Reproduction in Domestic Animals

Reproduction in Domestic AnimalsOfficial Organ of European Society for Domestic Animal Reproduction, European VeterinarySociety of Small Animal Reproduction and Spanish Society of Animal ReproductionThe Organising Committee wishes to acknowledge and expressits gratitude to the main sponsors of ICAR 2008 Congress:Editor-in-ChiefProf. Dr. H. Rodriguez-MártinezDivision of ReproductionDepartment of Clinical SciencesFaculty of Veterinary Medicine and Animal ScienceSwedish University of Agricultural Sciences (SLU)Ullsvägen 14C, Clinical Centre. P.O. Box 7054,Ultuna SE-750 07 Uppsala,SwedenTelephone: +46-(0) 18672172Fax: +46-(0) 18673545E-mail: heriberto.rodriguez@kv.slu.seAssociate EditorsProf. Dr. W.A. KingUniversity of GuelphDept. of Biomedical SciencesGuelph, Ont. N1G 2W1, CanadaProf. Dr. E. Martinez-GarciaVeterinary Teaching HospitalDept. of Vet. Pathology, University of MurciaCampus de Espinardo, 30500 Murcia, SpainProf. Dr. M. McGowanSchool of Veterinary ScienceUniversity of QueenslandSt. Lucia QLD 4072l, AustraliaProf Dr. José Luiz Rigo RodriguesUniversidade Federal do Rio Grande do SulFaculdade de VeterináriaCaixa Postal 1500491501-970 Porto Alegre RS, BrazilProf. Dr. E. SatoLaboratory of Animal ProductionGraduate School of Agricultural ScienceTohoku UniversityAoba-Ku, Sendai 981-8555, JapanPrinciple sponsorEditorial Advisory BoardW. R. Allen, Newmarket – J. Aurich, Vienna – F. W. Bazer, Texas – H. Bertschinger, Pretoria – P. Bols, Antwerp – G. Brem, Vienna – B. Brück, Frederiksberg –F. Camillo, Pisa – D. Cavestany, Montevideo – M. A. Crowe, Dublin – H. Dobson, Liverpool – G. Evans, Sydney – A. Fontbonne, Maisons–Alfort – G. Foxcroft, Alberta –R. Geisert, Oklahoma – T. Greve, Frederiksberg – B. Hoffman, Gießen – T. Katila, Helsinki – B. Kemp, Wageningen – G. Kilian, Pennsylvania – A. Kunavongkrit,Bangkok – C. Linde–Forsberg, Uppsala – C. Maxwell, Sydney – K. Niwa, Okayama – J. Nöthling, Pretoria – K. Okuda, Okayama – A. R. Peters, Potters Bar –D. Rath, Neustadt – L. Renato de França, Belo Horizonte – E. Seidel, Ft. Collins – L. Solti, Budapest – T. A. E. Stout, Utrecht – S. S. Suarez, Ithaca – C. Tamanini,Bologna – J. Vasquez, Murcia – E. Watson, Edinburgh – G. v.d. Weijden, Utrecht – A. Ziecik, OlsztynThe Journal is indexed by: AgBiotech News and Information; Agricola C R I S; Animal Breeding Abstracts; BIOBASE; Biological Abstracts; BIOSIS Previews;Chemical Abstracts; Current Contents; Dairy Science Abstracts; Derwent Biotechnology Abstracts; Food Science and Technology Abstracts; ForestProducts Abstracts; Forestry Abstracts; Index Veterinarius; MEDLINE; Nutrition Abstracts and Reviews; Personal Alert; Pig News & Information; PoultryAbstracts; Protozoological Abstracts; Referativnyi Zhurnal; Review of Medical and Veterinary Mycology; Science Citation Index; Soybean Abstracts (OnlineEdition); Veterinary Bulletin; Wildlife Review Abstracts; World Agricultural Economics and Rural Sociology Abstracts.Reproduction in Domestic Animals is the official organ for the European Society for Domestic Animal Reproduction, the European Veterinary Society ofSmall Animal Reproduction and the Spanish Society of Animal Reproduction.Reproduction in Domestic Animals publishes comprehensive information concerning physiology, pathology and biotechnology of reproduction, paying specialattention to applied and clinical research. The evaluation of papers is made by a team of specialists who decide about acceptance, changes or rejection. For author’sinstructions please visit the journals homepage www.blackwellpublishing.com/rdaResponsible according to Berlin press law: Prof. Dr. H. Rodriguez-Mártinez.Publication Rights Contributions published in this journal are protected by copyright. All rights, whether of the whole or part of the published material are reserved,specifically those of translation, reprinting, re-use of illustrations, reproduction on microfilms or by other means. Duplication of this journal or parts thereof is forbidden.Copying one article or parts thereof is only permitted under the provisions of the German Copyright Law of September 9,1965, in its version of June 24, 1985, and acopyright fee must always be paid. Violations fall under the prosecution act of the German Copyright Law. Legally permitted copies are to be marked with the sourceof the original as well as with the name of the copier.Disclaimer The Publisher, Societies and Editors cannot be held responsible for errors or any consequences arising from the use of information contained in this journal;the views and opinions expressed do not necessarily reflect those of the Publisher, Societies and Editors, neither does the publication of advertisements constituteany endorsement by the Publisher, Societies and Editors of the products advertised.Author material archive policy Please note that unless specifically requested, Blackwell Publishing will dispose of all hardcopy or electronic material submitted twomonths after publication. If you require the return of any material submitted, please inform the editorial office or production editor as soon as possible if you have notyet done so.Copyright and Photocopying Journal compilation © 2008 Blackwell Verlag. All rights reserved. No part of this publication may be reproduced, stored or transmitted inany form or by any means without the prior permission in writing from the copyright holder. Authorization to photocopy items for internal and personal use is granted bythe copyright holder for libraries and other users registered with their local Reproduction Rights Organisation (RRO), e.g. Copyright Clearance Center (CCC), 222Rosewood Drive, Danvers, MA 01923, USA (www.copyright.com), provided the appropriate fee is paid directly to the RRO. This consent does not extend to other kindsof copying such as copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale. Special requests should beaddressed to: journalsrights@oxon.blackwellpublishing.comSubscription information Reproduction in Domestic Animals is published six times a year. Subscription prices for 2008 are: Premium institutional: £594 (Europe),£653 (Rest of World), $1098 (The Americas); Personal: £177 (Europe, non Euro Zone), €266 (Europe, Euro Zone), £195 (Rest of World), $328 (The Americas).Customers in the UK should add VAT at 7%; customers in the EU should also add VAT at 7%, or provide a VAT registration number or evidence of entitlement toexemption. Customers in Canada should add 6% GST or provide evidence of entitlement to exemption. The Premium institutional price includes online access to thecurrent and all available previous year electronic issues. For other pricing options including access information and terms and conditions, please visit www.blackwellpublishing.com/rdaBack issues Single issues from current and recent volumes are available at the current single issue price from Blackwell Publishing Journals. Earlier issues may beobtained from Periodicals Service Company, 11 Main Street, Germantown, NY 12526, USA. Tel: +1 518 537 4700, Fax: +1 518 537 5899, Email: psc@periodicals.com.PublisherReproduction in Domestic Animals is published by Blackwell Verlag GmbH, Kurfürstendamm 58, 10707 Berlin, Germany. Tel: (+49 30) 327 906-65, Fax:(+49 30) 327 906-77. Email: verlag@blackwell.de. Internet homepage: www.blackwell.deBlackwell Verlag is now part of Wiley-BlackwellDelivery Terms and Legal TitlePrices include delivery of print journals to the recipient's address. Delivery terms are Delivered Duty Unpaid (DDU); the recipient is responsible for payingany import duty or taxes. Legal title passes to the customer on despatch by our distributors.Journal Customer Services:For ordering information, claims and any enquiry concerning your journal subscription please contact your nearest office:UK: Email: customerservices@blackwellpublishing.com; Tel: +44 (0) 1865 778315; Fax: +44 (0) 1865 471775USA: Email: customerservices@blackwellpublishing.com; Tel: +1 781 388 8599 or 1 800 835 6770 (Toll free in the USA & Canada); Fax: +1 781 388 8232or Fax: +44 (0) 1865 471775Asia: Email: customerservices@blackwellpublishing.com; Tel: +65 6511 8000; Fax: +44 (0) 1865 471775Production editor Jovencio Laoag Jr. Tel: (+65) 6511 8188, Fax: (+65) 65118288, Email: RDA@oxon.blackwellpublishing.comJournal publication manager Suzanne Albrecht. Tel: (+49 30) 327 906-65, Fax: (+49 30) 327 906-77. Email:suzanne.albrecht@ber.blackwellpublishing.comAccess to this journal is available free online within institutions in the developing world through the AGORA initiative with the FAO and the HINARI initiativewith the WHO. For information, visit www.aginternetwork.org or www.healthinternetwork.org.REPRODUCTION IN DOMESTIC ANIMALS (ISSN 0936-6768) is published six times a year. US mailing agent: Mercury Airfreight International Inc., 365 Blair Road, Avenel,NJ 07001, USA. Periodical postage paid at Rahway, NJ. Postmaster: Send all address changes to REPRODUCTION IN DOMESTIC ANIMALS, Blackwell Publishing Inc.,Journals Subscription Department, 350 Main St., Malden, MA 02148-5020.Information for Authors Please send your manuscript to the Editor-in-Chief Prof. Dr. H. Rodriguez-Mártinez and find instructions for authors on the journal homepagewww.blackwellpublishing.com/rdaOur policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp that is processed using acidfreeand elementary chlorine-free practices. Furthermore, we ensure that the text paper and cover board used in all our journals has met acceptable environmentalaccreditation standards.Note to NIH Grantees Pursuant to NIH mandate, Wiley-Blackwell will post the accepted version of contributions authored by NIH grant-holders to PubMed Centralupon acceptance. This accepted version will be made publicly available 12 months after publication. For further information, see www.wiley.com/go/nihmandate.ISSN 0936–6768 (Print) ISSN 1439-0531 (Online)Printed in Singapore by Markono Print Media Pte Ltd.For submission instructions, subscription and all other information visit: www.blackwellpublishing.com/rdaThis journal is available online at Blackwell Synergy. Visit www.blackwell-synergy.com to search the articles and register for table of contents e-mail alerts.Minitüb Abfüll- und Labortechnik GmbH & Co. KGGold sponsorsIMV TechnologiesIntervet/Schering-PloughSilver sponsorAlltech Hungary Ltd.

Reproductionin Domestic AnimalsVol. 43 • Supplement 2 • July 2008Proceedings of the16th International Congress on AnimalReproduction13–17 July 2008 – Budapest, HungaryGuest Editor:Prof. Gyula HuszeniczaGuest Editorial Board:Dr. Judit BarnaProf. Andras DinnyesDr. Ferenc FlinkProf. Gyorgy GaborProf. Miklos MezesProf. Jozsef RatkyProf. Laszlo SoltiProf. Laszlo Wekerle

Reproductionin Domestic AnimalsTable of Contents Volume 43 · Supplement 2 · July 2008 · 1-422ORIGINAL ARTICLESC. GALLI, G. LAZZARIThe Manipulation of Gametes and Embryos in Farm Animals 1-7D. BLACHE, G. B. MARTIN, S. K. MALONEYTowards Ethically Improved Animal Experimentation in the Study of Animal Reproduction 8-14S. M. RHINDEndocrine Disruptors and Other Food-contaminating Environmental Pollutants as Risk Factors inAnimal Reproduction 15-22J. E. P. SANTOS, T. R. BILBY, W. W. THATCHER, C. R. STAPLES, F. T. SILVESTRELong Chain Fatty Acids of Diet as Factors Influencing Reproduction in Cattle 23-30M. C. LUCYFunctional Differences in the Growth Hormone and Insulin-like Growth Factor Axis in Cattle and Pigs:Implications for Post-partum Nutrition and Reproduction 31-39P. CHEMINEAU, D. GUILLAUME, M. MIGAUD, J. C. THIÉRY, M. T. PELLICER-RUBIO, B. MALPAUXSeasonality of Reproduction in Mammals: Intimate Regulatory Mechanisms and Practical Implications 40-47M. MIHM, A. C. O. EVANSMechanisms for Dominant Follicle Selection in Monovulatory Species: A Comparison of Morphological,Endocrine and Intraovarian Events in Cows, Mares and Women 48-56D. J. SKARZYNSKI, G. FERREIRA-DIAS, K. OKUDARegulation of Luteal Function and Corpus Luteum Regression in Cows: Hormonal Control, ImmuneMechanisms and Intercellular Communication 57-65H. J. BERTSCHINGER, D. G. A. MELTZER, A. VAN DYKCaptive Breeding of Cheetahs in South Africa – 30 Years of Data from the de WildtCheetah and Wildlife Centre 66-73M. DEHNHARD, S. NAIDENKO, A. FRANK, B. BRAUN, F. GÖRITZ, K. JEWGENOWNon-invasive Monitoring of Hormones: A Tool to Improve Reproduction in Captive Breeding of theEurasian Lynx: Hormone Monitoring in Breeding Programmes of Mammals 74-82J. A. LONGReproductive Biotechnology and Gene Mapping: Tools for Conserving Rare Breeds of Livestock 83-88B. BERGLUNDGenetic Improvement of Dairy Cow Reproductive Performance 89-95J. L. M. R. LEROY, T. VANHOLDER, A. T. M. VAN KNEGSEL, I. GARCIA-ISPIERTO, P. E. J. BOLSNutrient Prioritization in Dairy Cows Early Postpartum: Mismatch Between Metabolism and Fertility? 96-103R. S. ROBINSON, A. J. HAMMOND, D. C. WATHES, M. G. HUNTER, G. E. MANNCorpus Luteum-Endometrium-Embryo Interactions in the Dairy Cow: Underlying Mechanisms andClinical Relevance 104-112N. C. FRIGGENS, M. BJERRING, C. RIDDER, S. HØJSGAARD, T. LARSENImproved Detection of Reproductive Status in Dairy Cows Using Milk Progesterone Measurements 113-121D. R. NOTTERGenetic Aspects of Reproduction in Sheep 122-128R. J. SCARAMUZZI, G. B. MARTINThe Importance of Interactions Among Nutritions, Seasonality and Socio-sexual Factors in theDevelopment of Hormone-free Methods for Controlling Fertility 129-136K. M. MORTONDevelopmental Capabilities of Embryos Produced In Vitro from Prepubertal Lamb Oocytes 137-143E. AXNÉRUpdates on Reproductive Physiology, Genital Diseases and Artificial Insemination in the Domestic Cat 144-149H. LINDEBERGReproduction of the Female Ferret (Mustela putorius furo) 150-156J. DE GIER, N. J. BEIJERINK, H. S. KOOISTRA, A. C. OKKENSPhysiology of the Canine Anoestrus and Methods for Manipulation of Its Length 157-164G. C. W. ENGLAND, K. M. MILLARThe Ethics and Role of AI with Fresh and Frozen Semen in Dogs 165-171A. CARATY, I. FRANCESCHINIBasic Aspects of the Control of GnRH and LH Secretions by Kisspeptin: Potential Applications forBetter Control of Fertility in Females 172-178R. FAYRER-HOSKENControlling Animal Populations Using Anti-Fertility Vaccines 179-185T. E. ADAMS, I. BOIMEThe Expanding Role of Recombinant Gonadotropins in Assisted Reproduction 186-192

T. A. L. BREVINI, S. ANTONINI, G. PENNAROSSA, F. GANDOLFIRecent Progress in Embryonic Stem Cell Research and Its Application in Domestic Species 193-199B. M. A. O. PERERAReproduction in Domestic Buffalo 200-206P. S. BRAR, A. S. NANDAPostpartum Ovarian Activity in South Asian Zebu Cattle 207-212M. J. R. PARANHOS DA COSTA, A. SCHMIDEK, L. M. TOLEDOMother-Offspring Interactions in Zebu Cattle 213-216B. S. PRAKASH, M. SARKAR, M. MONDALAn Update on Reproduction in Yak and Mithun 217-223F. X. DONADEU, H. G. PEDERSENFollicle Development in Mares 224-231M. A. HAYES, B. A. QUINN, N. D. KEIRSTEAD, P. KATAVOLOS, R. O. WAELCHLI, K. J. BETTERIDGEProteins Associated With the Early Intrauterine Equine Conceptus 232-237Z. ROTHHeat Stress, the Follicle, and Its Enclosed Oocyte: Mechanisms and Potential Strategies to ImproveFertility in Dairy Cows 238-244K. - P. BRÜSSOW, J. RÁTKY, H. RODRIGUEZ-MARTINEZFertilization and Early Embryonic Development in the Porcine Fallopian Tube 245-251S. PYÖRÄLÄMastitis in Post-Partum Dairy Cows 252-259M. G. DISKIN, D. G. MORRISEmbryonic and Early Foetal Losses in Cattle and Other Ruminants 260-267N. MANABE, F. MATSUDA-MINEHATA, Y. GOTO, A. MAEDA, Y. CHENG, S. NAKAGAWA, N. INOUE, K. WONGPANIT, H. JIN, H. GONDA, J. LIRole of Cell Death Ligand and Receptor System on Regulation of Follicular Atresia in Pig Ovaries 268-272F. F. BARTOL, A. A. WILEY, C. A. BAGNELLEpigenetic Programming of Porcine Endometrial Function and the Lactocrine Hypothesis 273-279K. C. CAIRES, J. A. SCHMIDT, A. P. OLIVER, J. DE AVILA, D. J. MCLEANEndocrine Regulation of the Establishment of Spermatogenesis in Pigs 280-287I. DOBRINSKIMale Germ Cell Transplantation 288-294N. RAWLINGS, A. C. O. EVANS, R. K. CHANDOLIA, E. T. BAGUSexual Maturation in the Bull 295-301A. DINNYES, X. C. TIAN, X. YANGEpigenetic Regulation of Foetal Development in Nuclear Transfer Animal Models 302-309R. C. BOTT, D. T. CLOPTON, A. S. CUPPA Proposed Role for VEGF Isoforms in Sex-Specific Vasculature Development in the Gonad 310-316B. K. WHITLOCK, J. A. DANIEL, R. R. WILBORN, T. H. ELSASSER, J. A. CARROLL, J. L. SARTINComparative Aspects of the Endotoxin- and Cytokine-Induced Endocrine Cascade InfluencingNeuroendocrine Control of Growth and Reproduction in Farm Animals 317-323C. R. BARB, G. J. HAUSMAN, C. A. LENTSEnergy Metabolism and Leptin: Effects on Neuroendocrine Regulation of Reproduction in the Gilt and Sao 324-330C. GALLI, I. LAGUTINA, R. DUCHI, S. COLLEONI, G. LAZZARISomatic Cell Nuclear Transfer in Horses 331-337D. RATH, L. A. JOHNSONApplication and Commercialization of Flow Cytometrically Sex-Sorted Semen 338-346J. M. VAZQUEZ, J. ROCA, M. A. GIL, C. CUELLO, I. PARRILLA, I. CABALLERO, J. L. VAZQUEZ, E. A. MARTÍNEZLow-Dose Insemination in Pigs: Problems and Possibilities 347-354C. B. A. WHITELAW, S. G. LILLICO, T. KINGProduction of Transgenic Farm Animals by Viral Vector-Mediated Gene Transfer 355-358A. C. O. EVANS, N. FORDE, G. M. O'GORMAN, A. E. ZIELAK, P. LONERGAN, T. FAIRUse of Microarray Technology to Profile Gene Expression Patterns Important for Reproduction in Cattle 359-367J. P. KASTELIC, J. C. THUNDATHILBreeding Soundness Evaluation and Semen Analysis for Predicting Bull Fertility 368-373G. C. ALTHOUSESanitary Procedures for the Production of Extended Semen 374-378B. LEBOEUF, J. A. DELGADILLO, E. MANFREDI, A. PIACÈRE, V. CLÉMENT, P. MARTIN, M. PELLICER,P. BOUÉ, R. DE CREMOUXManagement of Goat Reproduction and Insemination for Genetic Improvement in France 379-385N. KOSTEREVA, M. -C. HOFMANNRegulation of the Spermatogonial Stem Cell Niche 386-392P. MERMILLOD, R. DALBIÈS-TRAN, S. UZBEKOVA, A. THÉLIE, J. - M. TRAVERSO, C. PERREAU,P. PAPILLIER, P. MONGETFactors Affecting Oocyte Quality: Who is Driving the Follicle? 393-400K. KIKUCHI, N. KASHIWAZAKI, T. NAGAI, M. NAKAI, T. SOMFAI, J. NOGUCHI, H. KANEKOSelected Aspects of Advanced Porcine Reproductive Technology 401-406B. OBACKClimbing Mount Efficiency – Small Steps, Not Giant Leaps Towards Higher Cloning Success in Farm Animals 407-416P. LOI, K. MATZUKAWA, G. PTAK, Y. NATAN, J. FULKA JR, A. ARAVNuclear Transfer of Freeze-Dried Somatic Cells into Enucleated Sheep Oocytes 417-422

Minitüb:ProductsforArtificial Insemination andEmbryo Transfer

WhitePartners in ReproductionResearch and Practice

Reprod Dom Anim 43 (Suppl. 2), 1–7 (2008); doi: 10.1111/j.1439-0531.2008.01136.xISSN 0936-6768The Manipulation of Gametes and Embryos in Farm AnimalsC Galli 1,2 and G Lazzari 11 Laboratorio di Tecnologie della Riproduzione, Istituto Sperimentale Italiano Lazzaro Spallanzani, CIZ srl, Cremona, Italy; 2 Dip. ClinicoVeterinario, Universita` di Bologna, Bologna, ItalyContentsThis paper summarizes the major advances in farm animalassisted reproduction in the last 20 years with particularattention to the contributions of the authors. A main emphasisis on the biology of the oocyte, including a description ofmethods for isolation of developing follicles and culture of thecorresponding oocytes. Milestones that have led to optimizationof procedures for maturation of fully grown oocytes,fertilization, intracytoplasmic sperm injection and embryoculture in sheep, cattle, pigs and horses are described. Thecurrent status of nuclear transfer, cloning and embryonic stemcell generation and culture is also summarized for all majorfarm animal species. It is concluded that the manipulation ofearly development in farm animals is of crucial importance foragricultural purposes and that reproductive biotechnologies infarm animals are expected to play an increasing role in the nextdecades due to the growing demand for agricultural productsfrom the emerging economies worldwide. In the biomedicalfield large animals represent increasingly important researchmodels especially in the stem cell field for creating geneticallymodified animals for specific purposes. Finally, the successfultranslation of large animal research in the applied contextrequires solid science, long-term resource commitment frominvolved institutions, and vision, dedication and entrepreneurialskills from the scientists involved.IntroductionDespite the many scientific advances that have characterizedreproduction research, the science of reproductionretains an aura of wonder. This feeling is captured inHarvey’s phrase: ‘ex ovo omnia’ and this fundamentaltruth explains the central role that reproduction has inbiology of all life including our own. In livestockbreeding, as well as for companion animals and wildlifeconservation, reproduction and selection allow maintenanceand propagation of species and breeds. Mostimportantly, farm animal reproduction is a centralphenomenon required for genetic selection to improveproductive ⁄ reproductive traits in the next generation.Genomics, proteomics and marker-assisted selectionwould all be limited technologies if not exploited throughreproduction. The development of assisted reproductiontechniques, as advanced tools for animal breeding, hasbeing instrumental for the acquisition of in-depth knowledgeof the molecular mechanisms of the reproductionprocesses, for studying gametes and embryos and forallowing the manipulation of early development in farmanimals. The strong driving force behind the advancementof reproductive technologies has been the need foragricultural and, more recently, biomedical applications.In many instances, the technical achievements havesurpassed the full understanding of the underlyingbiological mechanisms for those achievements.In this paper, we will discuss the in vitro reproductivetechnologies that have had the greatest impact onanimal breeding and highlight our contributions to theirdevelopment. For this reason, it was not the aim of thepaper to make an extensive review of the publishedliterature but rather to mention the work and publicationsthat have directly inspired our work. We apologizein advance for those citations that are not included.At the beginning of our career in science, in the mid-1980s, the only embryo technology that was applied inpractice was superovulation and embryo transfer [multipleovulation embryo transfer (MOET)]. At that time,soon became apparent that the MOET technique hadmany limits and other ways could be developed forproducing embryos required by the cattle industry. Aseminal paper published by Staigmiller and Moor (1984)drew our attention, as it demonstrated that immatureoocytes, after in vitro maturation (IVM), could beconverted into viable embryos and offspring at a level ofefficiency that was remarkably high. Our interest wasparticularly attracted to the female germ line. On theone hand, we dreamed about the possibility of exploitingthe large pool of growing oocytes in the ovaries and,on the other hand, we aimed at rescuing those fullygrown oocytes present in medium to large size folliclesthat exist throughout the reproductive lifetime of femalecattle. The aim was to overcome the physiologicallimitation of conventional breeding, especially in themonotocycous Bos taurus, and make a better use offemale gametes that were otherwise destined to wastageby atresia. In this way, the genetic impact of females ofsuperior genotype and performance on a given populationcould be greatly amplified. Indeed the oocyte is themost interesting cell in biology and may play the mostcrucial role in successful breeding. A significant proportionof successful embryo development can be relatedback to the oocyte (Staigmiller and Moor 1984) and theproduction of large number of good quality oocytesusing in vitro techniques opened a new era in embryotechnologies for farm animals.In this paper, we will review our work of the last20 years. Efforts started with a specific interest in theoocyte and evolved into building novel approaches toembryo-related biotechnologies ranging from control ofmeiosis, in vitro fertilization (IVF), embryo culture,nuclear transfer, stem cells and genetic modification oflarge domestic animals. We always worked in anultrashort feedback loop between fundamental researchand technology application, in close contact with theend users of the technologies that we were working on.Such contacts have served as a strong reminder thatany new technology has to have application. In ourÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

2 C Galli and G Lazzaricircumstances, the final result was the birth of liveoffspring at an acceptable rate, with several differenttechnologies and across a number of different speciesthat included cattle, sheep, horses, pigs and buffalo.From the Very Beginning: The Growing OocyteIn mammals, only a tiny fraction of the large number ofprimordial follicles generates progeny; the vast majorityare lost during through atresia. For many years, severalinvestigators isolated and investigated the biology ofintact primordial follicles in rodents (Roy and Greenwald1985) and eventually grew them successfully inculture (Eppig and O’Brien 1996). This task provedmuch more difficult in large animals because of the largesize and extensive connective tissue of the ovary.Nonetheless, we developed (Lazzari et al. 1992) aprotocol for isolation of large numbers of primordialoocytes from the ovary of newborn piglets based onenzymatic digestion and separation by centrifugal elutriation.Later, oocytes were obtained from secondaryfollicles of cattle and successful development with thebirth of live calves was achieved, albeit at a low rate(Miyano and Manabe 2007). Culture requirements forgrowing a complex tri-dimensional structure such as anovarian follicle are very complex. The only progress thathas been made involves culture in vivo of thin slices ofthe ovary following transplantation in humans (Meirowet al. 2007). This technology has potential applicationsfor patients undergoing certain therapies that causeoocyte damage.The hypothesis that the ovarian reservoir of germ cellsis replenished by haematopoietic stem cells moving intothe ovaries (Johnson et al. 2005) is an intriguing onethat suggests new approaches for developing largenumbers of oocytes. However, no progeny has beenobtained from such migrating cells, and thus the longestablishedview that the number of oocytes is fixed atbirth is true (Eggan et al. 2006). A more realisticapproach for the production of new female gametes isthrough differentiation of embryonic stem (ES) cellsand, indeed, development of oocyte-like cells from EScells has been reported (Hubner et al. 2003; Ko andScholer 2006). In spite of these astonishing results, thecomplexity of epigenetic regulation of germ cell developmentin mammals (Sasaki and Matsui 2008) sets thisambitious target very far off.Oocyte Maturation In Vitro across Species andTimeFrom the late 1970s, it became clear that it was possibleto exploit the oocyte reservoir present in secondaryfollicles of farm animals. Early on, successful developmentwas obtained by maturing the oocyte inside thefollicle (Moor and Trounson 1977). The subsequentunderstanding of the crucial role of the surroundingfollicular ⁄ cumulus cells during the initial phases ofmaturation (Moor et al. 1981) leads to the developmentof a co-culture system for IVM of extrafollicular oocytesthat resulted in a high proportion of such oocytesdeveloping into live lambs after in vivo transfer (Staigmillerand Moor 1984). After the successful achievementof IVM in sheep, the work was expanded byinvestigating the effects of gonadotropins on oocytematuration (Galli and Moor 1991b). In addition, themodel was transferred to the cow at Animal BiotechnologyCambridge Ltd in UK, at Ovamass in Dublin(Lu and Polge 1992; Carolan et al. 1994) and elsewhere(Xu et al. 1987; Goto et al. 1988). Similar developmentsoccurred in the pig (Moor et al. 1990), horse (Zhanget al. 1989) and other species as well. When IVF wasperfected, allowing a deeper understanding of oocytedevelopmental competence, other factors, such as folliclesize, were identified as important indicators of oocytecompetence in the cow (Galli and Moor 1991a) andhorse (Galli et al. 2007). Studies on the protein synthesisactivity within the oocyte and on the interactionsbetween follicular cells and oocytes (Staigmiller andMoor 1984; Downs et al. 1988; Buccione et al. 1990)contributed to the identification of active compoundssecreted by somatic cells (epidermal growth factor,insulin growth factors, fibroblast growth factors, etc.)that promote competent maturation.The development of procedures for oocyte retrievalfrom live bovine donors (Pieterse et al. 1991; Hasleret al. 1995; Galli et al. 2001), referred to as ovum pickup, opened an important area for the application ofoocyte maturation and embryo production in vitro forbreeding purposes in cattle as well as other large speciesas buffalo (Galli et al. 1998) and horses (Galli et al.2002).Despite the high rate of nuclear maturation obtainedin vitro, the developmental competence of maturedoocytes is variable. One likely source for much of thisvariation is the intrinsic quality of the oocytes recoveredfrom the ovaries. To overcome the problem caused byharvesting of suboptimal oocytes, it was tested whethera period of prematuration in vitro to maintain oocytesarrested in prophase I could increase their developmentalcompetence following maturation. Butyrolactoneand roscovitine, two specific protein kinases inhibitors,were used for keeping oocytes arrested in GerminalVesicle (GV) for a 24-h culture period (Ponderato et al.2001). This meiosis block was fully reversible as demonstratedby resumption of meiosis and normal embryonicdevelopment following transfer of embryos derivedfrom prematured oocytes (Ponderato et al. 2002). Yet,this strategy did not improve oocyte competence andfound very limited application.Manipulating the Mature Egg for EmbryoProductionFollowing the seminal work in laboratory animals(Yanagimachi and Chang 1964), IVF was later developedin farm animal species. The birth of the first IVFcalf derived from in vivo matured oocytes (Brackettet al. 1982) was followed by significant advances in IVFwhen heparin was used as capacitating agent for bullsperm (Parrish et al. 1986). At about the same time, IVFbecame a reality in other large species except the horsewhere success has been only exceptional. Indeed, onlytwo foals have been reported as the result of IVF(Palmer et al. 1991). Assisted reproduction of thehorse eventually benefited from the development ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Embryo Biotechnologies in Farm Animals 3intracytoplasmic sperm injection (ICSI) in humans(Palermo et al. 1992). The use of intracytoplasmicsperm injection (ICSI) has been very efficient as a wayto circumvent failure of IVF (Galli et al. 2007). In otherfarm animals like cattle (Goto and Yanagita 1995; Galliet al. 2003c), sheep (Catt et al. 1996) and pigs (Kolbeand Holtz 2000), the efficiency of ICSI for embryoproduction remains lower than for IVF and thereforethe technique is not used routinely. In pigs, IVF ischaracterized by a high incidence of polyspermy thatcompromises embryonic development. This problem hasbeen overcome in part by improving the quality ofin vitro matured oocytes rather than by changing IVFconditions.In recent years, separation of X and Y-bearing spermby flow cytometry is finding wide application in reproductivetechnologies. When combined with IVF or ICSI,a synergy can occur to increase the number of embryosof the desired sex combination (Seidel 2003; Cran 2007)produced by assisted reproductive techniques moreeffectively than older technologies based on sexing ofcells collected by embryo biopsy.Manipulaton of Embryo DevelopmentThe goal of increasing efficiency of in vitro embryoproduction (IVP), especially in cattle, has been thedriving force for much of the applied research in embryobiology and culture; scores of papers have beenpublished on methods for improving the yield of IVPembryos in cattle. Yet, it became soon obvious thatmerely counting the number of blastocysts was not anaccurate measure of the quality of the overall procedureand of the viability of the embryos (Gandolfi and Moor1987). For many years, these constraints on the cultureof viable cattle embryos in vitro were overcome by usingin vivo culture in the oviduct of surrogate sheep (Galliand Lazzari 1996). In particular, a major concern forembryos produced by IVP was the so-called largeoffspring syndrome (LOS), first in sheep and then incattle (Young et al. 1998). The use of serum supplementationand coculture were recognized as the primarycause of LOS. Later, an extensive field study (vanWagtendonk-de Leeuw et al. 2000) demonstrated thatthe incidence of LOS was greatly reduced using a culturemedium based on Synthetic oviduct Fluid (Tervit et al.1972) and the industry took up this method in mostinstances. These observations were accompanied by alarge set of studies looking at the effects of in vitroculture on molecular and cellular markers of embryodevelopment, metabolism, gene expression and viability.Together with others, we investigated the cause of LOSby comparing different in vitro culture systems within vivo culture of bovine embryos in the sheep oviduct.We set up an experiment comparing in vivo culture in thesheep oviduct with in vitro culture in the presence ofserum or high levels of bovine serum albumin (BSA)(Lazzari et al. 2002). Results indicated that embryosgrown in serum but also embryos grown in high levels ofBSA have alterations in the gene expression levels ofseveral developmentally important transcripts. In addition,we demonstrated that these embryos have morecells at the blastocyst stage on day 7 than IVM–IVFembryos developed in the sheep oviduct. Followingtransfer of day 7 blastocysts to recipients and recoveryafter 5 days, we showed that in vitro cultured embryosare larger on day 12 than those derived from sheepoviduct cultured up to day 7. Moreover, when the day12 embryos derived from in vitro culture, werere-transferred to new recipients, they gave rise tooffspring with an average birth weight higher than thecorresponding offspring derived from embryos culturedin the sheep oviduct. These findings clearly indicatedthat in vitro culture can alter development at very earlystages and that the LOS is correlated to abnormallyadvanced embryonic growth and gene expressionpatterns at very early stages.Besides LOS, another important issue that has limitedthe use of in vitro produced embryos is their inconsistentsurvival to freezing and thawing. This limitation wasanother reason why we used for many years the sheepoviduct culture system because it allowed production ofIVM–IVF embryos that were undistinguishable fromin vivo produced embryos in terms of viability andfreezability (Galli and Lazzari 1996; Lonergan and Fair2008). At present, recent improvements in mediaformulation (Gardner et al. 1994) have removed almostentirely the barrier of embryo freezability; pregnancyrate following transfer of frozen-thawed in vitro producedcattle embryos is commercially viable. Importantvariations still exist across laboratories but those thathave really mastered the technology are now wellestablishedcommercial producers of IVP embryos.In vitro produced embryos account for approximatelyone-third of the cattle embryos produced worldwide(http://www.iets.org).A slightly different story has unfolded in the horse.For many years, equine-assisted reproduction laggedbehind other species. This was due in part to the lack ofinterest of the horse industry in developing this technologybut also to a sort of reluctance to really translatequestions that had been asked and answered before inother species. Our work on equine reproduction datesback to the Cambridge period, in the late 1980s, whenwe obtained equine embryos from in vitro maturedoocytes transferred to the oviduct of inseminated mares(Zhang et al. 1989). The most recent work is summarizedin a paper (Galli et al. 2007) where all the relevantsteps have been described including oocyte maturation,fertilization by ICSI, embryo culture and freezing, nonsurgicaltransfer, and somatic cell nuclear transfer.Prometea, the first cloned horse, was derived totallyfrom in vitro procedures that had been optimized step bystep with dedication and perseverance, taking advantageand translating the experience derived from the establishedbovine technology. Today, viable and freezableequine embryos can be produced efficiently by in vitrotechnologies (Colleoni et al. 2007). This success story isexpected to finally open up the horse industry to thesame opportunities that the cattle industry has exploitedfor several years.Nuclear Transfer and CloningThe birth of Dolly the sheep through cloning by nucleartransfer (Wilmut et al. 1997) is an astonishing accom-Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

4 C Galli and G Lazzariplishment that has attracted the attention of the generalpublic towards farm animal embryo-technologies. Thisis one of the best examples where the search of newreproductive techniques in farm animals has providedbasic knowledge not previously obtained in laboratoryanimals. As with many other areas of research, newconcepts and new techniques develop over time throughthe contribution of many investigators. The fundamentalwork on cloning was done in amphybia in the 1960s(Gurdon 2006) followed by significant advancementsachieved with embryo cloning of farm animals (Willadsen1986) when matured oocytes, instead of zygotes,became the recipients of the donor nuclei. Ten yearslater, it was again farm animals that, with somatic cellnuclear transfer, gave a major contribution to the basicunderstanding of cell reprogramming and epigeneticcontrol of mammalian development, and thereby openinga new era in cell biology. Several farm animals werecloned before cloning was demonstrated in laboratoryspecies. Most major domestic mammals have now beencloned albeit at a low efficiency. The success with whichdifferent mammals have been cloned was directlycorrelated to the availability of good quality matureoocytes and good quality embryos either after in vivo orin vitro culture. It was as early as 1990 at Cambridgethat we were generating embryos by nuclear transferwith somatic cells although the two pregnancies thatwere obtained did not develop to term (Galli et al.1991). In those early days, these kinds of experimentswere looked at with high degree of scepticism until itwas, in fact, accomplished (Campbell et al. 1996; Wilmutet al. 1997). Thereafter, it was merely a combinationof already available technologies of gamete andembryo manipulation that lead to success in a variety ofspecies. In the horse, for example, it was not until suchtechnologies were developed that cloning was achieved(Galli et al. 2003b; Lagutina et al. 2005) in a consistentand reproducible way.In the cloning of farm animals, there are significantspecies differences that have not been fully explained.For example, we never observed in cloned horses(Lagutina et al. 2005) and pigs (Lagutina et al. 2006)the abnormalities, mainly associated with placentadysfunction, described in cattle and sheep. From apractical perspective, over the last 10 years, the efficiencyof somatic cell nuclear transfer as a whole has notimproved much and its practical application is limited tothe production of animals that have high added valuelike breeding stock (Galli et al. 1999, 2003a; Lagutinaet al. 2005) or for generating transgenic founder animalswhen nuclear transfer is combined with genetic engineeringof somatic cells (Brophy et al. 2003; Kuroiwaet al. 2004; Wall et al. 2005; Lombardo et al. 2007).An understanding of the constraints of nucleartransfer and development of the concept of reprogrammingof fully differentiated somatic cells and thederivation of stem cells from cloned embryos (Wakayamaet al. 2001; Lazzari et al. 2006) has led to the ideaof reprogramming differentiated cells in vitro directlywithout the need of the oocyte. A set of four genes hasbeen demonstrated to be at the core of pluripotencywhen transfected somatic cells gave rise to inducedpluripotent stem cells (iPS cells) (Takahashi and Yamanaka2006). These cells, of somatic origin, are extraordinarybecause they carry all the properties of ES cellsderived from the early embryo. This research representsa major breakthrough for overcoming the limited oocyteavailability in humans.Understanding the molecular basis of pluripotencywill be a major step forward that is expected to lead toadvances in nuclear transfer technology and cloning offarm animals for agricultural and biomedical applicationson a larger basis.Stem CellsAt this writing, we still await the first definitive report ofthe derivation of farm animal ES cells. Over the years,from 1981, when mouse ES cells were first reported, tothis time, several laboratories and scientists all over theworld, including us, have attempted ES cell derivationfrom cattle, pig and sheep embryos (Galli et al. 1994;Keefer et al. 2007). The original mouse approach hasbeen used extensively and finally proven unsuccessfulalthough a few reports have been published on thederivation of so-called ES-like cells (Notarianni et al.1991). Yet, the stemness (staminality) of these cellsappeared to be very limited and most likely theyrepresent trophoblastic cells given their epithelial nature,loss of OCT4 expression and limited differentiationpotential (Notarianni et al. 1991; Iwasaki et al. 2000;Keefer et al. 2007). A major and often reported problemin ES cell derivation from farm animal embryos is thetendency of the trophoblast and ⁄ or the endoderm toovergrow the inner cell mass cells in culture even aftercareful dissection or isolation by immunosurgery (Keeferet al. 2007). This is a peculiarity of ruminant and pigembryos and reflects the exuberant growth of thetrophoblast and the underlying endoderm that occursin vivo before implantation. The other well-knownproblem is the tendency for neural differentiation,generally followed by apoptosis, that is often observedespecially after plating of inner cell mass cells in serumfreemedia. This tendency, which is a problem, alsorepresents an interesting biological system to exploreearly neurulation events in mammals (Lazzari et al.2006). We showed that bovine inner cell mass cells, bothfrom fertilized and nuclear transfer embryos, can bedifferentiated in an orderly manner towards the neuroectodermgiving rise to neural rosettes resembling thedeveloping neural tube. From these rosettes highlyproliferating neural precursors can be obtained andfrom those a large variety of nervous system derivatives.In light of these reports and observations, it appearsthat the search for ES cells in farm animals shouldabandon the original mouse procedure and pay moreattention to the recent findings for ES derivation inmouse and human. Advances in mouse and, morerecently, in human embryonic stem cell culture havedemonstrated that a number of different culture conditionscan support pluripotency of embryo-derived stemcells. Mouse ES cells can be grown not only in serumsupplementedculture media containing leukaemiainhibitory factor (LIF) and feeders but also in serumfreeand feeder-free culture by the addition of bonemorphogenetic protein (BMP) molecules (Ying et al.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Embryo Biotechnologies in Farm Animals 52003a). This second method has been developed followingthe finding that mouse ES cells grown in a serumsupplementedmedium and LIF can be differentiatedinto neuroectodermal cells by serum and LIF withdrawal(Ying et al. 2003b). The role of BMP moleculesis to counteract and block the induction of neuraldifferentiation and fix the undifferentiated state in aclever balance between conflicting inductive signallingpathways. Other recent protocols, based on the stimulationof the nodal-activin signalling pathways, havebeen shown to maintain the undifferentiated proliferationof human and mouse ES cells (James et al. 2005;Vallier et al. 2005; Brons et al. 2007; Ogawa et al. 2007).A particular attention should be paid to very recentreports of a novel stem cell type derived from mouseembryos and called epi stem cells (EpiSCs) (Brons et al.2007; Tesar et al. 2007). Interestingly, mouse EpiSCshave been shown to be very similar to human ES cells inmorphology, growth factor requirement and geneexpression (Brons et al. 2007; Tesar et al. 2007) whilemouse ES cells differ considerably from human ES inculture requirements for the maintenance of the undifferentiatedstate, growth rate and response to inductivesignals. Another important difference between mouseES cells and EpiSCs is the fact that only the formerare capable of giving rise to chimeric offspring followingblastocyst injection. Under the present status ofresearch, epiSC derivation in farm animals may proveeasier than true ES cell derivation and could representa significant step forward towards the understandingof the requirements for maintaining pluripotencyin cultured inner cells mass cells of farm animalembryos.From an applied perspective, embryonic stem cells infarm animals are important for several reasons but themost relevant is to provide a method to introduceprecise genetic modification into animals by homologousrecombination of ES cells (Lombardo et al. 2007)followed by blastocyst injection for chimera derivationand breeding, or by somatic cell nuclear transfer. Asecond important objective is to provide large animalmodels in which the ES cell technology can be tested fortissue-specific differentiation (Brown et al. 2007) and celltherapy of various tissues and organs. Conventional EScells, carrying all the properties of mouse ES cells, canserve both purposes but epiSCs can serve at least thesecond. Therefore, ES and epiSC research still deservesmajor research efforts and represents the frontier ofreproductive technologies in farm animals both foragriculture and biotechnological applications.ConclusionsThe manipulation of early embryonic development infarm animals is of foremost importance both foragriculture and biotechnology applications but also asa model to study some basic aspects of reproduction anddevelopmental biology. In the agricultural context,reproductive biotechnologies in farm animals areexpected to play an increasing role in the next decadesdue to the growing demand for agricultural productsfrom the emerging economies worldwide (see ‘TheEconomist’, The end of cheap food, 7–14 December2007). Modern reproductive biotechnologies, togetherwith the most recent molecular techniques of genomics,proteomics and transgenesis are the powerful tools thatwill provide the answers to the rapidly changing worldscenario of food demand. In the biomedical field, largeanimals represent increasingly important research modelsespecially in the stem cell field, for creating diseasemodels and genetically modified animals as potentialdonors of tissues and organs for xenotransplantation.These are the reasons why large animal research, thatshort-sighted worldwide research policies have neglectedin recent years, will regain a leading role in the comingfuture.Finally, successful translation of large animal researchinto commercial enterprise requires solid science, longtermresource commitment, and extensive steps ofvalidation to reach the thresholds of reproducibilityand profitability. Therefore, strong scientific drive,vision and entrepreneurial skills are all needed forcontributing to progress in large animal reproductivesciences.AcknowledgementsA successful career in animal reproduction is the result of a full-timetotal dedication to the work but also of good training, a stimulatingenvironment and interaction with peers around the world. For thisreason, we want to acknowledge the many colleagues and collaboratorswhose support over the years has been instrumental to success inour work. In particular, we want to acknowledge Professor A Lauria,who during our years as undergraduate students transmitted to us hisenthusiasm for animal reproduction, and especially Dr. Robert Moor,our mentor for many years, who gave us vision and inspired most ofthe work that we have been doing in the last 20 years and we willprobably be doing for the next 20 years.ReferencesBrackett BG, Bousquet D, Boice ML, Donawick WJ, EvansJF, Dressel MA, 1982: Normal development followingin vitro fertilization in the cow. Biol Reprod 27, 147–158.Brons IG, Smithers LE, Trotter MW, Rugg-Gunn P, Sun B,Chuva de Sousa Lopes SM, Howlett SK, Clarkson A,Ahrlund-Richter L, Pedersen RA, Vallier L, 2007: Derivationof pluripotent epiblast stem cells from mammalianembryos. Nature 448, 191–195.Brophy B, Smolenski G, Wheeler T, Wells D, L’Huillier P,Laible G, 2003: Cloned transgenic cattle produce milk withhigher levels of beta-casein and kappa-casein. Nat Biotechnol21, 157–162.Brown BD, Gentner B, Cantore A, Colleoni S, Amendola M,Zingale A, Baccarini A, Lazzari G, Galli C, Naldini L, 2007:Endogenous microRNA can be broadly exploited to regulatetransgene expression according to tissue, lineage anddifferentiation state. Nat Biotechnol 25, 1457–1467.Buccione R, Schroeder AC, Eppig JJ, 1990: Interactionsbetween somatic cells and germ cells throughout mammalianoogenesis. Biol Reprod 43, 543–547.Campbell KH, McWhir J, Ritchie WA, Wilmut I, 1996: Sheepcloned by nuclear transfer from a cultured cell line. Nature380, 64–66.Carolan C, Monaghan P, Gallagher M, Gordon I, 1994: Effectof recovery method on yield of bovine oocytes per ovary andtheir developmental competence after maturation, fertilizationand culture in vitro. Theriogenology 41, 1061–1068.Catt SL, Catt JW, Gomez MC, Maxwell WM, Evans G, 1996:Birth of a male lamb derived from an in vitro maturedÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

6 C Galli and G Lazzarioocyte fertilised by intracytoplasmic injection of a singlepresumptive male sperm. Vet Rec 139, 494–495.Colleoni S, Barbacini S, Necchi D, Duchi R, Lazzari G, GalliC, 2007: Application of ovum pick-up, intracytoplasmicsperm injection and embryo culture in equine practice. ProcAm Assoc Equine Pract (AAEP) 53, 554–559.Cran DG, 2007: XY sperm separation and use in artificialinsemination and other ARTs. Soc Reprod Fertil Suppl 65,475–491.Downs SM, Daniel SA, Eppig JJ, 1988: Induction of maturationin cumulus cell-enclosed mouse oocytes by folliclestimulatinghormone and epidermal growth factor: evidencefor a positive stimulus of somatic cell origin. J Exp Zool 245,86–96.Eggan K, Jurga S, Gosden R, Min IM, Wagers AJ, 2006:Ovulated oocytes in adult mice derive from non-circulatinggerm cells. Nature 441, 1109–1114.Eppig JJ, O’Brien MJ, 1996: Development in vitro of mouseoocytes from primordial follicles. Biol Reprod 54, 197–207.Galli C, Lazzari G, 1996: Practical aspects of IVM ⁄ IVF incattle. Anim Reprod Sci 42, 371–379.Galli C, Moor RM, 1991a: Development of immature bovineoocytes into viable embryos in vitro. Bull l’AssociationAnatom 228, 67–71.Galli C, Moor RM, 1991b: Gonadotrophin requirements forthe in vitro maturation of sheep oocytes and their subsequentembryonic development. Theriogenology 35, 1083–1093.Galli C, Laurie S, Lazzari G, Moor RM, 1991: Nucleartransplantation (by electrofusion) of cultured embryoniccells with Met II cytoplasts in the sheep. Annu Conf Ital SocVet Sci XLV, 299–303.Galli C, Lazzari G, Flechon JE, Moor RM, 1994: Embryonicstem cells in farm animals. Zygote 2, 385–389.Galli C, Duchi R, Crotti G, Lazzari G, 1998: Embryoproduction by Ovum Pick Up in Water Buffalo. Theriogenol49, 400.Galli C, Duchi R, Moor RM, Lazzari G, 1999: Mammalianleukocytes contain all the genetic information necessary forthe development of a new individual. Cloning 1, 161–170.Galli C, Crotti G, Notari C, Turini P, Duchi R, Lazzari G,2001: Embryo production by ovum pick up from livedonors. Theriogenology 55, 1341–1357.Galli C, Crotti G, Turini P, Duchi R, Mari G, Zavaglia G,Duchamp G, Daels P, Lazzari G, 2002: Frozen- thawedembryos produced by Ovum Pick Up of immature oocytesand ICSI are capable to establish pregnancies in the horse.Theriogenology 58, 705–708.Galli C, Vassiliev I, Lagutina I, Galli A, Lazzari G, 2003a:Bovine embryo development following ICSI: effect ofactivation, sperm capacitation and pre-treatment with dithiothreitol.Theriogenology 60, 1467–1480.Galli C, Duchi R, Crotti G, Turini P, Ponderato N, Colleoni S,Lagutina I, Lazzari G, 2003b: Bovine embryo technologies.Theriogenology 59, 599–616.Galli C, Lagutina I, Crotti G, Colleoni S, Turini P, PonderatoN, Duchi R, Lazzari G, 2003c: Pregnancy: a cloned horseborn to its dam twin. Nature 424, 635.Galli C, Colleoni S, Duchi R, Lagutina I, Lazzari G, 2007:Developmental competence of equine oocytes and embryosobtained by in vitro procedures ranging from in vitromaturation and ICSI to embryo culture, cryopreservationand somatic cell nuclear transfer. Anim Reprod Sci 98, 39–55.Gandolfi F, Moor RM, 1987: Stimulation of early embryonicdevelopment in the sheep by co-culture with oviductepithelial cells. J Reprod Fertil 81, 23–28.Gardner DK, Lane M, Spitzer A, Batt PA, 1994: Enhancedrates of cleavage and development for sheep zygotescultured to the blastocyst stage in vitro in the absence ofserum and somatic cells: amino acids, vitamins, and culturingembryos in groups stimulate development. Biol Reprod50, 390–400.Goto K, Yanagita K, 1995: Normality of calves obtained byintracytoplasmic sperm injection. Hum Reprod 10, 1554.Goto K, Kajihara Y, Kosaka S, Koba M, Nakanishi Y,Ogawa K, 1988: Pregnancies after co-culture of cumuluscells with bovine embryos derived from in-vitro fertilizationof in-vitro matured follicular oocytes. J Reprod Fertil 83,753–758.Gurdon JB, 2006: From nuclear transfer to nuclear reprogramming:the reversal of cell differentiation. Annu Rev CellDev Biol 22, 1–22.Hasler JF, Henderson WB, Hurtgen PJ, Jin ZQ, McCauleyAD, Mower SA, Neely B, Shuey LS, Stokes JE, TrimmerSA, 1995: Production, freezing and transfer of bovine IVFembryos and subsequent calving results. Theriogenology 43,141–152.Hubner K, Fuhrmann G, Christenson LK, Kehler J, ReinboldR, De La Fuente R, Wood J, Strauss JF III, Boiani M,Scholer HR, 2003: Derivation of oocytes from mouseembryonic stem cells. Science 300, 1251–1256.Iwasaki S, Campbell KH, Galli C, Akiyama K, 2000:Production of live calves derived from embryonic stem-likecells aggregated with tetraploid embryos. Biol Reprod 62,470–475.James D, Levine AJ, Besser D, Hemmati-Brivanlou A, 2005:TGFbeta ⁄ activin ⁄ nodal signaling is necessary for the maintenanceof pluripotency in human embryonic stem cells.Development 132, 1273–1282.Johnson J, Bagley J, Skaznik-Wikiel M, Lee HJ, Adams GB,Niikura Y, Tschudy KS, Tilly JC, Cortes ML, Forkert R,Spitzer T, Iacomini J, Scadden DT, Tilly JL, 2005: Oocytegeneration in adult mammalian ovaries by putative germcells in bone marrow and peripheral blood. Cell 122, 303–315.Keefer CL, Pant D, Blomberg L, Talbot NC, 2007: Challengesand prospects for the establishment of embryonic stem celllines of domesticated ungulates. Anim Reprod Sci 98, 147–168.Ko K, Scholer HR, 2006: Embryonic stem cells as a potentialsource of gametes. Semin Reprod Med 24, 322–329.Kolbe T, Holtz W, 2000: Birth of a piglet derived from anoocyte fertilized by intracytoplasmic sperm injection (ICSI).Anim Reprod Sci 64, 97–101.Kuroiwa Y, Kasinathan P, Matsushita H, Sathiyaselan J,Sullivan EJ, Kakitani M, Tomizuka K, Ishida I, Robl JM,2004: Sequential targeting of the genes encoding immunoglobulin-muand prion protein in cattle. Nat Genet 36, 775–780.Lagutina I, Lazzari G, Duchi R, Colleoni S, Ponderato N,Turini P, Crotti G, Galli C., 2005: Somatic cell nucleartransfer in horses: effect of oocyte morphology, embryoreconstruction method and donor cell type. Reproduction130, 559–567.Lagutina I, Lazzari G, Galli C, 2006: Birth of cloned pigs fromzona-free nuclear transfer blastocysts developed in vitrobefore transfer. Cloning Stem Cells 8, 283–293.Lazzari G, Galli C, Moor RM, 1992: Centrifugal elutriation ofporcine oocytes isolated from the ovaries of newborn piglets.Anal Biochem 200, 31–35.Lazzari G, Wrenzycki C, Herrmann D, Duchi R, Kruip T,Niemann H, Galli C, 2002: Cellular and molecular deviationsin bovine in vitro-produced embryos are related to thelarge offspring syndrome. Biol Reprod 67, 767–775.Lazzari G, Colleoni S, Giannelli SG, Brunetti D, Colombo E,Lagutina I, Galli C, Broccoli V, 2006: Direct derivation ofneural rosettes from cloned bovine blastocysts: a model ofearly neurulation events and neural crest specification invitro. Stem Cells 24, 2514–2521.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Embryo Biotechnologies in Farm Animals 7Lombardo A, Genovese P, Beausejour CM, Colleoni S, LeeYL, Kim KA, Ando D, Urnov FD, Galli C, Gregory PD,Holmes MC, Naldini L, 2007: Gene editing in human stemcells using zinc finger nucleases and integrase-defectivelentiviral vector delivery. Nat Biotechnol 25, 1298–1306.Lonergan P, Fair T, 2008: In vitro-produced bovine embryos:dealing with the warts. Theriogenology 69, 17–22.Lu K, Polge C, 1992: A summary of two-year’s results in largescale in vitro bovine embryo production. Proceedings of the12th International Congress on Animal Reproduction, TheHague, pp. 1315–1317.Meirow D, Levron J, Eldar-Geva T, Hardan I, Fridman E,Yemini Z, Dor J, 2007: Monitoring the ovaries afterautotransplantation of cryopreserved ovarian tissue: endocrinestudies, in vitro fertilization cycles, and live birth. FertilSteril 87, 418 e417–418 e415.Miyano T, Manabe N, 2007: Oocyte growth and acquisition ofmeiotic competence. Soc Reprod Fertil Suppl 63, 531–538.Moor RM, Trounson AO, 1977: Hormonal and follicularfactors affecting maturation of sheep oocytes in vitro andtheir subsequent developmental capacity. J Reprod Fertil49, 101–109.Moor RM, Osborn JC, Cran DG, Walters DE, 1981: Selectiveeffect of gonadotrophins on cell coupling, nuclear maturationand protein synthesis in mammalian oocytes. J EmbryolExp Morphol 61, 347–365.Moor RM, Mattioli M, Ding J, Nagai T, 1990: Maturation ofpig oocytes in vivo and in vitro. J Reprod Fertil Suppl 40,197–210.Notarianni E, Galli C, Laurie S, Moor RM, Evans MJ, 1991:Derivation of pluripotent, embryonic cell lines from the pigand sheep. J Reprod Fertil Suppl 43, 255–260.Ogawa K, Saito A, Matsui H, Suzuki H, Ohtsuka S,Shimosato D, Morishita Y, Watabe T, Niwa H, MiyazonoK, 2007: Activin-Nodal signaling is involved in propagationof mouse embryonic stem cells. J Cell Sci 120, 55–65.Palermo G, Joris H, Devroey P, Van Steirteghem AC, 1992:Pregnancies after intracytoplasmic injection of single spermatozooninto an oocyte. Lancet 340, 17–18.Palmer E, Bezard J, Magistrini M, Duchamp G, 1991: In vitrofertilization in the horse. A retrospective study. J ReprodFertil Suppl 44, 375–384.Parrish JJ, Susko-Parrish JL, Leibfried-Rutledge ML, CritserES, Eyestone WH, First NL, 1986: Bovine in vitro fertilizationwith frozen-thawed semen. Theriogenology 25, 591–600.Pieterse MC, Vos PL, Kruip TA, Wurth YA, van Beneden TH,Willemse AH, Taverne MA, 1991: Transvaginal ultrasoundguided follicular aspiration of bovine oocytes. Theriogenology35, 857–862.Ponderato N, Lagutina I, Crotti G, Turini P, Galli C, LazzariG, 2001: Bovine oocytes treated prior to in vitro maturationwith a combination of butyrolactone I and roscovitine atlow doses maintain a normal developmental capacity. MolReprod Dev 60, 579–585.Ponderato N, Crotti G, Turini P, Duchi R, Galli C, Lazzari G,2002: Embryonic and foetal development of bovine oocytestreated with a combination of butyrolactone I and roscovitinein an enriched medium prior to IVM and IVF. MolReprod Dev 62, 513–518.Roy SK, Greenwald GS, 1985: An enzymatic method fordissociation of intact follicles from the hamster ovary:histological and quantitative aspects. Biol Reprod 32, 203–215.Sasaki H, Matsui Y, 2008: Epigenetic events in mammaliangerm-cell development: reprogramming and beyond. NatRev Genet 9, 129–140.Seidel GE Jr, 2003: Sexing mammalian sperm – intertwining ofcommerce, technology, and biology. Anim Reprod Sci 79,145–156.Staigmiller R, Moor R, 1984: Effect of follicle cells on thematuration and developmental competence of ovine oocytesmatured outside the follicle. Gamete Res 9, 221–229.Takahashi K, Yamanaka S, 2006: Induction of pluripotentstem cells from mouse embryonic and adult fibroblastcultures by defined factors. Cell 126, 663–676.Tervit HR, Whittingham DG, Rowson LE, 1972: Successfulculture in vitro of sheep and cattle ova. J Reprod Fertil 30,493–497.Tesar PJ, Chenoweth JG, Brook FA, Davies TJ, Evans EP,Mack DL, Gardner RL, McKay RD, 2007: New cell linesfrom mouse epiblast share defining features with humanembryonic stem cells. Nature 448, 196–199.Vallier L, Alexander M, Pedersen RA, 2005: Activin ⁄ Nodaland FGF pathways cooperate to maintain pluripotency ofhuman embryonic stem cells. J Cell Sci 118, 4495–4509.van Wagtendonk-de Leeuw AM, Mullaart E, de Roos AP,Merton JS, den Daas JH, Kemp B, de Ruigh L, 2000: Effectsof different reproduction techniques: AI MOET or IVP, onhealth and welfare of bovine offspring. Theriogenology 53,575–597.Wakayama T, Tabar V, Rodriguez I, Perry AC, Studer L,Mombaerts P, 2001: Differentiation of embryonic stem celllines generated from adult somatic cells by nuclear transfer.Science 292, 740–743.Wall RJ, Powell AM, Paape MJ, Kerr DE, Bannerman DD,Pursel VG, Wells KD, Talbot N, Hawk HW, 2005:Genetically enhanced cows resist intramammary Staphylococcusaureus infection. Nat Biotechnol 23, 445–451.Willadsen SM, 1986: Nuclear transplantation in sheepembryos. Nature 320, 63–65.Wilmut I, Schnieke AE, McWhir J, Kind AJ, Campbell KH,1997: Viable offspring derived from fetal and adult mammaliancells. Nature 385, 810–813.Xu KP, Greve T, Callesen H, Hyttel P, 1987: Pregnancyresulting from cattle oocytes matured and fertilized in vitro.J Reprod Fertil 81, 501–504.Yanagimachi R, Chang MC, 1964: In Vitro fertilization ofGolden Hamster Ova. J Exp Zool 156, 361–375.Ying QL, Nichols J, Chambers I, Smith A, 2003a: BMPinduction of Id proteins suppresses differentiation andsustains embryonic stem cell self-renewal in collaborationwith STAT3. Cell 115, 281–292.Ying QL, Stavridis M, Griffiths D, Li M, Smith A, 2003b:Conversion of embryonic stem cells into neuroectodermalprecursors in adherent monoculture. Nat Biotechnol 21,183–186.Young LE, Sinclair KD, Wilmut I, 1998: Large offspringsyndrome in cattle and sheep. Rev Reprod 3, 155–163.Zhang JJ, Boyle MS, Allen WR, Galli C, 1989: Recent studieson in vivo fertilization of in vitro matured horse oocytes.Equine Veterin J 8, 101–104.Author’s address (for correspondence): C Galli, Laboratorio diTecnologie della, Riproduzione, Via Porcellasco 7 ⁄ f, 26100 Cremona,Italy. E-mail: cesaregalli@ltrciz.itConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 8–14 (2008); doi: 10.1111/j.1439-0531.2008.01137.xISSN 0936-6768Towards Ethically Improved Animal Experimentation in the Study of AnimalReproductionD Blache 1 , GB Martin 1 and SK Maloney 21 UWA Institute of Agriculture M082, The University of Western Australia; 2 Physiology-Biomedical, Biomolecular and Chemical Sciences,The University of Western Australia, Crawley, WA, AustraliaContentsThe ethics of animal-based research is a continuing area ofdebate, but ethical research protocols do not prevent scientificprogress. In this paper, we argue that our current knowledge ofthe factors that affect reproductive processes providesresearchers with a solid foundation upon which they canconduct more ethical research and simultaneously producedata of higher quality. We support this argument by showinghow a deep understanding of the genetics, nutrition andtemperament of our experimental animals can improve compliancewith two of the ‘3 Rs’, reduction and refinement,simply by offering better control over the variance in ourexperimental model. The outcome is a better experimentaldesign, on both ethical and scientific grounds.IntroductionReproduction has a long history as a field of researchand, if anything, there is actually an acceleration ofinterest in the main goals of better reproductivetechnology and better fundamental understanding ofthe biological processes. There is little to distinguishbetween these goals because reproductive technologiesare frequently essential as tools for basic research, andbasic research underpins reproductive technology. Indealing with the topic of this paper, we will illustrate ourarguments primarily with examples for farmed ruminants,particularly sheep. We will not dwell on distinctionsbetween applied and fundamental research despitethe focus on production animals.Over the last 200 years, we have seen an increase inthe concerns of society about the ethics of animal-basedresearch, particularly the potential suffering of experimentalanimals. There has been intense discussion of theethical dilemmas arising generally from the use ofanimals and this has led to a utilitarian framework,the cost ⁄ benefit analysis, for ethical assessment anddecision-making. Scientists have no trouble arguing thebenefits of animal experimentation and augmenting thedenominator of a cost ⁄ benefit analysis. However froman animal welfare viewpoint, the best way to decreasethe cost ⁄ benefit ratio is to decrease the numerator, i.e. todecrease the cost of experimentation to the subjectanimals.In this paper, we are not aiming to provide a recipefor improving the welfare of animals used in research,but to show that ‘ethical science is good science’(Somerville 2007). We will argue that, in the field ofreproduction, our current knowledge of the factors thataffect reproductive processes provides researchers with asolid foundation upon which they can comply withethical guidelines and simultaneously produce data ofhigher quality. This paper is therefore not a manifestofor animal rights or animal liberation. We will onlydiscuss ethics and how to be ethical whilst doing betterscience. We will begin by defining ethics and discussingwhy it has become more important and relevant inanimal research. We will briefly review the ‘Principle of3 Rs’ and how it should be applied to research inreproduction, for which we will need to summarize ourknowledge of the complex of processes that control thereproductive system. Finally, we will describe thepotential for improving the ethics of research in reproductionby considering the models used, particularly thegenetics, nutrition and temperament of the experimentalanimals. We will illustrate most of our points usinginformation about sheep because it is the focus of ourstudies, but the principles are broad so can be applied tomost species.Ethics and Animal-based Research?Everyday we make decisions about what we are willingor not willing to do. Most of such decisions are purelypersonal in that they do not impact on others. When ourdecisions impact on others, they are reasoned andassessed, at least in part, on moral grounds (LaFolette2007). When we plan experiments using animals, ourdecisions cannot be based on personal choice because, asa society, we have decided that animals are not objects,so we need to engage in moral debate to decide whetheran experiment is worth doing. Ethics is a philosophicalprocess involving critical evaluation of our actions andleads to moral truth (morally right or wrong; Bryantet al. 2005; Mepham 2005), so it deals with the moraldimension of decision-making.A major point to consider with respect to ethics is thatmoral decisions are taken in reference to ethical theories.Why? Because ethical theories, such as consequentialismor deontology, provide ground rules for thinking aboutmoral issues. Ethical theory, like any theory, provides aframework for analysing a question or fact critically.Animal research is mainly discussed in the context of theutilitarian theory, deontology being only used sparsely.The utilitarian theory, in short, leads to an assessment ofthe ethical value of a given experiment based oncost ⁄ benefit analysis (Vallentyne 2007). It is not alwayseasy to estimate accurately both the cost and the benefit,especially when considering experiments outside thelaboratory or with long-term consequences (Schmidtz2007). However, the benefits of animal experimentationcan be, for example, societal (medical breakthrough),scientific (new knowledge) or educational (training ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Ethical Models for Studying Reproduction 9research students), and they may even benefit theanimals themselves (new cures, better managementpractices). The cost side of the equation includesimpositions made on animals for the purpose of theexperimentation and can include the pain of procedures,lack of environmental stimulation (e.g. lack of contactwith conspecifics), or food restriction. Final assessmentof the cost–benefit ratio is fairly subjective becausesuffering is not easily quantified and the benefits areusually a ‘best-guess’. However, any attempt to decreasethe cost to the animals for a given benefit will increasethe ethical value of an experiment. A good way todecrease the cost is to adopt more ‘humane’ approachesto animal experimentation. In this context, the ‘Principleof 3 Rs’ (3Rs) originally proposed by Charles Hume, thefounder of the University Federation for AnimalWelfare (UFAW) in 1954, and later defined in greatdetail in the seminal book by Russell and Burch (1959),a valuable resource. In brief, the 3 Rs’ areReplacement – substitution of insentient material forconscious animals;Reduction – get the same amount of information fromfewer animals;Refinement – decrease the imposition on the animalsused in the research.The 3Rs is widely used by animal ethics committeesaround the world to assess animal-based research,including experimentation in the field of reproductivebiology. Replacement is seen as the ultimate ethicalsolution, but that does not mean that cell culture andmathematical models should replace live animal experimentation.Many theories can ultimately only be testedin whole organisms, as is very well demonstrated inresearch for drug discovery, even if some drugs tested onanimals turn out to be dangerous to humans (Festingand Altman 2002). In fact, cell lines, such as immortalizedGnRH neurons, have been intensively used to studythe regulation of the reproductive function (Wetsel1995).Knowledge in Reproductive Physiology: A Curseor a Blessing?Regardless of the level of the investigation of thereproductive system, from cell physiology to behaviour,it is important for the researcher to be cognizant of themultiplicity of factors that influence the reproductiveprocess, as illustrated in Fig. 1. The complexity of theinputs is both a curse because perfect experimentaldesign requires the control of all of these factors, and ablessing because a deep knowledge of how these factorsinfluence experimental outcomes gives us a strong graspof the factors that we need to control in our experimentaldesign. In accounting for all factors that affectreproduction, we account for more variability and thepower and precision of our experiments is increased,improving the ethical outcomes of refinement andreduction (Howard 2002). Some factors are easier tocontrol than others, especially when large groups ofanimals are required. Importantly, many of these factorsare not independent so interactions among them (e.g.nutrition, photoperiod and genotype; Blache et al. 2006,2007) can lead to complex outcomes but, because theyPhotoperiodicsignalsPinealmelatoninGenotypeAnnualrhythmgeneratorOestrogenAromataseNegativefeedback‘Emotional reactivity’AppetitePhotoperiod-driven filterNutritionalsignalsMetabolicSensor“Pulse generator”GnRH pulsesLH, FSHSex steroidsSexual behaviourSocio-sexualsignalsOestrogenAromataseNegativefeedbackFig. 1. A schema describing the proposed relationships betweenemotional reactivity and external inputs (photoperiod, nutrition andsocial signals) and the ways that they interact with genotype andsteroid feedback in the control of hypothalamo–pituitary–testicularaxis in the male sheep. The interactions indicated with broken lines areeither hypothetical or demonstrated (see text) but the nature of thesepathways is not fully understood in ruminants. Redrawn after Blacheet al. (2007)have been extensively studied, we have excellent opportunitiesto apply the 3Rs. Below, we will illustrate thispoint by showing how controlling for genotype, nutritionand temperament can lead to more ethical animalexperimentation.Genetics and ethical experimentationDuring the last decade, there has been a dramaticincrease in the use of genetically modified animals, suchas conventional inbred lines, or lines carrying deletedgenes (knock-out), additional genes (knock-in), or overexpressedgenes (transgenic). In the UK, the number ofgenetically modified animals used in research more thanquadrupled between 1995 and 2006 and are they nowused in 34% of all procedures (Home Office 2007). Theyare considered better experimental models than theirwild-type counterparts because they exhibit less variabilityand ⁄ or they have a modification that allowsexperiments to target a specific function, gene ormolecule (Festing 1990; Festing and Altman 2002). Inreproduction, the knock-out rodent models have confirmedand informed our understanding of, for example,the role of sex steroid hormones in many of the stages ofthe reproductive process. This has been clearly demonstratedfor the oestrogen receptor in females and theandrogen receptor in males (De Gendt et al. 2004;Hewitt et al. 2005). Similarly, animals with an engineereddeficiency in FSH and LH production, or theirreceptors, or in metabolic hormone receptors, have alsocontributed to advances in our knowledge of reproductiveprocesses and how those processes interact withother functions and systems (Kumar 2005). However,these models cannot completely substitute for ‘geneticallyintact’ animals because of the plasticity,Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

10 D Blache, GB Martin and SK Maloneyadaptability and redundancy of physiological systems,and because most systems are polygenic. In addition, inethical terms, the creation of knock-outs is costly for theanimals themselves because a proportion of geneticallymodified animals will not be viable and will have to bedestroyed before being part of an experiment (NuffieldCouncil on Bioethics 2005; NHMRC Animal WelfareCommittee 2007). Estimating the relative ratio betweencost and benefit for the use of genetically modifiedanimals is not an easy task and, to our knowledge, hasnot been addressed, as recognized in a recent reviewabout the use of transgenic farm animals to improveproduction (Bacci 2007).A different strategy for reducing inter-individualvariability, and thus addressing refinement and reduction,is the use of monozygotic twins (Biggers 1986).Twin studies have been used successfully in fields ofresearch such as health and immunology, but it does notseem to be as advantageous in reproductive biology. Insheep, for example, we have studied gonadotrophinresponses following an increase in nutrition, ovarianresponses to exogenous inhibin and the responses ofhypothalamo–pituitary axis to an opioid. For mostreproductive traits (ovulation rate, secretion of gonadotrophinsand sex steroids), randomly selected animalswere just as efficient for experimentation as geneticallyidentical twins (Celi et al. 2007). Monozygotic twinsshowed an advantage only for live weight and scrotalcircumference. In addition, the reproductive biotechnologiesthemselves need to address ethical questionsconcerning the overall imposition on animals neededto gain the result (Church 1988). Again, there is nodirect estimate of the reduction in the use of animalsfollowing cloning. On the contrary, the variancebetween clones in both behavioural and physiologicalparameters can be increased because of epigenetic andearly-life effects (Seidel 2001; Archer et al. 2003a,b).Nutritional input into reproduction and ethicalexperimentationThe relationship between the level of nutrition or, moregenerally, metabolic status and reproductive capacity isevident for all stages in the reproductive process:puberty, gamete production, conception and the survivalof embryos, fetuses and newborn (review: Robinsonet al. 2006). The responses can be rapid, with anacute increase in nutrient supply inducing multipleovulation in females (Vinoles et al. 2005) and anincrease in GnRH output in adult males (Zhang et al.2004). The responses can also be delayed in time, such asthe effect of peri-conception under-nutrition on thelength of gestation (Bloomfield et al. 2003). In addition,nutrition of the mother can affect reproductive processesin her offspring (‘fetal programming’), such as thenumber of Sertoli cells (Bielli et al. 2002) and length ofgestation and ovulation rate (Rae et al. 2001). Foetalprogramming is a relatively new field of investigationand many of the impacts of nutritional imbalance inutero on the reproductive physiology of the offspring arenot yet known (Rhind 2004). Thus, their potentialimpact on experimental design in, for example, a studydone with randomly selected animals, is currentlyimpossible to assess, although it is most likely that theyincrease between-animal variation.Therefore, if we are to ensure that the individuals usedin an experiment are homogenous, we need to considerthe metabolic status of not only the experimentalanimals but also their ancestors. Obviously, this phenomenonmight explain discrepancies among the outcomesof comparable experiments. We have beenconfronted with this problem in our own research intothe brain sites where steroids exert negative feedback onGnRH pulse frequency in mature Merino rams (Blacheet al. 1997). During a preliminary trial, done in WesternAustralia, implantation of oestradiol into the ventromedialnucleus of the hypothalamus decreased thefrequency of LH pulses. We then decided to repeat theexperiment with a larger number animals for statisticalvalidity, but in our collaborator’s laboratory in easternAustralia. The same treatment in the same brainlocation had no effect. A few months after this apparentfailure, we repeated the experiment a third time in thewest, and obtained the same results as in the first study.The contrary results could not be attributed to photoperiodor genotype. The answer came from anotherstudy, done at the same time, in which we demonstratedthat insulin was a very powerful stimulator of GnRHneuronal activity (Miller et al. 1995). When we measuredinsulin in samples from the steroid implantationstudies, we discovered that the concentration was aboutthree times higher in the eastern rams (31.4 ±1.4 ng ⁄ ml) than in the western rams (9.1 ± 0.6 ng ⁄ ml).This was not due to differences in their nutrition duringthe experiments, but because the eastern rams had abetter body condition (fatter) than the western ramsbefore the start of the experiment. We have thusconcluded, a posteriori, that the pre-experimental,nutritional history of the animals, and thus their tonicinsulin concentrations, prevented the oestradiol implantfrom inhibiting GnRH secretion in the eastern experiment.Thus, extensive knowledge of the history of theanimals can decrease the probability of conflictingexperimental outcomes. As we learn more about, forexample, the epigenetic effects of nutrition (Rhind 2004),this situation will become even more complex. Clearly,this is a particularly difficult issue for research done withfarm animals or wildlife.Animal emotion and ethical experimentationA major potential imposition on experimental animals is‘stress’ associated with the experimental protocol. Stressis an integral part of everyday life for all animals, withthe challenges of psychological or physical stressorsevoking the normal adaptive responses that maintainhomeostasis (Matteri et al. 1984). However, these sameresponses can not only interfere with an experimentaloutcome but also weigh heavily on the ‘cost’ side of thecost-benefit ratio. Thus, welfare can be improved byreducing any stress caused by experimental conditions,such as housing or the interactions between experimentalanimals and human experimenters. Our aim shouldbe to increase the freedom from fear and distress. This isvery relevant to studies in reproduction becausethe reproductive endocrine axis can be profoundlyÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Ethical Models for Studying Reproduction 11influenced by stress, as has been clearly demonstratedfor farm animals (e.g. Martin et al. 1981; Caraty et al.1997; von Borell et al. 2007).If reduced stress in experimentalanimals can improve the cost ⁄ benefit ratio, thenit might be useful to select experimental animals on thebasis of their temperament or fear reaction (Mills et al.1997).What is temperament and how is it assessed?Temperament has been defined for humans as ‘naturecontrolling the way he behaves, feels and thinks’(Cannon 1927). For animals, many attempts have beenmade to apply similar criteria on the basis of ourperceptions of their ‘temperament’. We cannot knowwhat is taking place in the mind of an animal, so weoften make assumptions about their ‘temperament’based on behavioural and neuroendocrine characteristics.Boissy (1995) concluded that emotional reactivityhas a significant impact on the relationship of an animalto its environment and that non-human animals havepersonality, temperament or individual behaviours,amongst which fearfulness plays an important role inan individual’s behavioural response to threateningsituations. Temperament has also been defined byMurphy (1999) as the emotivity of ‘fearfulness andreactivity of an animal in response to humans andstrange, novel or threatening environments’. Measuresof temperament, or fearfulness, in domesticated speciesare thus far restricted mostly to sheep, cattle, chickensand horses, and have been designed on the basis of theabove two definitions (Forkman et al. 2007). For cattle,such assessments have included subjective measuressuch as observations of crush behaviour or objectivemeasures such as flight time. (Voisinet et al. 1997; Fellet al. 1999). For sheep, we have used a combination oftwo objective behavioural tests, the ‘arena test’ and the‘box test’ (Murphy et al. 1994). Briefly, the ‘arena test’ isa motivational choice test that measures approach ⁄avoidance behaviour to humans (Murphy 1999). The‘box test’ is similar to isolation tests used in sheep andcattle (Cockram et al. 1994; Grignard et al. 2000) andprovides an objective measure of the degree of anxietyassociated with isolation. At the University of WesternAustralia, we have designed a selection index using bothtests and then genetically selected two lines of sheepusing males from the extremes of the variation. Withinthis framework, temperament has proven to be moderatelyheritable (h 2 0.3) and thus responds favourablyto selection (Murphy 1999) so, after more than 15generations, we now have ‘calm’ and ‘nervous’ lines.How can selection for temperament improve ourexperiments on reproduction?In mammals, temperament, or emotional reactivity ortameness, affects reproductive capacity and reproductivephysiology, either directly or indirectly (Price 2002). Insheep, direct effects of temperament have been demonstratedin the early stages of gestation and on thesurvival of newborn lambs. Early studies in our laboratoryshowed that calm ewes were better mothers thannervous ewes (Murphy et al. 1994) because they spendmore time with their lamb(s) and, when disturbed, havea shorter flight distance and return to their lamb(s)sooner. Consequently, the incidence of lamb mortality,birth to weaning, was about half for calm ewescompared to nervous ewes, and independent of thephysical status of the lambs (Murphy et al. 1994;Murphy 1999). Fecundity, and perhaps embryo survival,are higher in calm than in nervous ewes (Hartet al. 2008). Ewes of calm temperament also carriedmore twin embryos than nervous ewes (1.39 embryos,n = 472 : 1.29 embryos, n = 302; p < 0.001). Selectionfor calm temperament also modifies the expressionof reproductive behaviour, leading to an increase inproceptivity in maiden ewes (Gelez et al. 2003).Temperament also has indirect effects on the outcomesin other types of experimentation, includingstudies with metabolic factors such as leptin, a hormoneinvolved in the regulation of metabolism and thesignalling of the metabolic status to the reproductivesystem (Blache et al. 2007). Because leptin is also acytokine and thus involved in immune signalling, wedesigned an experiment to study the leptin response toan immune challenge (Fig. 2). A response was observedonly in sheep that had been selected for nervousness.When the data from both calm and nervous sheep werecombined, the response to the immune challenge wasnot significant, presumably because the temperamentvariance was now incorporated into residual (background)variance, reducing the power of the analysis.This example illustrates perfectly how selection fortemperament could improve the level of refinement andreduction in an animal model. Minimizing temperamentdifferences thus can reduce the number of animals usedChange in plasmaleptin (%)(a) LPS injected at 0 h (b) Vehicle injected at 0 h (c) Calm plus Nervous140130120110100900 1 2 3 4 5 6 0 1 2 3 4 5 6 0 1 2 3 4 5 6Time (h)Time (h)Time (h)Fig. 2. Change in plasma concentrations of leptin in sheep selected for calm (closed circles) or nervous temperament (closed squares) (a) duringfever induced by injection with lipopolysaccharide (0.4 lg ⁄ kg) or (b) after injection with vehicle. The black bar indicates the period during whichthe selected lines differed (p < 0.05). In the composite graph (c), calm and nervous animals have been pooled and there was no difference betweenvehicle (open circles) and induced fever (closed circles) (Blache and Maloney, unpublished data)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

12 D Blache, GB Martin and SK MaloneyVoluntary feedintakeMetabolic fuelEnergy reserveEmotional reactivityHPA axisMetabolicHormones:leptin, orexinsinsulin,IGF-1, GH,T3, T4ThermogenesisImmune functionCellular maintenanceGrowth, locomotionReproductionFig. 3. Central role of metabolic hormones in the interactions amongemotion, reproductive function, immune function, metabolic status,thermogenesis and other bodily functions. The sympathetic pathwayshave been omitted for clarityand limit the risk of conflicting results since, bydefinition, a group of animals showing a homogenousresponse to any stimulus will have less variability in anyparameter associated with that response.As described previously, the amount of availablemetabolic fuel needs to be ‘assessed’ by a control systemand then expenditure ‘directed’ towards reproductivefunction if and when required. For all systems, theassessment of energy levels and their partitioning isunder the control of metabolic hormones (Fig. 3).Thyroid concentrations were higher in a sub-populationof ‘calm’ than in a sub-population of ‘nervous’ animalsin our selection lines (Blache et al. 2002). We have alsoshown that the cortisol response to a novel object ishigher in nervous ewes than in calm ewes (Bickell et al.2008). These considerations apply equally to stressassociated with experimental protocols (Martin et al.1981; Adams et al. 1993). The effects of stress andcortisol on the reproductive axis are numerous anddiverse – in the female, they can either inhibit orstimulate ovulation (Dobson and Smith 2000). Theeffect of temperament genetics on reproductive processeslikely involves hormonal pathways. These linksare still to be demonstrated experimentally, but temperamentaffects the hormonal systems involved in thecontrol of energy partitioning and this may be how itinfluences the outcomes of experimentation targeting therelationship between nutrition and reproduction.Aside from improving the experimental design andoutcomes, the welfare of experimental animals could beimproved by selection for temperament. Less reactiveanimals will have an improved freedom from fear anddistress since animals that are less reactive to novelty willpresent reduced physiological and behavioural reactionsto novel experiences (Korte 2001), including exposure toa research environment.ConclusionIn inference science, hypothesis testing involves manipulatingone or a few parameters, while other factors thatcould influence the dependent parameters are controlled.In this paper, we have shown that multiplefactors that affect and control reproduction could affectthe outcomes of animal experimentation because theyaffect reproductive physiology, over both short and longtime scales, and even across generations. These factorsneed to be controlled. In addition, emotional reactivityhas the potential to alter experimental outcomes, just asgenotype can affect the outcomes of experimentation inwhich photoperiod or nutrition is manipulated. Byconsidering the nutritional history and emotional characteristicsof the individual subjects, the variabilitywithin experimental groups can be reduced and two ofthe 3Rs (reduction, refinement) would be applied,leading to more ethical experimentation. This creates awin–win situation for researchers and experimentalanimals because more ethical science results in betterscience.AcknowledgementsWe would like to thank the generous assistance of the students andstaff of Animal Science (University of WA). Funding was supplied bythe National Health and Medical Research Council, the AustralianResearch Council, the Ada Bartholomew Medical Research Trust, andthe University of Western Australia.ReferencesAdams NR, Atkinson S, Martin GB, Briegel JR, Boukhliq R,Sanders MR, 1993: Frequent blood sampling changes theplasma concentration of LH and FSH and the ovulationrate in Merino ewes. J Reprod Fertil 99, 689–694.Archer GS, Dindot S, Friend TH, Walker S, Zaunbrecher G,Lawhorn B, Piedrahita J, 2003a: Hierarchical phenotypicand epigenetic variation in cloned swine. Biol Reprod 69,430–436.Archer GS, Friend TH, Piedrahita J, Nevill CH, Walker S,2003b: Behavioural variation among cloned pigs. ApplAnimal Behav Sci 82, 151–161.Bacci ML, 2007: A brief overview of transgenic farm animals.Veterin Res Commun 31(Suppl. 1), 9–14.Bickell SL, Ferguson DM, Boissy A, Blache D, 2008: Sheepselected on their behavioural response to a social and humanchallenge differ also in their response to novelty. PhysiolBehav (in press).Bielli A, Pérez R, Pedrana G, Milton JTB, Lopez A,Blackberry M, Duncombe G, Rodriguez-Martinez H, MartinGB, 2002: Low maternal nutrition during pregnancyreduces the numbers of Sertoli cells in the newborn lamb.Reprod Fertil Dev 14, 333–337.Biggers JD, 1986: The potential use of artificially producedmonozygotic twins for comparative experiments. Theriogenology26, 1–25.Blache D, Tjondronegoro S, Blackberry MA, Anderson ST,Curlewis JD, Martin GB, 1997: Gonadotrophin and prolactinsecretion in castrated male sheep following subcutaneousor intracranial treatment with testicular hormones.Endocr 7, 235–243.Blache D, Blackberry MA, Van Cleeff JK, 2002: Does thethyroid axis play a role in domestication? 56th AnnualMeeting of the American Institute of Biological Sciences,Washington, DC, 22–24 March 2002, Abstract 27.Blache D, Zhang S, Martin GB, 2006: Dynamic and integrativeaspects of the regulation of reproduction by metabolicstatus in male sheep. Reprod Nutr Dev 46, 379–390.Blache D, Chagas LM, Martin GB, 2007: Nutritional inputsinto the reproductive neuroendocrine control system – amultidimensional perspective. in: Juengel JI, Murray JF,Smith MF (eds), Reproduction in Domestic Ruminants VI,Vol. Suppl 64. Nottingham University Press, Nottingham,UK, pp. 123–139.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Ethical Models for Studying Reproduction 13Bloomfield FH, Oliver MH, Hawkins P, Campbell M, PhillipsDJ, Gluckman PD, Challis Jr G, Harding JE, 2003: Apericonceptional nutritional origin for noninfectious pretermbirth. Science 300, 606.Boissy A, 1995: Fear and fearfulness in animals. Q Rev Biol70, 165–191.von Borell E, Dobson H, Prunier A, 2007: Stress, behaviourand reproductive performance in female cattle and pigs.Hormones Behav 52, 130–138.Bryant J, Baggott la Velle L, Searle J, 2005: Introduction toBioethics. Wiley, Chichester.Cannon WB, 1927: The James–Lange theory of emotions: acritical examination and an alteration. Am J Psychol 39,106–124.Caraty A, Miller DW, Delaleu B, Martin GB, 1997: Stimulationof LH secretion in sheep by central administration ofcorticotrophin-releasing hormone. J Reprod Fertil 111, 249–257.Celi P, Walkden-Brown SW, Blache D, Szell AZ, WilkinsonHM, Martin GB, 2007: Twin efficiency for reproductivevariables in monozygotic twin sheep. Theriogenology 68,663–672.Church RB, 1988: Are problems posed by genetically engineeredanimals? Appl Anim Behav Sci 20, 73–82.Cockram MS, Ranson M, Imlah P, Goddard PJ, Burrells C,Harkiss GD, 1994: The behavioural, endocrine and immuneresponses of sheep to isolation. Anim Production 58, 389–399.De Gendt K, Swinnen JV, Saunders PT K, Schoonjans L,Dewerchin M, Devos A, Tan K, Atanassova N, Claessens F,Le´cureuil C, Heyns W, Carmeliet P, Guillou F, Sharpe RM,Verhoeven G, 2004: A Sertoli cell-selective knockout of theandrogen receptor causes spermatogenic arrest in meiosis.Proc Natl Acad Sci U S A 101, 1327–1332.Dobson H, Smith RF, 2000: What is stress, and how does itaffect reproduction? Anim Reprod Sci 60, 743–752.Fell LR, Colditz IG, Walker KH, Watson DL, 1999: Associationsbetween temperament, performance and immunefunction in cattle entering a commercial feedlot. Aust J ExpAgric 39, 795–802.Festing MFW, 1990: Use of genetically heterogeneous rats andmice in toxicological research: a personal perspective.Toxicol Appl Pharmacol 102, 197–204.Festing M, Altman DG, 2002: Guidelines for the design andstatistical analysis of experiments using laboratory animals.ILAR 43, 244–258.Forkman B, Boissy A, Meunier-Salaun M-C, Canali E, JonesRB, 2007: A critical review of fear tests used on cattle, pigs,sheep, poultry and horses. Physiol Behav 91, 531–565.Gelez H, Lindsay DR, Blache D, Martin GB, Fabre-Nys C,2003: Temperament and sexual experience affect femalesexual behaviour in sheep. Appl Anim Behav Sci 84, 81–87.Grignard L, Boissy A, Boivin X, Garel JP, Le Neindre P, 2000:The social environment influences the behavioural responsesof beef cattle to handling. Appl Anim Behav Sci 68, 1–11.Hart K, Van Lier E, Vin˜ oles C, Paganoni B, Blache D, 2008:Calm Merino ewes have more multiple pregnancies thannervous Merino ewes due to higher ovulation rate. Proc 16thInternational Congress on Animal Reproduction (in press).Hewitt SC, Harrell JC, Korach KS, 2005: Lessons in estrogenbiology from knockout and transgenic animals. Annu RevPhysiol 67, 285–308.Home Office, 2007: Statistics of Scientific Procedures on LivingAnimals, Great Britain, 2006. The Stationery Office, London,p. 47.Howard BR, 2002: Control of variability. ILAR 43, 191–193.Korte SM, 2001: Corticosteroids in relation to fear, anxietyand psychopathology. Neurosci Biobehav Rev 25, 117–142.Kumar RT, 2005: What have we learned about gonadotropinfunction from gonadotropin subunit and receptor knockoutmice? Reproduction 130, 293–302.LaFolette H, 2007: The Practice of Ethics. Blackwell Publishing,Oxford.Martin GB, Oldham CM, Lindsay DR, 1981: Effect of stressdue to laparoscopy on plasma cortisol levels, the preovulatorysurge of LH, and ovulation in the ewe. Theriogenology16, 39–44.Matteri RL, Watson JG, Moberg GP, 1984: Stress or acuteadrenocorticotrophin treatment suppresses LHRH-inducedLH release in the ram. J Reprod Fertil 72, 385–393.Mepham B, 2005: Bioethics. Oxford University Press, Oxford.Miller DW, Blache D, Martin GB, 1995: The role ofintracerebral insulin in the effect of nutrition on gonadotrophinsecretion in mature male sheep. J Endocrinol 147,321–329.Mills AD, Beilharz RG, Hocking PM, 1997: Genetic selection.In: Appleby MC, Hughes BO (eds), Animal Welfare. CABInternational, Wallingford, pp. 219–231.Murphy PM, 1999: Maternal Behaviour and Rearing Ability ofMerino Ewes can be Improved by Strategic Feed SupplementationDuring Late Pregnancy and Selection for CalmTemperament. The University of Western Australia, Perth.Murphy PM, Purvis IW, Lindsay DR, Le Neindre P, OrgeurP, Poindron P, 1994: Measures of temperament are highlyrepeatable in Merino sheep and some are related to maternalbehaviour. Proc Austral Soc Anim Prod 20, 247–250.NHMRC Animal Welfare Committee, 2007: Guidelines forthe Generation, Breeding, Care and Use of GeneticallyModified and Cloned Animals for Scientific Purposes.Commonwealth of Australia, Canberra, p. 41.Nuffield Council on Bioethics, 2005: The Ethics of ResearchInvolving Animals. Nuffield Council on Bioethics, London,p. 376.Price EO, 2002: Animal Domestication and Behavior. CABInternational, Wallingford, p. 320.Rae MT, Palassio S, Kyle CE, Brooks AN, Lea RG, MillerDW, Rhind SM, 2001: Effect of maternal undernutritionduring pregnancy on early ovarian development andsubsequent follicular development in sheep fetuses. Reproduction,122, 915–922.Rhind SM, 2004: Effects of maternal nutrition on fetal andneonatal reproductive development and function. AnimReprod Sci 82–83, 169–181.Robinson JJ, Ashworth CJ, Rooke JA, Mitchell LM, McEvoyTG, 2006: Nutrition and fertility in ruminant livestock.Animal Feed Sci Technol 126, 259–276.Russell WMS, Burch RL, 1959: The Principle of HumaneExperimental Technique. Methuen, London.Schmidtz D, 2007: A place for cost–benefit analysis. In:LaFolette H (ed), Ethics in Practice: An Anthology.Blackwell Publishing, Oxford, pp. 669–679.Seidel GE, 2001: Cloning, transgenesis, and genetic variance inanimals. Cloning Stem Cells 3, 251–256.Somerville M, 2007: Thinking Ethics,Doing Ethics. EthicallyChallenged. The Miegunyah Press, Melbourne, pp. 68–98.Vallentyne P, 2007: Consequentialism. In: LaFolette H (ed),The Practice of Ethics. Blackwell Publishing, Oxford, pp.22–30.Vinoles C, Forsberg M, Martin GB, Cajarville C, Repetto J,Meikle A, 2005: Short-term nutritional supplementation ofewes in low body condition affects follicle development dueto an increase in glucose and metabolic hormones. Reproduction129, 299–309.Voisinet BD, Grandin T, Tatum JD, O’Connor SF, StruthersJJ, 1997: Feedlot cattle with calm temperaments have ahigher average daily weight gains the cattle with excitabletemperaments. J Anim Sci 75, 892–896.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

14 D Blache, GB Martin and SK MaloneyWetsel WC, 1995: Immortalized hypothalamic luteinizinghormone-releasing hormone (LHRH) neurons: a new toolfor dissecting the molecular and cellular basis of LHRHphysiology. Cell Mol Neurobiol 15, 43–78.Zhang S, Blache D, Blackberry MA., Martin GB, 2004:Dynamics of the responses in secretion of LH, leptin andinsulin following an acute increase in nutrition in maturemale sheep. Reprod Fertil Dev 16, 823–829.Conflict of interest: D Blanche has received grants from AustralianResearch Council, Medical Research Council and Meat & LivestockAustralia; Much of G.B. Martin’s research is funded by Meat &Livestock Australia; S. Maloney declares no conflict of interest.Author’s address (for correspondence): D Blache, School of AnimalBiology M085, Faculty of Natural and Agricultural Sciences, TheUniversity of Western Australia, 35 Stirling Highway, Crawley, WA6009, Australia. E-mail: dbla@animals.uwa.edu.auÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 15–22 (2008); doi: 10.1111/j.1439-0531.2008.01138.xISSN 0936-6768Endocrine Disruptors and Other Food-contaminating Environmental Pollutants asRisk Factors in Animal ReproductionSM RhindMacaulay Institute, Aberdeen, UKContentsPollutants of many chemical classes, derived primarily fromanthropogenic activities, are ubiquitous in the environment,persistent, biologically available and can exert adverse effectson the reproductive and other, indirectly related, physiologicalsystems. Food is generally considered to be the major route ofanimal exposure in vertebrate species but the relative contributionsof other routes of exposure such as through lungs, gillsor skin are not well studied and may be of importance forcertain animal groups, depending on their immediate environment.Animals are particularly sensitive to exposure duringdevelopmental stages but the pattern of exposure to chemicalsis likely to be different to that of adults. Quantification of therisk posed by the ingestion of pollutants in food is complexand depends on many factors including species, diet composition,duration of exposure to the food, efficiency of pollutantabsorption, subsequent metabolism, sensitivity of targetorgans and stage of development. While the effects of highdoses of single chemicals are proven, dietary exposure topollutants generally involves prolonged, low-level exposure toa large number of compounds, each of which has differentchemical characteristics, exerts different biological effects andis present at varying concentrations. Thus, while exposure topollutants through feed is undoubtedly a significant risk factorfor many species and may be the most important one for manyterrestrial vertebrates, other routes of exposure may be moreimportant in other groups.IntroductionExposure to pollutants, and in particular to endocrinedisrupting compounds (EDCs) and potentially toxicmetals (PTMs) can induce adverse physiological changesin animals of every group so far studied, from bacteria(Fox 2004) to humans (Toppari et al. 1996) and animalsof many phyla in between (Institute for Environmentand Health 1999). Many of the effects pertain to thereproductive system (Toppari et al. 1996; Paul et al.2005; Uzumcu and Zachow 2007, Fowler et al. 2008) butother physiological systems can be perturbed by exposure.These include the thyroid gland (Hansen 1998),neuroendocrine system (Lein et al. 2007) and thereforebehaviour (Erhard and Rhind 2004), the immune system(Vine et al. 2000; Daniel et al. 2001) and systems thatcontrol adipogenesis and nutrient partitioning (Newboldet al. 2007); disruption of these physiologicalsystems may also impinge on reproductive fitness.Evidence of relationships between exposure to pollutantsand reproductive effects is derived from observationalstudies of wildlife exposed, unintentionally, topollutants (Smith 1981; Guillette et al. 1994; Institutefor Environment and Health 1999) and from empiricalstudies of laboratory and domestic animals or cellcultures (Gray et al. 2000; Meerts et al. 2001; Evanset al. 2004). The reproductive systems of animals ofmany phyla have been found to be perturbed throughmany mechnisms (Table 1). The first type of studyserved to demonstrate associations between EDC exposureand adverse effects on reproductive function whilethe latter has begun to elucidate mechanisms of action.However, even if they have involved administration ofchemicals in feed, some controlled studies are poorrepresentations of normal real-world exposure throughfood, being based on single chemicals, often at very highconcentrations, and administered for short periods.The aim of this paper is to review what is known ofthe risks of exposure to pollutants through the diet andto identify some principles that pertain to their effects onreproductive function in general, irrespective of speciesand habitat.What Classes of Chemical are Involved andHow Do They Operate?The acute toxicity of high levels of PTMs has probablybeen recognized since metals were first smelted. However,recent reports indicate that they may also exertmore subtle effects and that very low, environmentallyrelevant, concentrations of cadmium, copper, cobalt,nickel, lead, tin and mercury can each activate theoestrogen receptor a (Martin et al. 2003; Henson andChedrese 2004) while methyl mercury can disrupt nonsteroidalreceptors (Klaper et al. 2006) and copper cancompromise olfactory or other senses in fish (Sandahlet al. 2007). Thus, each has the potential to disruptreproductive function.Organic EDCs, unlike PTMs which can also actthrough endocrine disruption, are primarily anthropogenicin nature and have been produced in largeamounts only for approximately 60 years, only. Classesand sources of EDCs have been described previously(Rhind 2005). They seldom induce an acute toxic effectbut can affect reproductive function at concentrationsmany orders of magnitude below the toxic dose. Theseeffects are expressed through many mechanisms includingbinding to the androgen receptor (DDT metabolites;Kelce et al. 1998), oestrogen receptor [alkyl phenols andpolychlorinated biphenyls (PCBs); Dodge 1998] and arylhydrocarbon receptor [polycyclic aromatic hydrocarbons(PAHs); Haque et al. 2005]. In addition, PCBs canact directly on cellular systems via non-receptor-mediatedsystems (Li and Hansen 1996). It is noteworthy thatthe DNA of the oestrogen receptor appears to be highlyconserved over a very wide range of phyla and species,suggesting that it is evolutionarily ancient (ThorntonÓ 2008 Macaulay Land Use Research Institute

16 SM RhindTable 1. Selected examples of reproductive perturbation, by environmental exposure to endocrine disrupting pollutants, in species representingdifferent classes and phylaSpecies Chemical Effect ReferenceWhite footed mice (Peromyscus leucopus) PCB and cadmium Reduced testis size Batty et al. (1990)Seagull (Larus californicus) DDT Feminization of embryos Fry and Toone (1981)American alligator (Alligator mississippiensis) Dicofol, DDT Abnormal testes, phalli and testosterone Guillette et al. (1994)Dogwhelk (Nucella lapillus) Tributyltin Imposex Gibbs and Bryan (1987)Mayfly (Cloeon dipterum) Esfenvalerate (insecticide) Reduced larval survival Beketov and Liess (2005)Annelids Various Multiple Krajniak (2005)Nematode (Caenorhabditis elegans) Various Multiple Ho¨ ss and Weltje (2007)et al. 2003). Since many EDC actions are mediatedthrough this receptor, it may explain why the influencesof these pollutants are present in so many animalgroups.Some EDCs act through multiple mechanisms andsince animals are usually exposed, simultaneously, tomany EDCs, those of different classes can act in concertwith each other (Rajapakse et al. 2002) and with PTMs(Bemis and Seegal 1999). Consequently, prediction ofrisk from environmental exposure is made more difficult.As a result of receptor-mediated signals, endocrineperturbations or direct effects of pollutants on DNA,changes in expression of a wide range of genes can beinduced. The changes in gene function are probablyinduced epigenetically (i.e. mitotically heritable changesin gene function may be caused without changes inDNA sequence) through processes such as DNA methylation(Crews and McLachlan 2007).What Factors Determine the Risk Associatedwith Pollutant Exposure?Risks associated with dietary pollutants in both vertebrateand invertebrate species depend on the propertiesof the chemicals involved since these may determine theamount present in food, the likelihood of ingestion,uptake rate and bioavailability following ingestion, andtarget organ sensitivity. Owing to the logistical difficultiesof measurements in small, invertebrate animals,much of the evidence pertaining to tissue levels ofpollutants is derived from studies of vertebrates andparticularly mammals. Many basic principles concerningexposure risk can be extrapolated, to an extent, tosmaller, invertebrate species but it should be noted thatthe contribution of diet to pollutant burden may berelatively less important than exposure through thebody surface, particularly for species that live in water,sediment or soil. Since assessment of the relativeimportance of the respective routes of exposure islogistically extremely difficult, this is poorly understood.However, since the health and reproductive efficiency ofsuch species is critical for ecosystem health and sustainability,potential effects of pollutants on these animalgroups are of great importance.Persistence and biological availabilityWhile environmental concentrations of PTMs andEDCs are highly variable, they are generally low.Many remain in the environment indefinitely but theirbiological availability determines whether or not theycan exert any adverse effects. Uptake of PTMs from thedigestive tract may be limited by sequestration withiningested soil or plants (Wilkinson et al. 2003) and differswith individual element and chemical speciation, animalspecies, diet composition (e.g. fibre content), vitaminstatus, dose, animal age, dietary chelators, the presenceof antagonistic elements, pregnancy and lactation (Institutefor Environment and Health 1998; Wilkinson et al.2003) (Fig. 1). Once absorbed, metallothionein, andother proteins produced in the liver and gut wall, bufferintracellular levels of metals and act as a storagemechanism, particularly in vertebrates (Wilkinson et al.2003).Gastrointestinal absorption of many EDCs is efficientin vertebrates (Wild and Jones 1992; Norstrom 2002)and once absorbed they remain, generally, biologicallyavailable being sequestered little (

Dietary Pollutants as Risk Factors in Reproduction 17Level ofexposureClass of compound – volatility, hydrophobicity,degradability, solubility, binding properties, etcTrophic level/biomagnificationEnvironmental concentrationUptakeDietary exposure(food/water)• Food type• Degradation in intestinal tract• Efficiency of absorption from tract• Dietary chelatorsNon-dietary exposure(skin/integument/lungs/gills, etc)• Structure/permability of barrier• Active transport processes (?)• Octanol/water or octanol/air coefficient• SequestrationPost-uptakeWithin animal• Metabolism• Excretion• Lipid contentFig. 1. Determinants of level ofexposure to environmentalpollutants, of uptake by animals,rates of exposure of their organsand effects on themIndividual organ• Permeability of organ membrane(e.g. brain; testis)• Sensitivity of target tissues/targetgenesOffspring• Susceptibility of gametes• Efficiency of trans-placentaltransfer• Fetal metabolism/excretionassociated gene expression. Furthermore, disruption ofthe germ cell line in one generation can be expressed insubsequent generations (Bøgh et al. 2001; Anway andSkinner 2006; Edwards and Myers 2007), probablythrough altered gene methylation, and therefore geneexpression, during the period of sex differentiation of thedeveloping gonad (Anway and Skinner 2006). Althoughdemonstrated in vertebrates, these fundamental mechanismsare likely to pertain to most species.How Important is Exposure Through Ingestion?The complex relationship between rates of pollutantingestion and effects exerted may be further complicatedby related factors such as additional uptake throughdrinking water, inhalation or, particularly in smaller,invertebrate species, through dermal absorption ofpollutants.Exposure through foodFood is generally considered to be the most importantroute of exposure in vertebrates, except, perhaps, at thelowest trophic levels (Fries 1995; Kavlock et al. 2002;Norstrom 2002). Since pollutants accumulate in thetissue of animals at each trophic level, the position of theanimal in the food chain determines the exposure rate;owing to the effects of biomagnification (increasedconcentrations as a result of food chain energetics),the highest tissue concentrations are frequently recordedin carnivorous species at or near to the top of the foodchain (Johnson et al. 1996; Darnerud et al. 2002; Davidand Gans 2003; Law et al. 2003; Naert et al. 2007). Forexample, in such species, concentrations of polybrominateddiphenyl ethers (PBDEs) often can be>100 lg ⁄ kg lipid (Darnerud et al. 2001) and PCBs>100 lg ⁄ kg dry matter based on the whole body(Johnson et al. 1996).Herbivores, being much closer to the primary sourceof food production, are not subject to multiple levels ofbiomagnification. Accordingly, tissue accumulation (theincrease in concentration tissues attributable to uptake)is frequently much lower with PBDEs being

18 SM Rhindingestion’ for these may be very different from that ofadults, e.g. mammalian foetuses are nourished viamaternal blood and developing birds by egg yolk andwhile young mammals consume milk, adults eat animalor plant material. Similarly, larval stages of insects andother invertebrates may have very different diets toadults. In addition to differences in food source,differences in metabolism with age mean that it isunwise to extrapolate from effects of exposure in adultsto those in immature animals, because the relationshipbetween the amounts of pollutants ‘ingested’ at thedifferent stages and the respective tissue concentrationsmay differ with differences in metabolism. Furthermore,indirect measures of exposure of immature animals topollutants, for example, through measurements inmammalian milk or umbilical cord blood may be ofless value than tissue concentrations in the foetus ⁄ immatureanimal for the prediction of physiological effectssince the former does not take account of effects ofmetabolism and excretion in the offspring.While concentrations of pollutants in human (Dekoningand Karmaus 2000; Darnerud et al. 2001) andsheep milk (Rhind et al. 2007b), and human umbilicalcord blood, amniotic fluid and meconium (Dallaireet al. 2003; Barr et al. 2007) have been measured, thereare relatively few reports of foetal tissue concentrations(Bosse et al. 1996; Dekoning and Karmaus 2000) in anyspecies and so direct assessment of risk to the foetus isnot possible. Concentrations of PTMs (Rhind 2005) andseveral PCBs and PBDEs (Rhind et al. 2007a) have beenshown to be lower in fetal sheep than in the correspondingmaternal tissue but individual concentrationsdiffer greatly with congener (PCB < 20 ng ⁄ kg to3.8 lg ⁄ kg; PBDE < 20 ng ⁄ kg to 30 lg ⁄ kg liver drymatter), as does the relationship between adult andfoetal levels. Schecter et al. (2007) provided the firstreport of PBDE levels in human foetuses; mean totalPBDE concentrations were 23 ng ⁄ kg lipid.Interestingly, sheep maternal and foetal concentrationsof neither PTMs (Rhind et al. 2005a) nor PCBsand PBDEs (SM Rhind, unpublished observations) weresignificantly correlated. However, related studies havealso shown that very small increases in maternalexposure to pastures fertilized with sewage sludge whichcontains multiple pollutants, result in changes in structureand function of the foetal testis (Paul et al. 2005)and ovary (Fowler et al. 2008). Collectively, theseobservations show that the relationships between pollutantexposure and effect are highly complex, that foetalexposure cannot be extrapolated from adult profiles,and that small increases in exposure to multiple pollutantscan adversely affect foetal reproductive development.Immediately after birth, the main source of nutrient inmammals is milk. The transfer of PTMs to the offspringthrough milk is probably relatively unimportant (Wilkinsonet al. 2003) but exposure to lipophilic organicpollutants through milk may be significant, particularlyin species such as marine mammals which have a highmilk fat content (Duinker and Hillebrand 1979). On theother hand, studies comparing milk concentrations atearly and late stages of lactation in sheep (periods of fatmobilization and deposition respectively) indicate thatmilk concentrations of phthalate and alkyl phenol arenot elevated by increased maternal fat mobilizationduring early lactation (Rhind et al. 2007b) suggestingthat maternal fat may not be a significant contributor toneonatal pollutant exposure in this species.Food-related exposureWhile food is generally regarded as the major source ofexposure in vertebrates, in some circumstances foodrelatedroutes may also be important. There is a risk forall species (vertebrate and invertebrate) of exposure tocontaminated water or air as a result of the processes offood acquisition and ingestion. For example, dairy cowsdrinking from canals in The Netherlands which arecontaminated with sewage, and therefore a cocktail ofpollutants, have been shown to have reduced reproductiveefficiency and lower milk production (Meijer et al.1999). Most of the pollutants that are likely to bepresent in food are relatively insoluble in water and soare not normally ingested with drinking water. However,if bound to suspended solids in the water, suchpollutants could be ingested. In addition, these pollutantsmay interact with steroids from the contraceptivepill and alkyl phenols which are typically present in theaqueous component of sewage (Institute for Environmentand Health 1999; Bowman et al. 2003).Similarly, surface application of sewage sludge topastures has been shown to be associated with compromisedfoetal testis and ovary development (Paul et al.2005; Fowler et al. 2008). Since systemic uptake ofpollutants by plants is very limited (Wild and Jones1992), it is likely that the associated exposure topollutants results from ingestion of soil along with theherbage, particularly when the herbage mass is small (upto 30% of dry matter intake of sheep may be soil;Thornton and Abrahams 1983), as a result of surfacecontamination of the herbage with soil particles.In specific circumstances, air-borne pollutants linkedto the food source could, at least theoretically, contributeto the body burden of pollutants, e.g. domesticanimals may inhale additional pollutants when grazingpastures fertilized with sewage sludge, owing to theretention of volatilized chemicals within the sward.Concentrations of many pollutant classes in normal airare measurable (David and Sandra 2001; U.S. CentralPollution Control Board Newsletter 2001; Lee et al.2004) and there is evidence that urban air-bornepollution can cause DNA damage in humans (Whyattet al. 1998). Thus, the potential for additional exposures,resulting from the process of acquiring food,should be taken into account.Age-related effectsThe degree of pollutant accumulation through fooddepends on animal age not only because of the differenceswith age in diet type but also because olderanimals may be exposed to pollutants for longer and, asshown in some species of fish, may accumulate more(Darnerud et al. 2001). Furthermore, it has been shownthat when exposure is prolonged, low, and apparentlyharmless, pollutant concentrations can exert biologicalÓ 2008 Macaulay Land Use Research Institute

Dietary Pollutants as Risk Factors in Reproduction 19effects at least equivalent to, and sometimes higher than,those induced by much higher doses administered overshort periods of time (Borman et al. 2000).Individual variation in exposureIn species as diverse as sheep (Rhind et al. 2005b) andinsects from lake sediments (Kovats and Ciborowski1989) differences in tissue concentrations of variouspollutants of at least 10-fold between individuals fromthe same source are commonly observed. Similar environmentalvariation was reported in soil samples (foodcontaminant and invertebrate habitat) collected fromwithin experimental plots (Rhind et al. 2002). Thisfurther complicates the prediction of responses ofindividual animals to pollutant exposure.In summary, there seems to be little doubt that thediet is a major source of exposure to pollutants in manyvertebrate species and while the effects on tissueconcentrations are greatest in carnivores near to thetop of the food chain, significant tissue accumulationcan occur in herbivores. However, the relative contributionsof other factors such as uptake from thesurrounding media (air, water, soil or sediment) are lessclear and, in certain circumstances and species, thesesources may be more important than dietary ones,particularly in small, soil-, sediment-, or water-dwellingspecies.What Risk is Associated with Pollutants inFood?Since many studies of effects of pollutants on reproductivefunction frequently involve transient, unusuallyhigh levels of exposure of wild or laboratory species ofanimals to a single chemical, it is sometimes concludedthat normal, low, levels of exposure to pollutants areof no biological consequence. This argument is reinforcedby the fact that environmental concentrationsof some, but not all, pollutants are declining (Darnerudet al. 2001; Norstrom 2002), as are milk (Solomonand Weiss 2002) and tissue concentrations (Noakeset al. 2006). In fact, even tissue concentrations thatare at or near to background environmental concentrationsmay exert biological effects, particularly whenthey act in conjunction with a mixture of otherchemicals; ‘real-world’ exposure normally involvesprolonged exposure to a mixture of compounds at lowconcentrations. Effects of such exposure patterns havebeen demonstrated by long-term studies of sheepmaintained throughout their breeding lives on pasturesfertilized with either conventional inorganic fertilizeror sewage sludge. The latter contains large amounts ofboth inorganic and organic pollutants, relative toenvironmental levels (Smith 1996; Stevens et al. 2003)and, when applied to pasture as a fertilizer, results invery small increases in environmental levels of thesechemicals (Rhind et al. 2002) and in the tissue levelsof sheep grazing the pasture (Rhind et al. 2005a,b,2007b). However, these small increases are associatedwith disruption of reproductive and other functions(Erhard and Rhind 2004; Paul et al. 2005; Fowleret al. 2008).Such studies clearly demonstrate that pollutants infeed can represent a risk to reproductive function but donot allow the quantification of that risk with respect toeither individual chemicals or individual animals. Thatwould require determination of dose response relationshipsbetween the concentrations of each chemical, ineach relevant tissue or organ and for each observedchange in structure or function. It is possible to achievethis, empirically, for a small number of individualchemicals but, at present, virtually impossible to do sofor the thousands of chemicals to which animals areexposed. Understanding of mixture effects, will requireintegration of the observations from the wide range ofempirical approaches that are available using powerfulpredictive computer models (Suk et al. 2002).ConclusionsThere is no doubt that exposure to pollutants canadversely affect reproductive function in animals andthat food is likely to be the most important route ofexposure for many or most species, particularly forthose near the top of the food chain. However,assessment of the risk is complicated by the fact thatin the ‘real world’ food contains many different classesof pollutants, each at low and variable concentrations,and exposure to them occurs throughout life, includingduring developmental stages which are particularlysensitive to their effects. In short, dietary pollutantsrepresent a significant risk factor for reproduction,under certain circumstances, but precise quantificationof these risks for individuals is extremely difficult withcurrent knowledge.ReferencesAlbro PW, Lavenhar SR, 1989: Metabolism of di-(2-ethylhexyl)phthalate. Drug Metab Rev 21, 13–34.Anway MD, Skinner MK, 2006: Epigenetic transgenerationalactions of endocrine disruptors. Endocrinology 147, S43–S49.Bachour G, Failing K, Elmadfa SG, Brunn H, 1998: Speciesand organ dependence of PCB contamination in fish, foxes,roe deer, and humans. Arch Environ Contam Toxicol 35,666–673.Barr DB, Bishop A, Needham LL, 2007: Concentrtions ofxenobiotic chemicals in the maternal-fetal unit. ReprodToxicol 23, 260–266.Batty J, Leavitt RA, Biondot N, Polin D, 1990: Anecotoxicological study of a population of the white footedmouse (Peromyscus leucopus) inhabiting a polychlorinatedbiphenyls-contaminated area. Arch Environ Contam Toxicol19, 283–290.Beketov MA, Liess M, 2005: Acute contamination withesfenvalerate and food limitation: chronic effects on themayfly, Cloeon dipterum. Environ Toxicol Chem 24, 1281–1286.Bemis JC, Seegal RF, 1999: Polychlorinated biphenyls andmethylmercury act synergistically to reduce rat brain dopaminecontent in vitro. Environ Health Perspect 107, 879–885.Bøgh IB, Christensen P, Dantzer V, Groot M, Thøfner ICN,Rasmussen RK, Schmidt M, Greve T, 2001: Endocrinedisrupting compounds: effect of octylphenol on reproductionover three generations. Theriogenol 55, 131–150.Ó 2008 Macaulay Land Use Research Institute

20 SM RhindBorman SM, Christian PJ, Sipes IG, Hoyer PB, 2000:Ovotoxicity in female Fischer rats and B6 mice induced bylow-dose exposure to three polycyclic aromatic hydrocarbons:comparison through calculation of an ovotoxic index.Toxicol Appl Pharmacol 167, 191–198.Bosse U, Bannert N, Niessen KH, Teufel M, Rose I1996: Fetalorgan and tissue content of organohalogens and polychlorinatedbiphenyl. Zbl Hyg 198, 331–339.Bowman JC, Readman JW, Zhou JL, 2003: Sorption of thenatural endocrine disruptors, oestrone and estradiol-17B inthe aquatic environment. Environ Geochem Health 25, 63–67.Crews D, McLachlan JA, 2007: Epigenetics, evolution, endocrinedisruption, health and disease. Endocrinology147(Suppl.), S4–S10.Crofton KM, Craft ES, Hedge JM, Gennings C, Simmons JE,Carchman RA, Carter WH, deVito MJ, 2005: Thyroidhormone-disruptingchemicals: evidence for dose-dependentadditivity or synergism. Environ Health Perspect 113, 1549–1554.Dallaire F, Dewailly E, Muckle G, Ayotte P, 2003: Timetrends of persistent organic pollutants and heavy metals inumbilical cord blood of Inuit infants born in Nunavik(Quebec, Canada) between 1994 and 2001. Environ HealthPerspect 111, 1660–1664.Daniel V, Huber W, Bauer K, Suesal C, Conradt C, Opelz G,2001: Associations of blood levels of PCB, HCHs, and HCBwith numbers of lymphocyte subpopulations, in vitrolymphocyte response, plasma cytokine levels, and immunoglobulinautoantibodies. Environ Health Perspect 109, 173–178.Darnerud PO, Eriksen GS, Johannesson T, Larsen PB,Viluksela M, 2001: Polybrominated diphenyl ethers: occurrence,dietary exposure, and toxicology. Environ HealthPerspect 109(Suppl. 1), 49–68.David RM, Gans G2003: Summary of mammalian toxicologyand health effects of phthalate esters. In: Staples CA(ed.),An Assessment of the Potential Environmental Risks Posedby Phthalates. Handbook of Environmental Chemistry 3.Springer, London, pp. 299–316.David F, Sandra P2001: Phthalate esters in the environment.Monitoring program for the determination of phthalates inair, vegetation, cattle feed, milk and fish in The Netherlands(1999–2001). ECPI -2001-10, Research Institute of Chromatography,Belgium.Dekoning EP, Karmaus W, 2000: PCB exposure in utero andvia breast milk. A review. J Exp Anal Environ Epidemiol 10,285–293.Dodge JA, 1998: Structure ⁄ activity relationships. Pure ApplChem 70, 1725–1734.Duinker JC, Hillebrand MThJ, 1979: Mobilisation of organochlorinesfrom female lipid tissue and transplacentaltransfer to foetus in a harbour porpoise (Phocoena phocoena)in a contaminated area. Bull Environ Contam Toxicol23, 728–732.Edwards TM, Myers JP, 2007: Environmental exposures andgene regulation in disease etiology. Environ Health Perspect115, 1264–1270.Erhard H, Rhind SM, 2004: Prenatal and postnatal exposureto environmental pollutants in sewage sludge alters emotionalreactivity and exploratory behaviour in sheep. SciTotal Environ 332, 101–108.Evans NP, North T, Dye S, Sweeney T, 2004: Differentialeffects of the endocrine-disrupting compounds Bisphenol-Aand Octylphenol on gonadotrophin secretion, in prepubertalewe lambs. Domest Anim Endocr 26, 61–73.Fry DM, Toone CK, 1981: DDT-induced ferninization of gullembryos. Science 213, 922–924.Fowler PA, Dora N, McFerran H, Amezaga MR, Miller DW,Lea RG, Cash P, McNeilly AS, Evans NP, Cotinot C,Sharpe RM, Rhind SM, 2008: In-utero exposure to lowdoses of environmental pollutants disrupts fetal ovariandevelopment in sheep. Mol Human Reprod doi:0-1093 ⁄molehr ⁄ gan020.Fox JE, 2004: Chemical communication threatened by endocrine-disruptingcompounds. Environ Health Perspect 112,648–653.Fries GF, 1995: A review of the significance of animal foodproducts as potential pathways of human exposures todioxins. J Anim Sci 73, 1639–1650.Gandolfi F, Pocar P, Brevini TAL, Fischer B, 2002: Impact ofendocrine disrupters on ovarian function and embryonicdevelopment. Domest Anim Endocr 23, 189–201.Gibbs P, Bryan G, 1987: TBT paints and the demise of thedog-whelk, Nucela lapillus (Gastropoda). Oceans 19, 1482–1487.Gray LE, Ostby J, Furr J, Price M, Veeramachaneni DNR,Parks L, 2000: Perinatal exposure to the phthalates DEHP,BBP, and DINP, but not DEP, DMP, or DOTP, alterssexual differentiation of the male rat. Toxicol Sci 58, 350–365.Guillette LJ, Gross TS, Masson JR, Matter JM, Percival HF,Woodward AR, 1994: Developmental abnormalities of thegonad and abnormal sex hormone concentrations in juvenilealligators from contaminated and control lakes in Florida.Environ Health Perspect 102, 680–688.Hansen L, 1998: Stepping backward to improve assessment ofPCB congener toxicities. Environ Health Perspect 106, 171–189.Haque M, Francis J, Sehgal I, 2005: Aryl hydrocarbonexposure induces expression of MMP-9 in human prostatecancer cell lines. Cancer Lett 225, 159–166.Henson MC, Chedrese PJ, 2004: Endocrine disruption bycadmium, a common environmental toxicant with paradoxicaleffects on reproduction. Exp Biol Med 229, 383–392.Ho¨ ss S, Weltje L, 2007: Endocrine disruption in nematodes:effects and mechanisms. Ecotoxicology 16, 15–28.Institute for Environment and Health, 1998: IEH Report onFactors Affecting the Absorption of Toxic Metals from theDiet (Report R8). MRC Institute for Environment andHealth, Leicester, UK.Institute for Environment and Health, 1999: Institute forEnvironment and Health Assessment of the EcologicalSignificance of Endocrine Disruption: Effects on ReproductiveFunction and Consequences for Natural Populations(Assessment A4). MRC Institute for Environment andHealth, Leicester, UK.Johnson MS, Leah RT, Conor L, Rae C, Saunders S, 1996:Polchlorinated biphenyls in small mammals from contaminatedlandfill sites. Environ Pollut 92, 185–191.Kavlock K, Boekelheide R, Chapin M, Cunningham E,Faustman E, Foster P, Golub M, Henderson R, HinbergI, Little R, Seed J, Shea K, Tabacova S, Tyl R, Williams P,Zacharewski T, 2002: NTP center for the evaluation of risksto human reproduction: phthalates expert panel report onthe reproductive and developmental toxicity of di(2-ethylhexyl)phthalate. Reprod Toxicol 16, 529–653.Kelce WR, Gray LE, Wilson EM, 1998: Anti-androgens asenvironmental endocrine disruptors. Reprod Fert Dev 10,105–111.Kelly BC, Ikonomou MG, Blair JD, Morin AE, Gobas FAPC,2007: Food web-specific biomagnification of persistentorganic pollutants. Science 317, 236–238.Klaper R, Rees CB, Drevnick P, Weber ED, Sandheinrich M,Carvan MJ, 2006: Gene expression changes related toendocrine function and decline in reproduction in fatheadÓ 2008 Macaulay Land Use Research Institute

Dietary Pollutants as Risk Factors in Reproduction 21minnow (Pimephales promelas) after dietary methylmercuryexposure. Environ Health Perspect 114, 1337–1343.Kovats ZE, Ciborowski JJH, 1989: Aquatic insect adults asindicators of organochlorine contamination. J Great LakesRes 15, 623–634.Krajniak K, 2005: Annelid endocrine disruptors and a surveyof invertebrate FMRFamide-related peptides1. Integr CompBiol 45, 88–96.Law RJ, Alaee M, Allchin CR, Boon JP, Lebeuf M, Lepom P,Stern GA, 2003: Levels and trends of polybrominateddiphenylethers and other brominated flame retardants inwildlife. Environ Int 29, 757–770.Lee RGM, Thomas GO, Jones KC, 2004: PBDEs in theatmosphere of three locations in Western Europe. EnvironSci Technol 38, 699–706.Lein PJ, Yang D, Bachstetter AD, Tilson HA, Harry GJ,Mervis RF, Kodavanti PRS, 2007: Ontogenetic alterationsin molecular and structural correlates of dendritic growthafter developmental exposure to polychlorinated biphenyls.Environ Health Perspect 115, 556–563.Li M-H, Hansen LG, 1996: Enzyme induction and acuteendocrine effects in prepubertal female rats receiving environmentalPCB ⁄ PCDF ⁄ PCDD mixtures. Environ HealthPerspect 104, 712–722.Martin MB, Reiter R, Pham T, Avellanet YR, Camara J,Lahm M, Pentecost E, Pratap K, Gilmore BA, Divekar S,Dagata RS, Bull JL, Stoica A, 2003: Estrogen-like activity ofmetals in Mcf-7 breast cancer cells. Endocrinology 144,2425–2436.Meerts IATM, Letcher RJ, Hoving S, Marsh G, Bergman A,Lemmen JG, van der Berg B, Brouwer A, 2001: In vitroestrogenicity of polybrominated diphenyl ethers, hydroxylatedPBDEs, and polybrominated bisphenol A compounds.Environ Health Perspect 109, 399–407.Meijer GAL, de Bree JA, Wagenaar JA, Spoelstra SF, 1999:Sewerage overflows put production and fertility of dairycows at risk. J Environ Qual 28, 1381–1383.Naert C, Van Peteghem C, Kupper J, Jenni L, Naegeli H,2007: Distribution of polychlorinated biphenyls and polybrominateddiphenyl ethers in birds of prey from Switzerland.Chemosphere 68, 977–987.Newbold RR, Padilla-Banks E, Snyder RJ, Phillips TM,Jefferson WN, 2007: Developmental exposure to endocrinedisruptors and the obesity epidemic. Reprod Toxicol 23,290–296.Noakes PS, Taylor P, Wilkinson S, Prescott SL, 2006: Therelationship between persistent organic pollutants inmaternal and neonatal tissues and immune responses toallergens: a novel exploratory study. Chemosphere 63,1304–1311.Norstrom RJ, 2002: Understanding bioaccumulation of POPsin food webs. Environ Sci Pollut Res 9, 300–303.Paul C, Rhind SM, Kyle CE, Scott H, McKinnell C, SharpeRM, 2005: Cellular and hormonal disruption of fetal testisdevelopment in sheep reared on pasture treated with sewagesludge. Environ Health Perspect 113, 1530–1587.Payne J, Scholze M, Kortenkamp A, 2001: Mixtures of fourorganochlorines enhance human breast cancer cell proliferation.Environ Health Perspect 109, 391–397.Pocar P, Perazzoli F, Luciano A, Gandolfi F, 2001: In vitroreproductive toxicity of polychlorinated biphenyls: effectson oocyte maturation and developmental competence incattle. Mol Reprod Dev 58, 411–416.Rajapakse N, Silva E, Kortenkamp A, 2002: Combiningxenoestrogens at levels below individual no-observed-effectconcentrations dramatically enhances steroid hormoneaction. Environ Health Perspect 110, 917–921.Reid BJ, Jones KC, Semple KT, 2000: Bioavailability ofpersistent organic pollutants in soils and sediments – aperspective on mechanism, consequences and assessment.Environ Pollut 108, 103–112.Rhind SM, 2002: Endocrine disrupting compounds and farmanimals: their properties, actions and routes of exposure.Domest Anim Endocrinol 23, 179–187.Rhind SM, 2005: Are endocrine disrupting compounds athreat to farm animal health, welfare and productivity?Reprod Dom Anim 40, 282–290.Rhind SM, Smith A, Kyle CE, Telfer G, Martin G, Duff E,Mayes RW, 2002: Phthalate and alkyl phenol concentrationsin soil following applications of inorganic fertiliser orsewage sludge to pasture and potential rates of ingestion bygrazing ruminants. J Environ Monit 4, 142–148.Rhind SM, Kyle CE, Owen J, 2005a: Accumulation ofpotentially toxic metals in the liver tissue of sheep grazedon sewage sludge-treated pastures. Anim Sci 81, 107–113.Rhind SM, Kyle CE, Telfer G, Duff EI, Smith A, 2005b: Alkylphenols and diethylhexyl phthalate in tissues of sheepgrazing pastures fertilised with sewage sludge or inorganicfertilizer. Environ Health Perspect 113, 447–453.Rhind SM, Kyle CE, Mackie C, Telfer G, 2007a: PCB andPBDE accumulation in adult and fetal sheep grazing sludgetreated pastures. Proc Soc Environ Toxicol Chem, Norwich.5–6 September, 2007.Rhind SM, Kyle CE, Mackie C, Telfer G, 2007b: Effects ofexposure of ewes to sewage sludge-treated pasture onphthalate and alkyl phenol concentrations in their milk.Sci Total Environ 383, 70–80.Sandahl JF, Baldwin DH, Jenkins JJ, Scholz NL, 2007: Asensory system at the interface between urban stormwaterrunoff and salmon survival. Environ Sci Technol 41, 2998–3004.Schecter A, Johnson-Welch S, Tung KC, Harris TR, Papke O,Rosen R, 2007: Polybrominated diphenyl ether (PBDE)levels in livers of U.S. human fetuses and newborns. JToxicol Environ Health 70, 1–6.Smith BS, 1981: Male characteristics on female mud snailscaused by antifouling paints. J Appl Toxicol 1, 22–25.Smith SR, 1996: Agricultural recycling of sewage sludge andthe environment CAB International. Wallingford, UK.Stevens JL, Northcott GL, Stern GA, Tomy GT, Jones KC,2003: PAHs, PCBs, PCNs, organochlorine pesticides, syntheticmusks, and polychlorinated n-alkanes in UK sewagesludge: survey results and implications. Environ Sci Technol37, 462–467.Suk WA, Olden K, Yang RSH, 2002: Chemical mixturesresearch: significance and future perspectives. EnvironHealth Perspect 110(Suppl. 6), 891–892.Thornton I, Abrahams P, 1983: Soil ingestion – a majorpathway of heavy metals into livestock grazing contaminatedland. Sci Total Environ 28, 287–294.Thornton JW, Need E, Crews D, 2003: Ancient origin ofsteroid receptors and reconstitution of the ancestral receptor.Science 301, 1714–1717.Tinwell H, Ashby J, 2004: Sensitivity of the immature ratuterotrophic assay to mixtures of estrogens. Environ HealthPerspect 112, 575–582.Toppari J, Larsen JC, Christiansen P, Giwercman A, GrandjeanP, Guillette LJ, Jegou B, Jensen TK, Jouannet P,Keiding N, Leffers H, McLachlan JA, Meyer O, Muller J,Meyts ER, Scheike T, Sharpe R, Sumpter J, Skakkebaek NE,1996: Male reproductive health and environmental xenoestrogens.Environ Health Perspect 104(Suppl. 4), 741–803.U.S. Central Pollution Control Board Newsletter, 2001: http://www.cpcb.nic.in/aug2001vi.htmU.S. Department of Health and Services, 2005: Report onCarcinogens, 11th edn. Public Health Service, NationalToxicology Program (http://ntp.niehs.nih.gov/ntp/roc/toc11.html).Ó 2008 Macaulay Land Use Research Institute

22 SM RhindUzumcu M, Zachow R, 2007: Developmental exposure toenvironmental endocrine disruptors: consequences withinthe ovary and on female reproductive function. ReprodToxicol 23, 337–352.Vine MF, Stein L, Weigle K, Schroeder J, Degnan D, Chiu-kitJT, Hanchette C, Backer L, 2000: Effects on the immunesystem associated with living near a pesticide dump site.Environ Health Perspect 108, 1113–1124.Watkins JB, Klaassen CD, 1986: Xenobiotic biotransformationin livestock: comparison to other species commonlyused in toxicity testing. J Anim Sci 63, 933–942.Whyatt RM, Santella RM, Jedrychowski W, Garte SJ, BellDA, Ottman R, Gladek-Yarborough A, Cosma G, YoungT-L, Cooper TB, Randall MC, Manchester DK, Perera FP,1998: Relationshp between ambient air pollution and DNAdamage in Polish mothers and newborns. Environ HealthPerspect 106(Suppl. 3), 821–826.Wild SR, Jones KC, 1992: Organic chemicals enteringagricultural soils in sewage sludges: screening for theirpotential to transfer to crop plants and livestock. Sci TotalEnviron 119, 85–119.Wilkinson JM, Hill J, Phillips JC, 2003: The accumulation ofpotentially-toxic metals by grazing ruminants. Proc NutrSoc 62, 267–277.Author’s address (for correspondence): SM Rhind, MacaulayInstitute, Craigiebucker, Aberdeen, AB15 8QH, UK. E-mail:s.rhind@macaulay.ac.ukConflict of interest: The author declares no conflict of interest.Ó 2008 Macaulay Land Use Research Institute

Reprod Dom Anim 43 (Supp. 2), 23–30 (2008); doi: 10.1111/j.1439-0531.2008.01139.xISSN 0936-6768Long Chain Fatty Acids of Diet as Factors Influencing Reproduction in CattleJEP Santos 1 , TR Bilby 2 , WW Thatcher 1 , CR Staples 1 and FT Silvestre 11 Department of Animal Sciences, University of Florida, Gainesville, FL; 2 Department of Animal Sciences, University of Arizona, Tucson, AZ, USAContentsCattle are fed moderate amounts of long chain fatty acids (FA)with the objective to enhance lactation and growth; however,recent interest on lipid feeding to cows has focused onreproduction, immunity and health. Increasing the caloricdensity of the ration by fat feeding has generally improvedmeasures of cow reproduction, but when milk yield and bodyweight losses were increased by fat supplementation, positiveeffects on reproduction were not always observed. Feeding fathas influenced reproduction by altering the size of thedominant follicle, hastening the interval to first postpartumovulation in beef cows, increasing progesterone concentrationsduring the luteal phase of the oestrous cycle, modulatinguterine prostaglandin (PG) synthesis, and improving oocyteand embryo quality and developmental competence. Some ofthese effects were altered by the type of FA fed. Thepolyunsaturated FA of the n-6 and n-3 families seem to havethe most remarkable effects on reproductive responses ofcattle, but it is not completely clear whether these effects aremediated only by them or by other potential intermediatesproduced during rumen biohydrogenation. Generally, feedingfat sources rich in n-6 FA during late gestation and earlylactation enhanced follicle growth, uterine PG secretion,embryo quality and pregnancy in cows. Similarly, feedingn-3 FA during lactation suppressed uterine PG release, andimproved embryo quality and maintenance of pregnancy.Future research ought to focus on methods to improve thedelivery of specific FA and adequately powered studies shouldbe designed to critically evaluate their effects on establishmentand maintenance of pregnancy in cattle.IntroductionRuminant diets are supplemented with fat primarily toincrease energy concentration and to enhance animalperformance. Dairy and beef cattle diets, without anysupplemental fat, contain approximately 2% long-chainfatty acids (LCFA) of vegetable origin that are predominantlypolyunsaturated. Because of the high energydensity, fats are usually incorporated into cattle rationsto improve production, growth and reproduction.During early lactation, when lactating cows undergo aperiod of nutrient deficit, it was initially thought thatincorporating supplemental fat to the diet wouldenhance energy intake and energy balance, which wasexpected to improve reproduction. Because early lactationcows mobilize large quantities of stored triacylglycerolsin adipose tissue, concentrations of fatty acids(FA) in blood are usually high during the first weeks oflactation (Drackley 1999). This has been suggested tocause an unbalance in substrate supply to the cow,which compromises appetite and overall energy intake(Drackley 1999). When fat is fed in early lactation, oftencows either consume less diet or production increases,therefore fat feeding early postpartum seldom altersenergy status even though a more energy dense ration isconsumed. Staples et al. (1998) indicated that feeding fatdid not alter the energy status of dairy cows andsuggested that reproductive responses were the result ofsupplying LCFA and altering substrate availability tothe cow rather than simply an energy effect.As with other nutrients, certain FA are essential formammals. In 1929, George O. Burr and his wife werethe first to describe the essentiality of FA in rats(Burr and Burr 1929, 1930). They observed that growingrats fed diets low in fat ceased growing and experiencedhealth problems and irregular ovulation, which werethen reversed after feeding fat sources rich in thepolyunsaturated FA C18:2 n-6 (linoleic acid) andC18:3 n-3 (a-linolenic acid) (Burr and Burr 1930).Therefore, the concept of essential FA was establishedand later understood that C18:2 n-6 and C18:3 n-3 couldnot be synthesized by mammalian cells because of lackof desaturase enzymes beyond the 9th C in the acylchain. Because of the essentiality of FA and the role ofspecific FA on reproductive processes, it is possible thatreproduction in cattle may be more influenced by thetype of fat fed than fat feeding per se. This is particularlyimportant and challenging as ruminants extensivelyhydrogenate polyunsaturated FA, thereby limiting thesupply of dietary unsaturated FA for absorption in thesmall intestine.Feeding Fat and Fatty Acids to CattleLipids are important molecules that serve as a source ofenergy and are critical components of the physical andfunctional structure of cells. Lipids present in cellmembranes such as FA in phospholipids play animportant role in regulating the properties and activitiesof cell membranes. Changes in chain length, degree ofunsaturation and position of the double bonds in theacyl chain of FA can have remarkable impacts on theirfunction and may play a role in reproduction in cattle(Staples et al. 1998; Mattos et al. 2000), although theexact mechanisms are still unclear. Potential mechanismsmay include improved dietary energy density(Ferguson et al. 1990), altered follicle development(Staples and Thatcher 2005), increased concentrationsof progesterone (Staples et al. 1998), suppressed luteolyticsignals around maternal recognition of pregnancy(Mattos et al. 2000), and improved embryo quality(Cerri et al. 2004).The use of fat in diets of dairy cattle usuallyincreases the energy density of the ration and improveslactation and reproduction, although improvements inreproduction occur in spite of provision of calories(Staples et al. 1998). These effects might be mediatedby the FA composition of the fat source; however, aÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

24 JEP Santos, TR Bilby, WW Thatcher, CR Staples and FT Silvestremajor impediment to the study of FA and reproductionin cattle is the inability to predict the delivery ofspecific lipids, particularly polyunsaturated FA to thesmall intestine for absorption, and the specific needs ofdifferent tissues for FA to modulate reproduction.Microbial activity in the rumen results in lipolysis oftriacylglycerols and biohydrogenation of unsaturatedFA which dramatically reduces the amount of polyunsaturatedFA reaching the small intestine forabsorption. In fact, Juchem (2007) demonstrated thatmore than 70% of the C18:2 n-6 and more than 85%of C18:3 n-3 fed to lactating cows were biohydrogenatedin the rumen when fed as unprotected oils oras Ca salts of long chain FA (Ca-LCFA), respectively.Therefore, if specific unsaturated FA are important forreproduction in cattle, it is critical that future researchwith lipids and reproduction aim to improve the extentof delivery of unsaturated FA for absorption. In spiteof the difficulties to deliver polyunsaturated FA toruminants, studies have generally indicated that thepolyunsaturated FA of the n-6 (linoleic acid) and n-3[a-linolenic acid; eicosapentaenoic (EPA), C20:5 n-3;docosahexaenoic (DHA), C22:6 n-3] families are themost beneficial to improving reproduction in cows.Fatty Acids and Postpartum Uterine HealthUterine health is an important risk factor for subsequentfertility in lactating dairy cows. During the process ofparturition, eicosanoids are produced in substantialamounts and play an important role in regulation andcontrol of parturition, and expulsion of the placenta anduterine contents through opening of the cervixand contractions of the uterus. Prostaglandin F 2a is animportant eicosanoid that regulates CL lifespan andmight influence retention of foetal membranes andsubsequent uterine health. Uterine synthesis of PGF 2ais regulated in part by substrate availability, andarachidonic acid (AA; C20:4 n-6) is the precursor forPGF 2a synthesis, so it is plausible to suggest thatincrements in AA content of endometrial tissue shouldenhance uterine PGF 2a secretion, which in turn mayinfluence uterine health.Burns et al. (2003) fed non-lactating beef cows n-3FA from fish meal and reduced the endometrialconcentration of AA and increased those of EPA andtotal n-3 FA. Similar effects have been observed withlactating dairy cows fed increasing amounts of fishmeal or Ca-LCFA enriched in fish oil (Bilby et al.2006b; Moussavi et al. 2007). Because of incorporationof n-6 and n-3 FA primarily in the phospholipidcomponent of endometrial tissue, it is possible thatchanges in FA content of the endometrial tissue mightmodulate endometrial secretion of PGF 2a in cows.Feeding approximately 2% of the ration as fish oil richin n-3 FA reduced the peripheral blood concentrationsof PGF 2a metabolite (PGFM) indicating reduceduterine secretion of PGF 2a (Mattos et al. 2004). Incontrast feeding supplemental fat pre-partum containingapproximately 30% of FA as C18:2 n-6 increaseduterine secretion of PGF 2a based on PGFM in blood(Cullens et al. 2004). Increased synthesis of PGF 2awhen cows were supplemented with n-6 FA pre-partummight enhance the potential for uterine and immunecells to secrete eicosanoids which may influencepostpartum uterine health and immuno-competenceof the cow. Collectively, these data indicate thatfeeding fat sources differing in FA profile during thetransition period can influence the natural release ofPGF 2a by the uterus of the cow.Three studies examined the effect of feeding fat prepartumon postpartum health of dairy cows (Cullenset al. 2004; Juchem 2007; Silvestre, unpublished data).When cows were supplemented with Ca-LCFA rich inn-6 FA pre-partum, incidence of postpartum diseasesincluding retained placenta, metritis and mastitis wasreduced (8.3% vs 42.9%) compared with cows not fedfat pre-partum (Cullens et al. 2004). Juchem (2007)supplemented the diet of 501 pre-partum dairy cowswith 2% Ca-LCFA of either palm oil or a blend ofC18:2 n-6 and trans-octadecenoic FA. Incidence ofretained placenta did not differ between treatments(6.6%). Risk of uterine disease was similar betweensources of FA, but cows fed the blend of C18:2 n-6and trans-octadecenoic FA had reduced the odds ofpuerperal metritis (8.8% vs 15.1%; adjusted oddsratio ¼ 0.53). Rate of uterine involution did not differ,and 91% of the cows had completed uterine involutionat the last ultrasonography on week 6 postpartum(Juchem 2007). In a similar attempt, Silvestre (unpublisheddata) fed 1167 pre-partum dairy cows 1.5% ofthe ration as Ca-LCFA of either palm oil or saffloweroil. Feeding a fat source rich in C18:2 n6 enhancedmeasures of innate immunity; however, incidences ofretained placenta (10.1%), metritis (17.4%), andpurulent cervical discharge (29%) did not differbetween treatments. These data suggest that, althoughfeeding fat sources rich in n-6 FA may enhanceimmune responses and have pro-inflammatory effects,its impacts on uterine health are subtle.Fatty Acids, Follicle Development andResumption of Postpartum CyclicityOne of the mechanisms by which fat feeding mightimprove fertility in cattle is by influencing folliclegrowth and ovulation (Lucy et al. 1993). Lucy et al.(1991) replaced corn with Ca-LCFA in the diet fed todairy cows beginning at parturition, and feedingCa-LCFA increased the number of medium (6–9 mm)sized follicles within 25 days postpartum, and that offollicles >15 mm in a synchronized oestrous cycle. Inaddition, diameter of the largest (18.2 vs 12.4 mm)follicle was greater in cows fed Ca-LCFA. When thisstudy was repeated with isocaloric diets, similar effectswere observed (Lucy et al. 1993). Staples and Thatcher(2005) summarized the effects of supplemental fats onthe size of the dominant follicle (Table 1). On average,dominant follicle diameter was 3.2 mm larger, whichrepresents a 23% increase in fat supplemented cows.Several studies have shown that dominant folliclediameter increased in cows fed diets enriched inpolyunsaturated FA compared with monounsaturatedFA, suggesting differential effects of FA on folliclegrowth (Staples et al. 2000; Bilby et al. 2006a). Folliclesfrom cows abomasally infused with yellow greaseÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Factors Influencing Reproduction in Cattle 25Table 1. Effect of supplemental fat on the diameter of the dominantovarian follicle of lactating dairy cows (from Staples and Thatcher2005)ReferenceFatsourceExperimental dietsControl (mm)Fat (mm)Ambrose et al. (2006) Rolled flaxseeds 14.1 16.9Beam and Butler (1997) Tallow, yellow grease 11.0 13.5Bilby et al. (2006a) Ca-LCFA or flaxseed oil 15.0 16.5Lucy et al. (1991) Ca-LCFA 12.4 18.2Lucy et al. (1993) Ca-LCFA 16.0 18.6Oldick et al. (1997) Yellow grease 16.9 20.9Robinson et al. (2002) Protected soybeans 13.3 16.9Staples et al. (2000) Soybean oil, fish oil 14.3 17.1Average 14.1 17.3Ca-LCFA = Ca salts of long chain fatty acids from palm oil or a blend of palmand soyabean oils.Control vs fat was P < 0.10 for each study.grew faster to a larger diameter than follicles fromcows infused with tallow (Oldick et al. 1997). Itappears that fat feeding, but more importantly typeof fat, stimulates follicle growth in cows. The impact oflarger ovarian follicles on fertility because of fatsupplementation has not been defined, but cowsexperiencing earlier postpartum ovulation have beenreported to have larger follicles (Beam and Butler1997). Therefore, it is possible that increasing thenumber and size of larger follicles by feeding fat canreduce the interval from calving to first postpartumovulation, which has been observed for postpartumbeef cows (Lammoglia et al. 1996, 1997; De Fries et al.1998).Although some studies have indicated that feeding fathastens follicle growth, which might influence resumptionof postpartum ovulation (Lammoglia et al. 1996,1997; De Fries et al. 1998), it is unclear whethersupplemental fats differing in FA profile have anydifferential effect on resumption of cyclicity. Juchem(2007) fed 699 multiparous cows either 400 g of FAfrom tallow or from Ca-LCFA containing palm and fishoils and observed no difference in proportion of cyclingcows at 65 days postpartum (83.2% vs 82.2%, respectively).Subsequently, dairy cows supplemented withCa-LCFA of palm oil or a blend of C18:2 n-6 and transoctadecenoicFA from 25 days pre-partum to 80 dayspostpartum experienced a similar mean interval to firstpostpartum ovulation (30.5 and 32.2 days, respectively;Juchem 2007). Recently, Silvestre (unpublished data) fedcows (n = 1055) either Ca-LCFA of palm oil or ofsafflower oil from 2 weeks pre-partum to 4 weekspostpartum, and then half of the cows in each transitiontreatment group were switched to either Ca-LCFA ofpalm oil or fish oil (264 ⁄ treatment). The proportions ofcyclic cows at 63 days postpartum were 84.2%, 79.5%,79.2% and 77.1% for cows fed palm oil ⁄ palm oil, palmoil ⁄ fish oil, safflower ⁄ palm oil, and safflower ⁄ fish oil,respectively, and they did not differ. Taken together,these data demonstrate that type of supplemental FA,whether more saturated or unsaturated does not influenceresumption of postpartum cyclicity in lactatingdairy cows.Fatty Acids and OestradiolOestradiol has stimulatory effects on uterine secretion ofPGF 2a (Knickerbocker et al. 1986), and can increase thesensitivity of the CL to PGF 2a (Howard et al. 1990)which may enhance regression of the CL. Thus loweredplasma oestradiol may help prevent pre-mature CLregression and early embryonic mortality. Oldick et al.(1997) reported that abomasal infusion of tallow oryellow grease reduced concentrations of plasma oestradiolon days 15 to 20 of a synchronized oestrous cyclecompared with cows infused with glucose, a responsethat also has been observed in beef cows supplementedwith lipids (Hightshoe et al. 1991). Also, oestradiolconcentration was reduced in the follicular fluid frombeef cows fed soybean oil (Ryan et al. 1992). Although areduction in follicular oestradiol caused by fat feedingmight potentially benefit CL lifespan, it may be detrimentalto expression of oestrus and uterine primingduring prooestrus.Fat and Luteal FunctionImproved fertility in cattle has been associated withincreased circulating concentrations of progesteroneduring the luteal phase before and after AI. Additionof fat to cattle diets has consistently shown to increaseplasma cholesterol and cholesterol content in follicularfluid and in the CL (Staples et al. 1998; Williams 1989;Ryan et al. 1992; Hawkins et al. 1995; Lammoglia et al.1996). Cholesterol serves as a precursor for the synthesisof progesterone by ovarian cells and both high and lowdensity lipoproteins deliver cholesterol to ovarian tissuesfor steroidogenesis (Grummer and Carroll 1991).Hypercholesterolemia may increase CL steroidogenesis;however, the increased plasma progesterone concentrationsin dairy and beef cows fed fat (Table 2) may beexplained possibly by reduced progesterone clearance,not by increased synthesis (Hawkins et al. 1995).Fatty Acids, Oocyte Quality and MembraneCompositionCompetence of the oocyte and embryo is related to FAcomposition; specifically, phospholipid content of thecellular membrane plays a vital role in developmentduring and after fertilization. The amount of lipid in theruminant oocyte is approximately 20-fold greater thanthat of the mouse (76 vs 4 ng) and consists (w ⁄ w) ofapproximately 50% triacylglycerol, 20% phospholipid,20% cholesterol and 10% free FA (McEvoy et al. 2000).Previous studies showed that C16:0 and C18:1 were themost abundant FA in the phospholipid fraction ofoocytes from cattle and may function as an energyreserve (Kim et al. 2001; Zeron et al. 2001). PolyunsaturatedFA comprised

