k el ) occur. A two-compartment model adds a peripheral compartment representing alltissues that do not rapidly equilibrate with blood plasma (Figure 13.1). The latter model does notimply equivalent compound concentrations in those tissues, but only that changes in compoundconcentration in those tissues occur proportionally to changes in compound concentration in theperipheral compartment.In a typical toxicokinetic study, blood samples are collected at various times after compoundadministration (oral in this example), compound concentration in the blood plasma samples ismeasured, and the data are presented graphically (Figure 13.2). Note how the aforementionedprocesses of absorption, distribution, and elimination conceptually relate to the shape of the curve.B. Basic Pharmacokinetic ParametersThe fundamental parameter used to describe systemic exposure is C (time) , compound concentrationat a given time. C (time) is given in units of concentration (e.g., µg/ml). C (0) indicates initial concentrationafter bolus intravenous administration (determined by extrapolation back to the moment ofinjection). For intravenous infusion studies when the rate of compound administration equals therate of elimination, C ss refers to the steady-state concentration. The basic parameters that describethe blood plasma drug concentration vs. time curve are:C max : the peak concentrationt max : the time (min or h) at which C max occursAUC 0-t : area-under concentration vs. time curve from time (zero) to time (t), in the units of the productof time and concentration (e.g, min × µg/ml)AUC 0-∞ : pertains to AUC from time zero until the concentration is virtually zero.© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 579CC maxDistributionAbsorptionAUCEliminationDoset maxTimeFigure 13.2Typical blood plasma drug concentration vs. time curve following a single oral dose, depictingthe maximum plasma drug concentration (C max ) achieved at time t max . Shading indicates the areaunder the concentration vs. time curve (AUC).C max and AUC can be determined graphically or computationally, with the AUC calculated asthe sum area of trapezoids defined by the sampling time points on the horizontal axis. C max indicatesthe highest systemic concentration achieved, whereas the AUC gives an indication of the totalityof systemic exposure. Depending upon route of administration and the rates of absorption, distribution,and elimination, a wide variety of blood plasma concentration vs. time profiles may beobserved. As will be discussed subsequently, a key application of toxicokinetics is to identify theexposure metric (e.g., C max , C ss , or AUC) that best correlates with toxicity potential. Experimentally,different routes of administration can be used to produce profiles having comparable AUC whileC max is significantly different, or vice versa (Figure 13.3).The net rate of compound elimination from the central compartment (blood plasma in theseexamples) is calculated from the semilogarithmic plot of compound concentration vs. time (Figure13.4). If such a plot is linear, the data are well described by the one-compartment model, and theslope of the resulting straight line gives the elimination rate constant (k el ) in the reciprocal unit oftime (e.g., min –1 or h –1 ):k el = 2.303 × slope (13.2)Another useful pharmacokinetic parameter is the apparent elimination half-life (t ½ ), the timefor the compound concentration in the central compartment to decrease by one-half, which can bedetermined graphically (Figure 13.4) or computationally. The units of half-life are time (e.g., minor h). In a one compartment model:t ½ = 0.693 / k el (13.3)Typically in a two-compartment model, the terminal elimination half-life (t ½ ) is quoted. Moreprecisely, there are two composite rate constants (k α and k β ) reflecting distribution (k 12 and k 21 ) andelimination (k el ). These can be used to calculate t ½α and t ½β , respectively, and one can compute an“effective” half-life for the central compartment from the AUC-weighted average of t ½α and t ½β ,i.e., effective half-life = (t ½α )(fAUC α ) + (t ½β )(fAUC β ), where fAUC α and fAUC β represent thefraction of the total AUC marked by lines α and β, respectively (see Figure 13.4).© 2006 by Taylor & Francis Group, LLC

580 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONCA. Single iv doseC max = 20 ug/mLAUC = 95 ug⋅h/mL_______________________________TimeCB. Continuous iv infusionC max = 4.7 ug/mLAUC = 95 ug⋅h/mL_______________________________TimeCC. Multiple iv doseC max = 4.7 ug/mLAUC = 27 ug⋅h/mL_______________________________TimeFigure 13.3Theoretical blood plasma compound concentration vs. time curves following various dosingregimens. In panels 3a (single bolus intravenous injection) and 3b (continuous intravenousinfusion), the AUC 0–t is comparable while the C max differs. In panel 3c (divided bolus intravenousinjections), the C max is comparable while the AUC 0–t differs from that in panel 3b.Clearance (Cl) is a concept related to half-life. Clearance from the body depends on biotransformationand excretion. It is the virtual volume of the central compartment that is completely“cleared” of the xenobiotic in a given unit of time, hence the units are in terms of flow (e.g.,liter/min). Just as the virtual volume V d can be experimentally determined following intravenousadministration:© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 581Log CC1 /2Cslope = k el /−2.303t t + t 1/2TimeLog CαβTimeFigure 13.4Semilogarithmic plots depicting one-compartment (top) and two-compartment (bottom) models.In the one-compartment model, first-order kinetics follow, and the plot reveals a straight line withdecreasing slope proportional to the elimination rate and inversely proportional to t ½ . In the twocompartmentmodel, the plot shows a decreasing exponential curve that can be resolved intotwo component straight lines (α and β). Their slopes reflect hybrid rate constants (α and β) andthat are inversely proportional to t ½α and t ½β , respectively.V d (l) = intravenous dose (mg) / C(0) (mg/l) (13.4)Cl can be computed from:Cl (l/min) = intravenous dose (mg) / AUC (0-∞) (min × mg/l) (13.5)V d and Cl are related by the expression:Cl (l/min) = [0.693 V d (l)] / t ½ (min) (13.6)Thus, more rapid elimination results in a shorter half-life and more rapid clearance, but clearancealso depends on distribution from the central compartment. Hence, a larger V d yields a higher rateof clearance from the central compartment.© 2006 by Taylor & Francis Group, LLC

582 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe fraction of the dose reaching the systemic circulation (the central compartment in theseexamples) is referred to as bioavailability (F). Oral bioavailability indicates the extent to which acompound is absorbed into blood plasma following oral administration. Experimentally, oral bioavailabilitycan be determined by comparing AUC following oral (AUC po ) and intravenous (AUC iv )administration of the identical dose:F = AUC po / AUC iv (13.7)where the same expression (0–t or 0–∞) of AUC is used for both.Some useful interrelationships among the foregoing pharmacokinetic parameters are:AUC iv = dose(i.v.) × t ½ / 0.693 V d (13.8)AUC po = F × dose(p.o.) × t ½ / 0.693 V d (13.9)These expressions illustrate that low bioavailability, a shorter half-life, and a greater volume ofdistribution lower the blood plasma AUC. Conversely, high bioavailability, slow elimination (longhalf-life), and a limited volume of distribution increase the blood plasma AUC. Because theserelationships reflect the underlying, fundamental processes of compound absorption, distribution,biotransformation, and elimination, they provide a rationale for explaining differences in toxicokineticsamong different compounds or among different conditions (e.g., pregnant vs. nonpregnant)for the same compound.IV. PHYSIOLOGICALLY BASED PHARMACOKINETICSPhysiologically based pharmacokinetic (PBPK) models depict individual or grouped organs (ororgan systems) as compartments with interconnections representing pathways for compound uptakeand disposition. Unlike classical pharmacokinetic models, PBPK models are not generic but arebased mechanistically on anatomical, physiological, biochemical, and physicochemical parameters.These parameters include, for example, body weight, organ weight or volume, cardiac output,regional blood flow, tissue permeability and tissue to blood partition coefficients, and metabolicrate constants. Mass balance differential equations are used to calculate rates of change in theamount of xenobiotic in each compartment. Consequently, PBPK models require more data andmore sophisticated computational methods than classical pharmacokinetic models. PBPK modelsoffer a unique benefit, in that they allow extrapolation beyond the exposure paradigm, physiologicalstate, and species used to generate the original model. This is possible because the PBPK parametersfundamentally relate to processes of absorption, distribution, biotransformation, and elimination.After defining a PBPK model for one set of conditions, the model parameters can be scaled toanother set of conditions (e.g., another species, another time in pregnancy) to predict blood andtissue concentrations. This application of PBPK models is particularly useful in studies of compounddisposition in pregnancy, including placental transfer, where the cost of performing experimentson multiple days of gestation and/or in multiple species can be prohibitive. An overview of selectedPBPK models of pregnancy is presented below, and the reader is referred to additional sources formore information on PBPK models and their applications in toxicology. 82–85A. Physiologically Based Pharmacokinetic Models of PregnancyEarly PBPK models of pregnancy described the disposition of drugs (tetracycline, 86 morphine, 87theophylline, 88 methadone, 89 and pethidine 90 ) among blood and various maternal and fetal tissues.These models successfully described drug distribution, identifying major reservoirs and the impact© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 583of physiological changes (e.g., blood flow) on drug elimination. In some cases, they were used tosimulate human tissue concentration vs. time curves in the mother and fetus.A PBPK model described by Fisher et al. 91 was among the first to incorporate gestation timedependentphysiological parameters, including maternal body weight, maternal body fat content,blood flow, and placental and fetal growth (the last two were not modeled prior to gestation day[GD]14). In addition the model employed different partition coefficients and metabolic constantsfor naive versus pregnant rats. Pregnant rats were exposed throughout gestation to trichloroethylene(TCE) by inhalation, bolus gavage or drinking water, and the maternal and fetal blood levels ofTCE and plasma levels of its oxidative metabolite trichloroacetic acid (TCA) were measured onGD 20 or 21. Values generated by computer simulation with the PBPK model compared favorablywith actual data generated at isolated time points near term. Although the “validation” data arescant, in principle this model could be used to simulate maternal and fetal exposure to TCE andTCA at any point from GD 14 to term following either inhalation or oral exposure.Further evolution of the rat pregnancy PBPK model was evident in the work of O’Flaherty etal. 92 in their pharmacokinetic studies of dimethyloxazolidine-2,4-dione (DMO). This model includesa more complete (entire period of gestation) and more extensive description, for both rat and mouse,of gestation time-dependent changes in maternal, placental, and embryo or fetal tissue volumes,blood flow, and pH. The disposition of DMO, a weak organic acid with high %fu, was primarilyinfluenced by compartment pH and volume. A variation of this PBPK model has been used to studythe disposition of 2-methoxyethanol (2-ME) and its metabolite 2-methoxyacetic acid (2-MAA), 93a rodent teratogen. Model simulation agreed well with actual concentrations in maternal plasma(2-ME, 2-MAA), the embryo (2-MAA), and extraembryonic fluid (2-MAA) following bolus gavageor subcutaneous injection of mice on GD 11. PBPK modeling was subsequently applied to studythe maternal-fetal disposition of 2-MAA in mice on GD 8 and 13 following intravenous administrationof 2-ME, 94 and in pregnant rats on GD 11 to 15 following oral, intravenous, 95 or inhalationexposure to 2-ME. 96 Furthermore, a PBPK model for human pregnancy was developed using thegeneral structure of the rat model but with partition coefficients and physiologic and metabolicconstants for humans. 96 PBPK modeling provided a rational basis for predicting systemic levels of2-MAA arising from 2-ME inhalation in humans and rats, taking into account several key interspeciesdifferences (e.g., hepatic clearance of 2-ME, half-life of 2-MAA, and ventilation rate). Thismodel was used to predict the 2-ME inhalation exposure concentration estimated to yield maternalplasma C max and AUC values in women corresponding to those associated with developmentaltoxicity in rats.The foregoing provides a foundation for PBPK pregnancy model representation and parameterization,including established values for gestation time-dependent anatomical and physiologicalparameters. Application of such models would require consideration of a given compound’s absorption,distribution, metabolism, and elimination, including the generation of chemical-specific parameters(e.g., partition coefficients, uptake, metabolism, and excretion rate constants). Given themagnitude of experimentation required to perform classical pharmacokinetics on multiple days ofgestation by multiple routes (and bearing in mind that the collection of fetal tissue requires terminalsampling procedures), there is high potential value for the application of PBPK modeling indevelopmental toxicology and risk assessment. 97,98 This is further enabled by the recent developmentof a comprehensive (including 27 maternal and 16 embryo-fetal compartments) PBPK model forpregnancy that can be fitted with either rat or human data, permitting comparisons across speciesand at multiple times of gestation. 99,100V. SAMPLING STRATEGIESGenerally six to eight time points over a 24-h period are required to produce an accurate plasmaconcentration vs. time curve. A serial sampling strategy obtains a sample from treated subjects at© 2006 by Taylor & Francis Group, LLC

584 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 13.2Example of a composite sampling strategySubject Identification NumberTime (h) 1 2 3 4 5 6 7 8 9 10 11 120 0.5 1.0 2.0 4.0 8.0 12 24 The general relationship is A = B, where A = (number of animals sampled) ×(number of time points sampled per animal) and B = (number of time points) ×(samples/time point)Table 13.3Sampling strategies for adult animalsSpecies Status a SiteFrequency(per 24 h)Sample BloodVolume (ml) bRat Live Tail vein 1 to 8 0.2Rat Terminal Vena cava 1 5.0Mouse Terminal Vena cava 1 0.5Rabbit Live Ear vein 1 to 8 0.5aWhere status = live, either a serial or composite sampling strategymay be used.bApproximate blood volumes offered for general reference only. Typicallywithin a 24-h period, the maximum blood withdrawn from adultanimals ranges from 7.5% to 20% of total blood volume, dependingupon the recovery period; for further guidance refer to Diehl, K.-H.,et al., J. Appl. Toxicol., 21, 15, 2001.each time point. This affords individual subject concentration vs. time curves, and pharmacokineticparameters can be calculated for individual subjects. These parameters can be compared withindividual subject toxicity data or can be averaged to obtain group mean pharmacokinetic parameters.This strategy can be used when the total sampling volume required to obtain all time pointsis compatible with the test conditions (species, test article, analytical sensitivity). In some situations,there can be a concern for the influence of subject restraint and/or blood or milk withdrawal ontoxicology study endpoints. In this case, a separate (“satellite”) group of animals, given the sametreatment regimen as the toxicology study animals, can be serially sampled to obtain toxicokineticdata only. Serially sampling a minimum of three animals is typically needed to characterize systemicexposure for a given treatment regimen.Some test articles have an effect on blood pressure or blood volume that limits the volume ofblood samples obtainable. Also, for some species or ages of animals, it may not be appropriate toremove the required amount of blood to cover all the desired time points (see Table 13.2, Table13.3, and Table 13.4). Both of these situations require a composite sampling strategy. A compositesampling strategy obtains samples from individual subjects at only one or a subset of the desiredtime points. For example, 12 subjects per group, with each sampled at 2 time points, yields 3independent blood samples at each of 8 time points (Table 13.2).The composite sampling strategy can be applied to toxicology study animals to minimize bloodsampling influences on study endpoints. Composite sampling can also be applied to satellite studyanimals when terminal sampling is required; for example when a relatively large volume of bloodper sample is needed or when taking tissue (target organs, embryos, fetuses) samples.© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 585Table 13.4Sampling strategies for embryos and fetusesSpecies Day Postcoitus Sample Sample Blood Volume (ml) aMouse 10 to 18 Tissue —Rat 9 to 20 Tissue —Rat 21 Blood (decapitation) 0.2Rabbit 10 to 18 Tissue —Rabbit 29 Blood (decapitation) 1.0aApproximate blood volumes offered for general reference.Table 13.5Sampling strategies for neonatal ratsAge (days) Status Site Sample Blood Volume (ml) a2 Terminal Decapitate 0.412 Terminal Vena cava 0.8aApproximate blood volumes offered for general reference.By use of a vacuum-assisted apparatus, 102 it is generally possible to collect 1 to 10 ml of milkfrom lactating rats injected with oxytocin. An alternative approach is to euthanize the suckling,neonatal animal, from which one can collect mother’s milk from the neonate’s stomach. Neonatalblood can then be obtained to assess neonatal systemic exposure.VI. REGULATORY EXPECTATIONSThe only comprehensive regulatory guidance relating to toxicokinetics in reproductive and developmentaltoxicology was written under the auspices of the International Conference on Harmonization(ICH). The topic is addressed both in the Guideline on Detection of Toxicity to Reproductionfor Medicinal Products (ICH Topic S5A) 103 and in the guideline entitled The Assessment of SystemicExposure in Toxicity Studies (ICH Topic S3A). 104 These guidelines are intended to apply to nonclinicaltoxicity studies of human medicinal products.A. ICH Reproduction Test GuidelineThis guideline uses the term “kinetics” equivalently to the term “toxicokinetics” as used in thischapter. The guideline discusses kinetics as it relates to study design and as it relates to studyinterpretation. Relevant passages of the guideline are quoted verbatim, along with this author’scomments reflecting practical experience.Kinetic information may be obtained in a variety of toxicology studies preceding the conductof reproductive or developmental toxicity studies. Such information can be extremely helpful inthe design of the latter studies. “It is preferable to have some information on kinetics before initiatingreproduction studies since this may suggest the need to adjust choice of species, study design, anddosing schedules. At this time the information need not be sophisticated nor derived from pregnantor lactating animals.”The following paragraphs describe the application of kinetics to species selection, study design,and dosing rationale.According to ICH guidelines, kinetic information could result in a justification for a single,relevant (to humans) species or for the use of two rodent species for embryo-fetal developmentstudies (normally both a rodent and a nonrodent species would be required). The rationale to justifya single species would include kinetic (including test compound metabolism) and pharmacologicaldata showing that the species is uniquely relevant to humans. Deselection of a nonrodent species© 2006 by Taylor & Francis Group, LLC

586 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmay be justified by kinetic data, for example, if the maximum practical or tolerated dose resultsin a systemic exposure level less than would be seen in humans.By ICH guidelines, a rodent embryo-fetal development study may be conducted as a standalonestudy (embryo-fetal development study, 4.1.3) or as an extension of the rodent fertility study(fertility, early embryonic, and embryo-fetal development study, 4.1.1 and 4.1.3). Furthermore,ICH Guidelines indicate: “For drugs where alterations in plasma kinetics are seen following repeatedadministration, the potential for adverse effects … may not be fully evaluated ….” It stands toreason that for drugs that induce their own metabolism (so-called “autoinducers”), the systemicexposure level during organogenesis might be much lower if dosing were initiated before mating(as in study 4.1.1 and 4.1.3) than if dosing began at the time of implantation (study 4.1.3). Othercompounds may reach a higher level of systemic exposure upon repeated administration (accumulate),in which case if dosing was initiated at the time of implantation (study 4.1.3), the systemicexposure level during organogenesis might be much lower than if dosing began before mating(study design 4.1.1 and 4.1.3).Kinetics often plays a decisive role in the rationale for dose selection. For poorly bioavailablecompounds, kinetic data can show that a plateau in systemic exposure is reached at a certain dose.That dose may be justifiable as a high dose because there is, “little point in increasing administereddosage if it does not result in increased plasma or tissue concentration.” Selection of lower dosagesis made, “depending on kinetic and other toxicity factors.” Such kinetic factors include the anticipatedlevel of systemic exposure in humans. The lowest animal test dose is generally selected toprovide a small multiple of the C max and AUC measured in humans at therapeutic dosages. Animaltest doses should also be selected to achieve good separation in the systemic exposure levels amongthe groups. Use of kinetics data can point to a dose selection surprisingly different from the typicalproportional spacing of administered doses because compounds frequently exhibit nonproportionalincreases in systemic exposure with increasing administered dose. For a compound with greaterthan dose-proportional systemic exposure, the AUC at the usual middose might be undesirablyclose to that at the high dose if kinetic information were not considered.Generally the route of administration to animals is selected to match the clinical route ofadministration. ICH guidelines indicate: “If it can be shown that one route provides a greater bodyburden, e.g., … (AUC), there seems little reason to investigate routes that would provide a lesserbody burden ….” This can be a practical advantage for the conduct of animal studies when theclinical route of administration is difficult to use in animals (e.g., intranasal) or when the clinicalroute of administration results in poor systemic exposure in the test animals.Some medicines are used clinically by multiple routes of administration. For the supportinganimal toxicology studies, “One route … may be acceptable if … a similar distribution (kineticprofile) results from different routes.” Therefore, a single route of administration in nonclinicalstudies may support multiple clinical routes if the selected route of administration results in animalsystemic exposure that sufficiently exceeds human C max and AUC by the clinical routes.According to ICH guidelines, “The usual frequency of administration is once daily but considerationshould be given to use either more frequent or less frequent administration taking kineticvariables into account.” This can apply if the drug’s half-life in the preclinical species is muchshorter than in humans, especially considering that embryonic development is more rapid in rodentsthan in humans. For example, if a drug with an elimination half-life less than 2 h is given oncedaily, there would be minimal systemic exposure for more than 12 h each day, and multiple dailyadministration (e.g., b.i.d. or t.i.d.) should be considered. As an alternative to more frequentadministration, another method of drug administration (e.g., continuous infusion) could be selectedto prolong the duration of exposure or a means (e.g., dietary or subcutaneous depot administration)that prolongs the absorption phase could be employed.In addition to the use of kinetics to assist with study design, kinetics is important for studyinterpretation: “At the time of study evaluation further information on kinetics in pregnant or© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 587lactating animals may be required according to the results obtained.” Also, kinetics in pregnantanimals “may pose problems due to the rapid changes in physiology. It is best to consider this asa two or three phase approach.” This may refer to the fact that kinetics evaluated at an a prioriselected time during gestation (e.g., near term) may not provide relevant toxicokinetic informationfor developmental toxicity induced at a different time in gestation (e.g., during organogenesis)because of changes in maternal physiology and placentation that occur throughout pregnancy.B. ICH Toxicokinetic GuidelineAccording to this guideline, toxicokinetics are not an absolute requirement in reproduction studies:“The limitation of exposure in reproductive toxicity is usually governed by maternal toxicity. Thus,while toxicokinetic monitoring in reproductive toxicity studies may be valuable in some instances,especially with compounds with low toxicity, such data are not generally needed for all compounds.”The interpretation of these statements could be that toxicokinetics in reproduction studies maynot add much value if: (1) dose-response relationships for toxicity are the same in nonpregnantand pregnant animals, (2) the high dose could be no higher due to maternal toxicity, and (3)preexisting toxicokinetic data are available for the same species, albeit in nonpregnant animals.However, “Where adequate systemic exposure might be questioned because of absence of pharmacologicalresponse or toxic effects, toxicokinetic principles could usefully be applied …,” and“Consideration should be given to the possibility that the kinetics will differ in pregnant and nonpregnantanimals.” So, when toxicity data from nonpregnant animals are used to select doses forpregnant animals and the expected maternal effects do not occur (rationale for dose selection did nothold true, possibly because of a difference in kinetics during pregnancy), toxicokinetic data may berequired to determine if the high dose was associated with an adequate level of systemic exposure.ICH guidelines do not absolutely require placental or milk transfer studies, but “In somesituations, additional studies may be necessary or appropriate in order to study embryo/foetaltransfer and secretion in milk.” Also, “Consideration should be given to the interpretation ofreproductive toxicity tests in species in which placental transfer of the substance cannot be demonstrated.”The second point could be raised when there is no effect on the embryo or fetus. It is not aconcern, however, if a maternally toxic dose was used and the maternal systemic exposure levelwas sufficiently higher than that observed in the clinic, where it is unlikely the extent of placentaltransfer would be known. Nonetheless, the point should be considered when presenting negativedevelopmental toxicity data for a compound in a therapeutic class previously associated withteratogenicity.ICH guidelines state, “Secretion in milk may be assessed to define its role in the exposure ofnewborns.” However, drug concentration in milk needs to be considered with respect to its bioavailabilityin milk. It is otherwise difficult to interpret the toxicological significance of the observedconcentration (see Section II.G.). In terms of validating the exposure of the neonate in a postnatalstudy, measurement of the drug in pup’s blood is more relevant than measurement of drug in milk.Finally, regardless of the preclinical data, product labeling will typically indicate (1) a drug hasthe potential to be secreted in breast milk or (2) the (more important) existence of human data onsecretion of the drug in milk.VII. TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGYThis section of the chapter combines scientific, regulatory, and practical aspects of toxicokineticsin developmental and reproductive toxicology to define how it can support nonclinical safetyassessment.© 2006 by Taylor & Francis Group, LLC

588 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. When Are Toxicokinetics Not Essential?When systemic exposure data at similar doses are available in the same species from preexistingstudies, it is generally not necessary• To generate exposure data in a rodent fertility study because preexisting data exist.• To generate exposure data in developmental toxicity studies with animals treated up to a doseproducing toxicity in the dams, given toxicokinetic data generated in nonpregnant animals up tothe same doses and the high dose resulted in similar toxicity.• To see if a compound crosses the placenta or to measure how much compound is in the embryowithout regard to any specific biological question. (Because human data for placental transfer orembryonic drug level are very rare, there is no value in generating placental transfer data in animalsthat cannot be interpreted in the absence of a specific question.)• To see if a drug is present in milk (see the preceding section).B. Use of Toxicokinetics in Study DesignAs mentioned previously, toxicokinetics can support study design with regard to selection of species,dose, and route of test material administration. For a species not commonly used in generaltoxicology studies but frequently used for developmental toxicity studies (e.g., rabbit), a preliminarytoxicokinetic evaluation (not necessarily in pregnant animals) provides assurance that an adequatelevel of systemic exposure can be achieved at a maximally tolerated dose. A preliminary study canalso look for evidence of major metabolites produced in humans. Such data may result in deselectionof a conventional species in favor of a more clinically relevant species. The use of toxicokineticsin dose selection was discussed in the preceding section.An alternative route of administration to animals can be selected to enhance systemic exposurerelative to that resulting from dosing by the intended clinical route. Drugs with low oral bioavailabilityin animals can be associated with C max and AUC values not substantially different fromthose seen in humans despite orally administered doses in animals that are high multiples of theclinical dose on a milligram per kilogram basis. However, an alternate route of administration (e.g.,intravenous) could provide a much higher C max and AUC in animals at a lower dose.C. Use of Toxicokinetics in Study InterpretationOften the doses for developmental toxicity studies are selected based on preceding data, usuallyincluding systemic exposure data collected in toxicology studies with nonpregnant animals. If adose producing clear evidence of toxicity in nonpregnant animals does not do so in the maternalanimals of the developmental toxicity study, it is useful to obtain toxicokinetic data in pregnantanimals to determine if the lack of maternal toxicity is attributable to lower than expected systemicexposure. Despite the absence of maternal toxicity, the additional data may indicate the study isadequate if the maternal exposure level is a sufficient multiple of that observed in humans. It isalso possible that an unexpectedly high degree of maternal toxicity can be observed relative to thatexpected from studies in nonpregnant animals. Here again, a comparison of systemic exposurebetween pregnant and nonpregnant animals given the same dose can help determine if the differencein toxicity is attributable to differences in kinetics or inherent susceptibility.The outcome of developmental and reproductive toxicity studies may include: species-specificdevelopmental toxicity, unique sensitivity of neonatal animals, and/or absence of a predicted classeffect. While it is important to consider mechanisms of toxicity in these cases, an evaluation oftoxicokinetics is the most practical first step and can aid in the design of mechanistic studies. Ageneral approach for investigating such outcomes involves© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 589Table 13.6Administered doses of 13-cis-retinoic acid (13-cis-RA) and systemic levels of13-cis-RA and 13-cis-4-oxoretinoic acid (13-cis-4-oxoRA) reported to beteratogenic in various speciesMaternal Plasma AUC (ng·h/ml) Embryo AUC (ng·h/g)Species Dose (mg/kg) 13-cis-RA 13-cis-4-oxoRA 13-cis-RA 13-cis-4-oxoRAMouse a 100 21205 N/A 449 N/ARat b 75 9226 10907 789 259Rabbit c 15 49100 58500 2390 1830Monkey d 2.5 2325 1784 956 590Note: AUCs determined on last day of treatment.aNMRI mice treated on day 11 postcoitus (pc) (Kochar, D.M., Kraft, J., and Nau, H., in Pharmacokineticsin Teratogenesis, Vol. 2, Nau, H. and Scott, W.J., Eds., chap. 16.); N/A indicates datanot available.bTeratology data in Sprague-Dawley rats treated on days 8–10 pc (Agnish, N.D., Rusin, G., andDiNardo, B., Fund. Appl. Toxicol., 15, 249, 1990), and kinetic data from Wistar rats treated ondays 7–12 pc (Collins, M.D., et al., Toxicol. Appl. Pharmacol., 127, 132, 1994)cRabbits treated on days 8–11 pc (Eckhoff, C., et al., Toxicol. Appl. Pharmacol., 125, 34, 1994).dCynomolgus monkey, 2.5 mg/kg once daily on days 16–26 of pregnancy then 2.5 mg/kg twicedaily on days 27-31 of pregnancy (Hummler, H., Korte, R., and Hendrickx, A.G., Teratology, 42,263, 1990; Hummler, H., Hendrickx, A.G., and Nau, H., Teratology, 50, 184, 1994).1. Identifying the period of susceptibility2. Evaluating maternal systemic exposure and compound metabolism3. Evaluating embryo-fetus, pup, or target organ (e.g., testes, ovary) exposure level, including majormetabolitesBecause of the potential for differences in toxicokinetics according to the period of gestationor lactation, it can be important to establish the sensitive period for toxicity before conducting atoxicokinetic evaluation. So-called critical window studies may be conducted to define the mostappropriate day of gestation or lactation for subsequent investigation.If developmental toxicity occurs in one species and not another, an initial evaluation of maternalsystemic exposure levels (parent molecule and major metabolites) in both species can help explainthe results. If maternal systemic exposure levels are comparable between sensitive and nonresponsivespecies, then studies of placental or lactational transfer in both species at the relevant timesof development are warranted. Because terminal sampling is required for such investigations andtherefore more animals must be used, it is logical to first determine that the developmental toxicityresults are not explainable simply in terms of the maternal systemic exposure data, which typicallyrequires many fewer animals to collect.The compound 13-cis-retinoic acid provides an illustration of how maternal systemic exposureand metabolism, as well as placental transfer, can be used to understand species differences interatogenicity. The reported teratogenic oral dose of 13-cis-retinoic acid varies 40-fold amongspecies (Table 13.6).Toxicokinetic evaluations in these species revealed significant differences in metabolism. Betaglucuronidationis the major pathway in rodents 107,111 (t ½ less than 1 h) 112,113 but not in rabbits ormonkeys (t ½ approximately 10 to 13 h). 108,110 Given the relationship between half-life and AUC, itis understandable that at teratogenic doses, the maternal AUCs for 13-cis-retinoic acid and theactive metabolite 13-cis-4-oxo-retinoic acid are more comparable (4- to 33-fold different) thanadministered dose (6- to 40-fold different). Moreover, placental transfer studies revealed significantlyless placental transfer of 13-cis-retinoic acid and 13-cis-4-oxo-retinoic acid in mice, rats, orrabbits (embryo to maternal AUC ratios range from 0.02 to 0.09) vs. monkeys (embryo to maternalconcentration AUC ratios of 0.41 and 0.33 for 13-cis-retinoic acid and 13-cis-4-oxo-retinoic acid,respectively). 105–108,110,111 Hence embryonic exposure levels at the teratogenic dose are much more© 2006 by Taylor & Francis Group, LLC

590 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 13.7Maternal plasma systemic exposure levels of valproic acidproducing approximately equal teratogenicity in mice 116Dosing Regimen Dose (mg/kg) C max (µg/ml) AUC (µg·h/ml)One injection a 356 445 652Four injections b 181 225 1369Infusion c 3809 248 5972aSingle subcutaneous injection on day 8 postcoitus (pc).bMultiple subcutaneous injections every 6 h on days 7–8.5 pc.cInfusion subcutaneously by implanted minipump on days 7.5–8.5 pc. C max= C ss .comparable (0.5 to 5-fold different) than administered dose. Indeed, given that 10 mg/kg is teratogenicin rabbits, 114 and assuming dose-proportional AUC values, the embryonic AUC values associatedwith teratogenicity are quite similar. Taken together, these toxicokinetic data provide arational basis for at least partly understanding species differences in teratogenic doses of 13-cisretinoicacid. These include humans, for which metabolism, half-life, and low teratogenic dosethreshold (0.5 to 1.5 mg/kg/day) are comparable with those of the monkey. 109,110To address unique sensitivity of neonatal animals (for example, unexpected pup deaths in apre- and postnatal study), it can be useful to determine if the pups received an unusually high doseof compound via milk. It might be assumed that measurement of drug concentration in milk canprovide an explanation for such events, but, as mentioned previously, drug concentration in milk(or milk to plasma concentration ratio) does not predict pup systemic exposure without taking intoconsideration the quantity of milk intake, oral bioavailability, and rate of drug clearance by theinfant. Assessment of the plasma drug level in pups can be more useful than assessment of excretionin milk for understanding marked toxicity to the neonate because such toxicity could have resultedin part from prenatal drug accumulation. Additionally, measurement of drug concentration in thepup’s blood in natural vs. cross-fostered litters can be used to determine the relative amount of apup’s drug body burden arising from prenatal versus lactational exposure.D. Use of Toxicokinetics in Risk AssessmentThe comparison of systemic exposure between test species and humans is an important elementof risk assessment. It is well established that interspecies extrapolation can be grossly inaccurateif made solely on the basis of administered dose (e.g., 13-cis-retinoic acid) because of significantspecies differences in the absorption, distribution, metabolism, and/or excretion of xenobiotics.Pharmacokinetic parameters can be used quantitatively in risk assessment, but this begs the questionas to which parameters are most influential for expression of toxicity. Experimentally, this can beexplored by measuring changes in the incidence or severity of toxicity in response to changes inthe mode of compound administration that dramatically alter systemic exposure. This approach isillustrated below by way of two examples.Valproic acid is an anticonvulsant drug for the treatment of certain seizure disorders. It producesneural tube defects in the mouse (exencephaly) but not in the rat, rabbit, or monkey. 115 Manipulationof dose and route was used by Nau 116 to investigate relationships among C max , AUC, and teratogenicpotential in the mouse. Analyses of equally teratogenic regimens suggested that peak drug level(C max ), but not administered dose or AUC (which varied by approximately 10-fold), was the mostpredictive dose metric, i.e., a threshold concentration must be exceeded (Table 13.7).The mouse C max associated with teratogenicity is not much beyond the range of plasma drugconcentration (up to 160 µg/ml) reported in the clinic following daily doses ranging up to 2400 mg, 117although serum protein binding of valproic acid is more extensive in humans than in the mouse© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 591(%fu = 8 and 65, respectively). 118 Nonetheless, prenatal exposure to valproic acid in humanpregnancy has been associated with an increased risk of spina bifida, 119–121 and the studies byNau 115,116 suggest that antiepileptic therapy that minimizes high peak concentrations of valproicacid could reduce the risk of teratogenicity.Another example is provided by studies with cyclophosphamide, a cytotoxic agent that producesa range of defects in all species tested. 122 Reiners and coworkers 123 studied developmental toxicityin mice administered cyclophosphamide subcutaneously (s.c.) by single injection vs. continuousinfusion by implanted minipumps. A relatively low maternal plasma cyclophosphamide concentration(0.09 µg/ml) maintained for 24 h was estimated to be comparably teratogenic to a singleinjection achieving a maternal plasma concentration of 13 µg/ml, suggesting that AUC (estimatedto be 2 to 3 µg·h/ml), rather than C max or administered dose, is the most predictive dose metric.Note that although a cyclophosphamide metabolite, phosphoramide mustard, may be the proximateteratogen, 124 it is expected to be present in the same proportion to the parent drug after injectionor infusion because the same route (s.c.) was used. The plasma half-life of cyclophosphamide inthe mouse is about 10 min 124 and in humans is about 9 h. 125 For a patient given a clinical dose (6.4mg/kg, i.v.), the observed C max was approximately 13 µg/ml, and, owing to the prolonged half-lifein humans, the cyclophosphamide AUC is greater than 400 µg·h/ml and that for phosphoramidemustard is about 9 µg·h/ml, 125 exposure levels significantly above those associated with developmentaltoxicity in mice.Whereas the teratogenic potential for some agents seemingly best correlates with C max (e.g.,valproic acid, caffeine) and for other agents it is AUC (e.g., cyclophosphamide, retinoids), 126,127 itis possible that different manifestations of developmental toxicity induced by the same agent canhave different predictive dose metrics. Following 2-ME exposure of mice on GD 8, there appearsto be a correlation between peak concentrations of 2-MAA and the occurrence of exencephaly, 128but on GD 11 there is a correlation between the AUC for 2-MAA and the incidence of digitmalformations. 129 The latter relationship holds true for the rat despite a very different plasmaelimination half-life for 2-MAA in rats (approximately 20 h) 130 compared with mice (approximately6 h). 129The litter incidence of dexamethasone-induced developmental toxicity (resorptions, death, cleftpalate) among rats given an identical maternal dose (0.8 mg/kg/day on GD 9 through 14) was notclosely correlated with either maternal plasma C max or AUC values. 131 Instead, the maternal plasmat ½α was positively correlated with the percentage of the litter affected, suggesting that elapsed timeduring which the dexamethasone concentration exceeded a threshold value was the key metric.Similar AUCs can be achieved with very different concentration vs. time profiles, for whichthe elapsed time during which a given concentration is exceeded can be markedly different (Figure13.5). Hence, an AUC value is a generic concentration-time product that per se does not preciselydefine a toxicity threshold in terms of magnitude or duration of a critical plasma concentration. Inthis context, a C max can be viewed as achieving a given concentration for a very brief period. Mostprecisely toxicokinetics should define both the critical concentration and time thresholds, whichare only approximated by C max or AUC.As the foregoing examples illustrate, a more accurate assessment of risk is made possible bycharacterization of maternal systemic exposure than by extrapolation of administered dose. Inprinciple, the risk of direct (as opposed to maternally mediated) developmental toxicity dependson the toxicants’ concentration in the embryo or fetus. However, such data are generally not availablefor humans, whereas dose-systemic exposure relationships for the mother can be obtained orextrapolated from similar data in men or nonpregnant women. Establishing correlations betweenmaternal systemic exposure and the incidence of animal teratogenicity is further substantiated whenembryonic and maternal concentrations have a similar profile, as has been shown for 2-MAA inrats following maternal 2-ME treatment. 132© 2006 by Taylor & Francis Group, LLC

592 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONCAC 1C 2B15 30 60Time (minutes)Figure 13.5Theoretical blood plasma compound concentration (C) vs. time curves following a single bolusintravenous injection (A) or continuous intravenous infusion (B). The AUCs are comparable,whereas the elapsed time exceeding a certain concentration (C 1 or C 2 ) is different. The higherof these concentrations (C 1 ) is only exceeded following the bolus dose, while the duration of timeduring which C 2 is exceeded is greater following infusion.VIII. CONCLUSIONSThe rate and extent of absorption, distribution, biotransformation, and elimination influence thetoxicity potential of xenobiotics. Pharmacokinetic parameters reflecting these events are used indevelopmental and reproductive toxicology study design, in species selection, and in the rationalefor route and dose selection. Although general assessments of systemic exposure and metabolismcan be collected in a relatively standardized manner, toxicokinetic assessments are ideally madein the context of having first identified the nature of the toxicity, critical “windows” of exposure,and sensitive vs. nonresponsive species. This is because toxicokinetics is not merely the assessmentof pharmacokinetics in the toxicology laboratory but is the determination of the relationshipsbetween the compound’s absorption, distribution, biotransformation, and elimination, and theexpression of toxicity. These relationships, based on classical pharmacokinetic parameters andphysiologically based models, can be used to explain experimental results and to extrapolate amongdifferent exposure conditions or between species. In that regard, toxicokinetics plays an integralrole in the risk assessment of developmental and reproductive toxicity.REFERENCES1. Crawford, J.S., The placenta, drugs, and the fetus, in Principles and Practice of Obstetric Anesthesia,5th ed., Blackwell Scientific Publications, Oxford, 1984, p. 132.2. Morriss, Jr., F.H. and Boyd, R.D.H., Placental Transport, in The Physiology of Reproduction, Knobil,E. and Neill, J., Eds., Raven Press Ltd., New York, 1988, chap. 50.3. Rozman, K.K. and Klaassen, C.D., Absorption, distribution, and excretion of toxicants, in Casarettand Doull’s Toxicology: The Basic Science of Poisons, 6th ed., Klaassen, C.D., Ed., McGraw-Hill,New York, 2001, chap. 5.4. Landas, G.R., et al., Placental p-glycoprotein deficiency enhances susceptibility to chemically inducedbirth defects in mice, Reprod. Toxicol., 12, 457, 1998.5. Goldstein, A., Aronow, L., and Kalman, S.M., Principles of Drug Action: The Basics of Pharmacology,2nd ed., John Wiley & Sons, New York, 1974, chap 2.6. Goldstein, A., Aronow, L., and Kalman, S.M., Principles of Drug Action: The Basics of Pharmacology,2nd ed., John Wiley & Sons, New York, 1974, p. 821.© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 5937. Dym, M. and Fawcett, D.W., The blood-testis barrier in the rat and the physiological compartmentationof the seminiferous epithelium, Biol. Reprod., 3, 300, 1970.8. Setchell, B.P., Vogimayr, J.K., and Waites, G.M.H., A blood-testis barrier restricting passage fromblood lymph into rete testis fluid but not into lymph, J. Physiol., 200, 73, 1969.9. Okumura, K., Lee, I.P., and Dixon, R.L., Permeability of selected drugs and chemicals across theblood-testis barrier of the rat, J. Pharmacol. Exp. Ther., 194, 89, 1975.10. Hendrikse, N.H., et al., Complete in vivo reversal of p-glycoprotein pump function in the blood-brainbarrier visualized with positron emission tomography, Br. J. Pharmacol., 124, 1413, 1998.11. Hughes, C.S., et al., Modulation of doxorubicin concentration by cyclosporin A in brain and testicularbarrier tissues expressing p-glycoprotein in rats, J. Neuro-Oncology, 37, 45, 1998.12. Choo, E.F., et al., Pharmacological inhibition of p-glycoprotein transport enhances the distribution ofHIV-1 protease inhibitors into brain and testes, Drug Metabol. Disposition, 28, 655, 2000.13. Sieber, S.M. and Fabro, S., Identification of drugs in the preimplantation blastocyst and in the plasma,uterine secretion and urine of the pregnant rabbit, J. Pharmacol. Exp. Ther., 176, 65, 1971.14. McLachlan, J.A., et al., Accumulation of nicotine in the uterine fluid of the six-day pregnant rabbit,Fertil. Steril., 27, 1204, 1967.15. Beck, F., Comparative placental morphology and function, in Developmental Toxicology, Kimmel,C.A. and Buelke-Sam, J., Eds., Raven Press, New York, 1981, p. 35.16. Amoroso, E.C., Placentation, in Marshall’s Physiology of Reproduction, 3rd ed., Parkes, A.S., Ed.,Longmans, London, 1952, Vol. II, chap. 15.17. Mirkin, B.L., Perinatal pharmacology: placental transfer, fetal localization, and neonatal dispositionof drugs, Anesthesiology, 43, 156, 1975.18. Williams, R.T., Detoxification Mechanisms, 2nd ed., Wiley, New York, 1971.19. Parkinson, A., Biotransformation of xenobiotics, in Casarett and Doull’s Toxicology: The BasicScience of Poisons, 6th Ed., Klaassen, C.D., Ed., McGraw-Hill, New York, 2001, chap. 6.20. Juchau, M.R., Mechanisms of drug biotransformation reactions in the placenta, Fed. Proc., 31, 48,1972.21. Juchau, M.R., Drug biotransformation in the placenta, Pharmacol. Ther., 8, 501, 1980.22. Heinrichs, W.L. and Juchau, M.R., Extrahepatic drug metabolism: the gonads, in Extrahepatic Metabolismof Drugs and Other Foreign Compounds, Gram, T.E., Ed., SP Medical and Scientific Books,New York, 1980, p. 313.23. Ishimura, K., et al., Ultrastructural and immunohistochemical studies on steroid-secreting cells oftestis and ovary of normal and 3-methylcholanthrene-treated mice, Cell & Tissue Research, 245, 681,1986.24. Batt, A.M., et al., Enzyme induction by drugs and toxins, Clinica Chimica Acta, 209, 109, 1992.25. Krishna, D.R. and Klotz, U., Extrahepatic metabolism of drugs in humans, Clin. Pharmacokinet., 26,144, 1994.26. Slikker, W. and Miller, R.K., Placental metabolism and transfer: role in developmental toxicology, inDevelopmental Toxicology, 2nd ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York,1994, p. 245.27. Ring, J.A., et al., Fetal hepatic drug elimination, Pharmacol. Ther., 84, 429, 1999.28. Thomas, J.A., Curto, K.A., and Thomas, M.J., MEHP/DEHP gonadal toxicity and effects on rodentaccessory sex organs, Environ. Health Perspect., 45, 85, 1982.29. Foster, P.M.D., et al., Testicular toxicity of 2-methoxyacetaldehyde, a possible metabolite of ethyleneglycol monomethyl ether in the rat, Toxicol. Lett., 32, 73, 1986.30. Boekelheide, K., 2,5-Hexanedione alters microtubule assembly: I. Testicular atrophy, not nervoussystem toxicity, correlates with enhanced tubulin polymerization, Toxicol. Appl. Pharmacol., 88, 370,1987.31. Gerber, N. and Lynn, R.K., Excretion of methadone in semen from methadone addicts; comparisonwith blood levels, Life Sci., 19, 787, 1976.32. Swanson, B.N., et al., Seminal excretion, vaginal absorption, distribution and whole blood kineticsof d-methadone in the rabbit, J. Pharmacol. Exp. Ther., 206, 507, 1978.33. Taylor, S.M., et al., The seminal excretion, plasma elimination, tissue distribution and metabolism ofnaltrexone in the rabbit, J. Pharmacol. Exp. Ther., 213, 289, 1980.© 2006 by Taylor & Francis Group, LLC

594 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION34. Beach, C.A., Bianchine, J.R., and Gerber, N., The excretion of caffeine in the semen of men:pharmacokinetics and comparison of the concentrations in blood and semen, J. Clin. Pharmacol., 24,120, 1984.35. Beach, C.A., et al., Metabolism, distribution, seminal excretion and pharmacokinetics of caffeine inthe rabbit, J. Pharmacol. Exp. Ther., 233, 18, 1985.36. Kershaw, R.A., et al., Excretion of acetaminophen in semen of men: comparison with concentrationin plasma, J. Am. Coll. Toxicol., 1, 113, 1982.37. Kershaw, R.A., et al., Disposition of aspirin and its metabolites in the semen of man, J. Clin.Pharmacol., 27, 304, 1987.38. Swanson, B.N., et al., Excretion of valproic acid into semen of rabbits and man, Epilepsia, 19, 541,1978.39. Lundquist, F., Aspects of biochemistry of human semen, Acta Physiol. Scand., 19, 66, 1949.40. White, M.A., Changes in pH of expressed prostatic secretion during the course of prostatitis, Proc.R. Soc. Med., 68, 511, 1975.41. Winningham, D.G., Nemoy, N.J., and Stamey, T.A., Diffusion of antibiotics from plasma into prostaticfluid, Nature, 219, 139, 1968.42. Winningham, D.G. and Stamey, T.A., Diffusion of sulfonamides from plasma into prostatic fluid, J.Urol., 104, 559, 1970.43. Paladine, W.J., et al., Possible sensitivity to vinblastine in prostatic fluid. N. Engl. J. Med., 292, 52, 1975.44. Lutwak-Mann, C., Schmid, K., and Keberle, H., Observations of the progeny of thalidomide treatedmale rabbits, Br. Med. J., 1, 1090, 1964.45. Anderson, P.O., Drug use during breast feeding, Clin. Pharm., 10, 594, 1991.46. Oo, C.Y., et al., Active transport of cimetidine into human milk, Clin. Pharmacol. Ther., 58, 548, 1995.47. Ito S. and Koren G., A novel index for expressing exposure of the infant to drug in breast milk, Br.J. Clin. Pharmacol., 38, 99, 1994.48. Atkinson, H.C., Begg, E.J., and Darlow, B.A., Drugs in human milk: clinical pharmacokinetic considerations,Clin. Pharmacokinet., 14, 217, 1988.49. Ito, S., Drug therapy for breast-feeding women, New Engl. J. Med., 343, 118, 2000.50. Matsuda, S., Transfer of antibiotics into maternal milk, Biol. Res. Pregnancy, 5, 57, 1984.51. Mattison, D.R., Physiological variations in pharmacokinetics during pregnancy, in Drug and ChemicalAction in Pregnancy, Fabro, S. and Scialli, A.R., Eds., Dekker, New York, 1986, p. 37.52. Mattison, D.R., Blann, E., and Malek, A., Physiological alterations during pregnancy: impact ontoxicokinetics, Fund. Appl. Toxicol., 16, 215, 1991.53. Krauer, B., Physiological changes and drug disposition during pregnancy, in Pharmacokinetics inTeratogenesis, Vol. 1, Nau, H. and Scott, W.J., Eds., CRC Press, Boca Raton, 1987, chap. 1.54. Hytten, F.E. and Thomson, A.M., Maternal physiological adjustments, in Biology of Gestation, Vol.1, The Maternal Organism, Assali, N.S., Ed., Academic Press, New York, 1968, chap. 8.55. Johnson, R.L., et al., Cardiac output distribution and uteroplacental blood flow in the pregnant rabbit:a comparative study, Am. J. Obstet. Gynecol., 151, 682, 1985.56. Koldovský, O., Digestion and absorption, in Perinatal Physiology, Stave, U., Ed., Plenum MedicalBook Company, New York, 1978, chap. 15.57. Calabrese, E.J., Age and Susceptibility to Toxic Substances, John Wiley & Sons, New York, 1986.58. Znamenacek, K. and Pibylová, H., The effect of glucose and insulin administration on the bloodglucose levels of the newborn, Cesk. Pediatr. (Prague), 18, 104, 1963.59 Angus, G.E. and Thurlbeck, W.M., Number of alveoli in the human lung, J. Appl. Physiol., 32, 483,1972.60. Besunder, J.B., Reed, M.D., and Blumer, J.L., Principles of drug biodisposition in the neonate: acritical evaluation of the pharmacokinetic-pharmacodynamic interface (part I), Clin. Pharmacokinet.,14, 189, 1988.61. Kurz, H., Michels, H., and Stickel, H.H., Differences in the binding of drugs to plasma proteins fromnewborn and adult man (part II), Eur. J. Clin. Pharmacol., 11, 469, 1977.62. Himwich, W.A., Physiology and pharmacology of the central nervous system, in Perinatal Physiology,Stave, U., Ed., Plenum Medical Book Company, New York, 1978, chap. 29.63. Neims, A.H., et al., Developmental aspects of the hepatic cytochrome P-450 monooxygenase system,Ann. Rev. Pharmacol. Toxicol., 16, 427, 1976.© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 59564. Klinger, W., Biotransformation of drugs and other xenobiotics during postnatal development, Pharmacol.Ther., 16, 377, 1982.65. Juchau, M.R., Fetal and neonatal drug biotransformation, in Drug Toxicity and Metabolism in Pediatrics,Kacew, S., Ed., CRC Press, Boca Raton, FL, 1990, p.15.66. Klaassen, C.D., Immaturity of the newborn rat’s hepatic excretory function for ouabain, J. Pharmacol.Exp. Ther., 183, 520, 1972.67. Ballatori, N. and Clarkson, T.W., Developmental changes in the biliary excretion of methylmercuryand glutathione, Science, 216, 61, 1982.68. Fetterman, G.H., et al., The growth and maturation of human glomeruli and proximal convolutionsfrom term to adulthood, J. Pediat., 35, 601, 1965.69. Gomez, R.A., et al., The maturing kidney: development and susceptibility, Renal Failure, 21, 283,1999.70. Morselli, P.L. and Provei, V., Placental transfer of pethidine and norpethidine and their pharmacokineticsin the newborn, Eur. J. Clin. Pharmacol., 18, 25, 1980.71. Rajchgot, P. and MacLeod, S., Perinatal pharmacology, in Developmental Pharmacology, MacLeod,S., Okey, A.B., and Spielberg, S.P., Eds., Alan R. Liss, New York, 1983, p. 14.72. Kauffman, R.E., The role of chloramphenicol in pediatric therapy, in Developmental Pharmacology,MacLeod, S., Okey, A.B., and Spielberg, S.P., Eds., Alan R. Liss, New York, 1983, p. 315.73. Goldstein, A., Aronow, L., and Kalman, S.M., Principles of Drug Action: The Basics of Pharmacology,2nd Ed., John Wiley & Sons, New York, 1974, pp. 291.74. Ambrose, P.J., Clinical pharmacokinetics of chloramphenicol and chloramphenicol succinate, Clin.Pharmacokinet., 9, 222, 1984.75. Aranda, J.V., et al., Pharmacokinetic aspects of theophylline in premature newborns, New Engl. J.Med., 295, 413, 1976.76. Boutroy, M.J., et al., Caffeine, a metabolite of theophylline during the treatment of apnea in thepremature infant, J. Pediat., 94, 996, 1979.77. Aranda, J.V., Turmen, T., and Sasyniuk, B.I., Pharmacokinetics of diuretics and methylxanthines inthe neonate, Eur. J. Clin. Pharmacol., 18, 55, 1980.78. Busto, U., Bendayan, R., and Sellers, E.M., Clinical pharmacokinetics of non-opiate abused drugs,Clin. Pharmacokinet., 16, 1, 1989.79. Wagner, J.G., Do you need a pharmacokinetic model and, if so, which one? J. Pharmacokinet.Biopharm., 3, 457, 1975.80. Rowland, M. and Tozer, T.N., Clinical Pharmacokinetics: Concepts and Applications, Lea and Febiger,Philadelphia, 1980.81. Gibaldi, M. and Perrier, D., Pharmacokinetics, 2nd ed., Marcel Dekker, New York, 1982.82. Himmelstein, K.J. and Lutz, R.J., A review of the applications of physiologically based pharmacokineticmodeling, J. Pharmacokinet. Biopharm., 7, 127, 1979.83. Gerlowski, L.E. and Jain, R.K., Physiologically based pharmacokinetic modeling: principles andapplications, J. Pharm. Sci., 72, 1103, 1983.84. Conolly, R.B. and Andersen, M.E., Biologically based pharmacodynamic models: tools for toxicologicalresearch and risk assessment, Annu. Rev. Pharmacol. Toxicol., 31, 503, 1991.85. Krishnan, K. and Andersen, M.E., Physiologically based pharmacokinetic modeling in toxicology, inPrinciples and Methods of Toxicology, 4th ed., Hayes, A.W., Ed., Taylor and Francis, Philadelphia,2001, chap. 5.86. Olanoff, L.S. and Anderson, J.M., Controlled release of tetracycline – III: A physiological pharmacokineticmodel of the pregnant rat, J. Pharmacokinet. Biopharm., 8, 599, 1980.87. Gabrielsson, J.L. and Paalzow, L.K., A physiological pharmacokinetic model for morphine dispositionin the pregnant rat, J. Pharmacokinet. Biopharm., 11, 147, 1983.88. Gabrielsson, J.L., Paalzow, L.K., and Nordstrom, L., A physiologically based pharmacokinetic modelfor theophylline disposition in the pregnant and nonpregnant rat, J. Pharmacokinet. Biopharm., 12,149, 1984.89. Gabrielsson, J.L., et al., Analysis of methadone disposition in the pregnant rat by means of a physiologicalflow model, J. Pharmacokinet. Biopharm., 13, 355, 1985.90. Gabrielsson, J.L., et al., Analysis of pethidine disposition in the pregnant rat by means of a physiologicalflow model, J. Pharmacokinet. Biopharm., 14, 381, 1986.© 2006 by Taylor & Francis Group, LLC

596 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION91. Fisher, J.W., et al., Physiologically based pharmacokinetic modeling of the pregnant rat: a multirouteexposure model for trichloroethylene and its metabolite, trichloroacetic acid, Toxicol. Appl. Pharmacol.,99, 395, 1989.92. O’Flaherty, E.J., et al., A physiologically based kinetic model of rat and mouse gestation: dispositionof a weak acid, Toxicol. Appl. Pharmacol., 112, 245, 1992.93. Clarke, D.O., et al., Pharmacokinetics of 2-methoxyethanol and 2-methoxyacetic acid in the pregnantmouse: a physiologically based mathematical model, Toxicol. Appl. Pharmacol., 121, 239, 1993.94. Terry, K.K., et al., Development of a physiologically based pharmacokinetic model describing 2-methoxyacetic acid disposition in the pregnant mouse, Toxicol. Appl. Pharmacol., 132, 103, 1995.95. Hays, S.M., et al., Development of a physiologically based pharmacokinetic model of 2-methoxyethanoland 2-methoxyacetic acid disposition in pregnant rats, Toxicol. Appl. Pharmacol., 163, 67, 2000.96. Gargas, M.L., et al., A toxicokinetic study of inhaled ethylene glycol monomethyl ether (2-ME) andvalidation of a physiologically based pharmacokinetic model for the pregnant rat and human, Toxicol.Appl. Pharmacol., 165, 53, 2000.97. Gabrielsson, J.L. and Larsson, K.S., Proposals for improving risk assessment in reproductive toxicology,Pharmacol. Toxicol., 66, 10, 1990.98. Welsch, F., Blumenthal, G.M., and Conolly, R.B., Physiologically based pharmacokinetic modelsapplicable to organogenesis: extrapolation between species and potential use in prenatal toxicity riskassessments, Toxicol. Letters, 82–83, 539, 1995.99. Luecke, R.H., et al., A physiologically based pharmacokinetic computer model for human pregnancy,Teratology, 49, 90, 1994.100. Luecke, R.H., et al., A computer model and program for xenobiotic disposition during pregnancy,Computer Methods Programs Biomed., 53, 201, 1997.101. Diehl, K.-H., et al., A good practice guide to the administration of substances and removal of blood,including routes and volumes, J. Appl. Toxicol., 21, 15, 2001.102. Smith Kline and French Laboratories, Philadelphia, Small Animal Milking Apparatus, U.S. Patent3,580,220.103. ICH Guideline S5A: Reproductive Toxicology: Detection of Toxicity to Reproduction for MedicinalProducts (CPMP/ICH/386/95).104. ICH Guideline S3A: Toxicokinetics: The Assessment of Systemic Exposure in Toxicity Studies(CPMP/ICH/384/95).105. Kochar, D.M., Kraft, J., and Nau, H., Teratogenicity and disposition of various retinoids in vivo andin vitro, in Pharmacokinetics in Teratogenesis, Vol. 2, Nau, H. and Scott, W.J., Eds., CRC Press, BocaRaton, 1987, chap. 16.106. Agnish, N.D., Rusin, G., and DiNardo, B., Taurine failed to protect against the embryotoxic effectsof isotretinoin in the rat, Fund. Appl. Toxicol., 15, 249, 1990.107. Collins, M.D., et al., Comparative teratology and transplacental pharmacokinetics of all-trans-retinoicacid, 13-cis-retinoic acid, and retinyl palmitate following daily administration in rats, Toxicol. Appl.Pharmacol., 127, 132, 1994.108. Eckhoff, C., et al., Teratogenicity and transplacental pharmacokinetics of 13-cis-retinoic acid inrabbits, Toxicol. Appl. Pharmacol., 125, 34, 1994.109. Hummler, H., Korte, R., and Hendrickx, A.G., Induction of malformations in the cynomolgus monkeywith 13-cis retinoic acid, Teratology, 42, 263, 1990.110. Hummler, H., Hendrickx, A.G., and Nau, H., Maternal toxicokinetics, metabolism, and embryoexposure following a teratogenic dosing regimen with 13-cis-retinoic acid (Isotretinoin) in the cynomolgusmonkey, Teratology, 50, 184, 1994.111. Creech Kraft, J., et al., Teratogenicity and placental transfer of all-trans-13-cis, 4-oxo-all-trans- and4-oxo-13-cis-retinoic acid after administration of a low oral dose during organogenesis in mice,Toxicol. Appl. Pharmacol., 100, 162, 1989.112. Kalin, J.R., Wells, J., and Hill, D.L., Disposition of 13-cis-retinoic acid and N-(2-hydroxyethyl)retinamide in mice after oral doses, Drug Metab. Dispos., 10, 391, 1982.113. Shelly, R.S., et al., Blood level studies of all-trans- and 13-cis-retinoic acid in rats using differentformulations, J. Pharm. Sci., 71, 904, 1982.114. Kamm, J.J., Toxicology, carcinogenicity, and teratogenicity of some orally administered retinoids, J.Am. Acad. Dermatol., 6, 652, 1982.© 2006 by Taylor & Francis Group, LLC

USE OF TOXICOKINETICS IN DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGY 597115. Nau, H., Species differences in pharmacokinetics, drug metabolism, and teratogenesis, in Pharmacokineticsin Teratogenesis, Vol. 1, Nau, H. and Scott, W.J., Eds., CRC Press, Boca Raton, 1987, chap. 6.116. Nau, H., Teratogenic valproic acid concentrations: infusion by implanted minipumps vs. conventionalinjection regimen in the mouse, Toxicol. Appl. Pharmacol., 80, 243, 1985.117. Meijer, J.W.A. and Hessing-Brand, L., Determination of lower fatty acid, particularly the antiepilepticdipropylacetic acid, in biological material by means of microdiffusion and gas chromatography,Clinica Chimica Acta, 43, 215, 1973.118. Nau, H. and Löscher, W., Valproic acid and metabolites: pharmacological and toxicological studies,Epilepsia, 25 (Suppl 1), S14, 1984.119. Robert, E. and Guibaud, P., Maternal valproic acid and congenital neural tube defects, Lancet, 2, 937,1982.120. Robert, E. and Rosa, F., Valproate and birth defects, Lancet, 2, 1142, 1983.121. Lindhout, D. and Meinardi, H., Spina bifida and in-utero exposure to valproate, Lancet, 2, 396, 1984.122. Mirkes, P.E., Cyclophosphamide teratogenesis: a review, Teratogenesis, Carcinogenesis, Mutagenesis,5, 75, 1985.123. Reiners, J., et al., Teratogenesis and pharmacokinetics of cyclophosphamide after drug infusion ascompared to injection in the mouse during day 10 of gestation, in Pharmacokinetics in Teratogenesis,Vol. 2, Nau, H. and Scott, W.J., Eds., CRC Press, Boca Raton, 1987, chap. 4.124. Nau, H., et al., Mutagenic, teratogenic and pharmacokinetic properties of cyclophosphamide and someof its deuterated derivatives, Mutation Res., 95, 105, 1982.125. Juma, F.D., Rogers, H.J., and Trounce, J.R., The pharmacokinetics of cyclophosphamide, phosphoramidemustard and nor-nitrogen mustard studied by gas chromatography in patients receiving cyclophosphamidetherapy, Br. J. Clin. Pharmacol., 10, 327, 1980.126. Nau, H., Species differences in pharmacokinetics and drug teratogenesis, Environ. Health Perspect.,70, 113, 1986.127. Nau, H., Pharmacokinetic considerations in the design and interpretation of developmental toxicitystudies, Fund. Appl. Toxicol., 16, 219, 1991.128. Terry, K.K., et al., Developmental phase alters dosimetry-teratogenicity relationship for 2-methoxyethanolin CD-1 mice, Teratology, 49, 218, 1994.129. Clarke, D.O., Duignan, J.M., and Welsch, F., 2-Methoxyacetic acid dosimetry-teratogenicity relationshipsin CD-1 mice exposed to 2-methoxyethanol, Toxicol. Appl. Pharmacol., 114, 77, 1992.130. Sleet, R.B., et al., Developmental phase specificity and dose-response effects of 2-methoxyethanolteratogenicity in rats, Fund. Appl. Toxicol., 29, 131, 1996.131. Hansen, D.K., et al., Pharmacokinetic considerations of dexamethasone-induced developmental toxicityin rats, Toxicological Sciences, 48, 230, 1999.132. Hays, S.M., et al., Development of a physiologically based pharmacokinetic model of 2-methoxyethanoland 2-methoxyacetic acid disposition in pregnant rats, Toxicol. Appl. Pharmacol., 163, 67, 2000.© 2006 by Taylor & Francis Group, LLC

CHAPTER 14Cellular, Biochemical, and Molecular Techniquesin Developmental Toxicology*Barbara D. Abbott, Mitchell B. Rosen, and Gary HeldCONTENTSI. Introduction ........................................................................................................................600A. Overview and Introduction........................................................................................600B. General Approach to the Study of Mechanisms of Developmental Toxicity ..........600II. Descriptive Approaches: Evaluating Protein and mRNA Expression Patterns ................601A. Immunohistochemistry ..............................................................................................602B. In Situ Hybridization .................................................................................................604III. Approaches for Quantifying mRNA Expression...............................................................607A. Semiquantitative Methods .........................................................................................6071. Northern Blot.......................................................................................................6072. Dot Blot or Slot Blot...........................................................................................608B. PCR and RT-PCR Quantitative Methods ..................................................................6091. Real-Time RT-PCR..............................................................................................6092. Quantitative RT-PCR Based on Image Analysis of PCR Products inElectrophoretic Gels ............................................................................................611C. Solution Hybridization: RNase Protection Assay.....................................................611D. RNA Transcription In Vitro: Nuclear Run-on Assay ................................................612IV. Characterizing the Role of Specific Genes and Pathways of Response...........................613A. Transgenic Animals ...................................................................................................613B. In Vitro Antisense RNA.............................................................................................615C. DNA Array.................................................................................................................615D. Genetic Library Screening.........................................................................................615E. Reverse Southern Blots and RT-PCR In Situ Transcription .....................................616References ......................................................................................................................................617* Disclaimer: This chapter has been reviewed by NHEERL and by U.S. Environmental Protection Agency, and has beenapproved for publication. Mention of trade names of commercial products does not constitute endorsement or recommendationfor use.599© 2006 by Taylor & Francis Group, LLC

600 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. Overview and IntroductionI. INTRODUCTIONCellular, molecular, and biochemical approaches vastly expand the possibilities for revealing theunderlying mechanisms of developmental toxicity. The increasing interest in embryonic developmentas a model system for the study of gene expression has resulted in a cornucopia of informationand approaches available to the toxicologist. The typical teratology screening test examines neartermfetuses after exposure throughout organogenesis and evaluates the potential for an exposureto disrupt development. However, this assay can also provide insight regarding the organs and/orcellular processes that might be targeted. More specific information can be derived from doseresponse studies, where exposure is restricted to one or only a few days during development.Examination of the target organs or cells at critical stages of morphogenesis, particularly at earlytime points after exposure, is important for identifying specific genes or proteins that may beinvolved in the response. This chapter presents an overview of some of these cellular, biochemical,and molecular biological methods and discuss their applicability to developmental toxicology,providing examples of applications of each method.The molecular biological techniques presented in this chapter may appear to be routine andstraightforward. However, most or all of these methods require a considerable investment of effort,time, patience, and laboratory resources to allow their introduction into a research program. Thiscautionary note is intended only to alert the investigator that these are not trivial undertakings.Even for experienced investigators, successful application of some of these techniques will notalways be straightforward but may require intuitive adjustment of specific protocol conditions tosuit the particular experimental situation. An investigator who has no previous experience withcellular and molecular biological methods should consider enrollment in a class that provides basicinformation regarding the principles and protocols. If possible, the best option is to learn a newtechnology in the laboratory of a mentor or colleague who is currently applying the method withsuccess. It is important to gain an understanding of the principles and pitfalls of each assay, andthere are a number of excellent manuals available, some general and others specifically addressinga single technique. Most of the major suppliers of molecular biology reagents and equipment havedetailed laboratory manuals and periodic newsletters available on request. New investigators shouldsearch the literature and compile and compare variations of each technique, particularly with regardto applications of the assay under conditions similar to those for their intended study. Contactingmore experienced researchers directly can also provide invaluable assistance, as most are verywilling to share their methods and expertise.Although it is beyond the scope of this chapter to provide detailed methods and protocols,several sources of such information will be cited. There are a great many excellent books andpublications available, including general molecular cell biology texts 1 and manuals for molecularbiological methods. 2–4 Because each technique is discussed in this chapter, examples of applicationsin developmental biology and toxicology are briefly described. These represent only a few of themany excellent studies that have been published, and readers are encouraged to seek out otherexamples from the literature that are relevant to their particular special interests.B. General Approach to the Study of Mechanisms of Developmental ToxicityThe developmental toxicity of a particular chemical or class of chemicals is generally characterizedinitially in a standard screening protocol that involves exposure throughout organogenesis andevaluation late in gestation. This protocol will identify particular patterns of developmental toxicity,malformations, or other effects of interest. To identify the mechanisms leading to these effects, itis generally important to narrow the exposure to a specific developmental period (one of only afew days of exposure) and to determine the dose-response for the effect within the narrow exposure© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 601period. Identifying the target tissue and the most sensitive developmental stage allows someinferences to be made regarding cellular processes or regulatory pathways that could be the targetof the toxicant. The etiology of these effects becomes the subject of detailed study at early timepoints following exposure, as well as throughout subsequent development and differentiation.In general, toxicants may disrupt development through maternally mediated effects, directeffects on the conceptus (placenta, extraembryonic membranes, or the embryo), or a combinationof effects on mother and conceptus. The mechanisms may involve disruption of gene expression,signal transduction, protein function, enzyme activity, membrane integrity, ionic balance, or othercellular activities critical to development. The cellular responses to these insults can include changesin cell cycle regulation, cell differentiation, cell-cell and cell-extracellular matrix interactions, cellmigration, cell survival, or cell death. Depending on the toxicant and the level and timing ofexposure, these biochemical and cellular alterations have the potential to affect pattern formation,organogenesis, and tissue and organ functions in later gestation and postnatally, and may result inprenatal lethality.An overview of the potential mechanisms of developmental toxicity and a discussion of 18signaling pathways that are considered critical to normal embryonic development was recentlypublished by the National Academy of Sciences. 5 Study of these complex, interactive signaltransduction pathways offers an approach for revealing the mechanisms involved in abnormaldevelopment and response to toxicants. The effects on gene transcription and protein synthesiscould involve altered spatial and temporal patterns of expression, changed rates of synthesis ordegradation, altered posttranscriptional and translational modifications, or altered activity andfunction in the cell. The methods described in this chapter can be applied to localize and/orquantitate the expression of specific mRNAs and proteins in these critical regulatory pathways andto search for novel protein and mRNA partners in the response pathways.The assays used to study embryos and/or small specimens must be highly sensitive to detectsubtle changes from a limited number of cells. It is also important to have methods that can evaluategene or protein expression in situ; because the specimens are histologically complex, it may benecessary to identify specific cells or regions that are affected. This chapter will discuss descriptiveapproaches to evaluate patterns and levels of protein and mRNA expression in situ, quantitativeand semiquantitative methods to evaluate mRNA transcription, and methods to characterize therole of specific genes and pathways or to identify novel genes that may be involved. Approachesdescribed in the chapter include evaluating the patterns of protein expression with immunohistochemistryin sectioned tissues or in whole mount preparations and localizing mRNA through insitu hybridization in sections or whole mounts. Additional approaches to determine effects onmRNA expression include Northern blot and dot blot, RNase (S1) nuclease protection assay, reversetranscriptase polymerase chain reaction (RT-PCR) amplification, and nuclear run-on assays fordetection of newly transcribed RNA. The role of specific genes in mediating a developmentalresponse can be studied using transgenic mice that have constitutive expression of a gene of interest,that “report” activation of transcription of a gene, or that fail to express the gene of interest(“knockout mice”). Such mice can provide insight into the function or role of a gene in themechanism of toxicity, as well as demonstrating redundancy of function within the embryo (if noeffects of gene knockout are detectable). Effects of reducing or eliminating mRNA translation canalso be studied in vitro by use of antisense RNA hybridization in the whole embryo, the organ, orcell culture.II. DESCRIPTIVE APPROACHES: EVALUATING PROTEINAND MRNA EXPRESSION PATTERNSMethods of particular value in mechanistic studies of developmental toxicology are those thatlocalize the effect within the embryo and have the resolution to detect effects on a particular subset© 2006 by Taylor & Francis Group, LLC

602 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof cells within the developing tissue. Two such techniques that are frequently applied in developmentalbiology and are valuable in mechanistic developmental toxicology are immunohistochemicallocalization of proteins and in situ hybridization to detect specific mRNAs. Both approaches localizethe target molecules within the embryo to specific regions, and some have the resolution to detectexpression in specific cell types. There are considerable advantages to applying these methods intandem to describe effects on both gene and protein expression to evaluate responses to a toxicant.A. ImmunohistochemistryFor immunohistochemical methods to be used, an antibody that specifically recognizes the proteinof interest must be available. There are numerous commercial antibody suppliers, and antisera arenow available to study peptides involved in signal transduction, proliferation, growth factors,cytokines, and cytoskeletal and membrane peptides, hormones, and receptors. A search of theInternet can reveal useful sources of information (e.g., http://www.antibodyresource.com orhttp://www.abcam.com) for contract antibody suppliers, books, educational resources about antibodies,online databanks and databases, online journals, and immunology and biotechnology sites.A useful reference for finding antibodies is Linscott’s Directory, which contains a list of more than17,000 monoclonal antibodies and the companies that sell them. Contacting investigators who havepreviously worked with the peptide of interest may also provide leads in the search for antisera.Preparation of antibody in your own facility or contracting for preparation by an outside companymay also be a possibility if resources and/or experienced personnel are available. Protocols forimmunohistochemical analysis are relatively simple and are available in handbooks 6 and fromcompanies providing antisera for these applications.Selection of fixative and frozen or paraffin embedded sections will depend to some extent onthe properties of a particular antibody. Some will give good results on both Western blots andsectioned tissues, but not all antisera are compatible with both protocols. The detection methodsare also diverse, and choice of method will depend on the desired application. Fluorescent labels(rhodamine 123 or fluorescein are commonly used) are sensitive and give outstanding color photographs,but quantitation is not generally an option because of the fading of the fluorescent signalthat occurs during illumination. Enzymatic detection systems are also commonly used and includehorseradish peroxidase or alkaline phosphatase to generate colored products. Some but not all ofthe substrates produce precipitates that are stable in the alcohols and solvents used to preparepermanently coverslipped slides. The enzyme may also be incorporated as part of a complex inwhich a biotinylated secondary antibody complexes with avidin-biotin-peroxidase (ABC) or phosphatase.This method is useful for amplification of low levels of bound primary antibody.Immunostaining of intact tissues or small embryos provides a three-dimensional view of proteinexpression and can be very helpful in evaluating changes over time during development or inresponse to treatment. The size of the tissue or embryo that can be successfully stained in thiswhole mount approach is limited, and generally only small specimens have been used (e.g.,presomite mouse embryos and stages just after neural tube closure, as well as some dissectedtissues, e.g., limb bud or palate). 7 The method typically involves fixation of the specimen followedby permeabilization with either a detergent, enzymatic digestion (with proteinase), or a salt solution.Incubation with hydrogen peroxide serves to inhibit endogenous peroxide and clears (makes moretransparent) the tissues. The incubation with primary antibody can require a prolonged period forpenetration (perhaps 4 to 5 days). The primary antibody is localized with secondary antibodycomplexed with peroxidase enzyme that then reacts with a substrate (e.g., 3,3′-diaminobenzidine[DAB]) to give a colored product or with a substrate such as 4-chloro-1-naphthol followed byethanolic incubations to give a fluorescent product. The immunostained whole mount specimensare likely to degrade over time and the color of the label can change, so it is important to obtainimages of the outcomes as soon as possible.© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 603Figure 14.1Vascular endothelial cell growth factor (VEGF) expression was detected immunohistochemicallyin trophoblastic giant cells of the GD 9.5 conceptus using a primary antibody complexed with abiotinylated secondary antibody that complexed with avidin-biotin-peroxidase. The substrate forthe enzymatic reaction was 3,3′-diaminobenzidine (DAB), which reacted to produce a red-brownprecipitate.A specialized version of the whole mount application exists for detection of apoptosis, i.e.,programmed cell death. Nuclear DNA is fragmented during programmed cell death through theapoptotic pathway, and that is diagnostic of this form of cell death. The TUNEL (terminal deoxynucleotidyltransferase [TdT]–mediated dUTP-biotin nick end labeling) method involves enzymaticlabeling of the ends of the DNA fragments and localization of these tagged ends with eitherfluorescent label or biotin-antibody complex methods (such as the avidin-biotin-peroxidaseapproach). This localization in early stage, whole embryos can be very informative regardingspecific sites of action and mechanisms involving cell death. The method is also applicable tosectioned tissues, and in either case there are a number of commercially available kits and reagents(Roche Diagnostics Corp, Indianapolis, IN; Gibco/BRL, Gaithersburg, MD; Oncor, Gaithersburg,MD, and others). Tissues are incubated with the enzyme to label the DNA and then processedthrough the detection step (fluorescent imaging or typical peroxidase enzymatic immunolabelingsteps). 8,9For developmental studies, the immunolocalization is generally not quantitated but is describedas patterns of expression (within specific cell types or certain organs) in the embryo. Many examplesof this application are available in the literature, and recent studies in our laboratory includeimmunolocalization of members of the vascular endothelial growth factor (VEGF) pathway inplacenta, extraembryonic membranes, and embryo. 10 The VEGF (vascular endothelial cell growthfactor) localization in trophoblastic giant cells of a gestation day (GD) 9.5 conceptus shown inFigure 14.1 illustrates the sensitivity and resolution that can be achieved using the ABC techniquewith DAB substrate. Other good examples include studies localizing expression of transforminggrowth factor (TGF)-β isoforms in the developing embryo 11–14 or in embryonic cultured palatalmesenchymal cells. 15 Mahmood et al. 13 examined the effects of retinoic acid on expression of TGFβin sections of early stage mouse embryos (GD 8.5 to 10.5) with the avidin-biotin-peroxidaselocalization method. Studies of the role of serotonin as a regulator of craniofacial and cardiacmorphogenesis used whole embryo culture to evaluate the effects of inhibitors of 5-hydroxytryptophan(5-HT)uptake in vitro, and immunohistochemical analysis of inhibition provided valuableinsight into the role of serotonin metabolism in morphogenesis and teratogenesis. 16,17In toxicology studies, the detection of differences between levels of expression in control versustreated tissues is an important goal, and measures of the intensity of immunohistochemical stainingcan provide some insight into the effects of a treatment. Semiquantitative analysis of stained sections© 2006 by Taylor & Francis Group, LLC

604 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONcan be performed with densitometry when the substrate is permanent (e.g., in our studies densitometrywas performed using avidin-biotin-peroxidase complex with diaminobenzidine). The sectionsare not counterstained with histological stains because staining interferes with the densitometricanalysis. For quantitative applications, extreme care must be exercised to expose all specimens toidentical solutions for precisely timed incubation periods, as well as to include appropriate controlsfor endogenous peroxidase and nonspecific binding of either primary or secondary antibodies. Also,substrate development should be monitored on test slides to determine the reaction time thatproduces a density allowing detection of either greater or lesser densities in the specimens. Theintensity of immunostaining can be ranked by the investigator based on viewing under the microscope(typically three levels are assigned: low = 1, medium = 2, high = 3). However, this methodshould be considered semiquantitative at best and susceptible to observer bias (the scorer shouldbe “blinded” to specimen identity).There are several computer software products available for densitometric analysis (e.g., Image-Pro Plus, NIH Image), and these can generate density values from images of immunostainedsections. Capture of images through a digital camera mounted on the microscope offers a mostconvenient method, although print or film images are also analyzed with these programs (generallythis requires scanning of the film or print to generate a digital file). In all of the densitometricmethods, it is critical to acquire images under tightly controlled, uniform illumination conditionsfor consistency across specimens. The darkest and lightest regions must not be “off scale” (notmaximal values). All images intended for densitometric analysis should be acquired with a monochrome(not color) camera, and sections should not be counterstained. The densitometric data arerelative values for a specific protein, and treatment effects are most appropriately expressed ascomparisons between the same cell type in the same region of treated and control sections. Suchdata may be statistically analyzed by ANOVA (analysis of variance), but consultation with astatistician is recommended because these data sets may require transformation or other approaches.B. In Situ HybridizationWith this technique, mRNA transcribed from the gene of interest can be localized on either sectionedtissue or in whole mount preparations. These protocols require preparation of antisense probes thathybridize to the sense strand (mRNA) in the tissue. Before this method can be used, it is thereforenecessary to have a gene sequence from which to synthesize the antisense probe. Through acomputer search of a gene database, it may be possible to determine if the gene of interest hasbeen sequenced and entered in the database. A literature search may identify a research group thathas the cDNA (complementary DNA) for the gene of interest, and most investigators will makethe material available on request after publication of its sequence.Alternatively, the American Type Culture Collection (ATCC) maintains an NIH repository ofhuman and mouse DNA probes and libraries. 18 cDNAs are placed in the repository by investigatorsand can be ordered from the catalog for a small handling fee. cDNAs from ATCC or private sourcesare not usually ready to use for making in situ hybridization probes. To prepare a quantity of thecDNA for use in probe preparation, it may be feasible to amplify the sequence with polymerasechain reaction (PCR). Alternatively, the cDNA can be cloned into a plasmid vector. Preparation ofRNA antisense–labeled probes for in situ hybridization will involve in vitro transcription of thecDNA sequence, which requires a promoter sequence for the polymerase (typically T3 or T7polymerase). Cloning or insertion of the cDNA into an appropriate vector and production of largequantities of that plasmid in bacterial culture is a standard, classical method. When received aspurified isolated cDNA or in an unsuitable vector, the cDNA must be cloned into a suitable vectorby standard protocols (as described in molecular biology reference texts). It is also possible to usesynthesized oligonucleotide probes that require no cloning.The choice of probe type (RNA, double strand DNA, single strand DNA, oligonucleotide),labeling method (in vitro transcription, end labeling, nick translation, random primer synthesis),© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 605and detection system (radioactive — 32 P, 35 S, 125 I, 3 H — or nonradioactive tag on nucleotideprecursor) will affect the sensitivity and resolution of the probe. The methods selected may bedetermined not only by the properties desired in a probe but also by individual experience, accessto necessary facilities, and level of resources and funding. There are several general reviews andsources of information that may be helpful in making these choices, in particular reviews by Angererand Angerer 19 and by Singer et al., 20 which give good overviews of the advantages and disadvantagesof the different approaches.A general consensus protocol has emerged based on experiences of many investigators, andthis version was used in our laboratory 11 and in most of the studies cited in this section. Briefly,the protocol for examination of embryonic gene expression uses 4% paraformaldehyde-fixed tissuesembedded in paraffin. In all the protocol steps, extreme care must be taken to prevent degradationof the RNA by RNase enzymes. The sections are placed on triethoxysilane coated (TESPA) slides.Probes are synthesized with [ 35 S]CTP incorporation, using in vitro transcription from an RNApolymerase promoter (T3, T7, or Sp6) on the linearized vector. Probe length may be adjusted to100 to 200 bases by alkaline hydrolysis at 60ºC.Prehybridization washes prepare the sections with proteinase K digestion, and additional fixationand acetylation steps. Hybridization is performed in humid chambers in an oven or incubatorovernight. Posthybridization treatment generally includes washes of various stringencies at elevatedtemperatures, with 50% formamide in some steps, as well as an RNase A digestion step to removeunhybridized probe (single stranded RNA). Slides are dipped in liquid photographic emulsion(Kodak NTB2), dried, and stored (4ºC) for days to weeks. After the emulsion is developed andcoverslipped, the slides can be viewed under the microscope by darkfield or (if they contain abundantlabel) brightfield illumination. This method has been successfully applied to embryonic tissues tocharacterize expression of specific genes, generating time, and tissue specific patterns of expression.An example from our laboratory is shown in Figure 14.2A, in which the expression of the arylhydrocarbon receptor (AhR) was characterized in the developing mouse embryo.Whole mount in situ hybridization offers the potential to visualize the expression of genes inthree dimensions in small specimens. 12,21 The power of this approach is illustrated in Figure 14.2B,in which bmp4 mRNA is detected in a GD 10 mouse embryo by use of a digoxigenin-labeledprobe. As was discussed for whole mount immunohistochemical staining, the size of the specimenis a limiting factor because the reagents and probe need to penetrate the tissue. This also mandatesthe use of a treatment to permeabilize the tissues, which is most often done by pretreating thespecimen with proteinase K. The length of pretreatment must be adjusted based on the age of theembryo and the activity of the enzyme. Following proteinase K digestion, the embryos are usuallyrefixed in an aldehyde fixative. The probe is synthesized as before, using in vitro transcription ofantisense RNA with a nonradioactive label. Probe length is generally not reduced by most laboratories,and experiments using riboprobes of 1 kilobase (kb) or greater are reported in the literature.Some protocols for whole mount in situ include incubations with methanol and hydrogen peroxideas a bleaching step. This incubation will also inhibit endogenous peroxidase and thus prevent anypotential for background with a peroxidase-based enzymatic detection method. The posthybridizationwashes may include RNase digestion and 50% formamide in low salt buffer at elevatedtemperature. The specimens may also be heated to 70ºC for 20 min prior to probe detection inorder to inhibit endogenous alkaline phosphatases. The hybridized probe is then detected withantidigoxigenin antisera conjugated to alkaline phosphatase. Alkaline phosphatase reacts withsubstrate to the enzyme, producing a colored precipitate. Alternatively, the antisera may carry afluorescent tag that can be visualized with a fluorescent dissecting microscope or by confocal laserscanning microscopy. Posthybridization incubations to render the specimen more transparent, thusallowing visualization of the label within the tissue and not just on the surface, may also be utilized.Other examples of applications of in situ hybridization to localize gene expression in thedeveloping embryo include studies of the effects of retinoic acid. Retinoic acid (RA) is an endogenouscompound important to embryonic development, but it also has the ability to disrupt morphogenesis© 2006 by Taylor & Francis Group, LLC

606 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFigure 14.2(A) Expression of the aryl hydrocarbon receptor (AhR) was examined in the GD 10 C57BL/6Nmouse embryo by in situ hybridization. Tissues were fixed in paraformaldehyde and embeddedin paraffin, and sections were incubated with radiolabeled RNA riboprobes. At this developmentstage, AhR mRNA was expressed at high levels in the neuroepithelial cells of the developingbrain. (B) Whole mount in situ hybridization in a GD 10 mouse embryo by use of a digoxigeninlabeledantisense riboprobe for bmp4. Arrows indicate regions of the nasal prominence and themaxillary and mandibular processes, which express bmp4. Note: np = nasal prominence, mx =maxillary process, mn = mandibular process. (Probe courtesy of Dr. Brigid Hogan.)when present inappropriately during development. A variety of molecular approaches have beenutilized in the study of this compound in the embryo. The differential distribution of the cellularretinoic acid binding proteins (CRABP I and II), the cellular retinol binding protein (CRBP), andthe retinoic acid receptors (multiple isoforms of RAR and RXR) provide excellent examples of thelocalization of mRNA spatially and temporally in developing embryos. 21–23 Hox gene expressionhas also been extensively studied with this technique, both in the normal embryo and in determinationof the effects of retinoic acid treatment on Hox and Krox gene expression. 24,25 Whole mountdetection was able to demonstrate that exogenous RA-induced expression of Hox-2 genes occurredin anterior regions of the embryo in a stage dependent manner. 26 The embryonic expression ofseveral TGFβs and growth and differentiation factors was also studied by several investigators 27,28using in situ hybridization. The role of growth factors in cardiac development was investigated© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 607using whole mount hybridization, sectioned tissue in situ, and immunohistochemistry, 12 and in ourlaboratory the expression of growth factors, including TGFβs, was examined in embryonic palatalshelves after exposure to chemicals that induce cleft palate. 11,29III. APPROACHES FOR QUANTIFYING MRNA EXPRESSIONA. Semiquantitative Methods1. Northern BlotNorthern blot assays are used to examine mRNA expression and can provide quantitative informationconcerning levels of expression. However, unlike the case with in situ hybridization, informationregarding localized expression within a tissue cannot be obtained because the samples arehomogenized for preparation of total RNA. Protocols for Northern blots are available in the generalreferences provided in the introduction, 2–4 and a brief outline of the basics follows. Investigatorsshould bear in mind that modifications and adjustments may be required to optimize conditionsfor a particular gene. In order to learn the method, it may be advisable to probe for a gene withwell characterized expression and that has been previously examined by Northern blot.Total or polyA + RNA is prepared from specific tissues or the whole embryo, and this RNA isseparated by electrophoresis on a denaturing gel. Northern blots allow multiple samples to beexamined simultaneously on a single gel. The method is reasonably sensitive, detecting mRNAsrepresenting as little as 0.003% of total mRNA. Nevertheless, the small size of most embryonicsamples generally requires pooling of multiple embryos (and in some cases pooling entire litters)to obtain sufficient total RNA. Also, when the target mRNA has very low expression, the treatmentdecreases expression, or relatively few cells in the tissue express the gene, this method may not besensitive enough to detect or quantitate the mRNA. After electrophoresis in the gel, the RNA isblotted onto a filter of nylon or nitrocellulose and bound to the filter by cross-linking with UVirradiation or baking in a vacuum oven. An antisense radiolabeled probe (RNA or DNA) is preparedagainst the specific mRNA being examined, and filter, probe, and hybridization buffer are incubatedovernight. After the excess probe is washed away, the filter is placed on x-ray film. After incubationat 80ºC, the exposed film is developed, and the bands are quantified by densitometry.The amount of probe hybridizing to the RNA is proportional to the amount of target RNA onthe blot. Thus, the intensity of the band on the film can be used to estimate differences betweensamples. Typically, the amount of RNA loaded in each lane is also estimated by hybridization ofa probe to a constitutively expressed gene (e.g., glyceraldehyde 3-phosphate dehydrogenase[GAPDH] or β-actin). This allows a correction between lanes for any variability in the amountloaded in different lanes. Hybridization to the mRNA of a constitutive gene can be performed onthe blot by either stripping off the first probe (this requires that the blot never be allowed to dryduring posthybridization manipulations, including exposure of the film) or by allowing the radioactivityto diminish to the point that the first probe no longer exposes the film (only a matter ofweeks for 32 P-labeled probes). The option to reprobe can also be very useful for screening for theexpression of several genes in the same samples. The size of the bands to which the probe hybridizescan be estimated by comparing the migration of the band to the migration of known molecularweight standards or by use of the 28S and 18S ribosomal RNA in total RNA samples. Ethidiumbromide staining allows the migration distances to be observed with a UV transilluminator andphotographed, either in the gel or on the blot.Titration on slot blots or RNase protection, as described in the following sections, is most oftenused for quantitation in conjunction with Northern blot data. Northern blots are more often usedfor semiquantitative, qualitative, or temporal studies rather than for quantitation. Quantitation withNorthern blots alone can be relatively difficult, and some additional considerations are important.© 2006 by Taylor & Francis Group, LLC

608 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAs is the case with protein staining on acrylamide gels, x-ray film exposure is linearly related to band(or spot) quantity only up to a point. As a result of a combination of intense signal and “flaring,”linearity is lost at high levels of signal. Therefore, it is generally advisable to use multiple exposuretimes and multiple amounts of standards to assure that the band intensities are in the linear range.Examples of Northern blot analysis of embryonic gene expression include studies of theexpression of retinoic acid receptor-β (RARβ) and the growth factor, TGFβ. Chick embryos wereexamined for expression of RARβ mRNA at developmental stages 22, 24, and 25. 30 The effects ofretinoic acid on the expression of RARβ were examined in stage-28 embryos. In this study, 10 µgtotal RNA was loaded per lane, and type I collagen expression was used to control for variabilityin loading. The embryos were dissected into frontonasal mass, mandibular primordia, head, limbbuds, and trunk, and each region was homogenized for total RNA preparation. In this study, theNorthern blots were not quantitated but were used to demonstrate that the mRNA is expressed inthese tissues. RARβ transcripts were localized with in situ hybridization, and the trends for increasedexpression found with Northern analysis were demonstrated over sections that also providedinformation on regional expression patterns.There have been numerous studies localizing the expression of the TGFβ isoforms in the mouseembryo with in situ hybridization. In Schmid et al., 31 Northern analysis was used to confirmspecificity of the probes and to determine transcript size in GD 13, 14, and 16 mouse embryos andin neonates.Northern blot analysis of cultured embryonic cells can also provide insight into the mechanismof action of a teratogen. Assays of the effects of retinoic acid on gene expression in cultured murineembryonic palate mesenchymal (MEPM) cells included Northern blot analysis of effects on TGFβ3,type III TGFβ receptor, CRABP I, histone H3, and tenascin. 32,33 This approach detected a differentialtime course of effects on transcription of these genes, and this information should be useful indesigning further studies. In a similar application, Northern blot analysis was used extensively todetermine the regulation of early immediate genes (c-jun, c-fos, junB) in developing lung and brainof mouse embryos. 34 In conjunction with the blots, immunohistochemistry localized the expressionof these transcription factors.There are nonradioactive detection methods for hybridization assays that can be used, but thesemethods may be unable to detect very low abundance mRNA. Northern blotting is less sensitivethan the RNase protection assay but may be useful for genes that are widely expressed or havehigh transcription rates.2. Dot Blot or Slot BlotThese blots are similar to Northern blots as binding of sample to a membrane support is followedby hybridization with a probe. This is a faster and easier method that may be useful for screeningfor specific target genes, but it is less accurate and sensitive than the Northern blot. The RNA orDNA sample is applied directly to restricted regions of the filter, often using a special vacuumapparatus. The hybridization, film exposure, and quantitation are similar to those described forNorthern blots. This method will not detect degradation of the RNA, and nonspecific hybridizationalso cannot be determined. This is because the range of mRNA band sizes within the sample cannotbe observed.In a study of 3-hydroxy-3-methylglutaryl coenzyme A (HGM-CoA) reductase mRNA in therat embryo, investigators localized expression with in situ hybridization, confirmed the specificityof their probe with Northern blots, and used dot blots to compare stimulated and unstimulatedcellular expression of the transcript. 35 A titration of total RNA from 5 to 0.1 µg on the dot blotdemonstrated the sensitivity of the response, as well as enhanced transcription in the stimulatedcell sample. The application of multiple techniques was not redundant in this study, as each providedslightly different information, and the combination proved advantageous in defining the responseto the treatment.© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 609B. PCR and RT-PCR Quantitative MethodsPolymerase chain reaction amplification is an enzymatic method to increase the amount of a specifictarget DNA or RNA available for study. Short oligonucleotide sequences (primers) are preparedthat bracket the DNA sequence that is to be amplified, i.e., they specifically recognize and bind tobases at the 5′ and 3′ ends of the target region. A thermal cycler is used to alternatively denaturethe double-stranded DNA (higher temperature) and promote primer hybridization and strand synthesis(lower setting), with repeated, automated cycling. Cycles typically require 5 min, and a fullrun may involve 30+ cycles. The template to which the primers bind must be DNA, so that theamplification of RNA requires the additional step of reverse transcription of the RNA to synthesizecDNA (RT-PCR). This is an extremely valuable technique for the study of embryos as qualitativeand quantitative information can be obtained from very small tissue samples, potentially even thoseconsisting of only a few cells. The thermostable enzymes required for these reactions are expensive,and large scale assays could become major components of a research budget. General informationand protocols are available in the three volume manual Molecular Cloning: A Laboratory Manual 2and in PCR Protocols, A Guide to Methods and Applications. 36RT-PCR can be used to detect gene expression in small specimens, and some studies indevelopmental biology included detection of gene transcripts in developing chick limb bud, duringmouse mandibular and tooth bud morphogenesis, and in developing mouse lung. 37–40 Detection ofthe presence of RXRγ mRNA in chick wing buds was achieved by reverse transcription andamplification of total RNA; β-actin was amplified as a positive control, and in some tubes theenzyme reverse transcriptase was omitted to confirm that the reaction was not contaminated bygenomic DNA. These are useful controls when attempting to exhibit the presence of a low-leveltranscript in embryonic tissues. In similar studies in the mouse, dissected tooth buds, mandibles,or lung buds were prepared from different stages of development. In the tooth bud, epithelial growthfactor (EGF) mRNA was detected by use of as little as 1 ng of RNA. The relative levels of expressionin each tissue were compared between developmental stages, and the findings were supplementedby immunolocalization of translated protein for the genes being studied. 38–401. Real-Time RT-PCRReal-time RT-PCR 41 is rapidly becoming the method of choice for determining the levels of mRNAexpression. Real time RT-PCR involves measuring the amount of PCR product produced after eachcycle as the reactions proceed. Several companies produce dedicated instruments to accomplishthese measurements. Although, these instruments differ widely in detail, they all are essentiallythermal cyclers combined with a fluorometer. The simplest approach to doing real time RT-PCRconsists of adding an intercalating dye, such as SYBR Green 1 (that fluoresces when bound toDNA) to the PCR reaction mixture. 42 The increase in fluorescence as the PCR product accumulatesis measured and is proportional to the amount of product. A disadvantage to this method is thatthe intercalating dyes will bind to all DNA synthesized, including primer dimers and any othernonspecific products produced.To improve specificity several related approaches have been developed. These methods rely onsynthesizing two probes, a reporter that will hybridize with the PCR product and is also labeledwith a fluorescent dye and a quencher, a dye that will absorb the light emitted by the fluorescentdye when they are in close proximity. Hydrolysis probes (i.e., Taqman) are end-labeled with thefluorescent dye and quencher on opposite ends of the probe. 43,44 Because the probes are short(typically 20 to 40 base pairs), the signal from the reporter is effectively quenched. This probe isdesigned to anneal to the specific amplicon (DNA product of the polymerase chain reaction) betweenthe forward and reverse primers. During the extension phase, the 5-exonuclease activity of Taqpolymerase degrades the probe, releasing the reporter and allowing it to fluoresce. Commonalternatives to hydrolysis probes are hybridization probes (i.e., molecular beacons). 45,46 Like the© 2006 by Taylor & Francis Group, LLC

610 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION2600Florescence24002200200018001600140012001000800600400200Threshold0Threshold Cycle353025y = 44.877 − 3.4109x R 2 = 0.999203.0 4.0 5.0 6.0 7.0Log Concentration (molecules)10 7.0 molecules10 6.0 molecules10 5.5 molecules10 5.0 molecules10 4.5 molecules10 4.0 molecules10 3.5 moleculesNegative Control−20010 20Cycle Number30 40Figure 14.3The plot illustrates a typical real-time RT-PCR experimental run in which the increase in fluorescenceis tracked by cycle number. The samples are DNA standards with 10 7 to 10 3.5 moleculesper reaction tube at ½ log intervals, except for the 10 6.5 interval, which was not run. The insetgraph is a plot of the initial concentration of DNA vs. the threshold cycle, the cycle at which eachtube exhibited the same level of amplification.hydrolysis probes, these are also end-labeled with a reporter and quencher. However, the ends ofthe probe contain short complementary sequences designed to allow the probe to form a stem-loopstructure when free in solution. This results in the reporter and quencher being in very closeproximity and having low background fluorescence when compared with hydrolysis probes. Whenthe probe hybridizes to the specific amplicon, the stem-loop structure opens, the distance betweenthe reporter and quencher increases, and the reporter fluoresces. As the PCR primers are extended,the probe is displaced rather than being degraded.Figure 14.3 illustrates a typical real-time RT-PCR experiment that measured the increase influorescence tracked by cycle number. The curve for each sample is characterized by a region whereinsufficient product has been synthesized to allow detection, followed by an exponential increasein the accumulation of product, and finally by the appearance of a plateau. The relative amount ofmRNA in each sample is determined by comparing the cycle number when the fluorescence reachesa threshold level. The threshold level is selected at a point sufficiently above the noise level of theinstrument to obtain accurate fluorescence measurements while amplification is still occurringexponentially. The amount of mRNA in the original sample is inversely proportional to the thresholdcycle number. Reproducibility between real-time RT-PCR runs is sufficient that for many purposesthe use of standards is not required. 47 Standards can be used to reduce run-to-run variability. Thesestandards can be either cDNA or RNA samples. If it is necessary to determine the actual numberof copies of a specific mRNA contained in a given sample, it is important to control for the efficiencyof the reverse transcription reaction. This can be accomplished by including a standard curve derivedfrom a pure RNA sample of the sequence being amplified with each run.© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 6112. Quantitative RT-PCR Based on Image Analysis of PCR Productsin Electrophoretic GelsQuantitative RT-PCR in gel-based applications requires that a method be used to standardize thedensity values between specimens and/or lanes on the gel. Typical approaches are to spike eachsample with an internal standard (synthesized DNA oligo or reverse transcribed RNA sequence)or to amplify a housekeeping gene for comparison purposes. One of these approaches is to comparethe RT-PCR product to an artificial mRNA internal standard or reference gene transcript that usesthe same primers as the target gene. 48 This involves adding increasing amounts of “competitor”(internal standard) to fixed amounts of RT-PCR product in a series of reaction tubes, with simultaneousamplification. The relative amounts of amplified internal standard and target cDNA can beidentified by electrophoretic separation on a gel and visualization with ethidium bromide. Densitometricanalysis produces a ratio of standard to target, and regression analysis identifies the pointat which both standard and target were equally amplified. By inference, the known amount ofstandard predicts an equimolar amount of target present in the tube. Other methods spike the PCRreactions with a mutant, modified, or shortened cDNA or oligo sequence of the target gene. Somestudies express mRNA levels relative to a housekeeping gene amplified from the same samples.Typically a constitutively expressed gene is amplified in a separate PCR reaction from the gene ofinterest. These approaches should be used with caution. Variation between the quality or quantitationof the original total RNA samples, as well as differences in amplification efficiency between differentgenes, may contribute to errors when comparisons are made to housekeeping genes or sequences thatdo not amplify using the same primers and in the same reaction tube as the target gene of interest.Methods using a recombinant mRNA internal standard (rcRNA) eliminate these problems as the internalstandard is reverse transcribed using primer sequences of the target gene, and this includes a T7 DNApolymerase promoter site and a polyA sequence. The internal standard would be added to the totalRNA sample prior to the reverse transcriptase reaction, and the amplified products vary in length fromthe target gene (50 bases approximately) to allow separation on the gel. 48 This protocol gives extremelysensitive detection using very low amounts of total RNA (100 ng total RNA per tube).This approach was used in our study of the expression of AhR, ARNT, and CYP1A1 in humanand mouse embryonic palatal tissues. 49,50 The mRNA levels were expressed in units of moleculesper 100 fg of total RNA. This common unit of expression allowed comparison of the expressionof these genes and the effects of treatment in the mouse after an in vivo exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) to expression in mouse and human palates exposed in vitro.This method was also used to examine the dose-response relationship for induction by TCDD ofa hepatic enzyme (CYP1A1) in adult rats, along with effects on PAI2 and TGFα. 51Another application of RT-PCR involves in situ transcription of mRNA from sectioned tissuesor single cells, followed by amplification in the presence of radiolabeled nucleotides. These transcriptsare used as probes on reverse Southern blots to which cDNAs of suspected target genes arebound. 52 This interesting method is discussed in more detail in the discussion of reverse Southernblots and RT-PCR.C. Solution Hybridization: RNase Protection AssayThe RNase protection assay is a solution hybridization–based method. The target RNA and probe(radiolabeled RNA or DNA) are allowed to hybridize in solution and are then exposed to S1nuclease (RNA-DNA hybrids) or RNase (RNA-RNA hybrids), which degrades all single stranded(nonhybridized) RNA. Electrophoresis on a denaturing polyacrylamide gel separates the samples,and the dry gel can be placed with film for autoradiography. As with Northern blots, the signalintensity is proportional to the amount of target mRNA. This assay is often performed as a titration,where the amount of sample increases while a constant known amount of probe is added to each© 2006 by Taylor & Francis Group, LLC

612 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONtube. This allows estimation of the amount of target RNA by means of linear regression of thehybridization densities for each ratio of target to probe. Solution hybridization is more sensitivethan Northern blots as the hybridization reactions go to completion (on blots a proportion of thetarget will be unavailable because of binding to the membrane).This method is particularly valuable for quantitation of mRNA in embryonic samples and hasbeen used effectively in studies of retinoic acid effects on RAR-β and CRABP-I. 53,54 It has alsobeen employed in developmental biology studies of glucocorticoid receptor (GR) regulation duringlung development 55 and to characterize expression of PDGF A (platelet derived growth factor A)and its receptor in early embryogenesis. 56 In the first example, the effects of retinoic acid exposureon various isoforms of RAR mRNA were determined in GD 11 mouse embryos (whole embryoor dissected limb buds, head, and body) and on GD 14 in whole embryo, dissected limb bud, brain,liver, body, and craniofacial region (including brain). 53 Treated embryos showed tissue and stagespecific increases (2- to 12-fold) in mRNA for various isoforms of RAR. The sensitivity of theassay allowed quantitation of the treatment effects by use of 32 P-labeled antisense cRNA probesand only 1 to 10 µg of total RNA.In a study of GR regulation, both Northern blots and solution hybridization were used to detectmRNA transcription during lung development in fetal rats on GD 16 to 21. 55 The probe for Northernblot analysis was 32 P-labeled antisense cRNA, with 20 µg total RNA loaded per lane. The solutionhybridization probe was a 35 S-labeled antisense cRNA, and β-actin was used as a control gene inboth assays. The protection assay confirmed qualitative information from the Northern blot analysisand provided statistically significant evidence for down-regulation of the gene in response toglucocorticoid exposure. The PDGF study also used a [ 32 P]cRNA probe with β-actin as a controlgene, and more detail was provided regarding the necessity to pool embryos to obtain sufficienttotal RNA to perform the analysis. 56 In this case, embryonic mice from GD 6.5 to 8.5 were examined,and 20 µg RNA per lane was used for each hybridization reaction. Whole embryos were pooledto prepare total RNA as follows: GD 6.5, n=10; GD 7.5, n=5; GD 8.5, n=3. This protocol indicatesthat even with a sensitive assay, the size of the embryos mandated pooling, and data were notobtained from individual embryos. In order to study the responses of single embryos, or even moreimportantly, the effects on particular target tissues, the analytical method of choice would involvePCR amplification of reverse transcribed mRNA (i.e., RT-PCR).D. RNA Transcription In Vitro: Nuclear Run-on AssayThe nuclear run-on assay measures the rate of RNA transcription (transcripts synthesized per geneper minute). The Northern blot, dot blot, solution hybridization, and in situ hybridization assaysall determine amounts of RNA present. Those techniques may not reflect transcription rates,however, because RNA levels may be affected by posttranslational processing and half-life in thecytoplasm. In the nuclear run-on assay, RNA is transcribed in vitro, by use of isolated nuclei andradiolabeled nucleotides. The labeled RNA transcripts are isolated and hybridized to cDNAs boundon filters (reverse Northern blot). The intensity of the autoradiographic signal is then measured andused to estimate the relative rates of transcription between different samples. Care must be takento have equal levels of radiolabeled precursor in each reaction tube. This method may not besensitive enough for poorly expressed mRNAs, and the experimental examples given below bothutilized cultured cells to define the levels at which genes were regulated.The nuclear run-on approach was used to define the up-regulation by retinoic acid of Hox-2 57and the PDGF α-receptor gene 58 in cultured F9 embryonal carcinoma cells. In the Hox-2 study,embryonic stem cells (which are more closely related to cells of the early embryo than the F9 cellsare) and Xenopus embryos were also examined. Exposure to RA increased levels of Hox-2.1 mRNA,as determined by Northern blots, and nuclear run-on experiments with the F9 cells demonstratedthat the increase in steady-state levels of Hox-2.1 was not accompanied by an increased transcription© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 613rate. Examination of the transcription rate by this method suggested that the effects of the chemicalexposure were at a posttranscriptional level. 57 The embryonal carcinoma cells also express PDGFα-receptor, and the combination of retinoic acid and dibutyryl cyclic AMP stimulated expressionof mRNA for that receptor. 58 The PDGF growth factors are expressed in embryos, and a lethalmutation in the patch locus correlates with failure to express the PDGF α-receptor. Exploring theregulation of this receptor with Northern blots and nuclear-run-on analyses demonstrated stronginduction by the RA-cAMP treatment. The study also showed that transcriptional rates increasedproportionally with mRNA levels, indicating transcriptional regulation was involved in the response.IV. CHARACTERIZING THE ROLE OF SPECIFIC GENESAND PATHWAYS OF RESPONSEMolecular techniques can be helpful in identifying potential biomolecules involved in response toa teratogen. This approach is conceptually the inverse of that discussed previously. In this alternativecase, the approach is to screen the genome of the embryo. Differences between treated and controlembryonic gene expression are determined, and a survey across the developmental period of interestmay be involved. The scope of these studies must of necessity be wider, and thus there is thepotential to identify contributors to a response that would otherwise go unrecognized. One approachinvolves screening of genetic libraries derived from embryos or specific organs in various stagesof development. Restriction fragment length polymorphism (RFLP) analysis has also been valuablein studies of human diseases and birth defects and has identified links between specific genes andmalformations. 59,60 Mutational analysis and targeted gene disruption allow the phenocopy approachto be used, in which a defect induced by a teratogenic agent can be compared to similar defectscaused by mutation or targeted gene disruption. Although potentially valuable for developmentaltoxicology studies, the RFLP and phenocopy approaches are not presented here. The followingdiscussion focuses on two methods, subtractive screening and in situ transcription RT-PCR reverseSouthern blot, which have the potential to detect differential gene expression across developmentand between experimental treatments.A. Transgenic AnimalsTransgenic mice are currently being generated to examine the regulation, function, and expressionpatterns of a wide range of genes that are implicated in controlling embryonic development. Thesetransgenic animals can be “reporter mice” that allow the expression of a particular gene to bemonitored during various stages of development. Reporter mice typically use the regulatorysequences from a gene of interest to drive the expression of a reporter gene, such as lacZ, to whichthe regulatory sequences are fused. This reporter transgene is activated in the regions that haveendogenous gene expression and is detected by the appearance of dark blue staining caused byβ-galactosidase activity.Other transgenic mice are generated to “knock out” expression of the gene of interest, generallythrough targeted mutation that results in a homozygous inactivation. These loss-of-function mutantsare produced by disruption of the gene by homologous recombination in embryonic stem (ES)cells. The mutant ES cells are then introduced into a mouse blastocyst to generate chimeric embryosthat are then placed in the hormonally primed uterus of a host, pseudopregnant female. The neonateswill be chimeric, and coat color is usually employed (host blastocyst being from the albino andES cells being from the pigmented strain) as a means of readily identifying the transgenic animals(ES cell homologous recombination was reviewed by Capecchi 61 ). The chimeric mice are bred togenerate mice homozygous for the disrupted gene, and the phenotype of the resulting embryos canbe informative concerning the role of that gene in development.© 2006 by Taylor & Francis Group, LLC

614 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONSimilarly, gain of function transgenic mice can be produced that inappropriately express thegene during development, either constitutively or under the regulation of an inducible promoter.These are typically dominant mutations that are produced in a manner similar to that just describedfor producing ES cell chimeric transgenic individuals, or they may be generated by directly injectingthe cloned DNA into fertilized mouse eggs. The presence of the transgene is detected in the resultingpups by Southern blot analysis of a tail-clip tissue sample.Analysis of a gene of interest, generation of either activating or inactivating mutations, andinsertion of these mutations or reporter constructs into the embryonic cells to produce transgeniclines of mice are not trivial endeavors and require a considerable investment of time and resources.There are increasing numbers of these transgenic lines available, and the developmental toxicologistmay be in a position to benefit from the wide interest of developmental biologists in this area.These mice are frequently obtainable from the laboratory that produced them if the researchersinvolved are contacted with an interesting proposal to examine effects of exogenous agents. Thismay be an option for projects in which the developmental toxicant is suspected to act through thegene modified in the transgenic mouse (or reported in the lacZ transgenic form) or in which thephenotype of the transgenic animal is homologous to that produced by the agent.The study of retinoic acid has taken advantage of transgenic technology to examine the effectsof RA exposure on Hox gene expression patterns, RA activation of retinoic acid receptors, andconsequences of constitutive expression of RAR in the lens of the developing eye. Hox genes areexpressed in the mouse embryo with spatially restricted but overlapping domains, and the combinationof expression of sets of Hox genes may represent a system for determining regional morphologicalspecificity. 62 Hox-lacZ reporter transgenic mice have provided spatial and temporal threedimensionalpatterning of Hox expression during development, 63 and these transgenic mice wereused to characterize the effects of RA on this pattern. 64 In this and other studies, Hox and Kroxgenes were shown to be responsive to RA and to exposure induced homeotic transformation ofhindbrain rhombomere identity.A similar approach was applied to the characterization of retinoic acid receptor gene expressionusing transgenic reporter mice to localize spatially and temporally the distribution of RARs andthe inducibility of the gene by RA. 65,66 The teratogenic RA exposure correlated with abnormalRARβ2 gene expression in the CNS, a region of the embryo sensitive to RA induced dysmorphology.A transgenic line was created in which constitutively expressed RAR-LacZ-fusion protein wastargeted to the ocular lens by linking the transgene to the α-crystallin promoter. The transgenicmice overexpressed RAR in the developing lens and exhibited eye defects similar to those observedafter RA exposure. 67 That particular application demonstrated an ability to target a particular systemfor overexpression of a gene suspected to have a role in the pathogenesis of teratogenic exposure.Growth factor genes are also being studied using transgenic mutations, and examples areavailable for overexpression or knockout of TGFα and TGFβ1. 68–72 TGFβ1 null mutants exhibit arapid wasting syndrome 2 to 3 weeks after birth, excessive inflammatory responses, and earlydeath. 71 The int-2 gene, a member of the FGF growth factor family, was disrupted in transgenicindividuals, and these mutants showed inner ear and tail defects and generally did not survive toadulthood. 72 TGFα appeared to promote tumor progression in overexpressing transgenic individuals,while embryogenesis seemed to progress normally. The transgenic mice with a knockout mutationfor TGFα also develop without major effects on organogenesis but have anomalies of the hair andskin. 69 Bryant et al. 73 used TGFα and EGF knockout mice to study the mechanism through whichTCDD produces cleft palate and hydronephrosis in embryonic mice. TCDD disrupted control ofcell proliferation and differentiation in the embryonic epithelial cells, and the responses in thepalate and ureter correlated with tissue specific altered expression of EGF and/or TGFα. Using thetransgenic knockouts revealed that EGF expression is important in mediating the responses inpalatal cells and that the ureteric cells were more responsive in the absence of EGF.A number of transgenic gene knockouts have minimal effects on the developing embryo, andit has been concluded that other closely related proteins are able to compensate functionally. Other© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 615hypotheses suggest that the proteins are expressed superfluously, where they have no function;eliminating these proteins thus has no discernible consequences. 70 On the other hand, knockout ofgenes involved in response to hypoxia or vascularization has proven to be lethal in utero (reviewedby Hanahan 74 ), and studies of the etiology of the lethality have given new insight into the developmentof the placenta and vascularization of the yolk sac and the embryo. 10,75,76 Interpretation ofthe developmental consequences of gene knockout or constitutive expression in transgenic miceappears to be a complex issue, and use of these models in developmental toxicology would alsorequire qualified interpretations.B. In Vitro Antisense RNAAntisense oligo probes can be designed to be reasonably stable in culture medium, to enter culturedembryos, organs, or cells, and to hybridize specifically with a target mRNA to inhibit translation. 77This approach is similar to transgenic knockout of genes for determining effects on developmentof failure to express a specific gene. The effectiveness of this approach depends on the ability ofthe oligo probe to reach the specific target in sufficient concentration. Size and stability in cultureare important factors, along with concentration and incubation temperature. These characteristicsmay vary with oligonucleotide class, modification of the oligos, and type of culture (e.g., wholeembryo vs. cell monolayer); the outcome may differ between different cell types.Use of antisense oligonucleotides directed against endogenous expression of the growth factorEGF in tooth bud organ culture resulted in abnormal tooth morphogenesis and inhibited growth ofthe cultures. 39 The relative levels of endogenous EGF and EGF receptor expressed in developingteeth were also determined in this study by use of RT-PCR and immunolocalization of expressionpatterns. The application of multiple methodologies provided strong support for the role of EGFin odontogenesis. The same research group applied this multitechnique approach (using RT-PCR,antisense in vitro, and immunolocalization) to determine the role of TGFβs in the morphogenesisof the mandible and tooth. 40,78 The antisense oligo inhibition of TGFβ1 resulted in dysmorphologyof Meckel’s cartilage, while treatment with antisense TGFβ2 had no such effect.In whole embryo culture, the role of wnt proto-oncogene family members in development wasexamined by injecting an antisense oligo into embryonic amniotic cavities. 79 PCR analysis detecteddecreased wnt mRNA, and malformations were induced in the treated embryos. This approach mayprovide an alternative to transgenic analysis when models are available that allow developmentalevents of interest to proceed in cultured cells, organ culture, or whole embryo culture. This mayallow the role of several genes to be investigated in culture with less investment of resources andtime than are required for generating transgenic animals. However, as with the data from knockouttransgenic mice, interpretation of these experiments should be approached with caution as multiplefactors can influence their outcomes.C. DNA ArrayThis is a powerful technique for assessing gene expression and has extensive applications in thearea of developmental and reproductive toxicity. For this reason, the full discussion of this methodis presented in a separate chapter (Chapter 15).D. Genetic Library ScreeningThis approach is directed toward detection of specific effects on regulation of gene expression andallows one to survey the entire genome to identify target genes. The method has been used toexamine gene expression at different stages of development or variable expression in differenttissues and to detect altered gene expression after treatment during development. Differential andsubtractive screening methods have been developed for these applications. The differential screen© 2006 by Taylor & Francis Group, LLC

616 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONis generally not as likely to detect transcripts from rarely expressed genes or to discern differencesbetween low-level expression mRNAs from a contrasted pair of samples. Subtractive hybridizationcross-hybridizes two mRNA or DNA samples that have been isolated from different experimentalsamples and then isolates the nonhybridized sequences. All nucleic acid sequences occurring inboth samples hybridize and are removed, leaving behind the unmatched mRNA or DNA. Thisisolated material represents the genes expressed in one sample and not the other. Using a higherconcentration of one sample than the other results in detection of either increased or decreasedexpression depending on which sample (control or treated, or different developmental stages) waspresent in excess.Branch et al. 80 adapted this approach to examine the effects of restraint of pregnant mice ongene expression in the embryos. Maternal restraint significantly increases the incidence of lumbarribs and encephalocele in mice. GD 8 embryonic mRNA was isolated from the treated group andreverse transcribed to give single stranded cDNA with c-tailed 3′ ends. Control mRNA wasbiotinylated and then cross-hybridized to the cDNA. Streptavidin binding removes all biotinylatedmRNA (hybridized and single stranded), subtracting or leaving behind the unmatched cDNA ofthe treated embryos. After second strand synthesis, this cDNA can be PCR amplified and clonedinto vectors for transformation of E. coli. The bacterial colonies with vector containing the cDNAinsert are isolated and sequenced, with the intent to identify the gene by comparing the sequenceto those in a gene data bank. In the stress-restraint study, seven cDNAs were isolated. One wasidentified as encoding a protein important in cell cycle regulation, and the others are under furtherstudy.Subtractive hybridization was also used to isolate mRNAs associated with programmed celldeath induced by exposure of thymoctyes to glucocorticoids. 81 One of the isolated genes wassequenced and found to contain a zinc-finger domain, suggesting involvement in DNA regulation,while a second gene appeared to encode a membrane-bound protein.When the isolated sequence is compared with known gene sequences, the isolation of a sequencethat is differentially expressed may lead to identification of the gene involved. The cDNA clonemay also be used to construct fusion proteins that can be synthesized in E. coli, purified, and usedto prepare antibodies. These antibodies may then be useful in isolating the protein that is encodedby the gene. Alternatively, the clone sequence may be used to synthesize a small peptide againstwhich to raise antibody.E. Reverse Southern Blots and RT-PCR In Situ TranscriptionIn situ transcription of mRNAs on a sectioned embryo coupled with RT-PCR and reverse Southernblotting has recently been proposed as a screen of differential gene activity. 52,82–84 With this method,specific regions of a tissue may be examined and compared with like regions from other developmentalstages, or controls can be compared with treated embryos. The in situ transcription usesoligo primers incubated directly on the sectioned tissue (or selected portion of the section followingremoval of unwanted regions) to synthesize cDNA directly from mRNA. Following separation ofthe mRNA-cDNA hybrids and second strand synthesis to generate double stranded cDNA, thismaterial can be amplified by PCR for use as a probe or it can be cloned into a vector. The riboprobesprepared from this cDNA can be used to screen Southern dot or slot blots, in which cDNA sequencesof genes suspected to be involved in the mechanism of toxicity or the developmental pathway forthat tissue are bound to the membrane. Hybridization intensity reflects the relative abundance ofthe mRNAs of interest. Gene expression in the Splotch mutant mouse, a model for neural tubedefects, was examined for a range of growth factors, transcription factors, Hox genes, and extracellularproteins. 82 This approach has detected effects on candidate genes after exposure to valproicacid, retinoic acid, or hyperthermia in SWV mice. 83 Similar studies examined changes in geneexpression after exposure of SWV mice to phenytoin. 84© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 617REFERENCES1. Lodish, H., Berk, A., Zipursky, L., Matsudaira, P., Baltimore, D., and Darnell, J., Molecular CellBiology, 4th ed., W.H. Freeman, New York, 2000.2. Sambrook, J., MacCallum, P., and Russell, D., Molecular Cloning: A Laboratory Manual, 3rd ed.,Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 2001.3. Perbal, B., A Practical Guide to Molecular Cloning, 2nd ed., John Wiley, New York, 1988.4. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., and Struhl, K.,Short Protocols in Molecular Biology, 4th ed., John Wiley, New York, 1999.5. Scientific Frontiers in Developmental Toxicology and Risk Assessment, National Academy Press,Washington, DC, 2000.6. Harlow, E. and Lane, D., Using Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory,Cold Spring Harbor, NY, 1999.7. Mark, M., Lufkin, T., Vonesch, J.L., Ruberte, E., Olivo, J.C., Dolle, P., Gorry, P., Lumsden, A., andChambon, P., Two rhombomeres are altered in Hoxa-1 mutant mice, Development, 119, 319, 1993.8. Stadler, B., Phillips, J., Toyoda, Y., Federman, M., Levitsky, S., and McCully, J.D., Adenosineenhancedischemic preconditioning modulates necrosis and apoptosis: effects of stunning andischemia-reperfusion, Ann. Thorac. Surg., 72, 555, 2001.9. Takagi, T.N., Matsui, K.A., Yamashita, K., Ohmori, H., and Yasuda, M., Pathogenesis of cleft palatein mouse embryos exposed to 2,3,7, 8-tetrachlorodibenzo-p-dioxin (TCDD), Teratog. Carcinog.Mutagen., 20, 73, 2000.10. Abbott, B.D. and Buckalew, A.R., Placental defects in ARNT-knockout conceptus correlate withlocalized decreases in VEGF-R2, Ang-1, and Tie-2, Dev. Dyn., 219, 526, 2000.11. Abbott, B.D., Harris, M.W., and Birnbaum, L.S., Comparisons of the effects of TCDD and hydrocortisoneon growth factor expression provide insight into their interaction in the embryonic mouse palate,Teratology, 45, 35, 1992.12. Dickson, M.C., Slager, H.G., Duffie, E., Mummery, C.L., and Akhurst, R.J., RNA and proteinlocalisations of TGF beta 2 in the early mouse embryo suggest an involvement in cardiac development,Development, 117, 625, 1993.13. Mahmood, R., Flanders, K.C., and Morriss-Kay, G.M., Interactions between retinoids and TGF betas in mouse morphogenesis, Development, 115, 67, 1992.14. Pelton, R.W., Saxena, B., Jones, M., Moses, H.L., and Gold, L.I., Immunohistochemical localizationof TGF beta 1, TGF beta 2, and TGF beta 3 in the mouse embryo: expression patterns suggest multipleroles during embryonic development, J. Cell Biol., 115, 1091, 1991.15. Linask, K.K., D’Angelo, M., Gehris, A.L., and Greene, R.M., Transforming growth factor-beta receptorprofiles of human and murine embryonic palate mesenchymal cells, Exp. Cell Res., 192, 1, 1991.16. Shuey, D.L., Sadler, T.W., and Lauder, J.M., Serotonin as a regulator of craniofacial morphogenesis:site specific malformations following exposure to serotonin uptake inhibitors, Teratology, 46, 367,1992.17. Yavarone, M.S., Shuey, D.L., Tamir, H., Sadler, T.W., and Lauder, J.M., Serotonin and cardiacmorphogenesis in the mouse embryo, Teratology, 47, 573, 1993.18. Nierman, W.C. and Maglott, D.R., American Type Culture Collection NIH Repository of Human andMouse DNA Probes and Libraries, American Type Culture Collection, Rockville, MD, 1992.19. Angerer, L.M. and Angerer, R.C., Localization of mRNAs by in situ hybridization, Methods Cell.Biol., 35, 37, 1991.20. Singer, R.H., Lawrence, J.B., and Villnave, C., Optimization of in situ hybridization using isotopicand non-isotopic detection methods., Biotechniques, 4, 230, 1986.21. Ruberte, E., Friederich, V., Morriss-Kay, G., and Chambon, P., Differential distribution patterns ofCRABP I and CRABP II transcripts during mouse embryogenesis, Development, 115, 973, 1992.22. Ruberte, E., Friederich, V., Chambon, P., and Morriss-Kay, G., Retinoic acid receptors and cellularretinoid binding proteins. III. Their differential transcript distribution during mouse nervous systemdevelopment, Development, 118, 267, 1993.23. Perez-Castro, A.V., Toth-Rogler, L.E., Wei, L.N., and Nguyen-Huu, M.C., Spatial and temporal patternof expression of the cellular retinoic acid-binding protein and the cellular retinol-binding proteinduring mouse embryogenesis, Proc. Natl. Acad. Sci. USA, 86, 8813, 1989.© 2006 by Taylor & Francis Group, LLC