26 JEP Santos, TR Bilby, WW Thatcher, CR Staples and FT SilvestreTable 2. Effect of supplemental fat on plasma progesterone concentrationsin dairy and beef cattleMeasurementProgesterone,ng ⁄ mlControlp

Factors Influencing Reproduction in Cattle 27Feeding n-3 FA can attenuate endometrial PGF 2aproduction. Feeding fish meal to dairy cows attenuatedthe decrease in plasma progesterone concentrations2 days after PGF 2a injection suggesting changes in CLregression because of fish oil FA (Burke et al. 1996).Mattos et al. (2002) demonstrated that FA from fishmeal reduced plasma PGFM concentrations comparedwith unsupplemented cows after an oestradiol ⁄ oxytocinchallenge. Dairy cows fed fish oil during the transitionperiod had greater EPA and DHA concentrations incaruncular tissues and reduced postpartum concentrationsof PGFM compared with cows fed olive oil(Mattos et al. 2004). Conversely, feeding fat sources richin n-6 FA increased plasma PGFM after an oxytocinchallenge (Robinson et al. 2002; Petit et al. 2004). Thussupplemental lipids can either inhibit or stimulate PGsecretion depending upon the specific FA.Incubation of bovine endometrial cells with AA stimulatedPGF 2a production compared with cells not supplementedwith FA. On the other hand, cellssupplemented with n-3 FA had reduced secretion ofPGF 2a (Mattos et al. 2003). The mechanism by which n-3FA inhibit PGF 2a secretion may involve decreasing theavailability of AA precursor, increasing the concentrationof FA that compete with AA for processing by PGHS-2,or inhibition of PGHS-2 (Mattos et al. 2000). Bilby et al.(2006b) concluded that n-3 FA supplementation tolactating dairy cows had little effect on the endometrialcomponents that regulate the PG cascade. Instead EPAand DHA exerted their regulatory effects as alternativesubstrates that reduced the lipid pools of AA. In supportof these conclusions, Burns et al. (2003) demonstratedthat feeding fish meal reduced the endometrial concentrationof AA and increased those of EPA and total n-3FA. Therefore, different FA can alter PGF 2a secretion byinfluencing FA availability in the endometrial tissue, andsupplying FA that inhibit PGF 2a release by the uterusmight improve the mechanism of embryo preservation,which may benefit embryonic survival in cattle.Fatty Acids and Fertility of CowsStudies evaluating the effects of supplemental fat onreproductive performance of beef cattle are limited. Toour knowledge, no controlled trials have been conductedwith adequate number of animals to evaluate thepotential for fat supplementation to impact establishmentand maintenance of pregnancy of beef cows. DeFries et al. (1998) observed a tendency (p = 0.09) forincreased pregnancy in Brahman cows fed 5.2% fatcompared with cows fed 3.7% fat in the diet; however,the number of cows used in this study was limited toonly 20 per treatment.Feeding fat to dairy cattle might improve pregnancyper AI (Table 3), although responses have not beenconsistent. When fat feeding increased postpartum bodyweight loss, primiparous cows fed fat had reducedpregnancy at first AI (Sklan et al. 1994). However,Ferguson et al. (1990) observed a 2.2-fold increased odds(odds ratio = 2.2) of becoming pregnant at first and allAI in lactating cows fed 0.5 kg ⁄ day of fat, which tended(p = 0.08) to enhance the proportion of pregnant cowsat the end of the study (93% vs 86.2%). In grazing cows,Table 3. Effect of fat supplementation on pregnancy at first postpartumAI in lactating dairy cowsReferenceCowsFatsource andamountPregnancy perAI, %ControlFerguson et al. (1990) 253 0.5 kg of saturated 42.6 59.1 *free FAMcNamara et al. (2003) 201 0.32 to 0.36 kg 35.5 51.1 *of FA fromCa-LCFASchingoethe153 Oilseeds 46.5 42.0and Casper (1991)Schneider et al. (1988) 181 0.5 kg of Ca-LCFA 43.1 60.5Scott et al. (1995) 443 0.45 kg of Ca-LCFA 49.3 45.7Sklan et al. (1991) 99 2.6% of ration 41.5 39.2as Ca-LCFASklan et al. (1994) 102 2.5 of ration asCa-LCFAPrimiparous 73.7 * 33.3Multiparous 42.1 33.3FA = fatty acid; Ca-LCFA = Ca salts of long chain fatty acids from palm oil.* Within a row, effect of supplemental fat (p < 0.05).supplementation with 0.35 kg of FA improved pregnancyafter the first postpartum AI, although theproportion of cows pregnant at the end of the study didnot differ (McNamara et al. 2003). Feeding Ca-LCFA ofpalm oil improved pregnancy of dairy cows (Schneideret al. 1988; Sklan et al. 1991), but the authors did notreport statistical significance. On the other hand, othersdid not observe improvements on fertility of dairy cowssupplemented with Ca-LCFA (Sklan et al. 1994; Scottet al. 1995) or oilseeds (Schingoethe and Casper 1991),which might be attributed to increased milk yield andbody weight losses (Sklan et al. 1991, 1994).Because the benefits of feeding fat may originate fromspecific FA (Staples et al. 1998; Staples and Thatcher2005), and absorption of unsaturated FA is limited inruminants because of microbial biohydrogenation in therumen (Juchem 2007), studies have evaluated whetherfeeding FA differing in the degree of saturation mightinfluence fertility of dairy cows (Table 4). When cowswere fed 0.75 kg of fat from flaxseed, a source rich inC18:3 n-3, or sunflower seed, a source rich in C18:2 n-6,pregnancy tended (p = 0.07) to be greater for cows fedn-3 FA (Ambrose et al. 2006). However, a similarresponse to flaxseed was not observed by others (Petitand Twagiramungu 2006; Fuentes et al. 2008). Similarly,feeding n-3 FA from fish oil as Ca-LCFA did notimprove pregnancy at first postpartum AI when comparedwith feeding beef tallow (Juchem 2007) or withCa-LCFA of palm oil (Silvestre, unpublished data),although pregnancy at second postpartum AI wasgreater for cows fed n-3 FA (Silvestre, unpublisheddata). Juchem (2007) evaluated the effect of feeding preandpostpartum cows Ca-LCFA of palm oil or a blendof C18:2 n-6 and trans-octadecenoic FA. Cows fedunsaturated FA were 1.5 times more likely to bepregnant at 27 or 41 days after AI compared with cowsfed palm oil. Improvements in pregnancy when cowswere fed Ca salts of a mix of C18:2 n-6 and transoctadecenoicFA were supported by increased fertilizationand embryo quality in non-superovulated lactatingdairy cows (Cerri et al. 2004).FatÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

28 JEP Santos, TR Bilby, WW Thatcher, CR Staples and FT SilvestreTable 4. Effect of fatty acid (FA) supplementation on pregnancy per AI in lactating dairy cowsFA source – Pregnancy per AI, %Reference Cows Amount ⁄ daySaturated n-6 FA n-3 FAAmbrose et al. (2006) 121 0.75 kg of fat – 32.2 48.4**Fuentes et al. (2008) 356 0.40 kg of fat – 39.2 38.8Juchem (2007) 699 0.40 kg of FA 40.7 – 35.9Juchem (2007) 323 0.40 kg of FA 22.8 – 24.8Juchem (2007) 344 2% of ration 28.6 37.9** –Petit and Twagiramungu (2006) 110 0.6–0.8 kg of FA 55.9 40.0 44.4Silvestre (unpublished data) – first AI 1055 1.5% of rationTransition 39.0 35.9 –>30 d postpartum 37.3 – 37.6Silvestre (unpublished data) – second AI 604 1.5% of rationTransition 29.0 34.5 –>30 d postpartum 27.2 – 37.0 *Saturated = mostly saturated and monounsaturated FA; n-6 FA = source rich in C18:2 n-6; n-3 FA = source rich in C18:3 n-3 or C20:5 n-3 + C22:6 n-3.* Within a row, effect of source of FA (p < 0.05).** Within a row, effect of source of FA (p < 0.07).Table 5. Effect of fatty acid (FA) supplementation on pregnancy losses after first postpartum AI in lactating dairy cowsFA source – Pregnancy loss, %Reference Pregnancies Amount ⁄ daySaturated n-6 FA n-3 FAAmbrose et al. (2006) 77 0.75 kg of fat – 27.3 9.8 *Juchem (2007) 257 0.40 kg of FA 20.4 – 23.5Juchem (2007) 77 0.40 kg of FA 5.4 – 10.0Juchem (2007) 114 2% of the ration 9.8 6.3 –Petit and Twagiramungu (2006) 51 0.6–0.8 kg of FA 21.1 12.5 0 *Silvestre (unpublished data) 388 1.5% of rationTransition 8.3 12.1 –>30 d postpartum 13.6 – 6.3 *Saturated = mostly saturated and monounsaturated FA; n-6 FA = source rich in C18:2 n-6; n-3 FA = source rich in C18:3 n-3 or C20:5 n-3 + C22:6 n-3.* Within a row, effect of source of FA (p < 0.05).Because n-3 FA can suppress uterine secretion ofPGF 2a (Mattos et al. 2002, 2003, 2004), it may improveembryonic survival in cattle (Mattos et al. 2000). Inthree of five experiments, feeding the n-3 FA C18:3 n-3(Ambrose et al. 2006; Petit and Twagiramungu 2006) orEPA and DHA (Silvestre, unpublished data) reducedpregnancy losses in lactating dairy cows (Table 5). Onthe other hand, when n-6 FA were fed as Ca-LCFA,pregnancy losses were similar to those observed for cowsfed Ca-LCFA of palm oil (Juchem 2007; Silvestre,unpublished data).Collectively, these data suggest that feeding fat to dairycows generally improves fertility and responses areobserved with the energy increment in the diet; also, thesedata suggest that fertility responses to fat feeding arealtered according to the type of dietary FA, althoughresponses are not always consistent. Feeding n-3 FA fromflaxseeds or as Ca-LCFA improved pregnancy per AI insome, but not all studies. Similarly, feeding Ca-LCFArich in n-6 FA improved pregnancy per AI in one of twoexperiments with lactating dairy cows. Although feedingn-3 FA has not consistently increased the risk ofpregnancy, it has reduced pregnancy losses in dairy cows.ConclusionsFat is recommended to be incorporated into dairy cattlediets at moderate amounts. Feeding fat to cattlegenerally improved establishment and maintenance ofpregnancy, but benefits to fertility can be negated whenweight losses are exacesbated by fat feeding. Potentialimprovements in fertility of cows caused by fat feedinghave generally been associated with enhanced follicledevelopment postpartum, increased diameter of theovulatory follicle, increased progesterone concentrationsduring the luteal phase of the cycle, altered uterine⁄ embryo cross-talk by modulating PG synthesis,and improved oocyte and embryo quality. Some ofthese effects have been more influenced by the type offatty acid than by fat feeding per se. Differentialresponses in vivo to FA feeding suggest that unsaturatedFA of the n-6 and n-3 families were most beneficial.AcknowledgementsPart of the research mentioned on this manuscript and conducted bythe authors was supported by the National Research InitiativeCompetitive Grant no. 2004-35203-14137 from the USDA CooperativeState Research, Education, and Extension Service.ReferencesAmbrose DK, Kastelic JP, Corbett R, Pitney PA, Petit HV,Small JA, Zalkovic P, 2006: Lower pregnancy losses inlactating dairy cows fed a diet enriched in a-linolenic acid.J Dairy Sci 89, 3066–3074.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Factors Influencing Reproduction in Cattle 29Balaguer SA, Pershing RA, Rodriguez-Sallaberry C, ThatcherWW, Badinga L, 2005: Effects of bovine somatotropin onuterine genes related to the prostaglandin cascade inlactating dairy cows. J Dairy Sci 88, 543–552.Beam SW, Butler WR, 1997: Energy balance and ovarianfollicle development prior to the first ovulation postpartumin dairy cows receiving three levels of dietary fat. BiolReprod 56, 133–142.Bilby TR, Block J, do Amaral BC, Sa Filho O, Silvestre FT,Hansen PJ, Staples CR, Thatcher WW, 2006a: Effects ofdietary unsaturated fatty acids on oocyte quality andfollicular development in lactating dairy cows in summerJ. Dairy Sci 89, 3891–3903.Bilby TR, Guzeloglu A, MacLaren LA, Staples CR, ThatcherWW, 2006b: Pregnancy, bST and omega-3 fatty acids inlactating dairy cows: II. Gene expression related to maintenanceof pregnancy. J Dairy Sci 89, 3375–3385.Burke JM, Staples CR, Risco CA, De La Sota RL, ThatcherWW, 1996: Effect of ruminant grade menhaden fish meal onreproductive and productive performance of lactating dairycows. J Dairy Sci 80, 3386–3398.Burns PD, Engle TE, Harris MA, Enns RM, Whittier JC,2003: Effect of fish meal supplementation on plasma andendometrial fatty acid composition in nonlactating beefcows. J Anim Sci 81, 2840–2846.Burr GO, Burr MM, 1929: A new disease produced bythe rigid exclusion of fat from the diet. J Biol Chem 82, 345–367.Burr GO, Burr MM, 1930: The nature and role of the fattyacids essential in nutrition. J Biol Chem 86, 587–621.Carroll DJ, Jerred MJ, Grummer RR, Combs DK, PiersonRA, Hauser ER, 1990: Effects of fat supplementation andimmature alfalfa to concentrate ratio on plasma progesterone,energy balance, and reproductive traits of dairy cattle.J Dairy Sci 73, 2855–2863.Cerri RLA, Bruno RGS, Chebel RC, Galva˜ o KN, RutglianoH, Juchem SO, Thatcher WW, Luchini D, Santos JEP, 2004:Effect of fat sources differing in fatty acid profile onfertilization rate and embryo quality in lactating dairy cows.J Dairy Sci 87(Suppl. 1), 297 (abstr).Cullens FM, Staples CR, Bilby TR, Silvestre FT, Bartolome J,Sozzi A, Badinga L, Thatcher WW, Arthington JD, 2004:Effect of timing of initiation of fat supplementation on milkproduction, plasma hormones and metabolites, and conceptionrates of Holstein cows in summer. J Dairy Sci 86(Suppl.1), 308 (abstr).De Fries CA, Neuendorff DA, Randel RD, 1998: Fatsupplementation influences postpartum reproductive performancein Brahman cows. J Anim Sci 76, 864–870.Drackley JK, 1999: ADSA foundation scholar award. Biologyof dairy cows during the transition period: the final frontier?J Dairy Sci 82, 2259–2273.Ferguson JD, Sklan D, Chalupa WV, Kronfeld DS, 1990:Effects of hard fats on in vitro and in vivo rumenfermentation, milk production, and reproduction in dairycows. J Dairy Sci 73, 2864–2879.Fouladi-Nashta AA, Gutierrez CG, Gong JG, GarnsworthyPC, Webb R, 2007: Impact of dietary fatty acids on oocytequality and development in lactating dairy cows. BiolReprod 77, 9–17.Fuentes MC, Calsamiglia SS, Sa´nchez C, Gonza´lez A,Newbold J, Santos JEP, Rodrı´guez-Alcala´ LM, FontechaJ, 2008: Effect of extruded linseed on productive andreproductive performance of lactating dairy cows. LivestSci 113, 144–154.Garcia-Bojalil CM, Staples CR, Risco CA, Savio JD, ThatcherWW, 1998: Protein degradability and calcium salts of longchainfatty acids in the diets of lactating dairy cows:productive responses. J Dairy Sci 81, 1374–1384.Grummer RR, Carroll DJ, 1991: Effects of dietary fat onmetabolic disorders and reproductive performance of dairycattle. J Anim Sci 69, 3838–3852.Hawkins DE, Niswender KD, Oss GM, Moeller CL, OddeKG, Sawyer HR, Niswender GD, 1995: An increase inserum lipids increases luteal lipid content and alters thedisappearance rate of progesterone in cows. J Anim Sci 73,541–545.Hightshoe RB, Cochran RC, Corah LR, Kiracofe GH,Harmon DL, Perry RC, 1991: Effects of calcium soaps offatty acids on postpartum reproductive function in beefcows. J Anim Sci 69, 4097–4103.Howard JJ, Scott RG, Britt JH, 1990: Associations amongprogesterone, estradiol-17b, and prostaglandin in cattletreated with hCG during diestrus to extend corpus luteumfunction. Prostaglandins 40, 51–70.Juchem SO, 2007: Lipid Digestion and Metabolism in DairyCows: Effects on Production, Reproduction and Health.PhD thesis, University of California Davis.Kim JY, Kinoshita M, Ohnishi M, Fukui Y, 2001: Lipid andfatty acid analysis of fresh and frozen-thawed immature andin vitro matured bovine oocytes. Reproduction 122, 131–138.Knickerbocker JJ, Thatcher WW, Foster DB, Wolfenson D,Bartol FF, Caton D, 1986: Uterine prostaglandin and bloodflow responses to estradiol-17b in cyclic cattle. Prostaglandins31, 757–776.Lammoglia MA, Willard ST, Oldham JR, Randel RD, 1996:Effects of dietary fat and season on steroid hormonalprofiles before parturition and on hormonal, cholesterol,triglycerides, follicular pattern, and postpartum reproductionin Brahman cows. J Anim Sci 74, 2253–2262.Lammoglia MA, Williard ST, Hallford DM, Randel RD,1997: Effects of dietary fat on follicular development andcirculating concentrations of lipids, insulin, progesterone,estradiol-17b, 13,14-dihydro-15-keto-prostaglandin F2a andgrowth hormone in estrous cyclic Brahman cows. J Anim Sci75, 1591–1600.Lim H, Gupta RA, Ma WG, Paria BC, Moller DE, MorrowJD, BuBois RN, Trzaskos JM, Dey SK, 1999: Cyclooxygenase-2-derivedprostacyclin mediates embryo implantationin the mouse via PPARd. Gene Dev 13, 1561–1574.Lucy MC, Staples CR, Michel FM, Thatcher WW, Bolt DJ,1991: Effect of feeding calcium soaps to early postpartumdairy cows on plasma prostaglandin F 2a , luteinizing hormone,and follicular growth. J Dairy Sci 74, 483–489.Lucy MC, De La Sota RL, Staples CR, Thatcher WW, 1993:Ovarian follicular populations in lactating dairy cowstreated with recombinant bovine somatotropin (sometribove)or saline and fed diets differing in fat content and energy.J Dairy Sci 76, 1014–1027.MacLaren LA, Guzeloglu A, Michel F, Thatcher WW, 2006:Peroxisome proliferator-activated receptor (PPAR) expressionin cultured bovine endometrial cells and response toomega-3 fatty acid, growth hormone and agonist stimulationin relation to series 2 prostaglandin production. DomAnim Endocr 30, 155–169.Mattos R, Staples CR, Thatcher WW, 2000: Effects of dietaryfatty acids on reproduction in ruminants. Rev Reprod 5, 38–45.Mattos R, Staples CR, Williams J, Amorocho A, McGuireMA, Thatcher WW, 2002: Uterine, ovarian, and productionresponses of lactating dairy cows to increasing dietaryconcentrations of menhaden fish meal. J Dairy Sci 85, 755–764.Mattos R, Guzeloglu A, Badinga L, Staples CR, ThatcherWW, 2003: Polyunsaturated fatty acids and bovine interferon-smodify phorbol ester-induced secretion of prostaglandinF2aand expression of prostaglandin endoperoxideÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

30 JEP Santos, TR Bilby, WW Thatcher, CR Staples and FT Silvestresynthase-2 and phospholipase-A2 in bovine endometrialcells. Biol Reprod 69, 780–787.Mattos R, Staples CR, Arteche A, Wiltbank MC, Diaz FJ,Jenkins TC, Thatcher WW, 2004: The effects of feeding fishoil on uterine secretion of PGF 2a , milk composition, andmetabolic status of periparturient Holstein cows. J Dairy Sci87, 921–932.McEvoy TG, Coull GD, Broadbent PJ, Hutchinson JSM,Speake BK, 2000: Fatty acid composition of lipids inimmature cattle, pig and sheep oocytes with intact zonapellucida. J Reprod Fertil 118, 163–170.McNamara S, Butler T, Ryan DP, Mee JF, Dillon P, O’MaraFP, Butler ST, Anglese D, Rath M, Murphy JJ, 2003: Effectof offering rumen-protected fat supplements on fertility andperformance in spring-calving Holstein–Friesian cows.Anim Reprod Sci 79, 45–56.Moussavi AR, Gilbert RO, Overton TR, Bauman DE, ButlerWR, 2007: Effects of feeding fish meal and n-3 fatty acids onovarian and uterine responses in early lactating dairy cows.J Dairy Sci 90, 145–154.Oldick BS, Staples CR, Thatcher WW, Gyawu P, 1997:Abomasal infusion of glucose and fat-effect on digestion,production, and ovarian and uterine function of cows. JDairy Sci 80, 1315–1328.Petit HV, Twagiramungu H, 2006: Conception rate andreproductive function of dairy cows fed different fat sources.Theriogenology 66, 1316–1324.Petit HV, Germiquet C, Lebel D, 2004: Effect of feedingwhole, unprocessed sunflower seeds and flaxseed on milkproduction, milk composition, and prostaglandin secretionin dairy cows. J Dairy Sci 87, 3889–3898.Robinson RS, Pushpakumara PGA, Cheng Z, Peters AR,Abayasekara DEE, Wathes DC, 2002: Effects of dietarypolyunsaturated fatty acids on ovarian and uterine functionin lactating dairy cows. Reproduction 124, 119–131.Ryan DP, Spoon RA, Williams GL, 1992: Ovarian follicularcharacteristics, embryo recovery, and embryo viability inheifers fed high fat diets and treated with follicle-stimulatinghormone. J Anim Sci 70, 3505–3513.Schingoethe DJ, Casper DP, 1991: Total lactational responseto added fat during early lactation. J Dairy Sci 74, 2617–2622.Schneider P, Sklan D, Chalupa W, Kronfeld DS, 1988:Feeding calcium salts of fatty acids to lactating cows. JDairy Sci 71, 2143–2150.Scott TA, Shaver RD, Zepeda L, Yandell B, Smith TR, 1995:Effects of rumen-inert fat on lactation, reproduction, andhealth of high producing Holstein herds. J Dairy Sci 78,2435–2451.Sklan D, Moallem U, Folman Y, 1991: Effect of feedingcalcium soaps of fatty acids on production and reproductiveresponses in high producing lactating cows. J Dairy Sci 74,510–517.Sklan D, Kaim M, Moallem U, Folman Y, 1994: Effect ofdietary calcium soaps on milk yield, body weight, reproductivehormones, and fertility in first parity and older cows.J Dairy Sci 77, 1652–1660.Son J, Grant RJ, Larson LL, 1996: Effects of tallow and escapeprotein on lactational and reproductive performance ofdairy cows. J Dairy Sci 79, 822–830.Spicer LJ, Vernon RK, Tucker WB, Wettemann RP, HogueJF, Adams GD, 1993: Effects of inert fat on energy balance,plasma concentrations of hormones, and reproduction indairy cows. J Dairy Sci 76, 2664–2673.Staples CR, Thatcher WW, 2005: Effects of fatty acids onreproduction of dairy cows. In: Garnsworthy PC, WisemanJ(eds), Recent Advances in Animal Nutrition. NottinghamUniversity Press, Nottingham, UK, pp. 229–256.Staples CR, Burke JM, Thatcher WW, 1998: Influence ofsupplemental fats on reproductive tissues and performanceof lactating cows. J Dairy Sci 81, 856–871.Staples CR, Wiltbank MC, Grummer RR, Guenther J, SartoriR, Diaz FJ, Bertics S, Mattos R, Thatcher WW, 2000: Effectof long chain fatty acids on lactation performance andreproductive tissues of Holstein cows. J Dairy Sci 83(Suppl.1), 278 (Abstr.)Thangavelu G, Colazo MG, Ambrose DJ, Oba M, Okine EK,Dyck MK, 2007: Diets enriched in unsaturated fatty acidsenhance early embryonic development in lactating Holsteincows. Theriogenology 68, 949–957.Thomas MG, Williams GL, 1996: Metabolic hormone secretionand FSH-induced superovulatory responses of beefheifers fed dietary supplements containing predominantlysaturated or polyunsaturated fatty acids. Theriogenology45, 451–458.Williams GL, 1989: Modulation of luteal activity in postpartumbeef cows through changes in dietary lipid. J AnimSci 67, 785–793.Zeron Y, Ocheretny A, Kedar O, Borochov A, Sklan D, AravA, 2001: Seasonal changes in bovine fertility: relation todevelopmental competence of oocytes, membrane propertiesand fatty acid composition of follicles. Reproduction 121,447–454.Zeron Y, Sklan D, Arav A, 2002: Effect of polyunsaturated fattyacid supplementation on biophysical parameters and chillingsensitivity of ewe oocytes. Mol Reprod Dev 61, 271–278.Author’s address (for correspondence): JEP Santos, LE ‘‘Red’’ LassonBuilding, Department of Animal Sciences, University of Florida,Gainesville, FL, 32611 USA. E-mail: jepsantos@ufl.eduConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 31–39 (2008); doi: 10.1111/j.1439-0531.2008.01140.xISSN 0936-6768Functional Differences in the Growth Hormone and Insulin-like Growth Factor Axisin Cattle and Pigs: Implications for Post-partum Nutrition and ReproductionMC LucyDivision of Animal Sciences, University of Missouri, Columbia, MO, USAContentsGrowth hormone (GH) and insulin-like growth factor-I(IGF-I) control growth and lactation in cattle and swine.Insulin participates in the endocrinology of growth andlactation because insulin and GH are antagonistic in theiractions. Dairy cows experience a period of negative energybalance during the first 4–8 weeks post-partum. During thisperiod, their somatotropic axis (comprised of GH, the GHreceptor and IGF-I) becomes uncoupled and there is elevatedGH and diminished IGF-I in the circulation. Blood insulinconcentrations are low as well. Sows are different from dairycows because their somatotropic axis remains coupled duringlactation and both GH and IGF-I are elevated. Nonetheless,sows that become catabolic during lactation will havereduced IGF-I concentrations. Sows are inseminated afterweaning so their metabolic state is different from post-partumbeef and dairy cows that are inseminated when they arelactating. Dairy cows are fed ad libitum and naturally havelow IGF-I during lactation. Sows have elevated IGF-I whenthey are well-fed. A threshold of IGF-I protein in follicularfluid may be met by local ovarian (paracrine ⁄ autocrine) andendocrine sources of IGF-I. Nutritionally induced changes ininsulin and in liver IGF-I secretion that arise from perturbationsof the somatotropic axis have a direct effect on theovary through the endocrine actions of insulin and IGF-I.Sows and cows that are nutritionally compromised have lowconcentrations of insulin and IGF-I in their blood and thistheoretically reduces ovarian responsiveness to gonadotropins.Although sows are inseminated after weaning, thereappear to be carry-over effects of the previous lactation onthe ovarian follicular populations that develop after the sowis weaned. Understanding the mechanisms through whichmetabolic hormones control ovarian function may lead toimproved reproductive management of both pigs and cattlebecause lactation and post-partum reproduction are closelytied in both species.IntroductionGrowth hormone (GH) is a pituitary hormone thatcontrols growth and lactation. Insulin-like growthfactor-I (IGF-I) is released from liver in response toGH and is believed to control growth and lactation aswell (Renaville et al. 2002). There are a series of wellstudiedIGF-binding proteins (IGFBP) that inhibit IGFactivity (Le Roith et al. 2001). In addition to causingIGF-I release, GH antagonizes insulin action (Ethertonand Bauman 1998). Antagonizing the actions of insulinhas a nutrient partitioning effect through which thedevelopment of lean tissue and the production of milkare favoured. A complete understanding of nutrientpartitioning during growth, lactation and reproductiontypically involves an appreciation of GH, IGF-I, IGFBPand insulin because collectively they act on the individualanimal (Chagas et al. 2007).Reproductive events are controlled primarily bygonadotropins (LH and FSH). Growth hormone,IGF-I and insulin can promote gonadotropin actionby potentiating gonadotropin receptor function (Webbet al. 2004; Scaramuzzi et al. 2006). Each hormone canalso act directly through their respective hormonereceptors on the ovary. Metabolic events such aslactation, therefore, are tied to reproductive eventsthrough a hormonal link that involves GH, IGF-I andinsulin. Although not the primary focus of this review,additional metabolic hormones including leptin andadiponectin should be considered when nutrition, reproductionand lactation are discussed.The comparison of cattle and swine is interesting andworthwhile because it may give insights into underlyingbiological mechanisms through which lactation andreproduction are linked. Beef and dairy cattle aredistinctly different in terms of lactation and nutrientpartitioning. The underlying biology of GH, IGF-I andinsulin are different as well. In many respects, swine mayshare more similarities to beef cattle than dairy cattlebecause beef cattle and swine nurse their young.Machine milking systems for dairy cows were neverintended to recapitulate the physiology or lactationcurve of the nursed mother.Cattle and swine are very different with respect to thecoincidence of lactation and post-partum breeding(insemination). Post-partum beef and dairy cattle areinseminated while they are lactating. Post-partum swine,however, are inseminated after weaning. Although bothcattle and swine share common physiology with regardto the local actions of GH, IGF-I, IGFBP and insulinon the ovary, breeding a weaned animal compared witha lactating animal represents a distinct physiologicaldifference, particularly when metabolic hormones areconsidered. This review will examine the somatotropicaxis within cattle and swine during lactation andsubsequent changes in the axis after weaning. Implicationsfor reproduction will be examined with an eyetoward understanding nutrition and reproduction collectively.The Somatotropic Axis in Post-partum CattleDairy cows are well-known for the period of negativeenergy balance that they experience during the first 4–8 weeks post-partum (Drackley 1999). This period ofnegative energy balance is caused by the tremendouscapacity of the mammary gland to sequester nutrientsand synthesize milk. Cows cannot consume enough feedto meet the energy demands for milk production so theyenter into a negative energy balance. Negative energyÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

32 MC Lucybalance is associated with adipose tissue mobilizationand elevated blood non-esterified fatty acid (NEFA)concentrations. During negative energy balance, bloodGH concentrations are increased. The increase in GHdrives nutrient partitioning and supports milk production.Actions of GH during lactationMultiple tissues are affected by GH but coordinatedevents in liver and adipose tissue may be most importantfor lactation (Bell 1995). In liver, the post-partumincrease in GH stimulates an increase in gluconeogenesisthat meets the requirement for mammary lactose synthesis.The increase in gluconeogenesis involves a directeffect of GH on the gluconeogenic pathway as well anindirect effect of GH through an antagonism of insulinaction (Etherton and Bauman 1998). In adipose tissue,GH promotes lipolysis while antagonizing lipogenesisand blocking insulin-dependent glucose uptake.Lactating dairy cows suffer from insulin resistance(insensitivity to insulin that is manifested at the tissuelevel) and this insulin-resistant state conserves glucosefor mammary lactose synthesis (Hayirli 2006). After aninitial post-partum period of elevated GH, there is asecond period where GH circulates at a lower bloodconcentration but remains elevated in high-producingdairy cows (Reist et al. 2003). The steady-state bloodGH concentrations may ultimately determine the relativepropensity for catabolism within the individual cow.Cows in United States herds are supplemented withrecombinant GH (Bauman 1999). The recombinant GHacts in a manner that is consistent with the normaleffects of GH (antagonizing insulin action, promotinggluconeogenesis and blocking lipogenesis). The neteffect on the cow is greater milk production throughthe nutrient partitioning effect of GH.The actions of GH are mediated by the GH receptor(GHR). The highest concentrations of GHR are foundin liver with the second most-abundant location beingadipose tissue (Lucy et al. 1998). Perhaps the mostwidely understood action of GH in liver is the increasein the synthesis and secretion of IGF-I. The IGF-I actsas an endocrine hormone that controls GH secretionthrough a negative feedback loop (Le Roith et al. 2001).Nutritional mechanisms that control GH secretion maymanifest themselves through this negative feedbackloop. For example, one mechanism through whichnutritional restriction for energy or protein elevatesGH is by blocking GHR action and reducing liver IGF-Isynthesis and secretion (Thissen et al. 1999). LesserIGF-I in blood relieves negative feedback and elevatesblood GH during undernutrition. In addition to its clearrole in GH negative feedback, IGF-I arising from theliver may also control the growth and function of cellsand tissues throughout the body via an endocrinemechanism (Jones and Clemmons 1995). The significanceof this endocrine mechanism of IGF-I action hasbeen questioned recently because conditional knock-outmice that do not synthesize IGF-I in liver have normalgrowth (Le Roith et al. 2001).There are three GHR mRNA variants (1A, 1B and1C, respectively) in cattle that differ within the 5¢untranslated region (Jiang and Lucy 2001). The GHRprotein is the same for each variant because the GHRprotein is encoded in exons 2 through 10 of the mRNA.The only tissue that expresses GHR 1A mRNA is theliver of adult animals and the GHR 1A comprises thebulk of liver GHR mRNA (Lucy et al. 1998). The GHR1A mRNA is different from the GHR 1B and 1CmRNA because the GHR 1A mRNA is regulated by avariety of biological signals (Butler et al. 2003; Rhoadset al. 2004; Radcliff et al. 2006). The GHR 1B and 1Care found in many tissues including liver but thesealternative GHR mRNA undergo less biological regulationand appear to be constitutively expressed at a lowlevel.Changes in GHR and IGF-I in post-partum dairy cowsIn dairy cows, liver GHR 1A expression changesdramatically during the period before and immediatelyafter calving (Lucy et al. 2001a; Radcliff et al. 2003a; b).The amount of GHR 1A mRNA decreases approximately2 days before calving, remains low for approximately1 week, and slowly increases during the secondweek after calving. The expression profile for IGF-ImRNA is similar to GHR 1A (decreased after calving)but the decrease in IGF-I post-partum occurs slightlylater than the decline in GHR 1A mRNA. The IGF-Idelay may reflect the dependence of liver IGF-I on GHand GHR 1A. There are post-partum changes in bloodIGFBP concentrations associated with the change inIGF-I (lesser IGFBP-3 and greater IGFBP-2; Skaaret al. 1991). The changes in IGFBP may affect thecapacity of the blood to carry IGF-I and the movementof IGF-I out of the vascular bed (Jones and Clemmons1995; Thissen et al. 1999). Blood insulin concentrationsdecrease as well after calving in dairy cows. The decreasein insulin occurs approximately 2–3 days after thedecrease in GHR 1A mRNA and is coincident withthe decrease in IGF-I. The decrease in post-partumGHR 1A and IGF-I is associated with an increase inblood GH and NEFA. There is a high correlationbetween the GHR 1A mRNA and GH binding inperiparturient liver (Radcliff et al. 2003a). Thus, thedecrease in GHR 1A mRNA coincides with a decreasein GH binding (protein expression) and the loss of GHaction in liver. One possibility is that the decrease inGHR 1A is indirectly causing the changes in GH andNEFA because lower GHR 1A may lead to lower liverIGF-I production and lesser GH negative feedback.Liver gluconeogenesis increases during this period ofdiminished GH action (Drackley et al. 2006). Onequestion that needs to be addressed is what controlsliver gluconeogenesis when the liver is refractory to GH.Changes in GHR in other cattle breeds and implicationsfor nutrient partitioningThe GHR studies described in the previous section weredone in Holstein dairy cows. We have also studiedGuernsey dairy cows and found that the Guernsey cowsunderwent nearly identical changes in GHR 1A whencompared with a contemporary control group of Holsteincows (Okamura et al. 2007). Guernsey cattle thatÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

GH and IGF-I in Cattle and Pigs 33have been selected for milk production independently ofHolstein, therefore, experience similar changes in GHR1A during the post-partum period. Angus beef cattle,however, did not undergo a decrease in GHR 1A atcalving (Jiang et al. 2005).One collective interpretation of these studies is thatGHR 1A expression partially coordinates the high milkproduction phenotype that typifies dairy cattle. Aftercalving there is a decrease in liver GHR 1A mRNA thatleads to low post-partum liver GHR expression. Theloss of liver GHR causes a GH refractory state in liverwhere liver does not produce IGF-I in response to GH.This refractory state represents an uncoupling of GHand IGF-I that is caused by the inactivation of theGHR. The subsequent decrease in blood IGF-I concentrationleads to diminished negative feedback andelevated pituitary production of GH. Elevated postpartumGH supports milk production.Beef cattle may not uncouple their somatotropic axisafter calving. The failure to uncouple this axis leads torelatively higher blood IGF-I concentrations and lowerblood GH concentrations. This endocrine state (higherIGF-I and lesser GH) is relatively anabolic whencompared with the endocrine state of low IGF-I andhigh GH. The uncoupling of the somatotropic axis inpost-partum dairy cows cannot be explained by a simplenutritional phenomena involving negative energy balancebecause feed restriction in later lactation dairycows did not decrease liver GHR 1A expression whenthe question was tested by two different laboratories(Kobayashi et al. 2002; Rhoads et al. 2007).Recoupling of the GH axis after calvingThe recoupling of the GH axis after calving in highproducingdairy cows has been linked to post-partumnutrition and energy balance. Insulin infusion into earlypost-partum dairy cows increased liver GHR 1AmRNA, liver IGF-I mRNA and blood IGF-I concentrations(Butler et al. 2003; Rhoads et al. 2004). Insulin,therefore, controls the recoupling through its positiveeffects on GHR 1A expression. There was also an effectof insulin on GHR in adipose tissue (reduced inresponse to insulin) but this effect was opposite to whatwas observed in liver (Butler et al. 2003). These data foran inhibitory effect of insulin on GHR mRNA need tobe reconciled with the fact that others have shown astimulatory effect of insulin on GHR in adipose tissue(Rhoads et al. 2004). Furthermore, GHR protein inadipose tissue was reduced in low-insulin states such asthose found in early post-partum dairy cows (Rhoadset al. 2004) and underfed dairy cows (Rhoads et al.2007).Greater nutrient intake and improved energy balancein later lactation dairy cows leads to elevated insulin anda direct effect on the GH axis via the effects of insulin onGHR expression (Lucy 2004). The recoupling of the axiscauses an increase in IGF-I that acts negatively on GHsecretion. The increase in insulin and IGF-I and thedecrease in GH switches the post-partum dairy cowfrom a catabolic state (low-insulin, low IGF-I andelevated GH) to an anabolic state (high insulin, highIGF-I and low GH).Little information is available on the recoupling of thesomatotropic axis in post-partum beef cows. The failureof post-partum beef cows to undergo a decrease in GHR1A at calving (Jiang et al. 2005) may imply that the axisis either never uncoupled or minimally uncoupled postpartum.Lake et al. (2006) reported that cows in lowerbody condition had greater GH and lesser IGF-I whencompared with cows in better body condition. Theuncoupling of the somatotropic axis for a beef cow inlow body condition, however, is conceptually differentfrom the homeorhetic mechanism that occurs in postpartumdairy cows. For beef cows, feeding additionalenergy or protein will typically recouple the somatotropicaxis and elevate IGF-I (Lalman et al. 2000). For thebeef cow, either the mammary gland is too small to usethe additional nutrients or the calf is too small toconsume the additional milk. In the latter case, failure tofully evacuate the gland promotes involution andessentially redirects nutrients back toward the mother.Dairy cows are fed ad libitum post-partum. Despite thisgenerous feeding their somatotropic axis is uncoupleduntil the demand of the mammary gland for nutrients isreduced (involution of the gland) or the capacity of thecow to consume nutrients (feed intake) increases. Ananalogous scenario existed when New Zealand andNorth American dairy cows were studied. Feeding 3 kgof concentrate improved body condition in post-partumNew Zealand Holstein–Friesian cows but the sameamount of concentrate feeding had no effect on bodycondition in North American Holstein cows (Rocheet al. 2006). In this comparison, the greater capacity ofthe North American (vs New Zealand) cows to producemilk affected the partitioning of nutrients to adiposetissue.The Somatotropic Axis in Post-partum SowsDifferences in animal management for pigs and cattleaffect the somatotropic axis within the respectivespecies. As mentioned above, lactating dairy cowsare fed ad libitum post-partum and continue to lactatethrough the breeding period. For the dairy cow,lactation is typically terminated when she is in the lasttrimester of gestation. Likewise for beef, calves areweaned in the autumn when cows have a well-establishedpregnancy. In the United States system, dairycows are fed a totally-mixed ration and this ration isbalanced for ad libitum intake. For beef cows and dairycows on pasture, pasture growth and quality maydetermine the overall energetics of the animal (Kolver2003).Obesity in sows creates farrowing difficulties andmetabolic problems, and also antagonizes post-partummilk production. Sows, therefore, are fed a maintenanceration during gestation and their body condition ismanaged carefully (Aherne and Williams 1992). Afterfarrowing and during lactation, ad libitum feeding ispracticed. Obviously, newborn piglets are small andtheir capacity to consume milk is low so sow milkproduction is lowest immediately after farrowing (Nobletand Etienne 1989). Greatest levels of milk productionare achieved near the time of weaning. The lengthof lactation differs between production systems and canÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

34 MC Lucybe 18–21 days (North American systems) or closer to28 days (European systems). After weaning, sows arereturned to a lower level of intake (approximately onehalfto one-third of their lactation ration).The preceding paragraph illustrates the complexity ofthe question when an attempt is made to understand theimpact of nutrition on the somatotropic axis in alactating sow. Feeding and management practices havea major effect on the outcomes of studies of metabolichormones. It is necessary to examine managementdetails carefully when inspecting the scientific literatureon the somatotropic axis in sows.Growth hormone, IGF-I and insulinFew studies have examined intensively the concentrationsof GH, IGF-I, IGFBP and insulin in the sowduring lactation. There is an increase in blood GHconcentrations after farrowing (Schams et al. 1994;Mejia-Guadarrama et al. 2002; Govoni et al. 2007).The increase in blood GH is associated with an increasein blood NEFA concentrations during lactation. Theincrease in NEFA implies that GH is mediating lipidcatabolism during lactation. The increase in GH duringlactation is caused partly by the suckling stimulus fromthe piglets (Rushen et al. 1993). The suckling-inducedGH release explains why a relatively high blood GHconcentration is sustained throughout lactation even inwell-fed sows. With regard to GH, the lactating sow issimilar to the lactating dairy cow because both specieshave elevated GH during lactation. The causes ofelevated GH may be different (lactating dairy cows arenot suckled) but they may also share some commonmechanisms because catabolic states typically elevateGH.One apparent difference between cows and sowsinvolves the recoupling of the somatotropic axis postpartum(Fig. 1). Post-partum sows have elevated IGF-Iafter farrowing (Schams et al. 1994; Govoni et al. 2007)and this endocrine state suggests that the increase in GHafter farrowing can stimulate the liver to synthesize andsecrete IGF-I. In the dairy cow there is a large decreasein IGF-I after calving that is associated with anuncoupling of the somatotropic axis (Radcliff et al.2003a; b). Little is known about changes in IGF-I inbeef cattle around calving but there appears to be atleast some reduction in IGF-I on day 3 post-partumrelative to day 6 post-partum (Lake et al. 2006).There are discrepancies with respect to insulin concentrationsin lactating sows. Some studies reportgreater insulin concentrations in sows after farrowing(Guedes and Nogueira 2001), whereas others reportlower insulin concentrations after farrowing (Revellet al. 1998). The latter case (lower insulin) is similar towhat is reported for cattle after calving and this lowerinsulin appears to be a consequence of metabolic glucoseconsumption during lactation. Sows may become insulin-resistantduring lactation (Quesnel et al. 2007).Sows fed more energy have greater blood IGF-I postpartum(de Braganca and Prunier 1999; van den Brandet al. 2001). The increase in IGF-I post-partum may bepartly explained by the fact that gestating sows are fedmaintenance diets during late pregnancy. In the gestatingsow, therefore, feeding level may limit liver IGF-Iproduction. The combination of elevated GH andad libitum feeding after farrowing drives the somatotropicaxis and increases liver IGF-I production. Theincrease in IGF-I may be a consequence of elevated GHbut also may arise from feed-dependent mechanismssuch as greater liver GHR concentration, an enhancedcapacity for GH to signal through the GHR orGH-independent mechanism through which feedingincreases IGF-I.During lactation, IGF-I concentration may remainhigh (well-fed sows) or may decrease over time (underfedsows) (van den Brand et al. 2001). Presumably, inthe underfed sow, IGF-I concentrations decrease duringlactation because the somatotropic axis becomes uncoupled.The uncoupling may occur in the second and thirdweek of lactation when litter milk consumption and sowmilk production are greater (Noblet and Etienne 1989).The sow enters into negative energy balance because sheFig. 1. Conceptual diagram for changes in blood GH, blood IGF-I and liver GHR after parturition (P; farrowing or calving) in sows (leftdiagram) and dairy cows (right diagram). In sows, the somatotropic axis remains coupled after farrowing and GH concentrations increasebecause of suckling. This leads to elevated GH and IGF-I during lactation. During the latter half of lactation, the somatotropic axis becomesuncoupled perhaps because there is negative energy balance and the amount of GHR in liver is less. Weaning (W) causes a decrease in GH andIGF-I. The blood IGF-I concentration increase after weaning as the sow approaches breeding (B). In dairy cows, the liver GHR decreases aftercalving and this causes uncoupling of the somatotropic axis. Loss of the GHR leads to a decrease in blood IGF-I. The decrease in blood IGF-Imay alleviate GH negative feedback and cause elevated GH in early lactation. Improved energy balance and insulin may increase the GHR andIGF-I and reduce GH concentrations as cows approach the breeding period (B).Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

GH and IGF-I in Cattle and Pigs 35has inadequate feed. In all cases, first parity sows are afocus of metabolic studies because their capacity for feedconsumption is less and they are more likely to becatabolic (van den Brand et al. 2001).Weaning is associated with changes in GH, insulin,glucose and IGF-I but these changes may be confoundedby the lower feed level after weaning andother management considerations. Weaning is stressfulfor sows and the combination of stress and thediscomfort of a fully engorged mammary gland maydepress appetite. Loss of appetite and decreased feedintake reduce blood insulin and IGF-I. Growthhormone concentrations typically decrease after weaning(Mejia-Guadarrama et al. 2002; Govoni et al.2007) because suckling stimulates GH release in pigs(Rushen et al. 1993). If sows are underfed duringlactation (negative energy balance) then weaning mayimprove insulin and IGF-I status (Koketsu et al. 1998;Mejia-Guadarrama et al. 2002). If sows are well-fedduring lactation then weaning may have lesser effectson insulin and IGF-I (Koketsu et al. 1998; Mejia-Guadarrama et al. 2002). Regardless of their nutritionalstatus during lactation, there appears to be anormalization of IGF-I within 3 days after weaning.Sows will typically undergo a short period of lowIGF-I immediate after weaning (associated with loss ofappetite, lower feeding level and stress) and then haveelevated IGF-I as they progress toward oestrus. Theincrease in IGF-I is not necessarily a metabolicresponse because there are stimulatory effects ofoestradiol on liver IGF-I synthesis and secretion(White et al. 2003).The porcine GHRSurprising little is known about the porcine GHR. Ina study of pigs from birth to 1 year of age, the liverGHR mRNA increased with increasing age but muscleGHR mRNA failed to change (Schnoebelen-Combeset al. 1996). The tissue-specific pattern of GHRexpression that was identified in the growing pigsuggested that GHR gene expression was controlledby different mechanisms in liver and muscle; asituation that is known to occur in cattle (see previoussection). Likewise, other investigators have reportedtissue-specific effects of nutrition and GH on theexpression of GHR in liver, muscle and adipose tissuebut a simple theory that explains the physiologicalresponse has not evolved (Dauncey et al. 1994;Brameld et al. 1996; Combes et al. 1997). To ourknowledge there has been only one study of GHRvariants in pigs. In that study, both GHR 1A andGHR 1B mRNA were found in pig liver (Liu et al.2000a). The GHR 1B mRNA but not the GHR 1AmRNA was detected in muscle. The pig, therefore,was like other species because GHR 1B was found ina variety of tissues, whereas GHR 1A was a liverspecificmRNA. Beyond this information, we knownothing about the regulation of specific GHR variantsin pig liver. We do not know if liver GHR expressionchanges during lactation or if negative energy balancein late lactation uncouples the somatotropic axisthrough a GHR-dependent mechanism.Unique aspects of the somatotropic axis during porcinelactationSows are typically catabolic during lactation and loseweight and body condition. Despite the catabolic natureof lactation, the somatotropic axis remains coupled inlactating sows and IGF-I is elevated during lactation.Greater blood IGF-I concentrations, however, do notsuppress blood GH during the sow lactation perhapsbecause nursing piglets stimulate GH secretion andsupersede the IGF-I negative feedback. In the dairycow, the act of machine milking will increase bloodprolactin concentration but, based on limited data, thereappears to be little change in blood GH in response tomachine milking (Negrao and Marnet 2002). The releaseof GH in response to suckling in the beef cow has notbeen studied extensively.Mechanisms that Link the Somatotropic Axis toReproductionSwine and cattle share many aspects of the underlyingbiology controlling ovarian follicular growth. Forexample, in both species follicular development iscontrolled by a combination of LH and FSH. Theinhibition of LH secretion by lactation prevents preovulatoryfollicular development in sows before weaning(Varley and Foxcroft 1990). After weaning, basal LHconcentrations and numbers of LH peaks increase(Shaw and Foxcroft 1985). The increase in LH pulsatilityis believed to drive follicular growth towardovulation. Concentrations of FSH initially increaseafter weaning, but then decline as preovulatory folliclesdevelop (Shaw and Foxcroft 1985). Although LH isclearly important for follicular growth, a high correlationbetween intensity of LH secretion and interval tooestrus has not been found for sows.In cows, follicular development begins shortly aftercalving with a transient increase in FSH and thedevelopment of a dominant follicle. A major componentof the process is the secretion of LH during the earlypost-partum period (Lucy 2003). Low LH pulsatility isassociated with anoestrus. The primary causes ofanoestrus are different for beef and dairy cows but theessential features are similar. In beef cows, sucklinginhibits LH pulsatility and the lack of LH pulsatilityleads to anoestrus (Williams and Griffith 1995). Thissituation is analogous to the sow (suckling inhibits LHsecretion). Anoestrus in dairy cattle is viewed as lesscommon (compared with beef) because the post-partuminhibition of LH may be less in animals that are notsuckled.Systemic interactions of metabolic hormones with thereproductive axisChanges in metabolic hormones like insulin and IGF-Iare dynamic in post-partum cows and sows. Insulin andIGF-I concentrations are initially depressed but graduallyincrease post-partum in cows (Lucy 2003). BloodIGF-I concentrations increase between weaning andoestrus in sows (Mejia-Guadarrama et al. 2002). Thefact that sows are weaned and have elevated IGF-IÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

36 MC Lucybefore breeding represents a major physiological differencebetween cows and sows. Cattle in poor bodycondition or cows failing to increase body conditionduring lactation have low blood IGF-I during thebreeding period (Chagas et al. 2007). Sows that areunderfed during lactation have low blood IGF-I concentrationsbut blood IGF-I normalizes rapidly afterweaning if sows are fed properly. Underfeeding afterweaning in sows will suppress IGF-I, compromisefollicular development (prolonged weaning to oestrusintervals and reduced ovulation rate), and reduceembryonic survival (Zak et al. 1997). Underfeedingafter weaning, however, would be an atypical managementpractice in swine herds. In lactating dairy cows,low IGF-I would be a typical situation because lowIGF-I and high GH are favourable for milk production(Chagas et al. 2007).Under normal physiological conditions, a thresholdof IGF-I protein in follicular fluid may be met by localovarian (paracrine ⁄ autocrine) and endocrine sources ofIGF-I. According to the somatomedin hypothesis,nutritionally induced changes in liver IGF-I secretionhave a direct effect on the ovary through the endocrineactions of IGF-I. Sows and cows that are nutritionallycompromised have low concentrations of insulin andIGF-I in their blood. The lower insulin and IGF-Iconcentrations theoretically reduce ovarian responsivenessto gonadotropins (see below). At the same time,post-partum cows and weaned sows have low blood LHconcentrations, in part because of the effects of metabolichormones on GnRH secretion from the hypothalamus.Thus, the effects of nutrition on reproduction aremanifested at the ovary and at the hypothalamus andpituitary. Overcoming one limitation will not necessarilyrecover ovarian function. Responses to metabolic hormonesand gonadotropins typically follow a plateaumodel where further improvement in reproductivefunction is not seen once a critical threshold of hormoneconcentration is achieved.Local expression of IGFs and their binding proteinsInsulin-like growth factors and their binding proteinsare expressed within specific cell layers [granulosa and(or) theca] of developing follicles. This pararine ⁄ autocrinesynthesis of the IGF has been well-studied andwill not be extensively reviewed here. The reader isreferred to previous reviews on this topic for pigs andcattle (Liu et al. 2000b; Lucy 2000, 2007; Prunier andQuesnel 2000; Lucy et al. 2001b; Webb et al. 2004).There are two types of IGF, IGF-I and IGF-II, butIGF-I is the primary IGF under metabolic control. Theporcine follicle expresses IGF-I (granulosa cell layer)and IGF-II (theca cell layer). Likewise for the cow,theca cells express IGF-II. Expression of IGF-I inbovine granulosa cells has been reported by some butnot all laboratories (Webb et al. 2004). The IGF-I and-II bind to specific receptors (type I and type II IGFreceptors). The type I IGF receptor is closely related tothe insulin receptor. Both insulin and type I IGFreceptors are receptor tyrosine kinases and bothreceptors signal through related intracellular pathways.The type II IGF receptor is an unusual receptorbecause its primary function is to bind, internalize anddegrade IGF-II. In the pig, granulosa and theca cellsexpress insulin receptors as well as type I and type IIIGF receptors. Thus, insulin and IGF-I can act directlyon ovarian cells in pigs. For the cow, there is type IIGF receptor expression in the granulosa cells and typeII IGF receptor expression in granulosa and thecacells. The IGFs (but not insulin) are bound to at leastsix different binding proteins (IGFBP). Each bindingprotein has its own unique physiology. Within thebovine and porcine ovary, IGFBP-2 and -4 areexpressed and are regulators of IGF function (Fortuneet al. 2004).Ovarian follicles contain GHR mRNA and protein(Lucy 2000). In liver, GH acts on the GHR and causesan increase in hepatic IGF-I synthesis and secretion. If asimilar mechanism exists in the ovary of cows and sowsthen GH, acting directly on ovarian follicles, couldaffect ovarian function by simulating the local productionof IGF-I. Growth hormone-dependent, ovarianIGF-I synthesis has been shown in the pig (Hsu andHammond 1987). In cattle, however, GH failed toincrease ovarian IGF-I synthesis (Kirby et al. 1996).The bovine study, however, measured IGF-I mRNA inwhole ovaries. Subtle changes in IGF-I gene expressionthat may have occurred in ovarian follicles may not havebeen detected.Synergism between insulin-like growth factors andgonadotropinsAt the ovarian level, follicular growth in both cows andsows depends on LH and FSH and also insulin andIGF-I. The independent contribution of each hormoneis difficult to establish, however, because there arecoordinated changes in each of these hormones whennutrition is improved. Insulin and IGF-I can act at thelevel of the hypothalamus to stimulate GnRH secretionand therefore control the release of LH and FSH (Lucy2003; Daftary and Gore 2005). Hypothetically, mechanismsthat increase LH pulsatility through their actionson the hypothalamus and pituitary also affect theresponsiveness of the ovary to gonadotropins. Forexample, LH pulsatility increases in cows during thepost-partum period. Both insulin and IGF-I concentrationsin blood increase as well. The collective effects ofLH, insulin and IGF-I acting on the ovary promotefollicular development and ovulation.The location of specific components of the IGFsystem corresponds to the location of LH and FSHreceptors within the developing follicles. The co-localizationof IGF and gonadotropin receptor genes suggestsa coordination of gonadotropin and IGF actionwithin ovary. There is a synergistic relationship betweenthe IGFs and insulin with gonadotropins for a variety ofcellular functions including mitogenesis and steroidogenesis(Lucy 2000, 2007; Webb et al. 2004). Thesynergism is caused by the ability of IGFs and insulinto increase gonadotropin receptor numbers and increasethe activity of gonadotropin receptor second messengersystems. At the same time, gonadotropins increase type IIGF receptor expression and may increase IGF-Isynthesis in granulosa cells.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

GH and IGF-I in Cattle and Pigs 37Both insulin and IGF-I may play a role in folliculardevelopment during periods of under-feeding or weightloss because undernutrition causes a decrease in bloodconcentrations of IGF-I and insulin. According to awidely held hypothesis, blood insulin and IGF-I act inan endocrine manner to affect ovarian cells. Thedecrease in blood IGF-I and insulin concentrationsduring undernutrition decreases the responsiveness ofthe ovary to gonadotropins which ultimately leads to adecrease in follicular growth. Cosgrove et al. (1992)found that re-alimentation of feed-restricted gilts couldincrease follicular growth in the absence of a change inLH secretion. The change in follicular development wasassociated with greater plasma IGF-I and insulin in there-alimented group. Thus, feeding increases the responsivenessof the ovary to LH through its effects on insulinand IGF-I.Carry-over effects of lactation in sowsBeef and dairy cows are inseminated while they arelactating so there may not be appreciable changes inblood insulin and IGF-I before the breeding period.Sows are weaned approximately 1 week before breedingand weaning has a large effect on the metabolism of thesow and blood concentrations of insulin and IGF-I(discussed in previous section). The blood IGF-I concentrationsare normalized within approximately 3 daysafter weaning. Nonetheless, there appear to be carryovereffects of lactation. Thus, a compromised state ofovarian follicular development that develops duringlactation can potentially influence follicular growth andthe time of oestrus after weaning. Evidence for thisevolved from a variety of studies including those offollicular development before and after weaning in sows(Lucy et al. 2001b). Sows with the shortest intervals toovulation had larger follicles shortly after weaning. Inthese same sows (those with the shortest interval toovulation), there was also greater estrogenic activity offollicles before weaning and an earlier rise in preovulatoryoestradiol after weaning. Based on these data, itappears that some component of the follicular developmentafter weaning is controlled by physiologicalprocesses during lactation that affect the ovarian follicularpopulations. These physiological processes probablyinvolve gonadotropins as well as key metabolichormones like GH, insulin and IGF-I that affectgonadotropin action at the level of the ovary.Summary and ConclusionsGrowth hormone, IGF-I and insulin are metabolichormones that control growth and lactation in cattleand swine. Cows and sows typically enter into negativeenergy balance during lactation and this negative energybalance is associated with low concentrations of insulinand IGF-I in the blood. Insulin and IGF-I affect thereproductive axis through direct effects on ovarian cellsand also through their effects on gonadotropin secretionand gonadotropin receptor function. Understanding themechanisms through which metabolic hormones controlovarian function may lead to improved reproductivemanagement of both cattle and pigs because lactationand post-partum reproduction are closely tied in bothspecies.AcknowledgementsThe author would like to thank Dr Tim Safranski and AmandaWilliams of the University of Missouri for their helpful suggestionsduring the preparation of this manuscript.ReferencesAherne FX, Williams IH, 1992: Nutrition for optimizingbreeding herd performance. Vet Clin N Am Food AnimPract 8, 589–608.Bauman DE, 1999: Bovine somatotropin and lactation: frombasic science to commercial application. Domest AnimEndocrinol 17, 101–116.Bell AW, 1995: Regulation of organic nutrient metabolismduring transition from late pregnancy to early lactation. JAnim Sci 73, 2804–2819.de Braganca MM, Prunier A, 1999: Effects of low feed intakeand hot environment on plasma profiles of glucose, nonesterifiedfatty acids, insulin, glucagon, and IGF-I in lactatingsows. Domest Anim Endocrinol 16, 89–101.Brameld JM, Atkinson JL, Saunders JC, Pell JM, Buttery PJ,Gilmour RS, 1996: Effects of growth hormone administrationand dietary protein intake on insulin-like growth factorI and growth hormone receptor mRNA expression inporcine liver, skeletal muscle, and adipose tissue. J AnimSci 74, 1832–1841.van den Brand H, Prunier A, Soede NM, Kemp B, 2001: Inprimiparous sows, plasma insulin-like growth factor-I canbe affected by lactational feed intake and dietary energysource and is associated with luteinizing hormone. ReprodNutr Dev 41, 27–39.Butler ST, Marr AL, Pelton SH, Radcliff RP, Lucy MC, ButlerWR, 2003: Insulin restores GH responsiveness duringlactation-induced negative energy balance in dairy cattle:effects on expression of IGF-I and GH receptor 1A.J Endocrinol 176, 205–217.Chagas LM, Bass JJ, Blache D, Burke CR, Kay JK, LindsayDR, Lucy MC, Martin GB, Meier S, Rhodes FM, RocheJR, Thatcher WW, Webb R, 2007: Invited review: newperspectives on the roles of nutrition and metabolic prioritiesin the subfertility of high-producing dairy cows. J DairySci 90, 4022–4032.Combes S, Louveau I, Bonneau M, 1997: Moderate foodrestriction affects skeletal muscle and liver growth hormonereceptors differently in pigs. J Nutr 127, 1944–1949.Cosgrove JR, Tilton JE, Hunter MG, Foxcroft GR, 1992:Gonadotropin-independent mechanisms participate in ovarianresponses to realimentation in feed-restricted prepubertalgilts. Biol Reprod 47, 736–745.Daftary SS, Gore AC, 2005: IGF-1 in the brain as a regulatorof reproductive neuroendocrine function. Exp Biol Med(Maywood) 230, 292–306.Dauncey MJ, Burton KA, White P, Harrison AP, Gilmour RS,Duchamp C, Cattaneo D, 1994: Nutritional regulation ofgrowth hormone receptor gene expression. FASEB J 8, 81–88.Drackley JK, 1999: ADSA Foundation Scholar Award.Biology of dairy cows during the transition period: the finalfrontier?. J Dairy Sci 82, 2259–2273.Drackley JK, Donkin SS, Reynolds CK, 2006: Major advancesin fundamental dairy cattle nutrition. J Dairy Sci 89, 1324–1336.Etherton TD, Bauman DE, 1998: Biology of somatotropin ingrowth and lactation of domestic animals. Physiol Rev 78,745–761.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

38 MC LucyFortune JE, Rivera GM, Yang MY, 2004: Follicular development:the role of the follicular microenvironment inselection of the dominant follicle. Anim Reprod Sci 82–83,109–126.Govoni N, Parmeggiani A, Galeati G, Penazzi P, De Iasio R,Pagotto U, Pasquali R, Tamanini C, Seren E, 2007: Acylghrelin and metabolic hormones in pregnant and lactatingsows. Reprod Domest Anim 42, 39–43.Guedes RM, Nogueira RH, 2001: The influence of parity orderand body condition and serum hormones on weaning-toestrusinterval of sows. Anim Reprod Sci 67, 91–99.Hayirli A, 2006: The role of exogenous insulin in the complexof hepatic lipidosis and ketosis associated with insulinresistance phenomenon in postpartum dairy cattle. Vet ResCommun 30, 749–774.Hsu CJ, Hammond JM, 1987: Concomitant effects of growthhormone on secretion of insulin-like growth factor I andprogesterone by cultured porcine granulosa cells. Endocrinology121, 1343–1348.Jiang H, Lucy MC, 2001: Variants of the 5¢-untranslatedregion of the bovine growth hormone receptor mRNA:isolation, expression and effects on translational efficiency.Gene 265, 45–53.Jiang H, Lucy MC, Crooker BA, Beal WE, 2005: Expressionof growth hormone receptor 1A mRNA is decreased indairy cows but not in beef cows at parturition. J Dairy Sci88, 1370–1377.Jones JI, Clemmons DR, 1995: Insulin-like growth factors andtheir binding proteins: biological actions. Endocr Rev 16, 3–34.Kirby CJ, Thatcher WW, Collier RJ, Simmen FA, Lucy MC,1996: Effects of growth hormone and pregnancy on expressionof growth hormone receptor, insulin-like growth factor-I, and insulin-like growth factor binding protein-2 and -3genes in bovine uterus, ovary, and oviduct. Biol Reprod 55,996–1002.Kobayashi Y, Boyd CK, McCormack BL, Lucy MC, 2002:Reduced insulin-like growth factor-I after acute feedrestriction in lactating dairy cows is independent of changesin growth hormone receptor 1A mRNA. J Dairy Sci 85,748–754.Koketsu Y, Dial GD, Pettigrew JE, Xue J, Yang H, Lucia T,1998: Influence of lactation length and feed intake onreproductive performance and blood concentrations ofglucose, insulin and luteinizing hormone in primiparoussows. Anim Reprod Sci 52, 153–163.Kolver ES, 2003: Nutritional limitations to increasedproduction on pasture-based systems. Proc Nutr Soc 62,291–300.Lake SL, Scholljegerdes EJ, Hallford DM, Moss GE, RuleDC, Hess BW, 2006: Effects of body condition score atparturition and postpartum supplemental fat on metaboliteand hormone concentrations of beef cows and their sucklingcalves. J Anim Sci 84, 1038–1047.Lalman DL, Williams JE, Hess BW, Thomas MG, KeislerDH, 2000: Effect of dietary energy on milk production andmetabolic hormones in thin, primiparous beef heifers.J Anim Sci 78, 530–538.Le Roith D, Bondy C, Yakar S, Liu JL, Butler A, 2001: Thesomatomedin hypothesis: 2001. Endocr Rev 22, 53–74.Liu J, Carroll JA, Matteri RL, Lucy MC, 2000a: Expression oftwo variants of growth hormone receptor messenger ribonucleicacid in porcine liver. J Anim Sci 78, 306–317.Liu J, Koenigsfeld AT, Cantley TC, Boyd CK, Kobayashi Y,Lucy MC, 2000b: Growth and the initiation of steroidogenesisin porcine follicles are associated with uniquepatterns of gene expression for individual components ofthe ovarian insulin-like growth factor system. Biol Reprod63, 942–952.Lucy MC, 2000: Regulation of ovarian follicular growth bysomatotropin and insulin-like growth factors in cattle. JDairy Sci 83, 1635–1647.Lucy MC, 2003: Mechanisms linking nutrition and reproductionin postpartum cows. Reprod Suppl 61, 415–427.Lucy MC, 2004: Mechanisms linking the somatotropic axiswith insulin: Lessons from the postpartum dairy cow. ProcN Z Soc Anim Prod 64, 19–23.Lucy MC, 2007: The bovine dominant ovarian follicle. J AnimSci 85, E89–E99.Lucy MC, Boyd CK, Koenigsfeld AT, Okamura CS, 1998:Expression of somatotropin receptor messenger ribonucleicacid in bovine tissues. J Dairy Sci 81, 1889–1895.Lucy MC, Jiang H, Kobayashi Y, 2001a: Changes in thesomatotropin axis associated with the initiation of lactation.J Dairy Sci 84, E113–E119.Lucy MC, Liu J, Boyd CK, Bracken CJ, 2001b: Ovarianfollicular growth in sows. Reprod Suppl 58, 31–45.Mejia-Guadarrama CA, Pasquier A, Dourmad JY, Prunier A,Quesnel H, 2002: Protein (lysine) restriction in primiparouslactating sows: effects on metabolic state, somatotropic axis,and reproductive performance after weaning. J Anim Sci 80,3286–3300.Negrao JA, Marnet PG, 2002: Effect of calf suckling onoxytocin, prolactin, growth hormone and milk yield incrossbred Gir · Holstein cows during milking. Reprod NutrDev 42, 373–380.Noblet J, Etienne M, 1989: Estimation of sow milk nutrientoutput. J Anim Sci 67, 3352–3359.Okamura CA, Bader JF, Cantley TC, Lucy MC, 2007: Growthhormone receptor expression in two dairy breeds during theperiparturient period. J Dairy Sci 90, 8–4 (abstract).Prunier A, Quesnel H, 2000: Influence of the nutritional statuson ovarian development in female pigs. Anim Reprod Sci60–61, 185–197.Quesnel H, Etienne M, Pere MC, 2007: Influence of litter sizeon metabolic status and reproductive axis in primiparoussows. J Anim Sci 85, 118–128.Radcliff RP, McCormack BL, Crooker BA, Lucy MC, 2003a:Growth hormone (GH) binding and expression of GHreceptor 1A mRNA in hepatic tissue of periparturient dairycows. J Dairy Sci 86, 3933–3940.Radcliff RP, McCormack BL, Crooker BA, Lucy MC, 2003b:Plasma hormones and expression of growth hormonereceptor and insulin-like growth factor-I mRNA in hepatictissue of periparturient dairy cows. J Dairy Sci 86, 3920–3926.Radcliff RP, McCormack BL, Keisler DH, Crooker BA, LucyMC, 2006: Partial feed restriction decreases growth hormonereceptor 1A mRNA expression in postpartum dairycows. J Dairy Sci 89, 611–619.Reist M, Erdin D, von Euw D, Tschuemperlin K, LeuenbergerH, Delavaud C, Chilliard Y, Hammon HM, Kuenzi N,Blum JW, 2003: Concentrate feeding strategy in lactatingdairy cows: metabolic and endocrine changes with emphasison leptin. J Dairy Sci 86, 1690–1706.Renaville R, Hammadi M, Portetelle D, 2002: Role of thesomatotropic axis in the mammalian metabolism. DomestAnim Endocrinol 23, 351–360.Revell DK, Williams IH, Mullan BP, Ranford JL, Smits RJ,1998: Body composition at farrowing and nutrition duringlactation affect the performance of primiparous sows. I:Voluntary feed intake, weight loss, and plasma metabolites.J Anim Sci 76, 1729–1737.Rhoads RP, Kim JW, Leury BJ, Baumgard LH, Segoale N,Frank SJ, Bauman DE, Boisclair YR, 2004: Insulinincreases the abundance of the growth hormone receptorin liver and adipose tissue of periparturient dairy cows. JNutr 134, 1020–1027.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

GH and IGF-I in Cattle and Pigs 39Rhoads RP, Kim JW, Van Amburgh ME, Ehrhardt RA,Frank SJ, Boisclair YR, 2007: Effect of nutrition on the GHresponsiveness of liver and adipose tissue in dairy cows. JEndocrinol 195, 49–58.Roche JR, Berry DP, Kolver ES, 2006: Holstein–Friesianstrain and feed effects on milk production, body weight, andbody condition score profiles in grazing dairy cows. J DairySci 89, 3532–3543.Rushen J, Foxcroft G, De Passille AM, 1993: Nursing-inducedchanges in pain sensitivity, prolactin, and somatotropin inthe pig. Physiol Behav 53, 265–270.Scaramuzzi RJ, Campbell BK, Downing JA, Kendall NR,Khalid M, Munoz-Gutierrez M, Somchit A, 2006: A reviewof the effects of supplementary nutrition in the ewe on theconcentrations of reproductive and metabolic hormones andthe mechanisms that regulate folliculogenesis and ovulationrate. Reprod Nutr Dev 46, 339–354.Schams D, Kraetzl WD, Brem G, Graf F, 1994: Secretorypattern of metabolic hormones in the lactating sow. ExpClin Endocrinol 102, 439–447.Schnoebelen-Combes S, Louveau I, Postel-Vinay MC, BonneauM, 1996: Ontogeny of GH receptor and GH-bindingprotein in the pig. J Endocrinol 148, 249–255.Shaw HJ, Foxcroft GR, 1985: Relationships between LH,FSH and prolactin secretion and reproductive activity in theweaned sow. J Reprod Fertil 75, 17–28.Skaar TC, Vega JR, Pyke SN, Baumrucker CR, 1991: Changesin insulin-like growth factor-binding proteins in bovinemammary secretions associated with pregnancy and parturition.J Endocrinol 131, 127–133.Thissen JP, Underwood LE, Ketelslegers JM, 1999: Regulationof insulin-like growth factor-I in starvation and injury.Nutr Rev 57, 167–176.Varley MA, Foxcroft GR, 1990: Endocrinology of thelactating and weaned sow. J Reprod Fertil Suppl 40, 47–61.Webb R, Garnsworthy PC, Gong JG, Armstrong DG, 2004:Control of follicular growth: local interactions and nutritionalinfluences. J Anim Sci 82, E63–E74.White ME, Johnson BJ, Hathaway MR, Dayton WR, 2003:Growth factor messenger RNA levels in muscle and liver ofsteroid-implanted and nonimplanted steers. J Anim Sci 81,965–972.Williams GL, Griffith MK, 1995: Sensory and behaviouralcontrol of gonadotrophin secretion during suckling-mediatedanovulation in cows. J Reprod Fertil Suppl 49, 463–475.Zak LJ, Cosgrove JR, Aherne FX, Foxcroft GR, 1997: Patternof feed intake and associated metabolic and endocrinechanges differentially affect post-weaning fertility in primiparouslactating sows. J Anim Sci 75, 208–216.Author’s address (for correspondence): MC Lucy, Division of AnimalSciences, University of Missouri, Columbia, MO 65211, USA. E-mail:lucym@missouri.eduConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Seasonality of Reproduction in Mammals 4140302010Harvest (10 6 liters)PricePrice (Euros/liter)Harvest0.610.530.460.38% females showing at least one ovulation/oestrus per month1008060402000246OvulationsOestrus81012141618202224262830320MonthsJ F M A M J J A S O N D0.30Oct. Feb. Oct. Feb. Oct. Feb.MonthsFig. 1. Seasonal variations in the harvest and farm price of goat milkin France (year 1992; adapted from Chemineau et al. 1996)post-partum ovulatory activity in beef cattle characterizedby a limited period of time for ovulations (Ingrandet al. 2003; Agabriel et al. 2004).In rams and he-goats, although spermatogenic activityand sexual behaviour do not stop, they deeply varywith season. In Soay rams, a primitive breed from NorthScotland, testicular size (which reflects spermatogenicactivity), plasma FSH and testosterone concentrationsas well as sexual ‘flush’ and aggressive behaviour reachtheir maximum between August and November, the‘rut’ season in this breed (Lincoln 1979). Ile-de-Francerams show testicular weight and sperm production pertestis (directly measured at its output) varying from lessthan 200 g and 1 billion per day in March, to more than300 g and 5 billion per day in September, respectively(Ortavant et al. 1985). Alpine bucks also display dramaticvariations in sexual behaviour (0–1.5 matings in10 min), sperm individual motility (2.5–3.5 over 5) andfertilizing ability (20%–70% of kiddings after AI)between the spring-summer and autumn-winter periods(Delgadillo et al. 1992). In this last breed, changes inejaculate volume and sperm concentration occur seasonally,which, in this species showing a deleteriouseffect of seminal plasma on in vitro sperm survival, hasimportant implications on semen technology. Finally,Fig. 3. Seasonal variations of the occurrence of ovulations andoestrous behaviour in Alpine goats (adapted from Chemineau et al.1992)stallions also show seasonal variations in sexual behaviourand sperm quality, the lowest season being inwinter and the highest in spring-summer (Magistriniet al. 1987).Within a given species, the various breeds may expressvariable degrees of seasonality. For example, Soay andTexel ewes are highly seasonal, while Merino andManchega ewes present more discrete expression ofseasonality (Hafez 1952; Santiago-Moreno et al. 2000).Breeds raised in the Subtropics and in the Tropicsgenerally present a low seasonality or cycle all the yearround with no anovulatory period (Gonzalez-Stagnaro1983; Khaldi 1984; Yenikoye 1984; Chemineau 1986;Mahieu et al. 1989; Arroyo et al. 2007). This is aninteresting trait of these breeds for local farmers whocan then organize the breeding season of their flock allthe year round, without expensive hormonal treatments.Unfortunately, a marked seasonality is expressed inthese breeds when subjected to the large photoperiodicvariations and temperate climates of northern countries(Chemineau et al. 2004), preventing the possible use ofthese breeds in flocks under temperate latitude. However,in temperate breeds maintained under environmentalconditions similar from which they originate,intra-breed variability also exists. Some reproductiveFig. 2. Timing of the annualreproductive cycle of some seasonalbreeders. Breeding season isindicated as a grey box and deliveringseason as a black box(adapted from Ortavant et al.1985)HorseFeral horseFeral cattleGoatTexel sheepSoay sheepMouflonCalifornian mountain sheepAlaska mountain sheepWild pigMinkSpringSummerAutumn WinterBreeding seasonBirth seasonJ F M A M J J A S O N D J F M A M J J A S O N DMonthsÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Seasonality of Reproduction in Mammals 43reproduction (Larkin et al. 2002; Lehman et al. 2002;Paul et al. 2008) or pelage moult (Paul et al. 2008). Thisendogenous rhythm is timed, in ewes, by discrete signalsgiven by external changes in photoperiod (Barrell et al.2000). However, the specific physiological mechanismsunderlying this circannual system remain largely unknown.The specific role of short days (SD) and long days(LD) days in timing reproductive activity varies amongspecies. In long-day breeders (i.e. animals which arefully sexually active during LD) like horses, LD arestimulatory and SD inhibitory of sexual activity. Incontrast, LD inhibit while SD stimulate sexual activityin short-day breeders such as sheep and goats. Howeverin mammals subjected to a constant photoperiod, thedaylength-specific effect is not permanent. Animals‘escape’ and become ’refractory’ to the prevailing photoperiod:LD are no longer stimulatory in mares orinhibitory in sheep and goats, SD are no longerinhibitory in mares or stimulatory in sheep and goats.This refractoriness could conceptually be considered asmerely the first step of the expression of the circannualendogenous rhythm. It can be ‘broken’ by transferringanimals into the opposite photoperiod: refractoriness toSD, which occurs naturally in sheep in late winter, isbroken by 2 months of exposure to LD in December–January, allowing the efficiency of stimulatory SD to berestored. Thus, by subjecting animals to oppositephotoperiods, it is possible to control seasonality ofreproduction. This property is now commonly used inphotoperiodic treatments applied on farms and ⁄ or inAI centres (see below). The definition of what is reallyLD and SD is not straightforward: it is possible todefine a threshold of photosensitivity based on thenumber of light hours per day, under which LD arestimulatory and below which SD are inhibitory (seereviews Chemineau et al. 1996 and Malpaux et al.1996). The ‘photoperiodic history’ of each individualshould also be taken into account. Thus, it is nowcommonly accepted that LD are days longer than thepreceding ones, and that SD are days shorter than thepreceding ones. This property is interesting under fieldconditions: following a period of artificial long days,animals perceive SD even though natural daylength islonger than 12 h of light per day. Another interestingproperty can be used under farm conditions to applylong days: the illumination of a specific phase of thenight, the so-called ‘photosensitive phase’, generallysituated 14–16 h after dawn in sheep and 9.5 h afterdusk in mares, allows animals to perceive LD eventhough real LD are not applied (see reviews of Malpauxet al. 1996; Chemineau et al. 1996 in sheep and goats,and Guillaume 1996 in mares).More generally, photoperiod, which entrains theendogenous circannual rhythms of reproduction, exertsits action through two different but complementary anddependent pathways by (1) adjusting the phases ofgonadal development with external natural conditionsand (2) by synchronizing the reproductive periodbetween individuals of the same species. In mammals,all the photoperiodic input is perceived exclusivelythrough the eyes then transmitted via a multi-synapticpathway to the pineal gland, which transduces thephotic signal into a chemical one by synthesizing andsecreting melatonin. Synthesized during the night by thepineal gland, it is thought that melatonin is delivered tothe brain via the cerebrospinal fluid, and to peripheraltissues by general circulation. To control reproductiveactivity in sheep, melatonin acts on the pre-mammillaryhypothalamus where specific receptors transcripts areexpressed (Migaud et al. 2005) and the membranereceptor is present, which stimulates, approximately45 days after the onset of daily impregnation (Viguie´et al. 1995), the pulsatile activity of LHRH-LH which inturn will drive gonadal and behavioural sexual activities(review by Malpaux 2006; Malpaux et al. 2001). Externalmelatonin can be given to ‘mimic’ short days and hasbeen of practical use in sheep and goats to stimulatereproduction in spring. Details of the neuroendocrinemechanisms responsible for the seasonal and lightcontrol of LHRH pulsatility could not be discussedextensively here (see Malpaux 2006 for an extensivereview), especially what differs between short-day breeders(sheep and goats) and long-day breeders (horses andhamsters) to explain their opposite change in LHRHneuron pulsatility to the same melatonin signal. However,it is likely that the difference lies in the neuralnetwork connecting hypothalamic melatonin receptorbearing cells and LHRH neurones, a network that isalso responsible for the long delay of action of melatoninon LHRH pulsatility (Viguié et al. 1995). Forexample, dopaminergic activity inhibits the LHRHpulsatility at the end of the neural network during longdays in sheep (Thiery et al. 1989). Interestingly, dopaminefrom the median eminence is also correlated withthe inhibition of gonadotropin release, which takes placeunder short days in these species (Steger et al. 1985;Glass et al. 1988).Although molecular mechanisms underlying the centralcontrol of gonadotrope system by melatonin stillconstitute a ‘black box’, numerous experiments usingonly photoperiodic variations have lead to proposals tofarmers and artificial insemination centres of specificlight devices, which may be able to control seasonalityof reproduction of their animals.Using Artificial Photoperiodic Treatments toControl Seasonality of Reproduction in FarmAnimalsPhotoperiodic treatments have been of practical interestfor controlling seasonal reproduction essentially insheep, goats and horses. Sheep and goat AI centres,equipped with dark housing, use alternate light regimeswith 1 month LD and 1 month SD which allowpermanent high semen production in rams and bucks,with no seasonal variations in sperm quality. Currently,in the French national genetic improvement scheme, allbucks (approximately 70 per year) are permanentlytreated with rapid alternation LD–SD, which increasesthe AI dose production (+40%) per buck and per year(Delgadillo et al. 1993) and reduces duration of thebreeding period of males (slaughtered after 18 monthsof production, or approximately 18 months earlier thatthose maintained under natural photoperiod). In otherAI centers which do not require all the year-roundÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

44 P Chemineau, D Guillaume, M Migaud, JC Thie´ry, MT Pellicer-Rubio and B Malpauxproduction of semen doses, rams raised in open barnsare subjected to 2–3 months extra ‘LD’ (from Decemberto February) followed either by return to naturalphotoperiod, or by subcutaneous treatment with melatoninimplants which mimics SD (‘SD’) (Fig. 5). Thissuccession ‘LD’–‘SD’ stimulates semen output in quantityand quality in spring, thus mimicking the normalsexual season which itself normally lasts 2–3 months(Malpaux et al. 1995).In goat farms, (always in open barns), males andfemales are subjected to the second type of treatmentperformed in AI centres (succession ‘LD’–‘SD’). Thistreatment should be associated with a ‘buck effect’(introduction of treated bucks for 45 days, after 35–75‘SD’) in order to induce ovulations and oestrousbehaviour and to get high fertility rate. Under suchconditions, out-of-season fertility and prolificacy can bekept at high levels (>75% kidding rate with approximatelytwo kids per kidding) (Chemineau et al. 1996).This type of treatment could be satisfactorily associatedwith AI on fixed-time basis with fertility results equivalentto ‘classical’ hormonal treatment with FGAsponge and eCG (Pellicer-Rubio et al. 2007, 2008).For local breeds in subtropical conditions where seasonalityis less marked than those raised under temperatelatitudes, the treatment of females is not necessary.When the LD treatment is applied only in bucks usedfor the ‘buck-effect’, the percentages of ovulatingfemales and their fertility after natural mating are high(Delgadillo et al. 2002, 2004).In ewes, a majority of out-of-season lambings are stillobtained using ‘classical’ hormonal treatments (FGAsponges + eCG), but the frequency of utilization ofmelatonin implants to get out-of-season breeding isincreasing, especially in Mediterranean countries. Theuse of implants also increases fecundity: + 0.20lamb ⁄ ewe treated ⁄ year has been almost always observed,mainly due to increased proportions of twinsrather than triplets (Chemineau et al. 1996).Photoperiodic treatments are also applied in mares toadvance the annual breeding season and to provide foalswith a decisive age-related advance when competingwith their contemporaries born the same year (seeabove). This is generally performed by exposing maresto LD or pseudo LD during autumn and winter. Suchtreatments allow mares to be fertilized approximately2–3 months earlier than females kept under a naturalphotoperiod (Guillaume 1996).Thus, photoperiodic treatments are now used in bothsexes of nearly all farm species to control seasonalreproduction. Whatever the species, they use commonproperties of alternations between inhibitory and stimulatoryphotoperiods, where durations are adapted tothe species and sex. When using pure light treatments(without melatonin), especially when applied in openbarns, they could be considered as non-invasive whichfully respects animal welfare considerations. It is veryprobable that these photoperiodic treatments will beused more extensively in the future as livestock productionsystems strive to be more sustainable.(a)Light-proof buildings:17h 16DuskDawn16LProgressivedecrease8L0 60JanuaryMarch150 daysJune(b)17h 16(c)(d)DuskDawn0 60JanuaryMarch17h 16DuskDawn0 60JanuaryMarchOpen barns:2420"Long days"2nd period of suppl. lightShort daysMelatonin"Short days"Natural dusk150 daysJune150 daysJunehours16128401st period of suppl. lightJanuary February March April May June"Long days"Natural lightorMelatonin"Short days"Natural dawnFig. 5. Photoperiodic treatmentsto control sexual activity in smallruminants raised in closed (a–c) oropen (d) barns (adapted fromChemineau et al. 1996)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Seasonality of Reproduction in Mammals 45ConclusionThe seasonality of animal products release is more theconsequence of interactions between the naturalendogenous rhythmicity of animals and environmentalconstraints than a deliberate choice of the farmer toproduce at a definite season. In most seasonal speciesand breeds involved in these productions, specificphotoperiodic treatments derived, at least in part,from the cumulative knowledge of physiological mechanismsinvolved in the control of the reproductivefunction, have progressively been proposed to overcomethe problem raised by the seasonal availability ofreproduction-derived products. In mammals, suchtreatments should be taken as an interesting alternativeto classical hormonal treatments in a general contextof reduction of hormonal utilization in more sustainableanimal production systems. The existence ofstrong genetic bases for seasonality of reproductiveactivity in the main farm animal species should befurther explored to propose selection criteria and ⁄ orgene markers accessible to primary breeder and producerorganizations willing to reduce seasonality intheir flocks.ReferencesAdams VL, Goodman RL, Salm AK, Coolen LM, Karsch FJ,Lehman MN, 2006: Morphological plasticity in the neuralcircuitry responsible for seasonal breeding in the ewe.Endocrinology 147, 4843–4851.Agabriel J, Blanc F, Egal D, Dhour P, 2004: Influencescombinées de la saison de mise bas et de l’exposition autaureau sur la venue en cyclicite´ de vaches Charolaises.Rencontres Recherches Ruminants Paris 8-9 de´c 11, 398.Al-Shorepy SR, Notter DR, 1997: Response to selection forfertility in a fall-lambing sheep flock. J Anim Sci 75, 2033–2040.Anderson GM, Connors JM, Hardy SL, Valent M, GoodmanRL, 2001: Oestradiol microimplants in the ventromedialpreoptic area inhibit secretion of luteinizing hormone viadopamine neurones in anoestrous ewes. J Neuroendocrinol13, 1051–1058.Arroyo LJ, Gallegos-Sanchez J, Villa-Godoy A, BerruecosJM, Perera G, Valencia J, 2007: Reproductive activity ofPelibuey and Suffolk ewes at 19° north latitude. AnimReprod Sci 102, 24–30.Barrell GK, Thrun LA, Brown ME, Viguie C, Karsch FJ,2000: Importance of photoperiodic signal quality to entrainmentof the circannual reproductive rhythm of the ewe. BiolReprod 63, 769–774.Billings HJ, Viguie C, Karsch FJ, Goodman RL, Connors JM,Anderson GM, 2002: Temporal requirements of thyroidhormones for seasonal changes in LH secretion. Endocrinology143, 2618–2625.Bronson FH, 1989: Mammalian Reproductive Biology. TheUniversity of Chicago Press Ltd, Chicago and London.Chemineau P, 1986: Sexual behaviour and gonadal activityduring the year in the tropical Creole meat goat. I. Femaleoestrous behaviour and ovarian activity. Reprod Nutr Dev26, 441–452.Chemineau P, Daveau A, Maurice F, Delgadillo JA, 1992:Seasonality of oestrus and ovulation is not deeply modifiedby submitting Alpine goats to a tropical photoperiod. SmallRuminant Res 8, 299–312.Chemineau P, Malpaux B, Pelletier J, Leboeuf B, DelgadilloJA, Deletang F, Pobel T, Brice G, 1996: Emploi desimplants de me´latonine et des traitements photope´riodiquespour maîtriser la reproduction saisonnie` re chez les ovins etles caprins. Prod Anim 9, 45–60.Chemineau P, Daveau A, Pelletier J, Malpaux B, Karsch FJ,Viguie C, 2003: Changes in the 5-HT 2A receptor system inthe pre-mammillary hypothalamus of the ewe are related toregulation of LH pulsatile secretion by an endogenouscircannual rhythm. BioMedCentral Neurosci 4, 1 (http://www.biomedcentral.com/1471-2202/4/1).Chemineau P, Daveau A, Cognie Y, Aumont G, Chesneau D,2004: Seasonal ovulatory activity exists in tropical Creolefemale goats and Black Belly ewes subjected to a temperatephotoperiod. BioMedCentral Physiol 4, 12 (http://www.biomedcentral.com/1472-6793/4/12).Delgadillo JA, Leboeuf B, Chemineau P, 1992: Abolition ofseasonal variations in semen quality and maintenance ofsperm fertilizing ability by short photoperiodic cycles inhe-goats. Small Ruminant Res 9, 47–59.Delgadillo JA, Leboeuf B, Chemineau P, 1993: Maintenanceof sperm production in bucks during a third year of shortphotoperiodic cycles. Reprod Nutr Dev 33, 609–617.Delgadillo JA, Flores JA, Veliz FG, Hernandez HF, Duarte G,Vielma J, Poindron P, Chemineau P, Malpaux B, 2002:Induction of sexual activity in lactating anovulatory femalegoats using male goats treated only with artificially longdays. J Anim Sci 80, 2780–2786.Delgadillo JA, Fitz-Rodriguez G, Duarte G, Veliz FG,Carrillo E, Flores JA, Vielma J, Hernandez H, MalpauxB, 2004: Management of photoperiod to control caprinereproduction in the subtropics. Reprod Fertil Dev 16, 471–478.Gallegos-Sa´nchez J, Delaleu B, Caraty A, Malpaux B,Thie´ry JC, 1997: Estradiol acts locally within the retrochiasmaticarea to inhibit pulsatile luteinizing-hormonerelease in the female sheep during anestrus. Biol Reprod56, 1544–1549.Glass JD, Ferrera S, Deaver DR, 1988: Photoperiodic adjustmentin hypothalamic amines gonadotrophin-relesaing hormoneand b-endorphin in white-footed mouse.Endocrinology 123, 1119–1127.Gonzalez-Stagnaro C, 1983: Comportamiento reproductivo delas razas locales de rumiantes en el tropico americano. In:Chemineau P, Gauthier D, Thimonier J (eds), Reproductiondes Ruminants en Zone Tropicale, Vol. 20. Inra Publications,Versailles, France, pp. 1–84.Guillaume D, 1996: Action de la photope´riode sur la reproductiondes équidés. Prod Anim 9, 61–69.Hafez ESE, 1952: Studies on the breeding season andreproduction of the ewe. J Agric Sci 42, 189–265.Hanocq E, Bodin L, Thimonier J, Teyssier J, Malpaux B,Chemineau P, 1999: Genetics parameters of spontaneousspring ovulatory activity in Merinos d’Arles sheep. GenetSel Evol 31, 77–90.Hanrahan JP, 1987: Genetic variation in seasonal reproductionin sheep. Proceedings 38th Annual Meeting of theEuropean Association for Animal Production, Lisbon,Portugal, 28 September–1 October, 14 pp.Ingrand S, Cournut S, Dedieu B, Antheaume F, 2003: Laconduite de la reproduction du troupeau de vaches allaitantes:mode´lisation des prises de décision. Prod Anim 16,263–270.Jansen HT, Jackson GL, 1993: Circannual rhythms in theewe: patterns of ovarian cycles and prolactin secretionunder two different constant photoperiods. Biol Reprod49, 627–634.Karsch FJ, Bittman EL, Foster DL, Goodman RL, Legan SJ,Robinson JE, 1984: Neuroendocrine basis of seasonalreproduction. Recent Progress Hormone Res 40, 185–232.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

46 P Chemineau, D Guillaume, M Migaud, JC Thie´ry, MT Pellicer-Rubio and B MalpauxKhaldi G, 1984: Variations saisonnie` res de l’activite´ ovarienne,du comportement d’oestrus et de la dure´e del’anoestrus post-partum des femelles ovines de raceBarbarine: influence du niveau alimentaire et de lapre´sence du mâle. PhD Thesis, University of Montpellier,France, 168 pp.Langlois B, Blouin C, 1997: Effect of a horse’s month of birthon its future sport performance. I. Effect on annualphenoypte indices. Ann Zootech 46, 393–398.Langlois B, Blouin C, 1998: Effect of a horse’s month ofbirth on its future sport performance. II. Effect on annualearning and annual earning per start. Ann Zootech 47,67–74.Langlois B, Blouin C, 2004: Statistical analysis of some factorsaffecting the number of horse births in France. Reprod NutrDev 44, 583–596.Larkin JE, Jones J, Zucker I, 2002: Temperature dependenceof gonadal regression in Syrian hamsters exposed to shortday lengths. Am J Physiol-Regul Integr Comp Physiol 282,R744–R752.Le Corre S, Segu L, Caldani M, Chemineau P, 1994:Differences in Ketanserin binding in the ventromedialhypothalamus of ewes responsive or refractory to shortdays. Neuroendocrinology 60, 589–600.Legan SJ, Karsch FJ, 1980: Photoperiodic control of seasonalbreeding in ewes: modulation of the negative feedbackaction of estradiol. Biol Reprod 23, 1061–1068.Lehman MN, Coolen LM, Goodman RL, Viguie C, BillingsHJ, Karsch FJ, 2002: Seasonal plasticity in the brain: the useof large animal models for neuroanatomical research.Reproduction Suppl 59, 149–165.Lincoln G, 1979: Photoperiodic control of seasonal breeding inthe ram: participation of the cranial sympathetic nervoussystem. J Endocrinol 82, 135–147.Magistrini M, Chanteloube P, Palmer E, 1987: Influence ofseason and frequency of ejaculation on production ofstallion semen for freezing. J Reprod Fertil Suppl 35, 127–133.Mahieu M, Jego Y, Driancourt MA, Chemineau P, 1989:Reproductive performances of Creole and Black-Belly ewesin the West Indies. A new major gene controlling ovulationrate? Anim Reprod Sci 19, 235–243.Malpaux B, 2006: Seasonal regulation of reproduction inmammals. In: Neill JD (ed.), Knobil and Neill’s Physiologyof Reproduction, 3rd edn. Elsevier, Amsterdam, pp. 2231–2281.Malpaux B, Maurice–Mandon F, Daveau A, Chemineau P,1995: Utilisation de la lumie` re et de la me´latonine pour lamaıˆ trise de la reproduction des ovins et des caprins.Rencontres Recherches Ruminants Paris 13–14 de´c 2, 379–386.Malpaux B, Viguie C, Thiery JC, Chemineau P, 1996:Controˆ le photope´riodique de la reproduction. Prod Anim9, 9–23.Malpaux B, Migaud M, Tricoire H, Chemineau P, 2001:Biology of mammalian photoperiodism and the critical roleof the pineal gland and melatonin. J Biol Rhythms 16, 336–347.Migaud M, Daveau A, Malpaux B, 2005: MTNR1A melatoninreceptors in the ovine premammillary hypothalamus:day–night variation in the expression of the transcripts. BiolReprod 72, 393–398.Ortavant R, Pelletier J, Ravault JP, Thimonier J, Volland P,1985: Photoperiod: main proximal and distal factor of thecircannual cycle of reproduction in farm mammals. OxfordRev Reprod Biol 7, 305–345.Palmer E, Driancourt MA, 1983: Some interactions of seasonof foaling, photoperiod and ovarian activity in the equine.Livest Prod Sci 10, 197–210.Paul MJ, Zucker I, Schwartz WJ, 2008: Tracking the seasons:the internal calendars of vertebrates. Philos Trans R SocLond B Biol Sci 363, 341–361.Pelletier J, Ortavant R, 1975: Photoperiodic control of LHrelease in the ram. II. Light-androgens interaction. ActaEndocrinol (Copenhagen) 78, 442–450.Pellicer-Rubio MT, Leboeuf B, Bernelas D, Forgerit Y,Pougnard JL, Bonne JL, Senty E, Chemineau P, 2007:Highly synchronous and fertile reproductive activity inducedby the male effect during deep anoestrus in lactatinggoats subjected to treatment with artificially long daysfollowed by a natural photoperiod. Anim Reprod Sci 98,241–258.Pellicer-Rubio MT, Leboeuf B, Bernelas D, Forgerit Y,Pougnard JL, Bonne JL, Senty E, Breton S, Brun F,Chemineau P, 2008: High fertility using artificial inseminationduring deep anoestrus after induction and synchronisationof ovulatory activity by the ‘male effect’in lactatinggoats subjected to treatment with artificial long days andprogestagens. Anim Reprod Sci (in press).Quirke JF, Hanrahan JP, Loughnane W, Triggs R, 1986:Components of the breeding and non-breeding seasons insheep: breed effects and repeatability. Irish J Agr Food Res25, 167–172.Ricordeau G, 1982: Selection for reduced seasonality and postpartumanoestrus. In: Second World Congress of GeneticsApplied to Livestock Production, Madrid, Spain, 5, 338–347.Salazar-Ortiz J, 2006: Bilan des effets du niveau d’alimentationsur la reproduction de la jument. PhD Thesis, University ofTours, France, 343 pp.Santiago-Moreno J, Lopez-Sebastian A, Gonzalez-Bulnes A,Gomez-Brunet A, Chemineau P, 2000: Seasonal changesin ovulatory activity, plasma prolactin, and melatoninconcentrations, in Mouflon (Ovis gmelini musimon) andManchega (Ovis aries) ewes. Reprod Nutr Dev 40, 421–430.Smith JF, Johnson DL, Reid TC, 1992: Genetic parametersand performance of flocks selected for advancedlambing date. Proc New Zealand Soc Anim Prod 52, 50(Abstr).Steger RW, Matt KS, Klemke HG, Bartke A, 1985: Interactionsof photoperiod and ectopic graphs on hypothalamicand pituitary function in male hamsters. Neuroendocrinology41, 89–96.Thiery JC, Malpaux B, 2003: Seasonal regulation ofreproductive activity in sheep: modulation of access ofsex steroids to the brain. Ann N Y Acad Sci 1007, 169–175.Thiery JC, Martin GB, Tillet Y, Caldani M, Quentin M,Jamain C, Ravault JP, 1989: Role of hypothalamic catecholaminesin the regulation of luteinizing hormone andprolactin secretion in the ewe during seasonal anestrus.Neuroendocrinology 49, 80–87.Thiery JC, Gayrard V, Le Corre S, Viguie C, Martin GB,Chemineau P, Malpaux B, 1995: Dopaminergic control ofLH secretion by the A15 nucleus in anoestrous ewes. JReprod Fertil Suppl 49, 285–296.Thiery JC, Lomet D, Schumacher M, Liere P, Tricoire H,Locatelli A, Delagrange P, Malpaux B, 2006: Concentrationsof estradiol in ewe cerebrospinal fluid are modulatedby photoperiod through pineal-dependent mechanisms. JPineal Res 41, 306–312.Thimonier J, Mauléon P, 1969: Variations saisonnie` res ducomportement d’oestrus et des activite´s ovariennes ethypophysaires chez les ovins. Ann Biol Anim Bioch Biophys9, 233–250.Viguie´ C, Caraty A, Locatelli A, Malpaux B, 1995: Regulationof luteinizing hormone-releasing hormone (LHRH)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Seasonality of Reproduction in Mammals 47secretion by melatonin in the ewe. I. Simultaneous delayedincrease in LHRH and luteinizing hormone pulsatile secretion.Biol Reprod 52, 1114–1120.Yenikoye A, 1984: Annual variations in estrual behavior, rateand possibilities for ovulation in Peulh ewes from Niger.Reprod Nutr Dev 24, 11–19.Author’s address (for correspondence): P Chemineau, UMR Physiologiede la Reproduction et des Comportements, INRA, CNRS,Université F. Rabelais, Haras Nationaux, 37380 Nouzilly, France.E-mail: philippe.chemineau@tours.inra.frConflict of interest: P Chemineau has received a research grant fromCEVA Animal Health and has no potential conflicts to declare.J-C Thiéry has received a research grant from ‘‘Institut de RechercheServier’’ and has no potential conflicts to declare.D Guillaume has no potential conflicts to declare.M Migaud has received a research grant from ‘‘Institut de RechercheServier’’ and has no potential conflicts to declare.M-T Pellicer-Rubio has no potential conflicts-to declare.B Malpaux has received a research grant from ‘‘Institut de RechercheServier’’, from ‘‘CEVA Animal Health’’ and has no potential conflictsto declare.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 48–56 (2008); doi: 10.1111/j.1439-0531.2008.01142.xISSN 0936-6768Mechanisms for Dominant Follicle Selection in Monovulatory Species: AComparison of Morphological, Endocrine and Intraovarian Events in Cows, Maresand WomenM Mihm 1 and ACO Evans 21 Division of Cell Sciences, Faculty of Veterinary Medicine, University of Glasgow, Glasgow, UK; 2 School of Agriculture, Food Science andVeterinary Medicine and the Conway Institute, University College Dublin, Belfield, Dublin 4, IrelandContentsThe selection of a single ovarian follicle for further differentiationand finally ovulation is a shared phenomenon inmonovulatory species from different phylogenetic classes. Thecommonality of dominant follicle (DF) development leads usto hypothesize that mechanisms for DF selection are conserved.This review highlights similarities and differences infollicular wave growth between cows, mares and women,addresses the commonality of the transient rises in FSHconcentrations, and discusses the follicular secretions oestradioland inhibin with their regulatory roles for FSH. In allthree species, rising FSH concentrations induce the emergenceof a follicle wave and cohort attrition occurs during decliningFSH concentrations, culminating in DF selection. Cohortsecretions are initially responsible for declining FSH, which issubsequently suppressed by the selected DF lowering it belowthe threshold of FSH requirements of all other cohort follicles.The DF acquires relative FSH-independence in order tocontinue growth and differentiation during low (cow, mare)or further declining FSH concentrations (women), and thusmay be the one cohort follicle with the lowest FSH requirementdue to enhanced FSH signalling. In all three monovulatoryspecies a transition from FSH- to LH-dependence ispostulated as the mechanism for the continued development ofthe selected DF. In addition, FSH and IGF enhance eachother’s ability to stimulate follicle cell function and access ofIGF-I and -II to the type 1 receptor is regulated by IGFbinding proteins that are in turn regulated by specificproteases; all of which have been ascribed a role in DFdevelopment. No fundamental differences in DF selectionmechanisms have been identified between the different speciesstudied. Thus functional studies of the selection of DFs incattle and mares are also valuable for identifying genes andpathways regulating DF development in women.IntroductionLarge antral follicle growth in ovaries does not occurrandomly. For example in cattle, detailed endocrine andquantitative morphological studies followed by thosewhich allowed the individual monitoring of antralfollicles from 3 mm in diameter [surgical (ink markingof follicles) and ultrasound studies] showed that growthis wave-like and coordinated, and is reflected in regularfluctuations of follicular hormone secretions (Evans2003; Mihm and Bleach 2003). In adults from monovulatoryspecies final antral follicle growth is characterizedby a shared phenomenon whereby a single follicle1 Please note that in this review the general term ‘cows’is used to include heifers and parous cows.[the dominant follicle (DF)] is selected from multipleand simultaneously growing cohort follicles to undergofinal differentiation and ultimately ovulate, or at leastacquire the ability for ovulation (Ginther et al. 2001a).Clearly, the development of a mechanism to guaranteesingle offspring increased the chances of its pre- andpostnatal survival in these species.There is commonality in follicle wave growth in verydifferent species from different classes, orders andfamilies of the phylogenetic tree. This leads us tohypothesize that conserved mechanisms for DF selectionexist. Identification of such shared mechanisms isof great importance to veterinary medicine and science,as well as human reproductive medicine. In all studiedmonovulatory species disruption of the selection phenomenonoccurs; for example, when twin or multipleovulations occur, or in polycystic ovary syndrome,which is diagnosed in more than 75% of women withanovulatory infertility (Franks et al. 2006). This canresult in temporary or long-term infertility due to cycleabnormalities, abortions, increased foetal and neonataldisease and mortality, or sub(in)fertility followingdystocia and post-partum disease when multiples aredelivered. Thus it is essential to develop a betterunderstanding and improved methods for diagnosisand treatment of any existing pathology.Conversely, controlled interference with DF selectionto improve the reproductive efficiency or fertility in thedifferent species (which include treatments for twinning,or superovulatory treatments for embryo transfer) mayhave enormous economic (agriculturally important species)or psychological (women) benefits. Interestingly,however, the likelihood of double ovulations ⁄ twinningdecreases even among known monovulatory speciesfrom mares (20% in thoroughbred mares, Ginther 1992)to cows 1 (

Dominant Follicle Selection in Cows, Mares and Women 49been determined in the three species. Subsequently, twoimportant intrafollicular functions regulated during DFselection (FSH and LH-responsiveness, insulin-likegrowth factor (IGF) binding protein production) areproposed as examples of conserved DF selection mechanisms.Finally, new candidate genes for attainment ofdominance are presented from genomic bovine studies.Dominant Follicle Selection: Follicle WaveMorphologyAt specific times during the luteal phase of the oestrouscycle in cows, but also at intervals of 7–10 days in mostother reproductive states (and this includes the prepubertal,post-partum and early pregnant anoestrousstate), rapid growth of a cohort of small (from largerthan 1 mm in diameter) antral follicles to 3–5 mm andbeyond is observed using daily ovarian ultrasound(Roche et al. 1998; Ireland et al. 2000). This observationhas been termed ‘emergence of a follicular wave’ and isfollowed by continued growth of a constantly reducingnumber of cohort follicles over the next 3 days, whereonly the minority will reach diameters of 6 mm, two orthree follicles will reach 7–8 mm, and generally only onefollicle is selected to be the DF and grows beyond 8 mmin diameter (Fig. 1). For a large number of in vivostudies, the time of DF selection in the bovine has beenroutinely defined using ultrasound criteria such as thetime of divergence between the DFs continued rapidgrowth vs the reduced growth shown by its closestcompetitor, the largest subordinate follicle (SF1) (onsetof deviation; Ginther et al. 1997), or a minimum size of8.5 mm and minimum diameter difference of 1 to>2 mm to the SF1 (Ginther et al. 1997; Mihm et al.1997, 2008). All other cohort (subordinate) folliclescease growth at different times after wave emergence,decrease in size, and eventually disappear. In general,two or three follicle waves emerge during the lutealFig. 1. The relationship between the transient rise in FSH, follicle wavegrowth, which culminates in selection of the DF, and the follicularsecretions oestradiol and inhibin in cows, mares and women. Note thatin women this is most likely the ovulatory follicle wave which emergesat luteolysis and grows during the early follicular phase, in mares theovulatory follicle wave emerges in the second half of the luteal phaseand the DF is selected at luteolysis, and in cows the most studied folliclewave is the first wave of the cycle, which is anovulatory and emerges atovulation, with DF selection occurring in the early luteal phase.Cohort, dominant and subordinate follicles in mares are generally twicethe size of follicles in the other two species. Note the longer FSH declinein mares compared with cows, and the particularly long FSH decline inwomen. DF, dominant follicle, SF, subordinate folliclesphase of the 21-day oestrous cycle in cattle, eachselecting a DF, and with the final DF selected atluteolysis or shortly after and thus becoming theovulatory DF in the relatively short follicular phase(Savio et al. 1988; Ginther et al. 1989). Very rarely theemergence of a wave has been detected which may growfor one day but does not lead to selection of a new DF(Mihm et al. 1999). There are several interesting featuresof DF selection in the bovine, apparent when studyingthis process closely by ultrasound: (1) the future DFgenerally seems to emerge first, 6–7 h before the largestSFI, and the size advantage is maintained throughoutthe common parallel growth phase until deviation(Ginther et al. 1997); (2) the number of cohort folliclesemerging over a 48–72 h period is highly variablebetween animals and can also vary within the cycle,ranging from only 2 to more than 50 small antralfollicles per wave (Burns et al. 2005); (3) follicle numbersper wave in animals representing the two extremes, thatis in animals with very low (25) numbers of follicles per wave are consistentwithin the cycle and from cycle to cycle (Burns et al.2005); (4) follicle numbers per wave do not usuallyinfluence the process of selecting one single DF, as nohigher incidence of co-dominant follicles has beenreported in animals showing very high follicle numbers(Burns et al. 2005), or in animals treated exogenouslywith bovine growth hormone which increases thenumber of small antral follicles (cohort follicles) peranimal (Gong et al. 1991). The exception are cows fromherds selected for twinning such as those described inEchternkamp 2000; (5) the diameter of the DF at thetime of selection (approximately 8.5 mm) is very consistentacross all (oestrous and anoestrous) reproductivestates, and across different breeds and types of cattle,and in cyclic animals is independent of when during theluteal phase (beginning, mid, end) the DF emerges(Roche et al. 1998; Ireland et al. 2000; Ginther et al.2001a); and (6) co-dominance in animals not geneticallyselected for twinning is generally rare (approximately5%) but can occur either temporarily with late selectionof a single follicle, or throughout the complete dominanceperiod, leading to twin ovulations if the last waveof the cycle is affected (Echternkamp 2000; Kulick et al.2001) Overall, these features indicate that the DFselection mechanism in the bovine is very robust andin unselected cows can only be inhibited by pharmacologicalinterference which elevates FSH or abolishesGnRH pulsatility (Prendiville et al. 1995; Gong et al.1996; Mihm et al. 1997).In contrast to cows, in mares and women not allfollicle waves detected during an interovulatory intervalwill result in the selection of a DF (these are termed‘minor waves’ where cohort follicles never reach the sizeof a DF; details of major and minor wave follicledynamics in both species are reported in Ginther 1993;Baerwald et al. 2003; Ginther et al. 2004a). The ovulatorywave in mares, which is the only follicle waveguaranteed to result in selection of a DF, emerges in thesecond half of the luteal phase leading to selection of theDF coincident with luteal regression. In women, however,the ovulatory wave only emerges during the firsthalf of the follicular phase, and again is the only waveÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

50 M Mihm and ACO Evansguaranteed to result in a DF. Thus, mares and womenshare very similar wave dynamics: (1) The occurrence ofmajor (leading to selection of a DF) anovulatory wavesduring the luteal phase in 20% (mares) or 30% (women)of cycles. (2) The occurrence of minor waves (whichdescribes the emergence and growth of cohort folliclesfollowed by their regression without DF selection)during both the luteal and follicular phases of the cycle,albeit the percentage differs between mares (almost100%) and women (only 57%). (3) The high percentageof pre-deviation follicles (mares 37%, women 48%)which are follicles that emerge with or ahead of the DFin ovulatory waves, reach DF size 1 day before theovulatory DF but begin to regress at the time ofdeviation of the ovulatory DF. At emergence anddeviation, respectively, cows and women share verysimilar cohort (3–6 mm), largest subordinate (7–8 mm)and DF diameters (8.5–10 mm) determined by ultrasound.Only women show an enhanced growth rate ofthe DF and regression rate of SF following deviation,while in mares and cows growth rates of DF aremaintained at rates seen before deviation (Ginther et al.2001b, 2004a). While ‘local’ effects of DF on antralfollicles from the ipsi- vs the contralateral ovary havenot been described in cows or mares, in women the DFaffects subordinate growth and atresia on the ispilateralovary following its selection, leading to ‘corona formation’around the DF towards the end of the follicularphase (Gore et al. 1997). Compared to women, twinovulations are frequent in mares (20%, Ginther 1992;Ginther et al. 2004a) and have recently been shown toreach similar levels in high-yielding dairy cows(Wiltbank et al. 2000). Similar to cows, follicle waveshave also been observed during the later half of theanovulatory season (Donadeu and Ginther 2002a;Watson and Al-zi’abi 2002) and during early pregnancyin mares (Ginther and Bergfelt 1992). However, theoccurrence of follicular waves during anovulatory states(such as during pregnancy or following weight loss) inwomen does not appear to have been investigated.The FSH Rise and DeclineIn all three monovulatory species, rising FSH concentrationsin systemic circulation induce the emergence ofa follicle wave which is usually detected at the time ofthe FSH peak (Austin et al. 2001; Ginther et al. 2005)(Fig. 1). Cohort attrition occurs during declining FSHconcentrations, and culminates in DF selection (cows:Adams et al. 1992; Sunderland et al. 1994; mares:Bergfelt and Ginther 1993; women: Van Santbrinket al. 1995). All growing cohort follicles contribute tothe FSH suppression, leading to a direct relationshipbetween the number of cohort follicles growing inresponse to the FSH rise and the degree of FSHsuppression (Gibbons et al. 1997; Burns et al. 2005;Ginther et al. 2005). The FSH decline causes DFselection and this has been shown functionally bypreventing the decline or elevating FSH in cows(Adams et al. 1993; Mihm et al. 1997), mares (Donadeuand Ginther 2001) and women (Schipper et al.1998; Hohmann et al. 2001). Therefore, DF selection isa systemic process simultaneously affecting cohortgrowth on both ovaries via reductions in FSH followinga transient rise. However, the incidence of transientFSH rises, the magnitude of each rise, and the lengthof the decline differ between the species. In cows,sequential FSH rises of similar magnitude as thegonadotrophin surge are associated with new folliclewaves during the luteal phase of the oestrous cycle(Adams et al. 1992; Sunderland et al. 1994), in thepost-partum period (Crowe et al. 1998; Stagg et al.1998), during pregnancy (Ginther et al. 1996) andbefore puberty (Evans et al. 1994). Functional in vivostudies showed that suppressing the first or second riseof FSH during the cycle or long-term suppression ofGnRH pulsatility will inhibit wave emergence, whichcan again be induced using exogenous FSH (Mihm andBleach 2003).Similar to cows, several transient FSH rises have beendemonstrated during the luteal phase in the mares (200–400% of basal levels), but interestingly, luteal rises arenot always associated with emergence of a follicle waveas detected by ultrasound (Gastal et al. 1997; Gintheret al. 2005). Three features are different in womencompared to cows and mares: (1) no minor or majorwaves emerge during the early to mid luteal phase(Baerwald et al. 2003); (2) very small magnitude FSHrises can induce emergence of the ovulatory waveindicating a relatively high sensitivity of the populationof small antral follicles to FSH (Baird 1987; VanSantbrink et al. 1995); and (3) DF selection occurs alsoduring the initial decline in FSH, but FSH continues todecline during the long dominance period of the DF.Thus FSH is elevated for more than 10 days in thefollicular phase (Van Santbrink et al. 1995; Baerwaldet al. 2003), which is in contrast to cows and mareswhere a final reduction of FSH to nadir levels occursjust after deviation (and this final decline is hypothesizedto induce SF1 atresia at deviation, Ginther et al. 1999;Bergfelt et al. 2001).The Follicular Secretions Oestradiol and InhibinThe FSH decline causes DF selection, and initialsuppression of FSH concentrations following the riseis contributed to the cohort follicles growing after waveemergence in all three monovulatory species (Fig. 1)(Ginther et al. 2001a). Thus, cohort secretions into thesystemic circulation must induce the decline in FSH.Cohort follicles, therefore, indirectly cause their ownatresia except for the DF which must acquire relativeFSH-independence in order to grow and differentiateduring low (cows, mares) or further declining (women)FSH concentrations. Dominant follicle secretions thencontinue to suppress FSH to prevent further waveemergence. Cohort and DFs in cows, mares and womenproduce the FSH-inhibitors oestradiol and inhibin-A(cows: Sunderland et al. 1996; Mihm et al. 1997; Bleachet al. 2001; mares: Bergfelt et al. 2001; Watson et al.2002; women: Schneyer et al. 2000; Laven and Fauser2004), and in cows and women cohort follicles produceinhibin-B (Beg et al. 2002; Laven and Fauser 2004). Inheifers and cows during growth of the first cohort of theoestrous cycle, inhibin-A increases in circulation coincidentwith small rises in systemic oestradiol reaching anÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Dominant Follicle Selection in Cows, Mares and Women 51early plateau before DF selection (inhibin-A) or a peakon the day of DF selection (oestradiol) while FSHconcentrations decline (Bleach et al. 2001; Kaneko et al.2002; Parker et al. 2003). Follicular inhibin-A secretionappears to be the most important regulator of FSHconcentrations during initial follicle wave growth incattle and is aided by oestradiol secreted from the DF atthe onset of dominance (Ginther et al. 2000, 2001a;Mihm and Bleach 2003): when atresia of growing cohortfollicles is caused using exogenous steroid treatmentsinhibin-A declines rapidly followed by a rise in FSHconcentrations despite high systemic oestradiol (Mihmet al. 2001). Similarly, FSH only rises again during lossof dominance of the first DF of the cycle when inhibin-Aconcentrations reach their nadir, despite oestradiolhaving declined two days previously (Bleach et al.2001). However, other follicular secretions may also beinvolved in the regulation of FSH during cohort growthsuch as follistatin (Mihm and Bleach 2003).In mares oestradiol concentrations rise in circulationapproximately 1 day before selection of the DF andluteolysis (Bergfelt et al. 2001). As FSH concentrationsdecline immediately after wave emergence which isseveral days from deviation, other follicular secretionsthan oestradiol must be responsible for the initial FSHsuppression. Inhibin-A is synthesized by cohort andDFs in the mares (Watson et al. 2002) and rises in totalinhibin have been observed associated with DF growth(Bergfelt et al. 2001). In women, similar to cows,oestradiol concentrations show a small but significantincrease after wave emergence. Very importantly, however,inhibin-B concentrations rise during cohort growthand appear to peak just before DF selection, decliningduring the long dominance period (Muttukrishna et al.2000). Thus, inhibin-B secretion from cohort folliclescontrols the initial FSH decline responsible for DFselection in women (Schneyer et al. 2000). Followingselection of the ovulatory DF in the follicular phase,oestradiol and inhibin-A concentrations rise rapidly,and together control the remaining protracted FSHdecline (Muttukrishna et al. 1994, 2000). In contrast tocattle, where oestradiol only exerts a transient inhibitionon FSH (O’Rourke et al. 2000), oestradiol is clearlyrequired for continued FSH suppression in women, asanti-oestradiol treatment (passive immunization orreceptor antagonist administration) during the midfollicularand the end-follicular phase will cause a rise inFSH and abolish DF selection or cause DF atresia(Zeleznik 2001).Intraovarian (Cellular) Mechanisms for DFSelection: Gonadotrophin Responsiveness andInsulin-like Growth Factor Binding Protein(IGFBP) ExpressionAntral follicles of 2–5 mm (cows, women) or 10–13 mm(mares) are clearly dependent on elevated FSH forcontinued development, firstly because they undergoatresia in the absence of FSH rises, and secondly, cohortfollicles undergo atresia after their emergence when FSHdeclines again following its transient rise. At deviationbovine DFs can maintain follicular cell proliferation andenhance oestradiol production despite declining FSHconcentrations, and the DF may be the one cohortfollicle with the lowest FSH requirement due toincreased or maintained high FSH receptor mRNAexpression and FSH-binding allowing it to pass the8.5 mm diameter threshold (Ireland and Roche 1983;Evans and Fortune 1997; Bao and Garverick 1998;Rozell et al. 2005). Interestingly, the follicle with highestintrafollicular oestradiol 33 h after the FSH peak (themost successful cohort follicle in the wave) also hashighest activin-A concentrations in follicular fluid(Austin et al. 2001), which is known to increase FSHreceptor expression in granulosa cells (Knight andGlister 2006). This information is extended to women,where it is postulated that the follicle which becomes theDF must first develop ‘activin-dominance’ before the‘inhibin-dominance’ characteristic for the DF (Schneyeret al. 2000).In all three monovulatory species a transition fromFSH- to LH-dependence is postulated as the mechanismwhich allows continued development of theselected DF (Ginther et al. 2001a), and in the animalspecies this is based on functional studies abolishing oradding LH pulses during follicle wave growth and afterDF selection (Gastal et al. 1999; Mihm and Bleach2003). Granulosa cells from newly selected bovine DFhave acquired an enhanced ability to bind LHcompared with SF1s and show increased mRNAexpression for the LH receptor during their earlygrowth phase (Ireland and Roche 1983; Evans andFortune 1997; Bao and Garverick 1998; Evans et al.2004; Mihm et al. 2006). Similarly, in mares theselected DF is characterized by its absolute dependenceon increased LH for enhanced oestradiol synthesis andcontinued growth (Bergfelt et al. 2001). This alsoapplies in women, as it is suggested that LH-inducedcAMP production in granulosa cells from DF replacesthe FSH-stimulated cAMP production seen in growingcohort follicles (Zeleznik 2001). However, at this pointit is not clear in any species whether the identifiedincrease in LH receptor expression in the DF is cause(will the first follicle that acquires LH receptor activitybecome selected?) or consequence of the selectionprocess. Recently a slight increase in LH receptormRNA expression was detected in granulosa cells fromthe largest bovine follicle a few hours before the onsetof deviation and the appearance of morphological andhormonal (follicular fluid oestradiol) differences, comparedwith the second largest follicle (Beg et al. 2001).This may indicate that acquisition of increased LHreceptor expression characterizes the follicle with thedevelopmental advantage even before other differencesto the cohort become apparent. As IGF1 has beenshown to enhance FSH-induced granulosa cell differentiation,particularly LH receptor acquisition (Hirakawaet al. 1999), the increased free IGF concentrationspostulated for the future DF during cohort growth (seebelow) may be responsible for such a developmentaladvantage.Because FSH is anti-apoptotic, proliferative anddifferentiative (it stimulates aromatase production andLH receptor acquisition), enhanced FSH-responsivenessor the ability to amplify FSH-stimulated granulosacell functions in one cohort follicle may lead toÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

52 M Mihm and ACO Evansits selection. FSH and IGF enhance each other’sability to stimulate proliferation, steroid, activin andinhibin production in bovine granulosa cells (Glisteret al. 2001). Access of IGF1 and 2 to the type 1receptor which is the main receptor conveying theirproliferative and steroidogenic activities (Spicer andAad 2007) is regulated by a group of differentmolecular weight IGF binding proteins with differingligand affinity and differential expression patterns inhealthy and atretic follicles (San Roman and Magoffin1993; Monget et al. 2002; Roberts and Echternkamp2003). In all three monovulatory species characteristicdifferences in the expression pattern of IGF bindingproteins have been observed between follicles growingwithin a wave, or between dominant and subordinateor cohort follicles (Schuller et al. 1993; Stewart et al.1996; Mihm et al. 1997). This has been attributed tothe ability of FSH-stimulated cohort follicles or theselected DF to downregulate expression of specificIGF binding proteins (e.g. IGF binding protein 2) orincrease their proteolytic processing (for example, IGFbinding proteins 4 and 5) by expressing a proteasesuch as PAPP-A, thus ‘freeing’ IGF for stimulation ofcellular proliferation and steroidogenesis duringdeclining or low FSH (Conover et al. 2001; Mongetet al. 2002). In vivo experiments in heifers have showna reduction in the lower molecular weight IGFbinding proteins 2, 4 and 5 in the largest cohortfollicles and the selected dominant compared withSF1s (Mihm et al. 1997; Austin et al. 2001); a specificreduction in IGF-BP4 (the non-glycosylated 24 kDaform) in the cohort follicle most likely to become theDF (Mihm et al. 2000); and increased proteolyticactivity ⁄ PAPP-A in the DF at deviation which isspecific for the two lowest molecular weight IGFbinding proteins 4 and 5 (Rivera et al. 2001; Riveraand Fortune 2003). Insulin-like growth factor bindingprotein 2 is increased in SF1 in mares at deviation(Donadeu and Ginther 2002b). In women amounts ofIGF binding protein 3 (the highest molecular weightform), the 28 kDa and the 24 kDa (most likely IGFbinding protein 4) in follicular fluid are reduced indominant compared with normal cohort or atreticfollicles, while IGF binding protein 2 increases inatretic follicles; increased PAPP-A is also considered amarker for DF selection (San Roman and Magoffin1993; Schuller et al. 1993; Amato et al. 1998; Conoveret al. 2001). Therefore, the ability to differentiallyreduce the expression of specific IGF binding proteinsor induce their proteolysis during declining FSHappears to be associated with DF selection in allthree monovulatory species. Recent in vivo studies inheifers and mares whereby IGF-1 and IGF bindingprotein 3 were administered into the SF1 and DF,respectively, confirm the differentiative and steroidogenicroles of IGF during the selection process(Ginther et al. 2004b; c). In addition, actions ofFSH and IGF are mediated (at least in part) viaAkt (protein kinase B) and Erk (extracellular regulatedkinase) in bovine granulosa and theca cells, anda greater activity of these pathways in newly selectedDFs compared to contemporary SF1s has beendemonstrated (Ryan et al. 2007). Thus a differencein these pathways may mark the future DF from SF1sbefore differences in follicle diameters or follicularfluid oestradiol and IGFBP concentrations appear.Because antral follicle growth and differentiationclearly depends on nutrient, hormone and growth factorsupply to the individual follicle, one other sharedmechanism recently proposed for DF selection in allthree monovulatory species is the possible differentialregulation of blood vessel formation and permeability inthe theca layers of cohort follicles. This hypothesis issupported by several studies, which have shownincreased expression of angiogenic factors such asvascular endothelial growth factor (VEGF) with increasedfollicle differentiation, increased blood flow inthe selected DF, and follicular arrest and atresiafollowing the experimental reduction of VEGF signalling(Acosta et al. 2004, 2005; Hunter et al. 2004; Fraserand Duncan 2005). Further functional studies are nowneeded to elucidate whether angiogenic factors arecausatively involved in DF selection.Candidate Genes for Regulation of DominantFollicle Selection Identified Using the BovineModelThe bovine follicle wave model has so far been used forfour studies using genomic approaches to identify genesexpressed by granulosa and theca cells which are regulatedat the time of DF selection, thus providing candidatesfor further functional studies of follicle survival anddifferentiation in monovulatory species (suppressionsubtractive hybridization, Sisco et al. 2003; Fayad et al.2004; custom cDNA microarray, Evans et al. 2004; serialanalysis of gene expression, SAGE, Mihm et al. 2008).Additional molecular studies have investigated transcriptionof candidate genes identified through genomicapproaches during growth of the follicle wave (Siscoet al. 2003; Zielak et al. 2007; Forde et al. 2008). Forexample, comparisons of granulosa cell mRNA expressionprofiles between the cohort follicle most likely fatedto become the DF, the newly selected DF and the SF1have led to the identification of a number of knowndifferentiation genes, but also of genes not previouslyassociated with bovine DF development or ovarianfunction, which are upregulated at dominance and areinvolved in oestradiol synthesis, anti-oxidant activity,LH-responsiveness, cell proliferation, anti-apoptoticactivity, and RNA splicing (Table 1; Mihm et al. 2008).Similar studies have not been carried out in mares andwomen. Because in these two species induction of higharomatase activity and increased LH responsiveness arealso functions of the selected DF (Ginther et al. 2001a;Zeleznik 2001), other cellular functions or pathwaysassociated with enhanced mRNA expression for thearomatase and LH receptor genes and identified usingthe bovine follicle wave model may be relevant for DFselection in all three monovulatory species, and thusneed to be tested further using functional in vivo and invitro approaches. For this the bovine DF selectionmodel is particularly suited, as difficulties in obtainingtissues or suitable technological platforms will prohibitthe study of follicle development in mares and women ascompletely as in cattle.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Dominant Follicle Selection in Cows, Mares and Women 53Table 1. Summary table of granulosa cell genes expressed at highest levels in the largest cohort follicle following emergence, or in the newlyselected dominant follicle compared to the most oestrogenic cohort follicle before selection and(or) the largest subordinate follicle (Evans et al.2004; Forde et al. 2008; Mihm et al. 2008; Zielak et al. 2007). Such identified genes are proposed as gene candidates for the regulation of DFselection in cowsBovine candidate genesfor DF selection Intracellular role Molecular approach ReferenceCell differentiationCYP19A1 a Estradiol synthesis, cell differentiation Custom microarray andqRT-PCR, SAGE andqRT-PCRDQ004742Uncharacterized, splice variant foraromataseGADD45BCell differentiation, response to cellularstress, apoptosisEvans et al. (2004),Mihm et al. (2008)SAGE and qRT-PCR Mihm et al. (2008)SAGE and qRT-PCR Mihm et al. (2008)GSTA2 Steroid synthesis, antioxidant SAGE and qRT-PCR Mihm et al. (2008)INHA Cell differentiation, inhibin synthesis SAGE and qRT-PCR Mihm et al. (2008)LHCGRGonadotrophin receptor, cell differentiationCustom microarray andEvans et al. (2004)qRT-PCRTBC1D1 Cell differentiation, proliferation Custom microarray andZielak et al. (2007)qRT-PCROvary-specific acidic proteinUncharacterized but linked to highSAGE and qRT-PCR Mihm et al. (2008)oestradiol synthesis in pre-ovulatoryfolliclesCell proliferation, apoptosisCCND2 Cell proliferation SAGE and qRT-PCR Mihm et al. (2008)MCL-1 Cell proliferation, antiapoptotic factor Custom microarray andEvans et al. (2004)qRT-PCRMIF Cell proliferation, antiapoptotic factor SAGE and qRT-PCR Mihm et al. (2008)Signal transductionANXA2 Tissue development, signal transduction SAGE and qRT-PCR Mihm et al. (2008)BCAR1 Signal transduction, cell proliferation Custom microarray andForde et al. (2008)qRT-PCRCALM2 Signal transduction SAGE and qRT-PCR Mihm et al. (2008)CLIC1 Signal transduction, chloride transport SAGE and qRT-PCR Mihm et al. (2008)Protein, DNA and RNA synthesisDICE-1Nuclear processes (DNA repair, transcription,Custom microarray andEvans et al. (2004)RNA splicing)qRT-PCRRFC4 Nuclear processes (DNA repair) SAGE and qRT-PCR Mihm et al. (2008)SFRS9 Nuclear processes (RNA splicing) SAGE and qRT-PCR Mihm et al. (2008)SLC22A17 Protein synthesis, ion transport SAGE and qRT-PCR Mihm et al. (2008)a Gene symbols are presented.ConclusionsThe similarities in follicle wave dynamics, the transientFSH rise and its regulation, follicular acquisition ofFSH and LH-responsiveness, and in the differentialpattern of IGF binding protein expression in cohort andDFs are striking between the three monovulatoryspecies. Differences exist mainly between cows andwomen in the follicle wave pattern, between womenand the two animal species in the extent of the FSHdecline, and in the inhibin forms secreted by cohortfollicles, and possibly also in the specific form of IGFbinding protein reduced in DF. Because of the considerableadvantage of large, relatively easily accessiblefollicles, mares present themselves as well-suited forfunctional experiments involving in vivo sampling oradministration of candidate substances. However, thebovine model is most easily monitored, extremely wellcharacterized,and has the advantage that its genomehas been sequenced allowing comparative molecularstudies with human or rodent models of follicle differentiationnot possible so far in other animal species. Inaddition, the bovine model has recently been used todetermine the relationship between low and high cohortfollicle numbers, the primordial follicle pool, andresponse to superovulatory treatment (Ireland et al.2007a,b). Thus, using selected heifers or cows as a modelof low antral follicle numbers may represent an excitingopportunity to study effects of reduced primordial poolsize on future cohort follicle or DF function simulatingevents in pre-menopausal women.ReferencesAcosta TJ, Gastal EL, Gastal MO, Ginther OJ, 2004:Differential blood flow changes between the future dominantand subordinate follicles precede diameter changesduring follicle selection in mares. Biol Reprod 71, 502–507.Acosta TJ, Hayashi K-G, Matsui M, Miyamoto A, 2005:Changes in follicular vascularity during the first follicularwave in lactating cows. J Reprod Dev 51, 273–280.Adams GP, Matteri RL, Kastelic JP, Ko JCH, Ginther OJ,1992: Association between surges of follicle-stimulatinghormone and the emergence of follicular waves in heifers.J Reprod Fertil 94, 177–188.Adams GP, Kot K, Smith CA, Ginther OJ, 1993: Selection ofa dominant follicle and suppression of follicular growth inheifers. Anim Reprod Sci 30, 259–271.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

54 M Mihm and ACO EvansAmato G, Izzo A, Tucker A, Bellastella A, 1998: Insulin-likegrowth factor binding protein-3 reduction in follicular fluid inspontaneous and stimulated cycles. Fertil Steril 70, 141–144.Austin EJ, Mihm M, Evans ACO, Knight PG, Ireland JLH,Ireland JJ, Roche JF, 2001: Alterations in intrafollicularregulatory factors and apoptosis during selection of folliclesin the first follicular wave of the bovine estrous cycle. BiolReprod 64, 839–848.Baerwald AR, Adams GP, Pierson RA, 2003: Characterizationof ovarian follicular wave dynamics in women. Biol Reprod69, 1023–1031.Baird DT, 1987: A model for follicular selection and ovulation– lessons from superovulation. J Steroid Biochem Mol Biol27, 15–23.Bao B, Garverick HA, 1998: Expression of steroidogenicenzyme and gonadotropin receptor genes in bovine folliclesduring ovarian follicular waves: a review. J Anim Sci 76,1903–1921.Beg MA, Bergfelt DR, Kot K, Wiltbank MC, Ginther OJ,2001: Follicular-fluid factors and granulosa-cell gene expressionassociated with follicle deviation in cattle. Biol Reprod64, 432–441.Beg MA, Bergfelt DR, Kot K, Ginther OJ, 2002: Follicleselection in cattle: dynamics of follicular fluid factors duringdevelopment of follicle dominance. Biol Reprod 66, 120–126.Bergfelt DR, Ginther OJ, 1993: Relationships between FSHsurges and follicular waves during the estrous-cycle inmares. Theriogenology 39, 781–796.Bergfelt DR, Gastal EL, Ginther OJ, 2001: Response ofestradiol and inhibin to experimentally reduced luteinizinghormone during follicle deviation in mares. Biol Reprod 65,426–432.Bleach ECL, Glencross RG, Feist SA, Groome NP, KnightPG, 2001: Plasma inhibin A in heifers: relationship withfollicle dynamics, gonadotropins, and steroids during theestrous cycle and after treatment with bovine follicular fluid.Biol Reprod 64, 743–752.Burns DS, Jimenez-Krassel F, Ireland JLH, Knight PG,Ireland JJ, 2005: Numbers of antral follicles during follicularwaves in cattle: evidence for high variation among animals,very high repeatability in individuals, and an inverseassociation with serum follicle-stimulating hormone concentrations.Biol Reprod 73, 54–62.Conover CA, Faessen GF, Ilg KE, Chandrasekher YA,Christiansen M, Overgaard MT, Oxvig C, Giudice LC,2001: Pregnancy-associated plasma protein-A is the insulinlikegrowth factor binding protein-4 protease secreted byhuman ovarian granulosa cells and is a marker of dominantfollicle selection and the corpus luteum. Endocrinology 142,2155–2158.Crowe MA, Padmanabhan V, Mihm M, Beitins IZ, Roche JF,1998: Resumption of follicular waves in beef cows is notassociated with periparturient changes in follicle-stimulatinghormone heterogeneity despite major changes in steroid andgonadotropin concentrations. Biol Reprod 58, 1445–1450.Donadeu FX, Ginther OJ, 2001: Effect of number anddiameter of follicles on plasma concentrations of inhibinand FSH in mares. Reproduction 121, 897–903.Donadeu FX, Ginther OJ, 2002a: Follicular waves andcirculating concentrations of gonadotrophins, inhibin andoestradiol during the anovulatory season in mares. Reproduction124, 875–885.Donadeu FX, Ginther OJ, 2002b: Changes in concentrationsof follicular-fluid factors during follicle selection in mares.Biol Reprod 66, 1111–1118.Echternkamp SE, Roberts AJ, Lunstra DD, Wise T, Spicer LJ,2004: Ovarian follicular development in cattle selected fortwin ovulations and births. J Anim Sci 82, 459–471.Echternkamp SE, 2000: Endocrinology of increased ovarianfolliculogenesis in cattle selected for twin births. Proc AmSoc Anim Sci. Available at http://www.asas.org/jas/symposia/proceedings/0935.pdf.Evans ACO, 2003: Characteristics of ovarian follicle developmentin domestic animals. Reprod Domest Anim 38, 240–246.Evans ACO, Fortune JE, 1997: Selection of the dominantfollicle in cattle occurs in the absence of differences in theexpression of messenger ribonucleic acid for gonadotropinreceptors. Endocrinology 138, 2963–2971.Evans ACO, Adams GP, Rawlings NC, 1994: Endocrine andovarian follicular changes leading up to the first ovulation inprepubertal heifers. J Reprod Fertil 100, 187–194.Evans ACO, Ireland JLH, Winn ME, Lonergan P, Smith GW,Coussens PM, Ireland JJ, 2004: Identification of genesinvolved in apoptosis and dominant follicle developmentduring follicular waves in cattle. Biol Reprod 70,1475–1484.Fayad T, Levesque V, Sirois J, Silversides DW, Lussier JG,2004: Gene expression profiling of differentially expressedgenes in granulosa cells of bovine dominant follicles usingsuppression subtractive hybridization. Biol Reprod 70, 523–533.Forde N, Mihm M, Canty MJ, Zielak AE, Baker PJ, Park SD,Lonergan P, Smith GW, Coussens PM, Ireland JJ, EvansAC, 2008: Differential expression of signal transductionfactors in ovarian follicle development; a role for betaglycanFIBP in granulosa cells in cattle. Physiol Genomics 33, 193–204.Franks S, McCarthy MI, Hardy K, 2006: Development ofpolycystic ovary syndrome: involvement of genetic andenvironmental factors. Int J Androl 29, 278–285.Fraser HM, Duncan WC, 2005: Vascular morphogenesis in theprimate ovary. Angiogenesis 8, 101–116.Gastal EL, Bergfelt DR, Nogueira GP, Gastal MO, GintherOJ, 1999: Role of luteinizing hormone in follicle deviationbased on manipulating progesterone concentrations inmares. Biol Reprod 61, 1492–1498.Gibbons JR, Wiltbank MC, Ginther OJ, 1997: Functionalinterrelationships between follicles greater than 4 mm andthe follicle-stimulating hormone surge in heifers. BiolReprod 57, 1066–1073.Ginther OJ, 1992: Reproductive Biology of the Mare, Basicand Applied Aspects, 2nd edn. Equiservices Publishing,Cross Plains, WI, USA.Ginther OJ, 1993: Major and minor follicular waves during theequine estrous-cycle. J Equ Vet Sci 13, 18–25.Ginther OJ, Bergfelt DR, 1992: Associations between FSHconcentrations and major and minor follicular waves inpregnant mares. Theriogenology 38, 807–821.Ginther OJ, Knopf L, Kastelic JP, 1989: Temporal associationsamong ovarian events in cattle during oestrous cycleswith two or three follicular waves. J Reprod Fertil 87, 223–230.Ginther OJ, Kot K, Kulick LJ, Martin S, Wiltbank MC, 1996:Relationships between FSH and ovarian follicular wavesduring the last six months of pregnancy in cattle. J ReprodFert 108, 271–279.Ginther OJ, Kot K, Kulick LJ, Wiltbank MC, 1997: Emergenceand deviation of follicles during the development offollicular waves in cattle. Theriogenology 48, 75–87.Ginther OJ, Bergfelt DR, Kulick LJ, Kot K, 1999: Selection ofthe dominant follicle in cattle: establishment of follicledeviation in less than 8 hours through depression of FSHconcentrations. Theriogenology 52, 1079–1093.Ginther OJ, Bergfelt DR, Kulick LJ, Kot K, 2000: Selection ofthe dominant follicle in cattle: role of estradiol. Biol Reprod63, 383–389.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Dominant Follicle Selection in Cows, Mares and Women 55Ginther OJ, Beg MA, Bergfelt DR, Donadeu FX, Kot K,2001a: Follicle selection in monovular species. Biol Reprod65, 638–647.Ginther OJ, Bergfelt DR, Beg MA, Kot K, 2001b: Follicleselection in cattle: relationships among growth rate, diameterranking, and capacity for dominance. Biol Reprod 65,345–350.Ginther OJ, Gastal EL, Gastal MO, Bergfelt DR, BaerwaldAR, Pierson RA, 2004a: Comparative study of the dynamicsof follicular waves in mares and women. Biol Reprod 71,1195–1201.Ginther OJ, Bergfelt DR, Beg MA, Meira C, Kot K, 2004b: Invivo effects of an intrafollicular injection of insulin-likegrowth factor 1 on the mechanism of follicle deviation inheifers and mares. Biol Reprod 70, 99–105.Ginther OJ, Gastal EL, Gastal MO, Beg MA, 2004c: Criticalrole of insulin-like growth factor system in follicle selectionand dominance in mares. Biol Reprod 70, 1374–1379.Ginther OJ, Beg MA, Gastal EL, Gastal MO, Baerwald AR,Pierson RA, 2005: Systemic concentrations of hormonesduring the development of follicular waves in mares andwomen: a comparative study. Reproduction 130, 379–388.Glister C, Tannetta DS, Groome NP, Knight PG, 2001:Interactions between follicle-stimulating hormone andgrowth factors in modulating secretion of steroids andinhibin-related peptides by nonluteinized bovine granulosacells. Biol Reprod 65, 1020–1028.Gong JG, Bramley T, Webb R, 1991: The effect ofrecombinant bovine somatotropin on ovarian function inheifers – follicular populations and peripheral hormones.Biol Reprod 45, 941–949.Gong JG, Campbell BK, Bramley TA, Gutierrez CG, PetersAR, Webb R, 1996: Suppression in the secretion of folliclestimulatinghormone and luteinizing hormone, and ovarianfollicle development in heifers continuously infused with agonadotropin-releasing hormone agonist. Biol Reprod 55,68–74.Gore MA, Nayudu PL, Vlaisavljevic V, 1997: Attainingdominance in vivo: distinguishing dominant from challengerfollicles in humans. Hum Reprod 12, 2741–2747.Hirakawa T, Minegishi T, abe K, Kishi H, Ibuki Y, MiyamotoK, 1999: A role of insulin-like growth factor I in luteinizinghormone receptor expression in granulosa cells. Endocrinology140, 4965–4971.Hohmann FP, Laven JSE, de Jong FH, Eijkemans MJC,Fauser BCJM, 2001: Low-dose exogenous FSH initiatedduring the early, mid or late follicular phase can inducemultiple dominant follicle development. Hum Reprod 16,846–854.Hunter MG, Robinson RS, Mann GE, Webb R, 2004:Endocrine and paracrine control of follicular developmentand ovulation rate in farm species. Anim Reprod Sci 82–83,461–477.Ireland JJ, Roche JF, 1983: Development of non-ovulatoryantral follicles in heifers: changes in steroids in follicularfluid and receptors for gonadotropins. Endocrinology 112,150–156.Ireland JJ, Mihm M, Austin EA, Diskin MG, Roche JF, 2000:Dominant follicle turnover in cattle: key concepts, studiesand terms. J Dairy Sci 83, 1648–1658.Ireland JJ, Ward F, Jimenez-Krassel F, Ireland JL, Smith GW,Lonergan P, Evans AC, 2007a: Follicle numbers are highlyrepeatable within animal but inversely correlated with FSHconcentrations and the proportion of good quality embryosafter superstimulation in cattle. Hum Reprod 22, 1687–1695.Ireland JLH, Jimenez-Krassel F, Smith G, Lonergan P, EvansA, Ireland J, 2007b: Variation in number of follicles growingduring follicular waves reflects size of the ovarian reserve incattle. Biol Reprod, Special Issue 128–129 (Abstract 218).Kaneko H, Noguchi J, Kikuchi K, Todoroki J, Hasegawa Y,2002: Alterations in peripheral concentrations of inhibin Ain cattle studied using a time-resolved immunofluorometricassay: relationship with estradiol and follicle-stimulatinghormone in various reproductive conditions. Biol Reprod67, 38–45.Knight PG, Glister C, 2006: TGF-b superfamily members andovarian follicle development. Reproduction 132, 191–206.Kulick LJ, Bergfelt DR, Kot K, Ginther OJ, 2001: Follicleselection in cattle: follicle deviation and codominance withinsequential waves. Biol Reprod 65, 839–846.Laven JSE, Fauser BCJM, 2004: Inhibins and adult ovarianfunction. Mol Cell Endocrinol 225, 37–44.Mihm M, Bleach ECL, 2003: Endocrine regulation of ovarianantral follicle development in cattle. Anim Reprod Sci 78,217–237.Mihm M, Good TEM, Ireland JLH, Ireland JJ, Knight PG,Roche JF, 1997: The decline in serum FSH concentrationsalters key intrafollicular growth factors involved in selectionof the dominant follicle in heifers. Biol Reprod 57, 1328–1337.Mihm M, Curran N, Hyttel P, Knight PG, Boland MP, RocheJF, 1999: Dominant follicle persistence in beef heifers resultsin alterations in follicular fluid estradiol and inhibin anddeviations in oocyte maturation. J Reprod Fertil 116, 293–304.Mihm M, Austin EJ, Good TEM, Ireland JLH, Knight PG,Roche JF, Ireland JJ, 2000: Identification of potentialintrafollicular markers involved in selection of the dominantfollicle in heifers. Biol Reprod 63, 811–819.Mihm M, Khan A, Bleach E, Knight PG, 2001: Health andatresia in first wave follicles is reflected in changing serumdimeric inhibin-A concentrations in dairy heifers. Mol CellEndocrinol 180, 200 (Abstract).Mihm M, Baker PJ, Ireland JLH, Smith GW, Coussens PM,Evans ACO, Ireland JJ, 2006: Molecular evidence thatgrowth of dominant follicles involves a reduction in folliclestimulating hormone – dependence and an increase inluteinizing hormone – dependence in cattle. Biol Reprod74, 1051–1059.Mihm M, Baker PJ, Fleming LM, Monteiro AM, O’ShaughnessyPJ, 2008: Differentiation of the bovine dominantfollicle from the cohort upregulates mRNA expression fornew tissue development genes. Reproduction 135, 253–265.Monget P, Fabre S, Mulsant P, Lecerf F, Elsen JM, MazerbourgS, Pisselet C, Monniaux D, 2002: Regulation ofovarian folliculogenesis by IGF and BMP system indomestic animals. Domest Anim Endocrinol 23, 139–154.Muttukrishna S, Fowler PA, Groome NP, Mitchell GG,Robertson WR, Knight PG, 1994: Serum concentrations ofdimeric inhibin during the spontaneous human menstrualcycleand after treatment with exogenous gonadotropin.Hum Reprod 9, 1634–1642.Muttukrishna S, Child T, Lockwood GM, Groome NP,Barlow DH, Ledger WL, 2000: Serum concentrations ofdimeric inhibins, activin A, gonadotrophins and ovariansteroids during the menstrual cycle in older women. HumReprod 15, 549–556.O’Rourke M, Diskin MG, Sreenan JM, Roche JF, 2000: Theeffect of dose and route of oestradiol benzoate administrationon plasma concentrations of oestradiol and FSH inlong term ovariectomised heifers. Anim Reprod Sci 59,1–12.Parker KI, Robertson DM, Groome NP, Macmillan KL,2003: Plasma concentrations of inhibin A and folliclestimulatinghormone differ between cows with two or threeÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

56 M Mihm and ACO Evanswaves of ovarian follicular development in a single estrouscycle. Biol Reprod 68, 822–828.Prendiville DJ, Enright WJ, Crowe MA, Finnerty M, HynesN, Roche JF, 1995: Immunization of heifers againstgonadotropin releasing hormone: antibody titres, ovarianfunction, body growth and carcass characteristics. J AnimSci 73, 2382–2389.Rivera GM, Fortune JE, 2003: Selection of the dominantfollicle and insulin-like growth factor (IGF)-binding proteins:evidence that pregnancy-associated plasma protein Acontributes to proteolysis of IGF-binding protein 5 inbovine follicular fluid. Endocrinology 144, 437–446.Rivera GM, Chandrasekher YA, Giudice LC, Evans ACO,Fortune JE, 2001: A potential role for insulin-like growthfactor binding protein-4 proteolysis in the establishment ofovarian follicular dominance in cattle. Biol Reprod 65, 102–111.Roberts AJ, Echternkamp SE, 2003: Insulin-like growth factorbinding proteins in granulosa and thecal cells from bovineovarian follicles at different stages of development. J AnimSci 81, 2826–2839.Roche JF, Mihm M, Diskin M, Ireland JJ, 1998: A reviewof regulation of follicle growth in cattle. J Anim Sci76(Suppl. 3), 16–29.Rozell TG, Monteiro AM, Baker PJ, Mihm M, 2005: Evidencefor differential expression of FSH receptor exons 10 and 11during follicular wave development in the cow. Biol Reprod,Special Issue, 179 (Abstract M445).Ryan KE, Casey SM, Canty MJ, Crowe MA, Martin F, EvansACO, 2007: Akt and Erk signal transduction pathways areearly markers of differentiation in dominant and subordinateovarian follicles in cattle. Reproduction 133, 617–626.San Roman GA, Magoffin DA, 1993: Insulin-like growthfactor-binding proteins in healthy and atretic follicles duringnatural menstrual cycles. J Clin Endocrinol Metab 76, 625–632.Savio JD, Keenan L, Boland MP, Roche JF, 1988: Pattern ofgrowth of dominant follicles during the oestrous cycle ofheifers. J Reprod Fertil 83, 663–671.Schipper I, Hop WCJ, Fauser BCJM, 1998: The folliclestimulatinghormone (FSH) threshold ⁄ window conceptexamined by different interventions with exogenous FSHduring the follicular phase of the normal menstrual cycle:duration, rather than magnitude, of FSH increase affectsfollicle development. J Clin Endocrinol Metab 83, 1292–1298.Schneyer AL, Fujiwara T, Fox J, Welt CK, Adams J,Messerlian GM, Taylor AE, 2000: Dynamic changes in theintrafollicular inhibin ⁄ activin ⁄ follistatin axis during humanfollicular development: relationship to circulating hormoneconcentrations. J Clin Endocrinol Metab 85, 3319–3330.Schuller AGP, Lindenberghkortleve DJ, Pache TD, ZwarthoffEC, Fauser BCJM, Drop SLS, 1993: Insulin-like growthfactorbinding protein-2, 28-kDa and 24-kDa insulin-likegrowth-factor binding-protein levels are decreased in fluid ofdominant follicles, obtained from normal and polycysticovaries. Regul Pept 48, 157–163.Sisco B, Hagemann LJ, Shelling AN, Pfeffer PL, 2003:Isolation of genes differentially expressed in dominant andsubordinate bovine follicles. Endocrinology 144, 3904–3913.Spicer LJ, Aad PY, 2007: Insulin-like growth factor (IGF) 2stimulates steroidogenesis and mitosis of bovine granulosacells through the IGF1 receptor: role of follicle-stimulatinghormone and IGF2 receptor. Biol Reprod 77, 18–27.Stagg K, Spicer LJ, Diskin MG, Sreenan JM, Roche JF, 1998:Effect of calf isolation on follicular wave dynamics, gonadotropinand metabolic hormone changes, and interval tofirst ovulation in beef cows fed either of two energy levelspostpartum. Biol Reprod 59, 777–783.Stewart RE, Spicer LJ, Hamilton TD, Keefer BE, Dawson LJ,Morgan GL, Echternkamp SE, 1996: Levels of insulin-likegrowth factor (IGF) binding proteins, luteinizing hormoneand IGF-I receptors, and steroids in dominant folliclesduring the first follicular wave in cattle exhibiting regularestrous cycles. Endocrinology 137, 2842–2850.Sunderland SJ, Crowe MA, Boland MP, Roche JF, Ireland JJ,1994: Selection, dominance and atresia of follicles during theoestrous cycle of heifers. J Reprod Fertil 101, 547–555.Sunderland SJ, Knight PG, Boland MP, Roche JF, Ireland JJ,1996: Alterations in intrafollicular levels of different molecularmass forms of inhibin during development of follicularandluteal-phase dominant follicles in heifers. Biol Reprod54, 453–462.Van Santbrink EJP, Hop WC, Van Dessel TJHM, De JongFH, Fauser BCJM, 1995: Decremental follicle-stimulatinghormoneand dominant follicle development during thenormal menstrual-cycle. Fertil Steril 64, 37–43.Watson ED, Al-zi’abi MO, 2002: Characterization of morphologyand angiogenesis in follicles of mares during springtransition and the breeding season. Reproduction 124, 227–234.Watson ED, Thomassen R, Steele M, Heald M, Leask R,Groome NP, Riley SC, 2002: Concentrations of inhibin,progesterone and oestradiol in fluid from dominant andsubordinate follicles from mares during spring transitionand the breeding season. Anim Reprod Sci 74, 55–67.Wiltbank MC, Fricke PM, Sangsritavong S, Sartori R,Ginther OJ, 2000: Mechanisms that prevent and producedouble ovulations in dairy cattle. J Dairy Sci 83, 2998–3007.Zeleznik AJ, 2001: Follicle selection in primates: ‘many arecalled but few are chosen’. Biol Reprod 65, 655–659.Zielak AE, Forde N, Park SD, Doohan F, Coussens PM,Smith GW, Ireland JJ, Lonergan P, Evans AC, 2007:Identification of novel genes associated with dominant follicledevelopment in cattle. Reprod Fertil Dev 19, 967–975.Author’s address (for correspondence): M Mihm, Division of CellSciences, Faculty of Veterinary Medicine, University of Glasgow,Bearsden Road, Glasgow, G61 1QH, UK. E-mail: m.mihm@vet.gla.ac.ukConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 57–65 (2008); doi: 10.1111/j.1439-0531.2008.01143.xISSN 0936-6768Regulation of Luteal Function and Corpus Luteum Regression in Cows: HormonalControl, Immune Mechanisms and Intercellular CommunicationDJ Skarzynski 1 , G Ferreira-Dias 2 and K Okuda 31 Department of Reproductive Immunology, Institute of Animal Reproduction and Food Research of PAS, Olsztyn, Poland; 2 C.I.I.S.A., Facultyof Veterinary Medicine, TU Lisbon, Portugal; 3 Laboratory of Reproductive Endocrinology, Graduate School of Natural Science andTechnology, Okayama University, Okayama, JapanContentsThe main function of the corpus luteum (CL) is production ofprogesterone (P4). Adequate luteal function to secrete P4 iscrucial for determining the physiological duration of theoestrous cycle and for achieving a successful pregnancy. Thebovine CL grows very fast and regresses within a few days atluteolysis. Mechanisms controlling development and secretoryfunction of the bovine CL may involve many factors that areproduced both within and outside the CL. Some of theseregulators seem to be prostaglandins (PGs), oxytocin, growthand adrenergic factors. Moreover, there is evidence that P4acts within the CL as an autocrine or paracrine regulator. Eachof these factors may act on the CL independently or maymodify the actions of others. Although uterine PGF 2a isknown to be a principal luteolytic factor, its direct action onthe CL is mediated by local factors: cytokines, endothelin-1,nitric oxide. The changes in ovarian blood flow have also beensuggested to have some role in regulation of CL development,maintenance and regression.IntroductionThe corpus luteum (CL) is a transient ovarian organthat is established by cells of a follicle after ovulation.The mammalian CL is composed of a heterogeneousmixture of cell types. There are at least two types ofsteroidogenic cells, large and small luteal cells, whichoriginate from the granulosa and thecal cells of thefollicle ruptured at ovulation, respectively. The CLconsists of not only steroidogenic luteal cells but alsonon-steroidogenic cells, i.e. vascular endothelial cells,fibroblasts, pericytes and immune cells such as lymphocytes,leucocytes and macrophages (Lei et al. 1991).Macrophages and endothelial cells infiltrate into thenewly formed CL concomitant with vascular angiogenesis(Reynolds and Redmer 1999). Most of the cellsproliferating during growth of the CL are endothelialcells under hypoxic conditions, although leucocytes havebeen shown to proliferate in CL as well (Reynolds andRedmer 1999; Towson et al. 2002). There is evidencethat steroids, protein hormones, growth factors, eicosanoidsand cytokines produced by the component cells ofthe CL play roles in establishing CL (Reynolds andRedmer 1999; Berisha and Schams 2005).Progesterone (P4), the primary product of the CL, isrequired for establishment and maintenance of pregnancy.P4 concentration in blood increases during thedeveloping luteal stage. The CL continues to secrete ahigh level of P4 until late luteal stage, and then its abilityrapidly decreases in the regressing luteal stage. Indomestic animals, luteinizing hormone (LH) releasedin a pulsatile fashion from the anterior pituitary is oneof the most potent regulators of synthesis and secretionof P4 in the CL (Niswender et al. 2007). Luteinizinghormone stimulates P4 production in small luteal cellsvia the LH receptor. In addition to pituitary hormones,intraluteal substances produced by the component cellsof the CL play important roles in regulating P4production during the luteal stages in bovine CL.Moreover, the changes in P4 concentration throughoutthe luteal phase are closely associated with blood flow aswell as steroidogenic capacity of luteal cells. Duringluteal regression, the decrease in P4 is concerned withreduced blood flow and loss of LH receptors in the CL(Niswender et al. 1976).When animals do not become pregnant, regression ofthe CL, termed luteolysis, is essential for normalcyclicity as it allows development of a new ovulatoryfollicle. In the cow, luteolysis is initiated by prostaglandin(PG)F 2a released from the uterus at the late lutealstage. Luteolysis consists of two phases, functionalluteolysis and structural luteolysis in mammals (McCrackenet al. 1999). A rapid functional regression of CLis characterized by a decrease of P4 production,followed by a phase of structural regression (McCrackenet al. 1999). During structural luteolysis, CL cellsundergo apoptosis, a process that has been describedby morphological and biochemical parameters in ruminants(Juengel et al. 1993; Rueda et al. 1995, 1997). It isgenerally accepted that the immune response playscentral roles in apoptosis of several tissues and celltypes (Nagata 1997). Recently, intraluteal mediatorsincluding cytokines and nitric oxide (NO) are thought toplay some roles as pro-apoptotic and anti-apoptoticfactors in the bovine CL, respectively (Petroff et al.2001; Skarzynski et al. 2005).This review focuses on mediators for regulatingbovine CL function throughout the oestrous cycle.Furthermore, some of our and others’ data on mediators(EDN1, cytokines, NO) of uterine PGF 2a action onbovine CL as well cell-to-cell contact and interactionsduring luteolysis have also been reviewed. Moreover,intra-luteal mechanisms controlling sensitivity of the CLto extragonadal PGF 2a are discussed.Development and Maintenance of the CorpusLuteumThe mechanism controlling development, maintenanceand secretory function of the CL may involve factorsthat are produced both within the CL and outside theovary (Fig. 1). Some of these regulators, that act asÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

58 DJ Skarzynski, G Ferreira-Dias and K Okuda(+)Prostaglandins(PGE 2, PGI 2, PGF 2a )LH/gonadotropins(+)P4Endothelial/luteal cellsdifferentiation and/orproliferationautocrine and ⁄ or paracrine factors, seem to be PGs andother arachidonic acid metabolites [PGE 2 , PGF 2a ,leucotrienes (LT)], neuropeptides [noradrenaline (NA)],peptide hormones [i.e. oxytocin (OT)], NO, growthfactors and hormones [vascular endothelial growthfactor (VEGF), fibroblastic growth factors (FGF),epidermal growth factor (EGF), growth hormone(GH), prolactin (PRL)] and steroids [P4 and 17boestradiol(E 2 )].Corpus luteum developmentThe physiological processes of CL growth and formationmight be regulated by many different factors, stillnot completely understood (Fig. 1) (Reynolds andRedmer 1999; Acosta and Miyamoto 2004; Schamsand Berisha 2004; Meidan et al. 2005; Ferreira-Dias andSkarzynski 2008). After ovulation, as the CL formsfrom the wall of the ruptured follicle, it grows andvascularizes rapidly. In fact, the rates of tissue growthand angiogenesis in the CL rival those of even thefastest growing tumours. The CL is a complex tissuecomposed of paranchymal (small and large steroidogenic)and non-parenchymal (fibroblast, vascularsmooth muscle, pericytes and endothelial) cells (Reynoldsand Redmer 1999; Lei et al. 2000; Towson et al.2002). In agreement with data from other tissues,VEGF seem to be a major angiogenic factor responsiblefor vascularization of the developing CL. Recent datasuggest that luteal expression of VEGF occurs primarilyin steroidogenic cells (granulose-lutein cells) and less inendothelial cells (specific perivascular cells, includingarteriolar smooth muscle and capillary pericytes), and isregulated primarily by oxygen levels (Berisha andSchams 2005; Ferreira-Dias and Skarzynski 2008).Soon after ovulation, pericytes derived from the thecalcompartment appear to be the first vascular cells toinvade the developing luteal parenchyma. The granulosa-derivedcells produce a factor that stimulatespericytes migration. Moreover, NO, which is a potentvasodilator and stimulates VEGF production andangiogenesis, is produced by endothelial cells of lutealarterioles and capillaries, often in association withexpression of VEGF by luteal perivascular cells (Berishaand Schams 2005).(+)Growth factors, cytokines(VEGF, FGF, IGF, EGF,TNFa)Fig. 1. Hypothetical model of the regulation of the CL development(see text for the details; adapted from Ferreira-Dias and Skarzynski2008)Basic FGF mRNA and protein and its receptors arepresent in the CL and may stimulate proliferation ofluteal endothelial cells (Reynolds and Redmer 1999;Berisha and Schams 2005). The IGF system plays a rolein CL development and may indirectly affect angiogenesisin the early CL by stimulating VEGF production(Schams and Berisha 2004; Berisha and Schams 2005).Thus, several growth factors act as auto-paracrineregulators affecting proliferation and differentiation ofthe developing CL cells (Reynolds and Redmer 1999;Acosta and Miyamoto 2004; Berisha and Schams 2005).Moreover, the luteotropic paracrine and ⁄ or autocrineeffects of OT, NO, PGF 2a and other metabolites ofarachidonic acid (PGE 2 , PGI 2 , leucotrienes) suggest thateikosanoids assume different and perhaps opposingroles depending on the cellular and hormonal milieu(Skarzynski et al. 2000, 2001; Weems et al. 2004; Meidanet al. 2005). Prostaglandins and leucotrienes mayregulate cell proliferation and stimulate P4 productionregulating ⁄ stimulating CL development and formation.Just after ovulation, the basement membrane breaksdown, and blood vessels from the theca interna invadethe avascular granulose-lutein cell layers (Acosta andMiyamoto 2004; Berisha and Schams 2005). Thesephenomena are thought to occur under hypoxic conditions.Thus, we have proposed a model for the initialprocess of luteal vascularization in which hypoxia playsa major role. In this model, a paracrine loop existsbetween the vascular endothelial cells which produceNO, and granulose-lutein cells which mainly produceVEGF, to ensure coordinated regulation of lutealvasodilation and angiogenesis. During this decade,hypoxia-inducible factor 1 (HIF1) has been recognizedto have critical roles in angiogenesis via transcriptionalregulation of angiogenic factors, such as VEGF. Ourrecent study indicates that HIF1 is essential for theVEGF-induced angiogenesis during luteal development,and suggests that formation of luteal vasculature incows is regulated by hypoxic condition following folliclerupture (Nishimura R and Okuda K, unpublished data).Maintenance of the corpus luteum: mandatory role ofprogesteroneIn cows, LH and GH are the primary hormones whichsupport the development and function of the CL (Hanseland Dowd 1986; Niswender et al. 2007). However, thelocal angiogenic growth factors (VEGD, EGF, FGF),PGs and peptide hormones (OT) should be considerednot only the potent regulators of luteal development, butalso important factors regulating P4 secretion andlifespan of CL (Skarzynski et al. 2001; Berisha andSchams 2005; Meidan et al. 2005; Fig. 1). Moreover, P4also has an effect on function of the bovine early and midCL in an autocrine and paracrine fashion (Skarzynskiand Okuda 1999; Duras et al. 2005). In order to removethe influence of P4 produced by cultured bovine lutealcells, we treated cells with a specific P4 antagonist(onapristone; Skarzynski and Okuda 1999; Okuda et al.2004). In early luteal cells, secretions of P4 OT, PGF 2aand PGE 2 were reduced by onapristone. Moreover, theP4 antagonist inhibited OT secretion by mid-cycle lutealcells, although it stimulated PGF 2a secretion (SkarzynskiÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Regulation of Luteal Function 59and Okuda 1999). These findings suggest that P4stimulates P4, OT and PGs secretion by early CL, butat mid-cycle CL, P4 inhibits PGF 2a secretion. Pate (1988,1996) also reported that P4 regulates PGF 2a secretion inthe mid-cycle CL, but not in the late CL. However,during pregnancy P4 did not affect its own secretion bybovine (Weems et al. 2002) and ovine luteal slices (Kimet al. 2001). Therefore, P4 may affect the secretoryfunction of the bovine CL in a stage-dependent fashionand P4 effects may depend on cell-to-cell contact andcomposition (Pate 1996; Skarzynski and Okuda 1999;Skarzynski et al. 2001; Weems et al. 2002).Recent studies have demonstrated that intra-luteal P4is one of the most important factors supporting maintenanceof the CL (Fig. 2). It has been shown that P4may suppress apoptosis in bovine luteal cells throughthe inhibition of Fas and caspase-3 mRNA expressionand inhibition of caspase-3 activation (Rueda et al.2000; Okuda et al. 2004). In addition to our previousfinding that P4 may stimulate its own production bybovine CL (Skarzynski and Okuda 1999), we havedemonstrated that blockage of the autocrine and ⁄ orparacrine action of P4 (by a specific antagonist –onapristone) reduced viability of the luteal cells(Fig. 2; Okuda et al. 2004). In conclusion, the overallresults showed that P4 may inhibit Fas L-mediatedapoptosis via a decrease of Fas and caspase-3 expressionand caspase-3 activity by bovine luteal cells. Themechanisms or signalling pathways by which P4 exertsits action on the bovine CL are unknown. P4 seems toact on the secretory function of bovine CL (Bogackiet al. 2000) non-genomically, i.e. through membranebinding sites (Rae et al. 1998) rather than throughnuclear mediation. In support of this idea, Cannon et al.(2003) have shown that P4 suppresses luteal cell-stimulatedT lymphocyte proliferation. However,classical-genomic P4 receptors were not expressed in Tlymphocytes showing that P4 may act non-genomicallyon the cells (Sugino et al. 1997). On the other hand, ithas been suggested that P4 plays an active role ininhibition of luteal regression by a direct effect onclassical P4 genomic binding sites in bovine luteal cells(Rueda et al. 2000). Moreover, P4 promotes survival ofbovine luteal cells by inhibiting apoptosis via theglucocorticoid receptors (GR; Rueda et al. 2000).All these findings confirmed the previous suppositionthat P4 is an universal autocrine and paracrine regulatorof luteal function in cows and plays one or more role(s)in the regulation of development and maintenance of thebovine CL (Pate 1996; Rae et al. 1998; Skarzynski andOkuda 1999; Rueda et al. 2000; Okuda et al. 2004).Thus, it could be assumed that intra-luteal P4 isimplicated in a survival pathway in the CL by stimulationPGs, OT and its own production as well as by theinhibition of apoptosis of the bovine CL.Intraluteal Factors Mediate Luteolytic Actionof PGF 2aIn the absence of an embryo(s) in the uterus, the processof CL regression begins on days 17–19 of the oestrouscycle in cows (McCracken et al. 1999). This process ischaracterized not only by functional but also structural(a)Cell viability (% of control)0(b)500Progesterone (ng/2x10 5 cells)(c)Caspase-3 activity(µmol pNA/min/2x10 5 cells)12510075502540030020010003020100aaaaabchanges that occur in the luteal (steroidogenic) and nonluteal(accessory) cells of the CL (Juengel et al. 1993;Meidan et al. 1999). In most species, uterine PGF 2a isknown to be a principal luteolytic factor, but its directaction within the CL is still debated (Skarzynski andOkuda 1999; Pate 2003; Schams and Berisha 2004;Meidan et al. 2005; Skarzynski et al. 2005). The actionof PGF 2a on bovine CL is mediated by local factors:EDN1, cytokines and NO (Fig. 3).Endothelin-1 and vasoactive peptidesA pivotal role for the endothelial cell product – EDN1in PGF 2a -induced luteal regression in cows has been welldocumented by Meidan’s group (Girsh et al. 1996;bbControl Fas L OP Fas L/OP TNF a/IFNg(6 nmol/l) (100 µmol/M) (by 3 nmol/l)Fig. 2. Effects of Fas L (6 nM; 05-351; Upstate Biotechnology, LakePlacid, NY, USA) and a specific progesterone antagonist (OP;100 lM; Schering AG, Berlin, Germany; #ZK98.299) on cell viability(a) progesterone secretion (b) and activity of caspase-3 (c) in luteal cellsobtained from the cows at the mid-luteal phase of the oestrous cycle(days 8–12 of the cycle) and cultured in medium supplemented with0.1% of BSA (#735078; Roche Diagnostics GmbH, Mannheim,Germany). Cytokines (TNF-a and INFc; both 3 nM; gift of DainipponSumitomo Pharma Co., Ltd, Osaka, Japan) were added as apositive control (adapted from Okuda et al. 2004). The data are shownas the mean ± SEM of values obtained in four separate experimentseach performed in triplicatebcccdbcÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

60 DJ Skarzynski, G Ferreira-Dias and K Okuda(+)(+)(+)EDN1(–)PGF 2aNO(–)(–)P4Apoptosis of steroidogenic andendothelial cells of the CLCytokines(TNFa,FAS-L,IFNg)Fig. 3. Hypothetical model of the structural and functional regressionof the CL (see text for the details; adapted from Ferreira-Dias andSkarzynski 2008)Mamluk et al. 1999; Meidan et al. 1999, 2005; Levyet al. 2000). EDN1 inhibited P4 secretion, in a dosedependent manner via selective EDN1 bindings sites(EBN A ; Girsh et al. 1996). The expression of membersof the EDN1 system (EDN1, EDN converting enzymes,and EDN A and EDN B receptors) increases during lutealregression (Ohtani et al. 1998; Klipper et al. 2004;Rosiansky-Sultan et al. 2006). Moreover, PGF 2a upregulatesEDN1 and EDN A expression within the CL(Mamluk et al. 1999). Besides EDN1, other vasoactivepeptides [i.e. angiothensin II (ANG-II), atrial natriureticpeptide (ANP)] are considered important factors inmediating PGF 2a luteolytic action (Schams and Berisha2004; Berisha and Schams 2005). These vasoactivepeptides decreased blood flow and triggered the luteolyticcascade and consequently inhibited P4 secretion(Kobayashi et al. 2002; Shirasuna et al. 2004). However,Flores with colleagues showed that the EDN1 system(EDN1, EDN A and EDN converting enzymes) exists inthe bovine CL through the oestrous cycle and PGF 2aincreased EDN1 mRNA expression in vivo only at 10 hafter treatment (Wright et al. 2001). In fact, Schamset al. (2003) showed that up-regulation of EDN1 andANG-2 occurred mainly during structural CL regression.Therefore, it has been suggested that EDN1 isinvolved in the process of structural CL regression bypromoting leucocyte migration and stimulating macrophagesto release cytokines [i.e. tumour necrosis factor a(TNF-a), interferon c (IFN-c); Meidan et al. 1999,2005].CytokinesImmune cells infiltrating the bovine CL play a centralrole in structural luteolysis of both steroidogenic andendothelial CL cells (Friedman et al. 2000; Pate andLandis Keyes 2001; Pate 2003). The number of leucocytes(i.e. T lymphocytes, macrophages) increased at thetime of structural luteolysis (Penny et al. 1999; Towsonet al. 2002). Shaw and Britt (1995) using a CL microdialysissystem showed that TNF-a is released duringspontaneous and PGF 2a -induced luteolysis in cows.Then, it was shown that the mRNAs for TNF-a and itsspecific receptors (TNFR type-I) are present in thebovine CL during luteolysis (Sakumoto et al. 2000;Neuvians et al. 2004). Tumour necrosis factor a incombination with IFN-c reduced P4 production andinduced apoptosis and PGF 2a production by luteal cellsin vitro (Sakumoto et al. 2000; Petroff et al. 2001;Korzekwa et al. 2006). Specific binding sites for TNFaare also present in endothelial cells derived frombovine CL (Okuda et al. 1999). Furthermore, TNF-ainduces EDN1 production by endothelial cells that maylead to structural regression of the CL (Okuda et al.1999; Friedman et al. 2000). Tumour necrosis factor aacting via TNFR type-I induces apoptotic death ofsteroidogenic (Petroff et al. 2001) and endothelial cells(Friedman et al. 2000) of bovine CL. However, some invitro studies indicated that TNF-a induces luteolysisonly in combination with INFc or other factors (i.e.EDN1; Petroff et al. 2001; Korzekwa et al. 2006).Therefore, we tested whether TNF-a acts as a luteolyticfactor in vivo, and whether it changes the lifespan ofbovine CL (Skarzynski et al. 2003a). Lower doses ofTNF-a increased PGF 2a and nitrite ⁄ nitrate (stablemetabolites of NO), decreased P4 level and consequentlyresulted in shortening of the oestrous cycle. Surprisingly,higher doses of TNF-a stimulated the synthesis of P4and PGE 2 and consequently resulted in prolongation ofthe oestrous cycle (Skarzynski et al. 2003a). However,inhibition of PG synthesis by indomethacin, a cyclooxygenase(COX; PTGS) inhibitor, injected into the aortaabdominalis, blocked the actions of TNF-a indicatingthat TNF-a acts mainly through mediation by arachidonicacid metabolites (Skarzynski et al. 2007). Inaddition to TNFR, other cytokine membrane receptors,second messengers, including calcium ions [Ca 2+ ] i andregulatory proteins are involved in apoptosis of steroidogenicand endothelial CL cells (Meidan et al. 1999;Petroff et al. 2001; Taniguchi et al. 2002). Fas ligand, amember of the TNF super family, primarily engages itsreceptors (Fas) to induce apoptosis (Taniguchi et al.2002; Okuda et al. 2004). The expression of Fas mRNAwas increased by IFN-c, and TNF-a augmented thestimulatory action of IFN-c on Fas expression (Taniguchiet al. 2002). Moreover, apoptotic bodies wereobserved in luteal cells treated with Fas L in thepresence of IFN-c and ⁄ or TNF-a, showing that leucocyte-derivedTNF-a and IFN-c play important roles inFas L-Fas-mediated luteal cell death in the bovine CL.Nitric oxideNADPH-d localization [a marker for nitric oxidesynthase (NOS)] and immunostaining of both isoformsof NOS [inducible (iNOS) and endothelial (eNOS)] weredetected in steroidogenic cells and in blood vessels of thebovine CL during the entire oestrous cycle withincreasing activity from the early to the late lutealphases (Skarzynski et al. 2003b). However, Rosiansky-Sultan et al. (2006) presented the highest level of eNOSand iNOS mRNA and protein expression in the earlyÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Regulation of Luteal Function 61bovine CL. These data suggest that NO produced bytwo NOS isoforms is involved in structural and functionalchanges that occur in the bovine CL through thewhole oestrous cycle. Luteolytic or luteotropic actionsof NO on the bovine CL is strictly dependent on thestage of CL, and cell interactions (cell-to-cell contact)and composition (Jaroszewski et al. 2003a; Klipperet al. 2004; Weems et al. 2004; Rosiansky-Sultan et al.2006). Nitric oxide donor (S-NAP) stimulated PGE 2secretion by steroidogenic CL cells in the early and midlutealphases (Skarzynski et al. 2000). During developmentand maintenance of the CL, PGE 2, which is bothluteotropic and antiluteolytic, stimulated NO production(Boiti et al. 2000). Nitric oxide donors and ET-1increased PGE 2 secretion by bovine CL slices in vitro ondays 13–14 of the cycle without any effect on P4secretion (Weems et al. 2004). Therefore, increased NOproduction during early stages of the cycle and pregnancyis likely to play a role in CL development andangiogenesis (Skarzynski et al. 2000; Weems et al. 2004;Vonnahme et al. 2005; Rosiansky-Sultan et al. 2006).In the late luteal phase PGF 2a might simulate a shearstress-like reaction of endothelial CL cells resulting incompensative NO release during the first steps ofluteolysis – up to 2–4 h after PGF 2a treatment (Li et al.2002; Skarzynski et al. 2003b; Acosta et al. 2007). Intralutealadministration of a NOS inhibitor (L-NAME)during the late luteal phase increased P4 secretion andprolonged the functional lifespan of bovine CL (Jaroszewskiand Hansel 2000; Sasahara et al. 2007). Whenan analogue of PGF 2a (aPGF 2a , cloprostenol) wasinjected on day 15 of the cycle in combination withL-NAME, the luteolytic effect of aPGF 2a was counteractedby the NOS inhibitor (Fig. 4; Jaroszewski et al.2003b; Skarzynski et al. 2003b). Nitric oxide has beenfound as the most potent inhibitor of P4 secretionin vitro (Skarzynski et al. 2003b; Korzekwa et al. 2004,2006) and in vivo (Sasahara et al. 2007). Moreover, aNO donor (Spermine NONOate) strongly stimulatedproduction of PGF 2a and LTC 4 by bovine CL bothin vitro and in vivo, showing that NO is involved in theprocess of luteal regression by bovine CL (Fig. 3;Progesterone (ng/ml)129630L-NAME/aPGF (n = 5)Saline/aPGF(n = 6)L-NAME/Saline (n = 4)Saline/saline (n = 4)aPGF0 3 6 9 12 15 18 0 4Day of the estrous cycle– EstrusEstrusFig. 4. The effect of 2 h infusion of saline or nitric oxide synthaseinhibitor-L-NAME (400 mg ⁄ h) and injection of saline or PGF 2aanalogue, cloprostenol (aPGF; 100 lg; Bioestrophan, Biowet, GorzowWielkopolski, Poland) at 30 min of infusion on progesterone concentrationsin peripheral blood plasma of heifers on day 15 of the oestrouscycle. Different subscript letters indicate significant differences(p

62 DJ Skarzynski, G Ferreira-Dias and K Okudato a luteolytic dose of aPGF 2a and suggested that thisincrease in blood flow is related to early luteolytic events(Acosta et al. 2002; Miyamoto et al. 2005). This increasein blood flow is thought to be caused by well-knownvasodilators, PGE 2 and ⁄ or NO (Miyamoto et al. 2005).In support, it has been recently shown that intralutealapplication of a NO donor drastically increase lutealblood flow (Sasahara et al. 2007).However, CL blood flow was reduced 8 h after oncePGF 2a i.m. injection in cows (Acosta et al. 2002) andwithin 4 h after two injections of PGF 2a (at 0 and 4 h)into the uterine lumen in ewes (Nett et al. 1976). Thus,PGF 2a may cause CL regression by depriving the gland ofnutrients, substrates for steroidogenesis, and luteotropicsupport (Nett et al. 1976; Niswender et al. 2000). Therefore,we hypothesized that hypoxic (low oxygen) conditionsin the CL induced by a decreased blood supply isrelated to the luteolytic cascade in cows. We examinedthe influence of hypoxia on the luteal P4 generatingsystem and luteal cell apoptosis using bovine luteal cellculture, and found that hypoxia decreased P4 synthesisby suppressing a steroidogenic enzyme P450scc activity(Nishimura et al. 2007) and also induced apoptotic celldeath by enhancing the expression of pro-apoptoticprotein BNIP3 and by activating caspase-3 (Nishimuraet al. 2008). Since P4 treatment under hypoxia decreasedluteal cell death, the induction of apoptosis by hypoxiaseems to be partially caused by a decrease in P4production (Nishimura et al. 2008). However, in ewesafter a single PGF 2a injection, levels of P4 decreased by12 h after PGF 2a , but total ovarian blood flow did notdecreased significantly until 30 h (Nett et al. 1976).Blood flow to ovine CL and luteal LH receptors didnot decrease until P4 declined in the peripheral blood andfunctional lutelysis was almost completed (Nett et al.1976; Niswender et al. 1976). Moreover, Watanebe et al.(2006) showed that P4 decreased even when blood flowremained high in bovine CL. Therefore, oxygen deficiencydue to decreased blood flow is not mandatory forthe induction of both functional and structural luteolysisin cows, but may play such a supporting role in the finalsteps of bovine CL regression.Resistance to PGF 2a Action During the LutealPhaseThe newly formed CL is resistant to exogenous PGF 2atreatment and action (Henricks et al. 1974; Beal et al.1980). Moreover, the sensitivity of CL to the luteolyticaction of extragonadal PGF 2a seems to increase progressivelytoward the end of the luteal phase (Skarzynskiet al. 1997; Pate 2003; Meidan et al. 2005). Since a singleclass of high-affinity PGF 2a receptors is present withinthe bovine CL by the second day after ovulation(Sakamoto et al. 1995; Wiltbank et al. 1995), neitherthe lack of responsiveness to PGF 2a in the early CL(Henricks et al. 1974; Beal et al. 1980) nor the lowersensitivity to PGF 2a in the mid-luteal CL (Skarzynskiet al. 1997) can be attributed to a deficiency of highaffinityPGF 2a receptors. Wiltbank et al. (1990) suggestedthat incomplete vascularization or incompletedifferentiation of degenerative mechanisms in the earlybovine CL are responsible for the lack of luteolyticcapacity. In recent years, Meidan et al. have explainedhow the products of endothelial cells are involved in theacquisition of luteolytic capacity by the bovine CL(Girsh et al. 1996; Mamluk et al. 1999; Meidan et al.1999, 2005; Levy et al. 2000). They showed that the lackof ET-1 synthesis and response to EDN-1 in the earlyand mid-luteal CL may make the CL unresponsive tothe luteolytic action of PGF 2a (Mamluk et al. 1999;Levy et al. 2000; Rosiansky-Sultan et al. 2006). In fact,PGF 2a upregulated the amount of mRNA encodingEDN-1 converting enzymes during the mid- and latelutealphase of the bovine CL and this action of PGF2awas cycle-phase specific, because it was not observed inthe early CL. Additionally, the steroidogenic cells of theCL may also possess autocrine mechanisms that protectbovine CL from premature luteolysis. PGF 2a was foundto induce mRNA for PTGS-2 on day 11 but not on day4 of the oestrous cycle, suggesting that such autoamplificationof luteal PGF 2a is also a key component ofluteolytic capacity by the bovine CL (Tsai and Wiltbank1998). Moreover, Silva et al. (2000) showed that upregulationof 15-hydroxyPG dehydrogenases in ovine CLduring early pregnancy may deactivate PGF 2a as amechanism for maternal recognition of pregnancy.We showed that the lack of response to PGF 2a by theearly bovine CL and the lower reaction of bovine CL toPGF 2a during the mid-luteal phase depended uponlocally produced PGs, OT and P4 (Skarzynski andOkuda 1999), NA and NO (Skarzynski et al. 2000). Thelack of response to PGF 2a in the early CL could be aconsequence of receptor desensitization and the biologicalwanes over time (Lohse 1993; Skarzynski et al.2001). Based on our findings, one could assume thatluteal neuropeptide, NO, OT, PGs and P4 are componentsof an auto ⁄ paracrine positive feedback cascade inbovine CL, and that they also play roles in regulatingthe function of PGF 2a receptors and the PGF 2a -intracellularcalcium ([Ca 2+ ] i )-protein kinase C cascade(Skarzynski and Okuda 1999; Skarzynski et al. 2000,2001). The auto ⁄ paracrine positive feedback loop can bea mechanism for protection against premature luteolysisduring the early and mid-luteal phase. Furthermore,each of the factors that directly affect PGs, OT and P4secretion by luteal cells may be indirectly involved inregulating function of PGF 2a receptors, and may beresponsible for the varying sensitivity of bovine lutealcells to PGF 2a . In contrast, locally produced NO maysensitize bovine CL to luteolytic action of PGF 2a(Skarzynski et al. 2000, 2003b). The amplified effectsof PGF 2a on luteal cells by pre-exposure to a NO donor(S-NAP) suggest that priming of bovine CL by NO isneeded for complete luteal regression (Skarzynski et al.2000). In support, NO up-regulated PGF 2a receptors inbovine CL endothelial cells (Rosiansky-Sultan et al.2006). Based on the overall findings, the differentresponse by CL to PGF 2a during the luteal phase andpregnancy should be explained not only by changes inthe concentration of PGF 2a receptors but also bychanges in their sensitivity as well as on the maturationof ET-1 and NO systems. Therefore, a number ofdifferent pathways may account for the lack of PGF 2a -induced luteolysis during the early luteal phase as well asduring early pregnancy.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Regulation of Luteal Function 63(+/–)Growth factorsEGFIGF-1(+)TGF(+/–)ET-1AVPOTConclusionThe mechanism controlling development, maintenanceand secretory function of the CL may involve factorsthat are produced both within the CL and outside theovary (Fig. 5). Some of these regulators seem to be PGsand other arachidonic acid metabolites (PGE 2 , PGF 2a ,LT), neuropeptides (NA), peptide hormones (OT, EDN-1), growth factors and hormones (VEGF, FGF, GH,PRL) and steroids (P4 and E2) that act as autocrineand ⁄ or paracrine factors. Although PGF 2a is known tobe the principal luteolytic factor, its action on the CL ismediated by other intra-ovarian factors: cytokines, NO,EDN-1. Nitric oxide, TNF-a in combination with IFN-creduced P4 secretion, increased luteal PGF 2a production,and induced apoptosis of the luteal cells.AcknowledgementsbFGF(?)(+/–)(+)(?)(+/–)PeptidesLHP4Neuropeptides(+)CytokinesIL-1IFNEicosanoidsTNFPGI2PGE2PGF2aCLThe work was supported by the grants-in-aid: from the Polish Ministryof Sciences and Higher Education (Scientific net; D.J. Skarzynski),international grant PORTUGALIA 78 ⁄ 2007 (G. Ferreira-Dias), andfrom the Japanese-Polish Joint Research Project under the agreementbetween Japan Society for Promotion of Sciences and Polish Academyof Sciences (K. Okuda).ReferencesAcosta TJ, Miyamoto A, 2004: Vascular control of ovarianfunction: ovulation, corpus luteum formation and regression.Anim Reprod Sci 82, 127–140.Acosta TJ, Yoshizawa N, Ohtani M, Miyamoto A, 2002:Local changes in blood flow within the early and midcyclecorpus luteum after PGF 2a injection in the cow. Biol Reprod66, 651–658.Acosta T, Bah M, Jaroszewski J, Skarzynski DJ, Okuda K,2007: Acute changes in circulating concentrations of oxygen,nitric oxide and progesterone. Biol Reprod 77 (Sp. Iss.),121–122.Beal WE, Milvae RA, Hansel W, 1980: Oestrous cycle lengthand plasma progesterone concentrations following administrationof PGF 2a early in the bovine oestrous cycle. JReprod Fertil 59, 393–396.Berisha B, Schams D, 2005: Ovarian function in ruminants.Domest Anim Endocrinol 29, 305–217.Bogacki M, Skarzynski DJ, Kotwica J, 2000: Non genomicaction of progesterone in luteal and endometrial epithelialcells in cattle. ESDAR Newslett 5, 23.(+)(–) (–)(–)(+)(+)(+)(+)Fig. 5. Schematic presentation of factors involved in the growth,maintenance and regression of the CL (see text for the details)(+)Boiti C, Zerani M, Zampini D, Gobbetti A, 2000: Nitric oxidesynthase activity and progesterone release by isolatedcorpora lutea of rabbits in the early and mid-luteal phasesof pseudopregnancy are modulated differently by PGE 2 andPGF 2a via AC and PKC. J Endocrinol 164, 179–186.Cannon MJ, Petroff MG, Pate JL, 2003: Effects of PGF 2a andprogesterone on the ability of bovine luteal cells to stimulateT lymphocyte proliferation. Biol Reprod 69, 695–700.Duras M, Brzosko E, Kotwica J, 2005: Influence of progesterone,pregnenolone and 17b-hydroxy-progesterone on thefunction of bovine luteal cells treated with luteinizinghormone, noradrenaline and PGE 2 . Pol J Vet Sci 8, 113–119.Ferreira-Dias GM, Skarzynski DJ, 2008: Some aspects ofregulation of luteal function and luteolysis in equine corporalutea. Pferdeheikunde 24, 10–14.Ford SP, Chenault JR, 1981: Blood flow to the corpus luteumbearingovary and ipsilateral uterine horn of cows duringoestrous cycle and early pregnancy. J Reprod Fertil 62, 555–562.Friedman A, Weiss S, Levy N, Meidan R, 2000: Role of TNFaand its type-1 receptor in luteal regression: induction ofprogrammed cell death in bovine corpus luteum-derivedendothelial cells. Biol Reprod 63, 1905–1915.Ginther OJ, Silva LA, Araujo RR, Beg MA, 2007: Temporalassociations among pulses of 13,14-dihydro-15-keto-PGF 2a ,luteal blood flow, and luteolysis in cattle. Biol Reprod 76,506–513.Girsh E., Milvae RA, Wang W, Meidan R, 1996: Effect ofendothelin-1 on bovine luteal cell function: role in PGF 2a -induced antisteroidogenic action. Endocrinology 137, 1306–1312.Hansel W, Dowd JP, 1986: New concepts of the control ofcorpus luteum function. J Reprod Fert 78, 755–768.Henricks DM, Long JT, Hill JR, 1974: The effect of PGF 2aduring various stages of the ooestrous cycle of beef heifers.J Reprod Fertil 41, 113–120.Jaroszewski JJ, Hansel W, 2000: Intraluteal administration ofa nitric oxide synthase blocker stimulates progesterone andoxytocin secretion and prolongs the life span of the bovinecorpus luteum. Proc Soc Exp Biol Med 224, 50–55.Jaroszewski JJ, Skarzynski DJ, Blair RM, Hansel W, 2003a:Influence of nitric oxide on the secretory function of thebovine corpus luteum: dependence on cell composition andcell-to-cell communication. Exp Biol Med 228, 741–748.Jaroszewski JJ, Skarzynski DJ, Hansel W, 2003b: Nitric oxideas a local mediator of prostaglandin F 2a -induced regressionin bovine corpus luteum: an in vivo study. Exp Biol Med228, 1057–1062.Juengel JL, Garverick HA, Johnson AL, Youngquist RS,Smith MF, 1993: Apoptosis during luteal regression incattle. Endocrinology 132, 249–254.Kim L, Weems YS, Bridges PJ, LeaMaster BR, Ching L,Vincent DL, Weems CW, 2001: Effects of indomethacin,luteinizing hormone, PGE 2 , trilostane, mifepristone, ethamoxytriphetolon secretion of PGE 2 , PGF 2a and progesteroneby ovine corpora lutea of pregnancy or the estrous cycle.Prostaglandins 63, 189–203.Klipper E, Gilboa T, Levy N, Kisliouk T, Soanel-Borowski K,Meidan R, 2004: Characterization of endothelin-1 and nitricoxide generating systems in corpus luteum-derived endothelialcells. Reproduction 128, 463–473.Kobayashi S, Acosta TJ, Ozawa T, Hayashi K, Berisha B,Ohtani M, Schams D, Miyamoto A, 2002: Intraluteal releaseof angiotensin II and progesterone in vivo during corporalutea development in the cow: effect of vasoactive peptides.Biol Reprod 66, 174–179.Korzekwa A, Jaroszewski JJ, Bogacki M, Deptua KM,Mas´lanka TS, Acosta TJ, Okuda K, Skar_zyn´ski DJ, 2004:Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

64 DJ Skarzynski, G Ferreira-Dias and K OkudaEffects of prostaglandin F 2a and nitric oxide on the secretoryfunction of bovine luteal cells. J Reprod Dev 50, 411–417.Korzekwa A, Shuko M, Jaroszewski J, Izabela Wocawek-Potocka, Okuda K, Skarzynski DJ, 2006: Nitric oxideinduces programmed cell dead in the bovine corpus luteum:mechanism of action. J Reprod Dev 52, 353–361.Lei ZM, Chegini N, Rao Ch V, 1991: Quantitative cellcomposition of human and bovine corpora lutea fromvarious reproductive states. Biol Reprod 44, 1148–1156.Levy N, Kobayashi S, Roth Z, Wolfenson D, Miyamoto A,Meidan R, 2000: Administration of PGF 2a during the earlybovine luteal phase does not alter the expression of ET-1and its type A receptors: a possible cause for corpus luteumrefractoriness. Biol Reprod 63, 377–382.Li H, Wallerath T, Forstermann U, 2002: Physiologicalmechanisms regulating the expression of endothelial-typeNO synthase. Nitric Oxide 7, 132–147.Lohse MJ, 1993: Molecular mechanisms of membrane receptordesensitization. Biochim Biophys Acta 1179, 171–188.Mamluk R, Levy N, Ruead B, Davis JS, Meidan R, 1999:Characterization and regulation of type A endothelinreceptor gene expression in bovine luteal cells types.Endocrinology 40, 2110–2116.McCracken JA, Custer EE, Lamsa JC, 1999: Luteolysis: aneuroendocrine-mediated event. Physiol Rev 79, 263–323.Meidan R, Milvae RA, Weiss S, Levy N, Friedman A, 1999:Intra-ovarian regulation of luteolysis. J Reprod Fertil Suppl54, 217–228.Meidan R, Levy N, Kisliouk T, Podlovny L, Rusiansky M,Klipper E, 2005: The yin and yang of corpus luteum-derivedendothelial cells: Balancing life and death. Domest AnimEndocrinol 29, 318–328.Miyamoto A, Shirasuna K, Wijayagunawardane MP, WatanabeS, Hayashi M, Yamamoto D, Matsui M, Acosta TJ,2005: Blood flow: a key regulatory component of corpusluteum function in the cow. Domest Anim Endocrinol 29,329–339.Nagata S, 1997: Apoptosis by death factor. Cell 88, 355–365.Nett TM, McClellan MC, Niswender GD, 1976: Effects ofprostaglandins on the ovine corpus luteum: blood flow,secretion of progesterone and morphology. Biol Reprod 15,66–78.Neuvians TP, Schams D, Berisha B, Pfaffl MW, 2004:Involvement of pro-inflammatory cytokines, mediators ofinflammation, and basic fibroblast growth factor in PGF 2a -induced luteolysis in bovine corpus luteum. Biol Reprod 70,473–480.Nishimura R, Sakumoto R, Tatsukawa Y, Acosta T, OkudaK, 2007: Oxygen concentration is an important factor formodulating progesterone synthesis in bovine corpus luteum.Endocrinology 147, 4273–4280.Nishimura R, Komiyama J, Tasaki Y, Acosta T, Okuda K,2008: Hypoxia promotes luteal cell death in bovine corpusluteum. Biol Reprod 78, in press.Niswender GD, Reimers TJ, Diekman MA, Nett TM, 1976:Blood flow: a mediator of ovarian function. Biol Reprod 14,64–81.Niswender GD, Juengel JL, Silva PJ, Rollyson MK, McIntushEW, 2000: Mechanisms controlling the function and lifespan of the corpus luteum. Physiol Rev 80, 1–29.Niswender GD, Davies TL, Griffith RJ, Bogan RL, Monser K,Bott RC, Bruemmer JE, Nett TM, 2007: Judge, jury andexecutioner: the auto-regulation of luteal function. ReprodDom Ruminants VII; Soc Reprod Fertil 64, 191–206.Ohtani M, Kobayashi S, Miyamoto A, Hayashi K, Fukui Y,1998: Real-time relationships between intraluteal andplasma concentrations of endothelin, oxytocin, and progesteroneduring PGF 2a -induced luteolysis in the cow. BiolReprod 58, 103–108.Okuda K, Sakumoto R, Uenoyama Y, Berisha B, MiyamotoA, Schams D, 1999: TNFa receptors in microvascularendothelial cells from bovine corpus luteum. Biol Reprod61, 1017–1022.Okuda K, Korzekwa A, Shibaya M, Murakami S, NishimuraR, Tsubouchi M, Woclawek-Potocka I, Skarzynski DJ,2004: Progesterone is a suppressor of apoptosis in bovineluteal cells. Biol Reprod 71, 2065–2071.Pate JL, 1988: Regulation of prostaglandins synthesis byprogesterone in the bovine corpus luteum. Prostaglandins36, 303–315.Pate JL, 1996: Intercellular communication in the bovinecorpus luteum. Theriogenology 45, 1381–1397.Pate JL, 2003: Lives in the balance: responsiveness of thecorpus luteum to uterine and embryonic signals. Reproduction61(Suppl), 207–217.Pate JL, Landis Keyes P, 2001: Immune cells in the corpusluteum: friends or foes? Reproduction 122, 665–676.Penny LA, Armstrong D, Bramley TA, Webb R, Collins RA,Watson ED, 1999: Immune cells and cytokine production inthe bovine corpus luteum throughout the ooestrous cycleand after induced luteolysis. J Reprod Fertil 115, 87–96.Petroff MG, Petroff BK, Pate JL, 2001: Mechanisms ofcytokine-induced death of cultured bovine luteal cells.Reproduction 121, 753–760.Rae MT, Menzis GS, McNeilly AS, Woad K, Webb R,Bramley TA, 1998: Specific non-genomic, membrane-localizedbinding sites for progesterone in the bovine corpusluteum. Biol Reprod 58, 1394–1406.Reynolds L, Redmer D, 1999: Growth and development of thecorpus luteum. J Reprod Fertil Suppl 54, 181–191.Rosiansky-Sultan M, Klipper E, Spanel-Borowski K, MeidanR., 2006: Inverse relationship between nitric oxide synthasesand endothelin-1 synthesis in bovine corpus luteum: interactionsat the level of luteal endothelial cell. Endocrinology147, 5228–5235.Rueda BR, Tilly KI, Hansen TR, Hoyer PB, Tilly JL, 1995:Expression of superoxide dismutase, catalase and glutathioneperoxidase in the bovine corpus luteum: evidencesupporting a role for oxidative stress in luteolysis. Endocr3, 227–232.Rueda BR, Tilly KI, Hansen TR, Jolly PD, Hoyer PB, TillyJL, 1997: Increased Bax and interleukin-1b-convertingenzyme messenger RNA levels coincide with apoptosis inthe corpus luteum during structural regression. Biol Reprod56, 186–193.Rueda BR, Hendry IR, Hendry WJ III, Fong HW, StormashakF, Slaydenlayden OD, Davis JS, 2000: Decreasedprogesterone level and progesterone receptors antagonistspromote apoptotic cell death in bovine luteal cells. BiolReprod 62, 269–276.Sakamoto K, Miwa K, Ezashi T, Okuda-Ashitaka E, OkudaK, Houtani T, Sugimoto T, Ito S, Hayaishi O, 1995:Expression of mRNA encoding the PGF 2a receptor inbovine corpora lutea throughout the ooestrous cycle andpregnancy. J Reprod Fertil 103, 99–105.Sakumoto R, Berisha B, Kawate N, Schams D, Okuda K,2000: TNFa and its receptors in bovine corpus luteumthroughout the oestrous cycle. Biol Reprod 62, 192–199.Sasahara K, Shirasuna K, Watanabe S, Asahi T, Nagai K,Miyamoto A, 2007: Nitric oxide stimulated by PGF 2ainduces the drastic increase of luteal blood flow and has arole in functional and structural luteolysis. Biol Reprod 77(Sp. Iss.), 225–225 (Abstract 578).Schams D, Berisha B, 2004: Regulation of corpus luteumfunction in cattle – an overview. Reprod Dom Anim 39,241–251.Schams D, Berisha B, Neuvians T, Amselgruber W, KraetzlWD, 2003: Real-time changes of the local vasoactive peptideÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Regulation of Luteal Function 65systems (angiotensin, endothelin) in the bovine corpusluteum after induced luteal regression. Mol Reprod Dev65, 57–66.Shaw DW, Britt JH, 1995: Concentrations of TNFa and P4within the bovine corpus luteum sampled by continuousflowmicrodialysis during luteolysis in vivo. Biol Reprod 53,847–854.Shirasuna K, Asaoka H, Acosta TJ, WijayagunawardaneMPB, Ohtani M, Hayashi M, Matsui M, Miyamoto A,2004: Real-time relationships in intraluteal release amongPGF 2a , endothelin-1, and angiotensin II during spontaneousluteolysis in the cow. Biol Reprod 71, 1706–1711.Silva PJ, Juengel JL, Rollyson MK, Niswender GD, 2000:Prostaglandin metabolism in the ovine corpus luteum:catabolism of PGF 2a coincides with resistance of the CLto PGF 2a . Biol Reprod 63, 1229–1236.Skarzynski DJ, Okuda K, 1999: Sensitivity of bovine corporalutea to PGF 2a is dependent on progesterone, oxytocin andprostaglandins. Biol Reprod 60, 1292–1298.Skarzynski DJ, Bogacki M, Kotwica J, 1997: Changes inovarian OT secretion as an indicator of corpus luteumresponse to PGF 2a treatment in cattle. Theriogenology 48,733–742.Skarzynski DJ, Kobayashi S, Okuda K, 2000: Influence ofnitric oxide and noradrenaline on PGF 2a -induced oxytocinsecretion and intracellular calcium mobilization in culturedbovine luteal cells. Biol Reprod 63, 1000–1005.Skarzynski DJ, Jaroszewski JJ, Okuda K, 2001: Luteotropicmechanisms in the bovine corpus luteum: role of oxytocin,PGF 2a , progesterone and noradrenaline. J Reprod Dev 47,125–137.Skarzynski DJ, Bah MM, Wocawek-Potocka I, Deptua K,Korzekwa A, Shibaya M, Pilawski W, Okuda K, 2003a:Roles of TNFa in the regulation of the oestrous cycle incattle: an in vivo study. Biol Reprod 69, 1907–1913.Skarzynski DJ, Jaroszewski JJ, Bah MM, Deptua KM, BarszczewskaB, Gawronska B, Hansel W, 2003b: Administrationof a nitric oxide synthase inhibitor counteracts PGFF 2a -induced luteolysis in cattle. Biol Reprod 68, 1674–1681.Skarzynski DJ, Jaroszewski JJ, Okuda K, 2005: Role of TNFaand nitric oxide in luteolysis in cattle. Domest AnimEndocrinol 29, 340–346.Skarzynski DJ, Woclawek-Potocka I, Korzekwa A, Bah MM,Piotrowska KK, Barszczewska B, Okuda K, 2007: Infusionof exogenous TNF dose dependently alters the length ofthe luteal phase in cattle: differential responses to treatmentwith indomethacin and a nitric oxide synthase inhibitor(L-NAME). Biol Reprod 76, 619–627.Sugino N, Telleria CM., Gibori G, 1997: Progesterone inhibits20a-HSD expression in the rat corpus luteum through theglucocorticoid receptor. Endocrinology 138, 1497–1500.Taniguchi H, Yokomizo Y, Okuda K, 2002: Fas-fas ligandsystem mediates luteal cell death in bovine corpus luteum.Biol Reprod 66, 754–759.Towson DH, O’Connor CL, Pru JK, 2002: Expression ofMCP-1 and distribution of immune cells populations in thebovine corpus luteum through the oestrous cycle. BiolReprod 66, 361–366.Tsai S-J, Wiltbank MC, 1998: PGF 2a regulates distinctphysiological changes in early and mid-cycle bovine corporalutea. Biol Reprod 58, 346–352.Vonnahme KA, Wilson ME, Li Y, Rupnow HL, PhernettonTM, Ford SP, Magness RR, 2005: Circulating levels ofnitric oxide and vascular endothelial growth factor throughoutovine pregnancy. J Physiol-Lond 565, 101–109.Watanebe S, Shirasuna K, Matsui M, Yamamoto D, BerishaB, Schams D, Miyamoto A, 2006: Effect of intralutealinjection of endothelin type A receptor antagonist onPGF 2a -induced luteolysis in the cow. J Reprod Dev 52,551–559.Weems YS, Lewis AW, Randel RD, Weems CW, 2002: Effectsof PGE 2 and PGF 2a , trilostane, mifepristone, palmitic acid(PA), indomethacin, ethamoxytriphetol, PGE 2 + PA, orPGF 2a + PA on PGE 2 , PGF 2a , and progesterone secretionby bovine corpora lutea of mid-pregnancy in vitro. Chin JPhysiol 45, 163–168.Weems YS, Randel RD, Tatman S, Lewis AW, NeuendorffDA, Weems CW, 2004: Effects of estrous synchronizationon the response to nitric oxide donors, nitric oxide synthaseinhibitors, and endothelin-1 in vitro. Prostaglandin OtherLipid Med 74, 45–59.Wiltbank MC, Gallagher KP, Christensen AK, Brabec RK,Keyes PL, 1990: Physiological and immunocytochemicalevidence for new concept of blood flow regulation in corpusluteum. Biol Reprod 42, 139–149.Wiltbank MC, Shiao TF, Bergfelt DR, Ginther OJ, 1995:PGF 2a receptors in the early bovine corpus luteum. BiolReprod 52, 74–78.Wright MF, Sayre B, Keith Inskeep EK, Flores JA, 2001:PGF 2a regulation of the bovine corpus luteum endothelinsystem during the early and midluteal phase. Biol Reprod65, 1710–1717.Author’s address (for correspondence): DJ Skarzynski, Department ofReproductive Immunology, Institute of Animal Reproduction andFood Research of PAS, 10-747 Olsztyn, Poland. E-mail: skadar@pan.olsztyn.plConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 66–73 (2008); doi: 10.1111/j.1439-0531.2008.01144.xISSN 0936-6768Captive Breeding of Cheetahs in South Africa – 30 Years of Data from the de WildtCheetah and Wildlife CentreHJ Bertschinger 1 , DGA Meltzer 2 and A van Dyk 31 Section of Reproduction, Department of Production Animal Studies, Faculty of Veterinary Science, University of Pretoria, Pretoria, South Africa;2 Emeritus Prof., Brooklyn, Pretoria, Gauteng Province, South Africa; 3 de Wildt Cheetah and Wildlife Trust, de Wildt, North West Province, SouthAfricaContentsThe de Wildt Cheetah and Wildlife Centre was established in1971 and the first cheetah cubs were born in 1975. During theperiod 1975–2005, 242 litters were born with a total of 785cubs. Mean cub survival from 1 to 12 months and greater than12 months of age was 71.3 and 66.2%, respectively. Themajority of losses (84.9%) occurred during the first monthpostpartum whereas only 15.1% deaths took place between 1and 12 months of age. Females were first bred at an age ofapproximately 3 years, reached maximum reproductive age at6–8 years, where after fertility declined. Males reached peakreproduction at 6 and maintained this for up to 12 years ofage. Male fertility was best correlated with sperm morphology.During recent years, for practical purposes, males wereallocated to ‘good’ (‡70% normal), ‘fair’ (40–70% normal)and ‘poor’ (

Captive Breeding of Cheetahs in South Africa 67Fig. 1. Layout of the camps at thede Wildt Cheetah and WildlifeCentre. C = quarantine; D =walkway ⁄ service road; E =camps for weaned cubs; G =camps for subadults; F = breedingcamps for females; M =maternity section within eachbreeding camp. Insert shows anexample of a maternity sectionwith den for cubs and triangularfeeding area with crush for catchingand restraining cheetahs andperforming minor procedures suchas collection of blood samplesAnnual selection of breeding animalsSelection based on pedigree and genetic markersEvery year, prior to the main breeding period (November–June),potential males and females are selectedaccording to pedigree and, from 2007 onwards, based onarea of origin and an unique breeding value calledWEDS (Oliehoek et al. 2006) that maximizes diversity.In addition, microsatellite genetic marker-based analysisof the de Wildt population has allowed us to select malesand females that represent the major breeding lines andclarify origin and relatedness between selected breeders.This has been done based on cluster analysis (Pritchardet al. 2000) and genetic marker-based relatedness(Lynch and Ritland 1999; Wang 2002) thus minimizinginbreeding levels as much as possible.Selection of males on the basis of their fertilityThe males that have been selected as described above arethen submitted to a breeding soundness examination. Acomplete spermiogram is performed using phase contrastmicroscopy (Meltzer et al. 1998). In addition tothis, the animal’s behaviour towards females is established.Males showing aggressive behaviour towardsfemales are not suitable. During the last 12 years, asimplified classification for the selection of breedingmales has been used at de Wildt. Males are classifiedinto ‘good’ (‡70% normal), ‘fair’ (40–70% normal) and‘poor’ (

68 HJ Bertschinger, DGA Meltzer and A van Dykcollected from six cheetah females two to three times perweek. The blood samples were assayed for progesteroneand 17b-oestradiol (Bertschinger et al. 1984).ResultsLitters and cubs produced 1975–2005During the period 1975–2005, 242 litters were born witha total of 785 cubs. It should be noted that in some yearssmall numbers of females were bred because of a limiteddemand. In 1982 as a result of a bush fire on theproperty no breeding was attempted. Mean gestationperiod was 93 days. Pseudopregnancy, which is shorterthan pregnancy, occurred sporadically. Mean cub survivalfrom 1 to 12 months and greater than 12 monthsof age was 71.3 and 66.2%, respectively. The majority oflosses (84.9%) occurred during the first month postpartumwhereas only 15.1% of deaths took placebetween 1 and 12 months of age.Female effectsBreeding ageFigure 2 shows the age specific birth rate of 55 litters incheetah females at de Wildt for the years between 1975and 1987 (Meltzer 1987). From this it is evident thatcheetah females started breeding when 2.5–3 years ofage but only reached maximum reproductive capabilitywhen 6 years old. First and even second litter femaleswere commonly poor mothers and often abandoned orate their newly born cubs. Maximum reproductive ratewas maintained until the age of eight after which itdeclined. This was the result of increased morbidity ratein females with diseases such as chronic renal failure andgastritis.Seasonal distribution of mating at de WildtAt de Wildt the practice was to breed cheetahs duringthe months November–February (Fig. 3). This maycreate the impression that the species is a seasonalbreeder. Births occurred from the end of March andcontinued until the beginning of June reflecting theduration of the breeding period.Fig. 3. Monthly distribution of conceptions of cheetah at de Wildtthat resulted in births »93 days later (Meltzer 1987)Oestrus cycleThe presence of males in the walk-way between thefemale camps appeared to stimulate oestrous cycles infemales. Oestrus was suspected when males accumulatedat the female’s enclosure. Females abandoned theirsecretive behaviour and moved towards males in thewalk-way and urinated close to the fence. Interactionbetween animals gives rise to overt signs of oestrus.They made chirping sounds in response to the stuttercalls of the males; tail lifting and rolling was seen. Theoestrus cycle length in six cheetahs that were allowed tocycle before they were bred ranged from 10 to 21 days.A typical progesterone and oestradiol profile observedin one of these females that was mated during the secondoestrus of the observation period is shown in Fig. 4(Bertschinger et al. 1998). Though difficult to determineother than by the period of receptivity, the length ofoestrus varied from 1 to 3 days. These figures agreefavourably with those reported for the domestic cat(Wildt et al. 1981). Oestradiol plasma concentrationsshowed cyclical peaks, which lasted approximately12 days. Progesterone plasma concentration only rosesubsequent to mating meaning that cheetahs, like someof the other members of the cat family, is an inducedFig. 2. Number of litters (total = 55) produced according to age offemales in years from 1975–1987 at de Wildt (Meltzer 1987)Fig. 4. Oestradiol and progesterone blood profiles during two oestrouscycles (cheetah Jean) also showing laparoscopic appearance of afollicle (F) and two corpora lutea (CLs). She was mated during thesecond oestrus after which a sharp rise in progesterone can be seen(Bertschinger et al. 1998).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Captive Breeding of Cheetahs in South Africa 69ovulator (Wildt et al. 1980). Plasma progesterone concentrationsare now used routinely to confirm mating.The laparoscopic examinations supported the findingsof the oestrus observations as well as the hormonalpatterns (Fig. 4; Bertschinger et al. 1984).Induction of oestrusSome females that did not show spontaneous heatduring the breeding period were induced for naturalmating using a modification of the regimen described byHoward et al. (1992). Serum progesterone concentrationswere measured to ensure basal levels beforeinductions were attempted. On day 0, 200 IU eCG[Folligon, Intervet SA (Pty) Ltd, Isando, South Africa]was injected and followed by 100 IU hCG (Chorulon,Folligon, Intervet SA (Pty) Ltd., Isando, South Africa)72 h later. A male was introduced into the females’enclosures soon after the hCG-injection. Of the 11females treated 5 (46%) were mated and showed a rise inplasma progesterone. Four (36%) gave birth to livecubs. Progesterone concentrations remained basal in the6 (56%) unmated females.Heterozygosity of de Wildt cheetahs and heritability offemale fertility and cub mortalityThe de Wildt captive population had a mix of ancestryfrom Namibia and South Africa and had lower levels ofinbreeding than free-ranging cheetahs analysed(Wright’s Fst values – 0. 019 vs 0.13, respectively; SPSasidharan, unpublished data). These findings wouldseem to be logical, as in the wild, with very lowpopulation densities, the degree of inbreeding should begreater due to the restricted availability of potentialpartners. The heritability for litter size was high at0.5848 (SE 0.078) in the de Wildt pedigree (532 progeny,1975–2007; Sasidharan, unpublished data). The maternalheritability for cub mortality was estimated to be0.596 (SE 0.131). As in the wild (Caro 1994), cubsurvival was not affected by season.Male effectsBreeding agePotential breeding males were first examined at an ageof 3.5–4 years. The proportion of males found to haveacceptable semen quality (‘good’ and ‘fair’ categories)increased from 55.6% (n = 18) to 71.5% (n = 14) inthe age categories 3.5 to 4.5 to

70 HJ Bertschinger, DGA Meltzer and A van DykTable 1. Summary characteristics of fresh cheetah semen samples(n = 160)pHVolume(ml)% Livesperm% Sperm withlinear motilitySpermcount · 10 6 ⁄ ml% NormalspermRange 6.4–8.0 0.3–2.1 10–95 8–90 1–211 0–86Mean 0.7 65.2 58.1 32.7 40.3SD 0.5 18.5 18.6 36.13 17.5Table 2 lists the more common sperm defects ofcheetahs. Abnormal heads, knobbed acrosomes(Fig. 7e, f), mid-piece and flagellar defects were themost common morphological abnormalities of cheetahsperm. It is accepted that a normal population ofcheetah sperm may vary moderately as far as head size isconcerned (pleiomorphism; Wildt et al. 1983, 1987).Compared to the sperm of species such as bulls, ramsand boars, cheetah sperm heads are very small and notas flat (Fig. 7b,c). Orientation on the slide probablyexacerbates the pleiomorphic appearance. Flagellardefects, (Fig. 7d,h,i), depending on their incidence, hada marked affect on linear sperm motility and generallyspeaking good motility was associated with at least fairmorphology. Where testicular function was severelyimpaired, the incidence of spermatogenic cells was oftenincreased (Fig. 7j).Figure 8 shows the relationship between sperm morphologyand number of litters born at de Wildt(Bertschinger and Meltzer 1998). As the % normalsperm increased so did the number of litters. As can beexpected, compared to % normal sperm, an inverserelationship between % major sperm defects and littersborn was seen. The low number of births on the lefthandside for % major defects and on the right for %normal sperm reflects the small numbers of male cheetahwith good sperm morphology. Approximately 30% ofall males tested fell into the ‘poor’ category (see below).Semen quality of wild-caught males appeared to be nobetter than that of captive-bred animals. Outbred freerangingsouthern African lions and leopards, by comparison,display good sperm morphology and the sizeand shape of sperm heads, which are flatter, areconsistent (Fig. 7k).During the period 1998–2007, 113 semen examinationswere carried out. The number and percentage ofmales falling into the ‘good’, ‘fair’ and ‘poor’ categorieswere 41 (36.3%), 27 (23.9%) and 45 (39.8%), respectively.Twenty-six males were examined over 2 years ormore during the period of 1998–2006. Of these, nineshowed an improvement by one or more categories, nineremained the same and seven showed a decrease insemen quality. Deterioration was often associated withdiseases such as gastritis. We analysed the litter datafrom 1999 to 2007 (excluding 2002) in terms of spermmorphology categories we use at de Wildt. The numberof litters produced and average litter sizes for categories‘good’, ‘fair’ and ‘poor’ males were 21 and 3.44, 18 and3.14 and 18 and 2.28, respectively. The differences inlitter size were not significant (p = 0.1449; Kruskall–Wallis test).DiscussionThe breeding results achieved at de Wildt over the last31 years prove that cheetah can be bred successfully incaptivity. The success, however, is dependent on thecorrect housing, day to day management and breedingmanagement. Other centres such as the Cheetah Projectat Kapama and the Wassenaar Wildlife Breeding Centrein The Netherlands have also bred captive cheetahssuccessfully. The breeding strategy at Wassenaar, whichhas resulted in 210 cubs born from 62 litters, is worthdescribing as it differs a little from the one practiced atde Wildt. Briefly, male and female utilize the same campin turns which means that they pick up the scents of theopposite sex (Beekman et al. 1997; Louwman andLouwman 2005). The behaviour of the cheetahs isobserved carefully and once the male and female showintense interest in the scents left behind contact throughthe fence of adjoining enclosures is allowed. When thefemale displays signs of oestrus she is given access to themale enclosure and left with him for approximately 48 hduring which two to five matings occur.From our earlier work (Meltzer 1987; Bertschingerand Meltzer 1998) a relationship was seen betweensperm morphology and number of litters born at deWildt. Despite poor semen quality, we showed morerecently (1999–2007) that some males are capable ofproducing litters albeit with a smaller average litter size(2.28 vs 3.14 and 3.44). A number of factors, however,may affect semen quality. Semen collected by means ofelectro-stimulation does not necessarily reflect the truepotential of a particular male. Firstly, urine contaminationaffects especially motility but maybe even tailHead defects Mid-piece defects Tail defectsTable 2. Common defects of cheetahspermKnobbed ⁄ lipped acrosome Aplasia of mitochondrial helix Mid-piece reflexLoose acrosome a Pseudodroplet Bent mid-pieceMacrocephalic with single or double tail Stump-tail Coiled tail – coiled belowor on top of the headMicrocephalicProximal dropletImmature headDistal dropletDegenerated or vacuolated headFree tail forms with aplasiaof the mitochondrial helix bPyriformDeformed headsLoose heada This defect, if the sperm is otherwise normal, was not added to the % abnormal sperm.b This defect should merely be noted and not added to the % abnormal sperm. It arises during meiosis and is associatedwith macrocephalic heads in the ejaculate.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Captive Breeding of Cheetahs in South Africa 71(a)(b)(c) (d) (e) (f)(g)(h)(i)Fig. 7. (a) Morphologically normalcheetah sperm (SEM) – MP =mitochondrial sheath, PP = principalpiece, E P = end piece; (b)lateral view of cheetah sperm head(SEM) – A = acrosome, CP =connecting piece, MS = mitochondrialsheath; (c) planar view ofcheetah sperm head (SEM); (d)SEM showing cheetah sperm withcoiled tails and bent midpiece; (e)lateral view of cheetah sperm withacrosomal defect (TEM) – A =acrosome, AC = acrosomal cyst,PC = proximal centriol; (f) lateralview of cheetah sperm with acrosomaldefect (SEM) – AC =acrosomal cyst, MP = mitochondrialsheath; (g) phase contrastmicrograph (PCM) of an underdevelopedcheetah sperm with anucleus and cytoplasm; (h) twocoiledtails and a ‘Dag’ defect incheetah sperm (PCM); (i) Cheetahsperm with bent midpieces (PCM);(j) three (arrows) spermatogeniccells with vacuoles in a semensmear; (k) typical head morphologyof normal lion sperm (PCM)(a–f: Coubrough et al. 1976, 1978;Coubrough and Soley 1982)(j)(k)morphology. Secondly, semen evaluation reflects awindow in time and may take place as many as4 months prior to actual mating of a female. Thirdly,semen quality, especially sperm count, is likely to beaffected by prior sexual activity. According to Amannand Almquist (1976) the daily sperm production in dairybulls is 11 million ⁄ g testicular tissue. Cheetahs havesmall testes with a mean combined volume of 8 ml(n = 28; »8 g), meaning that the daily sperm productionis limited to approximately 88 million per day. Inaddition the epididymal tails are small with a limitedcapacity for sperm storage. The mean number of spermper cheetah ejaculate was found to be 22.9 million,which is approximately 26% of daily estimated production.Nevertheless, semen evaluation remains the mostvaluable criterium for selection of males. An investigationinto poor fertility in a pair of Gir forest lions at theZurich Zoological gardens illustrates the importance ofsemen quality and fertility (Bertschinger et al. 2006a).Asian lions are known to less heterozygous than mostfree-ranging African lions (Gaur et al. 2006) and Wildtet al. (1987b) showed that inbreeding of large catscauses poor semen quality. The Asian lioness in questionproduced her last cubs in July 1998 and then had eightpresumed pseudopregnant cycles (35–74 days in length)and one anovulatory cycle (16 days long) before sheproduced cubs 27 months later. The male was examinedduring the same month that she conceived and found toÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

72 HJ Bertschinger, DGA Meltzer and A van DykFig. 8. Number of litters sired in relation to % normal sperm andmajor sperm defects (Bertschinger and Meltzer 1998)have only 12% normal sperm (80.9% major defects) butwith good motility (75% linear) and total sperm countof 390 million. Another factor that needs to be evaluatedis the heritability of male fertility or semen qualityin cheetah.Given the fact that natural breeding can be highlysuccessful, in our opinion, there seems to be littleindication for the use of assisted reproductive techniquesin captive cheetahs. In any case, reports of success incheetahs and other large cats with assisted reproductivetechniques are scant, despite the fact that the firstreports date back to 1992 (Howard et al. 1992). Artificialinsemination and in vitro fertilization have notimproved fertility and certainly cannot increase the rateof reproduction or reduce the generation interval. Whilethey may be of use for exchange of genetic material andcontrol of spread of diseases, their role for cheetahconservation is limited.The future of cheetah conservation, as for most if notall wildlife species, will depend on availability of habitatand control of illegal hunting and trading with wildcaughtcheetahs. According to Marnewick et al. (2007)there were 650 cheetahs in 44 captive facilities in SouthAfrica in 2006. Eleven of these facilities were activelybreeding cheetahs. Excluding 2000, the average annualnumber of cheetah exported legally from South Africafrom 1996 to 2002 was 30.5. In 2000, 129 cheetahs wereexported. The most likely source for the majority of theanimals exported in 2000 is from the wild. As far asillegal hunting or killing of cheetahs is concerned thereport indicates that in the Thabazimbi district of theLimpopo Province alone, 26 cheetahs were killed over a3-year period from 1999 to 2001. The main reason forthe killing of cheetahs is that farmers perceive them as athreat to their stock - both wild and domestic stock. Inorder to protect or save free-ranging cheetahs in nonprotectedareas, the de Wildt Cheetah and Wildlife Trustestablished the de Wildt Wild Cheetah Project in 2000.The goals were to establish numbers of free-rangingcheetahs outside of protected areas, assist farmers intrapping cheetahs in areas where they were not wanted,relocate such animals to fenced game reserves and toeducate farmers and communities about the need forcarnivore conservation in general. The project has beenmost successful and many farmers now see cheetahs asan asset rather than a liability. According to Marnewicket al. (2007), the first cheetahs were caught in 2000 andby December 2006, 137 had been removed fromfarmlands. Sixteen of these could not be relocatedbecause of serious injuries sustained before or at thetime of capture. Ninety-two animals could be finallyreleased (58 males and 33 females) into areas rangingfrom 1500 to 70 000 ha in size. The first cubs were bornin 2002 and by August 2007 94 cubs (average litter size3.9 cubs) had been born to 23 females. The data showthat, given the habitat and opportunity, cheetah reproducewell in the wild.The main role of captive breeding of cheetahs inSouth Africa should be to curtail illegal trade in wildcheetahs. Captive-bred animals from de Wildt have alsobeen successfully released into the wild (Pettifer 1981).Three 5-year-old males were released onto a game farmin the Hoedspruit area of the Limpopo Province andhunted spontaneously despite being born in captivityand not having been ‘taught’ to hunt. They wereeventually recaptured because of their lack of fear forhumans. This latter problem can be overcome if cubs areraised out of contact with humansConclusionsThe de Wildt Cheetah and Wildlife Centre and othershave shown that cheetahs can be bred successfully andin a sustained manner in captivity. In order to achievethis, the correct breeding management needs to beapplied. In some smaller South Africa reserves it hasalso become necessary to contracept cheetahs (eithermales or females) to prevent inbreeding. Deslorelinimplants (Suprelorin Ò , Peptech Animal Health, Sydney,NSW, Australia) provide a safe and reversible methodof contraception for both sexes (Betschinger et al.2002a; Bertschinger et al. 2006b). Zoos or sanctuariesthat do not have the facilities, capacity or the necessarycritical mass in terms of cheetah numbers, should rathernot attempt breeding. Live animal displays are animportant part of public education and can make avaluable contribution to conservation of a species likethe cheetah. It is advisable to consider taking animals onloan for display purposes in zoos and safari parks. Thesecould be pre-breeding age animals, males with poorsemen quality or animals that are past their prime. It isalso clear that populations in zoos and smaller gamereserves need to be managed genetically in order toreduce the risks of inbreeding depression. In order to dothis, a standardized approach should be adopted so that,internationally, valid comparisons can be made.ReferencesAmann RP, Almquist JO, 1976: Bull management to maximizesperm output. Proceedings of 6th Conference on ArtificialInsemination and Reproduction, Columbia, pp. 1–10.Beekman SPA, de Wit M, Louman J, Louman H, 1997:Breeding and observations on the behaviour of cheetah(Acinonyx jubatus) at the Wassenaar Wildlife BreedingCentre. Int Zoo YB 35, 43–50.Bertschinger HJ, Meltzer DGA, 1998: Reproduction in malecheetahs. 2: Sperm morphology. In: Penzhorn BL(ed.),Proceedings of a Symposium on Cheetahs as Game RanchÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Captive Breeding of Cheetahs in South Africa 73Animals. Onderstepoort, Republic of South Africa, WildlifeGroup, South African Veterinary Association, Onderstepoort,23–24 October, pp. 153–158.Bertschinger HJ, Meltzer DGA, van Dyk A, Coubrough RI,Soley JT, Collett FA, 1984: Cheetah life-line. Nuclear Active30, 2–7.Bertschinger HJ, Meltzer DGA, van Dyk A, Strachan A, 1998:Breeding female cheetahs in captivity. In: Penzhorn BL(ed.),Proceedings of a Symposium on Cheetahs as Game RanchAnimals. Onderstepoort, 23–24 October, pp. 168–174.Bertschinger HJ, Isenbu¨ gel E, Jannett F, Ossent P, Wild P,2006a: Infertility in a pair of Indian lions in Zurich Zoo.Proceedings of the Hungarian Society of Animal Reproduction,Budapest, 26–27 June, pp. 44–46.Bertschinger HJ, Jago M, No¨ thling JO, Human A, 2006b:Repeated use of the GnRH analogue deslorelin to downregulatereproduction in male cheetahs (Acinonyx jubatus).Theriogenology 66, 1762–1767.Bertschinger HJ, No¨ thling JO, Nardini RM, Hemmelder S,Broekhuisen MH, 2002b: Collection of semen in cheetahs(Acinonyx jubatus) using electro-ejaculation: attempts toavoid urine contamination. Adv Ethol 37, 122.Betschinger HJ, Trigg TE, Jo¨ chle W, Human A, 2002a:Induction of contraception in some African wild carnivoresby down-regulation of LH and FSH secretion using theGnRH analogue deslorelin. Reproduction Suppl 60, 41–52.Brown JL, Wildt DE, Wielebnowski N, Goodrowe KL,Graham LH, Wells S, Howard JG, 1996: Reproductiveactivity in captive female cheetahs (Acinonyx jubatus)assessed by faecal steroids. J Reprod Fertil 106, 337–346.Caro TM, 1994: Cheetahs of the Serengeti Plains: GroupLiving in and Asocial Species. The University of ChicagoPress, Chicago and London.Coubrough RI, Bertschinger HJ, Soley JT, Meltzer DGA,1976: Some aspects of normal and abnormal spermatozoa incheetah (Acinonyx jubatus). Proc Elect Microscopy SocSouthern Africa 6, 5–6.Coubrough RI, Bertschinger HJ, Soley JT, 1978: Scanningelectron microscopic studies on cheetah spermatozoa. ProcElectron Microscopy Soc Southern Africa 8, 57–58.Eaton RL, 1970: Notes on the reproductive biology of thecheetah. Int Zoo YB 10, 86–89.Gaur A, Shailaja K, Singh A, 2006: Twenty polymorphicmicrosatellite markers in the Asiatic lion (Panthera leopersica). Conserv Gene 7, 1005–1008.Howard JG, Donoghue AM, Barone MA, Goodrowe KL,Blumer ES, Snodgrass K, Starnes D, Tucker M, Bush M,Wildt DE, 1992: Successful induction of ovarian activity andlaparoscopic intrauterine artificial insemination in the cheetah(Acinonux jubatus). J Zoo Wildl Med 23, 288–300.Kelly MJ, Laurenson MK, FitzGibbon CD, Collins DA,Durant SM, Frame GW, Bertram BCR, Caro TM, 1998:Demography of the Serengeti cheetah (Acinonyx jubatus)population: the first 25 years. J Zoo London 244, 473–488.Labuschagne W, 1979: ‘n Bio-ekologiese en gedragstudie vandie jagluiperd, Acinonyx jubatus jubatus (Schreber, 1776).MSc Thesis, University of Pretoria.Louwman JWW, Louwman JCM, 2005: Cheetah breedingprogram at Wassenaar Wildlife Breeding Centre. AnimKeeper’s Forum 7 ⁄ 8, 368–369.Lynch M, Ritland K, 1999: Estimation of pairwise relatednesswith molecular markers. Genetics 152, 1753.Marker-Kraus L, Grisham J, 1993: Captive breeding of cheetahsin North American zoos: 1987–1991. Zoo Biol 12, 5–18.Marnewick K, Beckhelling A, Cilliers D, Lane E, Mills G,Herring K, Caldwell P, 2007: The status of the cheetah inSouth Africa. In: Breitenmoser C and Durant S (eds.), Thestatus and conservation needs of the Cheetah in SouthernAfrica. Cat News Special Edition, 22–31.Meltzer DGA, 1987: Reproduction in male cheetah, Acinonyxjubatus jubatus (Schreber, 1776). MSc Dissertation (Pretoria),p. 118.Meltzer DGA, Mu¨ lders MS, 1998: Reproduction in malecheetahs: III Reproductive hormones. In: Penzhorn BL(ed.), Proceedings of a Symposium on Cheetahs as GameRanch Animals. Onderstepoort, 23–24 October, pp. 159–167.Meltzer DGA, Bertschinger HJ, van Dyk A, 1998: Reproductionin male cheetahs: I. Breeding management and semenevaluation. In: Penzhorn BL (ed.), Proceedings of a Symposiumon Cheetahs as Game Ranch Animals. Onderstepoort,23–24 October, pp. 145–152.O’Brien SJ, Roelke ME, Marker L, Newman A, Winkler CA,Meltzer DA, Colly L, Everman JF, Bush M, Wildt DE,1985: Genetic basis for species vulnerability in the cheetah.Science 227, 1428–1434.O’Brien SJ, Wildt DE, Bush M, 1986: The cheetah in geneticperil. Sci Am 254, 84–92.Oliehoek PA, Windig JJ, van Arendonk JA, Bijma P, 2006:Estimating relatedness between individuals in general populationswith a focus on their use in conservation programs.Genetics 173, 483–496.Pettifer HHL, 1981: Experimental release of captive-bredcheetah (Acinonyx jubatus) into the natural environment.In: Chapman JA, Punsman P(eds), Worldwide FurbearerConference Proceedings, Donnelly, VA, pp. 1–9.Pritchard JK, Stephens M, Donnelly P, 2000: Inference ofpopulation structure using multilocus genotype data. Genetics155, 945–959.Sarri KJ, 1994: Estrous behaviour of the female cheetah(Acinonyx jubatus) and the male cheetahs’ response to anestrous female. In: Caro TM (ed.), Cheetahs of the SerengetiPlains: Group Living in and Asocial Species. The Universityof Chicago Press, Chicago and London.Ulmer FA, 1957: Cheetahs are born. America’s First Zoo 9, 7.Wang J, 2002: An estimator for pairwise relatedness usingmolecular markers. Genetics 160, 1203–1215.Wildt DE, Seager SWJ, Chakraborty PK, 1980: Effect ofcopulatory stimuli on incidence of ovualtion and serumluteinizing hormone in the cat. Endocrinology 107, 1212–1217.Wildt DE, Chan SY, Seager SWJ, Chakraborty PK, 1981:Ovarian activity, circulating hormones and sexual behaviourin the cat. I. Relationships during coitus-induced lutealphase and the oestrus period without mating. Biol Reprod25, 15–28.Wildt DE, Bush M, Howard JG, O’Brien SJ, Meltzer D, VanDijk A, Ebedes H, Brand DJ, 1983: Unique seminal qualityin the South African cheetah and a comparative evaluationin the domestic cat. Biol Reprod 29, 1019–1025.Wildt DE, O’Brien SJ, Howard JG, Caro TM, Roelke ME,Brown JL, Bush M, 1987a: Similarity in ejaculate-endocrinecharacteristics in captive versus free-ranging cheetahs of twosubspecies. Biol Reprod 36, 351–360.Wildt DE, Bush M, Goodrowe KL, Packer C, Pusey AE,Brown JL, Joslin P, O’Brien SJ, 1987b: Reproductive andgenetic consequences of founding isolated lion populations.Nature 329, 328–31.Author’s address (for correspondence): HJ Bertschinger, Section ofReproduction, Department of Production Animal Studies, Faculty ofVeterinary Science, University of Pretoria, Private Bag X04, SouthAfrica. E-mail: henk.bertschinger@up.ac.zaConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 74–82 (2008); doi: 10.1111/j.1439-0531.2008.01145.xISSN 0936-6768Non-invasive Monitoring of Hormones: A Tool to Improve Reproduction inCaptive Breeding of the Eurasian LynxHormone Monitoring in Breeding Programmes of MammalsM Dehnhard 1 , S Naidenko 2 , A Frank 1 , B Braun 1 ,FGo¨ritz 1 and K Jewgenow 11 Leibniz-Institute for Zoo Biology and Wildlife Research, Berlin, Germany; 2 A.N. Severtzov Institute of Ecology and Evolution, Moscow, RussiaContentsThe survival of many critical endangered mammal species isoften depending on successful captive breeding programmeswhich include the future option of reintroduction to the wild.Breeding in captivity also demands the application of modernassisted reproductive techniques to ensure maximal biodiversity,but knowledge on reproductive physiology is oftenlimited. Therefore, non-invasive monitoring of urinary andfaecal hormones has become an important tool for reproductivemanagement. To exemplify the importance of noninvasivehormone monitoring, we choose the Eurasian lynxas a model for the world’s most endangered felid species, theIberian lynx. We analysed faecal samples of pregnant andpseudo-pregnant female Eurasian lynxes during a 3-year studyperiod. Compared to pre-mating levels faecal progesteronemetabolite profiles revealed a tendency towards higher levels inpregnant and pseudo-pregnant females with no differencebetween both categories. Oestrogen levels raised in bothpregnant and pseudo-pregnant females with a tendency to bemore elevated and prolonged in pregnant females. Surprisinglyboth E2 and P4 metabolites were highly correlated (r 2 =0.8131, p < 0.0001) showing a postpartum increase both inpregnant and pseudo-pregnant females. The results from theEurasian lynx revealed that the measurement of faecalprogesterone metabolites led to profiles dissimilar to profilesshown in other felid species, but similar to those from faecalgestagen metabolite analysis in the Iberian lynx. To identifyfaecal gestagen and oestrogen metabolites a radio-metabolismstudy was performed. Using the progesterone immunoassaytwo major progesterone metabolites were detected demonstratingthat the assay indeed tracks the relevant metabolites.The oestrogen assay measured authentic 17b-oestradiol andoestrone, and their conjugates. The analysis of the faecalmetabolite composition in samples from early and latepregnancy and lactation particularly revealed a distinct shiftin the relation between 17b-oestradiol and oestrone thatchanged in favour of oestrone. This might indicate differenthormone sources during and after pregnancy (corpus luteum,placenta). We hypothesize, that placental steroid analysis incombination with other highly sophisticated analytical techniques,like liquid chromatography mass spectrometry orurinary relaxin analysis may led to analytical options toconfirm pregnancy and to differentiate this from pseudopregnancyin lynx species.IntroductionThe main threat of mammalian species is human activityand many in situ populations are critically endangeredbecause of landscape development has reduced andfragmented their habitat (Ceballos et al. 2005; Cardilloet al. 2006). Due to the rapid decline of the species, it isapparent that some mammal populations will alreadynot be able to recover on its own. Therefore, successfulcaptive breeding programmes of managed zoo populationsmight result in animals that could be returned topart of their original range under protected regimes asdescribed for several species; notably the Arabian andscimitar horned oryx (Mesochina et al. 2003) and thePrzewalski horse (Ryder 1993). To achieve these objectivesnew approaches to improve the success of mammalconservation is urgently required and captive populationshave become of prime importance for the conservationof species in terms of research and education.Concomitantly there is a high priority in maintaining orincreasing genetic diversity in small populations to avoidinbreeding (Earnhardt et al. 2004). Thus, successfulcaptive breeding programmes of endangered speciesmust be performed on scientific basis and depend moreand more on modern assisted reproduction techniques(ART). The last 20 years are characterized by sporadicapplication of artificial insemination, gamete bankingand ⁄ or embryo production in non-domestic species(Masui et al. 1989; Hermes et al. 2001, 2007; Penfoldet al. 2005; Pukazhenthi et al. 2006; Swanson 2006;Hildebrandt et al. 2007; Stoops et al. 2007; Swansonet al. 2007). However, the most striking limitation ofusing ART in breeding programmes is limited knowledgeon reproduction biology, in particular knowledgeon hormone regulation of female reproductive cycle andpregnancy. Therefore, the purpose of the first part of thepaper is to give a brief outline on the current state of theart regarding non-invasive hormone monitoring techniques.Hormones control reproductive success. Knowing ananimal’s hormone patterns allows understanding itsbiology, and diagnosing and correcting infertility. Whilemany hormones interact in the reproductive process, thesteroid hormones oestradiol, progesterone and testosteroneare the most important regulators of reproductivebehaviours and functions across vertebrates. They areproduced by ovaries and testes, and their concentrationin blood serum is used to validate reproductive activityof mammals. Single blood samples, which are usuallyavailable from captive animals, however, can hardly beused to assess reproductive status of an exotic species.Hormone values depend on many factors, like seasonality,reproductive silence, reproductive suppression,pseudo-pregnancy, which are simply not known orstudied before in several exotic species.Presuming the knowledge of a particular reproductivepattern of a mammalian species, the analysis ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Non-invasive Monitoring of Hormones 75hormones is essential to assist husbandry of animals incaptivity. These information’s contributed to the reproductivesuccess by increasing the understanding ofreproductive cycling and breeding behaviour (Moreiraet al. 2001; Paris et al. 2002; Cavigelli et al. 2003;Pereira et al. 2006), investigations on reproductiveseasonality (Scheibe et al. 1999; Mooring et al. 2004;Jewgenow et al. 2006), the establishment of successfulbreeding pairs in captivity (Heistermann et al. 2004),improvement of oestrous synchronization (Penfold et al.2005), and ovulation induction protocols with gonadotrophicreleasing hormone (GnRH) agonists to achievethe goals of successful artificial insemination and tocorrect infertility (Asa et al. 2006; Graham et al. 2006).Another important factor to be considered is theknowledge of hormone producing organs in a particularspecies. Beside gonads, the placenta in pregnant females,and also the adrenal glands are capable of producingsteroids. Stress caused by catching and restrainingbefore blood sampling, together with venipuncturemight activate the HPA axis and thus confound theeffects of the experimental treatment. Repeated bloodsampling might also induce anticipatory stress reactionsof the animals involved. To avoid this, during the lastdecades methods have been developed to measuresteroid hormones non-invasively. Collected urine andfaecal samples have become a substitute for analysinghormones in the serum or plasma (see review Schwarzenberger2007). Steroid hormones are extensivelymetabolized in various organs, mainly in the liver andare excreted via the bile into the gut and via the kidneyinto the urine. These metabolites are the source for noninvasivemethods of measuring an animal’s endocrinestatus. Therefore, what was once considered a daily‘waste product’ now has become a valuable ‘researchresource’, a virtual pool of biological information onreproductive cyclicity, the time of ovulation, pregnancyand impending parturition. Particularly exciting arerecent advances to measure glucocorticoid metabolitesas indicators of stress, which can be used to help enrichzoo environments, thereby improving reproduction andanimal well-being.Despite the enormous progress of using faecal andurine hormone analysis for research on reproductionbiology of exotic species, each analytical method needsto be validated each time for a particular species and aparticular hormone. Steroid metabolism by the liver aswell as microbial impact during the intestinal passage,and re-absorbtion into the enterohepatic circulationgenerate a vast number of faecal steroid metabolites, yetdifferent in even closely related species (Palme et al.1996; Heistermann et al. 2006).This necessary species-specific approach depends onthe determination of the major faecal hormone metabolite(s)of a steroid. The easiest way for identification ofthese metabolites is a radio-metabolism study, whichinclude the infusion of radiolabelled steroids and detectionof radioactivity within faecal or urine hormoneextracts. The disadvantage of radioinfusion studies isthat they typically require special permits and equipmentregarding animal care facilities and the generationof radioactive waste, but it should be performed whenever possible. Since animals in captivity are idealresearch subjects and such a study is considered to besafe (level of radioactivity is clearly below the maximumpermitted level), zoos had to be encouraged to performradiometabolism studies as often as possible to contributeto the increasing data base of steroid hormonemetabolites applicable for monitoring reproductivephysiology of exotic species.Currently, different techniques are applied fromresearchers worldwide leading to diverse concentrationsbetween studies regarding their absolute concentrationsreferred to faecal wet and dry weight, respectively. Themain reason is that different antibodies often generatedagainst the parent steroid were used with unknowncross-reactivities for the relevant steroid metabolites.Even commercially available antibodies advertised ashormone specific have been used with success for faecalsteroid analysis in several studies (Brown et al. 1996;Wasser et al. 2000). Nevertheless, comparability existsbetween their physiologic outcome, demonstrating thatfaecal and urinary steroid metabolites are reliableindicators of gonadal and adrenal activity whilst themolecular structures of the metabolites remainedunknown in most cases. This so called biologicalvalidation of a non-invasive hormone measurementmust be performed before a particular assay can beapplied for monitoring reproductive activity in a valuableanimal.The purpose of the second part of this paper is todemonstrate how non-invasive hormone techniques canbe used to monitor reproduction in a non-domesticmammal species and which steps had to be performedbefore it can be successfully applied for breedingroutine. We choose the Eurasian lynx as a model forbasic research and assay development. Although hormonemetabolites are widely described and validated formonitoring reproduction and pregnancies in female wildcats (Brown et al. 1994, 2001; Graham et al. 2005), theexample of the Eurasian lynx will show that validationof hormone metabolites must be performed for eachindividual species.Materials and MethodsAnimalsFemale Eurasian lynx were housed at the scientific fieldstation ‘Tchernogolovka’ of the A.N. Severtzov Institute,situated 50 km north-east of Moscow. Faecalsamples were collected once a month from individualfemales and stored at )20°C within 1 h after defecationuntil analyses. From February to April (prospectivemating season) the frequency of collection was increasedto one to two times a week.Faecal samples (0.5 g) were extracted for 30 min byshaking with 4.5 ml of 90% methanol. After centrifugation(15 min at 1200 · g) the supernatant was transferredinto a new tube. Aliquots of the faecal extractswere subjected either to high-performance liquid chromatography(HPLC) analysis, or diluted 1 + 1 withwater and added directly to the respective steroid enzymeimmune assay (EIA). All hormone measurements werecarried out in duplicates to assess a coefficient ofvariation. The results were expressed as immunoreactivesteroid metabolites in lg ⁄ g of wet faeces.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

76 M Dehnhard, S Naidenko, A Frank, B Braun, F Go¨ ritz and K JewgenowRadiometabolism studyTo identify relevant progesterone metabolites, whichreflect the corpus luteum (CL) activity, a radiometabolismstudy was performed: for that purpose a solution(0.25 ml) containing 250 lCi [ 3 H] progesterone(70–110 Ci ⁄ mmol, TRK413; Amersham Bioscience,Freiburg, Germany) in ethanol was used. Sterile 0.9%NaCl solution (2.25 ml) was added to the radiolabelledsolution and the total volume was injected into thecephalic vein of a 15-year-old female lynx. Prior toinjection, the animal was sedated by an i.m. injectionwith a mixture of 3 ml of Rometar (2% solution ofXylazine hydrochloride) and 1 ml of 10% Ketaminehydrochloride (both from Spofa, Prague, Czech Republic)corresponding to a dosage 2.4 and 4 mg ⁄ kg bodyweight for Xylazine and Ketamine, respectively.Following radiolabel injection, all voided faecal sampleswere collected separately for 4 days in plastic bagsfrom the cage and the floor of the enclosure immediatelyafter defecation and stored at )20°C. Aliquots of eachsample were extracted for progesterone and oestrogendetermination and for radioactivity counting. The secondsample, collected on day 2 after injection, containedthe highest amount of radioactivity (89%) and it wasused for HPLC analyses. All radioactive counting wasconducted in a Perkin Elmer MicroBeta Trilux counter(Perkin Elmer, Rodgau, Germany).Tritiated steroids are usually of very low radiotoxicitybecause of very low energy beta emission. The specificradio activity of injected 3 H-steroids, however, is quitehigh, so the additional mass of these radioactivehormones will not substantially increase the totalconcentration of hormone in the body. Furthermore,the amount of radiolabel used for this study is far belowmaximum permitted level. Nevertheless, in Germanyradiometabolism studies in animals require a priorpermission of the local Animal Experiment EthicalCommittee.HPLC analysis of metabolitesFor separation and characterization of faecal steroidmetabolites, 50 ll portions of faecal extracts were used.For gestagen metabolite analysis a reversed-phase UltrasepES100 ⁄ RP – 18 ⁄ 6 lm HPLC column (4 · 250 mm;Sepserv, Berlin, Germany) was used. Metabolites wereseparated with a methanol + water mixture (78 + 22)at a flow rate of 1 ml ⁄ min. Fractions of 0.33 ml werecollected at 20 s intervals and diluted with one volumeof water, before 20 ll of the fractions were added intothe assay systems. The elution positions of authenticprogesterone (4-pregnen-3,20-dione; P4), 5a-pregnan-3,20-dione (DHP), 5a-pregnan-3b-ol-20-one (5a-P), 5bpregnan-3a,20a-diol(pregnanediol, PD) on this columnhad been previously determined in separate HPLC runs.For faecal oestrogen metabolite separation an AllureBiphenyl 5 lm HPLC column (3.2 · 150 mm; Restek,Bad Homburg, Germany) was used. Metabolites wereseparated with a acetonitrile + water mixture (43 + 57)at a flow rate of 1 ml ⁄ min. Fractions were collected andadded to the assay as described above. The elutionpositions of authentic 1,3,5(10)-oestratrien-3,17-one(oestrone), 1,3,5(10)-oestratrien-3,17a-diol (17a-oestradiol),and 1,3,5(10)-oestratrien-3,17b-diol (17b-oestradiol)on this column had been determined in separateHPLC runs after their injection.ProgesteroneProgesterone (P4) analyses were carried out with anin-house microtitre plate enzyme immunoassay asdescribed earlier (Go¨ ritz et al. 2001) using a commercialP4 antibody (Sigma P1922, raised in rats to progesterone)and 4-pregnen-3,20-dione-3-CMO-peroxidaselabel. The cross-reactivities to progestins were asfollows: 4-pregnen-3,20-dione (progesterone), 100%;5a-pregnan-3,20-dione, 31%; 5a-pregnan-3b-ol-20-one,18%; 5-pregnen-3b-ol-20-one, 12%; 4-pregnen-3aol-20-one,4.2%; 0.05). Intra- and inter-assay coefficients of variationfor two biological samples with low and highconcentrations were 5.0 and 5.1% (n = 10) and 13.7 and22.8% (n = 10), respectively.OestrogensFaecal oestrogen analyses were carried out also with anin-house microtitre plate enzyme immunoassay using apolyclonal antibody raised in rabbits to 1,3,5(10)-oestratrien-3,17b-diol-17-HS-BSA and 1,3,5(10)-oestratrien-3,17b-diol-17-HS-peroxidaselabel (Meyer et al.1997). The cross-reactivities to oestrogens were asfollows: 1,3,5(10)-oestratrien-3,17b-diol (17b-oestradiol)100%, 1,3,5(10)-oestratrien-3,17-one (oestrone) 100%,1,3,5(10)-oestratrien-3,17a-diol (17a-oestradiol) 66%,1,3,5(10)-oestratrien-3,16a,17b-triol (oestriol) 1.5%, and 0.05).Intra- and inter-assay coefficients of variation for twobiological samples with low and high concentrationswere 5.8 and 12.3% (n = 10) and 17.0 and 9.7% (n =11), respectively.Results and DiscussionMonitoring of faecal steroid hormonesWe analysed faecal samples of pregnant (n = 15),pseudo-pregnant (n = 7) female Eurasian lynxes duringa 3-year study period. Figure 1a,b shows the faecalprogesterone and oestrogen metabolite profiles in apregnant and a pseudopregnant female. There is atendency towards higher gestagen and oestrogen metaboliteconcentration during pregnancy. However, nodistinct difference between profiles from the pregnant(a) and the pseudo-pregnant (b) female were obtained.In addition, both steroid metabolites showed a postpartumincrease with no difference between the pregnantand the pseudo-pregnant female. Surprisingly a highlyÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Non-invasive Monitoring of Hormones 77Fig. 1. Course of fecal progesteroneand oestrogen metabolites in apregnant (a) and pseudopregnant(b) Eurasian lynx. Samples werecollected over one year beginningwith October (M: mating; P: parturition)significant (n = 310, r 2 = 0.8131, p < 0.0001)correlation between E2 and P4 metabolites wasobtained, when calculated for all females involved inthe study. Altogether, the faecal steroid metaboliteprofiles obtained here are untypical compared to temporalpatterns measured in other felids. In pregnantcheetahs and clouded leopards P4 metabolite concentrationsdropped to baseline coincident with parturitionwhereas the duration of elevated faecal P4 metabolitesduring pseudopregnancy was approximately half thatobserved during pregnancy (Brown et al. 1994). Inaddition, as shown for both female lynxes, a distinctoestrogen peak at mating time was absent.The results from the Eurasian lynx revealed thatfaecal progesterone metabolites are a poor indicator ofreproductive status in this species as they fail to showclear elevations following copulation or during pregnancy.Such unusual endocrine profile conflicts withcourses reported from other felid species (Brown et al.1994, 2001), but is similar to faecal gestagen metaboliteanalysis in the Iberian lynx (Pelican et al. 2006).Faecal oestrogen and gestagen metabolites wereunreliable for oestrus and pregnancy diagnosis inEurasian lynx, wherefore a radiometabolism study wasnecessary to determine whether our enzyme immunoassaywould ‘catch’ the relevant metabolites reflecting thebiological active hormone. In addition a biologicalvalidation of the presumed luteal activity wasdemanded. It includes the characterization of immunoreactivemetabolites during and after pregnancy, thedetermination of blood serum hormones and, finally, anultrasound examination to verify the functional activityof CL outside breeding season.Radiometabolism studyTo identify the relevant progesterone metabolites, whichreflect CL activity in the Eurasian lynx, a radiometabolismstudy was performed. For the development oftechniques for faecal steroid analysis, experiments onthe metabolism of radiolabelled steroids have provided avaluable insight into the metabolism and the excretionof hormone metabolites via faeces and urine. Afterinjection of 3 H (tritiated) labelled hormone, the excretedmetabolites of a particular steroid can be analysed byHPLC.Figure 2a shows the distribution of radiolabelledprogesterone metabolites in a faecal extract of a femaleÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

78 M Dehnhard, S Naidenko, A Frank, B Braun, F Go¨ ritz and K JewgenowFig. 2. Elution profile of radiolabelledprogesterone metabolitesand immunoreactive progesteronemetabolites in a female Eurasianlynx. Faecal extracts chosen fromdays 18 and 36 of pregnancy and14 days p.p. were subjected toHPLC separation. Arrows indicatethe elution positions of progesterone(4-pregnen-3,20-dione, P4),5a-pregnane-3,20-dione (DHP),5-pregnen-3b-ol-20-one (pregnenolon,PD), and 5a-pregnane-3bol-20-one(5a-P). For methodologicaldetails see Jewgenow et al.(2006). Radiometabolism studyand detection of immunoreactivitieswere carried out in differentanimalsEurasian lynx. When passing through the non-polar(reversed phase) column the metabolites are retainedand separated based on differences in their polarity. Theextract of female lynx faeces is composed of four majorpolar radiolabelled gestagen metabolites were detectablein fractions 6–8, 10, 12–13, and 16–17. Only minoramounts of radiolabelled progesterone metabolites weredetectable at positions of non-polar substances correspondingto progesterone and one of the other gestagenstandards. This suggests that the circulating hormoneitself is merely present in minor quantities in faeces.Usually the pattern of radiolabelled metabolites inHPLC-analysis and those of immunoreactive hormonemetabolites does not coincide due to different immunologicalcross-reactivities of used commercial,custom-made or in-house antibodies, which are directedagainst original hormones or known derivates. It isextremely important to consider that the peak seen in anHPLC immunogram does not reflect the quantity offaecal hormone metabolites but is a result of %crossreactivity of the antibody together with the amountof metabolite in a particular HPLC fraction. In contrastto that, the radioactivity within a fraction directlyreflects the quantitative amount, but cannot be relatedto any chemical structure.The progesterone immunoassay demonstrated fivedifferent immunoreactive progesterone metabolites(Fig. 2b–d). Two of them were consistent with the twomajor radiolabelled metabolites (fractions 6–8 and12–13), whereas three immunoreactive metabolitesÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Non-invasive Monitoring of Hormones 79eluting later than fraction 20 corresponded to progesterone,DHP and 5a-P, respectively. The broad-shoulderedpeak at fractions 7–9 indicates a cluster of polarconjugated steroid metabolites. Enzymatic hydrolysis(Jewgenow et al. 2006) of samples changed the elutionpattern resulting in the disappearance of conjugatedmetabolites towards an increase of free steroids (datanot shown).Figure 3a–c shows the elution profile of immunoreactiveoestrogen metabolites from the different reproductivephases (same samples as in Fig. 2). Theoestradiol assay detected four different immunoreactivemetabolites. As described for progesterone, the firstpeak at fractions 4–5 indicates polar conjugated steroidmetabolites followed by a peak of unknown metabolites.Two additional metabolites co-elute with authentic17b-oestradiol and oestrone. Due to the peak broadeningon their basis both a proportion of 17a-oestradioland an unknown metabolite eluting close to oestronecan be assumed.The partial identity of immunoreactive and radiolabelledgestagen metabolites and the detection of distinctproportions of oestrone and oestradiol allows theconclusion, that the assay systems used are able todetect the relevant faecal steroid hormone metabolitesreflecting ovarian steroid secretion. A chemical identificationof faecal metabolites is not possible, since noneof the assays used is able to identify novel steroidmetabolites. However, to design a more specific assay(antibody), particularly the identification of majormetabolites will be attainable using the sensitivity andthe selectivity of modern LC-MS instrumentation.Fig. 3. Elution profile of immunoreactiveoestrogen metabolites in afemale Eurasian lynx. Faecalextracts chosen from days 18 and36 of pregnancy and 14 days p.p.were subjected to HPLC separation(see also fig. 2). Arrows indicatethe elution positions of17b-oestradiol, 17a-oestradiol,and oestroneÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

80 M Dehnhard, S Naidenko, A Frank, B Braun, F Go¨ ritz and K JewgenowBiological validationBiological validations are aimed to demonstrate thathormone metabolite measurements reflect the physiologicalevent in question. One possibility to validate anon-invasive hormone assay is to compare and correlatefaecal metabolite levels with hormone levels in blood(Heistermann et al. 1993; Barrett et al. 2002; Capezzutoet al. 2008). Even if blood sampling contradicts the noninvasiveapproach, veterinary check-ups and treatmentsshould be used to collect blood and fresh faecal samplessimultaneously.Appropriate techniques to physiologically validatenon-invasive methods are pharmacological stimulationsor inhibitions of steroid hormone release. These methodstypically involve the administration of high doses ofreleasing hormones such as GnRH to stimulate theproduction of gonadal sex steroids (Kretzschmar et al.2004) and adrenocorticotrophic hormone to stimulatethe adrenal gland to produce corticosteroids (Wasseret al. 2000).A biological validation can also be based on dataanalysis. The measured hormone pattern in the lynx(Fig. 1) should mirror the reproductive events of afemale. The predicted hormone pattern (derived fromother related felid species) should include an oestradiolpeak around mating, elevated progesterone levels duringpregnancy, followed by a decrease towards basal after4 weeks in pseudopregnant females and a more or lessimmediate decrease to baseline prior to parturition inpregnant females.Our results, however, revealed differences from thesepredicted patterns: oestrogens did not reflect follicularactivity peaking around ovulation (mating), but werestrongly correlated to the excretion of gestagens.Therefore, we assume that faecal oestradiol immunoreactivityreflect the activity of corpora lutea. Theprogesterone profiles of Eurasian lynxes were also notin accordance to typical felid hormone patterns. Wefound elevated progesterone (and oestradiol) metabolitelevels throughout pregnancy and thereafter. However,the composition of hormone metabolites after parturitionwas different from those during pregnancy. Particularlythe relation between the progesterone metaboliteseluting between fractions 7–9 and 13–14 differed markedlyduring pregnancy and the postpartum (p.p.) period(Fig. 2b–d). Comparing the oestrogen profiles from thedifferent reproductive phases (Fig. 3a–c) the relation ofconjugates vs the fractions including the unpolarmetabolites remained relatively constant whereas therelation between 17b-oestradiol and oestrone changeddramatically in the p.p. period in favour of oestrone(Fig. 3c).This might be contributed by different hormonesources during and after pregnancy (CL, placenta, oradrenals). To confirm this, an additional validation forCL function had to be performed. In case of theEurasian lynx, we performed a transrectal ultrasoundinvestigation by which we found corpora lutea. This is inagreement with the above mentioned high p.p. progesteronevalues, altogether supporting the hypothesis of ap.p. luteal activity. At present it is uncertain whether theCL are the result from retained CL of pregnancy orfrom a p.p. ovulation. In conclusion there is still noanalytical parameter available in faeces which can beused to monitor luteal activity in the lynx. Onealternative might be an extensive metabolite screeningusing LC-MS to detect suitable pregnancy markers inthe lynx. Then, these specific markers can be measuredcombining the selectivity of high pressure liquid chromatographywith mass spectrometric detection, in orderto be able to distinguish between pregnancy andpseudopregnancy.Currently the only pregnancy-specific method is thetransabdominal ultrasonographic or radiographic imagingof the uterus (Davidson et al. 1986), but thistechnique usually requires handling and anesthesia ofthe female, potentially stressing both the queen anddeveloping offspring.A second option are proteohormones produced by theplacenta, like relaxin. The cat placenta produces largequantities of relaxin, beginning approximately 20 daysof gestation (Addiego et al. 1987). As with dogs, relaxinhas not been detected in the serum of cycling orpseudopregnant cats (Stewart and Stabenfeldt 1985).However, the development of a urine-based relaxinpregnancy test would prove extremely useful for breedingmanagement of wildlife species. A radioimmuneassay (RIA) for relaxin was recently validated for domescat urine. Urinary relaxin was first detected betweendays 14 and 21 of gestation, whereas the levels peaked at42–49 days, followed by a decline during the last2 weeks prior to parturition (de Haas van Dorsser et al.2006). A similar urinary relaxin profile was demonstratedin the leopard (de Haas van Dorsser et al. 2006).These results indicate that measurement of urinaryrelaxin might turn out to be a reliable method forpregnancy determination in felids from as early as 3–4 weeks of gestation. First results analysing urinaryrelaxin in the Iberian lynx showed that relaxin is anappropriate analyte to differentiate between pregnantand pseudopregnant females particularly during thesecond half of pregnancy (BC Braun et al., unpublisheddata). Urinary relaxin-based pregnancy diagnosis mayprove useful in the breeding management of other felidspecies, and provides a foundation for future studies onpregnancy in captive exotic felids.Perhaps one of the greatest uses of steroid metabolitemonitoring will be in assisting reproductive managementby identification of ovarian cycle, oestrous, andpregnancy of females. Early pregnancy diagnosis, however,is complicated by the phenomenon of pseudopregnacnyin carnivoes. Particularly in felids, steroidanalyses did not provide an early pregnancy-specificdiagnosis. In the lynx, it is even impossible to use faecalP4 and estrogen metabolite analyses as an index ofpregnancy. However, there are evidences that theplacenta might contribute to steroid biosynthesis.Therefore, LC-MS based metabolite screenings comparingpregnant and pseudopregnant females offer anoption to identify pregnancy specific placental steroids.Their identification might allow the development of aspecific assay whose cross-species applicability had to beproved for practice in other felid species.Alternatively, sensitive urinary relaxin assays mightbe used to confirm pregnancy in the lynx. For theÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Non-invasive Monitoring of Hormones 81management of felids in zoos, urinary relaxin analysiswill be an important research tool supposed urinecollections could be arranged. Due to species specificbarriers further research is necessary to assess whetherurinary relaxin could be used to diagnose pregnancyin other felids. For free-ranging populations,however, techniques based on faecal samples areessential for studying animals within their naturalhabitat.ReferencesAddiego LA, Tsutsui T, Stewart DR, Stabenfeldt GH, 1987:Determination of the source of immunoreactive relaxin inthe cat. Biol Reprod 37, 1165–1169.Asa CS, Bauman K, Callahan P, Bauman J, Volkmann DH,Jo¨ chle W, 2006: GnRH agonist induction of fertile estruswith either natural mating or artificial insemination, followedby birth of pups in gray wolves (Canis lupus).Theriogenology 66, 1778–1782.Barrett GM, Shimizu K, Bardi M, Mori A, 2002: Fecaltestosterone immunoreactivity as a non-invasive index offunctinal testosterone dynamics in male Japanese macaques(Macaca fuscata). Primates 43, 29–39.Brown JL, Wasser SK, Wildt DE, Graham LH, 1994:Comparative aspects of steroid hormone metabolism andovarian activity in felids, measured noninvasively in faeces.Biol Reprod 51, 776–786.Brown JL, Terio KE, Graham LH, 1996: Faecal androgenmetabolite analysis for non invasive of testicular steroidogenicactivity in felids. Zoo Biol 15, 425–434.Brown JL, Graham LG, Wielebnowski N, Swanson WF,Wildt DE, Howard JG, 2001: Understanding the basicreproductive biology of wild felids by monitoring of faecalsteroids. J Reprod Fertil Suppl 57, 71–82.Capezzuto A, Chelini MOM, Felippe ECG, Oliveira CA, 2008:Correlation between serum and fecal concentrations ofreproductive steroids throughout gestation in goats. AnimReprod Sci 103, 78–86.Cardillo M, Mace GM, Gittleman JL, Purvis A, 2006: Latentextinction risk and the future battlegrounds of mammalconservation. Proc Natl Acad Sci U S A 103, 4157–4161.Cavigelli SA, Dubovick T, Levash W, Jolly A, Pitts A, 2003:Female dominance status and faecal corticoids in a cooperativebreeder with low reproductive skew: ring-tailedlemurs (Lemur catta). Horm Behav 43, 166–179.Ceballos G, Ehrlich PR, Sobero´ n J, Salazar I, Fay JP, 2005:Global mammal conservation: what must we manage?Science 309, 546–547.Davidson AP, Nyland TG, Tsutsui T, 1986: Pregnancydiagnosis with ultrasound in the domestic cat. Vet Radiol27, 109–114.Earnhardt JM, Thompson SD, Schad K, 2004: Strategicplanning for captive populations: projecting changes ingenetic diversity. Anim Conserv 7, 9–16.Go¨ ritz F, Quest M, Hildebrandt TB, Meyer HHD, Kolter L,Jewgenow K, 2001: Antiprogestins – a new approach tocontrol reproduction in captive bears. J Reprod Fertil Suppl57, 249–254.Graham LH, Goodrowe KL, Raeside JI, Liptrap RM, 2005:Non-invasive monitoring of ovarian function in several felidspecies by measurement of faecal estradiol-17 and progestins.Zoo Biol 14, 223–237.Graham LH, Byers AP, Armstrong DL, Loskutoff NM,Swanson WF, Wildt DE, Brown JL, 2006: Natural andgonadotropin-induced ovarian activity in tigers (Pantheratigris) assessed by fecal steroid analyses. Gen Comp Endocrinol147, 362–370.de Haas van Dorsser FJ, Swanson WF, Lasano S, Steinetz BG,2006: Development, validation, and application of a urinaryrelaxin radioimmunoassay for the diagnosis and monitoringof pregnancy in felids. Biol Reprod 74, 1090–1095.Heistermann M, Tari S, Hodges JK, 1993: Measurement offaecal steroids for monitoring ovarian function in NewWorld primates, Callitrichidae. J Reprod Fertil 99, 243–251.Heistermann M, Ademmer C, Kaumanns W, 2004: Ovariancycle and effect of social changes on adrenal and ovarianfunction in Pygathrix nemaeus. Int J Primatol 25, 689–708.Heistermann M, Palme R, Ganswindt A, 2006: Comparison ofdifferent enzyme immunoassays for assessment of adrenocorticalactivity in primates based on fecal analysis. Am JPrimatol 68, 257–273.Hermes R, Go¨ ritz F, Maltzan J, Blottner S, Proudfoot J,Fritsch G, Fassbender M, Quest M, Hildebrandt TB, 2001:Establishment of assisted reproduction technologies infemale and male African wild dogs (Lycaon pictus). JReprod Fertil Suppl 57, 315–321.Hermes R, Go¨ ritz F, Streich WJ, Hildebrandt TB, 2007:Assisted reproduction in female rhinoceros and elephants –current status and future perspective. Reprod Domest Anim42(Suppl. 2), 33–44.Hildebrandt TB, Hermes R, Walzer C, So´ s E, Molnar V,Mezo¨ si L, Schnorrenberg A, Silinski S, Streich J, SchwarzenbergerF, Go¨ ritz F, 2007: Artificial insemination in theanoestrous and the postpartum white rhinoceros usingGnRH analogue to induce ovulation. Theriogenology 67,1473–1484.Jewgenow K, Naidenko SV, Goeritz F, Vargas A, DehnhardM, 2006: Monitoring testicular activity of male Eurasian(Lynx lynx) and Iberian (Lynx pardinus) lynx by fecaltestosterone metabolite measurement. Gen Comp Endocrinol149, 151–158.Kretzschmar P, Gansloßer U, Dehnhard M, 2004: Relationshipbetween androgens, environmental factors and reproductivebehavior in male white rhinoceros (Ceratotheriumsimum simum). Horm Behav 45, 1–9.Masui M, Hiramatsu H, Nose N, Nakasato R, Sagawa Y,Tajima H, Saito K, 1989: Successful artificial inseminationin the giant panda (Ailuropoda melanoleuca) at Ueno Zoo.Zoo Biol 8, 17–26.Mesochina P, Bedin E, Ostrowski S, 2003: Reintroducingantelopes into arid areas: lessons learnt from the oryx inSaudi Arabia. CR Biol 326(Suppl. 1), S158–S165.Meyer HHD, Rohleder M, Streich WJ, Go¨ ltenboth R, Ochs A,1997: Sexualsteroidprofile und Ovaraktivita¨ ten des PandaweibchensYAN YAN im Berliner Zoo. Berl Mu¨ nchTiera¨ rztl Wschr 110, 143–147.Mooring MS, Patton ML, Lance VA, Hall BM, Schaad EW,Fortin SS, Jella JE, McPeak KM, 2004: Fecal androgens ofbison bulls during the rut. Horm Behav 46, 392–398.Moreira N, Monteiro-Filho ELA, Moraes W, Swanson WF,Graham LH, Pasquali OL, Gomes MLF, Morais RN, WildtDE, Brown JL, 2001: Reproductive steroid hormones andovarian activity in felids of the Leopardus genus. Zoo Biol20, 103–116.Palme R, Fischer P, Schildorfer H, Ismail MN, 1996:Excretion of infused 14C-steroid hormones via faeces andurine in domestic livestock. Anim Reprod Sci 43, 43–63.Paris MCJ, White A, Reiss A, West M, Schwarzenberger F,2002: Faecal progesterone metabolites and behaviouralobservations for the non-invasive assessment of oestrouscycles in the common wombat (Vombatus ursinus) and thesouthern hairy-nosed wombat (Lasiorhinus latifrons). AnimReprod Sci 72, 245–257.Pelican KM, Wildt DE, Pukazhenthi B, Howard J, 2006:Ovarian control for assisted reproduction in the domesticcat and wild felids. Theriogenology 66, 37–48.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

82 M Dehnhard, S Naidenko, A Frank, B Braun, F Go¨ ritz and K JewgenowPenfold LM, Monfort SL, Wolfe BA, Citino SB, Wildt DE,2005: Reproductive physiology and artificial inseminationstudies in wild and captive gerenuk (Litocranius walleriwalleri). Reprod Fertil Dev 17, 707–714.Pereira RJG, Polegato BF, De Souza S, Negrao JA, DuarteJMB, 2006: Monitoring ovarian cycles and pregnancy inbrown brocket deer (Mazama gouazoubira) by measurementof fecal progesterone metabolites. Theriogenology 65, 387–399.Pukazhenthi B, Laroe D, Crosier A, Bush LM, Spindler R,Pelican KM, Bush M, Howard JG, Wildt DE, 2006:Challenges in cryopreservation of clouded leopard (Neofelisnebulosa) spermatozoa. Theriogenology 66, 1790–1796.Ryder O, 1993: The Przewalski’s horse: prospects for reintroductioninto the wild. Conserv Biol 7, 13–15.Scheibe KM, Dehnhard M, Meyer HHD, Scheibe A, 1999:Noninvasive monitoring of reproductive function by determinationof faecal progestagens and sexual behaviour in aherd of Przewalski mares in a semireserve. Acta Theriol 44,451–463.Schwarzenberger F, 2007: The many uses of non-invasivefaecal steroid monitoring in zoo and wildlife species. Int ZooYB 41, 52–74.Stewart DR, Stabenfeldt GH, 1985: Relaxin activity in thepregnant cat. Biol Reprod 32, 848–854.Stoops MA, Bond JB, Bateman HL, Campbell MK, LevensGP, Bowsher TR, Ferrell ST, Swanson WF, 2007: Comparisonof different sperm cryopreservation procedures on postthawquality and heterologous in vitro fertilisation success inthe ocelot (Leopardus pardalis). Reprod Fertil Dev 19, 685–694.Swanson WF, 2006: Application of assisted reproduction forpopulation management in felids: the potential and realityfor conservation of small cats. Theriogenology 66, 49–58.Swanson WF, Magarey GM, Herrick JR, 2007: Spermcryopreservation in endangered felids: developing linkageof in situ-ex situ populations. Soc Reprod Fertil Suppl 65,417–432.Wasser SK, Hunt KE, Brown JL, Cooper K, Crockett CM,Bechert U, Millspaugh JJ, Larson S, Monfort SL, 2000: Ageneralized fecal glucocorticoid assay for use in a diversearray of nondomestic mammalian and avian species. GenComp Endocrinol 120, 260–275.Author’s address (for correspondence): M Dehnhard, Leibniz-Institutefor Zoo Biology and Wildlife Research, PF 601103, D-10252 Berlin,Germany. E-mail: dehnhard@izw-berlin.deConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 83–88 (2008); doi: 10.1111/j.1439-0531.2008.01146.xISSN 0936-6768Reproductive Biotechnology and Gene Mapping: Tools for Conserving Rare Breedsof LivestockJA LongAnimal Biosciences and Biotechnology Laboratory, Beltsville Agricultural Research Center, US Department of Agriculture, Beltsville, MD, USAContentsToday’s livestock diversity originated from the wild ancestorspecies and was subsequently shaped through the processes ofmutation, genetic drift, and natural and human selection. Onlya subset of the diversity present in the ancestral speciessurvives in the domestic counterparts. A 2007 report releasedby UN Food and Agriculture Organization ‘The State of theWorld’s Animal Genetic Resources’, compiled from surveysconducted in 169 countries, found that nearly 70% of theworld’s remaining livestock breeds live in developing countries.The UN report was presented to more than 300 policy makers,scientists, breeders, and livestock keepers at the First InternationalTechnical Conference on Animal Genetic Resources,held in September 2007 in Interlaken, Switzerland. Theconference aims were to adopt a global plan of action forconserving animal genetic resources as its main outcome. Inthis paper, the current and potential contributions of reproductiveand molecular biotechnology are considered as tools ofconserving rare breeds of livestock.IntroductionOf the 50 000 known mammalian and avian species, arelatively small proportion has been domesticated.Approximately 40 livestock species, shaped by a longhistory of domestication and development, contribute totoday’s agriculture and food production. Livestockbreed development has been and continues to be adynamic process of genetic change driven by selectionpressures, including environmental factors and humanintervention through controlled breeding and husbandry,which has resulted in a great variety ofgenetically distinct breeds. The livestock breeds developedover thousands of years has, until recently, causeda net increase in genetic diversity over time. During thepast 100 years, however, there has been a net loss ofdiversity because of an increased rate of extinction oflivestock breeds and varieties [UN Food and AgricultureOrganization (FAO) (2007)]. The number of breedslost over the past 8 years is rapidly approaching the rateof extinction that occurred from 1900 to 1999 (Table 1).Losses have been accelerated by the rapid intensificationof livestock production, a failure to evaluate localbreeds, and inappropriate breed replacement or crossbreedingfacilitated by the availability of high performingbreeds (FAO 2007). As an example of inappropriatebreed replacement, Uganda’s indigenous, droughthardyAnkole cattle could face extinction within20 years because they are being rapidly supplanted byHolstein-Friesians, a breed which produces much moremilk. During a recent drought, however, farmers whohad Ankole cattle were able to walk them long distancesto reach water sources while those who had traded theAnkole for imported breeds lost entire herds. Thephysiology and grazing behaviour of the importedbreeds typically are not adapted to the natural pasturesand climate of Africa, especially when drought strikes(Kay 1997). In another example, cross-breeding hasalmost decimated purebred populations of the EastAfrican Red Maasai sheep, which is renowned for itsdisease resistance to gastrointestinal parasites and highproductivity under extremely challenging environments.In the mid-1970s, as a result of a subsidized disseminationprogram, many farmers in Kenya cross-bred theirRed Maasai flocks with the less-hardy Dorpers sheep,which subsequently proved unsuitable in many productionareas. In 1992, the International LivestockResearch Institute (ILRI) undertook an extensive searchin Kenya and northern parts of the United Republic ofTanzania, and was only able to locate a very smallnumber of purebred animals, which later showed somelevels of genetic contamination (Gibson and Candiff2000).The extinction of a breed or population means theloss of its unique adaptive attributes, which are underthe control of many interacting genes and are the resultsof complex interactions between the genotype and theenvironment. Figure 1 illustrates the current status oflivestock breeds (FAO 2007). The regions with thehighest proportion of their breeds classified as at risk areEurope (28% of mammalian breeds; 49% of avianbreeds) and North America (20% of mammalian breeds;79% of avian breeds). Europe and North America arethe regions that have the most highly specializedlivestock industries, in which production is dominatedby a small number of breeds. In recent years, many ofthe world’s small farmers have abandoned their traditionalanimals in favour of higher yielding stockimported from Europe and the USA. For example, in1994 local breeds comprised 72% of the sow populationin northern Vietnam; within 8 years, however, this figurehad dropped to just 26%. Of the country’s 14 local pigbreeds, five are now listed as vulnerable, two areconsidered critical state and three are facing extinction.The 2007 report, ‘The State of the World’s AnimalGenetic Resources’, compiled by the FAO, with contributionsby the ILRI and other research groups,surveyed farm animals in 169 countries. Nearly 70%of the entire world’s remaining unique livestock breedsare found in developing countries, which as describedabove, are at risk from the importation and farming ofexogenous livestock breeds. Renowned organizationssuch as the ILRI and FAO have spearheaded the firstimportant step for conservation by conducting livestockÓ 2008 No claim to original government works

84 JA LongTable 1. Livestock breed extinction rates over timeYear span Number of breeds Percent of breedBefore 1900 15 21900–1999 111 16After 1999 62 9Unspecified a 502 73Total 690 100a Unspecified = no year of extinction indicated. Adapted from FAO (2007).breed surveys. These surveys are time-consuming andlogistically complex; however, ILRI scientists havedeveloped and integrated the Domestic Animal GeneticResources Information System with the FAO’s DomesticAnimal Diversity Information System to streamlinethe process. These web-based information systemsprovide a means for curation and dissemination ofvaluable information that will support development ofconservation priorities, as well as provide a database foruse of reproductive and molecular biotechnologies tomaintain valuable genetic resources.Impact of Reproductive Biotechnology onLivestock ConservationModern reproductive biotechnologies, such as artificialinsemination, embryo transfer, in vitro fertilization,gamete ⁄ embryo micromanipulation, semen sexing, genomeresource banking and somatic cell nuclear transfer(cloning) have enormous potential for conserving rarebreeds of livestock. The advent of artificial inseminationin the 1940s with cattle and semen cryopreservation inthe 1950s with poultry was instrumental in the successfultransfer of genetic material between and among livestockpopulations and breeds. There are several anecdotaland published reports of applying thesetechnologies to conserving rare livestock breeds. TheRare Breeds Program in Colonial Williamsburg, VA(USA) makes use of semen cryopreservation for livestockbreeds dating from the 17th and 18th centuries,including Devon cattle, Leicester sheep, Ossabaw pigsand American Cream horses, although actual numbersof offspring from artificial insemination with frozen⁄ thawed semen have not been documented to date.The Hamilton Rare Breeds Foundation in Hartland, VT(USA) has pioneered in the use of frozen semen in thePoitou Donkey, an ancient breed dating back over2000 years, and is the first group to have produced foalsfrom the rarest breed of donkey in existence today.Embryo transfer, pioneered in agricultural species inthe 1930s, also has been reported in use with heritagelivestock breeds. In collaboration with the Swiss VillageFarm (SVF) Foundation in Newport, RI (USA),frozen ⁄ thawed embryos from the Tennessee Myotonicor fainting goat breed were surgically transferred intotwo surrogate Nubian does (a common domestic breed)and resulted in the birth of one healthy buck (Matsaset al. 2005). In 2006, the SVF Foundation reported asecond birth from interspecies embryo transfer: anendangered Gulf Coast lamb born to a Santa Cruzewe. The Gulf Coast sheep provides a good example forthe importance of preserving the unique genetic attributesof heritage livestock breeds, as the Gulf Coast issheep is extremely resistant to parasites and nearlyimpervious to the foot rot that plagues many other ovinebreeds.One of the most promising areas of reproductivebiotechnology is the creation of genetic resource banksas a conservation tool for rare livestock breeds. Theconcept of banking gametes, embryos and DNA materialfor conservation purposes is not new, as the idea hasRisk status of the world's mammalian breeds in January 2006: absolute (table) and percentage (chart) figures by region100%80%60%40%20%0%Africa Asia Europe Latin americaNear & middleEastNorth americaSouthwestpacificInternationalTransboundrybreedsUnknown 384 469 459 304107 79 8058 1940Critical 13 23182 9012 97 255Critical-maintained 0 451 4 000059Endangered 26 50249 21622 11 22 407Endangered-maintained 4 3142 9011 0160Extinct 35 45 481 21 549 6 1643Not-at-risk 187 776 664 81 85 13 17 312 2135WorldFig. 1. Proportion of the world’s breeds by risk status category. Adapted from FAO (2007)Ó 2008 No claim to original government works

Biotechnology Methods for Preserving Rare Livestock 85Table 2. Rare and historic breed livestock genetic resource banksNameAvian Resource CenterCenter for Genetic ResourcesNational Bureau of Animal Genetic ResourcesINRA’s National CryobankInternational Livestock Research InstituteLivestock Research InstituteNordic Genebank for Farm AnimalsRare Breeds Gene BankSwiss Village Farm FoundationUSDA’s National Animal Germplasm ProgramCountryCanada ⁄ British ColumbiaThe NetherlandsIndiaFranceKenyaTaiwanNorwayNew ZealandUSAUSAbeen widely discussed for use in preserving endangeredwildlife populations (Wildt 1992, 2000; Long et al. 1996;Wildt et al. 1997; Holt and Pickard 1999; Andrabi andMaxwell 2007). Table 2 lists the pre-dominant germplasm⁄ genetic repositories for rare livestock breedsaround the world. One notable success story thatillustrates the potential benefits of genetic resourcebanks involves the Dutch Friesian cattle breed. In1879, the cattle population in the province of Frieslandconsisted mainly of Red Pied cattle registered as a redand white phenotype in the Friesian Cattle herd book.Black and white cattle progressively became morepopular than the original red and white; by 1970, therewere only 50 farmers registered as owning a total of2500 Red and White Friesian cattle. The sustainedimport of Holstein-Friesians from the United States andCanada further eroded the population, to where only 21Red and White cattle (4 males and 17 females) remainedin 1993. A group of owners started the Foundation forNative Red and White Friesian Cattle and, in collaborationwith the Nordic Genebank for Animals, developeda breeding program. Frozen ⁄ thawed semen thatwas preserved the 1970s and 1980s and subsequentlystored in the Genebank was used to breed females undera contract system. Resulting male progeny were raisedby breeders, and semen from these males was collected,frozen and later used under new contracts. The breedincreased in number, reaching 256 registered livingfemales and 12 living males in 2004. Currently, a total of11 780 semen doses from 43 bulls are stored in theGenebank and kept available for artificial insemination(FAO 2007). Another example of successful usage ofsemen from storage repositories involves the endangeredGauloise dore´e chicken, the oldest patrimonial poultrybreed in France. Using frozen ⁄ thawed semen and anintensive breeding program, current stocks have proventhe restoration of more than 96% of the initial genome(Blesbois et al. 2007). This point is particularly importantfor avian breeds, as neither the female gamete norembryo has not been successfully cryopreserved and,unlike mammals where the male gamete determinesgender, birds have a ZZ male ⁄ ZW female sex-determiningsystem.More recent reproductive biotechnologies such assomatic cell nuclear cloning also have enormouspotential for conserving rare breeds of livestock.Another rare breed success story involves the cloningof the Enderby Island Cow, the last survivor of theworld’s rarest cattle breed. In 1992, members of NewZealand’s Rare Breeds Conservation Society found freshhoof-prints of two cattle on Enderby Island. Fivemonths later, the world’s only surviving Enderby Islandcow and her heifer calf were captured. Unfortunately,the calf subsequently died of unknown causes, leavingthe cow as the only survivor of her breed in the world. Atotal of 35 embryo transfers were conducted, andresulted in the birth of a single male calf. In a lasteffort to save the breed, somatic cell nuclear transfer wasused to produce heifer clones from the cow (Wells et al.1998). To date, two of the three surviving clones havesince given natural birth to two heifer calves.Somatic cell cloning also was used to produce liveoffspring from the rare European mouflon sheep, abreed found on Sardinia, Corsica and Cyprus wherethere is thought to be fewer than 1000 mature individualsin the wild. Loi et al. (2001) injected enucleatedsheep oocytes from a closely-related domestic breed withsomatic granulosa cells recovered from the ovaries oftwo adult female mouflons found dead in the pasture.Blastocyst-stage cloned embryos were transferred intosheep foster mothers and two pregnancies were established,one of which produced an apparently normalmouflon lamb. What is remarkable about this example isthat although the nuclear donor cells were recoveredfrom dead animals and considered non-viable, thesepost-mortem cells were able to generate normal embryosand offspring. This example supports the use of cloningfor the expansion of critically endangered populations,both within a concerted conservation program and inextreme situations involving sudden death (Loi et al.2001). Despite these reports of the positive impact ofreproductive biotechnology on the conservation of rarelivestock breeds, there are too few examples of artificialinsemination, germplasm cryopreservation, or embryotransfer being used in conjunction with rare livestock.Application of Molecular Biotechnology Toolsfor Livestock ConservationAt the molecular level, the genetic diversity presentwithin a livestock species is a reflection of differences inDNA sequences, or allelic diversity, across the functionalDNA regions, or genes affecting animal developmentand performance. The complete and partialsequencing of major livestock genomes (chicken, 2004;bovine, 2005; rabbit, 2006; pig, 2007) provides a wealthof information useful for many aspects of livestockbreed conservation from identifying ancestral breeds tounderstanding disease resistance.Gene mapping has been used to as a tool tounderstand livestock origin and diversity in severallivestock species. For example, 5 distinct maternalmitochondrial major lineages have been identified indomestic goats (Luikart et al. 2001; Sultana et al. 2003;Joshi et al. 2004); while the Asian mouflon is purportedto be the only progenitor of domestic sheep (Hiendlederet al. 1998). The ancestor of the domestic pig is the wildboar (Sus scrofa), with at least 16 distinct subspecies ofwild boar have been described in Eurasia and NorthAfrica. A recent survey of mitochondrial DNA diversityamong Eurasian domestic pigs and wild boar revealed acomplex picture of pig domestication, with at least fiveÓ 2008 No claim to original government works

86 JA Longor six distinct centres across the geographical range ofthe wild species (Larson et al. 2005).Domestication of cattle has been particularly welldocumented through gene mapping, with clear evidenceof three distinct initial domestication events for threedistinct aurochs (Bos primigenius) subspecies. Bos primigeniusprimigenius and B. p. opisthonomous, are theancestors of the humpless B. taurus cattle of the NearEast and Africa, respectively, with domestication occurringapproximately 9000 years ago (Wendorf and Schild1994). Humped Zebu cattle (B. indicus) are believed tohave been domesticated at a later date, approximately7000–8000 years ago (Loftus et al. 1994; Bradley et al.1996; Bradley and Magee 2006). Finally, the domesticchicken (Gallus domesticus) is descended from the wildred jungle fowl (G. gallus). While previous molecularstudies suggested a single domestic origin in SoutheastAsia (Fumihito et al. 1994, 1996), at least six distinctmaternal genetic lineages have now been identified (Liuet al. 2006).In genetic diversity studies, the most frequently usedmarkers are microsatellites and these are the mostpopular markers in livestock genetic characterizationstudies (Sunnucks 2001). Their high mutation rate andco-dominant nature permit the estimation of within- andbetween-breed genetic diversity, and genetic admixtureamong closely related breeds. There are a few examplesof large-scale analyses of the genetic diversity oflivestock species. For example, chicken and pig diversitythroughout Europe have been reported (Hillel et al.2003; SanCristobal et al. 2006). Sheep diversity wasassessed at a large regional scale in northern Europeancountries (Tapio et al. 2005); while Can˜ on et al. (2006)studied goat diversity in Europe and the Middle East.Probably the most comprehensive study of this type inlivestock is a continent-wide study of African cattle(Hanotte et al. 2002), which revealed the genetic signaturesof the origins, secondary movements and differentiationof African cattle. For most livestock breeds,however, a comprehensive review is still lacking.Single nucleotide polymorphisms (SNPs) are used asan alternative to microsatellites in genetic diversitystudies (Marsjan and Oldenbroek 2007). Single nucleotidepolymorphisms are variations at single nucleotideswhich do not change the overall length of theDNA sequence in the region and occur throughoutthe genome. With this perspective, large-scale projectsare ongoing in several livestock species to identifymillions (Wong et al. 2004) and validate severalthousands of SNPs, and identify haplotype blocks inthe genome.Mitochondrial DNA (mtDNA) polymorphisms havebeen extensively used in phylogenetic and geneticdiversity analyses. The haploid mtDNA, carried by themitochondria in the cell cytoplasm, has a maternal modeof inheritance (individuals inherit the mtDNA fromtheir dams and not from their sires) and a high mutationrate; it does not recombine. These characteristics enablebiologists to reconstruct evolutionary relationshipsbetween and within species by assessing the patterns ofmutations in mtDNA. MtDNA markers may alsoprovide a rapid way of detecting hybridization betweenlivestock species or subspecies (Nijman et al. 2003).An alternative approach to the identification ofgenome regions carrying relevant genes has recentlybeen proposed. It consists of the detection of ‘selectionsignatures’ via a ‘population genomics’ approach (Blacket al. 2001; Luikart et al. 2003). Population genomicsutilizes phenotypic data at the breed level (or subpopulationswithin a breed), rather than at the individuallevel. The population genomics approach also canidentify genes subjected to strong selection pressureand eventually fixed within breeds and, in particular,genes involved in adaptation to extreme environmentsor disease resistance. Population genomics relies on theprinciple that loci across the genome are influenced bygenome-wide evolutionary forces (e.g. genetic drift, geneflow), whereas locus-specific forces, such as selection,imprint a particular pattern of variability on linked locionly (Luikart et al. 2003). By comparing the geneticdiversity of many loci across the genome, it is thenpossible to reveal loci displaying an atypical variationpattern, which are likely to be linked to those genomicregions affected by selection (Black et al. 2001). Therefore,in contrast to candidate-gene-based methods,strategies making use of population genomics do notfocus on a few loci only, but rather depict the effect ofselection over the whole genome (Storz 2005).Another new frontier emerging from the concept ofpopulation genomics is landscape genomics. Livestockby definition are adapted to the landscape (e.g. temperature,altitude, rainfall, disease challenge, nutritionalchallenge and human selection). The aim of landscapegenomics is to learn from the co-evolution of livestockand production systems and use the knowledge gainedto better match different breeds with production circumstances.A novel approach for evaluating populationgenomics is based on a spatial analysis methoddesigned to detect signatures of natural selection withinthe genome of domestic and wild animals (Joost et al.2007). Spatial analysis method goes a step furthercompared to classical approaches, as it is designed toidentify environmental parameters associated withselected markers (FAO 2007). By overlaying populationgenomic analyses (e.g. ‘signatures of selection’) withother sets of information such as agro-ecological mapsor other environmental ⁄ production information, it canbe determined what genetic materials are candidates foruse in which parts of the globe. The concept oflandscape genomics is promising, as this combines georeferencingof breed distributions, spatial ⁄ global geneticdiversity, climatic, ecological, epidemiological and productionsystem information which will facilitate anddirect priority decisions for breed conservation.Future Challenges and OpportunitiesLack of information on the world’s livestock resources,such as what livestock breeds ⁄ populations exist, theirgeographical location and their genetic characteristics, isa major impediment to their sustainable use. The currentdocumented numbers of breeds is likely an underestimation,as a large proportion of indigenous livestockpopulations are in the developing world and have yetto be described at phenotypic and genotypic levels(Hanotte and Jianlin 2005). Additionally, the geneticÓ 2008 No claim to original government works