618 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION24. Hunt, P., Wilkinson, D., and Krumlauf, R., Patterning the vertebrate head: murine Hox 2 genes markdistinct subpopulations of premigratory and migrating cranial neural crest, Development, 112, 43,1991.25. Morriss-Kay, G.M., Murphy, P., Hill, R.E., and Davidson, D.R., Effects of retinoic acid excess onexpression of Hox-2.9 and Krox-20 and on morphological segmentation in the hindbrain of mouseembryos, EMBO J., 10, 2985, 1991.26. Conlon, R.A. and Rossant, J., Exogenous retinoic acid rapidly induces anterior ectopic expression ofmurine Hox-2 genes in vivo, Development, 116, 357, 1992.27. Lehnert, S.A. and Akhurst, R.J., Embryonic expression pattern of TGF beta type-1 RNA suggestsboth paracrine and autocrine mechanisms of action, Development, 104, 263, 1988.28. Millan, F.A., Denhez, F., Kondaiah, P., and Akhurst, R.J., Embryonic gene expression patterns of TGFbeta 1, beta 2 and beta 3 suggest different developmental functions in vivo, Development, 111, 131,1991.29. Abbott, B.D. and Birnbaum, L.S., TCDD alters medial epithelial cell differentiation during palatogenesis,Toxicol. Appl. Pharmacol., 99, 276, 1989.30. Rowe, A., Richman, J.M., and Brickell, P.M., Retinoic acid treatment alters the distribution of retinoicacid receptor-beta transcripts in the embryonic chick face, Development, 111, 1007, 1991.31. Schmid, P., Cox, D., Bilbe, G., Maier, R., and McMaster, G.K., Differential expression of TGF beta1, beta 2 and beta 3 genes during mouse embryogenesis, Development, 111, 117, 1991.32. Nugent, P. and Greene, R.M., Interactions between the transforming growth factor beta (TGF beta)and retinoic acid signal transduction pathways in murine embryonic palatal cells, Differentiation, 58,149, 1994.33. Nugent, P., Sucov, H.M., Pisano, M.M., and Greene, R.M., The role of RXR-alpha in retinoic acidinducedcleft palate as assessed with the RXR-alpha knockout mouse, Int. J. Dev. Biol., 43, 567, 1999.34. Molinar-Rode, R., Smeyne, R.J., Curran, T., and Morgan, J.I., Regulation of proto-oncogene expressionin adult and developing lungs, Mol. Cell Biol., 13, 3213, 1993.35. Brewer, L.M., Sheardown, S.A., and Brown, N.A., HMG-CoA reductase mRNA in the post-implantationrat embryo studied by in situ hybridization, Teratology, 47, 137, 1993.36. Innis, M.A., Gelfand, D.H., Sninsky, J.J., and White, T.J., PCR Protocols, A Guide to Methods andApplications, Academic Press, New York, 1989.37. Thaller, C., Hofmann, C., and Eichele, G., 9-cis-retinoic acid, a potent inducer of digit patternduplications in the chick wing bud, Development, 118, 957, 1993.38. Hu, C.C., Sakakura, Y., Sasano, Y., Shum, L., Bringas, P., Jr., Werb, Z., and Slavkin, H.C., Endogenousepidermal growth factor regulates the timing and pattern of embryonic mouse molar tooth morphogenesis,Int. J. Dev. Biol., 36, 505, 1992.39. Warburton, D., Seth, R., Shum, L., Horcher, P.G., Hall, F.L., Werb, Z., and Slavkin, H.C., Epigeneticrole of epidermal growth factor expression and signalling in embryonic mouse lung morphogenesis,Dev. Biol., 149, 123, 1992.40. Chai, Y., Mah, A., Crohin, C., Groff, S., Bringas, P., Jr., Le, T., Santos, V., and Slavkin, H.C., Specifictransforming growth factor-beta subtypes regulate embryonic mouse Meckel’s cartilage and toothdevelopment, Dev. Biol., 162, 85, 1994.41. Heid, C.A., Stevens, J., Livak, K.J., and Williams, P.M., Real time quantitative PCR, Genome Res.,6, 986, 1996.42. Gentle, A., Anastasopoulos, F., and McBrien, N.A., High-resolution semi-quantitative real-time PCRwithout the use of a standard curve, Biotechniques, 31, 502, 2001.43. Giulietti, A., Overbergh, L., Valckx, D., Decallonne, B., Bouillon, R., and Mathieu, C., An overview ofreal-time quantitative PCR: applications to quantify cytokine gene expression, Methods, 25, 386, 2001.44. Holland, P.M., Abramson, R.D., Watson, R., and Gelfand, D.H., Detection of specific polymerasechain reaction product by utilizing the 5′-3′ exonuclease activity of Thermus aquaticus DNA polymerase,Proc. Natl. Acad. Sci. USA, 88, 7276, 1991.45. Piatek, A.S., Tyagi, S., Pol, A.C., Telenti, A., Miller, L.P., Kramer, F.R., and Alland, D., Molecularbeacon sequence analysis for detecting drug resistance in Mycobacterium tuberculosis, Nat. Biotechnol.,16, 359, 1998.46. Tyagi, S. and Kramer, F.R., Molecular beacons: probes that fluoresce upon hybridization, Nat. Biotechnol.,14, 303, 1996.© 2006 by Taylor & Francis Group, LLC

CELLULAR, BIOCHEMICAL, AND MOLECULAR TECHNIQUES IN DEVELOPMENTAL TOXICOLOGY 61947. Pfaffl, M.W., A new mathematical model for relative quantification in real-time RT-PCR, NucleicAcids Res., 29, E45, 2001.48. Vanden Heuvel, J.P., Tyson, F.L., and Bell, D.A., Construction of recombinant RNA templates foruse as internal standards in quantitative RT-PCR, Biotechniques, 14, 395, 1993.49. Abbott, B.D., Held, G.A., Wood, C.R., Buckalew, A.R., Brown, J.G., and Schmid, J., AhR, ARNT,and CYP1A1 mRNA quantitation in cultured human embryonic palates exposed to TCDD andcomparison with mouse palate in vivo and in culture, Toxicol. Sci., 47, 62, 1999.50. Abbott, B.D., Schmid, J.E., Brown, J.G., Wood, C.R., White, R.D., Buckalew, A.R., and Held, G.A.,RT-PCR quantification of AHR, ARNT, GR, and CYP1A1 mRNA in craniofacial tissues of embryonicmice exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin and hydrocortisone, Toxicol. Sci., 47, 76, 1999.51. Vanden Heuvel, J.P., Clark, G.C., Kohn, M.C., Tritscher, A.M., Greenlee, W.F., Lucier, G.W., andBell, D.A., Dioxin-responsive genes: examination of dose-response relationships using reverse-transcriptionpolymerase chain reaction, Cancer Res., 54, 62, 1994.52. Eberwine, J., Spencer, C., Miyashiro, K., Mackler, S., and Finnell, R., Complementary DNA synthesisin situ: methods and applications, Methods Enzymol., 216, 80, 1992.53. Harnish, D.C., Jiang, H., Soprano, K.J., Kochhar, D.M., and Soprano, D.R., Retinoic acid receptorbeta 2 mRNA is elevated by retinoic acid in vivo in susceptible regions of mid-gestation mouseembryos, Dev. Dyn., 194, 239, 1992.54. Harnish, D.C., Soprano, K.J., and Soprano, D.R., Mouse conceptuses have a limited capacity to elevatethe mRNA level of cellular retinoid binding proteins in response to teratogenic doses of retinoic acid,Teratology, 46, 137, 1992.55. Bronnegard, M. and Okret, S., Regulation of the glucocorticoid receptor in fetal rat lung duringdevelopment, J. Steroid Biochem. Mol. Biol., 39, 13, 1991.56. Mercola, M., Wang, C.Y., Kelly, J., Brownlee, C., Jackson-Grusby, L., Stiles, C., and Bowen-Pope,D., Selective expression of PDGF A and its receptor during early mouse embryogenesis, Dev. Biol.,138, 114, 1990.57. Papalopulu, N., Lovell-Badge, R., and Krumlauf, R., The expression of murine Hox-2 genes isdependent on the differentiation pathway and displays a collinear sensitivity to retinoic acid in F9cells and Xenopus embryos, Nucleic Acids Res., 19, 5497, 1991.58. Wang, C., Kelly, J., Bowen-Pope, D.F., and Stiles, C.D., Retinoic acid promotes transcription of theplatelet-derived growth factor alpha-receptor gene, Mol. Cell Biol., 10, 6781, 1990.59. Sassani, R., Bartlett, S.P., Feng, H., Goldner-Sauve, A., Haq, A.K., Buetow, K.H., and Gasser, D.L.,Association between alleles of the transforming growth factor-alpha locus and the occurrence of cleftlip, Am. J. Med. Genet., 45, 565, 1993.60. Chenevix-Trench, G., Jones, K., Green, A.C., Duffy, D.L., and Martin, N.G., Cleft lip with or withoutcleft palate: associations with transforming growth factor alpha and retinoic acid receptor loci, Am.J. Hum. Genet., 51, 1377, 1992.61. Capecchi, M.R., The new mouse genetics: altering the genome by gene targeting, Trends Genet., 5,70, 1989.62. Hunt, P., Gulisano, M., Cook, M., Sham, M.H., Faiella, A., Wilkinson, D., Boncinelli, E., and Krumlauf,R., A distinct Hox code for the branchial region of the vertebrate head, Nature, 353, 861, 1991.63. Sham, M.H., Hunt, P., Nonchev, S., Papalopulu, N., Graham, A., Boncinelli, E., and Krumlauf, R.,Analysis of the murine Hox-2.7 gene: conserved alternative transcripts with differential distributionsin the nervous system and the potential for shared regulatory regions, EMBO J., 11, 1825, 1992.64. Marshall, H., Nonchev, S., Sham, M.H., Muchamore, I., Lumsden, A., and Krumlauf, R., Retinoicacid alters hindbrain Hox code and induces transformation of rhombomeres 2/3 into a 4/5 identity,Nature, 360, 737, 1992.65. Balkan, W., Colbert, M., Bock, C., and Linney, E., Transgenic indicator mice for studying activatedretinoic acid receptors during development, Proc. Natl. Acad. Sci. USA, 89, 3347, 1992.66. Zimmer, A., Induction of a RAR beta 2-lacZ transgene by retinoic acid reflects the neuromericorganization of the central nervous system, Development, 116, 977, 1992.67. Balkan, W., Klintworth, G.K., Bock, C.B., and Linney, E., Transgenic mice expressing a constitutivelyactive retinoic acid receptor in the lens exhibit ocular defects, Dev. Biol., 151, 622, 1992.68. Lee, G.H., Merlino, G., and Fausto, N., Development of liver tumors in transforming growth factoralpha transgenic mice, Cancer Res., 52, 5162, 1992.© 2006 by Taylor & Francis Group, LLC

620 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION69. Mann, G.B., Fowler, K.J., Gabriel, A., Nice, E.C., Williams, R.L., and Dunn, A.R., Mice with a nullmutation of the TGF alpha gene have abnormal skin architecture, wavy hair, and curly whiskers andoften develop corneal inflammation, Cell, 73, 249, 1993.70. Erickson, H.P., Gene knockouts of c-src, transforming growth factor beta 1, and tenascin suggestsuperfluous, nonfunctional expression of proteins, J. Cell Biol., 120, 1079, 1993.71. Kulkarni, A.B., Huh, C.G., Becker, D., Geiser, A., Lyght, M., Flanders, K.C., Roberts, A.B., Sporn,M.B., Ward, J.M., and Karlsson, S., Transforming growth factor beta 1 null mutation in mice causesexcessive inflammatory response and early death, Proc. Natl. Acad. Sci. USA, 90, 770, 1993.72. Mansour, S.L., Goddard, J.M., and Capecchi, M.R., Mice homozygous for a targeted disruption ofthe proto-oncogene int-2 have developmental defects in the tail and inner ear, Development, 117, 13,1993.73. Bryant, P.L., Schmid, J.E., Fenton, S.E., Buckalew, A.R., and Abbott, B.D., Teratogenicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in mice lacking the expression of EGF and/or TGF-alpha, Toxicol.Sci., 62, 103, 2001.74. Hanahan, D., Signaling vascular morphogenesis and maintenance, Science, 277, 48, 1997.75. Maltepe, E., Keith, B., Arsham, A.M., Brorson, J.R., and Simon, M.C., The role of ARNT2 in tumorangiogenesis and the neural response to hypoxia, Biochem. Biophys. Res. Commun., 273, 231, 2000.76. Kozak, K.R., Abbott, B., and Hankinson, O., ARNT-deficient mice and placental differentiation, Dev.Biol., 191, 297, 1997.77. Crooke, R.M., In vitro toxicology and pharmacokinetics of antisense oligonucleotides, AnticancerDrug Des., 6, 609, 1991.78. Slavkin, H.C., Sasano, Y., Kikunaga, S., Bessem, C., Bringas, P., Jr., Mayo, M., Luo, W., Mak, G.,Rall, L., and Snead, M.L., Cartilage, bone and tooth induction during early embryonic mouse mandibularmorphogenesis using serumless, chemically-defined medium, Connect. Tissue Res., 24, 41,1990.79. Augustine, K., Liu, E.T., and Sadler, T.W., Antisense attenuation of Wnt-1 and Wnt-3a expression inwhole embryo culture reveals roles for these genes in craniofacial, spinal cord, and cardiac morphogenesis,Dev. Genet., 14, 500, 1993.80. Branch, S., Francis, B.M., Rosen, M.B., Brownie, C.F., Held, G.A., and Chernoff, N., Differentiallyexpressed genes associated with 5-aza-2′-deoxycytidine-induced hindlimb defects in the Swiss Webstermouse, J. Biochem. Mol. Toxicol., 12, 135, 1998.81. Owens, G.P., Hahn, W.E., and Cohen, J.J., Identification of mRNAs associated with programmed celldeath in immature thymocytes, Mol. Cell Biol., 11, 4177, 1991.82. Bennett, G.D., An, J., Craig, J.C., Gefrides, L.A., Calvin, J.A., and Finnell, R.H., Neurulation abnormalitiessecondary to altered gene expression in neural tube defect susceptible Splotch embryos,Teratology, 57, 17, 1998.83. Finnell, R.H., Van Waes, M., Bennett, G.D., and Eberwine, J.H., Lack of concordance between heatshock proteins and the development of tolerance to teratogen-induced neural tube defects, Dev. Genet.,14, 137, 1993.84. Musselman, A.C., Bennett, G.D., Greer, K.A., Eberwine, J.H., and Finnell, R.H., Preliminary evidenceof phenytoin-induced alterations in embryonic gene expression in a mouse model, Reprod. Toxicol.,8, 383, 1994.© 2006 by Taylor & Francis Group, LLC

CHAPTER 15Functional Genomics and Proteomics inDevelopmental and Reproductive ToxicologyOfer Spiegelstein, Robert M. Cabrera, and Richard H. FinnellCONTENTSI. Introduction ........................................................................................................................622II. Gene Expression ................................................................................................................622A. Background................................................................................................................622B. Microarrays................................................................................................................6231. Complementary DNA (cDNA) Arrays................................................................6232. Oligonucleotide Microarrays...............................................................................6283. Data Analysis and Bioinformatics.......................................................................6294. Experimental Considerations...............................................................................631C. Serial Analysis of Gene Expression..........................................................................6331. General.................................................................................................................6332. Methodology........................................................................................................6333. Limitations ...........................................................................................................6354. SAGE and Developmental Biology.....................................................................635III.Proteomics..........................................................................................................................636A. Background................................................................................................................636B. Protein Expression Analysis......................................................................................6361. Protein Isolation...................................................................................................6362. Protein Identification ...........................................................................................6383. Protein Arrays ......................................................................................................640IV. Predictive Values ................................................................................................................640V. Relevant Websites and Databases......................................................................................641A. Microarrays................................................................................................................641B. SAGE .........................................................................................................................641C. Proteomics..................................................................................................................641References ......................................................................................................................................642621© 2006 by Taylor & Francis Group, LLC

622 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONI. INTRODUCTIONOngoing worldwide efforts continue to produce sequences for genomic assembly and single nucleotidepolymorphism (SNP) typing of DNA, from model organisms to humans, with the initial draftof the human genome having been published in 2001. 1,2 The significance of genetic information isderived from the transcription of genomic DNA into messenger ribonucleic acids (mRNA) thatencode the amino acid sequences of translated proteins. The process culminates in posttranslationalmodifications that are required to form biologically active and functional proteins that participateas enzymes in metabolic pathways. To understand the factors involved in dynamics of biologicalactivity, the intricate relationships between DNA sequences (genomics), RNA levels (functionalgenomics), and expressed proteins (functional proteomics) must be examined. 3,4 Developmentaland reproductive toxicological phenomena in model organisms can also be discerned by carefulexamination at these three fundamental levels. Environmental and pharmaceutical agents can induceabnormal expression of genes, which, should it occur at specific gestational time points of highsusceptibility, may result in abnormal developmental processes that culminate in the birth of aninfant with congenital defects. Such deleterious effects may be examined by techniques that aredescribed in this chapter. Another level of topical organization to study involves protein-toxicantinteractions. This level of inquiry seeks to elucidate altered protein presence, altered binding, orconformational changes that interfere with or damage developmental biological processes.This chapter provides an overview of information and procedures used by contemporary investigatorsto study functional genomics and proteomics. These new cutting edge tools provide themeans to explore the etiology of developmental and reproductive toxicants. Each topic is presentedin such a way as to provide the reader not only with the background information but also with thetheory and assay limitations that can help make practical decisions about the application of thesenew techniques.A. BackgroundII. GENE EXPRESSIONGene expression studies may be approached using several different RNA-based technologies:• In situ hybridization (ISH)• Northern blotting• Ribonuclease protection assays (RPA)• Gene reporter assay• Branched DNA amplification• Reverse transcription–polymerase chain reaction (RT-PCR)• Real-time quantitative RT-PCR• Differential display• Scintillation proximity assay• Rapid analysis of gene expression (RAGE)• Serial analysis of gene expression (SAGE)• MicroarraysEach of these methods provides the means to gather information on gene expression levels.Specific experiments must be designed around the limitations of each assay and the type of dataneeded to fulfill the aims of a given study. To study complex biological processes at the molecularlevel, these approaches are best used in combination. For instance, microarrays can initially beused to screen thousands of both previously characterized and unknown genes for changes inexpression following exposure to a model compound. Once genes of interest have been identified,© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 623the data should be confirmed and validated by additional approaches, such as real time-PCR orNorthern blotting.With these tools now readily available to the scientific community, gene expression profilesthat are characteristic of modifications caused by developmental or reproductive toxicants can andshould be identified. Experimental designs presently adhere to traditional teratogen testing methods,i.e., comparing patterns of gene expression in experimental samples vs. those observed in controls.This section specifically reviews two recent methods for global gene expression profiling, namelymicroarrays and SAGE, that may be used in toxicology studies. With these tools, global changesin gene expression profiles can be identified and characterized. These are critical first steps forunderstanding the mechanisms of action of reproductive or developmental toxicants or for predictivepurposes in the identification of potential teratogens.B. MicroarraysThe platform typically associated with gene expression profiling is the microarray, also known asthe “DNA chip” or “gene chip.” 5 Microarrays come on different platforms and can be custommanufactured in-house or purchased from selected biotechnology companies. The common themeconcerning microarrays is that scientists are able to study the expression of a large number of genessimultaneously in a single experiment. Experiments that in the past would have taken several yearsto complete can now be performed within a matter of days. Aside from the high throughput natureof the microarray technology, one of the most prominent features is the ability to study novel andpreviously uncharacterized genes. Thereby previously unknown genes or genes never before linkedto the experimental questions being pursued may be implicated. This, in turn, provides opportunitiesto target these newly implicated genes as part of drug discovery processes or to study their function,regulation, and cellular networks. In toxicological studies, discovery of novel genes that are perturbedin response to toxicological stimuli may enhance understanding of the underlying mechanismsof toxicity associated with a specific compound.In retrospect, microarray methodology is merely a recent extension of well-established methodsto assess gene expression by hybridization. Such methods have been successfully applied for manyyears. The first hybridization-based method, Southern blotting, 6 was used to study only a limitedset of genes. Subsequent methodologies such as “dot-blots,” enabled investigators to explore alarger set of genes that are deposited on porous nylon membranes. 7 What distinguishes microarraysfrom the aforementioned methods is primarily the use of glass as the solid support. Contrary tonylon membranes, glass (being a nonporous material) enables high-density and spatially reproducibledeposition of target DNA, resulting in the ability to screen thousands of genes simultaneously.Using a nonporous support has also simplified hybridization kinetics and improved image acquisitionand processing procedures. 8 Concurrent to developments in fluorescent labeling of nucleicacids, microarrays enable scientists to evaluate gene expression from both experimental and controlsamples on the same slide simultaneously. There are two major types of DNA arrays: complementaryDNA arrays and oligonucleotide arrays. Several features, such as the length of target DNA,the method for DNA deposition, and the slide chemistry, distinguish the two types of arrays.1. Complementary DNA (cDNA) ArraysComplementary DNA arrays, better known as cDNA arrays, were initially developed in the laboratoryof Patrick Brown and colleagues from Stanford University. Their first publication describingthe use of this technology with a small subset of Arabidopsis genes appeared in 1995. 9 Since then,cDNA arrays have become the most popular microarray platform, and it is now quite common forinstitutions and departments to have their own core microarray manufacturing facilities. The mainreasons for the popularity of cDNA arrays are:© 2006 by Taylor & Francis Group, LLC

624 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1. The DNA deposited on the solid support can be easily prepared by investigators using standardPCR amplification.2. For producers of their own arrays, sets of cloned genes can be purchased from biotechnologycompanies, obtained from the National Institutes of Health (NIH), or prepared in-house. Thus,researchers have a great deal of control over the content of the arrays they make and use.3. The cost of preparing cDNA arrays in-house is substantially lower than the cost of commerciallyavailable arrays. Once the basic equipment for array fabrication has been obtained and the targetDNA prepared, hundreds of microarray slides can be produced in a relatively short time.a. Array FabricationThe choice of which genes or expressed sequence tags (ESTs) to use for cDNA array fabricationdepends on their availability; however, all deposited target DNA is prepared by standard PCRamplification of cDNA cloned into bacterial plasmids. Individually cloned genes and ESTs comein sets of several hundred to several thousand. They are either commercially available, obtainedfree of charge (e.g., from the NIH), or they may even be developed within the institution. SeeFigure 15.1 for an outline of cDNA microarray fabrication.Comprehensive collections of genes and ESTs are particularly useful for applications such asgene discovery, when the goal is to associate specific genes with a specific toxicological process.Several cDNA clone sets are commercially available. 10 These include: the BMAP clone set (Res-Gen, Carlsbad, CA), which was derived from adult C57BL6J mouse brain, spinal cord, and retinaltissues; arrayTAG (Lion Biosciences, San Diego, CA), which is a family of cDNA clones (mouse,rat, and dog) that were specifically designed for microarray applications by use of several uniquemethodologies; the full-length mouse cDNA clones from the RIKEN Genomic Sciences Center(Yokohama, Japan); and more.Two of the most useful clone sets for developmental biologists and developmental toxicologistsalike, the mouse 15K and 7.4K sets, were assembled at the National Institute of Aging. The 15Kclone set was constructed to specifically represent developmentally important mammalian genes. 11,12The approximately15,000 genes in this set were derived from pre- and peri-implantation embryosand from placenta. Between a third to a half of the 15,000 set contains known and previouslycharacterized genes, whereas the reminder is composed of ESTs of currently unknown function. 11The second clone set, released in 2002, is the 7400 clone set that contains approximately 7400genes with no redundancy within the set or with the original 15,000 set. The 7400 clone set isespecially valuable for developmental biologists and toxicologists as it includes genes expressedin a variety of stem cells, early embryos, and newborn organs, such as the heart and brain.Preparing in-house clone sets requires considerable molecular biology expertise and resources;however, it enables researchers to construct microarray slides with a tissue-specific set of genes orwith genes that are expressed during specific developmental processes, such as palatal shelf fusion.Using the in-house approach, investigators studying eye development created a library of approximately15,000 ESTs that were obtained from mouse retina at embryonic day 14.5. 13 By use ofmicroarray slides that were prepared from this clone set, a potential regulatory association betweenthe transcription factor Brn-3b and GAP-43, a protein associated with axonal growth, has beenidentified.Similarly, at the National Center for Toxicogenomics (NCT), which is part of the NationalInstitute of Environmental Health Sciences (NIEHS), a toxicological human cDNA chip called theToxChip has been developed. 14,15 The 2000 genes arrayed on the early version of the ToxChip(version 1.0) have been previously implicated in mechanisms of toxicity and cellular processes inresponse to different types of toxic insults. The latter include genes involved in apoptosis, DNArepair, and drug metabolism, transcription factors, tumor suppressor genes, and more. A similarapproach was used to evaluate the effects of the environmental toxicant β-naphthoflavone on miceby use of a toxicological gene array and Northern blotting. 16 The authors compared the two methods© 2006 by Taylor & Francis Group, LLC

Figure 15.1IAAAAAAAEmbryo mRNA mRNA : cDNAIIMicroarray SlideTTTTTTTAAAAAAA#1 #2 #3Cloned cDNATransformed BacteriaOutline of cDNA microarray preparation. (I)Isolation of mRNA from the tissue of interest, (II) reverse transcription of mRNAto cDNA, (III) cloning of cDNA into expression vectors (plasmids), (IV) transformation of bacteria to contain single plasmidinserts, (V) ampli cation of cDNA from plasmids b y PCR, and (VI) high density deposition of puri ed target DNA on glassslides.IIIVIIVVAmpliconFUNCTIONAL GENOMICS AND PROTEOMICS 625© 2006 by Taylor & Francis Group, LLC

626 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONand concluded that the sensitivity of microarrays is comparable to that of Northern blotting. Thedata variability introduced by the printing process was less than the interanimal (biological)variability observed in the study.After the target DNA has been amplified and purified by use of standard protocols, the DNAis dissolved in an arraying buffer. Several types of arraying buffers are commonly used, usuallybased on saline sodium citrate (SSC) and sodium dodecyl sulfate (SDS). However, some prefer touse printing buffers based on dimethyl sulfoxide (DMSO); evaporation from both the printed glasssurface and the microtiter plates is greatly reduced, thereby improving spot homogeneity. 17–19 Thetarget DNA solution, stored in either 96- or 384-well microtiter plates, is printed (arrayed) bymicroarray printers on specially treated glass microscope slides. Commonly used slide-coatingmaterials are poly-L-lysine, epoxysilane, aminosilane, and aldehyde. 17,18 All types of pretreatedmicroarray slides are commercially available, but they can also be prepared in-house with regularglass microscope slides. These special slide coatings are important for optimizing the slide’s surfacearea to promote the adhesion of target DNA, thereby preventing lateral dispersion or detachmentduring hybridization.Microarray printers are equipped with specialty printing tips that have the capability to pickup a submicroliter volume of target DNA solution and deposit nanoliter droplets with exceptionallyhigh accuracy and precision. A typical slide may contain several thousands of target DNA spots,100 to 200 µm in diameter, spaced 100 to 250 µm apart. When a large (usually exceeding 10,000)set of genes is to be printed, each target DNA can be arrayed once to maximize the total numberof genes represented on the slide. However, when a relatively small set of genes is to be printed,each target DNA can be arrayed multiple times at separate locations of the slide. This facilitatesstatistical analysis of the data, thereby increasing sensitivity and confidence in the results. 20After the slides have been printed, several additional steps are necessary prior to their use inexperiments. cDNA slides are often rehydrated to distribute the DNA more evenly on the glasssurface. Slides are also irradiated by UV light to cause cross-linking of DNA so that it remainsattached to the surface during hybridization and washing. In the final step of array fabrication, theslide’s surface is deactivated (blocked) in order to minimize nonspecific binding of the probe DNA,thereby decreasing background signal levels, increasing the signal-to-noise ratio (SNR), andimproving the overall sensitivity of detection.b. Probe PreparationFrom the RNA sample, a labeled cDNA probe is prepared by standard reverse transcriptionmethodology. A radiolabeled probe is often used with microarrays printed on membranes becausethe probe enables high detection sensitivity. However, membrane microarrays are not as popularas glass microarrays in which the probes are usually labeled with a fluorophore (fluorescent tag).Use of fluorescently labeled dNTPs is the most popular approach, primarily because of the abilityto cohybridize two samples simultaneously, which is not possible with radiolabeled targets. Theability to hybridize two samples simultaneously is a tremendous advantage. For example, one couldlabel a sample from the developing neural tube of a drug-treated mouse embryo with one fluorophoreand label a control sample with a different fluorophore. Thus, on a single microarray slide, adevelopmental toxicologist can examine the isolated effects of in utero exposure to a compoundand distinguish the altered patterns of gene expression that may result. See Figure 15.2 for anoutline of probe preparation.The two most widely used fluorescent dNTPs are the CyDyes (Amersham Biosciences Corp,Piscataway, NJ), namely, Cy-3-dUTP (Cy-3) and Cy-5-dUTP (Cy-5). In both dyes, the fluorophoresattached to the dUTP backbone have unique chemical structures that are spectrally distinctfrom one another. This enables the simultaneous detection of both dyes without cross interference,as it allows the concomitant measurement of signal from both the control and experimental samples.© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 627Treated EmbryoControl EmbryoIImRNAAAAAAAAAAAAAAAmRNAIIIIAAAAAAATTTTTTTmRNA : cDNAAAAAAAATTTTTTTmRNA : cDNAIIITTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTIIIIVFigure 15.2Outline of probe preparation and hybridization to microarrays. (I) Extraction of RNA from controland drug-treated embryos, (II) reverse transcription of mRNA to fluorescent cDNA probes, (III)probes are mixed and cohybridized to a microarray slide, and (IV) differential gene expressionis visualized after the slide is processed and scanned.The relative spectral intensities of the two samples are then obtained, and the transcriptional changesthat occurred in response to the experimental treatment are determined by calculating the signalratios for each gene. Other fluorescently tagged dNTPs are also available, such as the Alexa Fluor ®series of dyes (Molecular Probes, Eugene, OR). The Alexa dyes seem to be somewhat superior tothe CyDyes because of their higher SNRs and better chemical stability. 21,22© 2006 by Taylor & Francis Group, LLC

628 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThese dyes can be incorporated using “direct labeling” of cDNA, i.e., the fluorescently taggeddUTPs are directly incorporated into the probe DNA as it is being synthesized from RNA. Bothof these tagged nucleotides have the ability to cause premature termination of target DNA synthesis.That is because the large size of the fluorophore may cause steric hindrance, resulting in detachmentof the reverse transcriptase enzyme. An additional problem with the direct labeling technique isthat the differential incorporation ratios of each dye have the potential to introduce bias in dataanalysis.Since the direct labeling methodology has several pitfalls, an alternative method called “indirectlabeling” has been developed. In this technique, the fluorescent tag is conjugated after the probeDNA has been synthesized. Indirect labeling is achieved by the use of aminoallyl-dUTP in theinitial reverse transcription step. This chemical modification creates an activated dUTP moleculeto which the fluorescent dye can later be covalently attached. The chemical modification of thedUTP does not interfere with reverse transcription and nor does it cause premature termination ofcDNA synthesis. In addition, the problem of differential incorporation ratios into cDNA is eliminated.23 After the probe DNA has been synthesized, monoreactive Alexa or CyDyes are conjugatedto the aminoallyl-dUTP groups to make the fluorescent probe.c. Hybridization, Washing, and Image AcquisitionAfter the two probe samples have been prepared from both an experimental and control sample,each labeled with a different fluorescent dye, they are mixed together in a hybridization buffer. Thehybridization buffer usually contains SSC or formamide with blocking agents. Blocking agents thatare typically used are Cot-1 DNA, poly-dA oligos, and/or salmon sperm DNA. The purpose ofthese blocking agents is to minimize nonspecific binding of probe DNA to the slide. Hybridizationis performed at elevated temperatures, ranging from 40°C to 70°C depending on the hybridizationbuffer’s composition, the average length of the probe DNA, and the stringency requirements. Inaddition, hybridization is best accomplished in specially designed chambers that maintain optimalhumidity to prevent evaporation of water from the hybridization solution. Typically, the volume ofthe hybridization solution is 10 to 40 µl; therefore, even a small loss of volume may result inrelatively large changes in salt concentration, which may adversely affect the quality of the hybridization.Collectively, buffer composition and hybridization temperature will determine the stringencyof hybridization and the resulting signal and background fluorescence intensities.Hybridization time can range from a few hours to overnight, after which the slide is washedto remove excess hybridization buffer and unattached probe DNA. Buffers of different stringenciesare used in this step to reduce background signal. After the slide is dried, it is scanned by speciallaser scanners that detect the emitted signals from two separate channels (532 and 635 nm in thecase of Cy-3-dUTP and Cy-5-dUTP, respectively), where each channel represents the signal fromone of the fluorescent dyes.2. Oligonucleotide MicroarraysThe second form of DNA microarrays is the oligonucleotide array, of which two types currentlyexist. Affymetrix® (Santa Clara, CA) pioneered the development of the first type of oligonucleotidearray, called the GeneChip ® , which is made by photolithography technology and solid-phase DNAsynthetic chemistry. 24 These combined methods are similar to techniques used in combinatorialchemistry synthesis. They enable the simultaneous synthesis of approximately 300,000 different25-mer oligos at high density and high spatial resolution, covalently attached to a solid glasssupport. 25 The oligomers’ sequences are the target DNA and are designed exclusively on DNAsequence information. They do not require any clone sets or PCR procedures, as is necessary withcDNA arrays. For each gene represented on the GeneChip, 20 different 25-mer sequences areselected. These represent different areas along the coding sequence of the gene that serve as sensitive© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 629sequence-specific sensors of expression. These sequences are optimized so that there is minimaloverlap with other genes, gene-family members, or interfering genomic elements, such as repetitivesequences. The rational for using this probe redundancy (multiple sequences within a single gene)is to create improved accuracy and SNRs, minimizing false positives, false negatives, and nonspecificbinding. 25 Another important feature of the GeneChip is the use of 20 “mismatch control”sequences. They are identical to their corresponding “true sequences” with the exception of a singlebase difference. These mismatch sequences act as specificity controls for the true sequences, therebyallowing reduction of background and cross-hybridization signals and allowing increased confidencein the results. It is important to note that while the GeneChip technology offers an excitingmicroarray platform with which to work, researchers must purchase proprietary hardware, software,and the GeneChips from Affymetrix. As previously mentioned, cDNA arrays offer the advantageof being amenable to customized, in-house design and the production of multiple slides, whereasusing GeneChips is restrictive in the gene content and is substantially more expensive.The second type of oligonucleotide microarray is a hybrid between cDNA and the previouslydescribed GeneChip oligonucleotide array platform. The target DNA oligomers are prepared usingstandard DNA synthetic methods, and the considerations for designing the oligomers’ sequencesare essentially the same as with the GeneChip. The oligos are then printed in-house on pretreatedglass slides, similar to cDNA arrays. The oligo array method offers the advantages of specificityinherent in preselected short DNA sequences, plus the ability to prepare multiple slides containingthe same set of DNA sequences, thus substantially reducing costs. However, this type of arrayplatform does not quite have the highly controlled and reproducible nature of the GeneChip andnor does it have the probe redundancy and specificity control sequences.3. Data Analysis and Bioinformaticsa. NormalizationNormalization of data can be performed in numerous ways, depending on the experimental design.The goal is to extract meaningful information and expand the investigator’s ability to make comparisonsbetween data sets. The immediate effect of normalization is to correct systematic hybridizationvariation so that the data more closely represent the true biological variation. Generally,normalization methods can be broken into three categories: global normalization, single gene(housekeeping gene) normalization, and external control normalization. 26Global normalization assumes that the expression of only a small proportion of the genes willvary significantly between the experimental and control samples. 27 Global normalization is the mostwidely used and is based on the overall median or mean of spot signal intensity. 28–30 It is worthnoting that global normalization is not effective when working with transcripts that are expressedat low levels. 31 There is also evidence of spatially or intensity dependent dye biases in numerousexperiments that are not considered and go uncorrected in global normalization. It is important tounderstand that while global normalization relies on the generally correct assumption that the twofluorescent dye intensities are related by a constant factor, there can be spatial variations in signaland background intensities that are not displayed evenly by both dyes. Furthermore, the dyes donot necessarily share common emission curves; therefore, the factor that relates the two dyes canchange according to their intensity. This results in a dye-bias that often becomes apparent in theupper and lower bounds of dye emission intensities.A set of housekeeping genes (e.g., β-actin, glyceraldehyde-3-phosphate dehydrogenase, ubiquitin)that are believed to maintain constant expression across a variety of conditions can also beused in normalization. 26 Although using single gene normalization does not require the assumptionthat only a small proportion of the genes varies significantly, it does require the assumption thatthe expression of the housekeeping gene used for normalization is not significantly altered by theexperimental conditions. However, not all housekeeping genes are equally expressed under all© 2006 by Taylor & Francis Group, LLC

630 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONconditions, and experiment-specific genes thus need to be identified and evaluated. Only after agene has been shown to have constant expression under the relevant experimental conditions canit be used to normalize microarray data. Another limitation of using housekeeping genes is thatthey tend to be highly expressed, and their normalization parameters, when applied to other genes,may not be representative of true variation. 31An alternative method for normalizing gene expression data is the use of spiked-in controls. 26,32These externally added genes could be used either as positive or negative controls or as a titrationseries, thereby enabling normalization to account for intensity-based variation. To incorporate thesecontrols into a microarray experiment, a cDNA or oligonucleotide target DNA must be added tothe array at the printing step. The spiked controls and their corresponding capture targets must beobtained from a different species than the organism of study so that they are composed of nonhomologoussequences. Common sources of spike control DNA are Saccharomyces cerevisiae andArabidopsis thaliana. Their use minimizes cross hybridization with, for example, DNA of themouse and rat, which are conventional species used in developmental and reproductive toxicologystudies. Generating external foreign target DNA often entails PCR amplification from genomicsamples by use of primers that contain a T7-RNA polymerase promoter sequence. That enablesthe subsequent synthesis of complementary RNA (cRNA) in an in vitro transcription (IVT) reaction.An additional advantage to using this method is that it introduces a basis for quantification of theabsolute abundance of individual mRNAs. That is not possible with any of the other normalizationmethods. 27,33,34 This approach has the added benefit of allowing the comparison of multiple arrayexperiments, if investigators use the same control sets. However, the use of spiked controls is notproblem-free, and it has been shown that there are a number of uncertainties when extrapolatingfrom the spiked control to sample derived mRNA. 27Debate on the most effective method or “hybrid” method of normalization continues. Afterconsiderable testing, it is currently accepted as being both plausible and feasible to incorporatemore than a single method to normalize microarray data. 26,35 Despite the debate, the primary goalof normalization remains the same. Assessment of alternatives to the widely used global normalizationprocedure depends on whether they expand the investigator’s ability to compare data sets.Ideally, a procedure should account for or correct systematic hybridization variation so that arraydata more closely represents actual biological activity.b. ClusteringAfter data have been generated and normalized, an appropriate analytic method must be selected.Presently, the acceptable basis for analyzing and organizing (clustering) gene expression data is bygrouping genes with similar patterns of expression, similar function, or similar regulatory elements.36–39 These grouping methods address different key questions, such as which genes sharecommon expression patterns, which are expressed or controlled in a coordinated manner acrossdifferent sets of conditions, and which share related biological pathways. The goal of clusteringculminates in the latter question: what are the dynamics of the genetic network?There are several different clustering methods that can be used to organize gene expressiondata: hierarchical clustering, 36,40 k-means clustering, 41 principal components analysis, 42 mixturemodel methodology, 43 multidimensional scaling, 44 self-organizing maps, 45 graph-theoretic techniques,46 and a few additional less-well known approaches. 47–49 These methods often providegraphical representations of the data in dendrograms, which are branched treelike graphical structuresused to represent a hierarchy. Ratio-based coloring is also used in dendrograms to quantitativelyand qualitatively represent the experimental observations, the observed changes in geneexpression. Such clustering methods are capable of demonstrating expression profiles of developmentaltime points, tissue specific expression, or the changes in expression that occur followingexposure to a drug. The basic format of the various clustering methods assembles genes in matricesaccording to criteria applied by an algorithm. For example, rows may represent the genes spotted© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 631on the microarray, while columns represent the samples used to hybridize to the array. The signalratios, as measures of relative changes in gene expression between experimental and controlsamples, are depicted according to a color-based scale. By convention, up-regulation of genes inthe experimental samples appears in red, whereas down-regulation is green. No change in geneexpression is represented by yellow, whereas gray or black cells indicate missing or excludeddata. 36,504. Experimental Considerationsa. Limited RNA AvailabilityTypically, running a single sample on a microarray slide requires 20 to 100 µg of total RNA. Forthose working with large tissue samples or cell cultures, these RNA requirements are usually easyto meet. However, if the biological sample of interest is derived from a very small tissue sample,such as a portion of the neural tube of a neurulation-stage mouse embryo, the amount of total RNAthat can be obtained is limited and will most likely be less than 1 µg. For those using laser-capturemicrodissection, single-sample RNA yields may be in the low nanogram range. To overcome suchproblems, it may be possible to pool a large number of individual samples so that the combinedRNA content is sufficient for labeling and hybridization. While appropriate under specific experimentalsituations, this approach is likely to mask small and subtle changes in gene expression (i.e.,have reduced sensitivity) due to a dilution effect. Most importantly, pooling of samples eliminatesthe ability to measure the biological variability that is a particularly important experimental parameterwhen working with microarrays.An alternative approach to overcome the problem is to use labeling methods that increase thespecific activity of the probes, thus compensating for the limited quantities of RNA. Examples ofsuch methods are the Micromax ® Tyramide Signal Amplification (TSA) technique (NEN LifeScience Products, Zaventem, Belgium), the HiLight Array Detection System (QIAGEN, Valencia,CA), and the 3DNA Dendrimer technology (Genisphere ® , Hatfield, PA), One must keep in mindthat none of these techniques has emerged over time as the single most powerful and reliableapproach. 23,51–53Another option to deal with limited RNA availability in gene expression experiments has beendeveloped by James Eberwine at the University of Pennsylvania. His method is based on linearamplification of mRNA by IVT from cDNA. 54,55 The cDNA used for IVT is made by reversetranscription of mRNA by use of a poly-dT oligonucleotide with a T7 RNA polymerase promotersequence. Thus, by utilizing the well-known ability of T7 RNA polymerase to bind to its promoterand synthesize RNA at a constant rate of approximately 100 bases/sec, linear amplification of RNAis achieved with high fidelity. The extent of amplification depends on the amount of template used.It is typically 40 to 200-fold with relatively large amounts of template and up to 2000-fold whenthe starting amount of template is very small, such as that obtained from single cells. 54,56This method has been successfully used for determining changes in gene expression in mouseembryos following administration of several known teratogens. In these studies, dot blots containingup to 50 genes of interest were prepared by immobilizing cDNA on nylon membranes. The probesused in these experiments were prepared by the T7-based IVT method and labeled with a radioactivenucleotide. With this approach, it has been shown that there is a significant down-regulation in theexpression of several folate pathway genes (e.g., folate binding protein-1 and methylene tetrahydrofolatereductase) in neurulation stage neural tube defect–sensitive SWV mouse embryos followingin utero exposure to valproic acid. 57 In contrast, the more resistant LM/Bc strain embryoshad an up-regulation of these same genes in response to the teratogenic valproate exposure. Theseobservations provided a plausible mechanism for relative resistance to valproate-induced neuraltube defects. Similar experimental designs were used to study additional teratogen-induced changesin expression of growth and transcription factors as well as cell cycle checkpoint genes. 58–61© 2006 by Taylor & Francis Group, LLC

632 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAs previously described, microarray technology has overtaken the laborious low-throughputdot-blot technology; however, the basic experimental principles and technical considerations remainthe same. Thus, T7 RNA polymerase–based IVT can be integrated into microarray experiments toexamine changes in gene expression following in utero exposure to a variety of known and suspectedteratogens. Currently, only a handful of published studies used T7 RNA polymerase amplificationin conjunction with microarrays, and all were nonreproductive toxicology studies; however, theyserve to emphasize the feasibility of coupling the two techniques to approach these criticallyimportant issues. 62–64b. Importance of ReplicatesThe data generated from microarrays are subject to the variability inherent in all experimentalsystems. One source of this variability is introduced by biological factors, such as the differencesin response between individual animals. Additional variables, such as slide-surface inconsistenciesand probe preparation variations, can arise from any one of the numerous steps in manufacturingmicroarray slides. Therefore, considering the variability that can be generated from microarraydata, replication is absolutely necessary to allow confidence in the results. In the early days ofusing microarrays, it was not uncommon for studies to include data from a single slide perexperimental group. This was largely due to the high costs involved and the limited availability ofmicroarray slides. However, this has changed in recent years, and it is no longer acceptable toperform microarray studies without adequate replicates.The primary method of introducing replication in microarrays involves repeated measurementsof expression. The most common way to achieve this is to print the same target DNA at multiplelocations on the same slide, thereby creating intra-array replicates. This provides the means toassess the average expression level of every gene on the array, as well as the variability of themeasurement, thus providing a much more reliable estimate of the measurement itself. An advantageof this type of replication is that spatial problems on the array surface can be minimized. If, forexample, a particular area of the slide is defective and meaningful data cannot be extracted fromit, expression data can be obtained from the other replicates. When spotting replicate targets, onemust take into account the limited surface area available for printing because printing replicatespots on the same slide will reduce the overall number of genes that can be arrayed on a givenslide. This may not pose any problem if the gene collection is relatively small or if the microarrayprinter is capable of spotting in very high densities. However, in cases where the gene collectionis large and each gene is spotted in replicates, multiple slides may be necessary to print the entirecollection. This adds to both cost and inconvenience. An additional way to obtain repeated measurementsis to hybridize probes derived from the same original RNA sample on multiple arrays,thereby creating interarray replicates. This is in many ways similar to intra-array replication, butit also provides information on the variability between slides.It is most important to account for the biological variability within experiments. This type ofvariability is present in all experimental models and is likely the single major contributor to thesystem’s “noise.” The only way to overcome the problem of biological variability is to includesufficient biological samples in each experimental group. Thus, in very noisy models, such asoutbred mice or heterogeneous cell populations with highly variable responses, a larger number ofsamples per experimental group will be necessary to achieve the experimental goals. However, ifthe model is relatively “clean,” as is the case when working with inbred mice strains or stable celllines, experimental questions may be properly addressed with a smaller number of slides per group(e.g., three to four slides). An advantage of performing sufficient biological replicates is thatinterpretation of the results can go beyond the experimental groups and can be extrapolated to thegeneral population.Applying the most basic statistical consideration that “more is better” implies that the reliabilityof data is increased as the number of replicates (regardless of which type) is increased. 65 The© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 633specific number of replicates to be used in any given experiment varies according to the aforementionedsources of variability. Therefore, these parameters must be assessed in each experiment. 66And it would seem most logical to print as many replicate target DNA spots onto the microarrayslide as is practical and to perform several replicate experiments whenever possible. However,performing replicate hybridizations is relatively expensive, and the appropriate number of replicationsshould be determined by proper statistical considerations. 65,66c. Data ValidationConsidering the limitations in reliability of data generated from microarrays, it is absolutely vitalto use other experimental methods to validate the data generated by both types of microarrayplatforms. One such technology considered by many to be the ‘gold-standard’ for gene expressiondata is RT-PCR, which is even now giving way to the more state-of-the-art and informative methodof quantitative (also known as real time) PCR. 67–69 This method is used to examine a limited numberof genes of interest that have been identified by other methods, such as microarrays. QuantitativePCR is based on the detection of fluorescence produced during the amplification of gene-specificPCR products, and it can even determine the copy numbers of specific genes within a tissue sample.Since microarrays tend to have a relatively small (three- to fourfold) dynamic range, their use canlead to under-representation of changes that occur. Real time PCR has a greater dynamic rangeand is able to validate observed trends identified by microarray studies. 13,70–73 A more detaileddescription of the use of RT-PCR, real time PCR, and other more conventional methods forvalidating microarray data can be found in Chapter 14.C. Serial Analysis of Gene Expression1. GeneralConcomitant to the development of microarrays, SAGE methodology has been developed as anothertechnique for analyzing and screening thousands of expressed genes. In contrast to microarrays,which totally depend on genes that have already been cloned and at least partially sequenced, SAGEis independent of such limitations. SAGE enables the simultaneous detection and quantification ofthousands of expressed genes in tissues or cells without any prior knowledge of their identities orsequences. 74SAGE relies upon short, unique 9- or 10-base pair (bp) sequences that serve as distinctivemarkers (tags) to distinguish between different gene transcripts within tissues or cells of interest. 74According to the mathematical probabilities, using 9- or 10-bp tags allows the generation of 262,144or 1,048,576 different sequences, respectively (i.e., 4 9 and 4 10 , respectively). Thus, these short tagsprovide random coverage throughout the genome, representing each expressed gene in a discriminativemanner. The final stage of SAGE involves sequencing the cloned tags; however, in contrastto conventional sequencing of cloned genes, a process that requires considerable time and effort,sequencing in SAGE is extremely efficient. Each clone used in SAGE sequencing is composed ofmany different tags serially aligned (concatenated) and cloned in a single plasmid vector. See Figure15.3 for an outline of the SAGE process.2. MethodologyThe SAGE procedure starts by isolation of mRNA from the tissue of interest and the synthesis ofdouble stranded cDNA by use of a biotinylated oligo-dT primer. The double stranded cDNA issubsequently cleaved with a restriction endonuclease (“anchoring enzyme”) that has a 4-bp recognitionsequence. Generally, a 4-bp recognition site will cleave DNA on average every 256 basepairs, which is significantly shorter than the average size of mRNAs. The most commonly used© 2006 by Taylor & Francis Group, LLC

634 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIAAAAAAATTTTTTTIIIIIIIIAAAAAAATTTTTTTGTACAAAAAAATTTTTTTGTACAAAAAAATTTTTTTIV1CATGGTACAAAAAAATTTTTTT2CATGGTACAAAAAAATTTTTTTVVCATGGTACXXCATGGTAC00VICATGGTACX0X0CATGGTACVIVIICATGGTACCATGGTACCATGGTACFigure 15.3Outline of the SAGE methodology. (I) RNA extraction from tissue of interest and reverse transcriptionto double stranded cDNA with a biotin tag, (II) cleavage of cDNA with anchoring enzyme,(III) affinity purification of biotinylated cDNA and division of sample, (IV) ligation of the two linkers,(V) cleavage with tagging enzyme and blunt ending, (VI) ligation of two cDNA samples to formthe ditag, and (VII) concatenate ditags, clone, and sequenceenzyme is NlaIII, although in actuality, almost any restriction enzyme with a 4-bp recognition sitecan be used for the procedure.Following cleavage, the resulting cDNA that contains a 3′-cohesive end is affinity purified withstreptavidin-coated magnetic beads. The purified cDNA is then divided into two portions, and twoseparate linkers are ligated to the cohesive ends. These linkers are designed to contain a type IISrestriction endonuclease recognition site near the 3′-NlaIII sequence. These two restriction© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 635sequences enable digestion by their respective restriction enzymes (usually FokI and BsmFI), alsoknown as tagging enzymes. These restriction endonucleases cleave the cDNA at a distance of 9 to10 bp from the recognition sequence, resulting in small tags of a linker sequence. Blunt ending isperformed on the two populations of cDNAs, which are then mixed and ligated. The resultingproduct, a ditag with endonuclease recognition sites at either end, is PCR amplified and separatedon a polyacrylamide gel. The endonuclease-anchoring site is cleaved, and concatamers are clonedinto plasmid vectors, which are then used in standard sequencing procedures. The resulting sequencedata provides a gene expression profile of the original mRNA population. For detailed SAGEprocedures, consult Yamamoto et al. 753. LimitationsAlthough SAGE offers an exciting and unique way to obtain global gene expression profiles ofany type of eukaryotic cell, it has several drawbacks that have thus far limited its use and popularity.Aside from several technical issues, the major disadvantage of this technology is that it is difficultto perform SAGE successfully. The problem is mainly due to the inherent difficulty in creating taglibraries. Several steps of the SAGE protocol are relatively inefficient and introduce contaminatingsequences and bias in the data.One of the first drawbacks to be recognized was the need for relatively large amounts of startingRNA; however, in the past few years several modifications have been made to enable the use ofSAGE with considerably less RNA. MicroSAGE requires 500 to 5000 times less RNA than theoriginal SAGE protocols demanded and has been simplified by combining several steps within asingle tube. 76 Similarly, SAGE-lite was developed to allow analysis from as little as 100 ng ofRNA. 77 Another technical difficulty with the SAGE was that contaminating linker moleculesgenerated from PCR were cloned into the plasmid vector, thus introducing confusing and erroneousdata. By using biotinylated PCR primers, the ditags synthesized in PCR can be removed by affinityto streptavidin beads. 78 Another major problem of SAGE is the difficulty of further analyzing theunknown tags. Originally, the way to isolate and clone specific genes based on tag sequences wasdone by oligo-based plaque lifting. However, even though this method was used successfully inseveral cases, it appears to produce a high rate of false positives. 74 By using oligo-dT primers withthe identified tag sequences, this problem was partially resolved, mainly as a result of truncationat the 5′ ends of the genes. 794. SAGE and Developmental BiologyThe majority of the published literature on the use of SAGE technology has been in the cancerand immunology fields, whereas only a handful of developmental biology studies utilized thistechnology for global gene expression profiling. One such study has addressed the issue of limbspecificgene expression in the developing embryo. 80 By using SAGE on the forelimbs and hindlimbsof developing mouse embryos, Margulies and colleagues have been able to identify novel genesinvolved in the determination of limb identity. In order to obtain a refined and more relevant tagcollection for limb identity, these investigators developed a new algorithm for sequence analysis.They also have improved the method for tag-identity determination, thereby increasing the specificityof the tag-UniGene match. By using these novel approaches, they observed that the mostdifferentially expressed gene was Pitx1, a transcription factor known to play a vital role in limbdevelopment. 81 The differential expression of additional transcription factors critical to limb development,such as Tbx4, Tbx5, and limb-specific Hox genes was also detected by SAGE. 82 The authorshave further refined their tag library by performing a “virtual subtraction” with nonlimb specifictag libraries. The number of candidate genes was thereby reduced by 74%, yet the previouslyidentified regulators of limb identity were preserved.© 2006 by Taylor & Francis Group, LLC

636 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTo date, there is no published literature on the use of SAGE to study developmental orreproductive toxicology issues, although several studies have addressed toxicological questions ingeneral, such as identification of genes involved in dioxin-induced hepatotoxicity and carcinogenesis.83 Given the advantages of SAGE, especially with the recent developments making the technologymore accessible to the scientific community and the data more reliable and informative, weexpect SAGE to become increasingly popular among developmental and reproductive biologistsand toxicologists.A. BackgroundIII. PROTEOMICSThe term ‘proteomics’ was first introduced in 1995 as the study of “the total protein complementof a genome”, and it focuses on the isolation, quantification, and identification of the biologicallyactive molecules coded by genes. 84 Typical questions in this field of study include queries intoproteins’ functions and how they relate to the larger scientific questions being pursued. One mustremember that proteins are the moieties that carry out the majority of the cells’ functions. Thus,the processing of DNA to RNA, followed by the synthesis of proteins, places proteomics downstreamof functional genomics in the study of biological systems. Conceptually, proteomics is inmany ways similar to gene expression profiling with microarrays or SAGE since it seeks to describeand study protein expression patterns during different stages of development or in response toexposure to toxicants. In essence, proteomics is a necessary complement to gene expression analysis.Although it presently lacks well developed and high-throughput platforms, such as microarrays,new technologies are constantly emerging that will soon overcome these notable deficits and enablehigh-throughput and sensitive profiling of protein expression.Protein expression analysis can be used to analyze the myriad of proteins expressed by cellsduring embryonic and fetal development, whereas a toxicology-oriented study can also utilizeproteomics to compare expression profiles of treated versus control samples. Similar to microarrays,where gene expression of an experimental sample is compared with that of a matched control, theunderlying assumption in proteomics is that the altered state induced by exposure to a toxicant willbe characterized by either the up- or down-regulation of certain proteins. These changes in proteincontent will reflect various aspects of the toxic insult. Such changes may involve proteins that havebeen inactivated by a reactive metabolite, proteins that transduce cellular signals that produce toxicend points, or even proteins that participate in detoxification. Once these indicators have beenidentified, they can be screened against expression profiles induced by exposure to other drugs ortoxicants, thereby allowing the deduction of mechanisms of action or providing predictive toxicologicalassessments. 85,86B. Protein Expression Analysis1. Protein Isolationa. Two-Dimensional Gel ElectrophoresisThe first method that was developed for quantitatively studying global changes in protein expressionwas two-dimensional gel electrophoresis (2DE). 87 In spite of the many years that have passed sinceit was originally developed, 2DE is still the predominant technology used in proteomics to separateand observe changes in expression of specific proteins within biological samples. This approachcan only provide separation between proteins, while providing a gross visual representation. 2DE© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 637Control EmbryoTreated EmbryoIIIIIIIIIIIIIVFigure 15.4Outline of protein expression analysis with two-dimensional gel electrophoresis. (a) Extraction ofproteins from tissues of interest, (b) separation of proteins by isoelectric focusing, (c) separationof proteins by mass, and (d) extraction and purification of differentially expressed proteins, followedby mass-spectrometry analysis.is not able to provide additional information on the identity of the proteins, and thus, supplementaltechnologies for protein identification, such as mass spectrometry, must be used in conjunctionwith 2DE. See Figure 15.4 for an outline of the 2DE process.© 2006 by Taylor & Francis Group, LLC

638 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe basis for the separation of the hundreds of proteins within a sample is primarily the protein’sisoelectric point (IP), but separation is also based on molecular weight, solubility, and abundance.88,89 Briefly, the IP of a protein is the pH at which no net charge is exhibited by the protein.Protein extracts are placed onto an immobilized IP gradient (IIPG) strip, and an electric field isapplied. The IP of a protein is based upon the amino acid content; therefore, according to the pHrange used, different proteins can be resolved. After this primary one-dimensional separation, theIIPG strips are loaded onto an SDS polyacrylamide gel for electrophoresis. This second dimensionalseparation is thus based on the molecular weight of each protein, so that those proteins with similarIP’s now become resolved according to their molecular weights. 90–92 Following 2DE, the proteinsare visualized by the traditional Coomassie Blue or silver staining, although the dynamic range ofthese detection methods is limited. Recently, new methods of detection with fluorescent dyes, whichincrease the dynamic range in which proteins can be distinguished and quantified, have beenintroduced. For instance, the use of SYPRO orange and SYPRO red (Molecular Probes, Eugene,OR) enable increased sensitivity and reproducibility. 93–96Although 2DE is regarded as a standard method in proteomics, there are several aspects of itsapplication that limit its use. Proteins expressed at low levels often cannot be visualized becauseof the detection limits of this technology. 97 The difficulty in detecting low abundance proteins isfurther exacerbated because of masking by other, more highly expressed proteins. 98 Additionaltechnological limitations include the minimal solubility of membrane proteins, which often haveimportant roles in transporting chemicals across the cell membrane. 97,99 It is estimated that membraneproteins make up approximately 30% of the total cellular protein content, and resolving thisparticular problem is still a major challenge facing proteomics research.In order to better facilitate the use of 2DE and address some of the previously mentionedlimitations, fractionation is now often performed prior to isoelectric focusing. Fractionations areperformed with respect to different biochemical properties such as isolation according to cellularcompartments and gradients of aqueous solubility. Chromatography and free flow electrophoresisdevices have also been applied, not only to reduce protein mixture complexity but also to enrichthe presence of low expression proteins. 100b. Multidimensional Liquid ChromatographyAlthough the majority of proteomic studies utilize 2DE, alternatives have been under development.One such method incorporates the use of multidimensional liquid chromatography to separate theproteins prior to employing protein identification steps, such as mass spectrometry. 101,102 Liquidchromatography is more adaptable to high throughput and robotics, and more importantly, it offersseveral advantages in terms of protein separation. As previously discussed, some of the limitationsof 2DE relate to the inability to detect some proteins. Liquid chromatography, on the other hand,has better capabilities to separate protein mixtures and thus may be more informative than 2DE.Separation may be performed based on the molecular charge by use of an ion-exchange column.It is also possible to separate proteins on the basis of their hydrophobic interactions with thestationary phase or by specific affinity. The eluted protein fractions are subsequently identified,primarily by mass spectrometry.2. Protein Identificationa. Mass SpectrometryOnce differentially expressed proteins have been visualized, the next step involves protein identification.This is accomplished with mass spectrometry (MS), an approach that has lead to majoradvancements in proteomics research. Briefly, MS is a method that identifies molecules based ontheir molecular mass and charge signature (m/z) after they have been ionized and fragmented. The© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 639fragmentation pattern of each molecule is unique and generates a spectrum that distinguishes itfrom others. As a result, proteins and peptides can be identified and quantified by comparisonagainst existing databases. By coupling Edman degradation (removal of single amino acids fromthe N-terminal of the protein for sequencing) along with MS, it is possible to identify the aminoacid sequences of the protein by comparison with established protein databases. Although this lattermethod is highly precise, it is also regrettably slow. There is clearly a need for new approachesthat offer the increased throughput needed to analyze the numerous resulting spots from 2DE. 1031) MALDI and MALDI-TOF — Matrix-assisted laser desorption ionization (MALDI), coupledwith MS, is accomplished by laser excitation of a protein that has been digested with a site specificprotease, such as trypsin. The protein digest is deposited on a special matrix, and upon laserexcitation, the digested product becomes ionized and vaporizes to a gaseous state. 91,104 In the gaseousstate, the protein fragments can be detected and analyzed by MS and their identity elucidated.Currently, MALDI is often coupled with time-of-flight (TOF), which separates peptides and proteinsaccording to the time it takes them to travel from the MALDI matrix to the MS detector. After thepeptides have been vaporized and ionized by the laser, they travel across a chamber toward the MSdetector. Analysis is then carried out on the ionized molecules as they enter the detector accordingto their TOFs, and similar to all other MS methods, the m/z signature is used for identification ofthe protein of interest.2) MS/MS and ESI-MS/MS — Tandem mass spectrometry (MS/MS) has also emerged as acommon approach for the identification of proteins after extraction from 2-DE. 105 Once extracted,the protein of interest is digested and undergoes electrospray ionization (ESI) that allows ionizationof peptides directly from the liquid phase. The sample next enters into a tandem MS consisting oftwo mass analyzers. The first MS is used to isolate peptides ionized from the original sample. Thefragments analyzed by the first MS are subsequently sent into a collision chamber, where peptidesare further fragmented by collision with an inert gas, such as argon. The second MS analyzes thesecondarily formed ionized fragments, and a unique m/z signature is obtained. This method allowsfor the partial determination of amino acid sequences within the protein (also called peptidesequence tags), which can be used to search protein databases and identify the protein of interest.b. Isotope-Coded Affinity TagThe recently developed isotope-coded affinity tag (ICAT) methodology is conceptually similar tomicroarray technology in the sense that it enables the simultaneous identification and quantificationof peptides derived from two biological samples of interest. 106–108 ICAT is an exciting technologythat will certainly have a tremendous affect on our ability to examine proteomewide changes inprotein expression.ICAT is based upon MS combined with site-specific, stable isotopic tagging of proteins, therebyenabling distinct and high throughput analysis of protein mixtures. The ICAT reagent containsseveral components: a protein reactive group, a biotin tag, and a linker. The protein reactive groupis specific for the thiol groups of cysteine residues, where it covalently attaches to the reduced thiolform. The biotin tag is used for isolation by affinity chromatography of tagged-peptides that canbe used for analysis. The linker group is composed of an ethylene glycol region that comes in twoforms, light or heavy. The heavy form has eight deuterium atoms, whereas the light reagent haseight hydrogen atoms; therefore, they are identical and differ only in mass. As a result of the abilityof MAS to distinguish the 8-Da difference in mass, two distinct populations of proteins may becompared against each other.For example, let us consider an experiment where the mechanism of 2,3,7,8-tetrachloro dibenzodioxin(TCDD)-induced palatal clefting in mice is being investigated. Pregnant dams at gestationalday (GD) 12 are treated with either a relevant oral dose of TCDD or the vehicle. Palatal© 2006 by Taylor & Francis Group, LLC

640 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONtissues are removed from fetuses 24 to 48 h after TCDD exposure, and the proteins are isolated.Proteins from vehicle-treated fetuses are labeled with the light reagent, whereas proteins fromTCDD-treated fetuses are labeled with the heavy reagent. Following labeling, the protein lysatesare subjected to digestion by trypsin, and the resulting peptide fragments are purified by avidinaffinity. Only those peptides with modified cysteine residues that have reacted with the ICATreagents will possess the biotin group necessary for retention. The complexity of the affinity purifiedpeptide mixture is substantially reduced in comparison with the original sample because it containsapproximately 10% the amount of peptides. If desired, the purified peptide mixture can be furtherseparated and purified by liquid chromatography. Eventually, peptides from both TCDD and controlsamples are pooled and analyzed simultaneously by MS. Tagged peptides from both groups arecomparable and will fragment and ionize similarly in the mass spectrometer; however, because ofthe differences in mass, the detector can easily distinguish between the two forms of identicalpeptides. In other words, proteins from both samples can be identified and comparatively quantifiedby the heavy to light labeling shifts. Hence, TCDD-specific effects on the palatal protein contentcould be identified, thereby providing unique insight into potential mechanisms of TCDD-inducedpalatal clefting in mice.3. Protein ArraysHigh bandwidth expression assays, such as microarrays and SAGE, which enable the screening ofnumerous genes simultaneously, have revolutionized genomics. Yet, it is the protein that is thefunctional motif of biological reactions in mammals. Thus, the next logical step is to implementprotein microarrays, so that large numbers of proteins may be coanalyzed on a single platform.Protein arrays are currently not commonly used; however, exploratory and revolutionizing technologiesand platforms are being developed and tested. Protein arrays can be used to detect proteinantibodyinteractions, protein–nucleic acid interactions, protein-protein interactions, protein modifications,enzymatic activities, and more. 109–111 Several technical obstacles complicate the developmentof protein arrays. These include the tendency of proteins to denature and interact in a threedimensionalfashion, thus complicating the maintenance of relevant interactions on a surface.Emerging technologies that will affect this field in the upcoming years may be found in the followingreferences. 109–113IV. PREDICTIVE VALUESOne of the areas that is expected to benefit substantially from the development of toxicogenomicsand proteomics is the ability to predict pathological processes induced by xenobiotics or otherpotentially harmful agents. Microarrays have already been shown to have predictive merit in tumorclassification, thereby improving the prognostic capabilities of clinicians as well as the optimizationof chemotherapeutic regimens. 69,114 Similar efforts are being made to increase the ability to predicta novel compound’s toxicity based on similarities in gene expression. 115 Such efforts are currentlybeing led by Phase-1 (Santa Fe, NM), a biotechnology company dedicated to developing predictivetoxicological tools based on gene expression profiles. The company has created TOXbank, adatabase made up of toxicology (blood chemistry, histopathology, and more) and gene expressiondata. All gene expression data were generated with the company’s proprietary toxicology-specificrat arrays, from animals that had been treated with classical toxicants. Thus, gene expression profilesof new investigational compounds can be compared against the database, and any significant matchin gene expression profile and histopathology is suggestive of potential toxicity that may be elicitedby the chemical in question. 116Gene expression profiling studies in developmental toxicology have focused on implantation, 117early-stage differentiation of embryonic stem cells, 118 and the development of the nervous system 37© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 641and the retina. 13 Given the ethical and practical complexities of obtaining relevant samples fromhuman embryos or fetuses, no studies on their gene expression have as yet been published. Evendevelopmental toxicology studies in laboratory animals are often difficult to perform and interpretgiven the complexity of experimental design, routes of exposure, and differences in metabolismbetween the different animal species and humans.Developmental and reproductive toxicity studies are designed to assess the potential of agentsto induce toxicity in both a qualitative and quantitative manner. Most studies encompass the useof several animal species under different exposure paradigms. When there are specific, cross-species,and dose-dependent developmental abnormalities, extrapolation to humans is considered relativelyreliable. On the other hand, when toxicity is species specific, extrapolation of adverse outcomes tohumans might be established by more mechanistic approaches. 119 Therefore, using global geneexpression or protein profiling data of developmental and reproductive toxicants and newly developedpharmaceuticals to establish potential hazards in humans is one such approach that is constantlygaining support from both the academic and industrial sectors. However, it is important thatthe limitations and advantages of each the technologies described be kept in mind so that whenplanning such experiments, the most appropriate technique is chosen. Similar to the advancementsrecently made in the functional genomics of cancer biology, increasing knowledge and understandingof the mechanisms of developmental and reproductive toxicity will enable us to predict thosecompounds that posses the greatest potential for harm. 120–122A. MicroarraysV. RELEVANT WEBSITES AND DATABASES• The National Center for Biotechnology Information (NCBI), Gene Expression Omnibus, themicroarray data repository– http://www.ncbi.nlm.nih.gov/geo• The National Institute of Aging (NIA) mouse cDNA project home page– http://lgsun.grc.nia.nih.gov/cDNA/cDNA.html• Microarray user group– http://tango01.cit.nih.gov/sig/home.taf?_function=main&SIGInfo_SIGID=58• Basic information and relevant links– http://www.gene-chips.com– A public source for microarray protocols and software http://www.microarrays.org/index.html• The microarrays Yahoo eGroup– http://groups.yahoo.com/group/microarrayB. SAGE• NCBI’s SAGEmap– http://www.ncbi.nlm.nih.gov/SAGE• The Cancer Genome Anatomy Project– http://cgap.nci.nih.gov/SAGE• SAGE home page– http://www.sagenet.org/C. Proteomics• NCBI’s Structure Group– http://www.ncbi.nlm.nih.gov/Structure• Protein Data Bank– http://www.rcsb.org/pdb/© 2006 by Taylor & Francis Group, LLC

642 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONREFERENCES1. Lander, E.S., et al., Initial sequencing and analysis of the human genome, Nature, 409, 860, 2001.2. Venter, J.C., et al., The sequence of the human genome, Science, 291, 1304, 2001.3. Freeman, W.M., Robertson, D.J., and Vrana, K.E., Fundamentals of DNA hybridization arrays forgene expression analysis, Biotechniques, 29, 1042, 2000.4. Brown, P.O. and Botstein, D., Exploring the new world of the genome with DNA microarrays, Nat.Genet., 21, 33, 1999.5. Gerhold, D., Rushmore, T., and Caskey, C.T., DNA chips: promising toys have become powerfultools, Trends Biochem. Sci., 24, 168, 1999.6. Southern, E.M., Detection of specific sequences among DNA fragments separated by gel electrophoresis,J. Mol. Biol., 98, 503, 1975.7. Kafatos, F.C., Jones, C.W., and Efstratiadis, A., Determination of nucleic acid sequence homologiesand relative concentrations by a dot hybridization procedure, Nucleic Acids Res.., 7, 1541, 1979.8. Southern, E., Mir, K., and Shchepinov, M., Molecular interactions on microarrays, Nat. Genet., 21,5, 1999.9. Schena, M., et al., Quantitative monitoring of gene expression patterns with a complementary DNAmicroarray, Science, 270, 467, 1995.10. Ko, M.S., Embryogenomics: developmental biology meets genomics, Trends Biotechnol., 19, 511,2001.11. Tanaka, T.S., et al., Genome-wide expression profiling of mid-gestation placenta and embryo usinga 15,000 mouse developmental cDNA microarray, Proc. Natl. Acad. Sci. USA, 97, 9127, 2000.12. Kargul, G.J., et al., Verification and initial annotation of the NIA mouse 15K cDNA clone set, Nat.Genet., 28, 17, 2001.13. Mu, X., et al., Gene expression in the developing mouse retina by EST sequencing and microarrayanalysis, Nucleic Acids Res., 29, 4983, 2001.14. Lovett, R.A., Toxicogenomics. Toxicologists brace for genomics revolution, Science, 289, 536, 2000.15. Nuwaysir, E.F., et al., Microarrays and toxicology: the advent of toxicogenomics, Mol. Carcinog., 24,153, 1999.16. Bartosiewicz, M., et al., Development of a toxicological gene array and quantitative assessment ofthis technology, Arch. Biochem. Biophys., 376, 66, 2000.17. Hegde, P., et al., A concise guide to cDNA microarray analysis, Biotechniques, 29, 548, 2000.18. Eisen, M.B. and Brown, P.O., DNA arrays for analysis of gene expression, Methods Enzymol., 303,179, 1999.19. Diehl, F., et al., Manufacturing DNA microarrays of high spot homogeneity and reduced backgroundsignal, Nucleic Acids Res., 29, E38, 2001.20. Wildsmith, S.E. and Elcock, F.J., Microarrays under the microscope, Mol. Pathol., 54, 8, 2001.21. Wildsmith, S.E., et al., Maximization of signal derived from cDNA microarrays, Biotechniques, 30,202, 2001.22. Panchuk-Voloshina, N., et al., Alexa dyes, a series of new fluorescent dyes that yield exceptionallybright, photostable conjugates, J. Histochem. Cytochem., 47, 1179, 1999.23. Manduchi, E., et al., Comparison of different labeling methods for two-channel high-density microarrayexperiments, Physiol. Genomics, 10, 169, 2002.24. Fodor, S.P., et al., Light-directed, spatially addressable parallel chemical synthesis, Science, 251, 767,1991.25. Lipshutz, R.J., et al., High density synthetic oligonucleotide arrays, Nat. Genet., 21, 20, 1999.26. Yang, Y.H., et al., Normalization for cDNA microarray data: a robust composite method addressingsingle and multiple slide systematic variation, Nucleic Acids Res., 30, e15, 2002.27. Hill, A.A., et al., Evaluation of normalization procedures for oligonucleotide array data based onspiked cRNA controls, Genome Biol., 2, RESEARCH0055, 2001.28. Alon, U., et al., Broad patterns of gene expression revealed by clustering analysis of tumor and normalcolon tissues probed by oligonucleotide arrays, Proc. Natl. Acad. Sci. USA, 96, 6745, 1999.29. Selinger, D.W., et al., RNA expression analysis using a 30 base pair resolution Escherichia coli genomearray, Nat. Biotechnol., 18, 1262, 2000.30. Cho, R.J., et al., A genome-wide transcriptional analysis of the mitotic cell cycle, Mol. Cell, 2, 65, 1998.© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 64331. Yang, M.C., et al., A statistical method for flagging weak spots improves normalization and ratioestimates in microarrays, Physio. Genomics, 7, 45, 2001.32. Li, C. and Wong, W.H., Model-based analysis of oligonucleotide arrays: expression index computationand outlier detection, Proc. Natl. Acad. Sci. USA, 98, 31, 2001.33. Lockhart, D.J., et al., Expression monitoring by hybridization to high-density oligonucleotide arrays,Nat. Biotechnol., 14, 1675, 1996.34. Ishii, M., et al., Direct comparison of GeneChip and SAGE on the quantitative accuracy in transcriptprofiling analysis, Genomics, 68, 136, 2000.35. Schuchhardt, J., et al., Normalization strategies for cDNA microarrays, Nucleic Acids Res., 28, E47,2000.36. Eisen, M.B., et al., Cluster analysis and display of genome-wide expression patterns, Proc. Natl. Acad.Sci. USA, 95, 14863, 1998.37. Wen, X., et al., Large-scale temporal gene expression mapping of central nervous system development,Proc. Natl. Acad. Sci. USA, 95, 334, 1998.38. Sherlock, G., Analysis of large-scale gene expression data, Curr. Opinion Immunol., 12, 201, 2000.39. Zhang, K. and Zhao, H., Assessing reliability of gene clusters from gene expression data, Funct.Integr. Genomics, 1, 156, 2000.40. Heyer, L.J., Kruglyak, S., and Yooseph, S., Exploring expression data: identification and analysis ofcoexpressed genes, Genome Res., 9, 1106, 1999.41. Tavazoie, S., et al., Systematic determination of genetic network architecture, Nat. Genet., 22, 281,1999.42. Hilsenbeck, S.G., et al., Statistical analysis of array expression data as applied to the problem oftamoxifen resistance, J Natl. Cancer Inst., 91, 453, 1999.43. Ghosh, D. and Chinnaiyan, A.M., Mixture modelling of gene expression data from microarrayexperiments, Bioinformatics, 18, 275, 2002.44. Shmulevich, I. and Zhang, W., Binary analysis and optimization-based normalization of gene expressiondata, Bioinformatics, 18, 555, 2002.45. Tamayo, P., et al., Interpreting patterns of gene expression with self-organizing maps: methods andapplication to hematopoietic differentiation, Proc. Natl. Acad. Sci. USA, 96, 2907, 1999.46. Ben-Dor, A., Shamir, R., and Yakhini, Z., Clustering gene expression patterns, J. Comput. Biol., 6,281, 1999.47. Lukashin, A.V. and Fuchs, R., Analysis of temporal gene expression profiles: clustering by simulatedannealing and determining the optimal number of clusters, Bioinformatics, 17, 405, 2001.48. Herrero, J., Valencia, A., and Dopazo, J., A hierarchical unsupervised growing neural network forclustering gene expression patterns, Bioinformatics, 17, 126, 2001.49. Chiang, D.Y., Brown, P.O., and Eisen, M.B., Visualizing associations between genome sequences andgene expression data using genome-mean expression profiles, Bioinformatics, 17(Suppl. 1), S49, 2001.50. Bittner, M., Meltzer, P., and Trent, J., Data analysis and integration: of steps and arrows, Nat. Genet.,22, 213, 1999.51. Wong, K.K., Cheng, R.S., and Mok, S.C., Identification of differentially expressed genes from ovariancancer cells by MICROMAX cDNA microarray system, Biotechniques, 30, 670, 2001.52. Stears, R.L., Getts, R.C., and Gullans, S.R., A novel, sensitive detection system for high-densitymicroarrays using dendrimer technology, Physiol. Genomics, 3, 93, 2000.53. Shchepinov, M.S., et al., Oligonucleotide dendrimers: stable nano-structures, Nucleic Acids Res., 27,3035, 1999.54. Van Gelder, R.N., et al., Amplified RNA synthesized from limited quantities of heterogeneous cDNA,Proc. Natl. Acad. Sci. USA, 87, 1663, 1990.55. Kacharmina, J.E., Crino, P.B., and Eberwine, J., Preparation of cDNA from single cells and subcellularregions, Methods Enzymol., 303, 3, 1999.56. Baugh, L.R., et al., Quantitative analysis of mRNA amplification by in vitro transcription, NucleicAcids Res., 29, E29, 2001.57. Finnell, R.H., et al., Strain-dependent alterations in the expression of folate pathway genes followingteratogenic exposure to valproic acid in a mouse model, Am. J. Med. Genet., 70, 303, 1997.58. Bennett, G.D., et al., Valproic acid-induced alterations in growth and neurotrophic factor gene expressionin murine embryos [corrected], Reprod. Toxicol., 14, 1, 2000.© 2006 by Taylor & Francis Group, LLC

644 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION59. Wlodarczyk, B.C., et al., Valproic acid-induced changes in gene expression during neurulation in amouse model, Teratology, 54, 284, 1996.60. Wlodraczyk, B., et al., Arsenic-induced alterations in embryonic transcription factor gene expression:implications for abnormal neural development, Dev. Genet., 18, 306, 1996.61. Wlodarczyk, B.J., et al., Arsenic-induced neural tube defects in mice: alterations in cell cycle geneexpression, Reprod. Toxicol, 10, 447, 1996.62. Wang, E., et al., High-fidelity mRNA amplification for gene profiling, Nat. Biotechnol., 18, 457, 2000.63. Ohyama, H., et al., Laser capture microdissection-generated target sample for high-density oligonucleotidearray hybridization, Biotechniques, 29, 530, 2000.64. Luo, L., et al., Gene expression profiles of laser-captured adjacent neuronal subtypes, Nat. Med., 5,117, 1999.65. Lee, M.L., et al., Importance of replication in microarray gene expression studies: statistical methodsand evidence from repetitive cDNA hybridizations, Proc. Natl. Acad. Sci. USA, 97, 9834, 2000.66. Pan, W., Lin, J., and Le, C.T., How many replicates of arrays are required to detect gene expressionchanges in microarray experiments? A mixture model approach, Genome Biol., 3, research0022, 2002.67. Ginzinger, D.G., Gene quantification using real-time quantitative PCR: an emerging technology hitsthe mainstream, Exp. Hematol., 30, 503, 2002.68. Salonga, D.S., et al., Relative gene expression in normal and tumor tissue by quantitative RT-PCR,Methods Mol. Biol., 191, 83, 2002.69. Snider, J.V., Wechser, M.A., and Lossos, I.S., Human disease characterization: real-time quantitativePCR analysis of gene expression, Drug Discov. Today, 6, 1062, 2001.70. Mayanil, C.S., et al., Microarray analysis detects novel Pax3 downstream target genes, J. Biol. Chem.,276, 49299, 2001.71. Wurmbach, E., et al., Gonadotropin-releasing hormone receptor-coupled gene network organization,J. Biol. Chem., 276, 47195, 2001.72. Han, E., et al., cDNA expression arrays reveal incomplete reversal of age-related changes in geneexpression by calorie restriction, Mech. Ageing Dev., 115, 157, 2000.73. Choi, J.D., et al., Microarray expression profiling of tissues from mice with uniparental duplicationsof chromosomes 7 and 11 to identify imprinted genes, Mamm. Genome, 12, 758, 2001.74. Velculescu, V.E., et al., Serial analysis of gene expression, Science, 270, 484, 1995.75. Yamamoto, M., et al., Use of serial analysis of gene expression (SAGE) technology, J. Immunol.Methods, 250, 45, 2001.76. Datson, N.A., et al., MicroSAGE: a modified procedure for serial analysis of gene expression inlimited amounts of tissue, Nucleic Acids Res., 27, 1300, 1999.77. Peters, D.G., et al., Comprehensive transcript analysis in small quantities of mRNA by SAGE-lite,Nucleic Acids Res., 27, e39, 1999.78. Powell, J., Enhanced concatemer cloning-a modification to the SAGE (Serial Analysis of GeneExpression) technique, Nucleic Acids Res., 26, 3445, 1998.79. van den Berg, A., van der Leij, J., and Poppema, S., Serial analysis of gene expression: rapid RT-PCR analysis of unknown SAGE tags, Nucleic Acids Res., 27, e17, 1999.80. Margulies, E.H., Kardia, S.L., and Innis, J.W., A comparative molecular analysis of developing mouseforelimbs and hindlimbs using serial analysis of gene expression (SAGE), Genome Res., 11, 1686, 2001.81. Lanctot, C., et al., Hindlimb patterning and mandible development require the Ptx1 gene, Development,126, 1805, 1999.82. Takeuchi, J.K., et al., Tbx5 and Tbx4 genes determine the wing/leg identity of limb buds, Nature,398, 810, 1999.83. Kurachi, M., et al., Identification of 2,3,7,8-tetrachlorodibenzo-p-dioxin-responsive genes in mouseliver by serial analysis of gene expression, Biochem. Biophys. Res. Commun., 292, 368, 2002.84. Wasinger, V.C., et al., Progress with gene-product mapping of the Mollicutes: Mycoplasma genitalium,Electrophoresis, 16, 1090, 1995.85. Hellmold, H., et al., Identification of end points relevant to detection of potentially adverse drugreactions, Toxicol. Lett., 127, 239, 2002.86. Aardema, M.J. and MacGregor, J.T., Toxicology and genetic toxicology in the new era of “toxicogenomics”:impact of “-omics” technologies, Mutat. Res., 499, 13, 2002.© 2006 by Taylor & Francis Group, LLC

FUNCTIONAL GENOMICS AND PROTEOMICS 64587. O’Farrell, P.H., High resolution two-dimensional electrophoresis of proteins, J. Biol. Chem., 250,4007, 1975.88. Righetti, P.G., Gianazza, E., and Bjellqvist, B., Modern aspects of isoelectric focusing: two-dimensionalmaps and immobilized pH gradients, J. Biochem. Biophys. Methods, 8, 89, 1983.89. Gorg, A., Postel, W., and Gunther, S., The current state of two-dimensional electrophoresis withimmobilized pH gradients, Electrophoresis, 9, 531, 1988.90. Aebersold, R. and Leavitt, J., Sequence analysis of proteins separated by polyacrylamide gel electrophoresis:towards an integrated protein database, Electrophoresis, 11, 517, 1990.91. Patterson, S.D., and Aebersold, R., Mass spectrometric approaches for the identification of gelseparatedproteins, Electrophoresis, 16, 1791, 1995.92. Molloy, M.P., Two-dimensional electrophoresis of membrane proteins using immobilized pH gradients,Anal. Biochem., 280, 1, 2000.93. Steinberg, T.H., Haugland, R.P., and Singer, V.L., Applications of SYPRO orange and SYPRO redprotein gel stains, Anal. Biochem., 239, 238, 1996.94. Steinberg, T.H., et al., SYPRO orange and SYPRO red protein gel stains: one-step fluorescent stainingof denaturing gels for detection of nanogram levels of protein, Anal. Biochem., 239, 223, 1996.95. Lopez, M.F., et al., A comparison of silver stain and SYPRO Ruby Protein Gel Stain with respect toprotein detection in two-dimensional gels and identification by peptide mass profiling, Electrophoresis,21, 3673, 2000.96. Yan, J.X., et al., Postelectrophoretic staining of proteins separated by two-dimensional gel electrophoresisusing SYPRO dyes, Electrophoresis, 21, 3657, 2000.97. Celis, J.E. and Gromov, P., 2D protein electrophoresis: Can it be perfected? Curr. Opin. Biotechnol.,10, 16, 1999.98. Fey, S.J. and Larsen, P.M., 2D or not 2D. Two-dimensional gel electrophoresis, Curr. Opin. Chem.Biol., 5, 26, 2001.99. Lognonne, J.L., 2D-page analysis: a practical guide to principle critical parameters, Cell Mol. Biol.(Noisy-le-grand), 40, 41, 1994.100. Madoz-Gurpide, J., et al., Protein based microarrays: a tool for probing the proteome of cancer cellsand tissues, Proteomics, 1, 1279, 2001.101. Issaq, H.J., The role of separation science in proteomics research, Electrophoresis, 22, 3629, 2001.102. Issaq, H.J., et al., Multidimensional high performance liquid chromatography—capillary electrophoresisseparation of a protein digest: an update, Electrophoresis, 22, 1133, 2001.103. Hille, J.M., Freed, A.L., and Watzig, H., Possibilities to improve automation, speed and precision ofproteome analysis: a comparison of two-dimensional electrophoresis and alternatives, Electrophoresis,22, 4035, 2001.104. Karas, M. and Hillenkamp, F., Laser desorption ionization of proteins with molecular masses exceeding10,000 daltons, Anal. Chem., 60, 2299, 1988.105. Lahm, H.W. and Langen, H., Mass spectrometry: a tool for the identification of proteins separatedby gels, Electrophoresis, 21, 2105, 2000.106. Gygi, S.P., et al., Quantitative analysis of complex protein mixtures using isotope-coded affinity tags,Nat. Biotechnol., 17, 994, 1999.107. Smolka, M.B., et al., Optimization of the isotope-coded affinity tag-labeling procedure for quantitativeproteome analysis, Anal. Biochem., 297, 25, 2001.108. Smolka, M., Zhou, H., and Aebersold, R., Quantitative protein profiling using two-dimensional gelelectrophoresis, isotope-coded affinity tag labeling, and mass spectrometry, Mol. Cell Proteomics, 1,19, 2002.109. Zhu, H., et al., Global analysis of protein activities using proteome chips, Science, 293, 2101, 2001.110. Zhu, H., et al., Analysis of yeast protein kinases using protein chips, Nat. Genet., 26, 283, 2000.111. Lueking, A., et al., Protein microarrays for gene expression and antibody screening, Anal. Biochem.,270, 103, 1999.112. MacBeath, G. and Schreiber, S.L., Printing proteins as microarrays for high-throughput functiondetermination, Science, 289, 1760, 2000.113. Zhang, H.T., et al., Protein quantification from complex protein mixtures using a proteomics methodologywith single-cell resolution, Proc. Natl. Acad. Sci. USA, 98, 5497, 2001.© 2006 by Taylor & Francis Group, LLC

646 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION114. Golub, T.R., et al., Molecular classification of cancer: class discovery and class prediction by geneexpression monitoring, Science, 286, 531, 1999.115. Burczynski, M.E., et al., Toxicogenomics-based discrimination of toxic mechanism in HepG2 humanhepatoma cells, Toxicol. Sci., 58, 399, 2000.116. Gant, T.W., Classifying toxicity and pathology by gene-expression profile--taking a lead from studiesin neoplasia, Trends Pharmacol. Sci., 23, 388, 2002.117. Hines, R.S., Molecular analysis of implantation, Semin Reprod. Med., 18, 91, 2000.118. Guo, X., et al., Proteomic characterization of early-stage differentiation of mouse embryonic stemcells into neural cells induced by all-trans retinoic acid in vitro, Electrophoresis, 22, 3067, 2001.119. Guittina, P., Elefant, E., and Saint-Salvi, B., Hierarchization of animal teratology findings for improvingthe human risk evaluation of drugs, Reprod. Toxicol., 14, 369, 2000.120. Pennie, W.D., et al., The principles and practice of toxigenomics: applications and opportunities,Toxicol. Sci., 54, 277, 2000.121. Gross, S.J., et al., Gene expression profile of trisomy 21 placentas: a potential approach for designingnoninvasive techniques of prenatal diagnosis, Am. J. Obstet. Gynecol., 187, 457, 2002.122. Miertus, J., and Amoroso, A., Microarray-based genetics of cardiac malformations, Ital. Heart J., 2,565, 2001.© 2006 by Taylor & Francis Group, LLC

CHAPTER 16In Vitro Methods for the Study of Mechanismsof Developmental ToxicologyCraig Harris and Jason M. HansenCONTENTSI. Introduction ........................................................................................................................648II. Use of In Vitro Developmental Systems in Studies of Mechanisms ................................649A. Mammalian Embryo Culture.....................................................................................6491. Preimplantation Embryo Culture.........................................................................6492. Peri-implantation Embryo Culture ......................................................................6553. Postimplantation Embryo Culture .......................................................................655B. Other Systems and Approaches Employing Whole Embryo Culture ......................6661. Transgenic Animals .............................................................................................6662. Proteomics and Genomics ...................................................................................668C. Nonmammalian Whole Embryo Cultures.................................................................6701. Caenorhabditis elegans Cultures .........................................................................6702. Drosophila Cultures.............................................................................................6703. Amphibian Embryo Culture ................................................................................6704. Fish Embryo Culture ...........................................................................................6715. Avian Embryo Culture.........................................................................................671D. Organ and Tissue Culture..........................................................................................6721. Limb Bud and Related Perinatal Bone Cultures ................................................6732. Palatal Cultures....................................................................................................6763. Visceral Yolk Sac Cultures ..................................................................................6774. Sex Organs...........................................................................................................6785. Heart.....................................................................................................................6786. Eye and Lens .......................................................................................................6787. Embryonic Tooth Germ.......................................................................................6788. Spinal Cord ..........................................................................................................6789. Lung Buds............................................................................................................678E. Cell Culture and Clonal Analysis..............................................................................678III. Summary ............................................................................................................................683References ......................................................................................................................................684647© 2006 by Taylor & Francis Group, LLC

648 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONI. INTRODUCTIONCurrent applications of in vitro developmental systems can be broadly divided into two areas: (1)prescreening for developmental toxicants and (2) testing for elucidation of mechanisms of normaland abnormal embryogenesis. Manifestations of abnormal development are, herein, described asdysmorphogenesis or developmental toxicity to identify the functional or morphological abnormalitiesproduced by chemical or environmental insults in vitro. The relative merits of these modelsfor their respective purposes have been discussed previously at great length. 1–9 It is generally agreedthat in vitro models, such as the rodent whole embryo culture system, may require additionalvalidation before they can be used universally as predictive teratogen prescreens. Nevertheless,clear advantages support the continued development of the various methods as in vitro tools tostudy mechanisms of developmental toxicity. This is especially true for developmental investigationsin viviparous species. Major obstacles encountered in understanding mechanisms of abnormaldevelopment and embryotoxicity include experimental inaccessibility, a paucity of material, andthe numerous confounding maternal, nutritional and physiological influences present in utero. Mostof these limitations can now be overcome through systematic study of the whole conceptus or itsdissociated component parts in vitro at various stages of gestation. A complete understanding of anymechanism, however, is unlikely to emerge from analyses using only a single approach. Ultimately,several in vivo and in vitro methods, utilizing different levels of biological organization and complexity,must be combined in a systematic manner to elucidate mechanisms of developmental toxicity. Inaddition, the likelihood that a single universal mechanism will be found to explain embryotoxicmanifestations caused by environmental extremes and chemical exposure is highly improbable.Various manifestations of embryotoxicity are not expressed uniformly throughout the complex,dynamic and interactive continuum of development and are exhibited during the species’ periodof sensitivity for that specific tissue. Identical maternal exposures to a given chemical agent atdifferent times in gestation may, for example, produce a spectrum of deleterious effects characteristiconly of disturbances in specific developmental processes particularly vulnerable at the timeof exposure. Lesions produced by chemical agents and environmental extremes will, therefore, bemanifest both temporally and spatially at all levels of biological organization. These range fromeffects on the entire conceptus to highly selective alterations at the cellular and molecular levels.In the discussion to follow, emphasis is placed on the use of mammalian cells, tissues, organs, andembryos, although nonmammalian species are highlighted as deemed appropriate.The study of mechanisms of developmental toxicity bears the close and expected resemblanceto similar in vitro methods used for investigating a broad spectrum of toxicological effects in matureorganisms. Systems employing decreasing levels of biological complexity are commonly used toisolate specific processes and probe for specific biochemical and molecular mechanisms that maybe involved in developmental toxicology. Endpoints for analysis can be developed and customized,based on the needs of a particular study. The range of possibilities for analytical endpoints mayinclude (1) an evaluation of gross malformations using visual examination, light microscopy,scanning and transmission electron microscopy, scanning laser confocal microscopy, and computerassisted morphometric analysis; (2) functional and physiological alterations in organs or tissues,including studies of transport, metabolism, biosynthesis, turnover, electrical and metabolic coupling,and maintenance of cellular redox status, and (3) molecular events, including studies of time andtissue-specific gene expression, differentiation, pattern formation, induction, receptor binding andsignal transduction, programmed cell death, and imprinting. Experimental models with higher levelsof biological organization and complexity are useful in the evaluation of mechanisms involvingorgan-organ interactions, pharmacokinetics, and systemic metabolic regulation. Recent advancesin molecular and developmental biology have created new possibilities for evaluation of the dynamicmolecular and biochemical processes of selected cell populations within an intact embryo. Thetechniques involved include the use of specific fluorescent antibody probes and ribonucleotidehybridization markers.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 649In vitro approaches to the study of mechanisms of developmental toxicity discussed in thischapter are classified according to a descending order of biological organization beginning withwhole embryo culture, followed by organ, cell, and tissue culture. Specific details of variousprocedures to be used as examples have been described in several excellent reviews and in hundredsof methodological texts to which the reader is referred. Only a small portion of the availableliterature has been cited as examples in this chapter and the omission of otherwise excellent workis, regrettably, unavoidable. The focus of the following discussion is to describe selected examplesof in vitro approaches for the study of mechanisms of developmental toxicity, to describe someadvantages and disadvantages of each, and to suggest where additional applications may be appropriatefor future studies. Many current advances in the fields of developmental and molecularbiology have yet to be exploited for studies of chemical embryotoxicity.A. Mammalian Embryo CultureII. USE OF IN VITRO DEVELOPMENTAL SYSTEMSIN STUDIES OF MECHANISMSIn vitro approaches used for mechanistic studies of developmental toxicity employ whole embryostaken from various stages of the developmental spectrum. Embryos for culture range from pronuclearand early cleavage stage embryos (preimplantation) to conceptuses grown from the posteggcylinder-head fold stage to late limb bud stages (postimplantation). Each of the approaches describedbelow provides certain advantages for the experimental isolation and study of a potential chemicalinteraction and its accompanying mechanism. Not the least of these advantages is the ability toisolate and focus on single stage–specific processes. The limitations of each method, however,should be borne in mind. Not the least of these is the restricted duration for which the culturedtissue retains growth and differentiation characteristicsin vitro that are comparable with those oflittermates still growing in vivo.As growth, differentiation, and complexity of function increase during development, the abilityto maintain normal growth rate is generally the first parameter to be compromised in vitro.Conceptuses of very young or advancing gestational age are likely to undergo additional morphologicaland functional differentiation, but they universally lag with respect to overall growth asdetermined by overall size and macromolecule (protein and DNA) content. The loss of optimalgrowth is a more important consideration in the whole embryo than with cell or organ culturesbecause of the more specific need in the latter instances to preserve only function and the capacityto differentiate. Isolated tissues and organs grown in culture can lose important physical andchemical stimuli normally found in their embryonic microenvironment while still maintainingfunction. For these and other reasons pertaining to development in general, the effective length ofculture has also been shown to decrease with advancing gestational age as has been delineated byNew. 10 Major in vitro systems that utilize whole embryos are divided into three categories, namely,those that include starting material from preimplantation, peri-implantation, and postimplantationstages of development. Each of these requires a different methodological approach for the isolationand maintenance of the embryo, and they do not allow for a continuous culture from fertilizationthrough to the close of organogenesis.1. Preimplantation Embryo CultureEvaluation of the teratogenic and embryotoxic effects of chemical during fertilization and thepreimplantation phase of development has relied primarily on the premise set forth by Austin 11 thatchemical exposure of the preimplantation conceptus has one of two consequences: (1) the embryotoxicagent will destroy all or a significant number of the existing embryonic cells, resulting in© 2006 by Taylor & Francis Group, LLC

650 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONearly embryonic death, i.e., reproductive failure, or (2) chemical exposure will result in the destructionof a lesser portion of the embryonic cells, and from those that remain, the embryo can recoverand grow normally without malformation or growth retardation. The latter result is due to thedemonstrated totipotency and general plasticity of early cleavage stage embryonic cells, where asingle viable cell is capable of regenerating an entire normal conceptus. Although generally correct,the universality of this “all-or-none” theory has recently been challenged, partly through work usingpreimplantation culture systems that show that persistent functional and dysmorphogenic lesionsmay persist without embryonic death.Initial studies employing preimplantation whole embryo cultures were initiated as early as 1912,utilizing the rabbit blastocyst. 12 The subsequent development and use of preimplantation embryocultures for the study of developmental mechanisms has been recently reviewed by Spielmann andVogel 13 and Lawitts and Biggers. 14 Fertilized or unfertilized preimplantation embryos can beobtained after synchronization and superovulation of the test animal, the most common experimentalspecies used are the rabbit, rat, mouse, and hamster, and the procedures follow the general outlinedepicted in Figure 16.1. The collection of sufficient numbers of eggs or fertilized embryos for usein studies of mechanisms often requires superovulation of the female and some means to controlgestational age and stage. Successful superovulation can depend on a number of factors, includingthe age, weight, and strain of the animal used, as well as the time of injection.Superovulation and synchronization can be accomplished in the mouse by use of an initial i.p.injection of 5 IU pregnant mare’s serum gonadotropin (PMSG), followed 44 to 48 h later by ani.p. injection of human chorionic gonadotropin (hCG, 5 IU) 14,15 These methods may be used toobtain fertilized and nonfertilized ova, and early cleavage and later preimplantation stage embryosdepending on the time at which material is removed from the oviduct and whether mating is allowedprior to recovery. For in vitro experimentation involving the culture of fertilized preimplantationembryos, dams are placed with proven stud males immediately after hCG injection. Successfulmating is determined by the appearance of a vaginal plug on the following morning. Pregnant damsare sacrificed at 17 to 28 h after hCG administration, and their oviducts are removed and flushedto release pronuclear- or two-cell–stage embryos. Flushing of embryos fertilized in vivo from theoviduct at later cleavage stages (two- to eight-cell; 48 h) also offers some advantages, in that certaincomplicating factors related to in vitro culture are eliminated. The second polar body (Figure 16.1B),which appears within hours after fertilization, allows fertilized eggs to be distinguished from thosethat have not been fertilized (Figure 16.1A). More accurate determination of fertilization can bemade by observation of the formation of a pronucleus or of the decondensing sperm head.Once collected, embryos are staged and cultured individually in microdrops of M-16 or T-6culture medium and bovine serum albumin. The embryos are cultured under paraffin or siliconeoil as specified by the in vitro fertilization (IVF) protocol, or as groups in rotating cultures containingvarious media. 15 Paraffin oil is used to prevent unnecessary drying and cross-contamination ofembryos during culture. Its use can present additional problems for studies involving exposures,especially when the chemical exposure involves addition of highly water-soluble compounds thatmay precipitate in the oil prior to reaching the embryo. To avoid this problem, preimplantationembryos may also be successfully cultured in 96-well microplates.During the preimplantation period of development, several properties of the conceptus provideadvantages for experimental manipulation and mechanistic studies. Two-cell to four-cell embryoscan be easily disaggregated and the cells stained with fluorescein isothiocyanate (FITC) (Figure 16.1G).Reaggregation of cells into control and experimental chimeras allows the fate of treated anduntreated cells to be easily followed and identified during periods of exposure and response. This“chimera assay” approach has been used to evaluate both direct and indirect effects on embryogenesis.The direct effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) on cell proliferation anddifferentiation were reported by Blankenship and coworkers. 16 They used FITC-labeled chimerasto show that TCDD exposure of two-cell embryos did not appear to cause subsequent alterations© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 651PreimplantationRodent Embryo Culture44–48 hrPMSG12–14 hrFertilizedTime-MatinghCGUnfertilized17–28 hrCBOviduct removedand flushedAD2-CellFEIn vitro Fertilization4-CellG8-CellCavitationMorulaChimeraZana PellucidaHatchingBlastocystFigure 16.1 Schematic diagram of the general strategy for preimplantation rodent embryo culture. Superovulationis initiated by the successive injection of pregnant mare’s serum gonadotropin (PMSG) and human chorionicgonadotropin (hCG). Unfertilized oocytes are collected 12 to 14 h after hCG administration by flushing themfrom the oviduct. (A) Oocytes at this stage have extruded the first polar body prior to ovulation and fertilizationand can be used for tests involving in vitro fertilization (IVF). (B) Mating of superovulated females after hCGadministration allows for collection of fertilized eggs. They are identified by extrusion of the second polar body.(C to E) Sperm nucleus decondensation, formation of the pronucleus, and migration of pronuclei to the centerof the egg occur in succession as cells prepare for the first cleavage division, which can occur either in vivo orin vitro. Depending on the length of time after mating, embryos flushed from the oviduct may be at the two- orthe four-cell stage. This would circumvent the two-cell block typically seen in embryos cultured in vitro. (F)Compaction of the embryo begins at the eight-cell stage. This is followed by cavitation at the 32-cell stage, asdevelopment proceeds through the morula stage to hatching of the late blastocyst from the zona pellucida. (G)Embryos can be dissociated at the four-cell stage and reassociated with other labeled cells; this procedureproduces chimeras to be evaluated in vitro at the blastocyst stage or reintroduced into a receptive female,allowing development to continue to term.© 2006 by Taylor & Francis Group, LLC

652 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONin the rate of cell proliferation but did accelerate the rate of differentiation. This was manifestedby an increase in the rate of cavitation, which serves as a measure of trophectoderm differentiation.In a study by Oudiz et al., 17 male mice were treated with ethylene glycol monomethyl ether(EGME) or remained untreated and were mated with superovulated females. Cells from the resultingembryos were disaggregated; those from embryos sired by control males were labeled with FITC,and chimeras were formed from the cells obtained from the control and experimental matings.Proliferation ratios calculated using this approach showed that embryos from EGME male matingshad significant proliferation disadvantages compared with embryos obtained entirely from controlmating.Large quantities of unfertilized eggs collected as cumulus masses after superovulation of thenonpregnant animal can also be used for experiments looking at direct toxicity to the gamete ormechanisms involving the process of IVF. The latter approach provides the distinct added advantageof being able to synchronize development for better dose and response fidelity in evaluation ofchemical effects. An example of the use of this approach in developmental toxicity studies wasreported by Kholkute et al. 18 using B6D2F1 mice. They describe the effects of polychlorinatedbiphenyl (PCB) mixtures (Arochlor 1221, 1254, and 1268) and the individual congener (3,3′,4,4′-tetrachlorobiphenyl) added directly to the IVF medium containing capacitated sperm. Direct effectson fertilization and increased incidence of oocyte degeneration and abnormalities were reported.Additional modifications of this approach are possible but have not been thoroughly exploited forthe evaluation of developmental toxicity mechanisms.Success rates of IVF are sometimes low but can be improved considerably through coculturewith oviductal epithelial cells and/or by supplementation from appropriate growth factors. Thisprocedure will yield higher rates of fertilization and more robust embryos because of the expectedcontribution of natural growth factors provided by the epithelial cells. Such an approach mayprovide important insights into the toxicological significance of oviduct response and regulation.The combination of these two in vitro methods may provide an even more relevant environmentfor mechanistic studies of early development. For example, the direct and interactive consequencesof chemical exposure in the oviduct may be addressed by use of primary cultures of oviductepithelium with early preimplantation embryos under IVF conditions.In vitro evaluations of the response of cultured oviduct epithelial cells to modulators of glutathione(GSH) status and a pesticide (lindane) have been completed. They show that oxidativestress and altered cellular thiol status in the oviduct may lead to altered function and subsequentlyinfluence embryonic growth and development. 19,20 This approach has the mechanistic advantage ofallowing evaluations of chemical effects on the developmental environment, including the nutritiveand signaling functions of the oviduct epithelium. Investigations are underway to elucidate functionalchanges elicited in oviduct function and subsequent effects on the process of fertilization invitro. These studies should provide valuable mechanistic information regarding the manner in whichtoxicants exert direct effects on fertilization and early embryogenesis and also regarding theenvironment in which these events take place.Eggs and embryos during the one- and two-cell stages are particularly sensitive to environmentalconditions, especially ambient temperature and oxygen concentration. Several species differ withrespect to the ease with which their embryos undergo cleavage, compaction, and blastocyst formation,and progress through development. Because of a much higher success rate for culturing twocellembryos through to the blastocyst stage, the mouse is generally the species of choice for thesetypes of experiments. A characteristic difficulty in culturing embryos of several species, includingmouse and human, is a developmental two- or four-cell “block,” beyond which the fertilized embryois unable to progress. Usually considered an artifact of in vitro growth and closely associated withoxidative stress, this condition may actually hold relevance to some mechanisms of early reproductivefailure elicited by chemical agents. Indeed, recent analyses have provided evidence thatreactive oxygen species generated in vitro are responsible for the developmental arrest. 21,22© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 653Understanding the mechanistic interactions of the preimplantation embryo with its environmentand possible chemical insults could provide some explanations for currently unknown causes ofinfertility. It is also conceivable, but untested, that xenobiotic exposures within the oviduct couldalter redox status and antioxidant functions and lead to similar consequences in vivo. In relationto possible disturbances of cellular homeostasis following chemical insult, Gardiner and Reed 23have begun to characterize the GSH status of the preimplantation embryo. These authors haveprovided evidence that this antioxidant and embryoprotectant may be extensively involved in themechanism of numerous chemical toxicants. They have shown that the oxidative stress producedby exposure to t-butylhydroperoxide (tBH) significantly alters GSH status, the overall manifestationsof which change during the course of development. Very little systematic mechanistic ordescriptive information exists regarding the exposure, toxicokinetics, biotransformation, and/ormanifestations of cellular embryotoxicity for the developing embryo while in the oviduct, includingpossible roles played by the oviduct in regulating chemical exposure, biotransformation, anddetoxication. A great deal of information may be gained by the use of in vitro methods to elucidatedevelopmental effects at various stages of early gestation.From another perspective, preimplantation embryo cultures have been used extensively formechanistic studies designed to understand the biochemical and physiological processes that regulatenormal development. Much of this early developmental work with preimplantation embryocultures to date has evaluated the uptake and incorporation of labeled precursors into embryonicmacromolecules and the functional and morphological effects of inhibitors of metabolism. 24–28 Theseprocedures provide valuable endpoints for use in the evaluation of chemical embryotoxicity andhave, in fact, been employed in studies designed to describe the embryotoxicity of agents such ascyclophosphamide and trypan blue. 13,29 Procedures for the estimation of embryonic cell number,repair capacity, biotransformation, and drug uptake have also been devised. 13,15,29–31Conditions for successful culture vary with respect to species and gestational age. They depend,to a large degree, on the presence of specific nutritive and trophic factors to ensure progression ofthe embryogenesis beyond the characteristic in vitro block of development. In vitro studies havebeen instrumental in describing amino acid uptake and utilization, particularly for use in proteinsynthesis, in the preimplantation embryo. 26–28,32 Borland and Tasca 24 demonstrated the inhibitionof L-methionine uptake and incorporation into four-cell stage mouse embryos. The inhibitionappeared to be a function of alterations in active methionine influx and not a result of acceleratedefflux or inhibition of protein synthesis. Physiological and biochemical functions, such as growthfactor interactions, 33,34 prostaglandins, 35 nutritive functions, 36,37 glucose uptake and utilization, 38–40intermediary metabolism, 41,42 steroid hormone function and metabolism, 43–45 receptor-mediatedcellular functions, 33,34 and amino acid uptake and transport functions, 42,46,47 have all been evaluatedin preimplantation cultures. Most have not been exploited for use in studies of developmentalembryotoxicity.More realistic evaluation of the maternal, embryonic, and chemical factors that affect preimplantationgrowth and development requires the ability to assess the influence of the environmentand maternal factors on the process. Extensions of the in vitro method to utilize other in vivo andin vitro approaches can greatly enhance the power of this approach. The use of IVF, coculture (withoviductal epithelial, stromal, or decidual cells), and transplantation have also been demonstratedin developmental and physiological studies, but these paradigms have not yet been thoroughlyexploited for use in mechanistic studies of embryotoxicity. Heterogenic cultures and in vivo–invitro procedures can be used to test whether a xenobiotic is interfering directly with embryonicdevelopment or whether a maternal effect (e.g., decidualization in the uterus) is responsible. Thefactors that contribute to the “two-cell block” of rodent and rabbit conceptuses cultured in vitrohave been associated with bursts of reactive oxygen species that may be controlled by intracellularprotective capacities and the soluble constituents found in culture media or oviductal fluid. 21–23 Theresults of these studies are reinforced by work involving inclusion of cysteine, glycine, hypotaurine,© 2006 by Taylor & Francis Group, LLC

654 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONand other amino acids or antioxidants in the culture medium. Such supplementation can significantlyimprove viability and reproductive success under IVF conditions and where embryos are subsequentlyreintroduced into a receptive female for continuation of development. 48 A major disadvantageof the transplantation approach is that success rates among laboratories doing the sameprocedures differ greatly. Also, such methods can be quite time consuming and costly, especiallyif it is necessary to establish a dose response for several different chemicals or modes of exposure.Preimplantation embryo culture is possible using a number of species and is not dependent onextraembryonic membrane structure, as is postimplantation embryo culture. A major disadvantageof this system is the inability to grow embryos optimally beyond the late blastocyst stages. Attemptsto extend cultures beyond the normal time of implantation and through organogenesis have, however,been described. 49,50 Although differentiation continues in a near normal fashion in vitro, therates of growth and differentiation are considerably slower in comparison with patterns of growthand differentiation in vivo. In general, embryos explanted before the egg cylinder stage are notlikely to develop into the somite stages in vitro, and most do not establish an active heartbeat orcirculation. A great deal of differentiation is possible with regard to function, establishment ofprimary germ layers, and microstructural and morphological form, although growth lags and thefailure rate is very high.The direct use of in vitro preimplantation culture methods for the study of teratogenicitymechanisms has been reviewed by Spielmann and Eibs, 29 with particular reference to in vivoexposure to chemicals and/or therapeutic agents and subsequent culture and evaluation of embryosin vitro. Many early studies on the sensitivity of preimplantation embryos to external insults wereconducted in vitro and focused on the effects of UV and x-ray irradiation. It was found that therate of development to the blastocyst stage is actually a less sensitive indicator of embryotoxicitythan is cell number in the blastocyst. 29Kaufmann and Armant 51 used mouse preimplantation cultures to evaluate the in vitro embryotoxicityof cocaine and its metabolite, benzoylecognine. They showed that these agents are bothable to directly block embryogenesis, particularly at the one and two-cell stages, and that metabolismof cocaine seems to abate the embryotoxic potential. Other approaches to evaluate the potentialcontribution of biotransformation to chemical embryotoxicity have utilized the inclusion of anexogenous biotransformation system to generate reactive intermediates in vitro. A rat hepatic S-9fraction and appropriate cofactors was used by Iyer et al. 31 to show that day-3 blastocysts hadaltered hatching, culture dish attachment, and trophoblastic outgrowth when exposed to naphthalenein vitro. The use of the exogenous bioactivation source demonstrated the need for bioactivation ofnaphthalene to produce embryotoxicity and helped to define the mechanism of action of this andpossibly other polycyclic aromatic hydrocarbons. Direct exposure to 5-bromodeoxyuridine (BrdU)in preimplantation cultures results in alterations in both cell division and differentiation. 52Preimplantation cultures have been used extensively for the study of mechanisms of developmentin general and the endogenous and exogenous factors that influence early cell cleavage anddifferentiation. Endpoints for evaluation include morphological, physiological, biochemical, andmolecular events, each of which can be used in conjunction with other determinations to probemechanisms of development and the chemical disturbances that may influence the normal process.The processes of regulated gene expression, including the events governing the transition frommaternal to embryonic genomic control, can also be investigated. A more detailed discussion ofthe use of molecular techniques in the study of developmental toxicology can be found in Chapters14 and 15 of this volume. Gene expression, posttranscriptional modifications, and protein synthesishave been studied and found to be critical components in the regulation of normal development aswell as sensitive targets for chemical agents that interfere with normal development. 53–55An example of a recent molecular approach to integrate the various expression events andunderstand their mechanistic controls is described in an evaluation of HSP 70.1 expression concurrentwith the activation of the zygotic genome. 56 This approach used transgenic mice fromC57BL6/CBA F1 embryos that contained two to three copies of the HSP70.1 Luc plasmid/haploid© 2006 by Taylor & Francis Group, LLC

Anatomical peculiarities of the developing rodent conceptus make it particularly well suited forwhole embryo culture during the teratogen-sensitive period of organogenesis. Unlike a majority ofmammalian species, including human beings, rodents (rats and mice) possess a visceral yolk sac(VYS) that completely surrounds the amnion and developing embryo through inversion and antimesometrialattachment via the presumptive ectoplacental cone (Figure 16.2). In this arrangement,the VYS serves as a physical protective barrier encompassing the embryo; it also serves in numerousvital detoxication, metabolic, and nutritional roles. This compact arrangement results in conceptusesthat are easy to observe, manipulate, and perturb as needed experimentally and represents theIN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 655genome. Transgenic offspring were identified by slot blot hybridization, using polymerase chainreaction (PCR) primers specific to the firefly luciferase gene. One-cell zygotes obtained fromtransgenic crosses were typed by the luciferase assay when the HSP70.1 gene was expressed. Thestudy concluded that this product was expressed only during the one-cell stage and did not continueinto the two-cell stage. Considering the expected relevance of HSP 70.1 to environmental insultand chemical exposure, it is easy to suppose that this methodology might be appropriately appliedto the study of the effects of pre- and postfertilization chemical insult on the same parameters.Similar molecular approaches could be used to evaluate the roles and potential perturbations ofmechanisms related to cell division and/or cleavage, gap junctions, second messengers systems,growth factors, receptors, and the expression and function of many other components.2. Peri-implantation Embryo CultureIn conjunction with concerns of mechanisms of preimplantation development that involve directeffects and consequences to the developing conceptus itself, chemically elicited effects on theprocess of uterine implantation are also of considerable importance in developmental toxicity. Theoverall process of implantation, especially in the rodent and human (where the blastocyst becomescompletely imbedded into the uterine endometrial mucosa) has been the least experimentallyaccessible process in the developmental continuum. Recent advances, however, have produced anin vitro primary culture system that can address the mechanistic bases of toxic events that can affectimplantation. 57The system that has been developed utilizes a primary cultured monolayer of endometrial cellsgrown on a Cellagen membrane matrix impregnated with stromal cells. This approach not onlyallows recognition of the trophoblast by the endometrium but results in the ingress of the conceptusinto the stromal layer, accompanied by displacement of resident cells. In addition, membranouscell culture dish inserts may be used to provide a proper environment for the attachment and normalpolar orientation of columnar epithelial cells. Unlike other approaches, this method preserves alarge number of functional and structural characteristics seen in the same cells in vivo. These includeabsorptive and excretory functions, appropriate orientation of polarized cell surface markers, polarorientation of intracellular organellar structures (such as the Golgi apparatus and endoplasmicreticulum), preservation of estrogen and progesterone responsiveness, and synthesis and release ofproteins. While not ideal in every sense of functional integrity, this system has been shown topromote the recognition and implantation of competent blastocysts. It may provide the best availablemodel for the mechanistic study of early embryonic growth as related to implantation and toxicityfactors that may affect the embryo during the peri-implantation period. Attachment of human,bovine, and mouse blastocysts has been compared by the use of this system, 58 but few, if any,systematic studies have been conducted using this approach to assess the direct effects of chemicalson the process of implantation. Utilization of such an in vitro approach could lead to importantmechanistic insights regarding the reproductive toxicity of environmental and therapeutic chemicalsduring implantation.3. Postimplantation Embryo Culture© 2006 by Taylor & Francis Group, LLC

656 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPostimplantationRodent WholeEmbryo CultureTime-MatingMaternal blood collection andremoval of the uterusRemoval of decidual massesfrom the uterusConceptus dissected freeof decidual massGD 9.5 GD 10Ectoplacental coneVisceral yolk sacAmnionEmbryoReicherl's membrane opened48 hr Control24–26 hrTreatmentControlTreatmentRoller CultureFigure 16.2Schematic diagram of rat whole embryo culture. Time-mated pregnant females are anesthetizedon GD 9.5 or 10, prior to opening of the abdomen and removal of maternal blood (to be usedfor preparation of culture media). The uterus is removed with the ovaries and cervix attached.Uteruses containing the conceptuses are pinned to a silicone-filled dissecting dish, and decidualmasses are cut free from the uterus with iridectomy scissors. Conceptuses are teased free ofthe decidual mass by use of fine watchmaker’s forceps, and the Reichert’s membrane is removedto complete the preparation for culture. Embryos are then placed for 24 to 72 h at 37°C in rollerbottles containing rat serum in the appropriate gas mixture for optimal growth at the givendevelopmental stage. Morphological, biochemical, physiological, and molecular endpoints can beevaluated throughout the culture period by removing selected embryos from the culture. Relativesize, degree of development, and identification of major structures for the GD 9.5 and GD 10conceptuses are shown.highest order of in vitro biological organization. Because of its versatility, the rodent whole embryoculture system has been used to evaluate the embryotoxicity of several hundred compounds. Itcontinues to grow in popularity as a method of choice for mechanistic evaluations of embryotoxicants,as well as for investigations of mechanisms of development in general.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 657The historical evolution of the rodent whole embryo culture technique has been described byNew. 10 It encompasses a review of the first successful attempts to transfer and culture rabbitpreimplantation embryos, including the attempts to devise an effective artificial uterus. In the initialapproach, a rodent embryo encased in an intact VYS is suspended in a 5-cm diameter watch glasscontaining heat-inactivated rat serum. The watch glass is placed in a petri dish lined with wet cottonto maintain humid conditions, and the dish is kept in a constant gas environment of 60% to 95%O 2 and 3% to 5% CO 2 . This method is effective for growing conceptuses from early somite to earlylimb bud stages but has not been used extensively for mechanistic studies. This is a result in partof the difficulties of culturing large numbers of conceptuses.By far the most popular and universally applicable method for the culture and study of wholeembryos for investigation of mechanisms is the rotator culture system, first described by New 59 in1971, and summarized in Figure 16.2. The specific details of embryo explantation and preparationfor culture have been described in detail in a number of excellent texts. 60–62 Briefly, a pregnant damis anesthetized or euthanized, and its abdomen is opened to expose the gravid uterus. Maternalblood is withdrawn from the abdominal aorta and further processed for heat-inactivated serum tobe used as a component of the culture medium. The uterus, with ovaries and cervix attached, isremoved, pinned to a dissecting surface submerged in Hanks’ balanced salt solution (HBSS) andthe decidual swellings are exposed by use of iridectomy scissors. The decidua is separated withfine watchmaker’s forceps, and intact conceptuses are removed. Reichert’s membrane is then tornaway to expose the VYS and the conceptus is ready for culture.Several types of culture media have been described, although culture for more than a few hoursappears to have an absolute requirement for 35% to100% heat inactivated rat serum. The compositionof the remainder of the medium is usually a balanced salt solution or other defined medium(e.g., Waymouth’s) in various combinations. Conceptuses can be cultured in serum-free definedmedia, such as HBSS, for several hours if the conceptus is then returned to the usual serum-repletemedium. Subsequent growth in this type of culture is without apparent deleterious long-term effects.Media are warmed to 37°C and equilibrated with stage-appropriate O 2 and CO 2 concentrations.The conceptuses are submerged in glass vials (approximately 3 ml) that can be inserted into arotating culture apparatus capable of continuous gassing and maintenance of constant temperatureor grown in sealed roller bottles placed on a roller in an incubator and gassed every 24 h. 63 Therotating culture apparatus and roller bottle apparatus can each be used and have distinct advantages.Roller bottles generally contain 60 to 175 ml of media and support growth of multiple conceptusesup to 24 h following gassing (1 ml media per conceptus is usually considered sufficient). However,a continuous gassing rotator apparatus requires smaller volumes of media and better facilitates theculture of individual conceptuses. In either case, these two approaches are able to maintain thecorrect stage specific O 2 and CO 2 concentrations required for optimal growth.The direct exposure of cultured conceptuses to chemical agents can usually be accomplishedby addition of the agent to the culture medium. This is especially simple for water-soluble chemicals,whereas, lipophilic agents require prior solubilization in an appropriate vehicle. Acetone, dimethylsulfoxide (DMSO), and ethanol have been used successfully for this purpose and do not appear toadversely affect development as long as concentrations are kept reasonably low (less than 0.1%).There is a range of stages of embryonic growth during which conceptuses can be easilymanipulated and for which in vitro growth closely parallels that seen in the intact uterus. Theserange from the egg cylinder-head fold and early somite stages (at which time the anterior neuraltube has not yet closed, the heart has not yet begun to beat, and the embryo has yet to rotate tothe ventral orientation) to the later stages of organogenesis (characterized by complete closure ofthe anterior neural tube, establishment of a functional heart and circulatory system, and completerotation of the embryo). 64 Vascularization of the VYS becomes much more evident over this periodof development and is used as a marker of development in vitro. It is also during this criticaljuncture of development that the embryo undergoes a number of important metabolic and functional© 2006 by Taylor & Francis Group, LLC

658 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIn vitro development7.5–day +24 h +48 h +72 h−Figure 16.3Scanning electron micrographs of the best mouse embryos cultured in one experiment, showingthe extent of growth and morphogenesis that may be achieved in vitro (bar: 100 µm). Day 8.5embryos (plug day = day 0) were cultured in rotating bottles containing DR50 medium (changedafter 24 h of culture) and gassed intermittently with 5% O 2 , 5% CO 2 , and 90% N 2 for the first 38h; they were then cultured in fresh DR50 medium in a gas phase with 20% O 2 , 5% CO 2 and 75%N 2 until 48 h. Embryos were then cultured in rotating bottles gassed continuously with 20% to40% O 2 until 72 h in vitro. (From Sturm, K. and Tam, P.P.L., Methods Enzymol., 225, 164, 1993.With permission.)transformations, including conversion from an essentially anaerobic, glycolysis-based metabolismto an aerobic, citric acid cycle–based dependence. In the mouse, the window of optimal in vitrogrowth and differentiation begins on gestational day (GD) 8 (plug day = GD 0). In the rat, thisperiod begins on GD 9.5. In each case, a period of 48 h in culture is believed to encompass theoptimal period wherein the in vitro conceptus will grow at almost the same rate as if it had remainedin the uterus, although earlier and later stages have also been grown successfully. The entirespectrum of in vitro growth, spanning a period of 72 h, is shown in Figure 16.3. 61While most of the early efforts to perfect whole embryo culture have utilized mouse and ratconceptuses, whole embryo culture of other species, such as the rabbit, is becoming more common.Rabbit whole embryo culture has not been widely used in toxicological investigations but has greatpotential. In the past decade, procedures for rabbit whole embryo culture have been developedlargely by the efforts of Naya and coworkers, 65 Ninomiya and coworkers, 66 and Pitt and Carney. 67,68Criteria for uniform evaluation of rabbit conceptuses have also been developed and used as astandard of reference during in vitro development. 67,68 While the logistics of rabbit whole embryoculture are similar to those for rodent embryo culture (maximum culture time, timing of explantation,periods of organogenesis, gassing conditions, etc.), some physiologic differences requirespecial attention and make rabbit embryo culture unique (Figure 16.4).Rabbit conceptuses require that a portion of the bidiscoid placenta remain attached throughoutthe culture period. Although not critical for conceptual viability, the placental bed serves the functionof keeping the conceptus intact during the course of incubation. A striking difference between therabbit and rodent conceptus is that the rabbit VYS only surrounds the caudal portion of the partiallyrotated embryo, unlike mouse and rat where the VYS is inverted and completely surrounds the© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 659PostimplantationRabbit WholeEmbryo CultureTime MatingMaternal bloodcollection and removalof the uterusCut alongantimesometrialedge with scissorsUterine WallAmnionVisceral yolk sacEmbryoPull back uterinewalls to exposeconceptusBlood vesselof yolk sacAllantoic placentaCarefully cut aroundconceptus withoutdamaging bloodvessels of the yolk sacTransfer conceptuses individuallyto continuous gassing cultureapparatusCulture 12–48 hr and evaluateFigure 16.4 Schematic diagram of rabbit whole embryo culture. Time-mated pregnant females are anesthetizedon GD 9 prior to opening of the abdomen and removal of maternal blood (to be used for preparation ofculture media). The uterus is removed with the ovaries and cervix attached. Each conceptual swelling isindividually removed and pinned to a silicone filled dish for dissection. A single cut is made along the antimesometrialedge of the uterus, and the overlying uterine tissue is peeled back to expose the embryo. To free theembryo, it is necessary to cut through the placenta, leaving a portion of the placenta attached to maintain theintegrity of the conceptus for culture. Special care must be taken not to damage the visceral yolk sac or anyrelated blood vessels. Embryos are then placed for 24 to 48 h at 37°C in a continuous gassing incubator foroptimal growth at the given developmental stage. Morphological, biochemical, physiological, and molecularendpoints can be evaluated throughout the culture period by removing selected embryos from the culture.Relative size, degree of development, and identification of major structures for the GD 9.5 and GD 10 conceptusesare shown.© 2006 by Taylor & Francis Group, LLC