Biotechnology Methods for Preserving Rare Livestock 87characterization of all remaining wild ancestral populationsand closely related species is critical as these are theonly remaining sources of putative alleles of economicvalues that might have been lost during domesticationevents. Moreover, the development and use of reproductivebiotechnology, particularly genetic resourcebanks, is critical for the preservation and managementof the remaining agricultural resources. There is a largegap between developed and developing countries in theability to use reproductive and molecular biotechnologyfor setting and maintaining conservation priorities. Therecent International Technical Conference on AnimalGenetic Resources was a timely event that presentedmany areas for global concern and provided leadershipfor setting conservation priorities. It is particularlyimportant to conserve the current livestock geneticresources because the ancestors of most of our existinglivestock species no longer exist. Genetically diverselivestock populations provide a greater range of optionsfor meeting future challenges, whether associated withenvironmental change, emerging disease threats, newknowledge of human nutritional requirements, fluctuatingmarket conditions or changing societal needs.ReferencesAndrabi S, Maxwell W, 2007: A review on reproductivebiotechnologies for conservation of endangered mammalianspecies. Anim Reprod Sci 99, 223–243.Black WC, Baer CF, Antolin MF, DuTeau NM, 2001:Population genomics: genome-wide sampling of insectpopulations. Annu Rev Entomol 46, 441–469.Blesbois E, Seigneurin F, Grasseau I, Limouzin C, Besnard J,Gourichon D, Coquerelle G, Rault P, Tixier-Boichard M,2007: Semen cryopreservation for ex situ management ofgenetic diversity in chicken: creation of the French aviancryobank. Poultry Sci 86, 555–564.Bradley DG, Magee D, 2006: Genetics and the origins ofdomestic cattle. In: Zeder MA, Emshwiller E, Smith BD,Bradley DG (eds), Documenting Domestication: NewGenetics and Archaeological Paradigm. University of CaliforniaPress, California, U S A, pp. 317–328.Bradley DG, MacHugh DE, Cunningham P, Loftus RT, 1996:Mitochondrial DNA diversity and the origins of African andEuropean cattle. Proc Natl Acad Sci U S A 93, 5131–5135.Can˜ on J, Garcıa D, Garcıa-Atance MA, Obexer-Ruff G,Lenstra JA, Ajmone-Marsan P, Dunner S, 2006: TheECONOGENE Consortium. Geographical partitioning ofgoat diversity in Europe and the Middle East. Anim Genet37, 327–334.FAO, 2007: Multiple chapters. In: Rischkowsky B, Pilling D(eds), The State of the World’s Animal Genetic Resourcesfor Food and Agriculture. FAO, Rome.Fumihito A, Miyake T, Sumi S, Takada M, Ohno S, KondoN, 1994: One subspecies of the red junglefowl (Gallus gallusgallus) suffices as the matriarchic ancestor of all domesticbreeds. Proc Natl Acad Sci U S A 91, 12505–12509.Fumihito A, Miyake T, Takada M, Shingu R, Endo T,Gojobori T, Kondo N, Ohno S, 1996: Monophyletic originand unique dispersal patterns of domestic fowls. Proc NatlAcad Sci U S A 93, 6792–6795.Gibson JP, Candiff LV 2000: Developing straight-breedingand cross-breeding structures for extensive grazing systemswhich utilize exotic animal genetic resources. ICAR TechnicalSeries No. 3. In: Galal S, Boyazoglu J, Hammond K(eds), Proceedings of Workshop on Developing BreedingStrategies for Low Input Animal Production Environments.ICAR, Villa del Ragno, Via Nomentana, Rome Italy, pp.207–241.Hanotte O, Bradley DG, Ochieng JW, Verjee Y, Hill EW,2002: African pastoralism: genetic imprints of origins andmigrations. Science 296, 336–339.Hiendleder S, Mainz K, Plante Y, Lewalski H, 1998: Analysisof mitochondrial DNA indicates that the domestic sheep arederived from two different ancestral maternal sources: noevidences for the contribution from Urial and Argali sheep.J Heredity 89, 113–120.Hillel J, Groenen MA, Tixier-Boichard M, Korol AB, DavidL, Kirzhner VM, Burke T, Barre-Dirie A, Crooijmans RP,Elo K, Feldman MW, Freidlin PJ, 2003: Biodiversity of 52chicken populations assessed by microsatellite typing ofDNA pools. Genet Sel Evol 35, 533–557.Holt WV, Pickard AR, 1999: Role of reproductive technologiesin genetic resource banks in animal conservation. RevReprod 4, 143–150.Joost S, Bonin A, Bruford MW, Despres L, Conord C,Erhardt G, Taberlet P, 2007: A spatial analysis method(SAM) to detect candidate loci for selection: towards alandscape genomics approach to adaptation. Mol Ecol 16,3955–3969.Joshi MB, Rout PK, Mandal AK, Tyler-Smith C, Singh L,Thangaray K, 2004: Phylogeography and origins of Indiandomestic goats. Mol Biol Evol 21, 454–462.Kay RNB, 1997: Responses of African livestock and wildherbivores to drought. J Arid Environ 37, 683–694.Larson G, Dobney K, Albarella U, Fang M, Matisoo-Smith E,Robins J, Lowden S, Finlayson H, Brand T, Willerslev E,Rowley-Conwy P, Andersson L, Cooper A, 2005: Worldwidephylogeography of wild boar reveals multiple centersof pig domestication. Sci 307, 1618–1621.Liu YP, Wu GS, Yao YG, Miao YW, Luikart G, Baig M,Beja-Pereira A, Ding ZL, Palanichamy MG, Zhang YP,2006: Multiple maternal origins of chickens: out of theAsian jungles. Mol Phylogene Evol 38, 12–19.Loftus RT, MacHugh DE, Bradley DG, Sharp PM, CunninghamP, 1994: Evidence for two independent domesticationof cattle. Proc Natl Acad Sci U S A 91, 2757–2761.Loi P, Ptak G, Barboni B, Fulka J, Cappai P, Clinton M, 2001:Genetic rescue of an endangered mammal by cross-speciesnuclear transfer using post-mortem somatic cells. NatureBiotechnol 19, 962–964.Long JA, Larson SE, Wasser SK, 1996: Safeguarding diversity:challenges in developing a genome resource bank forCalifornia sea otters. Endangered Species Update 13, 57–60.Luikart GL, Gielly L, Excoffier L, Vigne JD, Bouvet J,Taberlet P, 2001: Multiple maternal origins and weakphylogeographic structure in domestic goats. Proc NatlAcad Sci U S A 98, 5927–5930.Marsjan PA, Oldenbroek JK2007: Molecular markers, a toolfor exploring genetic diversity. In: Rischkowsky B, Pilling D(eds), The State of the World’s Animal Genetic Resourcesfor Food and Agriculture, Section C, Part 4. FAO, Rome,pp. 359–379.Matsas D, Huntress V, Levine H, Ayres S, Amini J, Duby R,Borden P, Saperstein G, Overstrom E2005: Preservation ofheritage livestock breeds: integrated program to cryopreservegermplasm from Tennessee Myotonic goats. ReprodFertil Dev 17, 195 (Abstract).Nijman IJ, Otsen M, Verkaar EL, de Ruijter C, Hanekamp E,2003: Hybridization of banteng (Bos javanicus) and zebu(Bos indicus) revealed by mitochondrial DNA, satelliteDNA, AFLP and microsatellites. Heredity 90, 10–16.SanCristobal M, Chevalet C, Haley CS, Joosten R, RattinkAP, Harlizius B, Groenen MA, Amigues Y, Boscher MY,Ó 2008 No claim to original government works

88 JA LongRussell G, Law A, Davoli R, Russo V, Desautes C,Alderson L, Fimland E, Bagga M, Delgado JV, Vega-PlaJL, Martinez AM, Ramos M, Glodek P, Meyer JN, GandiniGC, Matassino D, Plastow GS, Siggens KW, Laval G,Archibald AL, Milan D, Hammond K, Cardellino R,2006: Genetic diversity within and between Europeanpig breeds using microsatellite markers. Anim Genet 37,189–198.Storz JF, 2005: Using genome scans of DNA polymorphismto infer adaptive population divergence. Mol Ecol 14,671–688.Sultana S, Mannen H, Tsuji S, 2003: Mitochondrial DNAdiversity of Pakistani goats. Anim Genet 34, 417–421.Sunnucks P, 2001: Efficient genetic markers for populationbiology. Tree 15, 199–203.Tapio M, Tapio I, Grislis Z, Holm LE, Jeppsson S, KantanenJ, Miceikiene I, Olsaker I, Viinalass H, Eythorsdottir E,2005: Native breeds demonstrate high contributions to themolecular variation in northern European sheep. Mol Ecol14, 3951–3963.Wells DN, Misica PM, Tervit HR, Vivanco WH, 1998: Adultsomatic cell nuclear transfer is used to preserve the lastsurviving cow of the Enderby Island cattle breed. ReprodFertil Dev 10, 369–378.Wendorf F, Schild R, 1994: Are the early Holecene cattle in theEastern Sahara domestic or wild? Evol Anthropol 3, 118–128.Wildt DE, 1992: Genetic resource banking for conservingwildlife species: Justification, examples and becomingorganized on a global basis. Anim Reprod Sci 28, 247–257.Wildt DE, 2000: Genome resource banking for wildliferesearch, management, and conservation. ILAR J 41, 228–234.Wildt DE, Rall WF, Crister JK, Monfort SL, Seal US, 1997:Genome resource banks: ‘living collections’ for biodiversityconservation. Bioscience 47, 689–698.Wong GK, Liu B, Wang J, Zhang Y, Yang X, Zhang Z, MengQ, Zhou J, Li D, Zhang J, Ni P, Li S, Ran L, Li H, Zhang J,Li R, Li S, Zheng H, Lin W, Li G, Wang XZhao W, Li J, YeC, Dai M, Ruan J, Zhou Y, Li Y, He X, Zhang Y, Wang J,Huang X, Tong W, Chen J, Ye J, Chen C, Wei N, Li G,Dong L, Lan F, Sun Y, Zhang Z, Yang Z, Yu Y, Huang Y,He D, Xi Y, Wei D, Qi Q, Li W, Shi J, Wang M, Xie F,Wang J, Zhang X, Wang P, Zhao Y, Li N, Yang N, DongW, Hu S, Zeng C, Zheng W, Hao B, Hillier LW, Yang SP,Warren WC, Wilson RKBrandstro¨m M, Ellegren H, CrooijmansRP, van der Poel JJ, Bovenhuis H, Groenen MA,Ovcharenko I, Gordon L, Stubbs L, Lucas S, Glavina T,Aerts A, Kaiser P, Rothwell L, Young JR, Rogers S, WalkerBA, van Hateren A, Kaufman J, Bumstead N, Lamont SJ,Zhou H, Hocking PM, Morrice D, de Koning DJ, Law A,Bartley N, Burt DW, Hunt H, Cheng HH, Gunnarsson U,Wahlberg P, Andersson L, Kindlund E, Tammi MT,Andersson B, Webber C, Ponting CP, Overton IM, BoardmanPE, Tang H, Hubbard SJ, Wilson SA, Yu J, Wang J,Yang H, International Chicken Polymorphism Map Consortium,2004: A genetic variation map for chicken with 2.8million single-nucleotide polymorphisms. Nature 432, 717–722.Author’s address (for correspondence): JA Long, Animal Biosciencesand Biotechnology Laboratory, Beltsville Agricultural Research Center,US Department of Agriculture, Beltsville, MD 20705, USA. E-mail: jlong@anri.barc.usda.govConflict of interest: The author declares no conflict of interest.Ó 2008 No claim to original government works

Reprod Dom Anim 43 (Suppl. 2), 89–95 (2008); doi: 10.1111/j.1439-0531.2008.01147.xISSN 0936-6768Genetic Improvement of Dairy Cow Reproductive PerformanceB BerglundDepartment of Animal Breeding and Genetics, Swedish University of Agricultural Sciences and Centre for Reproductive Biology in Uppsala, Uppsala,SwedenContentsThe welfare of cow along with profitability in production areimportant issues in sustainable animal breeding programmes.Along with an intense ⁄ intensive selection for increased milkyield, reproductive performance has declined in many countries,in part due to an unfavourable genetic relationship. Thelargely unchanged genetic trend in female fertility and calvingtraits for Scandinavian Red breeds shows that it is possible toavoid deterioration in these traits if they are properly consideredin the breeding programme. Today’s breeding is internationalwith a global selection and extensive use of the best bulls.The Nordic countries have traditionally recorded and performedgenetic evaluation for a broad range of functional traitsincluding reproduction. In recent years many other countrieshave also implemented genetic evaluation for these traits. Thus,the relative emphasis of dairy cattle breeding objectives hasgradually shifted from production to functional traits such asreproduction. Improved ways of recording traits, e.g. physiologicalmeasures, early indicator traits, assisted reproductivetechniques and increased knowledge of genes and their regulationmay improve the genetic selection strategies and havelarge impact on present and future genetic evaluation programmes.Extensive data bases with phenotypic recordings oftraits for individuals and their pedigree are a prerequisite.Quantitative trait loci have been associated to the reproductivecomplex. Most important traits, including reproduction traitsare regulated by a multitude of genes and environmental factorsin a complex relationship, however. Genomic selection mighttherefore be important in future breeding programmes. Informationon single nucleotide polymorphism has already beenintroduced in the selection programmes of some countries.IntroductionThere are strong motives for including reproduction ingenetic selection programmes. A good reproductiveperformance is crucial for economic as well as ethicalreasons. Without reproduction there will be no animalproduction. The unfavourable genetic correlation withmilk production has led to a decline in reproduction indairy cattle, at least in part due to an insufficientconsideration of this trait when selecting for a highermilk production. There are several reports that theincreasing use of Holstein genetics has caused decliningfertility as well as calving performance during the latestdecades (e.g. Berglund and Philipsson 1992; Royal et al.2000; Lucy 2001; Hansen et al. 2004). This has causeddeterioration in fertility in countries that heavily used thisbreed even if they have had reproductive performanceboth in the breeding goal and in the selection criteria. Forinstance, in Swedish Holsteins (SH), daughter fertilityhas fallen by approximately 1.2 index units per year overthe last fifteen years until now (Lindhe´ B, 2007: SvenskAvel, Skara, Sweden, personal communication), whereasthe Swedish Red breed (SRB), with similar milk yieldlevels and genetic trend, has largely maintained fertility.A national breeding programme considering reproductiontraits for more than three decades may explain this.However for Holsteins the inclusion of fertility in theSwedish breeding goal has not been enough to withstandthe large imports of genetic material from countries thathave low, or no weighting on fertility in their breedingobjective.In 2006 the World Holstein Friesian Federationinitiated a survey of the status on fertility in the Holsteinpopulation around the world. The conclusion from thissurvey was that fertility is a problem and actions need tobe taken both internationally and within each country(Sørensen et al. 2007). Many countries have implementedgenetic evaluation for fertility traits in recentyears. More traits are gradually being evaluated andmore sophisticated evaluation methods are being implemented.Maybe the decline in fertility now has levelledout and reached a plateau, as was concluded from aninternational conference on fertility in dairy cows held inLiverpool, 2007 (Crowe 2007, personal communication).Still the low level of reproductive performance is aproblem. Reproduction problems are among the mostcommon reason for culling in dairy production. This isthe major cause for involuntary culling, e.g. in Swedishdairy cattle (Swedish Dairy Association 2007).Today’s breeding is international, intensive and usesmodern reproductive and molecular genetic techniques.The welfare of cows along with profitability in productionare important issues in sustainable animal breedingprogrammes. Nielsen et al. (2006) suggested that whendefining breeding goals for sustainable production,breeding organizations should predict the selectionresponse based on market economic value and addnon-market value for traits with unacceptable selectionresponses. New ways of measuring, recording andanalysing traits, new reproductive techniques andincreased knowledge of genes and their regulation mayimprove the genetic selection strategies and have a largeimpact on genetic evaluation programmes. This paperwill focus on the genetic improvement of the femalefertility and calving traits.Difficulties in Selection for ReproductionMany reproduction traits are difficult to handle inparameter estimation and genetic evaluation. In generalheritabilities are low, usually less than 5%, mainly due toa large influence of management and environmentaleffects. Most traits are not normally distributed and havecensored records, which complicates the analysis of thetraits. In addition to the other features of the reproductiveÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

90 B Berglundtraits calving performance traits are influenced by aneffect of the mother and a direct effect of the young(Steinbock et al. 2003; Hansen et al. 2004) and it may bedifficult to correctly estimate these (co)variances. To getselection response in reproduction traits a well developedrecording scheme is needed and a breeding structureallowing large daughter group sizes.Several sub-traitsReproduction, especially female fertility, is a complextrait composed of several sub-traits. Thus it is importantto include all important aspects of fertility to achieve agood and expected selection response. A challenge indoing this is the data collection and the quality of data.Traditionally, most fertility traits are based on calvingand insemination data and each trait has its strengthsand weaknesses. For female fertility there are measuresreflecting the ability to resume oestrous cycles aftercalving and the ability to conceive, or measurescombining these abilities like calving to last insemination(CLI), often also called days open (DO). Informationabout treatments and culling for reproductivedisorders are recorded and used mainly in the Nordiccountries. In Sweden scores for heat symptoms arerecorded as well. Heat detection plays a considerablerole for the economy. A visible manifestation of oestrusand a high oestrus detection rate is important especiallywhen using AI and in countries where hormones foroestrus synchronization are not generally used. Dobsonet al. (2007) reported from the UK that the number ofoestrous animals standing-to-be-mounted has declinedfrom 80% to 50% over the past 30–50 years.Antagonistic genetic correlations to other traitsThe antagonistic relationship between fertility and milkproduction is well-known and was shown already byJanson and Andre´ asson (1981). These correlations wereconfirmed in studies by Roxstro¨ m et al. (2001b), whereunfavourable genetic correlation for, e.g. number ofinseminations per service period, interval from calvingto first AI and treatments for reproductive disorders tomilk production was shown. The correlations rangedfrom 0.2 to 0.4, increasing with lactation number,possibly as a consequence of a higher energy demandwith increasing production level and steeper lactationcurves. Due to the antagonistic genetic correlation tomilk production, undesirable trends are expected forthe reproductive traits if they are not included in thebreeding goal. Even if included in the breeding objectivethere is a risk for deterioration of this trait due to thelow heritability if too small a breeding goal weight ordaughter groups are used. Ethic values (cow welfare,consumer preference, etc.) should be put into thebreeding goal weight and those are not easily accessible.Moreover, there are unfavourable correlations amongthe reproduction traits that are important to consider.Negative energy balance and reproductionA top producing cow might not cope with the energyrequirements for both milk production and maintainedreproduction and health even though the cows have acapacity to mobilize from their energy reserves. In ourstudies of Swedish Red (SRB) and Swedish Holsteins(SH) we have seen that SH has a thinner layer ofsubcutaneous fat than SRB which may be important inthis aspect (Hjertén 2006). A negative energy balance isassociated with inferior embryo quality. A review paperon metabolic changes and embryo quality was given byChagas et al. (2007). Hayhurst et al. (2007a) found asignificant genetic variation in embryo quality andestimated a heritability of 0.13 for this trait, implyinga possibility of genetically selecting cattle with inherentquality to produce high quality embryos.Body condition score (BCS) is an internationallyaccepted, rapid, inexpensive and non-invasive method ofestimating body energy reserves in dairy cattle (Berryet al. 2007). Even though mainly used for managementpurposes, BCS may also add to the information ingenetic evaluations, especially when data on more directtraits are lacking. BCS is moderately heritable (0.09–0.45) and favourably correlated to fertility and survival.BCS is used as a predictor trait for genetic merit forfertility in some countries, e.g. the Netherlands (De Jong2005) and in Irish and UK cattle (Berry et al. 2007).Current Status in Breeding ProgrammesThe general breeding goal for the reproduction of cows israther similar in different countries and may be formulatedas follows: cows that return to normal cyclicity earlyafter calving show strong and regular heats, and whichconceive when inseminated at a correct time. Furthermore,the cow should carry her pregnancy to term, have agood calving ability and give birth to viable calves.Female fertilityEarly studies by Janson (1980) showed that even thoughheritability for fertility traits was low, the additivegenetic variation was shown to be substantial. Furthermore,the importance of including both interval andpregnancy measures at different ages in the geneticevaluation for fertility was underlined. These resultswere later confirmed in large field studies by Roxstro¨ met al. (2001a,b). The Nordic countries have traditionallyrecorded and performed genetic evaluation for a broadrange of functional traits including reproduction, but inrecent years many other countries have also implementedgenetic evaluation for these traits. Thus, therelative emphasis of dairy cattle breeding objectives hasgradually shifted from production to functional traitssuch as reproduction during the past couple of decades(Miglior et al. 2005).Breeding values for daughter fertility were introducedin Sweden as early as 1972, and have since been used inselection. In the Nordic countries, the integration of cowdata bases (pedigree, milk recording, AI and diseasedata) has facilitated selection for reproductive traits.Three of the Nordic (Denmark, Finland and Sweden)breeding organizations have had a joint genetic evaluationand breeding programme since 2005 (for moreinformation see http://www.nordicebv.info). The Nordicfertility index includes the traits number of AI per serviceÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Genetic Improvement of Dairy Cow Reproductive Performance 91period for heifers (h) and cows (c), interval from calvingto first AI (c), interval from first to last AI (h, c) andreproductive treatments (c). Since the mid 1990s geneticevaluation for fertility has gradually been introduced inseveral countries, e.g. the Netherlands (Hoekstra et al.1994), UK (Wall et al. 2003) and US (VanRaden et al.2004). An international genetic evaluation for fertilitytraits was introduced for Holstein populations inFebruary 2007 by Interbull and an evaluation of fertilityin the other main breeds of the Interbull membercountries is under development (Jorjani 2007) (for moreinformation see http://www.interbull.org).Genetic trend for fertilityLindhe´ and Philipsson (2001) found a clear unfavourablegenetic trend in female fertility for SwedishHolsteins and a slightly favourable genetic trend forthe Swedish Red breed. In the Norwegian Red breed notrend or a slightly favourable genetic trend for fertilitytraits was found (Chang et al. 2006). Traditional breedingstrategies have been very successful in selecting highyielding dairy cows (Fig. 1a,b). While functional traitssuch as reproductive performance have declined in manycountries, especially for Holsteins, the largelyunchanged genetic trend in female fertility and calvingtraits for Danish, Finnish and Swedish Red (Ayrshire(a) 115DFS/SWE scale DFS/SWE scale1059585(b) 115751985 1990 1995 2000Years1059585751985 1990 1995 2000YearsMIFAPRSCLOCECFDOCMFig. 1. (a) Genetic trends for functional traits and milk productiontraits for Red (Ayrshire type) breeds. DFS ⁄ SWE, Danish, Finnish andSwedish Red breeds on a Swedish scale for breeding values (Jorjani2007, personal communication); MI, milk yield; FA, fat yield; PR,protein yield; SC, somatic cell scores; LO, longevity; CE, direct calvingease; CF, calving to first insemination; DO, days open; CM, clinicalmastitis. (b) Genetic trends for functional traits and milk productiontraits for HolsteinsMIFAPRSCLOCECFDOCMtype) breeds (Fig. 1a,b) shows that it is possible to avoida deterioration in these traits if they are properlyconsidered in the breeding programme. While forDanish, Swedish and Finnish Holsteins the breedingprogramme could not fully compensate for the use offoreign bull father genetic material for which functionaltraits such as reproduction were not known.Calving performance traitsFor first-parity Holstein cows, calf mortality is a greatproblem (Berglund and Philipsson 1992) and nowadaysreports are many of their high stillbirth rates, 10–13% atfirst calving. Stillbirths are approximately twice as high forHolstein first-calvers as for Swedish Red (Fig. 2). Theincreasing trend for stillbirth in Holsteins has not beenaccompanied by an increase in calving difficulty and thedivergent phenotypic trends may be an illustration of avitality problem. For second-calvers hardly any differencesin calving difficulty and stillbirth exist between the breeds.In a post-mortem examination of 76 calves from firstcalvingHolsteins, one-third of the calves were bornwithout any visible defects or injuries (Berglund et al.2003) and only half of the calves were born with signs of adifficult calving, indicating a vitality problem. The geneticcorrelations for stillbirth at first and second calving were0.45–0.48 for Swedish Holsteins (Steinbock et al. 2003),but 0.83–0.85 for the Swedish Red breed (Steinbock et al.2006). Thus there seem to be a genetic difference invitality of calves between these breeds at first calving.There are probably multifactorial reasons behindincreasing stillbirth rates. Adamec et al. (2005) recentlyshowed a consistently unfavourable effect of inbreedingon calving difficulty and stillbirth in US Holsteins, withlargest effects for first parity births. McParland et al.(2007) found that inbreeding had a deleterious effectupon most of the traits studied such as dystocia, stillbirthand calving interval in Irish Holstein-Friesians. A maximizedgenetic gain should be balanced against a minimizedinbreeding. There are now programmes to be builtinto the genetic evaluation programmes that maximizegenetic gain while minimizing inbreeding.Because calving difficulty and stillbirth mainly is aproblem for first parity cows, this category should be themain source of information for genetic evaluation. Forcertain breeds such as the Swedish Red breed the reliabilityin breeding values for calving ability could be increasedby including higher calving numbers (Steinbock et al.2006). Genetic evaluations should consider both calvingdifficulty and stillbirth as calf and maternal traits.Calving performance has been recorded in Swedensince the 1960s. The number of countries recordingcalving traits is increasing (Mark et al. 2005), and aninternational genetic evaluation was introduced in 2005(Jacobsen and Fikse 2005). Many of the member countriesof Interbull genetically evaluate calving difficulty andmost of them now also evaluate bulls for stillbirth, e.g. aroutine evaluation for stillbirth in Holsteins was implementedin the US in 2006 (Cole et al. 2007). Difficultcalving and stillbirth reduce reproductive performance(Bicalho et al. 2007), and genes associated with difficultbirth also reduce reproductive success (Lo´ pez de Maturanaet al. 2007).Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

92 B Berglund12108Percent64201986 1989 1992 1995 1998 2001 2004YearsStillbirths SH Calving diff. SH Stillbirths SRB Calving diff. SRBFig. 2. Phenotypic trends in stillbirth rate and calving difficulty for first calvers of SLB ⁄ Swedish Holsteins (SH) and Swedish Red (SRB) (Swedishmilk recording statistics)Genetic trend for calving traitsMark et al. (2005) reported a slightly negative genetictrend from the Interbull genetic evaluations for maternalstillbirth in the Holstein breed since the early 1990swhile there was no clear trend in the other calving traits.For the Swedish and Norwegian Red breeds, no orslightly positive genetic trends for calving performancehave been observed (Heringstad et al. 2007; Lindhé,2007, personal communication). The genetic trends fordirect calving ease for Danish, Finnish and Swedish Redbreed and Holsteins can be seen as one of the functionaltraits in Fig. 1a,b.Genetic defects causing reproduction lossMutations in single genes occur regularly and may causecongenital genetic defects. These are commonly recessivelyinherited (http://www.omia.angis.org.au). Onerecent example is the complex vertebral malformation(CVM) in the Holstein breed, first described by Agerholmet al. (2001). Complex vertebral malformation has had amajor impact on the reproductive performance inHolsteins (Agerholm 2007). In a study of SwedishHolsteins, carriers of the CVM-gene had inferior NRrate compared with non-carrier bulls reflecting a higherintra-uterine mortality (Berglund et al. 2004). It isimportant to report all kinds of malformations and tohave national control programmes for congenital defectsin order to avoid multiplication of deleterious genescausing reproductive losses and animal welfare problems.Crossbreeding as a tool to enhance fertilityDecreasing fertility and increasing calf mortality inHolsteins have become a large problem. In some countries,e.g. in the USA, Scandinavian Red and other fertilebreeds are now crossed into Holsteins to improve theHolstein reproduction. Crossbreeding may help elevatethe level of traits combining breeds with favourable traitsand by exploring heterosis effects, but the continuousgenetic improvement has to be done in the pure breeds.Heins et al. (2006a) reported that calves from Holsteinfirst-calf heifers sired by Scandinavian Red (NorwegianRed and Swedish Red) bulls had significantly less calvingdifficulty (5.5%) and lower stillbirth rate (7.7%) thancalves sired by Holstein bulls (16.4% difficult calving and15.1% stillbirth). Moreover, DO were significantlyshorter for all crossbred groups studied and were129 days for Scandinavian Red ⁄ Holstein crossbredsand 150 days for pure Holsteins (Heins et al. 2006b).New Tools for Genetic Improvement ofReproductionImproved ways of recording traits, e.g. direct measuringof physiological measures may offer valuable opportunitiesto improve the genetic evaluation of fertility.Indicator traits are valuable because fertility has a lowheritability and is expressed late in life. More advancedand expensive recording technologies may be used innucleus herds and the genetic progress may be enhancedby using modern reproductive techniques such asMOET. Reproductive techniques like sexing, cloningand transfer of genetic material will impact present andfuture selection strategies in breeding programmes.The amount of information on the molecular level israpidly increasing. The sequencing of the bovine genomewas completed in 2003 (http://www.hgsc.bcm.tmc.edu/projects/bovine). Markers are connected to the phenotypicexpression of several traits in the reproductioncomplex. Extensive data bases with phenotypic recordingsof traits for individuals and their pedigree are aprerequisite. Gene expression profiles may increase ourunderstanding of the mechanisms behind reproductivefunctions and their phenotypic expression. As genes fortraits are identified, genetic selection strategies can beimproved. Most important traits, including reproductiontraits are regulated by a multitude of genes and environmentalfactors in a complex relationship, however.Progesterone-based measures of female fertilityEarly measures of the reproductive function of cow canbe obtained from progesterone profiles. Progesterone-Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Genetic Improvement of Dairy Cow Reproductive Performance 93based measures of fertility have higher heritability thanthe traditional measures of fertility (Royal et al. 2002;Petersson et al. 2007). Veerkamp et al. (1998) suggestedthat selection on days to first luteal activity (CLA) basedon monthly analysis of progesterone in milk may addfurther accuracy to the genetic evaluation for fertility.Van der Lende et al. (2004) suggested to select siresbased on a measure when 50% of the daughters of a sirehad an active corpus luteum (CLA 50%) based on 3- to6-week intervals of progesterone sampling. Peterssonet al. (2007) found that direct selection on progesteronebased measures of fertility may increase the accuracy inthe genetic evaluation for an early start of cyclic ovarianactivity after calving compared to the commonly usedmeasure CFI, even with an infrequent sampling such asin the regular milk recording system. Progesteroneanalysis in the first monthly collected milk samplescould also be used as a management tool. Peterssonet al. (2008) showed that four out of five cows withdelayed cyclicity could be predicted within 60 days aftercalving enabling an earlier treatment. The cost forprogesterone analysis of milk samples could thereforebenefit from management returns as well as fromimprovements in genetic gain for fertility.Automization of recordingsAutomatic milking systems, e.g. robotic milking systemsmay allow automization of recordings. Commercialsystems for on-line recordings allowing herd-profiles offertility (e.g. based on progesterone analysis) and healthparameters are underway (Friggens and Løvendahl2007). These may offer recordings with a high accuracythat could be built into selection programmes. Løvendahland Chagunda (2006) used activity meters forautomatic heat controls and estimated a heritability of0.17 for this trait.Juvenile predictorsThe genes controlling fertility are present and potentiallyexpressed early in life, but it takes at least four yearsbefore a bull has milking daughters and can be progenytested for fertility. Thus early fertility predictors would bevery valuable. These could be reproductive hormones ormetabolic traits. Hayhurst et al. (2007b) suggested apossibility of using the correlation between the prepubertalresponse to gonadotrophin releasing hormone inbull calves and the fertility of their daughters as a possibleselection tool. In another study by Hayhurst et al.(2007c), it was suggested that selection for bull calveswith lower concentrations of glucose and FFA couldresult in female offspring with genetically better fertility.Gene mapping studiesThe identification of quantitative trait loci (QTL) is afirst step towards novel selection methods based on bothphenotypic and molecular information. Using QTL inselection is most beneficial for low heritability traits, sexlimited traits and traits expressed late in life such asdaughter fertility. Holmberg and Andersson-Eklund(2006) found regions with several QTLs on chromosome9 and 11 for both reproduction and health traits. A QTLfor non-return rate was fine mapped to an interval ofless than 3 cM on chromosome 9 (Holmberg et al.2007). Marker assisted selection (MAS) can be used forpre-selection among full-sibs before progeny testing byaccounting for the Mendelian sampling and also toavoid genetic defects for which there are availablemarkers. For quantitative traits the benefit from MAS islimited by the proportion of the genetic varianceexplained by known QTL. The marker density in theQTL region can be increased by use of single nucleotidepolymorphism (SNP) markers. Single nucleotide polymorphismsare single base-pair differences betweenindividuals within a species and where the differentvariants (most often only two allelic forms) are relativelycommon in a population.Expression profiles of genes regulating reproductionGene expression profiling is a relatively recent toolcontributing to our knowledge about the various processesunderlying reproduction such as the function ofgenes and their products determining the phenotype forreproduction. Beerda and Veerkamp (2006) summarizedgene expression studies related to reproduction in cattle,sheep and swine in a paper at the World Congress onGenetics Applied to Livestock Production in Brazil.They concluded that a vast amount of information hasalready been achieved in this area and as informationwill grow exponentially over the next few years theyunderlined the importance of a major effort in bringingall information together in a ‘broadly accessible ontology’.The need for compiling and analysing largeamounts of molecular data has created a new field ofscience referred to as bioinformatics. An EU integratedproject SABRE was started in 2006, which aims toprovide fundamental knowledge on the genomics andepigenetics on, e.g. reproduction traits in dairy cattle tobe used in selection for improved reproduction efficiency.An update on studies on expression profiles ofgenes regulating dairy cow fertility was given at theInternational conference on fertility in dairy cows inLiverpool, 2007 (Beerda and Veerkamp 2007).Genomic selection and integration of molecular data intogenetic evaluation programmesGenomics is the study of variation of base pairs in thenucleic acids and genomic selection means using thisinformation in genetic selection programmes. The availabilityof large arrays of SNPs is changing the approachof predicting breeding values from molecular information.Genomic selection (GMAS) uses all markers, orrather haplotypes consisting of a pair of contiguousSNPs, spanning the genome for prediction of breedingvalues. Thus, by summing the effects of all markerhaplotypes in the genome of a bull calf, a breeding valueis obtained directly at birth whereby the generationinterval can be considerably shortened. This informationcan be used directly, hence theoretically no progenytesting is needed. Muir (2007) showed that by usingGMAS for traits of high (0.5) or low heritability (0.1) theaccuracy of selection increased between 10% and 30%.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

94 B BerglundVan Raden (2007) calculated measures of relationshipand inbreeding by use of genomic measures and demonstratedgains in reliability as compared to using traditionalmethods, and comparatively more for lowheritability traits than high heritability traits. A disadvantageis that the information may only be used forsome generations, maybe 7–8 generations, because theestimated effects of SNP haplotypes will change overtime. Furthermore, without progeny testing in eachgeneration the risk for late detection of undesired sideeffectsof selection might increase.Along with increasing knowledge at the genomic level,effort is put on how to integrate these data in the geneticevaluation systems. Gengler and Verkenne (2007) statedthat it will be a challenge to integrate molecular andphenotypic data and that SNPs and genomic selectionwill not change this situation.Genomic selection based on 3000 SNP markers toselect young bulls was introduced in the Netherlands inOctober 2006 (Van der Beek 2007). Denmark andNorway have also recently introduced SNP informationin their genetic selection programmes (MS Lund andS Lien, personal communication). Large scale projectsare ongoing in several livestock species to identify andvalidate several thousands of SNPs in haplotype blocksin the genome. SNP chips with 60 000 SNPs will soonbe available.Concluding RemarksThe genetic trend for functional traits in the Red(Ayrshire) breeds shows that it is possible to maintaina good dairy cow fertility and calving performancealong with selection for increased levels of milk production,if the traits are properly considered. It ispossible to make improvements in breeding, as well as innutrition and management to keep a good reproductiveability of the cow and a good health and well-being insustainable breeding programmes. Along with increasingknowledge at a molecular level, large changes maybe expected in genetic evaluation programmes.AcknowledgementsI would like to acknowledge Professor Erling Strandberg at theDepartment of Animal Breeding and Genetics, SLU, for valuablecomments on the manuscript. Associate Professor Hossein Jorjani atthe Interbull office (Jorjani 2007, communication), SLU, for kindlyproviding Interbull genetic trends and Associate Professor BengtLindhé at Svensk Avel (Lindhé 2007, communication) for kindlyproviding national genetic trends.ReferencesAdamec V, Cassell BG, Smith EP, Pearson RE, 2005: Effectsof inbreeding in the dam on dystocia and stillbirths in USHolsteins. J Dairy Sci 89, 307–314.Agerholm JS, 2007: Complex vertebral malformation inHolstein cattle: the story so far. Perinatal death in domesticanimals. Proc 20th NKVet Symp, Reykjavik, Iceland, p. 88.Agerholm JS, Bendixen C, Andersen O, Arnbjerg J, 2001:Complex vertebral malformation in Holstein calves. J VetDiagn Invest 13, 283–289.Beerda B, Veerkamp RF, 2006: Functional genomics of femalereproduction. 8th World Congress on Genetics Applied toLivestock Production, Belo Horizonte, Brazil. Proc Commun11–10 (http://www.wcgalp8.org.br).Beerda B, Veerkamp RF, 2007: Expression profiles of genesregulating dairy cow fertility: recent findings, ongoingactivities and future possibilities. Proc Fertility in DairyCows – bridging the gaps, 30–31 August 2007, LiverpoolHope University, Liverpool, UK, p. 33.Berglund B, Philipsson J, 1992: Increasing stillbirth rates in theSwedish Friesian population. Paper in 43rd Annual Meetingof the European Association on Animal Production,Madrid, 14–17 September 1992.Berglund B, Steinbock L, Elvander M, 2003: Causes ofstillbirth and time of death in Swedish Holstein calvespost-mortem examined. Acta Vet Scand 44, 111–120.Berglund B, Persson A, Stålhammar H, 2004: Effects ofcomplex vertebral malformation on fertility in SwedishHolstein cattle. Acta Vet Scand 45, 161–165.Berry D, Roche J, Coffey M, 2007: Body condition score andfertility – more than just a feeling. Proc Fertility in DairyCows – bridging the gaps, 30–31 August 2007, LiverpoolHope University, Liverpool, UK, p. 28.Bicalho RC, Galvao KN, Cheong SH, Gilbert RO, WarnikLD, Guard CL, 2007: Effect of stillbirths on dam survivaland reproduction performance in Holstein dairy cows.J Dairy Sci 90, 2797–2803.Chagas LM, Bass JJ, Blache D, Burke CR, Kay JK, LindsayDR, Lucy MC, Martin GB, Meier S, Rhodes FM, RocheJR, Thatcher WW, Webb R, 2007: Invited review: Newperspectives on the roles of nutrition and metabolic prioritiesin the subfertility of high-producing dairy cows. J DairySci 90, 4022–4032.Chang YM, Andersen-Ranberg IM, Heringstad B, Gianola D,Klemetsdal G, 2006: Bivariate analysis of number of servicesto conception and days open in Norwegian Red using acensored threshold-linear model. J Dairy Sci 89, 772–778.Cole JB, Wiggans GR, VanRaden PM, Miller RH, 2007:Stillbirth (co)variance components for a sire-maternalgrandsire threshold model and development of a calvingability index for sire selection. J Dairy Sci 90, 2489–2496.De Jong G, 2005: Usage of predictors for fertility in the geneticevaluation application in the Netherlands. Interbull Bull 33,69–73.Dobson H, Walker S, Morris M, Routly J, Smith R, 2007:Why is it getting more difficult to successfully artificiallyinseminate cows? Proc Fertility in Dairy Cows – Bridgingthe Gaps, 30–31 August 2007, Liverpool Hope University,Liverpool, UK, p. 4.Friggens N, Løvendahl P, 2007: Where did she come from?Where is she going? The potential of on-farm fertility profiles.Proc. Fertility in Dairy Cows – Bridging the Gaps, 30–31August 2007, LiverpoolHope University, Liverpool,UK, p.23.Gengler N, Verkenne C, 2007: Bayesian approach to integratemolecular data into genetic evaluations. Interbull Bull 37,37–41.Hansen M, Lund MS, Pedersen J, Christensen LG, 2004:Genetic parameters for stillbirth in Danish Holstein cowsusing a Bayesian threshold model. Livest Prod Sci 91, 23–33.Hayhurst C, Firth M, Christie MF, Royal MD, 2007a:Estimation of genetic variation in embryo quality: Is therepotential to genetically select cattle with the inherent abilityto produce high quality embryos? Proc Fertility in DairyCows – Bridging the Gaps, 30–31 August 2007, LiverpoolHope University, Liverpool, UK, p. 13.Hayhurst C, Flint APF, Lovendahl P, Wolliams JA, RoyalMD, 2007b: Prepubertal predictors for fertility in dairycattle: potential use of metabolic hormones. Proc BSASAnnual Conference 2007: Southport, UK, 58.Hayhurst C, Sorensen MK, Royal MD, Løvendahl P, 2007c:Metabolic regulation in Danish bull calves and the relation-Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Genetic Improvement of Dairy Cow Reproductive Performance 95ship to the fertility of their female offspring. J Dairy Sci 90,3909–3916.Heins BJ, Hansen LB, Seykora AJ, 2006a: Fertility andsurvival of pure Holsteins versus crossbreds of Holstein withNormande, Montbeliarde, and Scandinavian Red. J DairySci 89, 4944–4951.Heins BJ, Hansen LB, Seykora AJ, 2006b: Calving difficultyand stillbirths of pure Holsteins versus crossbreds ofHolstein with Normande, Montbeliarde, and ScandinavianRed. J Dairy Sci 89, 2805–2810.Heringstad B, Chang YM, Svendsen M, Gianola D, 2007:Genetic analysis of calving difficulty and stillbirth inNorwegian Red cows. J Dairy Sci 90, 3500–3507.Hjerte´ n J, 2006: Relationships between body condition score,subcutaneous fat, live weight and reproduction in SwedishHolstein and Swedish Red and White Cattle. In Swedish.Summary in English. MSc Thesis 281, SLU, Department ofAnimal Breeding and Genetics, Uppsala, Sweden.Hoekstra J, van der Lugt AW, van der Werf JHJ, OuweltjesW, 1994: Genetic and phenotypic parameters for milkproduction and fertility traits in upgraded dairy cattle.Livest Prod Sci 40, 225–232.Holmberg M, Andersson-Eklund L, 2006: Quantitative traitloci affecting fertility and calving traits in Swedish dairycattle. J Dairy Sci 89, 3664–3671.Holmberg M, Sahana G, Andersson-Eklund L, 2007: Finemapping of a quantitative trait locus on chromosome 9affecting non-return rate in Swedish dairy cattle. J AnimBreed 124, 257–263.Jacobsen JH, Fikse F, 2005: Feasibility of MACE for calvingtraits for non-Holstein breeds. Interbull Bull 33, 28–31.Janson L, 1980: Studies on fertility traits in Swedish dairycattle. II. Genetic parameters. Acta Agr Scand 30, 427–436.Janson L, Andre´ asson B, 1981: Studies on fertility traits inSwedish dairy cattle. IV. Genetic and phenotypic correlationbetween milk yield and fertility. Acta Agr Scand 31, 313–322.Jorjani J, 2007: International genetic evaluation of femalefertility traits in five major breeds. Interbull Bull 37, 144–147.Lindhe´ B, Philipsson J, 2001: Genetic trends in the twoSwedish dairy cattle breeds SRB and SLB in 1985–1999.SLU, Department of Animal Breeding and Genetics, Publ.138, Uppsala, Sweden.Lo´ pez de Maturana E, Legarra A, Varona L, Ugarte E, 2007:Analysis of fertility and dystocia in Holsteins using recursivemodels to handle censored and categorical data. J Dairy Sci90, 2012–2024.Løvendahl P, Chagunda MGG, 2006: Assessment of fertility indairy cows based on electronic monitoring of their physicalactivity. 8th World Congress on Genetics Applied toLivestock Production, Belo Horizonte, Brazil. Proc Commun18-04 (http://www.wcgalp8.org.br).Lucy MC, 2001: Reproductive loss in high-producing dairycattle: where will it end? J Dairy Sci 84, 1277–1293.Mark T, Jakobsen JH, Jorjani H, Fikse WF, Philipsson J,2005: International trends in recording and genetic evaluationof functional traits in dairy cattle. 56th Annual Meet ofthe European Association for Animal Production, Uppsala,Sweden, Abstract no. 536.McParland S, Kearney JF, Rath M, Berry DP, 2007:Inbreeding effects on milk production, calving performance,fertility and conformation in Irish Holstein-Friesians.J Dairy Sci 90, 4411–4419.Miglior F, Muir BL, Van Doormaal BJ, 2005: Selection indicesin Holstein cattle of various countries. J Dairy Sci 88, 1255–1263.Muir WM, 2007: Genomic selection: a break through forapplication of marker assisted selection to traits of lowheritability, promise and concerns. Paper at the 58th AnnualMeet of the European Association for Animal Production,Dublin, Ireland.Nielsen HM, Christensen LG, Ødega˚ rd J, 2006: A method todefine breeding goals for sustainable dairy cattle production.J Dairy Sci 89, 3615–3625.Petersson K-J, Berglund B, Strandberg E, Gustafsson H, FlintAPF, Wolliams JA, Royal MD, 2007: Genetic analysis ofpostpartum measures of luteal activity in dairy cows.J Dairy Sci 90, 427–434.Petersson K-J, Strandberg E, Gustafsson H, Royal MD,Berglund B, 2008: Detection of delayed cyclicity in dairycows based on progesterone content in monthly milksamples. Prev Vet Med (in press).Roxstro¨ m A, Strandberg E, Berglund B, Emanuelson U,Philipsson J, 2001a: Genetic and environmental correlationsamong female fertility traits and milk production in differentparities of Swedish Red and White dairy cattle. Acta AgricScand A Anim Sci 51, 7–14.Roxstro¨ m A, Strandberg E, Berglund B, Emanuelson U,Philipsson J, 2001b: Genetic and environmental correlationsamong female fertility traits and the ability to show oestrus,and milk production. Acta Agric Scand A Anim Sci 51, 192–199.Royal MD, Mann GE, Flint APF, 2000: Strategies forreversing the trend towards subfertility in dairy cattle. VetJ 160, 53–60.Royal MD, Flint APF, Wolliams JA, 2002: Genetic andphenotypic relationships among endocrine and traditionalfertility traits and production traits in Holstein-Friesiandairy cows. J Dairy Sci 85, 958–967.Sørensen AC, Lawlor T, Ruiz F, 2007: A survey on fertility inthe Holstein populations of the world. Proc Fertility inDairy Cows – Bridging the Gaps, 30–31 August 2007,Liverpool Hope University, Liverpool, UK, p. 17.Steinbock L, Näsholm A, Berglund B, Johansson K, PhilipssonJ, 2003: Genetic effects on stillbirth and calving difficultyin Swedish Holsteins at first and second calving. J Dairy Sci86, 2228–2235.Steinbock L, Johansson K, Näsholm A, Berglund B, PhilipssonJ, 2006: Genetic effects on stillbirth and calving difficultyin Swedish Red dairy cattle at first and second calving. ActaAgric Scand A, 56, 65–72.Van der Beek S, 2007: Effect of genomic selection on nationaland international genetic evaluations. Interbull Bull 37, 115–118.Van der Lende T, Kaal LMTE, Roelofs RMG, Veerkamp RF,Schrooten C, Bovenhuis H, 2004: Infrequent milk progesteronemeasurements in daughters enable bull selection forcow fertility. J Dairy Sci 87, 3953–3957.VanRaden PM, 2007: Genomic measures of relationship andinbreeding. Interbull Bull 37, 33–36.VanRaden PM, Sanders AH, Tooker ME, Miller RH,Norman HD, Kuhn MT, Wiggans GR, 2004: Developmentof a national genetic evaluation for cow fertility. J Dairy Sci87, 2285–2292.Veerkamp RF, Oldenbroek JK, van der Lende T, 1998: The useof milk progesterone measurements for genetic improvementof fertility traits in dairy cattle. Interbull Bull 18, 62–67.Wall E, Brotherstone S, Wolliams JA, Banos G, Coffey MP,2003: Genetic evaluation of fertility using direct andcorrelated traits. J Dairy Sci 86, 4093–4102.Author’s address (for correspondence): B Berglund, Department ofAnimal Breeding and Genetics, Swedish University of AgriculturalSciences, Box 7023, SE-750 07, Uppsala, Sweden. E-mail: britt.berglund@hgen.slu.seConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 96–103 (2008); doi: 10.1111/j.1439-0531.2008.01148.xISSN 0936-6768Nutrient Prioritization in Dairy Cows Early Postpartum: Mismatch BetweenMetabolism and Fertility?JLMR Leroy 1 , T Vanholder 1 , ATM Van Knegsel 2 , I Garcia-Ispierto 3 and PEJ Bols 11 Veterinary Centre Someren, Someren; 2 Adaptation Physiology Group and Animal Nutrition Group, Wageningen Institute of Animal Sciences,Wageningen University, Wageningen, The Netherlands; 3 Department of Animal Health and Anatomy, Autonomous University of Barcelona,Barcelona, SpainContentsFor several decades, researchers worldwide report a decreasein fertility in high-yielding dairy cows, most probably based onconflicting metabolic and reproductive needs. The dairy herdmanager’s success at improving milk production has beenaccompanied by a negative trend for the most visible reproductiveparameters such as calving intervals, number of daysopen and number of inseminations needed per pregnancy. Inparallel, many research groups studied the metabolic andendocrine factors that influence follicular growth and thedevelopmental competence of oocytes and embryos. In thepast, herd managers and reproductive biologists each tried totackle the same problems with limited consultation. Morerecently, the situation has improved significantly and theriogenologists,nutritionists and veterinarians now conductresearch in multidisciplinary teams. This review paper startsin a general way by discussing nutrient prioritization towardsthe udder to guarantee milk production and by describinginteractions between the somatotropic and gonadotropic axis.It then focuses on the consequences of the negative energybalance on follicular growth and environment, oocyte andembryo quality, not only by summarizing the currentlyaccepted hypotheses but also based on clear scientific evidenceat the follicular level. All this, with one question in mind: isthere a mismatch between metabolism and fertility and whatcan the dairy manager learn from research to tackle theproblem of reduced fertility?IntroductionDisappointing reproductive performance in high-producingdairy herds is a global problem, characterized asmultifactorial and urged a multidisciplinary approach inwhich animal scientists, veterinarians and molecularbiologists were required to unravel the complex pathogenesisof this ‘subfertility syndrome’. After all, producinga calf at regular intervals is considered a prerequisitefor profitable lactational performance (Royal et al.2000; Huirne et al. 2002). After giving birth, the processof becoming pregnant again in dairy cows starts withclearance and involution of the uterus followed byresumption of ovarian activity. This should result in thecompletion of the growth of a healthy follicle, enclosinga competent oocyte, and ultimately in oestrus, ovulation,fertilization and uterine attachment by a viableembryo. Any upset of these balanced and fine-tunedbiological and mechanical events leads to failing reproduction– and this is exactly where the shoe pinches inour modern dairy herds.The subfertility syndrome can be divided into twomajor sub-problems. First of all, up to 50% of moderndairy cows display abnormal oestrus cycles postpartumleading to prolonged calving to first inseminationintervals (Opsomer et al. 1998). In this context, especiallyinstability within the hypothalamo–pituitary–ovarian–uterine axis has been studied thoroughly (Lucy2001; Butler 2003). The concomitant reduced oestrusexpression or even anoestrus, cyst formation anddelayed first ovulation have been extensively documented(Beam and Butler 1997; de Vries and Veerkamp2000; Lopez et al. 2004; Vanholder et al. 2006a).Secondly, attention was focussed on disappointingconception rates (Bousquet et al. 2004) and the increasinglyhigh incidence of early embryonic mortality(Dunne et al. 1999; Mann and Lamming 2001; Bilodeau-Goeseelsand Kastelic 2003). Fertilization ofoocytes from high-genetic merit cows resulted in significantlylower blastocyst yields in vitro, irrespective ofmilk production as such (Snijders et al. 2000). Embryoquality was also reduced in high-producing dairy cowscompared with non-lactating counterparts (Wiltbanket al. 2001; Leroy et al. 2005a). A high proportion ofnon-viable embryos were found in lactating cowscompared with non-lactating cows (Sartori et al.2002). Approximately 70–80% of the total embryonicand foetal losses typically occur during the earlyembryonic, pre-attachment period (Santos et al. 2004a)(for review, see Leroy et al. 2007).Modern dairy cows, albeit sub-fertile, produce vastamounts of milk mainly because of significant geneticimprovements, combined with nutritional managementoptimized towards lactation. Based on almostunchanged heifer fertility, we can conclude that thereproductive processes of modern dairy cattle areessentially normal when lactation demands are notimposed (Lucy 2007). Why do modern dairy cowsprioritize milk production at the expense of sustainedreproductive efficiency? In this review, we aim to answerthis question. Are high milk yields and good fertilityoutcomes conflicting interests metabolically speaking?From Phylogenetically Driven to GeneticallyEnforced Nutrient Prioritization: TheConsequences on MetabolismFrom a biological point of view, it makes sense formammals in early lactation to favour milk productionover fertility: this we can refer to as nutrient prioritization(Lucy 2003). As nutrition becomes scarce, thelactating dam will preferentially invest the limitedresources in the survival of living offspring rather thangambling on the oocyte that is yet to be ovulated,Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Nutrient Prioritization and Fertility in Dairy Cows 97fertilized and cared for during an entire gestation. Thismaternal catabolic mechanism, also genetically programmed,should maximize the chance of survival of thenewborn offspring (Silvia 2003). Over the past 40 years,the focus of dairy industry has been on maximizing milkyield, thereby creating a ‘nutrient highway’ from thedaily ration and body reserves (estimated on 74% bodyfat and 6% body protein: Tamminga et al. 1997) directlyto the udder to sustain milk production.Nutrient requirements of the gravid uterus late ingestation impose a catabolic status on the dairy cow.Following parturition, an additional demand for glucose,fatty acids and protein is established as milkproduction starts. During this transition period, cowsare unable to compensate for such increased energydemands by increasing feed intake, and this results innegative energy balance (NEB). Drastically reducedinsulin concentrations bring approximate energy mobilizationand partitioning of energy to the udder.Hypoinsulinaemia promotes gluconeogenesis in the liver(up to 4 kg glucose each day) and acts as a massivelipolytic trigger. The mobilized non-esterified fatty acids(NEFAs) serve as an alternative energy source for othertissues to preserve glucose, which is preferentially usedby the mammary gland to form lactose (Vernon 2002).NEFAs are predominantly transported to the liverwhere they are oxidized to provide energy or transformedinto ketone bodies, again an alternative energysource elsewhere in the body. An aberrant over-load ofthe liver by NEFAs can induce steatosis and disturbedliver function (Herdt 2000). Hormone-sensitive lipasesin adipose tissue of high-yielding dairy cows have anincreased sensitivity to lipolytic stimuli (such as lowinsulin, and high catecholamines or glucocorticoidsconcentrations). In other words, high-yielding dairycows have been genetically selected to partition evenmore energy reserves into milk production (Coffey et al.2004). A higher dietary energy intake will thereforeresult in greater milk production, but a similar energyimbalance remains, with no beneficial effects on bodycondition score (BCS) at all (Patton et al. 2006).A series of biological mechanisms bring an approximateprioritization for milk production at the cost ofbody reserves in early postpartum dairy cows. First ofall, the udder benefits because it does not need insulin tofacilitate glucose uptake into cells by the glucosetransportmolecules, GLUT 1 and 3, while most othertissues predominantly express insulin-dependent GLUT4 (Zhao et al. 1996). Secondly, using repeatedly intravenousglucose-tolerance tests, we recently found atemporary suppression of pancreatic function in earlypostpartum high-yielding dairy cows and this wascorrelated with elevated NEFA concentrations (Bossaertet al. 2007). In vitro, high NEFA levels have toxiceffects on pancreatic cells (Cnop et al. 2001; Maedleret al. 2001). Thirdly, in the early postpartum period, lowinsulin concentrations uncouple the growth hormone(GH)–insulin like growth factor 1 (IGF-I) axis in theliver because of down-regulation of GH 1A receptorsand this can be restored by increasing insulin (Butleret al. 2003). As IGF-I production in the liver issuppressed, the negative feedback of IGF-I is removedat the level of the hypothalamus ⁄ pituitary, and GHconcentrations increase. High GH concentrations notonly stimulate milk production but also provoke livergluconeogenesis and lipolysis in adipocytes. The resultinghigh blood NEFA and GH concentrations antagonizeinsulin action and create a further state ofperipheral insulin resistance (Lucy 2007; Pires et al.2007). In this way even more glucose is conserved to beavailable for lactose synthesis.Fatter cows tend to mobilize more body fat because ofreduced appetite (Garnsworthy and Topps 1982). It isbroadly accepted that genetic selection for milk productionresults in greater BCS loss, further suggesting thatenergy is partitioned towards the udder (Roche et al.2006). An excessive BCS loss during the transitionperiod is a major risk factor for health and fertilitydisorders (Roche et al. 2007), which stresses the importanceof BCS monitoring early postpartum as a managementtool (Chagas et al. 2007).Interactions Between the Somatotropic and theGonadotropic AxisExtensive scientific research has shown that mechanismsthat regulate energy and nutrient distribution in thesomatotropic system may affect the reproductive systemat different levels of the hypothalamo–pituitary–ovarianaxis (Roche 2006; Chagas et al. 2007). Within thehypothalamus, interactions between the gonadotropicand somatotropic systems may occur in the pre-opticarea (Blache et al. 2006, 2007). This region produces thereleasing hormones that control the secretion of bothgonadotropins and somatotropin (Kacsoh 2000). Inaddition, it plays a crucial role in integrating appetite(Wynne et al. 2005), oestrus behaviour (Pfaff 2005) andsensing of the nutritional status (Wade and Jones 2004).Consequently, metabolic inputs in the hypothalamusmay have divergent effects on the gonadotropic andsomatotropic axis, i.e. stimulation of GH productionmay be accompanied by inhibition of GnRH secretion(Zieba et al. 2005). The hormones ⁄ metabolites that aremost likely to exert a signalling function are glucose andinsulin. Low postpartum insulin and glucose concentrationssuppress hypothalamic GnRH secretion andsubsequent pituitary LH release (Diskin et al. 2003;Ohkura et al. 2004). By activation of specific neurons inthe forebrain, peptides such as neuropeptide Y andcatecholamines are released, which suppress the hypothalamicGnRH pulse generator (Ichimaru et al. 2001;Diskin et al. 2003; Wade and Jones 2004). Othermetabolic signals may involve leptin and NEFA,although their role currently remains unclear (Lieferset al. 2003; Wade and Jones 2004; Amstalden et al.2005).At ovarian level, follicular growth and developmentseems to be directly influenced by altered insulin, IGF-I,leptin and NEFA levels. Because insulin locally stimulatesfollicular growth, maturation and steroidogenesis,reduced postpartum concentrations are linked to ovariandysfunction (Gutierrez-Aguilar 1997; Landau et al.2000; Armstrong et al. 2002a; Butler et al. 2004; Vanholderet al. 2005a; Kawashima et al. 2007). Long-termtreatment with exogenous bovine somatotropin clearlyincreased the number of small follicles (Bols et al. 1998).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

98 JLMR Leroy, T Vanholder, ATM Van Knegsel, I Garcia-Ispierto and PEJ BolsGong (2002) showed that the beneficial effects of insulinon ovarian functions are independent of changes inGnRH ⁄ LH release. In ovarian cells, insulin-independentGLUT-1 and GLUT-3 are the major glucose transporters,while the insulin-dependent GLUT-4 only plays asupportive role (Nishimoto et al. 2006). Hence, insulinmay exert its effects through mechanisms other thanmediating glucose uptake.Together with insulin, the IGF system plays animportant role in follicle growth and development byacting directly on ovarian cells (Spicer and Echternkamp1995; Gutierrez-Aguilar 1997; Beam and Butler 1999;Webb et al. 1999; Gong 2002). Consequently, lowcirculating IGF-1 concentrations negatively influencethe onset of postpartum ovarian activity and seeminvolved in the development of cystic ovarian follicles(Beam and Butler 1997; Kawashima et al. 2007; Ortegaet al. 2007). The effects of leptin on steroidogenesis andcell proliferation are, yet, dependent on the circulatingconcentrations of IGF-1, LH and insulin (Spicer andFrancisco 1997, 1998; Spicer et al. 2000).At ovarian level, NEFAs may affect follicular growthand development by acting directly on follicle cells.Adding NEFAs in vitro, at concentrations measured infollicular fluid (FF) during NEB has detrimental effectson follicle cell viability and function (Leroy et al. 2005b;Vanholder et al. 2005b, 2006b).In conclusion, metabolic changes induced by thesomatotropic system to sustain a high milk yield alsoaffect the reproductive system. By acting at differentlevels of the hypothalamo–pituitary–ovarian axis,altered hormone and metabolite levels exert a negativeeffect on follicle growth, development and probablyovulation.Consequences for Oocyte QualityResearchers assume the existence of a carry-over effectof the adverse metabolic conditions during primaryfollicle growth early postpartum on the health of thepre-ovulatory follicle 2–3 months later (Britt 1992).Such follicles may be less capable of producing adequateamounts of oestrogens and progesterone (followingovulation) and might be doomed to contain an oocyteof inferior quality (Britt 1992; Roth et al. 2001a). Thedevelopmental capacity of the oocyte is intrinsicallylinked to the growth phase and health of the developingfollicle (Bilodeau-Goeseels and Panich 2002; Suttonet al. 2003; Lequarre et al. 2005). Last but not least, dietcomposition can also alter the endocrine and metabolicmicro-environment of the developing oocyte (Bolandet al. 2001; McEvoy et al. 2001; Kenny et al. 2002).A deviant endocrine environment, because of or as aconsequence of a NEB can alter oocyte quality throughvarious mechanisms such as spindle formation, prolongedfollicular growth and resumption of meioticprogression, all of which has been extensively reviewedearlier (Leroy et al. 2007). Only a few studies haveexamined possible effects of NEB-associated low glucose,elevated b-hydroxybutyrate (b-OHB) or NEFAconcentrations on oocyte quality. Apart from indirecteffects of hypoglycaemia in early postpartum dairy cows(through an effect on LH secretion or ovarian responsivenessto gonadotrophins), hypoglycaemic conditions(e.g. clinical ketosis) are reflected in the microenvironmentof the pre-ovulatory oocyte, and can compromisethe oocyte’s developmental capacity, because glucose isan indispensable molecule for proper oocyte maturation(Bilodeau-Goeseels 2006; for review see Sutton et al.2003; Leroy et al. 2004, 2006). Kruip and Kemp (1999)suggested possible direct toxic effects of high NEFAconcentrations at the ovarian level of the ovary. Indeed,in an in vitro maturation model, saturated long-chainfatty acids, reduced rates of maturation, fertilization,cleavage and blastocyst formation. Apoptosis, and evencumulus cell necrosis, during maturation could explainthese observations (Leroy et al. 2005b). Finally, elevatedammonia and urea concentrations in the FF, because ofan unbalanced diet and protein catabolism were toxicfor the oocyte (Sinclair et al. 2000; De Wit et al. 2001;Leroy et al. 2004).Early Pregnancy in High-producing DairyCows: Embryo QualityEarly embryonic death is a major cause of reproductivefailure in dairy cows accounting for up to a total 80%pregnancy losses (Santos et al. 2004a). There are fourmajor factors impinging on embryo quality in thespecific case of high-producing dairy cows: gametequality, corpus luteum quality combined with thecirculating progesterone concentration, uterine involutionand nutrition. Yet, only those that are related toNEB will be discussed.Adverse pre-ovulatory conditions, such as NEB, mayhave carry-over effects on embryo metabolism andviability resulting in early embryonic mortality (Yaakubet al. 1999; Lozano et al. 2003; Rhoads et al. 2006).Distinguishing oocyte effects on embryo quality frompost-fertilization influences is extremely difficult. Only atleast a 6-day-old embryo can be transferred to arecipient to assess the impact of uterine environmenton embryo quality. Lucy (2007) supports the conceptthat fertility could be improved in dairy cows by usingembryo transfer and thus circumventing the period ofoocyte and early embryonic development.Well-timed and balanced post-mating progesteroneconcentrations are vital for zygote viability as progesteronemodulates the endometrial secretions, and thusoptimal uterine receptivity (McEvoy et al. 1995). It hasbeen suggested that disappointing pregnancy results inmodern dairying are partially caused by the retardedonset of the progesterone rise and suboptimal progesteroneconcentrations during the luteal phase (Mannand Lamming 2001). Furthermore, the typical NEBobserved early postpartum can reduce the number ofovulatory oestrous cycles preceding AI which mayhamper adequate uterus preparation (Butler 2003).Villa-Godoy et al. (1988) showed that cows in NEBpostpartum had lower progesterone concentrationsduring the first three ovarian cycles following calving.Despite larger volumes of luteal tissue, compared withnon-lactating heifers, maximal progesterone concentrationsin lactating cows are lower, possibly because of ahigher rate of degradation in the liver (Sangsritavonget al. 2002; Sartori et al. 2004; Wiltbank et al. 2006).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Nutrient Prioritization and Fertility in Dairy Cows 99Good postpartum uterine involution, comprisingendometrium repair and evacuation of bacteriallycontaminated contents, is of critical importance forreproductive performance (for review, see Roche et al.2006). Because of a reduced immune response,negative energy status can dramatically delay thisprocess, jeopardizing future fertility (Wathes et al.2007).Apart from interest in the consequences of NEB andassociated endocrine and metabolic imbalances, there isa growing focus towards the effect of milk yieldpromotingdiets that are rich in energy and protein(for review, see Leroy et al. 2007). From these studies wecan learn that optimum nutritional conditions forfollicle growth and ovulation are not compatible withembryo survival and maintenance of pregnancy(O’Callaghan and Boland 1999).Finally, heat stress can cause a reduction in drymatter intake which prolongs the period of NEB,decreasing plasma concentrations of insulin, glucoseand IGF-I, while increasing GH and NEFA (Drew1999; Butler 2001). In addition, there is a direct effect ofheat stress on FSH (increased) and oestradiol(decreased) plasma concentrations (Wolfenson et al.1997; Wilson et al. 1998). This causes not only poorexpression of oestrus, but also delayed follicle selectionand thus has potentially adverse effects on oocytequality (Roth et al. 2001a,b). Heat stress may furthermoreincrease uterine temperature by decreasing bloodflow to the uterus. These changes inhibit embryonicdevelopment (Rivera and Hansen 2001), increase earlyembryonic loss and reduce the proportion of successfulinseminations (Garcia-Ispierto et al. 2007). The impactof this heat effect decreases as the embryo develops(Paula-Lopes et al. 2003).All the factors described above, potentially affectingfertility, are diagrammatically presented in Fig. 1.Some Clues to Modify the Somatotropic Axis inOrder to Generate Acceptable Fertility ResultsTackling the multifactorial problem of subfertility indairy cows is a real challenge as the ‘skewed somatotropicaxis’ described above is not the only reason forthe decline in fertility. Evolving farm systems, growingherd sizes, increased managerial demands combinedwith a reduced labour input per animal and increasedsusceptibility to diseases, all interfere with the ability ofthe dairy cow to be successfully bred (Mallard et al.1998). All these factors should be carefully addressedwhen formulating management advice to the dairy farmmanager, but this is beyond the scope of the presentpaper.Metabolic disorders (hypocalcaemia, ketosis andacidosis) and infectious diseases during the puerperiumare all key risk factors for efficient reproductive performance(Grohn and Rajala-Schultz 2000; Santos et al.2004b). Balanced and sophisticated birth managementcombined with strict follow-up of cow health statusearly postpartum is vital to prevent a drop in theanimal’s appetite. Accurate and repeated assessment ofBCS to estimate changes in body reserves is critical.Minimizing BCS changes and thus the ‘exhaustion ofthe energy reserves’ early postpartum requires anoptimal dietary strategy by reducing energy intakeduring the first weeks of the dry period then anincreased energy supply (carbohydrates) shortly prepartum(for review, see Overton and Waldron 2004).Yet, the beneficial effect of extra pre-partum energy onpostpartum energy balance is a matter of debate(Grummer 2007). Furthermore, adequate dietary modulationsduring the demanding early postpartum periodare a promising approach although difficult to achieve(Grummer 2007). Therefore, Van Knegsel et al. (2007a)fed on an isocaloric and isonitrogenous basis, a mainlyglucogenic (by-pass starch) or a mainly lipogenic diet(beetpulp, MEGALAC and palm oil) and showed thatthe glucogenic (or ‘insulinogenic’) diet, stimulates energypartitioning towards body reserves in early lactation.Cows fed the glucogenic diet had lower NEB andreduced body fat mobilization, which led to milk fatdepression and less energy partitioned to milk (VanKnegsel et al. 2007a). In a follow-up study (Van Knegselet al. 2007b), multiparous cows fed with glucogenic dietalso had higher plasma insulin concentrations andtended to resume ovarian activity earlier(20.4 ± 2.1 days) compared to cows fed with morelipogenic diet (26.1 ± 2.1 days).Fig. 1. Interaction between geneticselection for milk production andfertility. The imposed metabolicand endocrine changes sustainingmilk production in combinationwith the increased susceptibility toheat stress, diseases and suboptimalmanagement conditions allnegatively affect reproductive performanceof the high-producingdairy cowBreeding priorityfor milkproduction traitsBreeding valuefor fertility+Increased susceptibility to:Metabolic/infectious diseasesHeat stressSuboptimal management:DietHousing conditionsHerd sizeLabour per animalHeat detection+NUTRIENT PRIORITIZATION :Genetically drivendepletion of body reservesMETABOLIC ADAPTATIONS TOCATABOLIC STATUSNEB, BCS loss, reduced appetiteLow insulin, low IGF-I, high GHLow leptinHigh NEFA, high ketones, high urea,low glucosePeriphere insulin resistanceFERTILITYHypothalamo-pituitary-ovarian-axisoocyte/embryo qualitycorpus luteum qualityfollicular, oviductal, uterine environmentDirect or carry over effects– –Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

100 JLMR Leroy, T Vanholder, ATM Van Knegsel, I Garcia-Ispierto and PEJ BolsIt is speculated that changes in management are morelikely to have a positive effect on EB. Shortening or evenskipping the dry period improves dry matter intakeperipartum, reduces milk production in early lactation,improves energy balance and reduces the number ofdays postpartum till resumption of ovarian activity(Gumen et al. 2005; Rastani et al. 2005).Besides this, growing attention is being paid to dietaryfatty acid content and composition provided by supplementedby-pass fats during the early postpartum period.Not the effect on energy balance as such but improvedsteroid secretion and alteration of the fatty acid profile(more x-3 poly-unsaturated fatty acids), resulting inmodified prostaglandin metabolism (Thatcher et al.2006). Suppression of milk fat synthesis by supplementationof rumen-protected conjugated linoleic acids(trans-10, cis-12) has been suggested to restrict energyloss through milk (Castaneda-Gutierrez et al. 2005).Yet, in spite of several recent interesting papers, theoutcome on energy balance and fertility are equivocal.An extensive description and clear overview of nutritionalstrategies supporting the metabolic demandsduring the transition period are given by Overton andWaldron (2004) and are beyond the scope of the presentpaper.Finally, genetic selection programmes in the dairyindustry have emphasized milk production traits byunintended mobilization of cow body reserves. This lossin BCS is not only dependent on the available mass ofadipose tissue but also on a genetically determined setpointfor BCS. This set-point is correlated with reproductiveoutcome (Lucy 2007). Therefore, not onlyfertility traits as such (Royal et al. 2000), but alsovariables comprising changes in BCS early postpartumshould be included in genetic selection criteria.ConclusionsIntense selection for milk production has resulted in animmense priority for the high-producing dairy cow topartition energy to milk, at the cost of body reserves.This has resulted in excessive NEB and poor reproductiveperformance. Thus, milk production and reproductiveperformance have conflicting interests inhigh-producing dairy cows. Metabolites and metabolichormones associated with energy prioritizing for milkproduction (NEFA, insulin, glucose, IGF-1, b-OH)influence fertility, indirectly by modulating the somatotropic⁄ gonadotropic axis, as well as directly at theovary, follicle or uterine environment. Strict follow-upperipartum to monitor health and BCS loss and directtreatment of (infectious or metabolic) disorders in earlylactation will limit fertility disorders postpartum. Furthermore,a series of promising management, geneticselection and nutritional strategies have been proposed,which have the potential to shift the somatotropic axisprioritizing energy partitioning of milk to a somatotropicaxis with an increased priority for body reserves toimprove fertility. Yet, research in this area is limited.Exploring such strategies, comparing their benefit oreven combining two or more strategies is an extremelyinteresting area of research, and essential to improvehealth and welfare of the modern dairy cow.ReferencesAmstalden M, Harms PG, Welsh TH Jr, Randel RD, WilliamsGL, 2005: Effects of leptin on gonadotropin-releasinghormone release from hypothalamic-infundibular explantsand gonadotropin release from adenohypophyseal primarycell cultures: further evidence that fully nourished cattle areresistant to leptin. Anim Reprod Sci 85, 41–52.Armstrong DG, Baxter G, Hogg CO, Woad KJ, 2002a:Insulin-like growth factor (IGF) system in the oocyte andsomatic cells of bovine preantral follicles. Reproduction 123,789–797.Beam SW, Butler WR, 1997: Energy balance and ovarianfollicle development prior to first ovulation postpartum indairy cows receiving three levels of dietary fat. Biol Reprod56, 133–142.Beam SW, Butler WR, 1999: Effects of energy balance onfollicular development and first ovulation in postpartumdairy cows. J Reprod Fertil Suppl 54, 411–424.Bilodeau-Goeseels S, 2006: Effect of culture media and energysources on the inhibition of nuclear maturation in bovineoocytes. Theriogenology 66, 297–306.Bilodeau-Goeseels S, Kastelic JP, 2003: Factors affectingembryo survival and strategies to reduce embryonic mortalityin cattle. Can J Anim Sci 83, 659–671.Bilodeau-Goeseels S, Panich P, 2002: Effects of quality ondevelopment and transcriptional activity in early bovineembryos. Anim Reprod Sci 71, 143–155.Blache D, Zhang S, Martin GB, 2006: Dynamic and integrativeaspects of the regulation of reproduction by metabolicstatus in male sheep. Reprod Nutr Dev 46, 379–390.Blache D, Chagas LM, Martin GB, 2007: Nutritional inputsinto the reproductive neuroendocrine control system – amultidimensional perspective. Reproduction, Supplement64, 124–139.Boland MP, Lonergan P, O’Callaghan D, 2001: Effect ofnutrition on endocrine parameters, ovarian physiology, andoocyte and embryo development. Theriogenology 55, 1323–1340.Bols PEJ, Ysebaert MT, Lein A, Coryn M, Van Soom A, deKruif A, 1998: Effects of long-term treatment with bovinesomatotropin on follicular dynamics and subsequent oocyteand blastocyst yield in an OPU-IVF program. Theriogenology49, 983–995.Bossaert P, Leroy JLMR, Cools S, Opsomer G, 2007: Impactof metabolic and endocrine parameters on ovulation in highyieldingdairy cows (abstract). Reprod Domest Anim 42,118.Bousquet D, Bouchard E, Du Tremblay D, 2004: Decreasingfertility in dairy cows: myth or reality? Le MédecinVéte´rinaire 34, 59–61.Britt JH, 1992: Impacts of early postpartum metabolism onfollicular development and fertility. Proc Annu ConventionAm Assoc Bovine Pract 24, 39–43.Butler WR, 2001: Nutritional effects on resumption of ovariancyclicity and conception rate in postpartum dairy cows.Anim Sci Occas Publ 26, 133–145.Butler WR, 2003: Energy balance relationships with folliculardevelopment, ovulation and fertility in pp dairy cows.Livestock Prod Sci 83, 211–218.Butler ST, Marr AL, Pelton SH, Radcliff RP, Lucy MC, ButlerWR, 2003: Insulin restores GH responsiveness duringlactation-induced negative energy balance in dairy cattle:effects on expression of IGF-I and GH receptor 1A. JEndocrinol 176, 205–217.Butler ST, Pelton SH, Butler WR, 2004: Insulin increases 17bestradiolproduction by the dominant follicle of the firstpostpartum follicle wave in dairy cows. Reproduction 127,537–545.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Nutrient Prioritization and Fertility in Dairy Cows 101Castaneda-Gutierrez E, Overton TR, Butler WR, BaumanDE, 2005: Dietary supplements of two doses of calcium saltsof conjugated linoleic acid during the transition period andearly lactation. J Dairy Sci 88, 1078–1089.Chagas LM, Bass JJ, Blache D, Burke CR, Kay JK,Lindsay DR, Lucy MC, Martin GB, Meier S, RhodesFM, Roche JR, Thatcher WW, Webb R, 2007: Newperspectives on the roles of nutrition and metabolicpriorities in the subfertility of high-producing dairy cows.J Dairy Sci 90, 4022–4032.Cnop M, Hannaert JC, Hoorens A, Eizirik DL, Pipeleers DG,2001: Inverse relationship between cytotoxicity of free fattyacids in pancreatic islet cells and cellular triglycerideaccumulation. Diabetes 50, 1771–1777.Coffey MP, Simm G, Oldham JD, Hill WG, Brotherstone S,2004: Genotype and diet effects on energy balance in the firstthree lactations of dairy cows. J Dairy Sci 87, 4318–4326.De Wit AAC, Cesar MLF, Kruip TAM, 2001: Effect of ureaduring in vitro maturation on nuclear maturation andembryo development of bovine cumulus-oocyte-complexes.J Dairy Sci 84, 1800–1804.Diskin MG, Mackey DR, Roche JF, Sreenan JM, 2003:Effects of nutrition and metabolic status on circulatinghormones and ovarian follicle development in cattle. AnimReprod Sci 78, 345–370.Drew B, 1999: Practical nutrition and management of heifersand high yielding dairy cows for optimal fertility. CattlePract 7, 243–248.Dunne LD, Diskin MG, Boland MP, O’Farrell KJ, SreenanJM, 1999: The effect of pre- and post-insemination plane ofnutrition on embryo survival in beef heifers. Anim Sci 69,411–417.Garcia-Ispierto I, Lopez-Gatius F, Bech-Sabat G, SantolariaP, Yaniz JL, Nogareda C, De Rensis F, Lopez-Bejar M,2007: Climate factors affecting conception rate of highproducing dairy cows in north-eastern Spain. Theriogenology67, 1379–1385.Garnsworthy PC, Topps JH, 1982: The effect of bodycondition score at calving on their food intake and performancewhen given complete diets. Anim Prod 35, 113–119.Gong JG, 2002: Influence of metabolic hormones and nutritionon ovarian follicle development in cattle: practicalimplications. Domest Anim Endocrinol 23, 229–241.Grohn YT, Rajala-Schultz PJ, 2000: Epidemiology of reproductiveperformance in dairy cows. Anim Reprod Sci 60,605–614.Grummer RR, 2007: Strategies to improve fertility of highyielding dairy farms: management of the dry period.Theriogenology 68, S281–S288.Gumen A, Rastani RR, Grummer RR, Wiltbank MC, 2005:Reduced dry periods and varying prepartum diets alterpostpartum ovulation and reproductive measures. J DairySci 88, 2401–2411.Gutierrez-Aguilar CG, 1997: The effect of nutrition andmetabolic hormones on follicular development in cattle.PhD Dissertation, University of Edinburgh, pp. 146–166.Herdt TH, 2000: Ruminant adaptation to negative energybalance. Vet Clin North Am Food Anim Pract 16, 215–230.Huirne RBM, Saatkamp HW, Bergevoet RHM, 2002: Economicanalysis of common health problems in dairy cattle.In: Kaske M, Scholz H, Ho¨ ltershinken M(eds), TwelfthWorld Buiatrics Congress, Hannover, Germany, 18–23August 2002, pp. 420–431.Ichimaru T, Mori Y, Okamura H, 2001: A possible role ofneuropeptide Y as a mediator of undernutrition to thehypothalamic gonadotropin-releasing hormone pulse generatorin goats. Endocrinology 142, 2389–2498.Kacsoh B, 2000: Endocrine Physiology. McGraw Hill BookCo., New York, NY.Kawashima C, Fukihara S, Maeda M, Kaneko E, MontoyaCA, Matsui M, Shimizu T, Matsunaga N, Kida K, MiyakeY, Schams D, Miyamoto A, 2007: Relationship betweenmetabolic hormones and ovulation of dominant follicleduring the first follicular wave post-partum in high-producingdairy cows. Reproduction 133, 155–163.Kenny DA, Humpherson PG, Leese HJ, Morris DG, TomosAD, Diskin MG, Sreenan JM, 2002: Effect of elevatedsystemic concentrations of ammonia and urea on themetabolite and ionic composition of oviductal fluid incattle. Biol Reprod 66, 1797–1804.Kruip TAM, Kemp B, 1999: Voeding en vruchtbaarheid bijlandbouwhuisdieren. Tijdschr Diergeneeskd 124, 462–467.Landau S, Braw-Tal R, Kaim M, Bor A, Bruckental I, 2000:Preovulatory follicular status and diet affect the insulin andglucose content of follicles in high-yielding dairy cows.Anim Reprod Sci 64, 181–197.Lequarre AS, Vigneron C, Ribaucour F, Holm P, Donnay I,Dalbies-Tran R, Callesen H, Mermillod P, 2005: Influenceof antral follicle size on oocyte characteristics and embryodevelopment in the bovine. Theriogenology 63, 841–859.Leroy JLMR, Vanholder T, Delange JR, Opsomer G, VanSoom A, Bols PEJ, Dewulf J, de Kruif A, 2004: Metabolicchanges in follicular fluid of the dominant follicle in highyieldingdairy cows early post partum. Theriogenology 62,1131–1143.Leroy JLMR, Opsomer G, De Vliegher S, Vanholder T,Goossens L, Geldhof A, Bols PEJ, de Kruif A, Van SoomA, 2005a: Comparison of embryo quality in high-yieldingdairy cows, in dairy heifers and in beef cows. Theriogenology64, 2022–2036.Leroy JLMR, Vanholder T, Mateusen B, Christophe A,Opsomer G, de Kruif A, Genicot G, Van Soom A, 2005b:Non-esterified fatty acids in follicular fluid of dairy cowsand their effect on developmental capacity of bovine oocytesin vitro. Reproduction 130, 485–495.Leroy JLMR, Vanholder T, Opsomer G, Van Soom A, deKruif A, 2006: The in vitro development of bovine oocytesafter maturation in glucose and beta-hydroxybutyrate concentrationsassociated with negative energy balance in dairycows. Reprod Domest Anim 41, 119–123.Leroy JLMR, Opsomer G, Van Soom A, Goovaerts IGF, BolsPEJ, 2007: Reduced fertility in high yielding dairy cows: arethe oocyte and embryo in danger? Part I: The importance ofnegative energy balance and altered corpus luteum functionto the reduction of oocyte and embryo quality in highyielding dairy cows (review). Reprod Domest Anim, inpress. DOI:10.1111/j.1439–0531.2007.00960.Liefers SC, Veerkamp RF, te Pas MFW, Delavaud C,Chilliard Y, van der Lende T, 2003: Leptin concentrationsin relation to energy balance, milk yield, intake, live weightand estrus in dairy cows. J Dairy Sci 86, 799–807.Lopez H, Satter LD, Wiltbank MC, 2004: Relationshipbetween level of milk production and oestrus behavior oflactating dairy cows. Anim Reprod Sci 81, 209–223.Lozano JM, Lonergan P, Boland MP, O’Callaghan D, 2003:Influence of nutrition on the effectiveness of superovulationprogrammes in ewes: effect on oocyte quality andpost-fertilization development. Reproduction 125, 543–553.Lucy MC, 2001: Reproductive loss in high-producing dairycattle: where will it end? J Dairy Sci 84, 1277–1293.Lucy MC, 2003: Mechanisms linking nutrition and reproductionin postpartum cows. Reprod Suppl. 61, 415–427.Lucy MC, 2007: Fertility in high-producing dairy cows:reasons for decline and corrective strategies for sustainableimprovement. Reprod Suppl. 64, 237–254.Maedler K, Spinas GA, Dyntar D, Moritz W, Kaiser N,Donath MY, 2001: Distinct effects of saturated and mono-Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

102 JLMR Leroy, T Vanholder, ATM Van Knegsel, I Garcia-Ispierto and PEJ Bolsunsaturated fatty acids on beta-cell turnover and function.Diabetes 50, 69–76.Mallard BA, Dekkers JC, Ireland MJ, Leslie KE, Sharif S,Vankampen CL, Wagter L, Wilkie BN, 1998: Alteration inimmune responsiveness during the peripartum period and itsramification on dairy cow and calf health. J Dairy Sci 81,585–595.Mann GE, Lamming GE, 2001: Relationship between maternalendocrine environment, early embryo development andinhibition of the luteolytic mechanism in cows. Reproduction121, 175–180.McEvoy TG, Robinson JJ, Aitken RP, Findlay PA, PalmerRM, Robertson IS, 1995: Dietary-induced suppression ofpre-ovulatory progesterone concentrations in superovulatedewes impairs the subsequent in vivo and in vitro developmentof their ova. Anim Reprod Sci 39, 89–107.McEvoy TG, Robinson JJ, Ashworth CJ, Rooke JA, SinclairKD, 2001: Feed and forage toxicants affecting embryosurvival and fetal development. Theriogenology 55, 113–129.Nishimoto H, Matsutani R, Yamamoto S, Takahashi T,Hayashi KG, Miyamoto A, Hamano S, Tetsuka M, 2006:Gene expression of glucose transporter (GLUT) 1, 3 and 4in bovine follicle and corpus luteum. J Endocrinol 188, 111–119.O’Callaghan D, Boland MP, 1999: Nutritional effects onovulation, embryo development and the establishment ofpregnancy in ruminants. Anim Sci 68, 299–314.Ohkura S, Ichimaru T, Itoh F, Matsuyama S, Okamura H,2004: Further evidence for the role of glucose as a metabolicregulator of hypothalamic gonadotropin-releasing hormonepulse generator activity in goats. Endocrinology 145, 3239–3246.Opsomer G, Coryn M, Deluyker H, de Kruif A, 1998: Ananalysis of ovarian dysfunction in high yielding dairy cowsafter calving based on progesterone profiles. Reprod DomestAnim 33, 193–204.Ortega HH, Palomar MM, Acosta JC, Salvetti NR, DallardBE, Lorente JA, Barbeito CG, Gimeno EJ, 2008: Insulinlikegrowth factor I in sera, ovarian follicles and follicularfluid of cows with spontaneous or induced cystic ovariandisease. Res Vet Sci 84, 419–427.Overton TR, Waldron MR, 2004: Nutrtional management oftransition dairy cows: strategies to optimize metabolichealth. J Dairy Sci 87, E105–E119.Patton J, Kenny DA, Mee JF, O’Mara FP, Wathes DC, CookM, Murphy JJ, 2006: Effect of milking frequency and diet onmilk production, energy balance and reproduction in dairycows. J Dairy Sci 89, 1478–1487.Paula-Lopes FF, Al-Katanani YM, Majewski AC, McDowellLR, Hansen PJ, 2003: Manipulation of antioxidant statusfails to improve fertility of lactating cows or survival of heatshockedembryos. J Dairy Sci 86, 2343–2351.Pfaff D, 2005: Hormone-driven mechanisms in the centralnervous system facilitate the analysis of mammalian behaviours.J Endocrinol 184, 447–453.Pires JAA, Souza AH, Grummer RR, 2007: Induction ofhyperlipidemia by intravenous infusion of Tallow emulsioncauses insulin resistance in Holstein cows. J Dairy Sci 90,2735–2744.Rastani RR, Grummer RR, Bertics SJ, Gumen A, WiltbankMC, Mashek DG, Schwab MC, 2005: Reducing dry periodlength to simplify feeding transition cows: milk production,energy balance, and metabolic profiles. J Dairy Sci 88, 1004–1014.Rhoads ML, Rhoads RP, Gilbert RO, Toole R, Butler WR,2006: Detrimental effects of high plasma urea nitrogen levelson viability of embryos from lactating dairy cows. AnimReprod Sci 91, 1–10.Rivera RM, Hansen PJ, 2001: Development of cultured bovineembryos after exposure to high temperatures in the physiologicalrange. Reproduction 121, 107–115.Roche JF, 2006: The effect of nutritional management of thedairy cow on reproductive efficiency. Anim Reprod Sci 96,282–296.Roche JR, Berry DP, Kolver ES, 2006: Holstein-Friesianstrains and feed effects on milk production, bodyweight andbody condition score profiles in grazing dairy cows. J DairySci 89, 3532–3543.Roche JR, Macdonald KA, Burke CR, Lee JM, Berry DP,2007: Associations among body condition score, bodyweight and reproductive performance in seasonal-calvingdairy cattle. J Dairy Sci 90, 376–391.Roth Z, Meweidan R, Shaham-Albalancy A, Braw-Tal R,Wolfenson D, 2001a: Delayed effect of heat stress on steroidproduction in medium-size and preovulatory bovine follicles.Reproduction 121, 745–751.Roth Z, Arav A, Bor A, Zeron Y, Braw-Tal R, Wolfenson D,2001b: Improvement of quality of oocytes collected in theautumn by enhanced removal of impaired follicles frompreviously heat-stressed cows. Reproduction 122, 737–744.Royal M, Mann GE, Flint APF, 2000: Strategies for reversingthe trend towards subfertility in dairy cattle. Vet J 160, 53–60.Sangsritavong S, Combs DK, Sartori R, Armentano LE,Wiltbank MC, 2002: High feed intake increase liver bloodflow and metabolism of progesterone and estradiol-17b indairy cattle. J Dairy Sci 85, 2831–2842.Santos JEP, Thatcher WW, Chebel RC, Cerri RLA, GalvaoKN, 2004a: The effect of embryonic death rates in cattle onthe efficacity of estrus synchronization programs. AnimReprod Sci 82 ⁄ 83, 513–535.Santos JE, Cerri RL, Ballou MA, Higginbotham GE, Kirk JH,2004b: Effect of timing of first clinical mastitis occurrence onlactational and reproductive performance of Holstein dairycows. Anim Reprod Sci 80, 31–45.Sartori R, Sartor-Bergfelt R, Mertens SA, Guenther JN,Parrish JJ, Wiltbank MC, 2002: Fertilization and earlyembryonic development in heifers and lactating cows insummer and lactating and dry cows in winter. J Dairy Sci 85,2803–2812.Sartori R, Haughian JM, Shaver RD, Rosa GJM, WiltbankMC, 2004: Comparison of ovarian function and circulatingsteroids in estrous cycles of Holstein heifers and lactatingcows. J Dairy Sci 87, 905–920.Silvia WJ, 2003: Addressing the decline in reproductiveperformance of lactating dairy cows: a researcher’s perspective.Vet Sci Tomorrow 3, 1–5.Sinclair KD, Kuran M, Gebbie FE, Webb R, McEvoy TG,2000: Nitrogen metabolism and fertility in cattle: II.Development of oocytes recovered from heifers offered dietsdiffering in their rate of nitrogen release in the rumen. JAnim Sci 78, 2670–2680.Snijders SE, Dillon P, O’Callaghan D, Boland MP, 2000:Effect of genetic merit, milk yield, body condition andlactation number on in vitro oocyte development in dairycows. Theriogenology 53, 981–989.Spicer LJ, Echternkamp SE, 1995: The ovarian insulin andinsulin-like growth factor system with an emphasis ondomestic animals. Domest Anim Endocrinol 12, 223–245.Spicer LJ, Francisco CC, 1997: The adipose obese geneproduct, leptin: evidence of a direct inhibitory role inovarian function. Endocrinology 138, 3374–3379.Spicer LJ, Francisco CC, 1998: Adipose obese gene product,leptin, inhibits bovine ovarian thecal cell steroidogenesis.Biol Reprod 58, 207–212.Spicer LJ, Chamberlain CS, Francisco CC, 2000: Ovarianaction of leptin: effects on insulin-like growth factor-I-Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Nutrient Prioritization and Fertility in Dairy Cows 103stimulated function of granulosa and thecal cells. Endocr 12,53–59.Sutton ML, Gilchrist RB, Thompson JG, 2003: Effects of invivo and in vitro environments on the metabolism of thecumulus-oocyte complex and its influence on oocyte developmentalcapacity. Hum Reprod Update 9, 35–48.Tamminga S, Luteijn PA, Meijer RGM, 1997: Changes incomposition and energy content of liveweight loss in dairycows with time after parturition. Livestock Prod Sci 52, 129–155.Thatcher WW, Bilby TR, Bartolome JA, Silvestre F, StaplesCR, Santos JEP, 2006: Strategies for improving fertility inthe modern dairy cow. Theriogenology 65, 30–44.Van Knegsel ATM, Van den Brand H, Dijkstra J, van StraalenWM, Heetkamp MJW, Tamminga S, Kemp B, 2007a:Dietary energy source in dairy cows in early lactation:energy partitioning and milk composition. J Dairy Sci 90,1467–1476.Van Knegsel ATM, Van den Brand H, Dijkstra J, van StraalenWM, Jorritsma R, Tamminga S, Kemp B, 2007b: Effect ofglucogenic vs. lipogenic diets on energy balance, bloodmetabolites, and reproduction in primiparous and multiparousdairy cows in early lactation. J Dairy Sci 90, 3397–3409.Vanholder T, Leroy JLMR, Dewulf J, Duchateau L, CorynM, de Kruif A, 2005a: Hormonal and metabolic profiles ofhigh-yielding dairy cows prior to ovarian cyst formation orfirst ovulation post partum. Reprod Domest Anim 40, 460–469.Vanholder T, Leroy JLMR, Van Soom A, Opsomer G, MaesD, Coryn M, de Kruif A, 2005b: Effect of non-esterifiedfatty acids on bovine granulosa cell steroidogenesis andproliferation in vitro. Anim Reprod Sci 87, 33–44.Vanholder T, Opsomer G, de Kruif A, 2006a: Aetiology andpathogenesis of cystic ovarian follicles in dairy cattle: areview. Reprod Nutr Dev 46, 105–119.Vanholder T, Leroy JLMR, Van Soom A, Maes D, Coryn M,Fiers T, de Kruif A, Opsomer G, 2006b: Effect of nonesterifiedfatty acids on bovine theca cell steroidogenesis andproliferation in vitro. Anim Reprod Sci 92, 51–63.Vernon RG, 2002: Nutrient partitioning, lipid metabolism andrelevant imbalances. Proceedings of the 12th World BuiatricsCongress, Hannover, Germany, 18–23 August, 2002.Villa-Godoy A, Hughes TL, Emery RS, Chapin LT, FogwellRL, 1988: Association between energy balance and lutealfunction in lactating dairy cows. J Dairy Sci 71, 1063–1072.de Vries MJ, Veerkamp RF, 2000: Energy balance of dairycattle in relation to milk production variables and fertility.J Dairy Sci 83, 62–69.Wade GN, Jones JE, 2004: Neuroendocrinology of nutritionalinfertility. Am J Physiol. Regul, Integrative Comp Physiol287, 1277–1296.Wathes DC, Fenwick M, Cheng Z, Bourne N, Llewellyn S,Morris DG, Kenny D, Murphy J, Fitzpatrick R, 2007:Influence of negative energy balance on cyclicity and fertilityin the high producing dairy cow. Theriogenology 68, S232–S241.Webb R, Garnsworthy PC, Gong JG, Robinson RS, WathesDC, 1999: Consequences for reproductive function ofmetabolic adaptation to load. Anim Sci Occasion Publ 24,99–112.Wilson SJ, Marion RS, Spain JN, Spiers DE, Keisler DH,Lucy MC, 1998: Effect of controlled heat stress on ovarianfunction in dairy cattle: I. Lactating cows. J Dairy Sci 1,2124–2131.Wiltbank MC, Sartori R, Sangsritavong S, Lopez H, HaughianJM, Fricke PM, Gumen A, 2001: Novel effects ofnutrition on reproduction in lactating dairy cows. J DairySci Suppl 84, 32 (abstract).Wiltbank M, Lopez H, Sartori R, Sangsritavong S, Gu¨ men A,2006: Changes in reproductive physiology of lactating dairycows due to elevated steroid metabolism. Theriogenology65, 17–29.Wolfenson D, Lew BJ, Thatcher WW, Graber Y, Meidan R,1997: Seasonal and acute heat stress effects on steroidproduction by dominant follicles in cow. Anim Reprod Sci47, 9–19.Wynne K, Stanley S, McGowan B, Bloom S, 2005: Appetitecontrol. J Endocrinol 184, 291–318.Yaakub H, O’Callaghan D, Boland MP, 1999: Effect ofroughage type and concentrate supplementation on folliclenumbers and in vitro fertilisation on follicle numbers and invitro fertilisation and development of oocytes recoveredfrom beef heifers. Anim Reprod Sci 44, 1–12.Zhao F, Dixon WT, Kennelly JJ, 1996: Localization and geneexpression of glucose transporters in bovine mammarygland. Comprehens Biochem Physiol 115, 127–134.Zieba DA, Amstalden M, Williams GL, 2005: Regulatoryroles of leptin in reproduction and metabolism: a comparativereview. Domest Anim Endocrinol 29, 166–185.Author’s address (for correspondence): J Leroy, Laboratory forVeterinary Physiology, Department of Veterinary Sciences, Facultyof Pharmaceutical, Biomedical and Veterinary Sciences, University ofAntwerp, Universiteitsplein 1, Gebouw U, B-2610 Wilrijk, Belgium.E-mail: jo.leroy@ua.ac.beConflict of interests: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 104–112 (2008); doi: 10.1111/j.1439-0531.2008.01149.xISSN 0936-6768Corpus Luteum–Endometrium–Embryo Interactions in the Dairy Cow: UnderlyingMechanisms and Clinical RelevanceRS Robinson 1 , AJ Hammond 2 , DC Wathes 3 , MG Hunter 2 and GE Mann 21 School of Veterinary Medicine and Science; 2 School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough, Leics;3 Department of Veterinary Basic Sciences, Royal Veterinary College, Hawkshead Lane, North Mymms, Hatfield, Hertfordshire, UKContentsConception rates of dairy cows are currently declining at anestimated 1% every year. Approximately, 35% of embryos failto prevent luteolysis during the first three weeks of gestation.Interactions between the corpus luteum, endometrium andembryo are critical to the successful establishment of pregnancyand inadequacies will result in the mortality of theembryo. For example, as little as a one day delay in the postovulatoryrise of progesterone has serious consequences forembryo development and survival. Recently, we found thatLH support, degree of vascularization and luteal cell steroidogeniccapacity were not the major factors responsible forthis luteal inadequacy, but are nevertheless essential for lutealdevelopment and function. Progesterone acting on its receptorin the endometrium stimulates the production of endometrialsecretions on which the free-living embryo is dependent.However, their exact composition and effects of inadequateprogesterone remains to be determined. The embryo isrecognized through its secretion of interferon tau (IFNT),which suppresses luteolytic pulses of prostaglandin F 2a . In thecow, it is most likely that IFNT inhibits oxytocin receptor upregulationdirectly and does not require the prior inhibition ofoestrogen receptor a (ESR1). Unravelling the precise lutealendometriumand embryo interactions is essential for us tounderstand pregnancy establishment and development ofstrategies to reverse the declining fertility of dairy cows.IntroductionOver the past 20 years, in the UK, calving rate to firstservice has fallen from approximately 60% to 40%(Royal et al. 2000), a decline of 1% every year. Similardeclines have been seen in the USA (Lucy 2001).However, while conception rates in lactating cows havefallen, conception rates in maiden heifers have not.While the decline in fertility in lactating cows is welldocumented, the precise causes are still the subject ofconsiderable debate, although involvement of increasedproductivity is irrefutable. The timing and extent ofpregnancy losses have been reviewed extensively bySreenan and Diskin (1986) and Peters (1996). Yet, whileit is clear that in a typical dairy herd

CL–Endometrium–Embryo Interactions 105mechanisms behind inadequate production of thesehormones.Pre-ovulatory Events and Subsequent EmbryoDevelopmentThe post-ovulatory embryo–endometrium–ovary interactionsbegin with the intimate inter-play between thefollicle and oocyte. Moreover, follicular and oocyte‘quality’ are undoubtedly linked (Wathes et al. 2003;Hunter et al. 2005). The mammalian oocyte undergoessignificant changes while enclosed within the follicle,particularly when it becomes dominant and ovulationapproaches. Oocytes recovered from large antral folliclesdevelop to blastocysts at significantly higher ratesthan those from smaller follicles (Lonergan et al. 1994),suggesting that the stage of follicle development atwhich the oocyte is removed is important in determiningits development. Various studies have shown thatnot only a higher proportion of in vivo-matured oocytesreach the blastocyst stage compared with their in vitromaturedcounterparts (Greve et al. 1987; Rizos et al.2002), but also that it is the late pre-ovulatory stagethat is critical for oocytes to obtain developmentalcapacity (Humblot et al. 2005). This suggests that thein vitro environment is inadequate to support oocytematuration and stimulate the correct pattern of geneexpression, as it occurs in vivo. Precisely, what is lackingin the in vitro environment is currently unknown, but isthe focus of much research (Lonergan 2007). Inaddition to affecting oocyte competence, abnormalitiesin pre-ovulatory follicle function (such as that whichoccurs, for example, during resumption of ovulation inpost-partum dairy cows) may also detrimentally affectoestradiol production and the number and differentiatedstate of granulosa and theca cells; this maysubsequently impact on the growth and developmentof the corpus luteum (CL) and its ability to produceprogesterone (Wathes et al. 2003; Robinson et al.2005).Post-ovulatory Period: The Initiation ofMaternal–Embryonic Interactions (Day 1–5)There are three critical events during the post-ovulatoryperiod. First, the cow must create an appropriateoviductal environment and secretions to nurture thedevelopment of the zygote to a morula. The oviductprovides support for the free-living embryo not only inthe form of nutrients (e.g. ions, amino acids andglucose) but also local growth factors such as insulinlikegrowth factors (IGF) 1 and 2. The IGFs may besecreted into the oviductal lumen, where they could actdirectly on the developing embryo. Alternatively, theymay stimulate oviductal secretions by acting on theIGF type 1 receptor (IGF1R), which is located in theglandular epithelial cells (Pushpakumara et al. 2002).These hypotheses are supported by the observationsthat IGF1 was able to stimulate the formation ofin vitro blastocysts (Moreira et al. 2002). Furthermore,the oviductal expression patterns of IGFBP2 andIGFBP6 are altered for cows in negative energy balance(Fenwick et al. 2008). As both these two bindingproteins have higher affinity for IGF2 than IGF1, thislikely alters the IGF signalling pathways within theoviductal environment. The ovary is the principalconductor for these changes to occur in a timely andsite-specific manner. For example, oestradiol up-regulatesIGF1 mRNA expression in the oviduct (Pushpakumaraet al. 2002) thereby creating an ‘IGF-rich’environment for the imminent arrival of the embryo. Inturn, the embryo regulates the bioavailability of theseIGFs by secreting IGF-binding proteins (IGFBP),thereby creating the appropriate environment for itsdevelopment (Watson et al. 1999). Secondly, embryonictransport through the oviduct requires the coordinatedaction of cilia and oviductal smooth muscle contractionsto enable the timely entrance of the morula intothe uterus.Thirdly, the formation of the CL and the rise in postovulatoryprogesterone are of critical importance. Thisfollicular-luteal transition is a dynamic process involvinga series of biochemical and morphological changesin the pre-ovulatory follicle induced by the LH surge(Reynolds and Redmer 1999; Niswender et al. 2000;Wathes et al. 2003). These include basement membranebreakdown, theca and granulosa differentiation intosmall and large luteal cells, respectively, tissue remodellingand growth. Concomitantly, there is increasedsteroidogenesis (through the up-regulation of 3b-HSDand p450scc) and a switch in steroidogenesis toprogesterone synthesis (through the down-regulationof aromatase). Progesterone production is also increasedthrough the up-regulation of key regulatoryproteins, such as steroid acute regulatory protein andperipheral-type benzodiazepine receptors, which controlthe rate-limiting transport of cholesterol from thecytoplasm to the inner mitochondrial membrane. Progesteroneproduction and its intricate control have beenextensively investigated and expertly reviewed (e.g.Niswender et al. 2000). Consequently, peripheral progesteroneconcentrations start to increase approximatelyby day 4 and by day 8–10 reach maximalconcentrations. It is this rapid decline in oestradiolconcentrations and subsequent rise in the progesteronethat provides the timely control of both oviductal andendometrial functions. Indeed, Green et al. (2005)showed that as early as day 5, the embryos from cowswith greater progesterone concentrations showed moreadvanced development.Formation of the Blastocyst and Progesterone(Day 5–12)Upon entry into the uterus, the blastocyst developswith the formation of the blastocoele surrounded bythe trophectoderm and it is in this window thatprogesterone is likely to have its greatest influence onthe development of the embryo (Mann et al. 1999).This is supported by the observation that progesteronereceptor (PGR) levels are maximal in both endometrialglands and sup-epithelial stroma from day 4 to day 10(Boos et al. 1996; Kimmins and MacLaren 2001;Robinson et al. 2001). Thus, it is likely that oestradiolinitiates PGR up-regulation during the follicularphase. PGR levels further increase in the early lutealÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

106 RS Robinson, AJ Hammond, DC Wathes, MG Hunter and GE Mannphase while progesterone concentrations are still lowand ⁄ or are further stimulated by the rise in plasmaoestradiol from the first dominant follicle on approximatelyday 4–5 of the oestrous cycle. Then, PGRconcentrations decline after a period of progesteroneexposure. Similar observations have been reported insheep (Wathes and Hamon 1993; Spencer and Bazer1995).The endometrial glands synthesize, secrete and ⁄ ortransport a complex mixture of amino acids, ions,glucose, transport proteins and growth factor calledhistotroph (Bazer 1975). These secretions are essentialfor the development of the blastocyst, which is freelivingduring the pre-attachment period. The requirementof these secretions has been demonstrated using anovine uterine gland knock out (UGKO) model (Grayet al. 2001). In this model, uterine gland development issuppressed by progesterone administration to the neonatalewe lamb. In UGKO sheep, embryos developinitially but are arrested at the blastocyst stage, indicatingthat glandular secretions are essential for posthatchingdevelopment. It is progesterone that controlsthese endometrial secretions acting through glandularPGR. This is supported by observations that rapidblastocyst development correlates with rising progesteroneconcentrations (Sreenan and Diskin 1986; Mannand Lamming 2001). Recent transcriptomic analyseshave helped in the identification of the molecularpathways that are involved in regulating endometrialfunctions during the bovine oestrous cycle (Bauersachset al. 2005; Mitko et al. 2008). These studies havehighlighted the genes associated with proliferation,cytoskeleton, extra-cellular matrix structure, remodellingand cell motility which were up-regulated duringoestrus. While, genes up-regulated during the lutealphase were involved in various cellular processesincluding the immune response, prostaglandin metabolism,cell adhesion, angiogenesis, tricarboxylic acid cycleenzymes, tight junctions and transport proteins. Thelatter group included those involved in glutamate, metalion, selenium and thyroid hormone transport. These areparticularly interesting as they are likely to control thecomposition of the histotroph. These observationssupport the hypothesis that luteal inadequacy adverselyaffects embryonic development through the deficient upregulationof these transporters. As well as havingfunctional effects, progesterone also exerts profoundstructural changes in the endometrium during thisperiod. For example, progesterone induces the reductionin endometrial area and increases the glandular ductdensity (Wang et al. 2007). This observation raises theexciting hypothesis that poor embryo development is theresult of reduced glandular formation. Finally, aninteresting observation from our studies in sheep andcows is the presence of immunostaining of non-nuclearprogesterone receptor along the apical surface of theluminal epithelium (Wathes and Hamon 1993; Robinsonet al. 2001). Furthermore, the appearance of thisstaining pattern was regulated through the oestrouscycle with the highest intensity being observed on day 12(Robinson et al. 2001). The function and regulation ofthis membrane-located progesterone receptor remains tobe elucidated.Blastocyst Elongation, Pregnancy Recognitionand Interferon tau (Day 12–20)Inhibition of the luteolytic mechanismThe embryo establishes pregnancy by blocking theluteolytic pulses of prostaglandin (PG) F 2a . The keyinitiation step in luteolysis is the up-regulation of theoxytocin receptor (OXTR) in the endometrial luminalepithelium (Wathes and Lamming 1995; Mann et al.1999). The exact mechanism by which OXTR are upregulatedand the role that oestrogen receptor a (ESR1)plays has been widely debated. However, it is acceptedthat progesterone initially suppresses OXTR and ESR1expression, but after a period of progesterone exposure(10 days) this block is lost (Lamming and Mann 1995;Spencer and Bazer 1995; Wathes et al. 1996). Also,during the follicular phase, rising oestradiol concentrationsinduce further up-regulation of OXTR and ESR1(Spencer and Bazer 1995; Robinson et al. 2001). Oestradiolis unlikely to have a direct effect on OXTR, as noclassical oestrogen response element (ERE) has beenlocalized in either the bovine (Telgmann et al. 2003) orovine (Fleming et al. 2006) OXTR promoter region.However, there are three ERE half-sites and oestradiolactivation of the OXTR is likely to require either steroidreceptor co-factors, such as SRC1e (Telgmann et al.2003) or the transcription factor SP-1 (Fleming et al.2006). It is always important to note that OXTR are notonly temporally, but also spatially regulated within theendometrium. However, the role of the ESR1 in theinitiation of OXTR up-regulation remains equivocal.Spencer et al. (1998) have indicated that ESR1 appearsbefore OXTR in sheep, while others have shown thatOXTR appears in the luminal epithelium before ESR1in both sheep (Wathes and Hamon 1993) and dairy cows(Robinson et al. 1999, 2001). The latter is supported byLeung and Wathes (2000) who showed that OXTR wereup-regulated spontaneously in luminal epithelial cellsin vitro, in the absence of oestradiol, but oestradiol didspeed-up the process. Interpretation of the relativetiming of the up-regulation of OXTR and ESR1 genesis further complicated by the variation of timing ofOXTR up-regulation (ranging from day 14 to day 18)and partially explains the variation in oestrous cyclelength. The possible mechanisms involved are thevariation in the timing of PGR down-regulation or thatOXTR up-regulation is influenced by the number offollicular waves (either two or three). The latter wouldindicate that OXTR expression is influenced by theovary through a mechanism other than progesterone oroestradiol.During early pregnancy, the blastocyst undergoesrapid elongation increasing from 10 cm by day 16, principally because of the rapidtrophoblast growth (Robinson et al. 2006a; Spenceret al. 2007). The elongation of the blastocyst initiatesIFNT production, which is detectable in uterine flushesfrom day 12 to day 25 (Roberts et al. 2003). Consequently,IFNT concentrations in uterine luminal fluiddramatically increase from day 14 to day 18 (Mannet al. 1999). This is because of the rapid growth of thetrophoblast during embryo expansion rather thanchanges in IFNT mRNA expression (Robinson et al.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

CL–Endometrium–Embryo Interactions 1072006a). IFNT is widely recognized as the maternalrecognition of pregnancy signal in ruminants (Robertset al. 2003) and establishes pregnancy by suppressingthe up-regulation of OXTR and ESR1 (Robinson et al.1999, 2001). Furthermore, recombinant IFNT caninhibit OXTR expression both in vivo (Spencer andBazer 1995, Spencer et al. 1998) and in vitro (Telgmannet al. 2003). However, what remains to be elucidated iswhether IFNT acts directly on the OXTR gene (Mannet al. 1999) or through suppression of the ESR1 gene(Spencer et al. 1996). The former hypothesis is supportedby the observations that the bovine OXTRpromoter region contains an interferon response element(IRE) and that interferon-regulatory factors(IRF)-1 and -2 bind to this site (Telgmann et al. 2003).Moreover, ESR1 was present in the luminal epitheliumof pregnant cows from day 12 to day 14 (Robinson et al.2001). Although Fleming et al. (2006) reported thatIRFs had no effect on the ovine OXTR promoter andmay represent a species difference.Interferon tau-stimulated genesInterferon tau also stimulates a plethora of endometrialgenes, which are likely to be important mediators forconceptus development and attachment to the endometrium.For example, IFNT suppresses the immunesystem by silencing genes, such as b2-microglobulinand major histocompatibility complex class 1 polypeptide-relatedsequence, thereby preventing the rejection ofthe conceptus allograft. IFNT also stimulates genesinvolved in cell proliferation (e.g. Wnt7a), uterinereceptivity and conceptus attachment (e.g. Galectin 15,mucins, cathepsin L). For extensive reviews on theeffects of IFNT on endometrial functions, see Demmerset al. (2001) and Spencer et al. (2007, 2008).Endometrial–embryonic interactionsThe endometrium and embryo secrete a number ofcytokines and growth factors (Martal et al. 1997). Theseinclude the IGF system, fibroblast growth factor (FGF)-1 and -2, transforming growth factor (TGF) a and b andepidermal growth factor (EGF). The precise role ofthese factors is currently unclear, but they are likely toplay an important role in the regulation of pre-attachmentblastocyst development, uterine receptivity andstimulation of IFNT secretion (Roberts et al. 2003;Imakawa et al. 2004). There is a strong link betweenmetabolic status and reproductive function (Watheset al. 2003). Thus, the IGF system is particularlyinteresting as IGF1 is an important intermediary ofmetabolic status. For example, systemic lower IGF1concentrations at two weeks post-partum were associatedwith a longer time to conception as a consequenceof impaired ovarian and uterine function (Wathes et al.2003, 2007; Taylor et al. 2004). IGF1 and its bindingproteins (IGFBPs) are expressed in the endometriumand are regulated during the oestrous cycle. IGF1mRNA is present in the endometrium through theoestrous cycle (Geisert et al. 1991; Robinson et al. 2000)with up-regulation during oestrus and during the midlutealphase (Robinson et al. 2000). Moreover, IGF1(Kirby et al. 1996; Robinson et al. 2000) and IGF2(Geisert et al. 1991; Bilby et al. 2006) mRNA expressionis increased during early pregnancy. This raises thepossibility that either the embryo up-regulates thesegenes or that embryo development was enhanced inanimals with increased levels of IGF1 and ⁄ or IGF2.Moreover, IGF1 in concert with IGF2 can stimulate theproduction of IFNT (Ko et al. 1991). These observationscombined with the fact that systemic IGF1concentrations are attenuated in cattle with poor fertility(Wathes et al. 2003), indicate that IGFs are likely toplay a key role in regulating endometrial function andsubsequent embryonic development. The endometrial–embryonic interactions through the IGF system aresummarized in Fig. 1.The bioavailability of IGFs is influenced by theIGFBPs, which are generally considered to sequesterIGFs, but can also promote IGF action. Interestingly,IGFBP levels are differentially regulated during pregnancywith IGFBP2 mRNA and IGFBP3 mRNA beingdown-regulated (Geisert et al. 1991; Robinson et al.2000), while IGFBP1 mRNA was up-regulated in thepresence of an embryo (Robinson et al. 2000). Theexpression pattern of IGFBP1 mRNA was particularlyinteresting as it was expressed in the luminal epitheliumIGFBP1IGFBP2IGF2MyometriumHistotrophIGF1RIGF1R (proliferation)IGF1IGF2(IGF1)IGFBP1 IGFBP2IGFBP3 IGFBP5IGF1R(proliferation)IGF1Systemic IGF1IGFBP5IGFBP3E 2ConceptusIFNTIGFBP1PGRCaruncleBlood vesselFig. 1. A working hypothesis of the inter-relationships between thecomponents of the insulin-like growth factor (IGF) system duringearly pregnancy in the cow. IGF-1 (endometrial-derived) and IGF2(embryonic-derived) stimulate conceptus growth and secretion ofinterferon tau (IFNT). These effects are likely to be modulated byIGF-binding protein (IGFBP) 1 (from luminal epithelium; dark-greybox) and IGFBP2 (from sub-epithelial stroma, light-grey box). Based onits expression patterns, IGFBP1 expression is stimulated by IFNT butsuppressed by progesterone. IGF1 acting through IGF type 1 receptors(IGF1R) also stimulates the secretion of required histotroph from theendometrial glands. IGF2 action, modulated by IGFBP3 and IGFBP5 islikely to stimulate caruncular development. IGFBP3 may also play arole in regulating the availability of systemic IGF1, while IGFBP5was localized to the myometrium. The dotted line indicates proposedregulatory mechanisms; the solid dot represents up-regulation_Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

108 RS Robinson, AJ Hammond, DC Wathes, MG Hunter and GE Mannand its up-regulation coincided with that of blastocystelongation. It is feasible that IGFBP1 increases IGFavailability for the embryo by regulating the transportof endometrial IGF1 to the uterine lumen. Moreover,the embryo may enhance this process by stimulatingIGFBP1 expression even more. In human implantation,there is considerable evidence for a specific role ofIGFBP1 in enhancing trophoblast invasiveness (Hamiltonet al. 1998). Also IGFBP1 mRNA and endometrialPGR were inversely correlated suggesting that progesteronesuppresses IGFBP1 mRNA expression and indicatingthat IGFBP1 transcription is switched on oncethe progesterone block has been lost. This has implicationsin cows with a delayed post-ovulatory rise inprogesterone in which the timing of this IGFBP1 upregulationwould be shifted effecting IGF bio-availability.The embryo may further increase the availability ofIGF1 and ⁄ or IGF2 by suppressing the expression of theinhibitory IGFBP2 and IGFBP3.Luteal Inadequacy: Consequences and CausesThe timing of the post-ovulatory progesterone rise iscritical to maintaining the appropriate synchronybetween the ovary-endometrium and embryo and anyluteal inadequacy has dramatic effects. For example,Mann and Lamming (2001) demonstrated that as littleas one day delay in the rise in post-ovulatory progesteronesignificantly reduced the subsequent developmentof the embryo. Similarly, Garrett et al. (1988) showedthat supplementation with progesterone on day 1 to day4 increased embryo development. More recently, Mannet al. (2006) demonstrated that early progesteronesupplementation from day 5 to day 9, but not latersupplementation from day 12 to day 16, increased bothtrophoblastic length fivefold and uterine IFNT concentrations.Pre-ovulatory follicleWhile the consequences of luteal inadequacy on embryonicdevelopment are well-characterized, the underlyingcause of this aberrant luteal function is poorly understood.A major limitation to investigating the aetiologyhas been the lack of a robust, physiological model inwhich, from the follicular phase, it can be predictedwhether the progesterone rise would be delayed ornormal. We have recently developed such a model bymanipulating the dynamics of the follicular phase byinducing luteolysis in the presence of either a large(>10 mm) or a small (30 h afteroestrus) also results in post-ovulatory luteal inadequacy(Mann and Lamming 2001; Kaim et al. 2003). Usingour model, with detailed endocrinology and ultrasonographicmonitoring of ovarian function, we found noevidence that delayed ovulation per se (LH surge toovulation time) resulted in a delayed rise in plasma postovulatoryprogesterone (Robinson et al. 2005).Luteal development, luteotrophic support andsteroidogenic capacityThe rates of luteal growth and angiogenesis are such thatthey are only equalled by the fastest growing tumours.This enables the CL to grow from 0.5 g immediatelypost-ovulation to >5 g within 10 days (Reynolds andRedmer 1999). Our group in two independent studies hasshown that the weight of CL is highly correlated withplasma progesterone concentrations on day 5 postoestrus(Robinson et al. 2006b; Green et al. 2007).However, this relationship is reduced by day 8 andcompletely lost by day 16 (Robinson et al. 2006b). Thus,it would appear that the rate of luteal development (e.g.compactness, luteal hypertrophy) influences luteal functionrather than size per se. Other potential causes ofluteal inadequacy were insufficient luteotrophic support(e.g. number of LH pulses, basal LH and LH pulseamplitude) and aberrant steroidogenic capacity (numberof small ⁄ large luteal cells, in vitro progesterone productionunder basal or LH-stimulated conditions). Wefound that none of these factors are associated withluteal inadequacy (Robinson et al. 2006b).AngiogenesisIn order to meet these demands, the growth of bloodvessels and establishment of a blood supply (angiogen-Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

CL–Endometrium–Embryo Interactions 109esis) is essential (Reynolds and Redmer 1999; Schamsand Berisha 2004). Indeed, the mature CL is so vascularthat the majority of luteal cells are adjacent to one ormore capillaries (Reynolds and Redmer 1999; Robinsonet al. 2006b) and the CL has one of the greatest bloodflow rates per unit of tissue. Moreover, luteal blood flowis closely associated with the increased plasma progesteroneconcentrations (Acosta et al. 2003). There isevidence that promoting angiogenesis in vivo improvesovarian function in gilts (Shimizu et al. 2003) andconversely that inhibiting angiogenesis suppresses thepost-ovulatory rise in progesterone (Yamashita et al.2008). Thus, angiogenesis is essential for the developmentand function of the CL. Using our luteal inadequacymodel, the degree of vascularization (vonWillebrandt factor; endothelial cell marker), pericytecoverage (a-smooth muscle actin) and proliferationindex (Ki67) were determined. While we found differencesbetween day 5 and day 8 (increased endothelialcell area and decreased proliferation index), no differencesbetween the high- and low-progesterone groupswere detected (Robinson et al. 2006b). However, theinitiation of tissue growth occurs on approximately day1–2, thus days 5 and 8 might be too late to detectFig. 2. The relationship between area of SMA (pericyte marker)staining and plasma progesterone on day 8 of the oestrous cycle. Thearea of SMA and progesterone were positively correlated (r 2 = 0.33;p < 0.05), suggesting that the degree of blood vessel maturationinfluences luteal production of progesteronedifferences. Interestingly, the area of pericyte stainingwas positively correlated with progesterone concentrationson day 8, indicating that degree of vasculaturematuration might influence progesterone production(Fig. 2). This adds to evidence that pericytes might bekey drivers of angiogenesis and vascular function.Vascular endothelial growth factor A (VEGFA) andfibroblast growth factor 2 (FGF2) are potent stimulatorsof endothelial cell proliferation and migration(Carmeliet 2003; Ferrara et al. 2003) and VEGFA isgenerally considered to be the principal growth factorcontrolling angiogenesis. Contrary to expectation, weshowed that FGF2 production, rather than VEGFA, isdynamic during the follicle-luteal transition and earlyCL development in the cow (Robinson et al. 2006b,2007). Follicular fluid concentrations of FGF2 increasedsixfold after the LH surge and were maintainedat high concentrations in the collapsed follicle. Thissuggested that the somewhat overlooked angiogenicfactor FGF2 plays a more important role in theinitiation of angiogenesis post-ovulation, as suggestedby Gospodarowicz et al. (1985), while VEGFA plays amore constitutive role in the maintenance of thedeveloping capillaries ⁄ blood vessels. This hypothesisand others are currently being tested using a physiologicallyrelevant luteal angiogenesis culture systemrecently developed in our laboratory (Fig. 3; Robinsonet al. 2008).This novel physiological system involves culture of amixed population of luteal cells (including small, largeluteal, endothelial and fibroblast cells) in a specializedendothelial cell medium on fibronectin-coated wells. Theendothelial cells develop tubule-like structures and formhighly organized intricate networks, which superficiallyresemble a capillary bed (Fig. 3; Robinson et al. 2008).Furthermore, we found that this endothelial networkformation was increased by both FGF2 and VEGFAindependently, but required the addition of both factorsfor maximal development (Robinson et al. 2008; Fig. 3-a,b). The strength of this system is that it will enable usto elucidate the role of these other cell types (e.g.pericytes) in promoting and directing the formation ofthese endothelial tubule-like structures with a long-termaim of manipulating angiogenesis and hence progesteroneproduction in vivo.(a)(b)Fig. 3. A novel culture system toinvestigate luteal angiogenesis.Endothelial cell networks (vonWillebrandt factor, green) oftubule-like structures developedafter 9 days in (a) the absence and(b) the presence of 1 ng ⁄ ml fibroblastgrowth factor (FGF) 2 +1ng⁄ ml vascular endothelialgrowth factor A (VEGFA). Theaddition of FGF2 and VEGFAincreased endothelial cell areaby 10-fold. The cells werecounterstained with DAPI(4¢,6-diamidino-2-phenylindole)(blue)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

110 RS Robinson, AJ Hammond, DC Wathes, MG Hunter and GE MannConclusionsThe establishment of pregnancy in cattle involves aseries of highly coordinated interactions between theovary, endometrium and embryo. First, the appropriatetiming of these interactions is critical. In particular, thetiming of the post-ovulatory rise in progesterone determinesboth the transport of the embryo and thestimulation of appropriate endometrial secretions requiredfor the rapid development of the embryo. Anyinadequacy in the post-ovulatory progesterone leads todelayed embryo development and embryo mortality.The ‘status’ of the pre-ovulatory follicle and rate ofsubsequent luteal development have profound effects onluteal function. Secondly, the precise interplay betweenthe embryo and the endometrium on which it isdependent are also crucial. The embryo must blockluteolysis to establish pregnancy. It does this in the cowby secreting IFNT, which in turn directly suppresses theup-regulation of the OXTR in the luminal epithelium.Embryonic development is influenced by a number ofendometrial growth factors, such as the IGFs. Thesefactors are likely mediators of the cow’s metabolic statusand are thus exciting targets for development ofstrategies to alleviate the declining fertility.AcknowledgementsThis work has been funded by Ministry of Agriculture, Fisheries andFood, DEFRA, BBSRC, the Wellcome Trust and the Milk DevelopmentCouncil. We greatly appreciate the technical assistance of staff atthe University of Nottingham and Royal Veterinary College withoutwhich this work would not have been possible.ReferencesAcosta TJ, Hayahsi KG, Ohtani M, Miyamoto A, 2003: Localchanges in blood flow within the pre-ovulatory follicle walland early corpus luteum in cows. Reproduction 125, 759–767.Bauersachs S, Ulbrich SE, Gross K, Schmidt SE, Meyer HH,Einspanier R, Wenigerkind H, Vermehren M, Blum H,Sinowatz F, Wolf E, 2005: Gene expression profiling ofbovine endometrium during the oestrous cycle: detection ofmolecular pathways involved in functional changes. J MolEndocrinol 34, 889–908.Bazer FW, 1975: Uterine protein secretions: relationshipto development of the conceptus. J Anim Sci 41, 1376–1382.Bilby TR, Sozzi A, Lopez MM, Silvestre FT, Ealy AD, StaplesCR, Thatcher WW, 2006: Pregnancy, bovine somatotropin,and dietary n-3 fatty acids in lactating dairy cows: I.Ovarian, conceptus, and growth hormone-insulin-likegrowth factor system responses. J Dairy Sci 89, 3360–3374.Boos A, Meyer W, Schwarz R, Grunert E, 1996: Immunohistochemicalassessment of oestrogen receptor and progesteronereceptor distribution in biopsy samples of the bovineendometrium collected throughout the oestrous cycle. AnimReprod Sci 44, 11–21.Carmeliet P, 2003: Angiogenesis in health and disease. NatMed 9, 653–660.Demmers KJ, Derecka K, Flint AP, 2001: Trophoblastinterferon and pregnancy. Reproduction 121, 41–49.Fenwick MA, Llewellyn S, Fitzpatrick R, Kenny DA, MurphyJJ, Patton J, Wathes DC, 2008: Negative energy balance indairy cows is associated with specific changes in IGFbindingprotein expression in the oviduct. Reproduction135, 63–75.Ferrara N, Gerber HP, LeCouter J, 2003: The biology ofVEGF and its receptors. Nat Med 9, 669–676.Fleming JG, Spencer TE, Safe SH, Bazer FW, 2006: Estrogenregulates transcription of the ovine oxytocin receptor genethrough GC-rich SP1 promoter elements. Endocrinology147, 899–911.Garrett JE, Geisert RD, Zavy MT, Morgan GL, 1988:Evidence for maternal regulation of early conceptus growthand development in beef cattle. J Reprod Fert 84, 437–446.Geisert RD, Lee CY, Simmen FA, Zavy MT, Fliss AE, BazerFW, Simmen RC, 1991: Expression of messenger RNAsencoding insulin-like growth factor-I, -II, and insulin-likegrowth factor binding protein-2 in bovine endometriumduring the estrous cycle and early pregnancy. Biol Reprod45, 975–983.Gospodarowicz D, Cheng J, Lui GM, Baird A, Esch F, BohlenP, 1985: Corpus luteum angiogenic factor is related tofibroblast growth factor. Endocrinology 117, 2383–2391.Gray CA, Taylor KM, Ramsey WS, Hill JR, Bazer FW, BartolFF, Spencer TE, 2001: Endometrial glands are required forpreimplantation conceptus elongation and survival. BiolReprod 2001 64, 1608–1613.Green MP, Hunter MG, Mann GE, 2005: Relationshipsbetween maternal hormone secretion and embryo developmenton day 5 of pregnancy in dairy cows. Anim Reprod Sci88, 179–189.Green MP, Mann GE, Hunter MG, 2007: Luteal characteristicsand progesterone production on day 5 of the bovineoestrous cycle. Reprod Domest Anim 42, 643–647.Greve T, Xu KP, Callesen H, Hyttel P, 1987: In vivodeveloment of in vitro fertilized bovine oocytes matured invivo versus in vitro. J IVF Embryo Transfer 4, 281–285.Hamilton GS, Lysiak JJ, Han VK, Lala PK, 1998: Autocrineparacrineregulation of human trophoblast invasiveness byinsulin-like growth factor (IGF)-II and IGF-binding protein(IGFBP)-1. Exp Cell Res 244, 147–156.Humblot P, Holm P, Lonergan P, Wrenzycki C, Lequarre´ AS,Joly CG, Herrmann D, Lopes A, Rizos D, Niemann H,Callesen H, 2005: Effect of stage of follicle growth duringsuperovulation on developmental competence of bovineoocytes. Theriogenology 63, 1149–1166.Hunter MG, Robinson RS, Mann GE, Webb R, 2005:Endocrine and paracrine control of follicular developmentand ovulation rate in farm species. Anim Reprod Sci 82 ⁄ 83,461–477.Imakawa K, Chang KT, Christenson RK, 2004: Pre-implantationconceptus and maternal uterine communications:molecular events leading to successful implantation. JReprod Dev 50, 155–169.Kaim M, Bloch A, Wolfenson D, Braw-Tal R, Rosenberg M,Voet H, Folman Y, 2003: Effects of GnRH administered tocows at the onset of estrus on timing of ovulation, endocrineresponses, and conception. J Dairy Sci 86, 2012–2021.Kimmins S, MacLaren LA, 2001: Oestrous cycle and pregnancyeffects on the distribution of oestrogen and progesteronereceptors in bovine endometrium. Placenta 22, 742–748.Kirby CJ, Thatcher WW, Collier RJ, Simmen FA, Lucy MC,1996: Effects of growth hormone and pregnancy on expressionof growth hormone receptor, insulin-like growth factor-I and insulin-like growth factor binding protein-2 and -3genes in bovine uterus, ovary and oviduct. Biol Reprod 55,996–1002.Ko Y, Lee CJ, Ott TL, Davis MA, Simmen RCM, Bazer FW,Simmen FA, 1991: Insulin-like growth factors in sheeputerine fluids: concentrations and relationship to ovinetrophoblastic protein-1 production during early pregnancy.Biol Reprod45, 135–142.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

CL–Endometrium–Embryo Interactions 111Koos RD, 1989: Potential relevance of angiogenic factors toovarian physiology. Semin Reprod Endocrinol 7, 29–40.Lamming GE, Mann GE, 1995: Control of endometrialoxytocin receptors and prostaglandin F2 alpha productionin cows by progesterone and oestradiol. J Reprod Fertil 103,69–73.Leung ST, Wathes DC, 2000: Oestradiol regulation of oxytocinreceptor expression in cyclic bovine endometrium. JReprod Fertil 119, 287–292.Lonergan P, 2007: State-of-the-art embryo technologies incattle. Soc Reprod Fert Suppl 64, 315–325.Lonergan P, Monaghan P, Rizos D, Boland MP, Gordon I,1994: Effect of follicle size on bovine oocyte quality anddevelopmental competence following maturation, fertilizationand culture in vitro. Mol Reprod Dev 31, 63–67.Lucy MC, 2001: Reproductive loss in high-producing dairycattle: where will it end? J Dairy Sci 84, 1277–1293.Mann GE, Lamming GE, 1999: The influence of progesteroneduring early pregnancy in cattle. Reprod Domest Anim 34,269–274.Mann GE, Lamming GE, 2001: Relationship between maternalendocrine environment, early embryo development andinhibition of the luteolytic mechanism in cows. Reproduction121, 175–180.Mann GE, Lamming GE, Robinson RS, Wathes DC, 1999:The regulation of interferon-tau production and uterinehormone receptors during early pregnancy. J Reprod FertSuppl 54, 317–328.Mann GE, Fray MD, Lamming GE, 2006: Effects of time ofprogesterone supplementation on embryo development andinterferon-s production in the cow. Vet J 171, 500–503.Martal J, Chene N, Camous S, Huynh L, Lantier F, HermierP, L’Haridon R, Charpigny G, Charlier M, Chaouat G,1997: Recent developments and potentialities for reducingembryo mortality in ruminants: the role of IFNs and othercytokine in early pregnancy. Reprod Fert Dev 9, 355–380.Mitko K, Ulbrich SE, Wenigerkind H, Sinowatz F, Blum H,2008: Dynamic changes in messenger RNA profiles ofbovine endometrium during the oestrous cycle. Reproduction135, 225–240.Moreira F, Paula-Lopes FF, Hansen PJ, Badinga L, ThatcherWW, 2002: Effects of growth hormone and insulin-likegrowth factor-I on development of in vitro derived bovineembryos. Theriogenology 57, 895–907.Niswender GD, Juengel JL, Silva PJ, Rollyson MK, McIntushEW, 2000: Mechanisms controlling the function and lifespan of the corpus luteum. Physiol Rev 80, 1–29.Peters AR, 1996: Embryo mortality in the cow. Anim BreedAbstr 64, 587–598.Pushpakumara PG, Robinson RS, Demmers KJ, Mann GE,Sinclair KD, Webb R, Wathes DC, 2002: Expression ofthe insulin-like growth factor (IGF) system in the bovineoviduct at oestrus and during early pregnancy. Reproduction123, 859–868.Reynolds LP, Redmer DA, 1999: Growth and development ofthe corpus luteum. J Reprod Fertil Suppl 54, 181–191.Rizos D, Lonergan P, Ward F, Duffy P, Boland MP, 2002:Consequences of bovine oocyte maturation, fertilization orearly embryo development in vitro versus in vivo: implicationsfor blastocyst yield and blastocyst quality. MolReprod Dev 61, 234–248.Roberts RM, Ezashi T, Rosenfeld CS, Ealy AD, Kubisch HM,2003: Evolution of the interferon tau genes and theirpromoters, and maternal-trophoblast interactions in controlof their expression. Reprod Suppl 61, 239–251.Robinson RS, Mann GE, Lamming GE, Wathes DC, 1999:The effect of pregnancy on the expression of uterineoxytocin, oestrogen and progesterone receptors during earlypregnancy in the cow. J Endocrinol 160, 21–33.Robinson RS, Mann GE, Gadd TS, Lamming GE, WathesDC, 2000: The expression of the IGF system in the bovineuterus throughout the oestrous cycle and early pregnancy.J Endocrinol 165, 231–243.Robinson RS, Mann GE, Lamming GE, Wathes DC, 2001:Expression of oxytocin, oestrogen and progesterone receptorsin uterine biopsy samples throughout the oestrouscycle and early pregnancy in cows. Reproduction 122, 965–979.Robinson RS, Hammond AJ, Hunter MG, Mann GE, 2005:The induction of a delayed post-ovulatory progesterone risein dairy cows: a novel model. Domest Anim Endocrinol 28,285–295.Robinson RS, Fray MD, Wathes DC, Lamming GE, MannGE, 2006a: In vivo expression of interferon tau mRNA bythe embryonic trophoblast and uterine concentrations ofinterferon tau protein during early pregnancy in the cow.Mol Reprod Dev 73, 470–474.Robinson RS, Hammond AJ, Nicklin LT, Mann GE, HunterMG, 2006b: Endocrine and cellular characteristics ofcorpora lutea from cows with a delayed post-ovulatoryprogesterone rise. Domest Anim Endocrinol 31, 154–172.Robinson RS, Nicklin LT, Hammond AJ, Schams D, HunterMG, Mann GE, 2007: FGF2 is more dynamic than VEGFAduring the follicle-luteal transition in the cow. Biol Reprod77, 28–36.Robinson RS, Hammond AJ, Mann GE, Hunter MG, 2008: Anovel physiological culture system that mimics lutealangiogenesis. Reproduction 135, 405–413.Royal MD, Darwash AO, Flint AP, Webb R, Woolliams JA,Lamming GE, 2000: Declining fertility in dairy cattle:changes in traditional and endocrine parameters of fertility.Anim Sci 70, 487–501.Schams D, Berisha B, 2004: Regulation of corpus luteumfunction in cattle: an overview. Reprod Domest Anim 39,241–251.Shimizu T, Jiang JY, Iijima K, Miyabayashi K, Ogawa Y,Sasada H, Sato E, 2003: Induction of follicular developmentby direct single injection of vascular endothelial growthfactor gene fragments into the ovary of miniature gilts. BiolReprod 69, 1388–1393.Spencer TE, Bazer FW, 1995: Temporal and spatial alterationsin uterine estrogen receptor and progesterone receptor geneexpression during the estrous cycle and early pregnancy inthe ewe. Biol Reprod 53, 1527–1543.Spencer TE, Mirando MA, Mayes JS, Watson GH, Ott TL,Bazer FW, 1996: Effects of interferon-tau and progesteroneon oestrogen-stimulated expression of receptors for oestrogen,progesterone and oxytocin in the endometrium ofovariectomized ewes. Reprod Fertil Dev 8, 843–853.Spencer TE, Ott TL, Bazer FW, 1998: Expression of interferonregulatory factors one and two in the ovine endometrium:effects of pregnancy and ovine interferon tau. Biol Reprod58, 1154–1162.Spencer TE, Johnson GA, Bazer FW, Burghardt RC, PalmariniM, 2007: Pregnancy recognition and conceptus implantationin domestic ruminants: roles of progesterone,interferons and endogenous retroviruses. Reprod FertilDev 19, 65–78.Spencer TE, Sandra O, Wolf E, 2008: Genes involved inconceptus-endometrial interactions in ruminants: insightsfrom reductionism and thoughts on holistic approaches.Reproduction 135, 165–179.Sreenan JM, Diskin MG, 1986: The extent and timing ofembryonic mortality in the cow. In: Sreenan JM, DiskinMG (eds), Embryo Mortality in Farm Animals. MartinusNijhoff, Dordrecht, pp. 1–11.Taylor VJ, Cheng Z, Pushpakumara PG, Beever DE, WathesDC, 2004: Relationships between the plasma concentrationsÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

112 RS Robinson, AJ Hammond, DC Wathes, MG Hunter and GE Mannof insulin-like growth factor-I in dairy cows and theirfertility and milk yield. Vet Rec 155, 583–588.Telgmann R, Bathgate RA, Jaeger S, Tillmann G, Ivell R,2003: Transcriptional regulation of the bovine oxytocinreceptor gene. Biol Reprod 68, 1015–1026.Wang CK, Robinson RS, Flint AP, Mann GE, 2007:Quantitative analysis of changes in endometrial glandmorphology during the bovine oestrous cycle and theirassociation with progesterone levels. Reproduction 134,365–371.Wathes DC, Hamon M, 1993: Localization of oestradiol,progesterone and oxytocin receptors in the uterus during theoestrous cycle and early pregnancy of the ewe. J Endocrinol138, 479–492.Wathes DC, Lamming GE, 1995: The oxytocin receptor,luteolysis and the maintenance of pregnancy. J Reprod FertSuppl 49, 53–67.Wathes DC, Mann GE, Payne JH, Riley PR, Stevenson KR,Lamming GE, 1996: Regulation of oxytocin, oestradiol andprogesterone receptor concentrations in different uterineregions by oestradiol, progesterone and oxytocin in ovariectomizedewes. J Endocrinol 151, 375–393.Wathes DC, Taylor VJ, Cheng Z, Mann GE, 2003: Folliclegrowth, corpus luteum function and their effects on embryodevelopment in postpartum dairy cows. Reprod Suppl 61,219–237.Wathes DC, Fenwick M, Cheng Z, Bourne N, Llewellyn S,Morris DG, Kenny D, Murphy J, Fitzpatrick R, 2007:Influence of negative energy balance on cyclicity and fertilityin the high producing dairy cow. Theriogenology 68S, S232–S241.Watson AJ, Westhusin ME, Winger QA, 1999: IGF paracrineand autocrine interactions between conceptus and oviduct.J Reprod Fert Suppl 54, 303–315.Yamashita H, Kamada D, Shirasuna K, Matsui M, Shimizu T,Kida K, Berisha B, Schams D, Miyamoto A, 2008: Effect oflocal neutralization of basic fibroblast growth factor orvascular endothelial growth factor by a specific antibody onthe development of the corpus luteum in the cow. MolReprod Dev, in press.Zelenik AJ, Schuler HM, Reichert JR, 1981: Gonadotrophinbindingsites in the rhesus monkey ovary: role of vasculaturein the selective distribution of human chorionic gonadotrophinto the preovulatory follicle. Endocrinology 109, 356–362.Author’s address (for correspondence): R Robinson, School ofVeterinary Medicine and Science, University of Nottingham, SuttonBonington Campus, Loughborough, Leics LE12 5RD, UK. E-mail:bob.robinson@nottingham.ac.ukConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 113–121 (2008); doi: 10.1111/j.1439-0531.2008.01150.xISSN 0936-6768Improved Detection of Reproductive Status in Dairy Cows Using Milk ProgesteroneMeasurementsNC Friggens 1 , M Bjerring 1 , C Ridder 1,2 , S Højsgaard 1 and T Larsen 11 Faculty of Agricultural Sciences, University of Aarhus, Research Centre Foulum, Tjele; 2 Lattec I ⁄ S, Hillerød, DenmarkContentsThis study tested a model for predicting reproductive statusfrom in-line milk progesterone ‘measurements. The model isthat of Friggens and Chagunda [Theriogenology 64 (2005)155]. Milk progesterone measurements (n = 55 036) representing578 lactations from 380 cows were used to test themodel. Two types of known oestrus were identified: (1)confirmed oestrus (at which insemination resulted in aconfirmed pregnancy, n = 121) and (2) ratified oestrus (wherethe shape of the progesterone profile matched that of theaverage progesterone profile of a confirmed oestrus, n = 679).The model detected 99.2% of the confirmed oestruses. Thisincluded a number of cases (n = 16) where the smoothedprogesterone did not decrease below 4 ng ⁄ ml. These cows hadsignificantly greater concentrations of progesterone, bothminimum and average, suggesting that between cow variationexists in the absolute level of the progesterone profile. Usingratified oestruses, model sensitivity was 93.3% and specificitywas 93.7% for detection of oestrus. Examination of falsepositives showed that they were largely associated with lowconcentrations of progesterone, fluctuating around the4ng⁄ ml threshold. The distribution of time from inseminationuntil the model detected pregnancy failure had a median of22 days post-insemination. In this test, the model was runusing limited inputs, the potential benefits of includingadditional non-progesterone information were not evaluated.Despite this, the model performed at least as well as otheroestrus detection systems.IntroductionProgesterone measurements have been used for sometime now, and are accepted as a valid indicator, forassessing the reproductive status of dairy cows (Bulmanand Lamming 1978; Royal et al. 2000). Implicit in thisprocess is a set of biological rules that are used toconvert the progesterone data into different reproductivestatuses (e.g. postpartum anoestrus, luteal andfollicular phases of the oestrous cycle and pregnancy).Traditionally, these rules were expert opinions usuallyapplied manually to progesterone data (e.g. Lammingand Darwash 1998). As measuring technology hasadvanced from the original radioimmunoassays,through laboratory based ELISA methods (Waldmann1993), to on-farm manual tests (Simersky et al. 2007),biosensors (Delwiche et al. 2001), and ultimatelytowards automated in-line systems for measuring milkprogesterone (Lattec 2007), these rules become bothexplicit and fixed. This is because automated systemsnecessarily incorporate a model to process and condensethe raw data, providing the end-user with the relevantbiological interpretation in a palatable form.The fact that these biological rules are codified allowsthem to be tested. Testing the ability of these rules tocorrectly categorize reproductive status becomes increasinglyimportant as automated in-line progesterone monitoringsystems are implemented on commercial farms.Clearly, the value of a progesterone monitoring systemdepends to a large extent upon how good the biologicalmodel is. The objective of this study was to test the abilityof a model based on in-line milk progesterone measurementsto predict reproductive status. The model tested isthat of Friggens and Chagunda (2005).Time-series models exist for detecting oestrus frommilk traits other than progesterone (de Mol et al. 1999;Firk et al. 2003), and decision strategies for interpretingprogesterone data have been suggested (Lamming andDarwash 1998; Opsomer et al. 1998; Prandi et al. 1999;Delwiche et al. 2001). However, to our knowledge, theonly published full biological model to predict reproductivestatus based on a time-series of milk progesteronemeasurements is that of Friggens and Chagunda(2005).The normal procedure for testing a new predictor of agiven state is to compare it with some ‘gold standard’reference measure. However, in the present case themodel is based on the measure, progesterone, which isthe accepted gold standard for assessing the reproductivestatus of dairy cows (Peters and Ball 1995; Cavalieriet al. 2003). Any conventional test of progesteroneagainst another indicator (e.g. visually determinedoestrus or pedometers) rapidly reverses polarity, becominginstead a test of the other indicator againstprogesterone. Thus, testing of such a model requires anon-conventional approach, which to a large extentbecomes an exploration of the shapes of the progesteroneprofiles. Consequently, this study also characterizeskey aspects of progesterone profiles according toreproductive status.Materials and MethodsTest data collectionMilk progesterone measurements were made on proportionalwhole milk samples collected from all milkingcows in one research herd (Danish Cattle ResearchCentre) during the period 12 September 2002 to 30September 2006. Cows were milked in a robotic milkingsystem with free traffic (mean no. visits ⁄ day = 2.4). Oneprogesterone measure was to be made daily during thefirst 120 days from calving. For the remainder oflactation, progesterone measurements were to be madeevery second day. Actual average intervals betweenprogesterone samples before and after 120 days fromcalving were: 1.4 and 2.4 days, respectively. ProgesteroneÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

114 NC Friggens, M Bjerring, C Ridder, S Højsgaard and T Larsenwas analysed using the Ridgeway ELISA-kit (RidgewayScience Ltd, Gloucestershire, UK). Milk samples werepipetted, diluted and distributed using a Biomek 2000 Ó(Laboratory Automation Workstation, Beckman Coulter,Fullerton, CA, USA). Milk samples (25 ll, diluted1 + 2 with water) were handled according to themanufacturer’s instructions, however incubation R4was increased to 1 h 30 min. Plates were read using aspectrophotometer ⁄ fluorometer, Fluostar Ó (BMG Labtechnologies,Offenburg, Germany) (575 nm). Analyseswere performed in 96-well plates; two sets of sevenstandards (0–30 ng ⁄ ml), locally made using milk froman ovariectomized cow and ethanolic progesteronesolutions, and two sets of two control samples wereused for every analysis and plate. For the low and highcontrols, the intra assay precision (CV%) was 14.9 and1.4, respectively; the inter-assay precision (CV%) was32.7 and 20.1, respectively; and the average inaccuracy(bias) was +0.82 and )0.60 ng ⁄ ml, respectively.The time series of milk progesterone measurementswas examined for gaps longer than 14 days with noprogesterone values. All data following such gaps wereexcluded. Further, the entire cow-lactation was excludedif there was insufficient data to identify the end of thepostpartum anoestrus period (i.e. the time series did notstart at calving and there were less than five measurementsbefore the cow was detected to have elevatedprogesterone and resumed oestrous cycles). The resultingdata represented information during 578 lactationsfrom 380 cows. The total number of milk progesteronemeasurements was 55 036.Timing of inseminations was based on visualdetection of oestrus backed up by activity meterindications (DeLaval, Tumba, Sweden). Farm staffhad no access to the progesterone data. Visualobservations for oestrus signs were made daily at06:00, 12:00 and 16:00, each period lasted 20 min. Cowsshowing oestrus earlier than 35 days from calving werenot inseminated to that oestrus. All cows wereinseminated to oestrus regardless of other managementfactors such as age, milk yield, and whether the cow wasdesignated for subsequent culling. Inseminations werecarried out the same day for oestruses detected at themorning observation, and the next morning foroestruses detected in the afternoon and evening.The model being testedThe model being tested here has been described in detailby Friggens and Chagunda (2005). Briefly, the model isbased on the cow always being in one of threereproductive states (Status), these are: postpartumanoestrus (Status 0), oestrus cycling (Status 1), andpotentially pregnant (Status 2). In the model thesestatuses are mutually exclusive. The Status the cow wasfound to be in by the previous run of the model is alwaysthe Status at the start of the current run. By default, thecow is assumed to be in Status 0 at calving. The model isdynamic and deterministic, designed to run each time anew trigger input occurs using both the current andprevious values. The main model inputs are a smoothed(extended Kalman filter) progesterone concentrationand slope. Inseminations and pregnancy determinationsare also included as inputs. To make use of other knowneffectors of reproductive performance that are notnecessarily reflected in progesterone concentrations,the model is designed to incorporate a number ofadditional inputs. However, the model is designed tofunction in the absence of these additional inputs(although with some loss of accuracy in some outputs).The model outputs are all reproductive Status specificwith the exception of days to next sample (DNS) whichis calculated in each model run regardless of reproductiveStatus. Days to next sample is an important outputwith respect to the progesterone measurements beingautomated. It is designed to feedback to the samplingsystem so that the frequency of milk sampling (i.e.progesterone measurement) can be varied according tothe predicted likelihood of a future reproductive eventsuch as onset of oestrus cycles. The other model outputsare: risk of prolonged postpartum anoestrus, risk andtype of ovarian cyst (i.e. prolonged luteal or follicularphase), onset of oestrus, likelihood of a potentialinsemination succeeding and likelihood of being pregnant(following oestrus).The shift from Status 0 (postpartum anoestrus) toStatus 1 (oestrus cycling) is caused by two consecutivesmoothed progesterone measurements above the thresholdvalue (LThresh) which is 4 ng ⁄ ml. (If EOD inputsare available these can also provoke this shift). Asdescribed by Friggens and Chagunda (2005), once a cowis in Status 1 the shift to Status 2 is caused by thesmoothed progesterone value decreasing belowLThresh. This shift generates a model-detected indicationof oestrus. If no insemination is recorded in thefollowing 5 days, the cow automatically reverts to Status1. If there is a timely insemination and subsequentprogesterone concentrations increase, then the cowremains in status 2 (potentially pregnant) as long asthe model predicted likelihood of pregnancy remainsabove a pregnancy threshold. Typically, if progesteroneconcentrations decrease, then the model shifts backfrom Status 2 to Status 1.The model tested in the present study contained thefollowing major modification relative to the modeldescribed by Friggens and Chagunda (2005). Visualinspection of the earliest progesterone profiles collectedshowed a number of oestrus-like decreases in progesteronethat did not decrease below LThresh. To dealwith this an additional threshold, HThresh was introduced(6 ng ⁄ ml). This allowed identification of these‘high-progesterone oestruses’ conditional on the maximumprogesterone concentration in the preceding cyclebeing greater than 15 ng ⁄ ml, and the present decreasebeing >10 days since a preceding oestrus.The model tested in the present study also containedthe following minor modifications relative to the modeldescribed by Friggens and Chagunda (2005). The slopeof the progesterone profile is used, for example in thefunction calculating the predicted likelihood of AIsuccess. The slope can have extreme values particularlywith short intervals between measurements, thus anadjusted slope (a sigmoid transformation) was usedinstead. Over the normal range this makes virtually nodifference but it caps the extremes. In Status 2, when thecalculated likelihood of pregnancy (LikePreg) decreasesÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reproductive Status Assessed by Milk Progesterone 115below a pregnancy threshold, the cow is deemed nolonger pregnant and reverts to Status 1. In the currentimplementation of the model, a modification was madesuch that once the cow was more than 30 days afterinsemination, 2 consecutive low LikePreg (i.e. lowprogesterone) values are required to cause the changeback to Status 1. Finally, the DNS functions in theluteal phase and during pregnancy were modified tobetter optimize the sampling frequency relative to timeof expected next oestrus (equations in Appendix).Because sampling frequency in the present study wasnot based on DNS, this change had no bearing on theresults presented here, except for those relating directlyto changing the sampling frequency.Model testingThe model has many features that warrant testing but inthe present study we have chosen to focus on thedetection of oestrus, end of postpartum anoestrus, andonset of pregnancy, because these are the major outputs,and those by which the usefulness of the model incommercial application will be primarily judged. Forthis study, the model was run in its most basic form withno additional inputs other than inseminations.Oestrus detectionTesting the ability of the model to identify oestrusimplies that there are a series of known oestruses withinthe data against which to compare the progesteronebasedmodel detected oestruses. Both visual observationsof oestrus and activity measurement-based oestrusalarms were available on the test farm. However, it iswell known from other studies (Rossing and Spahr1992), and was patently clear from inspection of thedata, that these external oestrus detection methods werenot sufficiently reliable for the purpose of testing themodel. Accordingly, an alternative approach to defininga reference set of known oestruses was made. Two typesof known oestrus were identified: confirmed and ratifiedoestruses. A confirmed oestrus was defined as an oestrusat which insemination resulted in a confirmed pregnancy.A ratified oestrus is one in which the shape of theprogesterone profile matches that of the average progesteroneprofile of the confirmed oestruses.Confirmed oestrusesInseminations that occurred within the time window±10 days around the time-point 284 days before asubsequent calving or 40 days before a positive pregnancydetermination (rectal palpation 6 weeks postinsemination)were assumed to be associated with a true,confirmed oestrus (because pregnancy resulted). Thestart of these confirmed oestruses was defined as thetime of the first smoothed milk progesterone measurement

116 NC Friggens, M Bjerring, C Ridder, S Højsgaard and T Larsenthresholds were applied. The probabilities had to begreater than 0.95 and the weighted number of observationshad to be greater than 3. The weighted number ofobservations was calculated as:X(absðt tmidpoint Þþ1Þ 1 ;where t is time from the midpoint of the segment for anygiven observation. For observations 7.5, 5, 2.5 and0 days from the midpoint the individual weights were;0.12, 0.17, 0.29 and 1, respectively. The threshold ofthree for the sum of the individual weights correspondsto an observation frequency of two progesterone measurementsper 3 days for evenly spread observations, ora hole of )3 to +3 days from the midpoint in anotherwise complete set of observations.In order to avoid the same oestrus being repeatedlyidentified as the rolling 15-day window moves throughthe progesterone time series, the following filtering ruleswere applied. In addition to p being >0.95, p for thepreceding observation had to be >0.90. Once anobservation was ratified, then subsequent observationsmeeting the above criteria were not accepted as ratifiedunless the probability had decreased to 8 ng ⁄ ml(Mahalanobis distance is scale invariant, i.e. it assessesthe shape of the segment and not the absolute level ofthe segment). The total number of ratified oestruses was679. The shape of the progesterone profile shown inFig. 1 is that which is associated with oestruses thatoccur after a prior period of luteal activity. Oestrusesthat occur in conjunction with the end of the postpartumanoestrus period do not match this profile andhave thus been excluded from the test of the models’ability to detect oestrus.Pregnancy determinationThe ability of the model to identify cows as pregnantfollowing insemination was evaluated in two ways. Bydefinition, the progesterone profile following confirmedoestruses should always be identified by the model as thecow being in Status 2 (pregnant). The proportion ofcases in which this was so was calculated, and anyexceptions were examined. For those inseminations thatdid not result in a confirmed pregnancy (n = 375), thedistribution of time intervals between insemination andmodel detected pregnancy failure (return from Status 2to 1) was examined.Sampling frequencyThe effects of reducing the progesterone measurementfrequency on model performance were evaluated bygenerating new data sets from the original, full, progesteronedata in which a proportion of the originalprogesterone measures were removed. This was done byimposing a minimum interval between consecutiveprogesterone measurements (within each cow-lactationtime-series). For minimum intervals of 1, 2, 3, 4 and5 days, the resulting data sets contained: 42 459, 26 455,16 756, 8713 and 1535 progesterone measurementsrespectively. With each of these reduced data sets, themodel was run and the proportion of ratified oestrusesthat was detected by the model using the original dataset was calculated. In addition, the model was run usingonly those progesterone measurements that it wouldhave requested if it had been running in real-timeaccording to the model output DNS function. This is avariable sampling frequency as described previously(and in Appendix). The total number of progesteronesamples used in this case was 11 957.Results and DiscussionConfirmed oestrusesOut of 121 confirmed oestruses, 104 were associatedwith a model-detected oestrus based on a decrease insmoothed progesterone below 4 ng ⁄ ml. An additional16 of the confirmed oestruses were model-detected onthe basis of the supplementary oestrus detection rulethat smoothed progesterone decreased below 6 ng ⁄ mlhaving previously been >15 ng ⁄ ml. Thus, the model asimplemented detected 99.2% of the confirmed oestruses,whereas the simpler model (4 ng ⁄ ml threshold) resultedin 85.6% of confirmed oestruses being detected.The majority of studies in the literature have used3ng⁄ ml as a threshold for detecting oestrus (Lammingand Bulman 1976). Thus, a discrepancy exists betweenthis definition and the fact that insemination of cowshaving these ‘high progesterone oestruses’ still resultedin conception. A possible explanation for this discrepancycould be related to the smoothing of the progesteroneprofiles. The very nature of smoothing is that it isresistant to extreme values. Indeed, this was one reasonfor using a threshold of 4 rather than 3 ng ⁄ ml in themodel. So, if insufficient low progesterone values areavailable, the smoothed profile may not decrease below4ng⁄ ml even when individual observations are

Reproductive Status Assessed by Milk Progesterone 1174 ng ⁄ ml (31 of 59 of false negatives). Of these, in 24cases smoothed progesterone did not decrease below6ng⁄ ml. This high-progesterone oestrus increasinglyseems to be a real biological phenomenon (Fig. 2). Ofthe remaining 28 false negatives, enlarging the window()3 to + 5 dfo) around oestrus for aligning modeldetected and ratified oestruses by ±3 days of the oestruscaptured 21 cases. Accepting these 21 cases as a matchreduced the number of false negatives to 38 resulting in amodel sensitivity of 93.3%. This sensitivity comparedvery favourably with the sensitivities reported for othermethods of oestrus detection (Firk et al. 2002), evenwhen these methods are used in very favourablecircumstances such as high intensity visual oestrusdetection (85.7% Van Eerdenburg 2006) or after oestrussynchronisation (91.3% for tail paint, 81.4% forpedometers Cavalieri et al. 2003). Only the combinedactivity, milk yield, milk temperature and conductivitymodel of de Mol et al. (1999) had a sensitivity (94%)that matched that for the ratified oestruses. Althoughthis sensitivity was less than the 99% found using theconfirmed oestruses.In addition to enlarging the number of oestrusesaccepted as ‘true’, the profile matching procedure allowsexamination of those cases in which the model indicatedoestrus in the absence of a ratified oestrus, and thusallows calculation of model specificity. There were 171model-detected oestruses (excluding first increases inprogesterone >4 ng ⁄ ml) that did not match a ratifiedoestrus (i.e. false positives). Thus, the model had apositive predictive value (proportion of model detectedoestruses that are true oestruses) of 72.2%[445 ⁄ (445 + 171)]. The total number of progesteronemeasures in the period of oestrus cycling (Status = 1and profile matching weight

118 NC Friggens, M Bjerring, C Ridder, S Højsgaard and T Larsenbasic specificity of 98.7%. However, this assumes thatfor any given progesterone observation oestrus could beindicated. This was not the case, because in the model,once oestrus was indicated, a waiting period of 5 dayswas mandatory before a new oestrus can be indicated(Friggens and Chagunda 2005). Allowing for thiswaiting period, the total number of progesterone measurementsfor which oestrus could be determined was2 723 resulting in a specificity of 93.7%. This specificityis similar to that found by de Mol et al. (1999).Examination of the false positives showed that theywere largely associated with low progesterone concentrationsfluctuating around the 4 ng ⁄ ml threshold(Fig. 3). Nearly 53% (90 ⁄ 171) of the false positivescame at the end of ‘oestrus’ cycles in which themaximum progesterone concentration did not exceed10 ng ⁄ ml. Only in 56 out of the 171 cases had themaximum progesterone concentration in the preceedingcycle exceeded 15 ng ⁄ ml. Clearly, every time such afluctuating profile decreased below 4 ng ⁄ ml, the modelindicated oestrus. This is the major drawback of asimple threshold-based rule, which cannot be overcomeby a simple smoothing of the progesterone profiles. Thenumber of false positives could be reduced by simplyusing a much greater threshold, thus increasing modelspecificity. Unfortunately, a greater threshold has asignificant unwanted cost in terms of reducing modelsensitivity (de Mol et al. 2001). Instead, each timeoestrus is detected the model calculates a likelihood of apotential insemination succeeding. Because this calculationincludes information about the height and length ofthe preceding oestrus cycle, these false positive oestrusesare associated with a very low likelihood of inseminationsucceeding. Average values of the predicted likelihoodof a potential insemination succeeding when the maximumprogesterone concentration in the preceding cycledid not exceed 5, 10 or 15 ng ⁄ ml were: 0.01, 0.17 and0.45 respectively (on a scale 0–0.9). Thus, false positivesProgesterone (ng/ml)30241812600 30 60 90 120 150 180Days from calvingFig. 3. An example of a progesterone profile showing a fluctuating lowprogesterone after calving. Because the smoothed progesterone profile(solid line) repeatedly decreases below 4 ng/ml during this period, themodel repeatedly detected false positive oestruses. Raw, unsmoothedprogesterone values are shown by the open circles. Pluses indicatemodel-determined Status 0 (postpartum an oestrus), solid trianglesindicate Status 1 (oestrus cycling), and solid squares indicate Status 2(potentially pregnant). The downward pointing open triangles indicateinseminations, the upwards pointing open triangles indicate pregnancydeterminations (1 ‘‘ng/ml’’ = not pregnant, 9 ‘‘ng/ml’’ = pregnant)are obvious to the end user of the system. The predictedlikelihood of a potential insemination succeeding alsoprovides the end user with helpful information formaking a decision about whether or not to inseminatethe cow. This has been explored further by Friggens andLøvendahl (2008).In the model, the first increase in the smoothedprogesterone profile above 4 ng ⁄ ml indicates the end ofthe postpartum anoestrus and coincidentally the firstoestrus. For those first oestruses that were ratified, theaverage difference between the time of model detectedfirst oestrus and ratified oestrus was 1.5 days(SD = 1.6). The distribution of the differences was leftskewed by those cases in which a period of fluctuationoccurred around the 4-ng ⁄ ml threshold (Fig. 3). Themedian difference was 2.2 days. Thus, the model is ingood agreement with the ratified oestruses. However,because the ‘Mahalanobis template’ was based onoestruses that occurred after a prior period of lutealactivity, it did not identify all first increases in progesteroneabove the threshold. The distribution of thesemodel-detected durations of postpartum anoestrus wassimilar to that reported by Royal et al. (2002).Pregnancy determinationOf the 121 confirmed pregnancies, 108 remained inStatus 2 (model-detected pregnant) from conceptionuntil at least positive pregnancy determination (by rectalpalpation), the sensitivity of the model for detectingpregnancy was thus 89.3%. Of the 11 cases in which themodel falsely detected the cow as no longer pregnant, 2were due to gaps in the progesterone data, 5 failedbecause the model judged that the insemination wasmistimed, and 6 were caused by decreases in progesteroneconcentration. These decreases were in some casesindistinguishable from those one would detect if the cowwas returning to oestrus following early embryo loss(e.g. a sudden decrease to < 10 ng ⁄ ml) but these cowsremained pregnant and subsequently calved. Giventhese types of cases, it is hard to envisage being ableto substantially improve pregnancy detection withoutsimultaneously increasing the number of false negatives.However, of greater important to the end user isdetecting true pregnancy failure.The distribution of model detected pregnancy lengths,for those pregnancies that did not result in a positivepregnancy determination or a calf, provides usefulinformation (Fig. 4). The distribution of time frominsemination to model-detected pregnancy failure isbimodal reflecting the two reasons for the modelassessing the cow to not be pregnant: (1) inappropriatetiming of insemination and (2) progesterone concentrationstoo low (first and second peaks, respectively). Themedian of the distribution was 22 days, reflecting thesecond peak. By 27 days 60% of pregnancy failures hadoccurred, and by 34 days 70% had occurred. Thesefigures reflect a typical time pattern of early embryonicloss (Sreenan et al. 2001). Reliable detection of pregnancyusing progesterone is not possible much earlierthan 18–23 days post-insemination because the progesteroneprofile during the luteal phase is not affected bypresence of an embryo (Bulman and Lamming 1978).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reproductive Status Assessed by Milk Progesterone 119Proportion of cases0.160.140.120.100.080.060.040.020.0010 20 30 40 50 60 70 80 90 100110120130140150160170Days from calvingFig. 4. Distribution of pregnancy lengths for cows in which pregnancyfailed (model detected reproductive status reverted from pregnancy tooestrus cycling)The present results indicate that after 23 days the modelreliably detected pregnancy failure. This is substantiallyearlier than the timing of reliable pregnancy determinationby rectal palpation at 6 weeks post-insemination(Peters and Ball 1995).Sampling frequencyThe above results were generated from a data set inwhich average time between collection of samples was1.4 days (for the period 0–120 days from calving).Because these results are expected to be influenced bysampling frequency, we examined the effect of reducingsampling frequency. As shown in Fig. 5, the relativeproportion of ratified oestruses detected by the modeldecreased with decreasing sampling frequency. Theseresults were achieved by simulating an evenly spreadreduction in samples and thus sampling frequency.However, this contrasts with the way the model is set upProportion of oestruses detected1.00.80.60.40.20.01.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0Interval between progesterone measurementsFig. 5. The effect of reducing progesterone sampling frequency(increasing the average interval between progesterone measurements[units = days]) on the proportion of ratified oestruses detected by themodel. The proportion is expressed relative to the case with the fullprogesterone data set. The solid line indicates the effect of reducedsampling frequency using fixed sampling schedules. The solid circleindicates the proportion of oestruses detected when using the modeloutput days to next sample function that results in a variable samplingfrequencyto request samples. In a situation in which the model ispart of an automated in-line progesterone measuringsystem, it controls sampling frequency through the DNSfunction. The DNS function implements a variablesampling frequency that is low immediately afteroestrus, and then increases as a function of bothduration of luteal phase and any evidence of a decreasein progesterone (Friggens and Chagunda 2005; see alsoAppendix). Using this variable function to reduce thesampling frequency in the test data resulted in anaverage sampling frequency of 3.3 days, but the proportionalreduction in oestrus detection was only 0.73.This is notably smaller than the 0.45 reduction for theequivalent equally distributed sampling frequency(Fig. 5). Even in this conservative test of samplingfrequency effects there is a clear benefit of having avariable sampling frequency.General considerationsThis paper provides a test of one particular model forinterpreting progesterone time-series data. In this testthe model was run using limited inputs, for simplicitythe potential benefits of including additional non-progesteroneinformation such as input from other externaloestrus-detection methods (e.g. pedometers), bodyenergy status, and urea concentrations were not evaluated.Despite this, the model performed at least as wellas other oestrus-detection systems (Firk et al. 2002;Cavalieri et al. 2003; Van Eerdenburg 2006). In thiscontext, it would have been relevant to compare thismodel with other models for interpreting progesteroneprofiles. At the time of writing, we could find no otherpublished progesterone models, thus comparison islimited to specific aspects such as oestrus detectionmeasured by indicators other than progesterone (virtuallyall tests use progesterone as the reference measure).Although the present model performed substantiallybetter in terms of sensitivity than the majority ofmethods in the literature (Firk et al. 2002; Cavalieri etal. 2003; Van Eerdenburg 2006), this is not in itself adefinitive assessment of the models quality. As demonstratedby numerous authors, sensitivity and specificitycan be altered by varying the thresholds used, andgenerally have an inverse relation to each other (de Molet al. 2001; Faustini et al. 2007). In the context of timeseriesmeasurements such as progesterone profiles,sensitivity and specificity measures are particularlyunsatisfactory because they are calculated for fixedtime-points or windows (Friggens et al. 2007). Forexample, Faustini et al. (2007) reported a sensitivity forprogesterone determination of pregnancy of 98.2%,substantially greater than the 89.3% reported in thepresent study, whereas Romano et al. (2006) reported asensitivity of 74.5%. The difference was in large partbecause of the time (days post-insemination) at whichthe sensitivity was calculated. Indeed, Romano et al.(2006) reported increasing sensitivity with increasingdays post-insemination. Clearly, factors such as thechosen time-point and sampling frequency make directcomparison of methods difficult.It should also be noted that in the present study,progesterone measurements were made in a laboratoryÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

120 NC Friggens, M Bjerring, C Ridder, S Højsgaard and T Larsenby a classical ELISA method. It may be expected thatsuch a method is more precise than the types ofmeasurement resulting from in situ biosensors. Conversely,milk samples used in the present study werecollected from cows milked robotically, resulting invariable inter milking intervals, and thus variable milkfat content (Friggens and Rasmussen 2001). Thisintroduces a greater variability in the milk progesteronemeasurements (Waldmann et al. 1999) than wouldbe expected from milk samples collected more conventionally.On balance, the results presented in thisstudy probably reflect what can be expected undercommercial conditions, but this remains to be quantified.ConclusionsThe type of model needed to condense and interpretprogesterone profile data in real-time on-farm has beentested and found to perform substantially better in termsof sensitivity of oestrus detection than other existingdetection systems. The model also was shown to providevaluable information about other aspects of reproductivestatus such as pregnancy determination and commencementof luteal activity.AcknowledgementsWe gratefully acknowledge the valuable contribution of the farm staffat KFC, Jens Clausen, Carsten Berthelsen, Peter Løvendahl, JesNielsen, and Connie Middelhede for their efforts in securing thismassive number of samples. This study, which was part of the Biosensproject, was funded by the Danish Ministry of Food, Agriculture andFisheries and the Danish Cattle Association.AppendixDays to next sampling functions for the luteal phase and duringpregnancyThe original days to next sampling (DNS) function in the lutealphase (Friggens and Chagunda 2005) decreased sampling frequencyas duration of oestrus cycle (cyclen) increased towards 21 but thenstayed low in prolonged luteal phases. The function was changed intwo ways. The default luteal DNS function (DNS1LdefR) waschanged from a declining sigmoid to a declining sigmoid plus asubsequent climbing sigmoid (i.e. it rises again after expectedcyclen). The slope modifier was adjusted to use the absolutedifference between the maximum progesterone concentration sincelast oestrus (CycMax) and current progesterone concentration ratherthan the slope:DayFromNextOest = abs((Date - DayOest) - Cyclen)MaxStep ¼ Cyclen DNSLPropDNS1Ldef ¼ MaxStep expð expðDNSLRatðDayFromNextOest DNSLlagÞÞÞSlopeMod ¼ expðexpðSModRat*((CycMax - Level)(CycMax/SModT))))DNS ¼ DNS1Ldef SlopeModwhere date is the current day, DayOest is the day of the precedingoestrus, and level is the smoothed progesterone concentration (ng ⁄ ml).DNSLProp, DNSLRat, DNSLlag, SModRat, and SModT are constantswith the following values: 0.25, )0.4, 5, 0.75 and 4, respectively.The DNS function during pregnancy was modified in a similar way toconcentrate sampling frequency around the time approximately22 days post-insemination and reduce it thereafter:PregStepT = Cyclen + PregStepLagMaxStepPreg ¼ðTopPregStep BotPregStepÞ expð expðPregStepRat ððRunTime AITimeÞPregStepTÞÞÞ þ BotPregStepDNS2def ¼ MaxStepPreg expð expðDNSLRatðDayFromNextOest DNSLlagÞÞÞTimeFromAIFun ¼ expðexpðTfAIRat*((RunTimeAITime) - TfAIT)))offsetLevTime ¼ expð expðPLevRat ðLevel ðPLevTþ DNS2LevToffsetÞTimeFromAIFunRÞÞÞDNS2LevTime ¼ð1 ð1 TimeFromAIFunÞÞ offsetLevTime þð1 TimeFromAIFunÞDNSR ¼ DNS2def DNS2LevTimewhere RunTime is the current time and AITime is the time ofinsemination. PregStepLag, TopPregStep, BotPregStep, PregStepRat,TfAIRat, TfAIT, PLevRat, PLevT and DNS2LevToffset are constantswith the following values: 6, 10, 5, )0.4, )0.3, 12, )0.5, 12 and 6,respectively.ReferencesBulman DC, Lamming GE, 1978: Milk progesterone levelsin relation to conception, repeat breeding and factorsinfluencing acyclicity in dairy cows. J Reprod Fertil 54,447–458.Cavalieri J, Eagles VE, Ryan M, Macmillan KL, 2003:Comparison of four methods for detection of oestrus indairy cows with resynchronised oestrous cycles. Aust Vet J81, 422–425.de Mol RM, Keen A, Kroeze GH, Achten JMFH, 1999:Description of a detection model for oestrus and diseases indairy cattle based on time series analysis combined with aKalman filter. Comput Electron Agr 22, 171–185.de Mol RM, Ouweltjes W, Kroeze GH, Hendriks MMWB,2001: Detection of estrus and mastitis: field performance ofa model. Appl Eng Agr 17, 399–407.Delwiche MJ, Tang X, BonDurant RH, Munro CJ, 2001:Estrus detection with a progesterone biosensor. TransASAE 44, 2003–2008.Faustini M, Battocchio M, Vigo D, Prandi A, Veronesi MC,Comin A, Cairoli F, 2007: Pregnancy diagnosis in dairycows by whey progesterone analysis: an ROC approach.Theriogenology 67, 1386–1392.Firk R, Stamer E, Junge W, Krieter J, 2002: Automation ofoestrus detection in dairy cows: a review. Livest Prod Sci 75,219–232.Firk R, Stamer E, Junge W, Krieter J, 2003: Improving oestrusdetection by combination of activity measurements withinformation about previous oestrus cases. Livest Prod Sci82, 97–103.Friggens NC, Chagunda MGG, 2005: Prediction of thereproductive status of cattle on the basis of milk progesteronemeasures: model description. Theriogenology 64,155–190.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reproductive Status Assessed by Milk Progesterone 121Friggens NC, Chagunda MGG, Bjerring M, Ridder C,Højsgaard S, Larsen T, 2007: Estimating degree of mastitisfrom time-series measurements in milk: a test of a modelbased on lactate dehydrogenase measurements. J Dairy Sci90, 5415–5427.Friggens NC, Løvendahl P, 2008: The potential of on-farmfertility profiles: in-line progesterone and activitymeasurements. In: Friggens NC, Royal MD, Smith RF(eds.), Fertility in Dairy Cows: Bridging the Gaps, BSASOccasional Publication. British Society of Animal Science,Edinburgh, UK (in press).Friggens NC, Rasmussen MD, 2001: Milk quality assessmentin automatic milking systems: accounting for the effects ofvariable intervals between milkings on milk composition.Livest Prod Sci 73, 45–54.Lamming GE, Bulman DC, 1976: Use of milk progesteroneradioimmunoassay in diagnosis and treatment of subfertilityin dairy cows. Br Vet J 132, 507–517.Lamming GE, Darwash AO, 1998: The use of milk progesteroneprofiles to characterise components of subfertility inmilked dairy cows. Anim Reprod Sci 52, 175–190.Lattec, 2007: http://www.herdnavigator.com/pages/id33.asp.Mahalanobis PC, 1936: On the generalised distance in statistics.Proc Natl Inst Sci India 12, 49–55.Opsomer G, Coryn M, Deluyker H, De Kruif A, 1998: Ananalysis of ovarian dysfunction in high yielding dairy cowsafter calving based on progesterone profiles. ReprodDomest Anim 33, 193–204.Peters AR, Ball PJH, 1995: Reproduction in Cattle, 2nd edn.Blackwell Science, Oxford, UK.Prandi A, Messina M, Tondoio A, Motta M, 1999:Correlation between reproductive efficiency, as determinedby new mathematical indexes, and the body condition scorein dairy cows. Theriogenology 52, 1251–1265.Romano JE, Thompson JA, Forrest DW, Westhusin ME,Tomaszweski MA, Kraemer DC, 2006: Early pregnancydiagnosis by transrectal ultrasonography in dairy cattle.Theriogenology 66, 1034–1041.Rossing W, Spahr SL, 1992: Automation in machine milking.In: Bramley AJ, Dodd FH, Mein GA (eds.), MachineMilking and Lactation. Queen City Printers Inc., Burlington,pp. 311–342.Royal MD, Darwash AO, Flint APF, Webb R, Woolliams JA,Lamming GE, 2000: Declining fertility in dairy cattle:changes in traditional and endocrine parameters of fertility.Anim Sci 70, 487–501.Royal MD, Pryce JE, Woolliams JA, Flint APF, 2002: Thegenetic relationship between commencement of luteal activityand calving interval, body condition score, production,and linear type traits in Holstein-Friesian dairy cattle. JDairy Sci 85, 3071–3080.Simersky R, Swaczynova J, Morris DA, Franek M, Strnad M,2007: Development of an ELISA-based kit for the on-farmdetermination of progesterone in milk. Vet Med 52, 19–28.Sreenan JM, Diskin MG, Morris DG, 2001: Embryo survivalrate in cattle: a major limitation to the achievement of highfertility. In: Diskin MG (ed), British Society of AnimalScience Occasional Publication, 26 (1). British Society ofAnimal Science, Edinburgh, pp. 93–104.Van Eerdenburg FJCM, 2006: Estrus detection in dairy cattle:how to beat the bull. Vlaams Diergen Tijds 75, 61–69.Waldmann A, 1993: Enzyme immunoassay (EIA) for milkprogesterone using a monoclonal antibody. Anim ReprodSci 34, 19–30.Waldmann A, Ropstad E, Landsverk K, Sørensen K, SølverødL, Dahl E, 1999: Level and distribution of progesterone inbovine milk in relation to storage in the mammary gland.Anim Reprod Sci 56, 79–91.Author’s address (for correspondence): NC Friggens, Faculty ofAgricultural Sciences, University of Aarhus, Research Centre Foulum,PO Box 50, 8830 Tjele, Denmark. E-mail: n.friggens@agrsci.dkConflict of interest: NC Friggens has received a research grant for thiswork, part-funded by Lattec I ⁄ S; C Ridder is employed by Lattec I ⁄ S;all remaining authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Supp. 2), 122–128 (2008); doi: 10.1111/j.1439-0531.2008.01151.xISSN 0936-6768Genetic Aspects of Reproduction in SheepDR NotterDepartment of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg, USAContentsMaintenance of high levels of realized fertility (defined as thepercentage of ewes that lamb) and appropriate levels offecundity are critical for efficient sheep production. Theoptimal level of fecundity in most situations is well below themaximum attainable level and can be targeted by combiningselection among and within breeds with use of an expandingarray of single-gene mutations affecting ovulation rate andlitter size. The heritability of litter size is approximately 0.10,allowing changes of up to 2% ⁄ year from simple mass selection.Mutations in several genes associated with the transforminggrowth factor b superfamily (BMPRIB, GDF9 and sex-linkedBMP15) can increase ovulation rates by 0.7–1.5 ova inheterozygous ewes. However, ewes that are homozygous forBMP15 or GDF9 mutations are sterile, so use of thesemutations requires carefully structured breeding programmes.Improvements in fertility may be critical for autumn lambingor programmes that aspire to lamb throughout the year.Selection to improve fertility in spring matings has beensuccessful; selected adult ewes have lambing rates of 80–85%in October and early November. The selected ewes have adramatically reduced seasonal anestrus, and many ewescontinue to cycle during spring and summer. Major genesaffecting seasonal breeding have not been identified in sheep.Polymorphisms in the melatonin receptor 1a gene appear to beassociated with seasonal breeding in some, but not all breeds.However, functional genomic studies of genes associated withcircadian and circannual rhythms have potential to revealadditional candidate genes involved in seasonal breeding.IntroductionImplementation of effective programmes of reproductivemanagement in commercial sheep productioninvolves synchronization of genetic potentials forreproductive ability with the production environment.The production objective in such situations is generallymaximization of farm profit. If meat is the main output,attainment of this objective commonly involvesmaximizing the annual number and ⁄ or total weight oflambs marketed. Wool may make an additionalcontribution to income, but unless wool quality is veryhigh, increases in wool production cannot compensatefor even minor losses in meat production. However,reproductive goals in sheep dairying, and particularlythe economic benefit from increasing litter size, may besecondary to the primary goal of increasing milkproduction.Opportunities to increase ovulation rates in sheep arevast, involving both well-established polygenic breeddifferences (Fahmy 1996) and access to a number ofsingle-gene mutations with major effects on ovulationrate and litter size. Mutant alleles may be introgressedinto different genetic backgrounds with relative ease,allowing large increases in ovulation rate without othermajor changes in genetic background or environmentaladaptation. Genetic management of ovulation rate andlitter size thus involves definition of the optimum littersize (which is commonly well below the potentiallyattainable maximum) and the development of breedingschemes to establish and maintain the desired level offecundity.Seasonal breeding limits both diversification andintensification of production. Under extensiveconditions, the breeding season of established, locallyadapted breeds generally has evolved to compliment theenvironmental conditions and desired production calendarand serves to ensure that lambing occurs at appropriatetimes. However, intensification of animalproduction to meet increased consumer demands formeat and other animal products, both locally and inexpanding global markets, provides both motivation andopportunity to modify the traditional lambing schedules.In some cases, these modifications require only a shift inthe breeding season, but there is also an opportunity toincrease the frequency of lambing, with 7- to 9-monthlambing intervals as an apparent practical minimum. Amassive literature exists cataloguing attempts to useexogenous hormone treatments or photoperiodic manipulationto induce fertile matings outside the normalbreeding season. Corresponding information now existsto show that useful levels of additive genetic variation induration of the breeding season are present among andwithin breeds, and the development of sheep with adramatically reduced breeding season will be discussed.The understanding of the genetic control of circadianand circannual timekeeping is likewise expanding rapidly,but access to quantitative trait loci (QTL) andfunctional mutations associated with seasonal breedingin sheep remains limited.This review will thus focus on genetic mechanismscontrolling ovulation rate and seasonal breeding insheep, and on strategies to synchronize geneticpotentials for reproductive traits with the productionenvironment.Genetic Control of Ovulation Rate and LitterSizePolygenic associationsA summary of heritability estimates for several reproductivetraits is shown in Table 1 (Safari et al. 2005).Estimates of heritability for litter size are generallyconsistent in the literature, averaging 0.11 (n = 102). Asdiscussed by Bradford (1985), these parameters wouldsuggest a maximum response to single-trait mass selectionfor litter size of approximately 2% per year. Thus, agenetic gain of 0.20 lambs born per ewe lambing fromsingle-trait mass selection would require approximatelyÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Genetic Aspects of Reproduction in Sheep 12310 years (or three sheep generations). These heritabilityestimates would generally be considered low, but evaluationof the potential for genetic change must alsoconsider the variation present in the population, quantifiedby the phenotypic coefficient of variation (CV) ofapproximately 35% for litter size. The predictedresponse to mass selection is proportional to the productof (1) the selection intensity (i), (2) the heritability of thetrait (h 2 ), (3) the phenotypic SD and (4) the inverse ofthe generation interval (t). For fixed i and t, the responseto mass selection as a percentage of mean performanceis thus directly proportional to the product (h 2 · CV).A comparison of values for h 2 and CV for litter size inTable 1 to values for other traits shows that the modestselection responses predicted for traits such as litter sizedo not arise directly from a shortage of heritablevariation. They instead reflect delays in age at measurementwhich limit annual selection response by extendingthe generation interval or, alternatively, necessitate thatselection be based on the litter size of the dam (therebyreducing accuracy of the evaluation). If we compare theanticipated selection response in litter size with that forpost-weaning body weight, we find that the heritabilityof body weight is much higher (0.28 vs 0.11), but thatthe phenotypic CV is considerably lower (12.5 vs 35%;Safari et al. 2005) leading to a slightly higher value forh 2 · CV for litter size (3.85) than for body weight (3.50).A particular powerful demonstration of the powerof selection to change litter size comes from anexperiment to study the response to selection fortwinning in cattle (Echternkamp et al. 2002). Thisexperiment was initiated in 1982 with an initialTable 1. Estimates of heritability (h 2 ), phenotypic coefficient ofvariation (CV), and relative selection response as a percentage of themean (h 2 .CV) for reproductive traits in sheep a Phenotypich 2 CV (%) h 2 CV bEwe fertility c 0.07 49 3.4Ovulation rate 0.19 28 5.3Embryo survival 0.01 27 0.0Litter size 0.11 35 3.9Lamb survival 0.04 ⁄ 0.05 47 1.9Ram scrotal circumference 0.23 10 2.3a Safari et al. (2005).b See text for interpretation of h 2 . CV values.c Percentage of ewes that lamb.incidence of twin births of less than 3%. By 1998,the incidence of twinning had increased to over 48%.Factors that contributed to success in this projectincluded initial intensive screening of the US cattlepopulation to identify cows with a history of twinbirths, use of repeated laparoscopic measures ofovulation rate at consecutive cycles to increase theaccuracy of assessment without correspondinglylengthening generation interval, and use of progenytesting and artificial insemination to maximize use ofsuperior sires. These genetic changes were achievedthrough the traditional quantitative strategies, but thepopulation has now also been used to detect QTLsthat affect twinning in cattle. However, the currentexperiment has reached a position of diminishingreturns because of increases in triplet pregnancies. Incontrast to the situation in sheep, the presence ofmore than one foetus in a single uterine hornnormally results in termination of pregnancy in cattle.Mutations affecting ovulation rate and litter sizeTable 2, derived from the review by Davis (2005), listsmutations that result in increased ovulation rates. Thesemutations all involve gene products and receptorsassociated with the transforming growth factor bsuperfamily (McNatty et al. 2005). Thus, the FecB Bmutation first reported in Booroola Merino is nowknown to result from a point mutation in the bonemorphogenic protein receptor IB (BMPRIB) gene(Wilson et al. 2001), which is also sometimes referredto as activin-like kinase 6 (ALK6). This mutation iswidespread in prolific Asian sheep breeds including theIndian Garole, Javanese Thin-tail, and Chinese Hu andHan as well as the Booroola Merino, but appears to beabsent in prolific European breeds (Davis et al. 2002,2006). The Garole has long been postulated to be thesource of the FecB B mutation in the Booroola Merino(Turner 1982), but the presence of mutation in ChineseHu and Han sheep suggests a central Asian origin. TheHu and Han are relatively closely related and derivedfrom Mongolian sheep breeds (Lu et al. 2005). Lu et al.(2005) likewise cite historical studies by Xie (1985)asserting that establishment of the Hu breed predatedemergence of the Han breed.Mutations in the X-linked bone morphogenic protein15 (BMP15) gene were first reported in New ZealandRomney, but mutations in this gene also exist in theTable 2. Single-gene mutationsaffecting ovulation rate (OR) insheep a Gene a Breed(s) Name ChromosomeOR effect1 copy 2 copiesBMPRIB Merino, Javenese Thin-tail,Booroola (FecB B ) 6 +1.5 3.0Garole, Hu, HanBMP15 Romney Inverdale ⁄ Hanna (FecX I ⁄ FecX H ) b X +1.0 – cBelclaire Belclaire (FecX B ) X +1.0 – cBelclaire, Cambridge Galway (FecX G ) X +0.7 – cLaucaune Laucaune (FecX L ) X +1.5 – cGDF9 Belclaire, Cambridge High fertility (FecG H ) 5 +1.4 – ca Derived from Davis (2005) and Bodin et al. (2007).b FecX I and FecX H represent two separate mutations with similar effects.c Ewes that are homozygous for these mutations are sterile.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

124 DR NotterBelclaire and Cambridge breeds in Ireland and the UnitedKingdom (Davis 2005) and in the French Laucaune(Bodin et al. 2007). Thus, five separate mutations inBMP15 are associated with increases in ovulation rate.The unique characteristic of BMP15 mutations is that anincrease in ovulation rate relative to the wild type isobserved only in individuals that are heterozygous for themutation. Homozygous individuals have abnormal ovariandevelopment and rudimentary ‘streak’ ovaries, andare sterile. A point mutation in the growth differentiationfactor 9 (GDF9) gene on chromosome 5 in Belclaire andCambridge sheep produces phenotypic effects that aresimilar to those of the X-linked BMP15 mutations, withan effect on ovulation rate of approximately +1.4 ova inheterozygotes, but abnormal ovarian development andsterility in homozygotes.Programmes to use single-gene mutations to increasefecundity usually must be carefully structured to avoidproduction of homozygous females. Under pastoralconditions, the relatively high incidence of quadrupletand higher births generally limits the desirability ofhomozygous FecB B ⁄ FecB B individuals. However, matingsof homozygous sires to ewes of breeds that do notcarry the mutation allows production of heterozygousdaughters. The average effect of such a substitution is toincrease the realized litter size by approximately 1.0 lambs ⁄litter, which may be useful in breeds that producepredominantly single lambs, provided the environmentand management system can be adjusted to accommodatetwin and occasional triplet births. The occurrence ofFecB B in phenotypically diverse breeds also providesopportunity to source FecB B genes for introgression frombreeds with similar levels of adaptation to the targetbreed. Utilization of BMP15 or GDF9 mutations requiresuse of DNA tests to identify males that are homozygousfor the mutation and can therefore be used to reliablyproduce heterozygous daughters. Maintenance of a flockthat is homozygous for these mutations is precluded bythe sterility of the homozygous females.Screening of ewes to identify exceptional individualspermits identification of animals that may carry mutantalleles within existing, adapted populations. However, itis likewise essential that the production system beconfigured in such a way that allows producers tobenefit from increases in litter size, and identification ofan appropriate target litter size distribution is a necessaryfirst step in developing more prolific lines. This taskis not as easy as it sounds, because effects of ewe age, theewe age distribution, effects of lambing season, and, inthe case of single-gene mutants, genetic backgroundmust all be considered. In many pastoral environments,twin births are considered ideal, and there is oftenconsiderable discrimination against triplet births. Borget al. (2007) demonstrated that attainment of more than60–65% twin births on a whole-flock basis generallyrequires a corresponding significant increase in thefrequency of triplets.Genetic Control of Seasonal BreedingPolygenic controlHeritability estimates for realized fertility (i.e. thepercentage of ewes that lamb) are very low (Table 1),almost always below 0.10 and often below 0.05. Thisresult is not particularly surprising. Natural selection forfertility is necessarily intense, and lowly fertile ewes leavefew progeny. In well-adapted populations maintainedunder long-established breeding conditions, the percentageof ewes that lamb is commonly high to very high(e.g. 85–95%) and potential for genetic improvement infertility is limited. However, this will not necessarily bethe case when animals are moved to different productionconditions or different lambing seasons.The potential for genetic improvement in fertility isconstrained by the current level of fertility. As discussedearlier, the potential rate of genetic change is proportionalto the CV of the trait. If fertility is expressed as abinary trait separating ewes that lamb from those thatdo not, the associated phenotypic variance is given asp(1 – p), where p is the proportion of ewes that lamb.For values of p between 0 and 1, p(1 – p) is maximizedat p = 0.5, and the anticipated selection responsedeclines as p increases towards the maximum valuep = 1.0. The phenotypic CV of 49% for fertility inTable 1 corresponds to a value of p = 0.87, and wouldbe expected to double at p = 0.5. Thus, the selection toimprove the fertility will become progressively lesseffective as mean fertility increases with little potentialfor further improvement in highly fertile breeds.Conversely, opportunities for genetic improvement infertility may be anticipated when animals are matedoutside their normal breeding season or managed invarious systems of accelerated lambing in order to allowmore than one lambing in a 12-month period (Notter2002). There have been several attempts to develop linesof sheep with reduced seasonality of breeding, but mostwere terminated before yielding measurable changes induration or timing of the seasonal anestrus. However, apopulation developed using classical selection proceduresat Virginia Tech (Al-Shorepy and Notter 1997;Notter et al. 1998; Notter and Cockett 2005) experienceda dramatic reduction in duration of the seasonalanestrus. This flock was derived from crosses among theDorset (50%), Rambouillet (25%) and Finnish Landrace(25%) breeds, all of which have some capacity tomate and conceive in spring and summer. Single-traitselection was imposed from 1988 to 1998 and was basedon the ability of the ewes to mate and conceive innatural, single-sire matings over an 8-week periodbeginning May 1.In phase 1 of the study (1988–1993; Al-Shorepy andNotter 1997), ewes were exposed to vasectomized ramsfor 2 weeks before the start of mating to attempt tostimulate ovulation, and selection of replacement ewesand rams was based on the mean fertility of the dam.The average fertility of adult ewes in the base populationwas approximately 60%, which would nearly maximizethe phenotypic SD and, therefore, anticipated selectionresponse, but the estimated heritability of spring fertilitywas only 0.08, indicating that the heritability of ewefertility in spring matings was not higher than theaverage estimate of 0.07 reported by Safari et al. (2005).The estimated cumulative and annual selectionresponses in fertility in phase 1 were 6.5% and1.4% ⁄ year, respectively. Ewe age influenced both meanfertility and selection response. Fertility of 7-month-oldÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Genetic Aspects of Reproduction in Sheep 125autumn-born ewe lambs placed with rams in springaveraged only 10% (vs approximately 60% for unselectedadult ewes) and did not respond to selection.In phase 2 of the study (1994–1998; Notter et al.1998), ewes were no longer exposed to vasectomizedrams before breeding. The distribution of lambingduring phase 2 was characterized by an immediate startat 140–145 days after the start of breeding, and mostewes (other than ewe lambs) that lambed did so within3–4 weeks after the start of lambing. This patternindicated that ewes were cycling when rams wereintroduced on May 1 and were not induced to cycle bythe ram effect (Martin et al. 1986). A system of breedingvalue prediction was also implemented to maximizeaccuracy of selection. Trends in estimated breedingvalues (EBV) in selected and control lines are shown inFig. 1, with a final difference in spring fertility in adultewes of 17% (87 vs 70%). Fertility of ewe lambs,however, remained unresponsive to selection and wasstill below 15% in selected ewes at the end of the study.The selection experiment ended in 1998, with subsequentstudies devoted to characterizing the selectedanimals. An area of concern was to determine whetherthe selection had truly lengthened the breeding season orhad simply shifted the period of anestrus to avoid thespring breeding season. To that end, Vincent et al.(2000) monitored mating behaviour of high and low BVewes that were continuously maintained with the samevasectomized rams from mid-January through late Julyof 1992, 1993 and 1995. Ewes were checked twice weeklyfor mating marks. They were identified as anestrouswhenever consecutive estruses were separated by morethan 21 days, and the numbers of days of anestrus weresummed over the period. High BV ewes (average EBVfor fertility of 12.6%) were anestrous for an average ofonly 28 days whereas ewes with average BV (mean EBVof 0.3%) were anestrous for an average of 70 days(p < 0.001). Each 1-day increase in EBV was associatedwith a decrease in days of anestrus of 2.15 ± 0.72 days(p < 0.01). Results of this study confirm that selectionfor ability to lamb in autumn resulted in a reduction inthe duration of the seasonal anestrus.A second study (J. M. Smith and D. R. Notter,unpublished) evaluated 68 selected ewes maintainedindoors with vasectomized rams from mid-January untilearly July. One half of the ewes were placed underconditions that mimicked ambient photoperiods andFall fertility EBV (%)20151050–5Selection lineControl line1988 1990 1992 1994 1996 1998Year of birthFig. 1. Changes in estimated breeding values (EBV) for fall fertility,defined as the percentage of ewes that lamb in fall, over time in selectedand control linesone half were maintained under constant 16-h days.Each group was maintained with three vasectomizedrams and the same rams remained with each groupthroughout the study. The number of days of anestruswas not affected by the photoperiod; results for 4-yearoldand older ewes were similar to those of Vincent et al.(2000) with an average of 33 days of anestrus, but theaverage duration of anestrus was longer for 3-year-old(56 days) and 2-year-old ewes (71 days). At the end ofthe study on July 6, 10 elite ewes that had missed at mostone cycle were identified. These ewes came from bothlight treatments and were combined, placed with fourvasectomized rams (two from each of the original lighttreatments), and maintained under 16-h days for anadditional 10 weeks (until approximately the fall equinox)to investigate their ability to continue cycling undersustained long days. Blood samples were collected twiceweekly and assayed for circulating progesterone toconfirm ovarian status. All ewes were found to becycling at the start of this period, but all but one of theewes subsequently missed at least one cycle whenmaintained under continuous long days and the relativelyhigh temperatures of mid-August in Blacksburg,Virginia. However, all but two of the 10 ewes reinitiatedcycles by September 7, when ewes were removed fromthe study and joined with intact rams outdoors forbreeding. These results suggest that even these elite ewesmay, under extended long days, have experienced aremnant anestrous period of 1–2 cycles in late summer.A study to assess late-summer ovarian activity in greaterdetail is currently underway.Selection to improve fertility in accelerated lambingsystems is particularly challenging because of thecomplexity of the systems (Notter 2002). The varioussystems used in practice often involve 3–5 matingseasons per year. Open ewes are commonly joinedwith rams at consecutive mating seasons until conceptionoccurs, and ewes re-enter the system shortlyafter weaning of their lambs. Ewes in these systemstend to become stratified, with high- and low-fertilityewes moving through the system in separate groups,but with periodic infusions of ewes with newly weanedlambs. Differences in fertility among seasons are oftenlarge (Lewis et al. 1996) and must be properlyconsidered in statistical models to ensure separationof genetic and seasonal environmental effects. Thechoice of the trait(s) that best characterizes geneticdifferences in performance in accelerated systems islikewise not clear.Lewis et al. (1996) and Banos et al. (2002) studiedgenetic control of reproductive performance in anintensive accelerated-lambing system developed at CornellUniversity and involving five breeding seasons peryear. Over 6 years, fertility varied from 15% in June to69% in October. Heritability estimates for the numberof exposures (mating seasons) per conception were0.09 ± 0.06 at first lambing, 0.15 ± 0.09 at secondlambing and 0.05 ± 0.06 at the third and subsequentlambings. Heritability estimates for age at first lambingand first lambing interval were 0.09 ± 0.06 and0.25 ± 0.11, respectively, but the heritability estimatefor the second and subsequent lambing intervals wasonly 0.03 ± 0.04. These results suggest some potentialÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

126 DR Notterto use early performance, and particularly the firstlambing interval, as a selection criterion, but recordsoccurring after the second lambing contributed littleadditional information.In a related study, Vanimisetti (2006) estimatedgenetic parameters for realized fertility in a commercialPolypay flock with three annual lambing seasons.Heritability estimates were near zero for all lambingintervals, but significant additive genetic effects wereobserved for ewe lamb fertility at first mating and forage at the first, second and third lambing, with heritabilityestimates of 0.14, 0.39, 0.28 and 0.36, respectively(all p < 0.01). These measures were strongly intercorrelated,with absolute values of genetic correlationsabove 0.85, suggesting that the ability of the ewes tomate and conceive at the first opportunity and thereforelamb at young ages was the best indicator of geneticmerit and could be relatively easily evaluated bymonitoring ewe ages at lambing.Possible major genes affecting seasonalityLoss-of-function mutations comparable to those thatdisrupt regulation of ovulation rate have not yet beenidentified for seasonal breeding. The most promisinggenetic marker has been a restriction fragment lengthpolymorphism (RFLP) polymorphism in the gene thatencodes melatonin receptor 1a (MTNR1A). This, andother, polymorphisms in MTNR1A were described byBarrett et al. (1997), Messer et al. (1997) and Pelletieret al. (2000), and their frequencies and effects werereviewed by Notter and Cockett (2005). Briefly,significant associations between seasonal breedingperformance and MTNR1A genotype were reportedin both the moderately seasonal Merino d’Arles breed(Pelletier et al. 2000) and the Virginia Tech out-ofseason-breedingline (Notter et al. 2003). In theVirginia Tech population, adult ewes with at leastone copy of the favourable RFLP allele had11.2 ± 5.1% higher spring fertility than ewes thatwere homozygous for the alternative allele (p = 0.03),and this marker accounted for 23.8% of the additivevariance in spring fertility.A review of allelic frequencies in various breeds foundthat the putative favourable MTNR1A allele was oftenpresent at relatively high frequencies in breedscommonly considered to be quite seasonal (Notter andCockett 2005). Hernandez et al. (2005) likewise reportedthat MTNR1A genotype had no association with lengthof the breeding season in the relatively seasonal Ile-de-France breed. These results suggest that effects of thepolymorphism may depend on genetic background andare not necessarily detectable in highly seasonal breeds,thereby limiting its usefulness to that of a breed-specificgenetic marker.Knowledge of functional genomic control of circadianand circannual rhythms is expanding in both laboratoryspecies and sheep. Several clock genes have beencatalogued and shown to both express circadianrhythms in gene expression and be inducible by changesin photoperiod. Studies of clock gene expression in thesuprachiasmatic nucleus and pars tuberalis (PT) ofsheep have shown that expression of the cryptochrome 1gene (cry1) is transiently stimulated by exposure toelevated melatonin levels at the onset of the dark phasewhereas expression of the period 1 gene (per1) coincideswith onset of the light phase (Lincoln et al. 2003a;Hazlerigg et al. 2004; Wagner et al. 2008). Dimerizationof the protein products of cry1 and per1 in thecytoplasm is required to allow their translocation intothe nucleus, where this protein complex is known toeffect expression of other clock genes (Reppert andWeaver 2001). Lincoln et al. (2003b) hypothesized thatthe photoperiodically mediated time interval betweenpeak expression of cry1 and per1 in the PT can act as anindicator of day length, allowing these cells to functionas ‘calendar cells’. Both cry1 and per1 are immediatelyresponsive to changes in photoperiod but expression ofa third clock gene in the ovine PT, RevErba, has beenshown to vary with photoperiodic history (Hazlerigget al. 2004), providing an as-yet hypothetical mechanismto explain the spontaneous changes in reproductive stateobserved in animals maintained under constant stimulatoryor inhibitory photoperiods.Endocrine regulation of seasonal breeding involveschanges in sensitivity of the brain to steroid(predominantly oestradiol) negative feedback. Amechanism for interaction of circulating steroids withmelatonin from the pineal gland is therefore required.Recent studies of effects of kisspeptin and its associatedGPR54 receptor in the hypothalamus may provide thismissing link. Kisspeptin, the product of the KiSS1 gene,stimulates GnRH release from the hypothalamus in anumber of species and nearly all GnRH neurons appearto co-express GPR54 (Smith 2008). Smith et al. (2007)demonstrated the presence of steroid receptors in KiSS1neurons in the arcuate nucleus of the ewe and found thatexpression of KiSS1 in the arcuate nucleus increasesduring transition from anestrus to the breeding season.Revel et al. (2006) reported that exogenous melatonincould override effects of photoperiod on KiSS1 expressionin Syrian hamster, indicating that melatoninmediates photoperiodic effects on kisspeptin secretion.The sensitivity of KiSS1 neurons to both melatonin andsex steroids may provide the missing link enablingseasonal changes in steroid sensitivity in the hypothalamus.These functional genomic discoveries have identifiedan array of candidate genes with potential effects onseasonality. Screening of large populations for ewes withunusual seasonal breeding patterns and screening ofthese candidate gene regions for functional sequencevariants may uncover useful mutations similar to thosefound for ovulation rate and litter size.ConclusionsThe establishment of appropriate levels of prolificacyand patterns of seasonal fertility is critical to efficientsheep production. The potential of classical selectionmethods to create change in lowly heritable reproductivetraits is often discounted, but has been shown to besubstantial. A number of single-gene mutations can alsobe used to enhance ovulation rates and litter size, butcomparable mutations affecting seasonal breeding havenot yet been detected.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Genetic Aspects of Reproduction in Sheep 127ReferencesAl-Shorepy SA, Notter DR, 1997: Response to selection forfertility in a fall-lambing sheep flock. J Anim Sci 75, 2033–2040.Banos G, Lewis RM, Notter DR, Hogue DE, 2002: Geneticprofile of fertility and prolificacy of maiden and mature ewesmanaged in a frequent lambing system. In: Proceedings ofthe 7th World Congress on Genetics Applied to LivestockProduction, CD-ROM communication No. 08-21, 4 pp.,Cedex, France.Barrett P, Conway S, Jockers R, Strosberg AD, Guardiola-Lemaitre B, Delagrange P, Morgan PJ, 1997: Cloning andfunctional analysis of a polymorphic variant of the ovineMel 1a melatonin receptor. Biochem Biophys Acta 1356,299–307.Bodin L, Di Pasquale E, Fabre S, Bontoux M, Monget P,Persane L, Mulsant P, 2007: A novel mutation in bonemorphogenic protein 15 gene causing defective proteinsecretion is associated with both increased ovulation rateand sterility in Lacaune sheep. Endocrinology 148,393–400.Borg RC, Notter DR, Kuehn LA, Kott RW, 2007: Breedingobjectives for Targhee sheep. J Anim Sci 85, 2815–2829.Bradford GE, 1985: Selection for litter size. In: Land RB,Robinson DW (eds), Genetics of Reproduction in Sheep.Butterworths, London, pp. 3–18.Davis GH, 2005: Major genes affecting ovulation rate in sheep.Genet Sel Evol 37(Suppl. 1), 11–23.Davis GH, Galloway SM, Ross IK, Gregan SM, Ward J,Nimbkar BV, Ghalsasi PM, Nimbkar B, Gray GD,Subandriyo, Inounu I, Tiesnamurti B, Martyniuk E, EythorsdottirE, Mulsant P, Lecerf F, Hanrahan JP, BradfordGE, Wilson T, 2002: DNA tests in prolific sheep from eightcountries provide new evidence on origin of the Booroola(FecB) mutation. Biol Reprod 66, 1869–1874.Davis GH, Balakrishnan L, Ross IK, Wilson T, Galloway SM,Lumsden BM, Hanrahan JP, Mullen M, Mao XZ, WangGL, Zhao ZS, Zeng YQ, Robinson JJ, Mavrogenis AP,Papachristoforou C, Peter C, Baumung R, Cardyn P,Boujenane I, Cockett NE, Eythorsdottir E, Arranz JJ,Notter D, 2006: Investigation of the Booroola (FecB) andInverdale (FecX I ) mutations in 21 prolific breeds and strainsof sheep sampled in 13 countries. Anim Reprod Sci 92,87–96.Echternkamp SE, Thallman RM, Kappes SM, Gregory KE,2002: Increased twinning and cow productivity in beef cattleselected for ovulation and twinning rate. In: Proceedings ofthe 7th World Congress on Genetics Applied to LivestockProduction. CD-ROM communication no. 08-12, 4, pp.,Cedex, France.Fahmy MH (ed.), 1996: Prolific Sheep. CAB International,Wallingford, UK.Hazlerigg DG, Andersson H, Johnston JE, Lincoln G,2004: Molecular characterization of the long-day responsein the Soay sheep, a seasonal mammal. Curr Biol 14,334–339.Hernandez X, Bodin L, Chesneau D, Guillaume D, Allain D,Chemineau P, Malpaux B, Migaud M, 2005: Relationshipbetween MT1 melatonin receptor gene polymorphism andseasonal physiological responses in Ile-de-France ewes.Reprod Nutr Dev 45, 151–162.Lewis RM, Notter DR, Hogue DE, Magee BH, 1996: Ewefertility in the STAR accelerated lambing system. J Anim Sci74, 1511–1522.Lincoln G, Andersson H, Hazlerigg D, 2003a: Clock genes andthe long-term regulation of prolactin secretion: evidence fora photoperiod ⁄ circannual timer in the pars tuberalis. JNeuroendocrinol 15, 390–397.Lincoln G, Andersson H, Loudon A, 2003b: Clock genes incalendar cells as the basis of annual timekeeping inmammals – a unifying hypothesis. J Endocrinol 179, 1–13.Lu S, Chang H, Du L, Tsunoda K, Sun W, Yang Z, Chang G,Ji D, 2005: Phylogenetic relationships of sheep populationsfrom coastal areas of East Asia. Biochem Genet 42, 251–260.Martin GB, Oldham CM, Cognie Y, Pearce DT, 1986: Thephysiological responses of anovulatory ewes to theintroduction of rams – a review. Livest Prod Sci 15,219–247.McNatty KP, Galloway SM, Wilson T, Smith P, Hudson NL,O’Connell A, Bibby AH, Heath DA, Davis GH, HanrahanJP, Juengel JL, 2005: Physiological effects of major genesaffecting ovulation rate in sheep. Genet Sel Evol 37(Suppl. 1),25–38.Messer LA, Wang L, Tuggle CK, Yerle M, Chardon P, PompD, Womack JE, Barendse W, Crawford AM, Notter DR,Rothschild MF, 1997: Mapping of the melatonin receptor1a (MTNR1A) gene in pigs, sheep, and cattle. MammGenome 8, 368–370.Notter DR, 2002: Opportunities to reduce seasonality ofbreeding in sheep by selection. Sheep Goat Res J 17, 20–32.Notter DR, Cockett NE, 2005: Opportunities for detectionand use of QTL influencing seasonal reproduction in sheep:a review. Genet Sel Evol 37(Suppl. 1), 39–53.Notter DR, Al-Shorepy SA, Vincent JN, McQuown EC, 1998:Selection to improve fertility in fall lambing. in: Proceedingsof the 6th World Congress on Genetics Applied to LivestockProduction, vol 27, pp. 43–46, NSW, Australia.Notter DR, Cockett NE, Hadfield TS, 2003: Evaluation ofmelatonin receptor 1a as a candidate gene influencingreproduction in a fall-lambing sheep flock. J Anim Sci 81,912–917.Pelletier J, Bodin L, Hanocq E, Malpaux B, Teyssier J,Thimonier J, Chemineau P, 2000: Association betweenexpression of reproductive seasonality and alleles of the genefor Mel 1a receptor in the ewe. Biol Reprod 62, 1096–1101.Reppert SM, Weaver DR, 2001: Molecular analysis of mammaliancircadian rhythms. Annu Rev Physiol 63, 647–676.Revel FG, Saboureau M, Masson-Pevet M, Pevet P, MikkelsenJD, Simmonneaux V, 2006: KiSS-1: a likely candidatefor the photoperiodic control of reproduction in seasonalbreeders. Chronobiol Int 23, 277–287.Safari A, Fogarty NM, Gilmour AR, 2005: A review of geneticparameter estimates for wool, growth, meat and reproductiontraits in sheep. Livest Prod Sci 92, 271–289.Smith JT, 2008: Kisspeptin signaling in the brain: steroidregulation in the rodent and ewe. Brain Res Rev 57, 288–298.Smith JT, Clay CM, Caraty A, Clarke IJ, 2007: KiSS-1messenger ribonucleic acid expression in the hypothalamusof the ewe is regulated by sex steroids and season.Endocrinology 148, 1150–1157.Turner HN, 1982: Origins of the CSIRO Booroola. in: PiperLR, Bindon BM, Nethery RD (eds), The Booroola Merino,Proceedings of a Workshop. CSIRO, pp. 1–7, Melbourne,Australia.Vanimisetti HB, 2006. Genetic evaluation of ewe productivityand its component traits in Katahdin and Polypay sheep.Ph.D. Dissertation, Virginia Polytechnic Institute and StateUniversity, Blacksburg, VA, USA.Vincent JN, McQuown EC, Notter DR, 2000: Duration of theseasonal anestrus in sheep selected for fertility in falllambing. J Anim Sci 78, 1149–1154.Wagner GC, Johnston JD, Clarke IJ, Lincoln GA, HazleriggDG, 2008: Redefining the limits of day length responsivenessin a seasonal mammal. Endocrinology 149, 32–39.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

128 DR NotterWilson T, Wu X-Y, Juengel JL, Ross IK, Lumsden JM, LordEA, Dodds KG, Walling GA, McEwan JC, O’Connell AR,McNatty KP, Montgomery GW, 2001: Highly prolificBooroola sheep have a mutation in the intracellular kinasedomain of bone morphogenic protein IB receptor (ALK-6)that is expressed in both oocytes and granulose cells. BiolReprod 64, 1225–1235.Xie CX, 1985: The History of Cattle, Sheep, and Goat Raisingin China Attached with the History of Deer Raising.Agricultural Publishing House of China, Beijing.Author’s address (for correspondence): David R Notter, Departmentof Animal and Poultry Sciences, Virginia Polytechnic Institute andState University, Blacksburg, VA 24061, USA. E-mail: drnotter@vt.eduConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 129–136 (2008); doi: 10.1111/j.1439-0531.2008.01152.xISSN 0936-6768The Importance of Interactions Among Nutrition, Seasonality and Socio-sexualFactors in the Development of Hormone-free Methods for Controlling FertilityRJ Scaramuzzi 1,2 and GB Martin 31 UMRPhysiologie de la Reproduction et des Comportements, L’Institut National de la Reproduction, Nouzilly, France; 2 Department of VeterinaryBasic Sciences, Royal Veterinary College, Hertfordshire, UK; 3 Animal Production Systems, UWA Institute of Agriculture, The University ofWestern Australia, Crawley, WA, AustraliaContentsAround the world, consumers are demanding animal productsthat are produced to agreed standards for human health,environmental management and animal welfare. This has led tothe development in Australia of the concept of ‘clean, green andethical’ (CGE) animal production based on the manipulation ofnutrition (‘focus feeding’) and the application of phenomena,such as the ‘male effect’, to provide ‘natural’ methods formanaging small ruminant production systems. With respect tothe management of fertility, CGE involves utilization of theinherited responses of animals to environmental factors tomanipulate their reproductive processes. The successfuldevelopment and implementation of this new generation ofmanagement tools depends on a thorough yet holisticunderstanding of the interactions among environmental factorsand the ways these interactions affect reproductive physiologyand behaviour of the animal. For sheep and goats, a centralaspect of CGE management is the way in which ovarian functionis affected by three major factors (nutrition, photoperiod andsocio-sexual signals) and by interactions among them. Nutritioncan exert two profound yet contrasting types of effect on ovarianactivity: (i) the complete inhibition of reproduction by undernutritionthrough the hypothalamic mechanism that controlsovulation and (ii) the enhancement of fecundity by nutritionalsupplementation, through a direct ovarian mechanism, infemales that are already ovulating. A similarly profound controlover ovarian function in female sheep and goats is exerted by thewell-known endocrine responses to photoperiod (seasonality)and to male socio-sexual signals. The ‘male effect’ already has along history as a valuable technique for inducing a synchronizedfertile ovulation during seasonal and post-partum anoestrus insheep and goats. Importantly, experimentation has shown thatthese three major environmental factors interact, synergisticallyand antagonistically, but the precise nature of these interactionsand their significance to reproductive outcomes are not wellunderstood. Most research to date has been with smallruminants but CGE principles can be applied to any species ina managed environment. For example, a male effect has beenreported for lactating cattle and, in the horse, the pattern ofseasonality of oestrus can be altered by nutrition. Well-fed mareshave a longer breeding season and some animals become nonseasonal.Similar observations have been reported for sheep andgoats. By working towards a holistic perspective of the physiology,nutrition, genetics and behaviour of our animals, we willbe able to formulate ways to manipulate the animals’environment that will improve management, productivity andprofitability and, simultaneously, promote a CGE industry.IntroductionAnimal industries around the world are being challengedby changing attitudes in the market place, andconsumers are increasingly demanding products that areproduced to agreed standards for human health,environmental management and animal ethics andwelfare. This is particularly evident in high-profit,discerning markets that are influenced by discretionaryspending power. Our challenge is to turn these developmentsto advantage, improve productivity and profitabilityand, simultaneously, promote ‘clean, green and ethical(CGE) production’ (Martin et al. 2004a; Kadokawa andMartin 2006; Martin and Kadokawa 2006). Importantly,these issues need not pose difficulties or increase costs – onthe contrary, they offer excellent opportunities. Forextensively managed sheep and goats, we are developingmanagement tools, such as ‘focus feeding’ and the ‘maleeffect’, for controlling reproduction and quantitativebreeding values for improving temperament (Martinet al. 2004a).This paper reviews current information, mainly forsheep and goats, on a central aspect of the CGE system –the way in which ovarian function is affected by threemajor factors (nutrition, photoperiod and socio-sexualsignals) and by interactions among them. We review theway in which nutrition can influence the ‘decision toreproduce’ through the hypothalamic mechanism thatcontrols ovulation and the ‘decision for fecundity’through a direct ovarian mechanism in females that arealready ovulating. In addition, we review the ways inwhich the endocrine responses to photoperiod(seasonality) and to male socio-sexual signals interact,synergistically and antagonistically, with nutritionalinputs. It is increasingly clear that the multi-dimensionalnature of these interactions are often far more importantthan the responses to the individual environmental input(Blache et al. 2007).Most research to date has been with small ruminantsbut the ‘CGE principle’, in which manipulation of theenvironment is used to control reproductive processes,can be applied to all species. For example, a male effecthas been reported for cattle (review: Ungerfeld 2007)and, in this review, we will describe nutrition–reproductioninteractions in the horse.Thus, the aim of this review is to outline the state ofour knowledge on the interactions among environmentalfacts that affect reproductive processes, and to pointout research that is needed for refining CGE managementof farm animals, particularly with respect tomanagement of the nutritional inputs.Nutrition and Ovarian FunctionAll major measures of reproductive performance (prolificacy,fertility and fecundity) are affected by geneticsÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

130 RJ Scaramuzzi and GB Martinand by a variety of environmental factors including:(i) nutrition, energy balance and diet; (ii) photoperiodand seasonality; (iii) socio-sexual factors; (iv) temperatureand relative humidity and (v) stress. Nutritionappears to have bi-directional effects on ovarian functionand the nature of these effects we suggest are inhibitory ata hypothalamic level (allowing or preventing ovulation)and stimulatory at an ovarian level (affecting ovulationrate) and, furthermore, may differ between ruminant andmonogastric species. With respect to prolificacy, smallruminants are generally regarded as seasonal monocotousspecies, although natural twinning is not uncommon,and there is considerable variation in the incidence oftwin ovulations and twin births among and within breeds.The component of the diet that is probably the mostimportant with respect to ovarian function is energyparticularly that derived from glucose, although theevidence in favour of any particular nutrient is stillcontroversial and the precise interrelationships betweenovarian function and the components of the diet areobscure.Nutritional effects on ovarian function are bi-directionalNutritional deprivation of the female, whether as aresult of an insufficient supply of energy in the diet, or areduced availability of energy because of excessivedemands for energy (extreme physical activity, energydemandingprocesses such as lactation), inhibits hypothalamicGnRH release and thus leads to reducedsecretion of pituitary LH and eventually to anovulationand anoestrus. The hypothalamic GnRH pulse-generatingsystem in monogastric species (rats, primates andhumans) appears to be more sensitive to nutritionaldeprivation than in ruminants (sheep, goats and cattle).In monogastric species, undernutrition and associatedhypoglycaemia quickly leads to suppression of theGnRH system and a cessation of LH pulsatility (Bronson1988; Cagampang et al. 1990). Farmed ruminantsrarely develop hypoglycaemia and thus appear to besomewhat protected against nutritional anovulation,although LH pulsatility is inhibited when they are madehypoglycaemic under experimental conditions (Downingand Scaramuzzi 1997; Szymanski et al. 2007) or as aresult of lactation (Rhodes et al. 1995). On the contrary,for animals in normal body condition, nutritionalsupplementation appears to have little direct influenceon hypothalamic GnRH secretion or pituitary gonadotrophinsecretion in the female and there is littleunambiguous evidence to show that nutritional supplementationstimulates gonadotrophin secretion by aprimary hypothalamic mechanism or pituitary mechanism.Thus, for ovulation, primarily a GnRH-dependentprocess, the effects of nutrient supplementation are notsimply the opposite to the effects of nutrient deprivation.On the contrary, nutritional supplementation can stimulatefolliculogenesis (Webb et al. 2004; Scaramuzziet al. 2006) and thus increase prolificacy (Knight et al.1975; Wang et al. 2004; Gaskins et al. 2005), but thisappears to involve direct nutritional influences on thefollicle itself. Follicular responses to increased nutrientsupply are not always positive and although increasedfolliculogenesis is associated with either short-termnutritional supplementation or moderate increases inbody weight, extreme increases in body weight (obesity)and rapid fluctuations in nutrient supply, such asrepeated cycles of ‘binge-eating’ followed by starvation,have inhibitory effects on ovarian function and reducefertility (Hartz et al. 1979; Lake et al. 1997; Wolfe 2005).Thus, there appears to be a nutritional thresholdbelow which reproduction is inhibited because of thesuppression of the GnRH pulse-generating system to alevel of activity below that required to maintain regularovarian cycles. At the hypothalamic level, nutritionalinfluences on reproduction are qualitative determiningfertility by assessing whether an animal ovulates or not.However, once this threshold is reached and ovariancycles are occurring, nutritional regulation becomesquantitative and operates at the level of the follicle todetermine the rate of reproduction – ovulation rate,prolificacy or litter size.Dietary supplementation, flushing and dietary signalsFlushing, a form of dietary supplementation, is apractice that has been used in ruminant productionsystems since at least the beginning of the 19th century(Youatt 1837). Despite, or perhaps because of, itswidespread use, there appears to be a variety ofnutritional practices covered by the term, ‘flushing’. Inthe absence of a precise definition of flushing, theinterpretation of published information and attempts toelucidate the mechanism of nutritional influences onovarian function have been problematic. The metabolicoutcomes of a period of short-term (3–7 days) nutritionalsupplementation are very different to those for a6- to 8-week period of supplementation. Similarly, themetabolic consequences of dietary supplementation ofanimals in negative energy balance will be very differentfrom those for animals in positive energy balance. It istherefore very likely that the effects on ovarian folliculogenesisand the GnRH pulse-generating system willalso depend on the metabolic state of the animal and thenature of the supplementation. For example, Nottleet al. (1997a) reported an experiment with ewes that hadbeen previously managed for 6 months as separateflocks on either a high or low plane of nutrition. The twoflocks were recombined 17 days before mating and givenshort-term nutritional supplementation with lupin grainfor the 10 days prior to mating. The short-term supplementincreased the ovulation rate by 54% in theunderfed group, compared with only 23% in the wellfedgroup, and the interaction between previous nutritionalhistory and short-term supplementation wassignificant (Nottle et al. 1997a). Another experiment(Leury et al. 1990) reported an opposite finding; lupingrain supplementation in ewes fed 1.2 times maintenanceincreased ovulation rate by 22% compared withewes fed at 0.8 times maintenance where lupin grainsupplementation increased ovulation rate by 10%. Thereasons for this difference are not immediately apparentand additional investigation is obviously required.There has been considerable speculation and someresearch attempting to identify the critical componentsof the diet that influence ovarian function. The diet ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Nutritional Interactions and Reproduction 131mammals is made up of complex mixtures of proteins,fats and simple and complex carbohydrates as well astrace elements, vitamins and other micronutrients, andall of these components have the potential to influencereproduction, either directly or indirectly. The currentconsensus suggests that dietary energy is a very importantcomponent of the diet with respect to nutritionalinfluences on ovarian function. The short-term administrationof glucose or other energy-yielding substratescan increase ovulation rate in ewes (Nottle et al. 1988;Teleni et al. 1989; Downing et al. 1995a,b; Vin˜ oles et al.2005). However, there are other published reports inwhich the increased supply of glucose did not increaseovulation rate in ewes (Iglesias et al. 1996) and furthermore,in the male, the published data suggest that theincreased spermatogenesis seen in response to feeding alupin grain supplement is not caused by an increasedavailability of glucose (Blache et al. 2002).From the above remarks it follows that, whenexamining the effects of nutrition on ovarian functionand folliculogenesis, there is an obvious need for a moredetailed description of both animals’ diets and metabolicstates than is customary.Sources of metabolic glucose that affect ovarian functionGlucose for the generation of ATP is derived from eitherdietary sources or from gluconeogenesis, either with orwithout accompanying glycogen synthesis. Monogastricspecies derive their glucose principally from dietarysources rather than from gluconeogenesis, whereas theruminant species reduce their dietary glucose in theanaerobic conditions of the rumen and derive theirglucose almost exclusively by gluconeogenesis, fromdiet-derived long-chain fatty acids and short-chainvolatile fatty acids. Compared with monogastrics, this isprobably a highly significant functional difference withrespect to the effects of nutrition on ovarian functionbecause ruminants, animals with highly efficient gluconeogenesisrarely become hypoglycaemic or, for thatmatter, hyperglycaemic. For example, dairy cows with ahigh genetic merit for milk production do not becomehypoglycaemic at the peak of lactation, although they canbe in severe negative energy balance. Indeed, hypoglycaemiais usually only seen in ruminants in the most extremesituations such as pregnancy toxaemia (twin lambdisease) in sheep and in lactating first-calf beef heifersthat are still growing. Hypoglycaemic beef heifers have anextended period of post-partum anovulation associatedwith reduced LH pulsatility, because the hypoglycaemiainhibits the GnRH pulse-generating system. The situationwith monogastric species appears to be quite differentbecause they depend on the diet for supplies of glucoseand have a limited capacity for gluconeogenesis comparedwith ruminants. Monogastric species readilybecome hypoglycaemic when deprived of dietary energyand the consequence is a reduced GnRH ⁄ LH pulsefrequency and thus anovulation.Seasonality and Ovarian FunctionSeasonal patterns of reproduction are relatively commonamong the mammals including the domesticlivestock – sheep, goats, horses and, under somecircumstances, cattle and pigs show seasonal rhythmsof ovarian function that are driven primarily by photoperiod.These patterns and the processes that underpinthem, especially in sheep, have been extensively investigatedand there are a number of detailed reviews on thesubject (e.g.: Yeates 1949; Hafez 1952; Karsch et al.1984; Malpaux 2006).Interactions between nutrition and seasonalityPhotoperiod and nutrition both exert major influenceson reproduction so it seems axiomatic that seasonalrhythms in ovulation will be influenced by nutrition. It ishighly probable that photoperiodic and nutritionalinputs to the hypothalamus interact at the level of theGnRH neuron (Ho¨ tzel et al. 2003; Blache et al. 2007),yet there are relatively few published investigations ofthe interaction between nutrition and photoperiodism infemales. There are few controlled studies with ewesinvestigating the relationship between nutrition and thepattern of seasonality, although there are reports ofalterations in the seasonal pattern of cyclic ovulatoryactivity associated with nutritional conditions in previousseasons (Hunter 1962; Smith 1965; Oldham et al.1990). In one study carried out over 13 months (Huletet al. 1986), ewes at pasture had a lower proportionovulating in the months of anoestrus (May, June andJuly) compared with ewes in a drylot that weresupplemented with alfalfa hay (Fig. 1). The pasturefedewes had a higher proportion of anovulatory ewes inFebruary suggesting that they were entering anoestroussooner. Similarly, in August, the pasture-fed ewes had alower proportion of anovulatory ewes, again suggestingthat they were slower to resume ovarian cyclicity at thestart of the new breeding season. In another study, theseasonal pattern of ovarian cyclicity was determined inRasa Aragonesa ewes maintained at two levels of bodycondition (Forcada et al. 1992; Forcada and Abecia2006). This study showed that the duration of anoestruswas reduced by approximately 2 months in the ewesEwes ovulating (%)1009080706050403020100FebPastureDry lotMayJun JulTime of the yearsAugSepFig. 1. The proportion of ewes ovulating in February (enteringanoestrus), May to July (during anoestrus) and August and September(the start of the next breeding season) in two groups of fine-wool ewes.One group was managed on rangeland and the other group wasmanaged with alfalfa hay in drylot. The within month treatment effectswere significant for all months except September. Data taken fromHulet et al. (1986)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

132 RJ Scaramuzzi and GB Martinmaintained in good body condition (BCS, 2.8) comparedwith ewes maintained in poor body condition(BCS, 2.3).More extensive data are available for two otherseasonal species, the horse and the goat. In a study justcompleted at the INRA laboratory at Nouzilly, Salazar-Ortiz and Guillaume studied the effect of nutrition onthe seasonal rhythm of ovarian activity in mares keptunder natural photoperiod (Salazar-Ortiz 2006). Thepattern of ovarian cyclicity over three consecutive yearswas determined from weekly blood progesterone measurementsin two groups of mares. One group was wellfed and the other group was received a diet calculated tokeep the mares thin but in good health. At the start ofthe experiment, the average body weight of the twogroups was approximately 300 kg. The well-fed maresgained approximately 20 kg in the first year after whichtheir mean live weight did not change for the remainderof the experiment. The restricted mares lost approximately65 kg in the first year and then their live weightalso stabilized for the remainder of the experiment. Thenutritional regime had a profound effect on the patternof ovarian cyclicity (Fig. 2) – the well-fed mares had amuch longer period of ovarian cyclicity and six of the 10mares had no period of winter anovulation in the 3-yearexperiment. This was in contrast with the restrictedgroup in which the winter period of anovulation waslonger and present in all the 10 mares.Zarazaga et al. (2005) studied the effect of nutritionon the seasonal pattern of sexual activity in femalePayoya goats kept under natural photoperiod over aperiod of 20 months. The does were maintained on twolevels of nutrition: a high group fed 1.5 times theircalculated maintenance requirements and a low groupfed to maintenance. The pattern of seasonality was thendetermined by daily monitoring for oestrus and forovulation using progesterone measured in blood twice aweek. The length of anoestrus was 32 days shorter indoes on the high diet and this difference was significantbecause of both a delayed entry into anoestrus and anearlier resumption of ovarian cyclicity at the end ofanoestrus.Although domestic Bos taurus cattle are non-seasonal,one study (Montgomery et al. 1985) reported an interactionbetween season and nutrition on the resumptionof ovarian cyclicity post-partum. In this study, an effectAnovulatory mares (%)100806040200Sum Aut Win Spr Sum Aut Win Spr Sum Aut Win Spr SumSeasonFig. 2. The weekly proportion of anovulatory mares over a 3-yearperiod. The data for the mares on a high plane of nutrition are shownby the solid line and that for the mares on a low plane of nutrition bythe dashed line. Redrawn from Salazar-Ortiz (2006)of the season of calving (winter versus spring at latitude45° south) was observed only under conditions of lownutrition. The same seems to apply to laboratory rats, inwhich undernutrition seems to be able to modulatereproductive responses to photoperiod through a pinealdependentpathway (Walker and Bethea 1977).What do all these observations tell us? Most importantly,that there is an interaction in which nutritionaffects the seasonal pattern of ovarian cyclicity in sheep,goats and horses; the variety of species suggests thatnutritional influences on patterns of seasonal reproductionare probably present in most, if not all, seasonalbreeding mammals, including non-seasonal species suchas beef cattle. However, the studies done to date withfemale animals do not shed any light on the mechanismsthat underlie this interaction or on how the interactioncan be usefully incorporated into systems of CGEanimal production.Socio-sexual Signalling and Ovarian FunctionThe ‘male effect’ is a well-studied phenomenon in sheepand goats and there are a number of excellent recentreviews of the subject (Martin et al. 2004b; Ungerfeldet al. 2004; Chemineau et al. 2006; Ungerfeld 2007). Insummary, sexually active males produce a pheromonethat stimulates the secretion of GnRH pulses and ovarianactivity and can induce anovulatory females to ovulate.Interactions between nutrition and socio-sexual signalsAs with photoperiod, socio-sexual signals are a majorfactor controlling ovulation in seasonal species and,because nutrition also has major influences onovulation, it also seems highly probable that nutritionwould influence the efficacy of the ‘male effect’ and thatthe site of the interaction is again the GnRH neuron inthe hypothalamus (Martin et al. 2004b; Blache et al.2007). When a group of female goats were dividedaccording to bodyweight, it was observed that theoestrus response to the ‘buck effect’ was reduced in thedoes with the lowest body weight compared with thosewith medium and heavy weights (Ve´liz et al. 2006).Similar findings have been reported for ewes (Lassouedand Khaldi 1990). Similarly, in beef cattle, the sociosexualinfluence of bulls can affect growth rate and theage of puberty in heifers (Roberson et al. 1991), andnutrition modifies the ability of bulls to reduce theduration of post-partum anoestrus (Monje et al. 1992;Stumpf et al. 1992).In a recent experiment, we have investigated theinteraction between nutrition and the ‘male effect’ inanoestrous female goats (De Santiago-Miramontes et al.2008). Two groups of 25 anoestrous female goats grazedvery poor quality natural vegetation from 09:00 to16:00 h daily. Overnight, they were housed in pens andone of the groups received a supplement (950 g lucernehay, 290 g rolled corn and 140 g soy bean per animaldaily) for 7 days before exposure to bucks. The proportionof does ovulating, the proportion in oestrus and theovulation rate at the first ovulation detected within5 days of exposure to males were all greater in supplementedthan in the control females (Fig. 3). The effect ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Nutritional Interactions and Reproduction 1332.01.610080OvulatingOestrusOvulation rate1.20.8Response (%)60400.420Fig. 3. The ovulation rate(mean ± SEM; n = 25), interval(mean ± SEM days) from theintroduction of bucks to the onsetof oestrus, and the proportions ofdoes ovulating, showing oestrusand having short cycles at the firstovulation following buckintroduction in does that wereeither given nutritional supplementationor not given nutritionalsupplementation for 7 days beforethe introduction of sexually activebucks. All between-treatmentcomparisons are significant. Datataken from De Santiago-Miramontes et al. (2008)Interval to oestrus (days)0.08.06.04.02.00.0Does with short cycles (%)0100806040200Control Supplemented Control Supplementedsupplementation did not persist and the proportion ofdoes in oestrus was not different if ovulation occurredafter 5 days (Fig. 4) and neither was the ovulation rate(De Santiago-Miramontes et al. 2008). We concludedthat short-term feed supplementation before the ‘maleeffect’ can increase the proportion of females respondingand their ovulation rate. However, the stimulatory effectof supplementation did not persist and was not observedat the subsequent ovulation. The increase in ovulationrate suggests that the supplement promoted the growthof a greater number of ovulatory follicles or reduced therate of atresia among the existing cohort of follicles.Somewhat surprisingly, supplementation also increasedthe proportion of does showing short cycles after thefirst male-induced ovulation and this response wasassociated with a reduced interval from the introductionof bucks to oestrus (Fig. 3). The effect of nutritionalsupplementation on the proportion of short cyclesinduced by the male effect could be due to the fact thatDoes in oestrus (%)6050403020100a0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15Time relative to the introduction of bucks (days)Fig. 4. The daily proportion of does showing oestrus for 15 days afterthe introduction of sexually active bucks on day 0 and following anutritional supplementation for 7 days before the introduction of thebucks. The data for the supplemented group are shown by the dashedline and for the non-supplemented (control) group by the solid line; theletter ‘a’ = p < 0.05 compared with the control group at the sametime. Redrawn from De Santiago-Miramontes et al. (2008)more of supplemented does ovulated over the first5 days compared with the non-supplemented doesbecause most does responding in the first 5 days havea short cycle (Chemineau et al. 2006; Delgadillo et al.2006). These findings are consistent with the suggestionthat a shorter delay from male introduction to the firstovulation is associated with an increased proportion ofshort cycles (Pearce et al. 1985). The nutritional supplementappears to have accelerated the follicular phaseinduced by the ‘buck effect’ and the shorter period offollicular development (Fig. 4) leads to either a nonfunctionalCL or, perhaps, to uterine entrainment for anearly luteolytic signal.Somewhat similar findings have been reported forewes. In one study, Nottle et al. (1997b) reported thatsupplementation of Merino ewes daily with 500 g lupingrain from Days 12 to 26 after the introduction of ramssignificantly increased the ovulation rate from 1.26 to1.46 at the second ovulation following ram introduction.Working with lactating Sarda ewes, Molle et al. (1995)also observed an increase in ovulation rate followingsupplementation starting 2 weeks before the ‘ram effect’,but only with a supplement of soya bean meal and notwith whole corn grain. In a later study, Molle et al.(1997) reported that short-term supplementation foronly 7 days before the introduction of rams increasedovulation rate from 1.36 to 1.66 but, in this case,the difference was not significant, possibly because of thegroup size of 15 ewes per treatment was too small. Theseoutcomes contrast with those from a factorial study with390 Merino ewes (Fisher et al. 1993) – there were nosignificant effects of short-term supplementation withlupin grain on the proportion of the flock responding,ovulation rate or the frequency of short cycles in ewesstimulated with the ‘ram effect’. This outcome wasindependent of the previous nutritional history of theewes, although the percentage of ewes ovulating washigher in ewes in good condition than in ewes in poorÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

134 RJ Scaramuzzi and GB Martincondition. There is no obvious explanation for thedifferences in outcomes among these various studies.Nutritional studies are difficult because they requirestrict control of previous nutritional history. The mechanismbehind the reproductive response to nutritionalsupplementation is still problematic because we do nothave experimental models to explain the variability inthe response to a nutritional supplementation, the maleeffect and their interaction.Overall, it is clear that the efficacy of the ‘male effect’is influenced by nutritional and metabolic factors in thefemales. Thus, optimization of the efficacy of the ‘maleeffect’ for CGE management requires a very clearunderstanding of how nutrition influences the responsivenessof females to socio-sexual signals.ConclusionsThis brief review of nutritional interactions with seasonalrhythms and the ‘male effect’ draws attention tothe importance of nutrition as a major factor influencingreproductive outcomes and points to critical areas offuture research. There are some reports, often onlydescriptively documenting the nature of these interactionsbut there has been very little research exploring theunderlying processes. We need molecular genetic explanationsof the reproductive, metabolic, behavioural andphysiological processes that underpin, for example, theinfluence of nutritional inputs on the melatonin-drivenseasonal rhythms of ovarian cyclicity, and on femaleresponsiveness to the ‘male effect’ at both brain andovarian levels. By working towards a holistic perspectiveof the physiology, nutrition, genetics and behaviour ofour animals, we will be able to formulate ways tomanipulate their environment that will improve management,productivity and profitability and, simultaneously,promote a CGE industry.AcknowledgementsRJ Scaramuzzi is the recipient of a Marie Curie Chair of Excellencefrom the European Union (SUSTREPRO: FP6-2005-Mobility-10 ⁄ 42499) and their support is gratefully acknowledged. GB Martinwas funded by Meat and Livestock Australia (Project MS.027;‘LambMax’). Thanks are due to Dr B Malpaux and Dr D Guillaumefor their constructive comments on the manuscript.ReferencesBlache D, Adam CL, Martin GB, 2002: The mature malesheep: a model to study the effects of nutrition on thereproductive axis. Reprod Suppl 59, 219–233.Blache D, Chagas LM, Martin GB, 2007: Nutritional inputsinto the reproductive neuroendocrine control system – amultidimensional perspective. In: Juengel JI, Murray JF,Smith MF (eds), Reproduction in Domestic Ruminants VI.Nottingham University Press, Nottingham, pp. 123–139.Bronson FH, 1988: Effect of food manipulation on the GnRH-LH-estradiol axis of young female rats. Am J Physiol 254,R616–621.Cagampang FR, Maeda K, Yokoyama A, Ota K, 1990: Effectof food deprivation on the pulsatile LH release in the cyclingand ovariectomized female rat. Horm Metab Res 22, 269–272.Chemineau P, Pellicer-Rubio MT, Lassoued N, Khaldi G,Monniaux D, 2006: Male-induced short oestrous andovarian cycles in sheep and goats: a working hypothesis.Reprod Nutr Dev 46, 417–429.De Santiago-Miramontes MA, Rivas-Mun˜ oz R, Mun˜ oz-Gutiérrez M, Malpaux B, Scaramuzzi RJ, Delgadillo JA,2008: Ovulation rate in female anoestrous goats exposed tothe male effect is increased by nutritional supplementationunder grazing conditions. Anim Reprod Sci 105, 409–416.Delgadillo JA, Flores JA, Véliz FG, Duarte G, Vielma J,Hernandez H, Fernandez IG, 2006: Importance of thesignals provided by the buck for the success of the maleeffect in goats. Reprod Nutr Dev 46, 391–400.Downing JA, Scaramuzzi RJ, 1997: The effect of the infusionof insulin during the luteal phase of the estrous cycle on theovulation rate and on plasma concentrations of LH, FSHand glucose in ewes. Theriogenology 47, 747–759.Downing JA, Joss J, Scaramuzzi RJ, 1995a: A mixture of thebranched chain amino acids, leucine, isoleucine and valine,increases ovulation rate in ewes when infused during the lateluteal phase of the oestrous cycle: an effect that may bemediated by insulin. J Endocrinol 145, 315–323.Downing JA, Joss J, Scaramuzzi RJ, 1995b: Ovulation rateand the concentrations of gonadotrophins and metabolichormones in ewes infused with glucose during the lateluteal phase of the oestrous cycle. J Endocrinol 147, 403–410.Fisher JS, Martin GB, Oldham CM, Gray SJ, 1993: Long-termeffects of nutrition on spontaneous ovulation in Merinoewes and their responses to the ‘ram effect’. Proc VII WorldConf Anim Prod 2, 32–33.Forcada F, Abecia JA, 2006: The effect of nutrition on theseasonality of reproduction in ewes. Reprod Nutr Dev 46,355–365.Forcada F, Abecia JA, Sierra I, 1992: Seasonal changes inoestrous activity and ovulation rate in Rasa Aragonesa ewesmaintained at two different body condition levels. SmallRumin Res 8, 313–324.Gaskins CT, Snowder GD, Westman MK, Evans M, 2005:Influence of body weight, age, and weight gain on fertilityand prolificacy in four breeds of ewe lambs. J Anim Sci 83,1680–1689.Hafez ESE, 1952: Studies on the breeding season andreproduction of the ewe. J Agric Sci (Camb) 42, 189–265.Hartz AJ, Barboriak PN, Wong A, Katayama KP, Rimm AA,1979: The association of obesity with infertility and relatedmenstrual abnormalities in women. Int J Obes 3, 57–73.Ho¨ tzel MJ, Walkden-Brown SW, Fisher JS, Martin GB, 2003:Determinants of the annual pattern of reproduction inmature male Merino and Suffolk sheep: responses to anutritional stimulus in the breeding and non-breedingseasons. Reprod Fertil Dev 15, 1–9.Hulet CV, Shupe WL, Ross T, Richards W, 1986: Effects ofnutritional environment and ram effect on breeding seasonin range sheep. Theriogenology 25, 317–323.Hunter GL, 1962: Observations on oestrus in Merinos. Proc SAfr Soc Anim Prod 1, 67–69.Iglesias RMR, Ciccioli NH, Irazoqui H, Giglioli C, 1996:Ovulation rate in ewes after single oral glucogenic dosageduring a ram-induced follicular phase. Anim Reprod Sci 44,211–221.Kadokawa H, Martin GB, 2006: A new perspective onmanagement of reproduction in dairy cows: the need fordetailed metabolic information, an improved selection indexand extended lactation. J Reprod Dev 52, 161–168.Karsch FJ, Bittman EL, Foster DL, Goodman RL, Legan SJ,Robinson JE, 1984: Neuroendocrine basis of seasonalreproduction. Recent Prog Horm Res 40, 185–232.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Nutritional Interactions and Reproduction 135Knight TW, Oldham CM, Lindsay DR, 1975: Studies in ovineinfertility in agricultural regions in Western Australia: theinfluence of a supplement of lupins (Lupinus angustifolius cv.Uniwhite) at joining on the reproductive performance ofewes. Aust J Agric Res 26, 567–576.Lake JK, Power C, Cole TJ, 1997: Women’s reproductivehealth: the role of body mass index in early and adult life.Int J Obes Relat Metab Disord 21, 432–438.Lassoued N, Khaldi G, 1990: Influence du niveau alimentaireavant et apre` s la mise bas sur la brebis de race Barbarine àl’effet mâle. Ann INRAT Tunisie 63, 1–16.Leury BJ, Murray PJ, Rowe JB, 1990: Effect of nutrition onthe response in ovulation arte in Merino ewes followingshort-term lupin supplementation and insulin administration.Aust J Agric Res 41, 751–759.Malpaux B, 2006: Seasonal regulation of reproduction inmammals. In: Neill JD (ed.), Knobil and Neill’s Physiologyof Reproduction, 3rd edn. Elsevier Academic Press,Amsterdam, pp. 2231–2281.Martin GB, Kadokawa H, 2006: ‘Clean, green and ethical’animal production. Case study: reproductive efficiency insmall ruminants. J Reprod Dev 52, 145–152.Martin GB, Milton JT, Davidson RH, Banchero HunzickerGE, Lindsay DR, Blache D, 2004a: Natural methods forincreasing reproductive efficiency in small ruminants. AnimReprod Sci 82–83, 231–245.Martin GB, Rodger J, Blache D, 2004b: Nutritional andenvironmental effects on reproduction in small ruminants.Reprod Fertil Dev 16, 491–501.Molle G, Branca A, Ligios S, Sitzia M, Casu S, Landau S,Zoref Z, 1995: Effect of grazing background and flushingsupplementation on reproductive performance in Sardaewes. Small Rumin Res 17, 245–254.Molle G, Landau S, Branca A, Sitzia M, Fois N, Ligios S,Casu S, 1997: Flushing with soybean meal can improvereproductive performances in lactating Sarda ewes on amature pasture. Small Rumin Res 24, 157–165.Monje AR, Alberio R, Schiersmann G, Chedrese J, Carou N,Callejas S, 1992: Male effect on the post-partum sexualactivity of cows maintained on two nutritional levels. AnimReprod Sci 29, 145–156.Montgomery GW, Scott IC, Hudson N, 1985: An interactionbetween season of calving and nutrition on the resumptionof ovarian cycles in post-partum beef cattle. J Reprod Fertil73, 45–50.Nottle MB, Hynd PI, Seamark RF, Setchell BP, 1988:Increases in ovulation rate in lupin-fed ewes are initiatedby increases in protein digested post-ruminally. J ReprodFertil 84, 563–566.Nottle MB, Kleemann DO, Seamark RF, 1997a: Effect ofprevious undernutrition on the ovulation rate of merinoewes supplemented with lupin grain. Anim Reprod Sci 49,29–36.Nottle MB, Kleemann DO, Grosser TI, Seamark RF, 1997b:Evaluation of a nutritional strategy to increase ovulationrate in Merino ewes mated in late spring-early summer.Anim Reprod Sci 47, 255–261.Oldham CM, Lindsay DR, Martin GB, 1990: Effects ofseasonal variation of live weight on the breeding activity ofMerino ewes. In: Oldham CM, Martin GB, Purvis IW(eds), Reproductive Physiology of Merino Sheep – Conceptsand Consequences. School of Agriculture (AnimalScience), The University of Western Australia, Perth, pp.41–58.Pearce DT, Martin GB, Oldham CM, 1985: Corpora luteawith a short life-span induced by rams in seasonallyanovulatory ewes are prevented by progesterone delayingthe preovulatory surge of LH. J Reprod Fertil 75, 79–84.Rhodes FM, Fitzpatrick LA, Entwistle KW, Kinder JE, 1995:Pulsatile hormone secretion during the first ovarian follicularwave in Bos indicus heifers. J Reprod Fertil Suppl 49,523–526.Roberson MS, Wolfe MW, Stumpf TT, Werth LA, Cupp AS,Kojima ND, Wolfe PL, Kittok RJ, Kinder JE, 1991:Influence of growth rate and exposure to bulls on age atpuberty in beef heifers. J Anim Sci 69, 2092–2098.Salazar-Ortiz J, 2006: Bilan des effets du niveau d’alimentationsur la reproduction de la jument. Thèse pour obtenir legrade de Docteur de l, Universite´ de Tours, Universite´François Rabelais, Tours.Scaramuzzi RJ, Campbell BK, Downing JA, Kendall NR,Khalid M, Mun˜ oz-Gutiérrez M, Somchit A, 2006: A reviewof the effects of supplementary nutrition in the ewe on theconcentrations of reproductive and metabolic hormones andthe mechanisms that regulate folliculogenesis and ovulationrate. Reprod Nutr Dev 46, 339–354.Smith ID, 1965: The influence of level of nutrition duringwinter and spring upon oestrous activity in the ewe. WorldRev Anim Prod 4, 95–101.Stumpf TT, Wolfe MW, Wolfe PL, Day ML, Kittok RJ,Kinder JE, 1992: Weight changes prepartum and presence ofbulls postpartum interact to affect duration of postpartumanestrus in cows. J Anim Sci 70, 3133–3137.Szymanski LA, Schneider JE, Friedman MI, Ji H, Kurose Y,Blache D, Rao A, Dunshea FR, Clarke IJ, 2007: Changes ininsulin, glucose and ketone bodies, but not leptin or body fatcontent precede restoration of luteinising hormone secretionin ewes. J Neuroendocrinol 19, 449–460.Teleni E, Rowe JB, Croker KP, Murray PJ, King WR, 1989:Lupins and energy-yielding nutrients in ewes. II.Responses in ovulation rate in ewes to increased availabilityof glucose, acetate and amino acids. Reprod FertilDev 1, 117–125.Ungerfeld R, 2007: Socio-sexual signalling and gonadalfunction: opportunities for reproductive management indomestic ruminants. In: Juengel JI, Murray JF, Smith MF(eds), Reproduction in Domestic Ruminants VI. NottinghamUniversity Press, Nottingham, pp. 207–221.Ungerfeld R, Forsberg M, Rubianes E, 2004: Overview of theresponse of anoestrous ewes to the ram effect. Reprod FertilDev 16, 479–490.Véliz FG, Poindron P, Malpaux B, Delgadillo JA, 2006:Positive correlation between the body weight of anestrusgoats and their response to the male effect with sexuallyactive bucks. Reprod Nutr Dev 46, 657–661.Vin˜ oles C, Forsberg M, Martin GB, Cajarville C, Repetto J,Meikle A, 2005: Short-term nutritional supplementation ofewes in low body condition affects follicle development dueto an increase in glucose and metabolic hormones. Reproduction129, 299–309.Walker RF, Bethea CL, 1977: Gonadal function in underfedrats. I. Effect of pineal gland and constant light onmaturation and fecundity. Biol Reprod 17, 623–629.Wang JX, Warnes GW, Davies MJ, Norman RJ, 2004:Overweight infertile patients have a higher fecundity thannormal-weight women undergoing controlled ovarianhyperstimulation with intrauterine insemination. FertilSteril 81, 1710–1712.Webb R, Garnsworthy PC, Gong JG, Armstrong DG, 2004:Control of follicular growth: local interactions and nutritionalinfluences. J Anim Sci 82 (E-Suppl), E63–74.Wolfe BE, 2005: Reproductive health in women witheating disorders. J Obstet Gynecol Neonatal Nurs 34,255–263.Yeates NTM, 1949: The breeding season of the sheep withparticular reference to modification by artificial means usinglight. J Agric Sci (Camb) 39, 1–43.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

136 RJ Scaramuzzi and GB MartinYouatt W, 1837: Sheep, Their Breeds, Management andDiseases. Baldwin & Craddock, London.Zarazaga LA, Guzma´n JL, Domı´nguez C, Pérez MC, PrietoR, 2005: Effect of plane of nutrition on seasonality ofreproduction in Spanish Payoya goats. Anim Reprod Sci87, 253–267.Author’s address (for correspondence): RJ Scaramuzzi, UMR Physiologiede la Reproduction et des Comportements, Centre INRA deTours, 37380 Nouzilly, France. E-mail: rscara@rvc.ac.ukConflicts of interest: Much of GBM’s research is funded by Meat &Livestock, Australia; RJS has not declared any conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 137–143 (2008); doi: 10.1111/j.1439-0531.2008.01153.xISSN 0936-6768Developmental Capabilities of Embryos Produced In Vitro from Prepubertal LambOocytesKM MortonThe Camel Reproduction Centre, Dubai, United Arab EmiratesContentsBreeding from prepubertal females, known as juvenile in vitroembryo transfer (JIVET), reduces the generation interval andincreases the rate of genetic gain in animal breeding programs.While the birth of the first lambs from prepubertal ewesoccurred nearly 30 years ago; and there is considerable interestin the commercialization of this technology, its efficiencyremains too low. The advent of in vitro production (IVP) ofembryo resulted in the more widespread use of JIVET.Morphologic and metabolic differences coupled with reducedin vitro and in vivo development of oocytes derived fromprepubertal animals have been reported. Research has beenundertaken to optimize donor selection and hormone stimulationmethods in an attempt to reduce the variability andincrease the proportion of donors responding to hormonestimulation and increase oocyte developmental competence.Yet, this variation persists and the development of oocytes andembryos from prepubertal animals remains reduced whencompared with adults. Recent improvements to JIVET,resulting from a modified hormone stimulation regime, haveeliminated the failure of donors to respond to hormonestimulation, and increased both the number and developmentalcompetence of oocytes harvested from very young prepubertallambs. This increased efficiency has facilitated theincorporation of other reproductive technologies such assperm sexing with JIVET, resulting in the birth of lambs of apre-determined sex from prepubertal lambs. Increased rates ofgenetic gain in sheep breeding programs can be achieved bycombining sexed sperm with oocytes obtained from lambs asyoung as 3–4 weeks of age. Continued increases in theefficiency of JIVET resulting from further improvements tohormone stimulation regimes and an increased understandingof the differences between oocytes from adult and prepubertalanimals will result in the commercialization of this technology.IntroductionThe use of prepubertal ewes for breeding programsprovides an exciting opportunity for sheep producers toreduce the generation interval and thereby, increase therate of genetic gain in their flocks (Nicholas 1996;Armstrong et al. 1997). This can be performed using aform of assisted reproductive technology, known asjuvenile in vitro embryo transfer (JIVET), to produceoffspring after the transfer of in vitro produced (IVP)embryos derived from oocytes obtained from prepubertal(juvenile) animals.Traditional sheep multiple ovulation and embryotransfer (MOET) schemes, which use adult ewes, resultin a generation interval of at least 12 months. Using 8–12-week-old lambs in a JIVET scheme, reduces thegeneration interval to 7 months (van der Werf 2005).Using oocytes obtained from 3- to 4-week-old lambsfurther reduces this generation interval to 6 months, andoffers producers a substantial increase in the rate ofgenetic gain when compared with other reproductivetechnologies such as MOET. Further increases in therate of genetic gain (of approximately 5%) can beachieved by combining sexed sperm with JIVET, whichis most advantageous at the commercial level (van derWerf 2005).During the early post-natal period in sheep, there isremarkable growth of the reproductive tract. Ovarianweight increased to 6.8 and 10.8 times the weight at birthby 4 and 8 weeks of age, respectively, and the number ofgrowing ovarian follicles increased significantly by4 weeks of age (Kennedy et al. 1974). Both the numberof growing follicles and ovarian weight declining by 8–12 weeks of age remained constant until 33 weeks of age(Kennedy et al. 1974). Ovarian follicles from 4-week-oldlambs began to show signs of atresia whilst the numberof follicles displaying signs of advanced atresia werehigher from 8- to 10-week-old lambs (Tassell et al. 1978).These results suggest this post-natal flourish of folliculargrowth provides a ‘window of opportunity’ to collectlarge numbers of oocytes from non-atretic follicles whenlambs are aged between 3 and 8 weeks of age.Initial attempts to produce embryos from prepubertallambs were achieved using MOET and, while prepubertallambs were capable of responding to exogenousgonadotrophins (Tassell et al. 1978; Worthington andKennedy 1979), the results were highly variable and thesuperovulatory response was generally low. In addition,embryos which were recovered from prepubertal lambsdisplayed a reduced development during in vitro culture(Wright et al. 1976) and after transfer to recipients(Quirke and Hanrahan 1977; McMillan and McDonald1985) when compared with those obtained from adultdonors. Despite these difficulties, a lamb was successfullyborn after the transfer of embryos derived from 10-to 16-week-old Welsh Mountain ewe lambs (Trounsonet al. 1977). The development of IVP, combined withjuvenile oocytes resulted in the birth of a calf (Armstronget al. 1992) followed closely by the birth of lambs(Armstrong et al. 1994; Earl et al. 1994).The use of JIVET was considerably expanded by thedevelopment of IVP and increased knowledge about thedifferences between oocytes from prepubertal and adultanimals. Despite considerable research, the efficiency ofJIVET has remained below commercially acceptablelevels, and increases in efficiency have been hindered bythe high proportion of lambs which do not respond tohormone stimulation, the highly variable response tohormone stimulation, the low proportion of oocytes thatreach the blastocyst stage and the high rates of foetalloss and malformations observed after the transfer ofIVP embryos derived from prepubertal animals (PtakÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

138 KM Mortonet al. 1999). Increases in both the number and in vitrodevelopmental competence of oocytes derived from veryyoung donors have been reported recently (Mortonet al. 2005c). This improvement has been attributed tothe hormone stimulation regime and an optimized IVPsystem, and has resulted in the birth of offspring fromIVP embryos derived from prepubertal lambs as youngas 3–4 weeks of age (Morton et al. 2004a, 2005d) andthe first pre-sexed JIVET offspring (Morton et al.2004a). The current status of JIVET and a number offactors affecting its efficiency including selection of thedonor, hormone stimulation prior to oocyte collectionas well as the developmental capabilities of prepubertaloocytes will be described.Donor SelectionThe ideal donor lamb will produce large numbers offollicles and developmentally competent oocytes inresponse to hormone stimulation. Prepubertal lambs,like their adult counterparts, display a wide variation inthe number of ovarian follicles (both pre- and posthormonestimulation) and in the number of oocytescollected (O’Brien et al. 1997a; Morton et al. 2005b; c,d). Approximately, 20% of lambs failed to respond tohormone stimulation (Ptak et al. 1999, 2003), which issimilar to adult animals (Cognie 1999). Oocyte developmentto the blastocyst stage is highly dependant onindividual lambs (Ptak et al. 2003; Morton et al. 2005c),with the number of embryos produced per lamb rangingfrom 0 to 55 (Morton et al. 2005b, c). Thus, selection oflambs which produce large numbers of follicles andoocytes with a high developmental potential becomesparamount.Various methods have been studied for selectinglambs although most have been inconclusive. Bodyweight prior to hormone stimulation and growth rateare not related to ovarian weight or the number ofovarian follicles after hormone stimulation (Mortonet al. 2005c). Donors have been selected on the basis ofovarian follicular status assessed by laparoscopy priorto hormone stimulation (Earl et al. 1995b); the numberof follicles observed was highly related (R = 0.98) tothe number post-hormone stimulation. Yet, this requirestwo laporscopic procedures within a fortnight and thereis no correlation between the number of ovarian folliclesand the in vitro developmental capabilities of theseoocytes. Ptak et al. (2003) observed similarities betweenfive sibling lambs derived from a lamb with oocytes of ahigh developmental capability, highlighting the role ofgenotype in oocyte developmental competence.Age of oocyte donorsArmstrong et al. (1997) suggest that the ideal age tocollect oocytes from prepubertal lambs is between 4 and6 weeks of age as this is the time of most follicularresponsiveness. Yet, in cattle oocytes obtained fromolder donors display an increased response to hormonestimulation (Armstrong et al. 1992, 1994), oocyte developmentin vitro (Presicce et al. 1995; Tervit et al. 1997)and pregnancy rate (Yang et al. 1997) when comparedwith oocytes from younger animals. Earl et al. (1995a)reported similar rates of in vitro development for oocytesfrom 8- to 9-week-hormone stimulated and 4-monthunstimulatedlambs and O’Brien et al. (1997a) reportedthat development of oocytes in vitro was higher from3- to 6-week-old hormone stimulated lambs than 16–24-week-old unstimulated lambs. Yet, Morton et al.(2005d) observed similar rates of oocyte developmentto the blastocyst stage for oocytes derived from 16- to24-week-unstimulated prepubertal lambs (29.0%) andadult ewes (39.3%) but was reduced for oocytes derivedfrom 3- to 6-week-hormone stimulated lambs (27.8%).These studies compared the development of oocytesfrom younger hormone stimulated lambs with olderunstimulated lambs, thereby precluding the identificationof the relative contributions of lamb age andhormone stimulation.Comparing hormone stimulated lambs of differingages (3–4 and 6–7 weeks), Morton et al. (2005b)observed that lamb age affected uterine weight, whileovarian weight, the number of ovarian follicles, thenumber of oocytes recovered or the proportion ofoocytes suitable for IVP were not affected. Yet, oocytedevelopment to the blastocyst stage was higher for 6–7 week (27.9%) than 3–4 week (20.4%) lambs demonstratingthe increase in oocyte development with donorage. Despite these results and those of Morton et al.(2005d), the greater follicular responsiveness and numberof oocytes obtained from younger lambs (4–6 weeksof age) outweights the increase in development whenoocytes are obtained from older lambs agreeing withArmstrong et al. (1997).In summary, a window of opportunity which existswhen lambs are aged between 3 and 6 weeks, to collectlarge numbers of oocytes which are capable of fulldevelopment in vitro and in vivo. Yet, the lack of efficientmethods for selecting oocyte donors lambs prevents fullexploitation of this opportunity and further research isrequired to develop an efficient method of selecting 3–6-week-old lambs for JIVET.Hormone Stimulation of DonorsHormone stimulation prior to oocyte collection is usedto increase the number of ovarian follicles and oocytesharvested (O’Brien et al. 1997a; Ledda et al. 1999;Morton et al. 2005b), collect oocytes at a synchronousstage of development (Armstrong et al. 1997) andincrease oocyte development in vitro (O’Brien et al.1997a; Morton et al. 2005b).While the precise mechanisms responsible for thisincrease in developmental competence are not known,the effects are not mediated solely through an increase infollicle size, but also by the activation of biosyntheticprocesses within the oocyte that are required forsuccessful embryo development (Armstrong 2001). Biochemicalchanges in the follicle cells and intrafollicularenvironment that would not have otherwise occurreduntil after puberty are also induced by hormonetreatment (O’Brien et al. 1997a), for example, increasingthe production of gonadotrophin receptors in granulosacells resulting in an enhanced ability of the follicle torespond to gonadotrophin stimulation, thereby improvingoocyte quality and function.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Developmental Capabilities of Prepubertal Lamb Embryos 139Most researchers do include hormone stimulation intheir JIVET regimes for lambs (Ptak et al. 1999, 2003,2006; Kelly et al. 2005a,b, 2007; Morton et al. 2005a; b,c, d) although Ledda et al. (1999) reported no differencein developmental competence of oocytes from unstimulatedor stimulated oocyte donors. The birth ofoffspring after in vitro maturated-in vitro fertilised(IVM-IVF) of oocytes from untreated 4–6-week-oldlambs (Ledda et al. 1999) demonstrates that hormonestimulation is not mandatory for such young donorsalthough it is usually applied for the aforementionedreasons.Considerable research has been undertaken to optimizethe regime for hormone stimulation of prepubertallambs. Hormone stimulation regimes used with prepubertalanimals have been based on adult superovulationregimes and consist of treatment over 3–4 days withFSH (single or multiple doses) or gonadotrophins with along half-life [i.e. equine chorionic gonadotrophin(eCG); Armstrong et al. 1997].Single vs multiple injections of FSH did not affect thenumber of oocytes recovered and cultured per lamb(Armstrong et al. 1994; Ptak et al. 1999). Yet, multipleFSH injections increased oocyte development to theblastocyst stage (Ptak et al. 1999). Furthermore, blastocystdevelopment was increased when oocytes werecollected 48 and 72 h after the initial FSH injection andlambs received multiple FSH injections when comparedwith a single FSH injection (Kelly et al. 2005b).Hormone stimulation of prepubertal donors hasrelied on lower doses of gonadotrophins than adultsbecause of their lower body mass and greater follicularresponsiveness before the intraovarian regulatory mechanismsdesigned to limit ovulation rate have come intoplay (Armstrong et al. 1997). Given that oocytes derivedfrom prepubertal animals have a relatively lower exposureto gonadotrophins when compared with oocytesderived from adults, and exposure to increased FSHduring IVM increases oocyte developmental competencein vitro (Morton et al. 2004b), larger gonadotrophindoses may confirm advantageous for cytoplasmic maturationof prepubertal oocytes. Recently, the use oflarger doses of FSH (130–160 mg) has resulted in anincrease in the efficiency of JIVET (Morton et al. 2004a,2005a; b, c, d; Kelly et al. 2005b, 2007).The inclusion of eCG in the regime did not increasethe number of oocytes collected but oocyte developmentto the blastocyst stage was increased (42.4% vs 43.9% ofoocytes; Kelly et al. 2005b) as did the administration ofeCG with the last FSH injection (29.6% vs 18.9%; Kellyet al. 2005b). Yet, Morton et al. (2005a,b,c,d) reportedsimilar rates of oocyte development in vitro when eCGwas administered with first FSH injection.Hormone stimulation regimes consisting of bothsteroids and gonadotrophins are known to increase theyield and in vitro development of oocytes (O’Brien et al.1997a; Morton et al. 2005b) although the relativecontribution of steroids to this increase remainsunknown and the majority of regimes do not utilizesteroids. Progestagen treatment for 7 days, followed byits withdrawal, resulted in a twofold increase in thenumber of ovarian follicles when compared with calvesand lambs where progestagen was not withdrawn(Armstrong et al. 1994), suggesting that folliculargrowth is suppressed in the presence of progestagen.Yet, oocytes subsequently collected after progestagenwithdrawal were observed to be at different stages ofmaturation suggesting that the increase in the number ofovarian follicles may be negated by a reduction inoocyte quality.While, progesterone treatment did not affect thenumber of oocytes collected, the efficiency of blastocystproduction dependent on the time of oocyte collection(48 h vs 60 h) and type of FSH administration(1 · 160 mg dose vs 4 · 40 mg dose; Kelly et al.2005b). These authors reported the highest oocytedevelopment to the blastocyst stage after no progesteronetreatment, four 40 mg injections of FSH, andcollection of oocytes 48 h after the last FSH treatment.While these results suggest that progesterone treatmentis not necessary, it has been successfully used byprevious authors (O’Brien et al. 1997a; Ptak et al.1999, 2003; Morton et al. 2004a, 2005a; b, c, d) withhigh ovarian responses reported (approximately 120–140 follicles per lamb; Morton et al. 2005b).GnRH treatment of lambs prior to oocyte collection,to facilitate the collection of in vivo matured oocytes, didnot alter the number of oocytes collected, oocytemorphology or development to the blastocyst stage(Kelly et al. 2007). Previously, Armstrong et al. (1997)administered GnRH to lambs and reported difficulty inaccurately controlling the timing of the endogenous LHsurge and asynchronous stages of maturity both withinand between donors for the oocytes collected. For thesereasons, GnRH is not routinely included in hormonestimulation regimes as most researchers aim to collectimmature oocytes and mature them in vitro.In summary, hormone stimulation increases thenumber of oocytes collected as well as oocyte developmentalcompetence, but the optimal regime for hormonestimulation of prepubertal lambs and the effects of therelative components of hormone stimulation regime onoocyte developmental competence have yet to be determined.Response to hormone stimulationThe ovarian response to hormone stimulation variesconsiderably between individual prepubertal donors(Earl et al. 1994; Ptak et al. 1999, 2003). Ptak et al.(2003) found that 50–72% of lambs aged 2–7 monthsdid not have a sufficient response to hormone stimulation(number and size of follicles) for embryo production.Morton et al. (2005b), on the contrary, observedthat all lambs responded to hormone stimulation and24.0% (6 ⁄ 25), 44.0% (11 ⁄ 25) and 32.0% (8 ⁄ 25) of 3–4-week-old lambs had low (100) responses, respectively,and a similar distribution was observed for 6–7-week-old lambs.Ptak et al. (1999, 2003) observed that oocytes fromlambs with a low response to hormone stimulation didnot develop to the blastocyst stage in vitro. On thecontrary, Morton et al. (2005b) reported that oocytesfrom low responding 3–4 and 6–7-week-old lambs werecapable of full development in vitro. Furthermore, whileÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

140 KM Mortonthe number of oocytes suitable for IVP was reduced tolow when compared with medium and high respondinglambs, the percentage of oocytes suitable for culture (%of total recovered) did not differ between low andmedium responding lambs, but was reduced in highresponding lambs (Morton et al. 2005b). Oocyte cleavageand blastocyst formation were not affected byresponse to hormone stimulation in 3–4-week-oldlambs, but oocytes from 6- to 7-week-old lambs with ahigh response to hormone stimulation displayed reducedrates of cleavage (Morton et al. 2005b). It remains to beseen whether this result was related to individual lambs,or the early development of the negative feedbackmechanisms normally established after puberty to limitovulation rate.The effect of hormone stimulation regime on thevariation in response to stimulation remains unknown.Most authors report considerable variation in theresponse to hormone stimulation regardless of theregime used. This variation is also accompanied by ahigh proportion of lambs which do not respond tostimulation and by poor in vivo survival of embryos.Yet, a modified regime, consisting of progestagenpriming, steroids and gonadotrophins (eCG and FSH)yielded a high average response to stimulation and highin vivo survival of IVP embryos (Morton et al. 2004a,2005a, b, c, d).In summary, modifications to hormone stimulationregimes have reduce the proportion of prepubertallambs which do not respond to stimulation, therebyincreasing the efficiency of JIVET. Using the modifiedhormone stimulation regime, the majority of lambsdisplayed a medium to high response to hormonestimulation which does not reduce oocyte developmentin vitro for 3–4-week-old lambs. Despite these improvements,the variation in response to hormone stimulationbetween lambs persists.Morphologic and Metabolic Characteristics ofPrepubertal OocytesMorphological differences have been observed in theorganelles of adult and prepubertal oocytes. Oocytesfrom prepubertal animals are smaller than those derivedfrom adult animals but in GV stage oocytes derivedfrom prepubertal and adult animals, the morphologywas similar for cytoplasmic organelles such as themitochondria, Golgi complexes, endoplasmic reticulaand lipid droplets (Ledda et al. 1997). Yet, oocytesderived from prepubertal compared with adult ewescontained a lower number of corona cell foot projections(Ledda et al. 2001) and there was a delay in themigration of the cortical granules in IVM oocytes(O’Brien et al. 1996).Protein synthesis during oocyte maturation, which isdependant on the correct coupling of the cumulus cellsand the oocyte, is essential for normal meiotic progressionand embryo development (Moor and Crosby 1986).Lower amino acid uptake (Ledda et al. 1996b, 2001;Kochhar et al. 2002) and protein synthesis (Kochharet al. 2002) during IVM, and different times of peakprotein synthesis have been reported for prepubertalwhen compared with adult oocytes (Kochhar et al.2002), and is related to the defective cumulus cell-oocytecoupling reported by Ledda et al. (1996b, 2001). Inaddition, the level and pattern of maturation promotingfactor (MPF) activity was similar in adult and prepubertaloocytes but after maturation, MPF activity waslower in oocytes from prepubertal than adult animals(Ledda et al. 2001).Metabolic differences between adult and prepubertaloocytes have also been reported. Glutamine metabolismduring IVM was lower for prepubertal than adultoocytes while glucose and pyruvate metabolism wassimilar (O’Brien et al. 1996). These results point tomorphologic, metabolic and ultrastructural differencesbetween oocytes from adult and prepubertal animals,although it must be remembered that oocytes used inthese studies were obtained from both unstimulated andhormone stimulated prepubertal lambs, ranging in agefrom 40 days to 6 months, which may have influencedthe findings.In summary, there are clear morphological and metabolicdifferences between oocytes derived from adultand prepubertal animals. Yet, comparisons betweenstudies are confounded by differing ages of the prepubertaldonors (ranging from days to many months) andwhether the donors were hormone stimulated prior tooocyte collection. The extend of these differences remainsunclear and studies describing the changes in morphologicand metabolic characteristics of oocytes from veryyoung (3–4 week) to older (16–24 week) are required.Development of Oocytes from PrepubertalanimalsCompared with adult oocytes, those from prepubertaldonors display similar rates of meiotic maturation(O’Brien et al. 1996, 1997b; Ledda et al. 2001) but ahigher prevalence of fertilization abnormalities (Leddaet al. 1996a, 1997; O’Brien et al. 1996, 1997b; Kochharet al. 2002) and lower development to the blastocyststage during in vitro culture (O’Brien et al. 1996, 1997b;Ptak et al. 1999; Kochhar et al. 2002; Leoni et al.2006b), attributed to perturbed cytoplasmic maturation(discussed below).Blastocysts produced in vitro, derived from prepubertaloocytes, are morphologically indistinguishable(O’Brien et al. 1997b), contain similar numbers of cells(O’Brien et al. 1996; Ledda et al. 1997; Kochhar et al.2002; Leoni et al. 2006b) and have a similar inner cellmass to total cell number ratio (Kochhar et al. 2002;Leoni et al. 2006b), when compared with those derivedfrom adult oocytes. Yet, altered speed of developmenthas been reported for oocytes derived from prepubertalsheep. Kochhar et al. (2002) reported altered kinetics ofIVM, observing that lamb oocytes required an additional2 h of IVM to ‘catch up’, but other studies havereported no difference (O’Brien et al. 1996; Ledda et al.1997). Delays in development of embryos derived fromprepubertal oocytes have been observed previously atthe two-cell (Leoni et al. 2006b), four-cell (Ptak et al.1999) and blastocyst stage (O’Brien et al. 1997b; Ptaket al. 1999; Leoni et al. 2006b). Yet, Morton et al.(2005d) reported no difference in the speed of embryodevelopment which may be related to higher FSH dosesÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Developmental Capabilities of Prepubertal Lamb Embryos 141during hormone stimulation, or differences in therespective culture systems, including media (KM Mortonand SL Catt, unpublished data) and oxygen tension(Leoni et al. 2007).The kinetics of both oocyte maturation (Dominko andFirst 1997) and embryo development (Van Soom et al.1997) indicate the developmental capabilities of embryos.The reduced speed of development of prepubertal derivedembryos may explain the low, or reduced in vivo survivalreported by numerous workers (Quirke and Hanrahan1977; McMillan and McDonald 1985; O’Brien et al.1997a; Ptak et al. 1999, 2003, 2006; Kelly et al. 2005b). Inaddition to this low survival rate, remarkably high ratesof embryonic ⁄ foetal loss (O’Brien et al. 1997a; Ptak et al.1999, 2003, 2006) and mummification (Ptak et al. 1999)have been reported. Ptak et al. (2006) reported severelyinterrupted pregnancy (between days 40 and 60) andfoetal losses (days 80–100) with only 6% of the 628embryos transferred surviving until full gestation.This reduced development is generally attributed toincomplete or perturbed cytoplasmic maturation, whichmay be expressed at many stages of embryonic development.Perturbed cytoplasmic maturation commonlypresents as a failure of sperm penetration and decondensation,inability to form normal male pronuclei,failure of the block to polyspermy, early cleavagefailure, failure to reach or survive the transition frommaternal to embryonic genome expression, and developmentalfailure leading to embryonic losses at laterpre-implantation and post-implantation stages of development(Armstrong 2001). All of these have beenreported for oocytes and embryos derived from prepubertalanimals, and the evidence regarding perturbedcytoplasmic maturation in prepubertal oocytes is nowoverwhelming.The perturbed cytoplasmic maturation is generallyattributed to the age of the donors and the lack of timeto sequester factors into the cytoplasm (Armstrong2001). Yet, Ptak et al. (2006) also demonstrated incompletenuclear maturation of prepubertal lamb oocytes,and the authors suggest that the previous definition ofnuclear competence (i.e. the ability to complete meiosis)was inadequate. Furthermore, the authors observeddiscordance between follicle and oocyte growth, leadingto incomplete or perturbed oocyte growth. This, accordingto Ptak et al. (2006), resulted in incomplete developmentof both the nuclear and cytoplasmiccompartments, which was responsible for the reduceddevelopmental competence of oocytes derived fromprepubertal animals.Reduction in global genome methylation (Ptak et al.2006) and mRNA storage (Leoni et al. 2006a) havebeen reported for oocytes derived from prepubertallambs. Ptak et al. (2006) observed that genome-widemethylation status (% of methylated vs total DNA)was lower for oocytes derived from prepubertal lambswhen compared with adult animals. The authorshypothesize that the reduced global methylation affectscertain imprinted genes as the patterns of foetal lossobserved by Ptak et al. (1999, 2006) are similar to micelacking oocyte-specific DNA methyltransferase-1(Dnmt1o) where the majority of foetus die in the lastthird of gestation (Howell et al. 2001). The reducedmethylation, mRNA storage and perturbed growth ofoocytes from prepubertal animals, provides strongevidence for the contribution of nuclear immaturityto their reduced developmental competence. Furtherstudies aiming to locate the specific sites where thedifferences in methylation occur will provide molecularexplanations for the differences in the developmentalcapabilities of oocytes from adult and prepubertalanimals.Most workers, most notably Ptak et al. (1999, 2006),have reported low rates of in vitro development forprepubertal oocytes accompanied by substantial embryonicand foetal loss. In other studies, by contrast,prepubertal oocytes have displayed a high developmentalcompetence in vitro not dissimilar to that of oocytesfrom adult sheep, suggesting that the use of large FSHdoses during hormone stimulation may attenuate manyof the problems associated with JIVET, specifically thehigh proportion of lambs which fail to respond tohormone stimulation, high rates of fertilization abnormalities,reduced and delayed development to the blastocyststage and compromised embryo survival aftertransfer to recipient ewes (Morton et al. 2005b; c, d).While Ptak et al. (1999, 2006) and Morton et al.(2005b,c,d) utilized oocytes from lambs of a similar age(4 weeks), the different breeds of the donors, hormonestimulation regimes (in particular, the FSH dose: 2.7 or7 mg and 130 mg pFSH, respectively) and IVP systemsconfound comparisons between these studies. The role ofin vitro culture on the perturbation of embryonic geneexpression is well-documented (Wrenzycki et al. 2005).Yet, the effects of different IVC systems and hormonestimulation with a high FSH regime during oocytegrowth on gene expression, have yet to be investigatedfor oocytes from prepubertal animals.Despite the reduced developmental competence ofoocytes from prepubertal animals, offspring have beenproduced from both fresh and frozen-thawed embryosderived from lambs as young as 3–4 weeks of age(Morton et al. 2004a, 2005d) and the improvements tothe efficiency of JIVET technology have facilitated theincorporation of JIVET with other emerging reproductivetechnologies such as sperm sexing (Maxwell et al.2004) for the production of pre-sexed embryos andoffspring.Future Directions and PerspectivesLimitations to the commercial application of JIVEThave included the high proportion of donors that fail torespond to hormone stimulation, the highly variableresponse to hormone stimulation and the reduceddevelopmental competence of oocytes derived fromprepubertal animals. Using a high FSH stimulationregime (Morton et al. 2004a, 2005a, b, c, d) all lambsresponded to hormone stimulation and oocyte developmentalcompetence was significantly improved. Developmentrate was similar for oocytes derived fromprepubertal and adult ewes, and in vivo survival afterembryo transfer was high and free from the highmalformation rates reported by other authors.The advent of genomic technologies has furtherincreased our knowledge on the differences betweenÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

142 KM Mortonoocytes derived from adult and prepubertal animals.Alterations in the expression of several developmentallyimportant genes have been observed in embryos derivedfrom prepubertal cattle (Oropeza et al. 2004) and sheep(Leoni et al. 2006a). Gene expression and epigeneticstudies will undoubtedly elucidate differences in oocytesand embryos derived from adult and prepubertalanimals. Furthermore, this technology will be used tore-evaluate hormone stimulation and IVP systems, sothat prepubertal oocytes and embryos may haveimproved competence.Two new techniques have great potential for combinationwith JIVET. Further reductions in the generationinterval on the female side could be achieved by incorporatingfoetal oocytes into IVP systems. Yet, foetal calfoocytes display lower rates of maturation, fertilization,cleavage and embryo development compared with oocytesfrom adult animals (Chohan and Hunter 2004) andfurther research is required. Utilization of gametes fromprepubertal male animals could dramatically reduce thegeneration interval on the male side. Spermatogenesiscan be initiated in prepubertal ram lambs by administeringeCG (Morton et al. 2004c). The injection of gametesfrom prepubertal ram lambs into IVM oocytes fromadult and prepubertal lambs resulted in the production of16–32 cell stage embryos (KM Morton, SL Catt, WMCMaxwell and G Evans, unpublished data). This technologyoffers, for the first time, a chance for producers toreduce the generation interval on the paternal side. Whilethe potential of this method has been demonstrated, thebirth of lambs produced from both prepubertal ram andewe lambs, and further refinement of treatment regimesfor ram lambs are required.In conclusion, oocytes from prepubertal animalsdisplay a reduced developmental competence whencompared with oocytes derived from their adult counterparts,resulting from perturbations to cytoplasmic andnuclear maturation. Despite these perturbations, oocytesfrom prepubertal lambs as young as 3–4 weeks of age arecapable of developing into viable offspring after transferto recipients. Recent advances in JIVET technology,which have significantly improved its efficiency, havebrought its commercialization closer and facilitatedincorporation with other emerging reproductive technologiessuch as sperm sexing. With the developments ingenomic technologies continuing to add to knowledgeregarding the differences between oocytes derived fromadult and prepubertal animals and the improvements inthe efficiency of JIVET this will bring, commercializationof this technology is foreseeable.AcknowledgementsThe Australian Research Council (ARC), XY Inc, Bioniche AnimalHealth Australasia and Sydney IVF for funding and research support.Professor W.M.C. Maxwell is thanked for his invaluable editorialassistance.ReferencesArmstrong DT, 2001: Effects of maternal age on oocytedevelopmental competence. Theriogenology 55, 1303–1322.Armstrong DT, Holm P, Irvine B, Petersen BA, Stubbings RB,McLean D, Stevens R, Seamark RF, 1992: Pregnancies andlive birth from in vitro fertilization of calf oocytes collectedby laparoscopic follicular aspiration. Theriogenology 38,667–678.Armstrong DT, Irvine BJ, Earl CR, 1994: In vitro fertilisationof follicular oocytes from juvenile lambs and their developmentalcompetence in vitro and in vivo. Biol Reprod50(Suppl.), 189.Armstrong DT, Kotaras PJ, Earl CR, 1997: Advances in theproduction of embryos in vitro from juvenile and prepubertaloocytes from the calf and lamb. Reprod Fertil Dev 9,333–339.Chohan KR, Hunter AG, 2004: In vitro maturation, fertilizationand early cleavage rates of bovine fetal oocytes.Theriogenology 61, 373–380.Cognie Y, 1999: State of the art in sheep-goat embryo transfer.Theriogenology 51, 105–116.Dominko T, First NL, 1997: Timing of meiotic progression inbovine oocytes and its effect of early embryonic development.Mol Reprod Dev 47, 456–467.Earl CR, Irvine B, Armstrong DT, 1994: Development oftechniques for the production of viable embryos from sixto seven week old lambs. Proc Aust Soc Anim Prod 20,428.Earl CR, Irvine BJ, Kelly JM, Rowe J, Armstrong DT, 1995a:Ovarian stimulation protocols for oocyte collection andin vitro embryo production from 8 to 9 week old lambs.Theriogenology 43, 203.Earl CR, Kotaras PJ, Rowe J, Kelly JM, DeBarro TM, 1995b:Laparoscopic selection of juvenile donors prior to stimulation.Proc Aust Soc Reprod Biol 26, 2.Howell CY, Bestor TH, Ding F, Latham KE, Mertineit C,Trasler JM, Chaillet JR, 2001: Genomic imprinting disruptedby a maternal effect mutation in the Dnmt1 gene.Cell 104, 829–838.Kelly JM, Kleeman DO, Walker SK, 2005a: The effect ofnutrition during pregnancy on the in vitro production ofembryos from resulting lambs. Theriogenology 63, 2020–2031.Kelly JM, Kleeman DO, Walker SK, 2005b: Enhancedefficiency in the production of offspring from 4- to 8-weekoldlambs. Theriogenology 63, 1876–1890.Kelly JM, Kleeman DO, Maxwell WMC, Walker SK, 2007:Effect of GnRH treatment on the maturation and in vitrodevelopment of oocytes collected from 4- to 6-week-oldMerino lambs. Reprod Fertil Dev 19, 947–953.Kennedy JP, Worthington CA, Cole ER, 1974: The postnataldevelopment of the ovary and uterus in the Merino lamb. JReprod Fertil 36, 275–282.Kochhar HP, Wu B, Morris LH, Buckrell BC, Pollard JW,Basrur PK, King WA, 2002: Maturation status, proteinsynthesis and developmental competence of oocytes derivedfrom lambs and ewes. Reprod Domest Anim 37, 19–25.Ledda S, Bogliolo L, Calvia P, Leoni G, Naitana S, 1996a:Developmental competence of follicular oocytes from juvenilelambs matured in vitro in different conditions. J ReprodFertil Abstr Ser 17, 28.Ledda S, Bogliolo L, Leoni G, Calvia P, Naitana S, 1996b:Influence of vasoactive intestinal peptide (VIP), atrialnatiuretic peptide (ANP) and insulin-like growth factors-I(IGF-I) on in vitro maturation of prepubertal and adultsheep oocytes. Zygote 4, 343–348.Ledda S, Bogliolo L, Calvia P, Leoni G, Naitana S, 1997:Meiotic progression and developmental competence ofoocytes collected from juvenile and adult ewes. J ReprodFertil 109, 73–78.Ledda S, Bogliolo L, Leoni G, Naitana S, 1999: Productionand lambing rate of blastocysts derived from in vitromatured oocytes after gonadotropin treatment of prepubertalewes. J Anim Sci 77, 2234–2239.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Developmental Capabilities of Prepubertal Lamb Embryos 143Ledda S, Bogliolo L, Leoni G, Naitana S, 2001: Cell couplingand maturation-promoting factor activity in in vitromaturedprepubertal and adult sheep oocytes. Biol Reprod65, 247–252.Leoni GG, Bebbere D, Succu S, Berlinguer F, Mossa F,Galioto M, Bogliolo L, Ledda S, Naitana S, 2006a:Relations between relative mRNA abundance and developmentalcompetence of ovine oocytes. Mol Reprod Dev 74,249–257.Leoni GG, Succu S, Berlinguer F, Rosati I, Bebbere D,Bogliolo L, Ledda S, Naitana S, 2006b: Delay on the in vitrokinetic development of prepubertal ovine embryos. AnimReprod Sci 92, 373–383.Leoni GG, Rosati I, Succu S, Bogliolo L, Bebbere L,Berlinguer F, Ledda S, Naitana S, 2007: A low oxygenatmosphere during IVF accelerates the kinetic of formationof in vitro produced ovine blastocysts. Reprod DomestAnim 42, 299–304.Maxwell WMC, Evans G, Hollinshead FK, Bathgate R, deGraaf SP, Eriksson B, Gillan L, Morton KM, O’Brien JK,2004: Integration of sperm sexing into the ART toolbox.Anim Reprod Sci 82–83, 79–95.McMillan WH, McDonald MF, 1985: Survival of fertilisedova from ewe lambs and adult ewes in the uteri of ewelambs. Anim Reprod Sci 8, 235–240.Moor RM, Crosby IM, 1986: Protein requirements forgerminal vesicle breakdown in ovine oocytes. J EmbryolExp Morph 94, 207–220.Morton KM, Catt SL, Hollinshead FK, Maxwell WMC,Evans G, 2004a: Production of lambs after the transfer offresh and cryopreserved in vitro produced embryos fromprepubertal lamb oocytes and unsorted and sex-sortedfrozen-thawed spermatozoa. Reprod Domest Anim 39,454–461.Morton KM, Maxwell WMC, Evans G, 2004b: The effect ofFSH concentration during IVM and gamete co-incubationlength during IVF on the development of unstimulatedprepubertal ewe oocytes. Reprod Fertil Dev 16, 205.Morton KM, Maxwell WMC, Evans G, 2004c: The effect ofPMSG administration on the growth and development ofthe reproductive tract of prepubertal ram lambs. In:Proceedings of the 15th International Congress on AnimalReproduction Porto Seguro, Brazil, 1, 233.Morton KM, Catt SL, Hollinshead FK, Maxwell WMC,Evans G, 2005a: The effect of gamete co-incubation timeduring in vitro fertilisation with frozen-thawed unsorted andsex-sorted ram spermatozoa on the development of in vitromatured adult and prepubertal ewe oocytes. Theriogenology64, 363–377.Morton KM, Catt SL, Maxwell WMC, Evans G, 2005b:Effects of lamb age, hormone stimulation and response tohormone stimulation on the yield and in vitro developmentalcompetence of prepubertal lamb oocytes. Reprod FertilDev 17, 593–601.Morton KM, Catt SL, Maxwell WMC, Evans G, 2005c: Anefficient method of ovarian stimulation and in vitro embryoproduction from prepubertal lambs. Reprod Fertil Dev 17,701–706.Morton KM, Catt SL, Maxwell WMC, Evans G, 2005d: Invitro and in vivo developmental capabilities and kinetics ofin vitro development of in vitro matured oocytes from adult,unstimulated and hormone stimulated prepubertal ewes.Theriogenology 64, 1320–1332.Nicholas FW, 1996: Genetic improvement through reproductivetechnology. Anim Reprod Sci 42, 205–214.O’Brien JK, Dwarte D, Ryan JP, Maxwell WMC, Evans G,1996: Developmental capacity, energy metabolism andultrastructure of mature oocytes from prepubertal and adultsheep. Reprod Fertil Dev 8, 1029–1037.O’Brien JK, Beck NFG, Maxwell WMC, Evans G, 1997a:Effect of hormone pre-treatment of prepubertal sheep on theproduction and developmental capacity of oocytes in vitroand in vivo. Reprod Fertil Dev 9, 625–631.O’Brien JK, Catt SL, Ireland KA, Maxwell WMC, Evans G,1997b: In vitro and in vivo developmental capacity ofoocytes from prepubertal and adult sheep. Theriogenology47, 1433–1443.Oropeza A, Wrenzycki C, Herrmann D, Hadeler K-G,Niemann H, 2004: Improvement of the developmentalcapacity of oocytes from prepubertal cattle by intraovarianIGF-I application. Biol Reprod 70, 1634–1643.Presicce GA, Jiang S, Simkin M, Yang X, 1995: Oocyte qualityand embryo development in prepubertal calves. Biol Reprod52(Suppl.), 127.Ptak G, Loi P, Dattena M, Tischner M, Cappai P, 1999:Offspring from one-month-old lambs: Studies on the developmentalcapability of prepubertal oocytes. Biol Reprod 61,1568–1574.Ptak G, Tischner M, Bernabo N, Loi P, 2003: Donordependantdevelopmental competence of oocytes fromlambs subjected to repeated hormonal stimulation. BiolReprod 69, 278–285.Ptak G, Matsukawa K, Palmieri C, Salda LD, Scapolo PA,Loi P, 2006: Developmental and functional evidence ofnuclear immaturity in prepubertal oocytes. Hum Reprod 21,2228–2237.Quirke JF, Hanrahan JP, 1977: Comparison of the survival inthe uteri of adult ewes of cleaved ova from adult ewes andewe lambs. J Reprod Fertil 51, 487–489.Tassell R, Chamley AW, Kennedy JP, 1978: Gonadotrophinlevels and ovarian development in the neonatal ewe lamb.Aust J Biol Sci 31, 267–273.Tervit HR, McMillan WH, McGowan LT, Smith JF, HallDR, Donnison MJ, 1997: Effect of juvenile calf age onfollicular dynamics and in vitro embryo production. Theriogenology47, 300.Trounson AO, Willadsen SM, Moor RM, 1977: Reproductivefunction in prepubertal lambs: ovulation, embryo developmentand ovarian steroidogenesis. J Reprod Fertil 49, 69–75.Van Soom A, Ysebaert MT, de Kruif A, 1997: Relationshipbetween timing of development, morula morphology andcell allocation to inner cell mass and trophectoderm in invitro produced bovine embryos. Mol Reprod Dev 47, 47–56.van der Werf J, 2005: Applying new technologies in sheepbreeding programs. Merinotech Best Practice Sheep BreedingForum Kojonup, Western Australia, 17–23.Worthington CA, Kennedy JP, 1979: Ovarian response toexogenous hormones in six-week-old lambs. Aust J Biol Sci32, 91–95.Wrenzycki C, Herrmann D, Lucas-Hahn A, Korsawe K,Lemme E, Niemann H, 2005: Messenger RNA expressionpatterns in bovine embryos derived from in vitro proceduresand their implications for development. Reprod Fertil Dev17, 23–35.Wright RW, Anderson GB, Cupps PT, Drost M, BradforsGE, 1976: In vitro culture of embryos from adult andprepubertal ewes. J Anim Sci 42, 912–917.Yang X, Presicce GA, Du F, Jiang S, 1997: Pregnancies andcalves derived from pre-pubertal calf oocytes. Theriogenology47, 163.Author’s address (for correspondence): Katherine M. Morton, TheCamel Reproduction Centre, P.O. Box 79914, Dubai, United ArabEmirates. E-mail: kmorton@vetsci.usyd.edu.auConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 144–149 (2008); doi: 10.1111/j.1439-0531.2008.01154.xISSN 0936-6768Updates on Reproductive Physiology, Genital Diseases and Artificial Insemination inthe Domestic CatE Axne´rDivision of Reproduction, Department of Clinical Sciences, Swedish University of Agricultural Sciences, Uppsala, SwedenContentsReproduction in the domestic cat is characterized by largeindividual variations in the female oestrous cycle and in malesemen quality. The female cat reproduction is stronglyinfluenced by season, and new data suggest that it is possiblethat season also affects male fertility. Repeated periods ofoestrus and spontaneous ovulation may lead to degenerativeendometrial changes causing infertility. For artificial insemination(AI), induction of ovulation is necessary in the absenceof the mating stimuli. The large variation in the relationshipbetween follicular growth and timing of expression of oestruscomplicates, however, timing of ovulation induction. Stressmay lead to progesterone secretion by the adrenal glands andpossibly have a negative impact on early pregnancy. The largeindividual variation in semen quality makes fertility evaluationand feline semen conservation a challenge. For AI, the semencan be deposited in the cranial vagina, the uterus or in theuterine tubes. Intrauterine insemination results in higherpregnancy rates than intravaginal but is more complicated.Surgical intrauterine or intratubal insemination has resulted inthe birth of kittens, but is an invasive procedure that is notallowed in all countries. Transcervical intrauterine inseminationwith frozen-thawed semen has, however, recently resultedin the birth of kittens.IntroductionThe domestic cat has been shown to be an interestingmodel for the development of reproductive biotechnologiesin wild felids with the aim to conserve geneticvariation in small populations. There is, however, also apotential for the use or semen preservation and artificialinsemination (AI) in the breeding of domestic cats. Inthe dog, these techniques can be considered routinetoday. In contrast, more research is necessary before AI,and semen conservation can be considered routineprocedures in the domestic cat. To be able to successfullyuse reproductive biotechnologies, a thoroughknowledge about reproductive physiology and pathologyis necessary.Physiology of the Female CatThe oestrous cycleThe female cat is seasonally polyoestrous with changesin daylength (increasing number of light hours) regulatingthe seasonality (Leyva et al. 1989; Tsutsui et al.2004a). Characteristic for reproduction in the domesticcat is the large normal variation in the length of theoestrous cycle, the length of the reproductive season andthe number of cycles ⁄ year between queens (Shille et al.1979; Tsutsui et al. 2004a). The presence of matureoocytes in the ovaries and the effect of ovulationinduction seem to be dependent on the day of the onsetof physiological oestrus during the follicular phase(Banks and Stabenfeldt 1982; Glover et al. 1985).Because there is a variation between queens in whichday during the follicular phase they will start to showoestrous behaviour, the day of oestrous behaviour inrelation to the progress of oocyte maturation is likely todiffer between queens (Shille et al. 1979; Malandainet al. 2002). Most queens will respond with an LHrelease,leading to ovulation of good quality oocytesafter a mating in midoestrus, while mating in earlyoestrus is not always followed by ovulation (Banks andStabenfeldt 1982; Glover et al. 1985). Expression ofFSH- and LH receptors in the ovarian follicles variesdepending on the stage of follicular growth or atresia(Saint-Dizier et al. 2007).Cyclical changes in the female tubular genitalia in relationto hysterographic appearance and cervical patencyThe outer diameter of the uterus and the appearance ofthe endometrial lining will vary depending on the stageof the oestrous cycle. During inactive stages, a singlelayer of cuboidal cells lines the endometrium, andthe endometrial glands are inactive. When the queenenters the follicular stage, the endometrial and myometrialdiameters increase and the endometrial glandsproliferate. The luteal phase endometrium is characterizedby a high number of elongated and active glands(Chatdarong et al. 2005). The shape of the uterine cavitycan be studied with hysterography. In the healthy cat,the endometrial outlining is smooth. During the inactivephase of the cycle, the uterine horns are straight.Because of the endometrial growth and increasedmuscular activity, the uterine horns become curved witha wavy luminal cavity during the follicular phase. Theluminal shape during the luteal phase varies fromstraight to irregular, wavy and coiled (Chatdarong et al.2005). The cervix is normally only patent (physiologicallyopen cervical canal) during oestrus and theduration of cervical patency does not depend onwhether or not the queen ovulates (Chatdarong et al.2002, 2006a). The period of cervical patency usuallycoincides with the period when the vaginal smear iscornified (Chatdarong et al. 2002).Sperm distribution after matingSperm deposition during mating is vaginal. Initial spermtransport in the female reproductive tract is very rapid.Spermatozoa can be found in the uterine tube 30 minafter mating. The uterotubal junction and uterine cryptsserve as initial sperm reservoirs but spermatozoa areÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproductive Physiology, Pathology and AI in Cat 145subsequently redistributed to the isthmus (Chatdaronget al. 2004), and the oocytes are fertilized in the oviductaround the time of ovulation (Swanson et al. 1994). As asingle mating or AI at the time of ovulation inductionmay result in conception, the fertile life of cat spermatozoamust be at least the same as the time betweenmating and ovulation (Chatdarong et al. 2007).Adrenal progesterone secretionProgesterone is produced by the corpora lutea afterovulation but can also be released from the adrenalglands after stress or injection with ACTH (Chatdaronget al. 2006b). A serum progesterone concentration of16.3 nmol ⁄ l has been observed after ACTH injection ofan ovariohysterectomized female, indicating that thefeline adrenals may produce substantial amounts ofprogesterone (Chatdarong et al. 2006b). ACTH administrationto sows has been shown to cause an alteredprogesterone profile, a possible decrease in the time fortransport of oocytes ⁄ embryos through the oviduct andloss of embryos (Brandt et al. 2007). Therefore, onecannot exclude that stress may have negative effects alsoon early pregnancy in cats, a phenomenon that warrantsfurther research in this species.Disorders of the Female Genital OrgansInfertility in the cycling queenInfertility in the queen with normal ovarian activitypresents a diagnostic challenge for the clinician. In thequeen with normal cycles, normal mating and matinginducedovulation, CEH or low-grade endometritis maycause infertility without any other obvious clinical signs.CEH is believed to be the result of cumulative changesin the endometrium induced by oestradiol and progesterone(Perez et al. 1999; Misirlioglu et al. 2006). Noefficient treatment for CEH except ovariohysterectomyhas been reported. CEH can be diagnosed with ultrasoundor hysterography (Chatdarong et al. 2005). Lowgradeendometritis is extremely difficult to diagnose.Uterine biopsy is an invasive procedure and does notguarantee that the whole endometrium is accessed andmay therefore give a false negative result if the pathologicalchanges are focal. Vaginal discharge is anindication of genital pathology, but may not be presentif the cervix is closed, the infection is mild, or in thepresence of a chronic inflammation. Leukocytes in avaginal swab indicate an acute inflammation, butleukocytes may be few or absent even if the queen hasa uterine infection. Vaginal bacterial culture is notdiagnostic because the normal vaginal bacterial floracannot be distinguished from that of the flora associatedwith uterine pathologies, such as endometritis (Stro¨mHolst et al. 2003). If an endometritis is suspected,bacterial culture can, however, be used to select anappropriate antibiotic. The likelihood of taking asample from the vagina that is truly representative ofthe uterine flora (i.e. finding the same bacteria in avaginal swab as those that are present in the uterus), isprobably higher during oestrus when the cervix is open,than at other stages of the cycle, although this has neverbeen confirmed. To decrease the risk of selectingbacterial strains resistant to antibiotics, the decision toput an infertile queen on antibiotics should only bemade after a thorough clinical examination to rule outother causes of infertility. Mild degenerative changes ofthe endometrium may be difficult to detect on ultrasoundbut can be a reason for treatment failure (Axne´ret al., 2008).Reproductive Physiology of the Male CatIn tomcats, spermatogenesis is established at 6–8 months of age, at which time an increase in testicularweight and testosterone production can be observed.The spermatogenic function in young toms is usuallynot comparable to that of mature males until after8 months of age (Tsutsui et al. 2004b; Siemieniuch andWoclawek-Potocka 2007). The daily sperm productionhas been estimated to be 16 · 10 6 spermatozoa ⁄ testis(Franc¸a and Godinho 2003). The effect of season onreproductive function in the male domestic cat isinsufficiently explored and reports are contradictory(Johnstone et al. 1984; Spindler and Wildt 1999). In aretrospective study on sperm morphology in privatelyowned cats, we found that the percentage of normalspermatozoa was higher in ejaculates collected from catsduring the breeding season than during the non-breedingseason, indicating a possible effect of season on malefertility in cats (Axnér and Linde Forsberg 2007).Blottner and Jewgenow (2007) found seasonal variationsin epididymal sperm quality and testosterone productionbetween spring and autumn. These recent findingsindicate that there might be seasonal variation also indomestic cat male fertility.Semen qualitySemen quality displays large individual variations in thecat. In a retrospective study, the median percentage ofmorphologically normal spermatozoa in an unselectedpopulation of cats was 44% (n = 48), (Axnér and LindeForsberg 2007). Of all cats with