660 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONembryo. It has been suggested that the VYS in the rat and mouse has a protective effect, as isevident by higher GSH concentrations and more biotransforming enzymes in the VYS than in theembryo proper. 69,70 The distinctive structure of the rabbit VYS is more closely related to humanVYS in that it does not envelop the embryo. While this is a distinct advantage of rabbit wholeembryo culture, it can prove to be an inconvenience during culture, and extra care must be takento avoid physically damaging the conceptus.The combined use of both rat and rabbit whole embryo cultures allow investigators to comparecommon developmental processes, and biochemical and molecular parameters that may contributeto mechanisms of toxicity and species sensitivity, and to resistance to chemical agents. Thisapproach was used to provide insight into thalidomide’s species selectivity and its mechanism ofaction. Hansen et al. 71 have shown that GSH and cysteine, two major contributors to intracellularredox status, are preferentially depleted in the rabbit conceptus by thalidomide, while rat conceptalGSH and cysteine remain relatively replete. This suggests that thalidomide-induced oxidative stressand possible misregulation of redox status occur only in the rabbit conceptus and may contributeto the mechanisms of teratogenesis. This approach, the comparison of both sensitive and resistantspecies in vitro, provides a means to determine species-specific differences and should prove usefulin the elucidation of specific developmental toxicity mechanisms.Whole embryo cultures have been used successfully to investigate the mechanisms of a numberof relevant toxicological endpoints. Some of the first and most significant studies attempting todescribe mechanisms of chemical embryotoxicity were investigations that dealt with conceptalbiotransformation of chemical teratogens. Inherent in these studies were attempts to understandthe relationships between maternal chemical biotransformation and the biotransformation capacitiesof the conceptal tissues. Some known in vivo teratogens, such as cyclophosphamide (CP) and 2-acetylaminofluorene (2-AAF), were found to produce little relevant dysmorphogenesis when addeddirectly to cultures of early somite-stage rat conceptuses. 72,73 That maternal bioactivation is aprerequisite for producing dysmorphogenesis was demonstrated in vitro by the addition of rathepatic 9000 × g supernatant fractions (S-9) and the appropriate cofactors (i.e., an NADPHgenerating system) for supporting cytochrome P450 monooxygenation activities. Because of theexperimental accessibility of the in vitro system and the ease with which environmental conditionscould be controlled, this approach was found to be useful in characterization of the maternal andconceptal biotransformation necessary to produce embryotoxicity. Some of the agents tested includecyclophosphamide, 72 2-AAF, 73 acetylsalicylic acid, 74 adriamycin, 75 acetaminophen, 76 retinoic acid(RA), 77,78 valproic acid, 79,80 niridazole, 81 cytochalasin D, 82 and rifampin. 83Hepatic S-9 fractions prepared from male rats exposed to chemical inducers of selected subclassesof cytochrome P-450 isozymes have helped to identify the requisite biotransformationsnecessary for making the chemicals in question dysmorphogenic. For example, it was demonstratedthat major phenobarbital (PB)-inducible isoforms were required to metabolize cyclophosphamideto its reactive and teratogenic forms; other classes of inducible enzymes were ineffective. 73,84Conversely, the 3-methylcholanthrene (3-MC)-inducible isoforms (predominantly CYP 1A1) 58 wereshown to be necessary for the conversion of 2-AAF to a variety of embryotoxic and dysmorphogenicproducts, including the putative proximal dysmorphogen, 7-hydroxy-acetylaminofluorene (7-OH-AAF). The latter metabolite was found to be dysmorphogenic without need for an exogenousbioactivation system. It was suggested to be the proximal teratogen because it produced neural tubedefects not altogether unlike those elicited when the parent 2-AAF was administered to the pregnantanimal. 73,76 Further studies revealed that, unlike PB-inducible isoforms, the major MC-inducibleisoform can also be induced in utero in conceptal tissues through maternal exposure to MC. Thesestudies further showed that conceptuses explanted from MC-exposed dams and grown in culturewere able to constitutively bioactivate AAF without an exogenous bioactivation system. 84 Subsequentin vitro studies using acetaminophen (APAP), a compound physically and chemically similarto 7-OH-AAF, allowed for direct testing of the prediction that similar manifestations of dysmorphogenesiswould result from both agents because of the common physical and chemical properties.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 661This was found to be the case, leading to additional studies. These studies concluded that theembryotoxic mechanism of both 7-OH-AAF and APAP was likely to involve additional metabolismby an endogenous bioactivation source to form more chemically reactive species, such as theirrespective catechols. 76The mechanistic implications from these studies were that constitutive and inducible cytochromeP450s exist in the embryo and VYS of the organogenesis-stage rodent conceptus. CytochromesP450 have been partially characterized in the VYS, using a series of resorufin ether analogs thathave been shown to be selectively metabolized by various cytochrome P450 isozymes. Thisapproach provides evidence that P450 activities are present and differentially distributed in varioustissue compartments of the conceptus. Total P450 specific activity is approximately 10-fold higherin the VYS than in the embryo proper. At least four constitutive isoforms and one inducible isoformwere postulated from these studies. Specific antibodies raised against purified adult P450 isozymeswere able to identify only the inducible form as CYP 1A1 by use of Western blot analysis, 85,86suggesting that the others may be uniquely expressed only in the conceptus.Subsequent work has sought to identify and further characterize the functions and metaboliccapacities of these enzymes. 87 Such studies using the whole embryo culture system have beencritical in the characterization of P450 function in the organogenesis-stage rat conceptus. Specificmechanistic studies into the molecular regulation of these activities and their consequences areunderway but may need to be combined with investigations of cell culture and cell free systemsto clearly define complex cellular mechanisms. In vitro models have been extremely useful inelucidating these pathways. They provide a means to systematically evaluate the mechanisms ofconceptual biotransformation in an experimentally friendly system that has been removed frompotentially confounding maternal factors.Other enzymes capable of the bioactivation and detoxication of chemical agents in the conceptushave also recently been characterized by use of the whole embryo culture system. Recent studiesof the biotransformation of phenytoin have provided compelling mechanistic evidence that thepreviously postulated generation of an arene oxide metabolite by the actions of cytochrome P450is most likely not responsible for the resultant embryotoxicity and dysmorphogenesis. 88,89 The datanow suggest that the majority of phenytoin’s embryotoxic and teratogenic effects are mediatedthrough the formation of organic free radicals or reactive oxygen species, generated through theactivity of endogenous lipoxygenases or prostaglandin H synthetase. 90,91 Significant lipoxygenaseactivities have been reported in the rodent conceptus. 92 Evidence for free radical–mediated embryotoxicityis supported by a concomitant reactive oxygen-mediated oxidation of DNA (8-OH-2′-deoxyguanosine) and findings that the embryonic lesions can be eliminated by addition of theantioxidants catalase and/or superoxide dismutase (SOD) to the culture medium. 93 Activities ofthese and related enzymes of the arachidonic acid pathway have been shown to be high in thedeveloping conceptus. They may well account for the bioactivation and subsequent embryotoxicityof a number of other known teratogens and in vitro dysmorphogens for which cooxidation to freeradicals and reactive intermediates has already been described.To use in vitro methods for reliable prediction of mechanisms operating in vivo, one must beable to address important pharmacokinetic factors that would determine the actual amount ofchemical reaching the conceptus in utero. Whole embryo cultures have been used for comparativein vivo–in vitro studies to define mechanisms related to the distribution, metabolism, and elicitedeffects in the whole animal. 94 Studies into the mechanisms of 2-methoxyethanol embryotoxicityhave utilized the whole embryo culture system to clarify the site and mode of metabolism of(methoxy- 14 C)-2-methoxyacetic acid (2-MAA) to 14 CO 2 , suggesting that the compound is metabolizedin the dam and not in the conceptus. 95 Short term in vitro cultures of CD-1 mouse embryoswere also used to demonstrate the inhibition of DNA synthesis by 2-MAA and the attenuation ofinhibition and embryotoxicity by compounds such as formate, acetate, and sarcosine. 96 Otherinvestigators have also integrated in vitro and in vivo approaches to study the pharmacokineticsand biotransformation of the teratogen, valproic acid. 79,94 These authors point out that in using in© 2006 by Taylor & Francis Group, LLC

662 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONvitro models, care must be taken to consider relevant pharmacokinetic behavior so as to avoidpossible misinterpretation of data. One significant property of the developing rodent conceptus thatalso should not be overlooked in pharmacokinetic and mechanistic considerations is the discoverythat the conceptal milieu during early organogenesis is significantly more alkaline than the surroundingmaternal fluids and tissues. This can result in the selective accumulation of weak acids,such as valproic acid, within the conceptus. 97,98 The cellular regulation of conceptal pH has notbeen adequately studied, especially from the perspective of possible deleterious alterations elicitedby chemical agents. Preliminary studies also indicate that the viable, intact conceptus may be usedin vitro with a customized perfusion apparatus and minute microfiberoptic (MFOP) sensors capableof monitoring real-time alterations in pH caused by environmental changes and chemical exposure.99,100 The ability to isolate the embryo in vitro and conduct evaluations of the specific effectsof individual chemical modulators and inhibitors provides a means to probe the mechanisms ofcomplex processes that are extremely difficult to unravel in the whole organism.An understanding of the entire spectrum of events leading to persistent anatomical and/orfunctional defects in the developing organism will not only require elucidation of the means bywhich chemicals are taken up and metabolized within the conceptus. It will also demand knowledgeof how the conceptal response to that insult affects dysmorphogenic or functional alterations. Therodent whole embryo culture system is an excellent model for the study of antioxidant, detoxication,and other protective systems. In particular, the studies described in the next sections have takenadvantage of the ability, in whole embryo culture, to selectively modulate several biochemical andphysiological pathways and to isolate and study specific effects caused by chemicals.Early reports indicated that organogenesis-stage rat conceptuses have very low or undetectableactivities of enzymes such as UDP-glucuronosyltransferase, sulfotransferase and glutathione-Stransferases,the predominant enzymes used by the adult for detoxication of reactive chemicalintermediates. 101 While these conclusions still appear to be valid, enzyme activities may be presentbut highly localized.Recent investigations also suggest that the conceptus maintains significant concentrations ofGSH and has ample activities of enzymes involved in its synthesis, turnover, and metabolism. 69,102,103The important role of GSH in embryo protection has been well established through both invitro 104–109 and in vivo studies. These have provided the basis for use of whole embryo culture toinvestigate the spatial and temporal mechanisms that control GSH synthesis, turnover, and metabolism.Chemical modulators used in the manipulation of GSH status for mechanistic and descriptivestudies include L-buthionine-S,R-sulfoximine (BSO), a selective and irreversible inhibitor of GSHsynthesis, 109 resulting in GSH depletion as a function of the intrinsic cellular rates of GSH turnover;102 diethyl maleate (DEM), an, α,β-unsaturated carbonyl compound that readily and nonezymaticallyforms GSH adducts, resulting in rapid GSH depletion; 69 a cysteine pro-drug, 2-oxothiazolidine-4-carboxylate(OTC), which is capable of supporting higher rates of GSH synthesis throughincreasing intracellular levels of the rate-limiting precursor cysteine; 104 glutathione monoethyl ester(GME), which produces selective elevations of intracellular GSH without the need for active GSHsynthesis; 110 and acivicin, which selectively inhibits the activity of γ-glutamyltranspeptidase (GGT),a plasma membrane glutathionase enzyme also believed to function in amino acid transport andmetabolism of GSH conjugates. 111 Microinjection of these and other selective modulators has alsobeen shown to result in selective spatial alteration of GSH within the conceptus. 110,112 Such abilitiesto selectively alter GSH status will be of considerable advantage in dissecting the tissue selectiveeffects of chemicals and the focal biotransformations that may define the basis for differential tissuesensitivity and resistance and the ability to respond to specific chemical agents. The ability tosynthesize new GSH is an important consideration in this context. In vitro studies have shown thatdepletion of conceptal GSH (to 30% to 40% of control in embryo and VYS) by addition of DEMdirectly to the culture medium is followed by a differential recovery in the two compartments.Visceral yolk sac recovery begins immediately after reaching the 30-min nadir and is replenished© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 663at a rate to exceed initial levels within 3 h. 69 The embryo, however, remains refractory until VYSconcentrations are restored to preexposure levels. This suggests that repletion of the VYS may beinvolved in the regulation of embryonic GSH synthesis through control of the supply of cysteine(the rate limiting precursor for GSH de novo synthesis) to the embryo or through inhibitory signals.Specific rates of GSH synthesis have been determined by measurement of the specific incorporationof [ 35 S]cysteine into GSH during periods of normal embryogenesis, and the parallel useof [ 35 S]methionine to show that no cystathionine pathway exists in the conceptus during this phaseof development. 69 Although the embryo has the enzymatic capacity to replenish GSH, embryonicsynthesis is not activated until the VYS is GSH-replete, perhaps due to regulation of the precursoramino acid supply. Investigations using whole embryo culture are currently underway to understandthe regulation of GSH precursors, their supply to the embryo, and possible interruptions resultingfrom xenobiotic exposure. The mechanisms of amino acid uptake and utilization have already beenaddressed, 113–115 but a lack of information still exists regarding the perturbation of these processesby chemical agents. The in vitro whole embryo culture model will also serve well for studiesdesigned to help understand the mechanisms of GSH stimulation, induction, and response duringchemical and oxidative stress.The ability to control conditions such as the time and concentration of exposure, oxygen tension,temperature, pH, and other environmental factors makes whole embryo culture a very useful systemfor evaluation of certain endpoints. These include effects of oxidative stress, production of reactiveoxygen species (ROS), 116 and the antioxidant functions that protect the conceptus from theseconditions. Direct additions of the enzymes superoxide dismutase, catalase, and glutathione peroxidasehave all been shown to affect outward manifestations of embryotoxicity in cases of exposureto agents such as glucose (hyperglycemia), 117 arsenic, 118 and phenytoin. 119 It is not yet understood,however, what the mechanisms of action are and/or whether the enzymes are exerting their effectsintra- or extracellularly. Superoxide dismutase, catalase, glutathione disulfide reductase, glutathioneperoxidase, γ-glutamyl transpeptidase, and other related protective enzymes have been postulatedto be expressed differentially in space and time throughout the continuum of development. Thismay well be responsible for some of the selective sensitivities and resistance seen at differentjunctures in gestation. Mechanistic biochemical and molecular investigations are currently underwayto probe the possibility that the regulation of programmed cell death is a function of ROSproduction and the cells ability to maintain the proper redox and antioxidant status. 120Closely related to all of the foregoing discussions of biotransformation, oxidative stress, celldeath, DNA repair, and maintenance of the proper intracellular environment are the roles ofextracellular calcium, 121 calcium channel blockers, 122,123 and the potential susceptibility of alteredcalcium homeostasis in the dysmorphogenesis of neural tube closure. 124 Regulation and maintenanceof pyridine nucleotide [NAD(H) and NADP(H)] status during development and the critical alterationsthat occur following chemical insult have also been identified as important factors inmechanisms of teratogenesis and embryonic development, but have not yet been systematicallystudied in the developing conceptus. The status and turnover of NAD(H) in aminothiadiazoleteratogenicity, 125 possibly related to DNA repair initiated through poly-ADP ribosylation andhypoxia, 126 have been addressed using in vitro systems. An important role of the major cellularreducing equivalents, NADP(H) and NAD(H), has been postulated for a number of critical biochemicalpathways and reactions involving normal developmental metabolism and their responseto xenobiotic insult, including the endogenous production of antimetabolites. 126–128 Because of theneed to monitor very rapid changes in redox status that may be altered due to chemical orenvironmental disruption of cellular function, Thorsrud and Harris 126 have used the whole embryoculture system and noninvasive, real-time MFOP fluorescence probes. These probes were used tomeasure pyridine nucleotide fluxes from the VYS surface of intact, viable rat and mouse conceptusesmaintained in a customized perfusion chamber (Figure 16.5). Chemical modulation of antioxidantfunctions and concurrent determinations of GSH status have shown that measured changes of© 2006 by Taylor & Francis Group, LLC

664 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONMercuryArc LampFiberoptic Sensor Tip366 nmPass FilterAuxillaryLight SourceRubberSeptum455 nmPass FilterPhotomultiplier250 µExcitation FiberEmission FibersAuxiliary FibersSpectrolluorometerChart RecorderPerifusate OutRat ConceplusGlass BeadsPerifusate inFigure 16.5Schematic diagram of the MFOP monitoring chamber, showing the location of the rat conceptusin the chamber and the placement of the sensor on the surface of the visceral yolk sac. Insertshows an amplified view of the microfiberoptic sensor tip. (From Thorsrud, B.A. and Harris, C.,Teratology, 48, 347, 1993. With permission.)surface fluorescence in the VYS can involve changes in different pyridine nucleotide pools. 129 Theseincluded the net reduction of NAD + following cyanide exposure and hypoxia. Another examplewas the net oxidation of NADPH following GSH oxidation and utilization by glutathione disulfidereductase after exposure to the diabetogenic agent, alloxan, and the glutathione oxidant, diamide.Further development of this and other similar technologies will aid in the study of other potentialmechanisms of embryotoxicity involving biotransformation, biosynthesis, mitochondrial function,intermediary metabolism, and macromolecular repair mechanisms.Since we are able to grow the intact conceptus (embryo proper plus associated extraembryonicmembranes) during critical periods of organogenesis in vitro, both spatial and temporal investigationsinto toxicological mechanisms have been possible. Whole conceptuses maintain the structuralintegrity and important physiological and functional interrelationships between embryo proper andextraembryonic membranes, which may be lost in disrupted systems. Careful selection of tissuespecificagents allows studies of how damage to one tissue may affect the function of another.Elegant studies 130–138 have demonstrated that several known teratogenic agents appear to act throughdisruption of the nutritional pathways of the rat VYS. The relatively unique placement of theinverted rodent VYS provides a physical as well as functional barrier between the embryo and thematernal milieu. During organogenesis, this organ serves an indispensable function in taking up© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 665maternal proteins. This is accomplished via active fluid phase pinocytosis and adsorptive endocytosis,followed by enzymatic degradation in VYS lysosomes, thus providing the major source of precursoramino acids for new protein synthesis. These proteins, incorporated into acidic lysosomal vacuoles,are rapidly degraded, principally by cathepsins D and L. The constituent amino acids are thentransported to the embryo for use in protein synthesis and as precursors for numerous otherbiosynthetic and regulatory processes. These studies showed that virtually all of the amino acidsincorporated into new protein synthesized in the developing conceptus were derived from VYSproteolytic digestion of maternal (or serum) proteins, hence the designation of this process ashistiotrophic nutrition.Chemical teratogens and embryotoxins known to specifically interfere with either pinocytosis(trypan blue) or proteolysis (leupeptin, chloroquine) are believed to exert their effects by theseprincipal means. Careful control of dose, exposure, and environmental conditions have allowed forthe study and further characterization of these processes as they relate to the mechanism of actionof a number of chemical toxicants. 130,131,133,134,137 In vitro methods focused on the function of theVYS have also shown that hyperglycemic embryopathies are accompanied by sequential yolk sacfailure. 138The development of new biochemical tools, such as selective fluorescent probes, has providedadditional insights into the mechanisms and systems that control proteolysis and their relationshipsto chemical embryotoxicity and dysmorphogenesis. Studies have also been conducted using intactconceptuses complemented by the culture of VYSs and the use of organ cultures, such as “giant”yolk sacs. 139 Gaining an understanding of which mechanisms of developmental toxicity involve theVYS and are directly dependent on the unique relationship between embryo and extraembryonicmembranes in the rodent is important for possible extrapolation to the human. That is because thehuman does not have an inverted VYS encompassing the embryo and entirely different functionsand responses to chemical insult may result.Other potential targets for developmental embryotoxicants related to VYS function that can beeffectively studied by use of the whole embryo culture system include the processes of intermediarymetabolism and energy production. Frienkel et al. 140 utilized the whole embryo culture system intheir evaluation of the “honeybee syndrome.” They proposed that the rodent embryo undergoes adramatic transition from an essentially anaerobic, glycolysis-dependent organism during earlyorganogenesis to an aerobically metabolizing, Krebs cycle–dependent organism following closureof the anterior neuropore and onset of an active cardiovascular system. 141–143 Addition of mannoseand other nutrient substrates to the culture medium allowed these and other investigators to eliminateconfounding maternal factors and determine the direct capabilities of the conceptus to metabolizeand respond to glucose. 144 Hiranruengchok and Harris 146–148 have incorporated methods used toassess intermediary metabolism and energy production into a chemical exposure paradigm in whichseveral events and responses can be correlated to help define the mechanisms of toxicity. Exposureto the thiol oxidant, diamide, revealed that GSH oxidation, protein mixed-disulfide formation, andaltered glycolytic and pentose phosphate pathway enzymatic activities all correlate temporally. Thatfinding suggests a role for disrupted metabolism in the mechanism of diamide embryotoxicity.A considerable amount of important work has also been conducted in the laboratories of Sadler,Eriksson, and others. Those workers described potential mechanisms involved in glucosemetabolism 148–150 and toxicity 151 and factors that may relate to maternal insulin-dependent diabetesmellitus (IDDM) and diabetic embryopathy, 142–144 as well the consequences of other maternalmetabolic disorders, such as phenylketonuria (PKU). 152 These studies provide considerable importantinformation as to the role of maternal and conceptal biotransformation and the need for specificnutrients in fuel processing. In vitro systems have been beneficial in determining that some diabeticfactors, such as ketosis and hypoglycemia, probably affect the embryo directly; others, such assomatomedin inhibitors, likely alter embryonic nutritional pathways by interference with the functionof the VYS. 130–135 It has also been suggested by Eriksson and Borg 151 that mechanisms of© 2006 by Taylor & Francis Group, LLC

666 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONembryotoxicity related to hyperglycemia include the production of reactive oxygen species andmay also involve GSH status and the antioxidant systems described above.B. Other Systems and Approaches Employing Whole Embryo CultureThe examples cited above represent only a small part of the body of work that has in common theuse of rodent whole embryo culture systems as a means to probe mechanisms of developmentaltoxicity. The application of these techniques and approaches has also been covered in some detailin other sections of this book in conjunction with specific subject areas. Numerous additionalapplications and techniques employing rodent embryo culture will surely be reported in the future.Several mechanistic theories not explicitly covered in the foregoing sections, but which havebenefited by current use of whole embryo culture or will benefit from its use in the future include:• Morphological events such as neural tube closure, craniofacial anomalies, limb defects, and oculardevelopment 153–158• Functional/physiological alterations involving growth factors, pH regulation, angiogenesis, andhemopoiesis 159–162• Macromolecule synthesis and turnover, including protein synthesis; DNA synthesis, damage, andrepair; cell death; and stress protein responses 163–166• Receptor binding and function (described in detail in Chapter 14)• Gene expression, including gene product regulation, genetic alterations associated with patternformation, gene expression, sense-antisense manipulations, in situ hybridizations, and a myriad ofother molecular approaches, several which are covered in detail in Chapters 14 and 15Some examples of the novel approaches being employed in vitro for studies of mechanisms ofdevelopmental embryotoxicity involve use of the rodent whole embryo culture system. For example,this system has been used to evaluate the causes and repercussions of programmed cell deathproduced by such chemical agents as N-acetoxy-acetylaminofluorene. Thayer and Mirkes 166 (Figure16.6) have adapted novel in situ procedures to identify and characterize cells undergoing programmedcell death. An in situ procedure identified as the TUNEL (TdT-mediated dUTP-biotinnick end labeling) method pinpoints apoptotic cells in histological sections of any region of treatedand control conceptuses. A vital dye, Nile blue sulfate, is used for identification of changing regionsof cell death at the gross light microscopic level. The morphological and cytological characteristicof the dying cell populations can be further characterized by transmission electron microscopy(TEM, Figure 16.6b). Such methods are valuable for the study of mechanisms because they allowinvestigators to focus on the specific cell populations that seem to be affected by chemical exposure.A second example of new approaches to study mechanisms of developmental toxicologyinvolves the incorporation of molecular biology techniques into studies of normal and abnormalmechanisms of embryogenesis. Sadler et al. 167 have microinjected antisense oligonucleotides thatare specifically targeted toward the engrailed-1 (En-1) gene into GD 9 mouse embryos. Theseauthors have demonstrated that En-1 is involved in developmentally determined pattern formationin the mouse and that dysmorphogenesis produced by the microinjection of antisense oligonucleotides(Figure 16.7) is most likely related to elicited reductions in En-1 protein levels. Theseexperimental models may allow direct evaluation of the interactions between chemicals and genesin multifactorial causes of developmental embryotoxicity.1. Transgenic AnimalsOver the past 25 years, advances in the manipulation of the mouse genome have opened doors tounderstanding specific developmental pathways and their relationship to mechanisms of developmentaltoxicity. Transgenic approaches provide a more in depth insight into the roles of individual© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 667(A)cfmhfChABffCcDcFigure 16.6(A) The TUNEL method was used on tissue sections of control embryos and embryos exposedto 50 µg/ml N-Ac-AAF, at the 10-h time point. Methyl green was used as a counterstain. A: Nearlymidsagittal section through a control embryo. Three regions of the brain are shown: forebrain (f),midbrain (m) and hindbrain (h). Between the heart (c) and brain a small portion of one of thevisceral arches can be seen (arrowhead). B: Sagittal section of treated embryo. C: Highermagnification of head of embryo shown in A. Arrow indicates DAB-positive (brown-stained) cellsor cellular particles. D: Higher magnification of section shown in B. (B) Transmission electronmicroscopy (TEM) of tissues from a control embryo and an embryo exposed to N-Ac-AAF, 5 hafter exposure. A: Typical appearance of mesenchymal cells in a control embryo. Cellular processeswere abundant, but there were no obvious apoptotic cells or debris. B: Embryo exposedto 50 µg/ml N-Ac-AAF. Occasional apoptotic bodies were noted (arrowhead), and electron-dense,intracellular inclusions were common (arrows). C: Cell from embryo shown in B, clearly undergoingapoptosis (arrowheads). D: Embryo exposed to 200 µg/ml N-Ac-AAF. Occasional apoptotic cells(arrowheads) were observed, and extracellular, heterogeneous, electron-dense bodies werecommon (open arrows). E: Section from embryo shown in D. In addition to apoptotic cells(arrowhead), some cells had discontinuous cell membranes and small electron-dense nuclei(large arrows), suggestive of necrosis. (Thayer, J.M. and Mirkes, P.E., Teratology, 51, 418, 1995.With permission.)© 2006 by Taylor & Francis Group, LLC

668 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(B)ABCDEFigure 16.6 (continued)genes during the development of specific tissues. There are numerous methods by which transgenicapproaches can modify eventual developmental outcomes, including the overexpression or misexpressionof a specific gene in an ectopic tissue, expression of antisense transgenes resulting inhybridization with endogeneous mRNA and effectively silencing gene expression, and gene knockouttechnologies. In developmental toxicology, the use of transgenic animals is still in its infancy,but the technique is being employed ever more frequently. An example is the use of antioxidantenzyme (SOD, GSH-peroxidase, etc.) deletions and overexpressions to assess their effects onchemical sensitivity and dysmorphogenesis. Clearly, the use of transgenic mice will aid in understandingthe relationships between specific gene expression and toxicant exposures and responses.These approaches are covered in greater detail in Chapter 14.2. Proteomics and GenomicsAs proteomic and genomic technologies are being perfected, microarray analyses of conceptuses,embryos, or developing tissues can potentially yield vast quantities of descriptive data aboutresponses to teratogenic insult. In vitro manipulation allows for precise culture and treatmentconditions and would allow for more exact, teratogen-specific determinations of gene and proteinexpression. Many teratogens presumably produce abnormalities by inhibiting or hyperstimulatingspecific tissues. In all likelihood, this is a result of misregulation of gene expression. Microarrayanalysis could provide information very useful for elucidating the initiation and manifestation ofbirth defects.Understanding the structural modification of specific conceptal macromolecules followingchemical exposure can aid greatly in elucidating the biochemical and molecular mechanisms oftoxicant action. Application of new time-of flight mass spectrometry techniques such as SELDIand MALDI offer the potential of identifying and characterizing alterations to target molecules andthe pathways responsible for toxicity. As the proteomic databases are expanded and developed, thissignificant knowledge can be applied to mechanisms of developmental toxicology.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 669AmBhfhtII1.0 mmCmDEmFhfhmf1.0 mmFigure 16.7Phenotypes of mouse embryos injected with En-1 sense (A) or antisense (B) oligonucleotidesat the 4- to 6-somite stage (inset) and cultured for 48 h. (A) Embryo injected with En-1 senseoligonucleotides (100 nl, 25 µM) showing normal morphology, including expansion of the 4thventricle (large asterisk) and establishment of the 1st and 2nd pharyngeal arches (small asterisks).(B) Embryo exposed to En-1 antisense oligonucleotides (100 nl, 25µM), exhibiting abnormaldevelopment of the 4th ventricle and reduction in caudal extension (arrow). (C) Embryo exposedto En-1 antisense probes (100 nl, 60 µM), exhibiting craniofacial defects (white arrowheads) andhypoplasia of the 4th ventricle (arrow). Reduction in caudal segments is evident from the forelimbbudto the tip of the tail. (D) Higher magnification of the tail region of the embryo in (C). Kinksare apparent in the neural tube (black arrows), and caudal segments are missing (white arrow).(E and F) Embryos exposed to sense (E) and antisense (F) En-1 oligonucleotides at the earlysomite stage, cultured for 48 h, and stained with anti-En antibodies. Note that staining is decreased(arrows) in the neural tube and somites of the antisense-treated embryo, but that a band ofstaining remains at the midbrain-hindbrain junction (asterisk). (A through F: f, forebrain; h,hindbrain; ht, heart; l, limb bud; m, midbrain.) (From Sadler, T.W., Liu, E.T., and Augustine, K.A.,Teratology, 51, 292, 1995. With permission.)© 2006 by Taylor & Francis Group, LLC

670 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONC. Nonmammalian Whole Embryo Cultures1. Caenorhabditis elegans CulturesThere are numerous advantages to the use of the relatively ubiquitous nematode, Caenorhabditiselegans (C. elegans). Because C. elegans can be hermaphroditic, self-fertilization is possible.Cultures are relatively easy to grow and maintain, with a generation time of only 3 days. The wormsare also transparent throughout their entire developmental period, making observations of internaldevelopment possible with the use of a light microscope. The C. elegans genome is very small (97megabases) and has been completely mapped. Another advantage is that the mature C. elegans hasrelatively few somatic cells, approximately 1000 in adults, and embryonic cell lineages have beencarefully defined, 168,169 so each mature cell can be traced back to its embryonic precursor. Thisadvantage was exploited in laser ablation experiments, where specific cells were killed and developmentwas followed to adulthood. This approach allows for understanding cell-cell interactionsduring development, an approach that is very difficult, if not impossible, in other animal modelswithout further genetic manipulations.While many developmental pathways that may be relevant to developmental toxicology can bestudied in C. elegans, this model has primarily been utilized in developmental biology. Pathwaysstudied in C. elegans have concentrated on receptor tyrosine kinases, 170,171 transforming-growthfactor β, 172–174 Notch and Delta, 175,176 Wnt, 177 and apoptosis 178,180 pathways. The use of C. elegansin developmental toxicology has predominantly been limited to embryo toxicity screens. Theorganism has not been widely used as a model for the evaluation of mechanisms of mammaliandevelopmental toxicity, despite its obvious advantages.2. Drosophila CulturesDrosophila larvae have been used extensively in studies of the control and regulation of geneexpression and pattern formation during embryogenesis. The virtues of mechanistic toxicity studiesin Drosophila mirror those of C. elegans toxicity testing. Drosophila are inexpensive and generallyeasy to maintain and grow, but the greatest advantage to their use is the wealth of informationabout them that has accrued since the 1920s. Drosophila genetics were first elucidated by T.H.Morgan at the University of Columbia, and in the 1970s Drosophila studies led to the discoveryof early egg organization and body patterning. Among the genes that were discovered during theseinitial experiments, the homeobox genes were found to be highly conserved in species rangingfrom flies to mammals. Later, it was found that many other Drosphila genes had mammalianhomologues, providing an easily manipulable nonmammalian model for developmental biology.Such knowledge and relatively well-understood developmental genes and specific pathways shouldrender Drosophila valuable for furthering an awareness of developmental toxicology mechanisms.However, much like C. elegans, Drosophila has not been used to elucidate developmental toxicologymechanisms but rather has been primarily employed as a potential developmental toxicity screeningtest organism.3. Amphibian Embryo CultureSome of the earliest studies of mechanisms of development were conducted in amphibians. Amphibiansare likely have somewhat more relevance to humans than C. elegans or Drosophila becausethey are vertebrates and thus develop more like humans. These simple organisms have been usedto study neural crest cells, osteogenesis, vasculogenesis, cardiogenesis, and development of gutderivatives (e.g., liver and pancreas), all of which have much in common with human developmentalprocesses.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 671In particular, the African clawed frog, Xenopus laevis, has figured prominently in developmentalstudies, as it has served as a model system for events such as fertilization, gene expression, patternformation, organogenesis, morphogenesis, and metabolism. Recent use has been made of the abilityto incubate and manipulate Xenopus eggs for developmental toxicity studies. The approach, identifiedas the FETAX system (Frog Embryo Teratogenesis Assay Xenopus) is intended for use as arapid screen for chemical teratogens but has recently been applied for use in more mechanisticstudies. Fort and colleagues have demonstrated thalidomide-induced limb reduction defects withthe FETAX system, 181 although, no mechanistic findings have come from these initial studies. Manyother chemicals have been tested using the FETAX system, including the environmental contaminantmercury, 182 the pesticide and redox cycler paraquat, 183,184 cadmium, 185 the carcinogenbenzo[a]pyrene, 186 and caffeine. 187 One of the potential advantages to the FETAX system, as withavian cultures, is that it allows xenobiotic effects to be followed throughout development fromfertilization to hatching and senescence because of the relatively short lifespan of X. laevis.Chemical-induced behavioral deficits can be determined in a shorter time span that would be thecase with mammals. Another advantage to the FETAX system, like C. elegans and Drosophila, isthe relatively low cost and the high output, as numerous concentrations of test chemicals can beassessed concurrently on many developing eggs, giving enormous amounts of data. In its currentform, however, the FETAX test system has a number of weaknesses, including inadequate interlaboratoryreproducibility, which must be addressed if it is ever to become a reliable mainstreamdevelopmental toxicity assay.4. Fish Embryo CultureThe use of zebrafish (Danio rerio) in developmental and developmental toxicity studies is relativelynew but may have advantages for modeling vertebrate development, including low cost and highoutput. Toxicology testing can be simplified, in that embryos are extremely permeable to toxicantsdissolved in water. A handful of toxicology studies have been performed on zebrafish, includingevaluations of cadmium, 188 TCDD, 189 and 4-chloroaniline. 190 Some genetic manipulations ofzebrafish have also been performed with excellent results, which have encouraged their increaseduse in toxicological studies. One unique benefit of zebrafish in toxicological assays is that duringdevelopment they are transparent, allowing specific organs and tissues to be visualized and moreeasily studied (i.e., with light microscopy). Zebrafish manipulation with reporter gene constructscoupled with this transparency during development allow for detailed study of a specific toxicant’seffects on certain tissues. The green fluorescent protein (GFP) transgenic fish was first constructedin 1995. Since then it has been utilized in studies analyzing gene expression patterns, and tissueand organ development, dissection of promoters and enhancers, cell lineage, axonal pathfinding,cellular localization of protein products, chimeric embryo and nuclear transplantation, and cellsorting. 191 The use of zebrafish in studies of mechanisms of developmental toxicology has, thusfar, been limited. Mattingly and coworkers 195 developed an aryl hydrocarbon receptor specific GFPconstruct and injected it into developing zebrafish. Following fertilization, zebrafish were exposedto TCDD (10 nM). Early in development (72 h), GFP was detected in the eye, nose, and vertebrae,while no GFP expression was detected in DMSO-treated controls. In longer-term examinations(5 days), terata coincided with the previously demonstrated regions of GFP expression and includedcraniofacial and vertebral defects.5. Avian Embryo CultureClassical studies in developmental biology have exploited the chick as a model for which a largenumber of developmental processes and mechanisms have been defined. Accessibility for experimentalmanipulation and observation have contributed to the success of these experiments. The© 2006 by Taylor & Francis Group, LLC

672 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONchick has also been used as a model for classic studies in teratology. Incubation of eggs with 6-amino-nicotinamide, 192 retinoic acid, 193,194 and other agents has provided fundamental knowledgeof developmental mechanisms and their perturbation with chemical agents. More basic studies,including exposure of chick embryos to ethyl alcohol and retinoic acid derivatives, followed by anevaluation of the function and behavior of cranial neural crest cells, have looked at mechanismsinvolving specific cell types and related processes. Coupled with use of in vitro cranial neural crestcell cultures, this approach has provided considerable information showing that the neural crestmay be sensitive to chemical agents that produce or increase the production of reactive oxygenspecies. This sensitivity may be due to deficiencies in intrinsic detoxication capacity, particularlyinvolving superoxide dismutase. In many respects, analogous studies can be conducted for earlypostimplantation embryogenesis in both avian and rodent species. The chick model has the distinctadvantage, however, of being able to continue the developmental process through hatching andsubsequent maturation.In general, most nonmammalian culture systems have numerous advantages, giving them greatpotential, which, unfortunately, has not been fully utilized in the study of toxicological mechanisms.When the genomes of these organisms are completely or nearly completely mapped, their utilizationfor mechanistic studies should increase. They may eventually be fashioned into screening systemsof choice, providing useful information regarding developmental toxicity mechanisms.D. Organ and Tissue CultureOrgan culture and tissue culture are terms often used interchangeably to describe the in vitro growthof structurally and functionally intact subsets of a complete organism. The levels of biologicalorganization encompassed by this definition range from intact, functional, normally vascularizedcomplex organs to essentially avascular, relatively homogeneous tissue fragments. Organ cultureprovides numerous advantages for the study of mechanisms in developmental toxicity. It allowsfor the maintenance of a functional unit of an organism in a controlled environment, while retainingan intact architectural structure and without the need of perfusion via an intact vasculature. 196,197Specific functional and morphological processes carried out by many organ systems are influencedby chemical and physiological factors produced endogenously and also those provided by otherorgans and tissues. The roles and functions of these factors can be evaluated by their selectiveaddition or removal from the culture medium.Successful elucidation of mechanisms of toxicity produced through chemical interactions ultimatelyrequires a system in which the temporal and spatial relationships of morphological, biochemical,and molecular alterations can be evaluated. This approach necessitates the removal oftissues and organs from confounding influences, so that their interactions can be identified andcharacterized. Several embryonic organs can now be dissected free and grown successfully inculture, still meeting a number of selected experimental criteria. These include maintenance ofsimilarities in growth and differentiation, compared with in vivo patterns, and retention of the timedsequence of production of organ-specific products. These techniques are useful in manipulating thesystem to probe mechanisms and are aided by the ability to determine the exact time and durationof exposure, the precise stage of development at the time of exposure, and the degree of chemicalmetabolism prior to reaching organ targets.Organs and tissues that have been used successfully in in vitro culture applications for the studyof mechanisms of developmental toxicity include limb buds, limbs, bone, heart, palatal shelves,visceral yolk sac, eyes, sex organs, and embryonic tooth germs. Many of the initial culture techniqueswere based on Trowell’s method, 198 where the organ is grown attached to a membrane in ahumidified atmosphere. Many of these techniques have been subsequently modified to use submergedcultures and a roller apparatus in defined media to improve morphological and biochemicalfidelity in comparison with parallel in vivo developmental events (Figure 16.8). Cocultures withother organs or cells, dissociation of specific cell types from the organ, and the use of nontoxic© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 673Trowell MethodOrgan CultureLimb budLimbBoneHeartPalateVisceral yolk sacEyeReproductive organsTooth germRoller CultureFigure 16.8Organ culture of numerous structures can be initiated from embryos obtained during a broadrange of developmental stages, depending on experimental needs. Organs can be grown inculture on static grids or filters using the Trowell method (Exp. Cell Res., 16, 118, 1959), asdepicted in the upper part of this illustration, or in roller cultures, as shown in the bottom portion.Growth and differentiation are generally superior in roller cultures.bioactivation systems extend the utility of these experimental models. They allow for the study oforgan-organ interactions and direct xenobiotic effects on interrelated cells and tissues. Endpointsfor evaluation of organ cultures range from gross morphogenetic changes to cellular effects,biochemical alterations, metabolic transitions, and physiological responses to the molecular basesof differentiation, as well as the genetic events that regulate or are regulated by xenobiotics.1. Limb Bud and Related Perinatal Bone CulturesA number of toxicants produce persistent malformations involving the developing limbs. Thissuggests that these organs may be particularly sensitive to chemicals during development and may,therefore, be good systems in which to study mechanisms of developmental toxicology. 199,200Undifferentiated limb buds from the chick were first successfully grown in culture by use of aplasma clot environment. 201 Later, chick, rat, and mouse limb buds were cultured by use of theTrowell method, 198 where limbs were grown attached to Millipore ® filters in a humidified environment.Plasma clot techniques are difficult to work with in studies involving the addition of chemicalagents or radioisotopes because the clot material interferes with chemical distribution and quantitation.Initial studies by Aydelotte and Kochhar 202 established the limb bud as an important earlytool for the mechanistic study of chemical induced limb malformations. Limb buds removed from40 to 45 somite rodent embryos provided the greatest degree of growth and differentiation andcould be cultured successfully for up to 6 d, when they reached optimal differentiation (Figure16.9). Organ cultures of this type produce the distinct patterns of chondrogenesis and establishmentof the skeletal anlage seen with normal limb growth in vivo, but they require a longer period tocomplete differentiation and undergo developmental growth. The expected sequence of developmentand the programmed production of unique products of differentiation, however, show that thecultured limb buds bear close morphological and biochemical resemblance to the intact limb invivo. Subsequent modifications to the Trowell method by Neubert and Barrach 203 utilized a submergedroller bottle culture system. Their modification effectively improved the cultured limbs’morphological appearance, decreased the time needed for optimal growth and differentiation, andprovided greater accessibility of organs and tissues to growth factors, nutrients, oxygen, and testchemicals. This approach effectively overcomes many of the morphological infidelities associatedwith growing organs attached to a solid filter substrate and produces differentiated structures in© 2006 by Taylor & Francis Group, LLC

674 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1 mmCO0,10,31,0Figure 16.9Effects of (bis[tri-n-butyltin])oxide (TBTO) on the differentiation of mouse forelimb buds in organculture. The cultures were initiated with explants from 12-d-old mouse embryos, and the cultureperiod was 6 d. TBTO was present in the culture medium for the entire culture period. (A) Controllimbs (CO); all cartilaginous bone anlagen are well developed. (B) 0.1 µg/ml TBTO in the medium;in comparison with controls, development of the phalanges was found to be inhibited, and thescapula was less well developed. (C) 0.3 µg/ml TBTO in the medium; drastic impairment ofdifferentiation of the paw skeleton and the scapula; in some cases ulna and radius were alsoaffected. (D) 1.0 µg/ml TBTO in the medium; almost no morphogenetic differentiation occurs;only the cartilage of the humerus and parts of the ulna and radius are recognizable. (FromKrowke, R.R., Bluth, U., and Neubert, D., Arch. Toxicol., 58, 125, 1986. With permission.)about half the time (3 d) when compared with the Trowell cultures. 198 An example of the degreeof growth and differentiation that can be obtained is shown in Figure 16.9. Controls (CO) haveundergone significant morphogenesis in vitro, while (bis[tri-n-butyltin])oxide (TBTO)-treatedorgans failed to grow or differentiate properly. 204Endpoints used for the effective study of mechanisms of limb malformation are intimatelyrelated to morphometric structure and to the patterns of gene expression and biochemical synthesisthat result in the formation of the cartilaginous anlage of long bones and digits. Investigations intomechanisms of limb teratogenesis have focused on the evaluation of two main processes importantfor normal and abnormal limb growth. First, the morphological and functional differentiations ofcells characteristic of the mature limb bud occur developmentally as a consequence of temporallyand spatially controlled gene expression. The patterned production of new enzymes, structuralproteins, and regulatory elements results in accumulation of the structural and functional elementsnecessary for altering, organizing, and establishing morphological form. Second, as a result ofprogrammed gene expression, the actual spatial and functional organization of the elements intocartilaginous anlage is accomplished. These events are accompanied by genetic regulation ofprogrammed cell death, resulting in the final sculpting of the finished limb. Selective alteration in© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 675any one or a combination of these processes as a result of exposure to chemical agents may representa relevant endpoint to be used as a probe for specific mechanisms of action of chemical embryotoxicity.In this context, the limb culture technique has been used to investigate the effects of over60 agents, such as retinoic acid, 205 retinol, 206 cyclophosphamide, 207,208 6-mercaptopurine riboside, 209sodium salicylate, cytosine arabinoside, eserine sulfate, 210 thiabendazole, 211 and many others, asreviewed by Friedman. 200 Several clues to the etiology of the toxicity were derived from thesestudies, including effects on mitochondrial function, maternal electrolyte and water balance, cartilageand extracellular matrix biochemistry, gene expression and regulation, receptor function andpattern formation, programmed cell death, and others, although no common mechanism wasidentified. It is highly unlikely that a common mechanism will be found. Chondrogenesis and thetissue localization of various forms of collagen, 212 glycosaminoglycans, 213 chondroitin sulfate, andother products, 199,207 specifically associated with cartilage formation have been studied. The advantagesof organ culture for quantitation and characterization of these biochemical products can nowbe effectively combined with other in vivo and in vitro techniques by which temporal and spatialexpression of selected genes can be related to functional outcome. Immunohistochemical localization,gene expression (pattern formation), in situ hybridization, and other molecular approachescan now be used in attempts to understand the effects of chemicals on limb development in organculture.The possible mechanisms to be evaluated and related endpoints to be measured in studies oflimb dysmorphogenesis are too numerous to be covered exhaustively in this chapter. The reader isreferred to several reviews and texts on this topic. 199,200,203 An overview of possibilities can be bestillustrated by drawing from a selected few examples from the vast variety of studies conductedwith this system using retinoic acid and its related metabolites. Earlier studies in mice establishedthe role of retinoic acid as a limb teratogen and defined gross morphological effects with respectto patterns of progressive differentiation. 202,214 Single maternal doses of retinoids produced phocomeliaor partial phocomelia in patterns that followed differential gradients established in foreandhindlimbs, depending on gestational age and time of exposure. The role of retinoids indevelopment and the endpoints and processes that may be investigated for studies of mechanismsof developmental toxicity have been reviewed in detail. 215 Experiments in vitro using mouse andchick limb culture systems determined that grafting a portion of the ZPA (zone of polarizing activity)of a region from the posterior limb bud to the anterior margin of another limb bud altered thegradient of RA distribution within the limb and resulted in a mirror-symmetrical duplication ofdigits. The relationship with RA distribution in the limb and a causal effect on the observed patternof development was confirmed by obtaining a similar result through grafting a bead coated withall-trans-retinoic acid to the anterior site normally reserved for the ZPA graft. These studiessuggested that RA was acting as a morphogen, a diffusible chemical distributed in a discreteconcentration gradient that dictates the final form of the organ through concentration-dependentinteractions. Such results also indicated that excess RA could be exerting its teratogenic effectsthrough disruption of the normal chemical gradient in the limb.Antibodies directed against a RA-bovine serum albumin antigen have allowed the immunohistochemicallocalization of endogenous and exogenously added RA in vitro. Subsequent studieshave shown, however, that although patterns of RA distribution correlate well with various aspectsof pattern formation, RA cannot be a complete morphogen. That is because of the requirement fora number of other factors essential for pattern formation and differentiation. Considerable interestand insight have been generated from the observations that exogenous application of retinoids tothe anterior margin of the chick wing bud has the same effect as transplantation of the ZPA orcoated beads. In vitro limb culture methods can be used in these ways for testing the effects ofchemical exposure on other morphogenetic and cellular processes.Other aspects of RA effects on the altered morphology and differentiation of the limb bud invitro have focused on characterization of the role and distribution of binding proteins and specific© 2006 by Taylor & Francis Group, LLC

676 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONclasses of retinoic acid receptors, and the mechanisms that regulate them. 215 Specific changes inthe types and distribution of retinoic acid receptors have been identified by in vitro methods.Approaches for the study of mechanisms of receptor and binding protein involvement in teratogenesishave included the histochemical and immunohistochemical localization of receptors andbinding proteins. They have also involved the use of in situ hybridization techniques to determinethe specific pattern-related distribution of receptor and cytosolic retinoic acid binding protein mRNAexpression under different exposure conditions. A suspected role for RA as a tissue morphogenand the probable role of this naturally occurring compound in regulating gene expression responsiblefor chondrogenesis and pattern formation have led to a number of studies aimed at understandingthe effects of RA on gene expression. Observations that embryotoxic agents can, indeed, act throughregulation or disruption of programmed gene expression have fueled investigations of these eventsas mechanisms for the action of teratogenic agents. These studies have provided considerablemotivation for the continued study of mechanisms for understanding RA teratogenicity. The abilityto coordinate specific enzyme activities, macromolecule synthesis, gene expression, and pharmacologicaleffects with observed patterns of differentiations in vitro makes limb culture a usefulresearch tool for studying mechanisms of limb development.Sequential expression of gene products directs the patterning of the limb and the delineationof the chondrogenic anlage. This must be followed by the events associated with programmed celldeath that result in the final sculpting of the digits to result in the functional hand, wing, or foot.In vivo and in vitro observations of incomplete or abnormal sculpting as a result of chemicalexposure make the mechanistic understanding of this process extremely interesting. Differentiationof specific cellular and molecular characteristics results in programmed cell death and alteredcellular function. Several important studies have been conducted in vivo and utilizing the culturedlimb bud to help delineate the changes that occur as cells commit to die. 2162. Palatal CulturesIncomplete closure of the primary and secondary palates is frequently seen in clinical manifestationsof malformation (cleft palate and cleft lip), both in humans and in experimental animal models. Anumber of chemical agents, including TCDD, 217 glucocorticoids, 218 phenytoin, 219 and retinoicacid, 220 have been shown to selectively induce incomplete palatal closure. Excision of the intactpalatal shelves and their subsequent development in vitro have aided greatly in understanding themechanisms that regulate palatal growth and differentiation in vivo. Palates were first successfullygrown in culture by use of the Trowell method, where palatal shelves were placed intact on aMillipore membrane and allowed to undergo fusion. 221 Several drawbacks of this method becameapparent in that only differentiation and morphology directly related to shelf fusion could beevaluated. The more global processes of shelf elevation, cell movement and related differentiation,and functional interactions could not be evaluated. The modified serum-free cultures of Saxen andSaxen, 221 Saxen, 222,223 and Abbot and Buckalew 224 improved growth and differentiation but fell shortof the more important advances that were realized by Shiota et al., 225 who grew whole palatalexplants in submerged, rotating cultures.Because exposure to dioxins in vivo, particularly 2,3,7,8-tetrachlorodibenzo-p-dioxin resultedin an increased incidence of cleft palate, an in vitro technique was developed to investigate themechanisms of palatal closure and biochemical, molecular, and morphological disruption by chemicalagents. Intact palatal shelves, removed from rodent embryos on GD 10 to12, have been shownto elevate and fuse in in vitro cultures in much the same manner that they do in vivo. Influenceson the palatal closure process have been characterized and have been found to include inductionthrough glucocorticoids, Ah-receptor responsive chemicals, and other metabolic inhibitors. 226,227The Ah responsiveness was determined by parallel experiments using selected strains of mice thatcharacteristically have different Ah receptor phenotypes. Retinoic acids have also been evaluated© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 677with this approach to elucidate the role of growth factor expression (EGF, TGF-α, and TGF-β) inaltered palatogenesis. 228,229 These studies could not demonstrate that perturbation of growth factorreceptors was the sole cause of the altered palatal growth and differentiation, but did indicate thattheir role is important. The perturbation of neural crest cell migration to the appropriate mesenchymalenvironments has also been investigated, suggesting additional possible mechanisms bywhich palatal development can be altered. 230 Molecular studies are easily conducted by use of thissystem but are not covered in this section.3. Visceral Yolk Sac CulturesThe visceral yolk sac plays a vital role in the nourishment, metabolism, and protection of thedeveloping conceptus, especially during the sensitive period of early organogenesis. The invertedVYS epithelium is structured to have the brush border facing the decidual blood lumen. Thisfacilitates the uptake by pinocytosis and the enzymatic degradation of maternal histiotroph (proteinsand other macromolecules derived from the decidual cell mass and from a transudate of the maternalserum) and the transport of carbohydrates, amino acids, and other nutrients. The rodent VYS isdistinct from those of other mammalian species because it retains its form and function until term.A number of embryotoxins and teratogens have been reported to alter VYS function, and theireffects on this organ have been suggested as central to their mechanisms of toxicity. The teratogentrypan blue was first shown to elicit its deleterious effects via inhibition of pinocytosis and/orproteolysis by the VYS. 231 Since these reports, other agents (e.g., leupeptin, anti-VYS antibody,chloroquine, diazomethyl ketones, E-64) have been shown to elicit their effects through inhibitionof pinocytosis or through inhibition of lysosomal proteolysis. 130–137,232–234Because of its direct contact with the culture environment, a number of direct exposures andmanipulations can be performed on the VYS in whole embryo culture without compromisingconceptal viability and function. Many of the characterizations of VYS function, however, used adetached, free-floating VYS in a roller culture system. These studies were used successfully todetermine the extent of proteolytic and pinocytotic activity in the VYS and to characterize changesin these activities elicited through chemical exposure or anti–yolk sac antibody recognition.114,115,233,235 Similar approaches were taken to evaluate the mechanisms of uptake of othermacromolecules, such as antibodies, ferritin, and enzymes. Relatively few studies have beenconducted to determine specific effects of chemicals on the VYS epithelium, endothelium, or stemcell populations, even though considerable evidence exists to support an extensive role of the VYSin several nutritional, metabolic, and physiologic functions.Organ culture of the VYS can also be accomplished at a level of biological organizationintermediate between whole embryo culture and the isolated VYS. These investigations utilize“giant yolk sacs,” in which conceptuses are explanted at an early somite stage (GD 9 in the rat),and the early or presomite-stage embryos are surgically ablated using dissecting needles. 139 Returnof the severed conceptus to culture results in the continued growth of only the VYS. Such preparationshave been shown to retain numerous vital functional and structural characteristics withoutinterference from the embryo proper or interactions associated with the vitelline circulation. Becauseof its ability to trap and sequester products in the exoceloemic fluid, the closed yolk sac that resultsfrom this procedure may have a number of advantages in studies of mechanism involving directionaltransport across the VYS epithelial barrier. It has not been clearly delineated, however, how a lackof vitelline circulation and possible hydrodynamic factors might affect VYS function or response.Nevertheless, such preparations could be of use in determining the mechanisms and effects ofchemical induction of avascular yolk sacs. 161Organ cultures similar to the examples cited above have been used to investigate probablemechanisms that are specific to the functions of other organ systems. Some examples are brieflymentioned in the following sections.© 2006 by Taylor & Francis Group, LLC

678 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION4. Sex OrgansInvestigations have looked into mechanisms of the normal and abnormal morphogenesis of theprimary sex organs. These have included in vitro studies of differentiation of the Wolffian andMüllerian ducts of avian and mammalian species. 2365. HeartDifferentiation and functional responses in heart that may be perturbed by chemical embryotoxicantsinclude coordinated muscular contractions, gap junctional communication, and electrophysiologicalcoupling. The latter aspect has been reviewed 237 in terms of reaggregation and electric coupling ofdissociated cells, although the direct effects of chemical embryotoxicants have not been systematicallyevaluated.6. Eye and LensThe in vitro approaches taken to study the effects of toxicants on this organ can be coordinatedwith other studies of neural tube closure and central nervous system development. 153 Sensitivity ofthe avian and human lens with regard to cataract formation following exposure to rubella virus wasevaluated in vitro by use of eyes collected from human embryos obtained from therapeutic abortions.196 Verification of direct viral involvement in the mechanism of cataract formation was possiblethrough use of the organ culture model.7. Embryonic Tooth GermEarly tooth germ formation and dentition can be studied in whole organ cultures in vitro. This canbe done on solid supports or in flotation cultures. 238,2398. Spinal CordPostmeiotic growth and differentiation of neuronal cells and structures was examined in culturesof spinal cord from the chick. 240 And with in vitro glutamate treatment, neurite outgrowth andgrowth cone motility from the mouse spinal cord were significantly decreased, implicatingglutamate as a regulator of neuronal differentiation, development, and function. 2419. Lung BudsBranching morphogenesis and N-myc expression were inhibited with TGFβ1 in cultures from theseorgans from the day 11.5 mouse. 242 Dihydrotestosterone (DHT) significantly increased lung budbranching, a phenomenon that was reduced by flutamide, an androgen receptor competitive inhibitor.243E. Cell Culture and Clonal AnalysisMorphological, histological, and ultrastructural analyses have demonstrated that teratogenicity orembryotoxicity may be elicited because of the particular sensitivity of a single cell type withincomplex tissues of the conceptus. The ultimate description of biochemical and molecular mechanismsin the complex organism may depend on an understanding of the specific functions andbehaviors of individual cell types and how they are induced to change in response to endogenousand exogenous cues. The utility of employing isolated primary cell cultures and clonal cell linesto meet this end will depend on the direct demonstration that they behave in vitro in a manner© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 679similar to their behavior in the intact organism. Some of the first successful attempts at clonalanalysis to demonstrate similar behavior in vivo and in vitro involved the culture of quail myoblastsfor which various degrees of success were obtained, depending on the substrate and the constituentsof the culture medium used. 244 A number of cell types can now be dissociated from the intactorganism, conceptus, organ, or tissue, and successfully grown in culture, where the proper choicesof medium and substrate are still the critical components in determining success. Techniques forprimary culture have been established for heart myoblasts, 245 chondrocytes, 246 cartilage, 247 pigmentedretina, 248 lens, 249 and cranial neural crest. 250,251Use of primary embryonic cell cultures and established cell lines in toxicology usually focuseson proliferation or on the progression of differentiation into a specific cell type, but few haveaddressed the effects of teratogens on embryonic stem (ES) cells. Since ES cells are pluripotent,effects of specific toxicants may perturb signal transduction pathways involved in specific cell fates,resulting in the promotion of cell death, decreased proliferation, or faulty differentiation. In anycase, such effects could result in severe birth defects and embryotoxicity. ES cells have been usedto study PCB toxicity in neurogenesis. Endosulfan and 2,2′,4,4′,5,5′-hexachlorobiphenyl (2,4,5-HCB) both caused an inhibition of neurite formation in a dose-dependent manner and decreasedintracellular gap junctions and connexin 43 expression, implicating intracellular communication inearly neurogenesis. 252 The use of ES cells in mechanistic studies may serve as a more appropriatein vitro model of teratogenesis than other established cell lines. Measurement could be assessedby morphologic, biochemical, or genetic markers following toxicant exposure. These approachescould yield a great deal of information about sensitive pathways of toxicant-induced dysmorphogenesisas well as highlight the role of these pathways in development.Hypotheses relating to the proposed mechanisms of several known teratogens have implied thatalterations in cranial neural crest cell (CNC) migration, differentiation, and function are probablecauses for observed developmental anomalies. The clonal primary culture of CNC was originallydescribed in Japanese quail, using procedures such as those reported by Loring et al. 250 andGlimelius and Weston. 251 With this method, the distal portion of the trunk (containing the last 8 to10 somites) is dissected free. It is then exposed to mild enzymatic digestion (pancreatin) to loosentissue adhesions and allow the neural tube to be physically pulled free from the notochord andectoderm. Additional treatments (trypsin, collagenase) may then be needed to remove other adherenttissues before placing the neural tubes in media-filled plastic or collagen-coated petri dishes forculture. Various culture media have been used successfully, although, all require an appropriateamount of serum, 10-d chick embryo extract, and glutamine supplementation for optimal growthand differentiation. The relative proportions of serum and chick embryo extract affect both cellproliferation and cell cluster formation. This methodology takes advantage of the inherent highproliferative and rapid migratory properties of CNC cells and produces a very homogeneouspopulation that will respond to a number of chemical and environmental stimuli (Figure 16.10).Chemical agents, such as the tumor promoter 12-O-tetradecanoyl phorbol-13-acetate (TPA),have been shown to promote cell proliferation while causing significant delays in differentiation,as manifested by alterations in melanocyte production. 251 Davis et al. 253–255 utilized the avian CNCculture system to investigate the role of free radical production and oxidative stress in the cellularpathogenesis of ethanol (EtOH) and retinoid (isotretinoin and 4-oxo-isotretinoin)-induced craniofacialabnormalities. They showed that treatment of CNC with these agents in vitro is accompaniedby the significant generation of reactive oxygen species, producing cellular blebs, loss of normalCa 2+ homeostasis, and decreased viability.This system allows investigators the flexibility to measure multiple biochemical and morphologicalendpoints and to alter several environmental conditions to probe mechanisms of ROSgeneration and cellular response. Using these approaches, Davis et al. 254 determined the CNC tobe devoid of superoxide dismutase and catalase, two important antioxidant enzymes critical forelimination of ROS. Addition of these enzymes directly to the culture medium provided significantprotection against ROS-mediated damage. In addition, those investigators were able to characterize© 2006 by Taylor & Francis Group, LLC

680 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1ABCDFigure 16.10Migration of neural crest cells from the neural tube is shown in this series of phase contrastmicrographs of a single living neural tube explanted in vitro. To indicate polarity, blood charcoal(arrows) was applied to the ventral (notochordal) surface of the neural tube before it wasexplanted. One of the lateral surfaces of the neural tube is in contact with the collagen substratumof the petri plate. (A) No cells migrate from the explant during the first 2 h. (B) After 8 h,mesenchymal cells being to migrate from the dorsal aspect of the tube. (C) Within 20 h, hundredsof cells have emigrated from the neural tube, primarily from the dorsal surface. The few mesenchymalcells seen along the ventral surface arose from an area damaged while the tube wasbeing explanted. In addition, one end of the tube has flattened to form an epithelial sheet. (D)The mesenchymal outgrowth was isolated by carefully dissecting away the neural tube and itsflattened epithelial end. The remaining primary outgrowth is grown for an additional 2 d untilenough cells are available for routine cloning procedures. (From Cohen, A.M. and Konigsberg,I.R., Devel. Biol., 46, 262, 1975. With permission.)related alterations in intracellular Ca 2+ homeostasis. This was done by use of ionophores andsulfhydryl reagents, and it provided detailed ultrastructural descriptions of the accompanyingmorphological changes.The primary culture of cerebrocortical neural cells has also been used to evaluate mechanismsof methyl iodide–induced neurotoxicity, as reported by Bonnefoi. 256 Endpoints for evaluation inthese studies focused on the role of intracellular cytosolic and mitochondrial GSH status in themechanism of methyl iodide toxicity. Addition of a series of antioxidants and correlations to cellinjury, as indicated by lactate dehydrogenase leakage, showed that manifestations of toxicity werecorrelated with a minimal 50% reduction in mitochondrial GSH and were not altered by cytosolicGSH depletion. Similar approaches and an expanded palette of measurable endpoints make theisolation and culture of clonal cell lines from the developing embryo a valuable tool for the studyof mechanisms of developmental toxicity, especially when effects can be correlated with the findingsof companion in vivo studies.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 681Micromass Cell CultureMidbrainABGD12Limb budsCell dissoclationABFigure 16.11Schematic diagram showing the procedure for midbrain or limb bud micromass cultures. Limbbuds or midbrains are dissected from GD 13 rat embryos or GD 12 rabbit embryos anddissociated to a suspension of single cells in medium containing trypsin and EDTA. (A) Highdensitycell droplets (micromasses) are cultured in small petri dishes or 24-well culture platesin a humidified incubator at 37°C with a 5% CO 2 atmosphere for 1 to 1.5 h. (B) After cells haveattached, cultures are flooded with medium, to which can be added test chemicals, inhibitors,modulators, bioactivation systems, and growth factors. (Adapted from Kistler, A. and Howard,W.B., Methods Enzymol., 190, 427, 1990. With permission.)A more recent advancement in the culture of mammalian embryonic cells for the study ofmechanisms has been the use of micromass cultures. This system, as with many other in vitroapproaches, has been touted as a possible rapid screen for teratogenic agents 257–259 but may findgreater utility in mechanistic studies. This method involves the removal of midbrains (CNS) andlimb buds (LB) of 34 to 36 somite rat or mouse embryos by microdissection, followed by successivewashing in calcium- and magnesium-free phosphate buffered saline solution and trypsin to weakencell adhesions (Figure 16.11). The cells are resuspended in an appropriate culture medium, dissociatedinto individual cells by repeated titration through a Pasteur pipette, and filtered on nylonmesh to remove clumps and provided a suitable starting cell suspension. Cell suspensions are platedon plastic petri dishes at concentrations of 1 × 10 5 to 4 × 10 5 cells per 20 µl in culture mediumdrops. Ninety minutes after initial plating, additional culture medium containing the test compound(and possibly an S-9 microsomal bioactivating system and additional modulators) is added. However,the S-9 mix can be toxic and therefore must be limited to a specific period of culture. 257 Atthe end of a 5-d culture period, the medium is removed, and the cells are fixed and stained formorphometric analysis.While midbrain cultures generally provide an excellent means to gather qualitative data, limbbud micromass cultures are typically stained with Alcian Blue to visualize regions of activechondrogenesis. Stained chondrogenic foci can be eluted and then measured spectrophotometricallyto provide quantitative data. In regions of initial plating of the LB micromass, massed mesenchymalcells generally differentiate in place, but other cell types will proliferate at the edge of the massand spread outward (Figure 16.12). This finding provides information about two distinct populationsin the same culture — a proliferating population and a differentiating population — which mayprove useful for various toxicological assays. When stained with mercury orange to assess smallbiothiol content (i.e., GSH and cysteine), cells in the differentiation zone are seen to contain muchless GSH than cells in the proliferative zone (Figure 16.13). This may predispose differentiatingcells to greater damage during oxidative stress. 260 Other biochemical analyses can be performedon these two populations and may be extrapolated to in vivo exposures.Numerous agents have now been tested with the micromass approach, but mostly duringscreening for teratogens, the intended use of the assay. 257–259 Validation of these procedures, especiallythe limb bud micromass technique, was attempted by Flint and Orton. 257 Eighteen differentknown teratogens were injected into time-mated rats. Sixteen hours later, the embryos treated inutero were removed and their limb bud cells were plated as micromass cultures. Of the 18 teratogens© 2006 by Taylor & Francis Group, LLC

682 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONABCDEFFigure 16.12Photomicrographs of rat and rabbit limb bud micromass cultures treated with the glutathionedepleting agents buthionine sulfoximine (BSO) and diethyl maleate (DEM). Limb bud micromassestreated with 100 µM BSO [(C) rat and (D) rabbit] or 25 µM DEM [(E) rat and (F) rabbit]showed a decrease in chondrogenesis in comparison with control micromasses from the respectivespecies (A) rat and (B) rabbit. In addition, a complete inhibition of chondrogenic focusformation was observed in rabbit micromasses, indicating that rabbit limb bud tissues may bemore sensitive to glutathione depletion and modulation. (Hansen, J.M., Carney, E.W., and HarrisC., Free Rad. Biol. Med., 31, 1582, 2001. With permission.)tested, 17 tested positive (6% false negative) for the inhibition of chondrogenesis. In a follow-upstudy, 27 known teratogens and 19 nonteratogens were used to treat rat limb bud micromass culturesvia addition to the culture medium. 261 Of the 27 known teratogens, only two (7% false negative)did not have inhibitory effects on chondrogenesis (2,4-dichlorophenoxyacetic acid and thalidomide).Conversely, only two of the nonteratogens (11% false positive) showed a decrease in chondrogenicfocus formation (glutethimide and dimenhydrinate). Other in vitro prescreens used to test chemicalshave shown some power to estimate the teratogenicity of specific chemicals, but the assessmentsdescribed above clearly support the micromass system as a prescreen for teratogenic potential.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 683200 µmFigure 16.13Limb bud micromass stained with mercury orange (HgO) for small biothiols, mainly cysteineand reduced glutathione. Regions of high chondrogenic differentiation near the center of themicromass (white arrow) contain much less glutathione than does the highly proliferative region(black arrow) surrounding the differentiative zone. (Hansen, J.M., Carney, E.W., and Harris C.,Free Rad. Biol. Med., 31, 1582, 2001. With permission.)While this procedure has been touted and primarily used as a screen for teratogenic potential,the utility of these assays have been exploited in developmental biology as well. Plating at highdensityallows for the formation of an artificial “slice” of the limb and can be easily manipulatedacross different scientific disciplines. Much work utilizing the micromass technique has revolvedaround the function of the different bone morphogenetic proteins (BMP-2 and -6) and sonichedgehog (Shh) and their role in chondrogenesis and limb differentiation. 262–265 Similarly, midbraincultures have been utilized in comparable fashion but to a much more limited manner. 266 Morespecific to developmental toxicology, LB micromass cultures treated with GSH depleting agents,such as buthionine sulfoximine (BSO) and DEM, have demonstrated a substantial decrease inchondrogenic focus formation (Figure 16.12). Such findings implicate GSH and oxidative stressas possible modulators of early cartilage development. 260 Although not employed as a developmentaltoxicology tool to the same extent as the limb bud micromass approach, midbrain micromasscultures may potentially provide similar information in regard to neurogenesis and developmentalneurotoxicology.III. SUMMARYA number of useful culture methods have been described for application in studies designed toprobe mechanisms of developmental toxicity. The systems used for these purposes span the entirespectrum of development, from fertilization to parturition (and, in some cases, beyond). Theyencompass virtually all levels of biological organization, from the intact, viable embryo or fetusto cellular organelles and individual cells. The ability to select a single methodological approach thatcan be used in the examination of a potential mechanism and that covers the entire developmental© 2006 by Taylor & Francis Group, LLC

684 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONspectrum is not currently possible. Rather, experimental protocols will likely require numerousmethodologies. Inconsistent rates of growth and differentiation between different systems, even atsimilar developmental stages and levels of biological complexity, also make absolute conclusionsregarding mechanisms difficult. Multiple experimental models must still be created in order toprobe various aspects of the mechanisms of developmental toxicity. Cocultures and combinedapproaches utilizing different in vitro methods and in vivo–in vitro models will further extend theutility of these approaches as research tools. Continued refinement of in vitro procedures and theadditional acquired knowledge regarding basic developmental mechanisms will continue to stimulateresearch in areas of developmental toxicity by providing a more complete understanding ofendpoints and their relevance to environmental insults.REFERENCES1. Schwetz, B.A., Morrissey, R.E., Welsch, F., and Kavlock, R.A., In vitro teratology, Environ. HealthPerspect., 94, 265, 1991.2. Welsch, F., In vitro approaches to the elucidation of mechanisms of chemical teratogenesis, Teratology,46, 3, 1992.3. Flynn, T.J., Teratological research using in vitro systems. I. Mammalian whole embryo culture.Environ. Health Perspect., 72, 203, 1987.4. Kimmel, G.L. and Kochhar, D.M., In Vitro Methods in Developmental Toxicology, CRC Press, BocaRaton, 1990.5. Sadler, T.W., Horton, W.E., Jr., and Hunter, E.S., Mammalian embryos in culture: a new approach toinvestigating normal and abnormal developmental mechanisms, Prog. Clin. Biol. Res., 171, 227, 1985.6. Sadler, T.W., and Warner, C.W., Use of whole embryo culture for evaluating toxicity and teratogenicity,Pharmacol. Rev., 36, 145S, 1984.7. Steele, C.E., Whole embryo culture and teratogenesis, Human Reprod., 6, 144, 1991.8. New, D.A.T., Whole embryo culture, teratogenesis, and the estimation of teratologic risk, Teratology,42, 635, 1990.9. Wilson, J.G., Survey of in vitro systems: their potential use in teratogenicity screening, in Handbookof Teratology, Wilson, J.G. and Fraser, F.C., Eds., Plenum Press, New York, 1978, p. 135.10. New, D.A.T., Introduction, in Postimplantation Mammalian Embryos: A Practical Approach, Copp,A.J. and Cockroft, D.L., Eds., IRL Press, Oxford, 1990, p. 1.11. Austin, C.R., Embryo transfer and sensitivity to teratogenesis, Nature, 244, 333, 1973.12. Brachet, A., Dévelopment in vitro de blastomères et jeunes embryon de mammifères, C.R. Acad. Sci.Paris, 155, 1191, 1912.13. Spielmann, H. and Vogel, R., Unique role of studies on preimplantation embryos to understandmechanisms of embryotoxicity in early pregnancy, Crit. Rev. Toxicol., 20, 51, 198914. Lawitts, J.A. and Biggers, J.D., Culture of preimplantation embryos, Methods Enzymol., 225, 153,1993.15. Pratt, H.P.M., Isolation, culture and manipulation of pre-implantation mouse embryos, in MammalianDevelopment: A Practical Approach, Monk, M., Ed., IRL Press, Washington, D.C., 1987, p. 13.16. Blakenship, A.L., Suffia, M.C., Matsumura, F., Walsh, K.J., and Wiley, L.M., 2,3,7,8-tetrachlorodibenzo-p-dioxin(TCDD) accelerates differentiation of murine preimplantation embryos in vitro,Reprod. Toxicol., 7, 255, 1993.17. Oudiz, D.J., Walsh, K., and Wiley, L.M., Ethylene glycol monomethyl ether (EGME) exposure ofmale mice produces a decrease in cell proliferation of preimplantation embryos, Reprod. Toxicol., 7,101, 1993.18. Kholkute, S.D., Rodriguez, J., Dukelow, W.R., Effects of polychlorinated biphenyls (PCBs) on invitro fertilization in the mouse, Reprod. Toxicol., 8, 69, 1994.19. McNutt, T.L. and Harris, C., Glutathione and cysteine modulation in bovine oviduct epithelium cellscultured in vitro, Teratology, 49, 413, 1994.20. McNutt, T.L. and Harris, C., Influence of lindane on cysteine (CYS) and glutathione (GSH) levels inbovine oviduct epithelium cells cultured in vitro, The Toxicologist, 15, 251, 1995.21. Goto, Y., Noda, Y., Mori, T., and Nakano, M., Increased generation of reactive oxygen species inembryos cultured in vitro, Free Rad. Biol. Med., 15, 69, 1993.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 68522. Goto, Y., Noda, Y., Narimoto, K., Umaoka, Y., and Mori, T., Oxidative stress on mouse embryodevelopment in vitro, Free Rad. Biol. Med., 13, 47, 1992.23. Gardiner, C.S. and Reed D.J., Status of glutathione during oxidant-induced oxidative stress in thepreimplantation mouse embryo. Biol. Reprod., 51, 1307, 1994.24. Borland, R.M. and Tasca, R.J., Inhibition of L-methionine uptake and incorporation by chlorpromazinein preimplantation mouse embryos, J. Reprod. Fert., 42, 473, 1975.25. Heyner, S., Rao, L.V., Jarett, L., and Smith, R.M., Preimplantation mouse embryos internalize maternalinsulin via receptor-mediated endocytosis: pattern of uptake and functional correlations, Dev. Biol.,134, 48, 1989.26. Sellens, M.H., Stein, S., and Sherman, M.I., Protein and free amino acid content in preimplantationmouse embryos and in blastocysts under various culture conditions, J. Reprod. Fert., 61, 307, 1981.27. Poueymirou, W.T. and Schultz, R.M., Regulation of mouse preimplantation development: inhibitionof synthesis of proteins in the two-cell embryo that require transcription by inhibitors of cAMPdependentprotein kinase, Dev. Biol., 133, 588, 1989.28. Murach, K., Frei, M., Gerhauser, D., and Illmensee, K., Protein synthesis in embryonic tissues duringmouse postimplantation development, J. Cell. Biol., 44, 19, 1990.29. Spielman, H. and Eibs, H.-G., Recent progress in teratology: a survey of methods for the study ofdrug actions during the preimplantation period, Arzneim.-Forsch., 28, 1733, 1978.30. Dumoulin, J.C.M., Evers, J.L.H., Bakker, J.A., Bras, M., Pieters, M.H.E.C., and Geraedts, J.P.M., Temporaleffects of taurine on mouse preimplantation development in vitro, Human Reprod., 7, 403, 1992.31. Iyer, P., Martin, J.E., and Irvin, R.T., Role of biotransformation in the in vitro preimplantationembryotoxicity of naphthalene, Toxicology, 66, 257, 1991.32. Winkle, L.J.V., Amino acid transport in developing animal oocytes and early conceptuses, Biochim.Biophys. Acta, 947, 173, 1988.33. Mattson, B.A., Rosenblum, I.Y., Smith, R.M. and Heyner, S. Autoradiographic evidence for insulinand insulin-like growth factor binding to early mouse embryos, Diabetes, 37, 585, 1988.34. Rao, L.V., Wikarczuk, M.L., and Heyner, S., Functional roles of insulin and insulinlike growth factorsin preimplantation mouse embryo development, In Vitro Cell. Dev. Biol., 26, 1043, 1990.35. Niimura, S. and Ishida, K., Immunohistochemical demonstration of prostaglandin E-2 in preimplantationmouse embryos, J. Reprod. Fert., 80, 505, 1987.36. Biggers, J.D. and Borland, R.M. Physiological aspects of growth and development in the preimplantationmammalian embryo. Ann. Rev. Physiol. 38, 95, 1976.37. Kane, M.T. and Bavister, B.D., Vitamin requirements for development of eight-cell hamster embryosto hatching blastocysts in vitro, Biol. Reprod., 39, 1137, 1988.38. Butler, J.E., Lechene, C., and Biggers, J.D., Noninvasive measurement of glucose uptake by twopopulations of murine embryos, Biol. Reprod., 39, 779,1988.39. Seshagiri, P.B. and Bavister, B.D., Glucose inhibits development of hamster 8-cell embryos in vitro,Biol. Reprod., 40, 599, 198940. McKiernan, S.H., Bavister, B.D., and Tasca, R.J., Energy substrate requirements for in vitro developmentof hamster 1- and 2-cell embryos to the blastocyst stage, Human Reprod., 6, 64, 1991.41. Brinster, R.L., Studies on the development of mouse embryos in vitro. II. The effect of energy source.J. Exp. Zool., 158, 59, 1965.42. Barbehenn, E.K., Wales, R.G., and Lowry, O.H., Measure of metabolites in single preimplantationembryos; a new means to study metabolic control in early embryos, J. Embryol. Exp. Morph., 43, 29,1978.43. Gupta, J., Dey, S.K., and Dickman, Z., Evidence that “embryonic estrogen” is a factor which controlsthe development of the mouse preimplantation embryo, Steroids, 29, 363, 1977.44. Wu, J.T., Metabolism of progesterone by preimplantation mouse blastocysts in culture, Biol. Reprod.,36, 549, 1987.45. Garris, D.R., Role of the preimplantation embryo in the timing of LH-dependent progesterone secretionfrom the rat corpus luteum, Proc. Soc. Exp. Biol. Med., 170, 443, 1982.46. Lamb, V.K. and Leese, H.J., Uptake of a mixture of amino acids by mouse blastocysts. J. Reprod.Fert., 102, 169, 1994.47. Epstein, C.J. and Smith, S.A., Amino acid uptake and protein synthesis in preimplantation embryonicdevelopment, Biol. Reprod., 12, 82, 1973.© 2006 by Taylor & Francis Group, LLC

686 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION48. Bavister, B.D. and McKiernan, S.H., Regulation of hamster embryo development in vitro by aminoacids, in Preimplantation Embryo Development, Bavister, B.D., Ed., Springer-Verlag, New York, 1993,p. 57.49. Hsu, Y., Baskar, J., Stevens, L. and Rash, J., Development in vitro of mouse embryos from the twocellegg stage to the early somite stage, J. Embryol. Exp. Morphol., 31, 235, 1974.50. Hsu, Y.-C. In vitro development of individually cultured whole mouse embryos from blastocyst toearly somite stage, Dev. Biol. 68, 453, 1979.51. Kaufman, R.A., and Armant, D.R., In vitro exposure of preimplantation mouse embryos to cocaineand benzoylecgonine inhibits subsequent development, Teratology, 46, 85, 199252. Pollard, D.R., Baran, M.M. and Bachvarova, R., The effect of 5-bromodeoxyuridine on cell divisionand differentiation of preimplantation mouse embryos, J. Embryol. Exp. Morph., 35, 169, 1976.53. Rothstein, J.L., Johnson, D., DeLoia, J.A., Skowronski, J., Solter, D., and Knowles, B., Gene expressionduring preimplantation mouse development, Genes Dev., 6, 1190, 1992.54. Bolton, V.N., Oades, P.J. and Johnson, M.H. The relationship between cleavage, DNA replication andgene expression in the mouse 2-cell embryo, J. Embryol. Exp. Morphol., 79, 139, 1984.55. Conover, J.C., Temeles, G.L., Zimmerman, J.W., Burke, B., and Schultz, R.M., Stage-specific expressionof a family of proteins that are major products of zygotic gene activation in the mouse embryo,Dev. Biol., 144, 392, 199156. Christians, E., Campion, E., Thompson, E.M., and Renard, J.-P., Expression of the HSP 70.1 gene,a landmark of early zygotic activity in the mouse embryo, is restricted to the first burst of transcription,Development, 121, 113, 1995.57. Kimber, S.J., Waterhouse, R., and Lindenberg, S., In vitro models for implantation of the mammalianembryo, in Preimplantation Embryo Development, Bavister, B.A., Ed., Springer-Verlag, New York,1993, 244-263.58. Lindenberg, S., Hyttel, P., Sjgren, A., and Greve, T., A comparative study of attachment of human,bovine and mouse blastocysts to uterine epithelial monolayer, Human Reprod., 4, 446, 1989.59. New, D.A.T., Methods for the culture of postimplantation embryos of rodents, in Methods in MammalianEmbryology, Daniel, J.C., Jr., Ed., W.H. Freeman & Co., San Francisco, 1971, p. 305.60. Cockroft, D.L., Dissection and culture of postimplantation embryos, in Postimplantation MammalianEmbryos: A Practical Approach, Copp, A.J. and Cockroft, D.L., Eds., IRL Press, Oxford, 1990, p. 15.61. Sturm, K. and Tam, P.P.L., Isolation and culture of whole postimplantation embryos and germ layerderivatives, Methods Enzymol., 225, 164, 1993.62. Freeman, S.J., Coakley, M.E., and Brown, N.A., Post-implantation embryo culture for studies ofteratogenesis, in Biochemical Toxicology: A Practical Approach, Snell, K. and Mullock, B., Eds., IRLPress, Oxford, 1987, p. 4.63. Tarlatzis, B.C., Sanyal, M.K., Biggers, W.J. and Naftolin, F., Continuous culture of the postimplantationrat conceptus, Biol. Reprod., 31, 415, 1984.64. Kaufmann, M.H., Morphological stages of postimplantation embryonic development, in PostimplantationMammalian Embryos: A Practical Approach, Copp, A.J. and Cockroft, D.L., Eds., IRL Press,Oxford, 1990, p. 81.65. Naya M., Kito Y., Eto K., and Deguchi T., Development of rabbit whole embryo culture duringorganogenesis, Cong. Anom., 31, 153, 1991.66. Ninomiya H., Kishida K., Ohno Y., Tsurumi K., and Eto K., Effects of trypan blue on rat and rabbitembryos cultured in vitro, Toxicol. In Vitro, 7, 707, 1993.67. Pitt J. and Carney, E.W., Development of a morphologically-based scoring system for postimplantationNew Zealand White rabbit embryos, Teratology, 59, 88, 1999.68. Pitt J. and Carney, E.W. Evaluation of various toxicants in rabbit whole-embryo culture using a newmorphologically-based evaluation system, Teratology, 59, 102, 1999.69. Harris, C., Glutathione biosynthesis in the postimplantation rat conceptus in vitro, Toxicol. Appl.Pharmacol., 120, 247, 1993.70. Hansen, J., Choe, Y., Carney, E.W., and Harris, C., Differential antioxidant enzyme activities andglutathione content between rat and rabbit conceptuses, Free Rad. Biol. Med., 30, 1078, 2001.71. Hansen, J., Carney, E.W., and Harris, C., Differential alteration by thalidomide of the glutathionecontent of rat vs. rabbit conceptuses in vitro, Reprod. Toxicol. 13, 547, 1999.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 68772. Fantel, A.G., Greenaway, J.C., Juchau, M.R., and Shepard, T.H., Teratogenic bioactivation of cyclophosphamidein vitro, Life Sci., 25, 67, 1979.73. Faustman-Watts, E.M., Greenaway, J.C., Namkung, M.J., Fantel, A.G., and Juchau, M.R., Teratogenicityin vitro of 2-acetylaminofluorene: role of biotransformation in the rat, Teratology, 27, 19, 1983.74. Greenaway, J.C., Bark, D.H., and Juchau, M.R., Embryotoxic effects of salicylates: role of biotransformation,Toxicol. Appl. Pharmacol., 74, 141, 1984.75. Fantel, A.G., Greenaway, J.C. and Juchau, M.R., The embryotoxicity of adriamycin in rat embryosin vitro, Toxicol. Appl. Pharmacol., 80, 155, 1985.76. Harris, C., Stark, K.L., Luchtel, D.L. and Juchau, M.R., Abnormal neurulation induced by 7-hydroxy-2-acetylaminofluorene and acetaminophen: evidence for catechol metabolites as proximate dysmorphogens,Toxicol. Appl. Pharmacol., 101, 432, 1989.77. Steele, C.E., Marlow, R., Turton, J., and Hicks, R.M., In vitro teratogenicity of retinoids, Br. J. Exp.Path., 68, 215, 1987.78. Creech Kraft, J. and Juchau, M.R., Conceptal biotransformation of 4-oxo-all-trans-retinoic acid, 4-oxo-13-cis-retinoic acid and all-trans-retinoyl-beta-glucuronide in rat whole embryo culture, Biochem.Pharmacol., 43, 2289, 1992.79. Klug, S., Lewandowski, C., Zappel, F., merker, H.-J., Nau, H. and Neubert, D., Effects of valproicacid, some of its metabolites and analogues on prenatal development of rats in vitro and comparisonwith effects in vivo, Arch. Toxicol. 64, 545, 1990.80. Bruckner, A., Lee, Y.J., O’Shea, K.S., and Henneberry, R.C., Teratogenic effects of valproic acid anddiphenylhydantoin on mouse embryos in culture, Teratology, 27, 29, 1983.81. Fantel, A.G., Greenaway, J.C., Walker, E., and Juchau, M.R., The toxicity of niridazole in rat embryosin vitro, Teratology, 33, 105, 1986.82. Fantel, A.G., Greenaway, J.C. Shepard, T.H., Juchau, M.R., and Selleck, S.B., The teratogenicity ofcytochalasin D and its inhibition by drug metabolism, Teratology, 23, 223, 1981.83. Greenaway, J.C. and Fantel, A.G., Enhancement of rifampin teratogenicity in cultured rat embryos.Toxicol. Appl. Pharmacol., 69, 81, 1983.84. Faustman-Watts, E.M., Namkung, M.J., Greenaway, J.C. and Juchau, M.R., Analyses of metabolitesof 2-acetylaminofluorene generated in an embryo culture system: relationship of biotransformationto teratogenicity in vitro, Biochem. Pharmacol., 34, 2853, 1985.85. Yang, H.-Y.L., Namkung, M.J., and Juchau, M.R. Cytochrome P-450-dependent biotransformation ofa series of phenoxazone ethers in the rat conceptus during early organogenesis: evidence for multipleP-450 isozymes. Mol. Pharmacol., 34, 67, 1988.86. Yang, H.-Y.L., Zelus, B.D. and Juchau, M.R., Organogenesis-stage cytochrome P450 isoforms: utilizationof PCR for detection of CYP1A1 mRNA in rat conceptal tissues, Biochem. Biophys. Res.Commun., 178, 236, 1991.87. Chapman, D.E., Yang, H.Y., Watters, J.J., and Juchau, M.R., Induction in vitro and complete codingregion sequence of cytochrome P4501A1 cDNA from cultured whole rat conceptuses during earlyorganogenesis, Biochem. Pharmacol., 48, 1807, 1994.88. Martz, F., Failinger, C., III, and Blake, D.A., Phenytoin teratogenesis: correlation between embryopathiceffect and covalent binding on putative arene oxide metabolite in gestational tissue, J. Pharmacol.Exp. Ther., 203, 231, 1986.89. Wells, P.G., Zubovits, J.T., Wong, S.T., Molinari, L.M., and Ali, S., Modulation of phenytoin teratogenicityand embryonic covalent binding by the cyclooxygenase inhibitor acetylsalicylic acid, caffeicacid, and α-phenyl-N-t-butylnitrone: implications for bioactivation by prostaglandin H synthease,Toxicol. Appl. Pharmacol., 97, 192, 1989.90. Kubow, S. and Wells, P.G., In vitro bioactivation of phenytoin to a reactive free radical intermediateby prostaglandin H synthase, horseradish peroxidase, and thyroid peroxidase, Mol. Pharmacol., 35,504, 1989.91. Yu, W.K. and Wells, P.G., Evidence for lipoxygenase-catalyzed bioactivation of phenytoin to ateratogenic reactive intermediate: in vitro studies using linoleic acid-dependent soybean lipoxygenase,and in vivo studies using pregnant CD-1 mice, Toxicol. Appl. Pharmacol., 131, 1, 1995.92. Roy, S.K., Mitra, A.K., Hilbelink, D.R., Dwornik, J.J., and Kulkarni, A.P., Lipoxygenase activity inrat embryos and its potential for xenobiotic oxidation, Biol. Neonate, 63, 297, 1993.© 2006 by Taylor & Francis Group, LLC

688 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION93. Liu, L. and Wells, P.G., DNA oxidation as a potential molecular mechanism mediating drug-inducedbirth defects: Phenytoin and structurally related teratogens initiate the formation of 8-hydroxy-2′-deoxyguanosine in vitro and in vivo murine maternal hepatic and embryonic tissues, Free Rad. Biol.Med., 19, 639, 1995.94. Nau, H., Pharmacokinetic aspects of in vitro teratogenicity studies: comparison to in vivo, in In VitroMethods in Developmental Toxicology: Use in Defining Mechanisms and Risk Parameters, Kimmel,G.L. and Kochhar, D.M., Eds., CRC Press, Boca Raton, 1990, p. 29.95. Mebus, C.A., Clarke, D.O., Stedman, D.B. and Welsch, F., 2-Methoxyethanol metabolism in pregnantCD-1 mice and embryos, Toxicol. Appl. Pharmacol., 112, 87, 1992.96. Stedman D.B. and Welsch, F., Inhibition of DNA synthesis in mouse whole embryo culture by 2-methoxyacetic acid and attenuation of the effects by simple physiological compounds, Toxicol. Lett.45, 111, 1989.97. Scott, W.J., Duggan, C.A., Schreiner, C.M., Collins, M.D., and Nau, H., Intracellular pH of rodentembryos and its association with teratogenic response, in Approaches to Elucidate Mechanisms inTeratogenesis, Welsch, F., Ed., Hemisphere, New York, 1987, p. 99.98. Nau, H. and Scott, W.J., Jr., Teratogenicity of valproic acid and related substances in the mouse: drugaccumulation and pHi in the embryo during organogenesis and structure-activity considerations, Arch.Toxicol. Suppl., 11, 128, 1987.99. Harris, C., Tan, W., Thorsrud, B.A., and Kopelman, R., Ultramicrofiberoptic pH sensors for measurementof real time responses to chemical insult in intact, organogenesis stage rat conceptuses in vitro,Teratology, 49, 413, 1994.100. Terlouw, G.D.C. and Bechter, R., Comparison of the metabolic activity of yolk sac tissue in the wholeembryo and isolated yolk sac culture, Reprod. Toxicol., 6, 85, 1992.101. Lucier, G.W., Lui, E.M.K., and Lemartiniere, C.A., Metabolic activation/deactivation reactions duringperinatal development, Environ. Health Perspect., 29, 7, 1979.102. Harris, C., Fantel, A.G., and Juchau, M.R. Differential glutathione depletion in rat embryo vs. visceralyolk sac in vivo and in vitro by L-buthionine-S,R-sulfoximine, Biochem. Pharmacol., 35, 4437, 1986.103. Harris, C., Stark, K.L., and Juchau, M.R. Glutathione status and the incidence of neural tube defectselicited by direct acting teratogens in vitro, Teratology, 37, 577, 1988.104. Harris, C., Namkung, M.J., and Juchau, M.R., Regulation of intracellular glutathione in rat embryosand visceral yolk sacs and its effects on 2-nitrosofluorene-induced malformations in the whole embryoculture system, Toxicol. Appl. Pharmacol., 88, 141, 1987.105. Harris, C., Juchau, M.R., and Mirkes, P.E. The role of glutathione and Hsp70 in the acquisition ofthermotolerance in postimplantation rat embryos, Teratology, 43, 229, 1991.106. Slott, V.L. and Hales, B.F., Enhancement of the embryotoxicity of acrolein, but not phosporamidemustard, by glutathione depletion in rat embryos in vitro, Biochem. Pharmacol., 36, 2091, 1987.107. Wong, M., Helston, L.M.J., and Wells, P.G., Enhancement of murine phenytoin teratogenicity by theγ-glutamylcysteine synthetase inhibitor L-buthionine-(S,R)-sulfoximine and by the glutathione depletordiethyl maleate, Teratology, 40, 127, 1089.108. Naya, M., Mataki, Y., Takanira, H. Deguchi, T., and Yasuda, M., Effects of phorone and/or buthioninesulfoximine on teratogenicity of 5-fluorouracil in mice, Teratology, 41, 275, 1990.109. Griffith, O.W. and Meister, A., Potent and specific inhibition of glutathione synthesis by buthioninesulfoximine (S-n-butylhomocysteine sulfoximine), J. Biol. Chem., 254, 7558, 1979.110. Lee, E. and Harris, C., Differential chemical modulation of glutathione and cysteine status in ratembryo and visceral yolk sac in vitro: microinjection and addition to the culture medium, In VitroToxicol. (in press), 1995.111. Stark, K.L., Harris, C., and Juchau, M.R., Embryotoxicity elicited by inhibition of gamma-glutamyltransferase by acivicin and transferase antibodies in cultured rat embryos, Toxicol. Appl. Pharmacol.,89, 88, 1987.112. Stark, K.L., Lee, Q.P., Namkung, M.J., Harris, C., and Juchau, M.R., Dysmorphogenesis elicited bymicroinjected acetaminophen analogs and metabolites in rat embryos cultured in vitro, J. Pharmacol.Exp. Ther., 255, 74, 1990.113. Rowe, P.B. and Kalaizis, A., Serine metabolism in rat embryos undergoing organogenesis, J. Embryo.Exp. Morphol., 87, 137, 1985.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 689114. Beckman, D.A., Pugarelli, J.E., Jensen, M., Koszalka, T.R., Brent, R.L., and Lloyd, J.B., Sources ofamino acid for protein synthesis during early organogenesis in the rat. I. Relative contributions offree amino acids and of proteins, Placenta, 11, 109, 1990.115. Beckman, D.A., Pugarelli, J.E., Jensen, M., Koszalka, T.R., Brent, R.L., and Lloyd, J.B., Sources ofamino acid for protein synthesis during early organogenesis in the rat. II. Exchange with amino acidand protein pools in embryo and yolk sac, Placenta, 12, 37, 1991.116. Jenkinson, P.C., Anderson, D., and Gangolli, S.D., Malformations induced in cultured rat embryosby enzymatically generated active oxygen species, Teratogen. Carcinogen. Mutagen., 6, 547, 1986.117. Eriksson, U.J. and Borg, L.A.H., Protection by free radical scavenging enzymes against glucoseinduced embryonic malformations in vitro, Diabetologia, 34, 325, 1991.118. Tabacova, S. and Hunter, E.S., Modulation of arsenic teratogenicity by free radical scavengers,Teratology, 47, 390, 1993.119. Wells, P.G., Leeder, J.S., and Winn, L.M., Phenytoin-initiated protein oxidation in murine embryoculture: a potential molecular mechanism mediating phenytoin teratogenicity, The Toxicologist, 15,276, 1995.120. Hockenberry, D.M., Oltvai, Z.N., Yin, X.M., Millman, C.L., and Korsmeyer, S.J., Bcl-2 functions inan antioxidant pathway to prevent apoptosis, Cell, 75, 241,1993.121. Smedley, M.J. and Stanisstreet, M., Calcium and neurulation in mammalian embryos, J. Embryol.Exp. Morph., 89, 1, 1985.122. Smedley, M.J. and Stanisstreet, M., Calcium and neurulation in mammalian embryos. II. Effects ofcytoskeletal inhibitors and calcium antagonists on the neural folds of rat embryos, J. Embryol. Exp.Morph., 93, 167, 1986.123. Stein, G., Srivastava, M.K., Merker, H.-J., and Neubert, D., Effects of calcium channel blockers onthe development of early rat postimplantation embryos in culture, Arch. Toxicol., 64, 623, 1990.124. Lee, H. and Nagele, R.G., Toxic and teratologic effects of verapamil on early chick embryos: evidencefor the involvement of calcium in neural tube closure, Teratology, 33, 203, 1986.125. Beaudoin, A.R., NAD precursors as antiteratogens against aminothiadiazole, Teratology, 13, 95, 1976.126. Thorsrud, B.A. and Harris, C., Real time, micro-fiberoptic monitoring of endogenous fluorescence inthe rat conceptus during hypoxia, Teratology, 48, 343, 1993.127. Krowke, R. and Neubert, D., Embryonic intermediary metabolism under normal and pathologicalconditions, in Handbook of Teratology, Wilson, J.G. and Fraser, F.C., Eds., Plenum Press, New York,1978, p. 117.128. Kitchin, K.T., Sanyal, M.K., and Schmid, B.P., A coupled microsomal-activating/embryo culturesystem: toxicity of reduced β-nicotinamide adenine dinucleotide phosphate(NADPH), Biochem. Pharmac.,30, 985, 1981.129. Thorsrud, B.A. and Harris, C., Real time microfiberoptic redox fluorometry: modulation of the pyridinenucleotide status of the organogenesis-stage rat visceral yolk sac with cyanide and alloxan, Toxicol.Appl. Pharmacol., 135, 237, 1995.130. Freeman, S.J. and Lloyd, J.B., Evidence that suramin and sodium aurothiomalate are teratogenic inrat by disturbing yolk sac-mediated embryonic protein nutrition, Chem. Biol. Interact., 58, 149, 1986.131. Freeman, S.J. and Lloyd, J.B., Inhibition of proteolysis in rat yolk sac as a cause of teratogenesis.Effects of leupeptin in vitro and in vivo, J. Embryol. Exp. Morphol., 78, 183, 1983.132. Ambroso, J.L. and Harris, C., Chloroquine embryotoxicity in the postimplantation rat conceptus invitro, Teratology, 48, 213, 1993.133. Ambroso, J.A. and Harris, C., In vitro embryotoxicity of the cysteine proteinase inhibitors benzyloxycarbonyl-phenylalanine-alanine-diazomethane(Z-Phe-Ala-CHN2) and benzyloxycarbonyl-phenylalanine-phenylalanine-diazomethane(Z-Phe-Phe-CHN2), Teratology, 50, 214, 1994.134. Ambroso, J.L. and Harris, C., Chloroquine accumulation and alterations of proteolysis and pinocytosisin the rat conceptus in vitro, Biochem. Pharmacol., 47, 679, 1994.135. Hunter, E.S., III, Phillips, L.S., Goldstein, S., and Sadler, T.W., Altered visceral yolk sac functionproduced by a low-molecular-weight somatomedin inhibitor, Teratology, 43, 331, 1991.136. Lerman, S., Koszalka, T.R., Jensen, M., Andrew, C. L., Beckman, D.A. and Brent, R.L., In vitrostudies on the effect of yolk sac antisera on functions of the viscera yolk sac. I. Pinocytosis andtransport of small molecules, Teratology, 34, 335, 1986.© 2006 by Taylor & Francis Group, LLC

690 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION137. Marlow, R. and Freeman, S.J., Differential effect of zinc on teratogen-induced inhibition of pinocytosisby cultured rat yolk sac, Life Sci., 40, 1717, 1987.138. Pinter, E., Reece, E.A., Leranth, C.Z., Sanyal, M.K., Hobbins, J.C., Mahoney, M.J., and Naftolin, F.,Yolk sac failure in embryopathy due to hyperglycemia: ultrastructural analysis of yolk sac differentiationassociated with embryopathy in rat conceptuses under hyperglycemic conditions, Teratology,33, 73, 1986.139. Dunton, A., Al-Alousi, L.A., Pratten, M.K., and Beck, F., The giant yolk sac: a model for studyingearly placental transport, J. Anat., 145, 189, 1986.140. Freinkel, N., Lewis, N.J., Akawawa, S., Roth, S.I., and Gorman, L., The honeybee syndrome —implications of the teratogenicity of mannose in rat-embryo culture, N. Engl. J. Med., 310, 223, 1984.141. Tanimura, T. and Shepard, T.H., Glucose metabolism by rat embryos in vitro, Proc. Soc. Exp. Biol.Med., 135, 51, 1970.142. Ellington, S.K.L., In vitro analysis of glucose metabolism and embryonic growth in postimplantationrat embryos, Development, 100, 431, 1987.143. Clough, J.R. and Whittingham, D.G., Metabolism of [14C]glucose by postimplantation mouseembryos in vitro, J. Embryol. Exp. Morph., 74, 133, 1983.144. Moore, D.C.P., Stanistreet, M., Beck, F., and Clarke, C.A., The effects of mannose on rat embryosgrown in vitro, Life Sci., 41, 1885, 1987.145. Hiranruengchok, R. and Harris, C., Formation of protein-glutathione mixed disulfides in the developingrat conceptus following diamide treatment in vitro, Teratology, 52, 196, 1995.146. Hiranruengchok, R. and Harris, C., Glutathione oxidation and embryotoxicity elicited by diamide inthe developing rat conceptus in vitro, Toxicol. Appl. Pharmacol., 120, 62, 1993.147. Hiranruengchok, R. and Harris, C., Diamide-induced alterations of intracellular thiol status and theregulation of glucose metabolism in the developing rat conceptus in vitro, Teratology, 52, 205, 1995.148. Horton, W.E., Jr. and Sadler, T.W., Effects on maternal diabetes on early embryogenesis. Alterationsin morphogenesis produced by the ketone body, β-hydroxybutyrate, Diabetes, 32, 610, 1983.149. Sadler, T.W., Hunter, E.S., III, Balkan, W., and Horton, W.E., Jr., Effects of maternal diabetes onembryogenesis, Am. J. Perinatol., 5, 319, 1988.150. Sadler, T.W., Denno, K.M. and Hunter, E.S., III, Effects of altered glucose metabolism duringgastrulation and neurulation stages of embryogenesis, Ann. New York Acad. Sci., 678, 48, 1993.151. Eriksson, U.J. and Borg, L.A.H., Diabetes and embryonic malformations: role of substrate-inducedfree-oxygen radical production for dysmorphogenesis in cultured rat embryos, Diabetes, 42, 411, 1993.152. Denno, K.M. and Sadler T.W., Phenylalanine and its metabolites induce embryopathies in mouseembryos in culture, Teratology, 42, 565, 1990.153. Sulik, K.K. and Sadler, T.W., Postulated mechanisms underlying the development of neural tubedefects. Insights from in vitro and in vivo studies, Ann. New York Acad. Sci., 678, 8, 1993.154. Bolon, B., Welsch, F., and Morgan, K.T., Methanol-induced neural tube defects in mice: pathogenesisduring neurulation, Teratology, 49, 497, 1994.155. Cuthbertson, R.A. and Beck, F., Postimplantation whole embryo culture: a new method for studyingocular development, Invest. Ophthalmol. Vis. Sci., 31, 1653, 1990.156. Seegmiller, R.E., Harris C., Luchtel, D.L., and Juchau, M.R., Morphological differences elicited bytwo weak acids, retinoic and valproic, in rat embryos grown in vitro, Teratology, 43, 133, 1991.157. Alles, A.J. and Sulik, K.K., Retinoic-acid-induced limb-reduction defects: perturbation of zones ofprogrammed cell death as a pathogenic mechanism, Teratology, 40, 163, 1989.158. Eto, K. and Takakubo, F., The role of the yolk sac in craniofacial development of cultured rat embryos,J. Craniofac. Genet. Dev. Biol., 5, 357, 1985.159. Baltz, J.M., Biggers, J.D., and Lechene, C., Intracellular pH regulation by the preimplantation embryo,in Preimplantation Embryo Development, Bavister, B.A., Ed., Springer-Verlag, New York, 1993, p. 97.160. Gressens, P., Hill, J.M., Gozes, I., Fridkin, M., and Brenneman, D.E., Growth factor function ofvasoactive intestinal peptide in whole cultured mouse embryos, Nature, 362, 155, 1993.161. Yasuda, Y., Nishi, N., Takahashi, J.A., Konishi, H., Ohara, I., Fujita, H., Ohta, M., Itoh, N., Hatanaka,M., and Tanimura, T., Induction of avascular yolk sac due to reduction of basic fibroblast growthfactor by retinoic acid in mice, Dev. Biol., 150, 397, 1992.162. Chen, L.T. and Hsu, Y.C., Hemopoesis of the cultured whole mouse embryo, Exp. Hemat., 7, 232, 1979.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 691163. Mirkes, P.E., Cyclophosphamide teratogenesis: a review, Teratogen. Carcinogen. Mutagen., 5, 75,1985.164. Mirkes, P.E. and Cornel, L., A comparison of sodium arsenite- and hyperthermia-induced stressresponses and abnormal development in cultured postimplantation rat embryos, Teratology, 46, 251,1992.165. Mirkes, P.E., Grace, R.H., and Little, S.A., Developmental regulation of heat shock protein synthesisand HSP 70 RNA accumulation during postimplantation rat embryogenesis, Teratology, 44, 77, 1991.166. Thayer, J. and Mirkes, P.E., Programmed cell death and N-acetoxy-2-acetylaminofluorene-inducedapoptosis in the rat embryo, Teratology, 51, 418, 1995.167. Sadler, T.W., Liu, E.T., and Augustine, K.A., Antisense targeting of engrailed-1 causes abnormal axisformation in mouse embryos, Teratology, 51, 292, 1995.168. Sulston, J.E. and Horvitz, H.R., Post-embryonic cell lineages of the nematode, Caenorhabditis elegans,Dev. Biol., 56, 110, 1977.169. Sulston, J.E., Schierenberg, E., White, J.G., and Thomson, J.N. The embryonic cell lineage of thenematode Caenorhabditis elegans, Dev. Biol., 100, 64, 1983.170. Kokel, M., Borland, C.Z., DeLong, L., Horvitz, H.R., and Stern, M.J., clr-1 encodes a receptor tyrosinephosphatase that negatively regulates an FGF receptor signaling pathway in Caenorhabditis elegans,Genes Dev., 12, 1425, 1998.171. Gleason, J.E., Korswagen, H.C., and Eisenmann, D.M., Activation of Wnt signaling bypasses therequirement for RTK/Ras signaling during C. elegans vulval induction, Genes Dev., 16, 1281, 2002.172. Krishna, S., Maduzia, L.L., and Padgett, R.W., Specificity of TGF-beta signaling is conferred bydistinct type I receptors and their associated SMAD proteins in Caenorhabditis elegans, Development,126, 251, 1999.173. Morita, K., Flemming, A.J., Sugihara, Y., Mochii, M., Suzuki, Y., Yoshida, S., Wood, W.B., Kohara, Y.,Leroi, A.M., and Ueno, N., A Caenorhabditis elegans TGF-beta, DBL-1, controls the expression of LON-1, a PR-related protein, that regulates polyploidization and body length, EMBO J., 21, 1063, 2002.174. Yoshida, S., Morita, K., Mochii, M., and Ueno, N. Hypodermal expression of Caenorhabditis elegansTGF-beta type I receptor SMA-6 is essential for the growth and maintenance of body length, Dev.Biol., 240, 32, 2001.175. Berset, T., Hoier, E.F., Battu, G., Canevascini, S., and Hajnal, A., Notch inhibition of RAS signalingthrough MAP kinase phosphatase LIP-1 during C. elegans vulval development, Science, 291, 2255,2001.176. Hermann, G.J., Leung, B., and Priess, J.R., Left-right asymmetry in C. elegans intestine organogenesisinvolves a LIN-12/Notch signaling pathway, Development, 127, 3429, 2000.177. Herman M., C. elegans POP-1/TCF functions in a canonical Wnt pathway that controls cell migrationand in a noncanonical Wnt pathway that controls cell polarity, Development, 128, 581, 2001.178. Liu, Q.A. and Hengartner, M.O., The molecular mechanism of programmed cell death in C. elegans.Ann. N. Y. Acad. Sci., 887, 92, 1999.179. Tzur, Y.B., Hersh, B.M., Horvitz, H.R., and Gruenbaum, Y., Fate of the nuclear lamina duringCaenorhabditis elegans apoptosis. J. Struct. Biol. 137, 146, 2002.180. Schumacher, B., Hofmann, K., Boulton, S., and Gartner, A., The C. elegans homolog of the p53 tumorsuppressor is required for DNA damage-induced apoptosis, Curr. Biol., 11, 1722, 2001.181. Fort, D.J., Stover, E.L., Bantle, J.A., and Finch, R.A., Evaluation of the developmental toxicity ofthalidomide using frog embryo teratogenesis assay-Xenopus (FETAX): biotransformation and detoxification,Teratog. Carcinog. Mutagen., 20, 35, 2000.182. Prati, M., Gornati, R., Boracchi, P., Biganzoli, E., Fortaner, S., Pietra, R., Sabbioni, E. and Bernardini,G., A comparative study of the toxicity of mercury dichloride and methylmercury, assayed by thefrog embryo teratogenesis assay — Xenopus (FETAX), Altern. Lab Anim., 30, 23, 2002.183. Vismara, C., Bacchetta, R., Cacciatore, B., Vailati, G. and Fascio, U., Paraquat embryotoxicity in theXenopus laevis cleavage phase, Aquat. Toxicol., 55, 85, 2001.184. Vismara, C., Battista, V.V., Vailati, G., and Bacchetta, R., Paraquat induced embryotoxicity on Xenopuslaevis development. Aquat. Toxicol., 49, 171, 2000.185. Kotyzova, D. and Sundeman, F.W., Maternal exposure to Cd(II) causes malformations of Xenopuslaevis embryos, Ann. Clin. Lab. Sci., 28, 224, 1998.© 2006 by Taylor & Francis Group, LLC

692 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION186. Propst, T.L., Fort, D.J., Stover, E.L., Schrock, B., and Bantle, J.A., Evaluation of the developmentaltoxicity of benzo[a]pyrene and 2-acetylaminofluorene using Xenopus: modes of biotransformation.Drug Chem Toxicol., 20, 45, 1997.187. Fort, D.J., Stover, E.L., Propst, T.L., Faulkner, B.C., Vollmuth, T.A., and Murray, F.J., Evaluation ofthe developmental toxicity of caffeine and caffeine metabolites using the frog embryo teratogenesisassay — Xenopus (FETAX), Food Chem. Toxicol., 36, 591, 1998.188. Van den Belt, K., Van Puymbroeck, S., and Witters, H., Toxicity of cadmium-contaminated clay tothe zebrafish Danio rerio, Arch. Environ. Contam. Toxicol., 38, 191, 2000.189. Henry, T.R., Spitsbergen, J.M., Hornung, M.W., Abnet, C.C., and Peterson, R.E., Early life stagetoxicity of 2,3,7,8-tetrachlorodibenzo-p-dioxin in zebrafish (Danio rerio), Toxicol. Appl. Pharmacol.,142, 56, 1997.190. Bresch, H., Beck, H., Ehlermann, D., Schlaszus, H., and Urbanek, M., A long-term toxicity testcomprising reproduction and growth of zebrafish with 4-chloroaniline, Arch. Environ. Contam. Toxicol.,19, 419, 1990.191. Gong, Z., Ju, B., and Wan, H., Green fluorescent protein (GFP) transgenic fish and their applications,Genetica, 111, 213, 2001.192. Iwama, M., Honda, A., Mori, Y., Alterations of glycosaminoglycans synthesized by chick embryocartilage treated with 6-amino-nicotinamide, J. Pharmacobiodyn., 6, 613, 1983.193. Shiga, T., Gaur, V.P., Yamaguchi, K., and Oppenheim, R.W., The development of interneurons in the chickembryo spinal cord following in vivo treatment with retinoic acid, J. Comp. Neurol., 360, 463, 1995.194. Richman, J.M. and Delgado, J.L., Locally released retinoic acid leads to facial clefts in the chickembryo but does not alter the expression of receptors for fibroblast growth factor, J. Craniofac. Genet.Dev. Biol., 15, 190, 1995.195. Mattingly, C.J., McLachlan, J.A., and Toscano, W.A., Green fluorescent protein (GFP) as a markerof aryl hydrocarbon receptor (AhR) function in developing zebrafish (Danio rerio), Environ. HealthPerspect., 109, 845, 2001.196. Karkinen-Jaaskelainen, M. and Saxen, L. Advantages of organ culture techniques in teratology, inTests of Teratogenicity In Vitro, Marois, M., Ed. North Holland Publishers, Amsterdam, 1976, p. 275.197. Shepard, T.H. and Pious, D., Cell, tissue and organ culture as teratologic tools, in Handbook ofTeratology, Wilson, J.G. and Fraser, F.C., Eds. Plenum Press, New York, 1978, p. 71.198. Trowell. O.A., The culture of mature organs in a synthetic medium, Exp. Cell Res., 16, 118, 1959.199. Lessmollmann, U., Neubert, D., and Merker, H.-J., Mammalian limb buds differentiating in vitro asa test system for the evaluation of embryotoxic effects, in New Approaches to the Evaluation ofAbnormal Embryonic Development, Neubert, D., Merker, H.J., Bedürftig, A., and Kreft, R., Eds.,Georg Thieme Verlag, Stuttgart, 1975, p. 99.200. Friedman, L., Teratological research using in vitro systems. II. Rodent limb bud culture system,Environ. Health Perspect., 72, 211, 1987.201. Strangeways, T.S.P. and Fell, H.B., Experimental studies on the differentiation of embryonic tissuesgrowing in vivo and in vitro. I. The development of the undifferentiated limb-bud (a) when subcutaneouslygrafted into the post-embryonic chick and (b) when cultured in vitro, Proc. Roy. Soc. LondonB, 99, 340, 1926.202. Aydelotte, M.B. and Kochhar, D.M., Development of mouse limb buds in organ culture: chondrogenesisin the presence of a proline analog, 1-azetidine-2-carboxylic acid, Dev. Biol., 28, 191, 1972.203. Neubert, D. and Barrach, H.J., Techniques applicable to study morphogenic differentiation of limbbuds in organ culture, in Methods in Prenatal Toxicology, Neubert, D., Merker, H.-J., and Kwasigroch,T.E., Eds., George Thieme Publishers, Stuttgart, 241, 1977.204. Krowke, R.R., Bluth, U., and Neubert, D., In vitro studies on the embryotoxic potential of (bis[tri-nbutyltin])oxidein a limb bud organ culture system, Arch Toxicol., 58, 125, 1986.205. Kochar, D.M. and Aydelotte, M.B., Susceptible stages and abnormal morphogenesis in the developingmouse limb, analyzed in organ culture after transplacental exposure to vitamin A (retinoic acid), J.Embryol. Exp. Morophol., 31, 721, 1974.206. Nakamura, H., Analysis of limb anomalies induced in vitro by vitamin A (retinol), Teratology, 12,61, 1975.207. Manson, J.M. and Smith, C. C., Influence of cyclophosphamide and 4-ketocyclophosphamide onmouse limb development, Teratology, 15, 291, 1977.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 693208. Stahlmann, R., Bluth, U., and Neubert, D., Effects of the cyclophosphamide metabolite acrolein inmammalian limb bud cultures, Arch. Toxicol., 57, 163, 1985.209. Neubert, D., Lessmollmann, U., Hinz, N., Dillman, I., and Fuchs, G., Interference of 6-metcaotopurineriboside, 6-methylmercaptopurine riboside and azathioprine with the morphogenic differentiation ofmouse extremities in vivo and in an organ culture, Naunyn-Schmiedebergs Arch. Pharmakol., 289,1093, 1977.210. Flint, O.P., The effects of sodium salicylate, cytosine arabinoside, and eserine sulphate on rat limbbuds in culture, in Teratology of the Limbs, Merker, H.-J., Hau, H., and Neubert, D., Eds., Walter deGruyter, Berlin, 1980, p. 325.211. Tsuchiya, T. and Tanaka, A., Effects of thiabendazole on the mouse limb-bud organ and cell in culture,Toxicol. Lett., 30, 19, 1986.212. Barrach, H.J., Mark, K.V.D., and Gay, S., Location of type I and type II collagen in developing limbbuds of mammalian, in New Approaches to the Elucidation of Abnormal Embryonic Development,Neubert, D. and Merker, H.-J., Eds., Georg Thieme Publishers, Stuttgart, 1975, p. 145.213. Lilja, S. and Schimmelpfennig, K., Studies on glycosaminoglycans in cultured mouse limb buds, inNew Approaches to the Elucidation of Abnormal Embryonic Development, Neubert, D. and Merker,H.-J., Eds., Georg Thieme Publishers, Stuttgart, 1975, p. 241.214. Kwasigroch, T.E. and Kochhar, D.M., Locomotory behavior of limb bud cells: effects of excess vitaminA in vivo and in vitro, Exp. Cell. Res., 95,269, 1975.215. Hoffman, C. and Eichele, G., Retinoids in development, in The Retinoids: Biology, Chemistry, andMedicine, 2nd Ed., Sporn, M.B., Roberts, A.B., and Goodman, D. S., Eds., Raven Press, Ltd., NewYork, 1994, p. 387.216. Knudsen, T.B., In vitro approaches to the study of embryonic cell death in developmental toxicity, inIn Vitro Methods in Developmental Toxicology: Use in Defining Mechanisms and Risk Parameters,Kimmel, G.L. and Kochhar, D.M., Eds., CRC Press, Boca Raton, 1990, p. 129.217. Pratt, R.M., Dencker, L., and Diewert, V.M., 2,3,7,8-tetrachlorodibenzo-pdioxin-induced cleft palatein the mouse: evidence for alterations in palatal shelf fusion, Teratogen. Carcinogen. Mutagen., 4,427, 1984.218. Pratt, R.M., Receptor-dependent mechanisms of glucocorticoid and dioxin-induced cleft palate, Environ.Health Perspect., 61, 35, 1985.219. Mino, Y., Mizusawa, H., and Shiota, K., Effects of anticonvulsant drugs on fetal mouse palates culturedin vitro, Reprod. Toxicol., 8, 225, 1994.220. Abbott, B.D., Harris, M.W., and Birnbaum, L.S., Etiology of retinoic acid-induced cleft palate varieswith the embryonic stage, Teratology, 40, 533, 1989.221. Saxen, I. and Saxen, L., Organ culture in teratology: closure of the palatal shelves as a model system,in New Approaches to the Evaluation of Abnormal Embryonic Development, Neubert, D., Merker,H.J., Bedürftig, A., and Kreft, R., Eds., Georg Thieme Verlag, Stuttgart, 1975, p. 84.222. Saxen, I., The effect of tetracycline on osteogenesis in vitro, J. Exp. Zool., 162, 269, 1966.223. Saxen, I., Effects of hydrocortisone on the development in vitro of the secondary palate in two inbredstrains of mice, Arch. Oral Biol., 18, 1469, 1973.224. Abbott, B.D. and Buckalew, A.R., Embryonic palatal responses to teratogens in serum-free organculture, Teratology, 45, 369, 1992.225. Shiota, K., Kosazuma, T., Klug, S., and Neubert, D., Development of the fetal mouse palate insuspension organ culture, Acta Anat., 137, 59, 1990.226. Pratt, R.M., Receptor-dependent mechanisms of glucocorticoid and dioxin-induced cleft palate, Environ.Health Perspect., 61, 434, 1985.227. Pratt, R.M., Receptor-dependent Mechanisms of craniofacial malformations, in Approaches to ElucidateMechanisms in Teratogenesis, Welsch, F., Ed., Hemisphere, New York, 1987, p. 149.228. Abbot, B.D. and Pratt, R.M., Retinoids and epidermal growth factor alter embryonic mouse palatalepithelial and mesenchymal cell differentiation in organ culture, J. Craniofac. Gen. Dev. Biol., 7, 219,1987.229. Abbott, B.D. and Pratt, R.M., Human embryonic palatal epithelial differentiation is altered by retinoicacid and epidermal growth factor in organ culture, J. Craniofac. Gen. Dev. Biol., 7, 241, 1987.230. Pratt, R.M., Goulding. E.H., and Abbott, B.D., Retinoic acid inhibits migration of cranial neural crestcells in the cultured mouse embryo, J. Craniofac. Gen. Dev. Biol., 7, 205, 1987.© 2006 by Taylor & Francis Group, LLC

694 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION231. Beck, F., Lloyd, J.B., and Griffiths, A., Lysosomal enzyme inhibition by trypan blue: a theory ofteratogenesis, Science, 157, 1180, 1967.232. Beck, F. and Lowy, A. The effect of cathepsin inhibitor on rat embryos grown in vitro, J. Embryol.Exp. Morphol., 71, 1, 1982.233. Duncan, R. and Lloyd, J.B., Pinocytosis in the visceral yolk sac. Effects of temperature, metabolicinhibitors and some other modifiers, Biochim. Biophys. Acta, 544, 647, 1978.234. Stevenson, G.B. and Williams, K.E., Ethanol-induced inhibition of pinocytosis and proteolysis in ratyolk sac in vitro, Development, 99, 247, 1987.235. Lerman, S., Koszalka, T.R., Jensen, M., Andrew, C.L., Beckman, D.A., and Brent, R.L., In vitrostudies on the effect of yolk sac antisera on functions of the visceral yolk sac. I. Pinocytosis andtransport of small molecules, Teratology, 34, 335, 1986.236. Haffen, K., Organ culture and the determination of sex organs, in Tests of Teratogenicity In Vitro,Marois, M., Ed., North Holland Publishers, Amsterdam, 1976, p. 285.237. DeHaan, R.L., Cell coupling and electrophysiological differentiation of embryonic heart cells, in Testsof Teratogenicity In Vitro, Marois, M., Ed. North Holland Publishers, Amsterdam, 1976, p. 225.238. Kollar, E. J., The use of organ cultures of embryonic tooth germs for teratological studies, in Tests ofTeratogenicity In Vitro, Marois, M., Ed. North Holland Publishers, Amsterdam, 1976, p. 303.239. Lasnitzki, I., Organ culture, In Animal Cell Culture: A Practical Approach, Freshney, R.I., Ed., IRLPress, Oxford, 1992, p. 213.240. Fisher, K.R.S. and Fedoroff, S., The development of chick spinal cord in tissue culture, In vitro, 14,878, 1978.241. Bird, M.M. and Owen, A., The effect of calcium ionophore A23187 on neurites from embryonicmouse spinal cord explants in culture, J. Electron Microsc. (Tokyo), 49, 379, 2000.242. Serra, R., Pelton, R.W., and Moses, H.L., TGFb1 inhibits branching morphogenesis and N-mycexpression in lung bud organ cultures, Development, 120, 2153, 1994.243. Levesque, B.M., Vosatka, R.J., and Nielsen, H.C., Dihydrotestosterone stimulates branching morphogenesis,cell proliferation, and programmed cell death in mouse embryonic lung explants, Pediatr.Res., 47, 481, 2000.244. Konigsberg, I.R., Clonal analysis of development, in Tests of Teratogenicity In Vitro, Marois, M., Ed.North Holland Publishers, Amsterdam, 1976, p. 275.245. Kasten, F.H., Mammalian myocardial cells, in Tissue Culture Methods and Applications, Kruse, Jr.,P.F., and Patterson, M.K., Jr., Eds., Academic Press, New York, 1973, p. 72.246. Zimmerman, B., Effects of matrix influencing substances on chondrocyte in monolayer culture, inTeratology of the Limbs, Merker, H.-J., Hau, H., and Neubert, D., Eds., Walter de Gruyter, Berlin,1980, p. 207.247. Coon, H.G., Clonal stability and phenotypic expression of chick cartilage cells in vitro, Proc. Nat.Acad. Sci. U.S.A., 55, 66, 1966.248. Cahn, R.D. and Cahn, M.B., Heritability of cellular differentiation: clonal growth and expression ofdifferentiation in retinal pigment cells in vitro, Proc. Nat. Acad. Sci. U.S.A., 55, 106, 1966.249. Eguchi, G. and Okada, T.S., Differentiation of lens tissue from the progeny of chick retinal pigmentcells cultured in vitro: a demonstration of a switch of cell types in clonal cell culture, Proc. Nat. Acad.Sci. U.S.A., 70, 1495, 1973.250. Loring, J., Glimelius, B., Erickson, C., and Weston, J. A., Analysis of developmentally homogeneousneural crest cell populations in vitro, Dev. Biol., 82, 86, 1981.251. Glimelius, B. and Weston, J. A., Analysis of developmentally homogeneous neural crest cell populationsin vitro. II. A tumor-promoter (TPA) delays differentiation and promotes cell proliferation, Dev.Biol., 82, 95, 1981.252. Kang, K.S., Park, J.E., Ryu, D.Y., and Lee, Y.S., Effects and neuro-toxic mechanisms of 2,2′,4,4′,5,5′-hexachlorobiphenyl and endosulfan in neuronal stem cells, J. Vet. Med. Sci., 63, 1183, 2001253. Davis, W.L., Crawford, L.A., Cooper, O.J., Farmer, G.R., Thomas, D.L., and Freeman, B.L., Ethanolinduces the generation of reactive free radicals by neural crest cells in vitro, J. Craniofac. Genet. Dev.Biol., 10, 277, 1990.254. Davis, W.L., Crawford, L.A., Cooper, O.J., Farmer, G.R., Thomas, D.L., and Freeman, B.L., Generationof radical oxygen species by neural crest cells treated in vitro with isotretinoin and 4-oxoisotretinoin,J. Craniofac. Genet. Dev. Biol., 10, 295, 1990.© 2006 by Taylor & Francis Group, LLC