146 E Axnérbetween, but also within individual cats, presents achallenge for the clinician wishing to perform a fertilityexamination or to conserve spermatozoa from anyparticular cat.Infertility in the Male CatThe causes of infertility in the male cat can be classifiedas developmental disorders, poor libido, testiculardegeneration, testicular hypoplasia and miscellaneous.These categories are, however, not always clearlydistinguished from each other. Chromosomal abnormalitiesare, for example, developmental disorders thatmay result in testicular hypoplasia, as seen in tortoiseshellmale cats with the Klinefelter’s syndrome (Axnéret al. 1996). Testicular degeneration is usually anacquired non-hereditary condition, and therefore has abetter prognosis than testicular hypoplasia, which is agenetic condition. Infertile cats often have a low spermconcentration and a high percentage of different spermdefects, or are completely azoospermic (Axne´r et al.1996; Axne´r and Linde Forsberg 2007). Both testicularhypoplasia and testicular degeneration will result in adecrease in semen quality but are not always possible todifferentiate between in a clinical situation. To make adefinitive diagnosis of infertility, a semen sample isusually needed. Repeated samples may, however, benecessary as fertility sometimes fluctuates over time(Axnér et al. 1996; Axne´r and Linde Forsberg 2007).Sperm Collection, Sperm Conservation andArtificial InseminationSperm collectionEjaculated spermatozoa can be collected by an artificialvagina or by electroejaculation (Sojka et al. 1970; Platzand Seager 1978; Platz et al. 1978; Howard et al. 1990).Collection by an artificial vagina generally yields ahigher total number of spermatozoa and allows repeatedsemen collections within short intervals. However,because the use of an artificial vagina is not alwayssuccessful, such use requires training and is unlikely tobe convenient in a clinical situation; electroejaculation isusually the method of choice (Platz and Seager 1978).Electroejaculation requires a surgical plane of anaesthesiabecause reactions in unanaesthethized animals indicatethat the procedure is associated with pain (Platzand Seager 1978; Stafford 1995). In wild felids, althoughketamine alone has been used previously, combinedregimens for anaesthesia are now preferred as theyprovide better sedation and analgesia (Swanson et al.2007). In our laboratory, we use a combination ofmedetomidine (80 lg ⁄ kg bw) and ketamine–HCl(5 mg ⁄ kg bw) (Axne´r et al. 1998). Propofol at a doseof 10 mg ⁄ kg bw is a suitable alternative for anaesthesiafor electroejaculation in domestic cats (Chatdaronget al. 2006c). Medetomidine, an a-adrenergic agent,was shown to increase sperm outflow when used at adose of 130–140 lg ⁄ kg bw (Zambelli et al. 2007). Thenumber of spermatozoa in an ejaculate collected byelectroejaculation, rarely exceeds the number of spermatozoathat are needed for one insemination only withthe techniques and knowledge available today (Sojkaet al. 1970; Platz et al. 1978; Howard et al. 1992;Tanaka et al. 2000; Tsutsui 2006; Chatdarong et al.2007). Refinement of methods for sperm cryopreservation,ovulation induction and AI allowing lower spermnumbers to be used for AI would increase the usefulnessof reproductive biotechnologies in cat breeding.As the domestic cat is routinely subjected to castration,studies on domestic cat epididymal spermatozoaare useful as a model for epididymal sperm preservationfrom wild felids threatened with extinction. Epididymalsperm preservation has, however, also a potential to beuseful in breeding programmes for domestic cats (Tsutsuiet al. 2003; Hermansson and Axne´r 2007). It has notbeen fully elucidated in the cat whether epididymalspermatozoa are comparable to ejaculated spermatozoain their function and reaction to sperm conservationprotocols. In other mammalian species, seminal plasmacontains, for example, decapacitation factors, which canbind to the spermatozoa and inhibit or reverse capacitation,but also factors that enhance capacitation(Oliphant et al. 1985; Boue´ et al. 1996; Therien et al.1998).It has been shown in several species that ejaculatedspermatozoa are more susceptible to cold shock thanepididymal spermatozoa (Gilmore et al. 1998). Spermmotility and viability of frozen-thawed epididymalcat spermatozoa was, however, similar to that ofejaculated spermatozoa (Tsutsui et al. 2003). Althoughejaculated spermatozoa had higher initial motility andplasma membrane integrity, there was no effect of spermsource (ejaculated or epididymal) in the sensitivity tocold shock at the cooling rates usually used in spermpreservation protocols (Hermansson and Axne´r 2007).The first birth of kittens after insemination with frozenthawedepididymal cat spermatozoa was reported byTsutsui et al. (2003).Semen conservationSemen can be used fresh, chilled or frozen-thawed forAI. Usually, the use of conserved semen is of moreinterest than the use of fresh semen as it allows forstorage for extended time periods. Each step in semenpreservation protocols may cause damage to the spermatozoa,thereby reducing their fertilizing potential andprobably also their longevity. Therefore, a highernumber of spermatozoa is needed for insemination withfrozen-thawed semen than with fresh spermatozoa(Tsutsui 2006). Cat spermatozoa do not seem to besusceptible to cold shock at the cooling rates usuallyapplied for sperm conservation (Hermansson and Axne´r2007) although there are conflicting reports regardingthe effect of cooling on acrosomal integrity (Pukazhenthiet al. 1999). Freezing, on the contrary, induces moresevere sperm damage (Axnér et al. 2004; Luvoni 2006).Different cryopreservation protocols have been published.In our laboratory, we use a protocol that hasbeen modified from a protocol that was originallydeveloped for preservation of canine spermatozoa (Rotaet al. 1997; Axne´r et al. 2004). Transcervical AI withejaculated spermatozoa preserved with this protocol hasresulted in the birth of live kittens demonstrating thefertilizing capacity of spermatozoa cryopreserved withÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproductive Physiology, Pathology and AI in Cat 147this method (Chatdarong et al. 2007). Addition ofEquex STM (Nova Chemical Sales, Scituate Inc., MA,USA) paste to a Tris–egg yolk extender reduces acrosomedamage after freezing-thawing, but may have atoxic effect during incubation, leading to a fasterdecrease in sperm viability (Axnér et al. 2004). Washingfrozen-thawed spermatozoa by centrifugation to removethe freezing extender including the Equex STM pastehad, however, no beneficial effect on sperm longevity ina preliminary study (Axne´r et al. 2006). Addition ofantioxidants to the freezing buffer has been shown toimprove post-thaw cat epididymal sperm quality(Luvoni et al. 2002; Thuwanut et al., 2008).Ovulation induction and timing of inseminationIn the absence of the mating stimuli, ovulation must beartificially induced before AI can be carried out. Thelarge variation in the female oestrous cycle complicatesdetermination of the optimal time for ovulation induction.Ultrasound evaluation may be a reliable method tofollow follicular maturation (Malandain et al. 2002).The occurrence of spontaneous ovulation may contributeto difficulties in timing the AI (Lawler et al. 1993;Chatdarong et al. 2007). In order to control the oestrouscycle, studies on different hormonal therapies have beenperformed (Pelican et al. 2006). Pharmacological inductionof oestrus and ovulation cause, however, abnormalhormonal profiles that may interfere with fertility(Pelican et al. 2006). Ovulation induction is usuallycarried out with one or two injections of hCG inmidoestrus (Pelican et al. 2006; Tsutsui 2006; Chatdaronget al. 2007). For induction of oestrus, a combinationof eCG with an injection of hCG 80–84 h later isoften used (Pelican et al. 2006). Both hCG and eCGstimulate the production of antibodies, hence, frequentadministration can lead to decreased treatment success(Pelican et al. 2006). The optimal protocol for ovulationinduction and timing of ovulation induction remains tobe elucidated.Spermatozoa should be inseminated within 49 hafter ovulation induction as no fertilizations have beenobserved to occur after this time (Sojka et al. 1970;Howard et al. 1992). Feline spermatozoa probably donot require a long time for capacitation. Niwa et al.(1985) reported that 100% of homologous oocyteswere penetrated by epididymal spermatozoa within30 min after insemination, while ejaculated cat spermatozoaseem to need 2.5–3 h for in vitro capacitation(Long et al. 1996). Considering the reduced longevityof frozen-thawed spermatozoa, they should probablybe inseminated close to ovulation. The optimal timefor AI in relation to ovulation has, however, not beenfully elucidated although some studies report AI atdifferent times with some conflicting results concerningthe effect of pre-ovulatory anaesthesia on ovulation(Howard et al. 1992; Tsutsui 2006; Chatdarong et al.2007).Techniques of semen depositionPregnancies have been achieved with intravaginal as wellas with surgical or laparoscopic intrauterine or intratubalsemen deposition (Sojka et al. 1970; Platz et al.1978; Howard et al. 1992; Tsutsui 2006; Tsutsui et al.2001). In most species as well as in the cat, intrauterinesemen deposition yields better pregnancy results thanvaginal deposition although the cat spermatozoa aredeposited in the vagina during natural coitus (Linde-Forsberg et al. 1999; Tanaka et al. 2000; Thomassenet al. 2006; Tsutsui 2006; Chatdarong et al. 2007).Surgical insemination is an invasive procedure that isnot allowed in all countries. Therefore, studies ontranscervical AI have been performed. Transcervicalcatheterization in the cat is, however, associated withdifficulties because of the vaginal anatomy. Abdominalfixation of the cervix as carried out with the Norwegiantranscervical catheter for the bitch is not possible in thequeen because of the position of the cervix and thenarrow cranial vagina makes endoscopy-guided transcervicalpassage impossible (Linde-Forsberg et al. 1999;Chatdarong et al. 2002; Zambelli et al. 2004; Thomassenet al. 2006). Zambelli and Castagnetti (2001)performed rectally guided transcervical catheterizationusing a rounded tip needle inserted at the cut end of a 3-Fr tomcat catheter.Birth of live kittens has resulted after transcervicalintrauterine deposition of frozen-thawed semen with thehelp of a catheter modified from a transcervical catheterfirst described by Swanson and Godke (1994) (Chatdaronget al. 2005, 2007). The result with birth of livekittens is promising for the use of non-surgical AI insmall felids, although the protocol still needs to berefined to improve pregnancy results and to allow theuse of a lower sperm number (Chatdarong et al. 2007).In conclusion, although more studies on evaluation offollicular maturation, protocols for artificial ovulationinduction, semen conservation and techniques of AI areneeded before an optimal protocol for routine AI in thecat can be developed, some promising progress has beenachieved in the development of semen conservation andAI in the cat.ReferencesAxne´r E, Linde Forsberg C, 2007: Sperm morphology in thedomestic cat, and its relation with fertility: a retrospectivestudy. Reprod Domest Anim 42, 282–291.Axne´r E, Stro¨m B, Linde-Forsberg C, Gustavsson I, LindbladK, Wallgren M, 1996: Reproductive disorders in 10 domesticmale cats. J Small Anim Pract 37, 394–401.Axne´r E, Stro¨m Holst B, Linde Forsberg C, 1998: Morphologyof spermatozoa in the cauda epididymidis beforeand after electroejaculation and a comparison with ejaculatedspermatozoa in the domestic cat. Theriogenology50, 973–979.Axne´r E, Hermansson U, Linde-Forsberg C, 2004: The effectof Equex STM paste and sperm morphology on post-thawsurvival of cat epididymal spermatozoa. Anim Reprod Sci84, 179–191.Axne´r E, Kronsell A, Linde-Forsberg C, 2006: Effect of postthawcentrifugation on viability of epididymal cat spermatozoa.in: Luvoni GC, Thuro´czy J (eds), Proc. 5th BiannualEVSSAR Conf. Budapest, April, P06.Axne´r E,A˚ gren E, Ba˚ verud V, Stro¨m Holst B. Infertility in thecycling queen: seven cases. Journal of feline medicine andsurgery, 2008: DOI: 10.1016/j.jfms.2008.04.005.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

148 E AxnérBanks DH, Stabenfeldt G, 1982: Luteinizing hormone releasein the cat in response to coitus on consecutive days of estrus.Biol Reprod 26, 603–611.Blottner S, Jewgenow K, 2007: Moderate seasonality in testisfunction of domestic cat. Reprod Domest Anim 42, 536–540.Boue´ F, Blais J, Sullivan R, 1996: Surface localisation ofP34H, an epididymal protein, during maturation, capacitation,and acrosome reaction of human spermatozoa. BiolReprod 54, 1009–1017.Brandt Y, Madej A, Rodrı´guez-Martı´nez H, Einarsson S,2007: Effects of exogenous ACTH during oestrus on earlyembryo development and oviductal transport in the sow.Reprod Domest Anim 42, 118–125.Chatdarong K, Rungsipipat A, Axne´r E, Linde-Forsberg C,2002: Investigation of cervical patency and uterine appearancein domestic cats by fluoroscopy and scintigraphy.Reprod Domest Anim 37, 275–282.Chatdarong K, Lohachit C, Linde Forsberg C, 2004: Distributionof spermatozoa in the female reproductive tract ofthe domestic cat in relation to ovulation induced by naturalmating. Theriogenology 62, 1027–1041.Chatdarong K, Rungsipipat A, Axne´r E, Linde Forsber C,2005: Hysterographic appearance and uterine histology atdifferent stages of the reproductive cycle and after progestagentreatment in the domestic cat. Theriogenology 64, 12–29.Chatdarong K, Lohachit C, Kiartmanakul S, Axne´r E, LindeForsberg C, 2006a: Cervical patency during non-ovulatoryand ovulatory estrus cycles in domestic cats. Theriogenology66, 804–810.Chatdarong K, Ponglowhapan S, Karlsson A, Linde-ForsbergC, 2006b: Exogenous ACTH induced progesterone levels infemale domestic cats. Theriogenology 66, 1482–1487.Chatdarong K, Ponglowhapan S, Manee-in S, Pongphet K,2006c: The use of propofol for electroejaculation in domesticcats. Theriogenology 66, 615–617.Chatdarong K, Axne´r E, Manee-In S, Thuwanut P, Linde-Forsberg C, 2007: Pregnancy in the domestic cat aftervaginal or transcervical insemination with fresh and frozensemen. Theriogenology 68, 1326–1333.França LR, Godinho CL, 2003: Testis morphometry, seminiferousepithelium cycle length, and daily sperm productionin domestic cats (Felis catus). Biol Reprod 68, 1554–1561.Gilmore JA, McGann LE, Ashworth E, Acker JP, Raath JP,Bush M, Critser JK, 1998: Fundamental cryobiology ofselected African mammalian spermatozoa and its role inbiodiversity preservation through the development ofgenome resource banking. Anim Reprod Sci 53, 277–298.Glover TE, Watson PF, Bonney RC, 1985: Observations onvariability in LH release and fertility during oestrus in thedomestic cat (Felis catus). J Reprod Fertil 75, 145–152.Hermansson U, Axne´r E, 2007: Epididymal and ejaculated catspermatozoa are resistant to cold shock but egg yolkpromotes sperm longevity during cold storage at 4°C.Theriogenology 67, 1239–1248.Howard JG, Brown JL, Bush M, Wildt DE, 1990: Teratospermicand normospermic domestic cats: ejaculate traits,pituitary-gonadal hormones, and improvement of spermatozoalmotility and morphology after swim-up processing. JAndrol 11, 204–215.Howard JG, Barone MA, Donoghue AM, Wildt DE, 1992:The effect of pre-ovulatory anesthesia on ovulation inlaparoscopically inseminated domestic cat. J Reprod Fertil96, 175–186.Johnstone IP, Bancroft BJ, McFarlane JR, 1984: Testosteroneand androstenedione profiles in the blood of domestic tomcats. Anim Reprod Sci 7, 363–375.Lawler DF, Johnston SD, Hegstad RL, Keltner DG, OwensSF, 1993: Ovulation without cervical stimulation in domesticcats. J Reprod Fertil Suppl 47, 57–61.Leyva H, Madley T, Stabenfeldt GH, 1989: Effect of lightmanipulation on ovarian activity and melatonin andprolactin secretion in the domestic cat. J Reprod FertilSuppl 39, 125–133.Linde-Forsberg C, Stro¨m Holst B, Govette G, 1999: Comparisonof fertility data from vaginal vs intrauterine inseminationof frozen-thawed dog semen: a retrospective study.Theriogenology 52, 11–23.Long JA, Wild DE, Wolfe BA, Critser JK, DeRossi RV,Howard JG, 1996: Sperm capacitation and the acrosomereaction are compromised in teratospermic domestic cats.Biol Reprod 54, 638–646.Luvoni GC, 2006: Gamete cryopreservation in the domesticcat. Theriogenology 66, 101–111.Luvoni GC, Guggiero C, Marinoni C, Kalchsmidt E, 2002:Effect of taurine-containing diluent for cryopreservation ofdomestic cat epididymal spermatozoa. Theriogenology 57,466.Malandain E, Rault D, Froment E, Ficheux C, Pichard JP,Fontbonne A, Begon D, Chastan-Maillard S, 2002: Followupof follicular growth during non ovulatory estral phases inqueen by ultrasonography and vaginal cytology. in: VerstegenJ, Onclin K, Linde Forsberg C (eds), Proc. 3rd EVSSARCongr. Lie` ge, 10–12 May, p. 141.Misirlioglu D, Nak D, Sevimli A, Nak Y, Ozyigit O, Akkoc A,Cangul IT, 2006: Steroid receptor expression and HER-2 ⁄ neu (c-erB-2) oncoprotein in the uterus of cats with cysticendometrial hyperplasia-pyometra complex. J Vet Med A53, 225–229.Niwa K, Oˆ hara K, Hosoi Y, Iritani A, 1985: Early events of invitrofertilization of cat eggs by epididymal spermatozoa. JReprod Fertil 74, 657–660.Oliphant G, Reynold AB, Thomas TS, 1985: Sperm surfacecomponents involved in the control of the acrosomereaction. Am J Anat 174, 269–283.Pelican KM, Wildt DE, Pukazhenthi B, Howard JG, 2006:Ovarian control for assisted reproduction in the domesticcat and wild felids. Theriogenology 66, 37–48.Perez JF, Conley AJ, Dieter JA, Sanz-Ortega J, Lasley BL,1999: Studies on the origin of ovarian interstitial tissue andthe incidence of endometrial hyperplasia in domestic andferal cats. Gen Comp Endocrinol 116, 10–20.Platz CC, Seager SWJ, 1978: Semen collection by electroejaculationin the domestic cat. J Am Vet Med Assoc 173,1353–1355.Platz CC, Wildt DE, Seager SW, 1978: Pregnancy in thedomestic cat after artificial insemination with previouslyfrozen spermatozoa. J Reprod Fertil 52, 279–282.Pukazhenthi B, Pelican K, Wildt DE, Howard JG, 1999:Sensitivity of domestic cat (Felis catus) sperm from normospermicversus teratospermic donors to cold-induced acrosomaldamage. Biol Reprod 61, 135–141.Rota A, Stro¨m B, Linde-Forsberg C, Rodriguez-Martinez H,1997: Effects of equex STM paste on viability of frozenthawed dog spermatozoa during in vitro incubation at 38°C.Theriogenology 47, 1093–1101.Saint-Dizier M, Malandain E, Thoumire S, Remy B, Chastan-Maillard S, 2007: Expression of follicle stimulating hormoneand luteinising hormone receptors during follicular growthin the domestic cat ovary. Mol Reprod Dev 74, 989–996.Shille VM, Lundstro¨m KE, Stabenfeldt GH, 1979: Follicularfunction in the domestic cat as determined by estradiol-17bconcentrations in plasma: relation to estrous behaviour andcornification of exfoliated vaginal epithelium. Biol Reprod21, 953–963.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproductive Physiology, Pathology and AI in Cat 149Siemieniuch MJ, Woclawek-Potocka I, 2007: Morphologicalfeatures of the seminiferous epithelium in cat (Felis catus, L.1758) testes. J Reprod Dev 53, 1125–1130.Sojka NJ, Jennings LL, Hamner CE, 1970: Artificial inseminationin the cat (Felis catus L). Lab Anim Care 20, 198–204.Spindler RE, Wildt DE, 1999: Circannual variations inintraovarian oocyte but not epididymal sperm quality inthe domestic cat. Biol Reprod 61, 188–194.Stafford KJ, 1995: Electroejaculation: a welfare issue. Surveillance22, 14–17.Stro¨m Holst B, Bergstro¨m A, Lagerstedt AS, Karlstam E,Englund L, Ba˚ verud V, 2003: Characterization of thebacterial population of the genital tract of adult cats. AmJ Vet Res 64, 963–968.Swanson WF, Godke RA, 1994: Transcervical embryo transferin the domestic cat. Lab Anim Sci 44, 228–291.Swanson WF, Roth TL, Wildt DE, 1994: In vivo embryogenesis,embryo migration, and embryonic mortality in thedomestic cat. Biol Reprod 51, 452–464.Swanson WF, Magarey GM, Herrick JR, 2007: Spermcryopreservation in endangered felids: developing linkageof in situ–ex situ populations. Soc Reprod Fertil Suppl 65,417–432.Tanaka A, Takagi Y, Nakagawa K, Fujimoto Y, Hori T,Tsutsui T, 2000: Artificial intravaginal insemination usingfresh semen in cats. J Vet Med Sci 62, 1163–1167.Therien I, Soubeyrand S, Manjunath P, 1998: Major proteinsof bovine seminal plasma and high-density lipoproteininduce cholesterol efflux from epdidiymal sperm. BiolReprod 59, 768–776.Thomassen R, Sanson G, Krogenaes A, Fougner JA, BergKA, Farstad W, 2006: Artificial insemination with frozensemen in dogs: a retrospective study of 10 years using a nonsurgicalapproach. Theriogenology 66, 1645–1650.Thuwanut P, Chatdarong K, Techakumphu M, Axne´r E,2008: The effect of antioxidants on motility, viability,acrosome integrity and DNA integrity of frozen-thawedepididymal cat spermatozoa. Theriogenology, doi: 10.1016/j.theriogenology.2008.04.005.Tsutsui T, 2006: Artificial insemination in domestic cats (Feliscatus). Theriogenology 66, 122–125.Tsutsui T, Tanaka A, Hori T, 2001: Intratubal inseminationwith fresh semen in cats. J Reprod Fertil Suppl 57, 347–351.Tsutsui T, Wada M, Anzai M, Hori T, 2003: Artificialinsemination with frozen-thawed epididymal sperm in cats. JVet Med Sci 65, 397–399.Tsutsui T, Nakagawa K, Hirano T, Nagakubo K, ShinomiyaM, Yamamoto K, Hori T, 2004a: Breeding season in femalecats acclimated under a natural photoperiod and intervaluntil puberty. J Vet Med Sci 66, 1129–1132.Tsutsui T, Kuwabara S, Kuwabara K, Kugota Y, Kinjo T,Hori T, 2004b: Development of spermatogenic function inthe sex maturation process in male cats. J Vet Med Sci 66,1125–1127.Zambelli D, Castagnetti C, 2001: Transcervical inseminationwith fresh or frozen semen in the domestic cat: newtechnique and preliminary results. In: Proc. 5th AnnualConf. ESDAR. Vienna, September, p. 34.Zambelli D, Buccioli M, Castagnetti C, Belluzzi S, 2004:Vaginal cervical anatomic modification during the oestruscycle in relation to transcervical catheterization in thedomestic cat. Reprod Domest Anim 39, 76–80.Zambelli D, Cunto M, Prati F, Merlo B, 2007: Effects ofketamine or medetomidine administration on quality ofelectroejaculated sperm and on sperm flow in the domesticcat. Theriogenology 68, 796–803.Author’s address (for correspondence): Eva Axne´r, Division ofReproduction, Department of Clinical Sciences, Swedish Universityof Agricultural Sciences, Uppsala, Sweden. E-mail: eva.axner@og.slu.seConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 150–156 (2008); doi: 10.1111/j.1439-0531.2008.01155.xISSN 0936-6768Reproduction of the Female Ferret (Mustela putorius furo)H LindebergUniversity of Kuopio, Department of Biosciences, Kuopio, FinlandContentsThe domestic ferret is a seasonally polyoestrous species.Females reach puberty at the age of 8–12 months. Femalesexhibit a constant oestrus between late March and earlyAugust if they are not bred. Increasing tumescence in thepink-coloured vulva is a sign of pro-oestrus. Oestrus canpersist for up to 5 months, but once ovulation is induced,either pregnancy or pseudopregnancy ensues. Folliculardevelopment and atresia overlap in such a manner that thereis a recent cohort of follicles available for ovulation whenevercopulation might occur. Copulation may last from 15 min to3 h, the average being 1 h. Ovulation is induced by pressureon the cervix associated with copulation. After sufficient LHrelease, the pre-ovulatory follicles mature and an average of12 oocytes (5–13) per female are ovulated 30–40 h aftercopulation into the ovarian bursa. The ferret oocytes aremost capable of being fertilized up to 12 h after ovulation,i.e. 42–52 h after copulation. Ferret oocytes ovulate at themetaphase of the second meiotic division (MII) embedded inthree layers of corona radiata cells. Embryos enter the uterusover a period of several days starting on day 5 after mating.Between days 12 and 13 after mating, the embryos havebecome implanted in the endometrium. Implantation in theferret is central with rapid invasion of the uterine epitheliumby the trophoblast over a broad area that eventually becomesa zonary band of endotheliochorial placenta. Gestationlength is 41 days (39–42 days). The domestic ferret givesbirth to an average of eight kits (1–18 kits), which weigh6–12 g at birth.IntroductionMustelids are polytocous carnivorous species demonstratinga variety of unique reproductive characteristics.The reproductive physiology of some species has beenquite thoroughly studied, while less is known aboutother species. This literature review was partly publishedas the PhD thesis of the author (Lindeberg 2003), and itconcentrates on the results obtained with the domesticferret. For comparative purposes, when necessary,results concerning other mustelids and carnivores arealso included.Photoperiods and seasonalityThe domestic ferret is considered to be a seasonallypolyoestrous species, but females exhibit a constantoestrus between late March and early August if they arenot bred (Mead 1989). Marshall (1904) classified the ferretas a mono-oestrous species. Offspring born the previoussummer reach puberty by the following spring at the ageof 8–12 months (Fox and Bell 1998). Initiation of thegonadal activity is totally dependent on the light-darkcycle, which stimulates or inhibits reproduction throughtransmission of information about the day length to thebrain (Bissonnette 1932; Turek and Van Cauter 1988).The ferret’s gonadal response to a given photoperioddepends both on the duration of the photoperiod andon the previous photoperiodic experience. Ferrets, likeother ‘long-day’ seasonal breeders, need alternatingperiods of increasing duration of daylight (long days),at which time they are sensitive to light, and days ofdecreasing duration of daylight (short days), duringwhich time they are insensitive to light, so that theannual cycle recurs normally (Herbert 1989). The stateof temporary unresponsiveness to photic stimuli isreferred to as the ‘photorefractory’ condition (Elliottand Goldman 1981), and it is required to induce areturn to the photosensitive state (Forsberg 1992).Ferrets exposed to artificially changing long and shortdaylight conditions (long days: 16 h light and 8 hdarkness; short days: 8 h light and 16 h darkness) startshowing oestrus approximately 3 weeks after the changefrom short days to long days (Fox and Bell 1998), andcease showing oestrus about the same time after achange from long days to short days. The repeatedchange from long days to short days after every6 months causes one period of gonadal activity peryear similar to a natural breeding season when theanimals are exposed to natural outdoor light conditions.A change in light conditions every 4 or 2 months causestwo or three periods of gonadal activity a year,respectively (Herbert 1989). Through the use of reverselight cycles, ferrets can be induced to breed any timeduring the year (Harvey and MacFarlane 1958).Oestrous cycle and vaginal cytologyThe increasing tumescence in the pink-coloured vulva isa sign of pro-oestrus in the ferret. The vulva enlarges upto 50 times its normal size during a 2- to 3-week periodat the beginning of the breeding season (Hammond andMarshall 1930), which extends from March to August(Marshall 1904). No change in the turgidity of the vulvaoccurs up to 36 h after copulation (Hammond andWalton 1934), but 3 or 4 days after mating, the vulvastarts regressing and regains its normal size in2–3 weeks. If the vulva does not recede, ovulation hasprobably not taken place, and the female may need to beremated (Lagerqvist 1992). Oestrus can persist for up to5 months, but once ovulation is induced, either pregnancyor pseudopregnancy ensues (Hammond andMarshall 1930).In conjunction with the vulval swelling during oestrus,there is a thickening of the uterine endometrium, andfollicles develop in the ovaries. The pattern of folliculardevelopment during a prolonged oestrus is unknown,but it is assumed to be continuous (Robinson 1918). Inthe absence of copulation, which results in a prolongedÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproduction of Domestic Ferret 151oestrus, a cohort of follicles develops, degenerates and isreplaced by a new cohort of follicles. Folliculardevelopment and atresia overlap in such a manner thatthere is a recent cohort of follicles available forovulation whenever copulation might occur (Murphy1989). After copulation, some pre-ovulatory folliclesmay persist in the ovaries and luteinise, but they do notrupture (Robinson 1918). The interval required fordevelopment from primordial to pre-ovulatory folliclehas not been reported for any mustelid. It has beenproposed that the time required for a wave of follicles todevelop prior to induced ovulation in ferrets isapproximately 6 days (Murphy 1989).Oestradiol secreted by the follicles controls vulvalswelling, uterine development, sexual receptivity andchanges in vaginal cells (Hamilton and Gould 1940).Ferrets with vulval diameter greater than 1 cm havemated successfully (Murphy 1989). Vaginal cytology iscommonly used to detect oestrus in some domesticcarnivores, especially dogs (Olson et al. 1984), and thetechnique has been described for domestic ferrets(Hamilton and Gould 1940; Williams et al. 1992), mink(Mustela vison) (Hansson 1947) and silver foxes (Vulpesvulpes) (Jalkanen et al. 1988). In ferrets, pro-oestrus ischaracterized by an increasing percentage of superficialepithelial cells together with enlargement of the vulva.During oestrus, usually >90% of epithelial cells aresuperficial cells, and after several days these cells arefully keratinized. Neutrophils are common during allstages of the oestrous cycle (Williams et al. 1992).Following copulation, which may last for 15 min up to3 h (Hammond and Walton 1934), the average being 1 h(Fox and Bell 1998), the percentage of superficial cells inthe vagina declines with vulval swelling. During mating,the male grips the female’s neck and makes repeatedpelvic thrusts until intromission is achieved (Carrollet al. 1985).OvulationOvulation is non-spontaneous, i.e. it is induced bypressure on the cervix associated with natural copulation,leading to endogenous LH release (Carroll et al.1985), or may be induced artificially by hCG treatment(Mead et al. 1988a). The increase in LH concentrationafter copulation is modest (only 3- to 4-fold elevation),and during oestrus, LH does not exhibit the usualmammalian pattern of a pulsatile release (Carroll et al.1985; Tritt 1986). After sufficient LH release, the preovulatoryfollicles mature and an average of 12 oocytes(5–13) per female are ovulated into the ovarian bursa30–40 h after copulation (Hammond and Walton 1934;Chang and Yanagimachi 1963; Carroll et al. 1985).The ferret ovary is encapsulated in a fatty bursacompletely enclosing the ovary, so that the ovulatedoocytes cannot be shed into the abdominal cavity(Marshall 1904). The oocytes remain capable of beingfertilized for a period of 30–36 h after ovulation(Hammond and Walton 1934; Chang and Yanagimachi1963), but mating during a period of 18–30 h afterovulation produce small litters (1–3 kits) (Hammondand Walton 1934). If the number of penetrated oocytesis considered as evidence of fertilization, the ferretoocytes are most capable of being fertilized up to 12 hafter ovulation, i.e. 42–52 h after copulation (Changand Yanagimachi 1963). Spermatozoa are found in theovarian bursa 6 hours after copulation in the ferret(Hammond and Walton 1934) or one hour aftercopulation in the stoat (Mustela erminea) (Amstislavsky,personal communication), but they require 3.5–11.5 h of capacitation time in the female reproductivetract to become fertile (Chang and Yanagimachi 1963).The spermatozoa apparently retain their capability toinduce fertilization for up to 126 h after copulation(Chang 1965b), but the rate of fertilization depends onthe short lifespan of the oocytes. The exact site wherefertilization takes place is not known (Murphy 1989),although the middle-third of the oviduct (Robinson1918) has been suggested. According to Chang andYanagimachi (1963), ferrets may sometimes ovulatespontaneously when they are handled. Generally,however, pre-intromission events do not lead to anyincrease in LH, and thus do not induce ovulation(Carroll et al. 1985).Oocyte maturation and fertilizationAfter copulation, a nuclear maturation process isinitiated, and the formation of the first polar body andthe appearance of the second maturation spindle occur(Hamilton 1934). Without coital stimuli, ferret oocytesremain immature in the germinal vesicle stage until theydegenerate. Ferret oocytes ovulate at the metaphase ofthe second meiotic division (MII) (Mainland 1931; Pilttiet al. 2003) embedded in three layers of corona radiatacells (Chang 1950), which detach after fertilization. Apositive correlation between the naked appearance offerret oocytes and fertilization has been reported (Chang1965b). Parthenogenetic cleavage of oocytes, i.e. theformation and division of blastomeres in the absence offertilization, is common up to 2–6 cells, and may occurin 43–60% of the unfertilized oocytes, most likely as aresult of oocyte ageing (Chang 1950, 1957). Themajority of unfertilized oocytes are embedded in thecorona radiata even 4–5 days after sterile mating(Chang and Yanagimachi 1963). If the oocyte isfertilized, the entire spermatozoon (head together withtail) passes through the zona pellucida (Hamilton 1934).The first pronuclei are discovered 40 ¾–43½ h aftercopulation (Mainland 1930; Hamilton 1934; Chang andYanagimachi 1963; Piltti et al. 2003). The second polarbody is extruded soon after oocyte penetration by aspermatozoon, as in most other mammals (Hamilton1934, Farstad et al. 2001). In the ferret, the morphologicaldifferences between the first and the second polarbody are not distinctive (Chang 1965a). The number ofpolar bodies that may change their location in thedistinct perivitelline space (Hamilton 1934) varies fromtwo to four (Mainland 1931; Chang and Yanagimachi1963), indicating that in both fertilized and unfertilizedoocytes the first polar body may either frequently divide(Chang and Yanagimachi 1963) or become fragmented.The thickness of the zona pellucida varies from 2 to20 lm (Mainland 1932; Chang 1950). The size of anunfertilized oocyte without a corona radiata has beenreported to be between 140 and 160 lm (Hamilton 1934;Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

152 H LindebergChang 1950). The oocyte is spherical or ovoid; itsnucleus is large, round and eccentric with one or, rarely,two nucleoli (Robinson 1918). It is rich in lipid particles,which in one-half of the oocyte are less numerous thanin the other half, making the oocyte look less dense inthis half (Hamilton 1934) and making it easier tovisualize the nuclear structures on the less dense half ofthe oocyte. This seems to be particular to the ferretoocyte compared with other lipid-rich oocytes observedin other carnivorous species, such as the canine andfeline oocytes.Early embryonic developmentThe first cleavage produces two blastomeres that aresimilar in appearance but unequal in size. The first twoblastomeres divide asynchronously, so that furtherstages of uneven numbers of blastomeres are produced(Hamilton 1934). Asynchronous divisions are a commonfeature for carnivore embryos (Amstislavsky et al. 1993;Lindeberg et al. 1993) but not for embryos of ruminants(Betteridge and Fle´ chon 1988). A rough estimation ofthe intervals between the first and second, and thesecond and third cleavages was 10–16 h for each stage(Chang 1965b). Asynchronous division of the cells anddifferences in sizes are detected up to the 16-cell stage,but no morphological differences can be detectedbetween the central (inner cell mass) cells and thesurrounding (trophoectoderm) cells at the morula stage,in which the cells are grouped closely together and adistinct perivitelline space is present. Fat is equallydistributed among the cells (Hamilton 1934).The degree of development of embryos of the sameage varies considerably (Rider and Heap 1986). At thesame time after mating, embryos from one ferret can beat the blastocyst stage, whereas those of another ferretmay still be at the morula stage (Chang 1969).Blastocysts may be further developed than those inwhich a longer time interval has elapsed after mating.Blastocysts appear to be almost spherical and completelyfill the zona pellucida, which has become thinned bythe presence of the expanding blastocyst. The inner cellmass appears as a dark mass at one pole of theblastocyst. The flattened cells of the trophoectodermand the inner cell mass are dotted with fatty granules(Robinson 1925; Hamilton 1934). Uterine blastocystsexpand from a size of 200 lm in diameter to more than2 mm in diameter during their pre-implantation development(Daniel 1970; Enders and Schlafke 1972).Progesterone has been reported to support blastocystexpansion up to a diameter of 1 mm, but for furtherblastocyst expansion additional ovarian factors arerequired (McRae 1992). Without the presence ofmaternal progesterone, neither cleavage of embryos inthe oviducts or in the uterus (Rider and Heap 1986) norexpansion of blastocysts in the uterus take place(Buchanan 1969; McRae 1992).Oviductal passageFerret embryos enter the uterus over a period of severaldays starting from day 5 after mating. The embryosstart entering the uterus on day 4, if females are treatedwith hCG (Chang 1969), and almost all embryos arefound in the uterus by day 7 as described in studies ofRobinson (1918), Hammond and Walton (1934), whoanalyzed Robinson’s (1918) data and Hamilton’s(1934). Lindeberg and Ja¨ rvinen (2003) reported thatuntil day 5 after the first mating, embryos of farmedEuropean polecats (Mustela putorius) were recoveredonly from the oviducts. Between days 6 and 9, embryoswere recovered both from the oviducts and the uteri.From day 10 onwards, the embryos were recoveredonly from the uteri. Ferret pre-implantation embryos(Fig. 1a and b) experience a prolonged period ofoviductal residence – a phenomenon that has also beendemonstrated in the cat (Swanson et al. 1994), the bluefox (Valtonen et al. 1985; Valtonen and Jalkanen 1993),the silver fox (Jalkanen 1992), the dog (Holst andPhemister 1971; Renton et al. 1991) and the horse(Betteridge et al. 1982). After unilateral or bilateralembryo transfer, ferret pre-implantation embryos havebeen reported to migrate from one horn to another (Liet al. 2006).ImplantationThree types of pregnancy have been identified in theMustelidae (Mead 1989; Ternovsky and Ternovskaya1994). All polecat species (Mustela putorius, Mustelaputorius furo, Mustela eversmanni, Mustela lutreola)have a short period of pregnancy of constant duration(37–44 days). Species such as the stoat (Mustelaerminea) and the sable (Martes zibellina) exhibit anobligatory embryonic diapause at the blastocyst stageand a long gestation period (7–10 months) (Amstislavskyand Ternovskaya 2000). In the American mink(Mustela vison), the gestation period is short butvariable (range 45–61 days), and may or may notinclude implantation delay (Mead 1989).Implantation in the ferret is central with rapidinvasion of the uterine epithelium by the trophoblast(a)(b)Fig. 1. (a) Freshly flushed day-9blastocysts; (b) Frozen-thawedday-9 blastocysts in the farmedEuropean polecatÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproduction of Domestic Ferret 153over a broad area that eventually becomes a zonaryband of endotheliochorial placenta (Strahl and Ballman1915; cited in Enders and Schlafke 1972). The endotheliochorialplacenta has three foetal (endothelium, connectivetissue, trophoblast) but only two maternal layers(connective tissue, endothelium) because the maternalepithelium is lost. The foetal trophoblast invades theendometrial epithelium, but does not destroy theendothelium of the maternal capillaries (Mossman1987; Valtonen 1992). Between days 12 and 13 aftermating, the embryos are implanted in the endometrium(Enders and Schlafke 1972; Mead et al. 1988b).Prior to implantation, the trophoblast differentiatesrapidly and gives rise to patches of syncytial trophoblast(Gulamhusein and Beck 1973). The first feature ofpenetration is the projection of a thin fold of syncytialtrophoblast between adjacent epithelial cells. As implantationprogresses, more of the blastocyst is involvedin implantation, and by day 14, a continuous sheet,either penetrating or overlying the area of the wall of theuterus, constitutes two-thirds of the circumference of thefuture zonary band placenta (Enders and Schlafke1972). At the cervical and ovarian ends of the chorionicvesicle and in the mesometrial region, the trophoectodermis non-invasive (Gulamhusein and Beck 1975). Inthese non-attached regions, the trophoectodermic cellsabsorb uterine milk (Gulamhusein and Beck 1973). Inthe immediate vicinity of implanting chorionic vesicles, ahighly localized increase in the permeability of uterineblood vessels is associated with the final stage ofattachment to the uterine epithelium, this being firstdetectable in the morning of day 12 after mating (Meadet al. 1988b). Prostaglandins are proposed to play animportant role in the process of implantation, but theprocess is unrelated to decidual formation because theferret is an adeciduate species (Mead et al. 1988b); thatis, ferret endometrium is not known to be capable ofdecidual transformation during implantation (Beck1974) in contrast to the situation of rodents and humansthat have a primary decidualization reaction beforeblastocyst(s) start penetrating the uterine epithelium(Johnson and Everitt 2000). Following implantation, awave of epithelial hypertrophy sweeps progressivelyfrom the uterine lumen towards the bases of the glands,and epithelial cells become extraordinarily enlarged withnuclei as large as 90–100 lm in diameter. Most of theluminal cells lose their integrity and form massescontaining whole or fragmented nuclei. This degeneratetissue is termed a symplasma because, although technicallyit is a syncytium, it is not an active tissue (Amoroso1952; Buchanan 1966). This degenerated tissue probablycontributes to the histiotrophe because it is ingested bythe syncytiotrophoblast (Gulamhusein and Beck 1975).In the ferret, the placental labyrinth is fully developedat day 18, when the greatly hypertrophied maternalcapillaries are completely surrounded by a layer ofsyncytiotrophoblast (Gulamhusein and Beck 1975). Atthe same time, accumulations of maternal blood, whichvary considerably in size and location and constitute the‘haemophagous organ’ (Creed and Biggers 1964),appear in the antimesometrial region between theplacental discs. This organ is fully formed by day 28,and it maintains its size almost to term (Gulamhuseinand Beck 1975). It is thought to act as an alternativesource of iron for the foetuses (Baker and Morgan1973).Gestation and function of corpora luteaGestation length is 41 days (39–42 days) in the domesticferret (Hammond and Marshall 1930; Ternovsky andTernovskaya 1994; Fox and Bell 1998). If fertilizationdoes not occur, pseudopregnancy lasting 40–42 daysensues, the functional life of corpora lutea (CL) beingsimilar to that in normal pregnancy (Hammond andMarshall 1930; Chang and Yanagimachi 1963). FerretCL start secreting progesterone immediately afterovulation, and the progesterone concentrations risecontinuously to peak at approximately days 12–14during the period of implantation (Daniel 1976), thendecrease steadily after approximately day 15 and leveloff by day 24 of pregnancy (Heap and Hammond 1974).Ferret CL consist of small (25 lm) luteal cells, with the smaller cells predominatingon day 6. A shift towards larger sizes is observed bydays 13 and 24 of pregnancy, and concurrently thepercentage of smaller cells declines (Joseph and Mead1988). Progesterone concentrations decline continuouslybetween day 24 and parturition at day 42 (Blatchley andDonovan 1976). Mustelids and other carnivores displaya protracted decline in circulating progesterone, andprogesterone levels reach basal concentrations ‡1 weekafter parturition (Møller 1973). The conceptuses haveno effect on the duration of the luteal phase, becausepregnancy and pseudopregnancy are indistinguishable(Hammond and Marshall 1930; Heap and Hammond1974; Agu et al. 1986). Removal of the uterus during theluteal phase has no effect on the life span of the CL(Deanesly 1967). Factors causing luteal regression in theferret are not known.The functional ferret CL secrete glucose-6-phosphateisomerase on days 6–9 of pregnancy, the time at whichimplantation-promoting activity has been found incorpora lutea. This isomerase has been identified to benecessary for embryo implantation in the domestic ferret(Schulz and Bahr 2003). It seems that oestrogen is notessential for implantation in the ferret (Foresman andMead 1978; Mead and McRae 1982). Experimentallyinduced maternal pregnancy reactions (for instance,traumatization with an intrauterine thread that resultsin the formation and subsequent necrosis of symplasmalnests of endometrial epithelial cells and hypertrophy ofthe maternal capillary endothelium) show that theendometrium is sensitive to implantation between days9 and 14 (Gulamhusein and Beck 1977).Parturition, rebreeding, induction of oestrus andsuperovulationThe domestic ferret gives birth to an average of eightkits (1–18 kits), which weigh 6–12 g at birth (Ternovskyand Ternovskaya 1994; Fox and Bell 1998). Females willreturn to oestrus within 2 weeks after weaning if theyare exposed to a stimulatory photoperiod. If the kits areremoved at birth, the mothers return to oestrus 8 weeksafter mating, as do pseudopregnant animals and femalesÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

154 H Lindebergthat have resorbed their foetuses (Marston and Kelly1969). If females give birth to only a small number ofkits (e.g. 1–5), they may return to oestrus while nursing(Fox and Bell 1998). Induction of oestrus in anoestrousferrets has been accomplished by treatment with crudepituitary extracts (Hill and Parkes 1930). Most anoestrousfemales maintained on a non-stimulatory lightcycle (10 h light and 14 h dark) showed oestrous signs,and were bred 6–13 days after treatment with 0.25 mgof FSH, administered twice daily (Mead and Neirinckx1989). For successful superovulation, ferret females havebeen treated with eCG followed by hCG 72 h later (Liet al. 2001, 2002).ReferencesAgu GO, Rajkumar K, Murphy BD, 1986: Evidence fordopaminergic regulation of prolactin and a luteotropiccomplex in the ferret. Biol Reprod 35, 508–515.Amoroso EC, 1952: Placentation. In: Parkes AS (ed.),Marshall’s Physiology of Reproduction, vol 2. Longman,London, pp. 127–311.Amstislavsky SY, Ternovskaya Y, 2000: Reproduction inmustelids. Anim Reprod Sci 60–61, 571–581.Amstislavsky SY, Maksimovsky LF, Ternovsky YG, TernovskyDV, 1993: Ermine reproduction and embryodevelopment (Mustela erminea). Scientifur 17, 293–298.Baker E, Morgan EH, 1973: Placental iron transfer in the cat.J Physiol 232, 485–501.Beck F, 1974: The development of a maternal pregnancyreaction in the ferret. J Reprod Fertil 40, 61–69.Betteridge KJ, Fle´ chon JE, 1988: The anatomy and physiologyof pre-attachment bovine embryos. Theriogenology 29, 155–187.Betteridge KJ, Eaglesome MD, Mitchell D, Flood PF, Be´ riaultR, 1982: Development of horse embryos up to twenty daysafter ovulation: observations on fresh specimens. J Anat135, 191–209.Bissonnette TH, 1932: Modification of mammalian sexualcycles; reactions of ferrets (Putorius vulgaris) of both sexes toelectric light added after dark in November and December.Proc R Soc Lond 110, 322–336.Blatchley FR, Donovan BT, 1976: Progesterone secretionduring pregnancy and pseudopregnancy in the ferret.J Reprod Fertil 46, 455–456.Buchanan GD, 1966: Reproduction in the ferret (Mustela furo)I. Uterine histology and histochemistry during pregnancyand pseudopregnancy. Am J Anat 118, 195–216.Buchanan GD, 1969: Reproduction in the ferret (Mustela furo)II. Changes following ovariectomy during early pregnancy.J Reprod Fertil 18, 305–316.Carroll RS, Erskine MS, Doherty PC, Lundell LA, Baum MJ,1985: Coital stimuli controlling luteinizing hormone secretionand ovulation in the female ferret. Biol Reprod 32, 925–933.Chang MC, 1950: Cleavage of unfertilised ova in immatureferrets. Anat Rec 108, 31–44.Chang MC, 1957: Natural occurrence and artificial inductionof parthenogenetic cleavage of ferret ova. Anat Rec 128,187–200.Chang MC, 1965a: Implantation of ferret ova fertilized bymink sperm. J Exp Zool 160, 67–80.Chang MC, 1965b: Fertilizing life of ferret sperm in the femaletract. J Exp Zool 158, 87–100.Chang MC, 1969: Development of transferred ferret eggs inrelation to the age of corpora lutea. J Exp Zool 171, 459–464.Chang MC, Yanagimachi R, 1963: Fertilization of ferret ovaby deposition of epididymal sperm into the ovarian capsulewith special reference to the fertilizable life of ova and thecapasitation of sperm. J Exp Zool 154, 175–188.Creed RFS, Biggers JD, 1964: Placental haemophagous organsin the Procyonides and Mustelidae. J Reprod Fertil 8, 133–137.Daniel JC Jr, 1970: Coincidence of embryonic growth anduterine protein in the ferret. J Embryol Exp Morphol 24,305–312.Daniel JC Jr, 1976: Plasma progesterone levels before and atthe time of implantation in the ferret. J Reprod Fertil 48,437–438.Deanesly R, 1967: Experimental observations on the ferretcorpus luteum of pregnancy. J Reprod Fertil 13, 183–185.Elliott JA, Goldman BD, 1981: Seasonal reproduction:photoperiodism and biological clocks. In: Adler NT (ed.),Neuroendocrinology of Reproduction, Physiology andBehaviour. Plenum Press, New York, pp. 377–423.Enders AC, Schlafke S, 1972: Implantation in the ferret:epithelial penetration. Am J Anat 133, 291–316.Farstad W, Hyttel P, Hafne AL, Nielsen J, 2001: Maturationand fertilization of blue fox (Alopex lagopus) oocytes invitro. J Reprod Fertil Suppl 57, 161–165.Foresman KR, Mead RA, 1978: Luteal control of nidation inthe ferret (Mustela putorius). Biol Reprod 18, 490–496.Forsberg M, 1992: Seasonal breeding and the significance oflight. In: Tauson A-H, Valtonen M (eds), Reproduction inCarnivorous Fur Bearing Animals. Nordiska JordbruksforskaresFo¨ rening, Rapport Nr. 75, Jordbrugsforlaget,Nordisk Kulturfond, Copenhagen, pp. 17–38.Fox JG, Bell JA, 1998: Growth, reproduction and breeding.In: Fox JG (ed.), Biology and Diseases of the Ferret, 2ndedn. Williams and Wilkins, Baltimore, pp. 211–227.Gulamhusein AP, Beck F, 1973: Light and electron microscopicobservations at the pre- and early post-implantationstages in the ferret uterus. J Anat 115, 159–174.Gulamhusein AP, Beck F, 1975: Development and structure ofthe extra-embryonic membranes of the ferret. A lightmicroscopic and ultrastructural study. J Anat 120, 349–365.Gulamhusein AP, Beck F, 1977: Determination of a sensitiveperiod for the induction of the maternal pregnancy reactionin the ferret. J Reprod Fertil 49, 127–128.Hamilton WJ, 1934: The early stages in the development of theferret: fertilisation to the formation of the prochordal plate.Trans R Soc Edinb B 58, 251–278.Hamilton WJ, Gould GH, 1940: The normal oestrous cycle ofthe ferret: the correlation of the vaginal smear and thehistology of the genital tract, with notes on the distributionof glycogen, the incidence of growth, and the reaction tointravitam staining by trypan blue. Trans R Soc Edinb B60, 87–106.Hammond J, Marshall FHA, 1930: Estrus and pseudopregnancyin ferret. Proc R Soc Lond B Biol Sci 105, 607–630.Hammond J, Walton A, 1934: Notes on ovulation andfertilization in the ferret. J Exp Biol 14, 307–319.Hansson A, 1947: The physiology of reproduction in mink(Mustela vison Schreb) with special reference to delayedimplantation. Acta Zool (Stockh) 28, 1–136.Harvey NE, MacFarlane WV, 1958: The effects of day lengthon the coat-shedding cycles, body weight, and reproductionof the ferret. Aust J Biol Sci 11, 187–199.Heap RB, Hammond J Jr, 1974: Plasma progesterone levels inpregnant and pseudopregnant ferret. J Reprod Fertil 39,149–152.Herbert J, 1989: Light as a multiple control system onreproduction in mustelids. In: Seal US, Thorne ET, BoganM, Anderson S (eds), Conservation Biology of theÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproduction of Domestic Ferret 155Black-footed Ferret. Yale University Press, New Haven,CT, pp. 138–159.Hill M, Parkes AS, 1930: On the relation between the anteriorpituitary body and the gonads. Part II. The induction ofovulation in the anoestrous ferret. Proc R Soc Lond B 107,39–49.Holst PA, Phemister RD, 1971: The prenatal development ofthe dog: preimplantation events. Biol Reprod 5, 194–206.Jalkanen L, 1992: Embryonal development and embryo lossesduring the preimplantation period in the silver fox. Norw JAgric Sci 9, 108–114.Jalkanen L, Valtonen M, Lukola A, 1988: Electrical resistanceand cytology of the vaginal tract in relation to ovarianhormone levels in the silver fox during estrus. In: Proceedingsof the 4th International Congress Fur Animal Production,Ontorio, Canada, pp. 5–9 .Johnson MH, Everitt BJ, 2000: Essential Reproduction, 5thedn. Blackwell Science Ltd, London, pp. 173–192.Joseph MM, Mead RA, 1988: Size distribution of ferret lutealcells during pregnancy. Biol Reprod 39, 1159–1169.Lagerqvist G, 1992: Reproductive features in the ferret(Mustela putorius). In: Tauson A-H, Valtonen M (eds),Reproduction in Carnivorous Fur Bearing Animals.Nordiska Jordbruksforskares Fo¨ rening, Rapport Nr. 75.Jordbrugsforlaget, Nordisk Kulturfond Copenhagen, pp.87–95.Li ZY, Jiang Q, Zhang YL, Liu XM, Engelhardt JF, 2001:Successful production of offspring after superovulation andin vitro culture of embryos from domestic ferrets (Mustelaputorius furos). Reproduction 122, 611–618.Li ZY, Jiang Q, Sabet MR, Zhang Y, Ritchie TC,Engelhardt JF, 2002: Conditions for in vitro maturationand artificial activation of ferret oocytes. Biol Reprod 66,1380–1386.Li ZY, Sun X, Chen J, Leno GH, Engelhardt JF, 2006:Factors affecting the efficiency of embryo transfer in thedomestic ferret (Mustela putorius furo). Theriogenology 66,183–190.Linderberg H, 2003: Embryo technology in the farmedEuropean Povelat (Mustela Putorius). PhD thesis, Universityof Kuopio, pp. 1–110.Lindeberg H, Ja¨ rvinen M, 2003: Early embryonic developmentand in vitro culture of in vivo produced embryos in thefarmed European polecat (Mustela putorius). Theriogenology60, 965–975.Lindeberg H, Jalkanen L, Savolainen R, 1993: In vitro cultureof silver fox embryos. Theriogenology 40, 779–788.Mainland D, 1930: The early development of the ferret: thePronuclei. J Anat 64, 262–287.Mainland D, 1931: A quantitative study of the polar body ofthe ferret, with a note on the second polar spindle. AmJ Anat 47, 195–240.Mainland D, 1932: The early development of the ferret: thezona granulosa, zona pellucida and associated structures.J Anat 66, 586–601.Marshall FHA, 1904: The estrous cycle in the common ferret.Quart J Microsc Sci 48, 323–345.Marston JH, Kelly WA., 1969: Contraceptive action of intrauterinedevices in the ferret. J Reprod Fertil 18, 419–429.McRae AC, 1992: Effect of ovariectomy on blastocystexpansion and survival in ferrets (Mustela putorius furo).Reprod Fertil Dev 4, 239–247.Mead RA, 1989: Reproduction in mustelids. In: Seal US,Thorne ET, Bogan MA, Anderson SH (eds), ConservationBiology and the Black-footed Ferret. Yale University Press,New Haven, CT, pp. 124–137.Mead RA, McRae M, 1982: Is estrogen required forimplantation in the ferret? Biol Reprod 27, 540–547.Mead RA, Neirinckx S, 1989: Hormonal induction of oestrusand pregnancy in anoestrous ferrets (Mustela putorius furo).J Reprod Fertil 86, 309–314.Mead RA, Joseph MM, Neirinckx S, 1988a: Optimal dose ofhuman chorionic gonadotrophin for inducing ovulation inthe ferret. Zoo Biol 7, 263–267.Mead RA, Bremner S, Murphy BD, 1988b: Changes inendometrial vascular permeability during the periimplantationperiod in the ferret (Mustela putorius). J Reprod Fertil82, 293–298.Møller OM, 1973: The fine structure of the lutein cells in themink (Mustela vison) with special reference to the secretoryactivity during pregnancy. Z Zellforsc Mikrosk Anat 138,523–544.Mossman HW, 1987: Vertebrate Fetal Membranes. RutgersUniversity Press, New Brunswick, NJ, pp. 250–262.Murphy BD, 1989: Reproductive physiology of femalemustelids. In: Seal US, Thorne ET, Bogan MA, AndersonSH (eds), Conservation Biology and the Black-footedFerret. Yale University Press, New Haven, CT, pp. 107–123.Olson PN, Thrall MA, Wykes PM, Husted PW, Nett TM,Sawyer HR, 1984: Vaginal cytology. Part I. A useful tool forstaging the canine estrous cycle. Compend Contin EducPract 6, 288–297.Piltti K, Aalto J, Ja¨ rvinen M, Korhonen H, Kuronen V,Lindeberg H, Amstislavsky S, Ternovskaya Y, Valtonen M,Halmekyto¨ M, 2003: Nuclear maturation of Europeanpolecat (Mustela putorius) in vivo oocytes and success inusing mustelid hybrids as recipients for embryo transfer.Theriogenology 59, 403. (abstract).Renton JP, Boyd JS, Eckersall PD, Ferguson JM, HarveyMJA, Mullaney J, Perry B, 1991: Ovulation, fertilizationand early embryonic development in the bitch (Canisfamiliaris). J Reprod Fertil 93, 1–10.Rider V, Heap RB, 1986: Heterologous anti-progesteronemonoclonal antibody arrest early embryonic developmentand implantation in the ferret (Mustela putorius). J ReprodFertil 76, 459–470.Robinson A, 1918: The formation, rupture, and closure ofovarian follicles in ferrets and ferret polecat hybrids, andsome associated phenomena. Trans R Soc Edinb B 52, 303–362.Robinson A, 1925: Lipoids in mammalian ova and theirzygotes. Proceedings of the Anatomical Society of GreatBritain and Ireland; J Anat 59, 109–111.Schulz LC, Bahr JM, 2003: Glucose-6-phosphate isomerase isnecessary for embryo implantation in the domestic ferret.Proc Natl Acad Sci USA 100, 8561–8566.Strahl H, Ballman E, 1915: Embryonalhu¨ llen und Plazentavon Putorius furo. Abhand Ko¨ nigl Akademie derWissenschaften Physik-Mathem Classe Nr. 4. Berlin, pp.1–69Swanson WF, Roth TL, Wildt DE, 1994: In vivo embryogenesis,embryo migration, and embryonic mortality in thedomestic cat. Biol Reprod 51, 452–464.Ternovsky DV, Ternovskaya YG, 1994: Ecology of Mustelids.Scientific Siberian Press, Novosibirsk Nauka, in Russian.Tritt SH, 1986: Regulation of luteinizing hormone (LH)secretion in the estrous ferret. Biol Reprod 34 (Suppl. 1) 198.Turek FW, Van Cauter E, 1988: Rhythms in reproduction. In:Knobil E, Neill JD (eds), The Physiology of Reproduction.Raven Press, New York, pp. 1789–1830.Valtonen M, 1992: Comparative aspects of reproductivephysiology in domestic animals. In: Tauson A-H, ValtonenM(eds), Reproduction in Carnivorous Fur Bearing Animals.Nordiska Jordbruksforskares Fo¨ rening, Rapport Nr. 75.Jordbrugsforlaget, Nordisk Kulturfond Copenhagen,pp. 1–15.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

156 H LindebergValtonen M, Jalkanen L, 1993: Species-specific features ofoestrous development and blastogenesis in domestic caninespecies. J Reprod Fertil Suppl 47, 133–137.Valtonen M, King WA, Gustavson I, Ma¨ kinen A, 1985:Embryonic development in the blue fox. Nord Vet Med 37,243–248.Williams ES, Thorne ET, Kwiatkowski DR, Lutz K,Anderson SL, 1992: Comparative vaginal cytology of theestrous cycle of black-footed ferrets (Mustela nigripes),Siberian polecats (M. eversmanni), and domestic ferret(M. putorius furo). J Vet Diagn Invest 4, 38–44.Author’s address (for correspondence): H Lindeberg, University ofKuopio, Department of Biosciences, P.O. Box 1627, FIN-70211Kuopio, Finland. E-mail: lindeber@messi.uku.fiConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 157–164 (2008); doi: 10.1111/j.1439-0531.2008.01156.xISSN 0936-6768Physiology of the Canine Anoestrus and Methods for Manipulation of Its LengthJ. de Gier, NJ Beijerink, HS Kooistra and AC OkkensDepartment of Clinical Sciences of Companion Animals, Faculty of Veterinary Medicine, University of Utrecht, Yalelaan, Utrecht, The NetherlandsContentsProgression from early to late anoestrus is characterized by theappearance of a larger number of gonadotrophin-releasinghormone (GnRH) pulses with a higher amplitude, an increasein the sensitivity of the pituitary to GnRH, an increase inovarian responsiveness to gonadotrophins, and an increase inbasal plasma follicle-stimulating hormone (FSH) concentration.A period of increased luteinizing hormone (LH) pulsatilityhas been observed shortly before the onset of pro-oestrus.Apart from these changes in the hypothalamus-pituitary-ovaryaxis, the initiation of a new follicular phase in the bitch is alsostimulated by dopaminergic influences other than the accompanyingplasma prolactin decrease. Metergoline, a drug whichin a low dosage lowers the plasma prolactin concentration viaa serotonin-antagonistic pathway, does not shorten theanoestrus; while bromocriptine, in a dosage insufficient tocause a decrease in the plasma prolactin concentration, doesprematurely induce a follicular phase. These observationsindicate that it is not the decrease in the plasma prolactinconcentration, but another dopamine-agonistic influence thatplays a crucial role in the transition to a new follicular phase.The dopamine-agonist induced oestrus is associated with arapid rise in the basal plasma FSH concentration, similar towhat is observed during the physiological late anoestrus.Administration of GnRH, eCG and oestrogens may also beused to induce oestrus but with variable results. Oestrus can beprevented surgically or medically, for which purpose progestagensare the most important drugs. The mechanism is stillunclear, although it has been demonstrated that with continuingmedroxyprogesterone acetate (MPA) treatment the FSHresponse to GnRH stimulation decreases and changes occur inthe pulsatile release of the gonadotrophins. In general, LHpulses coincide with a FSH pulse, but during MPA treatment,LH pulses were observed while there was such a small increasein FSH that it was not recognized as significant FSH pulse.Physiology of the Canine AnoestrusHypothalamus-pituitary-ovary axisThe duration of the oestrous cycle of the bitch isconsiderably longer than that of most other domesticanimals. The follicular phase and spontaneous ovulationsare followed by a luteal phase having an average durationof approximately 2 months, irrespective of pregnancy. Anon-seasonal anoestrus, with a duration of 2 to10 months, follows each oestrous cycle (Schaefers-Okkens1996). Follicle-stimulating hormone (FSH) andluteinizing hormone (LH) play an essential role in theinduction of folliculogenesis and ovulation. The mainregulator of FSH and LH secretion is gonadotrophinreleasinghormone (GnRH). The spontaneous GnRHrelease from excised hypothalamic fragments, thatinclude the ‘mediobasal hypothalamic-preoptic areasuprachiasmaticnucleus units’ derived from beaglebitches at different stages of the oestrous cycle, is pulsatilethroughout all stages of the oestrous cycle. Progressionfrom early to late anoestrus in the bitch is characterized byan increased release of GnRH by the hypothalamus.Especially during late anoestrus GnRH pulse frequency issignificantly increased (Tani et al. 1996). The sensitivityof the pituitary to GnRH and the indirect response of theovary to GnRH of bitches in early and late anoestrus havebeen investigated by way of the intravenous administrationof graded doses of GnRH. The responses, expressedas circulating LH and oestradiol concentration profilesover time, were significantly dose-dependent and higherin late anoestrus than in early anoestrus. In addition,GnRH-induced LH and oestradiol profiles were positivelycorrelated (van Haaften et al. 1994). These resultsindicate that during the course of anoestrus, there is anincrease in the sensitivity of the pituitary to GnRH and inovarian responsiveness to gonadotrophins (Jeffcoate1993; van Haaften et al. 1994).In order to study the role of gonadotrophins duringthe transition from anoestrus to the follicular phase, thepulsatile plasma profiles of LH and FSH were determinedduring early-, mid- and late anoestrus in beaglebitches (Kooistra et al. 1999a). During anoestrus, eachFSH pulse coincided with a LH pulse. The mean plasmaLH concentration of the smoothed baseline and themean area under the curve (AUC) for LH did not differsignificantly when the different phases of anoestrus werecompared. In contrast, the mean plasma FSH concentrationof the smoothed baseline and the AUC for FSHduring late anoestrus was significantly higher than thoseduring mid- and early anoestrus (Kooistra et al. 1999a;Onclin et al. 2001). These observations suggest that anincrease in circulating FSH levels is a key event inovarian folliculogenesis in the dog and consequently forthe termination of anoestrus. Indeed, in most mammalsstudied, FSH is regarded as the most important factor inthe early stages of follicular development, whereas LH isregarded as the primary regulatory factor in the moremature follicles (Moyle and Campbell 1995; Monniauxet al. 1997). Consequently, it may be hypothesized thatat a certain moment during anoestrus, the rising plasmaFSH concentration will exceed the threshold value of themost sensitive follicles of the canine ovarian antralfollicle pool leading to an enhancement of the developmentof these follicles. One of the main effects of FSH isthe acquisition of LH receptors in the granulosa cells.Beyond this stage, LH is progressively able to replaceFSH in supporting follicular maturation (Monniauxet al. 1997). Although FSH pulses appear to occurconcomitantly with LH pulses in all stages of the canineoestrous cycle and anoestrus (Kooistra et al. 1999a),differential regulation of FSH and LH has also beenreported in this species, both during anoestrus, duringthe follicular phase and during the period of ovulationÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

158 J de Gier, NJ Beijerink, HS Kooistra and AC Okkensand fertilization (Kooistra et al. 1999a; de Gier et al.2006). Differential regulation of FSH and LH secretioncan at least partly be explained by the frequency andamplitude of GnRH pulses (Haisenleder et al. 1991;Vizcarra et al. 1997). In addition, gonadal feedback(Mann et al. 1992; Shupnik 1996) and a specific hypothalamicFSH-releasing factor (Yu et al. 1997) may playa role in the differential or non-parallel secretion of FSHand LH. Furthermore, the intracellular mechanisms forthe storage and release differ for FSH and LH (Chowdhuryand Steinberger 1975; Moyle and Campbell 1995;Nicol et al. 2004).In addition to a rise in hypothalamic GnRH release, anincrease in basal plasma FSH concentration, andincreased sensitivity of the pituitary and ovaries duringprogression of anoestrus, several other factors that maybe involved in the initiation of folliculogenesis andtermination of anoestrus have been reported. In somebitches, an increased LH pulsatility has been observedshortly before the start of pro-oestrus (Concannon et al.1986; Kooistra et al. 1999b; Tani et al. 1999; Beijerinket al. 2004). The exact role of increased LH pulsatility inthe termination of anoestrus in the bitch remains elusive.One of the main effects of the rising FSH level is theacquisition of LH receptors in the granulosa cells(Monniaux et al. 1997). It is therefore possible that theincrease in LH pulsatility at the end of anoestrus providesa stimulus to follicles which are no longer receptive toFSH but have acquired sufficient LH receptors.In addition, enhancement of the expression of thegenes encoding for the oestrogen receptor (Tani et al.1997) and P450 aromatase (which catalyses oestrogenbiosynthesis in the canine hypothalamus) during anoestrushas been reported in dogs (Inaba et al. 2002).However, although sporadic elevations are observed,plasma oestradiol concentrations are usually low duringanoestrus and do not begin to rise until approximately amonth before the LH peak (Jeffcoate 1993). Furthermore,there are no manifestations of oestrogen influenceson the reproductive tract or on sexual behaviourduring anoestrus, and vaginal endoscopy or cytologyreveals no evidence of oestrogenic stimulation until lateanoestrus.In addition, there is some evidence that factorscausing a decrease in opioidergic activity promote LHrelease and the termination of anoestrus (Concannon1993). Treatment with naloxone, an opioid antagonist,stimulated LH release at nearly all stages of the oestrouscycle (Concannon and Temple 1988). Finally, nochanges in the expression of the FSH receptor havebeen found during canine anoestrus (McBride et al.2001). This study demonstrated that anoestrus in bitchesis neither due to failure of expression of the canine FSHreceptor nor due to a change in splicing pattern to aninactive form.Dopaminergic influences and induction of a prematureoestrusIn addition to the earlier-mentioned changes in thehypothalamus-pituitary-ovary axis, there is evidence ofinvolvement of dopaminergic influences in the initiationof a new follicular phase in the bitch. Administration ofthe dopamine-agonists bromocriptine and cabergoline isassociated with both inhibition of prolactin release andshortening of the interoestrous interval (Okkens et al.1985; van Haaften et al. 1989; Concannon 1993; Onclinet al. 1995; Kooistra et al. 1999b; Verstegen et al. 1999;Gobello et al. 2002; Beijerink et al. 2003). If bromocriptinetreatment is started during the luteal phase,shortening of the interoestrous interval is primarily theresult of a shortening of the anoestrus (Okkens et al.1985), but is also due to shortening of the luteal phase(Okkens et al. 1985, 1990). The shortening of the lutealphase is probably caused by a decrease in the secretionof prolactin, the main luteotrophic factor in the bitch(Okkens et al. 1990; Onclin and Verstegen 1997).It has been hypothesized that the shortening of theanoestrus by dopamine-agonists is also the result of thesuppression of prolactin secretion, as prolactin mayinhibit gonadotrophin release. Indeed, it has beendemonstrated in various mammalian species that highcirculating levels of prolactin in different pathologicalsituations inhibit LH pulsatility (Sauder et al. 1984;Yazigi et al. 1997) or are associated with decreased LHsecretion (Park et al. 1993). In addition, decreasedplasma LH levels were observed during physiologicalhyperprolactinaemia in lactating sows, while lowering ofthe plasma prolactin concentration by bromocriptineadministration led to a rise in plasma LH levels in theseanimals (Bevers et al. 1983). Yet, under physiologicalconditions plasma prolactin concentrations are lowduring canine anoestrus (Olson et al. 1982; Kooistraand Okkens 2001), and no obvious changes in plasmaprolactin concentration have been observed during thetransition from anoestrus to the follicular phase in thebitch (Olson et al. 1982). Furthermore, anoestrus wasnot shortened in dogs treated with low dosages of theserotonin receptor antagonist metergoline (Okkens et al.1997), although the plasma prolactin levels were lowerthan in bromocriptine-treated dogs. The results of thelatter study suggest that the induction of the follicularphase is not initiated by suppression of prolactinsecretion. This raised the question whether administrationof a dopamine agonist in a dosage that is too low tosuppress prolactin secretion still will result in a shorteningof anoestrus in the bitch. To investigate thishypothesis, bitches were daily treated twice with 5(5-group), 20 (20-group), or 50 (50-group) microgramsof bromocriptine per kg body weight orally, starting28 days after ovulation (Beijerink et al. 2003). In thebitches receiving 5 lg bromocriptine per kg body weighttwice daily, there was no difference between the averageplasma prolactin concentration prior to and duringtreatment. This is in contrast with the bitches whichreceived 20 or 50 lg bromocriptine per kg body weighttwice daily, in which the average plasma prolactinconcentration prior to bromocriptine treatment wassignificantly higher than that during treatment. Themean ± SEM interoestrous interval was 136 ± 16 daysin the 5-group, 96 ± 6 days in the 20-group, and92 ± 11 days in the 50-group. Each of these intervalswas significantly shorter than the mean interoestrousinterval in control cycles, 216 ± 9 days. The meaninteroestrous intervals in the 20- and 50-groups weresimilar and significantly shorter than that of theÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Canine Anoestrus, Oestrous Induction and Prevention 1595-group. This study provided further evidence for theassumption that a premature oestrus is not induced by adecrease in plasma prolactin concentration.Besides an inhibition of prolactin release, the bromocriptine-inducedshortening of anoestrus is also associatedwith a quick rise in the basal plasma FSHconcentration without a concomitant increase in thebasal plasma LH concentration (Fig. 1) (Kooistra et al.1999b). These results give further support to the notionthat in the bitch, an increase in circulating FSH shouldbe considered to be a critical event required for ovarianfolliculogenesis. Treatment with bromocriptine maycause an increase in plasma FSH concentration to alevel that enhances development of follicles. This issimilar to the endocrine events during late anoestrus innon-treated bitches (Kooistra et al. 1999a).A role for dopamine in the control of reproductionhas been demonstrated in different mammalian species,although the effects of dopamine in other species differfrom those in the bitch (Beck et al. 1978; Havern et al.1994; Besognet et al. 1997). In species other than thedog, dopamine-agonists may inhibit gonadotrophinsecretion during anoestrus and dopamine-antagonistsmay induce reproductive activity; whereas in the bitch,dopamine-agonists induce the onset of oestrus.As opposed to dopamine agonists, low dosages of theserotonin receptor antagonist metergoline do not resultin shortening of anoestrus (Okkens et al. 1997).Therefore, it may be expected that the changes ingonadotrophin release associated with dopamineagonist-induced shortening of anoestrus will not beobserved during treatment with low dosages of theserotonin receptor antagonist metergoline. Toinvestigate the effects of a low dose of a serotoninantagonist on the pulsatile secretion patterns of FSH andLH, 0.1 mg metergoline per kg body weight orally twicedaily, was administered to bitches, starting 100 days afterovulation (Beijerink et al. 2004). The mean interoestrousinterval in the eight bitches treated with the serotoninantagonist metergoline was not shortened as expecteddespite a decreased plasma prolactin concentration.Moreover, during the first weeks of treatment with theserotonin antagonist metergoline, there were no significantchanges in the pulsatile plasma profiles of FSH orLH. These findings indicate that serotonin antagonistinducedlowering of plasma prolactin does not lead toincreased secretion of FSH.In addition to administration of dopamine-agonists,induction of a follicular phase may also occur via theadministration of GnRH, gonadotrophins, eCG, hCG,porcine LH and oestrogens (for review see Kutzler2007). Although many protocols exist for oestrousinduction in bitches, the fertility results, with theexception of dopamine agonists, are variable andgenerally poor. Some methods are also too costly orlabour intensive to be suitable for veterinary clinicalpractice. In addition, aglepristone, a progesteronereceptorantagonist and prostaglandin F 2a shorten theinteroestrous interval, if administered during the lutealphase (Romagnoli et al. 1993; Galac et al. 2004).Oestrus preventionOestrus can be prevented medically or surgically.Ovariectomy has several advantages. It is effective afterone procedure. It considerably lowers the risk formammary cancer if performed before approximately2.5 years after the first oestrus. It also prevents thedevelopment of pyometra and progesterone-inducedgrowth hormone excess. There are, however, also severaldisadvantages such as the risk of complications duringanaesthesia and surgery, and the irreversibility of theprocedure. In addition, there is also the possibility ofside-effects such as changes in the coat and urinaryincontinence. There are indications that the risk forurinary incontinence is greater if the procedure isperformed prior to the first oestrus. Furthermore, ithas been shown that early-age gonadectomy is associatedwith an increased rate of cystitis and that age atgonadectomy is negatively correlated with the rate ofurinary incontinence (Spain et al. 2004). Urinary incontinenceoccurs mainly in large breeds and some breedse.g. the Boxer, Dobermann, Bouvier des Flandres, GiantSchnauzer, Irish Setter, Miniature Poodle, Old EnglishSheepdog, Weimaraner and Rottweiler appear to beespecially at risk for developing urinary incontinence(Thrushfield et al. 1998).Medical oestrous prevention can be accomplishedwith several types of drugs. The progestagens are themost important drugs for oestrous prevention.Fig. 1. The mean (± SEM) plasma FSH and LH concentrations of the smoothed baseline in six beagle bitches the day before and 14, 28 and42 days after the start of bromocriptine treatment (* indicates significant difference) (Kooistra et al. 1999b)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

160 J de Gier, NJ Beijerink, HS Kooistra and AC OkkensAndrogens can also be used for this purpose, butprimarily for short-term prevention. As androgens arenot currently registered for this use, they are notincluded in this paper.GnRH agonists or GnRH antagonistsGnRH agonists or GnRH antagonists may also be usedfor oestrus prevention. At this stage, the GnRHantagonists are unfit for clinical use, because ofeconomic and delivery problems for long-term use(Vickery et al. 1989). Gonadotrophin-releasing hormoneagonists administered at high dosages over a longperiod of time prevent oestrus by pituitary downregulation.However, the early stimulatory effect ofGnRH analogues may cause signs of oestrus, if administeredduring anoestrus and sometimes even if administeredduring the luteal phase. It is possible to preventthe occurrence of oestrus by administering a GnRHagonist in pre-pubertal bitches at approximately4 months of age, whereas implantation of a GnRHagonist at 7 months of age does not prevent theoccurrence of an oestrous cycle (Rubion et al. 2006;Trigg et al. 2006). However, in the latter dogs, oestrus isdelayed for a long time. Rubion et al. (2006) showedthat GnRH agonist implants administered beforepuberty prevented reproductive function for 1 year inpre-pubertal bitches. Following removal of the implant,oestrus occurred naturally in 7 of 10 bitches and wasinduced in three bitches after 1.2–14.3 months. Anotherstrategy to prevent oestrous induction during anoestrusafter implantation of a GnRH agonist is administrationof a progestagen before implantation. The efficacy ofthis initial treatment, however, appears to depend on theinterval between start of progestagen treatment andimplantation, the stage of anoestrus, the dosage of theused progestagen or the type of GnRH analogue(Wright et al. 2001; Sung et al. 2006). Furthermore,adding progestagens to prevent the initial stimulation ofGnRH agonists cancels the advantage of not administeringprogestagens for oestrous prevention. Additionally,it is sometimes a problem to remove these implants,because they cannot be found.ProgestagensThe mechanism of the contraceptive activity of progestagensis still unclear. In many species, there is evidencethat contraceptive progestagens reduce serum concentrationsof gonadotrophins. However, high doses ofmedroxyprogesterone acetate (MPA) administered tobeagle bitches for several months did not reduce theincreased circulating concentrations of LH in ovariectomizedbitches, nor did it lower LH concentrations inintact bitches (McCann et al. 1987). In another study,high contraceptive doses of megestrol acetate (MA) didnot suppress basal gonadotrophin secretion duringanoestrus, nor was the pituitary hypersecretion of LHand FSH that occurs in ovariectomized bitches suppressed(Colon et al. 1993). Beijerink (2007) examined 6-h plasma profiles of FSH and LH in five bitches beforeand 3, 6, 9 and 12 months after the start of MPAtreatment. The results of this study demonstrate thattreatment with MPA affects the hypothalamic-pituitaryovarianaxis. Oestrus, ovulation and a subsequent lutealphase did not occur in any of the bitches duringtreatment with MPA. However, the prevention ofoestrus by MPA could not be ascribed to a significantreduction in circulating levels of either FSH or LH. Onthe contrary, during the first months of MPA treatment,there was an increase in basal plasma FSH and LHconcentrations. This progestagen-induced increase ingonadotrophin concentration was not observed in thestudies of McCann et al. (1987) and Colon et al. (1993),and its recognition may be explained by the repeatedsampling employed in the studies by Beijerink (2007).The elevated plasma gonadotrophin levels during thefirst months of MPA treatment may be due to a directinhibitory effect of MPA at the ovarian level, resulting insuppression of the ovarian secretion of oestradiol orinhibin (Mann et al. 1992; Shupnik 1996).With continuing MPA treatment, basal plasmagonadotrophin concentrations returned to pre-treatmentlevels. In addition, the pituitary FSH response toGnRH stimulation decreased, suggesting that MPAtreatment attenuated pituitary FSH sensitivity to endogenousGnRH (Fig. 2) (Beijerink et al. 2007). PulsatileFSH and LH release was maintained during MPAtreatment, but there were indications that changesoccurred in the pulsatile release of the gonadotrophins.In general, LH pulses coincide with a FSH pulse, butduring MPA treatment several LH pulses were accompaniedby such small increases in FSH that theseincreases were not recognized as significant (Fig. 3)(Beijerink 2007).The progestagens most frequently used for oestrousprevention in the dog are proligestone and MPA. Thesingle subcutaneous injection dosage recommended byLH (µg/l)FSH (µg/l)12010080604020050403020100Before 2 5 8 11Months of treatmentBefore 2 5 8 11Months of treatmentFig. 2. Plasma LH and FSH responses (mean ± SEM) in five dogs ina combined anterior pituitary function test according to Meij et al.(1996) before and subsequently 2, 5, 8 and 11 months after the start ofthe treatment with MPA (10 lg ⁄ kg, every 4 weeks). Blood sampleswere collected at -15, 0, 5, 10, 20, 30 and 45 min following the injectionof the releasing hormones at 0 min (arrow) (Beijerink et al. 2007)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Canine Anoestrus, Oestrous Induction and Prevention 161Fig. 3. The 6-h plasma profiles of FSH (black squares) and LH (white triangles) in a 3-year-old beagle bitch before, and 3, 6, 9 and 12 monthsafter the start of treatment with MPA (10 lg ⁄ kg, every 4 weeks). *Significant pulses of both FSH and LH identified by the Pulsar programme;^significant pulse of LH without a concurrent significant pulse of FSH (Beijerink 2007)the manufacturer for proligestone 1 ranges from10 mg ⁄ kg for a dog of approximately 60 kg, to30 mg ⁄ kg for one of 3 kg and for MPA 2 , the single1 DelvosteronÒ, Mycofarm, Delft, The Netherlands2 DepopromoneÒ, Upjohn, Ede, The Netherlands; PerlutexÒ, LeoPharma, Ballerup, Denmarksubcutaneous injection dosage is 2 mg ⁄ kg (maximum60 mg per animal) (Schaefers-Okkens 1996). Thesedrugs should be administered during anoestrus approximately1 month before the onset of the expectedfollicular phase. The first oestrus after the use ofproligestone can be expected in the majority of bitcheswithin 9–12 months, using MPA it may take up toÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

162 J de Gier, NJ Beijerink, HS Kooistra and AC Okkens2–3 years. For that reason, the manufacturer does notrecommend using MPA intended for breeding at a laterdate. Medroxyprogesterone acetate 3 can also be administeredorally, 5 mg once daily (10 mg for large dogsduring the first 5 days) for a maximum of 21 days. Thereturn of oestrus may vary from 2 to 9 months.Use of progestagens for oestrus prevention may leadto the following side-effects:• Development of cystic endometrial hyperplasia(CEH)-endometritis (Sokolowski and Zimbelman1974).• Prolonged pregnancy. This occurs if progestagensare administered subcutaneously at the onset of thefollicular phase and the bitch becomes pregnant. Thegestation will be prolonged, and a caesarian sectionmay be needed.• Hypersecretion of growth hormone, which maylead to diabetes mellitus (Selman et al. 1997). Thishypersecretion of growth hormone as a result ofprogestagen administration can be successfully treatedby the progesterone receptor blocker aglepristone(Bhatti et al. 2006).• An increased risk of neoplastic transformation ofmammary tissue. This may range from hyperplasia,adenomatous hyperplasia and adenomas to malignanttumours. The progestagen-induced neoplastic transformationof mammary tissue starts with the proliferationof undifferentiated terminal ductal structures,so-called terminal end buds (Russo and Russo 1991).This proliferation increases the susceptibility of themammary tissue to malignant transformation.The occurrence of these side-effects is, with theexception of ‘prolonged pregnancy’, largely dependentupon the total progestagen exposure. With the adviseddosage regimens, the exposure may be higher with MPAand MA than with proligestone, the latter being a ratherweak progestagen.3 ProveraÒ, Upjohn, Ede, The Netherlands; OvuconÒ, Intervet,Boxmeer, The NetherlandsReferencesBeck W, Hancke JL, Wuttke W, 1978: Increased sensitivity ofdopaminergic inhibition of luteinizing hormone release inimmature and castrated female rats. Endocrinology 102,837–843.Beijerink NJ, 2007: Endocrinology of physiological andprogestin-induced canine anoestrus. Ch. 8 Pulsatile plasmaprofiles of FSH and LH before and during medroxyprogesteroneacetatetreatment in the bitch. PhD Thesis, UtrechtUniversity.Beijerink NJ, Dieleman SJ, Kooistra HS, Okkens AC, 2003:Low doses of bromocriptine shorten the interestrous intervalin the bitch without lowering plasma prolactin concentration.Theriogenology 60, 1397–1386.Beijerink NJ, Kooistra HS, Dieleman SJ, Okkens AC, 2004:Serotonin-antagonist-induced lowering of prolactin secretiondoes not affect the pattern of pulsatile secretion offollicle-stimulating hormone and luteinizing hormone in thebitch. Reproduction 128, 181–188.Beijerink NJ, Bhatti SFM, Okkens AC, Dieleman SJ, Mol JA,Duchateau L, van Ham LML, Kooistra HS, 2007: Adenohypophysealfunction in bitches treated with medroxyprogesteroneacetate. Domest Anim Endocrinol 32, 63–78.Besognet B, Hansen S, Daels PF, 1997: Induction of reproductivefunction in anestrous mares using a dopamineantagonist. Theriogenology 47, 467–480.Bevers MM, Willemse AH, Kruip TAM1983: The effect ofbromocriptine on luteinizing hormone levels in the lactatingsow: evidence for a suppressive action by prolactin and thesuckling stimulus. Acta Endocrinol 104, 261–265.Bhatti SFM, Duchateau L, Okkens AC, van Ham LML, MolJA, Kooistra HS, 2006: Treatment of growth hormoneexcess in dogs with the progesterone receptor antagonistaglépristone. Theriogenology 66, 797–803.Chowdhury M, Steinberger E1975: Biosynthesis and secretionof follicle-stimulating hormone. Endocr Rev 4, 369–374.Colon J, Kimball M, Hansen B, Concannon PW, 1993: Effectsof contraceptive doses of the progestagen megestrol acetateon luteinizing hormone and follicle-stimulating hormonesecretion in female dogs. J Reprod Fertil Suppl 47, 519–521.Concannon PW, 1993: Biology of gonadotrophin secretion inadult and prepubertal female dogs. J Reprod Fertil Suppl47, 3–27.Concannon PW, Temple M, 1988: Periovulatory decline andlate anestrus increase in naloxone-inducible LH release inadult domestic bitches. Biol Reprod Suppl 36, 101.Concannon PW, Whaley S, Anderson SP, 1986: Increased LHpulse frequency associated with termination of anestrusduring the ovarian cycle of the dog. Biol Reprod Suppl 34,119 (Abstract).Galac S, Kooistra HS, Dieleman SJ, Cestnik V, Okkens AC,2004: Effects of aglépristone, a progesterone receptorantagonist, administered during the early luteal phase innon-pregnant bitches. Theriogenology 66, 494–500.de Gier J, Kooistra HS, Djajadiningrat-Laanen SC, DielemanSJ, Okkens AC, 2006: Differential regulation of the secretionof luteinizing hormone and follicle-stimulating hormonearound the time of ovulation in the bitch. Theriogenology66, 1419–1422.Gobello C, Castex G, Corrada Y, 2002: Use of cabergoline totreat primary and secondary anestrus in dogs. J Am VetMed Assoc 11, 1653–1654.van Haaften B, Dieleman SJ, Okkens AC, Bevers MM,Willemse AH, 1989: Induction of oestrus and ovulation indogs by treatment with PMSG and ⁄ or bromocriptine.J Reprod Fertil Suppl 39, 330–331.van Haaften B, Bevers MM, van den Brom WE, Okkens AC,van Sluijs FJ, Willemse AH, Dieleman SJ, 1994: Increasingsensitivity of the pituitary to GnRH from early to lateanoestrus in the beagle bitch. J Reprod Fertil 101, 221–225.Haisenleder DJ, Dalkin AC, Ortolano GA, Marshall JC,Shupnik MA, 1991: A pulsatile gonadotropin-releasinghormone stimulus is required to increase transcription ofthe gonadotropin subunit genes: evidence for differentialregulation of transcription by pulse frequency in vivo.Endocrinology 128, 509–517.Havern RL, Whisnant CS, Goodman RL, 1994: Dopaminergicstructures in the ovine hypothalamus mediating estradiolnegative feedback in anestrous ewes. Endocrinology 134,1905–1914.Inaba T, Namura T, Tani H, Matsuyama S, Torii R, KawataN, Tamada H, Hatoya S, Kumagai D, Sugiura K, SawadaT, 2002: Enhancement of aromatase gene expression in themediobasal hypothalamus during anestrus in the beaglebitch. Neurosci Lett 333, 107–110.Jeffcoate IA, 1993: Endocrinology of anoestrous bitches.J Reprod Fertil Suppl 47, 69–76.Kooistra HS, Okkens AC, 2001: Secretion of prolactin andgrowth hormone in relation to ovarian activity in the dog.Reprod Domest Anim 36, 115–119.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Canine Anoestrus, Oestrous Induction and Prevention 163Kooistra HS, Okkens AC, Bevers MM, Popp-Snijders C, vanHaaften B, Dieleman SJ, Schoemaker J, 1999a: Concurrentpulsatile secretion of luteinizing hormone and folliclestimulatinghormone during different phases of the estrouscycle and anestrus in beagle bitches. Biol Reprod 60, 65–71.Kooistra HS, Okkens AC, Bevers MM, Popp-Snijders C, vanHaaften B, Dieleman SJ, Schoemaker J, 1999b: Bromocriptine-inducedpremature oestrus is associated with changes inthe pulsatile secretion pattern of follicle-stimulating hormonein beagle bitches. J Reprod Fertil 117, 387–393.Kutzler MA, 2007: Estrus induction and synchronization incanids and felids. Theriogenology 68, 354–374.Mann GE, Campbell BK, McNeilly AS, Baird DT, 1992: Therole of inhibin and oestradiol in the control of FSHsecretion in the sheep. J Endocrinol 133, 381–391.McBride MW, Aughey E, O’Shaughnessy PJ, JeffcoateIA2001: Ovarian function and FSH receptor characteristicsduring canine anoestrus. J Reprod Fertil Suppl 57, 3–10.McCann JP, Altszuler N, Hampshire J, Concannon PW, 1987:Growth hormone, insulin, glucose, cortisol, luteinizinghormone, and diabetes in beagle bitches treated withmedroxyprogesterone acetate. Acta Endocrinol 116, 73–80.Meij BP, Mol JA, Hazewinkel HAW, Bevers MM, Rijnberk A,1996: Assessment of a combined anterior pituitary functiontest in beagle dogs: rapid sequential intravenous administrationof four hypothalamic releasing hormones. DomAnim Endocrinol 13, 161–170.Monniaux D, Huet C, Besnard N, Cle´ment F, Mariana JC,Bosc M, Monget P, 1997: Follicular growth and ovariandynamics in mammals. J Reprod Fertil Suppl 51, 3–23.Moyle WR, Campbell RK1995: Gonadotropins. In: AdashiEY, Rock JA, Rosenwaks Z(eds), Reproductive Endocrinology,Surgery and Technology. Lippincott-Raven Publishers,Philadelphia, pp. 683–724.Nicol L, McNeilly JR, Stridsberg M, McNeilly AS, 2004:Differential secretion of gonadotrophins: investigation ofthe role of secretogranin II and chromogranin A in therelease of LH and FSH in LbetaT2 cells. J Mol Endocrinol32, 467–480.Okkens AC, Bevers MM, Dieleman SJ, Willemse AH, 1985:Shortening of the interoestrus interval and the lifespan ofthe corpus luteum of the cyclic dog by bromocriptinetreatment. Vet Quart 7, 173–176.Okkens AC, Bevers MM, Dieleman SJ, Willemse AH, 1990:Evidence for prolactin as the main luteotrophic factor in thecyclic dog. Vet Quart 12, 193–203.Okkens AC, Kooistra HS, Dieleman SJ, Bevers MM, 1997:Dopamine agonistic effects as opposed to prolactin concentrationsin plasma as the influencing factor on duration ofanoestrus in bitches. J Reprod Fertil Suppl 51, 55–58.Olson PN, Bowen RA, Behrendt MD, Olson JD, Nett TM,1982: Concentrations of reproductive hormones in canineserum throughout late anestrus, proestrus and estrus. BiolReprod 27, 1196–1206.Onclin K, Verstegen JP1997: In vivo investigation of lutealfunction in dogs: effects of cabergoline, a dopamine agonist,and prolactin on progesterone secretion during mid-pregnancyand -diestrus. Dom Anim Endocrinol 14, 25–38.Onclin K, Verstegen J, Silva LDM, Concannon P1995:Patterns of circulating prolactin, LH and FSH duringdopamine-agonist induced termination of anestrus in beagledogs. Biol Reprod Suppl 52, 314.Onclin K, Lauwers F, Verstegen JP, 2001: FSH secretionpatterns during pregnant and nonpregnant luteal periodsand 24 h secretion patterns in male and female dogs. JReprod Fertil Suppl 57, 15–21.Park SK, Keenan MW, Selmanoff M, 1993: Graded hyperprolactinemiafirst suppresses LH pulse frequency and thepulse amplitude in castrated male rats. Neuroendocrinology58, 448–453.Romagnoli SE, Camillo F, Cela M, Johnston SD, Grassi F,Ferdeghini M, Aria G, 1993: Clinical use of prostaglandinF2a to induce early abortion in bitches: serum progesterone,treatment outcome and interval to subsequent oestrus. JReprod Fertil Suppl 47, 425–431.Rubion S, Desmoulins PO, Riviere-Godet E, Kinziger M,Salavert F, Rutten F, Flochlay-Sigognault A, DriancourtMA, 2006: Treatment with a subcutaneous GnRH agonistcontaining controlled release device reversibly preventspuberty in bitches. Theriogenology 66, 1651–1654.Russo IH, Russo J1991: Progestagens and mammary glanddevelopment: Differentiation versus carcinogenesis. ActaEndocrinol 125 (Suppl. 1) 7–12.Sauder SE, Frager M, Case GD, Kelch RP, Marshall JC, 1984:Abnormal patterns of pulsatile luteinizing hormone secretionin women with hyperprolactinemia and amenorrhea:responses to bromocriptine. J Clin Endocrinol Metab 59,941–948.Schaefers-Okkens AC1996: Ovaries. In: Rijnberk A(ed.),Clinical Endocrinology of Dogs and Cats. Kluwer AcademicPublishers, Dordrecht, pp. 131–156.Selman PJ, Mol JA, Rutteman GR, van Garderen E, van denIngh TSGAM, Rijnberk A1997: Effects of progestin administrationon the hypothalamic-pituitary-adrenal axis andglucose-homeostasis in dogs. J Reprod Fertil Suppl 51, 345–354.Shupnik MA, 1996: Gonadal hormone feedback on pituitarygonadotropin genes. Trends Endocrinol Metab 7, 272–276.Sokolowski JH, Zimbelman RG, 1974: Canine reproductioneffects of multiple treatments of medroxyprogesteroneacetate on reproductive organs of the bitch. Am J Vet Res35, 1285–1287.Spain CV, Scarlett JM, Houpt KA, 2004: Long-term risks andbenefits of early-age gonadectomy in dogs. J Am Vet MedAssoc 224, 380–386.Sung M, Armour AF, Wright PJ, 2006: The influence ofexogenous progestin on the occurrence of proestrous orestrous signs, plasma concentrations of luteinizing hormoneand estradiol in deslorelin (GnRH agonist) treated anestrousbitches. Theriogenology 66, 1513–1517.Tani H, Inaba T, Tamada H, Sawada T, Mori J, Torii R, 1996:Increasing gonadotropin-releasing hormone release by perifusedhypothalamus form early to late anestrus in the beaglebitch. Neurosci Lett 207, 1–4.Tani H, Inaba T, Matsuyama S, Takamor Y, Torii R, TakanoH, Tamada H, Sawada T, 1997: Enhancement of estrogenreceptor gene expression in the mediobasal hypothalamusduring anestrus in the beagle bitch. Neurosci Lett 227,149–152.Tani H, Inaba T, Nonami M, Natsuyama S, Takamor Y, ToriR, Tamada H, Kawate N, Sawada T, 1999: Increased LHpulse frequency and estrogen secretion associated withtermination of anestrus followed by enhancement of uterineestrogen receptor gene expression in the beagle bitch.Theriogenology 52, 593–607.Thrushfield MV, Holt PE, Muirhead RH, 1998: Acquiredurinary incontinence in bitches: its incidence and relationshipto neutering practices. J Small Anim Pract 39, 559–566.Trigg TE, Doyle AG, Walsh JD, Swangchan-uthai T, 2006: Areview of advances in the use of the GnRH agonistdeslorelin in control of reproduction. Theriogenology 66,1507–1512.Verstegen JP, Onclin K, Silva LD, Concannon PW, 1999:Effect of stage of anestrus on the induction of estrus by thedopamine agonist cabergoline in dogs. Theriogenology 51,597–611.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

164 J de Gier, NJ Beijerink, HS Kooistra and AC OkkensVickery BH, McRae GI, Goodpasture JC, Sanders LM, 1989:Use of potent LHRH analogues for chronic contraceptionand pregnancy termination in dogs. J Reprod Fertil Suppl39, 175–187.Vizcarra JA, Wettemann RP, Braden TD, Turzillo AM, NettTM, 1997: Effect of gonadotropin-releasing hormone(GnRH) pulse frequency on serum and pituitaryconcentrations of luteinizing hormone and follicle-stimulatinghormone, GnRH receptors, and messenger ribonucleicacid for gonadotropin subunits in cows. Endocrinology 138,594–601.Wright PJ, Verstegen JP, Onclin K, Jo¨ chle W, Armour AF,Martin GB, Trigg TE, 2001: Supression of the oestrusresponses in bitches to the GnRH analogue deslorelin byprogestin. J Reprod Fertil Suppl 57, 263–268.Yazigi RA, Wuintero CH, Salameh WA1997: Prolactindisorders. Fertil Steril 67, 215–225.Yu WH, Karanth S, Walczewska A, Sower SA, McCann SM,1997: A hypothalamic follicle-stimulating hormone-releasingdecapeptide in the rat. Proc Natl Acad Sci 94,9499–9503.Author’s address (for correspondence): Auke C. Okkens, Departmentof Clinical Sciences of Companion Animals, Faculty of VeterinaryMedicine, University of Utrecht, Yalelaan, Utrecht, The Netherlands.E-mail: a.c.schaefers-okkens@uu.nlConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 165–171 (2008); doi: 10.1111/j.1439-0531.2008.01157.xISSN 0936-6768The Ethics and Role of AI with Fresh and Frozen Semen in DogsGCW England 1 and KM Millar 21 School of Veterinary Medicine and Science; 2 Centre for Applied Bioethics, School of Biosciences, University of Nottingham, Nottingham, UKContentsThe use of artificial insemination (AI) with fresh semen hasresulted in many benefits for the management of dog breeding,but there are disadvantages that can sometimes be overlooked.Furthermore, poorer quality semen arising as a result ofcryopreservation necessitates uterine insemination, whichraises the potential for surgical insemination. A number ofsignificant ethical concerns have been raised by key stakeholders(such as The Kennel Club and the Royal College ofVeterinary Surgeons) about AI per se, but particularly aboutthe use of surgical insemination. This paper discusses thetechnological development of AI and explores a number of theethical issues raised by its application to dog breeding. AnEthical Matrix method is used to map the potential ethicalissues for key interest groups, namely dogs, breeders, owners,veterinarians and wider society. There are national variationsin the way in which institutions have evaluated potentialethical impacts, and this is reflected in the different regulatoryframeworks governing the use of AI in dogs. In order tofacilitate decision-making and reduce some of the ethical risksassociated with this technology, the veterinary research communitycould take several proactive steps including: (i)clarifying clinical decision-making processes, (ii) enhancinginformed choice among clients and (iii) increasing the knowledge-baseof potential impacts of AI.IntroductionArtificial insemination (AI) is the process of placingsemen into the female reproductive tract without anatural mating. AI has become popular as a result ofthe importing and exporting of frozen dog semen,however, semen can be inseminated fresh, or cooled andstored for a few days prior to insemination, or frozenand stored indefinitely before being thawed prior to use.The first specific report of mammalian AI was in a dogin the 1700s, and the first AI following storage of dogsemen was reported in the 1950s. In the ensuing50 years, there have been huge developments in reproductivebiology and biotechnology enabling the use,and potential misuse, of AI. There are national variationsin the ethical and legislative frameworks governingthe use of AI in dogs, and the purpose of this paperis to review the techniques and consider some of thepotential ethical questions posed by increasing use ofAI in this species.Role of Artificial Insemination in DogsAdvantages and disadvantages of AIThere is no doubt that AI with fresh semen has anumber of potential advantages in the management ofbreeding in dogs including that it: (i) may allow the useof males or females that are unable to breed because ofanatomical or pathological reasons, (ii) may overcomerefusal to breed because of psychological reasons, (iii)may allow the splitting of an ejaculate so that morefemales can be bred, (iv) is an acceptable method of linebreeding, (v) can be highly efficient at facilitating geneticimprovement, (vi) may allow the control of someinfectious diseases either by removing physicalcontact between animals or by allowing treatment ofsemen prior to insemination, (vii) enables examinationof semen quality prior to insemination and if necessarythe selection of an alternative stud, (viii) can be aconvenient and rapid method of breeding, and (ix) bycryopreservation enables conservation and storage ofvaluable genes from male animals almost indefinitely.When combined with semen preservation, there areadditional potential advantages including; (i) makingshipping of semen possible such that genetic material isavailable from outside of a breeding colony, (ii) overcomingquarantine restrictions, (iii) overcoming theneed to transport the animal (reducing transportationstress ⁄ disease risks), and (iv) the ability to utilize semenafter death of the male. For breeders producing dogs forprofessional purposes, such as bomb ⁄ mine detectiondogs, drug detection dogs and guide dogs for the blind,etc., semen freezing can enable castration of males at anearly age but maintaining availability of their geneswhile their performance can be evaluated. The need tokeep males in the breeding kennels is also reduced,which will reduce costs.Despite these clear benefits, the possible veterinarydisadvantages are often overlooked. These can include;(i) causing physical or psychological trauma during theAI process, (ii) undertaking AI for inappropriatereasons (e.g. where reluctance to breed is a manifestationof underlying hereditary disease such as hipdysplasia or anatomical abnormality of the reproductivetract), (iii) potential for introduction of heritablediseases or abnormalities, (iv) potentially allowingoveruse of a male within a programme or breed, and(v) possibly allowing confusion of parentage (for reviewsee Linde-Forsberg 2002). In all the cases, it would beprudent to ensure a controlled and certified processinvolving clinical and ⁄ or molecular examination ofmales to ensure that they do no spread genetic orinfectious disease.Technique of inseminationAt the time of natural mating, semen is deposited withinthe vagina but sperm are transported into the uterusbecause of significant vaginal and active uterine contractions(Evans 1933). While vaginal insemination[AI(V)] is a simple procedure that is well tolerated inthe bitch (Seager et al. 1975), vaginal contractions areÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

166 GCW England and KM Millarscant even when efforts are made to cause vaginaldistension, and active uterine contractions do notappear to develop at least not as to the same extent aswith natural mating (GCW England, unpublishedobservations). The lack of a bitch’s physiological responseat the time of AI results in relatively few spermbeing transported into the uterus and this situation isworse for cryopreserved semen, because the spermthemselves have poor motility. Furthermore, cryopreservedsperm have a short longevity compared withfreshly ejaculated sperm (Olar 1984). These two effectsprobably explain the relatively poor fertility observedwhen cryopreserved semen is inseminated into thevagina (Olar 1984).For these reasons, a number of techniques have beendeveloped to place the semen at the cervix or into theuterus. This is difficult in the bitch, because the vagina islong and narrow and the cervix is placed at anunusual angle (Lindsay 1983). However, Takeishi et al.(1976) reported success from intra-cervical insemination[AI(C)], and various commercial devices have beenproduced to achieve AI(C) whereby semen is forced intothe cervix using a special catheter, which forms a tightseal at the cranial vagina. The catheter can be left inplace to simulate a copulatory ‘tie’.Transcervical intrauterine insemination [AI(TC)] ispossible using a method described initially by Fougneret al. (1973). An outer catheter sheath is placed into thevagina, the cervix can then be realigned by palpatingthrough the abdominal wall and a central catheter isinserted through the cervix. Relatively few reports detailthe success rates in achieving AI(TC) or the incidence ofcomplications (England and Verstegen 1996); however,although overall the technique is thought to be minimallyinvasive, it is reported that the transabdominalpalpation is resented by a significant proportion ofbitches (Wilson 1993, 2001). The AI(TC) was furtherdeveloped by Wilson (1993) with a rigid endoscopicmethod and using a wire to guide a catheter through thecervix. Wilson (1993) reported that 97% of bitches couldbe relatively easily catheterized using this method, andwhile it is clear that significant training is required, todate no reports of adverse effects have been reported(Wilson 2001).A simple way to overcome the requirement fortraining and to ensure that uterine AI can be reliablyperformed is to undertake surgical intrauterine insemination[AI(S)] at laparotomy or laparoscopy (Smith1984; Wildt 1986). Clearly, these methods are invasiveand require general anaesthesia, and although there iswide applicability in some countries, especially the USA,the ethics of a surgical insemination has been the subjectof some debate (Royal College of Veterinary Surgeons(RCVS) 2005). Interestingly, there are no published dataon complications of surgical insemination.Success rates of artificial inseminationThere are many reports of pregnancy rates after AI, butwith a few notable exceptions (Linde-Forsberg andForsberg 1993; Thomassen et al. 2006), most involvevery small numbers of animals and lack adequatecontrol groups. Furthermore, variations in pregnancyrates may be the result of differences in timing ofinsemination, quality of the semen inseminated, site ofsemen deposition, number of inseminations and theinherent fertility of the female and the male (includingeffects of age). Linde-Forsberg and Forsberg (1993)developed a simple scoring scheme in an attempt toquantify these variables, but application of their methodto other published studies is difficult as many authorsfail to report these important factors. Nevertheless, it iscommonly agreed that; (i) fresh semen AI has a greatersuccess rate than cryopreserved semen AI regardless ofthe site of insemination, (ii) increasing the number ofsperm inseminated improves the success rate regardlessof the site of AI, (iii) multiple AIs have a greater successrate than single AIs, (iv) AI(TC) and AI(S) producehigher pregnancy rates than AI(V), especially whenusing frozen-thawed semen (see Fontbonne and Badinand1993; Linde-Forsberg and Forsberg 1993; Thomassenet al. 2006).The success with fresh semen AI depends upon itsquality and the fertility of the bitch, but pregnancy ratesare approximately 80 ± 16 (SD)% for AI(V) and97 ± 4 (SD)% for AI(TC) and AI(S). Reported pregnancyrates for chilled semen inseminations are onaverage 47 ± 9 (SD)% for AI(V) and 81 ± 19 (SD)%for AI(TC). Finally, pregnancy rates for frozen-thawedsemen are on average: 45 ± 24 (SD)% for AI(V);60 ± 15 (SD)% for AI(C); 70 ± 11 (SD)% for AI(TC);and 95 ± 7 (SD)% for AI(S) (data collatedfrom: Seager et al. 1975; Olar 1984; Smith 1984;Fontbonne and Badinand 1993; Linde-Forsberg andForsberg 1993; Wilson 2001; Thomassen et al. 2006). Itis important that these data are interpreted cautiouslybecause of significant variations in methodology and thesmall numbers of animals used in many of the studies.Conducting an Ethical AnalysisWith technological improvements and an increasinginterest in the use of AI(TC) and AI(S), a number ofsignificant ethical concerns have been raised (RCVS2005). One approach for exploring these ethical issuesis to conduct a structured ethical analysis. A numberof methods have been developed to facilitate ethicalanalysis and stakeholder engagement, including theEthical Matrix (EM) method (Mepham 2000). TheEM is applied to facilitate the assessment of a proposedstrategy (i.e. the use of reproductive technology) interms of respect (or lack of respect) for three ethicalprinciples; wellbeing, autonomy and fairness, as appliedto a defined set of interest groups. The ‘weight’ orsignificance assigned to each ethical impact is determinedby the evaluation of evidence. The EM methodhas been previously applied to a number of biotechnologycases (e.g. Mepham 2000; Mepham et al. 2006;Millar and Tomkins 2007). It should be noted that thismethod is not prescriptive and therefore will notproduce ‘an answer’, but the method can make theethically relevant issues transparent and thus facilitateinformed decision-making. The value of the approach isthat, it makes explicit the evidence used to justify aposition and encourages ethical reflection on the impactsfor all ethically relevant interest groups. The method canÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

The Ethics and Role of AI in Dogs 167Modified ethical matrix(Translation of the ethical principles for the corresponding interest group)WELLBEING AUTONOMY FAIRNESSDOGS(dog, bitch andpuppies)Welfare(DW)Behaviouralfreedom(DA)Intrinsicvalue(DF)BREEDERSSatisfactory incomeand workingconditions (BW)Managerialfreedom(BA)Fair regulations andtrade(BF)OWNERSSafety andquality of life(OW)Choice(OA)Affordability ofproducts(OF)VETERINARIANSSatisfactory incomeand workingconditions(VW)Professionalfreedom(VA)Equitable standardsof practice(VF)Fig. 1. Modified ethical matrixSOCIETYSafety and socialharmony(SW)Democratic choice(SA)Fair resourceallocation(SF)also act as a starting point for ethical deliberation inpublic policy decision-making.When considering the ethical issues raised by the useof AI in the dog, an adapted form of the Matrix methodcan be applied. This adapted EM (Fig. 1) will be appliedto map the issues raised and to analyse whether theapplication of AI approaches might infringe upon orsupport broadly-defined ethical principles for the specifiedinterest groups. Within the limits of this paper it isnot possible to comprehensively explore all of the ethicalissues raised by the various forms of AI; however, thismethod can be used to highlight some of the moreprominent ethical considerations and clarify whereconflicts may arise. It should be noted that ethicalanalysis will highlight differences between the differentforms of AI (with fresh or frozen semen).The modified Matrix, which will be applied here,incorporates five interest groups; dogs, breeders, owners,veterinarians and wider society. Here, breeders aredefined as those individuals responsible for the matingof the bitch and dog, whereas owners are the recipientsof the pups. In a number of cases, this can be the sameperson, but this distinction is important when consideringthe ethical issues raised.Ethical Analysis of the Use of AI by InterestGroupDogs (dog, bitch and puppies)Welfare (DW)It is necessary to consider two welfare aspects whenassessing the use of AI in dogs; firstly, what are thegeneric benefits and risks for the bitch or ⁄ and dog? andsecondly, what are the risks associated with the differentAI methods?Many organizations classify natural breeding as thepreferable means of producing a pregnancy. In order torespect the wellbeing particularly of the bitch, anydecision to intervene must be carried out in her bestinterests. Some benefits can be conferred from intervention;specifically, forced mating may be traumatic initself and can lead to physical and psychological harm.In such cases, the use of AI can respect animal wellbeingand prevent harm, yet, the selection of an alternativesexual partner would represent greater respect. Wellbeingmay be respected if the use of AI protects against thetransmission of infectious disease. However, it mightbe argued that the existence of a disease risk may bejustification in itself not to breed from these animals. Incountries where import regulations are strict andwhere the breeding population is very small, AI maybe necessary to prevent inbreeding, which may compromisehealth.It has been claimed that the bitch gains no welfarebenefit from becoming pregnant. However, beyond abehavioural need, some authors contend that bitchesmay gain a number of physiological benefits frompregnancy and lactation such as reducing the risk ofdeveloping pyometra and mammary neoplasia, althoughevidence to support these assertions is very anecdotal.Even if reproductive intervention is not considered inthe best interest of the dog, intervention may representonly a neutral or very minor infringement of wellbeing(although any intervention must be explained andclearly justified). It is, therefore, important to ascertainthe documented impact on animal wellbeing and theperceived risks from the three forms of AI.Complications from AI appear to be rare and thewelfare risks associated with AI(V) often relate only tothe need to physically restrain the bitch in order toinseminate. It might be suggested that any form ofuterine AI has a greater welfare risk than AI(V). Of themethods available, AI(TC) is by its nature a reducedwelfare burden than AI(S). However, the use ofAI(TC) is seen as a procedure requiring a skilledoperator and in inexperienced hands may result in localÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

168 GCW England and KM Millartrauma or a prolonged and potentially stressful procedure.The AI(S) is an infringement of wellbeing even ifthe procedure is conducted without complications.Justification for the use of AI(S) is that, adverse effectsare minimal and therefore the welfare risk is equallysmall; however, there is limited data to support orrefute this statement. Some have also claimed thatAI(S) may be less stressful to the animal than a difficultAI(TC), and that AI(TC) may not be possible in somesmaller breeds as well as in some large or obese bitches;however, both these difficulties may be due to operatorinexperience or inappropriate equipment (i.e. endoscopesize). The use of low-dose sedatives, may reducehandling stress.When examining the risks of AI and the animalwelfare burden, the use of non-surgical approaches,namely AI(V) and AI(TC), respectively, appear torepresent far smaller welfare risks than AI(S).Behavioural freedom (DA)When AI is used over natural mating, dogs and bitchesare denied the ability to engage in normal sexualbehaviour (a negative infringement). Studies have indicatedthat some bitches can be highly selective in theirmating choices. If breeder intervention occurs, eitherforced mating or insemination removes the ability toself-determine the mate and this infringes the instinctivebehavioural pattern. In addition, forced mating can betraumatic for an unreceptive bitch. Continued attemptsto mate could reinforce a negative behavioural experience.In these circumstances, the use of AI would be theonly option for breeding.Intrinsic value (DF)When considering whether the use of AI may infringeor respect the notion of intrinsic value of the animal, itis important to consider whether the application ofthese technologies is potentially increasing the objectificationof the dog or bitch. It could be argued thatthe need to produce pups results in the dog being seenmerely as a ‘reproductive vessel’. This may in turndetract from the relationship between dog and owner⁄ breeder and deny the dog respect as an animal thathas value in its own right, beyond its instrumentalvalue, for example, as a working dog or breeding dog.However, it may be argued that the dog has an innatedrive to reproduce and raise young and that bydenying the bitch the ability to do so would be aninfringement of her ‘telos’ or innate purpose (linked torespect for behavioural freedom). There is no indicationfrom the literature or from anedoctal experienceby dog breeders that, bitches that are artificiallyinseminated as opposed to naturally mated femalesshow less maternal behaviour towards their pups.BreedersSatisfactory income and working conditions (BW)Breeders are not a homogenous group; many breeders,particularly in the UK, are individuals who operate ona non-commercial basis and are driven by theiradmiration of a specific breed. In contrast, a numberof breeders focus on economic factors and a need toproduce pedigree dogs for sale. However, when consideringthe conditions that drive requests to veterinariansto use AI, a number of overarching aspects can beidentified for all breeders as well as specific aspects thatrelate exclusively to commercially-oriented breeders.The availability of assisted reproductive technologies(ARTs) allows breeders to manage their breedingprogrammes and maximize fertility rates for their bestbitches. The availability of AI prevents breeding delaysand should not reduce litter size, which might prohibituse. In addition, breeders may wish to introducebreeding programmes for animals that have exaggeratedphysical or temperamental features, which canmake natural mating difficult. However, this may resultin breeders sustaining problem behaviour or undesirablephysical features within a line. The use of AIcould, therefore, inadvertently perpetuate the selectionof undesirable traits within a breed which wouldnegatively impact on breeders’ wellbeing as well asbeing detrimental for the breed. In contrast, thetechnology could allow the introduction of desirabletraits and ensure the preservation of endangered breedswith the secondary benefit of potentially enhancing thewellbeing of breeders.Managerial freedom (BA)Breeders are free to apply and optimize all availablereproductive technologies. Some breeders may have nooption other than to request AI, as quality breedingdogs may only realistically be available from otherbreeders outside of their geographic region. AI alsoallows breeders to use those dogs that are not accustomedto mating bitches (e.g. inexperienced dogs) or thedogs that are rejected by the bitch. The technology canensure that these quality animals are still able to breedand produce litters. However, the free availability of thistechnology may result in breeders opting for AI forvirgin dogs and bitches, rather than investing time intraining these animals to mate naturally. This may leadto an ‘intervention treadmill’ where AI is alwaysrequested for high value animals to ensure timelypregnancies.Fair regulations and trade (BF)The increasing availability of new reproductive technologiesmay undermine the ability of smaller breeders totrade in an increasingly competitive environment, asonly large commercial breeders will have the financialmeans to access these technologies. This differencebetween commercial and leisure breeders may be positivelyreinforcing over time and could eventually changethe composition of the breeding community. However,this argument could be advanced for all new innovationssuch as nutritional advances, veterinary treatment, etc.,therefore representing a minor or neutral infringementof the principle of justice for all breeders. The breedingof certain dogs, particularly sporting dogs and servicedogs, may be a highly commercial enterprise, andtherefore, improving breeding efficiency is a naturalÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

The Ethics and Role of AI in Dogs 169progression for an industry that is developing within amarket economy.Owners (recipients of pups)Safety and quality of life (OW)Using AI may reduce the risk of disease transmissionthat will not only benefit the dog but could also havepotentially positive impacts on the puppies’ owner(s).Use of AI may ensure a high-quality supply of pupsand allow owners to receive a pup on demand, ratherthan being affected by breeding difficulties. However,AI use with problem dogs (male and female) mayresult in progeny that are unsuitable as companionanimals (e.g. behavioural problems) or which will notreproduce themselves without intervention (e.g. anatomicalor physiological limitations). The overuse ofmales through AI may also facilitate the incorporationof undesirable traits within a line that are notapparent until later generations. Any form of ARTuse may be seen by some to infringe the dog–ownerrelationship. The risks to animal welfare associatedwith the use of AI and the perceived view of theunnaturalness of the process may be seen to have anegative impact on the companion animal–humanrelationship and, hence, infringe the wellbeing of theowner.Choice (OA)In order to respect the principle of autonomy forpotential owners, it is important that all techniquesand procedures applied to produce pups are disclosed.This will allow owners to make an informed decisionabout whether they want a pup that has beenconceived using AI, particularly AI(S); as this maybe an important differentiation for some owners.Potential owners should also be privy to the justificationfor using AI, i.e. such as for behavioural oranatomical reasons. In order to respect the autonomyof potential owners, this information should be loggedin the breeding record and should be offered (i.e. thisis a positive duty) to owners rather than being madeavailable on request. The use of AI may compromisethe reliability of the parentage information as semenmay be ‘mixed’ or the origin may be difficult to verify.However, this risk could be managed through currentsystems such as breeder certification and veterinarysupervision of AI (e.g. the use of stud books andmandatory DNA sampling at the time of semencollection etc.).Affordability of products (OF)Recently, the International Association of Human-Animal Interaction Organizations declared that ‘it is auniversal, natural and basic human right to benefit fromthe presence of animals’. The introduction of AI technologiescould result in a number of specialized breedsbeing prohibitively expensive for some members ofsociety. If that was the case, an increase in the use of thistechnology within the dog-breeding community may beseen as an infringement, and therefore, unfair to someeconomically disadvantaged members of the dog-owningcommunity.VeterinariansSatisfactory income and working conditions (VW)Unlike human health professionals where services aresupported via public funding, veterinarians operate in amarket environment where the quality and diversity oftheir services influence their income streams. Veterinarianswho invest in new skills and are able to offeradditional services can enhance the profitability of theirpractice and their personal income. In a marketenvironment, this can result in a competitive advantageover rival practices. The ability to respond to clients’needs and offer new services, such as AI, enhances(respects) veterinarians’ wellbeing.Professional freedom (VA)Freedom to innovate is an important driver of change inmany fields and this is no less the case for the veterinaryprofession. By developing and applying new diagnosticmethods, surgical techniques and veterinary products,veterinarians have improved the wellbeing of theirpatients as well as their clients (e.g. dog owners). Theopportunity to use AI allows veterinarians to determinethe best course of treatment for their patients sorespecting their professional autonomy, particularly ifthe cause of the bitches’ infertility is a barrier that canonly be overcome by AI. Artificial insemination mayprovide the only option for treating dogs that arecompromised because of anatomical, physiological orbehavioural problems.Equitable standards of practice (VF)It may be claimed that the availability of noveltechnologies, which are cost-effective but perhaps poserisks for animal welfare (e.g. surgical risks), couldunduly influence the market and decrease the use ofother non-invasive techniques that may require additionalskills or be more time-demanding. In order toensure the principle of justice is not infringed, adequateinformation would need to be provided to breeders andowners on the options for treatment. This could includeadvising against the use of an intervention method orpossibly even advising against breeding from a particulardog. In addition, when considering the use of thedifferent AI methods, the risks of each technique shouldbe clearly stated and, if necessary, referral to otherqualified veterinarians should be offered. This wouldensure that trading and professional standards were notunduly influenced by lack of transparency in the market.SocietySafety and social harmony (SW)The routine use of reproductive technologies may affectthe human–dog relationship by increasing society’sinstrumental view of companion animals. However,the clinical ability to intervene and facilitate a successfulpregnancy may further enhance the positive nature ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

170 GCW England and KM Millarthe animal–human bond. The impact on society’swellbeing will be influenced by the justification putforward for using the technique and whether the stated‘reasons’ enhance or infringe the human–dog relationship.Democratic choice (SA)The development and application of these techniquesmay result in the ability to conserve valuable genetictraits for the canine species and this could be seen as apositive impact. This will ensure society’s ability tochoose the breed traits that it wishes to preserve andpotentially allow future generations a degree of autonomythat may not have been possible without theapplication of these technologies. However, this argumentis relevant to the use of the technique only whenapplied to specific breeding strategies, rather than forroutine treatment purposes.Fair resource allocation (SF)Because the financial cost of using any form of AI willbe born by the breeder and the potential owner of thenew pup, the impact for the principle of justice forsociety will be neutral.Ethical Evaluation and DiscussionAlthough, because of the limits of this paper, the ethicalanalysis has only initially mapped the possible impactsof the application of AI technologies in dogs, theanalysis has highlighted a number of important issues.In order to come to a final judgement or position, theimpacts for the various interest groups need to beweighed against each other (ethical evaluation). It is thisweighing that can help identify the key areas ofdisagreement or value conflict between stakeholders.This process can also help identify knowledge gaps andareas of uncertainty. For some groups, the use of ascoring system (e.g. +1, 0, )1) can aid this weighingprocess, but it should be noted that the EM is a decisionsupportframework and not a decision-making tool. Theuse of numerical weighing in ethical evaluation has itslimitations.It is clear from the analysis that the use of AI mayinfringe ethical principles, particularly for the affectedanimal, but that it may respect other principles forbreeders, owners and veterinarians. Any veterinaryprocedure has a risk of infringing an animal’s wellbeingand autonomy as well as potentially being unjust.However, veterinarian intervention is repeatedly justifiedon the basis of clinical ‘need’ (i.e. inflicting acutepain for long-term benefit). One of the key issues in thedebate that surrounds the use of AI is the interpretationof ‘need’, as any ‘unnecessary procedure’ would representan infringement of the dog’s wellbeing and wouldnot be an ethically acceptable intervention under themajority of EU veterinary codes of practice. This impliesthat before AI is considered appropriate a comprehensivereproductive assessment must be carried out by aveterinarian. It also implies that a sequential approachshould be applied with the justification for each decisionstep [i.e. ruling out natural service and proceeding to theuse of AI(V)]. This information should then beproactively offered to future puppy owners and kennelclubs, etc.While the use of AI can result in a number of positiveethical impacts (for example for disease control orpreservation of genetic material), which outweigh potentialrisks, such benefits appears to be predicated onfour conditions: (i) a sequential decision-making process,which ensures that the use of AI is applied after naturalmating options are ruled out for clinical reasons, (ii)informed choice for breeder and owner is ensuredthrough proactive information provision and appropriaterecord keeping, (iii) veterinarian competence in use ofAI technologies is ensured, and (iv) the welfare consequencesfor the bitch are measured as negligible.If AI per se is acceptable under these conditions, italso appears that the judgment on whether surgical AI isacceptable (when all other options are excluded) ismodulated by evidence on (i) the incidence and nature ofthe reproductive failure and (ii) welfare consequencesfor the bitch. What appears to confound the assessmentof these two issues is a paucity of data. The literaturecontains very few reports on the consequences for thebitch and there are even fewer studies that record thefrequency of use and the form of the decision-makingprocess. There is a significant need for peer-reviewedevidence in these areas in order to facilitate an informedethical evaluation.Without explicitly conducting an ethical analysis, anumber of national bodies and professional organizationshave articulated their approach to ‘weighing’ theconflicting impacts and set out their ethical positions(e.g. through guidelines, regulations, etc.). It is usefulhere to examine a few of these positions and to reflect onthe risks associated with these strategies. In the UK, theRoyal College of Veterinary Surgeons (RCVS) set outtheir position (RCVS, 2005), stating that, although theuse of surgical AI ‘is unlikely to be carried out in the bestinterests of any particular dog’, veterinary surgeons mayperform the procedure when justified (e.g. ‘for example,the incorporation of new genetic traits’). The RCVSindicated that the reasons for not using other approaches,e.g. transcervical insemination, should berecorded. The RCVS approach is more conservativethan that applied in the USA where AI(S) is morecommonly used. It is important to note that the RCVS’sethical position is based on the assumption that, inexceptional circumstances, the potential welfare consequencesfor the dog are acceptable only if a sequentialdecision-making process and good record keepingoccur. However, no explicit advice is given on ensuringinformed choice [i.e. information provision on theavailability of AI(TC)] or whether the ‘recorded justification’for the procedure will be audited and reviewedby the RCVS. Because the surgical procedure is invasive,there may also be a proactive duty of care for theattending veterinarian to monitor and report on thewelfare outcomes for the bitch. These three aspects mayrepresent a significance risk to the RCVS’s principledapproach (ethical position) on the use of AI(S).Some EU countries, for example, Sweden, haveprohibited the use of AI(S). In the light of uncertaintyÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

The Ethics and Role of AI in Dogs 171regarding the welfare consequences for the dog, thesecountries appear to have applied a much moreprecautionary approach and have advanced theirresponsibility to the dog (duty of care) over theautonomy of breeders and veterinarians. The decisionalso appears to be predicated on the availability ofviable alternatives. Although some clinicians haveargued that this type of ‘ban’ is unworkable, as theavailability of viable alternatives is extremely limited(i.e. there are few surgeons who can perform theAI(TC) procedure successfully), others have arguedthat such a position may enhance veterinarianwellbeing and autonomy in the long-term by encouraging:(i) innovation through research (i.e. developmentof new techniques and enhanced semenpreservation) and (ii) personal investment in inseminationtraining.In conclusion, the use of AI raises some importantethical questions which cannot be fully analysed andevaluated here, however, the veterinary and researchcommunity can take several proactive steps to reducethe ethical risks associated with this reproductivetechnology by: (i) clarifying clinical decision-making,(ii) enhancing informed choice among clients and (iii)increasing the knowledge-base on potential impacts andthe use of all methods of AI.AcknowledgementsThe authors would like to thank Prof. Catharina Linde Forsberg(Swedish University of Agricultural Sciences) and Sandy Tomkins(University of Nottingham) for their provision of valuable literaturethat aided the development of this paper.ReferencesEngland GCW, Verstegen JP, 1996: Radiographic contrastmedium for uterine insemination in the bitch, and its effectupon the quality and fertility of fresh dog semen. Theriogenology46, 1233–1241.Evans EI, 1933: The transport of spermatozoa in the dog. AmJ Physiol 105, 287–293.Fontbonne A, Badinand F, 1993: Canine artificial inseminationwith frozen semen: comparison of intravaginal andintrauterine deposition of semen. J Reprod Fertil Suppl 47,325–327.Fougner JA, Aamdal J, Andersen K, 1973: Intrauterineinsemination with frozen semen in the Blue Fox. NordicVet Med 25, 144–149.Linde-Forsberg C, 2002: Ethical aspects of artificial insemination(AI). In: Proceedings of the 3rd EVSSAR Congress,Liege, Belgium, pp. 41–42.Linde-Forsberg C, Forsberg M, 1993: Results of 527 controlledartificial inseminations in dogs. J Reprod FertilSuppl 47, 313–323.Lindsay FEF, 1983: The normal endoscopic appearance of thecaudal reproductive tract of the cyclic and non-cyclic bitch:post-uterine endoscopy. J Small Anim Pract 24, 1–15.Mepham TB, 2000: A framework for the ethical evaluation ofnovel foods: the Ethical Matrix. J Agric Environ Ethics 12,165–176.Mepham TB, Kaiser M, Thorstensen E, Tomkins S, Millar K,2006: Ethical Matrix Manual. Agricultural EconomicsResearch Institute (LEI), The Netherlands.Millar K, Tomkins S, 2007: Ethical analysis of the use of GMin aquaculture: emerging issues for aquaculture development.J Agric Environ Ethics 20, 437–453.Olar TT, 1984: Cryopreservation of dog spermatozoa. PhDThesis, Colorado State University.Royal College of Veterinary Surgeons (RCVS), 2005: AdviceNote 8 – Canine Surgical Artificial Insemination. RCVS,London.Seager SWJ, Platz C, Fletcher WS, 1975: Conception rates andrelated data using frozen dog semen. J Reprod Fertil 45,189–192.Smith FO, 1984: Cryopreservation of canine semen: techniqueand performance. PhD Thesis, University of Minnesota.Takeishi M, Mikami T, Kodama Y, Tsunekane T, Iwaki T,1976: Studies on reproduction in the dog. VII Artificial inseminationusing frozen semen. Jpn J Anim Reprod 22, 28–33.Thomassen R, Sanson G, Krogenaes A, Fougner J, Berg K,Farstad W, 2006: Artificial insemination with frozen semenin dogs: a retrospective study of 10 years using non-surgicalapproach. Theriogenology 66, 1645–1650.Wildt DE, 1986: Diagnostic procedures, laparoscopy. In:Burke TJ (ed.), Small Animal Reproduction and Infertility,A Clinical Approach to Diagnosis and Treatment. Lea &Febiger, Philadelphia, pp. 121–140.Wilson MS, 1993: Non-surgical intrauterine artificial inseminationin bitches using frozen semen. J Reprod Fertil Suppl47, 307–311.Wilson MS, 2001: Transcervical insemination techniques in thebitch. Vet Clin North America 31, 291–303.Author’s address (for correspondence): Gary CW England, Dean,School of Veterinary Medicine and Science, The University ofNottingham, Sutton Bonington Campus, Loughborough, LE125RD, UK. E-mail: Gary.England@nottingham.ac.ukConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 172–178 (2008); doi: 10.1111/j.1439-0531.2008.01158.xISSN 0936-6768Basic Aspects of the Control of GnRH and LH Secretions by Kisspeptin: PotentialApplications for Better Control of Fertility in FemalesA Caraty and I FranceschiniPhysiologie de la reproduction et des comportements, INRA, Nouzilly, FranceContentsThe neuronal control of fertility and sterility has been a subjectof research for years. However, nowadays, in spite ofconsiderable literature about GnRH during the last fewdecades, the precise cellular and molecular mechanismswhereby gonadal steroids and other peripheral signals convergein the brain to achieve the fine regulation of GnRHsecretion remains partially unknown. In this scenario, a majorbreakthrough in our understanding of the neuronal signalsgoverning reproduction took place in 2003 with the discoveryof metastin ⁄ kisspeptin as a major player in the control ofGnRH secretion. This molecule, first described as having acrucial role in triggering the onset of puberty, is involved in allphases of reproductive life and hence has attracted the interestof many reproductive neuroendocrinologists. Administeredeither centrally or peripherally, kisspeptin strongly induces thesecretion of gonadotropin in many species, mainly throughstimulation of GnRH secretion. Kisspeptin cells involved inthe control of GnRH secretion are located in two regions ofthe brain: the preoptic area and the arcuate nucleus. Carryingoestradiol receptor alpha, kisspeptin cells of these regionsappear to be the main integration centres for the expression ofboth the positive and negative feedback of steroid on GnRHsecretion. More recently, this molecule has been shown to beable to synchronize preovulatory surges in cyclic ewes andcause ovulation in seasonally acyclic ewes. This reviewsummarizes the most relevant aspects of the role of kisspeptinin GnRH ⁄ LH release and the potential application of thismolecule in new strategies for controlling female fertility.IntroductionSuccessful reproduction requires the completion of acomplex cascade of events, which are the substrate of a‘fine dialog’ between the brain, the pituitary and thegonads. Early studies on the regulation of the hypothalamo-pituitary-gonadotropicaxis have emphasizedthe pivotal role of pulsatile GnRH release fromhypothalamic neurons for the secretion of the gonadotropinsLH and FSH by the pituitary, and then theproduction of sexual hormones and gametogenesis bythe gonads (see review by Knobil 2005). Followingintense research, our knowledge of GnRH regulationhas greatly improved in recent years, but the neuroendocrinemechanisms leading to the fine modulationof GnRH pulsatility remain to be defined. Forexample, in most if not all mammals, it is clear thatoestradiol induces a strong increase in GnRH releaseduring the follicular phase of the cycle, leading to apreovulatory GnRH surge (rat: Sarkar et al. 1976; ewe:Clarke et al. 1989; Moenter et al. 1991; monkey: Xiaet al. 1992), but the pathways and the hierarchy of themechanisms involved in this regulation are onlypartially understood.In 2003, this field of research was given new impetusby a conceptual breakthrough, with the discovery of anew central player, the Kiss-1 ⁄ GPR54 system, operatingupstream of the GnRH structure. The unsuspected linkbetween Kiss-1 ⁄ GPR54 and reproductive physiologywas identified in two independent studies, reporting thatisolated hypogonadotropic hypogonadism in humans isassociated with a genetic defect of the GPR54 gene, andthat the loss-of-function by a mutation of the GPR54gene in mice leads to a deficiency in sexual maturationand hence infertility (de Roux et al. 2003; Seminaraet al. 2003). This discovery of a link between GPR54(a G-protein-coupled receptor) and sexual maturationled to immediate interest in its ligand. This molecule,first identified as a metastasis suppressor molecule, wasnamed metastin or kisspeptin (Kotani et al. 2001; Muiret al. 2001; Ohtaki et al. 2001). Kisspeptin is the productof the Kiss-1 gene, which encodes a 145-amino acidpeptide that is further processed to generate biologicallyactive peptides of various lengths (10–54 amino acids)termed kisspeptins. These appear to be well conserved indifferent species, with only minor structural changes forthe last 10 amino acids of the C-terminal end, whichbear nearly all the biological activity of the moleculein vivo (Gottsch et al. 2004) and in vitro (Orsini et al.2007). Kisspeptins bind to a receptor previouslydescribed as a galanin-related orphan receptor, GPR54(Lee et al. 1999), and display similar high-affinitybinding in heterelogous cell systems (Kotani et al.2001). There have been considerable advances in kisspeptinresearch, and this peptide family is now recognizedas a major player in the control of thehypothalamic-pituitary-gonadal axis.Kisspeptins Stimulate GnRH⁄LH ReleaseThe very potent stimulatory effect of kisspeptin ongonadotropin secretion is unequivocal. First, kisspeptinwas shown to elicit the release of gonadotropin wheninjected into the cerebral ventricle of the mouse brain(Gottsch et al. 2004). Using either intracerebroventricularor peripheral injections (intravenous, subcutaneous…) of kisspeptin, this finding was widely confirmed inmany species: the rat (Matsui et al. 2004; Navarro et al.2005), sheep (Messager et al. 2005), pig (Lents and Barb2007), bovine (Kadokawa et al. 2008a), monkeys (Shahabet al. 2005) and humans (Dhillo et al. 2005). Thestimulatory effect of kisspeptin on gonadotropin secretionis mostly mediated by a stimulatory effect onGnRH release, as evidenced in the following studies: (1)local administration of kisspeptin near the GnRH cellÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Control of Fertility in Females by Kisspeptin 173Fig. 1. Effect of central kisspeptin administration (50 nmol given as acontinuous administration into the lateral ventricle over 4 h) onGnRH release into the cerebrospinal fluid of the third ventricle (filledsquare) and LH release in the peripheral blood (open lozenge) of anovariectomized oestradiol-treated ewe during the anoestrous season.Redrawn from Messager et al. (2005)bodies in the medial preoptic area stimulated LH releasein the rat (Patterson et al. 2006); (2) intracerebroventricularadministration of kisspeptin in sheep induced alarge and sustained release of GnRH into the cerebrospinalfluid (Messager et al. 2005; Fig. 1); and (3) thestimulatory effect on gonadotropin secretion has beenfound to be totally blocked in the presence of a GnRHantagonist (Gottsch et al. 2004; Irwig et al. 2004;Matsui et al. 2004). This hypothesis is further supportedby evidence that a large proportion of GnRH neuronsexpress GPR54 (Irwig et al. 2004; Han et al. 2005), thatkisspeptin-immunoreactive fibres have been observed inclose apposition to GnRH cell bodies (Kinoshita et al.2005; Clarkson and Herbison 2006), and that administrationof kisspeptin increases excitability of GnRHneurons (Han et al. 2005).While the action of kisspeptin in regulating GnRHrelease at the level of the hypothalamus is unequivocal,the possible role of this peptide at the pituitary levelremains a matter of controversy. Kisspeptin has beenshown to stimulate LH and FSH release from ratpituitary explants (Navarro et al. 2005), LH releasefrom bovine and porcine pituitary cells (Suzuki et al.2007), and dose-related LH and growth hormonesecretory responses from dispersed pituitary cells (rat:Gutierrez-Pascual et al. 2007; bovine: Kadokawa et al.2008b). In contrast, other studies have found no effect ofkisspeptin on pituitary response (Matsui et al. 2004;Thompson et al. 2004) or in altering GnRH-stimulatedLH secretion in hypothalamo-pituitary disconnectedewes (Smith et al. 2007a). On a morphological basis,GPR54 is expressed in the human pituitary (Kotaniet al. 2001; Muir et al. 2001), and kisspeptin-immunoreactivefibres are located in the external zone of themedian eminence in sheep (Franceschini et al. 2006;Pompolo et al. 2006). Moreover, as kisspeptin isreleased into the hypophyseal portal blood (Smith et al.2007a), it is more than likely that kisspeptin has a role atthe pituitary level, even if this is not presently known. Ina recent study, peripheral administration of kisspeptinwas clearly demonstrated to lead to a differentialresponse in terms of LH and FSH release (Caraty et al.2007). After rapid initial increase in both LH and FSHrelease in response to the kisspeptin-induced GnRHsecretion, plasma FSH levels were found to remain highFig. 2. Effect of peripheral administration of kisspeptin (100 nmolgiven iv) on LH (filled lozenge) and FSH release (open square) in theperipheral blood of an ovariectomized oestradiol-treated ewe duringthe anoestrous season. LH and FSH concentration increase within10 min and then begin to decrease according to the half-life of eachhormone. After 2 h, FSH reaches a plateau and stays above the preinjectionlevel until the end of the sampling period. Redrawn fromCaraty et al. (2007)for several hours (Fig. 2), suggesting a secondary effectof the peptide at the gonadotrope level enhancingconstitutive FSH release. Further studies are clearlyneeded to clarify whether and how the pituitarycontributes to the full gonadotropin response to kisspeptinadministration.Kisspeptin Neurons in the Brain, Distributionand Regulation per Sex SteroidsIn the mouse, Kiss-1 mRNA is found in the anteroventralperiventricular nucleus (AVPV), the periventricularnucleus, the anterodorsal preoptic nucleus, themedial amygdala and the arcuate nucleus (ARC)(Gottsch et al. 2004). In the ewe, the distribution ofthe Kiss-1 mRNA cells is similar, with a densepopulation in the ARC and an additional populationof positive cells in the POA (Smith et al. 2007b). Inline with these mRNA data, immunocytochemistryusing an antiserum raised against the amino acidresidues 43–52 of mouse kisspeptin (Franceschini et al.2006) identified kisspeptin-immunoreactive cells in theAVPV and ARC in the mouse (Clarkson and Herbison2006) and in the POA and ARC in the ewe(Franceschini et al. 2006; Fig. 3). In the mouse andsheep, a few immunoreactive cells for kisspeptin havealso been localized in the dorsomedial hypothalamicnucleus, an area devoid of Kiss-1 mRNA (Smith et al.2007b). This may indicate a slight cross reactivity ofthe antisera with other members of the relatedC-terminal RF amide family.Kiss-1 mRNA expression appears to be stronglyregulated by sex steroids. In the rat, mouse and ewe,steroid removal by ovariectomy greatly increasesmRNA expression in the ARC, an effect reversed byoestradiol replacement (Smith et al. 2005, 2006, 2007b;Maeda et al. 2007). Accordingly, Kiss-1 mRNA expressionis found to be lower at proestrus and oestrus thanat dioestrus 1 (Navarro et al. 2004). This picture differsclearly from what happens at the POA level. A decreaseof Kiss-1 mRNA expression in the AVPV occurs afterovariectomy in the female rat and mouse, an effectwhich is reversed by oestradiol administration (SmithÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