IN VITRO METHODS FOR THE STUDY OF MECHANISMS OF DEVELOPMENTAL TOXICOLOGY 695255. Davis, W.L., Jacoby, B.H., Farmer, G.R., and Cooper, O.J., Changes in cytosolic calcium, blebformation, and cell death in neural crest cells treated with isotretinoin and 4-oxo-isotretinoin, J.Craniofac. Genet. Dev. Biol., 10, 295, 1990.256. Bonnefoi, M.S., Mitochondrial glutathione and methyl iodide-induced neurotoxicity in primary neuralcell cultures, Neuro. Toxicol., 13, 401, 1992.257. Flint, O.P. and Orton, T.C., An in vitro assay for teratogens with cultures of rat embryo midbrain andlimb bud cells, Toxicol. Appl. Pharmacol., 76, 383, 1984.258. Kistler, A. and Howard, W.B., Testing of retinoids for teratogenicity in vitro: use of micromass limbbud culture, Methods Enzymol., 190, 427, 1990.259. Kistler, A., Limb bud cell cultures for estimating the teratogenic potential of compounds. Validationof the test system with retinoids, Arch. Toxicol., 60, 403, 1987.260. Hansen, J.M., Carney, E.W., Harris, C., Altered differentiation in rat and rabbit limb bud micromasscultures by glutathione modulating agents, Free Radic. Biol. Med., 31, 1582, 2001261. Flint, O.P., Orton, T.C., and Ferguson, R.A., Differentiation of rat embryo cells in culture: responsefollowing acute maternal exposure to teratogens and non-teratogens, J. Appl. Toxicol., 4, 109, 1984.262. Roark, E.F. and Greer, K., Transforming growth factor-beta and bone morphogenetic protein-2 act bydistinct mechanisms to promote chick limb cartilage differentiation in vitro, Dev. Dyn., 200, 103, 1994.263. Lee, Y.S. and Chuong, C.M., Activation of protein kinase A is a pivotal step involved in both BMP-2- and cyclic AMP-induced chondrogenesis, J. Cell Physiol., 170, 153, 1997.264. Boskey, A.L., Paschalis, E.P., Binderman, I., and Doty, S.B., BMP-6 accelerates both chondrogenesisand mineral maturation in differentiating chick limb-bud mesenchymal cell cultures, J. Cell Biochem.,84, 509, 2002.265. Kruger, M., Mennerich, D., Fees, S., Schafer, R., Mundlos, S. and Braun T., Sonic hedgehog is asurvival factor for hypaxial muscles during mouse development, Development, 128, 743, 2001.266. Whittaker, S.G., Wroble, J.T., Silbernagel, S.M. and Faustman, E.M. Characterization of cytoskeletaland neuronal markers in micromass cultures of rat embryonic midbrain cells, Cell. Biol. Toxicol., 9,359,1993.© 2006 by Taylor & Francis Group, LLC

CHAPTER 17Statistical Analysis for Developmental andReproductive ToxicologistsJames J. ChenCONTENTSI. Introduction ........................................................................................................................697II. Analysis of Continuous Response Data ............................................................................699A. Litter-Based ANOVA Method ...................................................................................699B. Linear Mixed-Effects Models....................................................................................700C. Example-Fetal Body Weight Data.............................................................................700III.Analysis of Binary Response Data....................................................................................701A. Independent Binary Response Data ..........................................................................701B. Correlated Binary Response Data .............................................................................7031. Litter-Based ANOVA Method .............................................................................7032. Likelihood and Quasi-Likelihood Approaches ...................................................7033. Example — Hydroxyurea Data...........................................................................704IV. Analysis of Count Data .....................................................................................................705A. Likelihood and Quasi-likelihood Methods................................................................706B. Example .....................................................................................................................706V. Quantitative Risk Assessment and Benchmark Dose........................................................707A. Dose-Response Model and Benchmark Dose...........................................................707B. Example .....................................................................................................................709VI. Discussion ..........................................................................................................................709References ......................................................................................................................................710I. INTRODUCTIONReproductive and developmental toxicity experiments are conducted in laboratory animals for theevaluation of potential adverse effects of chemical compounds on fertility, reproduction, and fetaldevelopment. Depending on the study design, a test compound is administered to either parentprior to conception, to the female during gestation, or postnatally. Various reproductive and developmentalendpoints are measured, recorded, and analyzed. These endpoints can be divided intotwo categories: (1) parental and (2) embryonic or fetal endpoints. Parental endpoints used to assessreproductive effects include: body weight and weight gain, mating index, fertility index, changes697© 2006 by Taylor & Francis Group, LLC

698 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONin gestation length, sperm count, numbers of corpora lutea, implantation sites, dead or resorbedimplants, viable fetuses, and target organ weights, as well as food and water consumption. Embryonicor fetal endpoints include: individual malformations (external or gross, visceral, and skeletal)and variations in viable fetuses, number of normal fetuses, fetal weight, fetal length, sex ratio,preweaning and postweaning survival and body weight, physical development, and functional testoutcomes (see Chapters 7, 8, 10, and 19).Mice, rats, and rabbits are the most commonly used species for reproductive and developmentaltoxicology studies. An experiment typically consists of an untreated control and three dosage groups.The highest dose is chosen to produce minimal maternal toxicity, ranging from marginal bodyweight reduction to not more than 10% mortality. The low dose should be a no-observed adverseeffectlevel for both maternal and developmental toxicity. The midlevel dose usually is halfwaybetween the low and high doses. A fourth dosage group may be added to avoid excessive dosageintervals and to assess dose-response relationships. Ideally the dose should be the amount of activesubstance at the affected tissue site. The estimate of the internal tissue-site dose requires data fromtoxicokinetic and/or toxicodynamic experiments. In the absence of this information, the amount ofthe internal dose is assumed proportional to the administered or exposure dose. Typically, dosageis measured in milligrams per kilogram body weight per day, and animals are randomly assignedto dosage groups. Each animal is monitored throughout the experiment to detect toxic and pharmacologiceffects of the test compounds, and all adult animals are necropsied at terminal sacrifice.Sample size determination is an important issue in detecting a dose effect. The probability ofdetecting a statistically significant effect increases with the increased magnitude of effect andincreased sample size. It is customary to specify a range of the magnitudes of the effect of interest;an optimal sample size then can be calculated to provide a reasonable power of detecting the effect.Sample size calculations and references for various tests are provided in Desu and Raghavarao. 1The U.S. regulatory guidelines generally recommend about 20 pregnant rodents and 15 nonrodentanimals per dosage group. The ICH guideline recommends use of 16 to 20 pregnant animals.Typical litter sizes (number of viable offspring) for the control animals range from 8 for rabbits to12 and 14 for mice and rats, respectively. However, it should be emphasized that sample sizedetermination is preferable to the arbitrary requirements for testing 10 or 20 animals as indicatedin various guidelines (cf., Chapter 19).In the analysis of animal developmental endpoints, the experimental unit is the entire litterrather than an individual fetus. This is because the test compound is administered to the adultanimal, so the effect of the test compound occurs in the female that receives the compound or ismated to a male that receives the compound. The treatment thus affects the embryo or fetusindirectly, via the dam. Thus, the development of individual offspring in a dam is not independent,and the fetal responses of individuals from the same litter are expected to be more alike thanresponses of individuals from different litters. This phenomenon is referred to as the “litter effect.”The proper experimental unit in the analysis should be the litter, with the fetal responses representingmultiple observations from a single experimental unit. Failure to account for the intralitter correlationsby using each fetus as the experimental unit can inflate the Type I error and reduce thevalidity of the test. The classical approach to the analysis of reproductive data is based on the littermean. An alternative approach is to model the fetal endpoints in a litter as correlated outcomes(clustered data). These two approaches are described in this chapter.Statistical analyses of various toxicity endpoints have been concerned with two aspects: qualitativetesting for adverse effects and quantitative estimation for risk assessment. The qualitativetesting is to determine if there is a statistically significant difference among the groups. Threestatistical tests are generally conducted:1. Homogeneity tests for detecting significant differences among groups2. Pairwise comparisons between two specific groups (usually control versus a dosage group)3. Trend tests to identify dose-related increases or decreases© 2006 by Taylor & Francis Group, LLC

STATISTICAL ANALYSIS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 699Trend tests are more sensitive (powerful) for detecting dose effects than are pairwise comparisonsbecause trend tests use information from all dosage groups. The quantitative estimation is todetermine a safe level of human exposure from an assessment of the dose-response relationship.However, the quantitative risk estimation is typically used for environmental or workplace exposuresand does not apply to pharmaceutical drugs.The reproductive and developmental toxicity endpoints described above can generally be dividedinto three categories: (1) binary (e.g., presence or absence of a malformation) or proportion (ofmalformation among the viable progeny), (2) count (e.g., number of implants), and (3) continuous(e.g., fetal body weight). This chapter describes statistical methods for the analysis of the threedata types.II. ANALYSIS OF CONTINUOUS RESPONSE DATAConsider a typical experiment of g groups, a control (d 1 = 0) and g – 1 dose groups ( d 2 , ⋯,d g ).Assume that the ith group contains n i female animals. Let y ijk be the response from the kth fetusout of n ij examined or tested for a particular developmental outcome, 1 ≤ i ≤ g, 1 ≤ j ≤ n i , and1 ≤ k ≤ n ij . For example, y ijk may be a fetal body weight or behavioral measurement conducted onoffspring following birth, and n ij is the number of viable fetuses from the jth female in the ithgroup. The continuous endpoints can also be measured in an adult animal (e.g., maternal body ororgan weight) denoted as y ij . The data can be summarized as the mean with a standard error of themean (SE) for each dosage group, using the litter as the experimental unit.A. Litter-Based ANOVA MethodAnalysis of variance (ANOVA) is the most commonly used procedure for the analysis of continuousresponse data. The ANOVA method assumes that data are independently and normally distributedwith homogeneous variance. Prior to ANOVA, Bartlett’s test 2 is conducted to test for homogeneityof variance. Transformations, such as the logarithmic, are often applied to satisfy the normalityassumption and stabilize the variance across groups. Transformations of the data are acceptablebecause the scale of measurement of variables is often arbitrary. Graphical normal plots and Boxplots 3 can be used to determine if the transformed data satisfy the normality and variance homogeneityassumptions. The SAS PROC UNIVARIATE 4 procedure provides various statistical testsand graphical plots for checking the underlying assumptions for ANOVA. The litter-based analysisis based on the per-litter response. For maternal endpoints, a simple one-way ANOVA on theobserved response y ij is used for testing homogeneity among groups. Developmental endpoints areanalyzed similarly but in terms of the average within each litter,y = y / nij∑A wide variety of post hoc tests is available for pairwise comparisons or trend analysis once asignificant result is obtained in an ANOVA. The Dunnett 5 and Williams 6 tests are the two mostlyused multiple test procedures for comparison of each dosage group with the control. The objectivesof interest often are to identify the largest dose that is not statistically significantly different fromthe control (NOEL) or the lowest dose that is significantly different from the control. The Williamstest assumes an increasing (or decreasing) dose-response trend. When this assumption is valid, theWilliams’ test is the more powerful to detect differences between control and dosage groups. Linearregression is the simplest method to assess dose-response trends. But contrast tests, with a properchoice of contrast coefficients in conjunction with ANOVA, are more commonly used for dose-responsekijkij© 2006 by Taylor & Francis Group, LLC

700 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 17.1Mean fetal weights (per litter) and dosage group means with their standard errors (SEs)Dose Mean fetal weight/litter Mean (SE)Control 6.845 6.296 7.355 6.289 6.885 6.966 6.235 5.894 5.365 6.535 6.467 (.183)Low 5.725 7.310 5.796 5.820 6.076 6.008 5.451 6.145 6.164 6.043 6.054 (.156)High 5.243 5.407 7.237 5.873 6.564 6.298 6.001 6.089 (.259)trend testing. A contrast test performs custom hypothesis tests by specifying coefficients for testinga univariate hypothesis or multivariate hypothesis. The trend analysis is simplified by examiningorthogonal contrasts among the treatment factor levels that measure the linear, quadratic, and higherlevel of polynomial effects. Table values of the orthogonal polynomial coefficients can be foundin experimental design textbooks. 2 Other trend tests are Tukey’s test, 7 which has been proposed toidentify the highest dose at which there is no statistically significant difference from the controltrend, and Bartholomew’s test 8 of ordered alternatives.Nonparametric methods are used when the assumption of normality fails. The nonparametricanalysis is initiated by ranking all observations of the combined groups. The nonparametric methodis then applied to ranked data. Further details may be found in general texts addressing nonparametricmethods. 9B. Linear Mixed-Effects ModelsA general approach to modeling fetal data can be carried out in terms of a mixed-effects model.Dempster et al. 10 proposed a normal mixed-effects model with two levels of variance in which littereffect is modeled by a nested random factor and dose by a fixed factor. The response y ijk in a litteris modeled with two sources of variation: the between litter γ ij and within litter variation e ijk ,The random components γ ij and e ijk are independently normally distributed with E( γ ij ) = 0,22Var( γij)= σ f , and E( eijk) = 0, Var( eijk) = σ . Thus, the mean and variance of y ijk are E( y ijk ) =µ i and2 22 2 2Var( y ijk ) = σ f + σ . The intralitter correlation between y ijk and y ijk′ , for k ≠ k′ is φ = σ f /( σ + σ f).The mean parameter µ ij is often modeled as a linear function of dose µ ij = β0 + β1dfor trend testsH o :β 1 (= 0). The maximum likelihood estimates of the parameters of linear mixed effects modelscan be obtained using the PROC MIXED procedure of SAS. 11 Alternatively, the generalizedestimating equations (GEE) approach 12 can be used to estimate the fixed effects parameters. Thesoftware for the GEE method is available publicly 13 or is found in the SAS PROC GENMODprocedure.C. Example-Fetal Body Weight Datay= µ + γ + eijk ij ij ijkAn analysis of fetal body weight data from a reproductive study by Dempster et al. 10 is used forillustration. The experiment consisted of three dosage groups: control, low, and high. The numbersof animals in the three groups were 10, 10, and 7, respectively. The litter means y ij s are given inTable 17.1. (The individual fetal weight data are listed in Table 17.2 of Dempster et al. 10 ) The threehypothesis tests, homogeneity, pairwise comparisons of dosage groups with control, and trend,were conducted using the litter-based analysis and linear mixed-effects model approaches. Table17.2 contains the summary of the analysis. The SAS PROC GLM and PROC MIXED procedureswere used for the litter-based analysis and the linear mixed-effects model, respectively.Using the method of litter-based ANOVA on the litter means, the homogeneity test of nodifference among the three means yields an F value of 1.50 (p = 0.244) with 2 and 24 degrees of© 2006 by Taylor & Francis Group, LLC

STATISTICAL ANALYSIS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 701Table 17.2p-values for testing homogeneity, control versus dosage groups, and trend,using the litter-based analysis and mixed-effects model approachesControl vs. DoseLSD (t-test)DunnettApproach Homogeneity µ 1 = µ 2 µ 1 = µ 3 µ 1 = µ 2 µ 1 = µ 3 TrendLitter-based analysis .244 .138 .198 .215 .332 .198Mixed-effect model .185 .093 .161 .163 .276 .161freedom. The p value is greater than the 5% significance level. The comparison could stop here,but for the purpose of illustration, the dosage group comparisons and trend test are performed. Forthe mixed-effects model, the maximum likelihood estimates of the variance parameters areˆσ 2 = 0.196 and ˆσ f 2 = 0.277. The estimated intralitter correlation is ˆφ = 0.277/(0.196+0.277) = 0.586,indicating the presence of litter effect. The two methods give very similar results; however, themixed-effect modeling approach appears more powerful because it uses individual fetal data thataccounts for litter effects in the analysis.III. ANALYSIS OF BINARY RESPONSE DATABinary endpoints can also be measured either at the parent level, such as success or failure ofpregnancy, or at the individual fetal level, such as presence or absence of a particular malformationtype. The binary responses from individual dams are independent, while the binary responses fromthe fetuses in a litter are correlated. Statistical methods for the analyses of the parental and fetalresponses are different.A. Independent Binary Response DataGiven a typical reproductive experiment, if n i represents the total number of animals from the ithgroup and y i is the number of animals found to present a particular response of interest, the datacan be summarized asDose d 1 d 2 … d g TotalNo. of responses y 1 y 2 … y g yTotal n 1 n 2 … n g nProportions p 1 p 2 … p g py and n are the row marginal totals, p = y/n, and p i = y i /n i is the observed incidence rate in the ithgroup.An asymptotic chi-square test for homogeneity is commonly used to compare incidence rates(proportions) among several groups:2χ g −1=g∑i=12( yi− nip)np( 1 − p)For pairwise comparisons, the chi-square test is often used with the continuity correction. It canbe simplified as( 05= | y − n p | − . )np( 1 − p)2 i 1χ 11i2© 2006 by Taylor & Francis Group, LLC

702 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe Cochran-Armitage (CA) test 14,15 is used to test for dose-related trend,Z =∑i∑yd − p ndiip ( − p )[ 2nd − ( nd ) 21/ n ]iii∑i∑iiiiiAlternatively, logistic regression provides a simple approach for analyzing binary data. The probabilityof response is modeled asexp( β0 + β1di)Pd ( i)=1 + exp( β + β d )0 1Logistic regression can be used for group comparisons and trend test. The model can include anyadditional explanatory variables (e.g., fetal weight, litter size).When the total number of occurrences is small, the asymptotic test may not be reliable, and anexact permutation test is recommended. The Fisher’s exact test is the best known permutation testfor comparing two groups. We describe a generalization of the Fisher’s exact test to the 2 × g tablefor trend test.Conditional on y . and n . , under the null hypothesis of no difference among groups, the distributionof ( y 1 , y 2 ,…,y g ) is the multivariate hypergeometric distributioniPy ( , y ,…y, ) =1 2g⎛ ⎞⎛⎜n1⎟⎜⎜ ⎟⎜⎜⎝ y ⎟ ⎜⎜1 ⎠⎝ y2narraycngThe trend score associated with ( y 1 , y 2 ,…, r g ) is defined by S = ∑ i=1diyi.The probability distributionfor the trend score statistic S is∑Ωggwhere Ω consists of all possible permutations of y i such that ∑ i=1 yi= y and ∑ i=1dyi i = s. Let s*denote the trend score associated with the observed outcome. The exact one-tailed p-value for apositive dose-related trend isAn example of a hypothetical animal experiment with four dosage groups, d1 = 0, d2 = 1, d3= 2,and d 4 = 4, with 50 animals in each group is used for illustration. Assume that the observed outcomefor the number of responses in the four groups is (0, 0, 3, 3), which corresponds to a trend scoreof S = 18. Note that two other permutations (0, 2, 0, 4) and (1, 0, 1, 4) also give the score 18. Theprobability for all possible permutations such that S >= 18 are given below:n⎛ ⎞⎜ . ⎟⎜ ⎟⎜⎝ y ⎟. ⎠⎞2 ⎟⎟⎟⎠⎟PS ( = s) = Py ( , y ,…y , g )∑1 2p − value = PS ( = s)s≥s∗⎛⎜⎜⎝⎜… n y⎞g ⎟⎟g ⎠⎟© 2006 by Taylor & Francis Group, LLC

STATISTICAL ANALYSIS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 703Response Pattern S Prob. Cumul. Prob.0 0 0 6 =24 0.00019 0.000190 0 1 5 =22 0.00129 0.001480 1 0 5 =21 0.00129 0.002771 0 0 5 =20 0.00129 0.004060 0 2 4 =20 0.00341 0.007470 1 1 4 =19 0.00699 0.014460 2 0 4 =18 0.00342 0.017880 0 3 3 =18 0.00466 0.022541 0 1 4 =18 0.00699 0.02953The one-sided exact p-value is p-value = 0. 00019 + 0. 00129 + ... + 0. 00699 = 0.02953. To comparewith the asymptotic test, the p-value is 0.01778 from the CA test and is 0.02735 from thelogistic regression Wald test. The exact permutation test generally gives a larger p-value (less likelyto be significant) than an asymptotic test. In a two group comparison, this method becomes theFisher’s exact test. That is, the p-value calculated from the Fisher’s exact test is usually larger thanthe p-value obtained from an asymptotic chi-square (Wald) test. But, the p-value from a chi-squaretest with the continuity correction and the Fisher’s exact p-value are comparable.The chi-square test and Fisher’s exact test are available in the SAS 11 PROC FREQ procedure. TheSAS PROC LOGISTIC procedure provides the p-values of the CA test (score test) and Wald test, andthe PROC MULTEST procedure provides the exact p-value for the permutation test. The exact p-valuecan also be obtained using StatXact software distributed by Cytel Software Corporation. 16B. Correlated Binary Response DataAs discussed, the fetal responses within a litter are not independent. The proper experimental unitin the analysis should thus be the litter. As was the case for the analysis of continuous responsedata, two methods can be described.1. Litter-Based ANOVA MethodThe conventional approach for the analysis of fetal binary endpoints is to consider the proportionper litter, such as the proportion of live fetuses with a certain type of malformation or the proportionof deaths and/or resorptions among the number of implants. If y ijk is the binary response from afetus out of n ij examined or tested for a particular developmental outcome, the total number ofresponses in a litter is yij = ∑k yijk. The litter-based analysis is based on the litter proportionpij = yij / nij. Typically, the proportions are transformed by an arcsine transformation, arcsin p ij .The transformed data are then analyzed as continuous data by use of the parametric ANOVAmethods described in Section II. Nonparametric methods, such as the Kruskal-Wallis 9 ANOVA testfor homogeneity, the Mann-Whitney 9 test for pairwise comparisons, and the Jonckheere 17 trendtest, can be used to analyze litter proportions by ranks. The nonparametric methods are availablein the SAS PROC NPR1WAY procedure. 11The litter-based approach does not use the data effectively because it does not account for littersize. For example, 1 out of 2 is treated as the same as 5 out of 10. Thus, the beta-binomial likelihoodand quasi-likelihood approaches have been developed recently to analyze correlated binary data.2. Likelihood and Quasi-Likelihood ApproachesWilliams 18 proposed the beta-binomial model for the analysis of developmental data. This modelassumes that responses within the same litter occur according to a binomial distribution, and theprobability of responses is assumed to vary among litters according to a beta distribution. Thedistribution for the number of responses in a litter y ij is© 2006 by Taylor & Francis Group, LLC

704 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 17.3The observed numbers of adverse developmental effects (death/resorption/malformation) (y)per litter from exposure to hydroxyurea and the numbers of implantations per litter (n)Control (y) 0 0 0 0 2 0 1 1 0 0 1 0 0 0 0 0 1(n) 10 10 12 13 11 9 9 12 11 8 11 13 12 6 8 11 9150 ppm (y) 1 0 0 5 1 2 2 4 1 0 6 1 0(n) 14 11 11 8 12 8 11 10 12 14 13 7 9200 ppm (y) 2 0 13 9 12 10 4 3(n) 14 11 16 11 13 13 13 11250 ppm (y) 8 11 7 10 9 4 4 9 8 5 11 9 10 10 6 9 9(n) 9 11 11 11 11 11 9 10 8 5 12 12 12 11 11 13 12(y) 8 11 6 11 3(c) 9 12 10 12 9⎛⎜nPy ( ij)= ⎜⎜ ywhere B ( ai, bi) = Γ( ai) Γ( bi) / Γ( ai + bi), where Γ() ⋅ is the gamma function, a i > 0, and b i > 0. Underthe reparameterization µ i = a i /( a i + b i ) and φ i = ( a i + b i + 1)−1, the parameters µ i and φ i are, respectively,the mean and the intralitter correlation parameters in the ith group. The mean of y ij isE( yij) = nijµ i, and the intralitter correlation is Corr( yijk , yijk ′) = φi. The variance of y ij is Var( yij) = [ φ i( nij − 1) + 1][ nijµ i( 1− µ i)].(The intralitter correlation coefficient is positive ( φ i > 0) for mostdevelopmental toxicity endpoints. Thus, the variance of the total response y ij in a litter is greaterthan the nominal binomial variance n ij µ i ( 1 − µ i ) (here the mean parameter µ i corresponds to the p iin the binomial model). The distribution of y ij is known as an overdispersion binomial, and whenφ i = 0, then y ij becomes a binomial.As for the binomial model, logistic regression can be applied to the beta-binomial model forgroup comparisons and trend tests. The parameters βs and φs are estimated by the maximumlikelihood method. The likelihood ratio or Wald test is often used to test for the significance ofparameters. However, the likelihood ratio test is preferable. One problem with the beta-binomialmodel is the bias and instability of the maximum likelihood estimates (MLEs) of the coefficientsin fitting the logistic dose-response model. Williams 19 proposed using the quasi-likelihood methodas an alternative approach to the beta-binomial model.In the quasi-likelihood approach, only assumptions regarding the mean and variance arerequired: E( yij | nij)= nijµ i and Var( yij | nij) = nijµ i( 1− µ i)[ φ( nij− 1) + 1]. Note that in the quasi-likelihoodestimation, the intralitter correlations typically are modeled as constant across groups. Thecoefficients of the βs can be obtained by solving the quasi-likelihood estimating (score) equations.The intralitter correlation coefficient is calculated by equating with the mean of Pearson chi-squarestatistics. Estimation details are given by Williams. 18 The Wald test is used to test for the significanceof βs in terms of the logistic regression model. Chen 20 and Chen et al. 21 compared several trendtests and homogeneity tests for the analysis of overdispersion binomial data from developmentalstudies.3. Example — Hydroxyurea Data⎞ij ⎟⎟⎝ ij ⎠⎟B(a + y , b + n − y )i ij i ij ijB ( a , b)A study of the effect of maternal exposure to hydroxyurea in mice is used for illustration. Theexperiment consisted of four treatment groups: 0, 150, 200, and 250 ppm. Table 17.3 shows thefrequency table of litters according to number of fetuses affected (including death, resorption, andmalformation) (y ij ) and number of implants (n ij ). Table 17.4 contains the means and standard errorsfrom the litter-based analysis, beta-binomial model, and binomial model. Note that the binomialmodel assumes that all fetuses are independent samples. Therefore, the total number of responsesii© 2006 by Taylor & Francis Group, LLC

STATISTICAL ANALYSIS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 705Table 17.4The mean (with standard error) from the litter-based analysis and themaximum likelihood estimates (with standard errors) from the beta-binomialand binomial modelsModel Parameter 0 100 200 250Litter-based µ 0.0340(.014) 0.1777(.056) 0.5058(.128) 0.7733(.044)Beta-binomial µ 0.0343(.014) 0.1720(.053) 0.4895(.111) 0.7700(.040)φ 0.0017(.040) 0.1758(.097) 0.3574(.125) 0.1136(.057)Binomial µ 0.0342(.014) 0.1642(.031) 0.5196(.049) 0.7768(.028)Table 17.5p-values for comparisons between the control and dosage groups, usingthe litter-based analysis, beta-binomial, quasi-likelihood, and binomialapproachesLSD (t-test)Control vs. DoseDunnettApproach µ 1 = µ 2 µ 1 = µ 3 µ 1 = µ 4 µ 1 = µ 2 µ 1 = µ 3 µ 1 = µ 4Litter-based analysis .010

706 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. Likelihood and Quasi-likelihood MethodsLet n ij be an observed count from the jth animal in the ith group 1 < i < g and 1 < j < m i . The countsn ij are often modeled by a Poisson distribution,nij −µ i epn ( ij) =n !ijµ i, n = 012 , , ,…,ijthe mean and variance of n ij are E( nij) = Var( nij)= µ i . A common complication in the analysis ofcount data is that the observed variation often exceeds or falls behind the variation that is predictedfrom a Poisson model. A generalized variance for n ij is of the form Var( n ij ) = µ i ( 1 + φ i µ i ).If φ i > 0,then n ij has an extra-Poisson variation; if φ i < 0, then n ij has a sub-Poisson variation; and if φ i = 0,then n ij becomes a Poisson. In the Poisson model, the mean function is often modeled by a loglinearfunction. The dose-response model for trend test is µ i = exp( β0 + β1di). This analysis isreferred to as the Poisson regression.The classical extra-Poisson modeling assumes that the mean of the Poisson has a gammadistribution, which leads to a negative binomial (gamma-Poisson) distribution for the observed data,−1ijΓ( nij+ φi) ⎛ φµ ⎞ ⎛i ipn ( ij)=−1Γ( n + 1) Γ( φ ) ⎝⎜ 1 + φµ ⎠⎟ 1 ⎞⎝⎜ 1 + φµ ⎠⎟ijiiinii1φiwhere φ i > 0. The maximum likelihood estimation of the negative binomial model was describedin detail by Lawless. 22 As in the beta-binomial model, the restriction on φ≥0 can limit its application.For example, the number of implantations or the number of corpora lutea may exhibit asub-Poisson variation. The quasi-likelihood approach 23,24 provides a method to model both extraorsub-Poisson variation data. The quasi-likelihood approach assumes that the mean and varianceof count data are of a negative binomial form, E( n ) = µ and Var( n ) = µ ( 1 + φ µ ).B. Exampleij i ij i i iThe example is the hydroxyurea data considered in Section III. In this example, the effect ofhydroxyurea exposure on the number of implants was analyzed. Because the exposure occurredafter implantation, the numbers of implants would not be expected to differ among the four groups.The means and variances for the four groups areµ ˆ = 10. 29 ( 3. 72) , µ ˆ = 10. 76 ( 5. 19) , µ ˆ = 12. 75 ( 3. 07) , µ ˆ = 10. 50 ( 3.21)1 2 34The means are much larger than the variance in all four groups. The data exhibited sub-Poissonvariation, so the negative binomial model is not appropriate for the analysis.Table 17.6 contains the summary of the analyses using the litter-based analysis, Poisson model(Poisson regression), and quasi-likelihood approach. In the litter-based analysis, the numbers ofimplants were transformed by a square root transformation for data normalization. The litter-basedand quasi-likelihood approaches gave similar results in all three tests at the 5% significance level.The Poisson model shows larger p-values in all tests because it does not adjust for sub-Poissonvariation. The SAS PROC GLM and PROC GENMOD procedures were used in the analysis.© 2006 by Taylor & Francis Group, LLC

STATISTICAL ANALYSIS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 707Table 17.6p-values for testing homogeneity, control versus dosage groups, and trend using the litterbasedanalysis, Poisson, and quasi-likelihood approachesControl vs. DoseLSD (t-test)DunnettApproach Homogeneity µ 1 = µ 2 µ 1 = µ 3 µ 1 = µ 4 µ 1 = µ 2 µ 1 = µ 3 µ 1 = µ 4 TrendLitter-based .044 .543 .007 .743 .881 .020 .978 .130Poisson .671 .691 .086 .843 .960 .206 .995 .587Quasi-likelihood .006 .502 .001 .743 .841 .003 .977 .404V. QUANTITATIVE RISK ASSESSMENT AND BENCHMARK DOSEA fundamental goal of reproductive and developmental experiments is to determine safe levels ofhuman exposure to a substance. Such estimated levels are commonly referred to as reference doses(RfDs) (see Chapter 21). The RfDs are presumed to cause no or negligible health risks in humans.Risk assessment for toxic effects other than cancer has traditionally been based on the concept ofa dosage “threshold.” At doses below such a threshold, adverse effects are not expected to occur.RfDs for reproductive and developmental effects are commonly set by dividing experimental noobserved-adverse-effect-levels(NOAEL) by appropriate safety (uncertainty) factors. Various limitationsfor the NOAEL plus uncertainty factor approach have been documented. 25,26 Crump 27proposed that the NOAEL be replaced by a benchmark dose estimated to be associated with anexcess risk in the range of 1% to 10%. This section describes a procedure for constructing RfDsfor correlated binary developmental endpoints.A. Dose-Response Model and Benchmark DoseDose-response models are usually formulated to establish a mathematical relationship between theobserved response and the dose. In addition to providing a mathematical framework for testinghypotheses as shown in the previous sections, an important use of dose-response models is todetermine RfDs for risk assessment. The standard approach involves fitting a dose-response functionto the experimental data and then using the fitted model to estimate the RfD. Because relativelysmall numbers of animals are exposed to only a few dose levels, the dose level used to produce aresponse is often much higher than the environmental exposure level of interest. It is well knownthat various dose response functions that fit the experimental data adequately can yield extrapolateddoses differing by several orders of magnitude. In general, one should aim for a model that is bothconsistent with the available biological knowledge and fits the observations reasonably well.For binary endpoints, such as malformation or death, the dose response function Pd ( ) is theprobability of occurrence of the adverse effect associated with the dose level d. The quantity ofinterest from the risk assessment viewpoint isπ( d) = P( d) −P( 0)which measures the excess risk (above the background risk) due to the added dose. A “safe” doseis defined as the dose for which the excess risk satisfies π( d ) = π∗ , where π ∗ is a predeterminedsmall number (e.g., 10 –5 ).Krewski and Van Ryzin 28 demonstrated that the risk estimates based on a variety of doseresponsemodels do not differ significantly above an incidence of 1%, although large relativedifferences can be derived at doses corresponding to lower risks. In general, the dose correspondingto a risk of 10% can be estimated with adequate precision and is not particularly dependent on the© 2006 by Taylor & Francis Group, LLC

708 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONdose-response model used to fit the data. The doses corresponding to an excess risk between 1%(ED 01 ) and 10% (ED 10 ) have been suggested as a point of departure (an anchor point) for riskassessment. 27,29,30 Thus, a “benchmark” dose is defined as a dose that is estimated to correspond toan excess risk in the range of 1% to 10%. With this benchmark dose as a starting point, uncertaintyfactors (or linear extrapolation) are then employed to establish RfDs. In order to accommodateexperimental variation, it is recommended that the lower confidence limit for the benchmark dosebe used. 26,28 For developmental effects, fetal death and malformation rates often exceeds 1% intypical developmental studies, and replacing the NOAEL with the ED 01 (1% benchmark dose) willoften result in a lower human exposure guideline. On the other hand, the ED 10 can be regarded asthe LOAEL (lowest-observed-adverse-effect-level) that can be detected to show a significant doseeffect. Thus, the lower confidence limit dose corresponding to excess risk between 1% and 10%(LED p ) is used as a benchmark dose. By use of an uncertainty factor of 10,000 (linear extrapolation)from a lower bound on the benchmark dose with an estimated risk of 10% (LED 10 ), the dosecorresponding to a risk of less than 10 –5 is estimated by LED 10 /10,000. Calculation of a benchmarkdose is illustrated below.As illustrated in Section III, the distribution of the total number of (adverse) responses in alitter corresponds to a beta-binomial model,⎛⎜nPy ( ij)= ⎜⎜ y⎞ij ⎟⎟⎝ ij ⎠⎟B(a + y , b + n − y )i ij i ij ijB ( a , b)The probability of the adverse response at the dose d i is P( di) = µ i = ai /( ai + bi). A parametricdose-response function is used to describe the relationship between the probability of response andthe exposure level. Chen and Kodell 31 proposed a Weibull dose response model,iiwPd ( ) = µ = −exp( −β − β d )i i 1 0 1 iThe power parameter on dose, w > 0, increases the flexibility in the shape of the dose responsefunction. The benchmark dose corresponding to 10% excess risk (ED 10 ) is the dose that satisfiesthe equation010 . = exp − 0 −exp− 0 − 1w( β ) ( β β d i )The estimate of ED 10 is computed from the coefficient estimates of the fitted Weibull dose-responsefunction. To estimate a lower confidence limit on the ED 10 , Chen and Kodell 31 proposed using thelikelihood confidence limit approach by solving this problem:subject to the constraintlog[ 1− 0.1exp( β0)]minimize−β2( Lmax − L1)= 1.645where 1.645 is the 95th percentile of the standard normal distribution (for a 95% confidence limit),L max is the maximum value of the log-likelihood subject to no constraint, and L 1 is the maximumvalue of the log-likelihood subject to the 0.10 constraint.12© 2006 by Taylor & Francis Group, LLC

STATISTICAL ANALYSIS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 709B. ExampleWe use the hydroxyurea data to illustrate the procedure. The maximum likelihood estimates of theWeibull dose-response model are ˆ−2 , ˆ −11 α = 3. 46 × 10 β = 9. 88 × 10 , and ŵ = 4.24. The point estimatesand the 95% lower confidence limits of benchmark doses areBenchmark dose 1% 5% 10%MLE estimate 78.05 114.67 135.92Confidence limit 49.15 84.62 107.51Based on the result from the quasi-likelihood method in Table 17.4, the NOAEL is 100 ppm. Thisnumber is very close to the LED 10 of 107.51 ppm. Benchmark doses can be calculated with softwaresuch as TOX_RISK 32 and ToxTools. 33VI. DISCUSSIONStatistical analysis of developmental endpoints presents a number of important issues. Developmentalendpoints consist of both continuous and discrete outcomes. The analysis should accountfor the correlations induced by the clustering effects within litters. The litter-based ANOVAapproach is the simplest method for data analysis. Once the data are normalized, the ANOVA canbe applied directly as described. The ANOVA is also used in more complex experiments involvingcrossed (e.g., dose levels, replicates) and nested (e.g., litter effects) factors. The repeated measuresANOVA is used for the analysis of postnatal behavioral data. The ANOVA and conventional posthoc test procedures such as the Dunnett test and linear regression analysis are all available instandard statistical software packages.The parametric likelihood-based and quasi-likelihood (generalization estimating equations)approaches are two alternative methods. The mixed-effects model is a general parametric approachto modeling continuous outcomes. The parametric model for the binary outcomes from fetalresponses is described in terms of the sum of correlated binary variables resulting in a generalizedbinomial distribution. The distribution for the common developmental endpoints, such as the numberof malformations per litter, is often an overdispersed binomial, but the distribution of sex combinationscan be an underdispersed binomial. For count data, the negative binomial distribution hasbeen used to model overdispersed count data, such as results from an ovarian toxicity experiment.However, the parametric approach has disadvantages in that it does not allow for underdispersedvariations and its maximum likelihood estimates can be biased or numerically unstable. The quasilikelihoodapproach is a semiparametric approach. It can be applied to modeling both over- andundervariation data. The estimates derived are generally more stable than the parametric maximumlikelihood estimates. The quasi-likelihood approach does not provide inference on the dispersionparameter.In a typical reproductive and developmental study, various endpoints are collected in order toobtain the maximum information from the study. As described, the standard approach for riskassessment of a compound has been based on the analysis of each reproductive and developmentalendpoint separately. Statistical analysis often involves a large number of tests, both in terms ofmultiple comparisons among groups and multiple tests on many endpoints. Multiple testing procedures,e.g., Dunnett’s test, are commonly applied to group comparisons, rather than to the analysisof test endpoints. In a single experiment, statistical tests are performed on 10 to 20 reproductiveand developmental endpoints. Because of the conduct of a large number of statistical tests, thechance of false positive findings is considerably increased. For example, the overall false positiverate is about 0.40 ≈1−( 1− 005 . )10 for tests of 10 independent developmental and reproductiveendpoints, all at α = 0.05.© 2006 by Taylor & Francis Group, LLC

710 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONInterpreting results of a reproductive or developmental experiment is a complex process, andthere are risks of both false negative and false positive results. The relatively small number ofanimals used and the low incidence rates can cause reproductive or developmental effects of acompound not to be detected. Because of the large number of comparisons involved, there is alsoa great potential for finding statistically significant positive trends or differences in some endpointsthat are due to chance alone. Furthermore, a significant result for an endpoint tested may bestatistical, biological, or both. The issues of false positive and false negative error rates in testingmultiple reproductive and developmental endpoints have not been addressed. The main considerationshould be to maximize the power of identifying every significant endpoint, while controllingthe overall Type I or familywise error rate. Several authors 34–38 have proposed joint modeling ofmultiple developmental outcomes. However, development of risk assessment methods for combinedanalysis of several reproductive and developmental outcomes will require further statistical research.REFERENCES1. Desu, M.M. and Raghavarao, D., Sample Size Methodology, Academic Press, New York, 1990.2. Winer, B.J., Brown, D.R., and Michels, K.M., Statistical Principles in Experimental Design, 3rd ed.,McGraw-Hill, New York, 1991.3. Hamilton, L.C., Regression with Graphics: A Second Course in Applied Statistics, Duxbury Press,Belmont, CA, 1991.4. SAS Institute Inc., Procedures Guide, Version 8, Cary, NC, 1999.5. Dunnett, C.W., A multiple comparison procedure for comparing several treatments with a control, J.Amer. Stat. Assoc., 50, 1096, 1955.6. Williams, D.A., The comparison of several dose levels with a zero dose control, Biometrics, 28, 519,1972.7. Tukey, J.W., Ciminera, J.L., and Heyse, J.F., Testing the statistical certainty of a response to increasingdoses of a drug, Biometrics, 41, 295, 1985.8. Barlow, R.E., et al., Statistical Inference Under Order Restrictions, John Wiley, New York, 1972.9. Lehmann, E.L., Nonparametrics: Statistical Methods Based on Ranks, Holden-Day, San Francisco,1975.10. Dempster, A.P., et al., Statistical and computational aspects of mixed model analysis, Appl. Stat., 33,203, 1984.11. SAS Institute Inc., SAS/STAT User’s Guide, Version 8, Cary, NC, 1999.12. Liang, K.Y. and Zeger, S.L., Longitudinal data analysis using generalized linear models, Biometrika,73, 13, 1986.13. Carey, V. and McDermott, A., gee() function, StatLib Archives, Carnegie Mellon University, Pittsburgh,PA. (S-PLUS function obtainable via anonymous FTP from lib.stat.cmu.edu., 1996.)14. Cochran, W.G., Some methods for strengthening the common χ 2 tests, Biometrics, 10, 417, 1954.15. Armitage, P., Tests for linear trends in proportions and frequencies, Biometrics, 11, 375, 1955.16. StatXact, Cytel Software Corporation, Cambridge, MA, 1992.17. Jonckheere, A.R., A distribution-free k-sample test against ordered alternatives, Biometrika, 41, 133,1954.18. Williams, D.A., The analysis of binary responses from toxicological experiments involving reproductionand teratogenicity, Biometrics, 31, 949, 1975.19. Williams, D.A., Extra-binomial variation in logistic linear models, Appl. Stat., 31, 144, 1982.20. Chen, J.J., Trend test for overdispersed proportions, Biometrical J., 35, 949, 1993.21. Chen, J.J., Ahn, H., and Cheng, K.F., Comparison of some homogeneity tests in analysis of overdispersedbinomial data, Environ. Ecolo. Stat., 1, 315, 1996.22. Lawless, J.F., Negative binomial and mixed Poisson regression. Canadian J. Stat., 15, 209, 1987.23. Breslow, N.E., Extra-Poisson variation in log-linear model, Appl. Stat., 33, 38, 1984.24. Chen, J.J. and Ahn, H., Fitting mixed Poisson regression models using quasi-likelihood methods,Biometrical J., 38, 81, 1996.© 2006 by Taylor & Francis Group, LLC

STATISTICAL ANALYSIS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 71125. Munro, I. and Krewski, D., Risk assessment and regulatory decision making, Food Cosmet. Toxicol.,19, 549. 1981.26. Kimmel, C.A. and Gaylor, D.W., Issues in qualitative and quantitative risk analysis for developmentaltoxicology, Risk Analysis, 8, 15, 1988.27. Crump, K.S., A new method for determining allowable daily intakes, Fundam. Appl. Toxicol., 854,1984.28. Krewski, D. and Van Ryzin, J., Dose response models for quantal response toxicity data, in Statisticsand Related Topics, George, M., Dawaon, D., Rao, J.N.K., and Saleh, E., Eds., North-Holland,Amsterdam, 1981, p. 201.29. U.S. Environmental Protection Agency, Proposed Guidelines for Carcinogen Risk Assessment: Notice,Fed. Regist., 56, 17995, 1996.30. Gaylor, D.W., et al., A unified approach to risk assessment for cancer and noncancer endpoints basedon benchmark doses and uncertainty/safety factors, Regulat. Toxicol. Pharmacol., 29, 151, 1999.31. Chen, J.J. and Kodell, R.L., Quantitative risk assessment for teratological effects, J. Am. Stat. Assoc.,84, 966, 1989.32. ICF/Clement, TOX_RISK, Ruston, LA. 1989.33. ToxTools, Cytel Software Corporation, Cambridge, MA, 1999.34. Lefkopoulou M, Moore D., and Ryan L.M., The analysis of multiple correlated binary outcomes:application to rodent teratology experiments, J Am. Stat. Assoc., 84, 810, 1989.35. Chen, J.J., et al., Analysis of trinomial responses from reproductive and developmental toxicityexperiments, Biometrics, 41, 47, 1991.36. Catalano, P.J. and Ryan, L.M., Bivariate latent variable models for clustered discrete and continuousoutcomes, J. Am. Stat. Assoc., 87, 651, 1992.37. Ryan, L.M., Quantitative risk assessment for developmental toxicity, Biometrics, 48, 163, 1992.38. Chen, J.J., A malformation incidence dose response model incorporating fetal weight and/or littersize as covariates, Risk Anal., 13, 559, 1993.© 2006 by Taylor & Francis Group, LLC

CHAPTER 18Quality Concerns for Developmental andReproductive ToxicologistsKathleen D. Barrowclough and Kathleen L. ReedCONTENTSI. Introduction ........................................................................................................................714II. Good Laboratory Practice Standards.................................................................................714A. Background................................................................................................................714B. Regulations ................................................................................................................714C. Regulations Specific to Electronic Records and Electronic Signatures ...................715D. Meeting 21 CFR Part 11 Requirements....................................................................715E. Current Trends ...........................................................................................................716III. The GLP Compliance Program .........................................................................................716A. Overview....................................................................................................................716B. Roles and Responsibilities.........................................................................................7171. Management.........................................................................................................7172. Quality Assurance................................................................................................7183. Study Directors ....................................................................................................719IV. Auditing Developmental and Reproductive Toxicity Studies ...........................................719A. Overview....................................................................................................................719B. Best Practices.............................................................................................................7201. Protocol Review...................................................................................................7202. In-Life Phase Audits............................................................................................7203. Records and Report Audit ...................................................................................725V. Quality Concerns and Quality Control Measures .............................................................727A. Overview....................................................................................................................727B. Prestudy......................................................................................................................727C. In-Life ........................................................................................................................727VI. Conclusions ........................................................................................................................730Acknowledgments ..........................................................................................................................730References ......................................................................................................................................731713© 2006 by Taylor & Francis Group, LLC

714 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONI. INTRODUCTIONIn today’s business and regulatory climate the ability to generate quality products in the mostefficient manner is critical to maintaining competitiveness. Regulators must rely on the scientificcommunity to provide the best possible data for input into their decision-making process. Thoseresponsible for submitting applications to the regulatory agencies must continually strive for themost efficient way to develop a quality product. This chapter will• Review the Good Laboratory Practice Standards (GLPs), which describe minimal standards forconducting nonclinical laboratory studies that support or are intended to support applications forresearch or marketing permits for products regulated by the Food and Drug Administration (FDA)or the Environmental Protection Agency (EPA).• Provide guidance for developing and managing an effective GLP compliance program.• Discuss specific procedures for auditing developmental and reproductive studies.• Summarize quality concerns as solicited from a number of toxicologists, study personnel, andquality assurance unit (QAU) personnel from multiple organizations.• Discuss recommendations for addressing the quality concerns.We hope that the information presented herein will provide the developmental and reproductivetoxicologist with useful tools to consider during all aspects of a study, from protocol preparationto final report issue.A. BackgroundII. GOOD LABORATORY PRACTICE STANDARDSIn the mid-1970s, inconsistencies and unacceptable as well as fraudulent practices were uncoveredin several studies that had been submitted to the FDA. 1 The kinds of issues that surfaced had thepotential to have a detrimental, if not harmful, effect on the consumer.Because of the FDA’s commitment to and responsibility for protecting the consumer, the agencypromulgated the FDA Good Laboratory Practice for Nonclinical Laboratory Studies, finalized in1979 and revised in 1984 and again in 1987. 2 By 1983, the EPA had promulgated the EPA ToxicSubstance Control Act (TSCA) and Federal Insecticide, Fungicide, Rodenticide Act (FIFRA) GLPs,both of which have since been revised. 3,4 The EPA’s TSCA and FIFRA GLPs were intended toconform as closely with the FDA GLPs as possible. 5Be aware that the issues prompting these responses by the federal regulatory agencies were notsimply minor judgmental oversights with little consequence. There were occurrences of underqualifiedpersonnel and supervision, selective reporting, poor animal care procedures, and inadequaterecordkeeping and missing data, sometimes a result of fraudulent activity. 3Over the years several international agencies have developed regulations or guidelines thatemulate the U.S. GLPs. For example, the Organization of Economic Cooperation and Development(OECD) Principles of GLP are based on the FDA GLPs and are similar though not identical tothem in content. 1B. RegulationsWhile the FDA and the EPA TSCA and FIFRA GLPs are not identical, there are enough similaritiesamong all three that if one adheres to the principles articulated by one GLP standard, very littleadjustment will be required to be compliant with another.© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 715If one considers the issues that prompted promulgation of GLPs, it is not difficult to envisionthe major thread running through the regulations – documentation:• Plan it well — protocol requirements• Conduct it well — standard operating procedures, documentation requirements• Report it well — reporting requirementsOther GLP requirements include:• Personnel training• Study director responsibilities• QAU responsibilities• Adequacy and operation of facilities• Calibration and maintenance of equipment• Handling and characterization of test, control, and reference substances• Retention of recordsC. Regulations Specific to Electronic Records and Electronic SignaturesWith the coming of the “computer age” came a desire to take advantage of the increased efficienciesand reduced opportunities for error in the generation and manipulation of data. What better wayto accomplish this than by computerizing one’s processes?Unfortunately, in the early 1990s when this proliferation of electronic record-keeping wasoccurring, there were no regulations to provide guidance for the industry or to provide assuranceto regulatory agencies that electronic records were as reliable as manually created records or thatelectronic signatures were irrefutable. Along came the Good Automated Laboratory Practices(GALP) 6 issued by the EPA in an effort to provide such guidance to regulated industry. UnlikeGLPs, the GALPs are guidance for industry, rather than regulation.In 1997, FDA promulgated the Electronic Records, Electronic Signatures Rule (21 CFR Part11). 7 This rule has not been without its controversies and issues. The regulated pharmaceuticalindustry, while striving to achieve compliance with 21 CFR Part 11, is not considered by most tohave reached this goal. The onset of 21 CFR Part 11, coupled with the increasing automation foundin testing entities, raises a new issue for the developmental and reproductive toxicologist: Are allthe systems validated, and are there appropriate controls to assure integrity of data in automateddata collection systems?FDA is not alone in its concern for the integrity of electronic records. The EPA has proposeda Cross-Media Electronic Reporting and Record-Keeping Rule (CROMERRR), 8 which at the timeof this writing has not been published as a final rule.D. Meeting 21 CFR Part 11 RequirementsThe pharmaceutical industry struggled with developing 21 CFR Part 11 compliant programs. Onthe one hand, there was a lack of available technology to meet the demands of the rule. However,the primary focus for industry in 1999 was the potential for significant glitches in computer systemswhen the new year arrived. It was believed that internal system clocks based on two digit years(e.g., 97, 98, 99) might not identify “00” in a manner that allowed computer operations dependenton dates to function normally. This concern, felt around the world, was labeled “Y2K.” Thepharmaceutical industry, like most industry was devoting significant resources to assuring Y2Keffects would not halt or disable business practices.Once the Y2K dilemma had come and gone, the pharmaceutical industry, as well as the FDA,began to tackle 21 CFR Part 11 in earnest. Recognizing the need to provide guidance for industry,© 2006 by Taylor & Francis Group, LLC

716 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFDA issued a Compliance Policy Guide in 1999. The purpose of the guide was to provide guidancefor both industry and regulators regarding what criteria would be used in making decisions onwhether to pursue regulatory actions. 9FDA decisions would be made based on the following:• Nature and extent of Part 11 deviations• Effect on product quality and data integrity• Adequacy and timeliness of planned corrective measures• Compliance history of the establishment 9Industry soon discovered that what FDA wanted to see was a good-faith plan, describing whatsteps would be taken to become compliant and establishing milestones that would measure progress.Some potential tasks were:• Develop an overall strategy and action plan• Develop policies and procedures that address 21 CFR Part 11 requirements, including the requirementfor validation• Train personnel on the meaning of electronic signatures and hold them accountable• Conduct an inventory of existing systems (referred to as legacy systems)• Prioritize the systems• Conduct evaluations to determine the Part 11 gaps in those existing systems• Develop plans to address the gaps• Revalidate systems, if warrantedThere is no “one way” to establish a 21 CFR Part 11 compliance program; however, a methodicalapproach that identifies and prioritizes the tasks may make this overwhelming project more manageable.E. Current TrendsIt is our belief that GLPs have played a significant role in improving consistency in the quality ofdata submitted to regulatory agencies. Not only have they established standards by which thescientist must conduct research, standards beyond scientific principles, but also they have provideda mechanism for research facilities to establish procedures for monitoring themselves. GLPs assignresponsibility for overall study conduct to the study director (SD). He or she must assure that thework is conducted properly, meets the GLP requirements, is documented clearly, is conducted withwell-maintained and appropriately calibrated equipment, and results in reliable, trustworthy data.A. OverviewIII. THE GLP COMPLIANCE PROGRAMAn effective GLP compliance program needs a proactive QAU and strong management support.Assuring management that studies have been conducted in compliance with the appropriate GLPregulations is but one facet of the QAU’s responsibility. “The Quality Assurance Unit is the onlyunit in the laboratory that has a total view of the personnel, facilities, equipment, and processesthat are occurring. From this perspective, the QAU can analyze the work and paper flow, identifydefective practices and procedures, and recommend suitable corrections.” 10 In order to realize thegreatest benefit, the GLP compliance program incorporates GLP regulations and draws from theorganization’s experiences (metrics) to define processes that build quality and efficiency into allphases of a study.© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 717B. Roles and Responsibilities1. ManagementManagement must clearly define its expectations for the overall organization. The GLPs provideminimal requirements but leave it up to an organization to determine how it will set the standards.Minimally, testing facility management must• Designate a study director before the study is initiated• Replace the study director promptly when necessary• Assure the presence of a QAU• Assure that test, control, and reference substances or mixtures have been appropriately tested foridentity, strength, purity, stability, and uniformity, as applicable• Assure that personnel, resources, facilities, equipment, materials, and methodologies are availableas scheduled• Assure that personnel clearly understand the functions they are to perform• Assure that any deviations from the GLP regulations reported by the QAU are communicated tothe study director and corrective actions are taken and documented 2–4One approach in setting quality expectations that are in alignment with the GLP regulations is todevelop high-level policies, which may be referred to as “quality directives.” These policy statementspresent the goals, expectations, requirements and standards of practice. They provide direction andguidance for meeting regulatory requirements as well as for supporting good business practices andare signed by the highest level of management in an organization. In addition, quality directives providea mechanism for assuring that the procedures used throughout the organization are consistent. An addedbenefit is that personnel may now move from one assignment to another without the additional burdenof learning new policies. Policies that may be addressed in quality directives are:• Creating, Revising, and Retiring Standard Operating Procedures (SOPs): “A testing facilitymust have standard operating procedures in writing setting forth study methods that managementis satisfied are adequate to ensure the quality and integrity of the data generated in the course ofa study.” 2–4 This quality directive provides consistent guidance for the creation, revision, andretirement of SOPs, including but not limited to formatting, numbering, and approval structure.• Best Practices for Meeting Electronic Records and Signature Requirements: Electronic recordsand signature requirements may be too extensive to address in one policy document. One approachis to issue two or three separate documents:– Define the organization’s policy on electronic signatures used in the electronic records subjectto 21 CFR 11, other applicable regulations, and internal policies. Explain the requirements,components, and controls specific to electronic signatures.– Address why validation is required, what needs to be validated in the GLP environment, whenvalidation needs to occur, and who is responsible for the validation process. The directives donot address how to validate, but do stress the importance of documentation for the process andmaintenance of the documentation.• Study Director and Principle Investigator Review of Data: Define expectations for regularreview of data for completeness, accuracy, and organization to ensure that the records demonstrateproper study conduct, adherence to the protocol, and compliance with applicable SOPs and GLPs.• Training Records: GLPs require that each individual engaged in the conduct of or responsiblefor supervision of a study shall have the education, training, and experience, or combination thereof,to enable that individual to perform the assigned functions. 2–4 This Directive provides the guidanceneeded to assure that the appropriate documentation exists to demonstrate compliance with theGLP requirement.As quality programs are developed and a need for management expectations is identified, otherquality directives may be created. Additional topics that may be addressed by policy documentsinclude data handling, security, and a metrology program.© 2006 by Taylor & Francis Group, LLC

718 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONWhile some individuals are concerned about yet another set of documents to be managed andaudited against, the benefits from having an organization that clearly understands managementexpectations far outweigh the perceived inconvenience of additional paperwork. In addition, havingmanagement approved policies that set expectations for a quality-driven, GLP compliant organizationgoes a long way toward assuring customers of management’s commitment to assuring thequality of the work product.2. Quality AssuranceGLPs require that management “assure the presence of a QAU.” 2–4 The responsibilities of the QAUinclude:• Assure management that facilities, personnel, practices, and records are in compliance with regulations• Maintain a master schedule sheet of studies• Inspect each nonclinical study at intervals to assure compliance• Report findings to the study director and management• Review the final report to assure that it accurately reflects the raw data• Prepare and sign a QA statement in the final reportThe regulations also require that procedures used by the QAU to fulfill its responsibilities bein writing. An added benefit of having these written procedures is assuring consistency among theauditors. One way to document the QAU procedures is to utilize a mixture of QAU SOPs andguidance policies. SOPs define the practices that must be adhered to for each study reviewed bythe QAU. Potential SOP topics include:• Activities and responsibilities of the QAU• Maintenance of the QAU master schedule• Maintenance of QAU files• Routine audit functions (protocol, conduct, records, reports)• Internal facility inspections• Distribution of audit reports to principal investigators, study directors, and managementGuidance policies may contain information on how to use a specific program (i.e., user’s manualfor the QAU master schedule) or how to conduct regulatory or customer inspections. Or they mayprovide guidance on auditor training. Guidance policies are not as prescriptive as SOPs and affordgreater flexibility.GLPs define the role of the QAU as “monitoring each study to assure management that thefacilities, equipment, personnel, methods, practices, records, and controls are in conformance withthe regulations in this part.” 2–4 The QAU can and should play a larger role in the success of abusiness organization. Regulators, technical management, and the scientific community have recognizedthe enormous potential of the QAU. “The role of the quality assurance (QA) auditor shouldtranscend simple verification of whether columns of numbers are added correctly or all of the crossoutshave been duly initialed and dated. The QA professional should be capable of bringing tolight considerations of scientific quality as well as data integrity.” 11 This sentiment is echoed byPaul Lepore, formerly of the FDA. 10 Consequently, many organizations today include the followingQAU roles and responsibilities in the GLP compliance pogram:• Liaison program• Compilation, analysis, and report back of metrics• GLP training for the technical staff (introductory and ongoing)• Consultation• Mentoring© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 719QAU auditors interact on a day-to-day basis with the individuals involved in study work, andwhile it is important for auditors to pay attention to what is happening with the current study, theyshould be continually looking at the work processes from a much broader perspective. As reportedby Caulfield, “Management expects the QAU to serve as the primary conduit through which theyreceive information regarding organizational needs like training, equipment quality and availability,inadequacies in facility, or gaps in study administration.” 12Though the QAU is primarily responsible for monitoring studies and facilities for GLP compliance,“it is clearly the QAU who is expected to provide much of the informational raw materialsneeded for them [management] to make informed decisions and to take any needed action.” 12 Themost effective organization will expect the QAU to identify and recommend changes to existingprocesses that will result in increased efficiency and effectiveness.3. Study DirectorsThe study director is defined as “the individual responsible for the overall conduct of a study” andis responsible “for the interpretation, analysis, documentation and reporting of results, and representsthe single point of control.” 2–4 The study director:• Assures the protocol is approved and followed. All changes in or revisions of an approved protocoland the reasons for the changes and revisions shall be documented, signed by the study director,dated, and maintained with the protocol.• Assures all data are accurately recorded and verified.• Assures unforeseen circumstances that may affect the quality and integrity of the study are notedwhen they occur, and corrective action is taken and documented.• Assures test systems are as specified in the protocol.• Assures all applicable GLPs are followed.• Assures all raw data, documentation, protocols, specimens and final reports are archived.• Assures all deviations from SOPs in a study are authorized by the study director and documentedin the raw data. 2–4While the list above may not seem exhaustive, it must be pointed out that the study directormust assure that all applicable GLPs are followed. This is a daunting task, and one that the studydirector cannot hope to successfully accomplish without assistance. “It is incumbent upon the QAUto assist them [study directors] in meeting the formidable responsibilities that have been placed onthem by the regulations.” 12 For instance, the study director needs to know that those personnelassigned to a study are adequately trained; however, management is responsible for assuringpersonnel clearly understand the functions they are to perform. 2–4 Additionally, the QAU can assistby reviewing training documentation for personnel observed working on a particular study. Managementmay designate supervision to be responsible for an annual review of training documentationand assignment of qualified personnel to a particular study function.Additionally, procedures should be in place to assure the study director that the facilities areappropriate, equipment functions as required, reagents and solutions are appropriately labeled, andthat appropriate SOPs are in place. The QAU should utilize its audit program to assist the studydirector in determining the compliance status of the resources to be used on his or her study andto provide recommendations for meeting the GLP requirements.A. OverviewThe QAU is required to:IV. AUDITING DEVELOPMENTAL ANDREPRODUCTIVE TOXICITY STUDIES© 2006 by Taylor & Francis Group, LLC

720 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION• Inspect each study at intervals adequate to ensure the integrity of the study.• Review the final report to assure that such report accurately describes the methods and standardoperating procedures, and that the reported results accurately reflect the raw data of the study. 2–4Nowhere in the regulations does it indicate how the QAU is to accomplish these tasks. Therefore,the QAU should:• Define audit functions in SOPs.• Determine an audit plan.• Decide who the responsible individuals are and how audit findings will be communicated andacted upon.The following sections provide one way of managing the audit function. The informationpresented incorporates best practices from various organizations. Keep in mind that as changesoccur in testing guidelines and technology, the most effective QAU continually looks for ways toimprove its audit practices — always looking for the “biggest bang for the buck.”B. Best Practices1. Protocol ReviewIdeally, the protocol is provided to the QAU as a draft document. The auditor reviews the protocolto assure that all items required by the GLPs have been incorporated and addressed. Whileresponsibility for the content of the protocol is clearly the responsibility of the study director, someorganizations expect the QAU to verify also that the protocol incorporates testing guideline requirements.In either case, use of a standard checklist may facilitate the review. An example of a protocolchecklist for GLP compliance is provided in Table 18.1.Once the review is complete, the auditor works with the study director to resolve any issues.A formal audit report requiring the study director to respond to findings may be issued if warranted.However, GLPs do not specifically require that an audit report be issued for the protocol review.Since the signed protocol needs to be submitted to the QAU, a review of that document providesan indication of whether issues have been addressed appropriately. Should the QAU not have anopportunity to review the draft document and should changes be required to the issued protocol,a protocol amendment will be requested.There may be instances where a protocol needs to be issued prior to having all the requisiteinformation available. For instance, a range finding study to determine dosage levels is ongoing,but the protocol for the main study must be issued to ensure that the resources necessary for thestudy are available. It is a requirement that the protocol include “each dosage level, expressed inmilligrams per kilogram of body or test system weight or other appropriate units, of the test, control,or reference standard to be administered and the method and frequency of administration.” 2–4 Inthis example, it is appropriate to indicate in the protocol that an amendment will be issued toaddress dosage levels once the range finding study has been completed.Another example of a protocol required item that may not be known at the time of protocoldevelopment or issue is the source of supply for the test system. In this case, it is acceptable toindicate that the exact source of supply will be documented in the study records. Care should betaken against using this approach indiscriminately. It is not acceptable to have a protocol riddledwith the “to be addressed in the study records” statement.2. In-Life Phase AuditsThe QAU “inspects each study at intervals adequate to ensure the integrity of the study.” 2–4 Specificsof what constitutes adequate intervals are not defined in the GLPs. The GLPs also do not define© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 721Table 18.1Protocol review checklistI. General Yes No Comment1. Descriptive title of the study2. Statement of purpose of the study3. Name and address of the sponsor4. Name and address of the testing facility5. Proposed experimental start and termination dates6. Records to be maintained7. Area for the signature and date of the study director8. Sponsors date of approval and/or area for sponsor signatureII. Test System (w/a a ) Yes No Comment9. Species (w/a; animal-rat, plant-apple tree, or soil- silt loam)10. Strain/substrain (w/a; Sprague-Dawley, McIntosh, or Keyport)11. Source of supply (w/a)12. Number (w/a)13. Sex, age, body weight range (w/a)14. Procedure for identification of the test system (pot, sample ID)15. Justification for selection of the test systemIII. Test and Control Substance Yes No Comment16. Identification of the test/control/reference substance17. Identification of solvents/emulsifiers/suspensions and carriers to be usedwith the test, control or reference substance(s)18. Specifications for acceptable levels of contaminants19. Route of administration and reason for its choiceIV. Methods (Study Conduct) Yes No Comment20. Description of the experimental design, including methods for control ofbias21. Each dosage level of the test, reference or control substance to beadministered, expressed in the appropriate units of measure22. Method and frequency of the administration of the test, control, or referencesubstance23. Type and frequency of tests, analyses, and measurements24. Proposed statistical methodsV. Feed and Water for Plant and Animal Studies (w/a) Yes No Comment25. Description and/or identification of diet used (w/a)26. Specifications for acceptable levels of contaminants (w/a)aWhen applicable.© 2006 by Taylor & Francis Group, LLC

722 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 18.2Critical phasesDose and diet preparation and sampling for analysisDose and diet analysisDose administration and exposure (test system identification and housing)Data collection (body weights, clinical observations, food consumption)MatingCulling proceduresBlood collectionVaginal lavage and slide preparation (estrous cycle evaluation)Epididymal and testicular sperm counts, morphology, motilityPup data collection (weighing and sexing)Vaginal patency, balanopreputial separation, anogenital distanceNecropsy and fetal examination (visceral and skeletal)Neurotoxicology endpoints (observe one or more):FOB* and motor activityAuditory startleWater mazeBrain collection and measurement of weightsNeuropathology* Functional observation battery.“critical phases.” Therefore, the QAU works with the technical staff to determine adequate intervalsand critical phases for each study type. Critical phases essential to the quality and integrity of thestudy are identified through dialogue between the study director and the QAU. The technical staffand QAU routinely review this list to assure that changes in testing guidelines or study design havenot affected the phases considered critical to a particular study. Table 18.2 identifies critical phasestypical for developmental or reproductive studies. Whether a phase relates to a particular study isdriven by the protocol.When determining adequate intervals for critical phase inspections, the following should beconsidered:• Duration of experimental phase of the study• Complexity of the study• Frequency with which the study is conducted• Experience of the personnel working on the studyOne critical phase per study is audited for studies that last less than 28 days and are routine innature. The procedure for assigning the critical phase to be inspected ensures that all critical phasesof a particular study type are subject to inspection within a reasonable time frame (e.g., each criticalphase for a particular study type is reviewed at least once per year). The identity of the criticalphase, date of inspection, identity of auditor, and dates the findings are reported to the study directorand management are entered into the audit log for the subject study. In addition to the critical phaseinspection, the protocol, records, and final report are also reviewed. Additional critical phases maybe reviewed at the discretion of the study director. The auditor may choose to review additionalphases if the personnel working on the study are fairly inexperienced or if the auditor is lookingto increase his or her understanding of the particular study type.Studies with duration of 28 days or greater and those studies that are nonroutine in naturefollow a different auditing scheme. Each critical phase is inspected at least once during the courseof the study. Inspections should be spread throughout the course of the study. For example, havingconducted all critical phase inspections during the first 3 months of a 2-year study without anyadditional monitoring would not be considered monitoring at “adequate intervals.”Auditors review the protocol and any pertinent SOPs prior to conducting the phase inspection.Checklists are used to provide guidance for the auditor and to ensure consistency between auditors.Table 18.3 is an example of an inspection checklist for test material preparation and administration.© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 723Table 18.3Phase inspection checklist for test material preparation and administrationTraining Records Yes N/A No Comment1. Personnel have been trained; training is documented. 2. Personnel are wearing the appropriate protective clothing. SOPs, Protocol, and Amendment(s) Yes N/A No Comment3. SOPs are readily available; the protocol and SOPs are followed. 4. Deviations from protocol or SOP are documented and authorizedby the study director. Protocol amendments are issued asnecessary.5. Unforeseen circumstances are noted when they occur; correctiveaction is taken and documented. Equipment Yes N/A No Comment6. Equipment is adequately calibrated and maintained according tothe appropriate SOP. Documentation is present.7. Appropriate instrument qualification and software validation hasbeen conducted for electronic data collection systems.8. Users have been trained in the use of the equipment and software;training is documented; users know whom to contact if there is asystem problem.9. Correct computer operation procedures are followed (login, security,etc). General Yes N/A No Comment10. All data are recorded accurately, directly, promptly, and legibly,signed and dated. Audit trails accompany any changes to the data.11. All reagents and solutions are labeled to indicate identity, titer, orconcentration, storage requirements, and expiration date.12. Beakers, vials, and storage containers of dose and diet formulationsare properly identified.13. The storage container for the test or control material is appropriatelylabeled.14. Test material preparation forms and usage logs are properlycompleted; all required information is recorded correctly. 15. Each test system is uniquely identified. 16. Test system housing units are appropriately labeled. 17. Feed bin stored in the animal room is properly labeled; feed iswithin its expiration date.18. The animal room is clean and well maintained. Accountability, rackchange, and stock records are adequately completed.19. Fish culture, feeding records, and purchase information areavailable. During the critical phase inspection, study personnel, including the study director, should beable to discuss the procedures that ensure the following:• Is there evidence that personnel have been trained and that the training is documented?• Are SOPs readily available?• Are the protocol and applicable SOPs being followed?© 2006 by Taylor & Francis Group, LLC

724 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION• Have deviations from the protocol or SOPs been documented and authorized by the study director?• Are protocol amendments issued as they are necessary?• Are unforeseen circumstances noted when they occur, and is corrective action taken and documented?• Has equipment been adequately calibrated and maintained according to the appropriate SOP? Isdocumentation present?• Has the appropriate instrument qualification or software validation been conducted for electronicdata collection systems, and is there documentation of user training?• Are all data recorded accurately, directly, promptly, and legibly; are entries signed and dated; doaudit trails accompany any changes to the data?“The QAU role is to assist the team with ensuring that sufficient documentation exists to allowfor full study reconstruction and to help study areas prepare for potential regulatory inspections.” 13Clear, concise documentation is essential for providing the tools needed to accurately reconstructa study. Regulators will expect study personnel to demonstrate proficiency in the tasks they areperforming as well as exhibit comprehensive understanding of the reasons for performing the tasks.Where appropriate, the items listed above are reviewed during each critical phase inspection,regardless of the phase assigned. Examples of items specific to a particular critical phase arediscussed below.Critical PhaseDose or diet preparation andsampling for analysisDose or diet analysisDose administration and exposure(includes test systemidentification and housing)Data collection (body weights,clinical observations, food orwater consumption)Mating, breeding, and cullingproceduresBlood collectionVaginal lavage and slidepreparationNecropsy and fetal examinationNeurotoxicology endpoints(functional observation battery[FOB] and motor activity, auditorystartle, water maze, braincollection and weightmeasurement, neuropathology)Audit ItemsLabeling of containers adheres to GLP requirements.Expired reagents and solutions are not in use.Appropriate equipment (e.g., balances, pipettes) is clean, calibrated, andappropriately maintained; documentation (maintenance and calibrationlogs) to support routine and nonroutine maintenance, and calibrationrequirements is readily available.Measurements are accurately recorded in the study records, includingunits of measure.Mixing times, where appropriate, are accurately recorded in the studyrecords.Test system is appropriately and uniquely identified.Housing units are uniquely identified.Test system identification is verified upon removal from and return to ahousing unit, and prior to administering the dosage and diet.Feed containers (if appropriate) are identified (minimally) by dose level.Amount of dose or diet is accurately recorded.Balance is calibrated, and calibration is documented.Test system identification is verified upon removal from and return to ahousing unit.Appropriate terminology is used for clinical observations.Care is taken not to mate siblings.Appropriate procedures for determining the gestation period are followed.Tubes are labeled with the appropriate information (e.g., animal ID, doselevel, date of collection, etc.).Injection site and size of needle and syringe are documented.Where required, blood is mixed in tubes appropriately.Appropriate chain-of-custody documentation for sample tracking isavailable.Animal identification is verified prior to blood collection.Slide is labeled uniquely with appropriate study information.Animal identification is verified against record.Balance is calibrated, and calibration is documented.Fetuses are identified with dam code.Test is conducted in a way that precludes tester from knowing dose level.Appropriate controls are in place to minimize disturbance (white noise formotor activity).Equipment is appropriately calibrated, and calibration is documented.Animal identification is verified.© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 725Critical PhaseDetermination of epididymal andtesticular sperm counts andmorphologyPup data collection (determinationof sex, weighing)Audit ItemsSlides are labeled appropriately.Timing and temperature are consistent with protocol and/or SOPrequirements.Equipment is appropriately calibrated, and calibration is documented.Expired reagents and solutions are not in use.Balance is calibrated, and calibration is documented.Dam identification is verified prior to data collection.3. Records and Report AuditAuditors are required to “review the final study report to assure that such report accurately reflectsthe methods and standard operating procedures, and that the reported results accurately reflect theraw data of the study.” 2–4For the purposes of this chapter, we assume the report is formatted in the following manner:Text — multiple sections:GLP compliance statementSummaryIntroductionMethods and materialsResults and discussionConclusionTables — generally a summary of the data presented in the appendixesAppendixes — consisting of individual animal data (e.g., body weights, clinical observations, foodconsumption, etc.)In order to review the report in the most efficient manner, an organized approach is recommended(see Table 18.4 for a final report checklist example):Review the study records (raw data) to verify that all endpoints specified in the protocol have beendocumented clearly and concisely, and that the date of entry and identity of the individual makingthe entry are known.If individual data are reported in appendixes, utilize a predetermined process to select and verify valuesfrom the report back to the raw data. For instance, when using validated, automated data collectionsystems, it might be sufficient to spot check a random sampling of data points and review the datato ensure that any out-of-the ordinary or unexpected notes and comments are captured in footnotes.However, if the data were collected manually, a much higher level of review would be required.Many organizations require that as much as 100% of the data is verified for manually collecteddata. The QAU SOPs should indicate the process used to audit the data.Verify summary tables to the appendixes and/or study records, as appropriate. Again, the appropriatelevel of data verification is defined in the QAU SOPs.Verify any values mentioned in the final report against summary tables, appendixes, and/or raw data,as appropriate.Check the following:• Is the introduction consistent with the protocol?• Are the methods as stated in the final report supported by the protocol, amendments, SOPs,and raw data?• Is the information presented in the results and discussion section supported by the summarytables, appendixes, and raw data?• Does the results and discussion section support the conclusion section?© 2006 by Taylor & Francis Group, LLC

726 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 18.4Final report checklistFinal Report Yes N/A No Comment1. Name and address of test facility 2. Dates on which the study was initiated and completed 3. Objectives and procedures as stated in the protocol andamendments 4. Statistical methods employed for analyzing the data 5. Identification of test and control materials (chemical name, uniquesample no., chemical abstracts no. or code no., strength, purity,and composition)6. Stability of the test and control materials under the conditions ofadministration 7. Methods used (e.g., generation in inhalation studies) 8. Test system used (no. of animals, species, strain, and substrain,source, age, sex, body weight range) 9. Methods of identification of test system (e.g., tattoo, cage card, etc.) 10. Dosage, dosage regimen, route and duration of administration 11. All circumstances that may have affected study quality or integrity 12. Name of study director, names of other scientists or professionals,and names of all supervisory personnel involved in the study13. Description of the transformations, calculations, or operationsperformed on the raw data, a summary and analysis of the data,and a statement of conclusions drawn from the analysis14. Signed and dated reports of other scientists or professionalsinvolved in the study (pathologist, clinical pathologist, statistician,chemist, ophthalmologist) 15. Storage location of specimens, raw data, and the final report 16. Dated signature of the study director 17. GLP compliance page signed by the study director • Is the information reported in the summary, results and discussion, and conclusion sectionsconsistent?• Are all items required by GLPs included in the final report?• Is the GLP compliance statement accurate?The time required to audit a final report depends upon a number of factors:• Complexity of the study• Experience of the auditor• Experience of the study personnel• Data collection — manual or automated• Quality control (QC) processesUse of validated, automated data collection systems has the potential to significantly reduce notonly the time spent during the experimental portion of the study, but also the amount of time anauditor needs to review the data. Likewise, having personnel responsible for QC of the data andreport before the package is sent to the QAU will have a positive affect on the time spent in review.© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 727A. OverviewV. QUALITY CONCERNS AND QUALITY CONTROL MEASURESDrawing from the experiences of scientists, technical support personnel, and quality assuranceprofessionals, critical quality issues encountered during the conduct of developmental and reproductivetoxicity studies have been identified. Based on “what makes sense” and “what has worked,”several alternatives for diminishing the potential of these issues to affect data integrity are presented.In no way should these approaches be considered the only means for correcting deficiencies. Whilemost scientists would agree that a well-designed study conducted by a well-trained staff has highpotential for success, experience shows that it requires much more.B. PrestudyConcernLack of understanding of studyprotocol and SOPs by studypersonnel and supportingscientists.Inaccurate database-drivencollection intervals for weights,food consumption, and clinicalobservations.ApproachHold a prestudy meeting with study personnel to review protocol and SOPs.Set clear expectations around communication.Have a QC process that requires data checks on a regular basis andincludes follow-through when consistent errors are observed.Establish adequate software validation procedures.Involve the study director, statistician, primary study technician, andapplication administrator in the prestudy meeting.Study director should issue and communicate amendments or changes tothe protocol in a timely fashion.C. In-LifeConcernDosing or dietpreparation errors.ApproachSD should review and sign documentation for calculations.Use dated version control measures to assure that the most current procedure is usedin diet or dosing preparation.Create a specialized, well-trained diet preparation or formulations group.Observe diet preparation to assure that all procedures are followed.Develop, follow, and document procedures for equipment cleaning that minimizepotential for cross-contamination, e.g., mixer, polytron, mortar and pestle.Develop and follow appropriate measuring and pipetting techniques. Use of disposableequipment (e.g., pipettes) is another option.Analyze each batch of test substance (GLP requirement) and recalculate diet or dosingformulas.Include a process for checking animal identification before and after dosing.Develop a process for comparing test substance identification in protocol, study records,documentation accompanying the test substance, and the final report.Verify that all test substance storage containers are labeled according to GLPrequirements — name, CAS number, batch number, expiration date, if any, and, whereappropriate, storage conditions necessary to maintain the identity, strength, purity, andcomposition. 2–4 This can prevent use of expired, incorrect, or improperly stored testsubstances.Follow well-documented test substance accountability practices.Assure that apparatus used in continuous i.v. dosing has been adequately calibratedso that the correct dose is injected over the correct timeframe.© 2006 by Taylor & Francis Group, LLC

728 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONConcernGavage misdosings.Blood collection —inadequate handlingof specimens andinsufficient samplevolume.Equipmentmalfunctions,operator error, andimproperlymaintainedequipment.Clinical observations,especially duringFOB testing.Observation andreporting of weightloss.Transcription errors.ApproachStudy director and QC reviewers should be on the alert for increased incidences ofmisdosings and investigate to determine their cause. Determine whether it is a trainingissue or if there is a possible systemic effect causing hyperactivity that leads tomisdosing. Pregnant animals can be more susceptible to toxic effects of the testsubstance or the dosing solution.Ensure that dosing solutions are neither too hot nor too cold. Provide specificinstructions and assure that they are followed. Severe discrepancies in temperatureof the dosing solution can cause pain to the animal. Inappropriate temperature of thetest substance may cause a change in the physical or chemical properties of the testsubstance, affecting delivered dose or making the delivery difficult.Training about proper rate of administration for gavage feedings is critical.Adherence to the protocol and SOPs must be verified.Recertification training, especially when there has been a time lapse, is one way toassure retention of the needed skill.Personnel training is critical.Create a blood collection form that includes collection requirements, such as volume,handling conditions (e.g., dry ice), spin time, and spin temperature.Assure that the appropriate preservatives are in the tubes by including a checkpoint inthe procedure.Properly maintain and calibrate the equipment.Document the appropriate size needle and syringe to reduce opportunity for animalinjury due to trauma.Develop adequate calibration and preventive maintenance procedures. There is agrowing trend toward establishing metrology programs to manage instrumentcalibration and maintenance. These programs can be a cost-effective means ofmaintaining an inventory, leveraging preventive maintenance programs with vendors,and providing appropriate maintenance routines that reduce equipment downtime andenable consistent equipment performance.Develop a cleaning schedule that includes follow-up swipe tests for equipment such asdiet mixers and blenders that have a high cross-contamination potential.To reduce variability in classifying observations, several trainees may observe the sameanimals and document their findings. Follow this by reconciling the differences overseveral trials until consistency among the trainees is reached. By recertifying at regularintervals, a high level of proficiency is maintained.Standardize terms used in describing clinical observations. Design the database usingcontrolled vocabulary to ensure consistency.Minimize injury and skewed results in the water maze battery by documenting watertemperature prior to the run.Prior to FOB testing, designate primary and secondary observers so that theobservations are verified.Prior to water maze tests, verify that the maze has been set up using the correct goals,i.e., learning retention vs. latency.When weight loss is observed, one should review the previous clinical observations,previous weight, and veterinarian and study director notes to determine whetheranything has been documented that supports the potential for the weight loss. Oneway of correlating weight loss with clinical observations is to conduct the activitiessimultaneously.Check the feed and water supply systems for indications of malfunction.Check food consumption data and feed bowl to evaluate the test system’s eating habits;diet aversion can be a real issue, generally in higher dose groups, but potentially inany group.Check the animal for injury or signs of illness.To assure appropriate measures are initiated, it is critical for the observer to bring allissues to the attention of the study director immediately.Personnel training that includes acceptable standards for data documentation and builtinQC checks can reduce the number of transcription and inputting errors.Use of validated automated data collection systems can significantly reduce data errors,calculation discrepancies, and protocol deviations.© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 729ConcernInaccurate recordingof gestational onset.For rodents —confusion overwhich day the maleshould be removedfrom cohabitation.Misdiagnosed malefertility effects.Test system issues.Inaccurate pup litterweights, pup mixups,anddisproportionateweights for culledlitters.Errors during fetalobservation anddissection.ApproachSometimes a state of “pseudo” pregnancy is provoked as a result of rough handlingduring vaginal lavage. Personnel training and awareness and observation of personnelduring the procedure reduce the incidence of such problems.Because there is a potential for vaginal plugs (sign of successful mating) to be formedwithout a successful mating, other means to confirm mating should be considered.One can conduct a wet reading of the vaginal lavage slide to determine the presenceof sperm, which is especially important if the plug was found in the cage rather thanin situ.Processing vaginal lavage samples for estrous cycle monitoring is one way to managea study. Accurate labeling of slides is essential. Some organizations create a grid slideto capture samples over several days. One way to assure correlation between theestrous cycle dates for gestational onset (as determined from slide samples) and thedate that was recorded in the database or notebook is to clearly mark the grid on theestrous cycle slide. For example, one could mark the slide with a “0” on the day thatthe plug was observed. Note: Verify that the mark will withstand contact with stainingchemicals.When checking cages for plugs in combination with conducting vaginal lavage forestrous cycle data, it is critical to check for plugs prior to conducting the lavage.Otherwise, the estrous cycle slide may be processed and result in a disparity betweenthe time to mating data gathered via slides and the day “0” of gestation date enteredwhen the plug was observed.Clearly define in the SOPs and/or protocol what constitutes the maximum duration ofthe cohabitation period, the criteria for evidence of mating, and when the male rodentshould be removed from cohabitation. For example, the first day of cohabitation usuallybegins in the afternoon. The following morning, each female is examined for evidenceof mating (a plug in situ or a sperm-positive vaginal lavage). When evidence of matingis observed, cohabitation is ended for that pair. Checking continues each morninguntil evidence of mating is observed. If no evidence of mating is evident, thecohabitation is continued (up to the maximum duration designated). For example, ifthe cohabitation period was defined as no more than a 7-d duration, and for sevenconsecutive morning checks no evidence of mating is observed, the male is removedfrom cohabitation.Personnel training and study director involvement in the study will minimize errors.Carefully identify and document the correct size restraint used for animals undergoingnose-only inhalation exposures. A too tight fit in the restraint can cause increasedtemperature, especially near the testes, resulting in hyperthermia with possiblereduced fertility.Artifacts on slides can cause inaccurate sperm count readings. Personnel training inappropriate techniques, maintaining slides in a manner that reduces potential forartifacts, and QC checks of slides combine to improve data quality.When sperm sample slides are not read in a timely fashion, erroneous sperm countsmay result. See that personnel are trained and focused when conducting the study.Temperature and timing are critical issues in sperm motility.If males are used over several matings, review the mating history to assure that themale has successfully mated and is not introducing bias into the study.Confirm that the breeder colony has not gone beyond the age prescribed in the protocol.Define the randomization process and follow it.Check the weigh bucket after weighing litters. Occasionally, pups may adhere to theinside of the weigh bucket. If the pup is not noticed, there is potential for interminglingwith the next weighed litter.If pups were mis-sexed, assure that appropriate editing techniques are used. Sex mustbe changed in all applicable locations, e.g., database, notebooks.Personnel training and recertification, along with the previously discussed approach(whereby trainees observe the same specimen but record their findings separatelyand then reconcile them), is an excellent way to reduce variability across workers.© 2006 by Taylor & Francis Group, LLC

730 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONVI. CONCLUSIONSDevelopmental and reproductive toxicity studies can be fragile. We have described specific approachesto handling known issues that are critical to the success of the study. However, it all comes down tohaving sound scientific practices, which in essence are tantamount to adhering to GLPs.As we mentioned earlier, there is no one way to conduct a successful developmental orreproductive toxicity study. Success has been achieved by following GLPs and by implementingprocedures that facilitate adherence to quality standards. Being open to new techniques and benchmarkingwith other scientists are valuable tools for the developmental and reproductive toxicologist.Pendergast has described the rewards of a quality study as follows: “We promise that sponsorswho produce high quality data with the highest integrity will be rewarded with reduced reviewingtime, quicker approvals, and a reduced threat of private lawsuits. In the end, good quality dataserves the interest of both the industry and the public.” 14The role of the study director is to manage the study in a way that results in a high qualityproduct that can be used in making the right decisions regarding substances with a potential toaffect human health or the environment. One of the questions we have posed to study directors,study personnel, and quality assurance auditors is: “What do you believe are the most significantfactors impacting study quality?” The top three answers:1. Study director involvement in the study2. An effective quality assurance unit3. A quality control processPlease do not confuse “quality assurance” with “quality control.” Quality assurance includesmonitoring to assure GLP compliance. It is not merely a data checking exercise but rather a programthat integrates all facets of GLPs. The QAU plays a resource and advisory role. Part of that roleis to assure that adequate training, facilities, equipment, procedures, protocol, study conduct, andreporting are combined to produce a GLP-compliant study.On the other hand, the role of the quality control reviewer is to assure that data at a given pointin time has been accurately recorded or transcribed according to procedures. When quality controlinspections are conducted at frequent intervals (typically at the end of the day or week), resolutionof errors is more expeditious.Reliable, trustworthy data should result when:Study personnel are knowledgeable and well trained.Study director involvement is adequate.Prestudy design has been given proper emphasis.Study protocol and procedures are followed.Facilities are adequate and well maintained.Equipment is adequately calibrated and maintained, and functions as intended.Quality control reviews are established and followed periodically during the course of the study.ACKNOWLEDGMENTSWe would like to express our appreciation to the following experienced individuals who providedvaluable insight into issues encountered when conducting and monitoring GLP studies: KimberlyB. Brebner, DuPont; Joyce L. Henry, B.S., Dow Corning Corporation; Deborah Little, RQAP-GLP,DLL Quality Consulting, LLC; Susan M. Munley, M.A., Independent Consultant; Eve Mylchreest,Ph.D., DuPont; Robert M. Parker, Ph.D., DABT, DuPont; Barbara J. Patterson, Argus Research;Patricia O’Brien Pomerleau, M.S., RQAP-GLP, CIIT Centers for Health Research; Ronald L. Poore,DuPont; and Deborah L. Tyler, DuPont.© 2006 by Taylor & Francis Group, LLC

QUALITY CONCERNS FOR DEVELOPMENTAL AND REPRODUCTIVE TOXICOLOGISTS 731REFERENCES1. Huber, L., Good Laboratory Practice, Hewlett-Packard Co., 1993.2. FDA GLP Regulations; Final Rule, 21 CFR Part 58, 33768, 1987.3. EPA TSCA GLP Standards, Final Rule, 40 CFR Part 792,34034, 1989.4. EPA FIFRA GLP Standards, Final Rule, 40 CFR Part 160,34052, 1989.5. Keener, S.K. and Hoover, B.K., Good laboratory practices: a comparison of the regulations, J. Am.Coll. Toxicol., 4 (6), 339, 1985.6. EPA Good Automated Laboratory Practices, Office of Information Resources Management, 1995.7. FDA, 21 CFR Part 11, Electronic Records: Electronic Signatures, Final Rule, 62, 54, 13430, 1997.8. EPA Establishment of Electronic Reporting: Electronic Records, Proposed Rule, 40 CFR Parts 3, 51,et al., 66, 170, 46162, 2001.9. FDA, Enforcement Policy: 21 CFR Part 11, Electronic Records: Electronic Signatures (CPG 7153.17),1999.10. Lepore, P., The 10 essential activities of the quality assurance unit, Quality Assurance: Good Practice,Regulation, and Law, 4(1), 29, 1995.11. Greenspan, B.J., Quality considerations in inhalation toxicology, Quality Assurance: Good Practice,Regulation, and Law, 2(1,2), 105, 1993.12. Caulfield, M., Customer focus and satisfaction, Quality Assurance: Good Practice, Regulations, andLaw, 3(1), 15, 1994.13. Usher, R.W., Developing an effective toxicology/QA partnership, Quality Assurance: Good Practice,Regulations, and Law, 4(4), 308, 1995.14. Pendergast, M. K., U. S. FDA. Integrity, accuracy, and quality of scientific information, QualityAssurance: Good Practice, Regulation, and Law, 2(½), 57, 1993.© 2006 by Taylor & Francis Group, LLC

CHAPTER 19Perspectives on the Developmental andReproductive Toxicity GuidelinesMildred S. Christian, Alan M. Hoberman, and Elise M. LewisCONTENTSI. Introduction ........................................................................................................................734II. Regulatory History.............................................................................................................734III. Roles of Various Regulatory Groups and Evolution of Testing Concerns .......................738A. IRLG Guidelines........................................................................................................738B. ICH Guidelines ..........................................................................................................738C. OECD Guidelines ......................................................................................................739D. EPA Guidelines..........................................................................................................739IV. Putting the Guidelines into the Final Protocol Design .....................................................740A. Fertility and General Reproductive Performance .....................................................740B. Embryo and Fetal Development(Developmental Toxicology or Teratology Studies) .................................................741C. Perinatal and Postnatal Evaluations ..........................................................................742D. ICH Stages.................................................................................................................7431. ICH Stage A — Premating to Conception..........................................................7432. ICH Stage B — Conception to Implantation......................................................7433. ICH Stage C — Implantation to Closure of the Hard Palate.............................7444. ICH Stage D — Closure of the Hard Palate to the End of Pregnancy..............7445. ICH Stage E — Birth to Weaning ......................................................................7446. ICH Stage F — Weaning to Sexual Maturity.....................................................744V. Differences in the Guidelines for Testing Pharmaceutical Products and the Effectof the ICH Harmonization Process....................................................................................744A. Compliance with Good Laboratory Practice (GLP) Regulations.............................744B. Species Selection .......................................................................................................744C. Numbers of Dosage Groups and Number of Animals per Dosage Group ..............745D. Dosage Selection .......................................................................................................745E. Timing of Gestation and Postnatal Development .....................................................7461. Exposure Intervals ...............................................................................................746F. Reproductive Performance ........................................................................................7471. Evaluation of the Ovaries....................................................................................752733© 2006 by Taylor & Francis Group, LLC

734 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONG. Maternal and Pup Interactions and Lactation...........................................................752H. Growth and Development of Conceptus through Puberty........................................7521. Fetal Evaluations .................................................................................................7522. Postnatal Evaluations for Gender and Sexual Maturation..................................752I. Endocrine-Mediated Findings ...................................................................................753J. Comparison of FDA and EPA Requirements and Methods for PostnatalBehavioral and Functional Evaluations.....................................................................754K. Developmental Neurotoxicity....................................................................................754References ......................................................................................................................................754Appendix 1.....................................................................................................................................759Appendix 2.....................................................................................................................................779Appendix 3.....................................................................................................................................792I. INTRODUCTIONThe purpose of this chapter is to describe the various regulatory guidelines specifically designedto address safety concerns regarding reproduction and development and to identify commonalitiesand differences in these guidelines. Regulatory guidelines issued for conduct of developmental andreproductive toxicology studies are essentially categorized into three sectors: pharmaceuticals, foodand food additives, and chemicals. Because the guidelines for development of pharmaceuticals areflexible, The Guidelines for Toxicity to Reproduction of Medicinal Products, an effort of theInternational Conference for Harmonization (ICH), 1 addresses all elements of reproductive anddevelopmental toxicology studies. The ICH guidelines are accepted by the U.S. Food and DrugAdministration (FDA), 2–4 the European Union (EU), and Japan. The ICH guidelines also providean overview of the various study designs and endpoints evaluated in the regulatory testing ofpharmaceuticals, biotechnology products, 5 indirect food additives and devices, 6 chemicals, 7–10 andpesticides, fungicides, and rodenticides. 7,11–13As the harmonization processes continue, it is likely that the currently available guidelines willbecome more, rather than less, similar. For instance, the ICH guidelines 2–4 replace the formerguidelines issued by the U.S. FDA Bureau of Drugs, 5 EU countries, 14,15 and Japan. 16–18 Those issuedby the U.S. Environmental Protection Agency (EPA) Office of Prevention, Pesticides and ToxicSubstances (EPA-OPPTS) replaced the U.S. EPA Toxic Substances Control Act (TSCA)guidelines 7–10 and the Federal Insecticide, Fungicide and Rodenticide (FIFRA) guidelines. 7,11–13International harmonization efforts are also ongoing between the EPA and the Organization forEconomic and Cooperative Development (OECD).Table 19.1 lists the principal regulatory guidelines currently in use. Additional information canalso be obtained from various websites, identified in Table 19.2 and Table 19.3, many of whichprovide updates on regulatory issues, information on professional society activities and projects,and access to historical databases.II. REGULATORY HISTORYThe reproduction and developmental toxicity regulatory guidelines were developed in response tohuman health concerns. Three human tragedies that resulted from in utero exposure to a drugprovided the momentum for initial development of the guidelines. These were: (1) 1961 —congenital malformations associated with the use of thalidomide; 32,33 (2) early 1970s — cancerassociated with maternal use of diethylstilbesterol; 34 and (3) 1976 — behavioral and functionalalterations associated with exposure to methyl mercury. 35 For a detailed history of the developmentof the guidelines associated with these issues, see Christian and Hoberman 36 in the previous editionof this book.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 735Table 19.1Current regulatory guidelines — ICH, FDA, OECD, and EPARegulatory Agency Ref. GuidelineMedical AgentsICH 1 Detection of toxicity to reproduction for medicinal productsICH 19 Detection of toxicity to reproduction for medicinal productsFDA 2 International Conference on Harmonisation: guideline on detection of toxicity toreproduction for medicinal productsFDA 3 International Conference on Harmonisation: guideline on detection of toxicity toreproduction for medicinal products: Addendum on toxicity to male fertilityFDA 4 International Conference on Harmonisation: maintenance of the ICH guidelineon toxicity to male fertility; An addendum to the ICH tripartite guideline ondetection of toxicity to reproduction for medicinal products, amended November9, 2000Indirect Food AdditivesFDA, Bureau ofFoodsFDA, Center for FoodSafety and AppliedNutritionFDA, Center for FoodSafety and AppliedNutrition20 Toxicological principles and procedures for priority based assessment of foodadditives (Red Book), guidelines for reproduction testing with a teratology phase6 Toxicological principles for the safety assessment of direct food additives andcolor additives used in food (Red Book II), guidelines for reproduction anddevelopmental toxicity studies21 Toxicological principles for the safety of food ingredients (Red Book 2000)ChemicalsEPA, OPPTS 22 Health effects test guidelines: prenatal developmental toxicity studyEPA, OPPTS 23 Health effects test guidelines: reproduction and fertility effectsEPA, OPPTS 24 Health effects test guidelines: developmental neurotoxicity studyEPA, Risk25 Guidelines for developmental toxicity risk assessmentAssessment ForumEPA 26 Guidelines for developmental toxicity risk assessmentOECD 27 Guidelines for testing of chemicals: teratogenicityOECD 28 Guidelines for testing of chemicals: one-generation reproduction toxicityOECD 29 Guidelines for testing of chemicals: two-generation reproduction toxicity studyOECD 30 Guidelines for testing of chemicals: combined repeated dose toxicity study withthe reproduction and developmental toxicity screening testOECD 31 Guidelines for testing of chemicals: developmental neurotoxicity studyTable 19.2 List of useful websites for obtaining regulatory documentsFederal Registerhttp://www.access.gpo.gov/su_docs/aces/aces140.htmlFDA guidance documentshttp://www.fda.gov/cder/guidance/index.htmICH guidelineshttp://www.ifpma.org/ich5s.htmlRed Book 2000http://vm.cfsan.fda.gov/~redbook/red-toct.htmlEPA (public drafts)http://www.epa.gov/OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/DraftsEPA (final guidelines)http://www.epa.gov/docs/OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/Series/EPA guidelines for reproductive toxicity http://www.epa.gov/ordntrnt/ORD/WebPubs/repro/index.htmlrisk assessmentEPA’s EDSTAChttp://www.epa.gov/scipoly/oscpendo/index.htmEPA Food Quality Protection Act of 1996 http://www.epa.gov/opppsps1/fqpaOECD guidelineshttp://www.oecd.org/ehs/test/testlist.htmOECD endocrine disrupter testing andassessmenthttp://www.oecd.org/EN/document/0,,EN-document-524-14-no-24-6685-0,00.htmlU.S. National Toxicology Program (NTP) http://ntp-server.niehs.nih.gov/main_pages/NTP_ALL_STDY_PG.htmltesting information and study resultsCalifornia Proposition 65 documents http://www.oehha.org/prop65.html© 2006 by Taylor & Francis Group, LLC

736 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 19.3List of useful research organizationsReproductive and Developmental ToxicologyTeratology SocietyMiddle Atlantic Reproduction andTeratology Association (MARTA)Midwest Teratology Association(MTA)Society of Toxicology Reproductiveand Developmental SpecialtySectionAustralian Teratology Society (ATS)European Teratology Society (ETS)Japanese Teratology Societyhttp://www.teratology.org/http://www.teratology.org/members/MARTAtext.htmhttp://www.midwest-teratology.org/http://www.toxicology.org/memberservices/specsection/specsection.htmlhttp://www.cchs.usyd.edu.au/bio/ats/atshome.htmhttp://www.etsoc.com/http://www.med.hiroshima-u.ac.jp/med/med/kiso/kaibo1/CongAnom/eWelcome.htmlToxicologySociety of Toxicology (SOT)American College of Toxicology(ACT)Mid-Atlantic Chapter of Society ofToxicology (MASOT)Italian Society of ToxicologyBritish Toxicology SocietyEurotoxAmerican Board of Toxicology (ABT)Academy of Toxicological Sciences(ATS)Society of Toxicology of Canadahttp://www.toxicology.org/http://www.landaus.com/toxicology/http://www.masot.org/http://users.unimi.it/~spharm/sit/engSIThome.htmlhttp://www.thebts.org/http://www.eurotox.com/http://www.abtox.org/http://www.acadtox.com/http://meds.queensu.ca/stcweb/index.htmlNeurotoxicologyNeurobehavioral Teratology Society(NBTS)Behavioral Toxicology SocietyThe Society for Neuroscience (SFN)The International NeurotoxicologyAssociationhttp://nbts.bsbe.umn.edu/http://www.behavioraltoxicology.org/BTS.htmlhttp://www.sfn.orghttp://www.neurotoxicology.org/Additional ResourcesReproductive Toxicology($850/year online)International Federation ofTeratology Societies (IFTS) Atlasof Developmental AbnormalitiesNational Institute of EnvironmentalHealth Sciences (NIEHS)Teratogen Information System(TERIS, $1000/year)http://www.elsevier.nl/locate/reprotoxhttp://www.ifts-atlas.org/http://www.niehs.nih.gov/http://depts.washington.edu/~terisweb/Recently, new guidelines have been developed to address specific political and economicconcerns, particularly regarding potential hazards to children. These guidelines are related toconcerns regarding treatment of neonatal and pediatric patients with pharmaceuticals and chronicand inadvertent exposure to food additives and other environmental chemicals, particularly late ingestation and/or before puberty, as described in the following information.The current FDA and ICH guidelines 2 for pharmaceuticals include postnatal evaluations inanimals for behavioral and functional alterations, in particular those concerning potential exposure© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 737of the central nervous system and other systems that continue to develop postnatally (e.g., pulmonary,immune, renal, osseous, and gastrointestinal) and thus may be potentially affected during lategestation and the early neonatal period. The FDA currently has a pediatric study requirement (1998Final Rule: Pediatric Studies). 37 To conform with this requirement, all new drugs, generally prescriptiondrugs (including biologics or drugs derived from living organisms), that are intended forcommon use in children or are important in medical treatment of children must include labelinginformation on safe pediatric use. Similar efforts are ongoing in the EU. 15More extensive increases in requirements for pediatric testing have been enacted or are underconsideration by the EPA and the OECD. In 1997, the EPA issued entirely new guidelines inresponse to children’s health concerns and established the Office of Children’s Health Protection(OCHP). 38 This federal agency identified four major environmental threats to children: (1) leadpoisoning; (2) pesticides in household chemicals and food, (3) asthma, and (4) drinking watercontaminants (principally microbial), polluted water, toxic waste dumps, PCBs, second-handtobacco smoke, UV light, and endocrine disruptors, and potential effects from particulate matterair pollution. These relatively new EPA guidelines are based on the perceived greater relativevulnerability of children to environmental threats. This perception is based on children havingimmature body organs and tissues, rapidly changing as they develop from infants to adolescents.They have weaker immune systems during infancy and are potentially exposed to more environmentalthreats while being least able to protect themselves from such hazards. One response tothese concerns was the issue of the EPA guidelines for developmental neurotoxicity. 24Another important response of the EPA is its endocrine disruption program, which was mandatedby the Food Quality Protection Act 39 and authorized by the amended Safe Water Act. 40 These twoacts resulted in formation of the Endocrine Disruptor Screening and Testing Advisory Committee(EDSTAC), which has completed its mandate. EDSTAC’s purpose was to develop recommendationsfor an endocrine disruptor screening program. 41 One of EDSTAC’s recommendations was expansionof the endocrine disruptor screening program to include not only pesticides but also chemicalsregulated under the TSCA, nutritional supplements, food additives, and cosmetics. EDSTAC alsorecommended that screens be performed to evaluate endocrine-mediated effects on the androgenand thyroid systems and other effects on fish and wildlife. 42,43The EDSTAC recommendations were used to design the EPA Endocrine Disruptor ScreeningProgram (EDSP). The Endocrine Disruptor Method Validation Subcommittee (EDMVS), which isunder a Federal Advisory Committee — the National Advisory Council for Environmental Policyand Technology (NACEPT) 44 — is one essential part of implementation of the EDSP. The EDMVSis responsible for evaluating the methods and the results of validation efforts for the new protocolsrecommended by EDSTAC. New protocols currently being considered for validation or in thevalidation process include a one-generation extension study 45 and a 20-day pubertal developmentand thyroid function study in juvenile male and female rats. 46,47Internationally, an OECD Task Force on Endocrine Disruptor Testing and Assessment (EDTA)was instituted in December, 1997 in an effort to: (1) consider and recommend priorities for thedevelopment of methods to detect estrogenic and androgenic effects and (2) harmonize test methodsand assessments of endocrine disruptors. Under the direction of the EDTA, OECD ValidationManagement groups were established to manage the technical conduct of validation work forscreening and testing for endocrine-mediated mammalian and ecotoxicological effects. The validationtesting phase has been completed for the rodent uterotrophic assay. 48 Revisions of the OECDTwo Generation Reproductive Toxicity 29 and Teratogenicity 27 guidelines have been issued as OECDTest Guidelines 416 and 414, respectively. Additional projects currently being addressed by theOECD include validation of the rodent Hershberger assay 48 and of studies to assess the feasibilityof enhancing the current OECD TG 407 (Repeated Dose Toxicity) guideline. 49 The EPA and OECDare working in close cooperation regarding the validation programs.© 2006 by Taylor & Francis Group, LLC

738 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIn toto, the concepts regarding the ability of agents to affect in utero and early postnataldevelopment, including sexual maturation and function, have changed remarkably since 1966, whenthe FDA issued its first guidelines. 5 In 1966, the focus was on identifying “teratogens.” Thus theconcerns and ways in which multiple endpoints, ranging from delayed or advanced sexual maturationto frank malformation, are to be addressed have evolved. As a result, it is now recognizedthat it is important to screen for adverse effects on the entire reproductive process as well as forin utero development and maturation.III. ROLES OF VARIOUS REGULATORY GROUPSAND EVOLUTION OF TESTING CONCERNSThe pharmaceutical and chemical industries are extensively regulated internationally. Remarkableinternational and local concern exists for both industries regarding regulatory requirements forreproductive and developmental toxicity testing. As a result, harmonization of guidelines and everincreasing appreciation of the multiplicity of the endpoints to be evaluated in reproductive anddevelopmental toxicity studies has been in progress for the last 20 years. In general, currentguidelines for use in the development of medicines for humans are the ICH guidelines. 2–4 FDAguidelines relevant to development of medicines for animals, foods, and food additives (“FDA RedBook”) are moving in the direction of the ICH guidelines, but have some differences. OECD andEPA guidelines relevant to the testing of chemicals are less flexible and often have different technicalrequirements, although the same endpoints are evaluated. Some information regarding the variousorganizations involved in guideline development and harmonization are provided below. Copies ofthe relevant ICH and EPA guidelines (with the exception of the EPA developmental neurotoxicityguideline, addressed in another chapter) are provided in Appendixes 1 and 2 of this chapter.A. IRLG GuidelinesThe Interagency Regulatory Liaison Group (IRLG), which is no longer active, was composed offive U.S. regulatory agencies: (1) Consumer Product Safety Commission; (2) Environmental ProtectionAgency; (3) Food and Drug Administration (U.S. Department of Human Health and Safety);(4) Food Safety and Quality Service (U.S. Department of Agriculture), and (5) the OccupationalSafety and Health Administration (U.S. Department of Labor). In 1980, the IRLG issued harmonizedguidelines for the conduct of developmental toxicity studies that were acceptable in these U.S.agencies. 50 Those guidelines became the basis for the early international OECD guidelines.B. ICH GuidelinesThe European Economic Community (EEC) guideline update 15 provided the basis for the ICHprocess. Beginning in 1990, 51,52 the International Federation of Pharmaceutical ManufacturersAssociations (IFPMA) began to bring together the regulatory authorities of Europe, Japan, and theUnited States, as well as experts from the pharmaceutical industry in these three major geographicalregions. Their task was to identify ways to eliminate redundant technical requirements. The objectivewas to expedite global development and availability of new medicines without sacrificing safeguardson quality, safety, or efficacy; the process continues to this day.The ICH effort had six cosponsors: the Commission of the European Communities (CEC), theEuropean Federation of Pharmaceutical Industries Associations (EFPIA), the Ministry of Healthand Welfare (MHW, Japan), the Japan Pharmaceutical Manufacturers Association (JPMA), theFDA, and the Pharmaceutical Manufacturers Association (PMA, United States), which has sincebecome PhRMA (Pharmaceutical Research and Manufacturers of America). The result of theseefforts was the international acceptance of the regulatory requirements of the ICH guidelines.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 739The ICH guideline on Detection of Toxicity to Reproduction for Medicinal Products was thefirst to be approved. 2 Since 1994, it has been amended twice 3,4 to clarify testing procedures for themale fertility portion of the studies. A history of the regulatory activities and the ICH programregarding these guidelines can be found in the forerunner to this book. 36C. OECD GuidelinesThe Organisation for European Economic Co-operation (OEEC) was formed to administer Americanand Canadian aid under the Marshall Plan for the reconstruction of Europe following WorldWar II. At that time it consisted of its European founder countries plus the United States andCanada. The OECD took over from the OEEC in 1961. The OECD currently includes 30 membercountries in Europe, North America, Asia, and the Pacific.In 1979 to 1980, OECD assisted in updating and standardizing testing procedures for chemicals.53 More recently the OECD Pesticide Working Group, part of OECD’s Environment Program,was established as the first forum for national pesticide regulators from developed countries todiscuss common issues. The OECD Pesticide Working Group was established in 1994 as thePesticide Forum. 54 The purpose of the Pesticide Working Group is to assist countries in managingtheir task of conducting new risk assessments for hundreds of established pesticides, as well asassessments of the active ingredients in newer pesticides. The major activity areas of the PesticideWorking Group include conventional, biological, and antimicrobial pesticides. 54D. EPA GuidelinesThe EPA regulates pesticides and other chemicals for reproductive and developmental hazard andrisk under multiple federal statutes, the most important of which are the Federal Insecticide,Fungicide, and Rodenticide Act 55 and the Federal Food, Drug, and Cosmetic Act (FFDCA). 56 Underthe authority of FIFRA, the EPA reviews and registers pesticides for specific uses. It can alsosuspend or cancel pesticide registration if it deems the pesticide to be hazardous. Section 408 ofthe Federal Food, Drug, and Cosmetic Act 56 allows the EPA to establish maximum residue levels,or tolerances, for pesticides used in or on foods consumed by humans or animals. In 1996, theFood Quality Protection Act (FQPA) was enacted, amending FIFRA and FFDCA. The FQPAestablished higher safety standards for all pesticides used on foods, required uniformity in foodprocessing, and changed pesticide regulations by requiring greater health and environmental protectionfor infants and children. 57 The Office of Pollution Prevention and Toxics (OPPT), formedin 1977, 58 has the primary responsibility for administering the Toxic Substances Control Act. 59 Thislaw covers production and distribution of commercial and industrial chemicals produced by orimported into the United States.The EPA OPPTS has developed uniform guidelines for testing pesticides and other substancesto meet the EPA’s data requirements. 60 These data requirements are regulated by two laws, theToxic Substances Control Act 59 and the Federal Insecticide, Fungicide, and Rodenticide Act. 55Before the harmonization process, slightly different testing procedures and requirements existed inseparate publications by the OPPT, the Office of Pesticide Programs (OPP), and the OECD. 60The EPA works closely with the OECD to develop and carry out OECD Pesticide Programactivities. OPP is involved in cooperative work to harmonize pesticide data requirements, focustest guidelines on pesticide regulatory needs, and harmonize industry data submissions and governmentdata review formats and contents. OECD encourages the exchange of review reports andhas developed an electronic database to facilitate such exchanges and collaboration on reviews.OPP is also working with OECD countries on electronic data submission for industry. As the resultof international agreements, the OECD guidelines may be applied to EPA-regulated chemicals indevelopment.© 2006 by Taylor & Francis Group, LLC

740 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIV. PUTTING THE GUIDELINES INTO THE FINAL PROTOCOL DESIGNAs previously mentioned, a copy of the ICH guideline and amendments 2–4 are provided as Appendix1 of this chapter. References for guidelines and other information generally considered in thedevelopment of drugs, food additives, pesticides, and other chemicals are provided as Appendix 3.To understand how to appropriately select and combine the various “ICH Stages” of the reproductivecycle, 2 it is necessary to understand the endpoints evaluated and some of the technical concernsregarding conduct of these studies. The following information cites the relevant endpoints and howthey are addressed in the various guidelines currently in use.A. Fertility and General Reproductive PerformanceThe evaluation of fertility and general reproductive performance includes examination of potentialadverse effects on gonadal function and mating behavior in male and female animals, conceptionrates, early and late stages of gestation, parturition, lactation, and development of the offspring.Thus, the study design is an abbreviated multigeneration study, usually ending at weaning of thesecond generation.For a full evaluation of male fertility and reproductive performance (i.e., a multigenerationstudy), the agent is given to sexually mature males for 60 to 70 days before mating, i.e., a completecycle of spermatogenesis, and then continues until mating occurs. If there is no evidence of impairedmale reproductive performance from other studies, males may be treated for 28-day or 14-dayintervals before mating with cohort females. Although not mandated, evaluation of the histopathologyof the testes and epididymides is generally performed, as well as evaluation of sperm for count,concentration, and morphology.To study female reproductive performance and fertility, sexually mature females are evaluatedfor estrous cycling for 14 days before treatment, for the first 14 days of treatment before cohabitation,and until mating occurs. Treatment continues through cohabitation and mating, and in thefemale animals until they are necropsied on day 13 of gestation or at completion of a 21-daylactation period. Parturition, lactation, maternal-pup interaction, and pup growth and developmentuntil weaning are monitored.Adult reproductive function may be evaluated by treating animals of only one sex and matingthese animals with untreated animals of the opposite sex, or by treatment of both males and females.A common refinement of the test is to use double sized control and high dose groups during thepremating period, with cross-mating of one-half of the animals in these two groups. This is doneto determine whether administration of the test agent to animals of only one sex causes differenteffects from those observed when both sexes are treated. 61Alternate procedures include:1. Extension of the treatment period through the remainder of gestation (e.g., for rats, until day 17or 19 of gestation, with caesarean-sectioning on day 20 or 21 of gestation), and evaluation of thefetuses for viability and gross (external), soft tissue, and skeletal alterations — a design addressingboth male and female fertility and reproductive performance and developmental toxicity.2. Extension of the treatment period through the remainder of gestation and then through parturitionand lactation (i.e., for rats until day 21 of lactation), and assessing the pups for developmentallandmarks and postnatal behavior and function (pups are selected at weaning for continuedevaluation through mating and gestation and examined for general sensory [auditory, visual] andbehavioral development and fertility. Additional functional evaluations are sometimes consideredfor inclusion 62,63 ).Canada is not one of the countries participating directly in the ICH process; however, it isusually considered in international marketing plans. The proposed guidelines for Canada 64,65 differ© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 741from the ICH guidelines 2 on the following points: (1) an 80-day dosage period before mating issuggested for male rodents, and (2) it was suggested that consideration be given to use of the UnitedKingdom (UK) protocol, 14 which is a design in which treatment of rats of both sexes continuesfrom before cohabitation until sacrifice after mating (males) or completion of a 21-day lactationperiod (females) and postnatal evaluation of the offspring through gestation. Because the UK ispart of ICH, Canada now generally accepts ICH guideline compliant studies.The guidelines in the FDA Red Book 21 require both “teratology” and multigeneration studies.However, the teratology study may have an exposure period extending from before mating of theparental generation through gestation and can be conducted separately or as part of the multigenerationstudy. The EPA 23 and OECD 28,29 guidelines also require multigeneration studies.B. Embryo and Fetal Development(Developmental Toxicology or Teratology Studies)The study design formerly identified as a “Segment II” or FDA “teratology” study is currentlygenerally described as a “developmental toxicity study,” i.e., a study of potential embryo and fetaltoxicity or embryotoxicity. Developmental toxicity studies are usually conducted in two species,one rodent and one nonrodent. Two species are tested because: (1) adverse effects on development,especially in the absence of maternal effects, remain the most sensitive regulatory concern; and (2)thalidomide did not cause phocomelia in the rat, although it can produce phocomelia in the rabbit. 66Despite Wilson 67 noting that one cannot assume mimicry of responses across species, and Palmer 68identifying that appropriate interpretation of rat data from a two-litter test would have identifiedthat thalidomide was selectively toxic to the development of the conceptuses (i.e., it increasedresorption and reduced live litter sizes at dosages that were not toxic to the dams), the rabbitcontinues to be the nonrodent species most commonly tested in this assay. However, rats and rabbitsshould not automatically be selected as the two most appropriate species. ICH guidelines emphasizethat the agent should have pharmacologic activity in the species tested. The increased considerationof the metabolism of test agents has resulted in the animal species whose metabolism of an agentis most like that of humans generally being considered the most appropriate species. Concernsabout pharmacological metabolic comparability and the ability to justify species choice haveincreased the use of nonhuman primate testing for biologics and the development of homologuesof human proteins for testing of biologics in rabbits, rats, or mice.Despite increased concern and appreciation of other endpoints, the developmental toxicity studycontinues to be commonly perceived as an evaluation of the potential of an agent to malform theconceptus. However, this study is actually an evaluation of embryo and fetal toxicity and differsfrom a true teratology study, in which a “pulse” exposure is typically the most effective treatmentmethod. 69 In the developmental toxicity study design, pregnant females are treated throughout theperiod of major organogenesis (embryogenesis), generally defined as the interval between implantationand palate closure, and necropsied shortly before expected parturition. The ovaries areexamined for the number of corpora lutea. The uterus is examined for implantation sites, live anddead fetuses, early and late resorptions, and placental abnormalities. The fetuses are weighed, thefetal sex is identified, and if indicated, anogenital distances are determined. Occasionally, crownrumplengths for the fetuses are recorded. In rodent studies, approximately one-half of the fetusesin each litter are examined for soft tissue alterations by using either Wilson’s sectioning technique 70or Staples’ visceral examination technique 71 (fully described in Christian, 72 possibly as adapted byStuckhardt and Poppe 73 ). The remaining fetuses are stained with alizarin red-S, 74 and if desired,also with alcian blue, 75–77 and examined for skeletal alterations. In nonrodent studies, all fetusesare examined for soft tissue and skeletal alterations.Day 0 of presumed gestation is generally recognized as the day spermatozoa, a vaginal or copulatoryplug (rodents) or insemination (rabbit) occurs. The period of organogenesis or embryogenesis is© 2006 by Taylor & Francis Group, LLC

742 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONidentified in the ICH guidelines 2 as occurring between implantation and palate closure. This periodwas extended from that in the FDA 1966 guidelines to include the fetal intervals when sexualdifferentiation and initial renal development occur, i.e., gestation days (GD) 15 through 18 or 19in rodents. The fetuses are caesarean-delivered one or two days before parturition is expected tooccur, i.e., usually on GD 18, 21, and 29 in mice, rats and rabbits, respectively.The ICH guidelines 2 recommend that there be 16 to 20 animals per dosage group, with a controland at least three dosage groups administered the test material. Generally, 20 rodents and 16nonrodents per dosage group have been considered appropriately sized populations for testing. TheFDA Red Book, 21 OECD, 27 and EPA guidelines, 22 however, recommend 20 pregnant animals pergroup, regardless of species. A frequent modification of this design is to use double-sized groups,with one-half of the pregnant animals caesarean-sectioned and the remaining pregnant animalsallowed to deliver and to nurse their litters until weaned. The pups may be examined for postnatalmaturation and physical development, including mating of the F 1 generation as adults and examinationof uterine contents of the female rats, using the same methods described for the firstgeneration animals.The FDA Red Book 21 allows conduct of a teratology study in rodents either within the multigenerationreproduction study, in which case exposure is continual in the parental generation beforemating and in the dams throughout gestation, or as a separate study. These guidelines also requireconduct of a rabbit study. FDA Red Book, 21 OECD, 27 and EPA 22 requirements differ from the ICH 2guidelines in terms of the numbers of animals per dosage group, as previously described, and alsoin the duration of exposure. These guidelines now require treatment to continue from implantation(i.e., GD 6 or 7) until the day before parturition is expected, i.e., through GD 17, 20, and 28 inmice, rats and rabbits, respectively. The EPA guidelines consider it appropriate to treat throughoutthe entire gestation period, although studies performed with treatment initiated at implantation aregenerally acceptable. The exception is when the percutaneous route is used, for which preimplantationexposure is often recommended so that appropriate blood levels can be achieved beforeembryogenesis begins. The EPA guidelines also recommend double-staining of fetuses. In addition,while the OECD 30 and EPA-OPPTS 78 guidelines require testing of two species, the two speciesneed not necessarily be a rodent and nonrodent species (i.e., two rodent species may be tested,when necessary). The FDA Red Book, 21 OECD, 27 and EPA 22 guidelines request uterine and maternalcarcass weights. On all other points, the EPA-FIFRA guidelines 11 allow performance of essentiallythe same evaluations as those recommended by ICH. 2C. Perinatal and Postnatal EvaluationsThe FDA Segment III study, 5 or perinatal and postnatal evaluation, was designed to evaluate theeffects of an agent on the late (fetal) stages of gestation and on parturition and lactation. This studydesign may still be used if it is warranted by the toxicity of an agent or a specific finding of interest,provided both embryo, fetal, and pup exposures are evaluated for postnatal effects. However, theperi-postnatal study is generally performed with treatment of the pregnant female rats beginningon GD 7 and continuing through gestation, parturition, and a 21-day lactation period, combiningembryo, fetal, and peri-postnatal exposures in one study. It is recognized that postnatal changesmay be induced by insult at any time during gestation. Thus, unless the developmental toxicitystudy includes a postnatal phase, potential adverse postnatal findings associated with embryoexposure could not be identified. Parturition and pup survival and development are monitored,including observations for prolonged gestation, dystocia, impaired maternal nesting or nursingbehavior, and postnatal mortality. If indicated by the results of this study or other studies, a crossfosteringprocedure may be included to demonstrate whether findings in the pups are direct effectsof exposure during gestation and/or lactation or if they are associated with abnormalities in maternalreproductive behavior or health. The peri-postnatal evaluation is not required by the FDA RedBook, OECD, or EPA guidelines. Rather, these evaluations are part of the multigeneration studies© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 743performed to meet the needs of these regulatory bodies that consider this design an appropriateoverall evaluation of reproductive capacity. 79,80The ICH guideline 2 acknowledges the overlap in the methods used to test chemicals andpharmaceuticals for adverse effects on the reproductive process, but recommends that segmentedstudy designs generally be used because humans usually take a pharmaceutical product for limitedintervals, rather than an entire lifetime. However, the ICH document 2 also notes that lifetimeexposure to a drug sometimes occurs, and that there may be drugs that can be more appropriatelytested by exposure throughout the entire reproductive life, i.e., through use of a multigenerationstudy design.Rather than a standard checklist approach, the ICH guideline 2 emphasizes flexibility in developingthe testing strategy. Although the most probable options are identified in the guideline, thedevelopment of the testing strategy is to be based on the following points:• Anticipated drug use, especially in relation to reproduction• The form of the substance and route(s) of administration intended for humans• Making use of any existing data on toxicity, pharmacodynamics, kinetics, and similarity to othercompounds in structure and activityThe notes in the ICH guideline 2 describe some methods appropriate for use in specific portionsof the tests; however, it emphasizes that the individual investigator is at liberty to identify the testsused and that the ICH guideline 2 is a guideline and not a set of rules. The overall aim of thedescribed reproductive toxicity studies is to identify any effect of an active substance on mammalianreproduction, with subsequent comparison of this effect with all other pharmacologic and toxicologicdata. The objective is to determine whether human risk of effects on the reproductive processis the same as, increased, or reduced, in comparison with the risks of other toxic effects of theagent. The document clearly states that for extrapolation of results of the animal studies, otherpertinent information should be used, including human exposure considerations, comparative kinetics,and the mechanism of the toxic effect.Because the objective of these tests is to assess all stages of reproduction, the total exposureperiod includes mature adults and all stages of development of the offspring, from conception toweaning. Observations should be made over one complete life cycle (from conception in the firstgeneration through conception and pregnancy in the second generation). The reproductive cycle issegmented into six ICH stages (ICH Stages A through F). To assure that there are no gaps inexposure when ICH stages are tested separately or in combination, it was suggested that there bea one-day overlap of treatment. The six ICH stages are described below.D. ICH Stages1. ICH Stage A — Premating to ConceptionThis ICH stage evaluates reproductive functions in adult males and females, including developmentand maturation of gametes, mating behavior, and fertilization. By convention, it was agreed thatpregnancy would be timed on the basis of when spermatozoa are identified in a vaginal smear ora copulatory plug is observed, and that these events would identify GD 0, even should mating occurovernight.2. ICH Stage B — Conception to ImplantationThis ICH stage examines reproductive functions in the adult female and preimplantation andimplantation stages of the conceptus. Unless data are provided that prove otherwise, it is assumedthat implantation occurs in rats, mice, and rabbits on days 6 to 7 of pregnancy.© 2006 by Taylor & Francis Group, LLC