174 A Caraty and I FranceschiniFig. 3. Two main populations of kisspeptin cells are found in thehypothalamus of the ewe: one is located into the preoptic area (POA)and the other one is located in the arcuate nucleus (ARC). Cells in thePOA are distributed close to the wall of the third ventricle (V). Thedensity of the cells is higher in the caudal part of the ARC, where theyare surrounded by a dense network of varicose fibres. Redrawn fromFranceschini et al. (2006)et al. 2005; Maeda et al. 2007). In contrast, no similarchanges are observed in the POA of the ewe (Smith et al.2007b).With regard to the regulation of peptide expression,few data are available to date. In sheep, ovariectomygreatly increases the number of kisspeptin cells of theARC (Pompolo et al. 2006), which is in line with in situhybridization studies. Similarly in rat, the changes innumber of kisspeptin cells observed in the AVPV andthe ARC by in situ hybridization between ovariectomizedand ovariectomized oestradiol-treated females aremirrored by similar changes in the number of cellsobserved by immunocytochemistry (Maeda et al. 2007).Yet, the changes in protein are not always parallel withthe changes in mRNA expression of kisspeptin. Forexample, Kiss-1 mRNA level is lower at proestrus andoestrus than at dioestrus 1 while the number ofkisspeptin neurons as well as the number of kisspeptinneurons expressing Fos are higher at proestrus than atdioestrus (Kinoshita et al. 2005).As sex steroids have a profound impact on Kiss-1mRNA expression, it is not surprising that most, if notall, kisspeptin cells have been found to express oestradiolreceptor alpha (ERa) in the mouse and rat (Smithet al. 2005, 2006). In the ewe, the majority of kisspeptincells of the ARC express ERa (Franceschini et al. 2006)as well as the progesterone receptor (Smith et al. 2007b).In contrast to rodents, only 50% of the kisspeptin cellsexpress ERa in the POA, suggesting that functionallydistinct subpopulations might exist in this species(Fig. 3), and this could be related to the positivefeedback effect of oestradiol on GnRH secretion (anissue addressed in the following sections).Kisspeptin and the Regulation of FemaleCyclicityThere is compelling evidence that the AVPV kisspeptincell population in the rat and mouse is critical forgenerating the preovulatory GnRH surge. The role ofAVPV in the induction of a GnRH ⁄ LH surge in rodentsis unequivocal (see review of Herbison 2007). Thekisspeptin cell population of this nucleus is highlysexually dimorphic. Both kisspeptin proteins and Kiss-1mRNA levels are higher in the female than the male(Clarkson and Herbison 2006; Gottsch et al. 2006;Kauffman et al. 2007). If female rats are neonatallytreated with testosterone during the critical period ofbrain differentiation, the number of Kiss-1 mRNA inthis nucleus is low, similar to that of male rats(Kauffman et al. 2007). In contrast, no similar sexuallydimorphic expression is observed for the kisspeptinpopulation of the ARC. Lesions of the AVPV block theLH surge, as does immunoneutralization of kisspeptinin the POA by a specific monoclonal antibody (Kinoshitaet al. 2005). More importantly, expression of Kiss-1mRNA is up-regulated concomitantly with the preovulatoryGnRH ⁄ LH surge (Smith et al. 2006), and amarked increase in c-Fos expression occurs at that timein these cells, proving transcriptional activation (Smithet al. 2006). The simplest and most plausible conclusionthat can be drawn from these data is that kisspeptinneurons of the AVPV, bearing ERa and having directcontact with GnRH cell bodies in the POA, are a majortarget for oestradiol to provoke preovulatoryGnRH ⁄ LH release. Unfortunately, the picture is notso simple. First, it is important to bear in mind thatthere is a circadian component in the induction of theGnRH ⁄ LH surge in rodents and that circadian signalsfrom the suprachiasmatic nucleus form additionalregulatory elements of the kisspeptin cells (Gu andSimerly 1997). Secondly, in ovariectomized oestradioltreatedfemale rats, Kiss-1 mRNA in AVPV is high,while LH is strongly suppressed (Smith et al. 2005).Finally, it has recently been shown that oestradiolinduces Fos expression in GnRH neurons and producesa GnRH-dependent LH surge in GPR54 KO mice(Dungan et al. 2007). This latter result may reflect acompensatory mechanism in GPR54 KO mice. Nevertheless,this finding questions the role of the kisspeptincells of the AVPV in the oestradiol-induced GnRHsurge mechanism.The picture is different for ewes, because oestradiolacts to induce a preovulatory GnRH ⁄ LH surge in themedio-basal hypothalamus and not in the POA (Blacheet al. 1991; Caraty et al. 1998). As mentioned above,immunocytochemistry has demonstrated that ovariectomyup-regulates the level of kisspeptin in the ARC ofthe ewe (Pompolo et al. 2006). More recently, upregulationof Kiss-1 mRNA has been found to occur inthe caudal region of the ARC during the preovulatoryperiod (Estrada et al. 2006), when peripheral oestradiolconcentration is high. Overall, these data suggest thatdifferent kisspeptin cell populations within the ARC areinvolved in negative and positive feedback regulation ofGnRH. Yet, we have recently found that Fos in intactcyclic ewes is activated in kisspeptin cells of the POAjust prior to the preovulatory GnRH ⁄ LH surge, andnot at the ARC level (Hoffman, personal communication).This transcriptional activation of the kisspeptincells of the POA may be secondary to previousactivation of cells in the ARC. Clearly, more experimentsare needed to determine by which mechanism(s),oestradiol acting in the medio-basal hypothalamusactivates GnRH cells to provoke the preovulatoryGnRH release in ewe.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Control of Fertility in Females by Kisspeptin 175Potential Applications of Kisspeptin in theControl of Fertility in FemalesThe unequivocal demonstration of a strong stimulatoryeffect of kisspeptin on GnRH and gonadotropin secretionhas already generated research to determine howthis type of molecule could provide new strategies forcontrolling the reproductive axis.An earlier study has shown that a single subcutaneousinjection of kisspeptin in prepubertal female rats is ableto induce ovulation (Matsui et al. 2004). However, itmust be pointed out that these animals had been primedwith pregnant mare’s serum gonadotropin, which stimulatesgonadal function and causes ovulation in sheepand goats (Ritar et al. 1984). A recent detailed analysisof the pattern of LH and FSH release in response tosystemic kisspeptin administration in ewes has defined aprotocol causing gonadotropin secretion similar to thatobserved during a preovulatory gonadotropin surge inlarger animals (Caraty et al. 2007). Acute intravenous orsubcutaneous administration of kisspeptin was observedto induce only a short-lived stimulation of LH release,with little relationship between the duration of thestimulation and the dose given. It must be recalled thaton a molecular weight basis, the potency to release LHfor kisspeptin is at least 10 times lower than for GnRHboth in vitro and in vivo (Thompson et al. 2004). Incontrast to acute kisspeptin administration, a low-doseinfusion of the peptide was shown to provide prolongedgonadotropin release. This protocol was then tested inewes during the breeding season, with the aim ofinducing well-timed ovulation. Ewes showing regularoestrous cycles were synchronized by the administrationof prostaglandin followed by insertion of an intravaginalprogesterone implant (CIDR) for 14 days, and aperfusion of kisspeptin (or of the vehicle as control) wasgiven 30 h after removal of the progesterone CIDR (afew hours before the time when the earliest spontaneousLH surge is usually observed). As illustrated in Fig. 4,an LH surge was found to occur within 2 h of the startof infusion in all kisspeptin-treated animals, whereas theLH surges were observed later and more widelyFig. 4. Left panel: during the breeding season, constant iv infusion ofkisspeptin (3.84 lmol for 8 h), starting 30 h after progesterone CIDRremoval in cyclic ewes, induces within 2 h an LH surge in all animals(9 ⁄ 9; closed circle). In animals receiving the vehicle, the LH surgeswere latter and more widely dispersed (open circle), hence only five outof nine animals show an LH surge during the period of bloodsampling. Right panel: progesterone plasma concentration increasesfor all animals over the next 5 days following the infusion, but the riseis significantly earlier for the kisspeptin-treated ewes. *p < 0.05,**p < 0.01. Redrawn from Caraty et al. (2007)dispersed in the vehicle-treated animals (from 42 h tomore than 65 h following progesterone removal). Theseperfectly timed preovulatory LH surges represent apharmacological intervention sending a kisspeptininducedGnRH surge signal to an oestrogen-primedpituitary. This result is in line with the observation thatkisspeptin stimulates gonadotropin release most potentlyduring the preovulatory phase of the menstrualcycle in women (Dhillo et al. 2007). This explosive LHrelease indicates that there is a window of time duringthe late follicular phase when kisspeptin synchronizesovulation perfectly and could therefore be used toenhance the efficacy of artificial insemination or in vitrofecundation programmes.The kisspeptin infusion protocol was also used toreinitiate cyclicity in animals with a quiescent gonadalaxis. A low-dose infusion of the peptide was given over48 h to anoestrus ewes. This treatment was carried outin two breeds of sheep and was shown to stimulategonadotropin secretion reliably and to induce LHsurges followed by successful ovulation in more than80% of the acyclic females. As illustrated in Fig. 5, aninitial LH response was seen within 1 h of starting theinfusion, but the elevated LH response was not maintainedwith continuous infusion. This gonadotropinincrease was, however, sufficient to ‘activate’ the systemas evidenced by a rise in peripheral oestradiol concentrations,and ovulatory surges were found to occurapproximately 20 h after the start of treatment, which isconsistent with positive oestrogen feedback. Accordingly,it can be concluded that constant low-doseadministration of kisspeptin to acyclic females is sufficientto activate the hypothalamo-pituitary axis andcause ovulation. This opens a very promising avenuewith regard to problems of fertility in large farmanimals such as the difficulty of obtaining well-timedovulation in equine species or chronic infertility inpostpartum dairy cows.ConclusionAdvances in kisspeptin research have been rapid, and ithas become evident that kisspeptin is a major player inthe control of the hypothalamo-pituitary-gonadal axisin numerous species. Kisspeptin is the most powerfulknown secretagogue of GnRH, and kisspeptin cells inthe hypothalamus are more than likely to be key playersin the steroid feedback control of GnRH secretion.Kisspeptin appears to be involved in most if not alltransitional steps of reproductive life, such as the onsetof puberty, initiation of the breeding season and thedynamic changes of gonadotropin secretion throughoutthe oestrous cycle. Using kisspeptin to manipulate thegonadotropin axis therefore appears to offer a verypromising avenue. The demonstration that continuousinfusion of kisspeptin can synchronize LH surges inprogesterone-primed cyclic ewes and reinitiate cyclicityin anoestrous animals suggests that other means ofcontinuous delivery of kisspeptin will be useful forsynchronizing ovulation. This will assist the design ofnew protocols for optimizing artificial inseminationand improving fertility in mammals. More than30 years after the discovery of GnRH, the possibilityÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

176 A Caraty and I FranceschiniFig. 5. LH secretion and ensuingovulation in ewes treated duringthe non-breeding season with anintravenous infusion of kisspeptin(20 lg ⁄ h for 48 h). The upperpanels show two examples of plasmaLH levels in kisspeptin-infusedewes, indicating that the treatmentled to increased secretion but thepreovulatory surge did not occuruntil 25 h later. This strongly suggeststhat the infusion initiated anartificial follicular phase and thatthe increased oestrogen resultingfrom this activated a positivefeedback response. The experimentwas repeated in France and inAustralia with similar results.Approximately 80% of the ewestreated with kisspeptin ovulated, asevidenced by resulting corporalutea on the ovaries. Adapted fromCaraty et al. (2007)of using kisspeptin for the control of the gonadotropinaxis provides hope for resolving all fertility problems.Who knows if this is the end of the story, or perhaps anew molecule activating kisspeptin might be discoveredin another 30 years’ time!ReferencesBlache D, Fabre-Nys CJ, Venier G, 1991: Ventromedialhypothalamus as a target for oestradiol action on proceptivity,receptivity and luteinizing hormone surge of the ewe.Brain Res 546, 241–249.Caraty A, Fabre-Nys C, Delaleu B, Locatelli A, Bruneau G,Karsch FJ, Herbison A, 1998: Evidence that the mediobasalhypothalamus is the primary site of action of estradiol ininducing the preovulatory gonadotropin releasing hormonesurge in the ewe. Endocrinology 139, 1752–1760.Caraty A, Smith JT, Lomet D, Ben Said S, Morrissey A,Cognie J, Doughton B, Baril G, Briant C, Clarke IJ, 2007:Kisspeptin synchronizes preovulatory surges in cyclical ewesand causes ovulation in seasonally acyclic ewes. Endocrinology148, 5258–5267.Clarke IJ, Cummins JT, Jenkin M, Phillips DJ, 1989: Theoestrogen-induced surge of LH requires a ‘signal’ pattern ofgonadotrophin-releasing hormone input to the pituitarygland in the ewe. J Endocrinol 122, 127–134.Clarkson J, Herbison AE, 2006: Postnatal development ofkisspeptin neurons in mouse hypothalamus; sexual dimorphismand projections to gonadotropin-releasing hormoneneurons. Endocrinology 147, 5817–5825.Dhillo WS, Chaudhri OB, Patterson M, Thompson EL,Murphy KG, Badman MK, McGowan BM, Amber V,Patel S, Ghatei MA, Bloom SR, 2005: Kisspeptin-54stimulates the hypothalamic-pituitary gonadal axis inhuman males. J Clin Endocrinol Metab 90, 6609–6615.Dhillo WS, Chaudhri OB, Thompson EL, Murphy KG,Patterson M, Ramachandran R, Nijher GK, Amber V,Kokkinos A, Donaldson M, Ghatei MA, Bloom SR, 2007:Kisspeptin-54 stimulates gonadotropin release mostpotently during the preovulatory phase of the menstrualcycle in women. J Clin Endocrinol Metab 92, 3958–3966.Dungan HM, Gottsch ML, Zeng H, Gragerov A, BergmannJE, Vassilatis DK, Clifton DK, Steiner RA, 2007: The roleof kisspeptin-GPR54 signaling in the tonic regulation andsurge release of gonadotropin-releasing hormone ⁄ luteinizinghormone. J Neurosci 27, 12088–12095.Estrada KM, Clay CM, Pompolo S, Smith JT, Clarke IJ, 2006:Elevated Kiss-1 expression in the arcuate nucleus prior tothe cyclic preovulatory gonadotropin-releasing hormone ⁄luteinising hormone surge in the ewe suggests a stimulatoryrole for kisspeptin in oestrogen-positive feedback. J Neuroendocrinol18, 139–149.Franceschini I, Lomet D, Cateau M, Delsol G, Tillet Y,Caraty A, 2006: Kisspeptin immunoreactive cells of theovine preoptic area and arcuate nucleus co-express estrogenreceptor alpha. Neurosci Lett 401, 225–230.Gottsch ML, Cunningham MJ, Smith JT, Popa SM, AcohidoBV, Crowley WF, Seminara S, Clifton DK, Steiner RA,2004: A role for kisspeptins in the regulation of gonadotropinsecretion in the mouse. Endocrinology 145, 4073–4077.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Control of Fertility in Females by Kisspeptin 177Gottsch ML, Clifton DK, Steiner RA, 2006: Kisspepeptin-GPR54 signaling in the neuroendocrine reproductive axis.Mol Cell Endocrinol 254 255, 91–96.Gu GB, Simerly RB, 1997: Projections of the sexuallydimorphic anteroventral periventricular nucleus in thefemale rat. J Comp Neurol 384, 142–164.Gutierrez-Pascual E, Martinez-Fuentes AJ, Pinilla L, Tena-Sempere M, Malagon MM, Castano JP, 2007: Directpituitary effects of kisspeptin: activation of gonadotrophsand somatotrophs and stimulation of luteinising hormoneand growth hormone secretion. J Neuroendocrinol 19, 521–530.Han SK, Gottsch ML, Lee KJ, Popa SM, Smith JT, JakawichSK, Clifton DK, Steiner RA, Herbison AE, 2005: Activationof gonadotropin-releasing hormone neurons by kisspeptinas a neuroendocrine switch for the onset of puberty. JNeurosci 25, 11349–11356.Herbison AE, 2007: Estrogen positive feedback to gonadotropin-releasinghormone (GnRH) neurons in the rodent: Thecase for the rostral periventricular area of the third ventricle(RP3V). Brain Res Rev 57, 277–287.Irwig MS, Fraley GS, Smith JT, Acohido BV, Popa SM,Cunningham MJ, Gottsch ML, Clifton DK, Steiner RA,2004: Kisspeptin activation of gonadotropin releasing hormoneneurons and regulation of KiSS-1 mRNA in the malerat. Neuroendocrinology 80, 264–272.Kadokawa H, Matsui M, Hayashi K, Matsunaga N, KawashimaC, Shimizu T, Kida K, Miyamoto A, 2008a:Peripheral administration of kisspeptin-10 increases plasmaconcentrations of GH as well as LH in prepubertal Holsteinheifers. J Endocrinol 196, 331–334.Kadokawa H, Suzuki S, Hashizume T, 2008b: Kisspeptin-10stimulates the secretion of growth hormone and prolactindirectly from cultured bovine anterior pituitary cells. AnimReprod Sci 105, 404–408.Kauffman AS, Gottsch ML, Roa J, Byquist AC, Crown A,Clifton DK, Hoffman GE, Steiner RA, Tena-Sempere M,2007: Sexual differentiation of Kiss1 gene expression in thebrain of the rat. Endocrinology 148, 1774–1783.Kinoshita M, Tsukamura H, Adachi S, Matsui H, UenoyamaY, Iwata K, Yamada S, Inoue K, Ohtaki T, Matsumoto H,Maeda K, 2005: Involvement of central metastin in theregulation of preovulatory luteinizing hormone surge andestrous cyclicity in female rats. Endocrinology 146, 4431–4436.Knobil E, 2005: Discovery of the hypothalamic gonadotropinreleasinghormone pulse generator and of its physiologicsignificance. 1992. Am J Obstet Gynecol 193, 1765–1766.Kotani M, Detheux M, Vandenbogaerde A, Communi D,Vanderwinden JM, Le Poul E, Brezillon S, Tyldesley R,Suarez-Huerta N, Vandeput F, Blanpain C, Schiffmann SN,Vassart G, Parmentier M, 2001: The metastasis suppressorgene KiSS-1 encodes kisspeptins, the natural ligands of theorphan G protein-coupled receptor GPR54. J Biol Chem276, 34631–34636.Lee DK, Nguyen T, O’Neill GP, Cheng R, Liu Y, HowardAD, Coulombe N, Tan CP, Tang-Nguyen AT, George SR,O’Dowd BF, 1999: Discovery of a receptor related to thegalanin receptors. FEBS Lett 446, 103–107.Lents C, Barb C, 2007: Stimulatory actions of kisspeptin ongonadotropin secretion in prepubertal swine. Biol ReprodSpecial Issue 105 (Abstract No. 125).Maeda K, Adachi S, Inoue K, Ohkura S, Tsukamura H, 2007:Metastin ⁄ kisspeptin and the control of estrous cycle in rats.Rev Endocr Metab Disord 8, 21–29.Matsui H, Takatsu Y, Kumano S, Matsumoto H, Ohtaki T,2004: Peripheral administration of metastin induces markedgonadotropin release and ovulation in the rat. BiochemBiophys Res Commun 320, 383–388.Messager S, Chatzidaki EE, Ma D, Hendrick AG, Zahn D,Dixon J, Thresher RR, Malinge I, Lomet D, Carlton MB,Colledge WH, Caraty A, Aparicio SA, 2005: Kisspeptindirectly stimulates gonadotropin-releasing hormone releasevia G protein-coupled receptor 54. Proc Natl Acad Sci USA102, 1761–1766.Moenter SM, Caraty A, Locatelli A, Karsch FJ, 1991: Patternof gonadotropin-releasing hormone (GnRH) secretion leadingup to ovulation in the ewe: existence of a preovulatoryGnRH surge. Endocrinology 129, 1175–1182.Muir AI, Chamberlain L, Elshourbagy NA, Michalovich D,Moore DJ, Calamari A, Szekeres PG, Sarau HM, ChambersJK, Murdock P, Steplewski K, Shabon U, Miller JE,Middleton SE, Darker JG, Larminie CG, Wilson S, BergsmaDJ, Emson P, Faull R, Philpott KL, Harrison DC, 2001:AXOR12, a novel human G protein-coupled receptor,activated by the peptide KiSS-1. J Biol Chem 276, 28969–28975.Navarro VM, Castellano JM, Fernandez-Fernandez R,Barreiro ML, Roa J, Sanchez-Criado JE, Aguilar E,Dieguez C, Pinilla L, Tena-Sempere M, 2004: Developmentaland hormonally regulated messenger ribonucleicacid expression of KiSS-1 and its putative receptor,GPR54, in rat hypothalamus and potent luteinizing hormone-releasingactivity of KiSS-1 peptide. Endocrinology145, 4565–4574.Navarro VM, Castellano JM, Fernandez-Fernandez R, TovarS, Roa J, Mayen A, Barreiro ML, Casanueva FF, Aguilar E,Dieguez C, Pinilla L, Tena-Sempere M, 2005: Effects ofKiSS-1 peptide, the natural ligand of GPR54, on folliclestimulatinghormone secretion in the rat. Endocrinology146, 1689–1697.Ohtaki T, Shintani Y, Honda S, Matsumoto H, Hori A,Kanehashi K, Terao Y, Kumano S, Takatsu Y, Masuda Y,Ishibashi Y, Watanabe T, Asada M, Yamada T, SuenagaM, Kitada C, Usuki S, Kurokawa T, Onda H, Nishimura O,Fujino M, 2001: Metastasis suppressor gene KiSS-1 encodespeptide ligand of a G-protein-coupled receptor. Nature 411,613–617.Orsini MJ, Klein MA, Beavers MP, Connolly PJ, MiddletonSA, Mayo KH, 2007: Metastin (KiSS-1) mimetics identifiedfrom peptide structure-activity relationship-derived pharmacophoresand directed small molecule database screening.J Med Chem 50, 462–471.Patterson M, Murphy KG, Thompson EL, Patel S, GhateiMA, Bloom SR, 2006: Administration of kisspeptin-54 intodiscrete regions of the hypothalamus potently increasesplasma luteinising hormone and testosterone in male adultrats. J Neuroendocrinol 18, 349–354.Pompolo S, Pereira A, Estrada KM, Clarke IJ, 2006: Colocalizationof kisspeptin and gonadotropin-releasing hormonein the ovine brain. Endocrinology 147, 804–810.Ritar AJ, Maxwell WM, Salamon S, 1984: Ovulation and LHsecretion in the goat after intravaginal progestagen sponge-PMSG treatment. J Reprod Fertil 72, 559–563.de Roux N, Genin E, Carel JC, Matsuda F, Chaussain JL,Milgrom E, 2003: Hypogonadotropic hypogonadism due toloss of function of the Kiss-1-derived peptide receptorGPR54. Proc Natl Acad Sci USA 100, 10972–10976.Sarkar DK, Chiappa SA, Fink G, Sherwood NM, 1976:Gonadotropin-releasing hormone surge in pro-oestrous rats.Nature 264, 461–463.Seminara SB, Messager S, Chatzidaki EE, Thresher RR,Acierno JS Jr, Shagoury JK, Bo-Abbas Y, Kuohung W,Schwinof KM, Hendrick AG, Zahn D, Dixon J, Kaiser UB,Slaugenhaupt SA, Gusella JF, O’Rahilly S, Carlton MB,Crowley WF Jr, Aparicio SA, Colledge WH, 2003: TheGPR54 gene as a regulator of puberty. N Engl J Med 349,1614–1627.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

178 A Caraty and I FranceschiniShahab M, Mastronardi C, Seminara SB, Crowley WF, OjedaSR, Plant TM, 2005: Increased hypothalamic GPR54signaling: a potential mechanism for initiation of pubertyin primates. Proc Natl Acad Sci USA 102, 2129–2134.Smith JT, Cunningham MJ, Rissman EF, Clifton DK,Steiner RA, 2005: Regulation of Kiss1 gene expression inthe brain of the female mouse. Endocrinology 146, 3686–3692.Smith JT, Popa SM, Clifton DK, Hoffman GE, Steiner RA,2006: Kiss1 neurons in the forebrain as central processorsfor generating the preovulatory luteinizing hormone surge. JNeurosci 26, 6687–6694.Smith JT, Rao A, Pereira A, Caraty A, Millar RP, Clarke IJ,2007a: Kisspeptin is present in ovine hypophysial portalblood, but does not increase during the preovulatoryluteinizing hormone surge: Evidence that gonadotropes arenot direct targets of kisspeptin in vivo. Endocrinology 149,1979–1986.Smith JT, Clay CM, Caraty A, Clarke IJ, 2007b: KiSS-1messenger ribonucleic acid expression in the hypothalamusof the ewe is regulated by sex steroids and season.Endocrinology 148, 1150–1157.Suzuki S, Kadokawa H, Hashizume T, 2007: Directkisspeptin-10 stimulation on luteinizing hormone secretionfrom bovine and porcine anterior pituitary cells. AnimReprod Sci 103, 404–408.Thompson EL, Patterson M, Murphy KG, Smith KL, DhilloWS, Todd JF, Ghatei MA, Bloom SR, 2004: Central andperipheral administration of kisspeptin-10 stimulates thehypothalamic-pituitary-gonadal axis. J Neuroendocrinol 16,850–858.Xia L, Van Vugt D, Alston EJ, Luckhaus J, Ferin M, 1992: Asurge of gonadotropin-releasing hormone accompanies theestradiol-induced gonadotropin surge in the rhesus monkey.Endocrinology 131, 2812–2820.Author’s address (for correspondence): A. Caraty, UMR 6175,Physiologie de la Reproduction et des Comportements,INRA ⁄ CNRS ⁄ Université Tours ⁄ Haras Nationaux, 37380 Nouzilly,France. E-mail: caraty@tours.inra.frConflict of interest: This work was supported by an ‘‘Agence Nationalede la Recherche’’ (ANR) Grant. The authors have no furtherdeclarations.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 179–185 (2008); doi: 10.1111/j.1439-0531.2008.01159.xISSN 0936-6768Controlling Animal Populations Using Anti-Fertility VaccinesR Fayrer-HoskenDepartment of Large Animal Medicine, College of Veterinary Medicine, University of Georgia, Athens, GA, USAContentsThe goal for fertility control of animal populations is thedevelopment of a safe, economical and effective contraceptive.One offshoot of the development of this technology is theacquisition of multiple therapeutic strategies for diseases, suchas immunotherapy probes for cancer. In the long run,successful population control requires multifactorial strategies.One component of population control is immunocontraception.Development of effective antigens for immunocontraceptivevaccines has been remarkable and has greatly advancedour understanding of the molecular mechanisms of fertilization.The chasm between the discovery of an antigen in thelaboratory, to the implementation of an effective field program,is immense. The zona pellucida (ZP) immunocontraceptivethat has been most extensively evaluated as a fertilityvaccine antigen and the porcine ZP has received particularattention. The long-term goal of population control would bethe use of a synthetic vaccine, e.g. the ZP, tailored to a targetspecies. In the future, if populations’ levels are to be controlledby fertility vaccines, we should consider that the vaccinatedanimals could receive other health protective agents at the sametime. For example, if a species were immunocontracepted, thenthey could be simultaneously vaccinated against habitatdiseases such as rabies (Plumb et al., Rev Sci Tech, 26, 2007,229).IntroductionThe ultimate goal of fertility control of a population isan agent that it is non-toxic, completely reversible, hasmulti-year efficacy, is easily administered, is devoid ofbehavioural side-affects, cost-effective and is efficaciousin multiple species and both sexes. This is an almostimpossible goal. The primary impetus for the developmentof a vaccine for fertility regulation is safety, so thatit has no adverse hormonal or toxic side-effects. Thesesame issues are of particular interest to researchers incompanies concerned with human immunocontraception.We must remember that evolution has assured thatthe survival of a species is protected and centres on themechanisms of fertilization and early embryonic development.For survival of the species, the reproductiveprocesses of fertilization are probably protected bynumerous alternative (redundant) mechanisms. Therefore,it stands to reason that nature will attempt tothwart or circumvent reproductive control strategies.There are multiple antigens that might be targets forfertility control with vaccines in animals. These targetsinclude hormones, oocyte proteins ⁄ glycoproteins, spermproteins and other molecules associated with fertilizationand early embryonic development. In addition, wemust consider that the individual animal may have aunique and variable immunological response to theseantigens. The presentation of these antigens to theanimals’ immune system will predicate the degree andseverity (Litscher and Wassarman 2007) of the animal’sfertility control. As part of the discussion, the developmentof an antigen from a complete product of purifiednative proteins, to portions of a native protein, to asynthetic protein with a limited or altered compliment ofcarbohydrates, to a synthetic polypeptide ⁄ carbohydrateantigen, to a DNA-vectored or virally vectored vaccineis included.Our experiences in immunocontraceptives, from laboratoryto field studies, have revealed a number ofsurprising and unexpected paradoxes. For vaccinedevelopment, the use of contraceptive data from amodel species should be very circumspectly extrapolatedto a different target species in the field. From laboratorydevelopment to field delivery of a vaccine, there areprofound practical problems. Vaccine administration inthe laboratory environment often has nominal relationshipto the intricacies of field administration. Policies forpracticality and techniques of field administration mustbe considered for both efficacy and safety. Nevertheless,we must remember that cost-effective administrationmight not be the best or safest strategy for administrationof an immunocontraceptive vaccine.Zona Pellucida (ZP) VaccinesThe conceptualization of the ZP as an antigen forfertility control was considered at the beginning of the20th century with reports from Metchnikoff (Shiverset al. 1972). In the early-1960s (Shivers and Metz 1962)to the mid-1970s (Yanagimachi et al. 1976), investigatorsconsidered the role of ZP-based immunocontraceptives.The most conspicuous advantage is thatimmunocontraceptive vaccines, by their mechanism,appear to have no hormonal effects or deleterious sideeffects(Barber and Fayrer-Hosken 2000a; b). As reproductivecyclicity generally remains normal in the shortterm, the immunocontracepted animals have minimalbehavioural or herd affects, as the inter-oestrous intervalwould be unaffected. Thus, herds of animals whosebehaviour and herd integrity is based on intact reproductivecyclicity are unaffected by porcine ZP (pZP)immunocontraceptives. Because the ZP glycoproteinsare a structurally unique and have very little homologywith other somatic proteins, they provide ideal targetsfor immunocontraception (Barber and Fayrer-Hosken2000a). The ZP protein structures are highly conservedacross several phyla and this facilitates interspecies use(Barber and Fayrer-Hosken 2000b; Litscher and Wassarman2007) of ZP antigens. This conserved homology(Hedrick 1996) of the ZP glycoproteins has allowedthe use of antigens derived from heterologous species(Barber and Fayrer-Hosken 2000b; Fayrer-HoskenÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

180 R Fayrer-Hoskenet al. 2000a,b, 2002; Barber et al. 2001). The use ofheterologous ZP antigens allows for standardization ofa vaccine’s dose and adjuvant as well as its deliverymethod. In addition, the use of heterologous ZPobviates the necessity for a homologous source ofantigens, which for most exotic species would beimpossible. The successful use of heterologous ZPglycoproteins suggests that the mechanism of sperm-ZP interaction might have some commonality betweenspecies. For the same reason, it works as an immunocontraceptivethat affects multiple species, it also questionsthe precise mechanism of the species-specificity ofsperm-ZP interaction. For much of the publishedliterature, the pZP has been the heterologous antigenof choice. To harvest enough native ZP for laboratoryor field studies has been labour-intensive and investigatorshave considered alternative sources of ZP. Thehorse is one of the best-studied mammals as a model forimmunocontraception. The prototypical studies in thehorse were completed by Dr Irwin Liu (Liu and Shivers1982; Liu et al. 1989), who demonstrated the efficacyand reversibility of the pZP vaccine. The horse has beenused as both a laboratory model (Liu et al. 1989; Williset al. 1994) and an ecological model (Kirkpatrick andTurner 2003). These immunocontraceptive studies in thehorse have shown that the pZP vaccine is both safe andeffective. Long-term vaccinations appear to have effectson the functionality of the ovary (Turner and Kirkpatrick2002), but these effects appear to be reversible withtime.Structure and origin of the ZPThe ZP matrix consists of 3–4 glycoproteins (ZP1, ZP2,ZP3 and ZP4) depending on the species. At a ratio of1 : 1 (Green 1997), ZP2 and ZP3 are inter-twined toform glycoprotein filaments that are cross-linked by ZP1(Wassarman and Mortillo 1991; Jovine et al. 2002). TheZP glycoproteins are combined into a microfibrillarmatrix. The ZP matrix is secreted around the oocyte andhas significant elastic properties. In some species, theindividual glycoproteins are produced in the ooplasmwithin the Golgi apparatus (Wassarman et al. 2005) andsubsequently moved through the cytoplasm and releasedoutside the oolemma into the perivitelline space.For the mare, the target antigens (eqZP) andvaccinational antigens (pZP) have been well characterized.The equine research shows that there are threeequine ZP (eqZP) glycoproteins using silver-stained2D-PAGE. The glycoprotein families have molecularweights of 93–120 K, 73–90 K and 45–80 K (Milleret al. 1992). Common epitopes on the eqZP (Milleret al. 1992) were detected by rabbit antiserum againstheat-solubilized pZP (RaHSPZ) (Dunbar and Raynor1980), and guinea pig antiserum against rabbit ZP(R55K) (Lee and Dunbar 1993; Lee et al. 1993).RaHSPZ recognized all three eqZP glycoprotein familiesand R55K only recognized the lowest molecularweight eqZP glycoprotein family eqZP3 (Miller et al.1992). This molecular data is complimented by thesuccess of mare immunocontraception in the field (Liuet al. 1989; Kirkpatrick et al. 1992). In addition, asanti-total-pZP antibodies and anti-R55K recognizeeqZP3, they could serve as a specific immunocontraceptiveantigen. This is important in targeting the ZP3molecule or portion of the molecule and would be thebest immunological approach for immunocontraceptionusing a synthetic ZP.The formation of the eqZP is a collaborative processbetween the oocyte and cumulus cells (Kolle et al. 2007),and the final external layer of eqZP is a function of thecumulus cells. The collaborative secretion of ZP glycoproteinsis also seen in pig, cow, dog, rabbit, marmosetand rhesus monkey (Kolle et al. 2007). In other species,e.g. the cat and mouse, only the oocyte secretes ZP. Thetemporal secretion of ZP glycoproteins is variablebetween species relative to oogenesis. In mouse, ZP2 isdeposited in the primordial follicle’s perivitelline space,but ZP3 and ZP4 only appear with the formation of theprimary oocyte and the initiation of its growth (Millaret al. 1993; Epifano et al. 1995). Therefore, there aredifferences in the cellular origin of the secretion of ZPglycoproteins, the temporal secretion of the ZP glycoproteinsand the manifestation of the antigenic determinantson the ZP glycoproteins.Although the precise molecular mechanism for theimmunocontraceptive effect of pZP remains to bedetermined, the carbohydrate portion of the ZP glycoproteinhas been implicated as an important aspect(Paterson et al. 1992). If immunocontraception is in partmediated by role of terminal carbohydrates, theirdistribution and configuration might hold the answer.Possible molecular explanations for the biochemicalmechanism of immunocontraception could includeblocking sperm binding sites (Barber and Fayrer-Hosken2000b) or structural changes in the ZP. When amare is successfully immunocontracepted, there are highcirculating levels of both IgG and IgM. The follicularIgG levels are dependent on the peripheral circulatinglevels (Hussein and Bourne 1984; Widders et al. 1984).IgG is probably the primary antibody to mediateimmunocontraception (Barber and Fayrer-Hosken2000b) as declining serum IgG levels are associated withreturn to fertility of the mare (Liu et al. 1989; Barberand Fayrer-Hosken 2000b). Another possible explanationfor the immunocontraceptive effects of pZP immunizationin the mare is that during the deposition offibrils of eqZP, anti-pZP IgG molecules might beintercalated into the matrix. The combination of theIgG and eqZP matrix could cause the anti-pZP antibodiesto act like cross-linkers (e.g. autoimmune reactions)and this could mimic the zona block. Theseimmunocontracepted zona-blocked oocytes would preventfertilization and this effect would remain untilsystemic levels of IgG fall low enough to allow conception.This mechanism would be most relevant inanimals, where cumulus cells participate in secretingZP. This would provide a reasonable explanation as towhy animals, e.g. cats whose ZP is derived completelyfrom the oocyte, are unaffected by heterologous pZPvaccination (Jewgenow et al. 2000).Another possible mechanism for the immunocontraceptiveeffect of pZP (Barber and Fayrer-Hosken 2000b)is an affect on the sperm-zona binding through therecognition of carbohydrate epitopes on the glycoproteinsof the ZP. It is clear that the ZP proteins areÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Controlling Animal Populations Using Anti-Fertility Vaccines 181strongly conserved among species (Benoff 1997) and it ispossible that species-specificity of fertilization may bemoderated by differences in carbohydrates on the zonasurface. Current data demonstrate that vaccination withthe three pZP glycoproteins induces immunocontraceptionin multiple species of mammals. To explainspecies-specificity of the carbohydrates and the speciesindependenteffect of pZP, the author suggests that thehypothesis still has some merit. Binding of anti-pZP IgGto surface carbohydrate epitopes probably preventssperm binding by direct and indirect interference. Thisis supported by the observation that 63% glycosidicrecognition capability of our anti-pZP antibodies(Barber 2001). The remainder of the polyclonal IgGantibodies recognizes the exposed protein matrix. Bindingof the ZP protein matrix by these IgGs probablycompletely blocks the induction of the acrosomereaction. Hence polyclonal IgGs block fertilization bypartial hindrance of binding and possible blockade ofacrosome reaction induction (Mahi-Brown et al. 1985;Henderson et al. 1988).Vaccine preparationThe ZP glycoproteins are secreted to combine with eachother in the perivitelline space in a coordinated process.However, in vitro, the protein harvesting of pZP forimmunocontraceptive vaccines varies from laboratoryto laboratory. We also know that the purification andstorage of the ZP glycoproteins varies from laboratoryto laboratory. The production and storage differencesmight also alter the site of presentation of the ZPglycoproteins. The purification and storage of theproteins might be extremely critical to antigen bioavailability.How have the glycoproteins recombined andwhat antigenic sites have been exposed? More importantly,does this represent the in vivo structure? Formouse proteins, we know that the individual glycoproteinsrecombine or repolymerize in an homomeric orheteromeric manner (Litscher et al. 2008). Repolymerizationoccurs under non-denaturing conditions evenafter the treatment of the glycoproteins with SDS, heat,urea or lyophilization. In the in vivo environment, thepolymerization of ZPs probably occurs in a coordinatedmolar integration. However, under in vitro conditions,especially with homomeric or heteromeric repolymerization,there could be novel aggregate presentations.The novel presentation could be in part an explanationof the effect on multiple species. The same rules ofantigen presentation also apply to synthetic vaccines.For the synthetic or recombinant vaccines, the antigenicityis also associated with the site of the source DNAsequence, the amount and quality of glycosylation, andthe tertiary assembly.The use of a complete vaccine containing all threepZP glycoproteins clearly works (Liu et al. 1989; Fayrer-Hoskenet al. 2000a,b, 2002; Delsink et al. 2002;Stoops et al. 2006) as an immunocontraceptive agent indifferent species. A more defined synthetic or recombinantvaccine might be safer and more effective. Thenecessity to standardize a vaccine product and have aready source has lead to the research into synthetic ZPvaccines (Choudhury et al. 2007; Duckworth et al.2007), but to date no synthetic vaccine has been usedextensively.Another consideration for vaccine usage is the role ofadjuvants. When ZP glycoproteins are adjuvented byconjugation, the ZP proteins are conjoined through thesulfhydryl linkages, and are presented to the immunesystem in a specific conformation. This might explainmore profound and unique effects of an adjuvant. Wehave used pZP with an adjuvant ImjectÒ (ThermoFisher, IL, USA), a maliamide-activated keyhole limpethaemocyanin. The conjugated product produced immunocontraceptionin gilts and showed significant degenerativepathology of the follicles and was probably due topresentation of B-cell epitopes. The other aspect of thevaccination is the role of the adjuvant. The role of addedadjuvants to the pZP appears to be vitally important. Todate, complete Freund’s (Delsink et al. 2007), incompleteFreund’s (Hannesdottir et al. 2004; Duckworth et al.2007), modified Freund’s (Lyda et al. 2005) and synthetictrehalose dicorynomycolate (STDCM) (Fayrer-Hoskenet al. 2002) have been used successfully. The adjuvantand pZP combination lead to stimulation of the immunemechanism. How much of it is due to the immunostimulantrole of the adjuvant and how much of it is dueto the pZP glycoprotein assembly after harvesting?Vaccination method and route of deliveryFor free-roaming populations, the antigen and adjuvantsgenerally will have to be delivered remotely. Thevaccine might be delivered with darts or in baits. Incase of dart delivery, the measure of success ispredicated by field conditions as this might includestalking on foot to helicopter delivery and the degree ofsuccess varies. From captive situations, e.g. elephants inzoos where there are controlled conditions, vaccinationsgenerally result in 100% efficacy (Fayrer-Hoskenet al. 1999). This is when compared with 80–100%vaccination titre efficacy in field when darting from ahelicopter (Fayrer-Hosken et al. 2000b). The practicalreality is that in the field, darting does not alwaysdeliver an intramuscular injection. Therefore, this is thepractical reality of field vaccination. The vaccineformulation also plays a role in the success of administration.The use of vaccine pellets (Turner et al. 2002;Lane et al. 2007) or lactide-glycolide (Kirkpatrick et al.1996; Liu et al. 2005) preparations for longer-actingvaccines makes the administration difficult. Oral baitingof a species-specific fertility vaccination would be veryexciting as a strategy, as seen with oral vaccinationagainst rabies (Tordo et al. 2006). The possibility oforal vaccines that employ plants (Tacket 2005) that areengineered to produce a glycosylated ZP product is anexciting possibility (Polkinghorne et al. 2005). Oralvaccines may provide a viable administration route,especially for populations that are difficult to reach.Species-specificity or bait-specificity to deliver theimmunocontraceptive to only the target species presentsthe most worrisome contraindication. While theremight be a possibility that baited or species-specificsites for intra-nasal (Corner et al. 2001; Costantinoet al. 2007) delivery of immunocontraceptives, this canonly be evaluated in situ. The most important facet ofÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

182 R Fayrer-Hoskenvaccine duration and efficacy is the magnitude andduration of serum antibody levels. For the nativeantigen, to all intents and purposes, its antigenicstructure is similar for preparations from the samelaboratory. This is not the case in synthetic preparationswhere the selected DNA sequence and expressionsystem affect the product.Duration of immunocontraceptionMost of the above-mentioned adjuvants provide at least12 months of immunocontraception with one or twoboosters. Some reports indicate longer effects with otherpreparations (Liu et al. 2005) with no boostering or onebooster. The 1-year duration vaccines can be administeredfor several years without any side-effect. However,after multiple seasons of administration, there areovarian effects as seen with the horses on AssateagueIslands. These cases that have an oestrogen phase ofoestrus, but did not appear to initiate a luteal orprogesterone phase. The initial (>5 years) effect oncyclicity of mares is reversed after several years when novaccines have been re-administered. Pregnant animals(Fayrer-Hosken et al. 2000b; Kirkpatrick and Turner2002) can be vaccinated with no deleterious effects onthe foetus and growth of the offspring. These offspringhave been shown to grow normally and are fertile asadults.Current vaccinesCurrent vaccines use synthetic or recombinant ZPfragments with or without glycosylation, DNA vaccinesor virally vectored vaccines. How does the structure ofthese antigens affect fertility and by what mechanism dothey function? The selection of the specific portions ofthe DNA for the production of ZP glycoproteins is thecurrent optimal technology. The selection of the ZPsequence is very important when an expression systemis selected. Immunocontraceptive antigens can beexpressed in bacterial systems (Hardy et al. 2008), yeastsystems (Tang et al. 2003), Chinese hamster ovary(CHO) systems (Zhao et al. 2004), plant systems (Tacket2005), viral systems (Hardy et al. 2003), insect systems(Hardy et al. 2003; Choudhury et al. 2007) and mammaliancell lines. The recombinant products of thesesystems have several exciting properties. The antigenpurity and structure are optimized. The purity (Tanget al. 2003) of purified native proteins has been aconcern and the effect of contaminating antigens has notbeen clearly documented. Once the appropriate antigenand expression system has been discovered for a specificspecies, abundant amounts of the antigen could beproduced in a cost-effective and repeatable process.These antigens and their production systems will be thefuture zenith of fertility control vaccines. Anotherstrategy is to use a species-specific antigen with aspecies-specific virus to produce a virally vectoredimmunocontraceptive (Redwood et al. 2007). Thevirally vectored immunocontraceptive would solvemany of the practical problems of fertility control ofpopulations. Nevertheless, there is a concern that viralor antigen mutation could lead to the doomsday virus.SummaryThe ZP vaccine is the most tested antigen fromlaboratory to field application. The vaccine appearssafe and effective. The fertility control vaccine does notappear to adversely affect the sociobiology of the targetspecies.Role of hormonal targets for vaccines to control fertilityFertility control can also be achieved by the vaccinationof male or female animals with antigens that arehormones. Several successful studies have been reportedusing gonadotropin-releasing hormone (GnRH) orluteinizing hormone releasing hormone (LHRH),human chorionic gonadotropin (hCG) and folliclestimulatinghormone (FSH).GnRH ⁄ LHRHThe hormonal vaccines can be used as fertility controls,and also as therapeutic agents. Treatment of endocrinedisorders and neoplasms is an important facet ofhormonal vaccines. A GnRH vaccine in humans as atherapeutic agent is increasingly important. The classicalGnRH or GnRH-I is the main hormone and alterationsto its secretions will affect ovaries, testis, prostate andplacenta. The GnRH vaccine has been used successfullyin multiple species. The GnRH vaccine must be conjugatedto a hapten or produced as a recombinant vaccinecontaining an immunogenic moiety in order to amplifythe antigenicity of the decapeptide. Conjugation toKLH, ovalbumin, more potent adjuvants or otheramplifying agents is the essential for efficacy. It hasworked for variable periods of time in both sexes in pigs(Killian et al. 2006), sheep (Earl et al. 2006), horses(Turkstra et al. 2005; Imboden et al. 2006; Elhay et al.2007), deer (Curtis et al. 2002) and bison (Miller et al.2004). The primary problem associated with GnRHvaccines is to produce durable titres.Over the years, several commercial preparations havebeen developed and marketed with greater or lesserfunctional product. Current preparations include GonaConÔ,ImprovacÔ (Dunshea et al. 2001), Equity andRepro-BlocÔ. The GnRH-analogue, GnRH-d6-Lys,was conjugated to recombinant Mycobacterium tuberculosishsp70 and adjuvanted with RIBI (STDCM) orIncomplete Freund’s was used in pre-pubertal mice.There was a statistically significant effect on the size ofthe urogenital complex and testosterone levels (RIBI).One of the most important animal uses of the GnRH-D6-Lys vaccine is in beef production. This product is animportant alternative to hormonal growth promotersand their side-effects. The vaccines have the desiredeffects through multiple treatments on both sexes withminimal side-effects. This is a quality product for thecaptive environment, but for wild populations, themultiple administrations pose the same problems aspZP vaccines. Depending on the presentation andadjuvantation, GnRH vaccines last from 1 to 2 years(Miller et al. 2000). The return to fertility with GnRHvaccines is more rapid than pZP vaccines. In the shortterm with GnRH vaccines, there is a cessation ofÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Controlling Animal Populations Using Anti-Fertility Vaccines 183oestrous behaviour and cyclicity. This is when comparedwith pZP, where there is cessation of cyclicity after longtermexposure. While cessation of cyclicity might beacceptable in some species, it is more controversial ormight be contra-indicated in herd-dependant species.Human chorionic gonadotropinHuman chorionic gonadotropin is another target forfertility vaccines primarily in humans. As the vaccinehas little effect on wild populations, the hCG ismentioned for completeness (Cui et al. 2007).Role of sperm antigens for vaccines to control fertilityThe presence of anti-sperm antigens is one of the mostpervasive causes of human infertility. So, there has beensignificant research on the antigenic causes of infertility.A series of excellent investigations has provided a seriesof candidate antigens that could be used for fertilitycontrol vaccines. Unfortunately, a limited number oftrials outside the mouse model have been reported forthe practical use of ASA for immunocontraception.Eppin (O’Rand et al. 2007) was used in monkeys and78% became infertile. The infertility was reversible, butas a practical field agent, the Eppin ⁄ Freund’s combinationwould not work as the monkeys had to bevaccinated every 3 weeks for the duration of the study.While many of the anti-sperm antigens have beendescribed, their use as fertility control vaccines islimited.ConclusionsThe need for fertility control in the various ecosystems isobvious. For fertility modulation of populations, thereis no ‘magic agent’ for comprehensive control. For theshear size of the problem, nature will constantly attemptto circumvent control strategies. Scientists have providedmany solutions over the last decade, but there isstill a gulf between dissection of the molecule in thelaboratory to delivery of the antigen to the target speciesin the field. Certainly, the most used and defined fertilityvaccine is the pZP glycoprotein-based immunocontraceptive.The vaccine in its native and recombinant formsis effective, safe and relatively easily delivered. Anti-GnRH and anti-sperm fertility control vaccines alsohold significant promise. In the future, the armoury oftarget antigens, adjuvants and delivery modalities usedwith critical evaluation of ecological and biodiversityissues, will allow safe and economical fertility control ofpopulations. Most importantly, the integrated use ofcontrol strategies within the realm of a damagedplanetary ecosystem will assist with conservation andpreservation of the beauties of nature.AcknowledgementsThe author would like to dedicate this manuscript to JJ van Altenawho is recovering from injuries obtained while darting wildlife forimmunocontraception. JJ van Altena has made the sacrifice for thetransition from laboratory science to field application. The authorwould also like to thank Irwin Liu and Barry Ball for their assistancein reviewing the manuscript.ReferencesBarber MR, 2001: Comparison of the Mechanism of Action ofthe pZP Contraceptive in the Pig, Horse, Elephant and Dog.PhD Thesis, University of Georgia, Athens.Barber MR, Fayrer-Hosken RA, 2000a: Evaluation of somaticand reproductive immunotoxic effects of the porcine zonapellucida vaccination. J Exp Zool 286, 641–646.Barber MR, Fayrer-Hosken RA, 2000b: Possible mechanismsof mammalian immunocontraception. J Reprod Immunol46, 103–124.Barber MR, Lee SM, Steffens WL, Ard M, Fayrer-HoskenRA, 2001: Immunolocalization of zona pellucida antigens inthe ovarian follicle of dogs, cats, horses and elephants.Theriogenology 55, 1705–1717.Benoff S, 1997: Carbohydrates and fertilization: an overview.Mol Hum Reprod 3, 599–637.Choudhury S, Srivastava N, Narwal PS, Rath A, Jaiswal S,Gupta SK, 2007: Feasibility and challenges in the developmentof immunocontraceptive vaccine based on zonapellucida glycoproteins. Soc Reprod Fertil Suppl 63, 479–493.Corner LA, Buddle BM, Pfeiffer DU, Morris RS, 2001:Aerosol vaccination of the brushtail possum (Trichosurusvulpecula) with bacille Calmette-Guerin: the duration ofprotection. Vet Microbiol 81, 181–191.Costantino HR, Illum L, Brandt G, Johnson PH, Quay SC,2007: Intranasal delivery: physicochemical and therapeuticaspects. Int J Pharm 337, 1–24.Cui C, Stevens VC, Schwendeman SP, 2007: Injectablepolymer microspheres enhance immunogenicity of a contraceptivepeptide vaccine. Vaccine 25, 500–509.Curtis PD, Pooler RL, Richmond ME, Miller LA, MattfeldGF, Quimby FW, 2002: Comparative effects of GnRH andporcine zona pellucida (PZP) immunocontraceptive vaccinesfor controlling reproduction in white-tailed deer (Odocoileusvirginianus). Reprod Suppl 60, 131–141.Delsink AK, van Altena JJ, Kirkpatrick J, Grobler D, Fayrer-Hosken RA, 2002: Field applications of immunocontraceptionin African elephants (Loxodonta africana). ReprodSuppl 60, 117–124.Delsink AK, van Altena JJ, Grobler D, Bertschinger HJ,Kirkpatrick JF, Slotow R, 2007: Implementing immunocontraceptionin free-ranging African elephants at Makalaliconservancy. J. S. Afr. Vet. Assoc. 78, 25–30.Duckworth JA, Wilson K, Cui X, Molinia FC, Cowan PE,2007: Immunogenicity and contraceptive potential of threeinfertility-relevant zona pellucida 2 epitopes in the marsupialbrushtail possum (Trichosurus vulpecula). Reproduction133, 177–186.Dunbar BS, Raynor BD, 1980: Characterization of porcinezona pellucida antigens. Biol Reprod 22, 941–954.Dunshea FR, Colantoni C, Howard K, McCauley I, JacksonP, Long KA, Lopaticki S, Nugent EA, Simons JA, Walker J,Hennessy DP, 2001: Vaccination of boars with a GnRHvaccine (Improvac) eliminates boar taint and increasesgrowth performance. J. Anim. Sci. 79, 2524–2535.Earl ER, Waterston MM, Aughey E, Harvey MJ, Matschke C,Colston A, Ferro VA, 2006: Evaluation of two GnRH-Ibased vaccine formulations on the testes function of entireSuffolk cross ram lambs. Vaccine 24, 3172–3183.Elhay M, Newbold A, Britton A, Turley P, Dowsett K, WalkerJ, 2007: Suppression of behavioural and physiologicaloestrus in the mare by vaccination against GnRH. AustVet J 85, 39–45.Epifano O, Liang LF, Familari M, Moos MC, Jr, Dean J,1995: Coordinate expression of the three zona pellucidagenes during mouse oogenesis. Development 121, 1947–1956.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

184 R Fayrer-HoskenFayrer-Hosken RA, Bertschinger HJ, Kirkpatrick JF, GroblerD, Lamberski N, Honneyman G, Ulrich T, 1999: Contraceptivepotential of the porcine zona pellucida vaccine in theAfrican elephant (Loxodonta africana). Theriogenology 52,835–846.Fayrer-Hosken RA, Dookwah HD, Brandon CI, 2000a:Immunocontrol in dogs. Anim Reprod Sci 60–61, 365–373.Fayrer-Hosken RA, Grobler D, Van Altena JJ, BertschingerHJ, Kirkpatrick JF, 2000b: Immunocontraception of Africanelephants. Nature 407, 149.Fayrer-Hosken RA, Barber MR, Crane M, Collins T, HodgdenR, 2002: Differences in immunocontraceptive responsesin white-tailed deer (Odocoileus virginianus) and goats(Capra hircus). Reprod Suppl 60, 125–129.Green DP, 1997: Three-dimensional structure of the zonapellucida. Rev Reprod 2, 147–156.Hannesdottir SG, Han X, Lund T, Singh M, Van Der Zee R,Roitt IM, Delves PJ, 2004: Changes in the reproductivesystem of male mice immunized with a GnRH-analogue conjugatedto mycobacterial hsp70. Reproduction 128, 365–371.Hardy CM, ten Have JF, Pekin J, Beaton S, Jackson RJ,Clydesdale G, 2003: Contraceptive responses of miceimmunized with purified recombinant mouse zona pellucidasubunit 3 (mZP3) proteins. Reproduction 126, 49–59.Hardy CM, Beaton S, Hinds LA, 2008: Immunocontraceptionin mice using repeated, multi-antigen peptides: immunizationwith purified recombinant antigens. Mol Reprod Dev75, 126–135.Henderson CJ, Hulme MJ, Aitken RJ, 1988: Contraceptivepotential of antibodies to the zona pellucida. J Reprod Fertil83, 325–343.Hedrick JL, 1996: Comparative structural and antigenicproperties of zona pellucida glycoproteins. J. Reprod. Fertil.Suppl. 50, 9–17.Hussein AM, Bourne FJ, 1984: Immunoglobulin concentrationsin pig follicular fluid. Int J Fertil 29, 54–57.Imboden I, Janett F, Burger D, Crowe MA, Hassig M, ThunR, 2006: Influence of immunization against GnRH onreproductive cyclicity and estrous behavior in the mare.Theriogenology 66, 1866–1875.Jewgenow K, Rohleder M, Wegner I, 2000: Differencesbetween antigenic determinants of pig and cat zonapellucida proteins. J. Reprod. Fertil. 119, 15–23.Jovine L, Qi H, Williams Z, Litscher E, Wassarman PM, 2002:The ZP domain is a conserved module for polymerization ofextracellular proteins. Nat Cell Biol 4, 457–461.Killian G, Miller L, Rhyan J, Doten H, 2006: Immunocontraceptionof Florida feral swine with a single-dose GnRHvaccine. Am J Reprod Immunol 55, 378–384.Kirkpatrick JF, Turner A, 2002: Reversibility of action andsafety during pregnancy of immunization against porcinezona pellucida in wild mares (Equus caballus). Reprod Suppl60, 197–202.Kirkpatrick JF, Turner A, 2003: Absence of effects fromimmunocontraception on seasonal birth patterns and foalsurvival among barrier island wild horses. J Appl Anim WelfSci 6, 301–308.Kirkpatrick JF, Liu IM, Turner JW, Jr, Naugle R, Keiper R,1992: Long-term effects of porcine zonae pellucidae immunocontraceptionon ovarian function in feral horses (Equuscaballus). J Reprod Fertil 94, 437–444.Kirkpatrick JF, Turner JW, Jr, Liu IK, Fayrer-Hosken R,1996: Applications of pig zona pellucida immunocontraceptionto wildlife fertility control. J Reprod Fertil Suppl 50,183–189.Kolle S, Dubois CS, Caillaud M, Lahuec C, Sinowatz F,Goudet G, 2007: Equine zona protein synthesis and ZPstructure during folliculogenesis, oocyte maturation, andembryogenesis. Mol Reprod Dev 74, 851–859.Lane VM, Liu IK, Casey K, van Leeuwen EM, Flanagan DR,Murata K, Munro C, 2007: Inoculation of female Americanblack bears (Ursus americanus) with partially purifiedporcine zona pellucidae limits cub production. ReprodFertil Dev 19, 617–625.Lee VH, Dunbar BS, 1993: Developmental expression of therabbit 55-kDa zona pellucida protein and messenger RNAin ovarian follicles. Dev Biol 155, 371–382.Lee VH, Schwoebel E, Prasad S, Cheung P, Timmons TM,Cook R, Dunbar BS, 1993: Identification and structuralcharacterization of the 75-kDa rabbit zona pellucidaprotein. J Biol Chem 268, 12412–12417.Litscher ES, Wassarman PM, 2007: Egg extracellular coatproteins: from fish to mammals. Histol Histopathol 22, 337–347.Litscher ES, Janssen WG, Darie CC, Wassarman PM, 2008:Purified mouse egg zona pellucida glycoproteins polymerizeinto homomeric fibrils under non-denaturing conditions. JCell Physiol 214, 153–157.Liu IK, Shivers CA, 1982: Antibodies to the zona pellucida inmares. J Reprod Fertil Suppl 32, 309–313.Liu IK, Bernoco M, Feldman M, 1989: Contraception inmares heteroimmunized with pig zonae pellucidae. J ReprodFertil 85, 19–29.Liu IK, Turner JW, Jr, van Leeuwen EM, Flanagan DR,Hedrick JL, Murata K, Lane VM, Morales-Levy MP, 2005:Persistence of anti-zonae pellucidae antibodies following asingle inoculation of porcine zonae pellucidae in thedomestic equine. Reproduction 129, 181–190.Lyda RO, Hall JR, Kirkpatrick JF, 2005: A comparison ofFreund’s Complete and Freund’s Modified Adjuvants usedwith a contraceptive vaccine in wild horses (Equus caballus).J Zoo Wildl Med 36, 610–616.Mahi-Brown CA, Yanagimachi R, Hoffman JC, Huang TT,Jr, 1985: Fertility control in the bitch by active immunizationwith porcine zonae pellucidae: use of different adjuvantsand patterns of estradiol and progesterone levels in estrouscycles. Biol Reprod 32, 761–772.Millar SE, Lader ES, Dean J, 1993: ZAP-1 DNA bindingactivity is first detected at the onset of zona pellucida geneexpression in embryonic mouse oocytes. Dev Biol 158, 410–413.Miller CC, Fayrer-Hosken RA, Timmons TM, Lee VH,Caudle AB, Dunbar BS, 1992: Characterization of equinezona pellucida glycoproteins by polyacrylamide gel electrophoresisand immunological techniques. J Reprod Fertil 96,815–825.Miller LA, Johns BE, Killian GJ, 2000: Immunocontraceptionof white-tailed deer with GnRH vaccine. Am J ReprodImmunol 44, 266–274.Miller LA, Rhyan JC, Drew M, 2004: Contraception ofbison by GnRH vaccine: a possible means of decreasingtransmission of brucellosis in bison. J Wildl Dis 40, 725–730.O’Rand MG, Widgren EE, Wang Z, Richardson RT, 2007:Eppin: an epididymal protease inhibitor and a target formale contraception. Soc Reprod Fertil Suppl 63, 445–453.Paterson M, Koothan PT, Morris KD, O’Byrne KT, BraudeP, Williams A, Aitken RJ, 1992: Analysis of the contraceptivepotential of antibodies against native and deglycosylatedporcine ZP3 in vivo and in vitro. Biol Reprod 46,523–534.Plumb G, Babiuk L, Mazet J, Olsen S, Rupprecht C, PastoretPP, Slate D, 2007: Vaccination in conservation medicine.Rev Sci Tech 26, 229–241.Polkinghorne I, Hamerli D, Cowan P, Duckworth J, 2005:Plant-based immunocontraceptive control of wildlife –‘‘potentials, limitations, and possums’’. Vaccine 23, 1847–1850.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Controlling Animal Populations Using Anti-Fertility Vaccines 185Redwood AJ, Harvey NL, Lloyd M, Lawson MA, Hardy CM,Shellam GR, 2007: Viral vectored immunocontraception:screening of multiple fertility antigens using murine cytomegalovirusas a vaccine vector. Vaccine 25, 698–708.Shivers CA, Metz CB, 1962: Inhibition of fertilization in frogeggs by univalent fragments of rabbit antibody. Proc. Soc.Exp. Biol. Med. 110, 385–387.Shivers CA, Dudkiewicz AB, Franklin LE, Fussell EN, 1972:Inhibition of sperm-egg interaction by specific antibody.Science 178, 1211–1213.Stoops MA, Liu IK, Shideler SE, Lasley BL, Fayrer-HoskenRA, Benirschke K, Murata K, van Leeuwen EM, AndersonGB, 2006: Effect of porcine zonae pellucidae immunisationon ovarian follicular development and endocrinefunction in domestic ewes (Ovis aries). Reprod Fertil Dev18, 667–676.Tacket CO, 2005: Plant-derived vaccines against diarrhealdiseases. Vaccine 23, 1866–1869.Tang J, Xie QX, Pan SP, Xiao LJ, Dong L, Zhang CX, SunCJ, 2003: [Expression of recombinant human zonapellucida-3 protein (rhZP3) in Pichia pastoris]. Sheng WuGong Cheng Xue Bao 19, 758–762.Tordo N, Bahloul C, Jacob Y, Jallet C, Perrin P, Badrane H,2006: Rabies: epidemiological tendencies and control tools.Dev Biol (Basel) 125, 3–13.Turkstra JA, van der Meer FJ, Knaap J, Rottier PJ, TeerdsKJ, Colenbrander B, Meloen RH, 2005: Effects of GnRHimmunization in sexually mature pony stallions. AnimReprod Sci 86, 247–259.Turner A, Kirkpatrick JF, 2002: Effects of immunocontraceptionon population, longevity and body condition in wildmares (Equus caballus). Reprod Suppl 60, 187–195.Turner JW, Jr, Liu IK, Flanagan DR, Bynum KS, RutbergAT, 2002: Porcine zona pellucida (PZP) immunocontraceptionof wild horses (Equus caballus) in Nevada: a 10 yearstudy. Reprod Suppl 60, 177–186.Wassarman PM, Mortillo S, 1991: Structure of the mouse eggextracellular coat, the zona pellucida. Int Rev Cytol 130,85–110.Wassarman PM, Jovine L, Qi H, Williams Z, Darie C, LitscherES, 2005: Recent aspects of mammalian fertilizationresearch. Mol Cell Endocrinol 234, 95–103.Widders PR, Stokes CR, David JS, Bourne FJ, 1984:Quantitation of the immunoglobulins in reproductive tractsecretions of the mare. Res Vet Sci 37, 324–330.Willis L, Heusner G, Warren R, Kessler D, Fayrer-Hosken R,1994: Equine immunocontraception using porcine zonapellucida: a new method for the remote delivery andcharacterization of the immune response. J Eq Vet Sci 14,364–370.Yanagimachi R, Winkelhake J, Nicolson GL, 1976:Immunological block to mammalian fertilization: survivaland organ distribution of immunoglobulin which inhibitsfertilization in vivo. Proc. Natl. Acad. Sci. USA 73, 2405–2408.Zhao M, Boja ES, Hoodbhoy T, Nawrocki J, Kaufman JB,Kresge N, Ghirlando R, Shiloach J, Pannell L, Levine RL,Fales HM, Dean J, 2004: Mass spectrometry analysisof recombinant human ZP3 expressed in glycosylationdeficientCHO cells. Biochemistry 43, 12090–12104.Author’s address (for correspondence): Richard Fayrer-Hosken,Department of Large Animal Medicine, College of Veterinary Medicine,University of Georgia, Athens, GA 30602-7385, USA. E-mail:rfh@vet.uga.eduConflict of interest: The author has not declared any conflict ofinterests.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 186–192 (2008); doi: 10.1111/j.1439-0531.2008.01160.xISSN 0936-6768The Expanding Role of Recombinant Gonadotropins in Assisted ReproductionTE Adams 1 and I Boime 21 Department of Animal Science, University of California, Davis, CA, USA; 2 Department of Molecular Biology and Pharmacology, WashingtonUniversity School of Medicine, St Louis, MO, USAContentsUsing recombinant gonadotropins for assisted reproduction ofdomestic species is still in its infancy. Yet, the purity, potencyand pathogen-free nature of recombinant gonadotropins makethem attractive alternatives to tissue-derived gonadotropicagents. In this study, the authors summarize the work to dateusing recombinant gonadotropins to enhance the - fertility ofdomestic animals and they discussed their recent studiesexamining the biopotency of single chain analogues of humangonadotropins. In these studies, single chain analogues offollicle stimulating hormone (Fca), chorionic gonadotropin(CGba) or a gonadotropin construct with dual activity(FcCGba) were administered to sheep pre-treated with antiseradirected against GnRH. Ovulation was induced 3 days afteranalogue administration using hCG (1000 IU, iv). AlthoughFca or CGba alone induced only modest oestradiol productionduring the pre-hCG period, serum concentrations of oestradiolwere markedly increased (p < 0.05) 3 days after administrationof FcCGba or the Fca +CGba combination. Finalovarian weight was significantly increased (p < 0.05) inanimals receiving Fca, Fca +CGba or FcCGba. Collectively,these observations demonstrate that the single chain analoguesof the human gonadotropins are active in sheep.IntroductionThe multiple ovulation-embryo transfer (MOET) technologyis one of the most important management toolsthat producers can use to accelerate the pace of geneticimprovement in commercial herds and flocks (Galliet al. 2003; Hasler 2003). Yet, the use of the MOETtechnology in the animal industries in North America isapproaching a plateau (Hasler 2003). One critical factorlimiting the more general acceptance of this technologyis its reliance on hormone-induced superovulation andthe variation and inconsistency inherent in that process(Kanitz et al. 2002). Until recently, the exogenoushormones used to induce the superovulatory responsewere gonadotropins derived from pituitary or placentaltissue. The heterogeneous nature of tissue-derivedgonadotropins and intrinsic difference between animalsin follicular development result in marked animal toanimal variation in the magnitude of the superovulatoryresponse (Monniaux et al. 1983). The purity andpotency of recombinant gonadotropins may reduce thevariation associated with hormone-induced superovulationand, thereby, facilitate more general acceptance ofthis valuable tool for genetic improvement.Non-Recombinant GonadotropinsConventional, tissue-derived gonadotropins used in theanimal industries (Boland et al. 1991; Kanitz et al. 2002)include ovine and porcine follicle stimulating hormone(FSH) and equine and human chorionic gonadotropin(eCG and hCG, respectively). Pituitary and placentalgonadotropins are composed of two subunits, commonlydesignated as the a and b subunits (Pierce andParsons 1981; Hearn and Gomme 2000). The a subunitis shared by all gonadotropins produced by a givenspecies, while the b subunit of each gonadotropin isunique. For example, the dimeric (ab) composition ofhuman LH, FSH and CG consists of a hormone specificb subunit and a common a subunit (Fig. 1). Althoughthe a and b subunits are not covalently linked, theconfiguration of each subunit is stabilized by severalintra-chain disulphide bridges.A feature of the gonadotropins that markedly influencespotency and duration of response is the degree ofglycosylation. Indeed, a significant portion of the massof gonadotropins resides in the oligosaccharide chainsadded during co- and post-translational processing.Pituitary-derived gonadotropins contain three or fourasparagine-associated (N-linked) oligosaccharidegroups. The a subunits of human, bovine and ovinegonadotropins carry oligosaccharide chains attached totwo asparagine residues. Similarly, the b subunits carryone (hLHb) or two (hFSHb) N-linked carbohydratechains. Human chorionic gonadotropin, like FSH, hastwo N-linked chains associated with the b subunit. Inaddition, the b subunit of hCG contains four oligosaccharidechains linked to serine residues (O-linkedchains) located in the carboxy-terminal portion of CGb.The N-linked glycans of LH and FSH are a heterogeneousmix of di- and tri-branched oligosaccharidechains (Green and Baenziger 1988a; b). Heterogeneity isalso evident at the branch termini, with terminalsulphate residues common in LH while sialylatedtermini predominate in FSH (Green and Baenziger1988b). The degree of sialylation also varies acrossspecies. For example, 88% of the oligosaccharide chainsof human FSH end with sialic acid, while only 38% oftermini in ovine FSH are sialylated. In contrast, theN-linked oligosaccharide chains in hCG are much moreuniform and generally carry terminal sialic acid residues(Kessler et al. 1979b). The O-linked glycans of hCG arealso highly sialylated (Kessler et al. 1979a).The carbohydrate portion of the gonadotropinsinfluences the folding, assembly, secretion, clearanceand biological activity of the gonadotropins (Thotakuraand Blithe 1995). The liver plays an active role in theclearance and degradation of the gonadotropins (Fieteet al. 1991). Indeed, hepatocytes contain high concentrationsof a receptor that specifically recognizes thesulphated, but not the sialylated, oligosaccharide chainsof the gonadotropins (Roseman and Baenziger 2000).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Recombinant Gonadotropins in Assisted Reproduction 187Fig. 1. Tissue-derived gonadotropins (LH, FSH and hCG) are composedof non-covalently linked a and b subunits each of which carryone (LHb) or two (a, FSHb and CGb) N-linked oligosaccharide chains(•). The carboxy-terminal peptide (CTP) of hCG (dashed line) alsocarries four O-linked glycans (*). In single chain gonadotropins, the aand b domains are linked by CTP. The chimeric single chaingonadotropin with dual activity (FCCGa) contains FSHb, CGb anda domains linked by CTPAs a consequence, LH, a sulphated gonadotropin, has ashort half-life. Conversely, the richly sialylated placentalcounterpart, hCG, has a much longer functional life.Deficiencies of Non-RecombinantGonadotropinsThe heterogeneous nature of tissue-derived gonadotropinpreparations results in significant variation amongbatches with respect to purity and biopotency (Murphyet al. 1984; Phillips et al. 1993; Kanitz et al. 2002). Inaddition, exogenous gonadotropins of equine or porcineorigin can induce an immune response that may limit therepetitive use of these preparations in sheep and cattle(Bodin et al. 1997; Drion et al. 2001). Furthermore, theshort half-life of ovine, porcine and human FSHpreparations necessitates continuous or repetitiveadministration, increasing the time and expense associatedwith animal sorting and handling (Alcivar et al.1992; Kanitz et al. 2002). Conversely, the long life ofeCG may compromise embryo viability (Martinuk et al.1991) and often requires the administration of anti-eCGsera during the post-ovulatory period (Gonzalez et al.1994). In addition, tissue-derived gonadotropins oftencontain differing proportions of LH and FSH activity(Kanitz et al. 2002). This heterogeneity likely contributesto variation in the physiological response (Murphyet al. 1984). A growing concern is the possibility thattissue-derived gonadotropins may harbour viruses, bacteriaor prions that may result in the inadvertentdevelopment of disease in treated animals (Reichl et al.2002; Galli et al. 2003; Matorras and Rodriguez-Escudero2003). Indeed, certain European countries havebanned the use of pituitary-derived gonadotropins insuperovulation protocols (Galli et al. 2003).Recombinant GonadotropinsRecombinant forms of LH and FSH are generated byimmortalized cells transfected with gene constructsencoding the a and b subunits (Kaetzel et al. 1985;Keene et al. 1989; Lunenfeld 2004). The transfected cellsare maintained in culture and represent a renewablesource of dimeric gonadotropin (Howles 1996). Olijveand co-workers (Olijve et al. 1996) transfected cells witha construct that incorporated the DNA encoding the aand b subunits of human FSH into a single vector. Inaddition, multiple copies of the composite constructwere stably inserted in the transfected cells (Olijve et al.1996), increasing the efficiency of dimeric FSH productionin the cell culture system. In addition, recentadvances in manufacturing and purification generaterecombinant gonadotropin preparations that are uniformin isoform profile and glycan composition (Driebergenand Baer 2003; Bassett and Driebergen 2005).These advances in recombinant gonadotropin technologyaddress many of the concerns that have developedregarding the use of tissue-derived gonadotropins.Current recombinant technology produces pharmaceuticalgrade LH and FSH of defined amino acid andcarbohydrate composition (Olijve et al. 1996). Currentpurification procedures ensure that the purity, biopotencyand batch-to-batch consistency are very high(Daya 2004). Furthermore, the generation of therecombinants in a serum-free system by immortalizeddisease-free cells ensures that the recombinant gonadotropinsare not vectors of disease. An additional benefitof the recombinant gonadotropins is that the high purityand consistent biopotency across batches permits dosingbased on mass (Driebergen and Baer 2003).The gene constructs encoding the gonadotropins aregenerally inserted into a cell line derived from ovariantissue of a Chinese hamster. The Chinese hamster ovary(CHO) cells are used from this purpose because theyaccept transfection easily (Olijve et al. 1996). In addition,CHO cells contain the intracellular machineryrequired to form the intra-chain disulphide bridges andcorrectly complete the folding and assembly of the dimer(Howles 1996). Furthermore, CHO cells constitutivelysecrete the recombinant protein into the culturemedium, a characteristic that facilitates gonadotropinrecovery and purification. Most importantly, the patternof glycosylation in CHO cells mirrors that found inpituitary-derived gonadotropins. Yet, CHO cells lackthe sulphotransferase contained in gonadotrope cells(Smith et al. 1992). As a consequence, the oligosaccharidechains of recombinant gonadotropins carry terminalsialic acid groups, rather than the terminal sulphateresidues that predominate in pituitary-derived gonadotropins(Smith et al. 1990; Baenziger et al. 1992). Thepractical consequence of sialic acid termini is that therecombinant gonadotropins are afforded protectionfrom the hepatic proteins that induce the rapid degradationof gonadotropins with sulphated termini (Baenzigeret al. 1992; Roseman and Baenziger 2000).Use of Recombinant Dimeric GonadotropinsThe efficacy of recombinant FSH in the treatment ofinfertility in humans was examined in a series of studiesconducted in the 1990s (Devroey et al. 1992; Germondet al. 1992) and the use of these recombinant gonadotropinpreparations has become common in in vitroÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

188 TE Adams and I Boimefertilization programs (Daya 2004). In the animalindustries, recombinant forms of bovine (Wilson et al.1993; Wehrman et al. 1996) and human (Takagi et al.2001) FSH (rbFSH and rhFSH, respectively) have beenused to induce superovulation in beef cattle. Recombinantporcine FSH (rpFSH) has also been produced andthe biological activity has been verified using an in vitroSertoli cell bioassay and an in vivo assay based on follicledevelopment in hypophysectomized mice treated withrpFSH (Inaba et al. 1997).The comprehensive studies of Wilson and co-workers(Wilson et al. 1993) demonstrated that rbFSH was aseffective as pituitary-derived FSH in promoting thesuperovulatory response in mature cows. The treatmentprotocol required rbFSH administration (im) at 12 hintervals for 3–5 days. The optimal response was notedat a total dose of 24 mg rbFSH. This level of rbFSH iscomparable to the total dose of purified pituitary FSHrequired to induce superovulation in beef and dairycows (Alcivar et al. 1992; Kanitz et al. 2002). Thedevelopment of multiple pre-ovulatory follicles was alsoinduced in heifers receiving twice daily injections ofrhFSH over a 4-day-treatment period (Takagi et al.2001). The magnitude of the rhFSH-induced responsewas comparable to the follicular response induced byeCG. Yet, when compared with eCG-treated heifers,serum concentrations of oestradiol during the preovulatoryperiod were significantly reduced in heifersreceiving rhFSH. As a consequence, the induction of thepre-ovulatory surge of LH was delayed. Takagi and coworkerssuggest that endogenous production of LH isnot sufficient to support oestradiol synthesis duringrhFSH-induced follicle development. Therefore, concurrentadministration of LH and rhFSH may berequired to achieve an ovulatory response that iscomparable to that induced by eCG.Dimeric CTP-Modified GonadotropinsIn contrast to pituitary-derived gonadotropins, placentalgonadotropins are cleared slowly and, as a consequence,the duration of response induced by eCG orhCG is much longer than the response induced by LH orFSH (Martinuk et al. 1991; Baenziger et al. 1992). Thelong half-life of the placental gonadotropins is due, inpart, to a carboxyl terminal peptide (CTP) extension ofthe b subunit that is subject to extensive O-linkedglycosylation (Kessler et al. 1979a; Bousfield et al.2001). The critical impact of the CTP on gonadotropinclearance was demonstrated by Matzuk and co-workers(Matzuk et al. 1990), who produced a truncated form ofhCGb lacking the CTP domain. Although the truncatedvariant of hCG was recognized with high affinity by theLH receptor, the in vivo potency was dramaticallyreduced by deletion of the CTP domain. Similarly,incorporating a CTP domain into the b subunits of FSHor TSH markedly increases the half-life of the CTPmodifiedforms of FSH and TSH (Fares et al. 1992;Joshi et al. 1995).Although the bioactivity of CTP-modified dimericFSH has not been examined in domestic species, theefficacy of an hFSH analogue incorporating the CTPdomain (hFSH-CTP) has been examined in humans.The hFSH analogue is effective in promoting spermatogenesisand follicle development (Bouloux et al.2001; Balen et al. 2004). Importantly, in comparisonwith rhFSH, the rate of clearance of hFSH-CTP ismarkedly reduced and development of pre-ovulatoryfollicles is induced by a single injection of the FSHanalogue. This is in contrast to the repetitive injectionschedule that is required to promoted follicle maturationusing pituitary-derived FSH. Collectively, thesedata demonstrate that the CTP domain extends thefunctional life of FSH and simplifies the ovulationinduction protocol by reducing the frequency ofadministration.Single Chain Analogues of LH and FSHA recent advance in the production of recombinantgonadotropins is the generation of single chain analoguesof the normally dimeric proteins (Garcia-Campayoand Boime 2001). The single chain technologyinvolves transfection of CHO cells with a plasmid thatcarries DNA encoding both subunits combined in asingle gene construct. The linker sequence that joins thea and b domains of the construct encodes the CTPportion of hCGb. The protein product secreted bytransfected CHO cells is a single peptide chain thatcontains a and b domains linked through the CTPsegment. To date, single chain analogues of hFSH(FSHb–CTP–a; Sugahara et al. 1996), hLH (LHb–CTP–a; Garcia-Campayo et al. 1997), equine LH(eLHb–CTP–a; Jablonka-Shariff et al. 2007) and hCG(CGb–a; Sugahara et al. 1995) have been developed.The single chain analogues of the gonadotropins arerecognized by the appropriate receptor (LH or FSHreceptor) with high affinity and the single chain analoguesinduce the normal intracellular second messengerresponse. The bioactivity of the single chain analoguesdemonstrates that the folding of the a and b componentsinto the configuration required to associate with thegonadotropin receptor is not impaired by the singlechain organization of the a and b domains. The correctfolding is likely facilitated by the intervening CTPsequence. Importantly, the high proportion of prolineand serine residues in the CTP sequence makes thissegment highly flexible and allows the a and b domainsof the single chain to fold into the active configuration.An additional benefit of incorporating the CTPsequence into the single chain is that, much like thedimeric CTP-modified gonadotropins discussed above,the CTP portion of the single chain reduces the rate ofclearance. An additional practical consequence of producinggonadotropins as single chain proteins is thatdimerization, the rate limiting step in the synthesis ofconventional gonadotropins, is bypassed. This uniquefeature of the single chain proteins enhances theefficiency of synthesis and secretion by transfectedCHO cells (Sugahara et al. 1996).The recent studies of Jablonka-Shariff and co-workersexamined the activity of a single chain analogue ofequine LH (Jablonka-Shariff et al. 2007). The analoguewas as potent as pituitary-derived eLH in promotingtestosterone production by equine Leydig cells inculture. The single chain eLH analogue also increasedÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Recombinant Gonadotropins in Assisted Reproduction 189the serum concentration of testosterone after administrationto stallions. The single chain analogue was alsoas effective as hCG in inducing ovulation in mares(Yoon et al. 2007). Indeed, ovulation was evident within48 h of analogue administration in mares with a maturefollicle (>35 mm) at the time of administration. Ovulationwas noted within the same temporal window inonly 15% of control mares receiving vehicle alone. Yoonand colleagues suggest that the single chain analogue ofeLH may be useful in accelerating ovulation and,thereby, facilitating timed mating and inseminationusing frozen ⁄ thawed semen.The bioactivity of single chain analogues of hFSH hasbeen assessed in rodent (Weenen et al. 2004) andprimate (Klein et al. 2002) models. Administration ofa single dose of the FSH analogue to prepuberal ratsinduced a significant increase in ovarian weight andinhibin secretion (Weenen et al. 2004) and promoted thedevelopment of multiple antral follicles (Trousdale et al.2007). In contrast, ovarian weight and inhibin productiondid not differ from saline-treated control animals inrats receiving a single injection of dimeric rhFSH. Thisindicates that the functional life of the single chainanalogue is greater than that of dimeric FSH. Indeed,direct assessment of the half-life indicated that theclearance of the single chain analogue was three timesslower than the clearance of dimeric FSH in the rodentmodel. Follicle development and oestradiol productionwere also increased in monkeys receiving a singleinjection of the single chain FSH analogue (Klein et al.2002). As in the rodent model, the half-life of the FSHanalogue was markedly increased relative to the half-lifeof recombinant dimeric FSH. Collectively, these studiesdemonstrate that the single chain form of FSH is activeand long-lived.Gonadotropins with Dual ActivityA novel variation of the single chain technology involvesthe incorporation of two or more different b subunitdomains into the single chain protein (Kanda et al.1999; Garcia-Campayo and Boime 2001). When examinedusing in vitro assay systems a single chaingonadotropin with a, hFSHb and hCGb domains(FSHb–CTP–CGb–a) interacted with LH and FSHreceptors and induced cAMP production that wascomparable to the level induced by dimeric LH andFSH (Kanda et al. 1999). The dual activity of thechimeric gonadotropin suggests that the single chainprotein can assume a secondary structure that isrecognized by both gonadotropin receptors. Yet, recentstudies indicate that the chimeric proteins assume twodifferent conformations, one that is recognized by theFSH receptor and a second configuration that permitsrecognition by the LH receptor (Garcia-Campayo et al.2004). In other words, the single chain proteins secretedfrom transfected CHO cells represent a heterogeneouspopulation of conformations, some with the capacity toactivate the LH receptor and others capable of activatingthe FSH receptor. The dual activity of the chimerichuman gonadotropins is similar to the bifunctionalpotential of eCG. Indeed, the dually active chimericgonadotropin was as effective as eCG in promotingfollicle development, enhancing of ovarian weight andaugmenting aromatase activity in immature mice (Jablonka-Shariffet al. 2006). The ovulation inductionaction of hCG was also mimicked by the dually activechimeric gonadotropin in eCG-primed immature mice.In our recent studies, the authors have compared thebiopotency of the dually active human gonadotropinwith single chain analogues of hFSH and hCG (Lemkeet al. 2008). The research model we used employed ewestreated with progesterone-containing vaginal inserts(CIDR’s) and anti-GnRH sera to suppress secretion ofendogenous gonadotropins (Fig. 2). Ewes received asingle injection of the dually active gonadotropin(FSHb–CTP–CGb–a; 5 IU⁄ kg, iv), single chain analoguesof hFSH (FSHb–CTP–a; Fca; 5IU⁄ kg, iv) orCG (CGb–a; CGba; 5IU⁄ kg, iv), or the FSH and CGanalogues in combination (Fca +CGba; both at5IU⁄ kg, iv) at the time of CIDR removal. Controlanimals received vehicle alone. Ovulation was inducedby hCG (1000 IU, iv) injection 3 days after administrationof the single chain gonadotropins.As oestradiol production requires the concertedaction of both LH and FSH, the authors used oestradiolsecretion as a measure of the bifunctional capacity of thesingle chain gonadotropins. Significant oestradiol secretionwas not evident in ewes receiving Fca or CGbaalone. Yet, serum levels of oestradiol were markedlyincreased in ewes receiving Fca and CGba in combination.The results of this study demonstrate that thesingle chain analogues of hFSH and hCG are active insheep. These data also clearly illustrate that the singlechain construct that incorporates FSHb and CGbdomains has both FSH and LH activity (Table 1).Similarly, follicle development was promoted by thedually active chimeric gonadotropins and the combinedtreatment with Fca and CGba. This augmented follicularresponse is indicated by the increase in ovarianweight and corpora lutea number that were noted inovarian tissue collected 11 days after administration ofFig. 2. The oestrous activity of yearling ewes was synchronized usingprogesterone-containing vaginal inserts (CIDRs). Secretion of endogenousgonadotropins was suppressed by passive immunization withantisera directed against GnRH (Anti-GnRH) 1 day prior to CIDRremoval. Animals (n = 6 ewes ⁄ group) received recombinant gonadotropin(CGba, Fca, FcCGba or CGba +Fca in combination) atCIDR removal. Chimeric gonadotropins were administered at a doseof 5 IU ⁄ kg (iv). Control animals received conditioned media of nontransfectedCHO-cells. Ovulation was induced by hCG (1000 IU, iv)on day 3 and ovarian tissue was collected on day 11Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

190 TE Adams and I BoimeTable 1. Effect of single chain human gonadotropins on pre-ovulatoryserum concentrations of oestradiol (concentration 3 days after administrationof single chain gonadotropin) and final ovarian weight andcorpora lutea per animalTreatment(5 IU ⁄ kg, iv) nPre-ovulatoryoestradiol (pg ⁄ ml)Ovarianweight (g)Corpora lutea(CLs ⁄ animal)Control 6 1.1 ± .6 a 1.3 ± .1 a 0 aCGba 6 3.2 ± 1.3 a 2.8 ± .2 b 1.0 ± .4 bFCa 6 3.7 ± 1.1 a 14.0 ± 1.4 c 9.5 ± 2.4 cFCa +CGba 6 34.4 ± 9.8 b 11.1 ± 1.0 c 11.3 ± 2.7 cFcCGba 6 44.0 ± 8.2 b 10.5 ± 1.0 c 6.7 ± 2.7 cOvarian parameters were assessed 11 days after administration of single chaingonadotropins. Values with different superscripts differed significantly(p < 0.05).the single chain gonadotropins. Although Fca alone wasunable to induce a significant increase in oestradiolproduction, this single chain analogue of hFSH didpromote the development of multiple follicles that wereinduced to ovulate by the hCG challenge administered3 days after single chain gonadotropin. These observationsindicate that follicle development and oestradiolbiosynthesis are separable phenomena. In addition, thisobservation illustrates that a single injection of the singlechain gonadotropins is sufficient to induce the superovulatoryresponse. This is in marked contrast to themultiple injection regimen required to induce the superovulatoryresponse using pituitary-derived or dimericrecombinant FSH.The Next Generation of RecombinantGonadotropinsThe single chain technology opens the possibility oflinking multiple bioactive domains in a single recombinantprotein. The development of recombinant proteinsincorporating portions of LH and FSH to generateproteins with dual activity is a clear example of thepotential of the single chain technology. Modificationsof this technology may result in the generation of singlechain gonadotropins specifically tailored to meet thefunctional requirements of a targeted physiologicresponse. For example, a single chain gonadotropinwith dual activity may be required to induce follicledevelopment and ovulation in the prepuberal period ornon-breeding season while a long-lived single chaingonadotropin with singular FSH activity may be optimalto induce a superovulatory response during thebreeding season. Similarly, a very long-lived gonadotropinmay be useful in promoting spermatogenesis inelite breeding males while a single chain with a moremoderate functional life may be optimal for follicledevelopment. Combining the potential of molecularbiology with the utility of the single chain technologywill allow scientists and production engineers to developselective analogues of the gonadotropins that aredesigned to address a specific application.The single chain technology may also be used togenerate functional proteins that include non-gonadotropindomains. For example, Trousdale and coworkers(Trousdale et al. 2007) recently generated a single chainconstruct that combined the functional domain ofvascular endothelial growth factor (VEGF) with adomain encoding the single chain variant of FSH. Theresulting bifunctional protein displayed angiogenic andFSH activity in in vitro assay systems and increasedfollicle development and perifollicular angiogenesisin vivo. Similarly, multi-functional proteins with LH,FSH and TSH activity have been generated using thesingle chain technology (Garcia-Campayo et al. 2002).Collectively, these observations demonstrate that thesingle chain technology is not limited to gonadotropinsbut may also be expanded to include other functionaldomains that may synergize with the gonadotropincomponent to enhance the fertility of domestic species.Another recent advance (Low et al. 2005) is thelinkage of the single chain form of FSH to the Fcdomain of immunoglobin G 1 (IgG 1 ). The single chainFSH-Fc protein retained FSH activity and the inclusionof the Fc domain markedly prolonged the functional lifeof the chimeric protein. In addition, the inclusion of theFc domain permitted Fc receptor mediated absorptionacross epithelial barriers in the intestine and lungs. Thisopens the possibility of administering multifunctionalgonadotropin constructs orally or by inhalation. IndeedLow and coworkers (Low et al. 2005) demonstrated thatthe physiological response to FSH (increased testisweight and increased inhibin production) was evident inrats and monkeys after oral or pulmonary administration.ConclusionTissue-derived gonadotropins have facilitated the developmentand application of reproductive technologiesthat increase the productivity of herds and flocks byreducing management and labour costs and increasingthe pace of genetic improvement. Yet, variation in thepotency, purity and availability of natural gonadotropinsmay limit further advances. The high purity,potency and longevity of recombinant dimeric gonadotropinsand novel single chain gonadotropin analoguesmake these gonadotropins practical and effective alternativesto tissue-derived preparations in superovulatoryprotocols and out-of-season breeding programs.AcknowledgementsThis project was supported by National Research Initiative CompetitiveGrant 5-35203-16274 from the USDA Cooperative StateResearch, Education, and Extension Service Animal ReproductionProgram and the California Agricultural Experiment Station.ReferencesAlcivar AA, Maurer RR, Anderson LL, 1992: Endocrinechanges in beef heifers superovulated with follicle-stimulatinghormone (FSH-P) or human menopausal gonadotropin.J Anim Sci 70, 224–231.Baenziger JU, Kumar S, Brodbeck RM, Smith PL, BeranekMC, 1992: Circulatory half-life but not interaction with thelutropin ⁄ chorionic gonadotropin receptor is modulated bysulfation of bovine lutropin oligosaccharides. Proc NatlAcad Sci USA 89, 334–338.Balen AH, Mulders AG, Fauser BC, Schoot BC, RenierMA, Devroey P, Struijs MJ, Mannaerts BM, 2004:Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Recombinant Gonadotropins in Assisted Reproduction 191Pharmacodynamics of a single low dose of long-actingrecombinant follicle-stimulating hormone (FSH-carboxyterminal peptide, corifollitropin alfa) in women with WorldHealth Organization group II anovulatory infertility. J ClinEndocrinol Metab 89, 6297–6304.Bassett RM, Driebergen R, 2005: Continued improvements inthe quality and consistency of follitropin alfa, recombinanthuman FSH. Reprod Biomed Online 10, 169–177.Bodin L, Drion PV, Remy B, Brice G, Cognie Y, Beckers JF,1997: Anti-PMSG antibody levels in sheep subjected annuallyto oestrus synchronisation. Reprod Nutr Dev 37, 651–660.Boland MP, Goulding D, Roche JF, 1991: Alternativegonadotrophins for superovulation of cattle. Theriogenology35, 5–17.Bouloux PM, Handelsman DJ, Jockenhovel F, Nieschlag E,Rabinovici J, Frasa WL, de Bie JJ, Voortman G, Itskovitz-Eldor J, 2001: First human exposure to FSH-CTP inhypogonadotrophic hypogonadal males. Hum Reprod 16,1592–1597.Bousfield GR, Butnev VY, Butnev VY, 2001: Identification oftwelve O-glycosylation sites in equine chorionic gonadotropinbeta and equine luteinizing hormone ss by solid-phaseEdman degradation. Biol Reprod 64, 136–147.Daya S, 2004: Follicle-stimulating hormone in clinical practice:an update. Treat Endocrinol 3, 161–171.Devroey P, Van Steirteghem A, Mannaerts B, CoelinghBennink H, 1992: First singleton term birth after ovariansuperovulation with rhFSH. Lancet 340, 1108–1109.Driebergen R, Baer G, 2003: Quantification of follicle stimulatinghormone (follitropin alfa): is in vivo bioassay stillrelevant in the recombinant age? Curr Med Res Opin 19, 41–46.Drion PV, DeRoover R, Houtain JY, McNamara EM, RemyB, Sulon J, Beckers JF, 2001: Increase of plasma eCGbinding rate after administration of repeated high dose ofeCG to cows. Reprod Nutr Dev 41, 207–215.Fares FA, Suganuma N, Nishimori K, LaPolt PS, Hsueh AJ,Boime I, 1992: Design of a long-acting follitropin agonist byfusing the C-terminal sequence of the chorionic gonadotropinbeta subunit to the follitropin beta subunit. Proc NatlAcad Sci USA 89, 4304–4308.Fiete D, Srivastava V, Hindsgaul O, Baenziger JU, 1991: Ahepatic reticuloendothelial cell receptor specific for SO4-4GalNAc beta 1,4GlcNAc beta 1,2Man alpha that mediatesrapid clearance of lutropin. Cell 67, 1103–1110.Galli C, Duchi R, Crotti G, Turini P, Ponderato N, Colleoni S,Lagutina I, Lazzari G, 2003: Bovine embryo technologies.Theriogenology 59, 599–616.Garcia-Campayo V, Boime I, 2001: Novel recombinantgonadotropins. Trends Endocrinol Metab 12, 72–77.Garcia-Campayo V, Sato A, Hirsch B, Sugahara T, Muyan M,Hsueh AJ, Boime I, 1997: Design of stable biologicallyactive recombinant lutropin analogs. Nat Biotechnol 15,663–667.Garcia-Campayo V, Kumar TR, Boime I, 2002: Thyrotropin,follitropin, and chorionic gonadotropin expressed as a singlemultifunctional unit reveal remarkable permissiveness inreceptor–ligand interactions. Endocrinology 143, 3773–3778.Garcia-Campayo V, Jablonka-Shariff A, Boime I, 2004: Asingle-chain bifunctional gonadotropin analog is secretedfrom Chinese hamster ovary cells as two distinct bioactivespecies. J Biol Chem 279, 44286–44293.Germond M, Dessole S, Senn A, Loumaye E, Howles C,Beltrami V, 1992: Successful in-vitro fertilisation andembryo transfer after treatment with recombinant humanFSH. Lancet 339, 1170.Gonzalez A, Wang H, Carruthers TD, Murphy BD, MapletoftRJ, 1994: Superovulation in the cow with pregnant mareserum gonadotrophin: effects of dose and antipregnant mareserum gonadotrophin serum. Can Vet J 35, 158–162.Green ED, Baenziger JU, 1988a: Asparagine-linked oligosaccharideson lutropin, follitropin, and thyrotropin. I. Structuralelucidation of the sulfated and sialylatedoligosaccharides on bovine, ovine, and human pituitaryglycoprotein hormones. J Biol Chem 263, 25–35.Green ED, Baenziger JU, 1988b: Asparagine-linked oligosaccharideson lutropin, follitropin, and thyrotropin. II. Distributionsof sulfated and sialylated oligosaccharides onbovine, ovine, and human pituitary glycoprotein hormones.J Biol Chem 263, 36–44.Hasler JF, 2003: The current status and future of commercialembryo transfer in cattle. Anim Reprod Sci 79, 245–264.Hearn MT, Gomme PT, 2000: Molecular architecture andbiorecognition processes of the cystine knot protein superfamily:part I. The glycoprotein hormones. J Mol Recognit13, 223–278.Howles CM, 1996: Genetic engineering of human FSH(Gonal-F). Hum Reprod Update 2, 172–191.Inaba T, Mori J, Ohmura M, Kato Y, Tomizawa K, Kato T,Ihara T, Sato I, Ueda S, 1997: Baculovirus-insect cellproduction of bioactive porcine FSH. Theriogenology 47,491–499.Jablonka-Shariff A, Kumar TR, Eklund J, Comstock A,Boime I, 2006: Single-chain, triple-domain gonadotropinanalogs with disulfide bond mutations in the alpha-subunitelicit dual follitropin and lutropin activities in vivo. MolEndocrinol 20, 1437–1446.Jablonka-Shariff A, Roser JF, Bousfield GR, Wolfe MW,Sibley LE, Colgin M, Boime I, 2007: Expression andbioactivity of a single chain recombinant equine luteinizinghormone (reLH). Theriogenology 67, 311–320.Joshi L, Murata Y, Wondisford FE, Szkudlinski MW, DesaiR, Weintraub BD, 1995: Recombinant thyrotropin containinga beta-subunit chimera with the human chorionicgonadotropin-beta carboxy-terminus is biologically active,with a prolonged plasma half-life: role of carbohydrate inbioactivity and metabolic clearance. Endocrinology 136,3839–3848.Kaetzel DM, Browne JK, Wondisford F, Nett TM, ThomasonAR, Nilson JH, 1985: Expression of biologically activebovine luteinizing hormone in Chinese hamster ovary cells.Proc Natl Acad Sci USA 82, 7280–7283.Kanda M, Jablonka-Shariff A, Sato A, Pixley MR, Bos E,Hiro’oka T, Ben-Menahem D, Boime I, 1999: Geneticfusion of an alpha-subunit gene to the follicle-stimulatinghormone and chorionic gonadotropin-beta subunit genes:production of a bifunctional protein. Mol Endocrinol 13,1873–1881.Kanitz W, Becker F, Schneider F, Kanitz E, Leiding C,Nohner HP, Pohland R, 2002: Superovulation in cattle:practical aspects of gonadotropin treatment and insemination.Reprod Nutr Dev 42, 587–599.Keene JL, Matzuk MM, Otani T, Fauser BC, Galway AB,Hsueh AJ, Boime I, 1989: Expression of biologically activehuman follitropin in Chinese hamster ovary cells. J BiolChem 264, 4769–4775.Kessler MJ, Mise T, Ghai RD, Bahl OP, 1979a: Structureand location of the O-glycosidic carbohydrate units ofhuman chorionic gonadotropin. J Biol Chem 254, 7909–7914.Kessler MJ, Reddy MS, Shah RH, Bahl OP, 1979b: Structuresof N-glycosidic carbohydrate units of human chorionicgonadotropin. J Biol Chem 254, 7901–7908.Klein J, Lobel L, Pollak S, Ferin M, Xiao E, Sauer M,Lustbader JW, 2002: Pharmacokinetics and pharmacodynamicsof single-chain recombinant human follicle-stimulatinghormone containing the human chorionicÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

192 TE Adams and I Boimegonadotropin carboxyterminal peptide in the rhesus monkey.Fertil Steril 77, 1248–1255.Lemke EP, Adams BM, Jablonka-Shariff A, Boime I, AdamsTE, 2008: Using novel monomeric gonadotropins withsingular (FSH-like) or dual (LH- and FSH-like) activity toinduce follicle development in sheep. J Endocrinol 196, 593–600.Low SC, Nunes SL, Bitonti AJ, Dumont JA, 2005: Oral andpulmonary delivery of FSH-Fc fusion proteins via neonatalFc receptor-mediated transcytosis. Hum Reprod 20, 1805–1813.Lunenfeld B, 2004: Historical perspectives in gonadotrophintherapy. Hum Reprod Update 10, 453–467.Martinuk SD, Manning AW, Black WD, Murphy BD, 1991:Effects of carbohydrates on the pharmacokinetics andbiological activity of equine chorionic gonadotropin in vivo.Biol Reprod 45, 598–604.Matorras R, Rodriguez-Escudero FJ, 2003: Prions, urinarygonadotrophins and recombinant gonadotrophins. HumReprod 18, 1129–1130.Matzuk MM, Hsueh AJ, Lapolt P, Tsafriri A, Keene JL,Boime I, 1990: The biological role of the carboxyl-terminalextension of human chorionic gonadotropin beta-subunit.Endocrinology 126, 376–383.Monniaux D, Chupin D, Saumande J, 1983: Superovulatoryresponses of cattle. Theriogenology 19, 55–81.Murphy BD, Mapletoft RJ, Manns J, Humphrey WD, 1984:Variability in gonadotrophin preparations as a factor in thesuperovulatory response. Theriogenology 21, 117–125.Olijve W, de Boer W, Mulders JW, van Wezenbeek PM, 1996:Molecular biology and biochemistry of human recombinantfollicle stimulating hormone (Puregon). Mol Hum Reprod2, 371–382.Phillips DJ, Hudson NL, Lun S, Condell LA, McNatty KP,1993: Biopotency in vitro and metabolic clearance rates offive pituitary preparations of follicle stimulating hormone.Reprod Fertil Dev 5, 181–190.Pierce JG, Parsons TF, 1981: Glycoprotein hormones: structureand function. Annu Rev Biochem 50, 465–495.Reichl H, Balen A, Jansen CA, 2002: Prion transmission inblood and urine: what are the implications for recombinantand urinary-derived gonadotrophins? Hum Reprod 17,2501–2508.Roseman DS, Baenziger JU, 2000: Molecular basis of lutropinrecognition by the mannose ⁄ GalNAc-4-SO4 receptor. ProcNatl Acad Sci USA 97, 9949–9954.Smith PL, Kaetzel D, Nilson J, Baenziger JU, 1990: Thesialylated oligosaccharides of recombinant bovine lutropinmodulate hormone bioactivity. J Biol Chem 265, 874–881.Smith PL, Skelton TP, Fiete D, Dharmesh SM, BeranekMC, MacPhail L, Broze GJ Jr, Baenziger JU, 1992: Theasparagine-linked oligosaccharides on tissue factor pathwayinhibitor terminate with SO4-4GalNAc beta 1,4GlcNAc beta 1,2 Mana alpha. J Biol Chem 267,19140–19146.Sugahara T, Pixley MR, Minami S, Perlas E, Ben-MenahemD, Hsueh AJ, Boime I, 1995: Biosynthesis of a biologicallyactive single peptide chain containing the human commonalpha and chorionic gonadotropin beta subunits in tandem.Proc Natl Acad Sci USA 92, 2041–2045.Sugahara T, Sato A, Kudo M, Ben-Menahem D, Pixley MR,Hsueh AJ, Boime I, 1996: Expression of biologically activefusion genes encoding the common alpha subunit and thefollicle-stimulating hormone beta subunit. Role of a linkersequence. J Biol Chem 271, 10445–10448.Takagi M, Kim IH, Izadyar F, Hyttel P, Bevers MM,Dieleman SJ, Hendriksen PJ, Vos PL, 2001: Impaired finalfollicular maturation in heifers after superovulation withrecombinant human FSH. Reproduction 121, 941–951.Thotakura NR, Blithe DL, 1995: Glycoprotein hormones:glycobiology of gonadotrophins, thyrotrophin and freealpha subunit. Glycobiology 5, 3–10.Trousdale RK, Pollak SV, Klein J, Lobel L, Funahashi Y,Feirt N, Lustbader JW, 2007: Single-chain bifunctionalvascular endothelial growth factor (VEGF)-follicle-stimulatinghormone (FSH)-C-terminal peptide (CTP) is superiorto the combination therapy of recombinant VEGF plusFSH-CTP in stimulating angiogenesis during ovarian folliculogenesis.Endocrinology 148, 1296–1305.Weenen C, Pena JE, Pollak SV, Klein J, Lobel L, TrousdaleRK, Palmer S, Lustbader EG, Ogden RT, Lustbader JW,2004: Long-acting follicle-stimulating hormone analogscontaining N-linked glycosylation exhibited increased bioactivitycompared with o-linked analogs in female rats. JClin Endocrinol Metab 89, 5204–5212.Wehrman ME, Fike KE, Kojima FN, Bergfeld EG, Cupp AS,Mariscal V, Sanchez T, Kinder JE, 1996: Development ofpersistent ovarian follicles during synchronization of estrusinfluences the superovulatory response to FSH treatment incattle. Theriogenology 45, 593–610.Wilson JW, Jones AL, Moore K, Looney CR, Bondioli KR,1993: Superovulation of cattle with a recombinant-DNAbovine follicle stimulating hormone. Anim Reprod Sci 33,71–82.Yoon MJ, Boime I, Colgin M, Niswender KD, King SS,Alvarenga M, Jablonka-Shariff A, Pearl CA, Roser JF,2007: The efficacy of a single chain recombinant equineluteinizing hormone (reLH) in mares: induction of ovulation,hormone profiles, and inter-ovulatory intervals. DomestAnim Endocrinol 33, 470–479.Author’s address (for correspondence): TE Adams, Department ofAnimal Science, University of California, Davis, CA, USA. E-mail:teadams@ucdavis.eduConflict of interest: TE Adams declares no conflict of interest; B Irvingis a paid consultant for AspenBio Pharma Inc.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 193–199 (2008); doi: 10.1111/j.1439-0531.2008.01161.xISSN 0936-6768Recent Progress in Embryonic Stem Cell Research and Its Application in DomesticSpeciesTAL Brevini, S Antonini, G Pennarossa and F GandolfiDepartment of Animal Science, Centre for Embryonic Stem Cell Research, Laboratory of Biomedical Embryology, University of Milan, ItalyContentsMany reports described cell lines derived in domestic species,which presented several important features typical of embryonicstem cells (ESCs). Such features unfortunately did notinclude the capacity to generate germ-line chimeras, thereforelimiting the possibility to use these cells as tools for the geneticmanipulation. However, farm animal ESCs may still be usefulfor the generation of transgenic animals as usually have a selfrenewalcapacity more prolonged than normal primarycultures thus increasing the possibility to transform and selectcells to be used as nucleus donors in cloning procedures. Farmanimal ESCs may also be an excellent experimental model inpre-clinical trials, assessing the feasibility of cell therapybecause of the close morphological and physiological resemblanceto humans of species like the pig. However, thepersistent lack of standard methods for the derivation,maintenance and characterization of ESCs in domestic speciesstimulated the search for alternatives. Embryonic germ cellsmay represent such an alternative. Indeed, these cells showed ahigher plasticity than ESCs as contributed to embryonicdevelopment forming chimeric newborns but, as for ESCs,standardization is still far away and efficiency is very low.Recent results indicated spermatogonial stem cells as possibletools for germ-line genetic modifications with some proof ofprinciple results already achieved. But, a real break throughcould arrive from the multipotent germ-line stem cells,virtually equivalent to ESC, derived from newborn and adultmouse testis.IntroductionThe important achievements already reached by embryonicstem cells (ESCs) research have been recentlycelebrated with the Nobel price awarded to the scientistswho derived these cells for the first time and used themto identify the function of genes (Evans and Kaufman1981; Doetschman et al. 1987; Thomas and Capecchi1987). However, this aspect of stem cell research hashardly reached the general public that, on the contrary,is eagerly awaiting for ESCs to fulfil the next bigachievement: to provide an unlimited source of replacementmaterial for the therapy of injured, degenerated orotherwise damaged tissues. It is difficult to tell, atpresent, if this ambitious goal will ever be reached byESCs but the simple hope has fuelled a great deal ofresearch efforts as well as of hot debate.Such efforts have involved domestic species from theearly days with the first reports on putative ESCs insheep, pig and cow being published in 1990 (Evans et al.1990; Piedrahita et al. 1990a,b).As for the other species, also in domestic animals, thefocus of the research has shifted during the years.Initially, the main aim of stem cell research was toimprove transgenesis efficiency and control of transgeneexpression (Seamark 1994). The vision was to obtain anabundance of totipotent ESC to be transformed accordingto plan and to take advantage of the variety ofprocedures available to ensure that those cells could beidentified, cloned and maintained as isogenetic cell lines.The publication, of evidence that pluripotent pig ESClines were able to form chimeras, a few years from thefirst reports in mice, (Wheeler 1994; Gerfen and Wheeler1995; Chen et al. 1999) sustained for a while this visionbut, to date, it has not been possible to demonstrate theintegration of embryo-derived cell lines into the germline either in pig or in any other domestic species (Keeferet al. 2007). Indeed, it has not been possible even toobtain cell lines able to consistently and significantlycontribute to embryo development upon injection intoblastocyst in farm animals including pig, with theexception of one report in rabbit (Schoonjans et al.1996).In the absence of cell lines capable to generate germlinechimeras, the hope to use embryo-derived cell linesas a tool to improve transgenesis in domestic animalsshifted towards the use of these cells as nuclear donorsin nuclear transfer procedures. This pathway immediatelyproved to be much more effective with the birth ofthe first lamb obtained from an embryonic cell line(Campbell et al. 1996). However, the following birth ofDolly (Wilmut et al. 1997) suggested that in order tomanipulate farm animal germ line, ESCs were notnecessary. In fact, the formal demonstration of transgenicsheep generated through the transfer of transformedsomatic cells followed in a short interval of time(Schnieke et al. 1997) and was soon applied to otherspecies (Cibelli et al. 1998).This diverted the interest from farm animal ESCs fora while. However, recent data in the mouse indicatingthat, in this species, ESCs are more amenable toreprogramming through nuclear transfer than theirsomatic counterparts (Hochedlinger and Jaenisch2006), may induce a renew interest in deriving stableESCs as nuclear donors. Even if this phenomenon maynot be true for other species or for all mouse strains(Oback and Wells 2007), the ability of putative ESCs toproliferate in vitro without undergoing through senescence,may represent per se an important potentialadvantage in comparison to primary cultures of somaticcells. A prolonged in vitro life span, if fact, represents adistinct advantage when genetic modifications need tobe inserted into cell lines as these procedure are usuallyfollowed by extensive selective processes that, in turn,require several cell replications. Therefore, it is importantthat the transformed cell line remains viable andÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

194 TAL Brevini, S Antonini, G Pennarossa and F Gandolfikaryotypically stable for as long as possible in order toallow its use as a source of transgenic nuclei.The derivation of ESCs from human embryos(Thomson et al. 1998) is another major reason for therecent renewed interest in farm animal ESCs. Thisimportant result brought the full realization of thetherapeutic potential of stem cells and made it clear thatthese cells are perfect candidates for regenerative medicine,tissue repair and gene therapy. However, it wassoon clear also that, before routinely applying ESCs as atherapeutic tool several issues need to be addressed.Reliable ESCs differentiation protocols are needed,enabling the isolation of highly purified cell lineages tobe used in transplantation and in combination withways to circumvent the immunological rejection oftransplanted cells. Ultimately, the safety of using thesecells in human patients has to be ascertained. To achievethis, in vitro tests are being performed in both mouseand human cell lines, while in vivo experiments are,understandably, limited to the mouse. However, markedmorphological and physiological differences, a reducedgenetic variability and a short life-span make the mousemodel in many respect rather unsatisfactory whenexperimental outcomes need to be extrapolated tohuman clinical applications. To this purpose, outbredlarge animal models will be increasingly instrumental insafety and efficiency assessment of ESC-based therapeuticmethodologies.What Is an Embryonic Stem Cell?At this point we must clarify that a formally correctdefinition of embryonic stem cells implies the satisfactionof several criteria including: derivation withouttransformation or immortalization; stable diploidkaryotype, to be clonogenic, unlimited self-renewalcapacity, ability to generate all foetal and adult celltypes in vitro and in teratomas, incorporation intoembryonic development and contribution to all germlayers in chimera, germ-line colonization and transmission(Smith 2001). Any cell line which fails to satisfy allthese requirements should be defined as stem cell-like orother dubitative descriptions.This cautionary nomenclature has been followed formany years when dealing with domestic animals celllines of embryonic origin (Galli et al. 1994; Prelle et al.1999). However, the definition of ESC has changedupon the isolation of human embryonic cell lines(Thomson et al. 1998) and it has recently been proposedas cells that can extensively self-renew in vitro, maintaina normal karyotype, can differentiate into derivatives ofall three germ layers, express Oct-4 and a panel of otherpluripotency genes, show telomerase activity (Hoffmanand Carpenter 2005).If this or similar definitions are taken into consideration,it is clear that many of the embryonic cell linesderived from mammalian embryos, different frommouse, can qualify as ESCs. This is not only a matterof appropriate nomenclature, but also support theconcept that large animal ESCs may be a useful,meaningful experimental and pre-clinical model forhuman ESCs (hESCs) and for their therapeutic applications.Are ‘True’ ESCs Unique to the Mouse?The genetic backgroundEmbryonic stem cells are most commonly derived fromthe epiblast, a specific cell subpopulation of the inner cellmass of the blastocyst (Ralston and Rossant 2005). It iswell known that, at least in the mouse, ESCs can be reintroducedinto the Inner Cell Mass (ICM) re-enteringthe process of embryonic development (Bradley et al.1984). However, the two cell types, ESCs and epiblast,are not equivalent, because the epiblast exists onlytransiently in the embryo and does not act as a stem cellcompartment in vivo whereas ESCs form a stable cell linein vitro. Therefore, ESCs are the result of a selection andadaptation process to the culture environment and, assuch, could be considered more an artefact than aphysiological cell type (Chambers and Smith 2004). Ifthis is the case, it has been suggested that the significantdifferences observed between species in their capacity tooriginate pluripotent cell lines from early embryos maydepend mainly on their ICM ability to adapt to anarbitrary set of artificial conditions (Smith 2001). Indeed,the standard protocols for deriving mouse ESCs(mESCs) work efficiently only with 129 and, to a lowerextent, C57BL ⁄ 6 strains (Robertson 1987) whereas thesame procedures are not equally successful with blastocystof different genetic background (Kawase et al.1994), confirming that there is a strong genetic componentto ESC derivation (Smith 2001). Since geneticbackground has such a strong influence on the initialfrequency of establishing ESCs and on their subsequentstability in culture, the differences observed betweenmESC and those derived in large animal species,including primates, may be caused by the impact ofgenetic heterogeneity typical of these species and which isvirtually absent in inbred mouse strains.Pre-implantation developmentDetailed studies in mouse embryos determined that thefirst differentiation process occurs at the late morulastage when the outer cells adopt an epithelial structure.This event is followed by the first appearance of theblastocoel that marks the divergence of the first twolineages: trophectoderm and inner cell mass. Uponblastocyst expansion differentiation continues with theICM forming two further cell lineages: the epiblast andthe primitive endoderm or hypoblast. Between days 3.5and 4.5, the epiblast will give rise to the embryo itselfand the hypoblast will evolve in the extra embryonicendoderm and later will contribute to the yolk sac. Inother species, these events take place over a longerperiod of time. Although the three early embryoniclineages are present in all eutherian mammals, the timebetween their formation and fertilization is substantiallyshorter in mouse and human than in domestic ungulates.Detailed studies in mice have established thatESCs are derivatives of the epiblast (Brook and Gardner1997). In humans, blastocyst formation of the threeearly lineages takes approximately 6 days (Dvash andBenvenisty 2004), whereas, for instance, in pig andbovine embryos epiblast formation begins at hatchingand is complete towards day 12 (Vejlsted et al. 2005,Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Farm Animals Embryonic Stem Cells 1952006). In practical terms, this translates in no definedepiblast being present in these species before hatchingwhich in vivo occurs on late day 6 or on day 7 ofdevelopment in pig and days 8–9 in cow (Hunter 1974;Renard and Heyman 1979; Betteridge and Flechon1988). At this stage the ICM is present, defined as thecells comprised between the Rauber’s layer, separating itfrom the uterine lumen, and the visceral endoderm.Which is to say that, in order to obtain in vivo producedpig and cow blastocysts equivalent to their murinecounterparts, the uterus must be flushed on days 7–8and 8–9, respectively, when embryos are completelyhatched and expanded (Betteridge and Flechon 1988;Vejlsted et al. 2006). Upon hatching, pig blastocystsretain a round shape and during days 8 and 9, thehypoblast is formed from the ICM and graduallyproliferates as a confluent cell layer surrounding theblastocoel cavity. At the same time and through day 10in pig and day 12 in cow, the polar trophectoderm thatcovers the epiblast, referred to as the Rauber’s layer,begins to degenerate until is completely lost leaving theepiblast directly exposed to the uterine lumen (Maddox-Hyttel et al. 2003; Flechon et al. 2004). The area wherethe epiblast has surfaced is spherical, has a whitishcolour under the stereomicroscope and is referred to asthe embryonic disc whose internal surface is covered bycuboidal hypoblast cells. Between days 11 and 13,development continues with the embryonic disk and thewhole conceptus begins to assume an ovoid shape. Atthis stage, the first signs of polarity become visible alsoat the stereo microscope in the form of a crescentshapedthickening on the posterior third (Vejlsted et al.2006). This thickening will originate, within a couple ofdays, the primitive streak which accompanies theappearance of defined mesoderm and endoderm layers.This differs from the mouse where mesoderm andendoderm differentiation follows rather than precedesthe primitive streak formation (Tam and Behringer1997). Around days 13 and 14 while the primitive streakis still clearly visible, the major part of the epiblasttransforms into neural ectoderm and forms the neuralplate. This corresponds to a gradual down regulation oftypical pluripotency maker OCT4 which is substitutedby b-tubulin III expression, a marker of neural differentiation.Therefore, it can be assumed that embryos atthis stage are no longer suitable for ESC derivation. Insummary, the extended pre-implantation period togetherwith the formation of an embryonic disk makesungulate embryo epiblast available for a much longertime than in rodents and primates. As a consequence,the ICM available in the pre-hatching blastocyst maynot be exactly equivalent to the mouse epiblast usuallyrecovered for ESC derivation.However, it must be noted that, in pig and cow, a bigvariation in size is observed in embryos collected at thesame time and embryos of the same size showsubstantial differences in the development of theembryo proper. This means that, when post-hatchingembryos are used for ESC derivation, the simpleindication of the day of collection may comprise awide range of development. This wide variation mayexplain why a recent survey of the literature hasshowed no obvious effect of embryo age on theefficiency of ESC line derivation, at least in pig (Breviniet al. 2007b).Although derivation efficiency was not affected by theage of the embryo, the quality of the cell lines mighthave been different as suggested by some recent findingsin mouse. Two independent groups, in fact, showed thatmESC derived from early post-implantation embryoscorrespond more closely to hESC rather than mESC, inall aspects tested so far (Brons et al. 2007; Tesar et al.2007) including being leukemia inhibitory factor (LIF)and bone morphogenetic protein-4 (BMP4) independentbut requiring activin and FGF for efficient self-renewal.Interestingly, these cell lines, named epiblast stem cells(EpiSCs) were unable to form chimeras (Tesar et al.2007) or did so at a very low efficiency (Brons et al.2007) closely resembling, in this respect, to domesticanimal cell lines. Moreover, EpiSCs did not showdependency from a specific genetic background similarlyto human and farm animal cell lines.Therefore, differences in pre-attachment developmentand timing of isolation may be another factor added togenetic variation explaining species-specific differencesin ESC derivation efficiency and pluripotency.It has been speculated that cells able to give rise to‘true’ ESCs are overgrown by more rapidly dividingslightly later cells (Lovell-Badge 2007) and this wouldexplain the differences between those few mouse strainswere ESC derivation is possible and all other speciesincluding human.How Different Are ESCs from DifferentSpecies?Recent evidence, based on the comparison between thetrascriptomes of mESC and hESC, leads to the conclusionthat mESCs are substantially different from humancell lines. In fact, combining together the results ofdifferent studies, it appears that only 13–55% oftranscripts enriched in mESCs are also enriched inhuman lines but when comparison is performed amongdifferent hESC lines the overlap raise to 85–99% (seeEckfeldt et al. 2005; for detailed references). Thesubstantial differences determined between hESC andmESC expression profiles are followed by some morphologicaland functional differences between the celllines of the two species (Laslett et al. 2003; Ginis et al.2004; Rao and Zandstra 2005). For instance, hESCstypically grow in tightly adherent, flattened groupsrather than in rounded clumps and tolerate physicalseparation very poorly. ESCs of both typically grow wellin presence of mouse embryonic fibroblasts; hESCs donot require the presence of LIF to activate JAK-STAT3transcription factors in order to maintain their pluripotencyin culture. ESCs of both the species can bemaintained in culture without a feeder layer, but in thiscondition hESCs require the presence of fibroblastgrowth factor-2 and activin, whereas mESCs need LIFand bone morphogenetic protein-4 (BMP4). In contrast,BMP4 induces trophoblast differentiation in hESCs(Ludwig et al. 2006; Brons et al. 2007; Tesar et al.2007). Furthermore, although they express the same setof factors known to be required for pluripotency inmouse ESCs (such as OCT4, SOX2 and NANOG),Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

196 TAL Brevini, S Antonini, G Pennarossa and F Gandolfisome other markers are not shared (Brivanlou et al.2003).Data currently available on ungulate ESC characteristics,morphology and properties are definitively morelimited than those available for mESC or hESC, but it isalready possible to delineate some common tracts aswell as some differences.For instance, classic hESC and mESC markers suchas OCT4, SSEA1, SSEA4 and alkaline phosphatase areindeed expressed by ungulates ICM and embryo-derivedcell lines; however, the same genes are also expressed inthe trophectoderm and endoderm (van Eijk et al. 1999;Kirchhof et al. 2000; Talbot et al. 2002; He et al. 2006).NANOG, another well-characterized ESC marker inhuman and mouse cell lines looks more reliable since ithas been shown to be expressed in pig ESC-like cell lines(Brevini et al. 2007a; b) and to be strongly downregulatedin caprine trophectoderm while being stronglyexpressed in the ICM (He et al. 2006).Spontaneous differentiation of embryo-derived celllines is a common problem when working with ungulatecells even in the presence of a STO feeder layer or ofother cell types (Brevini et al. 2007a; b; Keefer et al.2007). The addition of cytokines and growth factors,such as EGF, LIF, SCF and FGF, that inhibit spontaneousdifferentiation of hESC and mESC have beenreported to be ineffective in other species (Talbot et al.1993, 1995; Moore and Piedrahita 1997). These additionsto the culture medium are done without evidenceof the presence on ungulate ESCs of the relatedreceptors and are mainly based on previous experiencein the murine species. In preliminary studies, we havefound that pig cell lines do not express LIF receptorindicating that the addition of this cytokine to theculture medium is not essential for the maintenance ofpluripotency. However, in our experience, the presenceof LIF in the culture medium seems to inhibit thedifferentiation process because it prevented embryoidbodies formation (Brevini et al. 2007a). On the contrary,preliminary information suggest that LIF receptor andgp130 are both expressed in bovine ICM and preliminaryoutgrowth (Pant and Keefer 2006).Although the fundamental properties of ESC arecommon to the different species, overall, these observationssuggest that several details are species-specific.Alternatives to ESCsEmbryonic germ cellsGiven the current limitations affecting the derivationand maintenance of large animal ESCs, it is worthinvestigating the availability of alternative sources ofcells with similar properties.The most obvious candidates are embryonic germcells (EGCs), originally derived from the primordialgerm cells isolated from 8.5 dpc mouse embryos(Matsui et al. 1992), these cells are in most aspectsindistinguishable from ESCs including the capability toform chimeras and to give germ-line transmission(Smith 2001), although at a lower rate than ESCs.The capacity of EGCs to erase imprints is the maindifference so far described between EGCs and ESCs(Tada et al. 1997). The derivation of EGCs fromhuman embryos (Shamblott et al. 1998) has ignited aconsiderable interest in this type of cells as it happenedwith ESCs.Primordial germ cell-derived cell lines have beenderived from rabbit, pig, cow and sheep, although withdifferent plasticity depending on the species (reviewed byPrelle et al. 1999). The best results, so far, have beenobtained with pig EGCs which were able to contributeto liveborn chimeric offsprings both normal (Shim et al.1997) and after transgenic transformation (Piedrahitaet al. 1998; Mueller et al. 1999). However, these initialencouraging results have not been sustained by furtherprogress and, at present, the efficiency of chimerageneration remains low with a weak contribution ofthe EGC to the newborn piglets (Rui et al. 2004). Inaddition, as for ESC, no germ-line transmission has everbeen described.The use of EGCs as nuclear donors has been tested incattle with moderate success (Zakhartchenko et al.1999); however, current data available in the literaturedo not allow to evaluate whether domestic animalEGCs have the characteristics of indefinite self-renewalin vitro that would be useful for using them as avaluable source of karyoplasts as mentioned previously.Pig EGC lines are capable of differentiating into a widerange of cell types in culture, including endodermal,trophoblast-like, epithelial-like, fibroblast-like and neuron-likecells. Moreover, when cultured in suspension,these cells formed embryoid bodies (Shim and Anderson1998). Benign teratomas were obtained after transplantationof days 34 and 37, bovine genital ridges under thekidney capsule of athymic mice (Choi and Anderson1998).The characterization of farm animal EGCs remainsmore superficial compared with that of ESCs, but theavailable data suggest that EGCs show a better plasticityand a comparable ability of self-renewal, thereforeencouraging further research with these cells that mayrepresent a useful tool both for genetic manipulationand as a model for cell therapies.Spermatogonial stem cellsAnother alternative is represented by the spermatogonialstem cells (SSCs) found in the testis which have theunique potential for both self-renewal and production ofdifferentiated daughter cells which will ultimately formspermatozoa (for an excellent review, see Dobrinski2005).It has been demonstrated that it is possible to fullyrestore male fertility by the transplantation of spermatogonialcells from fertile donor mice to the testes ofinfertile recipient mice and this worked also betweenspecies when rat SSCs were transplanted into mice testis(reviewed by McLaren 1998). Unfortunately, this onlyworks within limited phylogenetic distance betweendonor and recipient species, since transplantation inthe mouse testis of germ cells from non-rodent donorsranging from rabbits, dogs, pigs, cattle, horses andincluding non-human primates and humans, resulted incolonization of the mouse testis but spermatogenesisbecame arrested at the stage of spermatogonial expansion(Dobrinski 2005). However, recent data suggestÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Farm Animals Embryonic Stem Cells 197that it may be possible to circumvent this limitation byco-transplanting SSCs together with the supportingSertoli cells (Shinohara et al. 2003).Although the ambitious goal to obtain mature spermof one species in the testis of a different one has not beenreached in domestic species, transplants of SSCsbetween different individuals of the same species hasbeen successful in pigs and goats (Honaramooz et al.2002, 2003a; b). Through this intriguing technique, ithas even been possible to transplant SSCs obtained froma transgenic buck and to transmit to the progeny thetransgene through the sperm ejaculated by a normalbuck who received the transgenic SSCs (Honaramoozet al. 2003b). These data support the vision that SSCscan represent the ideal way to propagate transgenesthrough the sperm, making the ideal substitute of ESCsfor this purpose.Multipotent germ stem cellsIndeed, this could be just the beginning of a moreextensive use of SSCs as suggested by a recent findingobtained during the prolonged culture of these cellsin vitro which was originally aimed at their transformation.It has been noticed that among mouse post-natalSSCs cultured in vitro for an extended period of time,ESC-like cells appeared upon treating the culture with amixture of cytokines and growth factors (Kanatsu-Shinohara et al. 2004). These cells have been namedmultipotent germ stem cells (mGSCs) as they arecompletely different from SSCs. In fact, whereas SSCscan give rise only to spermatozoa, mGSCs show all theproperties typical of mESCs including pluripotency andthe ability to generate germ-line chimeras (Shinoharaand Kanatsu-Shinohara 2007).Recently, the isolation of mGSCs has been describedalso from adult mouse testis (Guan et al. 2006) furtherenhancing the potential utility of these cells as a viablealternative to ESCs both as a way to genetically modifythe germ line and a source of undifferentiated cells fortherapeutic purposes.Extensive efforts have been spent in the attempt toobtain similar cell lines in other species and in human inparticular, but, at present, no positive results have beenannounced. Work on farm animals has lead to theidentification of germ cell-specific markers in sheep(Rodriguez-Sosa et al. 2006) and pig (Luo et al. 2006;Goel et al. 2007) which should be instrumental inreaching an highly purified cell population.ConclusionsThe long quest for farm animal ESCs has not lead yet tothe derivation of a bona fide ESC line, convincinglydisplaying all the properties of analogous cells in mouse.Many reports have described cell lines in domesticspecies, which presented several important featurestypical of ESCs. Such features may be sufficient in orderto use these cells as tools for the genetic manipulation astheir self-renewal may be more prolonged than that ofnormal primary cultures, thus enabling the possibility totransform and select the cells to be used as nucleusdonors in cell transfer experiments. An alternative use offarm animal ESCs is as an excellent experimental modelin pre-clinical trial assessing the feasibility of cell therapybecause of the closer morphological and physiologicalresemblance to humans of species like the pig whencompared with the mouse.However, the persistent lack of standard methods forthe derivation, maintenance and characterization ofESCs in domestic species stimulates the search foralternatives.Embryonic germ cells may represent such an alternative.Indeed, these cells showed a higher plasticity thanESCs as they were able to contribute to embryonicdevelopment forming chimeric newborns. However, asfor ESCs, standardization is still far away and efficiencyis very low.Recent results indicated SSCs as possible tools forgerm-line genetic modifications with some proof ofprinciple results already achieved. But, a real breakthough could arise from the isolation of mGSCs,virtually equivalent to ESCs, from newborn and adultmouse testis. Unfortunately, at present, we do not knowwhether such cells are, once again, only typical of themouse with the other species only having a pale copy ofthe original.AcknowledgementsThis work is supported by MUSRT PRIN 2005, 2006, FIRST 2005,2006, 2007.ReferencesBetteridge KJ, Flechon J-E, 1988: The anatomy and physiologyof pre-attachement bovine embryos. Theriogenology 29,155–187.Bradley A, Evans M, Kaufman MH, Robertson E, 1984:Formation of germ-line chimaeras from embryo-derivedteratocarcinoma cell lines. Nature 309, 255–256.Brevini T, Antonini S, Cillo F, Crestan M, Gandolfi F, 2007a:Porcine embryonic stem cells: facts, challenges and hopes.Theriogenology 68S, S206–S213.Brevini TAL, Tosetti V, Crestan M, Antonini S, Gandolfi F,2007b: Derivation and characterization of pluripotent celllines from pig embryos of different origins. Theriogenology67, 54–63.Brivanlou AH, Gage FH, Jaenisch R, Jessell T, Melton D,Rossant J, 2003: Stem cells. Setting standards for humanembryonic stem cells. Science 300, 913–916.Brons IG, Smithers LE, Trotter MW, Rugg-Gunn P, Sun B,Chuva de Sousa Lopes SM, Howlett SK, Clarkson A,Ahrlund-Richter L, Pedersen RA, Vallier L, 2007: Derivationof pluripotent epiblast stem cells from mammalianembryos. Nature 448, 191–195.Brook FA, Gardner RL, 1997: The origin and efficientderivation of embryonic stem cells in the mouse. Proc NatlAcad Sci U S A 94, 5709–5712.Campbell KH, McWhir J, Ritchie WA, Wilmut I, 1996: Sheepcloned by nuclear transfer from a cultured cell line. Nature380, 64–66.Chambers I, Smith A, 2004: Self-renewal of teratocarcinomaand embryonic stem cells. Oncogene 23, 7150–7160.Chen LR, Shiue YL, Bertolini L, Medrano JF, BonDurant RH,Anderson GB, 1999: Establishment of pluripotent cell linesfrom porcine preimplantation embryos. Theriogenology 52,195–212.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