744 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION3. ICH Stage C — Implantation to Closure of the Hard PalateThis ICH stage evaluates adult female reproductive functions, embryonic development, and major organformation. The period of embryogenesis is completed at closure of the hard palate, which by conventionis considered to occur on days 15 through 18 of pregnancy in rats, mice, and hamsters, respectively.4. ICH Stage D — Closure of the Hard Palate to the End of PregnancyAdult female reproductive function continues to be examined in this ICH stage, as well as fetaldevelopment and growth and organ development and growth.5. ICH Stage E — Birth to WeaningAgain adult female reproductive function is examined; this ICH stage also evaluates adaptation ofthe neonate to extrauterine life, including preweaning development and growth of the neonate. Byconvention, the day of birth is considered postnatal day 0, unless specified otherwise. It is alsonoted that when there are delays or accelerations of pregnancy, postnatal age may be optimallybased on postcoital age.6. ICH Stage F — Weaning to Sexual MaturityThis ICH stage is generally treated only when the intended use of the pharmaceutical product isin children. The ICH guideline 2 recognizes that it may sometimes be appropriate to also treat theneonate during this ICH stage, in addition to the required evaluations of the offspring. This ICHstage provides observations of postweaning development and growth, adaptation to independentlife, and attainment of full sexual development.V. DIFFERENCES IN THE GUIDELINES FOR TESTING PHARMACEUTICALPRODUCTS AND THE EFFECT OF THE ICH HARMONIZATION PROCESSA. Compliance with Good Laboratory Practice (GLP) Regulations 81–85In general, all studies submitted in support of the safety of a pharmaceutical or chemical productshould be conducted in conformance with the GLPs of one or more regulatory organizations.Although preliminary dosage range-finding studies need not be conducted in complete conformancewith GLPs, the results of such studies should be submitted.B. Species SelectionThe various guidelines generally identify testing in a rodent and nonrodent species, usually the ratand rabbit. To avoid additional studies to the extent possible, the ICH guideline 2 recommends useof the same species and strain of animals tested in other toxicology studies and in kinetic studies.It is also recommended that kinetic, pharmacological, and toxicological data be considered todemonstrate that the selected species provides a model relevant for the human. As mentioned earlier,the OECD 30 and EPA-OPPTS 78 combined guidelines allow use of two rodent species.Animals should have comparable ages, weights, and parity when assigned to the study. Historicalincidences of abnormalities in fertility, fecundity, fetal morphology, and resorption, and informationregarding consistency from study to study should be available. For EPA studies, positive controldata are often requested to ensure that the laboratory has appropriate experience in the conduct of© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 745the studies. In addition, for EPA regulated studies, it is preferred that historical data be restrictedto the 2-year period during which the study is conducted. For pharmaceuticals the same speciesand strain should be used to the extent possible in all toxicologic and kinetic studies. Although thisguideline is not consistently followed in EPA regulated studies, this would assist in producingmeaningful data.Because of the increased interest in identifying specific genes that may be affected by an agent,transgenic models and knock-out models of certain disease states are available. There are timeswhen it may be appropriate to test an agent in an animal expressing the disease state to be treated;such tests would usually supplement studies in outbred laboratory species. For example, testing anantidiabetic agent in a normal animal may result in unexpected and irrelevant findings for humandiabetics, as the result of abnormally lowering blood sugar. Testing the same agent in a diabeticanimal has the potential to demonstrate whether any developmental abnormalities produced werethe result of abnormal blood sugar levels in normal animals and whether such states would beattained in a diabetic animal.C. Numbers of Dosage Groups and Number of Animals per Dosage GroupThe various guidelines require different numbers of dosage groups and animals per dosage group.The ICH guideline 2 identifies that all studies should include a control and at least three groupsgiven the drug, although additional groups may be used if considered scientifically necessary. Itrecommends use of 16 to 20 animals per dosage group for most studies (pregnant with litters, whenappropriate), based on the observations that biological responses can generally be identified byusing these numbers of animals and that use of additional animals does not greatly enhance theability to detect changes, except in the case of infrequent events. The general practice is to use 20pregnant animals per group, regardless of species, with the exception of studies performed innonhuman primates or other large animal species. When these are used, it is recommended thatthe protocol design be discussed with the regulatory agencies before conduct of the study to ensurethat there is consensus regarding the sensitivity of assay.D. Dosage SelectionThe basis for identifying an appropriate high dosage is not clearly defined in the various existingguidelines. In general, for pharmaceuticals, a 10% reduction in maternal body weight gain was thecriterion used, unless the drug had so little toxicity that a very high multiple of the clinical dosagewas considered appropriate. For chemicals, a dose that did not result in greater than 10% mortalitywas the usual criterion. Obviously, this subject is one of great concern, especially in considerationof labeling issues. The ICH guideline 2 recognizes that the high dosage should produce minimalmaternal (adult) toxicity and be selected on the basis of data from all available studies, includingpharmacologic, acute, and subchronic toxicity and kinetic studies. Dosage-range studies are recommendedwhen other studies do not provide sufficient information for selection of an appropriatehigh dosage. Issues of toxicity in the high dosage group dam may include:• Reduction in body weight gain.• Increased body weight gain, particularly when related to perturbation of homeostatic mechanisms.• Toxicity to specific target organs.• Changes in hematology and clinical chemistry results.• Exaggerated pharmacological response, which may or may not be reflected as marked clinicalreactions (e.g., sedation, convulsions).• Practical limitations in the dose that can be administered because of the physico-chemical propertiesof the test substance, dosage formulation, and/or the route of administration. Under most circumstances,1 g/kg/day should be an adequate “limit dose.”© 2006 by Taylor & Francis Group, LLC

746 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 19.4ICH treated and evaluated stagesFDA UK EEC JAPANSegment ITreated A–E a A–E A–E A, BEvaluated A–E A–F A–F A–DSegment IITreated C C C C, ½ DEvaluated C, D C, D C, D C–FSegment IIITreated D, E D, E D, E ½ D, EEvaluated D, E D, E D, E ½ D, E, FaA = premating to conception, B = conceptionto implantation, C = implantaion to closure ofthe hard palate, D = closure of the hard palateto the end of pregnancy, E = birth to weaning,F = weaning to sexual maturity.• Kinetics can be useful in determining high dose exposure for low toxicity compounds. There is,however, little point in increasing the administered dosage if it does not result in increased plasmaor tissue concentrations.• Marked increase in embryo-fetal lethality in preliminary studiesSelection of the appropriate high dosage for studies regulated by OECD and EPA remainsproblematic because unless required as the result of negotiations with the regulatory body, parameterscommonly studied in development of pharmaceuticals are omitted from developmental andreproductive toxicology studies, and toxicokinetics are not usually assessed unless problems areidentified. Thus, interaction with the regulatory agency is recommended in this activity.E. Timing of Gestation and Postnatal DevelopmentUnless otherwise identified, GD 0 is defined as the day spermatozoa are observed in a smear ofthe vaginal contents and/or a copulatory plug is observed in situ, and day 0 of lactation is definedas the day of birth. Should alternate timing be used, such as GD 1 or day 1 of lactation, such achange should be clearly identified in the data and reports.1. Exposure IntervalsAs previously mentioned, the ICH guideline 2 segments, the entire reproductive process into sixICH stages, each of which has specific endpoints evaluated. As shown in Table 19.4, earlierguidelines that used a three-segment approach combined these ICH segments in various ways, justas did the various OECD and EPA approaches. For example, in the FDA, 5 UK, 14 and EEC 15 SegmentI studies, treatment of the animals occurred in ICH stages A through E. The UK 14 and EEC 15 studiesalso evaluated the animals in ICH Stage F. The Japanese Segment I study 16,17 treated the animalsin ICH Stages A and B but included evaluations of ICH Stages C and D. The FDA, 5 UK, 14 andEEC 15 Segment II studies treated the animals in ICH Stage C and also evaluated ICH Stage D. TheJapanese Segment II study 16,17 treated the animals in ICH Stage C and part of ICH Stage D, withevaluations continued into ICH Stage F. The FDA, 5 UK, 14 and EEC 15 Segment III studies treatedand evaluated the animals in ICH Stages D and E. The Japanese Segment III study 16,17 treated theanimals beginning near the end of ICH Stage D and through ICH Stage E, with evaluationscontinuing through ICH Stage F.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 747Table 19.5Comparison of current guidelines for the assessment of fertility and early embryonicdevelopmentGuideline ICH 4.1.3 a EPA OPPTS b FDA Red Book c OECD dSpecies (No.)MatingRoute of testcompoundadministrationExposure to testcompoundProcedures ontest animalsAt least one species, preferablyrats (16 to 20 litters)1:1 advisable; mating period of2–3 weeks or until observationof plugSimilar to the route used byhumansPremating treatment interval of 2weeks for females and 2, 4, or10 weeks for males; treatmentshould continue throughoutmating to termination of malesand at least through implantationfor femalesSacrifice females aftermidgestation and males anytime after matingDams: clinical signs; body weight;feed; gross necropsy; vaginalsmears; preserve ovaries, uteriand organs with macroscopicfindings; count corpora lutea andimplantation sites (live and deadconceptuses)Males: gross necropsy; preservetestes, epididymides, andorgans with macroscopicfindings; assess sperm countand viability— — —— — —— — —— — —— — —aFDA, Fed. Regist., 59(183), 1994; FDA, Fed. Regist., 61(67), April 5, 1996.bReproduction and fertility assessments are covered in the EPA OPPTS 870.3800 multigenerationalguideline.cReproduction and fertility assessments are covered in the FDA “Redbook” two generation guideline.dReproduction and fertility assessments are covered in the OECD 415 and 416 multigenerationguidelines.The current ICH guideline 2 allows any combination of intervals A through F, provided all theintervals are evaluated. This procedure eliminates differences in durations of the various segmentsand allows evaluation of the conceptuses at any time in gestation. Additional flexibility was addedto the fertility assessments by allowing the duration of the premating treatment period in maleanimals to be reduced from 60 to 28 days (4 weeks) 3 and later to 14 days 4 when the results of a4-week subchronic study in the same strain and species do not identify adverse effects on malereproductive organs. Most studies continue to be performed using the 28-day exposure period.Should adverse effects be present in the subchronic (multidose) study, the 60-day prematingtreatment period may be used, or other studies may be performed to characterize the effects. Incontrast, the FDA Red Book 21 and the OECD 28,29 and EPA 23 approaches are to provide more fixedprotocol designs to evaluate the same endpoints as those in studies performed for evaluation ofpharmaceuticals. A full comparison of the ICH endpoints and the multiple protocols that evaluatethese endpoints is provided in Table 19.5, Table 19.6, Table 19.7, and Table 19.8.F. Reproductive PerformanceMating and fertility are essential endpoints of reproductive studies; the techniques used in theseevaluations are the same as those used for impregnating animals to be used in developmental© 2006 by Taylor & Francis Group, LLC

Table 19.6Comparison of current guidelines for the assessment of embryo and fetal developmentGuideline ICH 4.1.3 2 EPA OPPTS 870.3700 22 FDA Red Book 21 OECD 27Species (No.)MatingRoute of testcompoundadministrationExposure to testcompoundProcedures on testanimalsOne rodent, preferably rats; onenonrodent, preferably rabbitsPlug, sperm, or insemination is day0Similar to the route used byhumansFrom implantation to the closure ofthe hard palateDam: clinical signs, body weight,feed; necropsy, preserve organswith macroscopic findings; countcorpora lutea; implantations (liveand dead conceptuses)Fetal: gender, body weight, fetalabnormalities (gross, soft tissue,and skeletal; rodents: half skeletaland half soft tissue; rabbits: allfetuses, gross evaluation ofplacentasRodent, e.g., rat, and nonrodent,e.g., rabbit (approximately 20animals with implantation sites atnecropsy)Mate females with males of thesame strain and species; plug orsperm is day 0Orally by intubation or potentialhuman exposure (needjustification)From at least the day ofimplantation to day beforeparturition, or from fertilization today before terminationDam: clinical signs, body weight,feed; terminal weight, grossnecropsy; examine uterinecontents (live and deadconceptuses), examine nongraviduteruses, weigh gravid uterus(with cervix), count corpora luteaFetal: gender, body weight, fetalabnormalities (external, skeletal,and soft tissue anomalies;rodents: half skeletal and half softtissue; rabbits: all fetuses)Rats, rabbits, hamsters, or mice(minimum 20 pregnant)In-house recommended; one maleto one or two females; plug orsperm is day 0Intended human exposure (diet ordrinking water)From implantation to day beforeparturition, or from fertilization toterminationDam: clinical signs, body weight,feed; terminal weight, grossnecropsy; gravid uterine weight,examine uterine contents (live anddead conceptuses), countcorpora luteaFetal: gender, body weight, fetalabnormalities (external, skeletal,and soft tissue anomalies;rodents: half for skeletal and halffor soft tissue; rabbits: all fetuses)Rats, rabbits, mice, or hamsters(20 pregnant)Mate females with males of thesame strain and species; plug orsperm is day 0Orally by intubation; alternate routewith justificationFrom preimplantation throughentire gestation period to the daybefore caesarean-sectioningDam: clinical signs, body weight,feed, terminal weight, grossnecropsy; examine nongraviduteri, gravid uterine weight (withcervix), examine uterine contents(live or dead conceptuses), countcorpora luteaFetal: gender, body weight, fetalabnormalities (external, skeletal,and soft tissue anomalies;rodents: half for skeletal and halffor soft tissue; rabbits: all fetuses)748 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 749Table 19.7Comparison of current guidelines for the postnatal and behavioral assessmentsGuideline ICH 4.1.2 a EPA OPPTS 870.6300 b Book cFDA RedSpecies (No.)MatingRoute of testcompoundadministrationExposure to testcompoundProcedures on testanimalsAt least one species,preferably ratsPlug, sperm, orinsemination is day 0Similar to the route usedby humansFrom implantation to theend of lactationFemales deliver and rearoffspring to weaning;select oneweanling/sex/litter toassess reproductivecompetenceMaternal observations:clinical signs, bodyweight, and feedconsumptionNecropsy: all adults,preserve organs withmacroscopic findingsfor possible histologicalevaluation;implantations,abnormalities inoffspring, live or deadoffspring at birth, bodyweight at birth, pre- andpostweaning survivaland growth and bodyweight, maturation andfertility, physicaldevelopment, sensoryfunctions, reflexes, andbehaviorOECD dApproximately 20 pregnant rats incontrol group; similar numbers formating pairs in treatment groups— —Mate females with males of the— —same strain and species (plug orsperm is day 0)Oral (diet, drinking water, or gavage) — —From implantation through day 10postnatallyDams: detailed clinical observations,including assessment of signs ofautonomic function; description ofabnormal movements, posture,gait, and behavior, etc.; bodyweights (at least weekly, day ofdelivery, and PDs 11 and 21)Observation of offspring: dailycageside observations (PDs 4, 11,21, 35, 45, and 60) for gross signsof mortality or morbidityDevelopmental landmarks: countand weigh live pups (PDs 0, 4, 7,14, and 21, then every 2 weeks)and assess sexual maturation; cullon PD 4Behavior: learning and memory(weaning and adulthood), motoractivity (PDs 13, 17, 21, and60 ± 2), auditory startle (PDs 22and approximately PD 60)Neuropathology: on PND 11 and attermination, remove and weighbrains of 1 pup/sex/litter (total10/sex/dosage group), then select6 pups/sex/dose forneuropathological evaluation— —— —aFDA, Fed. Regist., 59(183), 1994.bPostnatal assessments are covered in the EPA OPPTS 870.6300 Development Neurotoxicity study and EPAOPPTS 870.3800 multigeneration guideline.cPostnatal assessments are covered in the FDA “Redbook” two generation guideline.dPostnatal assessments are covered in the OECD 415 and 416 multigeneration guidelines PND(s) = day(s)postnatal.toxicity studies. As noted for male animals, reduced fertility may have many causes, and theproduction of a litter is not sufficient to assume the absence of an effect on reproductive performance.Rats and mice generally have multiple intromissions in one copulatory interval, and such activitygenerally is associated with improved fertility in the females. Should it appear that aberrant copulatorybehavior is present, videotaping the animals and grading the numbers of copulations (matings) andintromissions, duration of intromission, and expected female receptivity (lordosis) is recommended.The ICH guidelines do not identify specific methods for performing the sperm evaluations,while EPA protocols require evaluation of specific parameters by recommended methodologies,generally including automated sperm evaluations. There are two generally used automated techniques,the Cellsoft System and the Hamilton Thorne System. While the Cellsoft system had themost advanced technical capacity, and hand methods are sufficient to meet the guideline requirements,86,87 the Hamilton Thorne System has become that most widely used.© 2006 by Taylor & Francis Group, LLC

Table 19.8Comparison of current guidelines for evaluating reproductive functionGuideline EPA OPPTS 870.3800 23 FDA Red Book 21Species (No.) Rats, both males and females 5 to 9 weeks in age (approximately 20 pregnant femalesin the control group); similar numbers for mating pairs in treatment groupsMinimum of 20 rats/sex; begin F 0 with 30 rats/sex/group; begin F 1 with 25rats/sex/groupMating Mate females with males of the same strain and species (plug or sperm is day 0) 1:1 from the same dose until pregnant or 2–3 weeks elapsed (plug or spermis day 0)Route of test Oral (diet, drinking water, or gavage)Diet, drinking water, or gavagecompoundadministrationExposure to testcompoundProcedures ontest animalsP: at least 10 weeks before mating and through terminationF 1 : begin at weaning and continue into adulthood, mating, and production of F 2 anduntil the F 2 offspring are weanedParental: clinical observations; body weights (first day of treatment, then weekly, GDs0, 7, 14, and 21, and during lactation on the same days as the weighing of litters);food and/or water consumption; estrous cycle length (vaginal smears for a minimumof 3 weeks prior to mating and throughout cohabitation)Sperm evaluation: (all P and F 1 males) sperm from one testis and one epididymis forenumeration of homogenization-resistant spermatids and cauda epididymal spermreserves, respectively; sperm motility, and sperm morphologyOffspring: number and sex of pups, stillbirths, live births, and the presence of grossanomalies; count and sex live pups, weigh (days 0, 4, 7, 14, and 21 of lactation, atsexual maturation, and at termination) and assess sexual maturation; cull on day 4(optional)Termination: all P and F 1 adult males and females when they are no longer neededfor assessment of reproductive effects; all unselected F 1 offspring and all F 2 offspringat comparable ages after weaningGross necropsy on all P and F 1 parental animals and at least three pups per sex perlitter from unselected F 1 weanlings and F 2 weanlings; observe reproductive organs;vaginal smears to determine stage of estrous; examination of uteruses of allcohabited females (for presence and number of implantation sites)Organ weights on all P and F 1 parental animals: weigh uterus (with oviducts andcervix), ovaries; testes, epididymides (total weights for both and cauda weight foreither one or both), seminal vesicles (with coagulating glands and their fluids) andprostate; brain, pituitary, liver, kidneys, adrenal glands, spleen, and known targetorgansOrgan weights on all F 1 and F 2 weanlings: brain, spleen, and thymusTissue preservation and histopathology on P and F 1 parental animals: vagina, uteruswith oviducts, cervix, and ovaries; one testis, one epididymis, seminal vesicles,prostate, and coagulating gland; pituitary and adrenal glands; target organs, grosslyabnormal tissue; for F 1 and F 2 weanlings selected for macroscopic examination:grossly abnormal tissue and target organs; primordial follicle countF 0 males: at least 10 weeks before mating and during matingF 0 females: at least 10 weeks before mating and during mating andpregnancy to weaning of F 1 litterF 1 litter: prenatal period through postnatal lifeRecommendation: two generations; randomly select F 1 generation by takingat least one rat pup/sex/litter to mate with a pup from the same dosagegroup but a different litterParental: clinical observations at least twice daily; record behavioralchanges, signs of toxicity, and mortality; evaluate all F 0 and F 1 females forestrous cycle length by daily vaginal smears (minimum of 3 weeks prior tomating and during cohabitation); body weights before first dosage, onceweekly thereafter, and at necropsy; feed consumption weekly, waterconsumption (if the test substance is given in water)Offspring: number and sex of pups, stillbirths, live births, and the presenceof gross anomalies; evaluate dead pups; body weights (PD 0, 4, 7, 14, and21); measure anogenital distance of F 2 pups (PD 0); and pubertalevaluations for F 1 weanlings (i.e., age and weight at vaginal opening orbalanopreputial separation)Necropsy on all F 0 and F 1 control and treated parental animals: observeand weigh brain, pituitary, liver, kidneys, adrenal glands, spleen, knowntarget organs, and reproductive organs (i.e., uterus and ovaries, bothtestes, seminal vesicles with coagulating glands, and the prostate)Tissue preservation: (females) vagina, uterus with cervix, ovaries withoviducts, adrenal and pituitary glands, target organs, and grossly abnormaltissue; (male) one testis, one epididymis, seminal vesicles, coagulatingglands, prostate, and adrenal and pituitary glands, target organs, andgrossly abnormal tissueHistopathology: male (e.g., corpus, cauda, and caput epididymis) and female(e.g., primordial follicles) reproductive organsAdditional evaluations: testicular spermatid numbers and sperm motility,morphology, and count; fertility, gestation, live-born, weaning, and viabilityindices; percentage by sexOptions: culling; a third generation (i.e., if overt reproductive, morphologic,and/or toxic effects of a test substance are observed) and a teratologyphase; neurotoxicity and immunotoxicity screens750 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION© 2006 by Taylor & Francis Group, LLC

Guideline OECD 415 (One Generation) 28 OECD 416 (Two Generation) 29Species (No.) Rats 5 to 9 weeks in age (approximately 20 pregnant/group) Rats 5 to 9 weeks of age (approximately 20 pregnant/group)Mating 1:1 or 1:2 male to female; plug or sperm is day 0 Parental: 1:1 until mating occurs or 2 weeks elapse; plug or sperm is day 0F 1 : one weanling/sex/litter with other pups of the same dose level but froma different litterRoute of testcompoundadministrationOral (diet, drinking water or gavage)Oral (diet, drinking water or gavage)Exposure to testcompoundProcedures ontest animalsParental males: at least 10 weeks before matingParental females: at least 2 weeks before mating; continuing during 3-week matingperiod, pregnancy, and until weaning the F 1 offspringParental: clinical observations (at least once daily); body weights (first day oftreatment, then weekly); weekly feed consumption (premating and mating) andduring lactation on the days litters are weighedOffspring: number and sex of pups, stillbirths, live births, and the presence of grossanomalies in the offspring; count and weigh live pups (days 0, 4, and 7, then weeklyuntil termination); culling optionalGross necropsy: all parental animals, with emphasis on reproductive organsHistopathology: ovaries, uterus, vagina, cervix, testes, epididymides, seminal vesicles,prostate, coagulating gland, pituitary, and target organs of all P parental animals(i.e., in control and high dosage group P and F 1 parental animals selected for mating,and rats that die)Parental males: daily beginning at 5 to 9 weeks of age, 10 weeks beforematingParental females: 2 weeks before mating; continuing during mating (3weeks), pregnancy, and through weaning of F 1 offspringF 1 : begin at weaning and continue until terminationParental: clinical observations (at least once daily); body weights (first dayof dosing, then weekly); weekly feed consumption (premating and mating)and during lactation on the days litters are weighedOffspring: number and sex of pups, stillbirths, live births, and the presenceof gross anomalies in the offspring; count and weigh live pups (days 0, 4,and 7, then weekly until term); culling optionalGross necropsy: all parental animals (all P and F 1 ), with emphasis onreproductive organsHistopathology: ovaries, uterus, vagina, cervix, testes, epididymides,seminal vesicles, prostate, coagulating gland, pituitary, and target organsof all control and high dose group P and F 1 parental animals selected formating, and rats that dieNote: P = parental generation, F 0 = parental generation, F 1 = first-generation offspring, F 2 = second-generation offspring, PD = postnatal day, GD = gestational day.PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 751© 2006 by Taylor & Francis Group, LLC

752 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1. Evaluation of the OvariesOvarian weights may be recorded close to the time the ovaries are removed and trimmed to avoiddehydration, or after trimming and fixation. These same techniques as those generally used forrecording organ weights.Although follicle number and size can be identified by histological evaluation of cross-sectionsof the ovary, 88,89 this procedure is extremely time consuming and expensive (30 to 60 sections perovary for mice and rats, resulting in 300 and 600 sections per group of 10 animals, respectively).A shorter screening method, described by Plowchalk et al. 90 for potential use when indicated byobservations in companion studies, e.g., pharmacologic action or reduced ovarian weight or atresia,has the potential to save time and expense. However, when based only on primordial follicles, itshould not be used as the sole endpoint to identify reproductive toxicity. 91 This statement conformswith the harmonization efforts between the EPA and OECD. The EPA draft guidance for reproductionand fertility effects 92 was modified to remove “triggers” that would be inappropriate toidentify chemicals requiring further evaluation and to include an assessment of F 1 females for everystudy. The modifications also added flexibility in methodological criteria (e.g., laboratory-basedselection of ovarian section, number of females to be evaluated, and methods to increase the speedin counting and the statistical power of the sample 93 ).G. Maternal and Pup Interactions and LactationLactation, maternal-pup interaction, and pup growth and development until weaning are monitoredduring ICH Stage E and in FDA Red Book 21 , OECD, and EPA multigeneration studies, 23,28,29 aswell as in the OECD/EPA Developmental Neurotoxicity Study 24,31 and the EPA One GenerationExtension 45 and Pubertal Assays. 46,47H. Growth and Development of Conceptus through Puberty1. Fetal EvaluationsThe various guidelines have differences in terms of the numbers of litters and fetuses evaluatedand the technical procedures recommended. The ICH guideline 2 requires evaluation of all nonrodentfetuses for gross external, soft tissue, and skeletal alterations. Although it is possible to similarlyevaluate all rodent fetuses, for practical reasons (the litters are large) only external alterations arenoted for all rodent fetuses. Subsequently rodent litters are assigned to soft tissue or skeletalevaluations, approximately one-half of the fetuses to each evaluation, regardless of the method usedto evaluate the fetal soft tissue. When a finding is of special interest, the full set of examinationsmay be performed, although such is not common practice.The ICH guideline 2 recognizes each of the four possible outcomes 67 of adverse exposure of theconceptus as important. Regardless of differences in nomenclature, these endpoints remain death(preimplantation loss, early or late resorption, and fetal death, or postimplantation loss, and in somecases, abortion resulting in postimplantation loss), reduced body weight, dysmorphology (malformation,variations associated with retarded growth or with the genetic background of the species),and functional alterations (e.g., behavioral teratology, developmental neurotoxicity, functional teratology,functional alterations of organs, transplacental carcinogenicity).2. Postnatal Evaluations for Gender and Sexual Maturationa. Culling of LittersIn the United States and Japan, litters are commonly culled to a standard number of pups of eachsex. Some investigators believe that this practice can bias study outcome. 94 The ICH guideline 2© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 753allows use of either procedure, provided justification is given for use of the selected procedure. Itis expected that additional study in this area may clarify whether and when culling should beperformed.b. Juvenile Evaluations, including Postnatal Behavioral and Functional EvaluationsThe ICH guideline 2 requires that behavioral and functional evaluations be performed after in uteroexposure of ICH Stages C, D, and E, i.e., embryogenesis, fetogenesis, and lactation. Thus, whenthe treatment interval includes these three ICH stages, postnatal evaluations need be performedonly once; when these ICH stages are tested separately, postnatal evaluations are necessary in eachstudy. This position was supported by the observation that for many agents, fetal and pup bodyweights were as sensitive or more sensitive indicators of developmental toxicity than behavioraland functional evaluations, 95 regardless of whether treatment occurred continually during gestationor was limited to either embryogenesis or fetogenesis and lactation. 96 It must also be noted thatassessment of body weight and other developmental parameters can be confounded by live littersize, sex ratio, and gestational age.Although the specific postnatal evaluations to be performed are not provided by the ICHguideline, 2 the testing facility must justify the appropriateness of the tests to be used and demonstratetheir historical experience and the value of the tests. Because developmental landmark tests highlycorrelate with body weights and gestational ages, 95 such factors should be considered when thesetests are performed. In general, the only physical landmarks considered to be necessary observationsin the behavioral and functional evaluations were those related to sexual maturation, i.e., age atvaginal patency or balanopreputial separation. It was recommended that tests evaluating learning,memory, and activity be included in the behavioral and functional evaluations, and that theseparameters be evaluated by using tests that are not subject to investigator bias (i.e., to the extentpossible, automated tests should be used). Because behavioral and functional testing proceduresare described elsewhere in this book, only those relevant to sexual maturation are described (seeTable 19.8).c. Balanopreputial Separation and Vaginal PatencyPups should be evaluated daily for balanopreputial separation and vaginal patency, beginning onapproximately postnatal days (PD) 39 and 28, respectively. Sexual maturation assessments are bestperformed on at least three randomly selected weanling rats per sex, although common practicehas been to evaluate only one per sex. The criteria for identifying these endpoints are achievedwhen complete preputial separation or vaginal opening are observed, i.e., when the prepuce completelyretracts from the head of the penis or when there is any visible break in the membranoussheath covering the vaginal orifice, respectively. Detailed descriptions of these procedures anddiagrams are available in Lewis et al. 97I. Endocrine-Mediated FindingsAs previously noted, the EPA, in combination with the OECD, has an ongoing program to developand validate new protocols for identifying endocrine-mediated effects. These new protocols werefirst cited in the Final Report of the Endocrine Disruptor Screening and Testing Advisory Committee.98 They are variations of the ICH peri-postnatal and juvenile studies and, in general, assessthe same endpoints. However, because they will be EPA-regulated studies, the designs identifiedafter validation of these protocols should be used. In other words, the flexibility and scientificdigression acceptable in ICH guideline-compliant studies will not be present in these OECD/EPAstudy designs. The three study designs currently undergoing the validation process are:© 2006 by Taylor & Francis Group, LLC

754 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION1. One-generation extension study. 452. Assessments of pubertal development and thyroid function in juvenile female rats on PD 22 through42 or 43. 47 The purpose of this assay is to evaluate the effects of estrogens on vaginal patency.The theoretical advantage of this assay over the uterotrophic assay is that both estrogen agonistsand antagonists are detected by the same test. Additionally, this pubertal assay detects xenoestrogens,aromatase inhibitors, and alterations in hypothalamic-pituitary-gonadal function.3. Assessment of pubertal development and thyroid function in juvenile male rats on PD 23 through52 or 53. 46 The purpose of this pubertal protocol is to assess the effects of androgens andantiandrogens on preputial separation and the entire endocrine system after approximately 3 weeksof treatment.J. Comparison of FDA and EPA Requirements and Methodsfor Postnatal Behavioral and Functional EvaluationsThe revised EPA multigeneration (two-generation) protocol has multiple new endpoints that havebeen historically performed in the development of pharmaceuticals. However, as previously noted,EPA protocols differ from those developed for compliance with the ICH guidelines 2–4 in that theEPA protocols are set procedures. Changes must be negotiated, rather than based solely on scientificjudgment and justification regarding design. Endpoints not previously assessed in EPA-compliantprotocols include:Estrous cyclicity data (vaginal smears) for 3 weeks prior to mating, during mating, and at terminationSperm parameters (count, motility, and morphology)Developmental milestones (age at vaginal opening and preputial separation for the F 1 generation,anogenital distance for the F 2 generation, if triggered by a treatment-related effect on the sexualmaturation of the F 1 generation)Gross pathology and selected histopathology of weanlings and adult organ weight data (uterus, ovaries,testes, one epididymis, including the total weight and cauda epididymal weight, seminal vesicleswith coagulating glands and fluids, prostate, brain, liver, kidneys, adrenal glands, spleen, thymus,and all known target organs)K. Developmental NeurotoxicityThe EPA Developmental Neurotoxicity Test (DNT) is an extension of the procedures used forbehavioral and functional testing during postnatal development. In this test, the vehicle is administeredto a control group, and at least three groups of pregnant animals are administered the testsubstance. Exposure of the mated dams occurs during gestation and early lactation; pups arerandomly selected from within litters and assessed for neurotoxicity. Observations are made todetect gross neurologic and behavioral abnormalities, as well as to assess motor activity, responseto auditory startle, and learning. A neuropathological evaluation is performed, and brain weightsand morphometry are determined. The DNT can be performed as a separate study, as a follow upto a standard developmental toxicity and/or adult neurotoxicity study, or as part of a two-generationreproduction study. Assessment of the offspring is conducted on the second (F 2 ) generation.References1. International Conference on Harmonisation (ICH) Harmonised Tripartite Guideline, Detection oftoxicity to reproduction for medicinal products (proposed rule endorsed by the ICH Steering Committeeat Step 4 of the ICH Process, 24 June 1993), in Proceedings of the Second InternationalConference on Harmonization, D’Arcy, P.F. and Harron, D.W.G., Eds., Greystone Books, Ltd. Antrim,N. Ireland, 1993, p. 557.2. U.S. Food and Drug Administration, International Conference on Harmonisation, Guideline on detectionof toxicity to reproduction for medicinal products, Fed. Regist., 59(183), 1994.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 7553. U.S. Food and Drug Administration, International Conference on Harmonisation, Guideline on detectionof toxicity to reproduction for medicinal products, Addendum on toxicity to male fertility, Fed.Regist., 61(67), April 5, 1996.4. U.S. Food and Drug Administration, International Conference on Harmonisation, Maintenance of theICH Guideline on Toxicity to Male Fertility, An addendum to the ICH tripartite guideline on detectionof toxicity to reproduction for medicinal products, amended November 9, 2000.5. U.S. Food and Drug Administration, Guidelines for Reproduction Studies for Safety Evaluation ofDrugs for Human Use, 1966.6. U.S. Food and Drug Administration, Draft: Toxicological Principles for the Safety Assessment ofDirect Food Additives and Color Additives Used in Food (Red Book II), Guidelines for Reproductionand Developmental Toxicity Studies, Center for Food Safety and Applied Nutrition, Washington, DC,1993, p. 123.7. Christian, M.S. and Voytek, P.E., In vivo reproductive and mutagenicity tests, in A Guide to GeneralToxicology, Homburger, F., Hayes, J.A., and Pelikan, E.W., Eds., Karger, Basel, Switzerland, 1982,p. 294.8. U.S. Environmental Protection Agency, Subpart E — specific organ/tissue toxicity, No. 798.4900:developmental toxicity study, 40 CFR Part 798, Toxic Substances Control Act test guidelines, finalrules, Fed. Regist. 50:39433-39434, 1985a.9. U.S. Environmental Protection Agency, Subpart E — specific organ/tissue toxicity, No. 798.4700:reproduction and fertility effects, 40 CFR Part 798, Toxic Substances Control Act test guidelines, finalrules, Fed. Regist.,, 50:39432-39433, 1985b.10. U.S. Environmental Protection Agency, Toxic Substances Control Act Test Guidelines (TSCA),(OPPTS-42193: FRL-5719-5), Final rule, 40 CFR Part 799, 1997.11. U.S. Environmental Protection Agency (EPA-FIFRA), Hazard Evaluation Division Standard EvaluationProcedure: Teratology Studies, Office of Pesticide Programs, Washington, DC, EPA-540/9-85-018, 1985.12. U.S. Environmental Protection Agency (EPA-FIFRA), Pesticide Assessment Guideline, SubdivisionF — Hazard Evaluation: Human and Domestic Animals, Addendum 10: Neurotoxicity, Health EffectsDivision, Office of Pesticide Programs, March 1991.13. U.S. Environmental Protection Agency (EPA-FIFRA), Health Effects Division Draft Standard EvaluationProcedure, Developmental Toxicity Studies, Office of Pesticides Programs, Washington, DC,1993.14. Committee on the Safety of Medicines, Notes for Guidance on Reproduction Studies, Department ofHealth and Social Security, Great Britain, 1974.15. European Economic Community, The rules of governing medicinal products in the European Community,in Guidelines on the Quality, Safety and Efficacy of Medicinal Products for Human Use, Vol.III, Office of Official Publications of the European Communities, Brussels, 1988.16. Ministry of Health and Welfare, Japan, On Animal Experimental Methods for Testing the Effects ofDrugs on Reproduction, Notification No. 529 of the Pharmaceutical Affairs Bureau, Ministry of Healthand Welfare, March 31, 1975.17. Ministry of Health and Welfare, Japan, Information on the Guidelines of Toxicity Studies Requiredfor Applications for Approval to Manufacture (Import) Drugs, Notification No. 118 of the PharmaceuticalAffairs Bureau, Ministry of Health and Welfare, February 15, 1984.18. Tanimura, T., Kameyama, Y., Shiota, K., Tanaka, S., Matsumoto, N., and Mizutani, M., Report on theReview of the Guidelines for Studies of the Effect of Drugs on Reproduction, Notification No. 118,Pharmaceutical Affairs Bureau, Ministry of Health and Welfare, Japan, 1989.19. International Conference on Harmonisation (ICH) Harmonised Tripartite Guideline, Male fertility studiesin reproductive toxicology, in Proceedings of the Third International Conference on Harmonization,D’Arcy, P.F. and Harron, D.W.G., Eds., Greystone Books, Ltd., Antrim, N. Ireland, 1995, p. 245.20. U.S. Food and Drug Administration, Toxicological Principles and Procedures for Priority BasedAssessment of Food Additives (Red Book), Guidelines for Reproduction Testing with a TeratologyPhase, Bureau of Foods, Washington, DC, 1982, p. 80.21. U.S. Food and Drug Administration, Toxicological Principles for the Safety of Food Ingredients (RedBook 2000), Sections issued electronically July 7, 2000 at http:// vm.cfscan.fda.gov/~Redbook/Redtcct.html,2000.© 2006 by Taylor & Francis Group, LLC

756 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION22. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Prenatal Developmental ToxicityStudy, Office of Prevention, Pesticides and Toxic Substances (OPPTS) 870.3700, 1998.23. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Reproduction and FertilityEffects, Office of Prevention, Pesticides and Toxic Substances (OPPTS) 870.3800, 1998.24. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Developmental NeurotoxicityStudy, Office of Prevention, Pesticides and Toxic Substances (OPPTS) 870.6300, 1998.25. U.S. Environmental Protection Agency (EPA-Risk Assessment Forum), Guidelines for developmentaltoxicity risk assessment, Fed. Regist., 56(234): 63798-63826, Dec. 5, 1991.26. U.S. Environmental Protection Agency, Guidelines for Reproductive Toxicity Risk Assessment, NTISPB No. PB97-100098, 1996a.27. Organization for Economic Cooperation and Development, Guidelines for Testing of Chemicals,Section 4, No. 414: Teratogenicity, adopted 22 January 2001.28. Organization for Economic Cooperation and Development, OECD Guidelines for Testing of Chemicals,Section 4, No. 415: One-Generation Reproduction Toxicity, adopted 26 May 1983.29. Organization for Economic Cooperation and Development, OECD Guidelines for Testing of Chemicals,Section 4, No. 415: Two-Generation Reproduction Toxicity, adopted 22 January 2001(b).30. Organization for Economic Cooperation and Development, OECD Guidelines for Testing of Chemicals,Section 4, No. 422: Combined Repeated Dose Toxicity Study with the Reproduction/DevelopmentalToxicity Screening Test, adopted 22 March 1996.31. Organization for Economic Cooperation and Development, OECD Guidelines for Testing of Chemicals,Section 4, No. 426: Developmental Neurotoxicity Study, drafted October 1999.32. Lenz, W. Kindliche micebildungen nach medikament wahrend der draviditat? Deutsch. Med. Wochenschr.,86, 2555, 1961.33. McBride, W.G., Thalidomide and congenital abnormalities, Lancet, 2, 1358, 1961.34. Herbst, A.L., Ulfelder, H., Poskanzer, D.C., Adenocarcinoma of the vagina: association of maternalstilbestrol therapy with tumor appearance in young women, New Engl. J. Med., 284, 878, 1971.35. Koos, B.J. and Longo, L., Mercury toxicity in pregnant women, fetus and newborn infant, Am. J.Obstet. Gynecol., 126, 390, 1976.36. Christian M.S. and Hoberman A.M., Perspectives on the U.S., EEC, and Japanese developmentaltoxicity testing guidelines, in Handbook of Developmental Toxicology, Hood, R.D., Ed., CRL Press,Inc., New York, 1997, p. 551.37. U.S. Food and Drug Administration, 1998 Final Rule: Pediatric Studies, adopted April 1, 1999.38. U.S. Environmental Protection Agency, Executive Order: Protection of Children from EnvironmentalHealth Risks and Safety Risks, available at http://www.epa.gov/children/whatwe/executiv.htm (verified11 June 2002).39. U.S. Food and Drug Administration, Food Quality Protection Act (FQPA): Public Law 104-170, 1996a.40. U.S. Environmental Protection Agency, Safe Drinking Water Act. (SDWA): Public Law 104-182, 1996b.41. U.S. Environmental Protection Agency, Endocrine Disruptor Screening Program, available athttp://www.epa.gov/scipoly/oscpendo/index.htm (verified 11 June 2002).42. Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC), EDSTAC Final Report,Compilation of EDSTAC Recommendations, chap. 7, available at http://www.epa.gov/scipoly/oscpendo/history/chap7v14.pdf(verified 11 June 2002).43. U.S. Environmental Protection Agency, Endocrine disruptor screening program: statement of policy notice,Fed. Regist., 63(248): 71541-71568, available at http://www.epa.gov/scipoly/oscpendo/fr122898_1.pdf(verified 11 June 2002).44. U.S. Environmental Protection Agency, Endocrine Disruptor Method Validation Subcommittee under theNational Advisory Council for Environmental Policy and Technology, Fed. Regist., 66(197): 51951–51952,available at http://www.epa.gov/oscpmont/oscpendo/fedregnotice.htm (verified 11 June 2002).45. Research Triangle Institute, Draft Protocol: One-Generation Extension Study of Vinclozolin and din-butylPhthalate Administered by Gavage on Gestational Day 6 to Postnatal Day 21 in CD ® (Sprague-Dawley) Rats, Study No. RTA-ED01, Research Triangle Institute, Research Triangle Park, NC, for theUS EPA, 2001.46. Endocrine Disruptor Screening and Testing Advisory Committee, Male 20-day thyroid/pubertal assayin rodents, EDSTAC Final Report 1:5–30, 1998.47. Endocrine Disruptor Screening and Testing Advisory Committee, Female 20-day thyroid/pubertalassay in rodents, EDSTAC Final Report 1:5–26, 1998.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 75748. Endocrine Disruptor Screening and Testing Advisory Committee, EDSTAC Final Report, Chap. 5,Appendix L, Protocols for Tier 1 screening assays, available at http://www.epa.gov/scipoly/oscpendo/history/app-lv14.pdf (verified 11 June 2002).49. Organization for Economic Cooperation and Development, OECD Guidelines for Testing of Chemicals,Section 4, No. 407: Repeated Dose 28-Day Oral Toxicity Study in Rodents, adopted 27 July 1995.50. Interagency Regulatory Group, Recommended Guideline for Teratogenicity Studies in the Rat, Mouse,Hamster or Rabbit, 1980.51. Christian, M.S., Harmonization of reproductive guidelines: a perspective from the International Federationof Teratology Societies, J. Am. Coll. Toxicol., 11(3), 299, 1993.52. D’Arcy, P.F. and Harron, D.W.G., Eds., Topic 2: Reproductive toxicology, in Proceedings of the FirstInternational Conference on Harmonisation, The Queen’s University of Belfast, Belfast, Ireland, 1992,p. 255.53. Organization for Economic Cooperation and Development (OECD), Final report — OECD short-termand long-term toxicology groups, Teratogenicity, 110, 1979.54. U.S. Environmental Protection Agency, Harmonization and Regulatory Coordination: OECD PesticideWorking Group, available at http://www.epa.gov/oppfead1/international/harmonization.html (verified18 June 2002).55. U.S. Environmental Protection Agency, Federal Insecticide, Fungicide, and Rodenticide Act: 7 U.S.C.s/s 136 et seq., available at http://www4.law.cornell.edu/uscode/7/ch6.html (verified 18 June 2002).56. U.S. Food and Drug Administration, The Federal Food, Drug and Cosmetic Act: §408, 21 U.S.C. 348a.57. U.S. Environmental Protection Agency, About the Office of Pesticide Programs (OPP), available athttp://www.epa.gov/pesticides/about.htm (verified 18 June 2002).58. U.S. Environmental Protection Agency, About OPPT, available at http://www.epa.gov/oppt/opptabt.htm(verified 18 June, 2002).59. U.S. Environmental Protection Agency, Toxic Substance Control Act (TSCA): 15 U.S.C. s/s 2601 etseq., available at http://www4.law.cornell.edu/uscode/15/ch53.html (verified 18 June 2002).60. U.S. Environmental Protection Agency, About Harmonized Test Guidelines, available athttp://epa.gov/docs/OPPTS_Harmonized/abguide.txt.html (verified 18 June 2002).61. PMA guidelines, in Reproduction, Teratology and Pediatrics. Christian, M., Diener, R., Hoar, R., andStaples, R, Eds., 1981 (revised).62. Sullivan, F.M., Behavioral teratology, in Proc. 11th Conf. of the European Teratology Society, Paris,1983.63. Christian, M.S., Postnatal alterations of gastrointestinal physiology, hematology, clinical chemistry,and other non-CNS parameters, in Handbook of Experimental Pharmacology: Teratogenesis and ReproductiveToxicology, Vol. 65, Springer, Berlin, 1983, p. 263.64. Canada, Bureau of Drugs, Health Protection Branch, Health and Welfare, Draft of Preclinical TxicologicGuidelines, Section 2.4, 35, 1979.65. Drug directorate guidelines, Toxicological Evaluation 2.4, Reproductive Studies, Ministry of NationalHealth and Welfare, Health Protection Branch, Health and Welfare, Canada, 1990.66. Somers, G.F, Letter to the editor, Lancet, 1, 912, 1962.67. Wilson, J.G., Principles of teratology, in Environment and Birth Defects, Academic Press, New York,1973, p. 11.68. Palmer, A.K., Some thoughts on reproductive studies for safety evaluations, Proc. Europ. Soc. StudyDrug Tox., 14, 79, 1973.69. Johnson, E.M. and Christian, M.S., When is a teratology study not an evaluation of teratogenicity?J. Am. Coll. Toxicol., 3, 431, 1984.70. Wilson, J.G., Methods for administering agents and detecting malformations in experimental animals,Teratology, Principles and Techniques, University of Chicago Press, Chicago, 1965, p. 262.71. Staples, R.E., Detection of visceral alterations in mammalian fetuses, Teratology, 9, 37, 1974.72. Christian, M.S., Test methods for assessing female reproductive and developmental toxicology, inPrinciples and Methods of Toxicology, Hayes, A.W., Ed., Taylor and Francis, Philadelphia, 2001, p.1301.73. Stuckhardt and Poppe, Fresh visceral examination of rat and rabbit fetuses used in teratogenicitytesting, Teratogenesis, Carcinogenesis and Mutagenesis, 4, 181, 1984.74. Staples, R.E. and Schnell, J.L., Refinement in rapid clearing technique in the KOH-alizarin red Smethod for fetal bone, Stain. Technol., 39, 61, 1963.© 2006 by Taylor & Francis Group, LLC

758 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION75. Peters, P.W.J., Double staining of fetal skeletons for cartilage and bone, in Methods in Prenatal Toxicology,Neubert, D., Merker, W.G., and Kwasigroch, T.E., Eds., Georg Thieme Publ., Stuttgart, 1977, p.153.76. Whitaker, J. and Dix, K.M., Double staining technique for rat foetus skeletons in teratological studies,Lab. Animals, 13:309, 1979.77. Webb G.N. and Byrd R.A., Simultaneous differential staining of cartilage and bone in rodent fetuses: analcian blue and alizarin red S procedure without glacial acetic acid, Biotech Histochem., 69:181, 1994.78. U.S. Environmental Protection Agency, Health Effects Test Guidelines: Combined Repeated DoseToxicity with the Reproduction/Developmental Toxicity Screening Test, Office of Prevention, Pesticidesand Toxic Substances (OPPTS) 870.3650, 2000.79. Johnson, E.M., The scientific basis for multigeneration safety evaluation, J. Am. Coll. Toxicol., 5(4),197, 1986.80. Christian, M.S., A critical review of multigeneration studies, J. Am. Coll. Toxicol., 5(2), 161, 1986.81. U.S. Food and Drug Administration, Good Laboratory Practice Regulations, Final Rule 21, Code ofFederal Regulations, Part 58.82. Organization for Economic Cooperation and Development, Laboratory Practice in the Testing of Chemicals,Final Rule of the OCED Expert Group on Laboratory Practices [C(81) 3.0 (Final), Annex 2], 1981.83. Japanese Ministry of Agriculture, Forestry and Fisheries (MAFF), Good Laboratory Practice Regulations,59 Noshan No. 3850, 1984.84. U.S. Environmental Protection Agency, Federal Insecticide, Fungicide and Rodenticide Act (EPA-FIFRA), Good Laboratory Practice Standards, Final Rule, 40 CFR Part 160.85. U.S. Environmental Protection Agency, Toxic Substances Control Act, Good Laboratory PracticeStandards, Final Rule, 40 CFR Part 792.86. Christian, M.S., The use of CASA systems for reproductive toxicology (CRYO Resources Ltd. CellSoftSeries 4000 Tox. System vs the Hamilton Thorne HTM-IVOS), presented at the 82nd Meeting of theMid-Atlantic Reproduction and Teratology Association, Lambertville, PA, October 14–15, 1993.87. Christian, M.S., Compliance with tripartite guideline for male reproductive performance, presentedat the Pharmaceutical Manufacturers Association Drug Safety Subsection Annual Meeting, Austin,TX, October 17–20, 1993c.88. Plowchalk, D.R. and Mattison, D.R., Phosphoramide mustard is responsible for the ovarian toxicityof cyclophosphamide, Toxicol. Appl. Pharmacol., 107, 472, 1991.89. Smith, B.J., Plowchalk, D.R., Sipes, I.G., and Mattison, D.R., Comparison of random and serialsections in assessment of ovarian toxicity, Reprod. Toxicol., 5, 379, 1991.90. Plowchalk, D.R., Smith, B.J., and Mattison, D.R., Assessment of toxicity to the ovary using folliclequantitation and morphometrics, in Methods In Toxicology, Vol. 3B, Female Reproductive Toxicology,Heindel, J.J. and Chapin, R.E., Eds., Academic Press, San Diego, 1993, p. 57.91. Christian, M.S. and Brown W.R., Control primordial follicle counts in multigeneration studies in CRLSprague-Dawley (“Gold Standard”) rats, The Toxicologist, 66(1S), A1140, 2002.92. U.S. Environmental Protection Agency, Proposed Testing Guidelines, Notice of Availability andRequest for Comments, 61 FR 31522, 1996c.93. U.S. Environmental Protection Agency, Overview and Summary of Changes Made in the Harmonizationof OPPTS 870 Toxicology Guidelines with OECD Guidelines, available at http://www.epa.gov/docs/OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/Summary/summchge.htm (verified18 June 2002).94. Palmer, A.K. and Ulbrich, B.C., The cult of culling, Fund. Applied Toxicol., 38, 7, 1997.95. Diener, R.M., Timing and utility of behavioral studies in developmental toxicology, in Proceedingsof the First International Conference on Harmonisation, D’Arcy P.F. and Harron, D.W.G., Eds., TheQueen’s University of Belfast, Antrim, Northern Ireland,, 1993, p. 273.96. Lochry, E.A. and Christian, M.S., Behavioral evaluations in reproductive safety assessments, J. Am.Coll. Toxicol., 10(5), 585, 1991.97. Lewis, E.M., Barnett Jr., J.F., Freshwater, L., Hoberman, A.M., and Christian, M.S., Sexual maturationdata for Crl Sprague-Dawley rats: criteria and confounding factors, Drug Chem. Toxicol., 2002.98. Endocrine Disruptor Screening and Testing Advisory Committee, EDSTAC final report, available athttp:// http://www.epa.gov/scipoly/oscpendo/history/finalrpt.htm (verified 11 June 2002).© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 759APPENDIX 1International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticalsfor Human UseICH HARMONISED TRIPARTITE GUIDELINEDETECTION OF TOXICITY TO REPRODUCTION FOR MEDICINAL PRODUCTSRecommended for Adoption at Step 4 of the ICH Process on 24 June 1993 by the ICH SteeringCommitteeThis Guideline has been developed by the appropriate ICH Expert Working Group and has beensubject to consultation by the regulatory parties, in accordance with the ICH Process. At Step 4 ofthe Process the final draft is recommended for adoption to the regulatory bodies of the EuropeanUnion, Japan and USA.CONTENTS1. Introduction1.1 Purpose of the Guideline1.2 Aim of Studies1.3 Choice of Studies2. Animal Criteria2.1 Selection and Number of Species2.2 Other Test Systems3. General Recommendations Concerning Treatment3.1 Dosages3.2 Route and Frequency of Administration3.3 Kinetics3.4 Control Groups4. Proposed Study Designs — Combination of Studies4.1 The Most Probable Option4.1.1 Study of Fertility and Early Embryonic Development to Implantation4.1.2 Study For Effects on Pre- and Postnatal Development, including Maternal Function4.1.3 Study for Effects on Embryo-Fetal Development4.2 Single Study Design (Rodents)4.3 Two Study Design (Rodents)5. Statistics6. Data Presentation7. Terminology1. INTRODUCTION1.1 Purpose of the GuidelineThere is a considerable overlap in the methodology that could be used to test chemicals andmedicinal products for potential reproductive toxicity. As a first step to using this wider methodologyfor efficient testing, this guideline attempts to consolidate a strategy based on study designs currentlyin use for testing of medicinal products; it should encourage the full assessment on the safety ofchemicals on the development of the offspring. It is perceived that tests in which animals are treatedduring defined stages of reproduction better reflect human exposure to medicinal products and© 2006 by Taylor & Francis Group, LLC

760 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONallow more specific identification of stages at risk. While this approach may be useful for mostmedicines, long term exposure to low doses does occur and may be represented better by a one ortwo generation study approach.The actual testing strategy should be determined by:anticipated drug use especially in relation to reproduction,the form of the substance and route(s) of administration intended for humans and making use of anyexisting data on toxicity, pharmaco-dynamics, kinetics, and similarity to other compounds instructure/activity.To employ this concept successfully, flexibility is needed (Note 1). No guideline can provide sufficientinformation to cover all possible cases, all persons involved should be willing to discuss and considervariations in test strategy according to the state of the art and ethical standards in human and animalexperimentation. Areas where more basic research would be useful for optimization of test designs aremale fertility assessment, and kinetic and metabolism in pregnant/lactating animals.1.2 Aim of StudiesThe aim of reproduction toxicity studies is to reveal any effect of one or more active substance(s)on mammalian reproduction. For this purpose both the investigations and the interpretation of theresults should be related to all other pharmacological and toxicological data available to determinewhether potential reproductive risks to humans are greater, lesser or equal to those posed by othertoxicological manifestations. Further, repeated dose toxicity studies can provide important informationregarding potential effects on reproduction, particularly male fertility. To extrapolate theresults to humans (assess the relevance), data on likely human exposures, comparative kinetics,and mechanisms of reproductive toxicity may be helpful. The combination of studies selectedshould allow exposure of mature adults and all stages of development from conception to sexualmaturity. To allow detection of immediate and latent effects of exposure, observations should becontinued through one complete life cycle, i.e., from conception in one generation through conceptionin the following generation. For convenience of testing this integrated sequence can besubdivided into the following stages.A. Premating to conception (adult male and female reproductive functions, development and maturationof gametes, mating behavior, fertilisation).B. Conception to implantation (adult female reproductive functions, preimplantation development,implantation).C. Implantation to closure of the hard palate (adult female reproductive functions, embryonic development,major organ formation).D. Closure of the hard palate to the end of pregnancy (adult female reproductive functions, fetaldevelopment and growth, organ development and growth).E. Birth to weaning (adult female reproductive functions, neonate adaptation to extrauterine life,preweaning development and growth).F. Weaning to sexual maturity (postweaning development and growth, adaptation to independent life,attainment of full sexual function). For timing conventions, see Note 2.1.3 Choice of StudiesThe guideline addresses the design of studies primarily for detection of effects on reproduction.When an effect is detected, further studies to characterise fully the nature of the response have tobe designed on a case by case basis (Note 3). The rationale for the set of studies chosen should begiven and should include an explanation for the choice of dosages.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 761Studies should be planned according to the “state of the art”, and take into account preexistingknowledge of class-related effects on reproduction. They should avoid suffering and should usethe minimum number of animals necessary to achieve the overall objectives. If a preliminary studyis performed the results should be considered and discussed in the overall evaluation (Note 4).2.. ANIMAL CRITERIAThe animals used must be well defined with respect to their health, fertility, fecundity, prevalenceof abnormalities, embryofetal deaths and the consistency they display from study to study. Withinand between studies animals should be of comparable age, weight and parity at the start; the easiestway to fulfill these criteria is to use animals that are young, mature adults at the time of matingwith the females being virgin.2.1 Selection and Number of SpeciesStudies should be conducted in mammalian species. It is generally desirable to use the same speciesand strain as in other toxicological studies. Reasons for using rats as the predominant rodent speciesare practicality, comparability with other results obtained in this species and the large amount ofbackground knowledge accumulated.In embryotoxicity studies only, a second mammalian species traditionally has been required,the rabbit being the preferred choice as a “non-rodent”. Reasons for using rabbits in embryotoxicitystudies include the extensive background knowledge that has accumulated, as well as availabilityand practicality. Where the rabbit is unsuitable, an alternative non-rodent or a second rodent speciesmay be acceptable and should be considered on a case by case basis (Note 5).2.2 Other Test SystemsOther test systems are considered to be any developing mammalian and non-mammalian cellsystems, tissues, organs, or organism cultures developing independently in vitro or in vivo. Integratedwith whole animal studies either for priority selection within homologous series or assecondary investigations to elucidate mechanisms of action, these systems can provide invaluableinformation and, indirectly, reduce the numbers of animals used in experimentation. However, theylack the complexity of the developmental processes and the dynamic interchange between thematernal and the developing organisms. These systems cannot provide assurance of the absence ofeffect nor provide perspective in respect of risk/exposure. In short, there are no alternative testsystems to whole animals currently available for reproduction toxicity testing with the aims set outin the introduction (Note 6).3. GENERAL RECOMMENDATIONS CONCERNING TREATMENT3.1 DosagesSelection of dosages is one of the most critical issues in design of the reproductive toxicity study.The choice of the high dose should be based on data from all available studies (pharmacology,acute and chronic toxicity and kinetic studies, Note 7). A repeated dose toxicity study of about 2to 4 weeks duration provides a close approximation to the duration of treatment in segmentaldesigns of reproductive studies. When sufficient information is not available preliminary studiesare advisable (see Note 4).© 2006 by Taylor & Francis Group, LLC

762 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONHaving determined the high dosage, lower dosages should be selected in a descending sequence,the intervals depending on kinetic and other toxicity factors. Whilst it is desirable to be able todetermine a “no observed adverse effect level”, priority should be given to setting dosage intervalsclose enough to reveal any dosage related trends that may be present (Note 8).3.2 Route and Frequency of AdministrationIn general the route or routes of administration should be similar to those intended for humanusage. One route of substance administration may be acceptable if it can be shown that a similardistribution (kinetic profile) results from different routes (Note 9). The usual frequency of administrationis once daily but consideration should be given to use either more frequent or less frequentadministration taking kinetic variables into account (see also Note 10).3.3 KineticsIt is preferable to have some information on kinetics before initiating reproduction studies sincethis may suggest the need to adjust choice of species, study design and dosing schedules. At thistime the information need not be sophisticated nor derived from pregnant or lactating animals. Atthe time of study evaluation further information on kinetics in pregnant or lactating animals maybe required according to the results obtained (Note 10).3.4 Control GroupsIt is recommended that control animals be dosed with the vehicle at the same rate as test groupanimals. When the vehicle may cause effects or affect the action of the test substance, a second(sham- or untreated) control group should be considered.4. PROPOSED STUDY DESIGNS — COMBINATION OF STUDIESAll available pharmacological, kinetic, and toxicological data for the test compound and similarsubstances should be considered in deciding the most appropriate strategy and choice of studydesign. It is anticipated that, initially, preference will be given to designs that do not differ tooradically from those of established guidelines for medicinal products (The most probable option).For most medicinal products the 3-study design will usually be adequate. Other strategies, combinationsof studies and study designs could be as valid or more valid as the “most probable option”according to circumstances. The key factor is that, in total, they leave no gaps between stages andallow direct or indirect evaluation of all stages of the reproductive process (Note 11). Designsshould be justified.4.1 The Most Probable OptionThe most probable option can be equated to a combination of studies for effects onFertility and early embryonic developmentPre- and postnatal development, including maternal functionEmbryo-fetal development© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 7634.1.1 Study of Fertility and Early Embryonic Development to ImplantationAIMTo test for toxic effects/disturbances resulting from treatment from before mating (males/females)through mating and implantation. This comprises evaluation of stages A and B of the reproductiveprocess (see 1.2). For females this should detect effects on the oestrous cycle, tubal transport,implantation, and development of preimplantation stages of the embryo. For males it will permitdetection of functional effects (e.g., on libido, epididymal sperm maturation) that may not bedetected by histological examinations of the male reproductive organs (Note 12).ASSESSMENT OFmaturation of gametes,mating behavior,fertility,preimplantation stages of the embryo,implantation.ANIMALSAt least one species, preferably rats.NUMBER OF ANIMALSThe number of animals per sex per group should be sufficient to allow meaningful interpretationof the data (Note 13).ADMINISTRATION PERIODThe design assumes that, especially for effects on spermatogenesis, use will be made of data fromrepeated dose toxicity studies of at least one month duration. Provided no effects have been foundthat preclude this, a premating treatment interval of 2 weeks for females and 4 weeks for malescan be used (Note 12). Selection of the length of the premating administration period should bestated and justified (see also chapter 1.1, pointing out the need for research). Treatment shouldcontinue throughout mating to termination of males and at least through implantation for females.This will permit evaluation of functional effects on male fertility that cannot be detected byhistologic examination in repeated dose toxicity studies and effects on mating behavior in bothsexes. If data from other studies show there are effects on weight or histologic appearance ofreproductive organs in males or females, or if the quality of examinations is dubious or if there areno data from other studies, then a more comprehensive study should be designed (Note 12).MATINGA mating ratio of 1:1 is advisable and procedures should allow identification of both parents of alitter (Note 14).TERMINAL SACRIFICEFemales may be sacrificed at any point after mid pregnancy. Males may be sacrificed at any timeafter mating but it is advisable to ensure successful induction of pregnancy before taking such anirrevocable step (Note 15).OBSERVATIONSDuring study:signs and mortalities at least once daily,body weight and body weight change at least twice weekly (Note 16),food intake at least once weekly (except during mating),© 2006 by Taylor & Francis Group, LLC

764 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONrecord vaginal smears daily, at least during the mating period, to determine whether there are effectson mating or precoital timeobservations that have proved of value in other toxicity studies.At terminal examination:necropsy (macroscopic examination) of all adults,preserve organs with macroscopic findings for possible histological evaluation; keep correspondingorgans of sufficient controls for comparison,preserve testes, epididymides, ovaries and uteri from all animals for possible histological examinationand evaluation on a case by case basis. Tissues can be discarded after completion and reporting ofthe study.sperm count in epididymides or testes, as well as sperm viabilitycount corpora lutea, implantation sites (Note 16),live and dead conceptuses.4.1.2 Study for Effects on Pre- and Postnatal Development,Including Maternal FunctionAIMTo detect adverse effects on the pregnant/lactating female and on development of the conceptusand the offspring following exposure of the female from implantation through weaning. Sincemanifestations of effect induced during this period may be delayed, observations should be continuedthrough sexual maturity (i.e., stages C to F listed in 1.2) (Notes 17, 18).ADVERSE EFFECTS TO BE ASSESSEDenhanced toxicity relative to that in non-pregnant females,pre- and postnatal death of offspring,altered growth and development,functional deficits in offspring, including behavior, maturation (puberty) and reproduction (F1).ANIMALSAt least one species, preferably rats;NUMBER OF ANIMALSThe number of animals per sex per group should be sufficient to allow meaningful interpretationof the data (Note 13).ADMINISTRATION PERIODFemales are exposed to the test substance from implantation to the end of lactation (i.e., stages Cto E listed in 1.2).EXPERIMENTAL PROCEDUREThe females are allowed to deliver and rear their offspring to weaning at which time one male andone female offspring per litter should be selected (document method used) for rearing to adulthoodand mating to assess reproductive competence (Note 19).OBSERVATIONSDuring study (for maternal animals):signs and mortalities at least once daily,body weight and body weight change at least twice weekly (Note 16),© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 765food intake at least once weekly at least until delivery,observations that have proved of value in other toxicity studies,duration of pregnancy,parturition.At terminal examination (for maternal animals and where applicable for offspring)necropsy (macroscopic examination) of all adults,preservation and possibly histological evaluation of organs with macroscopic findings; keep correspondingorgans of sufficient controls for comparison,implantations (Note 16),abnormalities,live offspring at birth,dead offspring at birth,body weight at birth,pre- and postweaning survival and growth/body weight (Note 20), maturation and fertility,physical development (Note 21),sensory functions and reflexes (Note 21),behavior (Note 21).4.1.3 Study for Effects on Embryo-Fetal DevelopmentAIMTo detect adverse effects on the pregnant female and development of the embryo and fetus consequentto exposure of the female from implantation to closure of the hard palate (i.e., stages C toD listed in 1.2).ADVERSE EFFECTS TO BE ASSESSEDenhanced toxicity relative to that in non-pregnant females,embryofetal death,altered growthstructural changes.ANIMALSUsually, two species: one rodent, preferably rat; one non-rodent, preferably rabbit (Note 5). Justificationshould be provided when using one species.NUMBER OF ANIMALSThe number of animals should be sufficient to allow meaningful interpretation of the data (Note 13).ADMINISTRATION PERIODThe treatment period extends from implantation to the closure of the hard palate (i.e., end of C) —see 1.2.EXPERIMENTAL PROCEDUREFemales should be killed and examined about one day prior to parturition. All fetuses should beexamined for viability and abnormalities. To allow subsequent assessment of the relationship betweenobservations made by different techniques fetuses should be individually identified (Note 22).When using techniques requiring allocation to separate examination for soft tissue or skeletalchanges, it is preferable that 50% of fetuses from each litter be allocated for skeletal examination.A minimum of 50% rat fetuses should be examined for visceral alterations, regardless of thetechnique used. When using fresh microdissection techniques for soft tissue alterations — which© 2006 by Taylor & Francis Group, LLC

766 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONis the strongly preferred method for rabbits — 100% of rabbit fetuses should be examined for softtissue and skeletal abnormalities.OBSERVATIONSDuring study (for maternal animals):signs and mortalities at least once daily,body weight and body weight change at least twice weekly (Note 16),food intake at least once weekly, andobservations that have proved of value in other toxicity studies.At terminal examination:necropsy (macroscopic examination) of all adults,preserve organs with macroscopic findings for possible histological evaluation; keep correspondingorgansof sufficient controls for comparison,count corpora lutea, numbers of live and dead implantations (Note 16),individual fetal body weight,fetal abnormalities (Note 22),gross evaluation of placenta.4.2 Single Study Design (Rodents)If the dosing period of the fertility study and pre- and postnatal study are combined into a singleinvestigation, this comprises evaluation of stages A to F of the reproductive process (see 1.2). Ifsuch a study, if it includes fetal examinations, provided clearly negative results at sufficiently highexposure no further reproduction studies in rodents should be required. Fetal examinations forstructural abnormalities can also be supplemented with an embryo-fetal development study (orstudies) to make a 2-study approach (Note 3, 11).Results from a study for effects on embryo-fetal development in a second species are expected(see also 4.1.3).4.3 Two Study Design (Rodents)The simplest two segment design would consist of the fertility study and the pre- and postnataldevelopment study, if it includes fetal examinations. It can be assumed, however, that if the preandpostnatal development study provided no indication of prenatal effects at adequate marginsabove human exposure, the additional fetal examinations (4.1.3.) are most unlikely to provide amajor change in the assessment of risk.Alternatively, female treatment in the fertility study (4.1.1.) could be continued until closure of thehard palate and fetuses examined according to the procedures of the embryo-fetal development study(4.1.3.). This, combined with the pre- and postnatal study (4.1.2.), would provide all the examinationsrequired in “the most probable option” but use considerably less animals (Note 3, 11).Results from a study for effects on embryo-fetal development in a second species are expected(see also 4.1.3).5. STATISTICSAnalysis of the statistics of a study is the means by which results are interpreted. The most importantpart of this analysis is to establish the relationship between the different variables and theirdistribution (descriptive statistics), since these determine how groups should be compared. The© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 767distributions of the endpoints observed in reproductive tests are usually non-normal and extendfrom almost continuous to the extreme categorical. When employing inferential statistics (determinationof statistical significance) the mating pair or litter, not the foetus or neonate, should beused as the basic unit of comparison. The tests used should be justified (Note 23).6. DATA PRESENTATIONThe key to good reporting is the tabulation of individual values in a clear concise manner to accountfor every animal that was entered into the study. A reader should be able to follow the history ofany individual animal from initiation to termination and should be able to deduce with ease thecontribution that the individual has made to any group summary values. Group summary valuesshould be presented in a form that is biologically plausible (i.e., avoid false precision) and thatreflects the distribution of the variable. Appendices or tabulations of individual values such asbodyweight, food consumption, litter values should be concise and, as far as possible, consist ofabsolute rather than calculated values; unnecessary duplication should be avoided.For tabulation of low frequency observations such as clinical signs, autopsy findings, abnormalitiesetc. it is advisable to group together the (few) individuals with a positive recording.Especially in the presentation of data on structural changes (fetal abnormalities) the primary listing(tabulation) should clearly identify the litters containing abnormal fetuses, identify the affectedfetuses in the litter and report all the changes observed in the affected fetus. Secondary listings bytype of change can be derived from this, if necessary.7. TERMINOLOGYBesides effects on the reproductive competence of adult animals toxicity to reproduction includes:Developmental toxicity: Any adverse effect induced prior to attainment of adult life. It includes effectsinduced or manifested in the embryonic or fetal period and those induced or manifested postnatally.Embryotoxicity, fetotoxicity, embryo-fetal toxicity: Any adverse effect on the conceptus resulting fromprenatal exposure, including structural or functional abnormalities or postnatal manifestations ofsuch effects. Terms like “embryotoxicity”, “fetotoxicity” relate to the timepoint/-period of inductionof adverse effects, irrespective of the time of detection.One, two or three generation studies: are defined according to the number of adult breeding generationsdirectly exposed to the test material. For example, in a one generation study there is direct exposureof the F0 generation and indirect exposure (via the mother) of the F1 generation and the study isusually terminated at weaning of the F1 generation. In a two generation study as used for agrochemicalsand industrial chemicals there is direct exposure of the F0 generation, indirect and directexposure of the F1 generation and indirect exposure of the F2 generation. A three generation studyis defined accordingly.Body burden: The total internal dosage of an individual arising from the administration of a substance,comprising parent compound and metabolites, taking distribution and accumulation into account.Kinetics: The term “kinetics” is used consistently throughout this guideline, irrespective of intendingto mean pharmaco- and/or toxicokinetics. No better single term was available.NOTESNote 1 (1.1) Scientific flexibilityThese guidelines are not mandatory rules, they are a starting point rather than an end point. Theyprovide a basis from which an investigator can devise a strategy for testing according to availableknowledge of the test material and the state of the art. For encouragement some alternative test© 2006 by Taylor & Francis Group, LLC

768 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONdesigns have been mentioned in this document but there are others that can be sought out or devised.In devising a strategy, the primary objective should be to detect and bring to light any indicationof toxicity to reproduction. Fine details of study design and technical procedures have been omittedfrom the text. Such decisions rightly belong in the field of the investigator since a technique thatmay be suitable for one laboratory may not be suitable in another. The investigator needs to utilizestaff and resources to do the best he or she can achieve and should know how to do this better thanany outsider; human attributes of attitude, ability, and consistency are more important than materialfacilities. For necessary compliance to GLP, reference is made to such regulations.Note 2 (1.2) Timing conventionsIn this guideline the convention for timing of pregnancy is to refer to the day that a sperm positivevaginal smear and/or plug is observed as day 0 of pregnancy even if mating occurs overnight.Unless shown otherwise it is assumed that, for rats, mice and rabbits implantation occurs on day6-7 of pregnancy, and closure of the hard palate on day 15-18 of pregnancy. Other conventions areequally acceptable but MUST be defined in reports. Also, the investigator must be consistent indifferent studies to assure that no gaps in treatment occur. It is an advisable precaution to providean overlap of at least one day in the exposure period of related studies.The accuracy of the time of mating should be specified since this will affect the variability offetal and neonatal parameters. Similarly, for reared litters, the day offspring are born will beconsidered as postnatal or lactation day 0 unless otherwise specified. However, particularly withregard to delays in, or prolongation of parturition, reference to a postcoital time frame may be useful.Note 3 (1.3) First pass and secondary testingTo a greater or lesser degree all first pass (guideline) tests are apical in nature, i.e., an effect onone endpoint may have several different origins. A reduced litter size at birth may be due to areduced ovulation rate (corpora lutea count), higher rate of preimplantation deaths, higher rate ofpostimplantation deaths or immediate postnatal deaths. In turn these deaths may be the consequenceof an earlier physical malformation that can no longer be observed due to subsequent secondarychanges and so on. Particularly for effects with a natural low frequency among controls, discriminationbetween treatment induced and coincidental occurrence is dependent upon association withother types of effects.A toxicant usually induces more than one type of effect in a dose dependent manner. Forexample, induction of malformation is almost invariably associated with increased embryonic deathand an increased incidence of less severe structural changes. Given an effect on one endpointsecondary investigations for possible associations should be considered, i.e., the nature, scope andorigins of the substance’s toxicity should be characterized. Characterization should also includeidentification of dose-response relationships to facilitate risk assessment; this is different from thesituation in first pass tests where the presence or absence of a dose-response assists discriminationbetween treatment related and coincidental differences.Note 4 (1.3) Preliminary studiesAt the time most reproduction studies are planned or initiated there is usually information availablefrom acute and repeated dose toxicity studies of at least one month’s duration. This informationcan be expected to be sufficient in identifying doses for reproductive studies. If adequate preliminarystudies are performed, they are part of the justification of the choice of dose for the main study.Such studies should be submitted regardless of their GLP-status in principle. This may avoidunnecessary use of animals.Note 5 (2.1) Selection of species and strainsIn choosing an animal species and strain for reproductive toxicity testing care should be given toselect a relevant model. Selection of the species and strain used in other toxicology studies may© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 769avoid the need for additional preliminary studies. If it can be shown — by means of kinetic,pharmacological and toxicological data — that the species selected is a relevant model for thehuman, a single species can be sufficient. There is little value in using a second species if it doesnot show the same similarities to humans. Advantages and disadvantages of species (strains) shouldbe considered in relation to the substance to be tested, the selected study design and in the subsequentinterpretation of the results.All species have their advantages. Rats, and to a lesser extent mice, are good, general purposemodels; the rabbit has been somewhat neglected as a “non-rodent” species for repeated dose toxicityand other reproduction studies than embryotoxicity testing. It has attributes that would make it auseful model for fertility studies, especially male fertility. For both rabbits and dogs (which areoften used as a second species for chronic toxicity studies) it is feasible to obtain semen sampleswithout resorting to painful techniques (electro ejaculation) for longitudinal semen analysis. Mostof the other species are not good, general purpose models and probably are best used for veryspecific investigations only. All species have their disadvantages, for example:Rats: sensitivity to sexual hormones, unsuitable for dopamine agonists due to dependence on prolactinas the primary hormone for establishment and maintenance of early pregnancy, highly susceptibleto non-steroidal anti-inflammatory drugs in late pregnancy.Mice: fast metabolic rate, stress sensitivity, malformation clusters (which occur in all species) particularlyevident, small fetus.Rabbits: often lack of kinetic and toxicity data, susceptibility to some antibiotics and to disturbanceof the alimentary tract, clinical signs can be difficult to interpret.Guinea pigs: often lack of kinetic and toxicity data, susceptibility to some antibiotics and to disturbanceof the alimentary tract, long fetal period, insufficient historical background data.Domestic and/or mini pigs: malformation clusters with variable background rate, large amounts ofcompound required, large housing necessary, insufficient historical background data.Ferrets: seasonal breeder unless special management systems used (success highly dependent onhuman/animal interaction), insufficient historical background data.Hamsters: intravenous route difficult if not impossible, can hide doses in the cheek pouches and canbe very aggressive, sensitive to intestinal disturbance, overly sensitive teratogenic response to manychemicals, small foetus.Dogs: seasonal breeders, inbreeding factors, insufficient historical background data.Non-human primates: kinetically they can differ from humans as much as other species, insufficienthistorical background data, often numbers too low for detection of risk. They are best used whenthe objective of the study is to characterize a relatively certain reproductive toxicant, rather thandetect a hazard.Note 6 (2.2) Uses of other test systems than whole animalsOther test systems have been developed and used in preliminary investigations (“pre-screening” orpriority selection) and secondary testing. For preliminary investigation of a range of analogue seriesof substances it is essential that the potential outcome in whole animals is known for at least onemember of the series to be studied (by inference, effects are expected). With this strategy substancescan be selected for higher level testing. For secondary testing or further substance characterisationother test systems offer the possibility to study some of the observable developmental processesin detail, e.g., to reveal specific mechanisms of toxicity, to establish concentration-response relationships,to select “sensitive periods,” or to detect effects of defined metabolites.Note 7 (3.1) Selection of dosagesUsing similar doses in the reproductive toxicity studies as in the repeated dose toxicity studies willallow interpretation of any potential effects on fertility in context with general systemic toxicity.Some minimal toxicity is expected to be induced in the high dose dams.According to the specific compound, factors limiting the high dosage determined from repeatdose toxicity studies or from preliminary reproduction studies could include:© 2006 by Taylor & Francis Group, LLC

770 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONreduction in bodyweight gainincreased bodyweight gain, particularly when related to perturbation of homeostatic mechanismsspecific target organ toxicityhaematology, clinical chemistryexaggerated pharmacological response, which may or may not be reflected as marked clinical reactions(e.g., sedation, convulsions)the physico-chemical properties of the test substance or dosage formulation which, allied to the routeof administration, may impose practical limitations in the amount that can be administered. Undermost circumstances 1 g/kg/day should be an adequate limit dose.kinetics, they can be useful in determining high dose exposure for low toxicity compounds. There is,however, little point in increasing administered dosage if it does not result in increased plasma ortissue concentrationmarked increase in embryo-fetal lethality in preliminary studiesNote 8 (3.1) Determination of dose-response relationshipsFor many of the variables in reproduction studies the power to discriminate between randomvariation and treatment effect is poor and the presence or absence of a dosage related trend can bea critical means of determining the probability of a treatment effect. It has to be kept in mind thatin these studies dose-responses may be steep, and wide intervals between doses would be inadvisable.If an analysis of dose-response relationships for the effects observed is attempted in a singlestudy, it is recommended to use at least three dose levels and appropriate control groups. If indoubt, a fourth dose group should be added to avoid excessive dosage intervals. Such a strategyshould provide a “no observed adverse effect level” for reproductive aspects. If not, the implicationis that the test substance merits a greater depth of investigation and further studies.Note 9 (3.2) Exposure by different routes of administrationIf it can be shown that one route provides a greater body burden, e.g., area under the curve (AUC),there seems little reason to investigate routes that would provide a lesser body burden or which presentsevere practical difficulties (e.g., inhalation). Before designing new studies for a new route of administrationexisting data on kinetics should be used to determine the necessity of another study.Note 10 (3.3) Kinetics in pregnant animalsKinetic investigations in pregnant and lactating animals may pose some problems due to the rapidchanges in physiology. It is best to consider this as a two or three phase approach. In planning studieskinetic data (often from non-pregnant animals) provide information on the general suitability of thespecies, and can assist in deciding study designs and choice of dosage. During a study kinetic investigationscan provide assurance of accurate dosing or indicate marked deviations from expected patterns.Note 11 (4) Examples for choosing other optionsFor compounds causing no lethality at 2 g/kg and no evidence of repeated dose toxicity at 1 g/kg,conduct of a single 2 generation study with one control and 2 test groups (0.5 and 1.0 g/kg) wouldseem sufficient. However, it might pose the question as to whether the correct species had beenchosen or whether the compound was an effective medicine.For compounds that may be given as a single dose, once in a life time, (e.g., diagnostics,medicines used in operations) it may be impossible to administer repeated dosages more than twicethe human therapeutic dosage for any length of time. A reduced period of treatment allowing ahigher dose would seem more appropriate. For females considerations of human exposure suggestlittle or no need for exposures beyond the embryonic period. For dopamine agonists or compoundsreducing circulating prolactin levels female rats are poor models; the rabbit would probably makea better choice for all the reproductive toxicity studies, but it does not appear to have been attempted.This also applies to other types of compound when the rabbit shows a pattern of metabolismconsiderably closer to humans than the rat.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 771For drugs where alterations in plasma kinetics are seen following repeated administration, thepotential for adverse effects on embryo-fetal development may not be fully evaluated in studiesaccording to 4.1.3. In such cases it may be desirable to extend the period of drug administrationto females in a 4.1.1 study to day 17. With sacrifice at term, both fertility and embryo-fetaldevelopment can be assessed.Note 12 (4.1.1) Premating treatmentThe design of the fertility study, especially the reduction in the premating period for males, is basedon evidence accumulated and re-appraisal of the basic research on the process of spermatogenesisthat originally prompted the demand for a prolonged premating treatment period. Compoundsinducing selective effects on male reproduction are rare; mating with females is an insensitivemeans of detecting effects on spermatogenesis; good pathological and histopathological examination(e.g., by employing Bouin’s fixation, paraffine embedding, transverse sections of 2-4 micronsfor testes, longitudinal sections for epididymides, PAS and haematoxylin staining) of the malereproductive organs provides a more sensitive and quicker means of detecting effects on spermatogenesis;compounds affecting spermatogenesis almost invariably affect post meiotic stages; thereis no conclusive example of a male reproductive toxicant the effects of which could be detectedonly by dosing males for 9-10 weeks and mating them with females.Information on potential effects on spermatogenesis can be derived from repeated dose toxicitystudies. This allows the investigations in the fertility study to be concentrated on other, moreimmediate, causes of effect. It is noted that the full sequence of spermatogenesis (including spermmaturation) in rats lasts 63 days. When the available evidence, or lack of it, suggests that the scopeof investigations in the fertility study should be increased, or extended from detection to characterization,appropriate studies should be designed to further characterise the effects.Note 13 (4.1.1, 4.1.2, 4.1.3) Number of animalsThere is very little scientific basis underlying specified group sizes in past and existing guidelinesnor in this one. The numbers specified are educated guesses governed by the maximum study sizethat can be managed without undue loss of overall study control. This is indicated by the fact thatthe more expensive the animal is to obtain or keep, the smaller the group size proposed. Ideally,at least the same group size should be required for all species and there is a case for using largergroup sizes for less frequently used species such as primates. It should also be made clear that thenumbers required depend on whether or not the group is expected to demonstrate an effect. For ahigh frequency effect few animals are required, to presume the absence of an effect the numberrequired varies according to the variable (endpoint) being considered, its prevalence in controlpopulations (rare or categorical events) or dispersion around the central tendency (continuous orsemi-continuous variables). See also Note 23.For all but the rarest events (such as malformations, abortions, total litter loss), evaluation ofbetween 16 to 20 litters for rodents and rabbits tends to provide a degree of consistency betweenstudies. Below 16 litters per evaluation, between study results become inconsistent, above 20–24litters per group consistency and precision is not greatly enhanced. These numbers relate toevaluation. If groups are subdivided for different evaluations the number of animals starting thestudy should be doubled. Similarly, in studies with 2 breeding generations, 16-20 litters would berequired for the final evaluation of the litters of the F1 generation. To allow for natural wastage,the starting group size of the F0 generation must be larger.Note 14 (4.1.1) MatingMating ratios: When both sexes are being dosed or are of equal consideration in separate male andfemale studies, the preferred mating ratio is 1:1 since this is the safest option in respect of obtaininggood pregnancy rates and avoiding incorrect analysis and interpretation of results.© 2006 by Taylor & Francis Group, LLC

772 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONMating period and practices: most laboratories would use a mating period of between 2-3 weeks,some remove females as soon as a positive vaginal smear or plug is observed whilst others leavethe pairs together. Most rats will mate within the first 5 days of cohabitation (i.e., at the firstavailable estrus), but in some cases females may become pseudopregnant. Leaving the female withthe male for about 20 days allows these females to restart estrus cycles and become pregnant.Note 15 (4.1.1) Terminal sacrificeFEMALESWhen exposure of the females ceases at implantation, termination of females between days 13-15of pregnancy in general is adequate to assess effects on fertility or reproductive function, e.g., todifferentiate between implantation and resorption sites. In general, for detection of adverse effects,it is not thought necessary, in a fertility study, to sacrifice females at day 20/21 of pregnancy inorder to gain information on late embryo loss, fetal death, and structural abnormalities.MALESIt would be advisable to delay sacrifice of the males until the outcome of mating is known. In theevent of an equivocal result, males could be mated with untreated females to ascertain their fertilityor infertility. The males treated as part of study 4.1.1 may also be used for evaluation of toxicityto the male reproductive system if dosing is continued beyond mating and sacrifice delayed.Note 16 (4.1.1, 4.1.2, 4.1.3) ObservationsDaily weighing of pregnant females during treatment can provide useful information. Weighing ananimal more frequently than twice weekly during periods other than pregnancy (premating, mating,lactation) may also be advisable for some compounds. For apparently non-pregnant rats or mice(but not rabbits), ammonium sulphide staining of the uterus might be useful to identify periimplantationdeath of embryos.Note 17 (4.1.2) Treatment of offspringConsequent to derivation from existing guidelines for medicines this guideline does not fully coverexposures from weaning through puberty, nor does it deal with the possibility of reduced reproductivelife span. To detect adverse effects for medicinal products that may be used in infants andjuveniles, special studies (case by case designs) involving direct treatment of offspring, at ages tobe specified, should be considered.Note 18 (4.1.2) Separate embryotoxicity and peri-postnatal studiesIf a pre- and postnatal study is separated into two studies, one covering the embryonic period theother the fetal period, parturition, and lactation, postnatal evaluation of offspring is required in bothstudies.Note 19 (4.1.2) F1-animalsThe guideline suggests selection of 1 male and 1 female per litter on the evidence that it is feasibleto conduct behavioral and other functional tests on the same F1 individuals that will be used forassessment of reproductive function. This has the advantage of allowing cross referencing ofperformance in different tests at the individual level. It is recognised, however, that some laboratoriesprefer to select separate sets of animals for behavior testing and for assessment of reproductivefunction. Which is the most suitable for an individual laboratory will depend upon the combinationof tests used and the resources available.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 773Note 20 (4.1.2) Reduction of litter sizeThe value of culling or not culling for detection of effects on reproduction is still under discussion.Whether or not culling is performed, it should be explained by the investigator.Note 21 (4.1.2) Physical development, sensory functions, reflexes, and behaviorThe best indicator of physical development is bodyweight. Achievement of preweaning landmarksof development such as pinna unfolding, coat growth, incisor eruption etc., is highly correlatedwith pup bodyweight. This weight is better related to postcoital time than postnatal time, at leastwhen significant differences in gestation length occur. Reflexes, surface righting, auditory startle,air righting, and response to light are also dependent on physical development.Two postweaning landmarks of development that are advised are vaginal opening of femalesand cleavage of the balanopreputial gland of males. The latter is associated with increasing testosteronelevels whereas testis descent is not. These landmarks indicate the onset of sexual maturityand it is advised that bodyweight be recorded at the time of attainment to determine whether anydifferences from control are specific or related to general growth.Functional tests: To date, functional tests have been directed almost exclusively to behavior.Even though a great deal of effort has been expended in this direction it is not possible to recommendspecific test methods. Investigators are encouraged to find methods that will assess sensory functions,motor activity, learning and memory.Note 22 (4.1.3) Individual identification and evaluation of fetusesIt must be possible to relate all findings by different techniques (i.e., body weight, externalinspection, visceral and/or skeletal examinations) to single specimen in order to detect patterns ofabnormalities. The examination of mid and low dose fetuses for visceral and/or skeletal abnormalitiesmay not be necessary where the evaluation of the high dose and the control groups did notreveal any relevant differences. It is advisable, however, to store the fixed specimen for possiblelater examination. If fresh dissection techniques are normally used, difficulties with later comparisonsinvolving fixed fetuses should be anticipated.Note 23 (5.) Inferential statistics“Significance” tests (inferential statistics) can be used only as a support for the interpretation ofresults. The interpretation itself must be based on biological plausibility. It is unwise to assumethat a difference from control values is not biologically relevant simply because it is not “statisticallysignificant”. To a lesser extent it can be unwise to assume that a “statistically significant” differencemust be biologically relevant. Particularly for low frequency events (e.g., embryonic death, malformations)with one sided distributions, the statistical power of studies is low.Confidence intervals for relevant quantities can indicate the likely size of the effect. When usingstatistical procedures, experimental units of comparison should be considered: the litter, not theindividual conceptus, the mating pair, when both sexes are treated, the mating pair of the parentgeneration in a two generation study.DETECTION OF TOXICITY TO REPRODUCTIONFOR MEDICINAL PRODUCTS:ADDENDUM ON TOXICITY TO MALE FERTILITYTABLE OF CONTENTSI. Introduction (1)II. Amendments (2)A. Introduction (Last Paragraph) (I.A)© 2006 by Taylor & Francis Group, LLC

774 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONB. Study of Fertility and Early Embryonic Development to Implantation (4.1.1)1. Administration period (IV.A.1.e. (4.1.1))2. Observations (IV.A.1.h (4.1.1))C. Note 12 (IV.A.1 (4.1.1)) Premating TreatmentThis guideline was developed within the Expert Working Group (Safety) of the InternationalConference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals forHuman Use (ICH) and has been subject to consultation by the regulatory parties, in accordancewith the ICH process. This document has been endorsed by the ICH Steering Committee at Step4 of the ICH process, November 29, 1995. At Step 4 of the process, the final draft is recommendedfor adoption to the regulatory bodies of the European Union, Japan and the USA. This guidelinewas published in the Federal Register on April 5, 1996 (61 FR 15360) and is applicable to drugand biological products. Although this guideline does not create or confer any rights for or on anyperson and does not operate to bind FDA or the industry, it does represent the agency’s currentthinking on the detection of toxicity to reproduction for medicinal products. For additional copiesof this guideline, contact the Drug Information Branch, HFD-210, CDER, FDA, 5600 Fishers Lane,Rockville, MD 20857 (Phone: 301-827-4573) or the Manufacturers Assistance and CommunicationStaff (HFM-42), CBER, FDA, 1401 Rockville Pike, Rockville, MD 20852-1448. Send one selfaddressedadhesive label to assist the offices in processing your request. An electronic version ofthis guidance is also available via Internet using the World Wide Web (WWW) (connect to theCDER Home Page at http://www.fda.gov/cder and go to the “Regulatory Guidance” section).I. INTRODUCTION (1)This text is an addendum to the ICH Tripartite Guideline on Detection of Toxicity to Reproductionfor Medicinal Products and provides amendments to the published text. At the time of adoption,it was accepted that the male fertility investigation, as included in the currently harmonizedguideline, would need scientific and regulatory improvement and optimization of test designs.The amendments are intended to provide a better description of the testing concept and recommendations,especially those addressing:FlexibilityPremating treatment durationObservationsThe general principles and background were contained in two papers published in the Journal ofAmerican College of Toxicology. These papers contain the necessary experimental data (prospectiveand retrospective) for reaching consensus and have been commented on. The individual data fromthe Japanese collaborative study were also published in the Journal of Toxicological Science.II. AMENDMENTS (2)A. Introduction (Last Paragraph) (I.A)To employ this concept successfully, flexibility is needed (Note 1). No guideline can providesufficient information to cover all possible cases. All persons involved should be willing to discussand consider variations in test strategy according to the state-of-the-art and ethical standards inhuman and animal experimentation.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 775B. Study of Fertility and Early Embryonic Development to Implantation (4.1.1)1. Administration period (IV.A.1.e. (4.1.1))The design assumes that, especially for effects on spermatogenesis, use will be made of availabledata from toxicity studies (e.g., histopathology, weight of reproductive organs, in some caseshormone assays and genotoxicity data). Provided no effects have been found in repeated dosetoxicity studies of at least 4 weeks duration that preclude this, a premating treatment interval of 2weeks for females and 4 weeks for males can be used (Note 12). Selection of the length of thepremating administration period should be stated and justified. Treatment should continue throughoutmating to termination for males and at least through implantation for females. This will permitevaluation of functional effects on male fertility that cannot be detected by histopathologic examinationin repeated dose toxicity studies and effects on mating behavior in both sexes. If data fromother studies show there are effects on weight or histology of reproductive organs in males orfemales, or if the quality of examinations is dubious, or if there are no data from other studies, theneed for a more comprehensive study should be considered (Note 12).2. Observations (IV.A.1.h (4.1.1))At terminal examination, the following should be done:perform necropsy (macroscopic examination) of all adults;preserve organs with macroscopic findings for possible histopathological evaluation; keep correspondingorgans of sufficient controls for comparison;preserve testes, epididymides, ovaries, and uteri from all animals for possible histopathological examinationand evaluation on a case by case basis;count corpora lutea,implantation sites (note 16);count live and dead conceptuses; andsperm analysis can be used as an optional procedure for confirmation or better characterization of theeffects observed (Note 12).C. Note 12 (IV.A.1 (4.1.1)) Premating TreatmentThe design of the fertility study, especially the reduction in the premating period for males, is basedon evidence accumulated and on re-appraisal of the basic research on the process of spermatogenesis.Compounds inducing selective effects on male reproduction are rare; compounds affectingspermatogenesis almost invariably affect postmeiotic states and weight of testis; mating withfemales is an insensitive means of detecting effects on spermatogenesis. Histopathology of thetestis has been shown to be the most sensitive method for the detection of effects of spermatogenesis.Good pathological and histopathological examination (e.g., by employing Bouin’s fixation, paraffinembedding, transverse section of 2094 microns for testes, longitudinal section for epididymides,PAS and hematoxylin staining) of the male reproductive organs provides a direct means of detection.Sperm analysis (sperm counts, sperm motility, sperm morphology) can be used as an optionalmethod to confirm findings by other methods and to further characterize effects. Sperm analysisdata are considered more relevant for fertility assessment when samples from vas deferens or fromcauda epididymis are used. Information on potential effects on spermatogenesis (and female reproductiveorgans) can be derived from repeated dose toxicity studies or reproductive toxicity studies.For detection of effects not detectable by histopathology of male reproductive organs and spermanalysis, mating with females after a premating treatment of 4 weeks has been shown to be at leastas efficient as mating after a longer duration of treatment (2 weeks may be acceptable in somecases). However, when 2 weeks treatment period is selected, more convincing justification should© 2006 by Taylor & Francis Group, LLC

776 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONbe provided. When the available evidence suggests that the scope of investigations in the fertilitystudy should be increased, appropriate studies should be designed to characterize the effects further.INTERNATIONAL CONFERENCE ON HARMONISATION OF TECHNICALREQUIREMENTS FOR REGISTRATION OF PHARMACEUTICALS FOR HUMAN USEICH HARMONISED TRIPARTITE GUIDELINE MAINTENANCEOF THE ICH GUIDELINE ON TOXICITY TO MALE FERTILITYAn Addendum to the ICH Tripartite Guideline onDETECTION OF TOXICITY TO REPRODUCTION FORMEDICINAL PRODUCTSRecommended for Adoption at Step 4 of the ICH Process on 29 November 1995and amended on 9 November 2000 by the ICH Steering CommitteeThis Guideline has been developed by the appropriate ICH Expert Working Group and has beensubject to consultation by the regulatory parties, in accordance with the ICH Process. At Step 4 ofthe Process the final draft is recommended for adoption to the regulatory bodies of the EuropeanUnion, Japan and USA.Background Note (not part of the guideline)In June 1993 an ICH Harmonised Tripartite Guideline on Detection of Toxicity to Reproductionfor Medicinal Products was adopted by the ICH Steering Committee and has since been implementedin the three ICH regions. Additional data on the duration of the premating treatment andon the parameters suited for the detection of effects on the male reproductive system and on fertilityhave been collected and are presented in two papers that were published in J. Amer. Coll. Toxicol.in August 1995 and also in J. Toxicol. Sci. as Special Issue in September 1995. The Japanesecollaborative study which was conducted came to the conclusion that histopathology of reproductiveorgans is the most sensitive method for detecting effects on spermatogenesis. Sperm analysis givessimilar information and could be a useful method where histopathology is not feasible, as well asfor confirmative purposes and further characterisation. Mating trials were found to be sensitive forthe detection of effects of sperm maturation, sperm motility and of behavioural changes (e.g.,libido). The Japanese parties to ICH compared premating treatments of 4 and 9 weeks durationand found no differences in detection rate, i.e., at least 4 weeks is necessary. Data from the Europeanliterature survey show equal detection efficiency with premating treatments of 2 and 4 weeksduration. Validation studies conducted in Japan recently indicated that the 2 week study with anappropriate dose setting has equal potential in detecting the toxicity on male reproductive organs(Sakai et al 2000).References1. Takayama S, Akaike M, Kawashima K, Takahashi M, Kurokawa Y; A Collaborative Study in Japanon Optimal Treatment Period and Parameters for Detection of Male Fertility Disorders in Rats Inducedby Medical Drugs. J. Amer. Coll. Toxicol., 14: 266-292, 1995.2. Ulbrich B, Palmer AK; Detection of Effects on Male Reproduction – A Literature Survey. J. Amer.Coll. Toxicol., 14; 293-327, 1995 3. J. Toxicol. Sci., Special Issue, Testicular Toxicity, 20: 173-374,1995.3. Sakai T, Takahashi M, Mitsumori K, Yasuhara K, Kawashima K, Mayahara H, Ohno Y; Collaborativework to evaluate toxicity on male reproductive organs by 2-week repeated dose toxicity studies inrats. Overview of the studies. J. Toxicol. Sci. 25, 1-21 (2000)© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 777TOXICITY TO MALE FERTILITYAn Addendum to the ICH Tripartite Guideline onDETECTION OF TOXICITY TO REPRODUCTION FOR MEDICINAL PRODUCTSIntroductionThis text is an addendum to the ICH Tripartite Guideline on Detection of Toxicity to Reproductionfor Medicinal Products 1 and provides amendments to the published text. At the time of adoption,it was accepted that the male fertility investigation, as included in the currently harmonisedguideline, would require scientific and regulatory improvement and optimisation of test designs.The amendments are intended to provide a better description of the testing concept and recommendations,especially those addressing:• flexibility• premating treatment duration• observationsThe general principles and background in two papers were published in the J. Amer. Coll. Toxicol.These papers contain the necessary experimental data (prospective and retrospective) for reachingconsensus, and have been commentated. The individual data from the Japanese collaborative studywere also published in the J. Toxicol. Sci.AMENDMENTSINTRODUCTION (Last Paragraph)To employ this concept successfully, flexibility is needed (Note 1). No guideline can providesufficient information to cover all possible cases. All persons involved should be willing to discussand consider variations in test strategy according to the state of the art and ethical standards inhuman and animal experimentation.4.1.1. Study of Fertility and Early Embryonic Development to ImplantationADMINISTRATION PERIODIt is assumed that, especially for effects on spermatogenesis, use will be made of available datafrom toxicity studies (e.g., histopathology, weight of reproductive organs, in some cases hormoneassays and genotoxicity data). Provided no effects have been found in repeated dose toxicity studiesof at least 2 weeks duration that preclude this, a premating treatment interval of 2 weeks for femalesand 2 weeks for males can be used (Note 12). Selection of the length of the premating administrationperiod should be stated and justified. Treatment should continue throughout mating to terminationof males and at least through implantation for females. This will permit evaluation of functionaleffects on male fertility that cannot be detected by histopathological examination in repeated dosetoxicity studies and effects on mating behaviour in both sexes. If data from other studies showthere are effects on weight or histology of reproductive organs in males or females, or if the qualityof examinations is dubious or if there are no data from other studies, then the need for a morecomprehensive study should be considered (Note 12).OBSERVATIONSAt terminal examination:• Perform necropsy (macroscopic examination) of all adults,• Preserve organs with macroscopic findings for possible histopathological evaluation; keep correspondingorgans of sufficient controls for comparison,• Preserve testes, epididymides, ovaries and uteri from all animals for possible histopathologicalexamination and evaluation on a case-by-case basis,• Count corpora lutea, implantation sites (Note 16),© 2006 by Taylor & Francis Group, LLC

778 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION• Count live and dead conceptuses.• Sperm analysis can be used as an optional procedure for confirmation or better characterisationof the effects observed (Note 12).Note 12 (4.1.1) Premating treatmentThe design of the fertility study, especially the reduction in the premating period for males, is basedon evidence accumulated and on re-appraisal of the basic research on the process of spermatogenesis.Compounds inducing selective effects on male reproduction are rare; compounds affectingspermatogenesis almost invariably affect post meiotic stages and weight of testis; mating withfemales is an insensitive means of detecting effects on spermatogenesis. Histopathology of thetestis has been shown to be the most sensitive method for the detection of effects on spermatogenesis.Good pathological and histopathological examination (e.g., by employing Bouin’s fixation, paraffinembedding, transverse section of 2-4 microns for testes, longitudinal section for epididymides, PASand haematoxylin staining) of the male reproductive organs provides a direct means of detection.Sperm analysis (sperm counts, sperm motility, sperm morphology) can be used as an optionalmethod to confirm findings by other methods and to characterise effects further. Sperm analysisdata are considered more relevant for fertility assessment when samples from vas deferens or fromcauda epididymis are used. Information on potential effects on spermatogenesis (and female reproductiveorgans) can be derived from repeated dose toxicity studies or reproductive toxicity studies.For detection of effects not detectable by histopathology of male reproductive organs and spermanalysis, mating with females after a premating treatment of 4 weeks has been shown to be at leastas efficient as mating after a longer duration of treatment. Since 2 week study was validated to beas effective as a 4 week study, 2 weeks treatment before mating is also acceptable. When theavailable evidence suggests that the scope of investigations in the fertility study should be increased,appropriate studies should be designed to characterise the effects further.© 2006 by Taylor & Francis Group, LLC

The Office of Prevention, Pesticides and Toxic Substances (OPPTS) has developed this guidelinethrough a process of harmonization that blended the testing guidance and requirements that existedin the Office of Pollution Prevention and Toxics (OPPT) and appeared in Title 40, Chapter I,Subchapter R of the Code of Federal Regulations (CFR), the Office of Pesticide Programs (OPP)which appeared in publications of the National Technical Information Service (NTIS) and theguidelines published by the Organization for Economic Cooperation and Development (OECD).The purpose of harmonizing these guidelines into a single set of OPPTS guidelines is tominimize variations among the testing procedures that must be performed to meet the data requirementsof the U.S. Environmental Protection Agency under the Toxic Substances Control Act (15U.S.C. 2601) and the Federal Insecticide, Fungicide and Rodenticide Act (7 U.S.C. 136, et seq.).Final Guideline Release: This guideline is available from the U.S. Government Printing Office,Washington, DC 20402 on disks or paper copies: call (202) 512–0132. This guideline is alsoavailable electronically in ASCII and PDF (portable document format) from EPA’s World WideWeb site (http://www.epa.gov/epahome/research.htm) under the heading ‘‘Researchers and Scientists/TestMethods and Guidelines/OPPTS Harmonized Test Guidelines.’’PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 779APPENDIX 2United States Environmental Protection AgencyPrevention, Pesticides and Toxic Substances(7101)EPA 712–C–98–207August 1998Health Effects Test GuidelinesOPPTS 870.3700Prenatal Developmental Toxicity StudyIntroductionThis guideline is one of a series of test guidelines that have been developed by the Office ofPrevention, Pesticides and Toxic Substances, United States Environmental Protection Agency foruse in the testing of pesticides and toxic substances, and the development of test data that must besubmitted to the Agency for review under Federal regulations.OPPTS 870.3700 Prenatal developmental toxicity study.(a). Scope(1) Applicability. This guideline is intended to meet testing requirements of both the Federal Insecticide,Fungicide, and Rodenticide Act (FIFRA) (7 U.S.C. 136, et seq.), as amended by the FoodQuality Protection Act (FQPA)(Pub. L.104–170), and the Toxic Substances Control Act (TSCA)(15 U.S.C. 2601).(2) Background. The source material used in developing this harmonized OPPTS test guideline is theOPPT guideline under 40 CFR 798.4900, OPP guideline 83–3, and OECD guideline 414.(b). Purpose. This guideline for developmental toxicity testing is designed to provide general informationconcerning the effects of exposure of the pregnant test animal on the developing organism; this mayinclude death, structural abnormalities, or altered growth and an assessment of maternal effects. Forinformation on testing for functional deficiencies and other postnatal effects, the guidelines for thetwo-generation reproductive toxicity study and the developmental neurotoxicity study should beconsulted.(c). Good laboratory practice standards. The study should be conducted in accordance with the laboratorypractices stipulated in 40 CFR Part 160 (FIFRA) and 40 CFR Part 792 (TSCA)—Good LaboratoryPractice Standards.© 2006 by Taylor & Francis Group, LLC

780 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(d). Principle of the test method. The test substance is administered to pregnant animals at least fromimplantation to one day prior to the expected day of parturition. Shortly before the expected date ofdelivery, the pregnant females are terminated, the uterine contents are examined, and the fetuses areprocessed for visceral and skeletal evaluation.(e). Test procedures(1) Animal selection(i) Species and strain. It is recommended that testing be performed in the most relevant species,and that laboratory species and strains which are commonly used in prenatal developmentaltoxicity testing be employed. The preferred rodent species is the rat and the preferred nonrodentspecies is the rabbit.(ii)(iii)(iv)Age. Young adult animals should be used.Sex. Nulliparous female animals should be used at each dose level. Animals should be matedwith males of the same species and strain, avoiding the mating of siblings, if parentage isknown. Day 0 in the test is the day on which a vaginal plug and/or sperm are observed in therodent or that insemination is performed or observed in the rabbit.Animal care. Animal care and housing should be in accordance with the recommendationscontained in the DHHS/PHS NIH Publication No. 86–23, 1985, Guidelines for the Care andHousing of Laboratory Animals, or other appropriate guidelines.(v) Number of animals. Each test and control group should contain a sufficient number of animalsto yield approximately 20 animals with implantation sites at necropsy.(2) Administration of test and control substances(i)(ii)Dose levels and dose selection.(A) At least three-dose levels and a concurrent control should be used. Healthy animalsshould be randomly assigned to the control and treatment groups, in a manner which resultsin comparable mean body weight values among all groups. The dose levels shouldbe spaced to produce a gradation of toxic effects. Unless limited by the physical/chemicalnature or biological properties of the test substance, the highest dose should be chosenwith the aim to induce some developmental and/or maternal toxicity but not deathor severe suffering. In the case of maternal mortality, this should not be more than approximately10 percent. The intermediate dose levels should produce minimal observabletoxic effects. The lowest dose level should not produce any evidence of eithermaternal or developmental toxicity (i.e., the no-observed-adverse-effect level, NOAEL)or should be at or near the limit of detection for the most sensitive endpoint. Two- orfour-fold intervals are frequently optimal for spacing the dose levels, and the additionof a fourth test group is often preferable to using very large intervals (e.g., more than afactor of 10) between dosages.(B) It is desirable that additional information on metabolism and pharmacokinetics of thetest substance be available to demonstrate the adequacy of the dosing regimen. This informationshould be available prior to testing.(C) The highest dose tested need not exceed 1,000 mg/kg/day by oral or dermal administration,or 2 mg/L (or the maximum attainable concentration) by inhalation, unless potentialhuman exposure data indicate the need for higher doses. If a test performed at thelimit dose level, using the procedures described for this study, produces no observabletoxicity and if an effect would not be expected based upon data from structurally relatedcompounds, then a full study using three-dose levels may not be considered necessary.Control group.(A) A concurrent control group should be used. This group should be a sham-treated controlgroup or a vehicle control group if a vehicle is used in administering the test substance.(B) The vehicle control group should receive the vehicle in the highest volume used.(C) If a vehicle or other additive is used to facilitate dosing, consideration should be givento the following characteristics: Effects on the absorption, distribution, metabolism, orretention of the test substance; effects on the chemical properties of the test substancewhich may alter its toxic characteristics; and effects on the food or water consumptionor the nutritional status of the animals.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 781(f)(iii) Route of administration.(A) The test substance or vehicle is usually administered orally by intubation.(B) If another route of administration is used, for example, when the route of administrationis based upon the principal route of potential human exposure, the tester should providejustification and reasoning for its selection, and appropriate modifications may be necessary.Further information on dermal or inhalation exposure is provided under paragraphs(h)(12), (h)(28), and (h)(29) of this guideline. Care should be taken to minimizestress on the maternal animals. For materials administered by inhalation, whole-bodyexposure is preferable to nose-only exposure due to the stress of restraint required fornose-only exposure.(C) The test substance should be administered at approximately the same time each day.(D) When administered by gavage or dermal application, the dose to each animal should bebased on the most recent individual body weight determination.(iv) Dosing schedule. At minimum, the test substance should be administered daily from implantationto the day before cesarean section on the day prior to the expected day of parturition.Alternatively, if preliminary studies do not indicate a high potential for preimplantation loss,treatment may be extended to include the entire period of gestation, from fertilization to approximately1 day prior to the expected day of termination.Observation of animals(1) Maternal.(i)(ii)(iii)(iv)(v)(vi)Mortality, moribundity, pertinent behavioral changes, and all signs of overt toxicity shouldbe recorded at this cageside observation. In addition, thorough physical examinations shouldbe conducted at the same time maternal body weights are recorded.Animals should be weighed on day 0, at termination, and at least at 3-day intervals duringthe dosing period.Food consumption should be recorded on at least 3-day intervals, preferably on days whenbody weights are recorded.Termination schedule.(A) Females should be terminated immediately prior to the expected day of delivery.(B) Females showing signs of abortion or premature delivery prior to scheduled terminationshould be killed and subjected to a thorough macroscopic examination.Gross necropsy. At the time of termination or death during the study, the dam should be examinedmacroscopically for any structural abnormalities or pathological changes which mayhave influenced the pregnancy. Evaluation of the dams during cesarean section and subsequentfetal analyses should be conducted without knowledge of treatment group in order tominimize bias.Examination of uterine contents.(A) Immediately after termination or as soon as possible after death, the uteri should be removedand the pregnancy status of the animals ascertained. Uteri that appear non-gravidshould be further examined (e.g., by ammonium sulfide staining) to confirm the nonpregnantstatus.(B) Each gravid uterus (with cervix) should be weighed. Gravid uterine weights should notbe obtained from dead animals if autolysis or decomposition has occurred.(C) The number of corpora lutea should be determined for pregnant animals.(D) The uterine contents should be examined for embryonic or fetal deaths and the numberof viable fetuses. The degree of resorption should be described in order to help estimatethe relative time of death of the conceptus.(2) Fetal.(i) The sex and body weight of each fetus should be determined.(ii) Each fetus should be examined for external anomalies.(iii) Fetuses should be examined for skeletal and soft tissue anomalies (e.g., variations and malformationsor other categories of anomalies as defined by the performing laboratory).(A) For rodents, approximately one-half of each litter should be prepared by standard techniquesand examined for skeletal alterations, preferably bone and cartilage. The remainder© 2006 by Taylor & Francis Group, LLC

782 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(g)should be prepared and examined for soft tissue anomalies, using appropriate serial sectioningor gross dissection techniques. It is also acceptable to examine all fetuses bycareful dissection for soft tissue anomalies followed by an examination for skeletalanomalies.(B) For rabbits, all fetuses should be examined for both soft tissue and skeletal alterations.The bodies of these fetuses should be evaluated by careful dissection for soft-tissueanomalies, followed by preparation and examination for skeletal anomalies. An adequateevaluation of the internal structures of the head, including the eyes, brain, nasalpassages, and tongue, should be conducted for at least half of the fetuses.Data and reporting.(1) Treatment of results. Data should be reported individually and summarized in tabular form, showingfor each test group the types of change and the number of dams, fetuses, and litters displayingeach type of change.(2) Evaluation of study results. The following should be provided:(i) Maternal and fetal test results, including an evaluation of the relationship, or lack thereof, betweenthe exposure of the animals to the test substance and the incidence and severity of allfindings.(ii)(iii)(iv)(v)Criteria used for categorizing fetal external, soft tissue, and skeletal anomalies.When appropriate, historical control data to enhance interpretation of study results. Historicaldata (on litter incidence and fetal incidence within litter), when used, should be compiled,presented, and analyzed in an appropriate and relevant manner. In order to justify its use asan analytical tool, information such as the dates of study conduct, the strain and source of theanimals, and the vehicle and route of administration should be included.Statistical analysis of the study findings should include sufficient information on the methodof analysis, so that an independent reviewer/statistician can reevaluate and reconstruct theanalysis. In the evaluation of study data, the litter should be considered the basic unit of analysis.In any study which demonstrates an absence of toxic effects, further investigation to establishabsorption and bioavailability of the test substance should be considered.(3) Test report. In addition to the reporting requirements as specified under 40 CFR part 792, subpartJ, and 40 CFR part 160, subpart J, the following specific information should be reported. Both individualand summary data should be presented.(i)(ii)(iii)(iv)(v)Species and strain.Maternal toxic response data by dose, including but not limited to:(A) The number of animals at the start of the test, the number of animals surviving, the numberpregnant, and the number aborting.(B) Day of death during the study or whether animals survived to termination.(C) Day of observation of each abnormal clinical sign and its subsequent course.(D) Body weight and body weight change data, including body weight change adjusted forgravid uterine weight.(E) Food consumption and, if applicable, water consumption data.(F) Necropsy findings, including gravid uterine weight.Developmental endpoints by dose for litters with implants, including:(A) Corpora lutea counts.(B) Implantation data, number and percent of live and dead fetuses, and resorptions (earlyand late).(C) Pre- and postimplantation loss calculations.Developmental endpoints by dose for litters with live fetuses, including:(A) Number and percent of live offspring.(B) Sex ratio.(C) Fetal body weight data, preferably by sex and with sexes combined.(D) External, soft tissue, and skeletal malformation and variation data. The total number andpercent of fetuses and litters with any external, soft tissue, or skeletal alteration, as wellas the types and incidences of individual anomalies, should be reported.The numbers used in calculating all percentages or indices.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 783(h)(vi) Adequate statistical treatment of results.(vii) A copy of the study protocol and any amendments should be included.References. The following references should be consulted for additional background information onthis test guideline:(1) Aliverti, V.L. et al. The extent of fetal ossification as an index of delayed development in teratogenicitystudies in the rat. Teratology 20:237–242 (1979).(2) Barrow, M.V. and W.J. Taylor. A rapid method for detecting malformations in rat fetuses. Journalof Morphology 127:291–306 (1969).(3) Burdi, A.R. Toluidine blue-alizarin red S staining of cartilage and bone in whole-mount skeltonsin vitro. Stain Technolology 40:45–48 (1965).(4) Edwards, J.A. The external development of the rabbit and rat embryo. In Advances in Teratology(ed. D.H.M. Woolam) Vol. 3. Academic, NY (1968).(5) Fritz, H. Prenatal ossification in rabbits as indicative of fetal maturity. Teratology 11:313–320 (1974).(6) Fritz, H. and R. Hess. Ossification of the rat and mouse skeleton in the perinatal period. Teratology3:331–338 (1970).(7) Gibson, J.P. et al. Use of the rabbit in teratogenicity studies. Toxicology and Applied Pharmacology9:398–408 (1966).(8) Inouye, M. Differential staining of cartilage and bone in fetal mouse skeleton by alcian blue andalizarin red S. Congenital Anomalies 16(3):171–173 (1976).(9) Igarashi, E. et al. Frequence of spontaneous axial skeletal variations detected by the double stainingtechnique for ossified and cartilaginous skeleton in rat fetuses. Congenital Anomalies32:381–391 (1992).(10) Kaufman, M. (ed.) The Atlas of Mouse Development. Academic Press, London (1993).(11) Kimmel, C.A. et al. Skeletal development following heat exposure in the rat. Teratology47:229–242 (1993).(12) Kimmel, C.A. and E.Z. Francis. Proceedings of the workshop on the acceptability and interpretationof dermal developmental toxicity studies. Fundamental and Applied Toxicology 14:386–398(1990).(13) Kimmel, C.A. and C. Trammell. A rapid procedure for routine double staining of cartilage andbone in fetal and adult animals. Stain Technology 56:271–273 (1981).(14) Kimmel, C.A. and J.G. Wilson. Skeletal deviation in rats: malformations or variations? Teratology8:309–316 (1973).(15) Marr, M.C. et al. Comparison of single and double staining for evaluation of skeletal development:the effects of ethylene glycol (EG) in CD rats. Teratology 37:476 (1988).(16) Marr, M.C. et al. Developmental stages of the CD (Sprague-Dawley) rat skeleton after maternalexposure to ethylene glycol. Teratology 46:169–181 (1992).(17) McLeod, M.J. Differential staining of cartilage and bone in whole mouse fetuses by Alcian blueand alizarin red S. Teratology 22:299–301 (1980).(18) Monie, I.W. et al. Dissection procedures for rat fetuses permitting alizarin red staining of skeletonand histological study of viscera. Supplement to Teratology Workshop Manual, pp. 163–173(1965).(19) Organization for Economic Cooperation and Development, No. 414: Teratogenicity, Guidelinesfor Testing of Chemicals. [C(83)44(Final)] (1983).(20) Salewski (Koeln), V.E. Faerbermethode zum makroskopischen nachweis von implantationsstellen am uterus der ratte. Naunyn-Schmeidebergs Archiv fu¨ r Pharmakologie und ExperimentellePathologie 247:367 (1964).(21) Spark, C. and A.B. Dawson. The order and time of appearance of centers of ossification in the foreand hind limbs of the albino rat, with special reference to the possible influence of the sex factor.American Journal of Anatomy 41:411–445 (1928).(22) Staples, R.E. Detection of visceral alterations in mammalian fetuses. Teratology 9(3): A37–A38(1974).(23) Staples, R.E. and V.L. Schnell. Refinements in rapid clearing technique in the KOH—alizarin redS method for fetal bone. Stain Technology 39:61–63 (1964).(24) Strong, R.M. The order time and rate of ossification of the albino rat (mus norvegicus albinus)skeleton. American Journal of Anatomy 36: 313–355 (1928).© 2006 by Taylor & Francis Group, LLC

784 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(25) Stuckhardt, J.L. and S.M. Poppe. Fresh visceral examination of rat and rabbit fetuses used in teratogenicitytesting. Teratogenesis, Carcinogenesis,and Mutagenesis 4:181–188 (1984).(26) U.S. Environmental Protection Agency. Guideline 83–3: Teratogencity Study. Pesticide AssessmentGuidelines, Subdivision F. Hazard Evaluation: Human and Domestic Animals. Office of Pesticidesand Toxic Substances, Washington, DC, EPA–540/9–82–025 (1982).(27) U.S. Environmental Protection Agency. Subpart E—Specific Organ/Tissue Toxicity, 40 CFR798.4900: Developmental Toxicity Study.(28) U.S. Environmental Protection Agency. Health Effects Test Guidelines, OPPTS 870.3200, 21/28-Day Dermal Toxicity, July 1998.(29) U.S. Environmental Protection Agency. Health Effects Test Guidelines, OPPTS 870.3465, 90-DayInhalation Toxicity, July 1998.(30) U.S. Environmental Protection Agency Guidelines for Developmental Toxicity Risk Assessment.FEDERAL REGISTER (56 FR 63798–63826, December 5, 1991).(31) Van Julsingha, E.B. and C.G. Bennett. A dissecting procedure for the detection of anomalies inthe rabbit foetal head. In: Methods in Prenatal Toxicology (eds. D. Neubert, H.J. Merker, and T.E.Kwasigroch). University of Chicago, Chicago, IL, pp. 126–144 (1977).(32) Walker, D.G. and Z.T. Wirtschafter. The Genesis of the Rat Skeleton. Thomas, Springfield, IL(1957).(33) Whitaker, J. and D.M. Dix. Double-staining for rat foetus skeletons in teratological studies. LaboratoryAnimals 13:309–310 (1979).(34) Wilson, J.G. Embryological considerations in teratology. In ‘‘Teratology: Principles and Techniques’’(ed. J.G. Wilson and J. Warkany). University of Chicago, Chicago, IL, pp 251–277(1965).(35) Wilson, J.G. and F.C. Fraser, ed. Handbook of Teratology, Vol. 4. Plenum, NY (1977).(36) Yasuda, M. and T. Yuki. Color Atlas of Fetal Skeleton of the Mouse, Rat, and Rabbit. Ace Art Co.,Osaka, Japan (1996).United States Environmental Protection AgencyPrevention, Pesticides and Toxic Substances(7101)EPA 712–C–98–208August 1998Health Effects Test GuidelinesOPPTS 870.3800Reproduction and Fertility EffectsIntroductionThis guideline is one of a series of test guidelines that have been developed by the Office ofPrevention, Pesticides and Toxic Substances, United States Environmental Protection Agency foruse in the testing of pesticides and toxic substances, and the development of test data that must besubmitted to the Agency for review under Federal regulations.The Office of Prevention, Pesticides and Toxic Substances (OPPTS) has developed this guidelinethrough a process of harmonization that blended the testing guidance and requirements that existedin the Office of Pollution Prevention and Toxics (OPPT) and appeared in Title 40, Chapter I,Subchapter R of the Code of Federal Regulations (CFR), the Office of Pesticide Programs (OPP)which appeared in publications of the National Technical Information Service (NTIS) and theguidelines published by the Organization for Economic Cooperation and Development (OECD).The purpose of harmonizing these guidelines into a single set of OPPTS guidelines is tominimize variations among the testing procedures that must be performed to meet the data requirementsof the U. S. Environmental Protection Agency under the Toxic Substances Control Act (15U.S.C. 2601) and the Federal Insecticide, Fungicide and Rodenticide Act (7 U.S.C. 136, et seq.).© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 785Final Guideline Release: This guideline is available from the U.S. Government Printing Office,Washington, DC 20402 on disks or paper copies: call (202) 512–0132. This guideline is alsoavailable electronically in ASCII and PDF (portable document format) from EPA’s World WideWeb site (http://www.epa.gov/epahome/research.htm) under the heading ‘‘Researchers and Scientists/TestMethods and Guidelines/OPPTS Harmonized Test Guidelines.’’OPPTS 870.3800 Reproduction and fertility effects.(a) Scope(1) Applicability. This guideline is intended to meet testing requirements of both the Federal Insecticide,Fungicide, and Rodenticide Act (FIFRA) (7 U.S.C. 136, et seq.), as amended by the FoodQuality Protection Act (FQPA)(Pub. L. 104–170) and the Toxic Substances Control Act (TSCA)(15 U.S.C. 2601).(2) Background. The source material used in developing this harmonized OPPTS test guideline is theOPPT guideline under 40 CFR 798.4700, OPP guideline 83–4, and OECD guideline 416.(b) Purpose. This guideline for two-generation reproduction testing is designed to provide general informationconcerning the effects of a test substance on the integrity and performance of the male andfemale reproductive systems, including gonadal function, the estrous cycle, mating behavior, conception,gestation, parturition, lactation, and weaning, and on the growth and development of the offspring.The study may also provide information about the effects of the test substance on neonatal morbidity,mortality, target organs in the offspring, and preliminary data on prenatal and postnatal developmentaltoxicity and serve as a guide for subsequent tests. Additionally, since the study design includes inutero as well as post-natal exposure, this study provides the opportunity to examine the susceptibilityof the immature/neonatal animal. For further information on functional deficiencies and developmentaleffects, additional study segments can be incorporated into the protocol, utilizing the guidelines fordevelopmental toxicity or developmental neurotoxicity.(c) Good laboratory practice standards. The study should be conducted in accordance with the laboratorypractices stipulated in 40 CFR Part 160 (FIFRA) and 40 CFR Part 792 (TSCA)—Good LaboratoryPractice Standards.(d) Principle of the test method. The test substance is administered to parental (P) animals prior to andduring their mating, during the resultant pregnancies, and through the weaning of their F1 offspring.The sub-stance is then administered to selected F1 offspring during their growth into adulthood,mating, and production of an F2 generation, until the F2 generation is weaned.(e) Test procedures(1) Animal selection(i) Species and strain. The rat is the most commonly used species for testing. If another mammalianspecies is used, the tester should provide justification/reasoning for its selection, andappropriate modifications will be necessary. Healthy parental animals, which have been acclimatedto laboratory conditions for at least 5 days and have not been subjected to previousexperimental procedures, should be used. Strains of low fecundity should not be used.(ii) Age. Parental (P) animals should be 5 to 9 weeks old at the start of dosing. The animals ofall test groups should be of uniform weight, age, and parity as nearly as practicable, andshould be representative of the species and strain under study.(iii) Sex.(A) For an adequate assessment of fertility, both males and females should be studied.(B) The females should be nulliparous and nonpregnant.(iv) Animal care. Animal care and housing should be in accordance with the recommendationscontained in DHHS/PHS NIH Publication No. 86–23, 1985, Guidelines for the Care and Useof Laboratory Animals, or other appropriate guidelines.(v) Number of animals. Each control group should contain a sufficient number of mating pairsto yield approximately 20 pregnant females. Each test group should contain a similar numberof mating pairs.(vi) Identification of animals. Each animal should be assigned a unique identification number. Forthe P generation, this should be done before dosing starts. For the F1 generation, this shouldbe done for animals selected for mating; in addition, records indicating the litter of originshould be maintained for all selected F1 animals.© 2006 by Taylor & Francis Group, LLC

786 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(2) Administration of test and control substances(i) Dose levels and dose selection.(A) At least three-dose levels and a concurrent control should be used. Healthy animalsshould be randomly assigned to the control and treatment groups, in a manner which resultsin comparable mean body weight values among all groups. The dose levels shouldbe spaced to produce a gradation of toxic effects. Unless limited by the physical/chemicalnature or biological properties of the test substance, the highest dose should be chosenwith the aim to induce some reproductive and/or systemic toxicity but not death orsevere suffering. In the case of parental mortality, this should not be more than approximately10 percent. The intermediate dose levels should produce minimal observabletoxic effects. The lowest dose level should not produce any evidence of either systemicor reproductive toxicity (i.e., the no-observed-adverse-effect level, NOAEL) or shouldbe at or near the limit of detection for the most sensitive endpoint. Two- or four-fold intervalsare frequently optimal for spacing the dose levels, and the addition of a fourthtest group is often preferable to using very large intervals (e.g., more than a factor of 10)between dosages.(B) It is desirable that additional information on metabolism and pharmacokinetics of thetest substance be available to demonstrate the adequacy of the dosing regimen. This informationshould be available prior to testing.(C) The highest dose tested should not exceed 1,000 mg/kg/day (or 20,000 ppm in the diet),unless potential human exposure data indicate the need for higher doses. If a test performedat the limit dose level, using the procedures described for this study, producesno observable toxicity and if an effect would not be expected based upon data fromstructurally related compounds, then a full study using three dose levels may not be considerednecessary.(ii) Control group.(A) A concurrent control group should be used. This group should be an untreated or shamtreated group or a vehicle-control group if a vehicle is used in administering the test substance.(B) If a vehicle is used in administering the test substance, the control group should receivethe vehicle in the highest volume used.(C) If a vehicle or other additive is used to facilitate dosing, consideration should be givento the following characteristics: Effects on the absorption, distribution, metabolism, orretention of the test substance; effects on the chemical properties of the test substancewhich may alter its toxic characteristics; and effects on the food or water consumptionor the nutritional status of the animals.(D) If a test substance is administered in the diet and causes reduced dietary intake or utilization,the use of a pair-fed control group may be considered necessary.(iii) Route of administration.(A) The test substance is usually administered by the oral route (diet, drinking water, or gavage).(B) If administered by gavage or dermal application, the dosage administered to each animalprior to mating and during gestation and lactation should be based on the individual animalbody weight and adjusted weekly at a minimum.(C) If another route of administration is used, for example, when the route of administrationis based upon the principal route of potential human exposure, the tester should providejustification and reasoning for its selection, and appropriate modifications may be necessary.Further in-formation on dermal or inhalation exposure is provided under paragraphs(g)(18) and (g)(19) of this guideline. Care should be taken to minimize stress onthe maternal animals and their litters during gestation and lactation.(D) All animals should be dosed by the same method during the appropriate experimentalperiod.(iv) Dosing schedule.(A) The animals should be dosed with the test substance on a 7–days–a–week basis.(B) Daily dosing of the parental (P) males and females should begin when they are 5 to 9© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 787weeks old. Daily dosing of the F1 males and females should begin at weaning. For bothsexes (P and F1), dosing should be continued for at least 10 weeks before the matingperiod.(C) Daily dosing of the P and F1 males and females should continue until termination.(3) Mating procedure(i) Parental.(A) For each mating, each female should be placed with a single randomly selected malefrom the same dose level (1:1 mating) until evidence of copulation is observed or either3 estrous periods or 2 weeks has elapsed. Animals should be separated as soon as possibleafter evidence of copulation is observed. If mating has not occurred after 2 weeksor 3 estrous periods, the animals should be separated without further opportunity formating. Mating pairs should be clearly identified in the data.(B) Vaginal smears should be collected daily and examined for all females during mating,until evidence of copulation is observed.(C) Each day, the females should be examined for presence of sperm or vaginal plugs. Day0 of pregnancy is defined as the day a vaginal plug or sperm are found.(ii) F1 mating. For mating the F1 offspring, at least one male and one female should be randomlyselected from each litter for mating with another pup of the same dose level but different litter,to produce the F2 generation.(iii) Second mating. In certain instances, such as poor reproductive performance in the controls,or in the event of treatment-related alterations in litter size, the adults may be re-mated to producean F1b or F2b litter. If production of a second litter is deemed necessary in either generation,the dams should be remated approximately 1–2 weeks following weaning of the lastF1a or F2a litter.(iv) Special housing. After evidence of copulation, animals that are presumed to be pregnantshould be caged separately in delivery or maternity cages. Pregnant animals should be providedwith nesting materials when parturition is near.(v) Standardization of litter sizes.(A) Animals should be allowed to litter normally and rear their offspring to weaning. Standardizationof litter sizes is optional.(B) If standardization is performed, the following procedure should be used. On day 4 afterbirth, the size of each litter may be adjusted by eliminating extra pups by random selectionto yield, as nearly as possible, four males and four females per litter or five malesand five females per litter. Selective elimination of pups, i.e., based upon body weight,is not appropriate. Whenever the number of male or female pups prevents having four(or five) of each sex per litter, partial adjustment (for example, five males and three females,or four males and six females) is acceptable. Adjustments are not appropriate forlitters of eight pups or less.(4) Observation of animals(i)Parental.(A) Throughout the test period, each animal should be observed at least once daily, consideringthe peak period of anticipated effects after dosing. Mortality, moribundity, pertinentbehavioral changes, signs of difficult or prolonged parturition, and all signs of overttoxicity should be recorded at this cageside examination. In addition, thorough physicalexaminations should be conducted weekly on each animal.(B) Parental animals (P and F1) should be weighed on the first day of dosing and weeklythereafter. Parental females (P and F1) should be weighed at a minimum on approximatelygestation days 0, 7, 14 and 21, and during lactation on the same days as theweighing of litters.(C) During the premating and gestation periods, food consumption should be measuredweekly at a minimum. Water consumption should be measured weekly at a minimum ifthe test substance is administered in the water.(D) Estrous cycle length and pattern should be evaluated by vaginal smears for all P and F1females during a minimum of 3 weeks prior to mating and throughout cohabitation; careshould be taken to prevent the induction of pseudopregnancy.© 2006 by Taylor & Francis Group, LLC

788 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(E) For all P and F1 males at termination, sperm from one testis and one epididymis shouldbe collected for enumeration of homogenization-resistant spermatids and cauda epididymalsperm reserves, respectively. In addition, sperm from the cauda epididymis (orvas deferens) should be collected for evaluation of sperm motility and sperm morphology.(1) The total number of homogenization-resistant testicular sperm and cauda epididymalsperm should be enumerated (see paragraphs (g)(3) and (g)(13) of this guideline).Cauda sperm reserves can be derived from the concentration and volume ofsperm in the suspension used to complete the qualitative evaluations, and the numberof sperm recovered by subsequent mincing and/or homogenizing of the remainingcauda tissue. Enumeration in only control and high-dose P and F1 malesmay be performed unless treatment-related effects are observed; in that case, thelower dose groups should also be evaluated.(2) An evaluation of epididymal (or vas deferens) sperm motility should be performed.Sperm should be recovered while minimizing damage (refer to paragraph (g)(13)of this guideline), and the percentage of progressively motile sperm should be determinedeither subjectively or objectively. For objective evaluations, an acceptablecounting chamber of sufficient depth can be used to effectively combine the assessmentof motility with sperm count and sperm morphology. When computer-assistedmotion analysis is performed (refer to paragraph (g)(13) of this guideline), thederivation of progressive motility relies on user-defined thresholds for average pathvelocity and straightness or linear index. If samples are videotaped, or images otherwiserecorded, at the time of necropsy, subsequent analysis of only control andhigh-dose P and F1 males may be performed unless treatment-related effects areobserved; in that case, the lower dose groups should also be evaluated. In the absenceof a video or digital image, all samples in all treatment groups should be analyzedat necropsy.(3) A morphological evaluation of an epididymal (or vas deferens) sperm sampleshould be performed. Sperm (at least 200 per sample) should be examined as fixed,wet preparations (refer to paragraphs (g)(7) and (g)(13) of this guideline) and classifiedas either normal (both head and midpiece/tail appear normal) or abnormal.Examples of morphologic sperm abnormalities would include fusion, isolatedheads, and misshapen heads and/or tails. Evaluation of only control and high-doseP and F1 males may be performed unless treatment-related effects are observed; inthat case, the lower dose groups should also be evaluated.(ii) Offspring.(A) Each litter should be examined as soon as possible after delivery (lactation day 0) to establishthe number and sex of pups, stillbirths, live births, and the presence of gross anomalies.Pups found dead on day 0 should be examined for possible defects and cause of death.(B) Live pups should be counted, sexed, and weighed individually at birth, or soon thereafter,at least on days 4, 7, 14, and 21 of lactation, at the time of vaginal patency or balanopreputialseparation, and at termination.(C) The age of vaginal opening and preputial separation should be determined for F1 weanlingsselected for mating. If there is a treatment-related effect in F1 sex ratio or sexualmaturation, anogenital distance should be measured on day 0 for all F2 pups.(5) Termination schedule.(i) All P and F1 adult males and females should be terminated when they are no longer neededfor assessment of reproductive effects.(ii) F1 offspring not selected for mating and all F2 offspring should be terminated at comparableages after weaning.(6) Gross necropsy.(i)At the time of termination or death during the study, all parental animals (P and F1) and whenlitter size permits at least three pups per sex per litter from the unselected F1 weanlings andthe F2 weanlings should be examined macroscopically for any structural abnormalities orpathological changes. Special attention should be paid to the organs of the reproductive system.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 789(ii) Dead pups or pups that are terminated in a moribund condition should be examined for possibledefects and/or cause of death.(iii) At the time of necropsy, a vaginal smear should be examined to determine the stage of theestrous cycle. The uteri of all cohabited females should be examined, in a manner which doesnot compromise histopathological evaluation, for the presence and number of implantationsites.(7) Organ weights.(i) At the time of termination, the following organs of all P and F1 parental animals should beweighed:(A) Uterus (with oviducts and cervix), ovaries.(B) Testes, epididymides (total weights for both and cauda weight for either one or both),seminal vesicles (with coagulating glands and their fluids), and prostate.(C) Brain, pituitary, liver, kidneys, adrenal glands, spleen, and known target organs.(ii) For F1 and F2 weanlings that are examined macroscopically, the following organs should beweighed for one randomly selected pup per sex per litter.(A) Brain.(B) Spleen and thymus.(8) Tissue preservation. The following organs and tissues, or representative samples thereof, shouldbe fixed and stored in a suitable medium for histopathological examination.(i) For the parental (P and F1) animals:(A) Vagina, uterus with oviducts, cervix, and ovaries.(B) One testis (preserved in Bouins fixative or comparable preservative), one epididymis,seminal vesicles, prostate, and coagulating gland.(C) Pituitary and adrenal glands.(D) Target organs, when previously identified, from all P and F1 animals selected for mating.(E) Grossly abnormal tissue.(ii) For F1 and F2 weanlings selected for macroscopic examination: Grossly abnormal tissue andtarget organs, when known.(9) Histopathlogy(i) Parental animals. Full histopathology of the organs listed in paragraph (e)(8)(i) of this guidelineshould be performed for ten randomly chosen high dose and control P and F1 animalsper sex, for those animals that were selected for mating. Organs demonstrating treatment-relatedchanges should also be examined for the remainder of the high-dose and control animalsand for all parental animals in the low- and mid-dose groups. Additionally, reproductiveorgans of the low- and mid-dose animals suspected of reduced fertility, e.g., those that failedto mate, conceive, sire, or deliver healthy offspring, or for which estrous cyclicity or spermnumber, motility, or morphology were affected, should be subjected to histopathologicalevaluation. Besides gross lesions such as atrophy or tumors, testicular histopathological examinationshould be conducted in order to identify treatment-related effects such as retainedspermatids, missing germ cell layers or types, multinucleated giant cells, or sloughing ofspermatogenic cells into the lumen (refer to paragraph (g)(11) of this guideline). Examinationof the intact epididymis should include the caput, corpus, and cauda, which can be accomplishedby evaluation of a longitudinal section, and should be conducted in order toidentify such lesions as sperm granulomas, leukocytic infiltration (inflammation), aberrantcell types within the lumen, or the absence of clear cells in the cauda epididymal epithelium.The postlactational ovary should contain primordial and growing follicles as well as the largecorpora lutea of lactation. Histopathological examination should detect qualitative depletionof the primordial follicle population. A quantitative evaluation of primordial follicles shouldbe conducted for F1 females; the number of animals, ovarian section selection, and sectionsample size should be statistically appropriate for the evaluation procedure used. Examinationshould include enumeration of the number of primordial follicles, which can be combinedwith small growing follicles (see paragraphs (g)(1) and (g)(2) of this guideline), forcomparison of treated and control ovaries.© 2006 by Taylor & Francis Group, LLC

790 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION(f)(g)(ii) Weanlings. For F1 and F2 weanlings, histopathological examination of treatment-related abnormalitiesnoted at macroscopic examination should be considered, if such evaluation weredeemed appropriate and would contribute to the interpretation of the study data.Data and reporting(1) Treatment of results. Data should be reported individually and summarized in tabular form, showingfor each test group the types of change and the number of animals displaying each type ofchange.(2) Evaluation of study results.(i)(ii)(iii)(iv)An evaluation of test results, including the statistical analysis, should be provided. Thisshould include an evaluation of the relationship, or lack thereof, between the exposure of theanimals to the test substance and the incidence and severity of all abnormalities.When appropriate, historical control data should be used to enhance interpretation of studyresults. Historical data, when used, should be compiled, presented, and analyzed in an appropriateand relevant manner. In order to justify its use as an analytical tool, information suchas the dates of study conduct, the strain and source of the animals, and the vehicle and routeof administration should be included.Statistical analysis of the study findings should include sufficient information on the methodof analysis, so that an independent reviewer/statistician can reevaluate and reconstruct theanalysis.In any study which demonstrates an absence of toxic effects, further investigation to establishabsorption and bioavailability of the test substance should be considered.(3) Test report. In addition to the reporting requirements as specified under 40 CFR part 792, subpartJ and 40 CFR part 160, subpart J, the following specific information should be reported. Both individualand summary data should be presented.(i)(ii)(iii)(iv)(v)(vi)(vii)(viii)(ix)(x)Species and strain.Toxic response data by sex and dose, including indices of mating, fertility, gestation, birth,viability, and lactation; offspring sex ratio; precoital interval, including the number of daysuntil mating and the number of estrous periods until mating; and duration of gestation calculatedfrom day 0 of pregnancy. The report should provide the numbers used in calculating allindices.Day (week) of death during the study or whether animals survived to termination; date (age)of litter termination.Toxic or other effects on reproduction, offspring, or postnatal growth.Developmental milestone data (mean age of vaginal opening and preputial separation, andmean anogenital distance, when measured).An analysis of P and F1 females cycle pattern and mean estrous cycle length.Day (week) of observation of each abnormal sign and its subsequent course.Body weight and body weight change data by sex for P, F1, and F2 animals.Food (and water, if applicable) consumption, food efficiency (body weight gain per gram offood consumed), and test material consumption for P and F1 animals, except for the periodof cohabitation.Total cauda epididymal sperm number, homogenization-resistant testis spermatid number,number and percent of progressively motile sperm, number and percent of morphologicallynormal sperm, and number and percent of sperm with each identified anomaly.(xi) Stage of the estrous cycle at the time of termination for P and F1 parental females.(xii) Necropsy findings.(xiii) Implantation data and postimplantation loss calculations for P and F1 parental females.(xiv) Absolute and adjusted organ weight data.(xv) Detailed description of all histopathological findings.(xvi) Adequate statistical treatment of results.(xvii) A copy of the study protocol and any amendments should be included.References. The following references should be consulted for additional background information onthis test guideline:(1) Bolon, B. et al. Differential follicle counts as a screen for chemically induced ovarian toxicity in mice:results from continuous breeding bioassays. Fundamental and Applied Toxicology 39:1-10 (1997).© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 791(2) Bucci, T.J. et al. The effect of sampling procedure on differential ovarian follicle counts. ReproductiveToxicology 11(5): 689-696 (1997).(3) Gray, L.E. et al. A dose-response analysis of methoxychlor-induced alterations of reproductive developmentand function in the rat. Fundamental and Applied Toxicology 12:92–108 (1989).(4) Heindel, J.J. and R.E. Chapin, (eds.). Part B. Female Reproductive Systems, Methods in Toxicology,Academic, Orlando, FL (1993).(5) Heindel, J.J. et al. Histological assessment of ovarian follicle number in mice as a screen of ovariantoxicity. In: Growth Factors and the Ovary, A.N. Hirshfield (ed.), Plenum, NY, pp. 421–426(1989).(6) Korenbrot, C.C. et al. Preputial separation as an external sign of pubertal development in the malerat. Biology of Reproduction 17:298–303 (1977).(7) Linder, R.E. et al. Endpoints of spermatoxicity in the rat after short duration exposures to fourteenreproductive toxicants. Reproductive Toxicology 6:491–505 (1992).(8) Manson, J.M. and Y.J. Kang. Test methods for assessing female reproductive and developmentaltoxicology. In: Principles and Methods of Toxicology, A.W. Hayes (ed.), Raven, New York (1989).(9) Organization for Economic Cooperation and Development, No. 416: Two Generation ReproductionToxicity Study, Guidelines for Testing of Chemicals. [C(83)44 (Final)] (1983).(10) Pederson, T. and H. Peters. Proposal for classification of oocytes and follicles in the mouse ovary.Journal of Reproduction and Fertility 17:555–557 (1988).(11) Russell, L.D. et al. Histological and Histopathological Evaluation of the Testis, Cache River,Clearwater, FL (1990).(12) Sadleir, R.M.F.S. Cycles and seasons, In: Reproduction in Mammals: I. Germ Cells and Fertilization,C.R. Auston and R.V. Short (eds.), Cambridge, NY (1979).(13) Seed, J., R.E. Chapin, E.D. Clegg, L.A. Dostal, R.H. Foote, M.E. Hurtt, G.R. Klinefelter, S.L.Makris, S.D. Perreault, S. Schrader, D. Seyler, R. Sprando, K.A. Treinen, D.N.R. Veeramachaneni,and L.D. Wise. Methods for assessing sperm motility, morphology, and counts in the rat, rabbit,and dog: a consensus report. Reproductive Toxicology 10(3):237–244 (1996).(14) Smith, B.J. et al. Comparison of random and serial sections in assessment of ovarian toxicity. ReproductiveToxicology 5:379–383 (1991).(15) Thomas, J.A. Toxic responses of the reproductive system. In: Casarett and Doull’s Toxicology,M.O. Amdur, J. Doull, and C.D. Klaassen (eds.), Pergamon, NY (1991).(16) U.S. Environmental Protection Agency. OPP Guideline 83–4: Reproductive and Fertility Effects.Pesticide Assessment Guidelines, Subdivision F, Hazard Evaluation: Human and Domestic Animals.Office of Pesticides and Toxic Substances, Washington, DC, EPA–540/9–82–025 (1982).(17) U.S. Environmental Protection Agency. Subpart E—Specific Organ/Tissue Toxicity, 40 CFR798.4700: Reproduction and Fertility Effects.(18) U.S. Environmental Protection Agency. Health Effects Test Guidelines, OPPTS 870.3250, 90-DayDermal Toxicity, July 1998.(19) U.S. Environmental Protection Agency. Health Effects Test Guidelines, OPPTS 870.3465, 90-DayInhalation Toxicity, July 1998.(20) U.S. Environmental Protection Agency. Reproductive Toxicity Risk Assessment Guidelines. FederalRegister 61 FR 56274–56322 (1996)(21) Working, P.K. and M. Hurtt. Computerized videomicrographic analysis of rat sperm motility.Journal of Andrology 8:330–337 (1987).(22) Zenick, H. et al. Assessment of male reproductive toxicity: a risk assessment approach. In: Principlesand Methods of Toxicology, A.W. Hayes (ed.), Raven, NY (1994).© 2006 by Taylor & Francis Group, LLC

792 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONAPPENDIX 3REFERENCES FOR GUIDELINES AND OTHER INFORMATIONPHARMACEUTICALS INFORMATION INTENDED FOR INTERNATIONALUSE INTERNATIONAL CONFERENCE ON HARMONIZATION (ICH)International Conference on Harmonisation of Technical Requirements for the Registration ofPharmaceuticals for Human Use. ICH Harmonised Tripartite Guideline; Detection of Toxicity toReproduction for Medicinal Products, June 24, 1993, Presented at the Second International Conferenceon Harmonisation, Orlando, Florida, October 27-29, 1993 (Provided in Appendix 1).Bass, R., Ulbrich, B., Hildebrandt, A., Weissinger, J., Doi, O., Baeder, C., Fumero, S., Harada, Y.,Lehmann, H., Manson, J., Neubert, D., Omori, Y., Palmer, A., Sullivan, F., Takayama, S., Tanimura,T., Draft guideline on detection of toxicity to reproduction for medical products, Adverse DrugReact. Toxicol., Rev.9, 127, 1991.World Health Organization (WHO)World Health Organization (WHO), Principles for the Testing of Drugs for Teratogenicity, WHOTech. Rept. Series, No. 364, Geneva, 1967.Information Intended for Local UseCanadaCanada, Health Protection Branch, The Testing of Chemicals for Carcinogenicity, Mutagenicity andTeratogenicity, Ministry of Health and Welfare, Canada, 1977.Canada, Bureau of Drugs, Health Protection Branch, Health and Welfare, Draft of PreclinicalToxicologic Guidelines, Section 2.4, 35, 1979.Drug Dictorate Guidelines, Toxicological Evaluation 2.4, Reproductive Studies, Ministry ofNational Health and Welfare, Health Protection Branch, Health and Welfare, Canada, 1990.European Economic Council (EEC)European Economic Community. Council recommendation of 26 October 1983 concerning testsrelating to the placing on the market of proprietary medicinal products. Official Journal of theEuropean Communities, No. L 332 (83/571/EEC), November 28, 1983.European Economic Community, The Rules Governing Medicinal Products in the European Community,Vol. III, Guidelines on the Quality, Safety and Efficacy of Medicinal Products for HumanUse. Office of Official Publications of the European Communities, Brussels, 1988.Great Britain/United Kingdom (U.K.) — Committee on the Safety of Medicines (CSM)Committee on the Safety of Medicines (CSM), Notes for Guidance on Reproduction Studies,Department of Health and Social Security, Great Britain, 1974.Japan — Ministry of Health and Welfare (MHW)Ministry of Health and Welfare, Japan (MHW), On Animal Experimental Methods for Testing theEffects of Drugs on Reproduction, Notification No. 529 of the Pharmaceutical Affairs Bureau,Ministry of Health and Welfare, March 31, 1975.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 793Ministry of Health and Welfare, Japan (MHW), Information on the Guidelines of Toxicity StudiesRequired for Applications for Approval to Manufacture (Import) Drugs, Notification No. 118 ofthe Pharmaceutical Affairs Bureau, Ministry of Health and Welfare, February 15, 1984.Tanimura, T., Kameyama, Y., Shiota, K., Tanaka, S., Matsumoto, N. and Mizutani, M., Report onthe review of the guidelines for studies of the effect of drugs on reproduction. Notification No.118, Pharmaceutical Affairs Bureau, Ministry of Health and Welfare, Japan, 1989.Pharmaceutical Manufacturers Association (PMA)Pharmaceutical Manufacturers Association (PMA), Second Workshop in Teratology. Supplement tothe Teratology Workshop Manual (A collection of lectures and demonstrations from the SecondWorkshop in Teratology held January 25-30, 1965, Berkeley, California), Pharmaceutical ManufacturersAssociation, Washington, D.C., 1965.PMA Guidelines: Reproduction, teratology and pediatrics. Contributors: Christian, M., Diener, R.,Hoar, R. and Staples, R., 1981 (revised)USA — Food and Drug Administration (FDA)Reproductive ToxicityU.S. Food and Drug Administration (1959) Appraisal of the Safety of Chemicals in Foods, Drugs,and Cosmetics, third printing: 1975, the Association of Food and Drug Officials of the UnitedStates, Washington, DC.U.S. Food and Drug Administration (FDA), Guidelines for Reproduction Studies for Safety Evaluationof Drugs for Human Use, Washington, D.C., 1966.U.S. Food and Drug Administration (FDA). International Conference on Harmonisation; Guideline ondetection of toxicity to reproduction for medicinal products. Federal Register, Vol. 59, No. 183, 1994.U.S. Food and Drug Administration (FDA). International Conference on Harmonisation; Guidelineon detection of toxicity to reproduction for medicinal products; Addendum on Toxicity to malefertility. Federal Register, April 5, 1996, Vol. 61, No. 67, 1996.U.S. Food and Drug Administration (FDA). International Conference on Harmonisation; Maintenanceof the ICH guideline on toxicity to male fertility; An addendum to the ICH tripartite guidelineon detection of toxicity to reproduction for medicinal products. Amended November 9, 2000.Good Laboratory Practices (GLPS)U.S. Food and Drug Administration. Good Laboratory Practice Regulations; Final Rule. 21 CFRPart 58.USA — Interagency Regulatory Group (IRLG)Interagency Regulatory Group (IRLG): Recommended guideline for teratogenicity studies in therat, mouse, hamster or rabbit, April, 1980.Recommended Guideline for Teratogenicity Studies in Rat, Mouse, Hamster or Rabbit. InteragencyRegulatory Liaison Group Testing Standards and Guidelines Work Group, US Consumer productsafety commission, US Environmental Protection Agency, Food and Drug Administration (US Dept.of HHS), Food Safety and Quality Service (US Dept. of Agriculture) and Occupational Safety andHealth Administration (US Dept. of Labor), 2, 1981.© 2006 by Taylor & Francis Group, LLC

794 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFoods — Information Intended for Local UseUSA — Food and Drug Administration (FDA)U.S. Food and Drug Administration (1959) Appraisal of the Safety of Chemicals in Foods, Drugs,and Cosmetics, third printing: 1975, the Association of Food and Drug Officials of the UnitedStates, Washington, DC.U.S. Food and Drug Administration (FDA), Advisory Committee on Protocols for Safety Evaluations:Panel on Reproduction. Report on reproduction studies in the safety evaluation of foodadditives and pesticide residues, Toxicol. Applied Pharmacol., 16, 264, 1970.U.S. Food and Drug Administration (FDA). Toxicological Principles and Procedures for PriorityBased Assessment of Food Additives. (Red Book), Guidelines for Reproduction Testing with aTeratology Phase, Bureau of Foods, U.S. Food and Drug Administration, Washington, D.C., 1982,pp. 80-117.U.S. Food and Drug Administration (FDA): Toxicological principles for the safety assessment ofdirect food additives and color additives uses in food. National Technical Information Services,PB83-170696, 1982, pp. 80-119.U.S. Food and Drug Administration (FDA). Draft — Toxicological Principles for the Safety Assessmentof Direct Food Additives and Color Additives Used in Food. (Red Book II), Guidelines forReproduction and Developmental Toxicity Studies, Center for Food Safety and Applied Nutrition,U.S. Food and Drug Administration, Washington, D.C., 1993, pp. 123-134.U.S. Food and Drug Administration (FDA). Toxicological Principles for the Safety of Food Ingredients(Redbook 2000): Sections issued electronically July 7, 2000 at http:// vm.cfscan.fda.gov/~Redbook/Red-tcct.html,2000.USA — National Academy of Sciences (NAS)National Academy of Sciences, Food Protection Committee, Evaluating the Safety of Food Chemicals,National Academy of Sciences, Washington, D.C., 1970.USA — National Toxicology Program (NTP)Collins, T.F.X., Reproduction and teratology guidelines: Review of deliberations by the NationalToxicology Advisory Committee’s Reproduction Panel, J. Environ. Path. Toxicol., 2, 141, 1978.Chemicals — Information Intended for International UseOrganization on Economic Cooperation and Development (OECD)Reproductive ToxicityOrganization for Economic Cooperation and Development (OECD): Final Report- OECD shorttermand long-term toxicology groups, Teratogenicity, December 31, 1979, 110.Organization for Economic Cooperation and Development. Guidelines for Testing of Chemicals.Section 4, No. 414: Teratogenicity, 22 January 2001.Organization for Economic Cooperation and Development. First Addendum to OECD Guidelinesfor Testing of Chemicals. Section 4, No. 415: One-Generation Reproduction Toxicity, 1983.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 795Organization for Economic Cooperation and Development. Guidelines for Testing of Chemicals.Section 4, No. 416: Two Generation Reproduction Toxicity Study, 22 January 2001.Organization for Economic and Co-operative Development (OECD) Screening Information DataSet (SIDS). Guideline for Testing of Chemicals, Section 4, No. 421: Reproduction/DevelopmentalToxicity Screening Test, 27 July 1995.Good Laboratory PracticesOrganization for Economic Cooperation and Development. Good Laboratory Practice in the Testingof Chemicals, Final Report of the OECD Expert Group on Laboratory Practices [C(81) 3.0 (Final),Annex 2], 1981.World Health Organization (WHO)Food and Agriculture Organization of the United Nations (FAO), Working Party on PesticideResidues. Pesticide Residues in Food, World Health Organization, Geneva, 1990.World Health Organization (WHO), Monograph of the Joint Meeting on Pesticide Residues,Geneva, 1975.INFORMATION INTENDED FOR LOCAL USEJapan — Ministry of Agriculture, Forestry and Fisheries (MAFF)Reproductive ToxicityJapanese Ministry of Agriculture, Forestry and Fisheries: Procedure for performing toxicity testson chemicals, 1982.Japanese Ministry of Agriculture, Forestry and Fisheries. Guidance on Toxicology Study Data forApplication of Agricultural Chemical Registration. Notification No. 4200, January 28, 1985.Good Laboratory Practices (GLPS)Japanese Ministry of Agriculture, Forestry and Fisheries. Good Laboratory Practice Regulations.59 Nohsan No. 3850, 1984.USA — Environmental Protection Agency (EPA)Reproductive ToxicityChristian, M.S. and Voytek, P.E., In Vivo Reproductive and Mutagenicity Tests. EnvironmentalProtection Agency, Washington, D.C. National Technical Information Service, U.S. Department ofCommerce, Springfield, VA 22161, 1982.U.S. Environmental Protection Agency: Guidelines for the Health Assessment of Suspect DevelopmentalToxicants. Federal Register, Vol. 51, No. 185 34027-34040. Sept. 14, 1986.U.S. Environmental Protection Agency Workshop on One vs. Two-Generation Reproductive EffectsStudies, Sponsored by the USEPA Risk Assessment Forum, Washington, D.C., 1987.U.S. Environmental Protection Agency Proposed Guidelines for Assessing Male Reproductive Risk.Fed. Reg. 53:24850-24869, 1988.© 2006 by Taylor & Francis Group, LLC

796 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONU.S. Environmental Protection Agency Proposed Guidelines for Assessing Female ReproductiveRisk. Fed. Reg. 53:24834-24847, 1988.U.S. Environmental Protection Agency Health Effects Division Peer Review of Procymidone, Officeof Pesticide Programs, October 31, 1990.U.S. Environmental Protection Agency (EPA-FIFRA), Pesticide Assessment Guideline. SubdivisionF — Hazard Evaluation: Human and Domestic Animals, Addendum 10, Neurotoxicity, HealthEffects Division, Office of Pesticide Programs, 540/09-91-123, PB 91-154617, March, 1991.U.S. Environmental Protection Agency Guidelines for Developmental Toxicity Risk Assessment.Fed. Reg. 56(234):63798-63826, 1991.U.S. Environmental Protection Agency Memorandum: Atrazine Two-Generation Study, HED Adhoc Committee for Atrazine Reproductive Issues, for a Discussion of the Setting of a NOEL in aReproduction Study with Varying Mean Litter Sizes, September 23, 1992.U.S. Environmental Protection Agency Draft Guidelines for Reproductive Toxicity Risk Assessment,1993.USA — Environmental Protection Agency (EPA) — Office of Pesticides and ToxicSubstances (OPTS) or Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA)Reproductive ToxicityU.S. Environmental Protection Agency, OPP Guideline 83-4: Reproductive and Fertility Effects.Pesticide Assessment Guidelines, Subdivision F, Hazard Evaluation: Human and Domestic Animals.Office of Pesticides and Toxic Substances, Washington D.C., EPA-540/9-82-025, 1982.U.S. Environmental Protection Agency, Pesticide Assessment Guidelines. Subdivision F. HazardEvaluation: Humans and Domestic Animals, Nov. 1984.U.S. Environmental Protection Agency, Pesticide Assessment Guidelines. Subdivision F — HazardEvaluation: Human and Domestic Animals, November, 1984 (Revised Edition), 1984.U.S. Environmental Protection Agency, Hazard Evaluation Division Standard Evaluation Procedure:Teratology Studies. Office of Pesticide Programs, Washington, DC. EPA-540/9-85-018, 1985.U.S. Environmental Protection Agency, PR Notice 86-5, Standard Format For Data SubmittedUnder the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) and Certain Provisions ofthe Federal Food, Drug and Cosmetic Act (FFDCA), Issued July 29, 1986.U.S. Environmental Protection Agency, Health Effects Division Standard Evaluation Procedure.Inhalation Toxicity Testing. Office of Pesticide Programs, Washington, DC, EPA-540/098-88-101,1988.U.S. Environmental Protection Agency, FIFRA Accelerated Reregistration Phase 3 Technical Guidance,Office of Pesticides and Toxic Substances, Washington, DC, EPA No. 540/09-90-078, 1988.U.S. Environmental Protection Agency, (EPA-FIFRA), Pesticide Assessment Guideline. SubdivisionF — Hazard Evaluation: Human and Domestic Animals, Addendum 10, Neurotoxicity, HealthEffects Division, Office of Pesticide Programs, March, 1991.U.S. Environmental Protection Agency, Pesticide Assessment Guidelines, Subdivision F, hazardEvaluation: Human and Domestic Animals, Addendum 10, Neurotoxicity, Series 81, 82 and 83,March, EPA 540/09-91-123, PB 91-154617, 1991.© 2006 by Taylor & Francis Group, LLC

PERSPECTIVES ON THE DEVELOPMENTAL AND REPRODUCTIVE TOXICITY GUIDELINES 797U.S. Environmental Protection Agency, Code of Federal Regulations, Title 40 (40CFR) — Protectionof Environment, Subchapter E — Pesticide Programs, Part 158, Data Requirements for Registration,88, 1992.U.S. Environmental Protection Agency, Health Effects Division Draft Standard Evaluation Procedure,Developmental Toxicity Studies, Office of Pesticide Programs, Washington, DC, 1993.Good Laboratory Practices (GLPS)U.S. Environmental Protection Agency, Federal Insecticide, Fungicide and Rodenticide Act(FIFRA); Good Laboratory Practice Standards; Final Rule. 40 CFR Part 160.USA — Environmental Protection Agency (EPA) — Toxic Substances Control Act(TSCA)Reproductive ToxicologyU.S. Environmental Protection Agency, Subpart E — Specific Organ/Tissue Toxicity, No. 798.4900:Developmental Toxicity Study. 40 CFR Part 798 — Toxic Substances Control Act Test Guidelines;Final Rules. Fed. Regist. 50:39433-39434, 1985.U.S. Environmental Protection Agency, Toxic Substances Control Act Test Guidelines, Final Rules.CFR 40 Part 798, 1985.U.S. Environmental Protection Agency, Subpart E — Specific Organ/Tissue Toxicity, No. 798.4700:Reproduction and Fertility Effects. 40 CFR Part 798 — Toxic Substances Control Act Test Guidelines;Final Rules. Fed. Regist. 50:39432-39433, 1985.U.S. Environmental Protection Agency, Toxic Substances Control Act, Part 798 — Health EffectsTesting Guidelines. Subpart E — Specific Organ/Tissue Toxicity, 798.4700, Reproduction andFertility Effects. Fed. Regist. 50(188):3942-39433, 1985.U.S. Environmental Protection Agency, Toxic Substances Control Act Test Guidelines; Final Rule.Fed. Regist., September 27, 1985. Part II, Vol. 50, No. 188, 1985.Good Laboratory Practices (GLPS)U.S. Environmental Protection Agency. Toxic Substances Control Act (TSCA); Good LaboratoryPractice Standards; Final Rule. 40 CFR Part 792.USA — Environmental Protection Agency (EPA) — Office of Drinking Water (ODW)U.S. Environmental Protection Agency Office of Drinking Water Health Advisories, Drinking WaterHealth Advisory: Pesticides, Lewis Publishers, Michigan, 1989.USA — National Research Council (NRC)National Research Council (NRC) Reproduction and Teratogenicity Tests. In Principles and Proceduresfor Evaluating the Toxicity of Household Substances, National Academy Press, Washington,D.C., 1977.National Research Council (NRC), Nutrient Requirements of Laboratory Animals, Vol. 10, NationalAcademy Press, Washington, D.C., 1978.© 2006 by Taylor & Francis Group, LLC

798 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONNational Research Council (NRC), Biologic Markers in Reproductive Toxicology: Subcommitteeon Reproductive and Neurodevelopmental Toxicology Committee on Biological Markers, NationalAcademy Press, Washington, D.C., 1989.National Research Council (NRC), Pharmacokinetics: Defining Dosimetry for Risk Assessment,National Academy of Sciences Workshop, Washington, D.C., March 4-5, 1992.National Research Council (NRC), Risk Assessment in the Federal Government: Managing theProcess. National Academy Press, Washington, DC, p 159. ISBN 0-309-03979-7, 1983.USA — Office of Technical Assessment (OTA)U.S. Congress, Office of Technical Assessment (OTA), Reproductive Health Hazards in the Workplace,Chapter 4, Evidence of Workplace Hazards to Reproductive Function, J.B. Lippincott Company,Philadelphia, PA, 1988.USA — President’s Science Advisory Committee (PSAC)President’s Science Advisory Committee, Use of Pesticides, Government Printing Office, Washington,D.C., 1963.© 2006 by Taylor & Francis Group, LLC

CHAPTER 20Human Studies — Epidemiologic Techniques inDevelopmental and Reproductive ToxicologyBengt KällénCONTENTSI. Introduction ........................................................................................................................801II. Adverse Reproductive Outcomes to Study........................................................................802A. Infertility and Subfertility..........................................................................................8021. Definition of Infertility and Subfertility..............................................................8022. Time to Pregnancy...............................................................................................8023. Clinical Infertility and Subfertility......................................................................8034. Study Design........................................................................................................803B. Spontaneous Abortion (Miscarriage).........................................................................8041. Definition .............................................................................................................8042. Rates and Confounding .......................................................................................8053. Study Design........................................................................................................806C. Infant and Child Death ..............................................................................................8061. Definition .............................................................................................................8062. Rates and Confounding .......................................................................................8073. Classification........................................................................................................8074. Statistical Analysis...............................................................................................808D. Birth Weight, Pregnancy Length, and Growth Retardation......................................8081. Definition .............................................................................................................8082. Rates and Confounding .......................................................................................8103. Analysis of Prenatal Growth ...............................................................................810E. Congenital Malformations.........................................................................................8101. Definition .............................................................................................................8102. Classification and Grouping ................................................................................8113. Mutagenesis and Teratogenesis ...........................................................................8114. Source of Information on Congenital Malformations ........................................8125. Rates and Confounding .......................................................................................813F. Mental Retardation and Behavioral Problems ..........................................................8131. Definition .............................................................................................................8132. Sources of Information........................................................................................8133. Confounding ........................................................................................................814799© 2006 by Taylor & Francis Group, LLC

800 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIII.G. Childhood Cancer ......................................................................................................8151. Definition and Classification ...............................................................................8152. Sources of Information........................................................................................815Methods to Record Exposures...........................................................................................815A. Embryonic and Maternal Exposure...........................................................................815B. Estimates of Maternal Exposure ...............................................................................8151. Concerns about Exposure Data ...........................................................................8152. Information on Drug Exposure ...........................................................................8173. Information on Occupational Exposure ..............................................................8184. Information on General Environmental Exposure ..............................................8185. Misclassification of Exposure .............................................................................819C. Prospective and Retrospective Exposure Information ..............................................819IV. Methods to Record Outcomes ...........................................................................................819A. Questionnaire and Interview Information .................................................................819B. Medical Record Studies.............................................................................................820C. Register Studies .........................................................................................................820V. Case Control Studies..........................................................................................................820A. Principles....................................................................................................................820B. Selection of Cases .....................................................................................................821C. Control Definition......................................................................................................821D. Selection of Controls .................................................................................................822E. Statistical Analysis of Case Control Materials .........................................................8231. Unmatched Studies ..............................................................................................8232. Stratified Analysis — A Method to Control for Confounding...........................8253. Logistic Multiple Regression Analysis ...............................................................8264. Analysis of Matched Pairs or Triplets ................................................................8265. The Multiple Testing Problem.............................................................................828VI. Cohort Studies....................................................................................................................828A. Principles....................................................................................................................828B. The Formation of Cohorts.........................................................................................829C. Comparison Material .................................................................................................829D. Statistical Analysis of Cohort Data...........................................................................830VII. Nested Case Control Studies .............................................................................................831VIII.Confounding.......................................................................................................................832A. General Considerations..............................................................................................832B. Maternal Age and Reproductive History ..................................................................832C. Maternal Diseases......................................................................................................832D. Social Characteristics ................................................................................................833E. Biological Characteristics..........................................................................................833F. Incomplete Compensation of a Confounder .............................................................833IX. Surveillance........................................................................................................................834A. Birth Defects Surveillance in the Population............................................................834B. Surveillance after Specific Exposures.......................................................................835C. Outcome-Exposure Surveillance ...............................................................................835D. Statistical Considerations ..........................................................................................835X. Cluster Analysis .................................................................................................................836A. Definition of a Cluster...............................................................................................836B. Clusters as Chance Phenomena or Artifacts .............................................................836C. Exposure Analysis of Clusters ..................................................................................836D. Clusters and Statistics................................................................................................836XI. Concluding Words..............................................................................................................837References ......................................................................................................................................838© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 801I. INTRODUCTIONEpidemiological studies of human reproduction are important sources of information on developmentaltoxicity for the human. Because of species variability in developmental processes and inresponses to different harmful factors, the ultimate answer to questions on risks to humans canonly be obtained by information on humans. Controlled experiments are usually out of the questionfor practical and ethical reasons. The most important source of information will therefore beepidemiology: studies of possible statistical associations between specific exposures and reproductiveoutcome. Such associations do not necessarily prove causality because they may be due tounidentified or uncontrolled confounders, to bias, or to chance.In rare situations, when a very high risk exists, formal epidemiological studies may not benecessary in order to identify a risk. Examples are thalidomide and rubella embryopathy. Both ofthese important human teratogens were detected by alert clinicians. For such effects to be found,however, the risk increase associated with the exposure must be huge, which means that the effectmust be highly unusual. In most instances, however, we are dealing with small or moderate riskincreases (e.g., anticonvulsant drugs, maternal smoking). It is then necessary to make strict epidemiologicalanalyses, but it is easy to draw wrong conclusions because most such studies have someimperfection and are burdened with uncertainty.High risks are cause for concern for the individual — examples are thalidomide, isotretinoin,rubella, and maternal alcoholism. Low risks may be of little concern for any one exposed individual,but if many women are exposed, the number of reproductive failures caused by the factor may belarge anyway. Maternal smoking is an example. It has been shown to cause reproductive anomalies,including an increased risk for perinatal deaths. But the risk increase is moderate and does notnoticeably affect the probability that a specific pregnancy will end with a perinatal death (perhapsan increase from 0.6% to 0.9% in Sweden). Nevertheless, it has been estimated that some 10% ofall perinatal deaths are related to maternal smoking. 1From the individual’s point of view, absence of a demonstrable risk may be enough. Epidemiologicalstudies can never prove a lack of reproductive toxicity; they can only show that the riskis below a certain level. A low risk, perhaps not detectable even with a number of well-performedepidemiological studies, may nevertheless represent a major cause of birth defects or other reproductivefailures in the population. This also means that society’s decision about possible risks fromspecific exposures must sometimes be based not only on epidemiological data (notably when theseare negative), but also on other types of data, including animal toxicological tests.The basic question which the epidemiological study tries to answer is this: Is there an associationbetween a specific exposure (in the widest sense of the word) and an outcome; if so, can thisassociation be explained by chance, by the action of confounding factors (for a discussion, seeSection VIII), or by bias? Such associations can be estimated with different techniques, the twoclassical ones being the case control (or case referent) and cohort approaches. The first comparesexposures in “cases” (the abnormal reproductive outcome under study) and “controls” (lacking thatabnormal reproductive outcome). With this design, many exposures can be studied simultaneouslyin relation to the selected outcome. The cohort approach studies reproductive outcome in parentswho were exposed in a specific way and the outcome is compared with that of unexposed cohortsor the total population; with this approach, many different abnormal outcomes can be studied aftera specific exposure. In both methods, two variables must be defined: reproductive outcome andexposure.In the analysis and interpretation of all epidemiological studies, two main problems exist:confounding and bias. Confounding means that a third factor covaries with both exposure andoutcome, resulting in a statistical association between the two. Bias means that the informationon exposure is influenced by the outcome or vice versa. These problems are discussed later inthe text.© 2006 by Taylor & Francis Group, LLC

802 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONII. ADVERSE REPRODUCTIVE OUTCOMES TO STUDYA disturbance of reproductive function can result in a multitude of different adverse outcomes orendpoints for study. It is not necessary that a specific exposure causes all types of adverse outcomes.Some result in a very broad spectrum of disturbances, e.g., maternal alcoholism, which causesembryonic, fetal, and perinatal death, growth retardation, major and minor congenital malformations,and mental disturbances, including mental retardation. 2 Other agents may have a ratherspecific effect on only one or a few types of congenital malformations, e.g., the possible effect oflithium in increasing the risk for a congenital heart defect. 3 The full evaluation of the reproductivetoxicology of an agent therefore necessitates study of many different aspects of reproductiveoutcome. Even after such studies, it is always possible that the exposure results in a late, unexpectedeffect, as was the case with DES (diethylstilbestrol). 4A. Infertility and Subfertility1. Definition of Infertility and SubfertilityThe fertility of a couple will be a function of both male and female fertility. The fertility rate in apopulation will be affected by voluntary reduction of family size with the aid of contraceptivepractice and induced abortions.Female fertility depends on normal ovulation of a normal egg, its unimpaired passage through thefallopian tube to the uterus, thinning of the mucus surrounding the mouth of the cervix enabling thesperm to pass into the uterine cavity, and changes in the uterine lining enabling the fertilized egg toimplant. These processes are under endocrine regulation from the pituitary gland and the ovaries.Male fertility depends on the production of adequate numbers of healthy sperm in the testesand on the ability to achieve erection and ejaculation. Sperm production is under endocrine controlfrom the pituitary gland, and testicular hormone production is necessary for normal male genitalfunction.Infertility occurs when a couple is unable (without medical intervention) to have a child; thisis a rather frequent phenomenon. Ten percent or more of all couples trying to conceive will seekmedical help. In about half of the couples, male infertility is the cause and in about half, it is femaleinfertility.Modern medical technology often makes it possible to help these couples have children.Treatment will depend on the cause of the infertility. Examples are surgical opening of occludedfallopian tubes, fertility drug stimulation of the ovary and/or ovulation, and various types of invitro fertilization (IVF). Notably for male infertility, a special form of IVF called ICSI (intracytoplasmicsperm injection) can be used. The egg is then fertilized by one sperm cell, introduced intothe egg in vitro with micromanipulation.Subfertile couples have a reduced capability to achieve a pregnancy. They may have a pregnancyafter a longer delay than is usually the case. It is a matter of clinical judgement when a subfertilecouple is judged to be infertile—often 2 to 3 years of involuntary childlessness is needed beforeIVF procedures, for instance, are recommended.Exogenous causes may harm either the male or the female reproductive organs. Classicalexamples are male occupational exposure to dibromochloropropane, which resulted in a (temporary)arrest of spermatogenesis, 5 and prenatal female exposure to progestational agents or diethylstilbestrol,which has been said to affect future fertility. 62. Time to PregnancyA popular way to study an effect on fertility is to use the time taken to achieve pregnancy (TTP).This means the time (usually measured in months or menstrual cycles) it takes for a woman who© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 803is actively trying to conceive to become pregnant. Usually this means that she becomes aware ofher pregnancy. We will later see that a pregnancy may cease as a very early miscarriage before thewoman realizes she is pregnant. Such pregnancies will not be detected, but the delay in conceptionwill be registered as an extended TTP.Usually estimates of TTP are based on prospective or retrospective interview or questionnaireinformation. A large number of variables, some of which are difficult to control for, will affectTTP measurement, and great care has to be used in the interpretation of such data. 7–93. Clinical Infertility and SubfertilityStudies on infertility are usually made on populations identified in fertility clinics. Information onclinical subfertility (e.g., having experienced involuntary childlessness for at least 1 year) is usuallybased on interview or questionnaire data, even though such information is available in sources suchas the Swedish Medical Birth Registry.Related to this problem is the use of fertility drugs and IVF. Couples treated in this way representselected subfertility groups. That is because they did not achieve pregnancy without medical help,and access to these methods may differ according to socioeconomic characteristics. Identificationof women who underwent IVF has been made in different ways, e.g., from existing registries ofall IVF pregnancies. 104. Study DesignFertility problems can be of interest both as outcome measures and as risk factors. Does the existenceof a fertility problem affect the outcome of a pregnancy that has occurred, with or without medicalintervention? Does the outcome of pregnancies occurring after fertility drug or IVF treatmentsdiffer from the expected outcome?The presence of information on length of involuntary childlessness in the Swedish MedicalBirth Registry made it possible to describe epidemiological characteristics of affected women. Theperiod of involuntary childlessness increases with maternal age and (of course) decreases withparity, and it is associated with maternal smoking, previous spontaneous abortions, extrauterinepregnancies, and stillbirths. Previous induced abortions appears as a “protective factor” becausesubfertile couples are less likely to have an unwanted pregnancy. 11When subfertility was looked upon as a risk factor and delivery outcome was studied, anincreased rate of preterm births, low birth weight infants, and intrauterine growth retardation wasnoticed among singleton infants. Perinatal mortality was increased, but this was mainly explainableby the maternal age and parity distribution. 11 An association between the risk for spontaneousabortion and previous subfertility has also been described. 12Studies of infants born after IVF show clear-cut deviations from the expected outcome. Themost obvious effect is an increase in multiple births (because of more than one embryo beingtransferred), but effects were also seen on preterm birth, low birth weight, intrauterine growthretardation, and presence of congenital malformations. These effects are explained mainly andperhaps exclusively by the subfertility factor itself and not by the IVF procedure. 10A theoretically interesting finding is an increased risk for infant hypospadias after ICSI but notafter standard IVF. 13 The link is probably male subfertility due to a genetic defect resulting in bothtesticular anomalies and an increased risk for infant hypospadias.One study from the Swedish Medical Birth Registry used data on the rate of women whoreported involuntary childlessness to demonstrate that an increase in fertility had occurred duringone decade (1983 to 1993). 14 Great care must be used in the interpretation of this type of information,however, as recording of involuntary childlessness is incomplete and may well vary with time.When the possible effects of various exposures have been studied, sperm counts ormorphology 15,16 or TTP 17 have sometimes been assessed. Other investigators have studied exposures© 2006 by Taylor & Francis Group, LLC

804 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION908070Number of miscarriages60504030201004 8 12 16Pregnancy week20 24 28Figure 20.1The number of miscarriages (from a total of 610) during each gestational week. Data from onehospital. (Sandahl, B., Europ. J. Obst. Gynecol. Reprod. Biol., 31, 23, 1989.)of men attending an infertility clinic and showing poor sperm production. Men attending the clinicwith normal sperm counts are used as controls, thereby avoiding possible bias from the fact thatan infertility clinic had been contacted. 18B. Spontaneous Abortion (Miscarriage)1. DefinitionA spontaneous abortion is the death (and expulsion) of an embryo or fetus before it has reachedthe legal status of a child, which is usually defined by gestational age and/or birth weight. Thelimit differs among countries. In some the upper age limit for a fetus is 28 completed gestationalweeks (calculated from the last menstrual period, LMP), but in many populations, lower limits areused, e.g., 20 or 22 weeks of gestation. To some extent, the upper limit will affect estimates ofspontaneous abortion rates, but rather few fetuses die between weeks 20 through 28. A dead fetusexpelled after the upper time limit of spontaneous abortion is referred to as a stillborn infant. Ifborn alive before that limit, it will be considered an infant.Figure 20.1 shows the gestational week distribution of spontaneous abortions treated in onehospital. 19 Early miscarriages that did not make the woman seek medical advice and miscarriagesthat occurred before the woman realized she was pregnant were not included. Some studies 20 haveindicated that in a high percentage of pregnancies, the embryo dies before or just after implantation,and the woman will never realize that she was pregnant. Such very early pregnancy losses areextremely difficult to identify and are therefore not very suitable for epidemiological studies. Afew such studies have been made 21 however. Various exposures were studied in pregnancies diagnosedwith sensitive pregnancy tests, and the possible association with very early pregnancy losswas determined.As an indirect method, TTP can be used (see above). If a pregnancy loss occurs briefly afterconception, a new pregnancy may occur in the next menstrual cycle, but it will result in a lengthening© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 805100908070Induced ab.60Percent504030Miscarriages20100Births20 25 30 35 40 45 50Age at pregnancy, yearsFigure 20.2Percent distribution of miscarriages, induced abortions, and deliveries among Swedish womenaccording to age at pregnancy. (Reprinted with permission from Källén, B., Epidemiology ofHuman Reproduction, CRC Press, Boca Raton, FL, 1988.)of the time from when the couple begins to try to achieve pregnancy until a pregnancy becomesclinically apparent. A delay caused by a very early pregnancy loss cannot, however, be distinguishedfrom a delay due to reduced fecundity. The measurement of time to pregnancy is therefore asummary of at least two effects.Because of the considerations described above, most studies assess clinically overt miscarriages.Information on the occurrence of a miscarriage can be obtained by interview or questionnaire, butexperience has shown that these data may be uncertain and sometimes biased. 22 Such data can alsobe obtained from medical records or registries based on such records. The study will then probablybe restricted to true miscarriages, but early events will be underrepresented. Furthermore, confoundingfactors (e.g., education and social class) may affect the probability that a woman with anearly miscarriage will seek medical help and may have an even greater effect on the likelihood thatshe will be hospitalized (many registries on miscarriages are based on discharge diagnoses fromhospitals). 232. Rates and ConfoundingThe actual rate of recorded clinically identified miscarriages varies with the method of ascertainment.Often figures around 10% to 15% are mentioned, but these are very crude and are influencedby many factors, including maternal age (Figure 20.2). The relatively high rate makes miscarriagessuitable for statistical analysis, but this is counterbalanced by the fact that spontaneous abortionsoccur for a variety of reasons. In individual cases, little is known about the etiology and pathogenesis.In early miscarriages, a high percentage of the conceptuses show gross chromosomal anomaliesthat made continued development impossible. Maternal infections may cause a miscarriage as maycertain immunological relationships between the parents, uterine anomalies, etc. Fetuses with somestructural malformations are aborted spontaneously at a high rate, but the aborted specimen isseldom investigated from this point of view. An increase in fetal anomalies caused by an exogenous© 2006 by Taylor & Francis Group, LLC

806 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONagent will result in an increase in miscarriage rate, but it will often be moderate in size and difficultto discern from the “background noise” of miscarriages from other causes.3. Study DesignThere are thus some inherent difficulties in counting the number of miscarriages occurring in agroup of women, and it is still more difficult to use the result for an estimate of miscarriage risk.To become meaningful, the number of miscarriages has to be related to the total number ofpregnancies occurring. Very often, the denominator is restricted to the number of deliveries, and alarge proportion of pregnancies are left out, i.e., those ending in a voluntarily induced abortion,legal or (in some populations) illegal. An induced abortion will prevent a delivery but will notprevent a spontaneous abortion if it occurs before the time of the induced abortion. The later thatabortions are induced, the stronger will be their effect on miscarriage rate. This phenomenon caneasily be demonstrated in pregnancies occurring in girls below the age of 15 in Sweden. Very fewsuch pregnancies end as deliveries. If they do not abort spontaneously, an abortion will be induced.If the spontaneous abortion rate is calculated in girls of this age class, without consideration ofinduced abortions, the risk for a miscarriage appears very high, while it is really quite low. Thisis an extreme situation, but less dramatic differences can be seen between different socioeconomicgroups of women, for example. Information on induced abortions is often missing or incomplete,making rates of miscarriage uncertain.In a case control approach, the same problem also arises. The control of a woman who miscarriesshould not be a woman who delivered a normal infant. It should be a woman with an intrauterinepregnancy (at the stage of pregnancy when the case woman miscarried) who did not miscarry anytime during the pregnancy. Her pregnancy may end with the birth of a normal or a malformedinfant or as a preterm birth, or it may end as an induced abortion. Representatives of all thesereproductive outcomes should be eligible as controls but are seldom included. Therefore, anyexposures that are somehow related to preterm birth or to induced abortion (e.g., maternal smoking)will appear as risk factors for spontaneous abortions if only normal infants are used as controls. 24,25The perfect analysis of miscarriage rates takes this into consideration and is based on life-tableprinciples, 19 but this necessitates access to data that are often unavailable. If the analysis comesshort of this standard, due caution is needed in its interpretation.All these factors make miscarriage less suitable as an endpoint for developmental toxicitystudies than would be thought from its relatively high occurrence rate. Especially when small effectson miscarriage rates (even when statistically significant) are described, a sound scepticism is neededin their interpretation.C. Infant and Child Death1. DefinitionFor practical reasons, child death is usually divided into different groups:Stillbirth — the birth of an infant (see definition above) who shows no signs of life.Early neonatal death — the death of a live born infant within the first 7 days (or 168 h) of postnatal life.Perinatal death — the sum of stillbirth and early neonatal death.Neonatal death — the death of a live born infant before the age of 1 month (or 28 days).Infant death — the death of a live born infant before the age of 1 year.The limits between the groups are often but not always sharp. An infant giving a few gasps after birthbefore dying may be classified either as stillborn or live born. The lower limit of stillbirth (even whenstrictly defined by gestational duration) may also be somewhat arbitrary, and a classification into© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 8071009590>37 w33–36 w28–32 wPercent surviving8580757065601d 7d

808 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmalformations, but in many, no clear-cut cause of death is apparent (especially in the case ofstillbirths). In contrast to miscarriages, perinatal (and later) deaths can be divided into subgroupsaccording to the main cause of death. Precision in the definition of the adverse outcome can thereforebe rather high. For instance, it is possible to see whether an increased perinatal death risk is dueto a shift in gestational age–birth weight distribution, if it is due to an increased rate of lethalcongenital malformations, or if it is due to other causes, e.g., to a high rate of abruptio placentae.Various classification systems of perinatal or neonatal deaths are available. Some are built oncause of death information from death certificates, a source of information which is notoriouslyuncertain. 28 Various hierarchic systems have been constructed, and Wigglesworth published themost well known. 29 A classification that can be used in studies based on medical birth registrieshas also been published. 30 However, for one large group, intrauterine deaths before the onset ofdelivery, the cause of death is often not known.4. Statistical AnalysisFor the statistical analysis of infant and child survival, life table analysis is often suitable. Figure20.3 shows the survival rate of term and preterm infants (dividing lines 29, 33, and 37 completedweeks of gestation). For each period, these rates are calculated based on the number of infants whocould have died then. The denominator for the calculation of the death rate during the second weekis therefore the mean number of infants surviving the first week and the number of infants alsosurviving the second week.In a certain period i, the number of infants entering that period is n i , and the number of infantsliving at the end of that period is n i+1 . The number of infants dead during the period is d i , and thedeath rate during the period isdip = 0.5( ni+ ni+1 )The standard error of that estimate iss p =p(1 – p)0.5( ni+ n )i+1assuming that a normal approximation can be made (d i is at least 5).D. Birth Weight, Pregnancy Length, and Growth Retardation1. DefinitionBirth weights can be analyzed in different ways. They can be described as continuous variables,with means, standard deviations, and standard errors of the means. They can also be described asdiscontinuous variables, either as the percentage of very low birth weight (VLBW) infants, that is,with a weight below 1500 g, or as the percentage of low birth weight (LBW) infants, below 2500g. There are advantages and disadvantages with both methods. From a clinical point of view, theVLBW infants are the most important group because they represent the highest risk for perinataland later death and sequelae. However, they make up a very small proportion of live births (some0.5%), and marked shifts may occur in this fraction without any major change in the mean birthweight. On the other hand, most environmental factors affecting birth weight do so by increasingthe proportion of infants in the range of 1500 to 2499 g.© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 80940003500Mean birth weight, g3000250020001500100028 29 30 31 32 33 34 35 36Pregnancy week37 38 39 40 41 42 43Figure 20.4Mean birth weight at different gestational ages in two groups of infants. Infants of mothers withonly compulsory school (- - - - -); and infants of mothers with higher education (———).The most important determinant of birth weight is gestational duration, and this variable isoften difficult to adequately ascertain. Information based on LMP is often inaccurate, and ultrasounddetermination of gestational duration is often not available. An infant is regarded as born at termif gestational duration exceeds 37 completed weeks, counted from the LMP, and preterm if bornbefore that time. Very preterm infants are born before 32 completed weeks — this group and stillmore infants born before 28 completed weeks represent the group with the highest demand onneonatal intensive care.The interaction between gestational duration and birth weight is twofold. A preterm birth of anormal baby will result in low birth weight. Alternatively, birth weight may be reduced at a givengestational duration. This reflects growth retardation during development, resulting in an infantwho is small for gestational age (SGA). In order to identify infants that are small or large forgestational age, standard birth weight graphs that are determined in different ways are used. Inprinciple, such graphs should be based on the same population as that studied, but when two ormore groups of infants are compared, the details of the growth curve are of less importance.It is likely that most environmental factors affecting birth weight will do so by retarding growth,but growth retardation in itself increases the risk for preterm birth. This phenomenon gives rise togreat problems in the determination of normal growth weight distribution curves for differentgestational weeks. 31 Much of the published data underestimate the “normal” weight in a given weekand overestimate the “normal” dispersion. Typically in epidemiological studies, however, theimportant matter is not to classify an individual infant as growth retarded or not but to comparemean birth weights at different gestational weeks in two or more groups of infants.Figure 20.4 shows an example: the mean birth weight at different gestational durations in twogroups of Swedish infants selected by maternal educational level. This figure also illustrates aparadoxical finding that is quite common; even though at term infants of mothers with low educationweighed less than the infants of women with high education do (as expected), the opposite is thecase in preterm births. The explanation is that in the former group a higher proportion of preterminfants are normal infants born too early, while in the latter group the majority of preterm infants© 2006 by Taylor & Francis Group, LLC

810 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONare preterm because they are growth retarded. This phenomenon gives rise to unexpected effectsif standardization is made for gestational duration in analyses of birth weight. 322. Rates and ConfoundingBirth weight information is usually very reliable, although in large birth registries errors will occurthat, even though rare, may be important. For example, some 9% of infants with a birth weightbelow 1500 g in the Swedish registry were misrepresentations of infants with normal birth weight. 33This falsely reduced perinatal mortality in infants with a birth weight below 1500 g. The incidenceof preterm and LBW infants varies between populations and races within populations. In Sweden,approximately 5% of singleton infants are born preterm, and 3.5% have a birth weight below 2.5 kg.The high accuracy of birth weight information and the large numbers available (practically allinfants have a birth weight recorded) offer good opportunities for statistical analysis. It is also wellknown that many environmental factors, e.g., maternal smoking, 34 affect birth weight. The problemis that an infant’s birth weight is influenced by a large number of variables. Some of them, suchas maternal age and parity (these two variables have to be crossed because of a strong interactionbetween them) 35 and infant sex are usually easily controlled. Others, including maternal weightand weight increase during pregnancy, parental height, social class, and smoking habits, are moredifficult to control.Similar problems exist in the analysis of gestational duration, but with the added uncertaintyof the accuracy with which duration may be determined.3. Analysis of Prenatal GrowthCrude comparisons of birth weight or gestational duration are easily made with standard statisticaltechniques, e.g., t-tests of means with their standard errors. When confounding factors are considered,a multiple regression analysis is often useful (although it is based on a supposition of nearnormalityin data distribution). Frequencies of preterm or LBW infants are also easily comparedusing suitable statistical tests, e.g., chi-square tests. When confounders are considered, a stratifiedMantel-Haenszel or a logistic multiple regression method can be employed. Details of thesetechniques are discussed later.Intrauterine growth retardation can also be analyzed either as continuous or dichotomous data.In the former situation, the deviation in birth weight for each infant is expressed as standarddeviations (SD) from the value of the relevant gestational week in the normal growth curve. As arule of thumb, SD can be approximated to 12% of the mean weight for a specific week. Thismeasurement (SDS, standard deviation score) can be used in the analysis in a similar way as, forinstance, birth weight. If a dichotomous analysis is preferred, the usual cutoff is more than ±2 SDfrom the mean (large and small for gestational age, respectively), but obviously other cutoffs (e.g.,2.5 or 3 SD) can be used.Easy access to gestational duration and birth weight information, and the large amount of dataavailable make statistical analysis simple. They also make it simple to demonstrate differencesbetween infants born, for example, in different geographical areas. 35 The main problem lies in theinterpretation of the findings because there are so many confounding factors, many of which aredifficult or impossible to control.E. Congenital Malformations1. DefinitionThere is no strict definition of a congenital malformation. Very loosely, it can be defined as acongenital structural anomaly of significance for the individual. Its severity may, however, vary© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 811from monstrous, usually lethal conditions, through severe and handicapping conditions, to conditionsthat can be repaired to more or less complete normality. Even conditions that only havecosmetic significance or that are so insignificant that the individual can live a completely normallife are also sometimes included.We can illustrate this with various results of a disturbance of the closure of the neural folds toform a neural tube, the embryonic process that leads to the formation of the central nervous system.In the most severe instances, an infant with anencephaly is born; remnants of a heavily malformedbrain lie exposed, and the infant is stillborn or dies during the first few days after birth. A disturbanceoccurring at a slightly later time and hitting more caudal parts of the nervous system may lead toa spina bifida aperta, with an open spine and herniation of the spinal cord and meninges. Whensevere, this leads to pareses and hydrocephaly, with severe brain damage and perhaps lifelonghandicap. In less severe forms, neurosurgery can correct the conditions, sometimes with little orno remaining handicap and normal mental development. In still less pronounced instances, a spinabifida occulta may develop, with an opening of the spine but with a normal or near normal spinalcord. In most instances, this gives no problems for the carrier, although it sometimes results indisturbances of bladder innervation.2. Classification and GroupingThe classification of congenital malformations is usually based on the ICD (International Classificationof Diseases) codes, possibly with extensions. It should be realized that the ICD code isconstructed for clinical use and is often not specific enough for detailed studies of congenitalmalformations. Thus, specially designed coding systems that enable a greater specification than ispermitted by the ICD code have sometimes been used.Congenital malformations are a very heterogeneous group of conditions from the point of viewof aetiology and pathogenesis. It is therefore rather useless to analyze this outcome as the totalmalformation rate, and some division into categories must be made. How such a division shouldbe made can be debated. A common method is to use organ system subdivisions, but the group“limb malformations,” for example, contains many very different subgroups. These could includelimb reductions, polydactyly, syndactyly, or positional defects, and a change in the rate of one type(e.g., radial reduction defects) can easily be hidden by the large numbers of more common defects.A problem is that the pathogenesis of specific defects is not always agreed upon and may evenvary for the same malformation. A small bowel atresia, for instance, may be the result of a veryearly disturbance in gut development, the result of faulty secondary canalization of the gut rudiment,or the result of an ischemic insult perhaps occurring rather late in development. A transverse limbreduction defect may be caused by an early disturbance in limb development, may be the secondaryconsequence of a vascular accident, or may occur as the result of a so-called amniotic constrictionband. It is therefore not possible to set up a strict categorization of congenital malformations thatmirrors pathogenesis, but it must be done in a more intuitive way. Nevertheless, if the subdivisionis carried too far, the number of infants with malformations in each subgroup becomes so low thatanalysis is meaningless.3. Mutagenesis and TeratogenesisAn environmental factor may cause congenital malformations by a mutagenic process affectinggerm cells, resulting either in DNA point mutations or in chromosomal anomalies (e.g., nondisjunctionresulting in trisomy or breaks resulting in translocation). To cause congenital malformations,mutagenesis usually has to affect the germ cell genome. This means that an effect can beseen from paternal or maternal exposures and that most likely the event has occurred beforeconception. It could also occur during very early embryonic cell divisions. Mutagenic events duringembryonic or fetal development are more likely to cause malignancy rather than malformation. In© 2006 by Taylor & Francis Group, LLC

812 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONstudies of mutagenic events, therefore, exposures before pregnancy should be studied, and exposuresof both prospective parents are of interest. On the other hand, a mutagenic exposure is probablyrelatively nonspecific, and a limited number of easily identified conditions can be used as indicators.For DNA point mutations, so-called sentinel phenotypes 36 can be used, i.e., conditions caused bydominant autosomal mutations (e.g., achondroplasia). Such conditions are rare and either very largestudy populations or very strong effects are needed in order to detect a mutagenic effect of anenvironmental factor.It has been demonstrated that obtaining very precise diagnoses and identifying actual dominantmutations is perhaps not so important. Equal power can be obtained by using large data sets withless diagnostic accuracy. 37 In a similar way, trisomy 21 (Down syndrome) is often used as the“sentinel phenotype” for chromosome anomalies (or at least trisomies). This condition is rathereasily identified and occurs at a reasonable rate (some 1 in 700 births), but it must be rememberedthat its rate is highly sensitive to maternal age distribution. Intensive prenatal diagnosis (especiallyamong older women) further affects the rate of infants born, and this is also a problem in the caseof other conditions which are readily diagnosed prenatally.The second way in which an environmental factor can cause a malformation is by teratogenesis.Teratogenesis is used here to mean a direct effect on embryonic development, acting during theactual development. In practice this occurs by exposure of the pregnant woman, and only exposuresduring pregnancy (or even during a short part of the pregnancy) are of interest.Both from a practical and theoretical point of view, mutagenic and teratogenic processes differ,and in the design and interpretation of epidemiological studies of environmental factors causingmalformations, this has to be taken into consideration. So far, most convincing evidence of a harmfuleffect of environment (in its widest sense) is suggestive of teratogenic processes.This discussion is also relevant for investigations of the possible role of paternal exposures andcongenital malformations. As stated above, mutagenic events can occur at sperm production, andthe most sensitive period is perhaps 3 months before conception. There is indirect evidence thatthe majority of new dominant DNA mutations causing birth defects are actually paternal in origin,but few, if any, specific risk factors have been identified, except for paternal age. On the other hand,the vast majority of nondisjunctional processes leading to trisomy 21 occur at maternal meiosis(usually the first meiosis, which takes place before the time of conception). Teratogenic effects dueto paternal exposure may occur, notably by secondary exposure of the embryo via the maternalorganism; examples are passive smoking and “carry-home” exposures, e.g., of toxic substancesfrom the working place.4. Source of Information on Congenital MalformationsInformation on congenital malformations in infants is best retrieved from medical records becauseinformation from parents is often incomplete and nonspecific. It should be realized that only aproportion of congenital malformations in infants are recorded at the time of delivery. Many aredetected during the first week of life, and others may give symptoms later, sometimes even yearsor decades after birth. The follow-up time is therefore important for the recorded rate, especiallyfor mild or internal malformations, but even severe cardiac defects may go undetected for sometime after birth.In many areas of the world, a special opportunity is offered by the existence of registries ofcongenital malformations. Such registries were usually set up for monitoring purposes but are alsovery useful as a basis for epidemiological investigations. Complete ascertainment is seldomachieved, and the possibility of a biased ascertainment must be kept in mind. The major strengthlies in the possibility of retrieving data for large numbers of infants with rare specific defects. Mostsuch registries are limited to recording malformations detected during the newborn period, butsome have an extended follow-up time, e.g., to the age of 1 year.© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 813A distinction is sometimes made between registries that are based on data collection by staffscrutiny of hospital records in a limited area (“active” programs) and registries that rely on thedirect reporting of the congenital malformations, often from a large area such as a whole country(“passive” programs). I think the terms are misnomers because the activity of a registry is bestevaluated by the intensity in the analysis of the incoming data, not the way in which data areobtained.5. Rates and ConfoundingThe inclusion or exclusion of especially mild forms will give very different total rates of congenitalmalformations. In the literature you can find total rates ranging from less than 1% to nearly 10%.Relatively severe malformations that result in infant death, handicap, or major medical interventionoccur at an approximate rate of 2%, but this rate is also sensitive to inclusion criteria.Rates recorded in a specific study related to one problem can never be compared with a generallyapplicable malformation rate; the rate used for comparison must be defined in the same way as therate in the study group. Specifically, consideration must be given to the hospital of birth becauserecording of congenital malformations varies among hospitals, especially for minor or variableconditions such as less severe congenital heart defects or positional foot defects.Compared with studies on birth weight, for example, studies on congenital malformations havefewer problems with confounding. Maternal age effects exist for some conditions (e.g., increasingrisk for Down syndrome with maternal age, high risk of gastroschisis at low maternal age), whileparity effects usually are weak. Maternal smoking can slightly increase the risk for some specificmalformations (e.g., orofacial clefts) but is by and large a weak confounder. Rather few drugs showa teratogenic effect of any significance. An effect of socioeconomic level exists in some countriesfor some malformations. In most studies, the strongest confounder is actually the recording hospital,which makes analyses of the geographical distribution of malformations difficult.F. Mental Retardation and Behavioral Problems1. DefinitionFor the individual, the family, and society, exposures capable of harmful effects on fetal braindevelopment are probably more important than most other hazards. They may result in a reductionof the intelligence quotient (IQ), perhaps leading to mental retardation and a need of social support.At the least, the intellectual capacity of the child, and therefore its social success, may be reducedbelow the capability defined by its genetic makeup. Mild mental retardation is usually defined asan IQ of 50 to 69, moderate mental retardation as an IQ of 35 to 49, severe mental retardation asan IQ of 20 to 34, and profound mental retardation as an IQ under 20.Other effects of brain disturbance may be less dramatic but nevertheless result in social consequences;minimum brain damage may result in lack of concentration and in behavioral disturbancesfollowed by difficulties in school and in social adaptation. It has been suggested that anincreased rate of non–right-handedness in a population can be an indicator of slight brain damage. 382. Sources of InformationEffects on mental development are important but also difficult to study, most notably the less markeddeviations from normality. Limited groups of exposed infants can be followed and studied withvarious psychological tests. For example, this has been done for maternal alcoholism 2 and leadexposure. 39 It is much more difficult to conduct such studies on large groups of infants in situationswhere weak or moderate effects are expected, e.g., after workplace exposures, environment pollution,© 2006 by Taylor & Francis Group, LLC

814 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION10OR with 95% CI310.30.1500 1000 1500 2000 2500 3000 3500 4000 4500 5000Birth weight, gFigure 20.5Odds ratio for mental retardation as a function of birth weight in singletons. Infants with mentalretardation identified from special register of mentally handicapped infants (———); and infantswith a diagnosis of mental retardation identified from a hospital discharge registry (- - - - -).or food contamination. Nevertheless, toxic levels of organic mercury (e.g., the Minamatacatastrophe 40 ) have been shown to cause major brain damage.In some countries or areas, registries of infants with mental retardation exist and can be usedfor large scale studies. Efforts have also been made to make use of hospital discharge registries,even though it must be realized that only a portion of children with mental retardation needhospitalization. Figure 20.5 compares the odds ratios for mental retardation according to birthweight, based on two different populations: a specific register of infants in social care because ofmental retardation and infants hospitalized with a diagnosis of mental retardation. As this figureshows, the results are relatively similar.In order to study mild forms of brain damage, central registries are of little help. Registries ofmedical characteristics of young men at the time of conscription have been used to demonstratean association between late prenatal ultrasound exposure and an increased risk for non–righthandedness.383. ConfoundingPerinatal complications, including short gestational duration, low birth weight, and intrauterinegrowth retardation, are major causes of brain damage. Obviously, such variables may be intermediary.Because IVF procedures increase the risk for preterm birth among singletons and for multiplebirths, they are likely to increase the risk of brain damage, including effects on mental development.However, detailed studies of small groups of children have a low power to demonstrate sucheffects. 41A more difficult confounder is parental mental characteristics because genetics play a large rolein the mental development of the child. A confounding of socioeconomic conditions is thereforeto be expected, most notably for mild effects. When registries are based on hospitalization, otherparental characteristics, including maternal age, parity, and subfertility, may affect the probabilityof hospitalization. 42© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 815G. Childhood Cancer1. Definition and ClassificationMany instances of childhood cancer are thought to be initiated prenatally, perhaps as a result ofmutations in somatic cells of the embryo or fetus. The peak incidence of childhood cancer (thatis, a malignancy developing before the age of 15) occurs at 3 to 4 years of age, and it is probablethat malignancies with an early onset are especially likely to result from a disturbance duringintrauterine development. In spite of this, rather few instances of factors resulting in intrauterinecarcinogenesis are known. The best known example is intrauterine exposure to diethylstilbestrol(DES) resulting in vaginal cancer, which, however, was not observed until after the age of 15. 4In a study of childhood cancer in Sweden, about 40% were leukemias (mainly acute lymphoblasticleukemia), 28% were tumors of the central nervous system, and 9% were kidney tumors. 432. Sources of InformationChildhood cancer is a relatively rare occurrence, with a rate of about 2 per 1000 children. Individualfollow-up of children is therefore tedious, and large groups are needed for such studies. Usually,case identification is through central cancer registries. This necessitates that each individual canbe identified with certainty, e.g., by the use of personal identification numbers, as is the case inNordic countries. The completeness of cancer registries is often good, although not perfect. Oneshould also realize that when a group of children is followed from birth, detection of the cancernecessitates that the child survive and also that the child does not emigrate from the area coveredby the cancer registry. It is thus easy to demonstrate that children born in Sweden of immigrantparents have a lower probability of appearing in the cancer registry than children born of Swedishparents. This is a result of a higher emigration rate among the former.A. Embryonic and Maternal ExposureIII. METHODS TO RECORD EXPOSURESIn most instances, the embryonic (or fetal) exposure is the critical information we seek. It is seldomknown for certain, even though such measurements as lead content in deciduous teeth 44 or cotininein infant hair (to estimate exposure due to parental smoking) 45 have been used. Instead, maternalexposure is usually used as a proxy for embryonic exposure, but maternal exposure is seldomknown exactly. In drug teratogenesis studies, for example, the use of a drug may be ascertained(although even this may be difficult), but actual maternal serum levels of the drug during those fewdays or weeks when a specific malformation is formed are seldom known or estimable. Nevertheless,such information can sometimes be obtained for chronic medication, where drug concentrationsare routinely monitored (e.g., anticonvulsant drugs, lithium).B. Estimates of Maternal Exposure1. Concerns about Exposure DataIn studies on occupational exposures or exposures from environmental pollution, actual exposuredata are usually rather unspecific. Very often exposure data mainly indicate with a reasonableprobability that the woman has been exposed, but the actual level is not known. This will result ina bias toward null effects if a large proportion of women regarded as exposed were not actuallyexposed or were exposed only to a very low degree.© 2006 by Taylor & Francis Group, LLC

816 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIn practice, occupational exposure usually means that the woman has had a reasonable probabilityof having been exposed during pregnancy. Such exposure information can have differentlevels of precision: occupation information (e.g., nursing), actual working situations (e.g., workingin anesthesiology), types of exposures (e.g., description of working place, including types of volatileanesthetics handled), or measurements of air content of the volatile anesthetics used combined withexact information on working hours. Moving along this list, better and better approximations ofthe actual embryonic exposure will be obtained; at the same time, most likely fewer and fewerpregnancies will have that level of exposure information.The increased precision in exposure information is often strongly counterbalanced by a decreasingpower of analysis due to the lower number of cases. This will be especially important in thestudy of rare events, such as perinatal deaths or congenital malformations. It is not very informativeto say that among 10 women with an exposure of so many ppm of halothane 6 h a day during 5working days during weeks 4 through 5 of pregnancy, none had a baby with spina bifida. Thisestimate can actually mean anything from 0% to 37% risk of such a pregnancy outcome. It maybe more informative to get an idea of the risk (with its confidence interval) of having an infantwith spina bifida for any pregnant woman working in anesthesiology.On the other hand, exposure information may be so crude that it is meaningless. For example,a study of all women working in industrial production will probably give more information on theeffect of social factors on reproductive outcome than on the effect of specific chemical or physicalexposures.Similar concerns are valid for studies of exposures from environmental pollution. Very often,the only available information is the place of residence of the woman (quite often only the placeof residence at delivery, which does not necessarily mean the place of residence in early pregnancy,when most malformations are initiated), and estimates of actual pollution (e.g., in air) are veryrough. Furthermore, exposure occurs only to some extent at the place of residence. Most womenspend a considerable part of their time in other places, such as their workplace, and sometimes theplace of residence is a better estimate of social situation than of actual exposure to pollution.Before starting data collection, it is useful to make a power analysis to see whether the designof the study may permit the identification of a specific risk. The exact structure of such an analysisdiffers according to the design of the study, but we can exemplify it with a typical situation: onewants to compare congenital malformation rates in two groups of infants, one exposed and one notexposed, for a suspected harmful factor.The power calculation is based on two concepts, α and β, where α is the probability that thefinding is false (due to chance), and β is the probability that the study reveals an existing effect.Very often in power calculations α is set to 0.05 and β to 0.80 (but other values can be chosen).The following parameters are of interest: the number of exposed infants (ne), the number ofunexposed infants (nc), the expected rate of the outcome (e.g., congenital malformations) (p), andthe risk increase caused by the exposure (r). The following equation shows the relationship but isbased on normal approximations, which means that the expected number of outcomes in the groupwith the smallest number should exceed 5.The probability of a malformation in the exposed group is then P 1 = rp. The probability fornot having a malformation is thenQ 1 = 1 – P 1The probability for a malformation in the unexposed group is P 2 = p, the probability of not beingmalformed is Q 2 = 1 – p. The probabilities in the two groups together are called P m and Q m ,respectively. The fraction of the total that is exposed ne/(ne + nc) is called f.C 1-β is then© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 817( P1– P2) f ( 1 – f )( ne + nc) – c1−α/2 PmQmIf p is small (e.g., malformations, perinatal deaths), 1-p approximates 1, and the equation can besimplified.The equation can be solved in different ways, e.g., how large is α (the probability of findingan effect) at other given parameters, what number of ne and nc are needed in order to detect therisk r, or what risk increase is likely to be detectable given the other parameters? Computer programsfor power calculations are available in most statistical software packages.Let us apply this to the specific study of nurses exposed to anesthetic gases. We want to beable to demonstrate a doubling of the congenital malformation risk (put as 2% in the unexposedpopulation), and we want to compare two groups of infants born of nurses. The mothers of onegroup worked as anesthesiology nurses, and the others did not. How many cases in each group (ifthe two groups are equal in size) are needed to have an 80% probability of identifying a doubledrisk in the exposed group with a 95% certainty? The equation above gives the answer: 1140. Astudy of only 50 women in each group, for example, would require a nearly 10-fold increase inmalformation risk (18% occurrence of malformations) to have a reasonable chance of detection.Such an increase is not a very likely event, even if exposure criteria are very strict.A doubling of the risk in a group of women who may have a 50% chance to have had a realexposure means that the actual risk in the exposed group is fourfold. It is thus easier to detect therisk in a large group of women where exposure is present in only half than to detect it in a smallgroup of women where exposure has been individually verified.2. Information on Drug Exposure( – f)PQ1 1 1 + fP2Q2Since the occurrence of the thalidomide-induced epidemic of severe congenital malformations,interest has been paid to the possible hazards associated with maternal use of drugs during earlypregnancy. With the continuous addition of new drugs to the therapeutic arsenal, this problem isalways relevant. Even though new drugs are extensively tested in animals, the possibility of ahuman-specific teratogenic effect exists. This necessitates continued studies on the associationbetween drug use and birth defects.In most classical studies on drug teratogenesis, exposure information is obtained by retrospectiveinterview or questionnaire studies, usually in a case control design. Sometimes the study has beenspecifically designed for the purpose of studying the association between a specific malformationand maternal use of drugs. In other studies, data are obtained from ongoing recording of drug useby mothers of malformed infants and control mothers without any restriction to malformation typeor drug type. 46 Sometimes data have been extracted from data retrieved for other purposes, and thestudy of drugs has just been a spin-off of another study, e.g., the Vietnam Veteran Study. 47Other studies have tried to record all drugs used and all birth defects present in a specificpopulation. 48 In these studies, the recording of drug use has usually been prospective, but no specifichypothesis is tested. Instead, all drugs have been crossed with all groups of malformations toidentify combinations showing an excess. Such an approach is good for hypothesis generation, butthe fact that multiple associations are tested makes it nearly impossible to distinguish causalassociations in a given study from random occurrences. Most of these studies are research projectsof a limited size.Groups of women who have used a specific drug or a group of drugs can be identified indifferent ways, and their pregnancy outcomes can be studied. An example is women with epilepsy© 2006 by Taylor & Francis Group, LLC

818 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONwho are using anticonvulsants. During the last decades, such information has been published fromTeratology Information Services (TIS). Women and doctors can approach such organizations toask for advice when a drug has been used in early pregnancy, and by following up such exposedpregnancies, information on possible hazards can be obtained. The main drawbacks are that onlyrelatively small numbers of exposures are usually identified, that pregnancy outcome is not alwaysknown, and that it is sometimes difficult to get control groups for comparison. A positive characteristicis that exposure data with respect to timing and drug dosage are often precise.During the past few years, there have been further opportunities. Prospective recording of druguse in early pregnancy has become part of the Swedish Medical Birth Registry. This provides agrowing database of prospectively recorded drug information with outcome data that can easily becompared with that of all births. 49 The main drawback is that there are few data on time of exposureand amounts of drug used. Other programs have used linkage between registries of drug prescriptionand birth registries. 50 Exposure information will be crude because it will not be known whetherthe woman actually used the drug during pregnancy, especially if it is not a chronic medication.In one system, prospectively recorded drug exposure data are combined with retrospective data,obtained by interview. 51In studies on drug teratogenesis, various confounders, such as maternal age and parity, subfertility,and smoking habits, should be taken into consideration. The most important confounder, thedisease or complaint for which the drug was used, is difficult to control. One way is to comparedifferent drug categories used for the same condition.3. Information on Occupational ExposureIn many studies on occupational hazards to reproduction, maternal occupation has been used as aproxy for exposure. 52 This will usually give rather crude exposure information, and within thestudied group a subgroup with markedly different exposure conditions that may represent a riskmight be hidden. The more detailed the occupational characterization, the better the study, butusually it is not possible to get accurate actual exposure levels. In some studies (usually designedas nested case control studies), the actual exposure for each involved woman has been evaluatedblindly by an occupational hygienist. In such a situation, the relevant period when the malformationin question could have been formed should also be considered. Direct measurements will furtherincrease exposure precision, but as stressed above, the necessary small number of cases will oftenreduce the study power too greatly.In other studies, exposure information is obtained by retrospective interviews or questionnaires.Except for the risks involved in all retrospective studies (see Section III.C), lack of knowledgeamong the women of their real exposure is a problem.4. Information on General Environmental ExposureIn studies of the impact of the environment on pregnancy outcome, one is usually dealing withrather generalized exposures of large populations (e.g., air pollution, drinking water contamination,radioactive fallout) where very strong effects are not expected. In studies of congenital malformations,the environment for the woman during her early pregnancy is of interest. Often only themother’s address at the time of delivery is available, and a change of address is not uncommonduring pregnancy, especially the first pregnancy. In most studies, a rather crude geographicallocalization is made, e.g., parish or postal code. In more sophisticated studies, the actual coordinatesof the address can be used and perhaps linked to air pollution measurements or information ondrinking water composition.When a source of environmental pollution is studied, the distance between the address of thewoman and the source of pollution is often used as a proxy for exposure. In some studies of airbornepollution, the dominating wind direction has been taken into consideration.© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 8195. Misclassification of ExposureExcept for small studies with very exact exposure information, some misclassification is expectedto occur. There may be two types of misclassification. One is due to biased exposure information,e.g., obtained in retrospective case control studies. This can result in wrong and usually exaggeratedrisk estimates. The second type of misclassification is obtained as a more or less random phenomenonamong prospectively recorded exposure information. For example, in a study of maternalsmoking, some smoking women may deny that they smoke. As such a behavior probably hasnothing to do with the outcome, the misclassification will be nondirectional, i.e., it will be equalamong mothers of infants with birth defects and mothers of normal infants. Such a nondirectionalmisclassification will in principle result in a reduction of the risk estimate, i.e., a bias toward unity.It will therefore be of greatest importance in studies where the outcome of the study is that theexposure in question has no harmful effect — this could be the effect of an underestimate of anactual risk. A woman who reported the use of a drug but used it before or after the critical periodwhen a malformation can arise or a woman working in a specific occupational setting who had aholiday during the critical period would be additional examples of this type of misclassification.If the actual risk increase after a specific exposure is a doubling of the population rate ofcongenital malformations and the outcome is studied in a group of women where only 75% wereactually exposed, the estimated risk increase will amount to 1.8, and if 50% were exposed, theestimate would be only 1.5.C. Prospective and Retrospective Exposure InformationIn general, exposure information is obtained by questionnaires or interviews. This can be doneprospectively, that is, before the outcome of the pregnancy is known, or retrospectively, after theoutcome is known. The latter is the most common method. The drawback is that the woman whohas had an adverse reproductive outcome is sometimes more careful in the reporting of exposures.She may even imagine that exposures have occurred or that they occurred during the relevant partof the pregnancy instead of earlier or later. This has been called “memory bias” or “recall bias.”Its existence has been denied 53,54 or has been thought to be insignificant, but there is ample evidencethat a recall bias may exist and may give misleading results. 55 When information is retrieved byinterview, the interviewer’s knowledge of the pregnancy outcome may also affect the thoroughnessof the interview and the recording of the answers, thus introducing an “interviewer bias.”Both of the foregoing types of bias have a tendency to increase the estimate of a risk, while areduction of the risk estimate will be obtained if the woman understates the exposure. This is oftenthe case when information on smoking, alcohol use, etc., is sought. As most women know thatthese factors are harmful for the pregnancy, they have a tendency to underreport them. This mayalso, however, be influenced by the outcome of the pregnancy.In some circumstances, exposure information can be obtained from independent sources toreduce information bias. At least crude exposure information may be obtained from central registries(e.g., concerning maternal and paternal occupation) 56 and more refined information may be receivedfrom independent sources, such as industrial hygienists. 57 Measurements of environmental pollutionmay give an unbiased estimate of exposure. 58IV. METHODS TO RECORD OUTCOMESA. Questionnaire and Interview InformationPregnancy outcome can be identified in different ways. One is by questionnaire or interview studiesof women who have had a specific exposure. Some care should be applied because experience has© 2006 by Taylor & Francis Group, LLC

820 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONshown that some types of data are well recalled and reported (e.g., birth weight), while others arenot. The self-reporting of miscarriages, for instance, has been shown to be relatively inexact and mayalso be biased by the woman’s apprehension regarding various exposures she has had. 22 Congenitalmalformations are often underreported, most notably lethal and minor ones. Self-reported data may beuseful in studies of long-term follow-up of children, e.g., for mental disturbances.B. Medical Record StudiesAnother possible source is medical record information. Often information retrieval is restricted todelivery hospital files, which may give a serious underestimate of congenital malformations andwill usually say nothing about survival beyond the first few days of life. A complete search of otherrelevant clinical records (pediatric, pediatric surgery, plastic surgery, etc.) will increase ascertainmentbut will obviously be very tedious. The data obtained are reasonably unbiased.C. Register StudiesIn many parts of the world, central registries are available. These include medical birth registries(which can sometimes be linked to death registries), special registries of congenital malformations,and even central registries of spontaneous and induced abortions in a few countries. 23,59 Thesesources give very easy access to large amounts of data, and are of great value for the selection oflarge number of infants with a specific congenital malformation. They are also valuable in theselection of controls in case control studies and as the basis for cohort studies. When it is possibleto link the medical birth registries with other outcome registries (e.g., hospital discharge registries),further opportunities exist for long term follow-up of specific groups of children.On the other hand, it must be realized that errors usually exist in large computerized registries,and a thorough knowledge of the registry and its shortcomings is necessary for its proper usage.When used in a correct way, the registries offer good opportunities for conducting epidemiologicalstudies; when misused, the opportunity to obtain false results is also great.A. PrinciplesV. CASE CONTROL STUDIESIn case control or case referent studies, exposure rates are compared among “cases” (in this context,women with an adverse reproductive outcome) and “controls” or “referents” (women without thatoutcome). This study design has both advantages and disadvantages. Among the advantages is thatthe number of cases to study can be relatively high, and therefore the exposure rate in cases canbe determined reasonably well if exposure is not too rare. The exposure rate is compared with thatamong individuals where the reproductive outcome did not contain the anomaly under study. Ifinformation is available on the total population (which may be the case when population registriesare available), the exposure rate among controls can be accurately estimated. Often such informationis not available, but control exposure rates are estimated from a set of control individuals. Thehigher the number of controls per case, the more exact the estimate of the control exposure rate,but the uncertainty in the case exposure rate of course remains. Therefore, only a small increasein power is obtained by increasing the number of controls per case above two or three.The case control approach makes it possible to use the same set of cases for a study of manydifferent exposures. A typical example is a case control study of maternal occupation and a certainmalformation. Each occupation can be studied separately, but it should be remembered that if manysuch tests are made, false positive associations can be obtained for one or even a few occupationsjust by chance. 34© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 821Table 20.1Maternal smoking and odds ratiosfor different types of facial clefts 60Facial Cleft Category Odds Ratio 95% Confidence IntervalCleft lip and /or palate, isolated 1.11 0.99–1.29Associated 1.41 0.98–2.09Median cleft palate, isolated 1.35 1.12–1.63Associated 0.88 0.51–1.54All facial clefts, isolated 1.18 1.06–1.31Associated 1.19 0.88–1.62B. Selection of CasesIt may seem easy to define a case in studies of reproductive epidemiology, but it sometimes offersdifficulties. If we are interested in the effect of environmental factors, obviously it would be goodto exclude all infants where the malformation has a nonenvironmental cause, e.g., a monogeniccondition or a chromosome anomaly. Even though environment may play a role in the etiology ofa specific malformation (e.g., duodenal atresia) in a Down infant, the major cause of the malformationin that infant is the nondisjunction process leading to trisomy 21. In many conditions, e.g.,cleft lip and palate, a genetic component occurs. It can then be debated whether infants with aknown family history should be excluded. The ideal is to treat them as a separate group, butunfortunately such splitting results in a reduced number of cases available for analysis. It can alsobe argued that the genetic background may act by increasing the susceptibility to environmentalagents, and an exclusion of infants with a known family history may reduce the proportion of“susceptible” cases.Case selection can be more or less strict. Should all neural tube defects be lumped together, orshould anencephaly and spina bifida be kept apart, and should the latter perhaps be divided intohigh and low forms? It is possible that only a subset of the group is associated with the environmentalexposure. If all are analyzed together, the risk is that this subgroup will be drowned among theremaining ones that are not associated with the exposure. Again, the ideal is to subdivide thematerial and look for heterogeneity in it, but the risk is that numbers and statistical power arediminished. A balance must be kept between large enough numbers to permit statistical analysisand groupings specific enough to prevent dilution of possible associations, thereby reducing power.One important distinction that is often made is between infants with isolated congenital malformationsand infants with two or more associated malformations. There is some good evidencethat these may be different etiological and pathogenetic entities that should be analyzed separately.In some circumstances, separating cases into subgroups that make biological sense does notseem to affect the association with an exposure. Table 20.1 exemplifies this from a study on thepossible association between maternal smoking and facial clefts. 60 The material was divided intocleft lip with or without cleft palate and isolated cleft palate; these are regarded as very differententities. Cases involving known chromosomal anomalies were deleted, and those remaining werecategorized as “isolated” and “associated.” The association with maternal smoking was seen inthree of the four subgroups, and the estimated odds ratios did not differ significantly. Statisticalsignificance was not reached in all groups separately but was attained when the groups were totalled.C. Control DefinitionControl definition can also offer problems. It is important to realize the definition of a control. Itis not necessarily a delivery of a normal infant; it is a pregnancy that could have ended as a casebut did not. A spontaneous abortion can never be a control to an infant with a congenital malformation,but an infant with a congenital malformation can be a control to a spontaneous abortion.© 2006 by Taylor & Francis Group, LLC

822 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe control pregnancy for a case of spontaneous abortion is an intrauterine pregnancy that did notmiscarry, but it can end as an induced abortion, a preterm infant, a malformed infant, a stillbirth,etc. If normal surviving infants are used as controls, all factors associated with other types ofreproductive outcome will appear as risk factors for spontaneous abortions, and studies based onsuch comparisons are invalid.Similarly, the appropriate control for a case with one malformation (e.g., isolated spina bifida)is a nonmalformed infant, while the control for a case with two or more malformations can beeither a normal infant or an infant with one malformation. The suitable control for an infantdeveloping mental retardation or childhood cancer is an infant alive at the time of diagnosis butwithout these characteristics. If dead, the control could not have received the diagnosis under study!If, for example, the birth weight distribution of infants developing mental retardation is comparedwith that of all infants born, very low birth weight will appear as a protective factor for mentalretardation. In a way it is, because the risk of death is high and a perinatally dead infant cannotdevelop mental retardation. But if adequate controls are selected (surviving infants), one finds thatvery low birth weight is a risk factor for mental retardation.D. Selection of ControlsIn principle there are two ways to select a control: randomly or by matching. The random controlis drawn by a randomization process, e.g., from a registry of all births or a subset of that registry.If no registries are available, some other type of randomization must be applied.For control matching, one looks for controls that share some characteristics with the case, e.g.,maternal age, parity, or geographical location. Matching can be so-called frequency matching,which means that in the set of controls, the variable in question (e.g., maternal age) occurs withthe same distribution as among the cases. If, for example, 10% of the cases have a maternal agebetween 15 and 19, then 10% of the controls are selected from that maternal age stratum. Thisensures that the cases and controls resemble each other as groups. Individual matching selectscontrols that have the same values as the cases for the matching variables. For example, for aspecific case with a certain maternal age and parity and born at a certain hospital, one or morecontrols are selected with the same or similar (e.g., the same 5-year age group) maternal age andthe same parity, born at the same hospital.The purpose of matching is to reduce the influence of confounding variables (see further on).In other circumstances, matching is more a question of easy control selection. A typical exampleis the “next baby born” control, perhaps with restrictions concerning life or sex. This will give avery heavy matching for date of birth (which is usually not very important) and for hospital ofbirth, but no further matching. This type of control is standard in some registries for congenitalmalformations, but it should be realized that the matching is limited in extent.Matching makes impossible an analysis of the variables for which matching is made. Manyinvestigators therefore prefer to use random controls and control for confounding in other ways(see further on). When some variables are very skewed (e.g., maternal age when the infant hasgastroschisis or Down syndrome), a lack of matching may reduce the information on controlsavailable for extreme groups (very young women in the case of gastroschisis and older women inthe case of Down syndrome), thus reducing power in the analysis. One should remember that thepurpose of the controls is to estimate exposure rate in the population, and if we are interested inexposure rate in very young women (who had infants with gastroschisis), we must use the exposurerate in control women of that age stratum for comparison. If few such women are included in thesample of controls, the estimate of the exposure rate will be uncertain, regardless of the statisticalmethod used.Quite often, one chooses an infant with a congenital malformation (sometimes called a “sickcontrol”) as a control for an infant with another congenital malformation. The beauty of thisapproach is that the estimate of exposure rates may be less biased than when “healthy” controls© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 823are used, but the problem is that the risk factor under study may affect both malformations. In mostinstances, this is not a major problem because teratogenic effects usually have a rather high degreeof specificity, but it may pose a problem.E. Statistical Analysis of Case Control Materials1. Unmatched StudiesTable 20.2Information on smoking in earlypregnancy in women who had an infantwith Down syndrome and in controlsSmoker Nonsmoker TotalDown infant 183 569 752Control infant 444 1060 1504Total 627 1629 2256Generalized Form of a 2×2 TableExposed Nonexposed TotalCases a b n 1Controls c d n 2Total n 3 n 4 nSource: Data from the Swedish Medical Birth Registry,1983–1990.In this situation, two groups of individuals are compared: cases and controls. Table 20.2 shows anexample, the smoking rate among women who had infants with Down syndrome and those whohad infants without a chromosome anomaly. The latter group was randomly selected among alldelivered women.Smokers and nonsmokers in the two groups form a 2×2 table with marginal sums (see bottomhalf of Table 20.2). For each of the four cells, the expected numbers can be calculated from themarginal sums. For example, the expected number for cell a is n 1 n 3 /n. The distribution is reallyhypergeometric, but if numbers are large enough, it can be approximated as normal. The smallestexpected number in each cell should exceed 5. If a normal approximation is made, the χ 2 -distributioncan be applied as a test.The χ 2 value calculated from the 2×2 table will be χ 2 = Σ(o – e) 2 /e, where o is the observedand e the expected value in each cell. There is an easier and quicker way to reach the same result:χ 2 = (a*d – b*c) 2 *n/(n 1 *n 2 *n 3 *n 4 )If this formula is applied to the data from Table 20.2, the resulting χ 2 = 6.72, and from a statisticaltable, it is found that the probability of this occurring by chance is only 0.0095. Mothers of infantswith Down syndrome were less likely to be smokers (24%) than mothers of chromosomally normalinfants (30%), and this difference is unlikely to have been due to chance.As stated above, this method supposes that numbers (as they were in the example) are reasonablylarge. The smallest expected number in the table (smokers among mothers with Down syndromeinfants) was 627*752/2256=209. Even so, the p value is not quite correct, but it is a reasonablygood estimate (the exact p value is 0.010 instead of 0.0095, which matters little). Let us suppose,however, that Table 20.2 had been based on only 10% of the sample. There would perhaps havebeen 18 smokers among 75 mothers of infants with Down syndrome and 44 among 150 controls.The proportion of smokers in the two groups would be the same, but χ 2 =0.71, which gives p =0.40. Even though the mothers of infants with Down syndrome smoke less than control mothers,© 2006 by Taylor & Francis Group, LLC

824 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 20.3MaternalAge ClassSmoking in early pregnancy among women who hadinfants with Down syndrome and age-matchedcontrols, divided according to maternal age classDown InfantsControlsSmoking Nonsmoking Smoking Nonsmoking15–19 5 8 11 1520–24 43 67 72 14825–29 44 144 94 28230–34 45 156 103 29935–39 35 137 72 27240–44 11 52 31 9545–49 0 5 1 9Source: Data from the Swedish Medical Birth Registry, 1983–1990.the restricted study could not exclude the possibility that this was due to chance. Note that astatistically nonsignificant result does not mean that the effect is lacking, only that it can be dueto chance, perhaps because the sample was too small.When we are dealing with very small numbers, the normal approximation, which is a prerequisitefor the χ 2 test, is not permissible. In a study, 4 cases among 10 were exposed to a specificdrug, with only two cases among 30 controls (3 controls per case); χ 2 = 5.68, which is statisticallysignificant (p = 0.02). The lowest expected number is 10*6/40=1.5, which is below 5, and thenormal approximation is not permissible. It is, however, possible to calculate the exact probabilitythat this distribution is random with the Fisher test.Using the symbols of Table 20.2, the exact p value for this distribution is:p = n 1 !*n 2 !*n 3 !*n 4 !/(n!*a!*b!*c!*d!)where n! means n factorial, that is, 1*2*3*...*(n1)*nSuch a p value must be calculated for this distribution and for all distributions that are stillmore deviant (in our case 5-5 vs. 1-29 and 6-4 vs. 0-30), and the sum of these p values tells theexact probability that the observed or still more deviant distributions were obtained by chance. Thep value obtained is one-tailed, and in order to compare this value of p with that obtained in a χ 2test (which is two-tailed), it must be doubled. One-tailed tests are permissible only when it ispossible to say, a priori, that if a deviation is obtained, it must be in one direction only (e.g., anincreased smoking rate). In the example above, no such prior statement can be made; smokingcould conceivably increase as well as decrease the risk to have an infant with Down syndrome.Exact tests are available in most statistical computer packages. When applied to the observationsjust described, a p value of 0.03 is obtained, which is also statistically significant.The p value tells how likely it is that the finding was obtained by chance but tells nothing aboutthe strength of the association between the exposure the outcome studied. This association can beexpressed as a ratio between the two groups. In case control studies, the odds for being exposedare usually compared between two groups, cases and controls. The odds are calculated as the ratiobetween exposed and unexposed individuals. Using the designations in Table 20.2, the odds forcases to be exposed are a/b and those for controls are c/d. These two odds are compared as anodds ratio (OR), which will be OR = a/b:c/d. If the OR = 1, cases and controls had the sameexposure rate, an OR > 1 means that cases were more often exposed than controls, and the oppositeis true for OR < 1. The OR in Table 20.2 is 0.77, which means that smoking occurs less oftenamong case women than among control women.The odds ratio has a confidence interval. Its 95% confidence interval (CI) means that with 95%probability the true OR lies within that interval. If the lower limit of the interval is above 1, itmeans that the OR is significantly increased, while if the upper limit is below 1, the OR issignificantly decreased. The value of the CI is such that one can evaluate not only the significance© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 825but also the magnitude of an effect. The 95% confidence interval of the odds ratio can be calculatedby using Miettinen’s method (test based estimate) according to the following formula, althoughother methods are available:95%CI of OR = OR ±Let us take the example in Table 20.2 — the OR is 0.77 and the 95% CI is 0.63 to 0.94, while inthe smaller material the OR is nearly the same (0.76), but the 95% CI is much larger (0.40 to 1.44).From the former estimate, it can be said that most likely the OR is reduced by at least 6% and possiblyby as much as 37%, while from the latter estimate, the OR could be increased by up to 44%.It should be observed that the odds ratio is not a direct estimate of the risk or the risk decreasefor a smoking woman to have an infant with Down syndrome, but when very rare occurrences(such as Down syndrome) are studied, the estimate will be quite adequate. If we suppose that thepopulation risk of having an infant with Down syndrome is 1 in 700, it means that the 752 caseswere drawn from a total of some 526,400 births. Among them (judged from the smoking rateamong controls), 155,400 smoked and 371,000 did not. The risk of having an infant with Downsyndrome among smokers is thus 1.18 per 1000, and among nonsmokers it is 1.53 per 1000. Therisk ratio will be 0.77, exactly the odds ratio that was estimated.2. Stratified Analysis — A Method to Control for ConfoundingIn the example in Table 20.2, we saw a difference in smoking rate among case and control women.Is this difference causal? Does smoking really prevent the birth of a Down infant, or is it a secondaryeffect of a so-called confounder, that is, a variable that affects both smoking rate and the risk ofhaving a Down infant?Table 20.3 addresses this problem. Here the material has been divided into 5-year maternal ageclasses. From the control data it is apparent that smoking declines with age, and it is well knownthat the risk of having an infant with Down syndrome increases with age. The seemingly protectiveeffect of smoking may therefore be secondary to these differences in smoking distribution andDown infant risk between age classes.There are different methods to control for such a confounder (or a set of confounders). One isto match cases and controls, so each case and control will form a pair with the same maternal age.Another is to make a stratified analysis. The data in Table 20.3 can be looked upon as a series of2×2 tables, one for each age class or stratum. The Mantel-Haenszel method makes it possible tosum up the differences over the strata and to calculate a χ 2 (based on one degree of freedom, df)common for all strata.The method is as follows, using the designations of Table 20.2 for each 2×2 table, with E(a)designating the expected value of a (= n 1 n 3 /n). The subscript i refers to the ith 2×2 table:s = Σ[a i – E(a i )]The variance estimate of a i is:Vaiv = Σ(Va i )R 1 = Σ(a i *d i /n i )R 2 = Σ(c i *b i /n i )2( χ1196 .)n * n n * n=2n * n −i( )( 1 1)1 2 3 4χ 2 = s 2 /v. This χ 2 is based on only 1 df.© 2006 by Taylor & Francis Group, LLC

826 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONThe odds ratio is OR = R 1 /R 2The analysis thus gives an odds ratio and a χ 2 . Note that the χ 2 can be calculated even if, inindividual 2×2 tables, the expected number for one of the cells is small.If we apply this method to Table 20.3, we find that the OR approximates 1, and its statisticalsignificance disappears: OR = 0.94 with a 95% CI of 0.76 to 1.13.It should be noted that the conclusion supposes that all 2×2 tables estimate the same risk. Thismay not be so: The effect of smoking may be stronger at young age or vice versa. In order to testthis possibility, one can test for homogeneity between the strata. If numbers in each stratum arelarge enough, this can be done with Breslow and Day’s χ 2 test. If a normal approximation is notpermissible, there are also exact tests available (e.g., Zelen’s test). In the present example, the pvalue for homogeneity is p = 0.71, which indicates homogeneity, that is, the different age stratabehave in a similar way with respect to the effect of smoking. We thus have reason to believe thatthe “protective” effect of smoking was due to the confounding effect of maternal age.3. Logistic Multiple Regression AnalysisThe most popular current method to control for confounders is to use logistic multiple regression.This is a complicated technique that necessitates the use of computer programs, and many suchprograms are available. Here we can only indicate the principle of the basic type of analysis.A linear logistic model can be set up from an equation of the following type. Let p be the rateof occurrence of a specific event in the studied material (e.g., maternal smoking), thenln(p/(1 – p)) = α + β 1 *X 1 + β 2 X 2 + ... + β n X nfor n different variables, where X 1 can be the control or case status (0,1). Using an iterative technique,the best fit of the equation to the available data can be made, and one can also get estimates (witherrors) of the coefficients for each term, independent of the effects of the other terms. Eachcoefficient can be statistically tested against 0, and if it differs from 0, the variable has an effect.An estimate of an OR for each term (including the case control property) can also be obtained.Two variables may interact. For example, the risk of having a Down syndrome infant might beaffected differently by smoking in different age classes (even though we did not find this in thehomogeneity study above). This can be controlled by adding a term for interaction, consisting ofthe product of the two variables (age and smoking). If the interaction term becomes statisticallysignificant, it shows that smoking does have different effects in different age classes.The method has many advantages. Its greatest value compared with the stratified analysis isthat exact values for the variables can be used, e.g., exact maternal ages instead of 5-year classes.Classification of the data always carries the risk that the distribution within the class differs betweencases and controls. In each maternal age class, mothers of infants with Down syndrome may averagea few years older than control mothers. The standard analysis is, however, based on first-degreelinear regressions, and these may not be valid. The effect of maternal age on birth weight is anexample. This effect is U-shaped. When analyzed in a linear regression model, no maternal ageeffect is found because the linear distribution forced on a U-shaped function becomes practicallyhorizontal. Obviously, regressions of higher order can be constructed to take care of such phenomena,or the variable distribution can be transformed to linearity.4. Analysis of Matched Pairs or TripletsWhen data have been collected from matched cases and controls, those confounders that werematched for will have been eliminated (if matching has been effective). Sometimes the matching© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 827Table 20.4CaseSmoking among women who hadan infant with Down syndromeand age-matched controlsControlNumberof PairsSmoking Smoking n 1 = 38Not smoking Not smoking n 2 = 411Smoking Not smoking n 3 = 145Not smoking Smoking n 4 = 158Source: Data from the Swedish Medical BirthRegistry, 1983–1990.is “broken,” and the two sets of data (cases and controls) are analyzed as described above. Thisusually reduces the efficiency (if the matching has been meaningful). If the matching is donecorrectly, the variation between individuals within the strata used for matching is less than the totalvariation in the population. Differences between cases and controls can be revealed against thebackground of the low variation within the stratum but may not be observable when compared withthe large variation in the total population. Effectively matched materials should therefore beanalyzed without breaking the matching.There are many methods to analyze such data sets. Table 20.4 shows a matched sample ofDown syndrome infants and maternal age-matched controls, with information on maternal smoking.The material can be divided into four groups: (1) both case and control mothers smoked (++), (2)neither smoked (--), (3) the case mother but not the control mother smoked (+-), and (4) the controlbut not the case mother smoked (-+). If smoking was unrelated to the risk of having a Down infant,n 3 = n 4 , and the probability can be estimated that the observed values of n 3 and n 4 are taken fromthe binomial distribution of 50% of the total, n 3 + n 4 (which can be expressed as Bin(n 3 + n 4 , 0.5)].This probability can easily be evaluated (e.g., from a binomial table).The exact formula is: p = n!*p x *(1 – p) n–x /(x!*(n – x)!), where n is the total number of pairs,p = 1 – p = 0.50, and x is the observed number of pairs with characteristics +-. Note that the pvalue also has to be calculated for still more skewed distributions (x + 1/n – x – 1; x + 2/n – x –2, etc., if x > n/2). It should be noted that the two concordant groups (++ and – –) do not contributeto the evaluation of the effect studied.A similar exact analysis can be made with triplets (or a mixture of pairs and triplets). Whennumbers are large, approximate methods can be used instead. One much used technique is theMcNemar test, which can be used both for pairs and triplets. The formula for pairs is: If there aren 3 +– pairs and n 4 –+ pairs, then χ 2 = (n 3 – n 4 ) 2 /(n 3 + n 4 ). The odds ratio is then OR = n 3 /n 4 . Fromthe example in Table 20.4, we see that χ 2 = (145 158) 2 /(145 + 158) = 0.50, which is far fromsignificant, and the odds ratio is 0.92. Its 95% CI, calculated as described above (see p. 827), is0.72 to 1.16.With triplets, the following will be informative (triplets with all exposed or all nonexposed arenot informative): +–– (n 1 ), ++- (n 2 ), –++ (n 3 ), –+– (n 4 ). Then χ 2 = (2*n 1 – n 4 + n 2 – 2*n 3 ) 2 /(2*n 1+ n 2 + n 3 + n 4 ). The estimate of the OR is complicated but can be made.Another possibility is to use the Mantel-Haenszel test described above. There is no minimumsize limit for a stratum in that technique, and in principle a stratum can be a pair (or a triplet or aquadruplet). If pairs are analyzed, a 2×2 table for a +- pair (see above) will be:Exposed Unexposed TotalCase 1 0 1Control 0 1 1Total 1 1 2© 2006 by Taylor & Francis Group, LLC

828 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFor a triplet with one case and one control exposed, the table will look like:Exposed Unexposed TotalCase 1 0 1Control 1 1 2Total 2 1 3There are also methods available to apply logistic multiple regression techniques to matchedmaterials.5. The Multiple Testing ProblemCase control studies are rarely made with one specific question in mind. Quite often, informationis collected on a number of exposures. This may be because some of these exposures may beimportant confounders to the one of real interest, but very often many different exposures arestudied simultaneously. For example, one may wish to know if there is any occupation or maternaldrug use that is associated with a specific outcome.It is important to realize that if many different statistical tests are used on the same material,“significant” results may be obtained merely by chance. Statistical significance merely means thatsuch a deviant outcome is likely to be obtained by chance only once in 20 tests (α = 0.05). Forexample, if 20 different drugs are studied, it is expected that one will deviate (either as a risk or aprotective factor), and more than one may well do so. Great care must be taken to avoid overinterpretationof statistical tests that are not made on prior hypotheses. Ideally, the project plan stateswhich test or tests will be made. All “significant” findings obtained by the use of any additionaltesting should be regarded as “fishing expedition” results that may suggest new studies on freshmaterial.A similar situation will occur when 20 different studies of one drug-malformation relation areperformed. Very likely one of them will end up with a “statistical significance” merely by chance.A. PrinciplesVI. COHORT STUDIESA cohort study can be thought of as a study of a group of individuals characterized by a specificsituation. They may be born in the same year (a birth cohort), they may have been similarly exposed(e.g., have worked in a specific occupation or have taken a specific drug), or they may even becharacterized by having had an infant with a specific malformation (e.g., mothers of Down syndromeinfants). The question which is asked is whether the reproductive outcome of these women differsfrom the expected one. For instance, will women who have taken an antiepileptic drug have anincreased risk of having a malformed infant? Or will mothers of infants with Down syndrome havean increased risk of having an infant with a malformation in future pregnancies, compared withthe risk of “normal” women?A cohort study can thus tell us about many different reproductive outcomes, e.g., infertility,spontaneous abortion, low birth weight, congenital malformations, mental retardation, or childhoodcancer. The multiple testing situation, commented upon in the section for case control studies, willalso be relevant in this situation. For example, if 20 different types of malformations are studied,it is rather likely that one or more will show a “significantly” high or low incidence.© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 829B. The Formation of CohortsIn the classical cohort study, a group of persons is identified because of an exposure and is thenfollowed with respect to the development of a disease, but this is seldom the case in reproductiveepidemiology. As exposures during pregnancy are of major interest, this approach necessitates thatwomen are followed into pregnancy with respect to their exposures. Even when this approachmakes it possible to obtain rather exact exposure information, it will usually result in small numbersof subjects; therefore, it remains difficult to study rare outcomes, such as congenital malformations.Very often, retrospective cohorts are formed instead, e.g., women who work at a specificworkplace, and information on previous pregnancies and their outcome is identified (ideally, thisshould include determination of whether exposure occurred during those pregnancies). There is aclear danger in this approach, notably with respect to occupational exposures. If women have atendency to stop working after childbirth (which is often the case and was previously even moreprevalent), women remaining in the workforce will be selected for previous poor reproductiveoutcome or subfertility. Any such cohort will thus have a tendency to show an abnormally highincidence of reproductive failures. This may also be true when a specific geographic area is studied;if the area is less suitable for families with small children, a differential emigration from the areamay occur after successful pregnancies, leaving couples with poor reproductive outcome.A third possibility is that the cohort is defined from recorded exposures during the pregnanciesto be studied, e.g., drugs taken, occupations held, place of living. The problem lies in identifyingsuch cohorts. Sometimes this can be done by use of prospective or retrospective records obtainedfor all pregnant women (e.g., by interviews in early pregnancy or after delivery), and sometimesit can be achieved by linkage of different registries, e.g., census information linked to medical birthregistry information.C. Comparison MaterialThe pregnancy outcome in the defined cohort must be compared with that of another group. Quiteoften, a comparison is made with the expected outcomes known from the entire population or areasonable sample of the population. If, for instance, the perinatal death rate is known in thepopulation, the rate in a specific cohort can be evaluated. The problem resembles that in case controlstudies; the cohort under study may have characteristics that by themselves influence pregnancyoutcome, e.g., maternal age distribution or socioeconomic level. This may be avoided by a methodsimilar to that used in the matched case control study, i.e., the comparison with another cohortshould be selected to ensure as many similarities with the study cohort as possible except for theexposure under study. If a specific industrial occupational exposure is of interest, women with otherindustrial jobs may form a suitable cohort for comparison. If nurses working in anesthesiology arestudied, nurses working in wards of internal medicine might be chosen as controls. 61 As when sickcontrols are used in case control studies, it must be remembered that the cohort of comparison mayalso be exposed to reproductive hazards.There are two main problems with the cohort approach: the difficulty in definition and selectionof cohorts, and the problem with size. One needs very large numbers to study rare outcomes, suchas specific congenital malformations, and even large cohorts may be insufficient. We made a studyof reproductive outcome among women working as dentists, dental assistants, or dental technicians,because of an alleged increased risk of having an infant with spina bifida. 62 We studied more than8000 such births and found 3 infants with spina bifida when the expected number was 4.2. Wecould thus not confirm the alleged risk, but in spite of the large size of the study group, it is possiblethat an increased risk exists. This is true because the observed number of 3 may be a randomlylow sample estimate of a true expected number of 8.8 (i.e., a doubling of the risk).© 2006 by Taylor & Francis Group, LLC

830 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 20.5OutcomeInfant mortality when the mother worked in thechemical industry with a comparison group of allwomen working in industryMother Workedin Chemical IndustryMother Workedin Any IndustryNumber of infants 221 21,605Stillbirths 3 123Early neonatal deaths 5 82Later deaths 2 70Total deaths 10 (4.5%) 275 (1.3%)Source: Swedish data, infants born in 1971, 1981, or 1986. Källén, B.and Landgren, O., J. Occup. Med., 36, 563, 1994.D. Statistical Analysis of Cohort DataIn principle, the statistical analysis is similar to that for unmatched case control studies, but risksare often expressed as risk ratios instead. In the exposed cohort, the risk is measured as the numberof adverse outcomes divided by the total number of pregnancies or infants. This risk can becompared with the risk in the total population and expressed as a risk ratio (the statistical uncertaintyof the risk ratio will be nearly completely due to the estimate from the exposed cohort); alternatively,a risk ratio can be calculated between the risk in the exposed cohort and that in a cohort ofcomparison. Table 20.5 shows an example: infant and child deaths recorded in two cohorts ofwomen working in industry, the first in the chemical industry and the second (control cohort) inother industries. 63 The table shows that among 221 infants born to women working in the chemicalindustry, 10 died during the observation period (4.5%), while in the comparison cohort of 21,605infants, 1.3% died. The risk ratio is thus 3.5.In all such comparisons it is important to be certain that outcome information is of similarquality in the cohorts or in the cohort and the population. If detailed follow-up has been made ofeach child in the cohort, risks from that cohort cannot be compared with risks derived from sourcessuch as birth registries or congenital malformation registries, where ascertainment is often incompleteand risks therefore underestimated.The statistical significance of risk differences is evaluated with χ 2 tests or exact tests, asdescribed in the section on case control studies. The effect of confounding must again be takeninto consideration. If the study cohort, for example, differs in maternal age or parity distributionfrom the material used for comparison, this must be compensated for. This can be done in a similarway as in the case control study, using the Mantel-Haenszel procedure or a regression method. Ifthe comparison group is very large (e.g., the total population or as in Table 20.5), it is possibleinstead to calculate the expected number of outcomes for the study population, taking maternalage, parity, or other confounders into consideration. The observed number can then be comparedwith the expected number as a ratio that will represent the risk ratio.The observed-to-expected ratio is sometimes expressed multiplied by 100 as an SMR (standardmorbidity or mortality rate). In this situation, the confidence interval is most easily estimated fromthe observed number of reproductive outcomes: if this is a rare event (less than 10% of all), it willbehave as a Poisson distributed variable, and the 95% confidence interval can be determined as±2 n , where n is the observed number. These two values can be divided by the expected numberto obtain the confidence interval of the risk ratio. If n is small (less than 5), an exact calculationshould be made instead (or a Poisson table can be used). In the example in Table 20.5, the expectednumber of deaths is 2.8, and a normal approximation is not allowed. The 95% CI of the observedvalue of 10 is 4.8 to 18.4, and that of the risk ratio therefore is 1.7 to 6.6.© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 831Table 20.6Smoking StatusPerinatal death rate according to maternalsmoking in early pregnancyTotalBirths (No.)PerinatalDeaths (No.) %Not smoking 62,993 361 0.57Smoking < 10 cigs/day 16,571 130 0.78Smoking ≥ 10 cigs/day 10,606 88 0.83Source: Swedish infants born in 1985; data from the MedicalBirth Registry.Sometimes cohorts with different amounts of exposure are compared. An example is nonsmokingwomen, women smoking less than 10 cigarettes per day, and women smoking 10 or morecigarettes per day. Table 20.6 shows an example, stating the number of perinatally dead infants ineach cohort. In order to see whether a trend exists in such a table, the Armitage-Cochran test fora trend in a frequency table can be used. In principle, this test compares the χ 2 in the total materialwith that around a straight line fitted to the data, and the difference is due to the linear regressionin the material. For low numbers, exact trend tests should be used. The trend analysis of Table 20.6gives χ 2 = 14.6 at 1 df, p < 0.001.VII. NESTED CASE CONTROL STUDIESWe have seen advantages and disadvantages with the two main epidemiological techniques: casecontrol and cohort studies. They can sometimes be combined in a rather efficient way in what maybe called a nested case control study. The principle is that by making a crude cohort with a highprobability of exposure and then making a case control study within the cohort, the size of thematerial needed for study can be considerably reduced. All types of reproductive outcomes ofrelevance to the exposure can be studied.The definition of the crude cohort can, for example, be an occupation, perhaps identified fromregistries or by linkage of different registries. Such a cohort may not be “true,” in that all womenidentified actually had the occupation during pregnancy, but they are relatively likely to have hadit. Furthermore, only some of the working women may have the exposure of interest. For example,one may be interested in women exposed to plastic monomers during pregnancy. One can thenform a crude cohort of all women working in the plastic industry, and some but not all will havebeen exposed to the monomers. 64This crude cohort is then analyzed, and adverse outcomes are identified. Cases are selectedamong these outcomes (perhaps after exclusion of inherited conditions or other causes that obviouslyhave no relationship to the workplace), and one or more controls is selected for each casefrom the crude cohort (using the control definition that was discussed above). Then a regular casecontrol study is made, retrieving exposure information in the best possible way. This leads to tworesults: (1) that the control group estimates the proportion of the crude cohort that was actuallyexposed and (2) a comparison of exposure rates can be made between the case and control groups.If close to 100% of controls were exposed, obviously no difference from the case group can beexpected (unless the exposure was protective), but then it can be concluded that the outcome ofthe total crude cohort is representative for the exposure under study. If (as is usually the case) itappears that only a proportion of the controls were exposed, an odds ratio can be calculated in thestandard way, comparing case and control exposures. In this situation, the study can be regardedas any case control study, but with the advantage that primary selection of the crude cohort hasmarkedly increased the exposure rate in comparison with that of the total population.© 2006 by Taylor & Francis Group, LLC

832 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. General ConsiderationsVIII. CONFOUNDINGWe have already mentioned the phenomenon of confounding: a variable that affects both theexposure rate and the outcome rate. An example mentioned was maternal age, which affects bothsmoking rate and the risk of having an infant with Down syndrome. Many possibly importantconfounders can exist in reproductive epidemiology; some thought should be given to the biologicalmeaning of the statistical relationships, otherwise completely misleading results may be drawn. Aclassical example is maternal smoking and perinatal death rate, with birth weight as a confounder.Smoking causes a decrease in birth weight and an increase in perinatal mortality. Low birth weightis an important determinant of perinatal death risk and could therefore be regarded as a confounder.By stratifying for birth weight, not only does the effect of smoking on perinatal death disappear,but also an apparent protective effect appears. This does not mean that maternal smoking does notaffect perinatal death risk, but it means that it does so to a large extent by changing the birth weightdistribution. After birth weight stratification, one will compare small but otherwise healthy infantsof smoking women with small but sick (growth retarded) infants of nonsmoking women. This willnaturally result in a seemingly protective effect of smoking on perinatal death rate. Similar effectswere seen in a study on socioeconomic characteristics and infant mortality. 27B. Maternal Age and Reproductive HistoryBoth of these variables are important confounders in reproductive epidemiology, especially instudies of spontaneous abortion rate, birth weight distribution, and perinatal mortality. They can,however, have a complex effect on outcome, and it is not absolutely clear that a compensation forthem should always be made. Let us look at the group of relatively old primiparous women (say,above 35 years of age). This group contains two main subgroups: women who have had difficultybecoming pregnant and therefore give birth for the first time late in reproductive life, and womenwho have voluntarily postponed pregnancy because of considerations such as education or career.These two groups will have different distributions in, for example, different occupational groups.Differences recorded in reproductive outcome will not be due to actual differences in exposuresrelated to occupation, but to the fact that incomparable groups are compared. If the presence oflong-standing subfertility is recorded in an adequate way, this can be entered into the model, butoften such information is not available.Previous reproductive history, including subfertility, may be an important confounder in studieson occupational exposures, as was pointed out above. It may also affect drug usage, smoking, andother possible exposures. It can be shown that there is a positive association between the use ofantidepressive drugs and the occurrence of hypospadias. This is completely explained by the excessuse of these drugs by subfertile women and an association between subfertility and hypospadias. 65C. Maternal DiseasesMaternal diseases are obvious confounders in the study of the reproductive effects of maternal useof drugs during pregnancy. If the disease directly or indirectly (e.g., by genetic linkage) increasesthe risk for a congenital malformation, any drug used for that disease will appear to be a teratogen.A classical example is insulin and maternal diabetes. It is generally thought that the increased riskfor malformations seen in infants of diabetic mothers 66 is due to the disease and not to insulin;actually, good control of the disease with insulin may reduce the risk. Similarly, epilepsy as suchmay (perhaps mainly because of linked genes) increase the risk for facial clefts, 67 even thoughmaternal anticonvulsant drug use has an effect of its own.© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 833When different drugs are used for the same disease or when drugs are not always used, it isoften possible to control for the effect of the disease in an analysis of drug effects. Sometimesstudies involving different populations with different therapeutic traditions can help.In the study of other environmental conditions, drugs usually appear as a very weak confounder.Very few drugs have any marked effect on the teratogenic risk, and in these few instances, typicaloutcomes that can be identified are usually seen. In studies of occupational risks or the risk ofexposures from the general environment, it is generally not worthwhile to record drug exposures.Some chronic diseases (e.g., diabetes and epilepsy) have distinct teratogenic effects, but they occurat low frequency. In studies of occupational exposures, such chronic diseases may give a “healthyworker” effect if the disease is not compatible with the occupation under study, but this effect isusually weak.D. Social CharacteristicsSocial characteristics may have an effect on gestational duration, birth weight, stillbirth rate, andinfant death rate. In a few circumstances, an effect on specific congenital malformations is seen(e.g., neural tube defects in some populations). 68 Highly skewed social distributions are commonlyobserved in studies of occupational exposures and general environmental exposures, and it is oftendifficult to distinguish between the effect of an occupation and the consequences of the socialcharacteristics of that occupation. When very specific exposures are studied, this can be done byselecting women with a similar but unexposed occupation as controls, but even then differencesmay remain. There are also difficulties in the characterization of the social level, as most suchgroupings are based on occupation classifications. One always has to consider the possibility thatdifferences demonstrated between groups, such as occupational groups, are not due to actualdifferences in exposure but to social characteristics, including lifestyle (e.g., smoking, alcohol anddrug abuse).As a proxy for socioeconomic level, parental education is sometimes used. In many societies,paternal education seems to have the strongest effect; in others (e.g., the Nordic) it appears thatmaternal education is the stronger determinant of, for instance, birth weight. Especially for youngwomen, who give birth before finishing their training, education level at the time of delivery maybe less informative than final education as a proxy of social level. When registries are available ofthe education level of the members of the population, it may sometimes be more effective to use“final” education, even if it is attained many years after the delivery.E. Biological CharacteristicsSome biological characteristics of the women may appear as confounders. As pointed out above,subfertility not only can affect reproductive outcome, but also can affect the probability of staying inthe workforce or living in a specific geographical area. Physical fitness may be a prerequisite for aspecific job (e.g., physiotherapist), and this may result in a positive reproductive outcome comparedwith the general population. 69 Exposures within the job may nevertheless carry a risk that is difficultto identify when comparisons are made between the exposed group and the population.F. Incomplete Compensation of a ConfounderIt is often thought that the possible effect of a confounder can be disregarded if the confoundingvariable is entered into the statistical model, but this is not always true. In a study on socioeconomiclevel and perinatal deaths, a standardization was made for maternal smoking as recorded in earlypregnancy. 32 Standardization for smoking did reduce the risk ratio, but only partly. There are atleast three possible explanations. The obvious one is that socioeconomic level plays a role, irrespective© 2006 by Taylor & Francis Group, LLC

834 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof smoking. Another is that women of different social classes have different tendencies to stopsmoking during pregnancy. The third is that women of different social classes have differenttendencies to underreport their smoking when answering an interviewer in early pregnancy. Thestatistical compensation of the confounder will be incomplete if the registration of the confounderis incomplete or biased! Further, it is most important to note that there is always the possibilitythat confounders exist that have not been realized or for which one cannot control.IX. SURVEILLANCEA. Birth Defects Surveillance in the PopulationPopulation surveillance of reproductive outcomes has concentrated on congenital malformations.Most such activities began after the thalidomide tragedy. The introduction of a new drug, thoughtto be low in toxicity, caused a large number of infants with previously very rare malformations,especially absent or strongly reduced limbs (amelia, phocomelia). It took some time before this“epidemic” was detected, and thousands of children were damaged. The congenital malformationregistries that have been built up usually had surveillance as a primary purpose, but — as pointedout above — are also excellent sources for case finding for epidemiological studies. A crudesurveillance of perinatal or neonatal deaths occurs in most developed countries, but more elaboratesystems aiming at the detection of changes in specific subgroups are rare. Such an effort waspublished from Sweden, based on data in computerized perinatal registries. 30A change in the prevalence at birth of a specific congenital malformation can occur as a resultof the introduction of a new teratogenic agent in the environment. If such an agent causes a relativelycommon malformation, it is extremely difficult to detect the change in rate. 70 Only if the agent hasa very strong teratogenic effect or exposure to it becomes widespread in the population can effectson the population rates of common malformations be detected. Minor changes in the rates will bemasked by the random fluctuations in the number of malformed babies born, and this maskingeffect is exacerbated by changes in ascertainment of malformations. Prenatal diagnosis followedby selective abortion will also complicate an analysis. It is rather unlikely that population surveillanceof common malformations will ever be able to detect a new teratogenic agent causing amoderate increase in the rate of common malformations.What population surveillance can detect is the appearance of previously rare congenital malformationsor combinations of malformations. Many of the detected human teratogens had sucheffects, e.g., amelia or phocomelia seen after thalidomide, congenital cataract seen after rubella,and the malformation syndrome typically seen after isotretinoin. This, of course, means that therisk increase caused by these agents is enormous. For amelia or phocomelia the “normal” populationrate is perhaps 3 per 100,000 births, while after thalidomide exposure something like 6% of infantshad such limb defects, a 2000-fold increase.Congenital malformation surveillance is therefore less a matter of statistical analysis of therecorded rates of common malformations than it is the search for previously unusual malformationsor combinations of malformations. It must be realized that the absence of a noticeable change inthe occurrence of a specific malformation does not show that no new teratogen has been introduced.This only implies that if there was such an introduction, its impact on the total malformation ratewas small. We can illustrate this with valproic acid, an anticonvulsant that causes spina bifida insome 1% to 2% of exposed embryos. 71 In the Swedish population, spina bifida occurs at anapproximate rate of 1 in 2000 (50 to 60 cases a year). If valproic acid were suddenly introduced,and 20% of all epileptic women used the new drug (a rather unlikely supposition), among 400epileptic women giving birth each year, the new drug would cause one new case of spina bifida.Such a modest increase would be completely impossible to detect because of the random variability© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 835of the number of spina bifida cases born each year (with 55 as a mean and a 95% CI of 41 to 72).On the other hand, if thalidomide had been used by 80 pregnant women, perhaps 5 would haveborne infants with amelia or phocomelia, more than doubling the rate of these very unusualmalformations.B. Surveillance after Specific ExposuresThe effectiveness of the surveillance process can be markedly increased if it is restricted to specificgroups of women characterized by exposure. Surveillance of the reproductive outcomes of womenwith specific diseases (e.g., epilepsy), with specific occupations (e.g., working in the plasticsindustry), or living in areas with specific pollution (e.g., exposed to radioactive fallout after theChernobyl accident) has a much greater chance to detect changes and identify the effect of a newdrug, a changed work environment, or a specific pollutant.Such specific surveillance can be made, but it should be realized that a very large number ofpossible objects of surveillance exist, and surveillance can only give a signal that something maybe wrong. Suppose, for example, that during the surveillance of infants born of epileptic women,the rate of infants with spina bifida suddenly doubled, and half of these cases had been exposedto a specific anticonvulsant drug; this could be due to a causal association, but specific investigationof the subject should be made on an independent population.C. Outcome-Exposure SurveillanceAnother way to increase the power of routine surveillance is to make an on-going surveillance ofoutcome-exposure associations, e.g., maternal drug usage and congenital malformations in theoffspring. It may be based on the reporting of adverse drug reactions or similar systems, but it canalso be based on a systematic data collection, e.g., the MADRE system run by the InternationalClearinghouse for Birth Defects Monitoring Systems. 72,73 The Swedish Medical Birth Registryprovides an opportunity for an ongoing surveillance similar to that just described.D. Statistical ConsiderationsThere is a statistical problem in all surveillance processes because no prior hypothesis exists. Manydifferent types of malformations or other adverse outcomes are monitored during an extended time,and an increase in the rate of any malformation will give a signal. It is therefore not possible torun statistical tests on the observed changes. For example, during a 3-month period 10 infants areobserved with omphalocele, and the expected number of such infants is only 3.7. One cannot thensay that the probability for that increase being due to chance is only p = 0.01 (based on a twotailedPoisson distribution). There are dozens of malformations that can show such a change,monitoring may have occurred for many years, and there may be many quarters in which such achange could occur. Therefore, such a large difference between the observed and expected numbersof infants with omphalocele can very well be due to chance, in spite of the “statistical significance.”Any such observation is only an indication that something may be wrong and that perhaps aspecific study should be conducted. Personally, I am less alarmed by a finding of this type than afinding of the birth of two or three infants with cyclopia during the period of a few weeks, eventhough this also may have been due to chance.A similar situation exists in surveillance of drug-malformation combinations. Many drugs aresurveyed, and many types of malformation are studied; thus, a seemingly strong associationobserved between a specific drug and a specific malformation may well be random. Such anobservation can only serve as a signal, to be supported or rejected by other studies. Continuedsurveillance is one way to get new data, although at the price of a delay.© 2006 by Taylor & Francis Group, LLC

836 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONA. Definition of a ClusterX. CLUSTER ANALYSISA cluster is a time-space aggregation of adverse outcomes, e.g., infertility, congenital malformations,perinatal deaths, or miscarriages. It is sometimes observed at surveillance (see the precedingsection) but usually is noted in the “periphery” where the cluster occurs.B. Clusters as Chance Phenomena or ArtifactsA cluster can be quite impressive but nevertheless be due to chance. According to the same reasoningas that presented above, even strong clusters not only may arise by chance but should arise due tothe random fluctuation of numbers. Technical explanations for a cluster may also exist, such aslocal peculiarities in the diagnosis or registration of congenital malformations. Other clusters maybe caused by local teratogenic factors; that is the obvious reason for studying clusters at all. It isseldom possible to directly say whether a cluster is random or actually caused by a local factor. Inmany circumstances possible harmful factors can be suggested, irrespective of whether they causedthe cluster, something else did, or the cluster merely occurred by chance.Clusters are of interest in the study of the affect of the environment on reproduction but mustbe handled with great care. The first task is to see whether the cluster can be explained by sometechnical anomaly. If not, one should analyze the cases entering the cluster in order to look forsimilarities. A cluster consisting of infants with very different types of malformations is lessimpressive than a cluster where the malformations of the infants seem to have something in common.The analysis is less a statistical exercise than an evaluation of the biological plausibility of theobserved events.C. Exposure Analysis of ClustersIf one decides to follow up on a cluster, exposure information becomes important. Sometimes atentative cause can be identified, such as emission from an industry. If so, a case control study canbe carried out in the area where the cluster appeared. The study might compare the distance to thesource of emission from the homes of the cluster cases and those of controls selected from thesame geographical area. Other examples of tentative exposure sources are local water supply oroccupational exposure in a local industry. No tentative explanation may exist, or the explanationis not supported by exposure data. In order to get new hypotheses to explain the cluster, detailedinterviews of the (usually few) cluster cases can be conducted. If an exogenous factor of moderatestrength caused the cluster, a majority of the cluster cases should have been exposed to it, andformal control materials are often not necessary. If a possible explanation is found, the hypothesishas to be tested on a new population, if that is possible.The most difficult situation is when the suspected teratogenic factor is ubiquitous, e.g., awidespread air or water pollutant. All pregnancies will then be exposed, and the resulting clustercan be due to a relatively low teratogenic property. Such a teratogen can never be detected in acase control study within the area. If it is possible to identify other areas with the same type ofexposure, follow-up studies can be made in them to support or reject the hypothesis, but it canalways be argued that the exposure situation was not identical.D. Clusters and StatisticsVery often one has tested the “statistical significance” of a cluster, e.g., by applying a Poissondistribution test. Suppose, for example, that in a specific geographic area, eight infants were born© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 837with spina bifida and only 2.8 were expected from the country rate of the malformation. A Poissonmodel gives p = 0.008, which seems rather convincing. This is the probability that eight casesoccurred (with 2.8 expected) if this area had been selected for analysis for some specific reason,e.g., that an industry emitted something that was thought to cause spina bifida.In cluster analysis the situation is reversed; one observes an accumulation of cases and secondarilytries to fit that to an exposure situation. There are numerous areas where an aggregation ofspina bifida cases could have occurred, and there are many different types of malformations orother adverse reproductive outcomes that could have been clustered. Thus, it is not unlikely thatan aggregation of the magnitude described will occur in one of the areas for one of the malformations.The p value reached therefore does not have the common meaning of a p value, estimatingthe probability for the occurrence to be due to chance.We once conducted a study of the distribution per municipality of infants with Down syndromein Sweden. The distribution exactly followed what could be expected from a Poisson distribution,but nevertheless when specific municipalities and time periods were scrutinized, rather impressiveclusters were found. During the years 1978 to 1986 in one such municipality, 11 cases were foundwhen the expected number was 4.0 (p = 0.003); in another municipality, 7 infants with Downsyndrome were born in 1978 against the expected number of 1.5 (p = 0.0009). These strongaggregations can be due to chance, but obviously they can also have a specific cause, and the lattercluster resulted in an intense follow-up (without any reasonable cause being identified).XI. CONCLUDING WORDSEpidemiological studies of human reproduction are burdened with many problems and difficulties.It is rare that single epidemiological studies give a final answer on hazards associated with specificexposure. Each epidemiological study should be regarded as a piece in a jigsaw puzzle. When manypieces have been put together, perhaps together with information from other sources, a picture ofthe truth develops. It is not necessarily the truth, but hopefully resembles the truth as much aspossible. If too much importance is paid to the statistical significance levels in a single study, thereis a definite risk that the importance of the findings will be overestimated. One must be careful indrawing practical conclusions from single studies if they do not unequivocally show very high risksas well as biological plausibility. One phenomenon should be realized — mass significance alsoexists with respect to scientific studies. If 20 different researchers investigate the same problem(e.g., a possible role between exposure to electromagnetic fields and child leukemia), it is quiteprobable that in one of the studies a “statistically significant” effect will be seen, even if no causalrelationship exists between exposure and outcome. To this is added the “publication bias” problem.Studies that show an association usually have an advantage in competing for publication in scientificjournals. Thus, there is a tendency for the first reports appearing in the literature to overestimaterisks, sometimes even when the risk does not actually exist.To the above is added the information problem, namely, how to convey to society and to thepublic the digested interpretation of the available evidence. There are special problems in informingthe public, who often cannot accurately appreciate the concept of risk. Individual risks may besmall, but the attributable risk for the outcome under study may be large in the total population.Individual risks may be high (and may influence decisions about termination of pregnancy), butthe total number of damaged infants born may be negligible. Epidemiological studies may be“negative,” but they do not prove that the exposure under study is harmless. To transform thescientific information into recommendations for society and the public is a delicate matter, andthe scientists who conducted the studies may not be the best persons to interpret their findingsin an unbiased way. Indeed, probably one of the strongest biases in the field is to favor one’sown study!© 2006 by Taylor & Francis Group, LLC

838 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONREFERENCES1. Cnattingius, S., Haglund, B., and Meirik, O., Cigarette smoking as a risk factor for late fetal deathand early neonatal death, Brit. Med. J., 297, 258, 1988.2. Jones, K.L., Smith, D.W., Streissguth, A.P., and Myrianthopoulos, N.C., Outcome in offspring ofchronic alcoholic women, Lancet, 1, 1076, 1974.3. Källén, B., Lithium therapy and congenital malformations, in Lithium in Biology and Medicine,Schrauzer, G.N. and Klippel, K.-F., Eds., Weinheim, New York, 1991, p. 121.4. Herbst, A.L., Ulfelder, H., and Poskanzer, D.C., Adenocarcinoma of the vagina. Association ofmaternal stilbestrol therapy and tumor appearance in young women, New Engl. J. Med., 284, 878, 1971.5. Whorton D., Krauss R., Marshall, S., and Milby, T., Infertility in male pesticide workers, Lancet, 2,1259, 1977.6. Palmer, J.R., Hatch, E.E., Rao, R.S., Kaufman, R.H., Herbst, A.L., Noller, K.L., Titusernstoff, L., andHoover, R.N. Infertility among women exposed prenatally to diethylstilbestrol, Am. J. Epidemiol.,154, 316, 2001.7. Joffe, M. and Barnes, I. Do parental factors affect male and female fertility? Epidemiology, 11, 6, 2000.8. Jensen, T.K., Scheike, T., Keiding, N., Schaumburg, I., and Grandjean, P., Selection bias in determiningthe age dependence of waiting time, Am. J. Epidemiol., 152, 565, 2000.9. Basso, O., Juul, A., and Olsen, J., Time to pregnancy as a correlate of fecundity: differential persistencein trying to become pregnant as a source of bias, Int. J. Epidemiol., 29, 856, 2000.10. Bergh, T., Ericson, A., Hillensjö, T., Nygren, K-G., and Wennerholm, U.-B., Deliveries and childrenborn after in-vitro fertilisation in Sweden 1982–1995: a retrospective cohort study, Lancet, 354, 1579,1999.11. Ghazi, H.A., Spielberger, C., and Källén, B. Delivery outcome after infertility — a registry study,Fert. Ster., 55, 726, 1991.12. Gray, R.H. and Wu, L., Subfertility and risk of spontaneous abortion, Am J Publ Health, 90, 1452, 2000.13. Ericson, A. and Källén, B., Congenital malformations in infants born after IVF: a population-basedstudy, Human Reprod., 16, 504, 2001.14. Akre, O., Cnattingius, S., Bergström, R., Kvist, U., Trichopoulos, D., and Ekbom, A., Human fertilitydoes not decline: evidence from Sweden, Fert. Steril., 71, 1066, 1999.15. Tielemans, E. Burdorf, A. Velde, E.R.T., Weber, R.F.A., Vankooij, R.J., Veulemans, H., and Heederik,D.J.J., Occupationally related exposures and reduced semen quality; a case-control study, Fert. Steril.,71, 690, 1999.16. Kolstad, H.A., Bonde, J.P., Spano, M., Giwercman, A., and Zschiesche,Y., Subfertility and risk ofspontaneous abortion, W., Kaae, D, Larsen, S.B., and Roeleveld, N., Change in semen quality andsperm chromatin structure following occupational styrene exposure, Int. Arch. Occup. Environ. Health,72, 135, 1999.17. Wennborg, H., Bodin, L., Vainio, H., and Axelsson, G., Solvent use and time to pregnancy amongfemale personnel in biomedical laboratories in Sweden, Occup. Environ. Med., 58, 225, 2001.18. Kurinczuk, J.J. and Clarke, M., Case-control study of leatherwork and male infertility, Occup. Environ.Med., 58, 217, 2001.19. Sandahl, B., Smoking habits and spontaneous abortion, Europ. J. Obst. Gynecol. Reprod. Biol., 31,23, 1989.20. Miller, J.F., Williamson, E., Glue, J., Gordon, Y.B., Grudzinskas, J.G., and Sykes, A., Fetal loss afterimplantation. A prospective study, Lancet, 2, 554, 1980.21. Juutilainen, J., Matilainen, P., Saarikoski, S., Läärä, E., and Suonio, S., Early pregnancy loss andexposure to 50 Hz magnetic fields, Bioelectromagnetics, 14, 229, 1993.22. Axelsson, G. and Rylander, R., Validation of questionnaire reported miscarriage, malformation andbirth weight, Int. J. Epidemiol., 13, 94, 1984.23. Hemminki, K., Niemi, M.-L., Soloniemi, I., Vainio, H., and Hemminki, E., Spontaneous abortion byoccupation and social class in Finland, Int. J. Epidemiol., 9, 199, 1980.24. Kline, J., Stein, Y.A., and Susser, M., Smoking: a risk factor for spontaneous abortion, New Engl. J.Med., 297, 793, 1977.25. Goldhaber, M.K., Pole, M.R., and Hiatt, R.A., The risk of miscarriage and birth defects among womenwho use visual display terminals during pregnancy, Amer. J. Industr. Med., 13, 695, 1988.© 2006 by Taylor & Francis Group, LLC

HUMAN STUDIES 83926. Haglund, B. and Cnattingius, S., Cigarette smoking as risk factor for sudden infant death syndrome:a population-based study, Amer. J. Pub. Health, 80, 29, 1990.27. Ericson, A., Eriksson, M., Källén, B., and Zetterström, R., Socio-economic variables and pregnancyoutcome. 2. Infant and child survival, Acta Paediatr. Scand., 79, 1009, 1990.28. Winbo, I.G.B., Serenius, F.H., and Källén, B.A.J., Lack of precision in neonatal death classificationsbased on the underlying cause of death stated on death certificates, Acta Paediatr., 87, 1167, 1998.29. Wigglesworth, J.S., Monitoring perinatal mortality. A pathophysiological approach, Lancet, 2, 684,1980.30. Winbo, I.G.B., Serenius, F.H., Dahlquist, G.G., and Källén, B., NICE, a new cause of death classificationfor stillbirths and neonatal deaths, Int. J. Epidemiol., 27, 499, 1998.31. Källén, B., A birth weight for gestational age standard based on data in the Swedish Medical BirthRegistry, 1985–1989, Europ. J. Epidemiol., 11, 601, 1993.32. Ericson, A., Eriksson, M., Källén, B., and Zetterström, R., Secular trends in the effect of socioeconomicfactors on birth weight and infant survival in Sweden, Scand. J. Soc. Med., 21, 10, 1993.33. Ericson, A., Gunnarskog, J., Källén, B., and Otterblad Olausson, P., A registry study of very lowbirthweight liveborn infants in Sweden, 1973–1988, Acta Obst. Gynecol. Scand., 71, 104, 1992.34. English, P.B. and Eskenazi, B., Reinterpreting the effects of maternal smoking on infant birthweightand perinatal mortality: a multivariate approach to birthweight standardizationm Int. J. Epidemiol.,21, 1097, 1992.35. Ericson, A., Eriksson, M., Källén, B., and Meirik, O., Birth weight distribution as an indicator ofenvironmental effects on fetal development, Scand. J. Soc. Med., 15, 11, 1987.36. Mulvihill, J.J., and Czeizel, A.A., 1983 view of sentinel phenotypes, Mutation Res., 123, 345, 1983.37. Källén, B., Knudsen, L.B., Mutchinick, O., Mastroiacovo, P., Lancaster, P., Castilla, E., and Robert,E., Monitoring dominant germ cell mutations using skeletal dysplasias registered in malformationregistries. An international feasibility study, Int. J. Epidemiol., 22, 107, 1993.38. Kieler, H., Cnattingius, S., Haglund, B., Palmgren, J., and Axelsson, O., Sinistrality — a side effectof prenatal sonography: a comparative study of young men, Epidemiology, 47, 618, 2001.39. Feldman, R.G. and White, R.F., Lead neurotoxicity and disorders of learning, J. Child Neurol., 7,354, 1992.40. Harada, M., Congenital Minimata disease: intrauterine methylmercury poisoning, Teratology, 18, 285,1978.41. Sutcliffe, A.G., Dsouza, S.W., Cadman, J., Richards, B., McKinlay, I.A., and Lieberman, B., Minorcongenital anomalies, major congenital malformations and development in children conceived fromcryopreserved embryos, Human Reprod., 10, 3332, 1995.42. Ericson, A., Nugren, K.G., Otterblad Olausson, P., and Källén, B., Hospital care utilization of infantsborn after in vitro fertilization. Human Reprod., 17, 929, 2002.43. Forsberg, J.-G. and Källén, B., Pregnancy and delivery characteristics of women whose infants developchild cancer, APMIS, 98, 37, 1990.44. Needleman, H.L., Gunnoe, C.G., Leviton, A., Reed, R., Peresie, H., Maher, C., and Barrett, P., Deficitsin psychologic and classroom performance of children with elevated dentine lead levels, New Engl.J. Med., 300, 689, 1979.45. Oryszczyb, M.P., Godin, J., Annesi, I., Hellier, G., and Kauffman, F., In utero exposure to parentalsmoking. Cotinine measurements and cord blood IgE, J. Allergy Clin. Immunol., 87, 6, 1991.46. Martínez-Frías, M.L. and Rodrigues-Pinilla, E., Epidemiologic analysis of prenatal exposure to coughmedicines containing dextromethorphan: no evidence of human teratogenicity, Teratology, 63, 38,2001.47. Boneva, R.S., Moore, C.A., Botto, L., Wong, L.Y., and Erickson, J.D. Nausea during pregnancy andcongenital heart defects: a population-based case-control study, Am. J. Epidemiol., 149, 717, 1999.48. Heinonen, O.P., Slone, D., and Shapiro, S., Birth Defects and Drugs in Pregnancy, Publishing SciencesGroup, Inc., Littleton, MA, 1977.49. Källén, B., Use of omeprazole during pregnancy — no hazard demonstrated in 955 infants exposedduring pregnancy, Obst. Gynecol., 96, 63, 2000.50. Nielsen, G.L., Sorensen, H.T., Larsen, H., and Pedersen, L., Risk of adverse birth outcome andmiscarriage in pregnant users of non-steroidal anti-inflammatory drugs: population based observationalstudy and case-control study, Brit. Med. J., 322, 266, 2001.© 2006 by Taylor & Francis Group, LLC

840 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION51. Czeizel, A.E., Rockenbauer, M., Sorensen, H.T., and Olsen, J., The teratogenic risk of trimethoprimsulfonamides:a population-based case-control study, Reprod. Toxicol., 15, 637, 2001.52. Hemminki, K., Niemi, M.-L., Saloniemi, I., Vainio, H., and Hemminki, E. Spontaneous abortions byoccupation and social class in Finland. Int. J. Epidemiol., 9, 149, 1980.53. Feldman, Y., Koren, G., Mattice, D., Shear, H., Pellegrini, E., and MacLeod, S.M., Determinants ofrecall and recall bias in studying drugs and chemical exposure in pregnancy, Teratology, 40, 37, 1989.54. Mackenzie, S.G., and Lippman, A., An investigation of report bias in a case-control study of pregnancyoutcome, Amer. J. Epidemiol., 129, 65, 1989.55. Källén, B., Castilla, E.E., Robert, E., Lancaster, P.A.L., Kringelbach, M., Mutchinick, O., Martínez-Frías, M.L., and Mastroiacovo, P., An international case-control study on hypospadias. The problemwith variability and the beauty of diversity, Europ. J. Epidemiol., 8, 256, 1992.56. Ericson, A., Eriksson, M., Källén, B., and Zetterström, R., Maternal occupation and delivery outcome:a study using central registry data, Acta Paediatr. Scand., 76, 512, 1987.57. Kurppa, K., Holmberg, P.C., Hernberg, S., Rantala, K., Riala, R., and Nurminen, T., Screening foroccupational exposures and congenital malformations. Preliminary results from a nationwide casereferentstudy, Scand. J. Work. Environ. Health, 9, 89, 1983.58. Ericson, A., Källén, B., and Löfkvist, E., Environmental factors in the etiology of neural tube defects:a negative study, Environ. Res., 45, 38, 1988.59. Knudsen, L.B., Possibilities in the study of adverse reproductive outcome in Denmark, based onregistries, in Proc. 1st Int. Meet. Genetic and Reproductive Research Society (GRERS), Mastroiacovo,O., Källén, B., and Castilla, E.E., Eds., Ghedini Editori, Milano, Italy, 1995, p. 35.60. Källén, K., Maternal smoking and orofacial clefts, Cleft Palate-Cran.fac. J., 34, 11, 1997.61. Ericson, A. and Källén, B., Hospitalization for miscarriage and delivery outcome among Swedishnurses working in operating rooms 1973-1978, Anesth. Analg., 64, 981, 1985.62. Ericson, A. and Källén, B., Pregnancy outcome in women working as dentists, dental assistants ordental technicians, Int. Arch. Occup. Environ. Health, 61, 329, 1989.63. Källén, B. and Landgren, O., Delivery outcome in pregnancies when either parent worked in chemicalindustry. A study with central registries, J. Occup. Med., 36, 563, 1994.64. Ahlborg, G., Jr., Bjerkedal, T., and Egenaes, J., Delivery outcome among women employed in theplastic industry in Sweden and Norway, Amer. J. Industr. Med., 12, 507, 1987.65. Källén, K., Role of maternal smoking and maternal reproductive history in the etiology of hypospadiasin the offspring, Teratology, 66, 185, 2002.66. Åberg, A., Westbom, L., and Källén, B. Congenital malformations among infants whose mothers hadgestational diabetes or preexisting diabetes, Early Human Dev., 61, 85, 2001.67. Friis, M.L., Broeng-Nielsen, B., Sindrup, E.H., Lund, M., Fogh-Andersen, P., and Hauge, M.,Increased prevalence of cleft lip and/or cleft palate among epileptic patients, in Epilepsy, Pregnancy,and the Child, Janz, D., Dam, M., Richens, A., Bossi, L., Helge, H., and Schmidt, D., Eds., RavenPress, New York, 1982, p. 297.68. Jones, W.H., Myelomeningocele and social class, Dev. Med. Child. Neurol., 22, 244, 1980.69. Källén, B., Malmquist, G., and Moritz, U., Delivery outcome among physio-therapists in Sweden: Isnon-ionizing radiation a fetal hazard? Arch. Environ. Health, 37, 81, 1982.70. Källén, B., Population surveillance of congenital malformations. Possibilities and limitations. Reviewarticle, Acta Paediatr. Scand., 78, 657, 1989.71. Lammer, E.J., Sever, L.E., and Oakley, G.P., Jr., Teratogen update: valproic acid, Teratology, 36, 465,1987.72. Robert, E., Handling surveillance types of data on birth defects and exposures during pregnancy.Reproductive toxicology review, Reprod. Toxicol., 6, 205, 1992.73. Arpino, C., Brescianini, S., Robert, E., Castilla, E.E,,Cocchi, G., Cornel, M.C., de Vigan, C., Lancaster,P.A., Merlob, P., Sumiyoshi, Y., Zampino, G., Renzi, C., Rosano, A., and Mastroiacovo, P., Teratogeniceffect of antiepileptic drugs; use of an international database on malformations and drug exposure,Epilepsia, 41, 1436, 2000.© 2006 by Taylor & Francis Group, LLC

CHAPTER 21Developmental and Reproductive Toxicity RiskAssessment for Environmental Agents*Carole A. Kimmel, Gary L. Kimmel, and Susan Y. EulingCONTENTSI. Introduction ........................................................................................................................841A. Definitions..................................................................................................................842B. Historical Perspective on Risk Assessment Approaches: 1983 to 2002 ..................842II. Framework for Developmental and Reproductive Toxicity Risk Assessment..................843A. Planning and Problem Formulation ..........................................................................844B. The Risk Assessment Process ...................................................................................8451. Hazard Characterization ......................................................................................8452. Dose-Response Assessment.................................................................................8563. Exposure Assessment ..........................................................................................8594. Risk Characterization ..........................................................................................860III. Emerging Issues for Risk Assessment...............................................................................864A. Research Efforts to Develop Human Data on Developmental and ReproductiveEffects of Environmental Exposures — Contributions to Risk Assessment............864B. Using Mode of Action and Pharmacokinetic Data in Risk Assessment ..................865C. Harmonization of Approaches for Human Health Risk Assessment .......................866D. Use of Genomics, Proteomics, and Other “Futures Technologies” .........................866E. Implications of Benefits Analysis for Risk Assessment ...........................................867References ......................................................................................................................................867I. INTRODUCTIONThe approach to risk assessment for environmental agents has progressed significantly over the lasttwo decades since the publication of the National Academy of Sciences landmark report: RiskAssessment in the Federal Government: Managing the Process. 1 Although risk assessmentapproaches can be applied to all classes of chemical or physical agents, this chapter only dealswith the practice of risk assessment for environmental agents. Likewise, rather than deal with risk* The views expressed in this chapter are those of the authors and do not necessarily reflect the views or policies of theU.S. Environmental Protection Agency. Mention of trade names of commercial products does not constitute endorsementor recommendation for use.841© 2006 by Taylor & Francis Group, LLC

842 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONassessment for all types of health effects, this chapter focuses on the process as it relates toreproductive and developmental toxicity, recognizing that other types of health effects, includingmutagenicity, neurotoxicity, carcinogenicity, and immunotoxicity, may also affect reproductive anddevelopmental toxicity.This chapter updates and expands on much of the material covered in the first edition of thisbook 2 but does not go into detail on many of the aspects of hazard characterization covered there.Instead, issues of interest and importance in the rapidly changing area of environmental riskassessment are discussed here.A. DefinitionsThe following definitions are used in this chapter. Other relevant definitions can be found in theappropriate risk assessment guidelines 3–5• Developmental toxicology — The study of adverse effects on the developing organism that mayresult from exposure prior to conception (either parent), during prenatal development, or postnatallyto the time of sexual maturation. Adverse developmental effects may be detected at any point inthe lifespan of the organism. The major manifestations of developmental toxicity include: (1) deathof the developing organism, (2) structural abnormality, (3) altered growth, and (4) functionaldeficiency.• Reproductive toxicology — The study of the occurrence of biologically adverse effects on thereproductive systems of females or males that may result from exposure to environmental agents.The toxicity may be expressed as alterations to the female or male reproductive organs, to therelated endocrine system, or to pregnancy outcomes.B. Historical Perspective on Risk Assessment Approaches: 1983 to 2002Developmental and reproductive toxicity testing procedures have been available for the past 40 to50 years. 6 However, it was not until the development of a risk assessment paradigm by the NationalResearch Council (NRC) 1 that approaches for assessing the test data with a focus on potentialhuman health risks began to be formalized.Using the NRC paradigm, the U.S. Environmental Protection Agency (EPA) was instrumentalin developing risk assessment guidance in a number of health- and exposure-related areas. Fordevelopmental toxicity risk assessment, the Guidelines for the Health Assessment of SuspectDevelopmental Toxicants were published in 1986, 7 and later updated as Guidelines for DevelopmentalToxicity Risk Assessment. 3 Following this, the Guidelines for Reproductive Toxicity RiskAssessment were published in 1996. 4 In addition, Guidelines for Neurotoxicity Risk Assessment,published in 1998, 5 addressed developmental effects on the nervous system.In the process of developing guidance for risk assessment, issues important to the consistentassessment of developmental and reproductive toxicity data were discussed, and some were specificallyincorporated into the default assumptions. For example, the guidance made it clear thatall manifestations of developmental toxicity are of concern in assessing potential risk. This broadenedthe traditional focus on fetal death and malformations (teratology) to other, sometimes moresensitive, indicators of growth and functional alteration. Likewise, the reproductive endpointsincluded not only effects on primary and secondary sexual organs and their function but also cyclenormality, sexual behavior, puberty, and reproductive senescence.In 1993, the NRC published an influential report Pesticides in the Diets of Infants and Children. 8While the focus of this report was specifically on childhood pesticide exposure, the effect of thereport in the general area of childhood risk assessment was dramatic. In 1996, the Food QualityProtection Act (FQPA) was passed and included a number of the recommendations from the NRCreport. FQPA amended the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) and the© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 843Federal Food, Drug, and Cosmetic Act (FFDCA) and included a number of special provisions forinfants and children. These provisions included the requirement for determining whether pesticidetolerances are safe for children, the use of an additional safety factor of up to 10-fold (the FQPA10X factor) to account for data uncertainty with regard to toxicity and exposure to children, andthe collection of improved pesticide exposure data for infants and children. In addition, amendmentsto the Safe Drinking Water Act (SDWA) in 1996 mandated research for new water standards onthe health effects of subpopulations at greater risk and thus had an effect on children’s health riskassessment. The FQPA and the SDWA amendment requirements are the drivers of a number ofcurrent developmental and reproductive toxicology risk assessment issues and activities. Theseinclude the development of a screening and testing program for potential endocrine disruptingchemicals (EDCs) and the development of approaches for cumulative risk assessment for exposureto multiple chemicals having a similar mode of action.As a result of these and other events, there has been increasing attention paid to the potentialfor environmental exposures to have an effect on development and reproductive health. Schmidt 9recently reported on the “growth spurt” in children’s health executive and legislative initiatives.These include Executive Order (EO) 13045 (issued in 1997, amended in 2001 by EO 13229), whichmandated the consideration of children in the development of health policy and created the TaskForce on Environmental Health Risks and Safety Risks to Children. Also of importance is theChildren’s Health Act of 2000 (PL 106-310). Among other things, this act established the planningfor the National Children’s Study, a large prospective longitudinal cohort study of environmentalfactors on children’s health. These and several other efforts will be discussed in greater detail below.The efforts surrounding endocrine disrupting chemicals (EDCs) have focused a great deal of attentionon issues of developmental and reproductive effects related to such environmental agents. 10–13Finally, the field of risk assessment is moving toward harmonization of approaches for healthrisk assessment. This trend results in large part from the NRC report Science and Judgment in RiskAssessment, 14 which pointed out the importance of a risk assessment approach that is less fragmented,more consistent in application of similar concepts, and more holistic than the currentendpoint-specific guidelines. Given that different approaches are used for various types of effects(cancer versus other health endpoints, inhalation versus oral exposures, children versus adults,single chemicals versus mixtures), a consistent set of principles and guidelines is needed for drawinginferences from scientific information. Coupled with the explosion of new technology and datacollection at the cellular and molecular level, the fields of developmental and reproductive toxicologywill be challenged to evolve further. For example, a recent EPA review of the reference dose(RfD) and reference concentration (RfC) processes 15 promoted the development of a broader viewof risk assessment. It contained recommendations for derivation of acute, short-term, and longertermreference values, as well as the chronic RfD and RfC. The review also considered potentiallysusceptible populations (including life stages), mode of action, pharmacokinetics, aggregate exposureand cumulative risk issues, and harmonization of approaches for all health effects.II. FRAMEWORK FOR DEVELOPMENTAL ANDREPRODUCTIVE TOXICITY RISK ASSESSMENTHazard identification and characterization of reproductive and developmental effects must be viewedas part of the larger framework of information considered for the regulatory decision-makingprocess. Thus, a framework for risk assessment and risk management includes: (1) the planningand problem formulation phase, (2) the risk assessment itself, which comprises hazard characterization,dose-response assessment, exposure assessment, and risk characterization, and (3) riskmanagement and risk communication (Figure 21.1). The first two components of this frameworkare discussed in the following sections.© 2006 by Taylor & Francis Group, LLC

844 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONPLANNING AND PROBLEMFORMULATIONScope of the AssessmentRegulatory InterestsPublic Health InterestsLegislative and Judicial MandatesRISK ASSESSMENTHazard CharacterizationDose-Response AssessmentExposure AssessmentRisk CharacterizationRISK MANAGEMENTDevelopment of RegulatoryOptionsPublic Health, Economic,Social, & Political ImplicationsRegulatory DecisionFigure 21.1Framework for risk assessment and risk management.A. Planning and Problem FormulationAt the same time that risk assessment is moving toward the goal of harmonization and consistency,it is also true that the application of the risk assessment paradigm cannot be “one size fits all.”Assessments vary in scope, depending on the regulatory and public health interests, and have rangedfrom covering all possible exposures and effects to focusing on a specific exposure and healthendpoint. Moreover, legislative and judicial mandates often direct the type of assessment that maybe appropriate. The planning and problem formulation process allows the risk assessor to understandthe needs of the risk manager, the interests of the stakeholders, and the extent and limitations ofthe available database and analytical tools. Most important, the process should result in clearlydefined questions and goals for the assessment.Specific to developmental and reproductive toxicity risk assessments, the planning and problemformulation process identifies which exposure scenarios are being considered. As defined, developmentaland reproductive toxicity may include exposure of almost any segment of the population(except perhaps nonreproducing adults). But identifying the exposure scenario(s) to be considered© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 845can help identify specific populations of interest (e.g., school children, pregnant women, migrantworkers), as well as exposure characteristics with respect to medium (e.g., air, water, soil), route,duration, frequency, and life stage.The planning and problem formulation process should also help define what information isavailable and applicable to the risk assessment process, as well as appropriate analytical methods.Over the years, the type of toxicity test data available, especially for pesticides and some highprofile agents, has not only increased in size but also in depth. For example, developmental andreproductive toxicity testing now includes such measurements as sperm motility, estrous cycling,and endocrine function, and may sometimes include measurements of behavior and specific organand system development. In addition, technological advances permit measurements at the cellularand molecular levels, and this has prompted discussion on how to incorporate these data into riskassessment. Coupled with improved methods of statistical analysis and mathematical model development,it is important to consider what will be included in the risk assessment and to discuss howit will be analyzed in the initial planning and problem formulation process.B. The Risk Assessment ProcessThe process of risk assessment for reproductive and developmental toxicity is based on generaland specific scientific principles but also includes a great deal of scientific judgment. As previouslymentioned, risk assessment consists of four major components, hazard characterization, doseresponseassessment, exposure assessment, and risk characterization, each of which is discussedbriefly in the following sections.1. Hazard CharacterizationHazard characterization involves determining whether and under what conditions exposure to anenvironmental agent causes adverse effects. This information often comes from animal data,although infrequently human data are available. Guidelines for Developmental Toxicity Risk Assessment,3 Guidelines for Reproductive Toxicity Risk Assessment, 4 Guidelines for Neurotoxicity RiskAssessment, 5 and Guidelines for Carcinogen Risk Assessment 16 all address some aspect of reproductiveand developmental toxicity risk assessment. Endocrine disrupting chemicals (EDCs) comprisea subclass of agents that alter the normal functioning of the endocrine system. The mode ofaction for endocrine disruption may be receptor dependent, e.g., through interaction with a particularreceptor, or receptor independent, e.g., through alteration in the synthesis or metabolism of ahormone. Reproductive and/or developmental endpoints are almost always affected. In particular,the development of the reproductive system is often affected by early developmental EDC exposure.Guidelines for Reproductive Toxicity Risk Assessment 4 addresses many of the types of endpointsthat may be affected by EDCs. The general evaluation and interpretation of data are discussed here.a. Default AssumptionsThe process of interpreting toxicity data requires a great deal of scientific judgment because acomplete database to conduct a risk assessment is rarely if ever available. Assumptions must bemade because of uncertainties due to limited or nonexistent information on toxicokinetics, mechanismof action, low-dose response relationships, and human exposure patterns. Several of theseassumptions are basic to the extrapolation of toxicity data from animals to humans, while othersare specific to reproductive and developmental toxicity. The general assumptions (in the followinglist), which are used in the absence of data to the contrary, are based on the assumptions discussedin the developmental toxicity risk assessment guidelines 3 as expanded upon in the reproductivetoxicity guidelines. 4© 2006 by Taylor & Francis Group, LLC

846 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION• An agent that produces an adverse reproductive and/or developmental effect in experimental animalstudies is assumed to pose a potential reproductive and/or developmental threat to humans followingsufficient exposure at the right stages of reproduction and development.– This assumption is based on the evaluation of data for agents known to cause human reproductiveand/or developmental toxicity; these data indicate that, in general, adverse reproductiveand/or developmental effects seen in experimental animals are also seen in humans.• Effects of xenobiotics on male and female reproductive processes and/or developmental outcomesare assumed generally to be predictive of similar effects across species, although outcomes seenin experimental animal species may not necessarily be the same as those seen in humans.– For example, the differences between species in developmental and reproductive outcomes mayoccur because of species-specific differences in timing of exposure relative to critical periodsof development or reproductive system function, pharmacokinetics (including metabolism),developmental patterns, placentation, or modes of action. Because of this, all of the four majormanifestations of developmental toxicity (death, structural abnormalities, growth alterations,and functional deficits) are considered to be of concern for hazard characterization. This is incontrast to the approach in the 1960s and 1970s, when there was a tendency to consider onlymalformations and death as endpoints of concern. More recently, a biologically significantincrease in any of the four manifestations of altered development is considered indicative ofan agent’s potential for disrupting development and producing a developmental hazard.• When sufficient data (e.g., pharmacokinetic, mode of action) are available to allow a decision, themost appropriate species for estimating risk to humans should be used for hazard characterization.In the absence of such data, it is assumed that the most sensitive species should be used.– This is based on observations that humans are as at least as sensitive or more so than the mostsensitive animal species tested for the majority of agents known to cause human reproductiveand/or developmental toxicity.• In the absence of specific information to the contrary, it is assumed that a chemical that affectsreproductive function or development of offspring as a result of exposure of one sex may alsoadversely affect reproductive function or offspring development as a result of exposure of the othersex.– This assumption is based on three considerations: (1) for most agents, the nature of the testingand the data available are limited, reducing confidence that the potential for toxicity to bothsexes and their offspring has been adequately examined, (2) exposure of either males or femaleshas been shown to result in developmental toxicity, and (3) many of the mechanisms controllingimportant aspects of reproductive system function are similar in males and females, andtherefore they could be susceptible to the same agents.• When detailed information on mode of action and pharmacokinetics is available, it should be usedto predict the shape of the dose-response curve at low doses. In the absence of such information,a threshold is generally assumed for the dose-response curve for agents that produce reproductiveand/or developmental toxicity.– This assumption is based on known homeostatic, compensatory, or adaptive mechanisms thatmust be overcome before a toxic endpoint is manifested and on the rationale that cells andorgans of the reproductive system and the developing organism are known to have some capacityto repair damage. In addition, because of the multipotency of cells at certain stages of development,multiple insults at the molecular or cellular level may be required to produce an effecton the whole organism. It should be recognized that background levels of toxic agents, thelifestage at exposure, preexisting disease conditions, and other factors may increase the sensitivityof some individuals in the population. Thus, exposure to a toxic agent may result in anincreased risk of adverse effects for some, but not necessarily all, individuals within thepopulation. Although a threshold may exist, it is usually not feasible to distinguish empiricallybetween a true threshold and a nonlinear low-dose response relationship.b. Types of Data Used for Hazard CharacterizationData used for hazard characterization include both laboratory animal and human data. Adversereproductive endpoints include altered reproductive organs, endocrine system function, or pregnancy© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 847Table 21.1Common types of human studiesEpidemiologicCohortCase controlEcologicCross-sectionalGroups defined by exposureFollow groups over timeGroups defined by selected health effectsExposure history determinedStudy groups defined by general characteristicExposure history assumedExposure and outcome examined at same timeOther Human DataClustersCase reports/seriesConcern generated because of perceived excess of health effectTypically based on clinical observationMay be endpoint or exposure basedoutcomes, such as effects on ovulation and menstruation, sexual behavior, fertility, onset of puberty,gestation, heritable reproductive effects, and premature reproductive senescence. 4 Adverse developmentalendpoints include structural malformations, altered growth, premature mortality, andfunctional deficits, such as mental retardation, sensory loss, or immunological effects. 31) Human Studies — Adequate human data are the most appropriate sources of informationfor determining an agent’s potential for causing human reproductive or developmental toxicity.However, ethical and practical reasons often limit collection of human exposure and effects data,especially on reproductive and developmental toxicity because of potential harm to the developingfetus or child or because of ethical issues or social taboos on collecting such types of data.Nevertheless, guidance for data evaluation has been developed for cases where human data areavailable. The human studies category includes a variety of data sources, ranging from epidemiologicstudies to individual case reports (Table 21.1). The general design characteristics of humanstudies are covered in detail in Chapter 20 and were also reviewed by Adams and Selevan. 17Well-designed epidemiologic studies provide the most relevant information for assessing humanrisk. There are several epidemiologic study designs, each having certain advantages and limitations.Cohort studies define study populations by exposure and then examine outcomes associated withdiffering exposures. This study design can examine multiple health effects and is most powerfulwhen more common developmental effects are the endpoints. Prospective cohort studies, whichexamine exposures before the outcomes are known, avoid many of the problems with recall bias.In the case referent design, populations are defined by a selected health effect and exposure historyis determined retrospectively. This design can be used to examine multiple exposures and is usefulwhen rare outcomes are being considered. Both the cohort and the case referent study design can havesignificant power or ability to detect a true effect, depending on population size, outcome frequency,and the level of excess risk to be identified. But both can be fairly expensive and resource intensive.The ecologic study is an epidemiologic study that defines the study population by some generalcharacteristic (e.g., zip code, county of residence), and the exposure history is assumed, based onthat characteristic. Detailed information on individuals is most often not collected. This makes theecologic study a less expensive method of data collection, and it can be very useful in developinghypotheses for further testing. However, the lack of individual exposure and outcome data limitthe usefulness of this design in establishing clear associations.The cross-sectional study combines some of the features of other study designs and examinesboth exposure and outcome at the same time point. Exposure data are generally more reliable thanthose obtained from an ecologic study, since individual exposures are assessed through interviewsor monitoring. Exposures are assumed to be similar to historical exposure levels that may haveoccurred at the critical period of development when the effect was induced. Since outcome is© 2006 by Taylor & Francis Group, LLC

848 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmeasured at the same time, the choice of an appropriate outcome can be influenced by such thingsas time-to-appearance and its transience or persistence. The risk assessment must consider thesefactors in drawing associations between exposure and effect from this type of study.Other data that can provide indications of a possible effect are reports of clusters and casereports or series. Reports of clusters indicate a concentration of an outcome in a particular geographicalarea or population. Case reports or series generally come from reports of individuals ortheir physicians and are recognized as a potential pattern only after the data are brought together.Rare events are much more likely to be reported as an outcome than are more common events. Forexample, the thalidomide tragedy was recognized by physicians in large part because of the raremalformation (phocomelia) that the drug produced. Had exposure been associated with a morecommon effect, recognition of a possible association between the effect and the exposure wouldhave been more difficult, if not impossible, from case reports. Tabacova et al. 18 have developed anapproach for combining multiple case reports as a case series for a particular exposure and patternof associated outcomes for use in generating hypotheses for further study. Case reports were definedby a particular exposure (e.g., enalapril, a hypotensive agent) during pregnancy. They were thenevaluated using several criteria to determine the strength of the association between an exposure(including the timing of exposure) and various pregnancy outcomes. Using this approach, theauthors were able to confirm previous reports of enalapril-related effects following second and thirdtrimester pregnancy exposures.The differences between human and laboratory animal developmental toxicity studies areconsiderable and must be recognized by the risk assessor. Human studies rarely, if ever, providethe controlled study population and exposures characteristic of the laboratory study. Often, the sizeof the human study population is small, limiting the statistical power of the study. The potentialfor bias in data collection is more likely a concern in human studies because the study populationis almost always heterogeneous and exposures are seldom uniform and often are not specific enoughto provide actual dose information. Initial and continued participation in a study can be influencedby an individual’s perceived exposure, concern over a health effect, verbal skills, education, etc.Health records and personal recall can differ considerably and can be influenced by a physician’sawareness or a patient’s memory.In spite of these and other issues, it is important to reiterate that sufficient human data are themost appropriate source of information for determining an agent’s potential for causing humandevelopmental toxicity. Moreover, there have been advances in the collection of both exposure andhealth data and in data analysis that have strengthened the utility of human data in the overalltoxicological profile. In addition, integration of human and animal data often lends strength to thedatabase that neither alone can provide. Studies to evaluate the comparability of animal and humandata have provided a good deal of information on the bases for similarities and differences inspecies-specific responses and have pointed the way toward additional animal or human studiesneeded to further define or clarify exposure-related effects. 19–212) Laboratory Animal Studies—Standard Testing Protocols — Three testing protocols arecommonly used in the regulatory setting for assessing the reproductive and developmental toxicityof environmental agents in laboratory animals: (1) the prenatal developmental toxicity study, (2)the reproduction and fertility effects study, and (3) the developmental neurotoxicity (DNT) study 22(Table 21.2). The first two studies are routinely required for pesticides and may be required forindustrial chemicals if there is cause for concern. The DNT study has been required on a case-bycasebasis and was included as part of a data call-in for organophosphate insecticides. 23 The prenataldevelopmental toxicity study involves exposure of pregnant females during the period of organogenesisthrough the end of pregnancy (typically gestation days [GD] 6 to 20 in the rat) and evaluationof fetuses at term. This study focuses on general reproductive indices of the maternal animals,survival of the fetuses to term, fetal weight at term, and a detailed structural evaluation of thefetuses. The reproduction and fertility study involves dosing the parental animals from 10 weeks© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 849Table 21.2OPPTS harmonized guideline protocols for reproductive and developmental toxicity testingProtocol Guideline Brief DescriptionPrenatal developmentaltoxicity studyReproduction and fertilityeffects studyDevelopmentalneurotoxicity study870.3700 a Exposure from implantation to termination just prior to parturition —gestation day (GD) 18 in mice, 20 in rats, and 29–30 in rabbits.Detailed evaluation of offspring morphology.870.3800 a Exposure of males and females for 10 weeks prior to mating andduring mating, exposure of females throughout gestation andlactation, and exposure of offspring after weaning — typicallyconducted in rats. Estrous cyclicity in parental (P) and F 1 females;semen quality in P and F 1 males; survival, growth, and reproductivedevelopment of offspring (age at vaginal opening and preputialseparation, anogenital distance in F 2 pups if an effect is seen onF 1 sex ratio or sexual maturation); organ weights (reproductive,brain, spleen, and thymus, other target organs); histology, includingprimordial follicle counts in F 1 female ovaries.870.6300 Exposure from implantation to postnatal day (PND) 10 or 21.Offspring evaluated pre- and postweaning for general growth anddevelopment, timing of puberty, clinical signs (modified functionalobservation battery), motor activity, auditory startle habituation,cognitive function, and neurohistology.aThese complete guidelines are presented in Chapter 19, Appendix II.before mating and during pregnancy and lactation, with continued exposure of selected F 1 offspringthat are mated and allowed to deliver and maintain their (F 2 ) pups to weaning. This study focuseson the evaluation of reproductive development (both structure and function), as well as on adultreproductive structure and function and the developmental effects of continuous exposure. Thedevelopmental neurotoxicity study protocol involves exposure during organogenesis, through theend of gestation, and into the postnatal period, either to midlactation or to weaning. The studyfocuses on the evaluation of the developing nervous system and includes measures of physical andreflex development, motor activity, auditory startle habituation, learning and memory, and neuropathologyin young animals and adults. This is the only study that evaluates aspects of latency toresponse and reversibility of effects following termination of exposure.These guideline animal testing protocols and their use in regulatory testing are described indetail by Kimmel and Makris 24 and in Chapters 7 to 9 and 19 of the current volume. From suchstudies, a number of endpoints of general adult toxicity and reproductive and developmental toxicityare used to determine the types of effects and the effective dose levels for a given agent. Thesedata are used in hazard characterization and dose-response assessment.3) Other Types of Dataa) Data on Functional Endpoints — Functional developmental and reproductive toxicityrefers to the effects of agents on the functioning of essentially any organ system. The two-generationreproduction study discussed above provides information on reproductive function in the form ofdata on fertility, semen and sperm quality, estrous cycling, and ovarian function (follicle numbers,number of ova produced and fertilized). However, it should be recognized that there are limitationsin assessing reproductive function by means of the two-generation protocol. For example, exposureis continuous, and effects induced during development may not be separable from those producedat later ages. Since both males and females are exposed, it is not clear which sex is primarilyaffected, although this can be ascertained by evaluating endpoints in individual sexes or by crossmatingcontrol and treated animals. Other aspects of the interpretation of reproductive functioninduced in males and females during development are discussed in Chapter 9 and in the reproductivetoxicity risk assessment guidelines. 4 More detailed evaluations of male and female reproductivefunction 25 were added to the guideline protocol in 1994 26 and were republished as harmonizedguidelines for pesticides and toxic substances in 1998. 22© 2006 by Taylor & Francis Group, LLC

850 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONStandard testing approaches to evaluate a wide range of functional effects have not beendeveloped, with the exception of the developmental neurotoxicity study discussed in the precedingsections. Moreover, this study has not been required on a routine basis; yet developmental neurotoxicityis one of the hallmarks of several environmental agents in humans, e.g., lead, methylmercury,and polychlorinated biphenyls (PCBs). Recent data on lead and neurological effects suggestthat the developing nervous system may be sensitive at blood lead levels as low as 5 µg/dl. 27Strong support for using animal models to evaluate the potential for developmental neurotoxicityin humans came from a workshop on the qualitative and quantitative comparability of human andanimal developmental neurotoxicity. 28 Newer more sensitive measures that can be used in bothanimals and humans are becoming available and should be considered for incorporation into testingprotocols (see Sharbaugh et al. 29 ).Effects of chemicals on the developing immune system have been reviewed by severalauthors. 30–34 Although there are no testing guidelines for the evaluation of such effects, a fewendpoints (spleen and thymus weights) were added to the fertility and reproduction study guideline 22to screen for possible effects on the immune system and to trigger further study. Development ofa testing guideline is needed and has been discussed in recent EPA reports. 15,35Lau and Kavlock 36 summarized the literature on functional developmental toxicity of the heart,lungs, and kidneys. Their review revealed a plethora of information on a wide variety of chemicalsand endogenous agents (hormones and neurotransmitters), but they found a lack of uniform treatmentparadigms or observations among studies, although protocols have been developed for evaluatingrenal function. Interpretation of data on these organ systems is hindered by the limitedunderstanding of mechanisms controlling normal development in these organs and by the lack oftoxicological evaluation of these organs in an integrated manner, despite their interdependenceduring development. The authors were able to summarize some common features emerging fromstudies on functional developmental toxicity of these three organs that are probably applicable tothe functional development of other organ systems. These include1. Assessment of multiple endpoints is needed to adequately characterize the spectrum of toxicity.2. Compensatory mechanisms may mask functional deficiencies, necessitating the use of tests thatevaluate maximal responses of organ systems.3 Because of physiological interactions among functional systems, examples of organ-specific toxicantsare rare.4. Functional systems display broad periods of developmental susceptibility to toxic insult, withaccelerations or delays having potential long term consequences.5. Within the limits of the database on functional developmental toxicity in humans, there is generallygood concordance of effects across species.6. Although toxicologic mechanisms are unknown, disruption of endogenous trophic factors andaltered cellular and molecular interactions are likely common pathways for the induction offunctional developmental toxicity (adapted from Lau and Kavlock 36 ).A number of publications have linked prenatal undernutrition and reduced birth weight inhumans and animals with later development of heart disease, 37–40 diabetes and obesity, 41–47 andosteoporosis. 48 In addition, several studies have shown a positive association between birth weightand later development of breast cancer (for example, see Michels et al., 49 Innes et al., 50 andMellemkjaer et al. 51 ). Together, these studies strongly suggest that prenatal factors can play animportant role in the function of adult systems and in the aging process and that environmentalchemicals that affect prenatal nutrition and growth may also affect later organ function and occurrenceof disease. Thus, developmental factors should be considered in the etiology of disease statesthat are not recognized until adulthood or old age.b) Data on Endocrine Disrupting Chemicals — EDCs can be defined as chemicals that positivelyor negatively alter hormone activity and thereby may affect reproduction or development.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 851For example, in rodent studies, prenatal exposure to the antiandrogen pesticide vinclozolin can leadto effects on the development of the male reproductive system, including femalelike anogenitaldistance, nipple retention, cleft phallus with hypospadias, suprainguinal ectopic scrota and/or testes,a vaginal pouch, epididymal granulomas, and small to absent sex accessory glands. 52Given the potential for developmental and reproductive effects in humans and wildlife, EDCshave emerged as a major international environmental issue. Some human population studies ofaccidental exposures have found associations between high EDC exposure and developmentaland/or reproductive outcomes. For example, in the Michigan exposures to polybrominated biphenyls(PBBs), girls who experienced a high (greater than 7 ppb) in utero exposure and were breastfedhad a significantly earlier age of menarche and more advanced pubic hair development than eithergirls that were not breastfed or had a lower in utero exposure (less than 7 ppb). 53 Among womenwho were exposed before menarche to high dioxin levels from the Seveso explosion, a tenfoldincrease in serum dioxin concentration was associated with a lengthened menstrual cycle. 54Findings from laboratory animal as well as human studies suggest that the timing of pubertyis a sensitive biomarker of effect. For example, in rodent studies, females showed delayed vaginalopening in the absence of body weight decrements after developmental lead exposures. 55–57 Tworecent epidemiologic studies found an association between higher concurrent blood lead levels anda later timing of puberty in girls, as indicated by one or more delayed puberty measures. 58,59Additionally, racial and ethnic background modified the association, e.g., the association was foundto be statistically significant for Tanner stages in Mexican-American girls and for Tanner stages aswell as onset of menarche in African-American girls; no measures were significant in Caucasiangirls. 58 The associations observed between blood lead level and puberty timing were present aftercontrolling for body size, 58 suggesting that lead affects puberty via other pathways, such as thehypothalamic-pituitary-gonadal axis.How to define EDCs, what to call them (e.g., “hormonally active agent,” “endocrine modulator”),whether environmental exposure levels are high enough to affect the health of humans in thegeneral population, and whether there are “low dose” effects are all hotly debated questions.Multiple EDCs with the same mode of action are of concern because individual exposures to eachEDC may be “low” but the total exposure to a mode-of-action class may be “high.” A recent studyfound that multiple “no effect” level exposures to EDCs with an estrogenic mode of action weredose additive, leading to observable effects. 60 Furthermore, the fetus and child are likely to be moresusceptible to EDC exposure than the adult because of critical windows of exposure in the endocrine,neural, and reproductive systems. 61The FQPA and the SDWA amendments passed by Congress required the EPA to assess theestrogenic potential of pesticides and water contaminants that may affect human health. As a resultof this legislation, the EPA formed the Endocrine Disruptor Screening and Testing Committee(EDSTAC), a multistakeholder federal advisory committee, to develop a screening program.EDSTAC expanded the scope of their charge to include testing as a second tier, and androgen andthyroid activity effects in addition to estrogenic effects. The EDSTAC broadened the chemicals tobe assessed to include industrial chemicals, pesticide “inert” ingredients, pharmaceutical chemicals,and environmental contaminants. A final consensus report of their recommendations was publishedin August 1998. 62A number of endpoints in the current EPA testing guidelines for reproduction and fertilityeffects 22 are sensitive to altered estrogen, androgen, or thyroid hormone activity. For example, thetiming of preputial separation (PPS), a marker of male puberty, can be affected by altered androgenactivity. A number of additional test endpoints sensitive to hormone activity were recommendedby EDSTAC, and these include brain histology and weight, some motor activity and memoryendpoints, TSH and T4 levels, and testis descent. 62 See Chapter 11 for detailed coverage of theEDSTAC test protocols.One of the challenges of developing such a screening and testing program was to include assaysthat would cover the numerous modes of action for EDCs. Similar to carcinogens, the EDC© 2006 by Taylor & Francis Group, LLC

852 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONclassification focuses on mode of action data in determining whether a chemical is an EDC. Theassays recommended by EDSTAC were designed to detect estrogenic, antiestrogenic, androgenic,antiandrogenic, thyroid, and antithyroid activity. Two tiers of testing batteries, one screening (T1S)and one testing (T2T), were recommended. The intention of T1S is to determine whether a chemicalis capable of affecting the activity of estrogens, androgens, or thyroid hormones. T1S in vitroscreening assays include an estrogen receptor binding or reporter gene assay, an androgen receptorbinding or reporter gene assay, and a steroidogenesis assay that uses minced testis. T1S in vivoscreening assays include a rodent 3-d uterotrophic assay, a rodent 20-d pubertal female assay withenhanced thyroid endpoints, a rodent 5- to 7-d Hershberger assay, a frog metamorphosis assay, anda fish reproductive screening assay. The analysis of the data from T1S will either conclude that thechemical be moved into T2T or that no further screening or testing is currently needed (placed “onhold”).The T2T battery was designed to determine whether a chemical may be an EDC and to quantifythe effects on estrogen, androgen, and thyroid hormone activity. The tests include dosing duringdevelopment and a larger dose range; most include testing in two generations. T2T in vivo testingassays include a two-generation mammalian reproductive toxicity study or a less comprehensivealternative mammalian reproductive toxicity test, an avian reproduction toxicity test, a fish lifecycle toxicity test, an opossum shrimp (Mysidacea) or other invertebrate life cycle toxicity test,and an amphibian development and reproduction test.* Information from the T2T battery is of thetype necessary for hazard assessments for human health, and since the outcome of tests is moredefinitive than that of screens, a negative outcome in T2T will supplant a positive finding from T1S.Interpreting the data from the EDC screening and testing program and then developing riskassessment methods are the next challenges. While the screening and testing tiers were not designedexpressly for risk assessment purposes, the data from the screens and tests will be available foruse in risk assessment. Specific guidance for the risk assessment of EDCs has not been developed.The attention on EDCs has raised a number of important issues for developmental and reproductivetoxicology risk assessment that will likely apply to other types of effects and chemical classes.One of the most contentious issues has been the potential for effects at “low doses” or “environmentallevels” (i.e., likely levels of human exposure). A National Toxicology Program (NTP)Workshop (Endocrine Disruptors Low Dose Peer Review) evaluated the existing evidence for lowdose reproductive and developmental effects for a number of EDCs. 63 The peer review defined “lowdose” as equal to human exposure levels, below those used in EPA’s standard reproductive anddevelopmental toxicity testing, or those below the current no effect level (NOEL). Based on thesedefinitions, the panel concluded that there is credible evidence for low dose effects on certainendpoints for estradiol, diethylstilbestrol, nonylphenol, methoxychlor, and genistein. 63 Other EDCrelatedissues of interest are: (1) the effective addition and subtraction (depending on whether theagent acts as an agonist or antagonist) from endogenous steroid hormone levels, 64 (2) critical andsensitive windows of EDC exposure that correspond to developmental events regulated by steroidhormones, 65 and (3) cumulative effects of exposure to chemicals with the same or opposing modesof action. 60 There is also interest in understanding the significance of nonmonotonic dose responsecurves, hormesis, and U-, J-, or inverted U-shaped dose response curves, 66–71 which may reflectmultiple modes of action for one chemical at different doses. 68–704) Toxicokinetics and Toxicodynamics — Toxicokinetic data provide information on theabsorption, metabolism, distribution (including placental transfer), and/or excretion (including viabreast milk) of an agent. When available, toxicokinetic data can be extremely useful in determiningtarget dosimetry and critical levels or duration of exposure at the target site (e.g., peak concentration,area under the curve) and can be related to expression of the toxic effect (toxicodynamics). Inaddition, toxicokinetic information provides support for defining similarities and differences among* Further information is available at http://www.epa.gov/scipoly/oscpendo/index.htm.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 853routes of exposure and among species, or in accurately extrapolating information from one speciesto another.At a minimum, absorption data may assist in the interpretation of developmental and reproductivetoxicity studies in animal models and will aid in the extrapolation of data from animal modelsto humans if human absorption data are available. Absorption data are essential for studies in whichdermal exposure is used, whether or not any developmental or reproductive effects are produced. 71A study showing no adverse effects without evidence of dermal absorption is considered insufficientfor risk assessment and is potentially misleading, especially if interpreted as a “negative” study.Toxicokinetic studies related to developmental toxicity are most useful if conducted in animalsat the stage when developmental insults occur, rather than at the end of gestation. In this way,specific toxicokinetic patterns can be correlated with developmental outcomes to define the parametersthat contribute most to the outcomes measured. 72–74 Certain drugs and chemicals have beenclassified according to the correlation between the malformations produced and either the integratedexposure area under the curve (AUC) or the peak concentration (C max ). 75 A study by Terry et al., 76however, showed that for the same agent (2-methoxyacetic acid, a metabolite of 2-methoxyethanol[2-ME]), the toxicokinetic pattern that correlated best with outcome differed depending on the stageof development at which treatment occurred. In this case, C max correlated better with exencephalyproduced by a dose of the parent compound, 2-ME, on gestation day (GD) 8, while AUC correlatedbetter with limb defects produced on GD 11. 73 Thus, developmental stage–specificity may be asimportant as or more important than the toxicokinetic characteristics of a particular agent. If true,this becomes especially important in extrapolation of data to humans, where the rate of developmentis much slower than that of most laboratory animal species used.The ontogeny of metabolic capabilities in the developing fetus and child is important forunderstanding potential susceptibility of children to chemical exposures. 77 Several recent studieshave contributed significantly to our understanding of these processes. For example, Cresteil 78published data on the activity of several P450 enzymes in human liver from birth to 10 years ofage. Koukouritaki et al. 79 showed the developmental ontogeny of the flavin-containing monooxygenase(FMO) 1 and 3 enzymes in human liver samples ranging from 8 weeks gestation to 18years of age.Several physiologically based pharmacokinetic (PBPK) models have been developed that takeinto account the changes occurring during pregnancy in rodents and nonhuman primates 80–82 andin humans. 83–85 Other efforts have been made to model lactational transfer 86,87 and pharmacokineticsin young children. 88,89 Ginsberg et al. 90 evaluated the published literature for data on child and adultpharmacokinetic differences for pharmaceutical agents, and Hattis et al. 91 used the database toanalyze the variability in children’s drug elimination half-lives for the purpose of building a PBPKmodel. These approaches are likely to improve the ability of researchers to extrapolate informationacross species because the approaches can accommodate the anatomic and physiological changesthat occur as pregnancy advances in different species. 92 An in-depth review of information availableon perinatal pharmacokinetics was developed recently. 93Physiologically based models for reproductive systems have also been developed. For example,Keys et al. 94 developed a PBPK model for di(2-ethylhexyl) phthalate (DEHP) and mono(2-ethylhexyl)phthalate (MEHP) to quantify the testis dose of MEHP. Plowchalk and Teeguarden 95 developeda PBPK model for estradiol in rats and humans to explore the importance of various physiologicalparameters on circulating levels of estrogen in reproductive tissues and responses toendocrine-active chemicals. Several researchers have studied 4-vinylcyclohexene to determine thebasis for species differences in ovarian toxicity in rats and mice. 96,97 For a more extensive discussionof pharmacokinetics, see Chapter 13.Understanding toxicodynamics (mechanism or mode of action) is an important aspect of hazardcharacterization and may influence significantly many aspects of the approach taken to risk assessment.Such data may come from a variety of short-term in vitro or in vivo studies or comparative© 2006 by Taylor & Francis Group, LLC

854 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONstudies in different species. Not a great deal is known as yet about the mechanisms of action ofreproductive and developmental toxicants, although there has been a strong research effort investigatingEDCs, especially those that act by binding and activating the estrogen or androgen receptor.Other endocrine modes of action, such as effects on steroidogenesis and thyroid hormone antagonism,have been identified for some chemicals, and more mode of action data are expected fromthe Endocrine Disruptor Screening and Testing Program. 98 The National Research Council Report, 99Scientific Frontiers in Developmental Toxicology and Risk Assessment, suggested areas of researchfor gaining a better mechanistic understanding of developmental toxicity.As toxicokinetic and mechanistic information becomes available for reproductive and developmentaltoxicity risk assessment, these types of information can be incorporated into biologicallybased dose-response models that can provide a more accurate assessment of risk. For example,information from studies on enzyme function, receptor-ligand interaction, metabolic and syntheticpathways, and cell growth, function, and control were incorporated into mathematical models todefine the overall dose-response relationship for 5-fluorouracil. 100–102 Such information can providea conceptual framework for defining the underlying events associated with cellular control and therelationship between exposure parameters (e.g., dose, frequency, duration, and timing) and theultimate health outcome. 103–106c. Characterization of the DatabaseOnce the toxicity data in both humans and animals have been thoroughly reviewed, the databaseis characterized as being sufficient or insufficient to proceed further in the risk assessment process.This characterization of hazard considers the context of exposure, e.g., dose, route, duration, andtiming, relative to the life stage(s) during which exposure occurred. Characterizing the hazardpotential in this manner allows for better determination of the strengths and weaknesses, as wellas the uncertainties, of the data associated with a particular exposure. This does not, however,address the nature and magnitude of the human health risks, which are determined after the exposureassessment is completed and the final characterization of risk is conducted. A great deal of scientificjudgment and expertise in the area of developmental and reproductive toxicity, including laboratorystudies, epidemiology, and statistics, may be necessary to characterize the hazard for a givendatabase.The EPA’s Guidelines for Developmental Toxicity Risk Assessment 3 and Guidelines for ReproductiveToxicity Risk Assessment 4 present similar categorization schemes for characterizing thedatabase for developmental and reproductive toxicity. The scheme from the latter document ispresented in Table 21.3. Two broad categories are included: sufficient evidence and insufficientevidence. Within the sufficient evidence category, there is a further breakdown, depending onwhether human data are available and the strength of those data. Human data are considered primaryin determining a potential hazard but are often limited or lacking, so that few agents can be classifiedas having sufficient human evidence. Often, however, limited human data and sufficient animaldata can be used together in characterizing the hazard of a particular agent.The text in the categories in Table 21.3 defines the minimum evidence necessary to determinesufficiency, and the strength of evidence for a database increases with replication of the findingsand with additional human studies or additional animal species tested. All data, whether indicativeof a hazard potential or not, are to be considered in a weight-of-evidence approach for makingjudgments about the strength of evidence available to support a complete risk assessment forreproductive and developmental toxicity. Thus, within each category, minimum criteria are detailedfor determining whether data are sufficient to judge whether an agent does or does not have thepotential to cause reproductive or developmental toxicity.Information on pharmacokinetics and/or pharmacodynamics (mechanisms of action), especiallyat the developmental stage or in the reproductive tissue of concern, may reduce uncertainties indetermining relevance to humans. More evidence (e.g., more studies, species, or endpoints) is© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 855Table 21.3Categorization of the health-related databaseSufficient EvidenceThe Sufficient Evidence category includes data that collectively provide enough information to judge whether areproductive hazard exists within the context of effect as well as dose, duration, timing, and route of exposure.This category may include both human and experimental animal evidence.Sufficient Human EvidenceThis category includes agents for which there is convincing evidence from epidemiologic studies (e.g., casecontrol and cohort) to judge whether exposure is causally related to reproductive toxicity. A case series inconjunction with other supporting evidence also may be judged as Sufficient Evidence. An evaluation ofepidemiologic and clinical case studies should discuss whether the observed effects can be consideredbiologically plausible in relation to chemical exposure.Sufficient Experimental Animal Evidence — Limited Human DataThis category includes agents for which there is sufficient evidence from experimental animal studies and/orlimited human data to judge if a potential reproductive hazard exists. Generally, agents that have been testedaccording to EPA’s two-generation reproductive effects test guidelines (but not limited to such designs) wouldbe included in this category. The minimum evidence necessary to determine if a potential hazard exists wouldbe data demonstrating an adverse reproductive effect in a single appropriate, well-executed study in a singletest species. The minimum evidence needed to determine that a potential hazard does not exist would includedata on an adequate array of endpoints from more than one study with two species that showed no adversereproductive effects at doses that were minimally toxic in terms of inducing an adverse effect. Information onpharmacokinetics, mechanisms, or known properties of the chemical class may also strengthen the evidence.Insufficient EvidenceThis category includes agents for which there is less than the minimum sufficient evidence necessary forassessing the potential for reproductive toxicity. Included are situations such as when no data are available onreproductive toxicity as well as for databases from studies on test animals or humans that have a limited studydesign or conduct (e.g., small numbers of test animals or human subjects, inappropriate dose selection orexposure information, other uncontrolled factors); data from studies that examined only a limited number ofendpoints and reported no adverse reproductive effects; or databases that were limited to information onstructure-activity relationships, short-term or in vitro tests, pharmacokinetic data, or metabolic precursors.Source: From EPA, 1996.necessary to judge that an agent is unlikely to pose a hazard than is required to support the findingof a potential hazard. This is because it is more difficult, both biologically and statistically, tosupport a finding of no apparent adverse effect than to support a positive effect.The insufficient evidence category may include situations in which no data are available orsituations in which data are available from studies with a limited study design or of a type (e.g.,some short-term studies) that are not yet considered sufficient to use for human risk assessment.Such data are often useful in recommending the need for additional studies. In some cases, asubstantial database may exist in which no one study is considered sufficient but the body of datamay be judged as meeting the sufficient evidence category.When the database is considered sufficient for risk assessment, the hazard potential along withinformation on other relevant health effects and dose-response, pharmacokinetic, pharmacodynamic,or other supporting information can be used to calculate reference values (e.g., the chronicRfD or RfC).A recent review of the RfD and RfC processes urges expansion of the data considered in hazardcharacterization and discusses criteria to be considered in a weight-of-evidence approach. 15 Forexample, a robust database would include evaluation of a variety of health endpoints (both structuraland functional), durations of exposure, timing of exposure, life stages, and susceptible subpopulations.In the absence of complete human data, mechanistic and other data that show the relevanceof the animal data for predicting human response would provide a strong basis for extrapolationof animal data to humans. The database for a reference value of less-than-chronic duration shouldalso address the issue of reversibility of effects and latency to response, taking into consideration© 2006 by Taylor & Francis Group, LLC

856 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONthe possibility that less-than-chronic exposure may lead to effects some time after exposure.Biological and chemical characteristics of the exposure and outcomes, as well as known limits onreserve capacities and repair of damage, may provide helpful information in determining whetherthe length of follow-up is sufficient to detect latent effects or the degree of reversibility.2. Dose-Response AssessmentThe evaluation of the dose-response relationship for developmental toxicity includes an evaluationof both laboratory animal and human data. Unfortunately, adequate exposure information (and lessfrequently, actual dose data) are rarely available from human studies, so that laboratory animal dataare most frequently relied upon. Evidence for a dose-response relationship is an important criterionfor identifying an agent as causing developmental or reproductive toxicity.The lowest effective doses in adults and young are often similar or may be the same, but thetype and severity of effects may be very different, e.g., the developmental effects may be permanentwhile the maternal effects may be reversible. The difference between the maternally toxic dose andthe developmentally toxic dose may at times be related to the relative thoroughness with whichendpoints are evaluated in dams and offspring. It is also important to reiterate a point madepreviously 2 regarding the relationship of maternal and developmental toxicity. From a risk assessmentperspective, developmental toxicity in the presence of maternal toxicity cannot be simplyconsidered “secondary to maternal toxicity” and discounted. This presumption was never consideredappropriate in evaluating data for risk assessment purposes, 3,7 is still not considered a valid argumentfor ignoring developmental toxicity in an animal study, and should have been put to rest long ago.In most cases, the lowest dose at which adverse developmental or reproductive effects are seen(lowest observed adverse effect level — LOAEL), and the dose at which no adverse effects can bedetected (no observed adverse effect level — NOAEL) are determined. These doses are oftenidentified based on statistical differences from controls, but may be determined by examining thetrend in response and certain biological considerations, such as rarity of the effect. When sufficientdata are available to do mathematical modeling of the dose-response relationship, this approach ispreferable to the NOAEL approach. Mathematical modeling may be performed to determine aquantitative estimate of response at low doses in the experimental range. This approach can beused to determine the benchmark dose (BMD) and its lower confidence limit (BMDL), which maybe used in place of the NOAEL. 107,108a. Duration Adjustment and Derivation of Human Equivalent Concentrationsand Human Equivalent DosesPrior to derivation of BMDs or NOAELs, data from toxicity studies are adjusted to convert theduration of exposure in a particular study to a continuous exposure scenario. For inhalationexposures, this adjustment also involves the application of a concentration × time (c × t) adjustment,while for oral studies, the adjustment is to a daily exposure (e.g., a 5-d/week exposure is convertedto 7 d/week). For inhalation exposures, this adjustment has not traditionally been done in the caseof developmental toxicity studies as it has for other types of health effects. That is because ofconcerns about peak versus integrated exposure and the likelihood of a threshold for effects. 3However, the EPA, in a review of the RfD and RfC processes, examined additional newer data aswell as the basis for the adjustment of other endpoints and recommended that inhalation developmentaltoxicity studies be adjusted in the same way as for other endpoints. 15A number of approaches have been developed for establishing short-duration (less than lifetime)exposure limits. 109,110 Each approach addresses specific exposure scenarios, but a common featureof most of these approaches is the assumption that the probability of response depends on thecumulative exposure, i.e., the product of exposure concentration and duration. The concept of aconstant relationship between concentration and duration is generally referred to as Haber’s© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 857Law. 111,112 And even though the approach was not intended for exposures of long durations and/orlow concentrations, it has come to be more generally applied and is used in extrapolating toinhalation exposures that are not covered by the experimental data. The reliability of this approachin estimating the effects of a particular exposure at various concentrations and durations is relativelyundefined. However, recent reviews of this area have begun to question its general applicabilityover wide ranges of concentration and duration. 113,114The development of new and more appropriate models should define the relationship betweenexposure concentration and duration. Initial application of such models will require broad assumptionsabout the mode of action. Ultimately, more refined models should permit interpolation of datato untested c × t combinations, reducing the uncertainty in assessing the data and increasingconfidence in the prediction of potential human risk. The duration of exposure, even for agentswith a short half-life, can influence the developmental effects that are produced, as demonstratedby recent studies on all-trans-retinoic acid 115 and ethylene oxide. 116 The effects of exposure tohyperthermia during a brief period of development (GD 10 in the rat) have also been shown to bea function of the level and duration of temperature elevation. 117 Thus, the default adjustment ofinhalation developmental toxicity studies for duration of exposure will likely be more protectivewith adjustment from shorter to longer durations of exposure. 116,117Derivation of the human equivalent concentration (HEC) for inhalation exposures is intendedto account for pharmacokinetic differences between humans and animals, while the potentialpharmacodynamic differences are usually accounted for by a portion of the interspecies uncertaintyfactor (typically 10 0.5 ). The application of dosimetric adjustment factors (DAFs) to derive the HECis discussed in detail in the EPA’s guidance for RfC derivation. 118Currently, there is no similar procedure available to account for pharmacokinetic differencesin deriving the human equivalent dose (HED), and the basis for deriving the dose metric for oralexposure differs between cancer and noncancer risk assessment. In the case of noncancer riskassessment, oral dose is expressed on the basis of body weight 1 (e.g., milligrams per kilogram perday), while for cancer risk assessment, dose is expressed on the basis of body weight to the 0.75power (e.g., milligrams per kilogram 0.75 per day). Harmonization of these approaches is needed formore consistent dose-response assessment.b. Benchmark Dose ModelingThe benchmark dose modeling (BMD) approach has several advantages over the NOAEL approach.For example, dose-response modeling with derivation of a BMD uses all the data in fitting a model,accounts for the variability in the data, and does not limit the BMD to one of the experimentaldoses. The BMD is defined as a dose that corresponds to a particular level of response — thebenchmark response (BMR). The BMR is defined as a particular response rate for a quantalresponse, e.g., 1%, 5%, 10%, or as a specified change from control values for a continuous response,e.g., 1 SD change from the control mean. The BMR must be selected for the derivation of a BMD,and this is not straightforward, particularly in the case of continuous data, because the decision isbased on selecting the magnitude of change from controls that is considered adverse.Studies have been conducted to evaluate several approaches to calculating BMDs for standardprenatal developmental toxicity data. 119–122 These studies applied several dose-response models,both generic and developmental toxicity–specific, to a large number of standard developmentaltoxicity studies in animals (rats, mice, rabbits) that were dosed throughout the period of majororganogenesis (or in some cases throughout pregnancy). These studies showed that such modelscan be used successfully with standard developmental toxicity data. The BMD for a 5% excessrisk above controls for binomial endpoints corresponded on average to the NOAEL. Incorporationof variables, such as intralitter correlation and litter size, appeared to enhance the fit of thedevelopmental toxicity–specific models, whereas inclusion of a threshold did not. Threshold in thiscase refers to a model-derived estimate of the point at which the model can no longer distinguish© 2006 by Taylor & Francis Group, LLC

858 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONtreatment effects from background control levels and has nothing to do with a biological threshold.Application of various models to fetal weight data, a continuous endpoint, also was successful. 122Several definitions of change from control values were found to give BMDs similar to the NOAEL,including a change of 1 SD from control values.The EPA has published a draft technical guidance document for BMD derivation, 108 whichrecommends deriving, as a default, a 10% BMD and lower confidence limit (BMDL) for quantaldata, and a BMD and BMDL for 1 SD from the control mean for continuous data. These defaultvalues allow comparison between endpoints and between chemicals. The BMDL is used as thepoint of departure (POD), i.e., the point on the dose-response curve used for deriving the referencevalues or for linear low-dose extrapolation and estimation of risk.c. Application of Uncertainty Factors and Derivation of Reference Values(RfDs and RfCs)In the absence of pharmacokinetic or mode of action data to the contrary, developmental andreproductive effects (as well as many other types of noncancer health effects) are generally assumedto have a threshold or nonlinear dose-response relationship at low dose levels. Because of thenonlinear or threshold assumption, extrapolation to low dose levels by use of mathematical modelsis not typically done. Instead, reference values, such as the RfD and RfC, are derived.RfDs and RfCs are estimates of a daily exposure to the human population assumed to be withoutappreciable risk of deleterious effects over a lifetime. A major disadvantage of the RfD and RfCapproach is that it does not give an estimate of risk at particular dose levels, so when exposureoccurs at levels above the RfD or RfC, the risk is unknown.RfDs and RfCs are derived by dividing the POD by uncertainty factors that account for suchthings as animal to human extrapolation (UF A ), variations in sensitivity within the human population(UF H ), LOAEL to NOAEL extrapolation when a NOAEL has not been determined (UF L ; not appliedwhen a BMDL has been derived), subchronic to chronic extrapolation (UF S ), and deficiencies inthe database (UF D ). RfDs and RfCs are set for chronic exposures. In some cases, less than lifetimereference values are set. For example, acute, short-term, longer-term, and chronic health advisoriesare set for drinking water contaminants, and acute and short-term RfDs are sometimes set forpesticide exposures, e.g., for acute dietary and short-term dermal exposures. The EPA review ofthe RfD and RfC process recommended that less than lifetime values, in addition to the chronicRfD and RfC, be set for all chemicals where appropriate data are available. 15 In setting less thanlifetime values, it is important to consider developmental and reproductive effects along with othertypes of toxicity. The EPA’s guidelines for developmental toxicity risk assessment 3 had recommendedsetting RfD DT s to account for the fact that a lifetime exposure is not necessary to producea developmental effect. However, this approach will not be necessary if shorter-term referencevalues are generally derived to be protective of all potential health effects for a given exposurescenario and to cover all lifestages.The default value for uncertainty factors is 10, but this may be reduced (often to 10 0.5 ) dependingon the confidence in the data or information that provides assurance of reduced intra- or interspeciesvariability (see Moore et al., 123,124 for examples). The UF A and UF H are considered to cover bothpharmacokinetic and pharmacodynamic components of inter- and intraspecies differences. Chemicalspecific data on pharmacokinetics and pharmacodynamics (i.e., mode of action) may be usedto replace part or all of these UFs. The methodology for deriving RfCs uses physiological andpharmacokinetic data to replace the PK component of the UF A , reducing it to 10 0.5 . 118 A numberof studies have been done to evaluate the adequacy of the various 10-fold uncertainty factors as adefault and to determine whether these factors adequately protect susceptible subpopulations. 125–129Peleki et al. 130 used a physiological model–based approach to examine the magnitude of thepharmacokinetic part of the intraspecies UF for VOCs for adults and children. They reported that© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 859no adult-children differences in the parent chemical concentrations were likely to be seen withinhalation exposures.Models that are biologically based should provide a more accurate estimate of low-dose riskto humans. The development of biologically based dose-response models has been limited by intraandinterspecies differences in developmental and reproductive effects, a lack of understanding ofthe biological mechanisms underlying developmental and reproductive toxicity, and other factors.The advent of the use of molecular techniques for understanding developmental and reproductivebiology gives promise for understanding mechanisms of toxicity 99 and will support the constructionof biologically based models.3. Exposure AssessmentThe exposure assessment describes the magnitude, duration, schedule, and route of human exposuresfrom various sources for comparison with the toxicity data. An important point to rememberis that based on the definition of developmental and reproductive toxicity used in the EPA guidelines,3,4 almost any segment of the population may be at risk for developmental or reproductivetoxicity. The Guidelines for Estimating Exposures 131,132 discusses the various approaches that canbe taken for estimating exposures through actual monitoring or by modeling various sources andexposure scenarios. Default values have been published 133 for use in estimating exposures, e.g.,from food and water consumption in adults and children, soil ingestion in children, and respirationrates in children and adults. In some cases, behavior patterns of importance in children, e.g., crawlingon the floor, pica, and food preferences or patterns, should be taken into account. 8 Recently, theChild-Specific Exposure Factors Handbook has been published. 134 This discussion does not attemptto provide definitive guidance for an exposure assessment but focuses on those issues of specificimportance to developmental and reproductive toxicity risk assessment.Several exposure conditions are unique for developmental toxicity. For example, during pregnancya woman is exposed directly and her developing conceptus is exposed indirectly. For physicalagents, exposure may be direct to the embryo or fetus, e.g., x-irradiation and heat, with only aminimal modification of the exposure by the maternal system. Exposure of the conceptus tochemical agents is dependent on maternal absorption, distribution, metabolism, and excretion, aswell as placental metabolism and transfer. Transit time in the conceptus also may depend on itsability to metabolize and/or excrete the chemical. In a few cases, a chemical agent may have itsprimary effect on the maternal system, 135 and the effect on the conceptus does not depend onexposure to the agent and/or its metabolites, but on some factor induced in the mother. Anotherunique exposure for infants is via breast milk, particularly for agents that are fat soluble and storedin maternal fat. Data suggest that lead may be mobilized along with calcium from bone stores inwomen during pregnancy and lactation 136 and be excreted in milk, 137–139 although the levels absorbedby the infant may be low and blood levels in the baby may actually decrease during nursing. 89 Inaddition, continuous bone remodeling in the child, in addition to lead exposure from house dust,paint, etc., may contribute to blood lead. 88 Thus, exposure of the conceptus and child may not bethe same as for the pregnant or lactating mother, and measurement of the agent in cord blood andin breast milk may give a better estimate of exposure.Considerations unique to exposures that affect the male reproductive system are related to thestages of spermatogenesis. An acute exposure may result in different effects on the germ cells thatare possible to detect only at certain periods after the exposure occurs. For example, germ cellsthat were spermatozoa, spermatids, spermatocytes, or spermatogonia at the time of acute exposuretake 1 to 2, 3 to 5, 5 to 8, or 8 to 12 weeks to appear in the ejaculate. Thus, an acute exposure willnot be detected if the affected endpoint is looked for too early or too late. With longer, morecontinuous exposures, the effects may be more apparent, but timing of exposure and expectedoutcomes must still be considered.© 2006 by Taylor & Francis Group, LLC

860 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONExposures that affect the female reproductive system are also important with respect to thestage of reproductive life or different physiological states (e.g., nonpregnant, pregnant, lactating,perimenopausal, postmenopausal). Exposure of pregnant women to diethylstilbestrol (DES) at atime when critical stages of reproductive system development were occurring resulted in a numberof cases of reproductive tract abnormalities and a rare form of cancer in young women exposed inutero. 140,141 Compounds that cause destruction of primordial follicles or primary oocytes or thatincrease the rate of apoptosis in the germ cells may result in infertility or alter time to pregnancy,disrupt follicular recruitment, ovulation, and ovum transport, and lead to changes in viability orfertilizability of the egg.Children and adults are exposed directly to contaminants via water, food, air, and soil. Dustand paint or anything that can be mouthed by a young child should also be considered a potentialsource of exposure. The period and duration of exposure should be related to life stage of developmentor reproduction. For example, exposures should be characterized as preconception, first,second, or third trimester of pregnancy (or more specific periods, if possible), infancy, early, middle,and late childhood (specific ages, if possible), adolescence (stage of puberty and development),adulthood (by age), pre- or postmenopausal, older adult, and the aged. These life stages may havedifferent sensitivities to an agent, and exposure estimates should be characterized for as many aspossible. In addition, exposure to either parent prior to conception should be considered as apossibility in the production of adverse developmental effects. 142–144Exposures that can cause developmental and reproductive toxicity may be for short or longdurations. Many experimental studies have confirmed for a number of agents that a single exposureis sufficient to cause developmental or reproductive toxicity, depending on the life stage, i.e.,repeated exposure is not a necessary prerequisite for adverse outcomes. Thus, agents with shorthalf-lives, as well as those that show accumulation with repeated exposures or that have long halflives,may cause effects. The pattern of exposure, whether intermittent or continuous, as well aspeak exposures and average exposure levels over the period of exposure, should be examined.Exposures affecting development and reproduction will often result in latent effects. For example,effects of exposure during pregnancy may not be manifest until after birth in the form ofstructural malformations, growth retardation, cancer, or mental retardation or other functionaldefects. Likewise, effects of exposures that enhance follicular degeneration early in developmentmay not become apparent until an early onset of menopause is noted. Other types of effects thatmay not become apparent until long after exposure during early life stages include cancer, heartdisease, neurodegenerative disease, and shortened lifespan.4. Risk CharacterizationRisk characterization is the final step in the risk assessment process and involves integration of thehazard characterization and dose-response assessment with the exposure assessment. Integral torisk characterization is information on the quality of the data for the hazard evaluation and theexposure assessment, relevancy to humans, strengths and weaknesses of the database, assumptionsmade, scientific judgments, and the level of confidence in determining potential risks to humans.All of these factors are part of the weight-of-evidence determination. Table 21.4 gives a list ofissues to be considered in each of the components of risk assessment and can serve as a guide forthe questions to be addressed in risk characterization.The weight-of-evidence information is communicated to the risk manager along with thequantitative information on the range of effective exposure levels, dose-response modeling estimates(e.g., the BMD and BMDL), appropriate uncertainty factors for each duration reference value, aswell as the reference values. Information on developmental and reproductive toxicity is consideredan important part of the database used in calculating chronic RfDs and RfCs and must be weighedappropriately relative to other adverse outcomes. When developmental and reproductive toxicity© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 861data are not available, the database is considered to be deficient, and an additional databaseuncertainty factor may be applied. In the case of pesticides, the Food Quality Protection Act requiresthe addition of a 10-fold margin of safety to take into account the potential for pre- and postnataltoxicity and the completeness of the toxicology and exposure databases. The EPA’s Office of PesticidePrograms has published a guidance document 35 that provides direction for applying the additional 10-fold FQPA safety factor and clarifies that there is overlap between this tenfold factor and the traditionaluncertainty factors, particularly the database deficiency factor. It presumes that in most cases thetraditional uncertainty factors will be sufficient to be protective of children’s health. Nevertheless, theremay be residual uncertainties about health effects not captured in the RfD process or in the exposureassessment that provide a basis for maintaining part or all of the FQPA factor.Three types of descriptors of human risk are especially useful and important. The first of theseis related to interindividual variability, i.e., the range of variability in population response to anagent and the potential for highly sensitive or susceptible subpopulations. A default assumption ismade that the most sensitive individual in the population will be no more than 10-fold more sensitivethan the average individual; thus, a default 10-fold uncertainty factor is often applied in calculatingreference values to account for this potential difference. When data are available on highly sensitiveor susceptible subpopulations, the risk characterization can be done on them separately or by usingmore accurate factors to account for the differences. When data are not available to indicatedifferential susceptibility between developmental stages and adulthood, all stages of developmentand reproduction may be assumed to be highly sensitive or susceptible. Certain age subpopulationscan sometimes be identified as more sensitive because of critical periods for exposure; for example,pregnant or lactating women, infants, children, adolescents, adults, or the elderly. In general, notenough is understood about how to identify sensitive subpopulations without obtaining specificdata on each agent, although it is known that factors such as nutrition, personal habits (e.g., smoking,alcohol consumption, drug abuse), quality of life (e.g., socioeconomic factors), preexisting disease(e.g., diabetes), race, ethnic background or other genetic factors may predispose some individualsto be more sensitive than others to the developmental effects of various agents.The second descriptor of importance is for highly exposed individuals. These are individualswho are more highly exposed because of occupation, residential location, behavior, or other factors.For example, children are more likely than adults to be exposed to agents deposited in dust or soil,either indoors or outdoors, both because of the time spent crawling or playing on the floor or groundand the mouthing behavior of young children. The inherent sensitivity of children may also varywith age, so that both sensitivity and exposure must be considered in the risk characterization. Ifpopulation data are absent, various scenarios representing high end exposures may be assumed byuse of upper percentile or judgment based values. This approach must be used with caution, however,to avoid overestimation of exposure.The third descriptor that is sometimes used to characterize risk is the margin of exposure (MOE).The MOE is the ratio of the NOAEL (or BMDL) from the most appropriate or sensitive speciesto the estimated human exposure level from all potential sources. Considerations for the acceptabilityof the MOE are similar to those for the uncertainty factors applied to the NOAEL or BMDLto calculate the reference values. Examples of the calculation of an MOE based on developmentaltoxicity data have been described for dinoseb, 145 lithium, 123 and boric acid and borax. 124 In the caseof dinoseb, the MOEs were very small, in some cases less than 1, indicating toxicity in the animalstudies at levels to which people were being exposed. The analysis of the data on dinoseb led toan emergency suspension of this pesticide in 1986 and ultimately to its removal from the market.The risk characterization is communicated to the risk manager, who uses the results along withtechnological factors (e.g., exposure reduction measures), as well as social and economic considerations,in reaching a regulatory decision. Depending on the statute involved and other considerations,risk management decisions are usually made on a case-by-case basis. This may result indifferent, but appropriate, regulation of an agent under different statutes.© 2006 by Taylor & Francis Group, LLC

862 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 21.4Guide for developing chemical-specific risk characterizations for developmentaland reproductive effects aPart One: Summarizing Major Conclusions in Risk CharacterizationI. Hazard CharacterizationA. What are the key toxicological studies that provide the basis for health concerns for developmental andreproductive effects?How good are the key studies?Are the data from laboratory or field studies? In one or several species?What adverse developmental and reproductive endpoints were observed, and what is the basis forthe effects at the lowest exposure levels?Describe other studies that support these findings.Discuss any valid studies that conflict with these findings.B. Besides the developmental and reproductive effects observed in the key studies, are there other healthendpoints of concern? What are the significant data gaps?C. What epidemiological or clinical data are available? For epidemiological studies:What types of studies were available (e.g., human ecologic, case control, or cohort studies, or casereports or series)?Describe the degree to which exposures were described.Describe the degree to which confounding factors were considered.Describe the major demographic and other personal factors examined (e.g., age, sex, ethnic group,socioeconomic status, smoking status, occupational exposure).Describe the degree to which other causal factors were excluded.Describe the degree to which the findings were examined for biological plausibility, internal andexternal consistency of the findings, and the influence of limitations of the design, data sources,and analytic methods.D. How much is known about how (through what biological mechanism) the chemical produces adversedevelopmental and reproductive effects?Discuss relevant studies of mechanisms of action or metabolism.Does this information aid in the interpretation of the hazard data?What are the implications for potential adverse developmental and reproductive effects?E. Comment on any nonpositive data in animals or humans and whether these data were considered inthe hazard characterization.F. If adverse health effects have been observed in wildlife species, characterize such effects by discussingthe relevant issues as in A through E above.G. Summarize the hazard characterization and discuss the significance of each of the following:Confidence in the conclusionsAlternative conclusions that are also supported by the dataSignificant data gapsHighlights of major assumptionsII. Characterization of Dose ResponseA. What data were used to develop the dose response curve? Would the result have been significantlydifferent if based on a different data set?If laboratory animal data were used:Which species were used?Most sensitive, average of all species, or other?Were any studies excluded? Why?If human data were used:Which studies were used?Only positive studies, all studies, or some other combination?Were any studies excluded? Why?Was a meta-analysis performed to combine the epidemiological studies? What approach was used?Were studies excluded? Why?B. Was a model used to develop the dose response curve and, if so, which one? What rationale supportsthis choice? Is chemical-specific information available to support this approach? What other models wereconsidered?How were the reference values (or the acceptable range) calculated?What assumptions and uncertainty factors were used?What is the confidence in the estimates?C. Discuss the route, level, timing (i.e., life stage), and duration of exposure observed, as compared toexpected human exposures.Are the available data from the same route of exposure as the expected human exposures? If not,are pharmacokinetic data available to extrapolate across route of exposure?What information was used to support duration adjustment and to calculate the human equivalentdose or concentration (HED or HEC)?How far does one need to extrapolate from the observed data to environmental exposures? One totwo orders of magnitude? Multiple orders of magnitude? What is the impact of such anextrapolation?D. If adverse health effects have been observed in wildlife species, characterize dose response informationusing the process outlined in A through C above.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 863Table 21.4Guide for developing chemical-specific risk characterizations for developmentaland reproductive effects a (continued)III. Characterization of ExposureA. What are the most significant sources of environmental exposure?Are there data on sources of exposure from different media?What is the relative contribution of different sources of exposure?What are the most significant environmental pathways for exposure?B. Describe the populations that were assessed, including the general population, highly exposed groups,and highly susceptible groups, and including different life stages, e.g., pregnant women, children,adolescents, adults, elderly.C. Describe the basis for the exposure assessment, including any monitoring, modeling, or other analysesof exposure distributions, such as Monte Carlo or krieging.D. What are the key descriptors of exposure?Describe the (range of) exposures to: “average” individuals, “high-end” individuals, general population,high exposure group(s), children, susceptible populations, males, females (nonpregnant, pregnant,lactating).How was the central tendency estimate developed?What factors and/or methods were used in developing this estimate?How was the high-end estimate developed?Is there information on highly exposed subgroups?Who are they?What are their levels of exposure?How are they accounted for in the assessment?E. Is there reason to be concerned about cumulative or multiple exposures to classes of agents with asimilar mechanism of action?Are there biological, behavioral, ethnic, racial, or socioeconomic factors that may affect exposures?F. If adverse developmental and reproductive effects have been observed in wildlife species, characterizethe wildlife exposure by discussing the relevant issues in A through E above.G. Summarize exposure conclusions and discuss the following:Results of different approaches, i.e., modeling, monitoring, probability distributionsLimitations of each, and the range of most reasonable valuesConfidence in the results obtained, and the limitations to the resultsPart Two: Risk Conclusions and ComparisonsI. Risk ConclusionsA. What is the overall picture of risk, based on the hazard, quantitative dose response, and exposurecharacterizations?B. What are the major conclusions and strengths of the assessment in each of the three main analyses(i.e., hazard characterization, quantitative dose-response, and exposure assessment)?C. What are the major limitations and uncertainties in the three main analyses?D. What are the science policy options in each of the three major analyses?What are the alternative approaches evaluated?What are the reasons for the choices made?II. Risk ContextA. What are the qualitative characteristics of the developmental and reproductive hazard (e.g., voluntaryvs. involuntary, technological vs. natural, etc.)? Comment on findings, if any, from studies of risk perceptionthat relate to this hazard or similar hazards.B. What are the alternatives to this developmental or reproductive hazard? How do the risks compare?C. How does this developmental or reproductive risk compare to other risks?How does this risk compare to other risks in this regulatory program or other similar risks that theEPA has made decisions about?Where appropriate, can this risk be compared with past agency decisions, decisions by other federalor state agencies, or common risks with which people may be familiar?Describe the limitations of making these comparisons.D. Comment on significant community concerns that influence public perception of risk.III. Existing Risk InformationA. Comment on other reproductive risk assessments that have been done on this chemical by the EPA,other federal agencies, or other organizations. Are there significantly different conclusions that meritdiscussion?IV. Other InformationA. Is there other information that would be useful to the risk manager or the public in this situation that hasnot been described above?Source: Modified from EPA, 1996.© 2006 by Taylor & Francis Group, LLC

864 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONIII. EMERGING ISSUES FOR RISK ASSESSMENTA. Research Efforts to Develop Human Data on Developmental and ReproductiveEffects of Environmental Exposures — Contributions to Risk AssessmentSeveral large research efforts are underway or being planned to address data gaps on environmentalexposures and links with children’s health effects. These include the NIEHS, EPA, and CDC Centersof Excellence in Children’s Environmental Health and Disease Prevention Research.* These centers(currently 12 in number) are focusing on research to address possible links between environmentalexposures and health effects of great concern for children, e.g., neurobehavioral effects, includingautism, and respiratory diseases, including asthma.Another large effort currently being planned is the National Children’s Study, a national cohortstudy of environmental effects on child health and development (www.nationalchildrensstudy.gov).The study is being organized by the U.S. Department of Health and Human Services (DHHS) andthe EPA to follow children from early gestation through infancy, childhood, and adolescence, intoadulthood. The National Institute of Child Health and Human Development (NICHD), togetherwith a consortium of federal agencies, was charged with conducting the study by the Children’sHealth Act of 2000. The EPA, together with NICHD, NIEHS, and CDC, has been engaged inplanning the study, which is scheduled to be launched in late 2005. The NCS will evaluate theeffects of environment, defined broadly to include chemical, physical, biological, and social factors,on growth and development of children. Because data on health and environment will be collectedat times throughout the study, there will be an opportunity to evaluate links between health andenvironment in a way that has not been possible in smaller, more limited duration studies. Inaddition, genetic factors will be analyzed to allow the evaluation of gene-environment interactions.Biological and environmental samples will also be collected and stored to serve as a nationalresource. This will allow the investigation of hypotheses in addition to those posed for the corestudy, including questions that may become important in the future. 146These data, which will help us understand the effects of environmental factors on human health,has a number of implications for risk assessment. For example, the issue of relative susceptibilityof adults and children to environmental factors can be more fully explored, as well as the questionof when and to what extent children are more highly exposed to environmental agents. Because ofthe longitudinal nature and large size of the National Children’s Study, a number of important riskassessment questions can be addressed. For example:What is the role of various environmental factors in accounting for children’s health problems? Doesexposure to environmental agents increase the burden of disease in children?Are there long-term health effects, e.g., asthma, cancer, cardiovascular disease, obesity, diabetes, orneurological diseases, from early exposures of children to environmental agents?What are the factors that increase or decrease the susceptibility of children from exposure to environmentalagents?Are there disparities in children’s environmental health because of race, ethnicity, poverty, housing,income, nutrition, location near heavy industry or toxic waste sites, etc?What genetic factors can increase the susceptibility of children to environmental exposures?What is the variability in the response of children at different ages or life stages to environmentalexposures?Are the default risk assessment approaches used to account for children’s health in both the toxicityand exposure assessments sufficient to protect children?* Information on each of the centers and their research focus can be found at http://es.epa.gov/ncer/centers/.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 865B. Using Mode of Action and Pharmacokinetic Data in Risk AssessmentThe mode of action for a chemical can be defined as the key steps after interaction with the targetsite upon which all other changes are dependent. 147 Incorporating mode of action (MOA) informationinto risk assessment will reduce uncertainties, because changes in the key steps can bemeasured, and endpoints relevant to the MOA, as opposed to biomarkers, can be assessed.There are several examples of developmental and reproductive toxic agents for which MOAdata exist. One example is the androgen antagonist pesticide vinclozolin (reviewed by Euling andKimmel 148 ). Another example is the use of MOA information to classify chemicals into equivalencygroups. For example, inclusion in the dioxin and dioxinlike chemical grouping was done using thetoxic equivalency factor approach, based on binding to the aryl hydrocarbon receptor. In individualrisk assessments, information supporting a particular MOA may be available for a chemical but isnot always considered conclusive. Even in cases where the MOA has been relatively well established,such as for perchlorate or ethylene oxide, chemical-specific or biologically based doseresponsemodels have yet to be established. For atrazine, a triazine herbicide, the MOA was acritical factor in eliminating the finding of cancer in the rodent as relevant to humans, leavingreproductive effects as the most sensitive endpoint. 149The concept that chemicals with the same mode of toxicity can be considered dose additivehas broad implications for risk assessment (e.g., for NOAEL and RfD determination). Cumulativerisk for endocrine disruptors has been shown to be of concern for estrogenic chemicals. 60 Specifically,these authors found that combinations of weak estrogen agonists, when combined at dosesbelow their no-observed-effect concentration (NOEC) or EC01, were dose additive (consideringpotency differences).Developmental stage sensitivity information for a developmentally toxic agent can help informMOA (and vice versa). For example, if a critical window of exposure for a chemical correspondsto a period of male sexual differentiation, then the chemical may be affecting androgen action atsome level. Improving the incorporation of developmental stage information into risk assessmentaddresses the changes in sensitivity to exposures that occur during development. An example ofthis for carcinogenesis was recently published by Ginsberg, 150 and the concept has been incorporatedinto the recent draft guidance for assessing risks of carcinogenesis to children. 151 Efforts to compileand model developmental stage sensitivity information are ongoing. 148,152,153 Information on criticalwindows during development for different organ systems in humans and model organisms wascompiled and compared at the Workshop to Identify Critical Windows of Exposure for Children’sHealth. 61FQPA requires that cumulative effects of chemicals with the same mechanism of toxicity beincorporated into the assessment when determining the adequacy of tolerances for infants andchildren. Thus, mode and mechanism of action data became the basis for determining the cumulativerisk of multiple chemicals. The U.S. EPA Office of Pesticide Programs (OPP) developed guidancethat relies on mechanism of action for cumulative risk assessment of pesticides 154 and a policy oncommon mechanism of toxicity, defined as the same sequence of major biochemical events (alsoconsidered the “mode of action”). 155 These guidance documents were informed by a case studyand a classification system for common mechanism and mode of toxicity (based on acetylcholinesteraseactivity) developed for organophosphate pesticides. 156,157 These and other efforts culminatedin the development of the integrated, multichemical, multipathway risk assessment for organophosphatepesticides. 158FQPA also requires that tolerances on food can only be established if they do not result in harmfrom aggregate exposure to the pesticide residue, including all anticipated dietary exposure as wellas all other exposures. This means that all nonoccupational exposures must be aggregated for asingle pesticide from all routes and pathways. OPP has developed operating guidance for conductingaggregate exposure assessments. 159© 2006 by Taylor & Francis Group, LLC

866 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONFinally, as MOA is defined either in general for a particular chemical or for specific biologicalparameters and responses, we will be able to examine dose-dependent transitions from homeostaticsituations to toxicologic responses and the critical elements in pathogenesis at levels relevant to likelyhuman exposure conditions. 160 This is an area of toxicology that is of growing interest, especially asthe field becomes more focused on the events that occur at the lower end of the exposure range. 70C. Harmonization of Approaches for Human Health Risk AssessmentHarmonization of risk assessment refers to developing a consistent set of principles and guidelinesfor drawing inferences from scientific information. While it is unlikely that a single method willbe developed for the assessment of all toxicities and exposures, it is likely that endpoint-specificrisk assessment guidance will give way to guidance that is applicable to many different endpoints,relies more on our understanding of mechanism (mode) of action, and models potential responseto a stressor at several levels of biological complexity.There is progress in this area with respect to environmental health risk assessment. Followingthe NRC report on Science and Judgment in Risk Assessment, 14 the EPA organized two internalcolloquia to discuss harmonization within the context of the agency. 161,162 Following these colloquia,the agency cosponsored a Society of Toxicology workshop, Harmonization of Cancer and NoncancerRisk Assessment. 163 These efforts saw consensus develop around the importance of movingaway from traditional risk assessment dichotomies, of better defining the mode of action andtoxicity, and of assessing the entire available database in the context of exposure and response.More recently, the agency has completed a review of the RfD and RfC process that identified anumber of specific elements where the consistency of the current approach needs to be furtherevaluated. 15 Among these were the modification of study protocols to provide more comprehensivecoverage of life stages for both exposures and outcomes, the collection of more information fromless-than-lifetime exposures to evaluate latency to effect and reversibility of effect, and furtherevaluation of appropriate adjustment of doses for duration of exposure. Of particular importanceto a more unified approach is the recommendation that health effects no longer be categorized as“cancer” or “noncancer” for the purposes of hazard characterization and dose-response analysis.The agency’s Risk Assessment Forum has established a Harmonization Steering Committee thatis beginning to address many of these issues.Another level of harmonization was addressed by the EPA’s Science Policy Council in itsGuidance on Cumulative Risk Assessment, Part 1, Planning and Scoping. 164 This document supportsa global shift toward aggregate and cumulative approaches to risk assessment, attempting to“harmonize” and integrate areas of risk assessment that are now considered separately (e.g.,ecological and human health assessments, as well as cancer and noncancer assessments).How do these issues relate to reproductive and developmental toxicity? Initially, a broaderperspective of the relationship of these fields to other areas of toxicology must be developed. Toa large extent, this is consistent with reproductive and developmental toxicology methods becausethe issues addressed are not limited to a particular organ system but include a wide variety ofendpoint parameters (e.g., many specific organ systems that include growth, structural, and functionalmanifestations). However, our knowledge of basic MOA in reproduction and developmentmust greatly expand. As risk assessment moves to a more unified approach, current testing andassessment approaches may have to be modified, and the biological reasoning behind endpointspecificapproaches must be made consistent within and across all exposures and responses.D. Use of Genomics, Proteomics, and Other “Futures Technologies”DNA microarrays can be used to measure global gene expression changes in tissues after toxicagent exposure. This technology has been used to illustrate that chemicals sharing a MOA havesimilar gene expression profiles. 165,166 Hamadeh et al. 167 suggest that a gene expression profile© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 867database for existing chemicals will be very useful for screening purposes. For example, the databasecould be queried about a chemical with DNA microarray data but other missing information (e.g.,unknown MOA). While this technology generates a lot of detailed information on gene expression,it is critical to establish the links between the microarray expression profile and the adverse effectsof a given chemical before this technique can be used routinely in risk assessment. EPA’s SciencePolicy Council Interim Policy on Genomics 168 states that the incorporation of genomics data intorisk assessment will be on a case-by-case basis.Toxicogenomic data will become increasingly available. Thus, it will be important to establishwhether observed gene expression changes are required for the adverse effects, are part of acompensatory mechanism, or are related to effects seen only at high doses. In the future, data onproteomics and metabonomics may ultimately be found more useful than gene microarray data.High-throughput techniques for these assays are becoming available, but there will continue to beissues with analysis and interpretation of the data for some time to come, as there have been withgenomics. A more detailed coverage of gene expression profiling can be found in Chapter 15.E. Implications of Benefits Analysis for Risk AssessmentBenefits analysis (assessing the benefits of reducing potential health effects) is an area that iscurrently being explored. This approach typically has been used for cancer, where a linear lowdoseresponse relationship is assumed and modeling of data can be used to estimate risks at lowlevels. Estimating the benefit of reducing exposure is relatively straightforward in this case. Benefitsanalysis for other types of health effects for which a nonlinear low dose response relationship isassumed also requires the development of quantitative estimates of risk at low levels. 169–175 However,an approach for low dose quantitative risk estimation has not been adopted, so the total benefit ofreducing exposures cannot be determined and is likely underestimated. Because the draft Guidelinesfor Carcinogen Risk Assessment 16 introduced the possibility of a nonlinear dose response relationshipfor some types of cancers (e.g., thyroid tumors resulting from long-term stimulation of thethyroid), efforts are underway to develop dose response models for extrapolating nonlinear doseresponse relationships to low levels of environmental exposure. This is an area that holds a gooddeal of promise for moving forward with dose response modeling of all types of health effects andmore accurate assessment of the financial impact of health effects on society.REFERENCES1. National Research Council, Risk Assessment in the Federal Government: Managing the Process,Committee on the Institutional Means for the Assessment of Risks to Public Health, Commission onLife Sciences, National Research Council, National Academy Press, Washington, D.C., 1983, p. 17.2. Kimmel, C.A. and Kimmel, G.L., Principles of developmental toxicity risk assessment, in Handbookof Developmental Toxicology, Hood, R.D., Ed., CRC Press, Boca Raton, 1996, p. 667.3. US Environmental Protection Agency, Guidelines for developmental toxicity risk assessment, Fed.Reg., 56, 63798, December 5, 1991.4. US Environmental Protection Agency, Guidelines for reproductive toxicity risk assessment, Fed. Reg.,61, 56274, October 31, 1996.5. U.S. Environmental Protection Agency, Guidelines for neurotoxicity risk assessment, Fed. Reg., 63,26926, May 14, 1998.6. Palmer, A.K., Regulatory requirements for reproductive toxicology: theory and practice, in DevelopmentalToxicology, Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York, 1981, p. 259.7. U.S. Environmental Protection Agency, Guidelines for the health assessment of suspect developmentaltoxicants, Fed. Reg., 51, 34028, 1986.8. National Research Council, Pesticides in the Diets of Infants and Children, Committee on Pesticidesin the Diets of Infants and Children, National Research Council, 408, p. 1993.© 2006 by Taylor & Francis Group, LLC

868 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION9. Schmidt, C. W., A growth spurt in children’s health laws, Environ. Health Perspect., 109, A270, 2001.10. Colborn, T. and Clement, C., Eds., Chemically-Induced Alterations in Sexual and Functional Development-- The Wildlife/Human Connection, Princeton Scientific, Princeton, New Jersey, 1992, p. 403.11. Colborn, T., Dumanoski, D., Myers, J.P. Our Stolen Future: Are We Threatening Our Fertility,Intelligence, and Survival? Penguin, New York, 1997, p. 316.12. National Research Council, Hormonally Active Agents in the Environment, Committee on HormonallyActive Agents in the Environment, National Academy Press, Washington, D.C., 2000, p. 452.13. U.S. Environmental Protection Agency, Special Report on Environmental Endocrine Disruption: AnEffects Assessment and Analysis, EPA/630/R-96/012, February 1997.14. National Research Council, Science and Judgment in Risk Assessment, National Academy Press,Washington, D.C., 1994, p. 672.15. U.S. Environmental Protection Agency, A Review of the Reference Dose and Reference ConcentrationProcesses, EPA/630/P-02/002F, 01 Dec 2002, Risk Assessment Forum, Washington, D.C., 2002, p.192; available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=55365.16. U.S. Environmental Protection Agency, Guidelines for Carcinogen Risk Assessment (March 2005),EPA/630/P-03/001F, April 7, 2005, Risk Assessment Forum, Washington, D.C., 2003, p. 125; availableat http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=116283.17. Adams, J. and Selevan, S.G., Issues and approaches in human developmental toxicology, in DevelopmentalToxicology, 2nd ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York, 1994, p. 287.18. Tabacova, S., Little, R., Tsong, Y., Vega, A., and Kimmel, C.A., Adverse pregnancy outcomes associatedwith maternal enalapril antihypertensive treatment, Pharmacoepidem. Drug Safety, publishedonline January 3, 2003, DOI: 10.1002/pds.796.19. Kimmel, C.A., Holson, J.F., Hogue, C.J., and Carlo, G., Reliability of Experimental Studies forPredicting Hazards to Human Development, National Center for Toxicological Research, Jefferson,AR, NCTR Technical Report for Experiment No. 6015, 1984.20. Tabacova, S.A. and Kimmel, C.A., Enalapril: pharmacokinetic/dynamic inferences for comparativedevelopmental toxicity, Reprod. Toxicol., 15, 467, 2001.21. Tabacova, S.A. and Kimmel, C.A., Atenolol: pharmacokinetic/dynamic aspects of comparative developmentaltoxicity, Reprod. Toxicol., 16, 1, 2002.22. U.S. Environmental Protection Agency, OPPTS Harmonized Test Guidelines, Health Effects, 870Series, August, 1998; available at http://www.epa.gov/opptsfrs/OPPTS_Harmonized/870_Health_Effects_Test_Guidelines/Series/.23. U.S. Environmental Protection Agency, Data Call-in Notice for Cholinesterase-Inhibiting OrganophosphatePesticides, letter from Lois A. Rossi to registrants, September 10, 1999.24. Kimmel, C.A. and Makris, S.L., Recent developments in regulatory requirements for developmentaltoxicology, Toxicol. Lett., 120, 73, 2001.25. Daston, G. and Kimmel C., An Evaluation and Interpretation of Reproductive Endpoints for HumanHealth Risk Assessment, ILSI Press, Washington, D.C., 1999, p. 103.26. U.S. Environmental Protection Agency, OPPTS Health Effects Test Guidelines, 870 Series, EPA 712-C-94-208, July 1994.27. Canfield, R.L., Henderson, C.R. Jr., Cory-Slechta, D.A., Cox, C., Jusko, T.A., and Lanphear, B.P.,Intellectual impairment in children with blood lead concentrations below 10 microg per deciliter, NewEngl. J. Med., 348, 1517, 2003.28. Kimmel, C.A., Rees, D.C., and Francis, E.Z., Eds., Proceedings of the Workshop on the Qualitativeand Quantitative Comparability of Human and Animal Developmental Neurotoxicity, Neurotoxicol.Teratol., 12, 173, 1990.29. Sharbaugh, C., Viet, S.M., Fraser, A., and McMaster, S.B., Comparable measures of cognitive functionin human infants and laboratory animals to identify environmental health risks to children, Environ.Health Perspect., 111, 1630, 2003, DOI:10.1289/ehp.6205, published online July 2, 2003.30. Dostal, M., Developmental immunotoxicology, Reprod. Toxicol. 6, 179, 1992.31. Holladay, S.D. and Luster, M.I., Developmental immunotoxicology, in Developmental Toxicology,2nd ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York, 1994, p. 93.32. Holladay, S.D. and Smialowicz, R.J., Development of the murine and human immune system: differentialeffects of immunotoxicants depend on time of exposure. Environ. Health Perspect., 108(Suppl.3), 463, 2000.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 86933. Holsapple, M.P., Developmental immunotoxicology and risk assessment: a workshop summary. Hum.Exp. Toxicol., 21, 473, 2002.34. Holsapple, M.P., Developmental immunotoxicity testing: a review, Toxicology, 185, 193, 2003.35. U.S. Environmental Protection Agency, Determination of the Appropriate FQPA Safety Factor(s) forUse in the Tolerance-Setting Process, Office of Pesticide Programs, 2002; available from: http://www.epa.gov/oppfead1/trac/science/#10_fold.36. Lau, C. and Kavlock, R.J., Functional toxicity in the developing heart, lung, and kidney, in DevelopmentalToxicology, 2nd ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York, 1994, p. 119.37. Osmond, C., Barker, D. J.P., Winter, P.D., Fall, C.H.D., and Simmonds, S.J., Early growth and deathfrom cardiovascular disease in women, Br. Med. J., 307, 1519, 1993.38. Langley-Evans, S.C., Phillips, G.J., Benediktsson, R., Gardner, D.S., Edwards, C.R., Jackson, A.A.,and Seckl, J.R., Protein intake in pregnancy, placental glucocorticoid metabolism and the programmingof hypertension in the rat, Placenta, 17, 169, 1996.39. Barker, D.J., Fetal programming of coronary heart disease, Trends Endocrinol. Metab., 13, 364, 2002.40. Eriksson, J.G. and Forsen, T.J., Childhood growth and coronary heart disease in later life, Ann. Med.,34, 157, 2002.41. Snoeck, A., Remacle, C., Reusens, B., and Hoet, J.J., Effect of low protein diet during pregnancy onthe fetal rat endocrine pancreas, Biol. Neonate, 57, 107, 1990.42. Gauguier, D., Bihoreau, M.T., Ktorza, A., Berthault, M.F., and Picon, L., Inheritance of diabetesmellitus as consequence of gestational hyperglycemia in rats, Diabetes, 39, 734, 1990.43. Van Assche, F.A., Aerts, L., and Holemans, K., Metabolic alterations in adulthood after intrauterinedevelopment in mothers with mild diabetes, Diabetes, 40, 106, 1991.44. Dahri, S., Snoeck A., Reusens-Billen B., Remacle, C., and Hoet, J.J., Islet function in offspring ofmothers on low-protein diet during gestation, Diabetes, 40, 115, 1991.45. Hoet, J.J., Ozanne, S., and Reusens, B., Influences of pre- and postnatal nutritional exposures onvascular/endocrine systems in animals, Environ. Health Perspect., 108(Suppl. 3), 563, 2000.46. Eriksson, J., Forsen, T., Tuomilehto, J., Osmond, C., and Barker, D., Size at birth, childhood growthand obesity in adult life, Int. J. Obes. Relat. Metab. Disord., 25, 735, 2001.47. Eriksson, J.G., Forsen, T., Tuomilehto, J., Osmond, C., and Barker, D.J., Early adiposity rebound inchildhood and risk of Type 2 diabetes in adult life, Diabetologia, 46, 190, 2003. [Epub 2003 Jan 08].48. Cooper, C., Javaid, M.K., Taylor, P., Walker-Bone, K., Dennison, E., and Arden N., The fetal originsof osteoporotic fracture, Calcif. Tissue Int., 70, 391, 2002. [Epub 2002 Apr 19].49. Michels, K.B., Trichopoulos, D., Robins, J.M., Rosner, B.A., Manson, J.E., Hunter, D.J., Colditz,G.A., Hankinson, S.E., Speizer, F.E., and Willett, W.C., Birthweight as a risk factor for breast cancer,Lancet, 348, 542, 1996.50. Innes, K., Byers, T., and Schymura, M., Birth characteristics and subsequent risk for breast cancer invery young women, Am. J. Epidemiol., 152, 1121, 2000.51. Mellemkjaer, L., Olsen, M.L., Sorensen, H.T., Thulstrup, A.M., Olsen, J., and Olsen, J.H., Birth weightand risk of early-onset breast cancer (Denmark), Cancer Causes Control, 14, 61, 2003.52. Gray, Jr. L.E., Ostby, J.S., and Kelce, W.R., Developmental effects of an environmental antiandrogen: thefungicide vinclozolin alters sex differentiation of the male rat, Toxicol. Appl. Pharmacol., 129, 46, 1994.53. Blanck, H.M., Marcus, M., Tolbert, P.E., Rubin, C., Henderson, A.K., Hertzberg, V.S., Zhang, R.H.,and Cameron, L., Age at menarche and tanner stage in girls exposed in utero and postnatally topolybrominated biphenyl, Epidemiology, 11, 641, 2000.54. Eskenazi, B., Warner, M., Mocarelli, P., Samuels, S., Needham, L.L., Patterson, D.G., Jr., Lippman,S., Vercellini, P., Gerthoux, P.M., Brambilla, P., and Olive, D., Serum dioxin concentrations andmenstrual cycle characteristics, Am. J. Epidemiol., 156, 383, 2002.55. Kimmel, C.A., Grant, L.D., Sloan, C.S., and Gladen, B.C., Chronic low-level lead toxicity in the rat.I. Maternal toxicity and perinatal effects, Toxicol. Appl. Pharmacol., 56, 28, 1980.56. Grant, L.D., Kimmel, C.A., West, G.L., Martinez-Vargas, C.M., and Howard, J.L., Chronic low-levellead toxicity in the rat. II. Effects on postnatal physical and behavioral development, Toxicol. Appl.Pharmacol., 56, 42, 1980.57. Dearth, R.K., Hiney, J.K., Srivastava, V., Burdick, S.B., Bratton, G.R., and Dees, W.L., Effects of lead(Pb) exposure during gestation and lactation on female pubertal development in the rat, Reprod.Toxicol., 16, 343, 2002.© 2006 by Taylor & Francis Group, LLC

870 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION58. Selevan, S.G., Rice, D.C., Hogan, K.A., Euling, S.Y., Pfahles-Hutchens, A., and Bethel, J., Bloodlead concentration and delayed puberty in girls, New Engl. J. Med., 348, 1527, 2003.59. Wu, T., Buck, G.M., and Mendola, P., Blood lead levels and sexual maturation in U.S. girls: The ThirdNational Health and Nutrition Examination Survey, 1988–1994, Environ. Health Perspect., 111, 737, 2003.60. Silva, E., Rajapakse, N., and Kortenkamp, A., Something from “nothing” — eight weak estrogenicchemicals combined at concentrations below NOECs produce significant mixture effects, Environ.Sci. Technol., 36, 1751, 2002.61. Selevan, S.G., Kimmel, C.A., and Mendola, P., Eds., Identifying critical windows of exposure forchildren’s health, Environ. Health Perspect., 108(Suppl 3), 449, 2000.62. U.S. Environmental Protection Agency, Endocrine Disruptor Screening and Testing Advisory CommitteeFinal Report, EPA/743/R-98/003, August 1998; available at http://www.epa.gov/scipoly/oscpendo/history/finalrpt.htm.63. National Toxicology Program, Report of the Endocrine Disruptors Low-Dose Peer Review, August2001; available at http://ntp-server.niehs.nih.gov/htdocs/liason/LowDosePeerFinalRpt.pdf.64. Euling, S.Y., Gennings, C., Wilson, E.M., Kemppainen, J.A., Kelce, W.R., and Kimmel, C.A.,Response-surface modeling of the effect of 5-dihydrotestosterone and androgen receptor levels on theresponse to the androgen antagonist vinclozolin, Toxicol. Sci., 69, 332, 2002.65. Wolf, C.J., LeBlanc, G.A., Ostby, J.S., and Gray, L.E., Jr., Characterization of the period of sensitivityof fetal male sexual development to vinclozolin, Toxicol. Sci., 55, 152, 2000.66. vom Saal, F.S., Timms, B.G., Montano, M.M., Palanza, P., Thayer, K.A., Nagel, S.C., Dhar, M.D.,Ganjam, V.K., Parmigiani, S., and Welshons, W.V., Prostate enlargement in mice due to fetal exposureto low doses of estradiol or diethylstilbestrol and opposite effects at high doses, Proc. Natl. Acad.Sci. USA, 94, 2056, 1997.67. Kohn, M.C. and Melnick, R.L., Biochemical origins of the non-monotonic receptor-mediated doseresponse,J. Mol. Endocrinol., 29, 113, 2002.68. Almstrup, K., Fernandez, M.F., Petersen, J.H., Olea, N., Skakkebaek, N.E., and Leffers, H., Dualeffects of phytoestrogens result in u-shaped dose-response curves, Environ. Health Perspect., 110,743, 2002.69. Gaido, K.W., Maness, S.C., McDonnell, D.P., Dehal, S.S., Kupfer, D., and Safe, S., Interaction ofmethoxychlor and related compounds with estrogen receptor alpha and beta, and androgen receptor:structure-activity studies, Mol. Pharmacol., 58, 852, 2000.70. Slikker, W. Jr., Anderson, M.E., Bogdanffy, M.S., Bus, J.S., Cohen, S.D., Conolly, R.B., David, R.M.,Doerrer, N.G., Dorman, D.C., Gaylor, D.W., Hattis, D., Rogers, J.M., Setzer, R.W., Swenberg, J.A.,and Wallace, K., Dose-dependent transitions in mechanisms of toxicity, Toxicol. Appl. Pharmacol.,15, 203, 2004.71. Kimmel, C.A. and Francis, E.Z., Proceedings of the workshop on the acceptability and interpretationof dermal developmental toxicity studies, Fundam. Appl. Toxicol., 14, 386, 1990.72. Kimmel, C.A. and Young, J.F., Correlating pharmacokinetics and teratogenic endpoints, Fundam.Appl. Toxicol., 3, 250, 1983.73. Clarke, D.O., Duignan, J.M., and Welsch, F., 2-Methoxyacetic acid dosimetry-teratogenicity relationshipsin CD-1 mice exposed to 2-methoxyethanol, Toxicol. Appl. Pharmacol., 114, 77, 1992.74. Nau, H., Teratogenic valproic acid concentrations: infusion by implanted minipumps vs conventionalinjection regimen in the mouse, Toxicol. Appl. Pharmacol., 80, 243, 1985.75. Nau, H., Species differences in pharmacokinetics, drug metabolism and teratogenesis, in Pharmacokineticsin Teratogenesis, Vol. 1, Nau, H. and Scott, W. J., Eds., CRC Press, Boca Raton, 1987, p. 81.76. Terry, K.K., Elswick, B.A., Stedman, D.B., and Welsch, F., Developmental phase alters dosimetryteratogenicityrelationship for 2-methoxyethanol in CD-1 mice, Teratology, 49, 218, 1994.77. McCarver, D.G. and Hines, R.N., The ontogeny of human drug-metabolizing enzymes: phase IIconjugation enzymes and regulatory mechanisms, J. Pharmacol. Exp. Ther., 300, 361, 2002.78. Cresteil, T., Onset of xenobiotic metabolism in children: toxicological implications, Food Addit.Contam., 15(Suppl), 45, 1998.79. Koukouritaki, S.B., Simpson, P., Yeung, C.K., Rettie, A.E., and Hines, R.N., Human hepatic flavincontainingmonooxygenases 1 (FMO1) and 3 (FMO3) developmental expression, Pediatr. Res., 51,236, 2002.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 87180. Gabrielsson, J.L., Johansson, P., Bondesson, U., and Paalzow, L.K., Analysis of methadone dispositionin the pregnant rat by means of a physiological flow model, J. Pharmacokinet. Biopharm., 13, 355,1985.81. O’Flaherty, E.J., Scott, W.J., Nau, H., and Beliles, R.P., Simulation of valproic acid kinetics in primateand rodent pregnancy by means of physiologically-based models, Teratology, 43, 457, 1992.82. Clarke, D.O., Elswick, B.A., Welsch, F., and Conolly, R.B., Pharmacokinetics of 2-methoxyethanoland 2-methoxyacetic acid in the pregnant mouse: a physiologically-based mathematical model, Toxicol.Appl. Pharmacol., 121, 239, 1993.83. Mattison, D.R., Physiologic variations in pharmacokinetics during pregnancy, in Drug and ChemicalAction in Pregnancy: Pharmacologic and Toxicologic Principles, Fabro, S. and Scialli, A.R., Eds.,Marcel Dekker, New York, 1986, p. 37.84. Androit, M.D. and O’Flaherty, E.J., Development of a Physiologically-Based Pharmacokinetic Modelfor Lead during Pregnancy, Report prepared for the U.S. Environmental Protection Agency underCooperative Agreement No. 820303-01-0, 1995.85. Luecke, R.H., Wosilait, W.D., Pearce, B.A., and Young, J.F., A physiologically based pharmacokineticcomputer model for human pregnancy, Teratology, 49, 90, 1994.86. Byczkowski, J.Z., Gearhart, J.M., and Fisher, J.W., “Occupational” exposure of infants to toxicchemicals via breast milk, Nutrition, 10, 43, 1994.87. Byczkowski, J.Z. and Lipscomb, J.C., Physiologically based pharmacokinetic modeling of the lactationaltransfer of methylmercury, Risk Anal., 21, 869, 2001.88. O’Flaherty, E.J., Physiologically based models for bone-seeking elements. V. Lead absorption anddisposition in childhood, Toxicol. Appl. Pharmacol., 131, 297, 1995.89. Manton, W.I., Angle, C.R., Stanek, K.L., Reese, Y.R., and Kuehnemann, T.J., Acquisition and retentionof lead by young children, Environ. Res., 82, 60, 2000.90. Ginsberg, G., Hattis, D., Sonawane, B., Russ, A., Banati, P., Kozlak, M., Smolenski, S., and Goble,R., Evaluation of child/adult pharmacokinetic differences from a database derived from the therapeuticdrug literature, Toxicol. Sci., 66, 185, 2002.91. Hattis, D., Ginsberg, G., Sonawane, B., Smolenski, S., Russ, A., Kozlak, M., Goble, R., Differencesin pharmacokinetics between children and adults--II. Children’s variability in drug elimination halflivesand in some parameters needed for physiologically-based pharmacokinetic modeling, Risk Anal.,23, 117, 2003.92. O’Flaherty, E.J. and Clarke, D.O. Pharmacokinetic/pharmacodynamic approaches for developmentaltoxicity, in Developmental Toxicology, 2nd ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press,New York, 1994, p. 215.93. U.S. Environmental Protection Agency, Exploration of Perinatal Pharmacokinetic Issues, EPA/630/R-01/004. May 10, 2001, Risk Assessment Forum, Washington, D.C.; available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29420.94. Keys, D.A., Wallace, D.G., Kepler, T.B., and Conolly, R.B., Quantitative evaluation of alternativemechanisms of blood and testes disposition of di(2-ethylhexyl) phthalate and mono(2-ethylhexyl)phthalate in rats, Toxicol. Sci., 49, 172, 1999.95. Plowchalk, D.R. and Teeguarden, J., Development of a physiologically based pharmacokinetic modelfor estradiol in rats and humans: a biologically motivated quantitative framework for evaluatingresponses to estradiol and other endocrine-active compounds, Toxicol. Sci., 69, 60, 2002.96. Smith, B.J., Carter, D.E., and Sipes, I.G., Comparison of the disposition and in vitro metabolism of4-vinylcyclohexene in the female mouse and rat, Toxicol. Appl. Pharmacol., 105, 364, 1990.97. Fontaine, S.M., Hoyer, P.B., and Sipes, I.G., Evaluation of hepatic cytochrome P4502E1 in the speciesdependentbioactivation of 4-vinylcyclohexene, Life Sci., 69, 923, 2001.98. U.S. Environmental Protection Agency, Endocrine Disruptor Screening Program Web Site,http://www.epa.gov/scipoly/oscpendo/index.htm, 2003.99. National Research Council, Scientific Frontiers in Developmental Toxicology and Risk Assessment,National Academy Press, Washington, D.C., 2000, p. 354.100. Shuey, D.L., Lau, C., Logsdon, T.R., Zucker, R.M., Elstein, K.H., Narotsky, M.G., Setzer, R.W.,Kavlock, R.J., and Rogers, J.M., Biologically-based dose-response modeling in developmental toxicology:biochemical and cellular sequelae of 5-fluorouracil exposure in the developing rat, Toxicol.Appl. Pharmacol., 126, 129, 1994.© 2006 by Taylor & Francis Group, LLC

872 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION101. Lau, C., Mole, M.L., Copeland, M.F., Rogers, J.M., Kavlock, R.J., Shuey, D.L., Cameron, A.M., Ellis,D.H., Logsdon, T.R., Merriman, J., and Setzer, R.W., Toward a biologically based dose-responsemodel for developmental toxicity of 5-fluorouracil in the rat: acquisition of experimental data, Toxicol.Sci., 59, 37, 2001.102. Setzer, R.W., Lau, C., Mole, M.L., Copeland, M.F., Rogers, J.M., and Kavlock, R.J., Toward abiologically based dose-response model for developmental toxicity of 5-fluorouracil in the rat: amathematical construct, Toxicol. Sci., 59, 49, 2001.103. Barton, H.A., Andersen, M.E., and Clewell, J.H., Harmonization: developing consistent guidelinesfor applying mode of action and dosimetry information to cancer and noncancer risk assessment,Human Ecol. Risk Assess., 4, 75, 1998.104. Barton, H.A., Deisinger, P.J., English, J.C., Gearhart, J.M., Faber, W.D., Tyler, T.R., Banton, M.I.,Teeguarden, J., and Andersen, M.E., Family approach for estimating reference concentrations/dosesfor series of related organic chemicals, Toxicol. Sci., 54, 251, 2000.105. Barton, H.A. and Clewell, J.H., Evaluating effects of trichloroethylene: dosimetry, mode of action,and risk assessment, Environ. Health Perspect., 108(Suppl. 2), 323, 2000.106. Clewell, J.H., Andersen, M.E., and Barton, H.A., A consistent approach for the application of pharmacokineticmodeling in cancer and noncancer risk assessment, Environ. Health Perspect., 110, 85, 2002.107. Crump, K.S., A new method for determining allowable daily intakes, Fund. Appl. Toxicol., 4, 854,1984.108. U.S. Environmental Protection Agency, Benchmark Dose Technical Guidance Document, ExternalReview Draft, EPA/630/R-00/001, October 2000; available at http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=20871.109. Jarabek, A.M., Consideration of temporal toxicity challenges current default assumptions, Inhal.Toxicol., 7, 927, 1995.110. Kimmel, G.L., Exposure-duration relationships: the risk assessment process for health effects otherthan cancer, Inhal. Toxicol., 7, 873, 1995.111. Witschi, H., The story of the man who gave us Haber’s Law, Inhal. Toxicol., 9, 201, 1997.112. Witschi, H., Some notes on the history of Haber’s Law, Toxicol. Sci., 50, 164, 1999.113. Eastern Research Group, Summary of the U.S. EPA Workshop on the Relationship Between ExposureDuration and Toxicity, Prepared for the U. S. Environmental Protection Agency, Risk AssessmentForum, Washington, D.C., under Contract No. 68-D5-0028, 1998; available from: Technical InformationStaff (202-564-3261).114. Pierano, W.B., Mattie, D., and Smith, P., Proceedings of the Conference on Temporal Aspects in RiskAssessment for Noncancer Endpoints, Inhal. Toxicol, 7, 837, 1995.115. Tzimas, G., Thiel, R., Chahoud, I., and Nau, H., The area under the concentration-time curve of alltrans-retinoicacid is the most suitable pharmacokinetic correlate to the embryotoxicity of this retinoidin the rat, Toxicol. Appl. Pharmacol., 143, 436, 1997.116. Weller, E., Long, N., Smith, A., Williams, P., Ravi, S., Gill, J., Henessey, R., Skornik, W., Brain, J.,Kimmel, C., Kimmel, G., Holmes, L. and Ryan, L., Dose-rate effects of ethylene oxide exposure ondevelopmental toxicity, Toxicol. Sci., 50, 259, 1999.117. Kimmel, G.L., Williams, P.L., Claggett, T.W., and Kimmel, C.A., Response-surface analysis ofexposure-duration relationships: the effects of hyperthermia on embryonic development of the rat invitro, Toxicol. Sci., 69, 391, 2002.118. U.S. Environmental Protection Agency, Methods for Derivation of Inhalation Reference Concentrationsand Application of Inhalation Dosimetry, EPA/600/8-90/066F, 1994; available from: NationalTechnical Information Service, Springfield, VA.119. Faustman, E.M., Allen, B.C., Kavlock, R.J., and Kimmel, C.A., Dose-response assessment for developmentaltoxicity: I. Characterization of data base and determination of NOAELs, Fundam. Appl.Toxicol., 23, 478, 1994.120. Allen, B.C., Kavlock, R.J., Kimmel, C.A., and Faustman, E.M., Dose-response assessment for developmentaltoxicity: II. Comparison of generic benchmark dose estimates with NOAELs, Fundam. Appl.Toxicol., 23, 487, 1994.121. Allen, B.C., Kavlock, R.J., Kimmel, C.A., and Faustman, E.M., Dose-response assessment for developmenttoxicity: III. Statistical models, Fundam. Appl. Toxicol., 23, 496, 1994.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 873122. Kavlock, R.J., Allen, B.C., Kimmel, C.A. and Faustman, E.M., Dose-response assessment for developmentaltoxicity. IV. Benchmark doses for fetal weight changes, Fundam. Appl. Toxicol., 26, 211, 1995.123. Moore, J.A. and an IEHR Expert Scientific Committee, An assessment of lithium using the IEHRevaluative process for assessing human developmental and reproductive toxicity of agents, Reprod.Toxicol., 9, 175, 1995.124. Moore, J.A. and an Expert Scientific Committee, An assessment of boric acid and borax using theIEHR Evaluative Process for assessing human developmental and reproductive toxicity of agents,Reprod. Toxicol., 11, 123, 1997.125. Dourson, M.L., Felter, S.P., and Robinson, D., Evolution of science-based uncertainty factors innoncancer risk assessment, Regul. Toxicol. Pharmacol., 24, 108, 1996.126. Renwick, AG., Toxicokinetics in infants and children in relation to the ADI and TDI, Food Addit.Contam., 15(Suppl), 17, 1998.127. Renwick, A.G. and Lazarus, N.R., Human variability and noncancer risk assessment — an analysisof the default uncertainty factor, Regul. Toxicol. Pharmacol., 27, 3, 1998.128. Calabrese, E.J., Assessing the default assumption that children are always at risk, Human Ecol. RiskAssess., 7, 37, 2001.129. Schilter, B., Renwick, A.G., and Huggett, A.C., Limits for pesticide residues in infant food: a safetybasedapproach, Regul. Toxicol. Pharmacol., 24, 126, 1996.130. Peleki, M., Gephart, L.A., and Lerman, S.E., Physiological-model-based derivation of the adult andchild pharmacokinetic intraspecies uncertainty factors for volatile organic compounds, Regul. Toxicol.Pharmacol., 33, 12, 2001.131. U.S. Environmental Protection Agency, Guidelines for estimating exposures, Fed. Reg., 51, 34042, 1986.132. U.S. Environmental Protection Agency, Guidelines for exposure assessment, Fed. Reg., 57, 22888,May 29, 1992.133. U.S. Environmental Protection Agency, Exposure Factors Handbook, Office of Health and EnvironmentalAssessment, Washington, D.C., EPA/600/8-98/043, 1990.134. U.S. Environmental Protection Agency, Child-Specific Exposure Factors Handbook (Interim Report),EPA-600-P-00-002B, 01 Sep 2002, Office of Research and Development, National Center for EnvironmentalAssessment, Washington Office, Washington, D.C., 2002, p. 448; available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=55145135. Daston, G.P., Relationships between maternal and developmental toxicity, in Developmental Toxicology,2nd ed., Kimmel, C.A. and Buelke-Sam, J., Eds., Raven Press, New York, 1994, p. 189.136. Manton, W.I., Angle C.R., Stanek, K.L., Kuntzelman, D., Reese, Y.R., and Kuehnemann, T.J., Releaseof lead from bone in pregnancy and lactation, Environ. Res., 92, 139, 2003.137. Silbergeld, E.K., Lead in bone: implications for toxicology during pregnancy and lactation, Environ.Health Perspect., 91, 63, 1991.138. Silbergeld, E.K., Schwartz, J., and Mahaffey, K., Lead and osteoporosis: mobilization of lead frombone in postmenopausal women, Environ. Res., 47, 79, 1988.139. Watson, L.R., Gandley, R., and Silbergeld, E.K., Lead: interactions of pregnancy and lactation withlead exposure in rats, Toxicologist, 13, 349, 1993.140. Herbst, A.L., Ulfelder, H., and Poskanzer, D.C., Adenocarcinoma of the vagina. Association ofmaternal stilbestrol therapy with tumor appearance in young women, New Engl. J. Med., 284, 878,1971.141. Herbst, A.L., Hubby, M.M., Blough, R.R., and Azizi, F., A comparison of pregnancy experience inDES-exposed and DES-unexposed daughters, J. Reprod. Med., 24, 62, 1980.142. Olshan, A.F. and Mattison, D.R., Eds., Male-Mediated Developmental Toxicity, Plenum Press, NewYork, 1994.143. Robaire, B. and Hales, B.F., Eds., Advances in Male Mediated Developmental Toxicity, Advances inExperimental Medicine and Biology, Vol. 518, Kluwer Academic, New York, 2003, p. 298.144. Selevan, S.G., Kimmel, C.A., and Mendola, P., Identifying critical windows of exposure for children’shealth, Environ. Health Perspect., 108(Suppl 3), 451, 2000.145. U.S. Environmental Protection Agency, Pesticide products containing dinoseb; notices, Fed. Reg., 51,36645, 1986.© 2006 by Taylor & Francis Group, LLC

874 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION146. The National Children’s Study Interagency Coordinating Committee (Branum, A.M., Collman, G.W.,Correa, A., Keim, S.A., Kessel, W., Kimmel, C.A., Klebanoff, M.A., Longnecker, M.P., Mendola, P.,Rigas, M., Selevan, S.G., Scheidt, P.C., Schoendorf, K., Smith-Khuri, E., and Yeargin-Alsopp, M.),The National Children’s Study of environmental effects on child health and development, Environ.Health Perspect., 111, 642, 2003.147. International Programme on Chemical Safety. IPCS Workshop on Developing a Conceptual Frameworkfor Cancer Risk Assessment, 16-18 February, Lyon, France, IPCS/99.6, final draft, 1999.148. Euling, S.Y. and Kimmel, C.A., Developmental stage sensitivity and mode of action information forandrogen agonists and antagonists, Sci. Total Environ., 274, 103, 2001.149. Stoker, T.E., Guidici, D.L., Laws, S.C., and Cooper, R.L., The effects of atrazine metabolites onpuberty and thyroid function in the male Wistar rat, Toxicol. Sci., 67, 198, 2002.150. Ginsberg, G.L., Assessing cancer risks from short-term exposures in children, Risk Anal., 23, 19, 2003.151. U.S. Environmental Protection Agency, Supplemental Guidance for Assessing Cancer Susceptibilityfrom Early-Life Exposure to Carcinogens, EPA/630/R-03/003F. April 7, 2005, Risk Assessment Forum,Washington, D.C., 2005, p. 126; available at http://cfpub.epa.gov/ncea/raf/recordisplay.cfm?deid=116283.152. Clewell, H.J., Teeguarden, J., McDonald, T., Sarangapani, R., Lawrence, G., Covington, T., Gentry,P.R., and Shipp, A.M., Review and evaluation of the potential impact of age and gender-specificpharmacokinetic differences on tissue dosimetry, Crit. Rev. Toxicol., 32, 329, 2002.153. Price, K., Haddad, S., and Krishnan, K., Physiological modeling of age-specific changes in thepharmacokinetics of organic chemicals in children, J. Toxicol. Environ. Health, 14, 66, 417, 2003.154. U.S. Environmental Protection Agency, Guidance on Cumulative Risk Assessment of Pesticide ChemicalsThat Have a Common Mechanism of Toxicity, January 16, 2002, Docket Number: OPP-00658B;available at http://www.epa.gov/pesticides/trac/science/cumulative_guidance.pdf.155. U.S. Environmental Protection Agency, Guidance for Identifying Pesticide Chemicals and OtherSubstances That Have a Common Mechanism of Toxicity (Revised), February 5, 1999, Docket Number:OPP-00542; available at http://www.epa.gov/pesticides/trac/science/cumulative_guidance.pdf.156. Mileson, B.E., Chambers, J.E., Chen, W.L., Dettbarn, W., Ehrich, M., Eldefrawi, A.T., Gaylor, D.W.,Hamernik, K., Hodgson, E., Karczmar, A.G., Padilla, S., Pope, C.N., Richardson, R.J., Saunders,D.R., Sheets, L.P., Sultatos, L.G., and Wallace, K.B., Common mechanism of toxicity: a case studyof organophosphorus pesticides, Toxicol. Sci., 41, 8, 1998.157. ILSI Risk Science Institute, A Framework for Cumulative Risk Assessment, Report of an ILSI RiskScience Institute workshop, 2001.158. US Environmental Protection Agency, Organophosphate Pesticides: Revised OP Cumulative RiskAssessment, June 10, 2002; available at http://www.epa.gov/pesticides/cumulative/rra_op/.159. U.S. Environmental Protection Agency, Guidance for performing aggregate exposure and risk assessments,Fed. Reg., 66, 59428, November 28, 2001.160. Slikker, W. Jr., Anderson, M.E., Bogdanffy, M.S., Bus, J.S., Cohen, S.D., Conolly, R.B., David, R.M.,Doerrer, N.G., Dorman, D.C., Gaylor, D.W., Hattis, D., Rogers, J.M., Setzer, R.W., Swenberg, J.A.,and Wallace, K., Dose-dependent transitions in mechanisms of toxicity: case studies, Toxicol. Appl.Pharmacol., 15, 226, 2004.161. U.S. Environmental Protection Agency, Summary of the U.S. EPA Colloquia on a Framework for HumanHealth Risk Assessment, Vol. 1; available at http://www.Cfpub.epa.gov/ncea/raf/index.cfm, 1997.162. U.S. Environmental Protection Agency, Summary of the U.S. EPA Colloquia on a Framework forHuman Health Risk Assessment, Vol. 2; available at http://www.Cfpub.epa.gov/ncea/raf/index.cfm,1998.163. Bogdanffy, M.S., Daston, G., Faustman, E.M., Kimmel, C.A., Kimmel, G.L., Seed, J., and Vu, V.,Harmonization of cancer and non-cancer risk assessment: proceedings of a consensus-building workshop,Toxicol. Sci., 61, 18, 2001.164. US Environmental Protection Agency, Guidance on Cumulative Risk Assessment, Part 1, Planningand Scoping, Science Policy Council, July 3, 1997; available at http://www.epa.gov/osp/spc/cumrisk2.htm.165. Hamadeh, H.K., Bushel, P.R., Jayadev, S., Martin, K., DiSorbo, O., Sieber, S., Bennett, L., Tennant,R., Stoll, R., Barrett, J.C., Blanchard, K., Paules, R.S., and Afshari, C.A., Gene expression analysisreveals chemical-specific profiles, Toxicol. Sci., 67, 219, 2002.© 2006 by Taylor & Francis Group, LLC

DEVELOPMENTAL AND REPRODUCTIVE TOXICITY RISK ASSESSMENT 875166. Waring, J.F., Jolly, R.A., Ciurlionis, R., Lum, P.Y., Praestgaard, J.T., Morfitt, D.C., Buratto, B., Roberts,C., Schadt, E., and Ulrich, R.G., Clustering of hepatotoxins based on mechanism of toxicity usinggene expression profiles, Toxicol. Appl. Pharmacol., 175, 28, 2001.167. Hamadeh, H.K., Bushel, P.R., Jayadev, S., DiSorbo, O., Bennett, L., Li, L., Tennant, R., Stoll, R.,Barrett, J.C., Paules, R.S., Blanchard, K., and Afshari, C.A., Prediction of compound signature usinghigh density gene expression profiling, Toxicol. Sci., 67, 232, 2002.168. U.S. Environmental Protection Agency, Science Policy Council Interim Policy on Genomics, July 25,2002; available at http://www.epa.gov/osp/spc/genomics.htm.169. Baird, S.J.S., Cohen, J.T., Graham, J.D., Shlyakhter, A.I., and Evans, J.S., Noncancer risk assessment:a probabilistic alternative to current practice, Human Ecol. Risk Assess., 2, 79, 1996.170. Brand, K.P., Rhomberg, L., and Evans, J.S., Estimating noncancer uncertainty factors: are ratiosNOAELs informative? Risk Anal., 19, 295, 1999.171. Evans, J.S., Rhomberg, L.R., Williams, P.L., Wilson, A.M., and Baird, S.J.S., Reproductive anddevelopmental risks from ethylene oxide: a probabilistic characterization of possible regulatory thresholds,Risk Anal., 21, 697, 2001.172. Gaylor, D.W. and Kodell, R.L., Percentiles of the product of uncertainty factors for establishingprobabilistic reference doses, Risk Anal., 20, 245, 2000.173 Maull, E.A., Cogliano, V.J., Scott, C.S., Barton, H.A., Fisher, J.W., Greenberg, M., Rhomberg, L.,and Sorgen, S.P., Trichloroethylene health risk assessment: a new and improved process. Drug Chem.Toxicol., 20, 427, 1997.174 Slob, W. and Pieters, M.N., A probabilistic approach for deriving acceptable human intake limits andhuman health risks from toxicological studies: general framework, Risk Anal., 18, 787, 1998.175. Swartout, J.C., Price, P.S., Dourson, M.L., Carlson-Lynch, H.L., and Keenan, R.E., A probabilisticframework for the reference dose (probabilistic RfD), Risk Anal., 18, 271, 1998.© 2006 by Taylor & Francis Group, LLC

CHAPTER 22Principles of Risk Assessment — FDA PerspectiveThomas F. X. Collins, Robert L. Sprando, Mary E. Shackelford, Marion F. Gruber, andDavid E. MorseCONTENTSI. Introduction and Background ............................................................................................878II. Food Safety ........................................................................................................................878A. Perspective .................................................................................................................878B. The Statutory Safety Standard for Additives ............................................................8791. Guidance for Food Additives ..............................................................................8802. Estimating Exposure to a Direct Food Additive.................................................8803. Safety Testing of a Food or Color Additive........................................................881III.Vaccine Safety....................................................................................................................890A. Perspective .................................................................................................................890B. General Considerations..............................................................................................891C. Specific Considerations .............................................................................................8911. Animal Models ....................................................................................................8912. Dose .....................................................................................................................8923. Immunization Interval .........................................................................................8924. Exposure Period and Follow-Up Period .............................................................8925. Study Endpoints...................................................................................................892D. Summary Remarks ....................................................................................................893IV. Drug Safety ........................................................................................................................893A. Perspective .................................................................................................................893B. Product Labels ...........................................................................................................893C. Pregnancy Labeling — Practical Considerations......................................................894D. Assessment of Reproductive and Developmental Toxicity ......................................8941. Nonclinical Reproductive Toxicity Data Assessment — Future ........................8952. Data Needed for the Integration Process ............................................................8953. Types of Reproductive and Developmental Toxicity Evaluated.........................8954. The Integration Process .......................................................................................8965. Section A. Overall Decision Tree Applicable to All Data-Sets..........................8966. Section B. No Signal (Applicable to Data-sets without Evidenceof Reproductive or Developmental Toxicity)......................................................8987. Section C. Applicable to Data-sets with One or More Positive Indicationsof Reproductive or Developmental Toxicity ......................................................900877© 2006 by Taylor & Francis Group, LLC

878 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION8. Factor 1. Signal Strength, Part I..........................................................................9009. Factor 2. Signal Strength, Part II ........................................................................90210. Factor 3. Pharmacodynamics...............................................................................90311. Factor 4. Concordance between the Test Species and Humans .........................90412. Factor 5. Relative Exposures...............................................................................90513. Factor 6. Class Alerts ..........................................................................................90614. Summary and Integration of Positive Findings ..................................................906Acknowledgments..........................................................................................................................907References ......................................................................................................................................907I. INTRODUCTION AND BACKGROUNDThe mission of the Food and Drug Administration (FDA) is “to promote and protect the publichealth by helping safe and effective products reach the market in a timely way, and monitoringproducts for continued safety after they are in use” (http://www.fda.gov). The FDA thus regulatesa multitude of diverse consumer products.Five centers within the FDA are responsible for regulating the use of a wide range of productsfor humans and animals. For humans, the safety of food is regulated by the Center for Food Safetyand Applied Nutrition (CFSAN) (excluding meat and poultry products regulated by the U.S.Department of Agriculture). Human drugs used for the treatment and prevention of disease areregulated by the Center for Drug Evaluation and Research (CDER). Biologics (a category thatincludes vaccines, blood products, biotechnology products, and gene therapy) are regulated by theCenter for Biologics Evaluation and Research (CBER). Medical devices (a category that includesitems that range from simple tongue depressors to complex items such as pacemakers and dialysismachines) are regulated by the Center for Devices and Radiological Health. The Center forVeterinary Medicine at FDA regulates drugs and devices used for animals (pets and also animalsthat produce food). Before manufacturers can market animal drugs, they must obtain FDA approvalby providing proof of their safety and effectiveness. Livestock drugs are also evaluated for theirsafety to the environment and to the people who eat the animal products. Of the many productsregulated by the FDA, only human food additives (regulated by CFSAN), drugs (regulated byCDER), and vaccines (regulated by CBER) are discussed here. The emphasis is on how reproductiveand developmental toxicity studies are utilized in risk assessment at the FDA. The chapter reflectsthe complexity of the utilization of these studies within the agency.A. PerspectiveII. FOOD SAFETYFood safety in the United States is intricately interwoven with legal decisions. From the mid-1800sto the early 1900s, use of chemical preservatives and toxic colors in foods was virtually uncontrolled.In 1938, Congress passed the Federal Food, Drug, and Cosmetic Act (FFDCA). Although imperfect,the law provided some measure of food safety. Some of the provisions were: poisonous substancescould no longer be added to foods except where unavoidable or required in production; factoriescould be inspected; food standards were required to be set up when needed in the interest ofconsumers; legal remedies against violations included Federal court injunctions.During the 1940s and early 1950s, with the increased needs of World War II and rapid advancesin chemical and food technology, many new food additives were introduced by the food industry.Congress and the general public began to be concerned and to ask questions about the safety ofsome of the additives. The Pesticide Amendments of 1956 were the first Congressional actions© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 879requiring the preclearance of safety studies for chemicals added to food and granted the FDAauthority to set tolerances for pesticides in foods (Section 408).In 1958, Congress enacted the Food Additives Amendment (Section 409) to the FFDCA. Thisamendment requires that a food additive be shown to be safe under its intended conditions of usebefore it is allowed in food. The amendment requires premarket safety evaluation of food additivesand provides the FDA with authority to issue tolerances and regulations approving the use of foodadditives. In establishing the standard for demonstrating the safety of a food additive, Congressrecognized that it is not possible to determine with absolute certainty that no harm will result fromthe intended use of an additive. Specifically, therefore, the legislation requires that the petitionerprovide proof of reasonable certainty that no harm will result from the intended use of the additive.Prior to 1958, upon request of the industry, FDA periodically would review the safety of variouschemicals and render letters of advisory opinion on their acceptability, under certain conditions,as safe additives in food. 1 These written FDA Advisory opinions were recognized by Congress inthe 1958 Food Additives Amendment as “prior sanctions.” In the food additives regulations, theseare categorized in 21 CFR Part 181.Under Section 409 of the act as originally established, food additives require a food additivepetition, premarket approval by FDA, and publication of a regulation authorizing their intendeduse. The Food Additives Amendment applies to both “direct” and “indirect” food additives. Directadditives (21 CFR Part 172) are substances added deliberately in finite amounts, for whatevertechnical purpose served, and remain in the food as consumed. Secondary food additives (21 CFRPart 173) represent substances added to food or food components during manufacture or processingbut removed before the final product is consumed. Indirect food additives (21 CFR Parts 175 and179) are not intentionally added to food; they are substances used as articles or components ofarticles that are intended for use in packaging, transporting, or holding food. Because indirect foodadditives are not intended to become components of the food, their presence in food may be theresult of migration or inadvertent extraction from the food contact surface. 1Section 309 of the Food and Drug Administration Modernization Act of 1997 (FDAMA) amendsSection 409 of the FFDCA (21 USC 348) to establish a notification procedure as the primarymethod by which the FDA regulates food additives that are food contact substances (FCS). Foodcontact substances include all substances that are intended for use as components of materials usedin manufacturing, packaging, transporting, or holding food if the use is not intended to have anytechnical effect in the food.The Delaney Clause, part of the 1958 Food Additives Amendment (Section 409) to the FFDCA,sets a zero-tolerance cancer-risk standard for substances added directly or indirectly to food. Thisfar-reaching standard is still the priority test for food additives, and any other risks or risk assessmentsare based on knowledge of the noncarcinogenic properties of the additive.In 1960, Congress passed the Color Additive Amendments (Section 721) of the FFDCA, whichrequired that color additives be shown to be safe for use in or on food, drugs, or cosmetics priorto approval. The amendments created an analogous premarket approval process for color additivesused in foods, drugs, and cosmetics, or medical devices. Approved color additives are listed in 21CFR Parts 73 and 74.The act also recognizes GRAS (Generally Recognized as Safe) status for certain food ingredients.GRAS substances are generally recognized by experts as safe, based either on their extensivehistory of use in food before 1958 or on generally available and accepted scientific evidence. 1 Theavailable data and information must establish that the intended use of the substance is safe.B. The Statutory Safety Standard for AdditivesThe Food Additives Amendment of the 1938 Act requires that a food additive be shown to be safeunder the intended conditions of use before it is allowed in food. This safety requirement is oftenreferred to as the general safety clause for food additives. Although what is meant by “safe” is not© 2006 by Taylor & Francis Group, LLC

880 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONexplained in the general safety clause for food additives, FDA regulations define “safety” as areasonable certainty in the minds of competent scientists that the substance is not harmful underthe intended conditions of use. It does not require proof beyond any possible doubt that no harmwill result under any conceivable circumstance. This concept of safety requires the considerationof available data on the additive in question, as well as additional relevant information including:(1) the probable consumption of the additive, (2) the cumulative effect of such additive in the dietof humans or animals, and (3) safety factors that qualified experts consider appropriate.1. Guidance for Food AdditivesThe act requires that chemicals added to food be tested and/or evaluated for safety. This is a generalrequirement for food additives and color additives. Safety evaluation usually involves toxicologytesting by a battery of studies in animals. Protocols for safety testing reflect the increased sophisticationof the science of toxicology as well as an increased concern about food safety. 1 Despitethe increased sophistication of knowledge and techniques, the four principles of toxicity testing forthe evaluation of food additives published by the National Research Council (NRC) 2 remainessentially the same. These principles are:• Animal test endpoints are appropriate for identifying human risk.• Extrapolation of safety information from animal test data to humans requires an understanding ofspecies differences in metabolism and disposition, sensitivity of target tissue, and rates of injuryand repair.• The principle of dose response relationship, that intensity, time of onset, and duration of biologicaleffects are related to dose, is central to toxicological evaluation. Generally, the greater the dose,the greater the likelihood of observing a toxic effect.• The potential exposure from the intended use must be assessed in the evaluation of the safety offood additives.Shortly after the NRC publication, the FDA issued Toxicological Principles for the Assessmentof Direct Food Additives and Color Additives Used in Food. 3 The publication was intended to serveseveral purposes: to assess a substance’s potential for causing toxicity, to determine if safe conditionsof use can be established for food additives, and to provide testing guidance, including a prioritybasedranking scheme for food additives that ranks compounds according to their level of healthconcern (based on the extent of human exposure and the compound’s known or presumed toxicologicaleffects). The 1982 guidelines provided general toxicity guidelines and promulgated individualtests to be used for determining a compound’s potential for causing specific types of toxicity,such as reproductive and teratological effects. Tests for reproductive and teratological effects hadbeen available before this time, but this was the first time they were issued together with other testsas part of a set of guidelines. Further expansion and refinement of the tests was done in the 1993Draft Redbook II. 4 As the individual guidelines are reviewed and updated, they are made availableonline on FDA’s Web site. 5 The testing guidelines for female reproduction and teratology studiesare posted online (http://www.cfsan.fda.gov/~redbook/redivc9a.html for reproduction studies andhttp://www.cfsan.fda.gov/~redbook/redivc9a.html for developmental toxicity studies) and arereadily available, so they will not be repeated here.2. Estimating Exposure to a Direct Food AdditiveSection 409 of the act describes the statutory requirements that are required to establish safety ofa food additive. The chemical and technological data encompass four general areas of information:(1) identity of the additive, (2) proposed use of the additive, (3) intended technical effect of theadditive, and (4) a method of analysis for the additive in food. Section 409 also requires thatprobable consumption of an additive and of any substance formed in or on food because of its use© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 881be considered in determining whether the proposed use is safe. An estimate of probable humanexposure to food additives or any of its by-products, resulting from the ingestion of food containingthe substance, is most often referred to as an estimated daily intake (EDI). A cumulative estimateddaily intake (CEDI) is calculated if a new use of a regulated additive results in an increase inconsumption. A detailed evaluation of the consumer exposure to direct food additives is posted atFDA’s Web site in the FDA Guidance for Industry Recommendations for Submissions of Chemicaland Technological Data for Direct Food Additives and GRAS Food Ingredient Petitions. 63. Safety Testing of a Food or Color AdditiveUnder the Food Drug and Cosmetic Act, the safety of food additives must be established prior tomarketing by evaluating the probable exposure to the new substance and appropriate toxicologicaland other scientific information. Thus, the approval of any new food additive used in food dependsin part upon the outcome of toxicity tests that are performed and evaluated prior to marketing.Through the years, the FDA has and continues to adjust testing recommendations for food andcolor additives used in food as necessary to reflect the steady progress of science and currentinformation about population exposure. 7 Protocols for safety testing reflect the increased sophisticationof the science of toxicology as well as an increased concern about food safety.a. Direct Food Additives and Color AdditivesThe FDA’s Toxicological Principles for the Assessment of Direct Food Additives and Color AdditivesUsed in Food 3 recommended a series of individual tests to be used to determine a compound’spotential for causing specific types of toxicity. The document sets out specific principles to determinethe likelihood that the test substance poses a potential for health risk to the public. Thedocument proposed that the development of information needed for the safety assessment processshould be done in a tiered procedure whereby the small number of additives with the highestprobable risk would receive the greatest amount of scrutiny and the large number of substanceswith minimal potential toxicity will receive a less comprehensive and detailed evaluation. Thisconcept was referred to as the “concept of concern,” and the safety evaluation was determined bythe “levels of concern.” This tiered approach to testing has been the subject of several reviews. 1,8-10Three concern levels are generally used, based on molecular structures combined with exposureinformation for the entire population. As the duration of exposure increases, the types of effectsare determined with greater sensitivity. Tests for Concern Level 1 are short-term tests and are the leastsensitive. Tests for Concern Level 2 are of intermediate sensitivity, and consist of subchronic feedingstudies, a reproduction study with a teratology phase, and short-term tests for carcinogenicity. Testsfor Concern Level 3 are the most demanding. In addition to chronic studies, a reproduction study witha teratology phase is required, with the possibility of testing additional generations.Expansion and refinement of the tests was done in the 1993 Draft Redbook II. 4 The FDAcontinues to propose the use of Concern Levels as the basis for determining minimum toxicologytesting recommendations. The tests for Concern Levels 2 and 3 still require multigenerationreproduction studies, each with a teratology phase. As the individual guidelines are reviewed andupdated, they are made available on FDA’s Web site.b. Food Contact SubstancesSection 309 of the Food and Drug Administration Modernization Act of 1997 11 amended Section409 of the FFDCA to establish a notification procedure as the primary method by which the FDAregulates food additives that are food contact substances. The most recent FDA Guidance forIndustry is Preparation of Food Contact Notifications for Food Contact Substances: ToxicologyRecommendations. 12 These recommendations are consistent with the general principle that the© 2006 by Taylor & Francis Group, LLC

882 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONpotential risk of a substance is likely to increase as exposure increases. Food contact substanceswith a dietary concentration at or greater than 0.5 ppb are now regulated under the food contactnotification process for food contact substances. 13Because the safety standard is the same for all food additives, the chemical and technologicalinformation needed for food contact substances is comparable to that recommended for a directfood additive. The probable consumer exposure may include the food contact substance itself andthe degradation products and impurities in the food contact substance. The cumulative estimateddaily intake from all regulated uses and effective food contact substance notifications are used bythe FDA in the safety evaluation of a food contact substance. The most recent information on theestimation of consumer exposure to food contact substances is posted at FDA’s Web site in FDAGuidance for Industry, Preparation of Premarket Notifications for Food Contact Substances: ChemistryRecommendations. 14The FDA Guidance for Industry recommends safety testing to assess the safety of the foodcontact substance and/or various constituents of the food contact substance. If the probable humandietary exposure for a single use is at or less than 0.5 ppb (1.5 µg/person/day), then no new safetystudies are recommended. If the cumulative human dietary exposure is greater than 0.5 ppb (1.5µg/person/day) but not exceeding 50 ppb (150 µg/person/day), then a battery of genotoxicity testsis recommended to evaluate the potential for carcinogenicity of the food contact substance and/orconstituent. If the cumulative human dietary exposure is between 50 ppb (150 µg/person/day) and1 ppm (3 mg/person/day), then the potential toxicity of the food contact substance and/or constituentshould be evaluated in a battery of genotoxicity studies as well as subchronic oral toxicity studiesin rodents and nonrodents. These studies are used to determine if longer term or specialized toxicitytests are needed to assess safety. Additional specialized toxicity tests include metabolism studies,teratology studies, reproductive toxicity studies, neurotoxicity studies, and immunotoxicity studies.Food contact substances that are used primarily for their antimicrobial or fungicidal effects aretoxic by design. The minimum safety testing recommendation is one-fifth the value of the cumulativeestimated daily intake (CEDI) used to determine the appropriate safety testing for other typesof food contact substances. 15c. Determination of the ADIThe FDA’s 1982 guidelines 3 defined a tiered system of tests to be conducted to determine the effectsof food additives. The goal of the testing is to assess the substance’s potential for causing toxicityand determine if safe conditions of use can be established, provided the additive serves its intendedfunction. Not only do the guidelines provide testing guidance, they also set out a priority-basedranking scheme for additives in food that ranks compounds according to their level of healthconcern. Health concern is, in turn, based on the extent of human exposure and the assessed orpresumptive toxicological effects.A fundamental principle used in setting tolerances (maximum acceptable limits of intake) forchemicals in food is that toxicity is a result of dose response and that there is no toxic effect atsome low dose. In ensuring safe conditions for use, the FDA must consider whether a tolerancelimit for a food additive should be specified by regulation. For this, the likely level of consumerexposure to an additive (from the estimated daily intake) and the maximum level that can beconsidered safe or noninjurious to health (from the maximum acceptable daily intake, or ADI) mustbe compared. The EDI is the likely level of consumer exposure to an additive from the daily useof a food containing it. The ADI is the daily intake of a chemical that, during an entire lifetime,appears to be without appreciable risk on the basis of all the known facts at the time. Thedefinition refers to average lifetime exposure, and thus it can be inferred that slightly exceedingthe ADI for a period may be compensated for by periods when consumption is below the ADI. 16The ADI is an estimate that depends on a number of statistically calculated levels, some of which© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 883may be quite primitive. It is a yardstick that should be used with care to set permitted levels foradditives in food.Both chemical and biological data are needed to establish an ADI, and all the toxicological dataare taken into account in deriving an ADI. The chemical information helps to determine the relevanceand importance of a study in estimating human risk. Chemical information and data required includeidentity and purity of the chemical, impurities, stability, metabolic products, products formed incombination with other food components, absorption, distribution, metabolism, etc.In deriving an ADI, the no effect level (NOEL) is used as determined in the most susceptiblebut appropriate animal and with the most sensitive indicator of toxicity. Determination of theappropriateness of a species requires in-depth knowledge of comparative physiology, biochemistry,and metabolism in laboratory animals and humans. Acute toxicity studies and short-term toxicitystudies are valuable, but the most sensitive indicator has historically been found in long-term studies.For substances that could affect pregnant women, special studies should be conducted, includingtests for reproductive function and developmental toxicity.If the probable level of consumer exposure to the additive (i.e., the EDI) is determined to bewell below the maximum safe level of consumption (i.e., the ADI) and typical use levels are unlikelyto be exceeded, the use of the additive could be considered self-limiting and would not require aquantitative tolerance. Use of the additive under conditions of good manufacturing practice isconsidered acceptable where there is a wide margin of safety (i.e., the EDI is well within the ADI)and the use is self-limiting. If, however, the exposure may exceed safe limits under certain conditionsof use, a quantitative tolerance may be required. The tolerance may be set to limit the amount ofadditive to be used in all foods, in a range of foods, in specific foods, or foods in certain foodcategories. 1d. Interpretation and Evaluation of Data from Developmental andReproductive Toxicity TestsThere are difficulties in interpreting animal studies, in defining a strict cutoff point between whatcan be regarded as adverse and extrapolating from this point to give acceptable levels for humans,as the regulatory process requires. To allow for these uncertainties, a safety factor is used toextrapolate from the no-effect level obtained in animal studies to an ADI. The safety factor isintended to allow for differences in sensitivity between the animal species and humans, variationsin susceptibility among the human population, and the fact that the number of animals tested issmall compared with the size of the human population that may be exposed. A safety factor of 100was proposed by WHO in 1958, 17 and this has been accepted as adequate by the FDA for somesubstances. The safety factor is influenced by the nature of the observed toxicity and the adequacyof the data. Larger factors are used to compensate for slight deficiencies in toxicity data, such assmall numbers of animals. The safety factor of 1000, as proposed by Jackson, 18 linked dosage withthe severity of anomalies seen in developmental toxicity studies.The interpretation of data from developmental toxicity studies is similar to the interpretationof data from other toxicity tests in that the animal tests must be well designed and properlyconducted. After the data have been collected, it must be determined whether any observed differencesbetween treated and control groups are real, whether they can be reproduced, and whetherthey can be attributed to treatment with the test chemical. 18Observed differences that are spontaneous must be distinguished from those that are inducedby exposure to the test chemical. Spontaneous changes may result from hereditary factors, disease,or environmental factors. If proper randomization has been done, these changes are as likely tooccur with the same frequency in treated and untreated animals. The use of concurrent experimentalcontrols provides the primary evidence of whether a particular change can be regarded as treatmentrelated. Colony control data and strain data collected from other laboratories can be useful when© 2006 by Taylor & Francis Group, LLC

884 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 22.1Types of adverse developmental effectsTypes Changes Developmental Toxicity ExamplesIIIPermanent (irreversible), possibly lifethreateningchanges, frequentlyassociated with gross malformationsNonpermanent (reversible) changes,not life-threatening, usually notassociated with gross malformationsSource: Jackson, B.A., 1988, p. 207.Decreased number live fetuses (litter size)Increased number resorptionsIncreased number fetuses with malformations, structuralchangesIncreased number fetuses with retarded developmentof skeletal or soft-tissuesevaluating the significance of the occurrence of low incidence or rare malformations. Statisticalanalyses are employed for identifying changes that are unlikely to be chance occurrences. 18The endpoints observed in the animal test system represent various manifestations of outcomesof altered development. These endpoints fit three general and well-known responses: (1) alteredrate of development, (2) abnormal development, and (3) death.In the past, attention was focused on the occurrence of developmental malformations. Now,neurological, immunological, and carcinogenic effects are recognized as possible manifestationsof developmental toxicity. In reproduction studies, it is sometimes more difficult to determinewhether intrauterine or postnatal lethality was a result of a lethal malformation or a manifestationof maternal or fetal toxicity. Examination of dead fetuses for structural changes or malformationsin many cases reveals only partial information regarding whether malformations preceded oraccompanied death.A multigeneration test with two generations and one litter per generation, as currently recommendedby FDA, can help evaluate the adverse effects of agents on the reproductive systems ofboth males and females, as well as postnatal maturation and reproductive capacity of the offspring.Also, and perhaps most importantly, the cumulative effects of a substance through several generationscan be evaluated.Of the multitude of developmental and reproductive effects, some changes are more seriousthan others. Jackson 18 proposed classifying observations into two types, according to severity andreversibility (Table 22.1). Type I changes are perceived as permanent and irreversible, possibly lifethreatening, and frequently associated with gross malformations. Type II changes are perceived asnonpermanent and reversible, nonlife threatening, and not necessarily associated with inducedmalformations. Type I changes include those associated with death and abnormal development,such as decreased litter size, increased resorptions, and gross malformations. Type II changesinclude retarded skeletal and soft-tissue development, decreased fetal weight, and decreased postnatalgrowth. In using the latter category, emphasis should be placed on acquiring sufficientinformation to rule out the possibility that the test chemical presents a special and unique hazardto the developing organism. 18A stepwise evaluation of developmental toxic effects should be done. Any combination of TypeI and Type II changes would result in the provisional classification of the test chemical as a TypeI compound. Disparities between the outcomes of two test systems might indicate metabolic,mechanistic, or other differences that could be used in clarifying its hazard to humans. 18Appropriate procedures are necessary to translate the information on developmental toxicityand reproductive effects into regulatory action. Such procedures must consider all the informationon a particular chemical, including the results from other toxicological evaluations. For example,if metabolic data indicate that humans detoxify a compound more rapidly than the test species thatexhibited an adverse effect, then this can be reflected in the safety assessment.Any attempt to predict the conditions under which a compound can be safely used brings sharplyinto focus those areas in which our knowledge is less well established. Two such areas can beidentified, and they merit some discussion here. One area that has been much studied but whichstill contains deficiencies is how to separate the range of abnormalities that are manifestations of© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 885interference with development. In the Type I and Type II classification scheme proposed above,the two extremes of the range of deficiencies are considered. Between the extremes, a determinationof safety can present difficulties. One possible link is the dose response relationship. Materials thatinduce frank malformations can be expected to induce other effects at the same and lower doses.Another problem area is whether a distinction can be made between substances that producemalformations at dosages that are toxic to the mother and those substances that produce thesechanges at doses that are not maternally toxic. The classic example of this phenomenon is thalidomide’sadverse effects on the fetus at levels that are not maternally toxic.Safety factors applied to the results of animal toxicity studies are employed to establish theacceptable exposure levels for the use of compounds by humans. A 1000-fold safety factor wouldbe applied in those cases where there was evidence of serious, irreversible developmental effects(Type I effects), and a 100-fold safety factor would be applied to a Type II hazard. In those instanceswhere the results in two test systems are discordant, consideration could be given to informationon maternal toxicity, metabolic differences, and mechanism as mitigating factors.Two-generation reproduction studies with a developmental toxicity phase are used as the basicscreening tool for identifying compounds with the potential for producing adverse developmentaleffects. Testing needs are identified in a stepwise fashion from the results of the initial screeningand information such as exposure and relation to known developmental toxicities. The combinationof endpoints is used to maximize test sensitivity.For developmental and reproductive effects, the primary measures of toxicity are generally theproportion of dead and resorbed implants per litter and the proportion of various types of malformationsamong the live fetuses in a litter. The variability of these proportions among litters withindose groups must be considered in testing for toxicity. Litter size and fetal weight are examinedas well and may be used as covariates or intermediate variables for dose response relationships toestimate the risk. 19Measurements of biological effects, such as body weight, organ weight, hormone concentrations,enzyme concentrations, etc., are measured on a continuous scale. For a biological effect measured ona continuous scale, there generally is not a sharp point of demarcation between a normal and an adverseresponse. For such cases, a normal range can be established in unexposed control animals or individuals,and the risk can be based upon the magnitude of the abnormal response.e. Interpretation and Evaluation of Data from Male Reproductive Toxicity TestsBetween the FDA guidelines published in 1982 and those published in1993, a major revisioninvolved the increased emphasis on reproductive effects in males. Compared with other animalspecies, the human male is relatively infertile and therefore is at greater reproductive risk fromexposure to various toxicants than laboratory models commonly used in testing protocols. Unfortunately,in the past, the ability of a compound to induce damage to the male reproductive systemwas assessed by examining fertility. It is now recognized that fertility, which measures the endresult of the reproductive process, is not a sensitive predictor of male reproductive toxicity in therodent. A comprehensive assessment of male reproductive toxicity, which requires information frommultiple endpoints, has been included in testing guidelines from national and international organizations(EPA, OECD, ICH, etc.).Data obtained from male reproductive toxicity studies must be evaluated in light of the interspeciesvariation between the human and the test species. At present, the extrapolation of animaldata to the human is troublesome, especially with respect to the male because of a lack of knowledgeof the physiological variation between the human male and males from various test species.Therefore, it is very important that regulatory animal testing protocols for regulatory purposes bewell designed and properly conducted. Testing protocols utilized in the regulatory process shouldbe designed to simulate the factors that affect the test results for reproductive risk assessment (i.e.,route of exposure, potential for bioaccumulation, and the comparability of the test species to the© 2006 by Taylor & Francis Group, LLC

886 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONhuman) for the compound in question. Several important parameters that must be considered indesigning a study to test for effects on the male reproductive system are the duration of exposure,appropriate routes of exposure, level of exposure associated with adverse risks in humans, choiceof the test species, and the endpoints selected for study. Other parameters that should be consideredinclude appropriate fixatives for the preservation of reproductive tissues and the most appropriatemethods to detect changes in the selected endpoints.The endpoints commonly observed in male reproductive toxicity testing include animal bodyweight, fertility, reproductive organ weights, sperm quality (morphology and motility), spermquantity (cauda epididymal sperm numbers and testicular homogenization-resistant spermatid numbers),histopathological observations on reproductive organs (and in particular the testis), anogenitaldistance, and when necessary, reproductive endocrine profiles. Many of the endpoints obtainedfrom these studies are interrelated. The strength of collecting this type of data lies in the fact thatdata obtained on one endpoint can help corroborate effects observed on other endpoints. It istherefore suggested that a weight-of-evidence approach be taken when interpreting male reproductivetoxicity data obtained in regulatory testing protocols.The endpoints observed in the animal test system represent various manifestations of outcomesof altered sperm production, sperm quality, and sperm quantity. A multitude of reproductive effectscan be observed from exposure to a toxicant, but the effects are unequal. Effects observed shouldnot only be classified as to their severity but also as to whether they are reversible. For example,compounds that destroy stem cells will have a long-term irreversible effect on the testis, whereasanother compound may have only a short-term effect on a different cell population.The following presents a brief discussion of some endpoints of male reproductive toxicitycommonly collected and their use in the risk assessment process. Among the endpoints of malereproductive toxicity commonly assessed, two endpoints have received the most attention in thelast 10 years: evaluation of testicular histopathology and assessment of sperm motility. Theseendpoints are discussed first and are followed by a discussion of other selected endpoints.1) Testicular Histopathology — Careful histopathological examination of the testis is recognizedas a sensitive method to identify effects on spermatogenesis. The sensitivity of this methodcan be increased or decreased depending on the quality of the tissue fixation and the embedmentof the testicular tissues. In the past, histopathological evaluation of the testis identified only majorchanges between control and treated groups. Subtle lesions (retained spermatids, missing germ celllayers) may not have been observed because the tissue preservation methodology was substandardand pathologists were not familiar with testicular morphology and the kinetics of spermatogenesisof the test species examined. Testicular morphological changes can be used to identify injury thatis not severe enough to cause a reduction in fertility or sperm production in the rat. Morphologicalchanges that occur in the absence of functional changes (i.e., libido, sperm motility, fertility) shouldbe interpreted as an indication of the potential for the compound to affect spermatogenesis adverselyin the human. Although histopathological changes may fail to reveal treatment related effects,consideration should be given to the possible presence of other testicular or epididymal effects notdetected histologically that may affect reproductive function.Excellent articles and books have been written describing methods that can be used to preservetesticular tissue as well as to evaluate testicular histopathology. 20,21 These resources provide variousapproaches for the qualitative and/or quantitative assessment of testicular lesions. Included in thebooks are detailed descriptions of the various stages of the cycle of the seminiferous epitheliumand extensive information on tissue preparation and examination, and interpretation of observationsfrom normal microscopic and ultrastructural histology of many experimental animals.The histological picture seen after toxic insult may provide insight into the site and/or mechanismof action of the toxicant on a particular reproductive organ. If similar targets or mechanismsexist in humans and other test species, the data can be utilized for extrapolation between speciesand may allow a prediction of the extent of damage and the chances for recovery. 22© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 8872) Sperm Motility — Sperm motility may be assessed either from the vas deferens or the caudaepididymis (rat and mouse) or from an ejaculate (rabbit and dog). In rats and mice, placing the vasor cauda epididymis in an appropriate medium and allowing sperm to swim out is a popular methodfor collecting sperm for analysis. 23 Sperm motion may be captured on videotape or digitallyformatted and analyzed at a later date, either manually or by computer-assisted sperm analysis(CASA) systems.Although there is no standard for the percentage of motile sperm, the general consensus is thatin a control sample the number of “progressively” motile sperm should be at least 70%. 24,25 Visualestimates of counts of moving sperm are straightforward. CASA systems have given us the abilityto quantify a number of sperm motion parameters. For the CASA systems, a variety of parametersmust be defined correctly for accurate image acquisition and analysis. It is important to rememberthat the information obtained from a CASA system is dependent upon the instrument settings. Twodifferent laboratories can use the same instrument but use different settings and obtain differentresults for various motion parameters. Therefore, data obtained by these instruments cannot becompared across laboratories, and all values obtained should be compared to concurrent controls,not historical controls.The use of the CASA system allows for the rapid and efficient collection of motility information,including velocity (straight-line and curvilinear) measurements, beat cross-frequency, and amplitudeof lateral head displacement measurements, on a large number of sperm. Recently, using CASAsystems, compound-induced changes in motility have been detected 26-29 and associated with changesin fertility 30-32 in experimental animals. Preliminary results have suggested that significant reductionsin sperm velocity are associated with reduced fertility, even when the percentage of motile spermis not affected. It is important to distinguish motility from progressive motility. Sperm may bemotile but not demonstrate a directed forward movement. A change in motility may be independentof changes in testicular histopathology because the agent may adversely affect epididymal function.Motility is also influenced by abstinence, the time between obtaining the sample and the evaluationof the sample, medium pH, sample chamber depth, and temperature. The investigator should becareful to avoid artifactual cell death during specimen preparation so that the percentage of motilesperm from control animals is consistently high. A reduction in the velocity or pattern of motilityof treated animals when compared with controls may indicate a need for further testing. Of all thesperm motion parameters, the percent motility is clearly related to fertility. Therefore, any changein the percent motility should be considered an adverse effect, especially if it is dose related.3) Other Selected Endpoints of Male Reproductive Toxicitya) Organ Weights — Other endpoints commonly evaluated are reproductive organ weights andhistopathological effects observed on male secondary sex organs. The male reproductive organsfor which weights may be useful for reproductive risk assessment include the testis, epididymis,pituitary, seminal vesicle with coagulating gland, and the prostate. Reproductive organ weightchanges may occur irrespective of a decrease in body weight; however, it should be rememberedthat changes in other endpoints related to reproductive function (i.e., motility) might be alteredwithout a concomitant change in reproductive organ weight. Organ weights should be reported asabsolute weights or relative weights (weights per gram of body weight; weights per gram of brainweight). 33 Care should be taken to ensure that deleterious reproductive effects are not maskedbecause the organ weight and body weight are both reduced by treatment. Testicular weights arerelatively stable and vary only modestly within a given test species. 34,35 Increases or decreases intestis weight indicate that a compound has significantly altered testis structure but do not help tocharacterize the effect.Depending on the treatment period, not all testicular effects will result in a decrease or anincrease in testicular weight. For example, short treatment periods may produce increased germcell degeneration that may not immediately be reflected in testicular weights. Additionally, blockage© 2006 by Taylor & Francis Group, LLC

888 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof the efferent ducts by cells sloughed from the seminiferous epithelium may result in increasedtestis weights, or slight increases in testis weight due to increased intratesticular fluid might offsetthe weight change induced by germ cell depletion. 36,37 Testicular effects may also be observed atdose levels higher than those required to affect other measures of gonadal status. 38-40 Although achange in testicular weight may not reflect certain adverse testicular effects and does not indicatethe nature of an effect, a significant increase or decrease is indicative of an effect.In conclusion, statistically or biologically significant changes in absolute or relative malereproductive organ weights may indicate an adverse reproductive effect and provide a basis forobtaining additional information on toxicity of an agent. It is important to note that significantalterations in other endpoints related to reproductive function may not be reflected in organ weightdata. Therefore, lack of a change in organ weight does not indicate a lack of effect in otherreproductive endpoints.b) Histopathology — The histopathological evaluation of the testis and other reproductive organs(epididymis, prostate, seminal vesicles with coagulating gland, and pituitary) plays a very importantrole in reproductive risk assessment. Histopathological observation can provide information on thetoxicity of the compound, the site and extent of toxicity, and the potential for recovery. It can beuseful as a tool for identifying low-dose effects. Significant histopathological damage in excess ofthe level seen in the control tissue of any male reproductive organ should be considered an adversereproductive effect. Of paramount importance in any histological examination is the proper fixationand embedding of the tissue. Appropriate fixatives and embedding materials that minimize theextent of fixation artifact should be utilized.c) Sperm Morphology — Sperm morphology profiles are relatively stable and characteristic ina normal individual over time. Sperm morphology is one of the least variable sperm measures innormal individuals, which may enhance its use in the detection of spermatotoxic events. 41 Duringspermiogenesis there is a progressive condensation of chromosomal material in the spermatid duringsuccessive stages of differentiation, and this results in the tight packaging of DNA in the heads ofthe spermatozoa. Although temporal changes in nuclear basic proteins are involved, the chemicalbasis for changes in chromatin packing is not well understood. In eutherian mammals, somatichistones present in cells at early stages of spermatogenesis are replaced entirely by protamine. Inthe rat it has been shown that three histonelike proteins first appear in primary spermatocytes butare not retained in late spermatids and that new proteins are synthesized just prior to chromatincondensations in elongating spermatid nuclei. Thus, chromatin condensation is a highly complexprocess that involves the appearance and disappearance of various nucleoproteins throughout thecondensation process. Therefore, changes in sperm head morphology indicate that the agent has insome way affected spermatogenesis and that the quality of the sperm produced has been in someway affected. A number of studies in test species and humans have suggested that abnormallyshaped sperm may not reach the oviduct or participate in fertilization, 42,43 suggesting that the greaterthe number of abnormal sperm in the ejaculate, the greater the probability of reduced fertility. Careshould be taken in evaluating abnormal sperm morphology because altered nuclear morphologyshould not be interpreted as indicating genetic damage, that is, not all dysmorphogens are mutagens.Traditionally the characterization of abnormalities in sperm morphology is based on the appearanceof the sperm head, midpiece, and tail. The sperm may be visualized as unfixed samples bybright field microscopy, 44 or fixed, unstained preparations can be examined by phase contrastmicroscopy. 24,45 This approach is useful in species in which the sperm population is homogeneous(i.e., rats and mice); however, in species in which sperm morphology is heterogeneous (human),sperm morphological alterations are difficult to assess. Data that categorize the types of abnormalitiesobserved and quantify the frequencies of their occurrences are preferred to estimation of theoverall proportion of abnormal sperm. Objective, quantitative approaches that are done properlyshould result in a higher level of confidence than more subjective measures. Recently CASA systems© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 889have been developed to assess sperm morphological alterations in both the head and tail for humansand other test species.At the present time it is not clear how alterations in sperm morphology influence fertility. Spermshape abnormalities have been correlated with poor fertility in humans and laboratory species.Therefore, because the events underlying the process of nuclear condensation in all species examinedto date are similar, it is believed that changes in sperm morphology in test species can bedirectly related to the same endpoint in humans. An increase in abnormal sperm morphology hasbeen considered evidence that the agent has gained access to the germ cells. 46 The implication isthat the greater the number of abnormal sperm in the ejaculate, the greater the probability of reducedfertility. Changes in sperm nuclear morphology should be considered as an adverse reproductiveeffect.d) Sperm and Spermatid Numbers — One frequently reported semen variable for the humanis sperm concentration. 47 The quantitation of sperm numbers can be assessed from either theepididymis (cauda epdidymal sperm counts), the testis (homogenization resistant spermatid numbers),or the ejaculate.Epididymal counts reflect both the production of sperm by the testis and the ability of theepididymis to store sperm. Epididymal sperm counts are directly related to fertility, and anystatistical reduction in epididymal sperm counts should be considered adverse. The ability of thecauda epididymis to store sperm is assessed by counting cauda epididymal sperm, which arecommonly used for both sperm morphology and motility assessments. Epididymal sperm numberscan be determined by performing counts on the sperm found in the suspension used to evaluatesperm motility and sperm morphology or the sperm obtained by the subsequent mincing and/orhomogenization of the remaining cauda tissue. Epididymal sperm numbers can be counted usingeither a hemocytometer or CASA systems. Typically, epididymal sperm counts are expressed inrelation to the weight of the cauda epididymis; however, this can be misleading because spermcontribute to the epididymal weight and the expression of the data as a ratio may mask a reductionin sperm numbers. Therefore, absolute cauda epididymal sperm numbers should be included withthe data obtained from toxicity studies. It should be remembered that cauda epididymal spermcounts are influenced by the level of sexual activity. 48,49The quantitation of testicular spermatid numbers is a rapid, inexpensive method commonlyused to evaluate the production of sperm by stem cells and to evaluate the ability of the developinggerminal cells to survive through the proliferative, meiotic, and spermiogenic phases of spermatogenesis.The quantitation of testicular spermatid numbers will identify agents that have a toxiceffect on spermatogenesis but will not provide information on the mechanism of action or thespecific cell types sensitive to the agent. The enumeration of spermatid numbers should primarilybe used in chronic studies where spermatid numbers have stabilized, but not in short term studieswhere treatment may not have affected the late spermatid population. If the evaluation was conductedwhen the effect of a lesion would be reflected adequately in the spermatid counts, thenspermatid count may serve as a substitute for quantitative histologic analysis of sperm production. 21However, spermatid counts may be misleading when the duration of exposure is shorter than thetime required for a lesion to be fully expressed in the count. A reduction in homogenization resistantspermatid head counts does not indicate that the agent is toxic to the germ cells. The agent mayhave adversely affected the Sertoli cell, Leydig cell, or other elements such as the hypothalamopituitary-gonadalaxis, and therefore a reduction in homogenization resistant spermatid countswould be a secondary response. Testicular spermatid counts may rise if spermatids are not beingreleased from the Sertoli cell. Small increases in spermatid retention will not be detected in thehomogenization resistant spermatid counts or in the epididymal sperm counts but should be pickedup in the histopathological evaluation of the testis. A total inhibition of sperm release will severelyreduce epididymal sperm counts. Remember that a normal sperm count does not provide informationon motility, fertilizing potential, or the genetic integrity of the cells that are important for© 2006 by Taylor & Francis Group, LLC

890 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONfertility. Homogenization resistant spermatid count data may be reported as numbers of homogenizationresistant spermatids per testis, efficiency of sperm production, or daily sperm productionrates.Sperm count determinations from ejaculates in dogs, rabbits, and other species are affected bythe frequency of collection. Therefore, care should be taken to collect these counts at appropriateintervals, to ensure that the same male is used, and to ensure that the entire ejaculate is collected.Baseline information from these species (humans and test species) is also valuable, as changesduring exposure and recovery can be better defined.In summary, sperm quantity and quality are both related to fertility. Although the preciserelationship between sperm quantity and fertility may not be the same for all species, there isnevertheless a distinct relationship. Fertility can be affected if sperm count is significantly reducedin the presence of normal sperm quality. 30,32e) Fertility — Fertility can be affected if sperm quality is significantly reduced in the presence ofadequate numbers of sperm. 50 Sperm shape has also been linked to infertility in the human and allspecies of laboratory animals examined. Because morphological effects are also observed in thepresence of decreased sperm count and decreased motility, it is not easy to separate effects ofabnormal sperm morphology from effects related to decreased sperm quantity or quality. Nevertheless,the underlying processes for nuclear condensation are similar in both test species and thehuman, and therefore data on sperm morphology are important in the risk assessment process.Finally, fertility can be impaired in the presence of small decreases in sperm quality and quantity.It is important to remember that small changes in either sperm quantity or quality may not impairfertility but can be predictive of infertility in the human.A. PerspectiveIII. VACCINE SAFETYThe FDA’s Center for Biologics Evaluation and Research (CBER) reviews a broad spectrum ofinvestigational vaccines. For the purpose of the following discussion a vaccine is a product, theadministration of which is intended to elicit an immune response that can prevent and/or lessenthe severity of one or more infectious diseases. A vaccine may be a live, attenuated preparation ofbacteria, viruses, or parasites, inactivated (killed) whole organisms, living irradiated cells, crudefractions, or purified immunogens, including those derived from recombinant DNA in a host cell,conjugates formed by covalent linkage of components, synthetic antigens, polynucleotides (suchas plasmid DNA vaccines), living vectored cells expressing specific heterologous immunogens, orcells pulsed with immunogen. It may also be a combination of vaccines listed above. Therapeuticvaccines, e.g., viral vector–based gene therapy and tumor vaccines, are not considered here.Although advances in analytical technologies have improved biological product characterization,these products are usually composed of, or extracted from, living organisms and tend to bemore complex than conventional drug products. Therefore, for many biotherapeutics and vaccines,unique issues arise that frequently require modifications to standard preclinical testing requirementsand designs that are used to evaluate drugs.There are numerous vaccines in clinical development for the prevention of infectious diseasesin adolescents and adults. Thus, the target population for vaccines often includes females in theirreproductive years. In addition, there are a number of vaccines in clinical development specificallyintended for maternal immunization, with the goal of preventing infectious disease in the pregnantwoman and/or young infant through passive antibody transfer from mother to fetus. There arespecial considerations, such as the potential for adverse effects of the vaccine on normal fetal© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 891development, that need to be employed in the assessment of the risks versus the benefits ofimmunization programs for pregnant women and/or females of childbearing potential.Currently, unless the vaccine is specifically indicated for maternal immunization, no data arecollected regarding the vaccine’s safety in pregnant women during the prelicensure phase. Thus, inmost cases, preclinical reproductive toxicity studies provide the only data source upon which to baseestimations of risk to the developing fetus. However, there is a scarcity of scientific literature on animalreproductive toxicity testing for preventive vaccines. The following outlines considerations for reproductivetoxicity study designs for preventive vaccines. It should be noted that approaches to reproductivetoxicity assessment for these products may evolve as new methods and test systems become available.B. General ConsiderationsIf the vaccine is indicated for females of childbearing potential, subjects may be included in clinicaltrials without reproductive toxicity studies, provided appropriate precautions are taken to excludepregnant women by the use of pregnancy testing and appropriate contraceptive methods. However,reproductive toxicity studies should be included in the initial Biologics License Application (BLA)when the vaccine is indicated for or may have the potential to be indicated for females of childbearingpotential. If the vaccine is indicated for maternal immunization and if pregnant women areenrolled in a clinical trial, data from preclinical reproductive toxicity studies should be availableprior to the initiation of the clinical trial.The International Conference on Harmonization (ICH) has published Detection of Toxicity toReproduction for Medicinal Products (ICHS5a), which should be consulted as a point of referenceto assist in the design of reproductive toxicity studies of a preventive vaccine. 51 However, aspreventive vaccines present a diverse class of biologics, product specific issues frequently arise thatmay necessitate modifications to standard study designs. It is, therefore, recommended that thevaccine manufacturer establish an early dialogue with CBER to reach agreement on specific designissues and study endpoints prior to the conduct of such studies.All data generated from prior preclinical toxicity studies should be reviewed for their possiblecontribution to the interpretation of any adverse developmental effects that appear in the reproductivetoxicology studies, e.g., fetal toxicity secondary to maternal toxicity. 52 In addition, availableclinical experience in pregnant women should be considered for any potential application to thedesign of reproductive toxicity studies in animals. Clinical experience derived from immunizationof pregnant women may be helpful in the evaluation of the potential for any adverse outcome onthe viability and development of offspring.C. Specific Considerations1. Animal ModelsAppropriate rationales should be provided in the selection of a relevant animal species for developmentaltoxicity testing for preventive vaccines. The primary criterion used to define a relevantanimal model is the ability of the vaccine to elicit an adequate immune response in the animalmodel. Species specific differences in the immune response and frequent lack of availability of a“standard species” model can be major obstacles that are encountered. It is recognized that responsesinduced in a particular animal species may not always be predictive of the expected human response.However, in the absence of alternative test methods, animal models remain the best option in theattempt to predict toxicities for humans. It is important to realize that such studies do not necessarilyneed to be conducted in the traditional species, i.e., rat and rabbit, nor is the use of two speciesmandatory. It is the responsibility of the sponsor to provide a scientific rationale for either theanimal model used or the lack thereof.© 2006 by Taylor & Francis Group, LLC

892 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION2. DoseTraditional approaches, i.e., the selection of low, intermediate, and high doses to identify potentialdose response relationships with respect to adverse effects may not always be applicable to thereproductive toxicity studies for preventive vaccines because immune-based reactions frequentlydo not follow dose response relationships. The dose level(s) selected should elicit an immuneresponse in the animal model and, when feasible, should include a full human dose equivalent (e.g.,1 human dose equals 1 rabbit dose). Injection volume is one consideration for feasibility. Datafrom other available preclinical toxicology studies may guide in the selection of the doses for thereproductive toxicity studies. Control animals should be dosed with the vehicle at the same frequencyas test group animals.3. Immunization IntervalThe immunization interval and frequency of immunization(s) should be based on the clinicallyproposed immunization interval. Similarly, the route or routes of administration should mimic theintended clinical routes, if possible. Since preventive vaccines indicated for females of childbearingpotential or pregnant females are administered infrequently, episodic dosing of pregnant animalsis likely to be more relevant than daily dosing. Thus, in order to mimic potential exposure in theclinical situation, several groups of pregnant animals should be dosed only at defined intervals duringsensitive periods of organogenesis. Modifications to the timing and frequency of dosing may benecessary, depending on the kinetics of the antibody response induced in the animal species. Thus, inaddition to immunizing the female animal during gestation, it may be necessary to also administer apriming dose prior to conception to allow for an immune response to occur during gestation.4. Exposure Period and Follow-Up PeriodOf primary concern are potential adverse effects of the vaccine on embryo-fetal development, asunintentional clinical exposure to the vaccine will likely occur during early stages of pregnancy.Thus, the pregnant animal should be exposed to the vaccine during the period of organogenesis,i.e., from implantation to birth, defined as stages C to D of the ICHS5A document. In addition,some females should be allowed to deliver and the offspring should be followed until weaning toevaluate the health of the neonates as well as look for evidence of maternal antibody transfer. Dataon reproductive performance (i.e., fertility for males and females) are usually not necessary.Additional data regarding the integrity of the reproductive organs for both sexes can be obtainedfrom histopathology data generated in the toxicology studies in nonpregnant animals.5. Study EndpointsThe parameters and endpoints for assessing the potential reproductive toxicity of a preventivevaccine will be based on the theoretical risk and concern associated with the particular productunder study. For selection of standard developmental toxicity endpoints, the ICHS5a documentshould be consulted. Briefly, parameters that should be included are maternal weight gain, clinicalobservations, implantation number, corpora lutea number, litter size, number of live fetuses, fetaland embryonic deaths, resorptions, fetal weight, and crown-rump length, as well as incidence anddescription of external, visceral, and skeletal malformations. Other parameters may also be included.Postnatal evaluations may include maternal-newborn relationship, neonate adaptation to extrauterinelife, preweaning development and growth, survival incidence, developmental landmarks, andfunctional testing. Again, other parameters may also be included.Since the immune system represents the primary target organ for preventive vaccines, thepharmacodynamic effects of the vaccine antigen may potentially result in unwanted findings in the© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 893mother and/or the offspring. Therefore, it is important to measure various immune response parametersthat may be induced by the vaccine. At a minimum, serum samples collected from pregnantanimals should be assessed for antibody production at designated time points. In addition, cordblood as well as blood samples from newborn animals should be evaluated for the presence ofantibodies (i.e., antibody transfer from the mother to the fetus or neonate). If adverse effects onembryo-fetal development are observed, additional studies, including studies for assessments ofpotential immunotoxicity (i.e., lymphocyte subset tests, immune function tests) may be necessaryto further evaluate the mechanism for the toxicity.For certain vaccines, there may be concerns that immunization of pregnant females may interferewith the ability of the offspring to mount an active immune response to either the same or a relatedvaccine antigen. Such concerns may need to be addressed on a case-by-case basis in clinicalimmunogenicity studies in infants born to mothers that have been immunized with the vaccineduring pregnancy.D. Summary RemarksCBER has formulated guidance for developmental toxicity studies for preventive vaccines in a draftguidance document for industry Considerations for Reproductive Toxicity Studies for PreventiveVaccines for Infectious Disease Indications. 52 However, regulatory policy with regard to reproductiveand developmental toxicity assessment of preventive vaccines will need to be viewed in thecontext of the evolving field of biotechnology and may change as new test systems and informationbecome available. It is important, therefore, to initiate a dialogue with CBER to discuss proposedprotocol designs for preclinical reproductive toxicity studies well in advance of implementation.Preclinical protocols for these products should follow general principles set forth in the CBERdraft guidance document, applicable ICH documents, and other available guidance. However, themost useful approach for reproductive and developmental toxicity testing of preventive vaccinesshould be based on rational science.A. PerspectiveIV. DRUG SAFETYCDER is currently involved in an attempt to change the way animal data are translated into riskassessment for humans. The primary focus is on molecular entities new to the market that havebeen tested in a relatively small series of clinical trials, sometimes 100 patients or less. When adrug first goes to market, it is likely that there will be very little information about its effects onmother or offspring when given during pregnancy, unless it is very specifically intended to be usedduring pregnancy. The most common situation with drugs at the FDA is that the molecular entityis new, animal data is available, and essentially no human experience data is available to predictwhat will happen to the general population. This is a very different situation from one in which aproduct has been on the market for several years, and product registries, behaviors, etc., exist forit. In a study, the use of a drug is tightly controlled, whereas drugs in the general population canbe taken in combination with other prescription medications, over-the-counter medications, andherbal and food supplements.B. Product LabelsAt the present time, the FDA uses established pregnancy labeling categories for drugs, but manycomplaints have been made against the use of the categories. The categories are defined by theavailability of human data (positive or negative), the availability of animal reproductive toxicity© 2006 by Taylor & Francis Group, LLC

894 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONTable 22.2Current pregnancy categoriesCategory Effects CriteriaA No adverse effects demonstrated in human clinical trials Most stringent criteriaB No effect in humans with evidence of adverse effects in Less stringent criteriaanimals or no effects in animals without adequatehuman dataC Adverse effects in animals in the absence of human Assumption of a level of risk without datadata or no data available for animals or humansD Adverse effects demonstrated in humans or adverseeffects in animals with strong mechanistic expectationof effects in humansExpectation of translation of effects inhumans based on conservative geneticpropertiesXAdverse effects in humans or animals without indicationfor use during pregnancyMay indicate high risk for adversereproductive effectsdata (positive or negative), and the indication for use (Table 22.2). The “limits” or “extremes” ofpregnancy labeling are generally defined by clinical data. Risk and benefit are codeterminants ofthe drug product category. Based on the 1996 Physicians’ Desk Reference (PDR), approximately25% of the products in the PDR have no information on reproduction.There is an ongoing attempt to provide the clinician with more information than is currentlyavailable with the five pregnancy categories. It is an attempt to deal with some of the inconsistenciesthat have occurred in the evaluation of some drugs. For example, historically, all oncology drugshave been given a pregnancy category of “D,” and some of the assignments may not be justified.It is time to improve the system. The changes being proposed are designed to reduce some of theassumptions that have been made up to now and to evaluate the risk of reproductive effects ofdrugs with a more objective series of questions. The changes will involve primarily the evaluationof new products.C. Pregnancy Labeling — Practical ConsiderationsThe important aspects of pregnancy category assignment and risk assessment are that nonclinical studydata generally define the pregnancy labeling for new products and systematically collected clinical dataare unlikely for agents with “significant” or “meaningful” positive effects in nonclinical studies.One issue to be addressed is that risk and benefit are codeterminants of the pregnancy categoryfor a drug or therapeutic. Risk ranges from none or minimal to significant, and benefit ranges fromminimal to significant. When risk is minimal and benefit significant, or risk is significant and benefitis minimal, the outcome of the codetermination process is easily discerned. However, when riskand benefit are roughly equivalent, the outcome of the analysis becomes less certain and the finaldecision is more difficult. When a particularly difficult situation arises in the evaluation process,the product may be referred to the FDA’s Reproductive Toxicity Committee and to a policy groupthat oversees the committee. One issue to be addressed in the new process is whether risk andbenefit should be addressed separately in drugs and therapeutics used in pregnancy or should theycontinue to be combined. This issue has been extensively debated within the agency.D. Assessment of Reproductive and Developmental ToxicityThere are three types of underlying animal studies: (1) Segment I (ICH Stages A and B): In thistype of study, the effects of a compound are evaluated from premating to conception or fromconception to implantation. In the first type, male and female reproductive function are tested, aswell as gamete maturation, mating, behavior, and fertilization. In the second type, female reproductivefunction, preimplantation development, and implantation are tested. (2) Segment II (ICHstage C): In this type of study, the effects of a compound are evaluated from implantation through© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 895closure of the hard palate. Female reproductive function is tested, as well as embryonic developmentand major organ formation during the period of organogenesis. (3) Segment III (ICH stages D, E,and F): The effects of a compound are evaluated perinatally or postnatally; from hard palate closureto parturition, from parturition to weaning, or from weaning to sexual maturity.Clearly, the aim of the underlying reproductive and developmental toxicity studies in animalsis to reveal possible effects of an agent on human reproduction. Both latent and immediate effectsshould be known, and multigeneration reproductive studies in animals are particularly valuablebecause it is unlikely that multigeneration effects in humans will be measurable.1. Nonclinical Reproductive Toxicity Data Assessment — FutureA Pregnancy Data Integration Working Group has been formed and an integration process is beingdeveloped to serve a threefold purpose: (1) to assist in interpretation and integration of reproductiveand developmental toxicity data, (2) to promote consistency in the interpretation of reproductivetoxicity study findings, and (3) to provide a common framework for the review, interpretation, anddiscussion of findings. The Working Group is developing a “process” or “tool” that reflects theconsensus thought processes of experts in reproductive toxicity and the regulatory sciences asapplied to the interpretation of findings from studies of reproductive and developmental toxicity.A draft guidance document was published in October 2001 by CDER to describe a process forestimating the increase in human developmental and reproductive risks as a result of drug exposurewhen definitive human data are unavailable. 53 The integration process would be used on drug labels.The draft document reflects current thinking on the topic; it is not final. The overall approachintegrates nonclinical information from a variety of sources (i.e., reproductive toxicology, generaltoxicology, and toxicokinetic and pharmacokinetic information, including absorption, distribution,metabolism, and elimination findings) and available clinical information to evaluate a drug’spotential to increase the risk of an adverse developmental or reproductive outcome in humans. Theintegration process focuses on the likelihood that a drug will increase the risk of adverse humandevelopmental or reproductive effects.2. Data Needed for the Integration ProcessThe integration process should be based on an evaluation of a complete set of the expected generaltoxicology, reproductive toxicology, and pharmacokinetics studies. This evaluation should includean assessment of the ability of the drug to produce a positive finding in relevant animal studies(e.g., whether doses used were large enough to induce toxicity of some kind). The evaluation shouldalso compare animal and human pharmacodynamic effects, animal and human metabolism anddisposition, animal and human pharmacological and toxic effects, and drug exposures in animalstudies in relation to the highest proposed dose in humans. The type and extent of availabletoxicology data may vary, depending on the biological actions of the product, test systems availablefor studying the compound, and other factors.3. Types of Reproductive and Developmental Toxicity EvaluatedFor purposes of the integrative process, two broad categories of toxicity are considered: reproductiveand developmental toxicity. In the reproductive toxicity category, three classes of toxicity areevaluated: effects on fertility, parturition, and lactation. In the developmental toxicity category, fourclasses of toxicity are evaluated: mortality, dysmorphogenesis (structural alterations), alterationsto growth, and functional toxicities. For a given drug, each class of toxicity should be assessed. Apositive signal in any class of reproductive or developmental toxicity should be evaluated to estimatethe likelihood of increased reproductive or developmental risk for humans.© 2006 by Taylor & Francis Group, LLC

896 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONa. Reproductive ToxicitiesReproductive toxicities include structural and functional alterations that may affect reproductivecompetence in the F 0 generation.1. Fertility: Male reproductive toxicity associated with administration of a drug may be due todegeneration or necrosis of the cells in the reproductive organs, and may be assessed by severalendpoints, including reduced sperm count, altered sperm motility or morphology, aberrant matingbehavior, altered ability to mate, altered endocrine function, and overall reduced fertility.Female reproductive toxicity may be due to damage to the reproductive organs or alterationsto endocrine regulation of gamete maturation and release, and may manifest as aberrant matingbehavior, altered ability to mate, or overall reduction in fertility. Diminished fertility in femaleanimals is typically detected by reductions in the fertility index, the number of implantation sites,time to mating, or fecundity.2. Parturition: Changes in the duration of parturition are frequently reported as mean time elapsedper pup or as total duration of parturition.3. Lactation: Drugs administered to lactating animals may be a source of unwanted exposure in thenursing neonate, may alter the process of lactation in the nursing mother (e.g., the quality orquantity of milk), or may alter maternal behavior toward the nursing offspring.b. Developmental ToxicitiesDevelopmental toxicities, generally those that affect the F 1 generation, include:1. Mortality: Mortality due to developmental toxicity may occur at any time, from early conceptionto postweaning. A positive signal may appear as pre- or peri-implantation loss, early or lateresorption, abortion, stillbirth, neonatal death, or periweaning loss.2. Dysmorphogenesis: Dysmorphogenic effects (structural alterations) are observed as malformationsor variations to the skeleton or soft tissues of the offspring.3. Alterations to growth: Growth retardation, excessive growth, and early maturation are consideredalterations to growth. Body weight is the most common measurement for assessing growth rate,but crown-rump length and anogenital distance are also frequently measured.4. Functional toxicities: Any persistent alteration of normal physiological or biochemical functionis considered a functional toxicity, but typically only effects on neurobehavior and reproductivefunction are measured. Common assessments include locomotor activity, learning and memory,reflex development, time to sexual maturation, mating behavior, and fertility.4. The Integration ProcessAccording to the October 2001 CDER Guidance Document, the data should be passed through theintegration process in a stepped procedure consisting of three sections that are applicable to alldata sets, applicable only to data sets without evidence of reproductive or developmental toxicity,or applicable to data sets with positive indications of reproductive or developmental toxicity.5. Section A. Overall Decision Tree Applicable to All Data-Sets (Figure 22.1)The decision tree process outlined here should be used for each class of reproductive or developmentaltoxicity. A single study may address several classes of toxicity. Each study should beevaluated individually. A sequential series of decisions should be made in evaluating the varioussituations that may be encountered and the next steps that should be taken where there are studiesthat can be evaluated.© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 8971. Availability of studies?YesNo“Unknown” or “Not Evaluable” RiskInformation is Not Available toAssess Risk2. Relevance of Studies(Test System and Route)No“Unknown” or “Not Evaluable” RiskDefine Non-Relevance of Test Systemor Study(Do Not Use Figure 22.3)Yes3. Presence of a Signalfor an Endpoint?No“No Observed Effect”Use Figure 22.2 for Integration of Datafor Endpoints with No SignalYes“Positive Effect Observed”Use Figure 22.3 for Integration of Datafor Endpoints with Positive ResultsFigure 22.1Overall decision tree for evaluation of reproductive/developmental toxicities.a. Availability of StudiesIf adequate studies were not done or the results are not available for comprehensive evaluation,risk to humans is considered unknown or not possible to evaluate, and the product labeling shouldreflect that conclusion. If studies were conducted and are available for comprehensive evaluation,the assessment should continue.b. Relevance of StudiesIf the test system was not relevant (e.g., nonrelevant route of administration) or not appropriate(e.g., inappropriate test protocol or species), the risk to humans is considered unknown or notamenable to evaluation, and the product labeling should reflect that conclusion. If the studies arerelevant to evaluating the risk to humans, the risk integration process should continue with the nextquestion.c. Presence or Absence of a SignalIf the test system is relevant and appropriate for assessing the risk of toxicity in humans, the nextquestion that should be asked is: “Was there a positive signal (suggesting toxicity)?” If no positivesignal (suggesting toxicity) was seen, the evaluation process should continue to Section B. If apositive signal was seen, the evaluation process should continue to Section C.© 2006 by Taylor & Francis Group, LLC

898 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION6. Section B. No Signal (Applicable to Data-sets without Evidenceof Reproductive or Developmental Toxicity) (Figure 22.2)Where there is no positive signal for one of the seven classes of reproductive or developmentaltoxicity, the risk assessment should be a stepwise process leading to a recommendation about therelevance of the nonfinding in humans. If multiple studies are available and all the results arenegative, the process in Section B should be used. If any study has a positive signal, the processin Section C should be used. The following factors should be considered during the evaluation ofeach class of reproductive or developmental toxicity for which there was no signal:a. The Model or Test Species Predictive AdequacyThe following questions bear on the determination of a model’s predictive adequacy. Do any ofthe model or test species (or systems) demonstrate or have the capability of responding to thepharmacodynamic effects of the drug? Do any of the model or test species (or systems) demonstratean overall toxicity profile that is relevant to the human toxicity profile? Do any of the model ortest species (or systems) demonstrate pharmacokinetic (including absorption, distribution, metabolism,and elimination) profiles for the drug that are qualitatively similar to those in humans? Ifthe responses to these questions suggest that the response of the test species is of little relevanceto humans, the review should explain why the test may have low predictive value. Even if the testsystem is determined to be of limited relevance, the review should consider the remaining factors2 to 4 (see following sections) and describe any additional uncertainties.b. Adequacy of Study Doses and ExposureThe following elements should be considered in assessing the relevance to humans of the drugdoses and exposure in the test system: Were adequate doses (concentrations) of the drug administeredto the test species or test systems (e.g., maximum tolerated dose [MTD])? Were the exposures(based on the area under the curve of plasma concentration time course [AUC], maximum achievedconcentration [C max ], or other appropriate systemic exposure metric) achieved in the test speciesor test systems adequate relative to those demonstrated in humans at the maximum recommendedhuman dose? If the answer to either of these questions is no, the evaluation should state that theanimal studies are inadequate and explain why. The evaluation should proceed to Section C below.c. Class AlertClass alerts should be based on adverse reproductive or developmental effects previously demonstratedin humans by closely related chemical entities or compounds with similar pharmacodynamiceffects. If there is a class alert for the drug, based on a related chemical structure of parent ormetabolite or related pharmacologic effect, the class-specific information relevant to the class oftoxicity reported to be negative should be included in the risk evaluation and discussion of thedrug. The basis for the class alert for adverse effects in humans should be reasonably applicableto the drug being evaluated.d. Signals for Related Types of Reproductive and Developmental ToxicityThe next step in evaluating the relevance is to assess findings for related reproductive and developmentaltoxicities. A positive signal for one class of toxicity may suggest some risk in humansfor other toxicities in the same category for which there were no findings in animals. The issue ofrelated toxicities is most relevant for developmental toxicities. For example, if there is no signalfor fetal mortality but there is a positive signal for alterations to growth or dysmorphogenesis in© 2006 by Taylor & Francis Group, LLC

Figure 22.2No Signal1. Adequacy of Model2. Adequate Study Doses/Exposure?3. Class Alert?4. Any Endpoint Positivein Related Repro/Devel.Category?“No Predicted Risk”Decision tree for reproductive/developmental toxicities with no signal.NoNoNoYesYes“Unknown Risk” OR “Possible RiskBased on Related Drug”Inadequate Information Exists toAssess Risk to HumanReproduction, Since... The TestSystem or test Conduct wereDeemed Lacking (Describe Situation)OR, the Compound wasSubject to a Class Alert (DescribeSituation and Include ClassInformation).“No Observed Effects on..”PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 899© 2006 by Taylor & Francis Group, LLC

There are six factors that may affect the level of concern for a positive signal: (1) signal strengthwith respect to cross-species concordance (where more than one species has been studied), presenceof multiple effects, and presence of adverse effects at several times; (2) signal strength analyzedwith respect to the coexistence of maternal toxicity, the presence of a dose response relationship,and the observation of rare events; (3) pharmacodynamics, (4) concordance (metabolic and toxicologicconcordance with the human); (5) relative exposure; and (6) class alerts.The positive signal may be from a reproductive or developmental toxicology study or an effectobserved on a reproductive tissue, system, or behavior in a general toxicology study. Each factorshould be analyzed independently. Guidance is provided on what types of observations for each ofthe six factors might increase, decrease, or leave unchanged the level of concern for that factor.These analyses should be a weighted integration that takes into account the quality and nature ofthe data under consideration. The assessments of concern for each of the six factors should beassigned values of +1, –1 or 0, respectively, if the factor is perceived as increasing, decreasing, orleaving unchanged the level of concern for a class of reproductive or developmental toxicity.Conclusions from the six analyses should be summed to arrive at a comprehensive evaluation ofthe potential increase in risk for each class of the seven reproductive or developmental classes forwhich there was a positive signal.Intra- and interspecies concordance of adverse effects in animals deserves some special considerationin this risk integration process. Positive signals for related types of reproductive ordevelopmental toxicity within the same species indicate intraspecies concordance of effects (e.g.,a reduction in normal growth and an increase in developmental mortality). Positive signals for thesame or a related type of toxicity across species indicate interspecies concordance. In general,findings for which there is intra- or interspecies concordance are more convincing than a positivesignal in only one toxicity class in only a single species.In evaluating potential human risk for adverse reproductive or developmental outcomes, if thereis interspecies concordance for a single adverse effect, it may be reasonable to conclude that asimilar effect is the most likely adverse event to be seen in humans treated with the drug. If differentbut related adverse effects are seen in multiple test species (e.g., alterations to growth in one speciesand developmental mortality in another, or parturition effects in one species with lactation effectsin the second), it may be reasonable to assume there is some level of risk for categorically relatedendpoints in humans.A detailed discussion of the overall integrative analysis, the six individual factors, the contributoryelements for each factor, and the assignment of the level of concern for each factor is presented here.900 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONone or more animal species, it may be inappropriate to conclude there is no risk of fetal mortalityfor humans. Related toxicities may also be relevant for reproductive toxicities where a hormonalmechanism is identified and the mechanism is relevant to humans.If positive signals for related classes of toxicity were observed in the animal studies, theevaluation should state that there was no observed effect on the type of toxicity being assessed,but positive signals were seen for related toxicities. If there is no positive signal for any classof reproductive or developmental toxicity, the evaluation should state that there is no expectedincrease in risk for reproductive or developmental toxicity in humans based on the results ofanimal studies.7. Section C. Applicable to Data-sets with One or More Positive Indicationsof Reproductive or Developmental Toxicity (Figure 22.3)8. Factor 1. Signal Strength, Part IA positive signal in any reproductive or developmental toxicity class should be analyzed withrespect to whether the finding is present in more than one setting: cross-species concordance (where© 2006 by Taylor & Francis Group, LLC

CLASSES OF TOXICITYOR SIGNALSReproductiveToxicity1. Fertility2. Parturition3. LactationPositivesignalDevelopmentalToxicity1. Mortality2. Dysmorphogenesis3. Alterations to growth4. Functional toxicityFigure 22.3SignalStrengthISignalStrengthIAnimal DataSignalStrengthIISignalStrengthIIPDPDADMETOXProgress to Next StepADMETOXExposureExposureData Integration ProcessClass,AlertsClass,AlertsIntegration tool for reproductive or developmental toxicites with a positive signal.Human DataPredictedto IncreaseHumanRepro. RiskMayIncreaseHumanRepro. RiskDoes NotAppear toIncreaseHumanRepro. RiskHumanDataConcernConcernHumanDataPRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 901© 2006 by Taylor & Francis Group, LLC

902 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONmore than one species has been studied), multiplicity of effects, and adverse effects seen at morethan one time.a. Cross-Species ConcordanceThe defining characteristic of cross-species concordance is a positive signal in the same class oftoxicity in more than one species. Cross-species concordance is most likely to be identified forstructural abnormalities (dysmorphogenesis) or developmental mortality because these toxicitiesare frequently detected in the organogenesis testing paradigm, in which multiple species aretypically evaluated. In addition, alterations to endocrine function or gonadal histopathology (whichmay alter fertility) may be indirectly detected in subchronic and chronic toxicity studies in rodentsand nonrodents. When cross-species concordance is observed, there is increased concern forreproductive or developmental toxicity in humans. In contrast, there is decreased concern when asignal is detected in only one species (with the proviso that the negative species is an appropriateanimal model and the studies were adequate in design, dosing, and implementation).b. Multiplicity of EffectsMultiplicity of effects refers to observation, in a single species or animal model, of two or more positivesignals within one of the two general categories of toxicity (reproductive or developmental) or withinone of the seven classes of reproductive or developmental toxicities. The observation of increasedembryo-fetal death and structural abnormalities in an animal test species is an example of multiplepositive signals within a general category. The observation of two or more positive signals for structuralabnormalities in tissues of multiple embryonic origin (e.g., defects affecting soft tissue, skeletal tissue,and/or neural tissue) is an example of multiple positive signals in a toxicity class.If all species examined demonstrate multiplicity of effects, there is increased concern forreproductive or developmental toxicity in humans. If there are positive signals in two or morespecies, but multiplicity of effects is observed in only one species, concern is unchanged for thiselement. If no species studied exhibits multiplicity of effects, there is decreased concern.c. Adverse Effects at Different Stages of the Reproductive or Developmental ProcessEvidence of toxicity may arise during any stage of the reproductive or developmental process. Ifa positive signal in animals is observed in multiple stages of development, there is generally greaterconcern for adverse human reproductive outcomes. If a positive signal is observed only during asingle, discrete interval, the level of concern is unchanged. If the positive signal occurs only duringprocesses that are of limited relevance to humans, there would be less concern for adverse humanreproductive outcomes. In addition to its relevance to this evaluation process, it is also importantto define the timing of the period of susceptibility for the observed positive signal to provide acontext for the human risk.9. Factor 2. Signal Strength, Part IIA positive signal should also be analyzed with respect to the coexistence of maternal toxicity,presence of a dose response relationship, and the observation of rare events.a. Maternal ToxicityIn weighing a signal of toxicity, the magnitude of adverse effects in the offspring versus the severityof maternal (and, for fertility studies, paternal) toxicity should be considered when drawing aconclusion about the relevance of the F 0 toxicity to effects observed in the offspring. A positive© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 903signal occurring at doses that are not maternally toxic increases concern for human reproductiveor developmental toxicity. If a positive signal is observed only in the presence of frank maternaltoxicity, there is decreased concern, provided that the positive signal may be reasonably attributedto maternal toxicity.When evaluating a positive signal in two or more species, assessment of the implications ofmaternal or paternal toxicity should be based on a composite analysis of the data from all adequatelystudied species. If a positive signal is seen in two or more species in the absence of maternaltoxicity, there is increased concern for adverse human reproductive outcomes. If a positive signalis seen only in the presence of clear relevant maternal toxicity in multiple species, there is decreasedconcern. If there is nonconcordance between test species as to the presence and relevance of maternaltoxicity, there may be no change in the overall level of concern for this contributory element.b. Dose Response RelationshipConcern for human reproductive or developmental toxicity is increased when a positive signal ischaracterized by any of the following: increased severity of adverse effects with an increase indose, increased incidence of adverse effects with an increase in dose, or a high incidence of adverseeffects across all dosed groups. Conversely, the absence of all three of these indicators of doseresponse would be cause for unchanged or decreased concern.If multiple species are evaluated, a clear dose response across all tested species increasesconcern. If a positive signal occurs in more than one species, only one of which demonstrates oneof the dose response relationships described above, the level of concern will generally be unchanged.If there is no dose response in any species, concern is decreased.c. Rare EventsDevelopmental toxicity studies usually lack the statistical power to detect subtle increases in rareevents. Thus, an increased frequency of positive signals for rare events in drug-exposed animalsincreases concern for reproductive or developmental toxicity in humans. The absence of an increasedfrequency of rare events, however, does not decrease concern. When multiple species are studied,an increased frequency of positive signals for rare events in more than one species increases concernfor adverse outcomes in humans.10. Factor 3. PharmacodynamicsA positive signal should be analyzed with respect to the therapeutic index, biomarkers as abenchmark, and similarity between the pharmacologic and toxicologic mechanisms.a. Therapeutic IndexThe therapeutic index (TI) is used to identify the extent to which there is overlap between therapeuticdoses and doses that cause reproductive or developmental toxicity. It is unusual to obtain welldefineddose–response curves for toxicity and efficacy from a single species. Thus, the use ofestimations or surrogate endpoints (related to the therapeutic mechanism) for this evaluation maybe warranted. To reduce the impact of variation in the slope of the dose response curves, estimationof the TI should generally be based on comparison of the TD10 and the ED90 concentrations, andresulting in the composite score called TI10/90.The TD10 (toxic dose or concentration) should be defined by the C max (or other appropriateexposure metric) that produced the toxic reproductive or developmental response in 10% of aresponsive or sensitive species, whereas the ED90 (efficacious dose or concentration) should bedefined by the C max (or other appropriate exposure metric) that produced the desired effect in 90%© 2006 by Taylor & Francis Group, LLC

904 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONof the test species. These parameters can be estimated. Preferably, both the TD10 and ED90 wouldbe defined in the same species. In some instances, estimation of the ED90 can be based on in vitrocell inhibition studies (frequently seen for antibiotics and antineoplastic agents). Although lessdesirable, efficacy data can be derived from another species, but caution should be exercised insuch situations. The same exposure metric should be used in the estimation of the TD10 and ED90values. Scientific justification for the drug exposure metrics used for comparison should be provided.If the TI10/90 is less than 5 (obtained by dividing TD10 by ED90), there is increased concernfor reproductive or developmental toxicity in humans, as there is limited separation in the dosescausing adverse effects from those responsible for efficacy. If the TI10/90 ratio falls between 5 and20, the level of concern is unchanged. If the TI10/90 ratio is greater than 20, there is decreasedconcern because of the wide separation in doses causing adverse effects from those resulting inefficacy.If there are data available to determine the TI10/90 ratio in multiple species, assessment of thelevel of concern for this element should be based on an integrated analysis of data from alladequately studied species. The extent of concordance in the size of the TI10/90 between speciesmay increase, decrease, or leave unchanged the level of concern (i.e., the greater the concordance,the more likely concern will be increased). In the event of nonconcordance of the TI ratios betweenmultiple test species, the nature of the positive signals observed and the relevance of the endpointand test species to the human condition should be considered before making an assessment. In theevent that one species is considered inappropriate to the analysis, the evaluation should be performedwithout reference to that species.b. Biomarkers as a BenchmarkThere may be circumstances in which an effect on a biomarker is consistently seen in multiplespecies at doses lower than the NOEL for demonstrable reproductive and developmental toxicity.If there is an effect on this biomarker at or below the therapeutic dose in humans, there is increasedconcern for reproductive or developmental toxicity in humans. If this biomarker is responsive tothe drug in humans, can be monitored, and is not affected at the therapeutic dose, there may bedecreased concern.c. Similarity between Pharmacologic and Reproductive DevelopmentalToxicologic MechanismsIf a positive signal is an extension of, progression of, or related response to the intended pharmacologiceffect of the drug (e.g., delay of parturition by drugs known to suppress uterine smoothmuscle contractility or hypotension in the offspring of dams treated during late gestation with adrug known to lower blood pressure), there is increased concern for reproductive or developmentaltoxicity in humans. There is less concern if the positive signal is attributed to an animal-specificpharmacological response, even though it may be an extension of the pharmacologic effect of thedrug (e.g., pregnancy loss in rats due to hypoprolactinemia).11. Factor 4. Concordance between the Test Species and HumansConcordance between the test species and humans should be evaluated with respect to the metabolicand drug distribution profiles, the general toxicity profiles, and biomarker profiles.a. Metabolic and Drug Distribution ProfilesDrug distribution, elimination, and biotransformation (pathways and metabolites) in the test speciesand in humans should be compared. Quantitative differences in metabolic and drug distribution© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 905profiles between the test species and humans are often seen, may not have important implications,and should not be overemphasized. Reproductive and developmental toxicities induced by compoundswhose metabolic and distribution profiles are very similar in animals and humans increasesconcern for reproductive or developmental toxicity in humans. For compounds with highly dissimilarmetabolic or tissue distribution profiles in animals and humans, there is less concern if thetoxic effect seen in the test species can be attributed to a metabolite or tissue distribution profilenot seen in humans. For any other scenario, the concern is unchanged.When there are significant differences in drug distribution or metabolic profiles between severalspecies, yet each test species demonstrates a positive signal for a reproductive or developmentaltoxicity, the toxicity is assumed to be attributable to the parent drug or a common biotransformedproduct, and concern is increased.b. General Toxicity ProfilesIf the overall toxicity profile of a drug in one or more test species with a positive signal is similarto that in humans, there is increased concern for reproductive or developmental toxicity in humans.If the overall toxicity profiles are dissimilar, there may be decreased concern. When generaltoxicology data are available for more than one species, the determination of the level of concern(increased, decreased, or unchanged) should be based on an assessment of each test species’ abilityto indicate human adverse effects in response to the drug.c. Biomarker ProfilesWhen biomarker profiles are available for comparison, an approach similar to that described in theprevious section on assessment of general toxicity profiles may be useful.12. Factor 5. Relative ExposuresWhen considering the relative exposure comparisons discussed in the next section, more emphasisshould be placed on a parameter within this factor when there is a scientifically plausible linkbetween the exposure metric (or biomarker) and the adverse reproductive (or developmental) effect.a. Kinetic Comparison of Relative ExposureComparison of systemic drug exposure at the NOEL for the reproductive or toxicity class in thetest species to that in humans at the maximum recommended dose is a critical determination. Thiscomparison should be based on the most relevant metric (e.g., AUC, C max , C min , BSA [body surfacearea] adjusted dose). In general, there is increased concern for reproductive or developmentaltoxicity in humans for relative exposure ratios (animal:human) that are 10 or less, decreased concernfor exposure ratios that are 25 or more, and no change in concern for ratios between 10 and 25.When applicable, the relative exposure ratio should consider both the parent compound and itsmetabolites. For example, it is appropriate to combine parent compound and metabolite when bothare pharmacologically active and the activity relates to the reproductive or developmental toxicity.Where there are exposure data for multiple test species, the NOEL exposure for each shouldbe compared to human exposure at the maximum recommended dose. If the exposure ratios arelow (10-fold or less) in multiple species with a positive signal, there is increased concern. If theexposure ratios are high (25-fold or more), there is decreased concern. In the event a significantdifference in relative exposures is observed between multiple test species, the appropriateness ofthe metric (e.g., AUC, C max ) being used to define the interspecies exposure comparisons should bereassessed. If an alternative metric fails to reduce the disparity between the species, the assessmentof concern should be based on the lowest ratio (i.e., in the most sensitive species).© 2006 by Taylor & Francis Group, LLC

906 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITIONRelative interspecies exposure data should be evaluated in light of species differences in proteinbinding (free drug concentration), receptor affinity (if related to the positive signal), or site-specificdrug concentrations. In the absence of meaningful differences between the test species and humansin these parameters, the interspecies comparisons should be based on total drug exposure.b. Biomarkers as a Measure of Relative ExposureThe purpose of this relative exposure metric is to compare the dose causing a reproductive toxiceffect in the test species to the therapeutic dose in humans, normalized to the doses causing aresponse common to both species. In practice, this is done by taking the NOEL for the adversereproductive or developmental effect and dividing by the dose at which the biomarker response isseen in the test species. This is compared with the human therapeutic dose divided by the dose atwhich the biomarker response is seen in the human. The ratio calculated for animals is then dividedby the ratio calculated for humans. When this ratio of relative biomarker exposure (animal tohuman) is 10 or less, there is increased concern for human reproductive or developmental toxicity.When this ratio is 25 or more, there is less concern. When this ratio falls between 10 and 25, thelevel of concern is unchanged.Where there are data to compute relative biomarker exposure ratios for multiple species, thelevel of concern assessment should be based on an integrated analysis of species. Where there arenonconcordant biomarker ratios between multiple test species, the relevance of the biomarker asexpressed in the various species should be considered before making an assessment. If there is noscientific rationale for the disparity between species, the biomarker, as a measure of exposure, willbe of questionable utility.13. Factor 6. Class AlertsConsideration of a class associated effect should be based on prior human experience for a drugwith related chemical structure (parent or metabolite) or related pharmacologic effect, and withknown reproductive or developmental outcomes in humans. There is increased concern for reproductiveor developmental toxicity in humans when the drug is from a class of compounds knownto produce adverse effects in the same toxicity class in humans and animals. There is decreasedconcern only in circumstances in which a class of compounds, although demonstrating adverseeffects in animals, has been previously shown to have no related adverse effects on human reproductionor development. In the absence of adequate human reproduction or developmental data fora class, the level of concern is unchanged.14. Summary and Integration of Positive FindingsThe factors discussed here are derived from a limited sample of pharmaceuticals where the clinicaloutcomes are reasonably well defined. CDER believes that using specific factors and benchmarkvalues to assess the potential to increase risk to humans for adverse reproductive and developmentaloutcomes will result in a more unbiased and uniform evaluation. CDER also believes this approachwill help identify specific areas of additional information about a pharmaceutical that would beuseful in more fully defining risk and will allow specific analysis of areas of disagreement thatinfluence the risk evaluation.• Where there is a positive finding in nonclinical or general toxicology studies for one of the sevenclasses of reproductive or developmental toxicity, there is a potential for increased human risk. Inevaluating the level of increased risk, positive findings from each of the seven classes of reproductiveand developmental toxicity should be assessed separately. All relevant information shouldbe considered.© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 907• In evaluating the level of concern for each of the six factors in the overall assessment, the analysisshould reflect the weight of evidence, taking into account the quality and type of data underconsideration for each factor (i.e., should not be merely an arithmetic summation of the contributoryelements for each factor). For each factor there should be a determination of increased (+1),decreased (–1), or no change (0) in the level of concern.• The values for the six factors should then be summed to arrive at one of the following overallconclusions for each class of reproductive or developmental toxicity: (1) the drug is predicted toincrease risk, (2) the drug may increase risk, or (3) the drug does not appear to increase risk ofthat class of reproductive or developmental toxicity in humans. Where there is sufficient informationabout the drug to assess each of the six factors within Figure 22.3, a net value of +3 or moresuggests that a drug is predicted to increase risk for that class of toxicity in humans, a valuebetween +2 and 2 suggests that the drug may increase the risk, and a value of –3 or less suggeststhe drug does not appear to increase the risk.The summary risk conclusions for the outcomes of analyses using Figure 22.3 are:• Does not appear to increase risk: The drug is not anticipated to increase the incidence of adversereproductive (or developmental) effects above the background incidence discussed in humans whenused in accordance with dosing information in the product label.• May increase risk: The drug may increase the incidence of adverse reproductive (or developmental)events above the background incidence in humans when used in accordance with the dosinginformation in the product label.• Predicted to increase risk: The drug is expected to increase the incidence of adverse reproductive(or developmental) events above the background incidence in humans when used in accordancewith the dosing information in the product label.ACKNOWLEDGMENTSMany thanks are due to the following persons from the various FDA centers who reviewed orhelped expedite the manuscript: Mitchell Cheeseman, Karen Goldenthal, David Hattan, KarenMidthun, Robert Osterberg, George Pauli, Mercedes Serabian, and John Welsh.REFERENCES1. Kokoski, C.J., et al., Methods used in safety evaluation, in Food Additives, Branen, A.L., Davidson,P.M., and Salminen, S., Eds., Marcel Dekker, New York, 1990, p. 579.2. Food and Nutrition Board, National Research Council, Risk Assessment/Safety Evaluation of FoodChemicals, NAS Press, Washington, D.C., 1980.3. FDA, Toxicological Principles for the Safety Assessment of Direct Food Additives and Color AdditivesUsed in Food, U.S. Food and Drug Administration, Bureau of Foods, Washington, DC, 1982.4. FDA, Toxicological Principles for the Safety Assessment of Direct Food Additives and Color AdditivesUsed in Food, “Draft Redbook II,” U.S. Food and Drug Administration, Center for Food Safety andApplied Nutrition, Washington, D.C., 1993.5. FDA, Toxicological Principles for the Safety of Food Ingredients, U.S. Food and Drug Administration,Center for Food Safety and Applied Nutrition (http://www.cfsan.fda.gov/~redbook/red-toca.html),2000.6. FDA, Recommendations for Submission of Chemical and Technological Data for Direct Food Additiveand GRAS Food Ingredient Petitions, U.S. Food and Drug Administration, Center for Food Safetyand Applied Nutrition, Washington, D.C., 1993.7. FDA, Color Additive Petitions: FDA Recommendations for Submission of Chemical and TechnologicalData on Color Additives for Food, Drugs, or Cosmetics, U.S. Food and Drug Administration,Center for Food Safety and Applied Nutrition (http://www.cfsan.fda.gov/~dms/opa-col1.html), 1997.© 2006 by Taylor & Francis Group, LLC

908 DEVELOPMENTAL REPRODUCTIVE TOXICOLOGY: A PRACTICAL APPROACH, SECOND EDITION8. Kokoski, C.J. and Flam, W.G., Establishment of acceptable limits of intake, Proceedings of the SecondNational Conference for Food Protection, Department of Health and Human Services, Washington,DC, 1984, p. 61.9. Kokoski C.J., Regulatory toxicology and food flavors, in Thermal Generation of Aromas, Parliament,T.H., McGorrin, R.J., and Ho, C.T., Eds. American Chemical Society, Washington, DC, 1989, p. 23.10. Rulis, A.M., Safety assurance margins for food additives currently in use, Regul. Toxicol. Pharmacol.,7, 160, 1987.11. FDA, The Food and Drug Modernization Act, P.L. 105-115, 105th Congress, 1997.12. FDA, Guidance for Industry. Preparation of Food Contact Notifications for Food Contact Substances:Toxicology Recommendations, Final Guidance, February 2002 (http://intranet.cfsan.fda.gov/ofas/internetprep/opa-fcnt.html), 2002.13. FDA, Food additives: food contact substance notification system, Fed. Regist., 67, 35724, May 21,2002.14. FDA, Guidance for Industry. Preparation of Premarket Notifications for Food Contact Substances:Chemistry Recommendations (http://www.cfsan.fda.gov/~dms/opa-pmnc.html), 2000.15. FDA, Guidance for Industry, Preparation of Premarket Notifications for Food Contact Substances:Toxicology Recommendations (http://www.cfsan.fda.gov/~dms/opa-pmnt.html), 1999.16. Copestake, P., The acceptable daily intake, Food Chem. Toxicol., 27, 273, 1989.17. World Health Organization (WHO), Procedures for the Testing of Intentional Food Additives toEstablish Their Safety for Use, Second Report, FAO Nutrition Meetings Report Series, 1958, No. 17;WHO Technical Report Series 1958, No. 144.18. Jackson, B.A., Regulatory implications of reproductive toxicity, in The Toxicology Forum 1988 AnnualWinter Meeting, Toxicology Forum, Washington, D.C., 1988, p. 207.19. Gaylor, D.W., et al., Health risk assessment practices in the U.S. Food and Drug Administration,Regulat. Toxicol. Pharmacol., 26, 307, 1997.20. Hess, R.A. and Moore, B.J., Histological methods for evaluation of the testis, in Male ReproductiveToxicology, Chapin, R.E. and Heindel, J., Eds., Academic Press, San Diego, CA, 1993, p. 52.21. Russell, L.D. et al., Histological and Histopathological Evaluation of the Testis, Cache River Press,Clearwater, FL, 1990.22. Russell, L.D., Normal testicular structure and methods of evaluation under experimental and disruptivecondition, in Reproductive Toxicity of Metals, Clarkson, T.W., Nordberg, G.F., and Sager, P.R., Eds.,Plenum, New York, 1983, p. 227.23. Klinefelter, G., Gray, E.L., and Suarez, J.D., The method of sperm collection significantly influencessperm motion parameters following ethane dimethanesulphonate administration in the rat, Reprod.Toxicol., 5, 39, 1991.24. Seed, J., et al., Consensus report: methods for assessing sperm motility, morphology, and counts inthe rat, rabbit and dog, Reprod. Toxicol., 10, 237, 1996.25. Chapin, R.E., et al., Methods for assessing rat sperm motility, Reprod. Toxicol., 6, 267, 1992.26. Toth, G.P., et al., The automated analysis of rat sperm motility following subchronic epichlorhydrinadministration: methodologic and statistical consideration, J. Androl., 10, 401, 1989.27. Toth, G.P., et al., Effects of three male reproductive toxicants on rat cauda epididymal sperm motion,Reprod. Toxicol., 6, 507, 1992.28. Slott, V.L., et al., Acute inhalation exposure to epichlorhydrin transiently decreases rat sperm velocity,Fundam. Appl. Toxicol., 115, 597, 1990.29. Klinefelter, G.R., et al., Chlorethylmethanesulfonate-induced effects on the epididymis seem unrelatedto altered Leydig cell function, Biol. Reprod., 51, 82, 1994.30. Toth, G.P., et al., Sources of variation in the computer-assisted motion analysis of rat epididymalsperm, Reprod. Toxicol., 5, 487, 1991.31. Oberlander, G., Yeung, C.H., and Cooper, T.G., Indication of reversible infertility in male rats by oralornidazole and its effects on sperm motility and epididymal secretions, J. Reprod. Fertil., 100, 551,1994.32. Slott, V.L. et al., Synchronous assessment of sperm motility and fertilizing ability in the hamsterfollowing treatment with alpha-chlorhydrin, J. Androl., 16, 523, 1995.33. Stevens, K.R. and Gallo, M.A., Practical considerations in the conduct of chronic toxicity studies, inPrinciples and Methods of Toxicology, Hays, A.W., Ed., Raven Press, New York, 1989, p. 237.© 2006 by Taylor & Francis Group, LLC

PRINCIPLES OF RISK ASSESSMENT — FDA PERSPECTIVE 90934. Schwetz, B.A., Rao, K.S., and Park, C.N., Insensitivity of tests for reproductive problems, J. Environ.Pathol. Toxicol., 3, 81, 1980.35. Blazak, W.F., Ernst, T.L., and Stuart B.E., Potential indicators of reproductive toxicity, testicular spermproduction and epididymis sperm numbers, transit time and motility in Fisher 144 rats, Fundam. Appl.Toxicol., 5, 1097, 1985.36. Hess, R.A., et al., The fungicide benomyl methyl-1-(butylcarbamoyl)-2-(benzimidazolecarbamate)causes testicular dysfunction by inducing the sloughing of germ cells and occlusion of efferentductules, Fundam. Appl. Toxicol., 17, 733, 1991.37. Nakai, M., Moore, B.J., and Hess, R.A., Epithelial reorganization and irregular growth followingcarbendazim-induced injury of the efferent ductules of the rat testis, Anat. Rec., 235, 51, 1993.38. Berndtson, W.E., Methods for quantifying mammalian spermatogenesis: a review, J. Anim. Sci., 44,818, 1977.39. Foote, R.H., Schermerhorn, E.C., and Simken, M.E., Measurement of semen quality, fertility andreproductive hormones to assess dibromochloropropane (DBCP) effects in live rabbits, Fundam. Appl.Toxicol., 6, 628, 1986.40. Ku, W.W., et al., Testicular toxicity of boric acid (BA): relationship of dose to lesion developmentand recovery in F344 rat, Reprod. Toxicol., 7, 305, 1993.41. Zenick, H.,et al., Assessment of male reproductive toxicity: a risk assessment approach, in Principlesand Methods of Toxicology, Hays, A.W., Ed., Raven Press, New York, 1994, p. 937.42. Nestor, A. and Handel, M., The transport of morphologically abnormal sperm in the female reproductivetract of mice, Gamete Res., 10, 119, 1984.43. Redi, C.A. et al., Spermatozoa of chromosomally heterozygous mice and their fate in male and femalegenital tracts, Gamete Res., 9, 273, 1984.44. Filler, R., Methods for evaluation of rat epididymal sperm morphology, in Methods in Toxicology:Male Reproductive Toxicology, Chapin, R.E. and Heindel, J.J., Eds., Academic Press, San Diego, CA,1993, p. 334.45. Linder, R.E., et al., Endpoints of spermatotoxicity in the rat after short duration exposures to 14reproductive toxicants, Reprod. Toxicol., 6, 491, 1992.46. U.S. Environmental Protection Agency (EPA), Guidelines for mutagenicity risk assessment, Fed.Regist., 51, 34006, 1986.47. Wyrobek, A.J., et al., An evaluation of the mouse sperm morphology test and other sperm tests innon-human mammals, Mutat. Res., 115, 1, 1983.48. Amann, R.P., A critical review of methods for the evaluation of spermatogenesis from seminalcharacteristics, J. Androl., 2, 37, 1981.49. Hurtt, M.E. and Zenick, H., Decreasing epididymal sperm reserves enhances the detection of ethoxyethanol-inducedspermatotoxicity, Fundam. Appl. Toxicol., 7, 348, 1986.50. Klinefelter, G.R. et al., The ethane dimethanesulfonate-induced decrease in the fertilizing ability ofcauda epididymal sperm is independent of the testis, J. Androl., 15, 318, 1994.51. International Conference on Harmonization (ICH) Harmonized Tripartite Guideline, Detection oftoxicity to reproduction for medicinal products, Fed. Regist., 59, 48746, 1994.52. FDA, Considerations for reproductive toxicity studies for preventive vaccines for infectious diseaseindications, Fed. Regist., 65, 54534, 2000.53. FDA, Reviewer Guidance. Integration of Study Results to Assess Concerns about Human Reproductiveand Developmental Toxicities (http://www.fda.gov/cder/guidance/4625dft.pdf), October 2001.© 2006 by Taylor & Francis Group, LLC