198 TAL Brevini, S Antonini, G Pennarossa and F GandolfiChoi SJ, Anderson GB, 1998: Development of tumors frombovine primordial germ cells transplanted to athymic mice.Anim Reprod Sci 52, 17–25.Cibelli JB, Stice SL, Golueke PJ, Kane JJ, Jerry J, Blackwell C,Ponce de Leon FA, Robl JM, 1998: Cloned transgeniccalves produced from nonquiescent fetal fibroblasts. Science280, 1256–1258.Dobrinski I, 2005: Germ cell transplantation and testis tissuexenografting in domestic animals. Anim Reprod Sci 89, 137–145.Doetschman T, Gregg RG, Maeda N, Hooper ML, MeltonDW, Thompson S, Smithies O, 1987: Targeted correction ofa mutant HPRT gene in mouse embryonic stem cells. Nature330, 576–578.Dvash T, Benvenisty N, 2004: Human embryonic stem cells asa model for early human development. Best Pract Res ClinObstet Gynaecol 18, 929–940.Eckfeldt CE, Mendenhall EM, Verfaillie CM, 2005: Themolecular repertoire of the ‘almighty’ stem cell. Nat RevMol Cell Biol 6, 726–737.van Eijk MJ, van Rooijen MA, Modina S, Scesi L, Folkers G,van Tol HT, Bevers MM, Fisher SR, Lewin HA, RakacolliD, Galli C, de Vaureix C, Trounson AO, Mummery CL,Gandolfi F, 1999: Molecular cloning, genetic mapping, anddevelopmental expression of bovine POU5F1. Biol Reprod60, 1093–1103.Evans MJ, Kaufman MH, 1981: Establishment in culture ofpluripotential cells from mouse embryos. Nature 292, 154–156.Evans MJ, Notarianni E, Laurie S, Moor RM, 1990: Derivationand preliminary characterization of pluripotent celllines from porcine and bovine blastocysts. Theriogenology33, 125–128.Flechon JE, Degrouard J, Flechon B, 2004: Gastrulationevents in the prestreak pig embryo: ultrastructure and cellmarkers. Genesis 38, 13–25.Galli C, Lazzari G, Flechon JE, Moor RM, 1994: Embryonicstem cells in farm animals. Zygote 2, 385–389.Gerfen RW, Wheeler DA, 1995: Isolation of embryonic celllinesfrom porcine blastocysts. Anim Biotechnol 6, 1–14.Ginis I, Luo Y, Miura T, Thies S, Brandenberger R, Gerecht-Nir S, Amit M, Hoke A, Carpenter MK, Itskovitz-Eldor J,Rao MS, 2004: Differences between human and mouseembryonic stem cells. Dev Biol 269, 360–380.Goel S, Sugimoto M, Minami N, Yamada M, Kume S, ImaiH, 2007: Identification, isolation, and in vitro culture ofporcine gonocytes. Biol Reprod 77, 127–137.Guan K, Nayernia K, Maier LS, Wagner S, Dressel R, Lee JH,Nolte J, Wolf F, Li M, Engel W, Hasenfuss G, 2006:Pluripotency of spermatogonial stem cells from adult mousetestis. Nature 440, 1199–1203.He S, Pant D, Schiffmacher A, Bischoff S, Melican D, GavinW, Keefer C2006: Developmental expression of pluripotencydetermining factors in caprine embryos: novel pattern ofNANOG protein localization in the nucleolus. Mol ReprodDev 73, 1512–1522.Hochedlinger K, Jaenisch R, 2006: Nuclear reprogrammingand pluripotency. Nature 441, 1061–1067.Hoffman LM, Carpenter MK, 2005: Characterization andculture of human embryonic stem cells. Nat Biotechnol 23,699–708.Honaramooz A, Megee SO, Dobrinski I, 2002: Germ celltransplantation in pigs. Biol Reprod 66, 21–28.Honaramooz A, Behboodi E, Blash S, Megee SO, Dobrinski I,2003a: Germ cell transplantation in goats. Mol Reprod Dev64, 422–428.Honaramooz A, Behboodi E, Megee SO, Overton SA,Galantino-Homer H, Echelard Y, Dobrinski I, 2003b:Fertility and germline transmission of donor haplotypefollowing germ cell transplantation in immunocompetentgoats. Biol Reprod 69, 1260–1264.Hunter RH, 1974: Chronological and cytological details offertilization and early embryonic development in thedomestic pig, Sus scrofa. Anat Rec 178, 169–185.Kanatsu-Shinohara M, Inoue K, Lee J, Yoshimoto M,Ogonuki N, Miki H, Baba S, Kato T, Kazuki Y, ToyokuniS, Toyoshima M, Niwa O, Oshimura M, Heike T, NakahataT, Ishino F, Ogura A, Shinohara T, 2004: Generation ofpluripotent stem cells from neonatal mouse testis. Cell 119,1001–1012.Kawase E, Suemori H, Takahashi N, Okazaki K, HashimotoK, Nakatsuji N, 1994: Strain difference in establishment ofmouse embryonic stem (ES) cell lines. Int J Dev Biol 38,385–390.Keefer CL, Pant D, Blomberg L, Talbot NC, 2007: Challengesand prospects for the establishment of embryonic stem celllines of domesticated ungulates. Anim Reprod Sci 98, 147–168.Kirchhof N, Carnwath JW, Lemme E, Anastassiadis K,Scholer H, Niemann H, 2000: Expression pattern of Oct-4in preimplantation embryos of different species. Biol Reprod63, 1698–1705.Laslett AL, Filipczyk AA, Pera MF, 2003: Characterizationand culture of human embryonic stem cells. TrendsCardiovasc Med 13, 295–301.Lovell-Badge R, 2007: Many ways to pluripotency. NatBiotechnol 25, 1114.Ludwig TE, Levenstein ME, Jones JM, Berggren WT, MitchenER, Frane JL, Crandall LJ, Daigh CA, Conard KR,Piekarczyk MS, Llanas RA, Thomson JA, 2006: Derivationof human embryonic stem cells in defined conditions. NatBiotechnol 24, 185–187.Luo J, Megee S, Rathi R, Dobrinski I, 2006: Protein geneproduct 9.5 is a spermatogonia-specific marker in the pigtestis: application to enrichment and culture of porcinespermatogonia. Mol Reprod Dev 73, 1531–1540.Maddox-Hyttel P, Alexopoulos NI, Vajta G, Lewis I, RogersP, Cann L, Callesen H, Tveden-Nyborg P, Trounson A,2003: Immunohistochemical and ultrastructural characterizationof the initial post-hatching development of bovineembryos. Reproduction 125, 607–623.Matsui Y, Zsebo K, Hogan BL, 1992: Derivation of pluripotentialembryonic stem cells from murine primordial germcells in culture. Cell 70, 841–847.McLaren A, 1998: Germ cells and germ cell transplantation.Int J Dev Biol 42, 855–860.Moore K, Piedrahita JA, 1997: The effects of human leukemiainhibitory factor (hLIF) and culture medium on in vitrodifferentiation of cultured porcine inner cell mass (pICM).In Vitro Cell Dev Biol Anim 33, 62–71.Mueller S, Prelle K, Rieger N, Petznek H, Lassnig C, LukschU, Aigner B, Baetscher M, Wolf E, Mueller M, Brem G,1999: Chimeric pigs following blastocyst injection of transgenicporcine primordial germ cells. Mol Reprod Dev 54,244–254.Oback B, Wells DN, 2007: Donor cell differentiation, reprogramming,and cloning efficiency: elusive or illusive correlation?Mol Reprod Dev 74, 646–654.Pant D, Keefer CL, 2006: Gene expression in cultures ofinner cell masses isolated from in vitro produced and invivo derived bovine blastocysts. Reprod Fertil Dev 18,110.Piedrahita JA, Anderson GB, BonDurant RH, 1990a: Influenceof feeder layer type on the efficiency of isolation ofporcine embryo-derived cell lines. Theriogenology 34, 865–877.Piedrahita JA, Anderson GB, Bondurant RH, 1990b: On theisolation of embryonic stem cells: Comparative behavior ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Farm Animals Embryonic Stem Cells 199murine, porcine and ovine embryos. Theriogenology 34,879–901.Piedrahita JA, Moore K, Oetama B, Lee CK, Scales N,Ramsoondar J, Bazer FW, Ott T, 1998: Generation oftransgenic porcine chimeras using primordial germ cellderivedcolonies. Biol Reprod 58, 1321–1329.Prelle K, Vassiliev IM, Vassilieva SG, Wolf E, Wobus AM,1999: Establishment of pluripotent cell lines from vertebratespecies – present status and future prospects. Cells TissuesOrgans 165, 220–236.Ralston A, Rossant J, 2005: Genetic regulation of stem cellorigins in the mouse embryo. Clin Genet 68, 106–112.Rao BM, Zandstra PW, 2005: Culture development for humanembryonic stem cell propagation: molecular aspects andchallenges. Curr Opin Biotechnol 16, 568.Renard JP, Heyman Y, 1979: Variable development ofsuperovulated bovine embryos between day 6 and day 12.Ann Biol Anim Biochim Biophys 19, 1589–1598.Robertson EJ, 1987: Teratocarcinomas and Embryonic StemCells, a Practical Approach. IRL Press, Oxford, WashingtonDC.Rodriguez-Sosa JR, Dobson H, Hahnel A, 2006: Isolation andtransplantation of spermatogonia in sheep. Theriogenology66, 2091–2103.Rui R, Shim H, Moyer AL, Anderson DL, Penedo CT, RoweJD, BonDurant RH, Anderson GB, 2004: Attempts toenhance production of porcine chimeras from embryonicgerm cells and preimplantation embryos. Theriogenology61, 1225–1235.Schnieke AE, Kind AJ, Ritchie WA, Mycock K, Scott AR,Ritchie M, Wilmut I, Colman A, Campbell KH, 1997:Human factor IX transgenic sheep produced by transfer ofnuclei from transfected fetal fibroblasts. Science 278, 2130–2133.Schoonjans L, Albright GM, Li JL, Collen D, MoreadithRW, 1996: Pluripotential rabbit embryonic stem (ES) cellsare capable of forming overt coat color chimeras followinginjection into blastocysts. Mol Reprod Dev 45, 439–443.Seamark RF, 1994: Progress and emerging problems inlivestock transgenesis: a summary perspective. Reprod FertilDev 6, 653–657.Shamblott MJ, Axelman J, Wang S, Bugg EM, Littlefield JW,Donovan PJ, Blumenthal PD, Huggins GR, Gearhart JD,1998: Derivation of pluripotent stem cells from culturedhuman primordial germ cells. Proc Natl Acad Sci U S A 95,13726–13731.Shim H, Anderson GB, 1998: In vitro survival and proliferationof porcine primordial germ cells. Theriogenology 49,521–528.Shim H, Gutierrez-Adan A, Chen LR, BonDurant RH,Behboodi E, Anderson GB, 1997: Isolation of pluripotentstem cells from cultured porcine primordial germ cells. BiolReprod 57, 1089–1095.Shinohara T, Kanatsu-Shinohara M2007: Culture and geneticmodification of mouse germline stem cells. Ann NY AcadSci, Annals. 1120, 58–71.Shinohara T, Orwig KE, Avarbock MR, Brinster RL, 2003:Restoration of spermatogenesis in infertile mice by Sertolicell transplantation. Biol Reprod 68, 1064–1071.Smith AG, 2001: Embryo-derived stem cells: of mice and men.Annu Rev Cell Dev Biol 17, 435–462.Tada M, Tada T, Lefebvre L, Barton SC, Surani MA, 1997:Embryonic germ cells induce epigenetic reprogramming ofsomatic nucleus in hybrid cells. EMBO J 16, 6510–6520.Talbot NC, Rexroad CE Jr, Pursel VG, Powell AM, Nel ND,1993: Culturing the epiblast cells of the pig blastocyst. InVitro Cell Dev Biol Anim 29A, 543–554.Talbot NC, Powell AM, Rexroad CE Jr, 1995: In vitropluripotency of epiblasts derived from bovine blastocysts.Mol Reprod Dev 42, 35–52.Talbot NC, Caperna TJ, Wells KD, 2002: The PICM-19 cellline as an in vitro model of liver bile ductules: effects ofcAMP inducers, biopeptides and pH. Cells Tissues Organs171, 99–116.Tam PP, Behringer RR, 1997: Mouse gastrulation: theformation of a mammalian body plan. Mech Dev 68, 3–25.Tesar PJ, Chenoweth JG, Brook FA, Davies TJ, Evans EP,Mack DL, Gardner RL, McKay RD, 2007: New cell linesfrom mouse epiblast share defining features with humanembryonic stem cells. Nature 448, 196–199.Thomas KR, Capecchi MR, 1987: Site-directed mutagenesisby gene targeting in mouse embryo-derived stem cells. Cell51, 503–512.Thomson JA, Itskovitz-Eldor J, Shapiro SS, Waknitz MA,Swiergiel JJ, Marshall VS, Jones JM, 1998: Embryonic stemcell lines derived from human blastocysts. Science 282,1145–1147.Vejlsted M, Avery B, Schmidt M, Greve T, Alexopoulos N,Maddox-Hyttel P, 2005: Ultrastructural and immunohistochemicalcharacterization of the bovine epiblast. BiolReprod 72, 678–686.Vejlsted M, Du Y, Vajta G, Maddox-Hyttel P, 2006: Posthatchingdevelopment of the porcine and bovine embryodefiningcriteria for expected development in vivo and invitro. Theriogenology 65, 153–165.Wheeler MB, 1994: Development and validation of swineembryonic stem cells: a review. Reprod Fertil Dev 6, 563–568.Wilmut I, Schnieke AE, McWhir J, Kind AJ, Campbell KH,1997: Viable offspring derived from fetal and adult mammaliancells. Nature 385, 810–813.Zakhartchenko V, Durcova-Hills G, Schernthaner W, StojkovicM, Reichenbach HD, Mueller S, Steinborn R, MuellerM, Wenigerkind H, Prelle K, Wolf E, Brem G, 1999:Potential of fetal germ cells for nuclear transfer in cattle.Mol Reprod Dev 52, 421–426.Author’s address (for correspondence): Fulvio Gandolfi, Departmentof Animal Science, Centre for Embryonic Stem Cell Research,Laboratory of Biomedical Embryology, University of Milan, ViaCecoria 10, 20133 Milano, Italy. E-mail: fulvio.gandolfi@unimi.itConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 200–206 (2008); doi: 10.1111/j.1439-0531.2008.01162.xISSN 0936-6768Reproduction in Domestic BuffaloBMAO PereraFaculty of Veterinary Medicine and Animal Science, University of Peradeniya, Peradeniya, Sri LankaContentsThe domestic buffalo is an indispensable livestock resource tomillions of smallholder farmers in developing countries,particularly in Asia. Although its reproductive biology isbasically similar to that of cattle, there are importantdifferences and unique characteristics that need to be consideredin order to apply modern reproductive technologies toimprove its productivity. Under most smallholder productionsystems, the reproductive efficiency of buffalo is compromisedby factors related to climate, management, nutrition anddiseases. However, when managed and fed properly, buffalocan have good fertility and provide milk, calves and draughtpower over a long productive life. The basic technicalproblems associated with artificial insemination in buffalowere largely overcome two decades ago, but the technologyhas not had the expected impact in some developing countries,because largely of infrastructural and logistic problems.Approaches involving the use of hormones for treatinganoestrus and for synchronizing oestrus have had varyingrates of success, depending on the protocols used and theincidence of underlying problems that cause infertility.Embryo technologies such as multiple ovulation embryotransfer, in vitro embryo production, cryopreservation andcloning are being intensively studied but have had far lowersuccess rates than in cattle. Improving the productivity ofbuffalo requires an understanding of their potential andlimitations under each farming system, development of simpleintervention strategies to ameliorate deficiencies in management,nutrition and healthcare, followed by judicious applicationof reproductive technologies that are sustainable withthe resources available to buffalo farmers.IntroductionThe world population of domestic water buffalo (Bubalusbubalis) is estimated to be approximately 172million, of which 166 million (96%) are in Asia, and theremainder mainly in the Mediterranean and LatinAmerican regions. Buffalo have been classified in totwo main ‘types’: the river type located in South Asiaand the swamp type spread across the South-East Asianregion. Systems of buffalo production vary widelythrough the different regions of the world (Perera et al.2005) and are determined by a matrix of severalinteracting factors that include climate (tropical ortemperate, humid or arid), location (rural, peri-urbanor urban), cropping systems (rain-fed or irrigated,annual or perennial crops), type of operation (small orlarge farm, subsistence or commercial) and the primarypurpose (milk, meat, draught, capital or mixed). Thebuffalo has been traditionally regarded as a poorbreeder with low reproductive efficiency, characterizedby late attainment of puberty and maturity, seasonalityof calving, long postpartum anoestrus, poor expressionof oestrous signs, low conception rates and long calvingintervals (Perera 1999; Barile 2005). This can be attributedto factors such as harsh environments, lack of yearroundfeed supply and minimal managerial inputs, in themajority of farming systems under which buffalo areraised. Studies in Pakistan (Usmani et al. 1990), SriLanka (Perera et al. 1987) and Brazil (Vale 1997) haveshown that they can have good fertility if they aremanaged and fed properly so as to overcome suchstresses.The reproductive physiology and endocrinology ofdomestic buffalo and their comparative aspects withcattle were comprehensively reviewed two decades agoby Dobson and Kamonpatana (1986) and a decadelater, subsequent advances in knowledge on buffaloreproduction were reviewed by Madan et al. (1996) andPerera (1999). A recent publication of the Food andAgriculture Organization edited by Borghese (2005)contains a series of comprehensive review papers on thebuffalo, including the production systems in the worldand their reproductive and productive characteristics.This paper will therefore limit its objectives to: (a)highlighting some of the reproductive characteristics ofbuffalo that have a bearing on their fertility andproductivity, (b) reviewing recent advances in modernreproductive technologies and (c) discussing the potentialapplications and limitations of these technologies forimproving buffalo production.Genetics and BreedingThe phylogeny of water buffalo is still a matter ofdebate. Based on studies of mitochondrial DNA(mtDNA) of swamp and river buffalo, together withanalysis of data published from South-East Asian andAustralian water buffalo, Kierstein et al. (2004) concludedthat both types of buffalo descend from onedomestication event, probably in the Indian subcontinent.They also found evidence for introgression of wildAsian buffalo (Bubalus arnee) mtDNA into domesticswamp buffalo. However, studies by Kumar et al. (2007)showed that river and swamp buffalo are distinguishedinto two distinct clades, indicating that the two typeswere domesticated independently. This was supportedby studies in China (Lei et al. 2007) that showed twomtDNA lineages with divergence estimated at18 000 years ago, indicating independent domesticationevents for the swamp buffalo from China and the riverbuffalo from the Indian subcontinent.From the practical aspects of buffalo breeding, thedisparity in the number of chromosomes in swamp(2n = 48) and river (2n = 50) buffalo has relevance(Huang et al. 2003). The F1 hybrids have 49 chromosomes,while the F2 hybrids have 48, 49 or 50 chromosomes.The backcrosses have two different karyotypeÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproduction in Domestic Buffalo 201categories each, with 2n = 48 and 2n = 49 in the threequarters swamp types and 2n = 49 and 2n = 50 in thethree quarters river types (Harisah et al. 1989). Thedistribution of chromosome categories among the F2hybrids and backcrosses suggests that only geneticallybalanced gametes of the F1 hybrids are capable ofproducing viable F2 and backcross generations, andthat crossbreds with 2n = 49 had lower fertility thancrossbreds with 2n = 50 (Huang et al. 2003).Reproduction in the FemaleBuffaloes are polyoestrus and are capable of breedingthroughout the year. Yet, in many countries, a seasonalpattern of calving has been observed. In some locations,where annual changes in rainfall determine the availabilityand quality of feed, ovarian activity commencessome 2–3 months after the onset of rains, followed byconceptions that result in a peak calving season10 months later (Perera et al. 1987). In Italy, however,where buffalo are fed with a constant balanced diet, adistinct seasonal reproductive pattern is also found, andthe inference from a series of studies is that seasonality isinfluenced by photoperiod, mediated by melatoninsecretion (Zicarelli 1997). In India, where summer heatstress can be severe, high prolactin secretion has beenidentified as a factor contributing to acyclicity and poorfertility by lowering progesterone secretion during thesummer months (Roy and Prakash 2007). The mainreproductive characteristics of buffalo, based on reportsfrom various countries on river and swamp buffalounder a wide range of management systems, aresummarized in Table 1.Puberty, ovarian characteristics and oestrous cyclesBuffalo heifers usually attain puberty when they reachapproximately 55–60% of adult body weight, but theage at which this occurs can be highly variable (Table 1),being influenced by genotype, nutrition, management,social environment, climate and year or season of birth.Although buffalo attain puberty later than cattle, theyhave a longer reproductive life, which compensates forthe early disadvantage. The ovaries of cyclic buffaloheifers have a reservoir of only 10 000 to 20 000primordial follicles (Danell 1987) compared withTable 1. Summary of reproductive characteristics of buffalo (adaptedfrom Perera et al. 2005)Parameter Mean RangeAge at puberty (months) 30 16–46Weight at puberty (kg) 275 200–350Length of oestrous cycle (days) 21 17–26Length of oestrus (h) 10 5–27Time of ovulationAfter onset of oestrus (h) 34 24–48After end of oestrus (h) 14 6–21Length of gestation (days)River type 310 300–320Swamp type 330 320–340Birth weight of calves (kg) 26 22–36Involution of uterus (days) 30 25–35approximately 150 000 in cattle. This could be one ofthe reasons for the lower yield of embryos when multipleovulation embryo transfer (MOET) is attempted inbuffalo (see section ‘Multiple ovulation embryo transfer’).Various treatments have been attempted for earlierinduction of puberty in buffalo heifers and one regimethat has been successful is the use of ProgesteroneReleasing Intravaginal Devices (PRID) left in place for12 days followed by treatment with Equine ChorionicGonadotrophin (eCG) at the time of withdrawal (Barileet al. 2001). This resulted in conception rates of 50–60%at 60 days after treatment, compared with 22% incontrols.Ovarian follicular growth during the oestrous cycle inbuffalo is similar to that observed in cattle and ischaracterized by waves of follicular recruitment, growthand regression. Studies in Brazil found that 63% ofanimals had two follicular waves during a cycle while33% had three follicular waves (Baruselli et al. 1997),and that the latter group had a longer luteal phase,resulting in a longer oestrous cycle, than the former(mean 24.0 vs 21.8 days). However, a recent study onsuckled Indian buffalo showed that five out of eightanimals showed a single wave of follicular growth, whilethe others had two waves (Awasthi et al. 2006). Thetemporal changes of progesterone in blood and milkduring the oestrous cycle are similar to those in cattle,but the concentration is relatively lower (Dobson andKamonpatana 1986; Perera et al. 1987).A major difference between buffalo and cattle is that,external signs of oestrus are less obvious in the former,with homosexual behaviour between females being rare(Perera 1987). The main behavioural signs are restlessness,bellowing and frequent voiding of small quantitiesof urine. Externally detectable physical changes includeswelling of the vulva, resulting in effacement of thehorizontal wrinkles that are present on its externalsurface and this, together with vestibular reddening, canbe detected by regular examination of individualanimals under confined systems. Mucus, secreted fromthe cervix during oestrus, is less copious than in cattleand does not usually hang as strands from the vulva buttends to accumulate on the floor of the vagina and bedischarged either when the animal is lying down or withthe urine. Thus, in females that are housed or tethered, agood practice is to examine the floor near the rear of theanimal each morning and evening for signs of mucus.These factors have contributed to the observation thatsilent ovulation (also termed as silent oestrus) is morecommon in buffalo than in cattle. Awasthi et al. (2007)found that buffalo with silent ovulation followingtreatment with prostaglandin had slower growth rateand smaller diameter of the ovulatory follicle, and alonger interval from treatment to ovulation, than thosewith overt oestrous signs. Studies are warranted todetermine whether such differences also occur in spontaneouslyovulating animals.The duration of oestrus and the intervals from theonset and end of oestrus to ovulation are shown inTable 1. In hot climates, the duration of oestrus tends tobe short and the signs may be exhibited only during thenight or early morning. In contrast, studies on Italianbuffalo show that they have longer duration of oestrousÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

202 BMAO Pererabehaviour and a longer interval from the beginning ofoestrus to ovulation, with the interval from peak LHconcentration to ovulation being 25 ± 13 h in animalsthat conceived to artificial insemination (AI) and46 ± 18 h in those that did not (Moioli et al. 1998;Borghese 2005). The results from most studies indicatethat the appropriate time for breeding, whether bynatural service or AI, is during the latter part of theoestrous period. Thus, the AM ⁄ PM rule, which isapplied in cattle, where females detected in oestrus inthe morning are bred in the afternoon of the same dayand those detected in the afternoon or evening are bredthe next morning, also applies in buffalo.Pregnancy, parturition and the postpartum periodThe duration of gestation in the buffalo ranges from 300to 330 days with a mean of approximately 310 days forriver types and 320 days for swamp types (Perera and deSilva 1985). Early diagnosis of pregnancy can be carriedout by ultrasound scanning from approximately 20 daysonwards (Presicce et al. 2001) and by rectal palpationfrom approximately 45 days onwards. The techniquesare basically similar to those routinely used in cattle, butthe longer gestation period has to be taken account of inassessing the stage of pregnancy. The main laboratorymethod of diagnosis is based on measurement ofprogesterone 20–23 days after breeding (Perera et al.1980). The process of parturition and the changesoccurring in hormones such as progesterone, oestroneand the main metabolite of prostaglandin F 2a aroundthe period of parturition in buffalo are similar to thosein cattle (Perera et al. 1981). Involution of the uterus isusually completed in 25–35 days after calving (Lohanet al. 2004), and the stimulus of suckling shortens theinvolution time (Usmani et al. 1990).The period of postpartum anoestrus or acyclicity ishighly variable in the buffalo and is usually longer thanin cattle under comparative management conditions.Under good conditions, buffalo can resume cyclicity by30–90 days, but factors such as poor nutrition and bodycondition (Baruselli et al. 2001), suckling management(Usmani et al. 1990) and climate (which also influencesnutrition through feed quality and availability) candelay this considerably. For example, indigenous buffaloin Sri Lanka raised under free-grazing conditions withabundant natural feed and suckling restricted to onceper day resumed cyclicity by 30–60 days after calving,but those under harsh conditions with free suckling bythe calves remained acyclic for 150–200 days (Pereraet al. 1987). A review of literature from Egypt, India andPakistan (El-Wishy 2007) showed that only 34–49% ofbuffalo showed oestrus during the first 90 days aftercalving and 31–42% remained anoestrus for more than150 days. Both postpartum ovulation and oestrusoccurred later in swamp than in river buffalo. The firstpostpartum ovulation was frequently followed by one ormore short oestrous cycles (

Reproduction in Domestic Buffalo 203Reproductive TechnologiesArtificial inseminationThe procedures for processing and using buffalo semenfor AI are based mainly on techniques developed forcattle with some modifications (Vale 1997; IAEA 2005).The main differences are in the semen diluents, with Trisbuffers, egg yolk, skim milk and coconut water beingcommonly used as ingredients for preserving semen inboth chilled (+4°C) and deep-frozen forms. The cryoprotectiveagent used for freezing buffalo semen isglycerol, at a final concentration of 6.5–7.0%. Variousadditives have been investigated for possible beneficialeffects on buffalo spermatozoa during freezing andthawing, and the two proteins that help to maintainhigher post-thaw motility are oviductal proteins (Kumaresanet al. 2005) and Bradykinin (Shukla and Misra2007). Buffalo spermatozoa, subjected to freezing andthawing, appear to have a shorter fertile lifespan insidethe female tract than those in fresh semen (Moioli et al.1998). Thus, proper detection of heat and timing of AIbecome critical when frozen semen is used and may beone reason for the lower conception rates, while anothercould be the narrow cervix of the buffalo, which makesAI more difficult.Apart from the technical aspects, there are manyconstraints which influence the success of AI in buffalo,especially in developing countries. These include factorsthat impede the effective and timely delivery of services,such as poor communication, lack of transport andinadequate remuneration for AI technicians, and limitationsof smallholder farming systems, such as poornutrition and reproductive management of cows.Separation of X and Y spermatozoaIdentification of X- and Y-bearing spermatozoa inbuffalo can be performed by fluorescence in situhybridization, using the cattle Y-chromosome-specificBC1.2 probe or the X- and Y-specific probes from theyak (Re´vay et al. 2003). Use of high-speed flowcytometric cell sorting, followed by deep AI carriedout near the utero-tubal junction (UTJ) with 2.5 millionlive frozen-thawed sperm, resulted in 43% conceptionrate with eight out of nine fetuses corresponding to thepredicted sex (Presicce et al. 2005). Semen depositionnear the UTJ of buffalo can be performed using theGhent device that was developed for cattle (Presicceet al. 2004). Sex-sorted sperm has also been usedsuccessfully for in vitro fertilization (IVF) and embryotransfer (Lu et al. 2007).Oestrous synchronizationThe procedures used in the past for oestrous synchronizationin buffalo were empirically based on thosedeveloped for cattle, aimed at either inducing prematureluteolysis using prostaglandins or prolonging the lutealphase using progestagens (Perera 1987). However, theefficacy of prostaglandins for synchronizing ovulation isnow known to be dependent upon progesterone concentrationand ovarian follicular status at the time oftreatment (Brito et al. 2002). Use of the two-dose regimeof prostaglandin overcomes some of the above limitations,but manipulation of follicular development isnecessary to achieve better synchrony and improvedfertility (De Rensis and Lo´pez-Gatius 2007). Therefore,most current protocols include GnRH or gonadotrophinsin combination with prostaglandins, progesteroneor oestradiol. The ‘Ovsynch’ protocol (GnRH, prostaglandin7 days later and second GnRH 2 days later) hasbeen successful in synchronizing ovulation in 70–90% ofbuffalo with conception rates ranging from 33% to 60%(Baruselli et al. 1999; Paul and Prakash 2005). Inaddition to the type of protocol selected, the followingfactors must also be addressed to achieve success inbuffalo: (a) selection of animals that are in goodnutritional condition and free from disease; (b) preventionof stress during the procedures for treatment andAI, especially under tropical conditions, where animalsmay be herded together or moved to other locations;and (c) where seasonal differences exist, schedulingtreatment for the more favourable periods when themajority of animals are cycling.Multiple ovulation embryo transferThe current status of MOET and other advancedreproductive technologies in buffalo has been recentlyreviewed (Drost 2007). Studies in many countries haveconfirmed that buffalo have a lower superovulatoryresponse than cattle, attributed mainly to the smallerpopulation of recruitable follicles in the ovary (Madanet al. 1996; Manik et al. 2002). A further limitationappears to be the relatively low rate of transfer ofoocytes to the oviduct and ⁄ or impaired transport of ovaand embryos in the reproductive tract (Baruselli et al.2000). In a large scale MOET operation in India (Misraet al. 1994), the total embryo yield per treatmentincreased from 1.77 to 3.83 over 5 years, and thenumber of viable embryos from 0.92 to 2.13; thetransfer of 469 embryos resulted in a pregnancy rateof 17% and a calving rate of 9.8%. In Italy, Campanileet al. (1995) transferred 76 buffalo embryos andobtained a pregnancy rate of 30%, with no significantdifference between fresh or frozen-thawed embryos andbetween morulae or blastocysts. Yet, the overall rate ofsuccess in terms of calves born per superovulation is stilltoo low for MOET to be widely applicable under fieldfarmingconditions. Two situations in which it mighthave a role are for establishing nucleus breeding stocksor for conservation of genetic resources.In vitro embryo production and cryopreservationThe low efficiency of MOET in buffalo has led to anincreased interest in in vitro embryo production (IVEP)technologies for achieving rapid genetic improvement.The sequence of procedures involving recovery ofoocytes from ovaries of slaughtered animals by directaspiration or from live animals by ultrasound-guidedtransvaginal ovum pick-up (OPU), followed by in vitromaturation, IVF and in vitro culture have been studiedin buffalo over the past decade (Boni et al. 1996; Madanet al. 1996; Hufana-Duran et al. 2004), but the successÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

204 BMAO Pererarates are lower than those achieved in cattle. Thecollection of immature oocytes from slaughtered buffaloby aspiration of antral follicles can yield one to threeoocytes per ovary that are suitable for IVEP. The OPUtechnique can be used for repeatedly collecting oocytesin cycling buffalo once or twice every week (Boni et al.1996; Manik et al. 2006) or in prepubertal heifers andcyclic cows after treatment with follicle-stimulatinghormone (FSH) (Techakumphu et al. 2004). The yieldof cumulus–oocyte complexes using OPU in non-superovulatedbuffalo is approximately 1–2 per ovary percollection, whereas after FSH treatment it is approximately2–3 per ovary (Boni et al. 1997; Promdireg et al.2005). The cleavage rate of oocytes selected as suitablefor IVEP is approximately 50%, but only approximately10% develop into morulae and blastocysts (Manik et al.2006). Maturation of buffalo oocytes in vitro occursearlier than in cattle (Neglia et al. 2003).For the cryopreservation of buffalo oocytes, the slowfreezing procedure has been less successful than themethod of vitrification, and solid surface vitrificationappears to be superior to in-straw vitrification, usingethylene glycol as the cryoprotectant (Boonkusol et al.2007). Recent studies in Philippines on the transfer ofvitrified embryos derived from oocytes collected fromslaughtered river type animals resulted in a pregnancyrate of 16% and a calving rate of 11% in river typerecipients (Hufana-Duran et al. 2004), and a calvingrate of 10% in swamp type recipients (Hufana-Duranet al. 2007). In China, transfer of fresh embryos derivedfrom repeated OPU to recipients after natural oestrusresulted in 35% calvings, while transfer of embryosderived from abattoir ovaries to synchronized recipientsresulted in 15% calvings (Liang et al. 2007).CloningResearch on cloning by somatic cell nuclear transfer inbuffalo is still in its early stages, but some basicunderstanding has been achieved regarding parthenogenetic,in vitro and in vivo development of embryosreconstructed by transferring donor nuclei from foetaland adult fibroblast cells to enucleated buffalo oocytes.Embryos reconstructed using foetal fibroblasts werecapable of developing to blastocyst stage (Meena andDas 2006) and some pregnancies were detected after thetransfer of cloned blastocysts into recipients, but nonewere carried to term (Saikhun et al. 2004). Clonedembryos that are capable of developing to the morulaand blastocyst stages have also been constructed usingenucleated rabbit oocytes as recipient cytoplasm andcow, swamp buffalo, pig and elephant fibroblasts asdonor nuclei (Numchaisrika et al. 2007). When cloningtechnology does become established in the buffalo,problems that have been encountered in other species,such as high incidence of developmental abnormalities,will need to be addressed before it can have widepractical applications.ConclusionsDomestic buffalo are generally regarded as having lowreproductive efficiency. This is largely because of theconditions under which the majority of them are raised,being smallholder farming systems with harsh environments,poor nutrition and minimal managerial inputs.However, they can have good fertility when managedand fed properly. Modern methods in molecular geneticsare helping to unravel the evolutionary and geneticstatus of the river and swamp types of buffalo, but froma practical viewpoint, the disparity in the number ofchromosomes in the two types needs to be considered incross-breeding programmes in order to avoid decline infertility. Buffalo, cows and bulls are capable of breedingthroughout the year but often show seasonal fluctuationsin fertility because of climatic and nutritionalfactors that modulate ovarian and testicular functions.The physiology and endocrinology of reproduction inbuffalo are basically similar to those in cattle, but someimportant differences exist that must be considered inattempts to improve reproductive efficiency through theuse of modern reproductive technologies. A majorfactor limiting wider uptake of AI by buffalo farmersis the difficulty in detecting oestrus. Although improvedprotocols of oestrous synchronization can overcome thisproblem, there are many other factors such as nutritionalstatus, seasonality and reproductive managementthat need to be addressed to achieve success. Embryotechnologies that include MOET and IVEP have beenvigorously studied over the past two decades, but thesuccess rates remain below that achieved in cattlebecause of many inherent biological features that areunique to the buffalo. Once the technological problemsare overcome, the successful practical application ofthese methods will need to be preceded by measure toovercome the managerial and nutritional causes ofinfertility that are common in the majority of currentbuffalo farming systems.ReferencesAwasthi MK, Khare A, Kavani FS, Siddiquee GM, PanchalMT, Shah RR, 2006: Is one-wave follicular growth duringthe estrous cycle a usual phenomenon in water buffaloes(Bubalus bubalis)? Anim Reprod Sci 92, 241–253.Awasthi MK, Kavani FS, Siddiquee GM, Sarvaiya NP,Derashri HJ, 2007: Is slow follicular growth the cause ofsilent estrus in water buffaloes? Anim Reprod Sci 99, 258–268.Bahga CS, Khokar BS, 1991: Effect of different seasons onconcentration of plasma luteinizing hormone and seminalquality vis-a` -vis freezability of buffalo bulls (Bubalus bubalis).Int J Biometeorol 35, 222–224.Barile VL, 2005: Review article: improving reproductiveefficiency in female buffaloes. Livest Prod Sci 92, 183–194.Barile VL, Galasso A, Marchiori E, Pacelli C, Montemurro N,Borghese A, 2001: Effect of PRID treatment on conceptionrate in Mediterranean buffalo heifers. Livest Prod Sci 68,283–287.Baruselli PS, Mucciolo R, Visintin GA, Viana VC, Arruda RP,Madureira EH, 1997: Ovarian follicular dynamics duringthe estrous cycle in buffalo (Bubalus bubalis). Theriogenology47, 1531–1547.Baruselli PS, Madureira EH, Visintin JA, Barnabe RC,Amaral R, 1999: Timed insemination using synchronizationof ovulation in buffalo. Rev Bras Reprod Anim 23, 360–362.Baruselli PS, Madureira EH, Visintin JA, Porto-Filho R,Carvalho NAT, Campanile G, Zicarelli L, 2000: Failure ofÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reproduction in Domestic Buffalo 205oocytes entry into oviduct in superovulated buffalo. Theriogenology53, 491.Baruselli PS, Barnabe VH, Barnabe RC, Visintin JA, Molero-Filho JR, Porto R, 2001: Effect of body condition score atcalving on postpartum reproductive performance in buffalo.Buffalo J 17, 53–65.Boni R, Roviello S, Zicarelli L, 1996: Repeated ovum-pick-upin Italian Mediterranean buffalo cows. Theriogenology 46,899–909.Boni R, Rovello S, Gasparrini B, Zicarelli L, 1997: Pregnanciesestablished after transferring embryos yielded by ovumpick-up and in vitro embryo production in Italian buffalocows. In: Proceedings of the Fifth World Buffalo Congress,Caserta, Italy, pp. 787–792.Boonkusol D, Faisaikarm T, Dinnyes A, Kitiyanant Y, 2007:Effects of vitrification procedures on subsequent developmentand ultrastructure of in vitro-matured swamp buffalo(Bubalus bubalis) oocytes. Reprod Fertil Dev 19, 383–391.Borghese A(ed.), 2005: Buffalo Production and Research.FAO Regional Office for Europe, Technical Series 67, Foodand Agriculture Organization, Rome.Brito LF, Satrapa R, Marson EP, Kastelic JP, 2002: Efficacyof PGF(2alpha) to synchronize estrus in water buffalo cows(Bubalus bubalis) is dependent upon plasma progesteroneconcentration, corpus luteum size and ovarian follicularstatus before treatment. Anim Reprod Sci 73, 23–35.Campanile G, Boni R, Esposito L, Spadetta M, Zicarelli L,1995: Embryo transfer in mediterranean buffalo cow bred inItaly. Bubalus bubalis 3, 63.Danell B, 1987: Oestrus Behaviour, Ovarian Morphology andCyclical Variation in Follicular System and EndocrinePattern in Water Buffalo Heifers. PhD Thesis, SwedishUniversity of Agricultural Sciences, Uppsala.De Rensis F, Lo´pez-Gatius F, 2007: Protocols for synchronizingestrus and ovulation in buffalo (Bubalus bubalis): areview. Theriogenology 67, 209–216.Dobson H, Kamonpatana M, 1986: A review of female cattlereproduction with special reference to a comparison betweenbuffaloes, cows and zebu. J Reprod Fertil 77, 1–36.Drost M, 2007: Advanced reproductive technology in thewater buffalo. Theriogenology 68, 450–453.El-Wishy AB, 2007: The postpartum buffalo. II. Acyclicity andanestrus. Anim Reprod Sci 97, 216–36.Harisah M, Azmi TI, Hilmi M, Vidyadaran MK, Bongso TA,Nava ZM, Momongan V, Basrur PK, 1989: Identification ofcrossbred buffalo genotypes and their chromosome segregationpatterns. Genome 32, 999–1002.Huang YJ, Shang JH, Liang MM, Zhang XF, Huang FX,2003: Studies of chromosomal heredity and fertility ofprogenies (2n = 49) crossed between river and swampbuffalo [in Chinese]. Yi Chuan 25, 155–159.Hufana-Duran D, Pedro PB, Venturina HV, Hufana RD,Salazar AL, Duran PG, Cruz LC, 2004: Post-warminghatching and birth of live calves following transfer of invitro-derived vitrified water buffalo (Bubalus bubalis) embryos.Theriogenology 61, 1429–1439.Hufana-Duran D, Pedro PB, Venturina HV, Duran PG, CruzLC, 2007: Full-term delivery of river buffalo calves(2n = 50) from in vitro-derived vitrified embryos by swampbuffalo recipients (2n = 48). Livest Sci 107, 213–219.IAEA, 2005: Improving Artificial Breeding of Cattle andBuffalo in Asia: Guidelines and Recommendations. IAEA-TECDOC 1480, International Atomic Energy Agency,Vienna. (http://www-pub.iaea.org/MTCD/publications/PDF/TE_1480_web.pdf)Kierstein G, Vallinoto M, Silva A, Schneider MP, Iannuzzi L,Brenig B, 2004: Analysis of mitochondrial D-loop regioncasts new light on domestic water buffalo (Bubalus bubalis)phylogeny. Mol Phylogenet Evol 30, 308–324.Kumar S, Nagarajan M, Sandhu JS, Kumar N, Behl V,Nishanth G, 2007: Mitochondrial DNA analyses of Indianwater buffalo support a distinct genetic origin of river andswamp buffalo. Anim Genet 38, 227–232.Kumaresan A, Ansari MR, Garg A, 2005: Modulation ofpost-thaw sperm functions with oviductal proteins inbuffaloes. Anim Reprod Sci 90, 73–84.Lei CZ, Zhang W, Chen H, Lu F, Liu RY, Yang XY, ZhangHC, Liu ZG, Yao LB, Lu ZF, Zhao ZL, 2007: Independentmaternal origin of Chinese swamp buffalo (Bubalus bubalis).Anim Genet 38, 97–102.Liang X, Zhang X, Yang B, Cheng M, Huang F, Pang C, QingG, Liao C, Wei S, Senatore EM, Bella A, Presicce GA, 2007:Pregnancy and calving rates following transfer of in-vitroproducedriver and F1 (river x swamp) buffalo (Bubalusbubalis) embryos in recipients on natural oestrus or synchronisedfor ovulation. Reprod Fertil Dev 19, 670–676.Lohan IS, Malik RK, Kaker ML, 2004: Uterine involutionand ovarian follicular growth during early postpartumperiod of Murrah buffaloes (Bubalus bubalis). Asian-australasJ Anim Sci 17, 313–316.Lu YQ, Liang XW, Zhang M, Wang WL, Kitiyanant Y, LuSS, Meng B, Lu KH, 2007: Birth of twins after in vitrofertilization with flow-cytometric sorted buffalo (Bubalusbubalis) sperm. Anim Reprod Sci 100, 192–196.Madan ML, Das SK, Palta P, 1996: Application of reproductivetechnology to buffaloes. Anim Reprod Sci 42,299–306.Malfatti A, Barbato O, Todini L, Terzano GM, DebenedettiA, Borghese A, 2006: Blood testosterone levels in ItalianMediterranean buffalo bulls managed in two differentbreeding conditions. Theriogenology 65, 1137–1144.Manik RS, Palta P, Singla SK, Sharma V, 2002: Folliculogenesisin buffalo (Bubalus bubalis): a review. Reprod FertilDev 14, 315–325.Manik RS, Chauhan MS, Gupta V, Singla SK, Palta P, 2006:In vitro fertilization of oocytes obtained through transvaginaloocyte retrieval from cyclic Murrah buffaloes (Bubalusbubalis). Reprod Fertil Dev 18, 219–219.Meena CR, Das SK, 2006: Development of water buffalo(Bubalus bubalis) embryos from in vitro matured oocytesreconstructed with fetal skin fibroblast cells as donor nuclei.Anim Reprod Sci 93, 258–267.Misra AK, Kasiraj R, Rao MM, Reddy NSR, Joshi BV1994:Embryo transfer in buffalo in India: Progress in the last fiveyears. In: Proceedings of the IVth World Buffalo Congress,Sao Paulo, Brazil, 3, 501–504.Moioli BM, Napolitano F, Puppo S, Barile VL, Terzano GM,Borghese A, Malfatti A, Catalano A, Pilla AM, 1998:Patterns of oestrus, time of LH release and ovulation andeffects of time of artificial insemination in Mediterraneanbuffalo cows. Anim Sci 66, 87–91.Nanda AS, Brar PS, Prabhakar S, 2003: Enhancing reproductiveperformance in dairy buffalo: major constraints andachievements. Reprod Suppl 61, 27–36.Neglia G, Gasparrini B, Di Palo R, De Rosa C, Zicarelli L,Campanile G, 2003: Comparison of pregnancy rates withtwo estrus synchronization protocols in Italian Mediterraneanbuffalo cows. Theriogenology 60, 125–133.Numchaisrika P, Thongphakdee A, Rungsiwiwut R, PruksananondaK, Virutamasen P, Techakumphu M, 2007: Thedevelopment of intra- and inter-species cloned embryosderived from rabbit oocytes: the effect of donor cell sources.Reprod Fertil Dev 19, 153–153.Paul V, Prakash BS, 2005: Efficacy of the ovsynch protocol forsynchronization of ovulation and fixed-time artificial inseminationin Murrah buffaloes (Bubalus bubalis). Theriogenology64, 1049–1060.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

206 BMAO PereraPerera BMAO, 1987: A review of experiences with oestroussynchronisation in buffaloes in Sri Lanka. Buffalo J Suppl 1,105–114.Perera BMAO, 1999: Reproduction in water buffalo: comparativeaspects and implications for management. J ReprodFertil Suppl 54, 157–168.Perera BMAO, de Silva LNA, 1985: Gestation length inindigenous (Lanka) and exotic (Murrah) buffaloes in SriLanka. Buffalo J 1, 83–87.Perera BMAO, Pathiraja N, Motha MXJ, Weerasekera DA,1979: Seasonal differences in plasma testosterone profiles inbuffalo bulls. Theriogenology 12, 33–38.Perera BMAO, Pathiraja N, Abeywardena SA, Motha MXJ,Abeygunawardena H, 1980: Early pregnancy diagnosis inbuffaloes from plasma progesterone concentration. Vet Rec106, 104–106.Perera BMAO, Abeygunawardena H, Thamotharam A, KindahlH, Edqvist LE, 1981: Peripartal changes of oestrone,progesterone and prostaglandin in the water buffalo. Theriogenology15, 463–467.Perera BMAO, de Silva LNA, Kuruwita VY, KarunaratneAM, 1987: Postpartum ovarian activity, uterine involutionand fertility in indigenous buffaloes at a selected villagelocation in Sri Lanka. Anim Reprod Sci 14, 115–127.Perera BMAO, Abeygunawardena H, Vale WG, ChantalakhanaC, 2005: Buffalo. In: Livestock and Wealth Creation –Improving the Husbandry of Animals Kept by Poor Peoplein Developing Countries. Owen E, Kitayi A, Jayasuriya N,Smith T (Eds). Livestock Production Programme, NaturalResources International Limited, Nottingham, UK, pp.451–471.Presicce GA, De Santis G, Stecco R, Senatore E, De MauroGJ, Terzano GM, 2001: Foetal gender determination byultrasound in the Mediterranean Italian buffalo (Bubalusbubalis). Theriogenology 55, 532.Presicce GA, Verberckmoes SB, Senatore EM, Pascale MD,Jeroen Dewulf JB, Van Soom AB, 2004: Assessment of anew utero-tubal junction insemination device in the MediterraneanItalian water buffalo (Bubalus bubalis) under fieldconditions. Bubalus Bubalis 1, 58–64.Presicce GA, Verberckmoes SB, Senatore EM, Klinc P, RathD, 2005: First established pregnancies in MediterraneanItalian buffaloes (Bubalus bubalis) following deposition ofsexed spermatozoa near the utero tubal junction. ReprodDomest Anim 40, 73–75.Promdireg A, Adulyanubap W, Singlor J, Na-Chiengmai A,Techakumphu M, 2005: Ovum pick-up in cycling andlactating postpartum swamp buffaloes (Bubalis bubalis).Reprod Domest Anim 40, 145–149.Re´vay T, Kova´cs A, Presicce GA, Rens W, Gustavsson I,2003: Detection of water buffalo sex chromosomes inspermatozoa by fluorescence in situ hybridization. ReprodDomest Anim 38, 377–379.Roy KS, Prakash BS, 2007: Seasonal variation and circadianrhythmicity of the prolactin profile during the summermonths in repeat-breeding Murrah buffalo heifers. ReprodFertil Dev 19, 569–575.Saikhun J, Kitiyanant N, Songtaveesin C, Pavasuthipaisit K,Kitiyanant Y, 2004: Development of swamp buffalo (Bubalusbubalis) embryos after parthenogenetic activation andnuclear transfer using serum fed or starved fetal fibroblasts.Reprod Nutr Dev 44, 65–78.Shukla MK, Misra AK, 2007: Effect of bradykinin on Murrahbuffalo (Bubalus bubalis) semen cryopreservation. AnimReprod Sci 97, 175–179.Sing C, 2003: Response of anestrus rural buffaloes (Bubalusbubalis) to intravaginal progesterone implants and PGF2alphainjection in summer. J Vet Sci 4, 137–141.Techakumphu M, Promdireg A, Na-Chiengmai A, PhutikanitN, 2004: Repeated oocyte pick up in prepubertal swampbuffalo (Bubalus bubalis) calves after FSH superstimulation.Theriogenology 61, 1705–1711.Usmani RH, Dailey RA, Inskeep EK, 1990: Effects of limitedsuckling and varying prepartum nutrition on postpartumreproductive traits of milked buffaloes. J Dairy Sci 73, 1564–1570.Vale WG, 1997: News on reproduction biotechnology inmales. In: Proceedings of the Vth World Buffalo Congress,Caserta, Italy, pp. 103–123.Zicarelli L, 1997: Reproductive seasonality in buffalo. BubalusBubalis Suppl 4, 29–52.Author’s address (for correspondence): BMA Oswin Perera, Faculty ofVeterinary Medicine and Animal Science, University of Peradeniya,Peradeniya 20400, Sri Lanka. E-mail: oswinp@pdn.ac.lkConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 207–212 (2008); doi: 10.1111/j.1439-0531.2008.01163.xISSN 0936-6768Postpartum Ovarian Activity in South Asian Zebu CattlePS Brar and AS NandaDepartment of Animal Reproduction, Gynaecology and Obstetrics, Guru Angad Dev Veterinary and Animal Sciences University, Ludhiana, Punjab,IndiaContentsTimely onset of postpartum ovarian activity is vital foroptimal reproductive performance of dairy cows. Muchdepends upon genetic constitution of an animal althoughseveral factors interplay to govern the onset of postpartumovarian activity. South Asian zebu cattle have much longerservice period when compared with other exotic or crossbredcattle reared in the same Asian environment, which suggestsdifferences in their genetic makeup. However, the cows withsame genetic configuration expressed better reproductivepotential when reared under different environment, such asin Brazil and Mexico, which suggests the role of extrinsicfactors such as management, nutrition, environment anddisease conditions. Better management of animals (provisionof proper shade, water and housing, efficient oestrous detectionand timely insemination), good quality nutrition supplementedwith appropriate minerals and vitamins, prevention ofdiseases (vaccination, deworming, suitable therapeutic interventions)and application of biotechnology have helped inimproving postpartum ovarian activity and, therefore, reproductiveperformance of zebu cattle in Asia. No comprehensivestudy appears to have been carried out on the various aspectsof reproduction in zebu cattle reared under South Asian socioagro-climaticconditions. This paper is a modest effort tocollect what ever information available and to critically reviewthe postpartum ovarian activity in zebu cattle with specialreference to the effect of the various managemental practicesand pharmacological interventions.IntroductionZebu cattle (Bos indicus), world’s oldest domesticatedcattle, originated in Western Asia and subsequentlymigrated to larger areas in Asia, Africa and South andCentral America. There are approximately 75 recognizedbreeds, of which 30 exist in India and Pakistanalone (Chenoweth 1994). Sahiwal, Gir, Amritmahal,Haryana, Rathi, Tharparkar and Nelore are some of themost prominent breeds found in South Asia. Originatingfrom the wild, most of these are dual purpose breeds;males being used as draft animals and females for milkproduction. Lately, however, with mechanization ofagricultural operations and commercialization of dairying,high producing milk breeds became preferred. Sointense has been the need and desire for enhanced milkproduction that crossbreeding of zebu with exotic milkbreeds became fashionable in the early sixties andonwards. With this, zebu cattle however faced neglectand their numbers were drastically reduced. Havingevolved under tropical conditions, zebu were betteradapted to hot and dry climatic conditions (Barcelos etal. 1989), can utilize low-quality roughages and havebetter disease resistance when compared with Bostaurus. Their quality of milk is comparable or evenbetter than in crossbred cattle (Joshi et al. 2001). Zebucontributes significantly to draft power and meatproduction in many South Asian countries.The reproductive efficiency of Asian zebu (Table 1),however, is much lower than in their counterparts in thewestern world, the Bos taurus. They have delayed onsetof puberty, poorly expressed oestrus, long inter-calvingintervals (ICI) and distinct seasonality of reproductivetraits. In particular, they suffer from prolonged postpartumanoestrous periods (Abeygunawardena andDematawewa 2004) which make their rearing uneconomical.Hence, it is important to understand their basicreproductive physiology, identify major impedimentsand devise proper interventions to improve postpartumovarian activity.Physiology of Ovarian ActivityThe developmental and the structural properties of thepre-antral follicles are similar in B. indicus and B. taurus.The follicular wave pattern studied in some zebu breeds(e.g. in Rathi cattle; Gaur and Purohit 2007) revealedtwo (78.57%) and three follicular waves (21.42%) withcharacteristics similar to Bos taurus. Two (6.67%), three(60.00%), four (26.67%) and five (6.67%) follicularwaves have also been reported in Gir cows (Viana et al.2000). Certain characteristics of the follicular wave suchas rate of growth and atresia in dominant and subordinatefollicles were similar in cows with differentnumbers of follicular waves (Table 2). There was nodifference of cycle length (21.11 ± 1.76 and22.25 ± 1.71 days) and circulatory progesterone levelsduring dioestrus (14.24 ± 4.61 and 16.15 ± 4.45 nM)incows with one or more follicular waves.Some scientists have considered postpartum pituitaryfunctions in zebu apparently equivalent to Europeancattle (Chenoweth 1994). However, their reproductiveendocrine profiles are generally lower. The basal (Bo etal. 2003) and GnRH-induced LH release is lower in zebuthan in exotic breeds of cattle and their crosses (Portilloet al. 2008), although the follicular size and ovulationrates between Angus cows and Angus-Brahman crossesdid not differ.Circulatory progesterone profiles are also normallylower in zebu (Baruselli et al. 2004), althoughthe pattern of changes in its concentrations duringdifferent stages of reproduction is similar to that inB. taurus. Average serum progesterone concentrationson days 0 (1.30 ± 0.57 nM), 7 (3.87 ± 0.47 nM) and 15(9.41 ± 1.93 nM) of oestrous cycle and on 22 ofpregnancy (11.16 ± 2.13 nM) have been reported(Chakurkar et al. 2004). Ovariectomized zebu cattlerelease progesterone following exogenous challengesÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

208 PS Brar and AS NandaZebu breed; country of observationsPostpartumanoestrousperiod (days)Services perconception ICI (days) ReferenceTable 1. Postpartum fertility traitsin some breeds of zebu cows inAsiaPunganur; India 152 ± 14 1.57 ± 0.1 473 ± 18 Bramhaiah et al. (2003)Ongole; India 113 ± 2 1.53 ± 0.1 407 ± 2 Baburao and Rao (1999)Sahiwal; India 292 ± 15 – 582 ± 15 Katiyar et al. (1993)Non-descript indigenous; Indonesia 151 ± 56 3.0 – Tjiptosumirat et al. (2007)Non-descript indigenous; Malaysia 114 ± 63 2.8 – Nordin et al. (2007)Non-descript indigenous; Myammar 125 ± 32 1.6 – Win et al. (2007)Table 2. Characteristics of follicular waves in Rathi cowsParameterFollicular wavesWave onset (days) 2.10 ± 0.4 10.55 ± 0.6Wave duration (days) 13.35 ± 1.7 10.45 ± 1.0Max. diameter (mm) of dominant follicle 11.75 ± 1.6 14.65 ± 1.2Day of max. diameter (days) 5.00 ± 0.6 17.5 ± 0.5Growth rate (mm ⁄ day) 1.81 ± 0.3 1.16 ± 0.2Growth period (days) 3.10 ± 0.5 8.90 ± 0.8Onset of atresia (days) 10.50 ± 1.1 –Length of atresia (days) 6.50 ± 0.8 –Atresia rate (mm ⁄ day) 0.55 ± 0.2 –Largest subordinate follicle diameter (mm) 7.42 ± 0.6 7.60 ± 0.9with ACTH, indicating that adrenal glands can be anextra ovarian source of progesterone (Bolanos et al.1997).The overall average oestradiol concentrations on days–1, 0, 5, 10, 14, 18 and 22 of oestrus in Rathi cows were29.36 ± 6.90, 37.25 ± 8.00, 18.24 ± 1.58,24.77 ± 4.22, 21.76 ± 5.50, 22.35 ± 3.81 and34.97 ± 7.74 pM, respectively (Purohit et al. 2000).These values are much lower than in western breeds,especially during oestrus (Dobson 1978). This may bethe cause for a very high incidence of silent oestrus inRathi cows under Indian conditions.Postpartum ReproductionAsian zebu cattle have high incidence of postpartumanoestrus (48.7%, Abeygunawardena and Dematawewa2004) leading to very long ‘open days’ (113–291 days;Table 1). Anoestrous zebu crosses continued to havebasal concentrations of FSH (0.75 ± 0.10 nM) andoestradiol-17 b (6.31 ± 0.62 pM; Mondhe et al. 1989)for very long-period postpartum and exhibited the firstprogesterone rise above 2.0 nM by 66.1 ± 6.8 dayspostpartum (Lyimo et al. 2004). In approximately 45%of zebu crosses, the first postpartum oestrous cycles wereof short duration (3.18 nM, which indicated ovulation followed by activeluteal functions, were noticed as early as 25 dayspostpartum in Ongole cattle and the first standingoestrus was observed between 40 and 45 dayspostpartum (Venkantanaidu et al. 2007). This suggeststhat under favourable conditions, the Asian zebu cattlehave the potential of breeding as early in postpartumperiod as their western counterparts. This has furtherbeen vindicated by the performance of zebu cattle rearedunder relatively better environmental and nutritionalconditions in North and South Americas (Chenoweth1994; Bo et al. 2003).Crossbreeding zebu with exotic dairy breedssubstantially improved their milk production but hadvariable effects on reproduction. While Akhtar et al.(2007) reported elongation of ICI in crosses of pure cowswith Pakistani indigenous breeds, others claimedsignificant improvement in their various reproductiontraits, at least in F1 generation. The F2 generation with>75% exotic blood, however, started having multipletype of reproductive problems. The average ICIincreased from 437 ± 28.9 days in F1 generation to451 ± 34.3 days in F2 onwards with >80% HFinheritance (Singh 2005). The gross uterine involutionin zebu cows was completed by approximately 30 days ofcalving but was delayed to 35.14 ± 1.21 days in cowswith >75% exotic blood (Deshmukh and Kaikini 1990).Endocrine Interventions to Hasten Onset ofPostpartum Ovarian ActivityAssessment of circulatory progesterone profiles has beensuccessfully used for early diagnosis of pregnancy incertain South Asian zebu cattle (Boettcher 2007). Thiswould allow early interventions in non-pregnant cows tohave overall beneficiary effects on cattle production.Progesterone radioimmunoassay (RIA) was 100accurate in diagnosing non-pregnancy and 84–90%pregnancy on 20–24 days after artificial insemination(AI) in Bangladesh, Malaysia, Indonesia and Myanmarzebu cattle (Nordin et al. 2007; Shamsuddin et al. 2007;Tjiptosumirat et al. 2007).Certain hormonal interventions that have commonlybeen used to improve reproductive efficiency in Bostaurus have also been investigated in certain Bos indicusbreeds. Exogenous administration of GnRH,progestogens and ⁄ or PGF 2µ hastened the onset ofovarian activity and induced ovulation in postpartumanoestrous zebu cows. Single injection of 250 lg GnRH(ReceptalÒ, Intervet, Holland) induced ovulation inAsian zebu cows having good body condition score(BCS) and well-grown follicles on the ovariesÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Postpartum Ovarian Activity in South Asian Zebu Cattle 209(Abeygunawardena et al. 1992). The results were evenbetter (75% ovulation) when GnRH was injected on theday of lowest vaginal impedance (Sahu et al. 2005a)which normally happens under oestrogen influence,again suggesting the presence of active follicles in suchcases. Administration of GnRH in Gir cows just beforeAI reduced early embryonic mortality and enhancedpregnancy rates by 20%. The concentration of circulatingprogesterone on days 0, 14 and 22 post-treatment,however, was similar to the untreated controls (0.73 vs0.86; 9.92 vs 10.78 and 16.28 vs 14.72 nM, respectively;Shelar et al. 2002)Application of progestogens, more often in the formof norgestomet ear implants (CrestarÒ, Intervet, Holland)embedded subcutaneously for 9–12 days in zebucattle yielded variable results. Approximately 35% and60% cows manifested synergistic sexual behaviourbetween 31 and 57 days after norgestomet implantremoval in dry and wet seasons, respectively (Medranoet al. 1996). Singh et al. (2002) reported oestrousresponses in approximately 71% zebu cows with overallconception rate of up to 83%. Ovulatory oestrus hasalso been reported in 90% Deoni cows with conceptionrate of 70% with 1.85 services per conception (Markendeyaand Bharkad 2004). Solano et al. (2001),however, could achieve ovulations in only 16% Bosindicus cows following norgestomet implants alone.Oestrogen administered along with norgestomet implantsincreased the expression of behavioural oestrusbut had no bearing on ovulation rate. Incorporation ofequine chorionic gonadotropin (eCG) to progestogentherapy induced oestrus in 100% anoestrous zebu cattlewith good BCS (Abeygunawardena et al. 1992).Kathiresan et al. (2001) advocated two simultaneousnorgestomet implants instead of one in Kangeyam cows.While one implant increased LH pulsatile release andmaintained the dominant follicle, two implants suppressedLH release with pattern akin to a mid-lutealphase and restored dominant follicle turn over whichwas useful for oestrous synchronization. Oestrus couldalso be synchronized in cycling zebu with two PGF 2µinjections 11 days apart. Serum progesterone profileshowed that 67% and 100% of the animals responded tothe first and the second injection, respectively (Obasi etal. 1999). Ovsynch oestrous synchronization protocol,which proved meritorious in European breeds, performedpoorly in zebu (Baruselli et al. 2004) but proveduseful in inducing ovarian cyclicity in some anoestrousBos taurus · Bos indicus crosses under tropical conditions(Ahuja and Montiel 2004).Zebu cows respond well to superovulatory protocolsbeing used in European cows. Treatment with FSH as acomponent of the superovulatory protocols generated11.62 ± 1.28 ovulations and recovered 6.25 ± 2.63embryos of which, 2.35 ± 0.96 were viable and transferable.Finally, the subsequent calving rate in the implantedrecipient cows was 27.28% (Chakurkar et al. 2004).Although the results from the use of the aforesaidhormonal regimes were encouraging, yet their adoptionremained minimal. Under field conditions in Asia, zebucattle are generally reared at zero input system. Poor basaldiet, BCS and general management normally prevalentunder these circumstances are bound to affect overallreproductive performance of the cattle. Besides, highcosts of hormonal interventions is a deterrent. This alongwith lack of interest in promotion of indigenous zebucows could be the cause of lack of any proved superovulationand embryo transfer programme in zebu cattle.Managemental Aspects of the Onset ofPostpartum Ovarian ActivityQuality of management of animals affects postpartumreproduction in all species of animals, and so is in zebu.Oestrous detection was missed once or several times in34.7% of the cows, most likely because of poor oestrousdetection system or high incidence of silent oestrus(Lyimo et al. 2004). Oestrous detection is even worst inzebu cows reared by nomads and kept on free grazingpastures. Prevalence of a strong phenomenon of hierarchyin zebu cattle may also affect oestrous expressionand mating in subordinate cows (Solano et al. 2004).Body weight, heart girth and height at withers play asignificant role in determining social ranks among Rathicows. Dominance ranks have also been positivelycorrelated (p < 0.05) to lactation yield (Prasad et al.1996). Environmental extremes, especially the summer,are stressful to zebu cows. Scorching sun restrictsgrazing, thereby affecting body condition and ultimatelyovarian activity (Mitloehner and Laube 2003), cowslower in social ranking being more vulnerable.Long-term exposure of Bos indicus to heat stress haddeleterious effects on follicular dynamics and oocytecompetence (de S Torres-Ju´nior et al. 2008). Provisionof appropriate shade and cooling practices can beemployed to decrease effects of heat stress in cattle(Nienaber and Hahn 2007).Dynamics of body weight at calving, and thereforenutritional status, have great bearing on postpartumovarian activity in zebu. Early gain in body weightpostpartum was associated with early resumption ofovarian activity in Gir cows (Patil and Deshpande1981). Nutritional deficiency, especially during latepregnancy and early postpartum period, results in majorloss of body weight and prolonged postpartum anoestrusin zebu cattle (Montiel and Ahuja 2005). Supplementaryfeeding of UMMB to small zebu cows ofBangladesh initiated ovarian cyclicity by25.50 ± 3.39 days postpartum when compared with>60 days in the control group (Ghosh et al. 1993).Similar observations have been reported followingUMMB supplementation in several other South Asiancountries (Boettcher 2007). Supplementary feeding oftrace minerals and vitamin E for a period of 2 weeksinduced oestrus within 10 days in 65% anoestrousDeoni cows and all conceived to 1.9 services perconception (Markandeya et al. 2002).There is a strong bull effect on postpartum fertility,especially in feed-supplemented zebu cows. Exposure tobulls shortened the average length of postpartumanoestrus in zebu cattle from 119 to 95 days. Nutritionalsupplementation to these further helped their fertility.All such cows exhibited oestrus within 150 days postpartumagainst only 69% controls (Rekwot et al. 2004).Suckling is known to suppress postpartum fertility inmost species of mammals and is commonly practiced inÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

210 PS Brar and AS Nandalactating animals throughout Indian subcontinent, especiallyunder small holding farming system (Singh et al.2006). The phenomenon of suckling appears to be moreintense in zebu cows than in crossbred cows. A study onrestricted suckling revealed that zebu suckled for 36% ofthe grazing time which was significantly longer than incrossbred cows. In addition, the duration of eachsuckling bout and number of suckling bouts weresignificantly higher in zebu cows than in crossbred.The duration and frequency of cross-suckling and intersucklingdecreased with increase in age of calves, andboth behaviour variables were significantly higher forcrossbred calves than for zebu calves. Although zebucows had lower milk yield, the total duration of sucklingand preferential bonding towards their own offspringswere higher than in crossbred cows (Das 1999). Therewas no significant difference in the growth rate ofdivergent follicles between suckling restricted and sucklingnon-restricted cows (0.87 vs 0.93 mm ⁄ day, respectively).However, with 72-h suckling restriction, 94.7%postpartum cows showed signs of oestrus when comparedwith 46.6% in the control group (p < 0.05;Mahecha et al. 2003). Temporary suspension of sucklingenhanced the oestrous synchronization success in Nelorecows (Moraes et al. 2003).Studies on improving success of AI in Ongole cowsrevealed better conception rates with semen depositeddeep intrauterine (73.63%) than in body of uterus(63.79%) or mid-cervix (53.68%) (Rao and Naidu2001). Wrong time insemination (9–15%), a majorcause infertility in zebu cows, has been reported in themost developing countries including Bangladesh, Chinaand Myanmar. Further, microbial endometritis in morethan 11% zebu cows forms other factor limiting thesuccess of AI (Shamsuddin et al. 2007).Zebu cattle in tropical Asia are normally kept bysmall farmers and often under poor managementalconditions. They are prone to parasitic infestations, animportant pathological variable causing prolongedpostpartum anoestrus in zebu cattle. In a study incentral parts of India, Singh and Saxena (2006) revealedbursate worms infestation in 100% anoestrous Haryanacows. Substantially, higher proportion of these cameinto oestrus within 20 days after deworming (26% vs14% untreated cows). Ecto- and endoparasitism ishighly prevalent in zebu cows in tropical Asia. Luckily,however, they have better innate immunity against theseand other tropical diseases compared with the crossbredcattle (Chenoweth 1994). Their reproduction is, therefore,less likely to be affected by such diseases than theircrossbred counterparts.It is thus evident that the reproductive efficiency ofzebu cows could be improved through provision ofproper nutrition, management, especially during periparum,calf weaning and improved oestrous detection(Mugerwa et al. 1991).Conservation of Zebu CattleConservation of zebu cattle is very important owingto their merits of better disease resistance, thermotoleranceand crude fibre utilization than in crossbredand exotic cows. Overall incidence of clinical reproductivehealth problems was lesser in zebu than intheir crosses with exotic blood (22% vs 36%). Studiesby Mandal et al. (2005) revealed that incidence ofabortion, still birth, premature birth, retention ofplacenta, dystocia and total parturient disorders were3.16%, 2.70%, 0.45%, 7.00%, 0.90% and 14.21%,respectively. More cows suffered from parturientdisorders during winter, followed by summer, andrainy seasons. Seasons, however, appeared to have nomajor effect on mortality rate in adult zebu cattle(Prasad et al. 2004), owing perhaps to their innatetolerance to extremes of tropical temperature andhumidity. Pre-implantation zebu embryos are lessvulnerable to elevated temperature than are embryosfrom European breeds (Hansen 2004). In response tothreats of global warming, further development andpropagation of zebu cattle are vital.From the foregoing, it is evident that the zebu cattlehas been and is an integral and vital part of socioeconomiclife, especially in rural tropical Asia. However,these animals did not attract the attention of thescientists and the policy makers alike, and postpartumreproductive physiology remained unexplored and thezebu unexploited. The fact that several breeds of Asianzebu have performed better in American and Australiancontinents, emphasis is needed on studies to developstrategies to exploit the full production potential ofthese breeds within Asia too. As the postpartum periodis the key window in the entire spectrum of reproductiveprocess, much more work needs to be carried out tounderstand the physiology and the limiting factors sothat appropriate interventions to enhance fertility inzebu cattle are developed.ReferencesAbeygunawardena H, Dematawewa CM, 2004: Pre-pubertaland postpartum anoestrus in tropical zebu cattle. AnimReprod Sci 82–83, 373–387.Abeygunawardena H, Abayawansa WD, Kaduwela SC, 1992:Postpartum anostrus in cattle and buffaloes;experiences withhormonal therapy. In: Proceedings of 45th Annual Conventionof Sri Lanka Veterinary Association, Colombo, SriLanka 11–12 December 1992, pp 21.Ahuja C, Montiel F, 2004: Induction of ovulation with GnRHand PGF2 alpha in lactating Bos taurus · Bos indicus cows.Acta Vet Hung 52, 501–508.Akhtar T, Lodhi LA, Khanum SA, Rashad MA, Hussain M,2007: Improving reproductive efficiency in an artificialinsemination programme through early non-pregnancydiagnosis, management and training. In: Boettcher P (ed.),Improving the Reproductive Management of Dairy CattleSubjected to Artificial Insemination, Vienna, Austria.IAEA-TECDOC-1533, pp. 103–112.Baburao K, Rao KS, 1999: Reproductive performance ofOngole cattle. Indian J Anim Reprod 20, 120–122.Barcelos AF, Garcia JA, Cardoso RM, Torres CAA, FerreiraBA, Americo GJ, Maciel CR, Alves CA, 1989: Effects ofclimate and diet on the physiological reactions of buffaloes,zebu cattle, Bos taurus and their crosses. 1. Rectal temperatureand respiration rate. Rev Soc Bras Zoot 18, 32–41.Baruselli PS, Reis EL, Marques MO, Nasser LF, Bo GA,2004: The use of hormonal treatments to improve reproductiveperformance of anoestrus beef cattle in tropicalclimates. Anim Reprod Sci 82–83, 479–486.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Postpartum Ovarian Activity in South Asian Zebu Cattle 211Bo GA, Baruselli PS, Martinez MF, 2003: Pattern andmanipulation of follicular development in Bos indicus cattle.Anim Reprod Sci 78, 307–326.Boettcher P, 2007: Improving the Reproductive Managementof Dairy Cattle Subjected to Artificial Insemination, Vienna,Austria. IAEA-TECDOC-1533.Bolanos JM, Molina JR, Forsberg M, 1997: Effect of bloodsampling and administration of ACTH on cortisol andprogesterone levels in ovariectomized zebu cows (Bosindicus).Acta Vet Scand 38, 1–7.Bramhaiah KV, Rao ST, Reddy YR, 2003: Reproduction inPunjanur cattle. Indian J Anim Reprod 24, 51.Chakurkar EB, Deopurkar VL, Bakshi SA, Chinchkar SR,2004: Success rate of embryo transfer and serum progesteronelevels in recipients of gir and its crosses under farmconditions. Indian Vet J 81, 250–254.Chenoweth PJ, 1994: Aspects of reproduction in female Bosindicus cattle: a review. Aust Vet J 71, 422–426.Das SM, 1999: Performance and behaviour of the cow and calfin restricted suckling and artificial rearing systems. Acta-Universitatis-Agri-Sueciae-Agraria 145, 36.de S Torres-Ju´nior JR, de FA Pires M, de Sa´ WF, de MFerreira A, Viana JH, Camargo LS, Ramos AA, FolhadellaIM, Polisseni J, de Freitas C, Clemente CA, de Sa´ Filho MF,Paula-Lopes FF, Baruselli PS, 2008: Effect of maternal heatstresson follicular growth and oocyte competence in Bosindicus cattle. Theriogenology 69, 155–166.Deshmukh AW, Kaikini AS, 1990: Studies on involution ofuterus in relation to postpartum oestrus in Sahiwal · Jerseycrossbreds with three levels of inheritance. Indian J AnimReprod 11, 122–124.Dobson H, 1978: Plasma gonadotropins and estradiol duringestrus in the cow. J Reprod Fert 52, 51–53.Gaur M, Purohit GN, 2007: Follicular dynamics in Rathi (Bosindicus) cattle. Vetrinarski Arhiv 77, 177–186.Ghosh A, Alam MGS, Akbar MA, 1993: Effect of ureamolasses-mineralblock supplementation on postpartumovarian activity in Zebu cows. Anim Reprod Sci 31, 61–67.Hansen PJ, 2004: Physiological and cellular adaptations ofzebu cattle to thermal stress. Anim Reprod Sci 82–83, 349–360.Joshi BK, Singh A, Gandhi RS, 2001: Performance evaluation,conservation and improvement of Sahiwal cattle in India.Anim Genet Res Inf 31, 43–54.Kathiresan D, Arosh AJ, Devanathan TG, Subramanian A,Nainar MA, Nachimuthu K, 2001: Dose of Norgestometand its effect on the secretary pattern of hormones inKangeyam cows. Indian J Anim Reprod 22, 1–5.Katiyar SS, Singh C, Katiyar MP, 1993: Effect of parity onsome productive and reproductive traits in sahiwal cattle(Bos indicus). Indian J Anim Reprod 14, 30–31.Lyimo ZC, Nkya R, Schoonman L, Eerdenburg FJCM, 2004:Postpartum reproductive performance of crossbred dairycattle on smallholder farms in sub-humid coastal Tanzania.Trop Anim Health Prod 36, 269–279.Mahecha LL, Henao D, Cardona M, Olivera AM, 2003: Effectof suckling restriction on heat presentation in postpartumzebu cows. Revista Colombiana de Ciencias Pecuarias 16,215–219.Mandal DK, Tayagi S, Mukherjee S, 2005: Seasonal variationin the incidence of parturient disorders of Sahiwal cattle.Indian J Anim Reprod 26, 58–59.Markendeya NM, Bharkad GP, 2004: Controlled breedingwith norgestomet ear implant for induction of postpartum oestrus in Deoni cows. Indian J Anim Reprod 25,53–54.Markandeya NM, Bhikane AU, Bharkad GP, 2002: Clinicalresponse to microminerals with vitamin E supplementationin anoestrus Deoni cows. Indian J Anim Reprod 23, 77–78.Medrano EA, Hernandez O, Lamothe C, Galina CS, 1996:Evidence of asynchrony in the onset of signs of oestrus inzebu cattle treated with a progestogen ear implant. Res VetSci 60, 51–54.Mitloehner FM, Laube RB, 2003: Chronobiological indicatorsof heat stress in Bos indicus cattle in the tropics. J Anim VetAdv 2, 654–659.Mondhe DM, Dopurkar VL, Hukeri VB, Mantri AM,Hingane VR, Kaikini AS, 1989: Serum FSH and oestradiol-17bconcentrations in postpartum anoestrus crossbredcattle. Indian J Anim Reprod 10, 6–9.Montiel F, Ahuja C, 2005: Body condition and suckling asfactors influencing the duration of postpartum anestrus incattle: a review. Anim Reprod Sci 85, 1–26.Moraes TT, Okuda HT, Bianchinni SE, 2003: The calvingafter the first estrous period observed after oestrus synchronizationprotocol associated with the temporary interruptionof suckling. Revista Brasileira de Reproducao Animal27, 666–673.Mugerwa EM, Tegegne A, Ketema H, 1991: Patterns ofpostpartum oestrus onset and associated plasma progesteroneprofiles in Bos indicus cows in Ethiopia. Anim ReprodSci 24, 273–284.Nienaber JA, Hahn GL, 2007: Livestock production systemmanagement responses to thermal challenges. Int J Biometeorol52, 149–157.Nordin Y, Zaini N, Wan Zahari WM (2007) Reproductivestatus following artificial insemination and factors affectingconception rate in dairy cows in smallholder productionsystems. In: Boettcher P (ed.) Improving the ReproductiveManagement of Dairy Cattle Subjected to ArtificialInsemination, Vienna, Austria. IAEA-TECDOC-1533, pp.79–92.Obasi OL, Ogwu D, Mai H, 1999: Efficiency of prostaglandinF2 alpha in the induction of physiological oestrus in Zebucattle. J Sus Agri Env 1, 114–118.Patil JS, Deshpande BR, 1981: Study of body weight changesduring antepartum, parturition and postpartum period inGir cows with special reference to exhibition of postpartumactions. Indian Vet J 58, 376–379.Portillo GE, Bridges GA, de Araujo JW, Shaw MKV, SchrickFN, Thatcher WW, Yelich JV, 2008: Response to GnRH onday 6 of the estrous cycle is diminished as the percentage ofBos indicus breeding increases in Angus, Brangus, andBrahman · Angus heifers. Anim Reprod Sci 103, 38–51.Prasad S, Mittal JP, Kaushish SK, Prasad S, 1996: Dominancepattern in free grazing zebu cattle. Indian J Anim ProdManag 12, 99–103.Prasad S, Ramachandran N, Raju S, 2004: Mortality patternsin dairy animals under organized herd managementconditions at Karnal, India. Trop Anim Health Prod 36,645–654.Purohit GN, Datt M, Shama SS, Upadhyaya RC, 2000:Estradol profile of Rathi cattle during estrus cycle. Indian JAnim Reprod 21, 6–7.Rao BK, Naidu VG, 2001: Effect of site of semen depositionon conception rate in Ongole cows. Indian J Anim Sci 22,60–62.Rekwot PI, Akinpelumi OP, Sekoni VO, Eduvie LO, OyedipeEO, 2004: Effects of nutritional supplementation andexposure to bulls on resumption of post-partum ovarianactivity in Bunaji (Bos indicus) cattle. Vet J 167, 67–71.Sahu SK, Tiwari RP, Mishra UK, 2005a: Treatment ofpostpartum anoestrus Sahiwal cows on the basis of VMIand ultrasonography using single injection of GnRH. IndianJ Anim Reprod 26, 99–101.Sahu SK, Tiwari RP, Mishra UK, Mukherejee K, 2005b:Ovarian activity and vaginal resistance in postpartumanoestrus Sahiwal cows. Indian J Anim Reprod 26, 109–112.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

212 PS Brar and AS NandaShamsuddin M, Hossein MS, Siddiqui MAR, Khan AHMSI,Bari FY, Alam MF, Rahman M, Sayem ASM, Momont H,2007: Use of milk progesterone radioimmunoassay andcomputer applications for community based reproductivehealth services in smallholder dairy farms of Bangladesh. in:Boettcher P (ed.) Improving the Reproductive Managementof Dairy Cattle Subjected to Artificial Insemination, Vienna,Austria. IAEA-TECDOC-1533, pp. 9–22.Shelar RR, Deopurkar VL, Bakshi SA, Gulavane SU, 2002:Efficacy of pre-insemination treatment with GnRH forimproving conception rate in repeat breeder cows. Indian JAnim Reprod 23, 69–70.Singh A, 2005: Crossbreeding of cattle for increasing milkproduction in India – A review. Indian J Anim Sci 75,383–390.Singh CP, Saxena A, 2006: Study of effect of anthelmintictreatment on metabolic profile of anoestrus Haryana cows.Indian J Anim Reprod 27, 91–93.Singh U, Singh I, Khar SK, Singh U, Singh I, 2002: Influenceof season, age and postpartum interval on reproductiveparameters of norgestomet treated anestrus zebu cattle.Indian J Anim Reprod 23, 113–116.Singh AK, Brar PS, Nanda AS, Parkash BS, 2006: Effect ofsuckling on basal and GnRH induced LH release inpostpartum dairy buffaloes. Anim Reprod Sci 95,244–250.Solano J, Orihuela A, Galina CS, Montiel F, 2001: Sexualbehavior of Zebu cattle (Bos indicus) following estrousinduction by Syncro-Mate B, with or without estrogeninjection. Physiol Behav 71, 503–508.Solano J, Galindo F, Orihuela A, Galina CS, 2004: The effectof social rank on the physiological response during repeatedstressful handling in Zebu cattle (Bos indicus). Physiol Behav82, 679–683.Tjiptosumirat T, Tuasikal BJ, Murni AP, Lelananingtyas N,Darwati S, Ariyanto A, Yunita F, Mondrida G, Triningsih ,Setyowati S, Sutari , Toleng AL, Arman C, Rizal Y, 2007:Improvement of the efficiency of artificial inseminationservices through the use of radioimmunoassay and acomputer database application. In: Boettcher P (ed.)Improving the Reproductive Management of Dairy CattleSubjected to Artificial Insemination. IAEA-TECDOC-1533,pp. 57–78.Venkantanaidu G, Rao SA, Rao BK, 2007: Progesteroneprofile in postpartum lactating Ongole (zebu) cows. Indian JAnim Reprod 28, 12–14.Viana JHM, Ferreira A de M, Sa WF de, Camargo LS de, AFerreira A de, Sa WF, de A Camargo LS, 2000: Folliculardynamics in zebu cattle. Pesqui Agropecu Bras 35, 2501–2509.Win N, Win YT, Kyi SS, Myatt A, 2007: Evaluation ofreproductive performance of cattle bred by artificial inseminationin Myanmar through the use of progesteroneradioimmunoassay. In: Boettcher P (ed.) Improving theReproductive Management of Dairy Cattle Subjected toArtificial Insemination, Vienna, Austria. IAEA-TECDOC-1533, pp. 93–102.Author’s address (for correspondence): PS Brar, Department ofAnimal Reproduction, Gynaecology and Obstetrics, Guru AngadDev Veterinary and Animal Sciences University, Ludhiana, Punjab,India. E-mail: parkashsingh@satyam.net.inConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 213–216 (2008); doi: 10.1111/j.1439-0531.2008.01164.xISSN 0936-6768Mother–Offspring Interactions in Zebu CattleMJR Paranhos da Costa 1,2 , A Schmidek 1,3 and LM Toledo 1,41 Grupo ETCO – Grupo de Estudos e Pesquisas em Etologia e Ecologia Animal; 2 Departamento de Zootecnia, FCAV-UNESP; 3 Programa de Po´sGraduaça˜o em Gene´tica e Melhoramento Animal, Faculdade de Cieˆncias Agra´rias e Veterina´rias, UNESP, Jaboticabal; 4 APTA, Po´lo Regional doVale do Ribeira, UPD Registro, Sa˜o Paulo, BrazilContentsThe knowledge of the interaction between mother andoffspring might contribute to enhance the welfare of theoffspring and to improve the reproductive efficiency of thecow. However, there is still little information available aboutsuch interaction in some cattle breeds. A series of observationalstudies were set up, addressing the mother–offspringrelationships of Nelore, Guzerat and Gyr cattle breeds. Firstly,the behaviour of cows and calves around the time ofparturition was described, and then, the underlying factorsthat affect the calves’ survival and development were studied.Special attention was given to the failure or delay in the firstsuckling. The results together are indicative of geneticvariability for some studied variables, indicating the possibilityof selection for calf vigour (using latency to stand up andlatency to suckle as its indicators) and maternal ability (usingpercentage of time in contact with the calves), in spite of theestimates of heritability were low and presented high standarddeviation for all variables. The individual variability in theirsuckling behaviour and the efficiency in first suckling cannot beexplained by a single isolated underlying factor. By now, thereare some results available, although there are many questionswithout answers. The field is still open for the development offuture research.IntroductionThe adaptation of a newborn calf to the extra-uterineenvironment is clearly dependent to the expression ofappropriate behaviour, presented by itself and by itsmother. It must stand up and suckle as soon as possible,while its mother must take care of it, licking, protectingagainst predators and remaining quiet, standing, whenthe calf tries to get the teat into its mouth and suckle(Lidfors 1994; Fraser and Broom 1997; Paranhos daCosta and Cromberg 1998). The importance of thesebehaviours for the calves’ survival has been reported(Schmidek 2003; Ribeiro et al. 2007), and when amother licks its offspring it also stimulates the calf’srespiratory and circulatory systems, removes the foetalmembranes from its body and stimulates the defecationand urination (Lidfors 1994; Fraser and Broom 1997).The first suckling must occur soon after calving, and thelack or delay of it increases the incidence of calfmortality (Paranhos da Costa and Cromberg 1998;Schmidek et al. 2008).Quality and intensity of maternal behaviour dependson a complex interaction among genetic, physiologicalfactors and maternal experience (Edwards and Broom1982; Lawrence and Fowler 1997; Paranhos da Costaand Cromberg 1998).After the recognition of its own offspring, a cowusually does not allow other calves to suckle. However,some previous studies described the occurrence ofallosuckling in cattle (when a cow allows other calf,than its own, to suckle), ranging from 3.0% (Lewandrowskiand Hurnik 1983; Das et al. 2000) to 19.02% ofthe total suckling events (Víchova´ and Bartosˇ 2005), butit seems to be rare in Zebu cattle (Paranhos da Costaet al. 2006a).Theoretically, these variations could be resultant frommany causal factors, genetic and environmental. This isa vague statement and does not help to understand thebiological phenomenon neither to solve the practicalproblems resulting from it. There are some empiricalevidences that, under restricted milk and space conditions,the probability of allosuckling increases in waterbuffalo (Murphey et al. 1991, 1995) and the fact thatallosuckling increases the competition for milk amongwater buffalo calves (Paranhos da Costa et al. 2000).This statement is, in some extension, supported by theresults with cattle, for example, Paranhos da Costa et al.(2006a) studied Zebu cattle kept in free-range conditions,the cows were not milked and the calves had freeaccess to suckle, and under these conditions, the authorsdid not record any single case of allosuckling. On thecontrary, the studies of Lewandrowski and Hurnik(1983), Das et al. (2000) and Vı´chova´ and Bartosˇ (2005)reported the occurrence of allosuckling in cattle, andunder their studies conditions, the cows and calves hadsome restriction in space and milk availability. In spiteof the research interests on this subject, for theoreticaland practical reasons, the underlying factors of allosucklingand its role on the rate of calf survival and calfperformance are still not well explained.The mother–offspring relationships of cattle are stillnot well understood in many aspects, probably becauseof the complex scenario and the challenging situationsinvolved on their expression. The aim of this article is topresent some results achieved through the behaviouralstudies of Zebu cattle in Brazil and to introduce somepractical recommendations about how to improve themanagement based on these studies.The Importance of the Behavioural Studies toIncrease Reproductive EfficiencyThe reproductive (and economic) efficiency of a cow isdirectly dependent of the survival of its offspring. Theknowledge of the mother and offspring behaviour isimportant to identify situations that could increase therisk of calf weakness, abandon or death, giving theopportunity for one to develop appropriate strategies tominimize these problems and the respective economicÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

214 MJR Paranhos da Costa, A Schmidek and LM Toledolosses (Cromberg and Paranhos da Costa 1997; Schmideket al. 2008).Cows usually isolate themselves from the herd whennear to calve. However, when they are in a smallpaddock, they are not able to do it, resulting in a higherrisk of fail or delay in the first suckling. This is because,under high density conditions, the social interactionsamong parturient cows are more frequent and sometimescalves spend time trying to suckle in other cowsthan in their own mothers (Ribeiro et al. 2004), justbecause these cows are close to the parturient cow andits respective calf, or because some cows show interesttowards a calf, behaving like its mother.Throughout the parturition process, cows usually lyedown and stand up just after delivering the calves(Paranhos da Costa and Cromberg 1998). The identificationof this simple behaviour has practical relevance, asthe risk of calf death is higher when the cow delivers in astanding (16.1%) rather than in a recumbent position(4.2%) (Paranhos da Costa et al. 2006b). Better understandingof this situation and its underlying factors mayassist to solve this problem. Some factors that affect thecows’ postures when delivering their calves were alreadyidentified, among them: lack of calving experience,presence of potential predators near the parturient cows(Toledo 2005; Paranhos da Costa 2008).The episodes of aggressive behaviour towards anewborn calf are also important. According to Schmideket al. (2006), they were more frequent in primiparousthan in multiparous cows (55.7% and 22.0%,respectively), resulting on higher frequency of calvesthat did not succeed to suckle after 3 h after birth(15.7% and 5.7%, respectively) and higher latency forfirst suckling (102.6 and 76.0 min, respectively). The riskof calf mortality also increase with the delay in the firstsuckling latency above 3 h (Schmidek et al. 2006), andby this, we recommend the need of reviewing the criticallimit for the cattle first suckling latency, usually reportedas 6 h after calving (Broom 1983).The presence of potential predators in the calvingplace also affects the mother–young relationship. Underour environmental conditions, special attention wasgiven to the black vultures (Coragyps atratus) becausethis species is more and more present at the cattlecalving places, disturbing the formation of the mother–offspring bond and attacking directly the calves. Themain effect of the presence of black vulture was todecrease the time that the cows spend in contact withtheir calves, as usually they spend more time invigilance. This situation resulted in negative effects onthe latencies to stand up and first suckling (Toledo2005).Weather conditions (Toledo et al. 2007), calf weightat birth, udder shape and teat size (Ventrop andMichanek 1992) are other identified factors that wouldresult in fail or delay in the first suckling.Failure in the first suckling increases the mortality rate ofcalvesIn animal husbandry, this subject has crucial importancebecause the death of one animal means an economicalloss (Cromberg and Paranhos da Costa 1997).It is expected that cow–calf behaviour during the firsthours after calving has a direct effect on the calfmortality; by this, the knowledge of the mother–offspring behaviour has practical significance, becauseit can be used to evaluate the strategies of managementand selection in beef cattle. In spite of this, there are notmuch data about this subject, especially when consideringcattle bred in Brazil.In one of our studies (Schmidek et al. 2008), the highpercentage of newborn Guzerat calves that fail in thefirst suckling stimulated us to study the genetic andenvironmental factors underlying this situation, and itseffect on calf mortality. Data from 1527 calves (bornbetween 1992 and 2004) revealed that 279 (18.3%) ofthem failed in first suckling and were helped to suckle,more than 6 h after birth. Two major factors wereidentified for the occurrence of the failure in the firstsuckling: cows with big udders or teats and calvesweighing less than 25 kg at birth. The risk of death forcalves that failed in the first suckling was 28% timeshigher (p < 0.01) than those that were able to suckle ontime. The estimated maternal effect presented lowheritability (0.08) and the correspondent direct effectwas null. These low heritability estimates indicate thatthe genetic progress through selection criteria againstfailure of the first suckling might be achieved only inlong term. However, the conformation of the udder andthe birth weight of the calf are two possible indicatorsfor immediate action to help the calf to suckle as soon aspossible.Routine inspections in the maternity area to detecteventual problems with parturient cows and neonates(ideally three times in the day: early morning, noon andlate afternoon) could minimize the incidence of thefailure in the first suckling of newborn calves (Paranhosda Costa 2008; Schmidek et al. 2008).Variability in the expression of cow and calf behavioursamong and within breedsThe differences in suckling behaviour seem to beproduced by a complex combination of genetic andenvironmental factors, which would result in a particularbehavioural relationship style of a mother–offspringpair (Paranhos da Costa et al. 2006a). There arescientific evidences of variability among Zebu breeds,among herds of the same breed and among individualswithin herd; for example, the latency to stand up afterparturition varied significantly between Nelore andGuzerat breed (Paranhos da Costa and Cromberg1998), between two herds of Nelore breed (Toledo et al.2007) and among several progenies of Nelore bulls(Schmidek 2003).It is usually accepted that the behavioural variationbetween species, breeds, populations and individuals hasa genetic bases, by this, the most common strategy toidentify genetic variability in behaviour is the comparisonamong breeds (Hohenboken 1986; Buchenauer1999). For example, the variation in maternal behaviourbetween beef and dairy cows was reported (Le Neindre1989; Buchenauer 1999), showing that beef cowsdisplayed better maternal behaviour than dairy cows.Besides, beef calves are usually more agile to stand upÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Mother–Offspring Interactions 215and suckle than the dairy ones. The reason for suchdifference is probably because of selection, as we expectthat dairy cows give milk without the offspring stimulation.In such conditions, most of the calves’ needs aresupplied by human beings, which result in the relaxationof the selection pressure, turning the survival of calvesless dependent on its own and mother behaviours.There are few studies estimating the genetic varianceof mother–offspring behaviour, probably due mainly tothe difficulties for data recording. Only three studiesaddressing this specific subject were found. Two of themwith lambs (Cloete et al. 1998, 2002), showed lowcoefficients of heritability (h 2 , direct and maternal) forthe latencies to stand up (h 2 = 0.10 ± 0.05 and0.09 ± 0.04, respectively) and to suckle after standingup (h 2 = 0.07 ± 0.04 and 0.19 ± 0.04, respectively),and one study with Nelore cattle (Schmidek 2003),which also showed low coefficients of heritability for thelatencies for calves to stand up, to suckle after stand upand to suckle (h 2 = 0.16 ± 0.17, 0.09 ± 0.16 and0.13 ± 0.18, respectively) and the percentages of timespent by cows in contact with the calves and disturbingthe calves (h 2 = 0.22 ± 0.14, 0.00 ± 0.12 and0.13 ± 0.18, respectively). These results are suggestive,indicating the existence of additive genetic variabilityfor, at least, three variables: the latencies to stand upand to suckle, (which could be used as an indicator ofcalf vigour) and the percentage of time that a cow was incontact with its calf (which is an indicator of maternalability).ConclusionsThe rate of calf survival is an important indicator ofreproductive efficiency in a cattle herd. As reported,there are many factors determining the risk of calfdeath, and the role of most of them is still not wellunderstood. It is necessary to keep on working to clarifythis subject and understand the relationship among cowand calf behaviour and their reproductive performance.The results presented in this paper bring some light onthe subject, and reinforce the idea that the appropriatedexpression of mother–offspring behaviour is essential toachieve a high level in the reproductive performance.Taking the results of the behavioural studies in account,one can adopt some practical handling procedures todecrease the risk of fail or delay in the first suckling. Thedefinition of the space availability in the birth paddock,for example, allows a cow to move away from the herdjust before birth, and this is important to establish anappropriated attachment between mother and offspring.Special attention should be given to the first calvingcows, including a special paddock for calving. Theyrequire better environment and more intensive assistanceto achieve good reproductive performance; doingthis, it is possible to decrease the risk of abandonedcalves and first suckling fail or delay.It is also recommended to organize periodic inspectionsin the calving paddock, ideally three times in theday (early morning, noon and late afternoon). Doingthis, it will be possible to assess the risk for cows andneonates, identifying situations that would result inproblems for both, and giving the opportunity to actpreventively, reducing deaths and optimizing the reproductiveefficiency of the herd.The results also indicate the possibility to improve thecalf survival through selective breeding, consideringbehavioural traits (percentage of time spent by a cow incontact with its offspring, and the latencies to stand upand to suckle) as indicators of maternal ability and calfvigour, respectively.In conclusion, there is some information availableabout the mother–offspring interactions that could beused to improve Zebu cattle reproductive performance.However, the picture is not complete; there are manyquestions without answers and problems to be solved.The field is still open for the development of futureresearch.AcknowledgementsFinancial support was provided by Fapesp (Fundação de Ampaso doEstado de Sa˜ o Paulo) and CNPq (Conselho Nacional de DesenvolvimentoCientifico e Technologico).ReferencesBroom DM, 1983: Cow-calf and sow-piglet behaviour inrelation to colostrum ingestion. Ann Rech Vet 14, 342–348.Buchenauer D, 1999: Genetics of behaviour in cattle. In: FriesR, Ruvinsky A(eds). The Genetics of the Cattle. CABInternational, Wallingford, pp. 365–390.Cloete SWP, Olivier JJ, Scholtz AJ, 1998: Aspects of lambingand neonatal behaviour of dual-purpose sheep. In: Proceedingsof the 6th World Congress on Genetic Applied toLivestock Production, Armidale, NSW, Australia, vol. 27,pp. 39–42.Cloete SWP, Scholtz AJ, Gilmour AR, Olivier JJ, 2002:Genetic and environmental effects on lambing and neonatalbehaviour of Dormer and AS Mutton Merino lambs. LivestProd Sci 78, 183–193.Cromberg VU, Paranhos da Costa MJR, 1997: Mamandologo, para crescer a receita. Anualpec 97, 215–217.Das SM, Redbo I, Wiktorsson H, 2000: Effect of age of calf onsuckling behaviour and other behavioural activities of Zebuand crossbred calves during restricted suckling periods.Appl Anim Behav Sci 67, 47–57.Edwards SA, Broom D, 1982: Behavioural interactions ofdairy cows with their newborn calves and the effects ofparity. Anim Behav 30, 525–535.Fraser AF, Broom DM, 1997: Parturient behaviour. In: FraserAF, Broom DM (eds), Farm Animal Behaviour andWelfare. CAB International, Wallingford, UK, pp. 208–218.Hohenboken WD, 1986: Inheritance of behavioural characteristicsin livestock: a review. Anim Breed Abstr 54, 623–639.Lawrence TLJ, Fowler VR, 1997: Growth of Farm Animals.CAB International, Wallingford, 330 pp.Le Neindre P, 1989: Influence of rearing conditions and breedon social relationships of mother and young. Appl AnimBehav Sci 23, 129–140.Lewandrowski NM, Hurnik JF, 1983: Nursing and crossnursingbehaviour of beef cattle in confinement. Can J AnimSci 63, 849–853.Lidfors L, 1994: Mother-young behaviour in cattle. Thesis(Doctoral). Swedish University of Agricultural Science,Skara, Sweden.Murphey RM, Paranhos da Costa MJR, Lima LOS, DuarteFAM, 1991: Communal suckling in water buffalo (Bubalusbubalis). Appl Anim Behav Sci 28, 341–352.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

216 MJR Paranhos da Costa, A Schmidek and LM ToledoMurphey RM, Paranhos da Costa MJR, Silva RG, SouzaRC, 1995: Allonursing in river buffalo, Bubalus bubalis:nepotism, incompetence or thievery? Anim Behav 49,1611–1616.Paranhos da Costa MJR, 2008: Improving the welfare ofcattle: practical experience in Brazil. In: Dawkins MS,Bonney R(eds), The Future of Animal Farming. BlackwellPublishing, Oxford, pp 145–152.Paranhos da Costa MJR, Cromberg VU, 1998: Relac¸o˜ esmaterno-filiais em bovinos de corte nas primeiras horas apo´so parto. In: Paranhos da Costa MJR, Cromberg VU(eds),Comportamento materno em mamı´feros: bases teo´ricas eaplicac¸o˜ es aos ruminantes domésticos. Sociedade Brasileirade Etologia, Sa˜ o Paulo, pp. 215–236.Paranhos da Costa MJR, Andriolo A, Oliveira JFS, SchmidekWR, 2000: Suckling and allosuckling in river buffalo calvesand its relation with weight gain. Appl Anim Behav Sci 66,1–10.Paranhos da Costa MJR, Albuquerque LG, Eler JP, Silva JAII, de V, 2006a: Suckling behaviour of Nelore, Gir andCaracu calves and their crosses. Appl Anim Behav Sci 101,276–287.Paranhos da Costa MJR, Schmidek A, Toledo LM, 2006b:Boas pra´ticas de manejo: bezerros ao nascimento. EditoraFunep, Jaboticabal-SP, 36 pp.Ribeiro L, Toledo LM, Paranhos da Costa MJR, 2004:Influeˆ ncia de locais do parto no comportamento de vacas ebezerros da raça Nelore. In: ZOOTEC, Brası´lia. Anais.Brası´lia. 1 CD Rom .Ribeiro ARB, Alencar MM, Paranhos da Costa MJR, Negra˜ oJA, 2007: Effects of sire breed-grazing system and environmentalparameters on the behaviour of beef calves just afterbirth. Appl Anim Behav Sci 107, 198–205.Schmidek A, 2003: Ana´lise de fatores genéticos e ambientaisrelacionados a caracterı´sticas de vigor e qualidade materna,para as raças Nelore e Guzera´. Dissertac¸ão de Mestrado,Faculdade de Cieˆ ncias Agrárias e Veterina´rias, UniversidadeEstadual Paulista, Jaboticabal-SP, Brasil.Schmidek A, Paranhos da Costa MJR, Mercadante MEZ,Toledo LM, 2006: The effect of newborn calves vigour intheir mortality probability. In: Proceedings of the 40thInternational Congress of the International Society ofApplied Ethology, Bristol, 221 pp.Schmidek A, Mercadante MEZ, Paranhos da Costa MJR,Figueiredo LA, 2008: Falhas na primeira mamada em umrebanho da rac¸a Guzera´: fatores predisponentes, reflexos nasobreviveˆ ncia do bezerro e paraˆ metros genéticos. Rev Brasde Zootec 37, 998–1004.Toledo LM, 2005: Fatores intervenientes no comportamentode vacas e bezerros do parto ate´ a primeira mamada. Tese deDoutorado, Faculdade de Cieˆ ncias Agrárias e Veterina´rias,Universidade Estadual Paulista, Jaboticabal-SP, Brasil.Toledo LM, Paranhos da Costa MJR, Titto EAL, Figueiredo LA,Ablas DS, 2007: Impactos de varia´veis clima´ticas na agilidadede bezerros Nelore neonatos. Ciência Rural 37, 1399–1404.Ventrop M, Michanek P, 1992: The importance of udder andteat conformation for teat seeking by the newborn calf.J Dairy Sci 75, 262–268.Vı´chová J, Bartosˇ L, 2005: Allosuckling in cattle: gain orcompensation? Appl Anim Behav Sci 94, 223–235.Author’s address (for correspondence): Mateus JR Paranhos da Costa,Departamento de Zootecnia, FCAV-UNESP, 14884-900 Jaboticabal-Sa˜ o Paulo, Brazil. E-mail: mpcosta@fcav.unesp.brConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 217–223 (2008); doi: 10.1111/j.1439-0531.2008.01165.xISSN 0936-6768An Update on Reproduction in Yak and MithunBS Prakash 1 , M Sarkar 2 and M Mondal 31 Dairy Cattle Physiology Division, NDRI, Karnal, Haryana; 2 National Research Centre on Yak, Dirang, Arunachal Pradesh; 3 National ResearchCentre on Mithun, Jharnapani, Nagaland, IndiaContentsYak and mithun are two domesticated herbivores of higheconomic importance to the farming community living inhighlands. Improved yak and mithun production couldsignificantly enhance the living standards of these highlanders.Over the years, their dwindling numbers have been a cause ofserious concern. In view of the lack of knowledge on thereproductive physiology of these ruminants, studies have beenundertaken to investigate their reproductive endocrinology inrecent years. This paper attempts to present the latestinformation on the endocrine changes associated with theirvarious reproductive processes viz. growth and puberty,oestrous cyclicity and oestrous behaviour, ovulation, and,pregnancy and parturition. The paper also provides the recentdevelopments on research done towards the enhancement ofyak and mithun reproductive efficiencies through endocrineand embryo biotechniques.IntroductionThe yak is a unique bovine species domesticated by thehighlanders inhabiting the high altitude inhospitableterrains in the middle and inner Himalayan region, andsurvives only at high altitudes between 2500 and6000 m. The world’s total yak population is estimatedto be approximately 14 million of which more than 90%is in China (Wiener et al. 2003). India possessesapproximately 70 000 yaks in the high reaches ofHimalayan region. The animal is a seasonal breederand the factors which are considered to be responsiblefor seasonality in breeding include nutrition, climate andaltitude. Mithun (Bos frontalis), believed to be thedescendent of wild gaur (Bos gaurus gaurus) of India andMyanmar, is an animal of special significance for thetribals of North Eastern-Hills region of India and isbasically used for beef purpose besides its use in bartertrade, marriage gift, etc. Probably, mithun is the onlyruminant that can browse on biomass most efficiently ondifficult steep hill slopes under adverse climatic conditionprevailing in the region. At present, India had atotal population of 0.27 million mithuns (Ministry ofAgriculture Department of Animal Husbandry, Dairying& Fisheries 2003).Survey reports reveal that majority of Indian yaks andmithuns suffer from various reproductive problems likelate maturity, long calving interval, poor oestrousexpression and repeat breeding; in addition yaks areseasonal breeders. The information related to basicendocrinology is, therefore, paramount not only forbetter understanding of these animals and improvinggrowth, inducing early puberty and improving itsfertility but also for application of endocrine biotechniquessuch as oestrus synchronization and set timeAI, superovulation and embryo biotechniques forimplementation of ex situ as well as in situ conservationof valuable germplasm. The present paper highlightssome of the recent research findings on reproductiveendocrinology and subsequent use of various biotechnologicaltools for augmenting fertility in these twowonderful species.Development of Hormone AssaysWe have developed sensitive enzyme immuno-assay(EIA) methods using the second antibody coatingtechnique on microtitre plates for the determinationof LH (Mondal et al. 2005a; Sarkar and Prakash2005a), oxytocin (Mondal et al. 2006d; Sarkar andPrakash 2006), GH (Mondal et al. 2005b; Sarkar et al.2007d), prolactin (Sarkar and Prakash 2006; Mondalet al. 2007a), total oestrogen, oestradiol-17b (Sarkaret al. 2006b; Mondal et al. 2007b), cortisol (Sarkar et al.2007b) and PGFM (Mondal et al. 2006a; Sarkar et al.2007c).Growth and Puberty in Yak and MithunThe role of GH in postnatal somatic growth is wellestablished (Breier and Gluckman 1991). Age at pubertyinfluences production efficiency in animals (Short andBellows 1971; Hoffman and Funk 1992). Factors modulatingthe endocrine regulation of onset of puberty arenot fully understood. Physiological mechanism thatcontrols puberty by changing interplay between hormonessecreted from the hypothalamus, the anteriorpituitary and the gonads lead to a transition from sexualquiescence to sexual function (Hull and Harvey 2001).Among these complex endocrine interaction, GH acts asa common hormonal link (Hull and Harvey 2001).Plasma progesterone level which increases with theadvancement of age and increasing body weight isconsidered to be an indicator of puberty in domesticanimals (El-Nouty 1971). We found that with increasingage and body weight (BW), GH and GH per 100 kg BWboth decreased and the animal with higher plasma GHand GH per 100 kg BW showed higher growth rates inmithuns (Mondal et al. 2007c) and yaks (Sarkar et al.2008).In growing and adult mithuns, we found that GHpatterns consisted of frequent pulses of varyingamplitude. Growth hormone pulses occurred at anaverage frequency of 0.69 and 0.48 ⁄ h in growing andadult mithuns, respectively (Mondal et al. 2004,2005d). Peak and mean GH levels were associatedpositively (r = 0.98, p < 0.001) with rates of weightgain. Results from these studies indicated that the GHÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

218 BS Prakash, M Sarkar and M MondalProgesterone, LH (ng/ml)4.03.53.02.52.01.51.00.50.0Progesterone LH GHGroup-I Group-II Group-III Group-IV Group-V Group-VIGroupsFig. 1. Changes in plasma progesterone, LH and GH for consecutive6 weeks in mithuns of six different age groups from birth throughadulthood. Group I (0–3 months; n = 9), group II (>3–6 months;n = 7), group III (>6–12 months; n = 11), group IV (>12–18months; n = 10), group V (>18–24 months; n = 10) and group VI(>24–31 months; n = 12). Blood samples were collected from allanimals for six consecutive weekspulses occur at frequent intervals throughout the dayand night and alterations in GH levels and patternsare elicited more by pulse amplitude than frequencymodulation. LH pulses occurred at an average rate of0.54 ⁄ h (5 pulses ⁄ 9 h). The mean plasma LH levelwas correlated with body weight (r = 0.82; p < 0.05)and pulse amplitude in growing mithuns (Mondalet al. 2005c). There was higher LH concentrationswith higher pulsatility and greater amplitude in prepubertalmithuns than exhibited in growing mithuns(Mondal et al. 2005e). Our results showed increasingtrend of plasma progesterone and LH concentrationsin the process of onset of puberty in mithun (Fig. 1).Plasma GH levels were recorded to be doubled duringpubertal process (Fig. 1). The optimum LH pulsefrequency and amplitude required for onset of pubertyin mithun were ‡9 pulses ⁄ 24 h and ‡1.6 ng ⁄ ml,respectively.Possible use of GH-releasing factor as an inducer ofpubertyKeeping in view the promising results obtained recentlyof using exogenous GH-releasing hormone (GHRH)for growth enhancement (Mondal and Prakash 2004)and the suggestion of Gelato and Merriam (1986) thatGHRH being active in a wide range of species andhence its treatment could potentially be used toaccelerate growth of animals of commercial importance,we first standardized the dosage of GHRH inmithun (Mondal et al. 2006c) and then went on to testthe effects of exogenous GHRH on secretion patternsof GH and LH as a preliminary study for the first timeever in this ruminant species (Mondal et al. 2006e).Our results suggest that 10 lg GHRH per 100 kg BWis the dosage, which can be used for augmentation ofmithun production. Using this dosage, GHRH wasfound to increase plasma GH and LH pulse frequencyand amplitude with higher mean LH levels in growingmithuns suggesting the possibility of using such strategyfor enhancement of maturity and ⁄ or pubertyprocess in this species.160140120100806040200GH (ng/ml)Endocrine Changes Associated withReproductive Processes in Yaks and MithunsRelationship of plasma oestradiol-17b, total oestrogenand progesterone to oestrous behaviour in mithun cowsWe carried out studies to (1) establish the characteristicsof oestrous behaviour in mithun cows and (2) determinethe relationships between this behaviour and theplasma concentrations of oestradiol-17b (E2), totaloestrogen and progesterone. Among the behaviouralsigns of oestrus, the cow to be mounted by bull (100%)was the best indicator of oestrus followed by standing tobe mounted by other females (92% Mondal et al.2006b). Rectal examination of the reproductive tractrevealed a relaxed cervix, turgid uterus and ovarieshaving palpable follicles in all animals (Mondal et al.2006g). The length of oestrous cycle and duration ofoestrus were recorded to be 21.8 ± 0.69 days and12.6 ± 1.34 h, respectively. E2 and total oestrogenprofiles during the peri-oestrous period (Mondal et al.2006f) showed that the mean highest peak concentrationsof E2 (27.29 ± 0.79 pg ⁄ ml) and total oestrogen(45.69 ± 2.32 pg ⁄ ml) occurred at 3.90 ± 2.27 and3.89 ± 2.26 h prior to the onset of oestrus, respectively(Fig. 2a). Plasma progesterone concentration was basal(0.14 ± 0.001 ng ⁄ ml) during the peri-oestrous period.Plasma E2 and total oestrogen were found to increasefrom 6 days before oestrus to reach a peak level on theday of oestrus and decline thereafter to basal level onday 3 of the cycle (Fig. 2b). The plasma progesteroneconcentration was the lowest on the day of oestrusshowing gradual increase to register a peak level on day15 of the cycle.Plasma prolactin and oxytocin profiles during cyclicity inmithun and yak cowsIn yaks the plasma prolactin concentration althoughhigh at oestrus, its level fluctuated throughout the cycle,without any definite trend (Sarkar 2004).In mithuns, the plasma prolactin concentration washigh at oestrus and showed fluctuation thereafter withno definite trend throughout the cycle (Mondal et al.2007a). Results on prolactin profile suggest an involvementof prolactin in the ovulation process in this species.The mean plasma oxytocin concentration duringdifferent days of the oestrous cycle was found to besignificantly different (p < 0.001; Mondal et al. 2006d).The plasma oxytocin concentration was low at oestrus.In mithun cows, two peaks of oxytocin were recorded ondays 6 and 18 of the oestrous cycle.Timing of ovulation in relation to onset of oestrus and LHpeak in mithun cowsMondal et al. (2006g) found that the preovulatory LHsurges in mithun cows consisted of several pulses(2.92 ± 0.26 pulses ⁄ animal). The peak level of LH forindividual mithun varied from 6.99 ± 0.44 to 12.69 ±2.10 ng ⁄ ml. The concentration of LH during surge was10.83 ± 0.76 ng ⁄ ml. The duration of LH surge was6.98 ± 0.22 h. LH surge started 1.23 ± 0.18 h afteronset of oestrus. Ovulation occurred at 26.92 ± 0.31Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reproduction Augmentation in Yak and Mithun 219(a)Estradiol-17b and totalestrogen (pg/ml)403020100E2TE(b)4.5 P4 TE4.03.53.02.52.01.51.00.50.0Progesterone (ng/ml)E2403020100Estradiol-17b, total estrogen(pg/ml)–24–18–12 –60 6 11 17 23 29Periestrous period (h)–6 –4 –2 0 2 4 6 8 10 12 14 16 18 20Days of estrous cycleFig. 2. Mean (±SEM) plasma oestradiol-17b (E2) and total oestrogen (TE) profiles of mithun cows during peri-oestrous period. Blood sampleswere collected hourly intervals for 24 h prior to onset of oestrus, at 15-min intervals for 9 h post-onset of oestrus and thereafter at 2 h intervals till31 h post-onset of oestrus (a). Mean plasma E2, TE and progesterone (P4) profiles during the different days of oestrous cycle in mithun cows(n = 12). Blood samples were collected daily for the entire cycle (b)LH (ng/ml)1086420LHProgesteroneOvulation0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40Hours after the onset of estrus1.000.750.500.250.00Fig. 3. Changes in the plasma LH and progesterone profile (mean ±SEM) in mithun cows (n = 12) after onset of oestrus. Blood sampleswere collected at 15 min intervals after the initial expression of heatsymptoms by the mithuns for 9 h period and thereafter at an intervalof 2 h till 4 h post-ovulation. Ovulation was confirmed by rectalpalpation at 2 h intervalsafter the onset of oestrus and 18.63 ± 0.35 h after theend of LH surge (Fig. 3).Seasonal breeding pattern in yaks and associatedendocrine changesYaks are considered seasonal breeders. However, informationabout the breeding season is rather conflicting.The onset and the end of the breeding season areaffected by ecological factors such as climate, grassgrowth, latitude and altitude. Following their longperiod of deprivation and weight loss over the winter,the female yaks come into the breeding season whentemperature and humidity start to rise, and grass beginsto grow which also improves their body condition. InWest Kameng and Tawang districts of ArunachalPradesh in India, the breeding season reaches its peakin July and August when temperature is at its highestand grass growth is at its best and lasts up to November.Sarkar et al. (2006b) estimated plasma progesterone,total oestrogen and oestradiol-17b profile during breeding(July to November) and non-breeding season(February to March) in yak. Plasma progesterone wasvery low (£0.2 ng ⁄ ml) at oestrus, thereafter started toProgesterone (ng/ml)rise with a sharp increase during the late luteal phaseand reached a peak on day 15–16 of the cycle, decliningrapidly thereafter to basal levels at oestrus.During non-breeding season, in 50% (4 ⁄ 8) of theanimals studied, plasma progesterone level was at basallevel as anticipated. However, cyclic changes in plasmaprogesterone were seen in three yaks while the plasmaprogesterone level stayed high throughout the samplingperiod in one animal. There were clear indications ofcyclic luteal activity in a large proportion of animalseven during the non-breeding season, although oestrussymptoms were not exhibited (Sarkar et al. 2006a). Inthe same study, plasma total oestrogen and oestradiol-17b concentrations were high at oestrus and thendeclined to basal level on day 2 of the cycle and anothersmall elevation was found between days 8 and 12 of thecycle (Sarkar et al. 2006a). This elevation of thehormones suggests the possibility of additional follicularwaves in yaks.Circadian rhythmicity in circulatory melatonin concentrationswas exhibited with low concentrations duringdaytime and high during night in both the periodsunder investigation. Similar observations have beenrecorded in cattle (Berthelot et al. 1990), buffalo(Borghese et al. 1994), sheep (Kennaway et al. 1977)and goats (Kloren and Norton 1995). A circadianrhythmicity of prolactin release was also seen with amaximum mean concentration at 04:00 h and a minimumat 14:00 h in breeding season and with a maximumvalue at 00:00 h and a minimum at 12:00 h in nonbreedingmonth. The observations on diurnal prolactinpattern in yaks seems to be in sharp contrast to thoserecorded in cattle and buffaloes where higher prolactinconcentrations have been recorded during daytime(Singh and Madan 1993; Gustafson 1994). Diurnalrhythmicity in hormonal or biochemical constituents isinfluenced by many factors including metabolic rate,nutrition and environmental conditions. Positive relationshipof prolactin secretion to stress related situationsis also well known (Raud et al. 1971; Tucker 1971).Under extremely cold conditions at night, yaks areexposed to greater stress. The condition is reversedduring the daytime, under the influence of sunlight. Thismay be the reason for higher prolactin concentrationÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

220 BS Prakash, M Sarkar and M Mondalduring night times. Sarkar and Prakash (2005c) alsoobserved that the overall plasma prolactin levels duringday as well as night time were found to be more in nonbreedingmonth than the corresponding time in breedingmonth. This marked increase of prolactin level in nonbreedingseason may be stress related and associatedwith harsh environmental condition particularly lowtemperature as well as nutritional stress during the nonbreedingseason. The relationship between prolactin andmelatonin secretion in blood plasma during breeding(November) and non-breeding (February) months wasinvestigated in yaks by Sarkar and Prakash (2005c). Thepositively correlated (0.8 in breeding season and 0.7 innon-breeding season) circadian rhythms of plasmamelatonin and prolactin is not clearly understood.Tindal (1974) reported that melatonin profoundlymodulates prolactin release from the adenohypophysis.However, different species interpret the pineal gland’shormonal signal in a fundamentally different manner.This preliminary investigation opens the gates forresearchers to unravel the interrelationship of thetwo hormones for more comprehensive and detailedinvestigation.Timing of ovulation and plasma LH in yakYaks ovulated at 30.5 ± 0.8 h (28–34 h) after theonset of spontaneous oestrus. Preovulatory LH surgeoccurred 3.0 ± 0.7 h after the commencement ofoestrus signs with the surge lasting 7.3 ± 0.6 h. Ovulationtook place 20.3 ± 1.0 h after LH surge (18–26 h)(Sarkar and Prakash 2005a).Hormone concentrations during pregnancy and periparturientperiod in yakProgesterone profiles were similar in pregnant and nonpregnantyaks up to 14 days after oestrus; however, inpregnant yaks concentrations were significantly higheron day 19 and tended to increase gradually thereafter.Plasma progesterone concentrations decreased rapidlyon day 120, then increased to reach a maximum level onday 210. Concentrations decreased again 20 days beforeparturition reaching basal levels at parturition. Oestradiol-17blevels in plasma and milk increased graduallyuntil day 23 after conception, decreased abruptly on day60, and increased again to reach a maximum level atparturition. Oestradiol concentrations decreased againafter parturition to reach similar levels as measuredduring mating (Yu et al. 1993).Profiles of progesterone, cortisol and 13,14-dihydro-15-keto-PGF 2a (PGFM) in blood plasma of yaks duringperi-parturient period were determined (Sarkar et al.2007b,c). Plasma PGFM level showed an increasingtrend beginning day 4 prior to parturition (0.73 ± 0.16ng ⁄ ml) and increased steeply thereafter to reach a peaklevel (17.16 ± 1.31 ng ⁄ ml) on the day of parturition.The levels, then, declined sharply on day 1 post-partumto reach 1.20 ± 0.40 ng ⁄ ml and thereafter declinedgradually over the days to reach 0.28 ± 0.09 ng ⁄ ml onday 20 post-partum and this level was maintained withfluctuation within narrow limits thereafter till conclusionof the blood sampling on day 90 post-calving.Plasma cortisol level showed an increasing trend beginningday 2 prior to parturition (2.36 ± 0.65 ng ⁄ ml) andincreased steeply thereafter to reach a peak level(26.65 ± 5.28 ng ⁄ ml) on the day of parturition. Thelevels, then, declined gradually over the days andreached 2.36 ± 0.25 ng ⁄ ml on day 3 post-partum, andthis level was maintained with fluctuation within narrowlimits thereafter till conclusion of the blood sampling onday 7 post-calving. The plasma progesterone concentrationon day 7 before parturition was high(2.10 ± 0.10 ng ⁄ ml). The hormone concentrationdecreased gradually followed by an abrupt fall on theday of parturition (0.24 ± 0.04 ng ⁄ ml).Endocrine Biotechniques to Enhance FertilitySilent or non-detected oestrus is one of the majorcontributory factors in both yak and mithun reproductionleading to poor reproductive efficiency in theseanimals (Sarkar and Prakash 2005a; Mondal et al.2006g).Synchronization of oestrus using PGF 2a in mithunsWe tried synchronization of oestrus using doubleinjection of PGF 2a at 11 days apart in cyclic mithuncows. Plasma LH characteristics and time of ovulationin respect to LH peak and onset of oestrus in mithuncows synchronized for oestrus with PGF 2a are presentedin Table 1. All the cyclic animals that were subjected tooestrus synchronization using PGF 2a responded to thetreatment. The intensity of oestrus in the mithun cowssynchronized with double injection of PGF 2a wascomparable to that of spontaneous oestrus (MondalM, Rajkhowa C, and Prakash BS., unpublished data).Synchronization of oestrus using CIDR in mithunsExperiments were conducted to synchronize oestrususing CIDR in cyclic and post-partum mithun cows. Inboth categories of animal, synchronized oestrus usingCIDR showed more prominent behavioural signs thanspontaneous heat (unpublished data). More interestingly,application of CIDR on day 45–50 after parturitioninduced first post-partum oestrus immediatelyafter uterine involution (day 53–58 post-parturition).Unlike other bovines, mithun cows exhibit firstpost-partum oestrus at approximately day 102 ± 19.6post-partum.Table 1. Plasma LH characteristics and timing of ovulation inmithuns subjected to two injection of PGF 2a at 11 days apart foroestrus synchronizationParameters Mean ± SEM RangeHighest LH peak concentration (ng ⁄ ml) 12.54 ± 1.37 7.34–22.65Duration of LH surge (h) 14.76 ± 1.47 10–19Time from:Second PGF 2a injection to onset of oestrus 43.52 ± 5.93 36–58Onset of oestrus to onset of LH surge (h) 2.45 ± 0.43 1.5–3.25Onset of oestrus to ovulation (h) 33.95 ± 1.41 27–39After end of LH surge to ovulation (h) 20.45 ± 0.89 17–24After end of LH surge to ovulation (h) 20.45 ± 0.89 17 to 24Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reproduction Augmentation in Yak and Mithun 221Application of Ovsynch protocol in yaks and mithunsThe Ovsynch protocol (Pursley et al. 1995) was developedas a breeding strategy to eliminate the need foroestrus detection. This method of oestrus synchronizationwas based on control of both ovulation andfollicular growth (Pursley et al. 1995). The protocol iscomposed of an injection of GnRH at random stages ofoestrous cycle to synchronize a new follicle waveemergence. Seven days later PGF 2a is given to regressthe original and the newly formed corpora lutea (CL),followed by a second GnRH injection 48 h later toinduce a synchronous ovulation 24–32 h later thatallows for fixed time artificial insemination 12–16 hafter the second GnRH injection.We tested the efficacy of ovulation synchronization inyaks (Sarkar and Prakash 2005b) and mithuns (unpublisheddata) using the Ovsynch protocol. Ovulation wasdetected by rectal palpation at 2 h intervals from theinitial signs of oestrus till ovulation. Ovsynch protocolfor synchronization of ovulation produced a response of87.5% and 100% in yak and mithun, respectively.Ovulation time and LH characteristics followingovsynch protocol in mithun is presented in Table 2.The LH peak values among individual yaks rangedfrom 10.2 to 40 ng ⁄ ml (n = 8) with a mean of22.8 ± 5.1 ng ⁄ ml. LH concentration (mean ± SEM)increased steeply from 0.98 ± 0.66 ng ⁄ ml at the time ofsecond GnRH injection to peak value of16.04 ± 4.30 ng ⁄ ml 2 h post-GnRH administrationdropping sharply thereafter to basal levels of£0.31 ng ⁄ ml. The duration of LH peak (mean ±SEM) was 4.69 ± 0.36 h with a range of 2.75 to 5.75 h.In seven yaks, the circulatory levels of plasmaprogesterone were basal (£0.2 ng ⁄ ml) at the time ofsecond GnRH administration, and remained basal tillovulation. It was concluded that the Ovsynch protocolcould be applied successfully for fixed time AI in yakand mithun.Application of Heatsynch protocol in yaksHeatsynch is a newly developed synchronization protocolthat uses the less expensive hormone oestradiolcypionate (ECP) in place of the second GnRH injectionin Ovsynch protocol (Lopes et al. 2000).We tested the efficacy of induction of oestrus,synchronization of ovulation and timed artificial inseminationin anoestrous yaks using the Heatsynch protocol(Sarkar et al. 2007e). All the animals responded toHeatsynch treatment for induction of oestrus andsynchronization of ovulation. The high degree ofTable 2. Plasma LH characteristics and timing of ovulation inmithuns (n = 23) subjected to Ovsynch protocolParameters Mean ± SEM RangeHighest LH peak concentration (ng ⁄ ml) 12.23 ± 0.66 9.03–17.22Duration of LH surge (h) 8.25 ± 1.38 6–12Time from:Onset of LH surge after second1.90 ± 0.44 1.25–2.75GnRH injection (h)Ovulation after second GnRH injection (h) 26.75 ± 2.02 19–33Ovulation after end of LH surge (h) 18.62 ± 1.69 15–27ovulation synchronization could be attributed to thehighly synchronized LH peaks in the treated animals. Inanoestrous yaks treated with the Heatsynch protocoland subjected to TAI 40% pregnancy rates wererecorded. Heatsynch protocol could therefore be successfullyutilized for improving fertility in anoestrusyaks too.Superovulation and embryo transferSuperovulatory treatments are widely used in embryotransfer programs to increase the supply of embryosfrom animals of superior genetic merit. In a trial byZagdsuren et al. (1997) three female yaks were superovulatedusing progesterone and FSH. Oestrus wasdetected by a teaser male and the females were inseminated24–36 h after the last FSH injection. Numbers ofCL and ovarian follicles detected by rectal palpationand laparoscopy averaged 5.0 ± 0.6 and 1.3 ± 0.9respectively, but embryo recovery was not reported.Davaa et al. (2000) used FSH and PMSG to inducesuperovulation in yak cows. Oestrus was detected34.1 ± 0.52 h after the prostaglandin treatment. Theaverage number of ovarian follicles was 5.4 ± 0.65, ofwhich 4.5 ± 0.43 ovulated.Sarkar et al. (2006a) used four yaks for superovulationfollowing Ovsynch program. On day 20 after thefirst GnRH injection, females received a total dose of200 mg Folltropin, twice daily, equal doses, over 4 daysperiod in association with two injections of prostaglandinanalogues on 48 and 60 h after initiation ofsuperovulation. The females were mated and the numberof CL was recorded on day 7 after the last injectionof FSH. Only one animal was flushed non-surgically forembryo recovery 7 days after mating. The averagenumber of palpable CL was 2 ± 0.71. A total of threeembryos were recovered on non-surgical flushing from asingle animal. One embryo was transferred to a recipientyak cow, which produced one female yak calf after258 days. This is the first yak calf reported throughembryo transfer technology. In another trial cyclic yaks(n = 10) were synchronized using Ovsynch protocol.On day 10 after expected oestrus, animals received atotal of 200 mg Folltropin, twice daily over 4 days.From 9 yaks, 27 ovulations were detected and 16embryos were recovered (Sarkar et al. 2007a). Plasmaprogesterone profiles from individual yaks suggestedthat a poor superovulatory response in terms of embryorecovery in some animals was caused by the lysis of CLbefore flushing which was carried out 7 days aftersuperovulatory ooestrus. It was suggested that flushing5 days post-superovulatory oestrus could improve thesuperovulatory response in this species.ConclusionsWe have presented here our major research efforts onsome aspects of reproduction in mithuns and yaks. Ourinitial efforts on developing hormone assay procedureshave helped us to successfully extend recent endocrinetechniques for fertility improvement in these species.The results of these studies hold promise for undertakingthese techniques on a larger scale.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

222 BS Prakash, M Sarkar and M MondalHowever, extensive research work is also required onelucidation of hormone receptor interplay at the targetsite to understand the precise molecular role of hormonesin regulation of reproductive cyclicity in thesespecies. An understanding of hormonal interplay atreceptor level is essential for alleviating reproductiveproblems of endocrine origin which in turn will furtherimprove the efficiency of reproduction.ReferencesBerthelot XM, Lautentie JP, Ravault J, Ferney J, Toutain PL,1990: Circadian profile and production rate of melatonin inthe cow. Domest Anim Endocrinol 7, 315–319.Borghese A, Barile VL, Tergaro GM, Pilla AM, ParmegganiA, 1994: Melatonin trend during seasons in heifers andbuffalo cows. Proceedings of the 4th Buffalo Congress, SaoPaulo, Brazil, 27–30 June, Vol. 3, pp. 528–530.Breier BH, Gluckman PD, 1991: The regulation of postnatalgrowth: nutritional influences on endocrine pathways andfunction of the somatotrophic axis. Lives Prod Sci 27, 77–94.Davaa M, Altankhuag N, Zagdsuren Y, 2000: The preliminaryexperiment to induce superovulation in female yaks. Int YakNews Lett 5, 78.El-Nouty F, 1971: The effect of different systems before andafter weaning on age at puberty and age at first conceptionin buffalo heifers. MSc Thesis, Faculty of Agriculture, AinShams University, Egypt.Gelato MC, Merriam GR, 1986: Growth hormone releasinghormone. Ann Rev Physiol 48, 569–591.Gustafson GM, 1994: Effect of changes in light on hormonalsecretion and milk production of dairy cows in earlylactation. Acta Agric Scand 44, 160–168.Hoffman PC, Funk DA, 1992: Applied dynamics of dairyreplacement growth and management. J Dairy Sci 75, 2504–2515.Hull KL, Harvey S, 2001: Growth hormone: roles in femalereproduction. J Endocrinol 168, 1–23.Kennaway DJ, Frith RG, Phillipou G, Seamark RW, 1977:Patterns of progesterone, melatonin, and prolactin secretionin ewes maintained in four different photoperiods. J Endocrinol97, 229–242.Kloren WRL, Norton BW, 1995: Melatonin and fleece growthin Australian cashmere goats. Small Rumin Res 17, 179–185.Lopes FL, Arnold DR, Williams J, Pancarci SM, Thatcher MJ,Drost M, Thatcher WW, 2000: Use of estradiol cypionate fortimed insemination. J Dairy Sci 83, 216 (Abstr).Ministry of Agriculture Department of Animal Husbandry,Dairying & Fisheries, 2003: Seventeenth Livestock Census,Government of India.Mondal M, Prakash BS, 2004: Effects of long-term growthhormone-releasing factor (GRF) administration on patternsof GH and LH secretion in growing female buffaloes(Bubalus bubalis). Reproduction 127, 45–55.Mondal M, Dhali A, Rajkhowa C, Prakash BS, 2004:Secretion patterns of growth hormone in growing captivemithuns (Bos frontalis). Zool Sci 21, 1125–1129.Mondal M, Prakash B, Rajkhowa C, Prakash BS, 2005a:Development and validation of a simple, sensitive, secondantibody format enzyme immunoassay (EIA) for LHdetermination in mithun (Bos frontalis) plasma. J ImmunImmunochem 26, 157–167.Mondal M, Rajkhowa C, Prakash BS, 2005b: Standardizationand validation of a simple, sensitive, second antibody formatenzyme immunoassay (EIA) for GH determination inmithun (Bos frontalis) plasma. Zoo Biol 24, 483–493.Mondal M, Rajkhowa C, Prakash BS, 2005c: Secretionpatterns of luteinizing hormone in growing captive mithuns(Bos frontalis). Reprod Biol 5, 227–235.Mondal M, Rajkhowa C, Prakash BS, 2005d: Twenty-fourhourrhythmicity of growth hormone in captive adultmithuns (Bos frontalis). Biol Rhythm Res 36, 255–262.Mondal M, Rajkhowa C, Prakash BS, 2005e: Twenty-fourhoursecretion patterns of luteinizing hormone in captiveprepubertal female mithuns (Bos frontalis). Gen ComEndocrinol 144, 197–203.Mondal M, Meyer HHD, Rajkhowa C, Prakash BS, 2006a:Highly sensitive second antibody format enzymeimmunoassayfor determination of 13,14-dihydro-15-keto-PGF 2a inblood plasma of mithun (Bos frontalis). Prost Other LipidMed 80, 100–109.Mondal M, Rajkhowa C, Prakash BS, 2006b: Behavioralestrous signs can predict the time of ovulation in mithun(Bos frontalis). Theriogenology 66, 1391–1396.Mondal M, Rajkhowa C, Prakash BS, 2006c: Determinationof effective dosage of GH-releasing factor for blood GHresponses in Mithun (Bos frontalis). J Anim Physiol AnimNutr 90, 453–458.Mondal M, Rajkhowa C, Prakash BS, 2006d: Development ofa biotin–streptavidin amplified enzyme immunoassay foroxytocin and its application during milk ejection andreproductive cycle in the mithun (Bos frontalis). Zool Sci23, 633–639.Mondal M, Rajkhowa C, Prakash BS, 2006e: Exogenous GHreleasinghormone increases GH and LH secretion ingrowing mithuns (Bos frontalis). Gen Com Endocrinol149, 197–204.Mondal M, Rajkhowa C, Prakash BS, 2006f: Relationship ofplasma estradiol-17b, total estrogen and progesterone toestrous behavior in mithun (Bos frontalis) cows. HormBehav 49, 626–633.Mondal M, Rajkhowa C, Prakash BS, 2006g: Oestrousbehavior and timing of ovulation in relation to onset ofoestrus and LH peak in mithun (Bos frontalis) cows. ReprodDomest Anim 41, 479–484.Mondal M, Rajkhowa C, Prakash BS, 2007a: Development andvalidation of a highly sensitive economic enzymeimmunoassayfor prolactin determination in blood plasma of mithun(Bos frontalis) and its application during milk let down andcyclicity. Anim Reprod Sci 99, 182–195.Mondal M, Rajkhowa C, Prakash BS, 2007b: Highly sensitivesecond antibody format enzymeimmunoassays for determinationof estradiol-17b and total estrogen in blood plasmaof mithun (Bos frontalis). Zool Sci 24, 408–413.Mondal M, Rajkhowa C, Prakash BS, 2007c: Plasma growthhormone concentration in mithun (Bos frontalis) of differentages: relation to age and body weight. J Anim Physiol AnimNutr 91, 68–74.Pursley JR, Mee MO, Wiltbank MC, 1995: Synchronization ofovulation in dairy cows using PGF 2a and GnRH. Theriogenology44, 915–923.Raud HR, Kiddy CA, Odell WD, 1971: The effect of stressupon the determination of serum prolactin by radioimmunoassay.Proc Soc Expt Biol Med 136, 689–704.Sarkar M, 2004: Endocrine changes during cyclicity and timingof ovulation during spontaneous and synchronised estrus inyaks. PhD Thesis, Submitted to National Dairy ResearchInstitute, Karnal, Haryana, India.Sarkar M, Prakash BS, 2005a: Timing of ovulation in relationto onset of estrus and LH peak in yak (Poephagus grunniensL.). Anim Reprod Sci 86, 353–362.Sarkar M, Prakash BS, 2005b: Synchronisation of ovulation inyaks (Poephagus grunniens L.) using PGF 2a and GnRH.Theriogenology 63, 2494–2503.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reproduction Augmentation in Yak and Mithun 223Sarkar M, Prakash BS, 2005c: Circadian variations in plasmalevels of melatonin and prolactin during breeding and nonbreedingseasons in yak (Poephagus grunniens L.). AnimReprod Sci 90, 149–162.Sarkar M, Prakash BS, 2006: Application of sensitiveenzymeimmunoassays for oxytocin and prolactin determinationin blood plasma of yaks (Poephagus grunniens L.)during milk let down and cyclicity. Theriogernology 65,499–516.Sarkar M, Borah S, Chakraborty P, Deka BC, Sharma BC,Duttaborah B, Borah S, Baruah K, Ramesha KP, PourochottamneR, Kataktakware M, Sarvanan BC, SenguptaDH, Das S, Prakash BS, Bhattacharya M, 2006a: Calvesborn from anestrus yak (Poephagus grunniens L.) usingovsynch and superovulation treatment. Zool Sci 23, 721–725.Sarkar M, Meyer HHD, Prakash BS, 2006b: Is the yak(Poephagus grunniens L.) really a seasonal breeder? Theriogenology65, 721–730.Sarkar M, Chakraborty P, Sharma BC, Deka BC, DuttaborahBK, Mohanty TK, Prakash BS, 2007a: Assessment ofsuperovulatory responses in terms of palpable corpora luteaand embryo recovery using plasma progesterone in yaks(Poephagus grunniens L.). Res Vet Sci, in press.Sarkar M, Das BC, Duttaborah BK, Kumar V, Mohan K,Bhattacharya M, Prakash BS, 2007b: Application of sensitiveenzymeimmunoassay for determination of cortisol inblood plasma of yaks (Poephagus grunniens L.). Gen CompEndocrinol 154, 85–90.Sarkar M, Duttaborah BK, Kumar V, Mohan K, Prakash BS,2007c: Plasma concentrations of 13,14-dihydro-15-keto-PGF 2a and progesterone during periparturient period inyaks (Poephagus grunniens L.). Zoo Sci (communicated).Sarkar M, Nandankar UA, Duttaborah BK, Bhattacharya M,Prakash BS, 2008. Plasma growth hormone concentrationsin female yak (Poephagus grunniens L.) of different ages:relations with age and body weight. Lives Sci 115, 313–318.Sarkar M, Sengupta DH, Duttaborah B, Rajkhoa J, Bora S,Bandopadhyay S, Ghosh M, Ahmed FA, Saikia P, MohonK, Prakash BS, 2007e: Efficacy of heatsynch protocol forinduction of estrus, synchronization of ovulation and timedartificial insemination in yaks (Poephagus grunniens L.).Anim Reprod Sci 104, 306–312.Short RE, Bellows RA, 1971: Relationships between weightgains, age at puberty, and reproductive performance inheifers. J Anim Sci 32, 127–131.Singh J, Madan ML, 1993: RIA of prolactin as related tocircadian changes in buffaloes. Buffalo J 9, 159–164.Tindal JS, 1974: Hypothalamic control of secretion and releaseof prolactin. J Reprod Fertil 39, 437–461.Tucker HA, 1971: Hormonal response to milking. J Anim Sci32, 137–146.Wiener G, Jianlin H, Ruijun L, 2003: The Yak, 2nd edn. RAPpublication 2003 ⁄ 06, FAO, Bangkok, Thailand.Yu SJ, Huang TM, Chen BX, 1993: Reproductive patterns ofthe yak. I. Reproductive phenomena of the female yak. BrVet J 149, 579–583.Zagdsuren Y, Davaa M, Magash A, 1997: Superovulation infemale yaks. In: Proceedings of the 2nd InternationalCongress on Yak, Xining, China, pp. 193–194.Author’s address (for correspondence): BS Prakash, Dairy CattlePhysiology Division, NDRI, Karnal-13200, Haryana, India. E-mail:bsp1001@rediffmail.comConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 224–231 (2008); doi: 10.1111/j.1439-0531.2008.01166.xISSN 0936-6768Follicle Development in MaresFX Donadeu 1 and HG Pedersen 21 Roslin Institute, Royal (Dick) School of Veterinary Studies, University of Edinburgh, Edinburgh, UK; 2 Veterinary Reproduction and Obstetrics,Department of Large Animal Sciences, Faculty of Life Sciences, University of Copenhagen, Copenhagen, DenmarkContentsThe mare provides a unique experimental model for studyingfollicle development in monovular species. Development ofantral follicles in horses is characterized by the periodic growthof follicular waves which often involve the selection of a singledominant follicle. If properly stimulated, the dominant folliclewill complete development and eventually ovulate a fertileoocyte. Regulation of follicular wave emergence and follicleselection involves an interplay between circulating gonadotropinsand follicular factors that ensures that individual folliclesare properly stimulated to grow (or to regress) at any givenstage of follicular wave development. Periodic development offollicular waves continuously occurs during most of post-natallife in the mare and is influenced by factors such as stage ofoestrous cycle, season, pregnancy, age, breed and individual sothat different types of follicular waves (minor or major,ovulatory or anovulatory) and different levels of activitywithin waves may develop under different physiologicalconditions. Changes in gonadotropin levels and ⁄ or in thesensitivity of follicles to circulating gonadotropins seemto account largely for these physiological variations in follicledevelopment.IntroductionAlthough scientific interest in equine follicles existed bythe 1920s (Seaborne 1925), detailed studies on equinefollicle dynamics did not begin until 50 years later underthe pioneering lead of OJ Ginther at the University ofWisconsin (reviewed in Ginther 1979). During the early1980s, ultrasonography became available for equinetheriogenology (Palmer and Driancourt 1980). Transrectalultrasonography combined with systemic hormonemeasurements were subsequently used extensivelyto understand equine follicle dynamics. The more recentintroduction (mid-1990s) of the technique of ultrasoundguidedtransvaginal ovarian puncture provided scientistsunprecedented access to the live equine ovary andallowed the targeting of individual follicles for collectionof follicle samples, injection of test substances orexperimental manipulation of follicles (Gastal et al.1997; Gerard and Monget 1998; Donadeu and Ginther2001; Martoriati et al. 2003). The experimental use ofthese procedures has led to extraordinary advances inthe knowledge of equine ovarian physiology during thelast 15 years (reviewed in Beg and Ginther 2006;Donadeu and Watson 2007). Two important outcomesof this knowledge have been an improvement of reproductiveefficiency in mares (Squires 2006) and a greaterunderstanding of follicular physiology in monovularspecies in general. In regard to the latter, considerationof the mare as a useful model for studying ovarianprocesses during human health and disease has recentlyincreased (Ginther et al. 2004c; Carnevale 2008).This review summarizes current knowledge on equinefollicle development. Information on pre-antral andearly antral follicles is particularly scarce in the horse;therefore, relatively more focus has been placed on thelate stages of antral follicle development up to the preovulatorystage. Details on physiological and practicalaspects of the ovulatory process and subsequent formationof the corpus luteum can be found elsewhere(Ginther 1992; Boerboom and Sirois 2001; Squires 2006).Preantral Stages of Follicle DevelopmentDuring early fetal life, primordial germ cells migrate tothe primitive gonadal ridge where they proliferate andsubsequently enter meiotic division to become arrestedin prophase I as primary oocytes (Ginther 1992).Meiotic arrest continues until atresia or until theprimary oocyte is stimulated by an LH surge duringpost-natal life. Based on histological studies of thedeveloping fetal horse ovary, meiotic divisions begin atabout day 70 of pregnancy (Deanesly 1975). Widespreadatresia characterizes the first oocytes to enter the meioticphase between days 73 and 150 of pregnancy with apeak at approximately day 100. Development of oocytesinto primordial follicles is not apparent until day 150(Deanesly 1975) and the number of primordial folliclescontinues to increase during subsequent fetal life so thatseveral thousand of these follicles are contained withinthe ovaries at birth (Ginther 1992). Proliferation of thesomatic cell layer surrounding the primary oocyte leadsto the sequential development of primordial folliclesinto primary, secondary and, finally, antral follicles,a process that begins in the fetus and continuesthroughout post-natal life. It is not known how folliclesare selected to grow from the pool of primordial, restingfollicles in the mare. Adult equine ovaries were reportedto contain approximately 36 000 primordial follicles and100 growing follicles at any one time, with largevariability in actual follicle numbers between individualmares (Driancourt et al. 1982). Comparatively higherfollicle numbers (threefold) were found in the adultbovine ovary. Proliferative activity of small pre-antralfollicles in pony mares was higher at the beginning of theovulatory season than during anoestrus or at the end ofthe ovulatory season, whereas follicular atresia wasunaffected by season (Driancourt et al. 1983).Antral Follicle Development and FollicularWavesThe equine follicle develops an antrum when it reachesapproximately 0.3 mm in diameter (Driancourt et al.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Follicles and Mares 2251982). Similar to pre-antral stages, little is known on thedevelopment of antral follicles before they reachapproximately 2 mm, the smallest diameter that canusually be detected by transrectal ultrasonography. Ithas been reported that the growth of an equine folliclefrom 0.1 to 1 mm takes approximately two oestrouscycles (Driancourt 1979) and that atresia duringthis early phase of antral development is rare (Ginther1992).As in other farm species and humans, the developmentof antral follicles in the horse is characterized bythe periodic growth of cohorts of follicles or follicularwaves (Sirois et al. 1989; Bergfelt and Ginther 1993).Characterization of follicular waves has involved theultrasonic day-to-day identification of individual follicles(Bergfelt and Ginther 1993; Gastal et al. 1997) andthe use of a statistical method that avoids the need tomaintain the identities of individual follicles duringserial ultrasound examinations (Ginther and Bergfelt1992; Donadeu and Ginther 2002b). Follicular wavesin the horse can be identified in relation to follicles2 mm in diameter and larger; however, it is not knownwhether earlier antral and pre-antral stages are alsocharacterized by wave-like patterns of growth,a question that has also not been clarified in otherspecies (Mizunuma et al. 1999; McGee and Hsueh2000). Follicular waves and their regulation with anemphasis on normal oestrous cycles will be described inthis and the next section. Follicular wave patternscharacteristic of other reproductive states will bedescribed in a separate section.Characteristics of a follicular wave. Follicular waveemergence has been normally defined for experimentalpurposes as occurring when the largest follicle reaches 6or 13 mm, depending on the study (Ginther et al.2003a). Identification of wave emergence often requiresday-to-day ultrasonic evaluation of the ovary usuallyafter aspiration of all follicles from previous waves.A follicular wave initially involves the simultaneousgrowth of a variable number of follicles at a commonrate of between 2 and 3 mm ⁄ day. A recent studyinvolving ablation of all follicles during the middle ofan oestrous cycle in pony mares reported a mean ofapproximately 12 follicles emerging during the commongrowth phase of the ablation-induced wave (Gastalet al. 2004). Approximately two follicles emerged eachday during the first 4 days after wave emergence with aprogressive decrease in the numbers of follicles emergingthereafter. The exact number of follicles emerging withinwaves is affected by factors such as season (Donadeuand Ginther 2003).The phase of common follicle growth is followed bythe selection of a single follicle (occasionally twofollicles) which is manifested as a deviation in diameterbetween the two largest follicles of the wave beginningwhen the largest follicle reaches approximately 22 mm(Ginther et al. 2003a). Deviation begins a mean of7 days before ovulation and is characterized by thecontinuous growth (at an unchanged rate) of the largest(selected) follicle as a dominant follicle and thesimultaneous cease in growth and subsequent regressionof smaller, subordinate follicles. Once it grows toapproximately 35–45 mm, the dominant follicle normallyeither ovulates or ceases to grow and begins toregress, depending on whether an ovulatory LH surgeoccurs. The establishment of dominance is mediated bya differential increase in trophic support to the largestfollicle of a wave by mechanisms that will be explainedin more detail in the next section. This leads to profoundchanges in follicular cell function that are necessary forthe eventual full maturation of the follicle to anovulatory-competent state and that are reflected indramatic changes in global gene expression, changesthat have begun to be characterized in other species,most notably cattle (Sisco et al. 2003; Fayad et al. 2004;Mihm et al. 2008).Dominance does not seem to be a pre-determinedtrait among the follicles growing in a wave becausefollicle ablation studies have demonstrated that allfollicles have similar capacity to become dominant,a capacity that in subordinate follicles is lost withinapproximately 48 h after the beginning of deviation(Gastal et al. 2004). The same studies showed that inapproximately 61% of waves the first follicle toemerge at 6 mm maintains its size advantage oversmaller follicles through the common growth phaseand becomes dominant. The likelihood of the largestfollicle becoming dominant increases as it approachesthe expected diameter at the beginning of deviation(Gastal et al. 2004). In a few instances, yet, the largestfollicle ceases or slows down its growth during thecommon growth phase and is replaced by the secondlargest follicle (or sometimes even a smaller follicle)which then becomes the dominant follicle.Types of follicular waves. Follicular waves have beenclassified as major waves (referred to in the literatureand throughout this review simply as follicular waves)or minor waves, depending on whether they involve thedevelopment of a readily identifiable dominant follicleor they produce only smaller follicles, respectively(Ginther 1993; Donadeu and Ginther 2002b). Thenumber of follicular waves during an oestrous cyclevaries between species. In the horse, as in humans, onlyone or two major follicular waves develop during eachcycle (Ginther et al. 2004c). A major wave (namedprimary wave) always emerges during the middle of theequine oestrous cycle and produces the ovulatoryfollicle. Approximately 25% of interovulatory intervalsinvolve an additional major wave (secondary wave) thatdevelops during the first half of the cycle (Sirois et al.1989; Bergfelt and Ginther 1993). The incidence ofsecondary waves is significantly higher in some breedssuch as Thoroughbreds, and some of these waves mayproduce ovulations. The ability to ovulate in thepresence of high progesterone levels during dioestrusseems to be unique to the horse (Ginther 1992). Minorwaves (largest follicle usually

226 FX Donadeu and HG PedersenRegulation of Follicular Wave DevelopmentEmergence of a follicular wave and subsequent selectionof a dominant follicle are under finely-tuned regulationby systemic mechanisms involving changes in gonadotropinlevels and local mechanisms that involve changesin follicular factor levels.Systemic regulation of follicular wave development.Unlike pre-antral follicles, antral follicles cannotdevelop without adequate gonadotropin stimulation andthis has been experimentally shown in the mare(Pedersen et al. 2002; Imboden et al. 2006).Follicular waves in the mare, as in other species, aretemporally preceded by a stimulatory surge in circulatingFSH (Ginther et al. 2003a). The acute dependence offollicular waves on FSH has been shown in mares by theinability of follicles to grow beyond a diameter of 15 mmafter suppression of circulating FSH by systemic injectionof follicular fluid (Bergfelt and Ginther 1985). Closefunctional relationships between FSH surges and follicularwaves involve not only positive effects of FSH onfollicles but also negative effects of follicles on FSH. Thewave-associated FSH surge reaches a peak when thelargest follicle is approximately 13 mm in diameter(Gastal et al. 1997; Donadeu and Ginther 2001). Thefollowing decline in FSH results from an increase incirculating inhibin, presumably inhibin-A (Watson andAl-zi’abi 2002), contributed mainly by the three largestfollicles of the wave as they grow above 13 mm(Donadeu and Ginther 2001). The declining FSHconcentrations continue to support growth of thefollicles of the wave until the largest (future dominant)follicle reaches the expected diameter at the beginning ofdeviation (approximately 22 mm). At that time, circulatingFSH levels become too low to maintain thegrowth of the follicles of the wave. The low FSH, yet,does not restrict the growth of the future dominantfollicle which by that time has acquired the ability tomore efficiently use circulating gonadotropins forgrowth (Ginther et al. 2003a). The differential responsesof the largest and smaller follicles of a wave togonadotropins result in the initiation of diameterdeviation. Continuous suppression of FSH throughoutdeviation ensures the morphological and functionaldemise of subordinate follicles and is attributable toproduction of high levels of inhibin and oestradiol bythe dominant follicle (Gastal et al. 1999; Donadeu andGinther 2001). The critical role of low FSH levels in thedeviation mechanism in mares is illustrated by thedisruption of the deviation mechanism after administrationof FSH (Squires 2006) or immunization againstinhibin (McCue et al. 1992) leading to the developmentof multiple ovulatory follicles.While FSH is particularly important for folliculargrowth before deviation, LH becomes more criticalduring deviation. This has been demonstrated by studiesshowing that experimental suppression of LH in cyclingmares leads to the regression of the dominant follicleearly during its development (Bergfelt et al. 2001), and isconsistent with an increase in circulating LH before thebeginning of deviation (Bergfelt et al. 2001). The studyof LH–follicle interrelationships during the anovulatoryseason has revealed a direct temporal relationshipbetween an increase in circulating LH, but not FSH,and the development of major waves, further indicatingthat an increase in LH plays a major role in stimulatingthe development of dominant follicles (reviewed inDonadeu and Watson 2007). An additional conclusionwas that circulating LH concentrations above thoserequired for growth of the dominant follicle are requiredfor the development of ovulatory competence, i.e. forthe dominant follicle to become fully responsive to anLH surge (Donadeu and Watson 2007).Studies in cattle have shown that LH receptorexpression differentially increases in granulosa and thecacells of the early dominant follicle (Bao and Garverick1998; Beg and Ginther 2006). These findings areconsistent with those in other species (Richards 2001).Although not critically examined in relation to thebeginning of deviation, higher LH receptor levels havebeen reported in dominant-size follicles than in smallerfollicles in the horse (Fay and Douglas 1987; Goudetet al. 1999). Taken together, the findings on LH levelsand LH receptor expression during a follicular wave inmares are consistent with the conclusion that deviationinvolves a critical increase in the dependence of thedominant follicle on LH.Limited data exist on the involvement of substancesother than gonadotropins in the endocrine regulation ofantral follicles in mares. Based on observed direct effectson follicle growth or on the presence of specific receptorsin equine ovaries, positive roles on follicle growth havebeen suggested for circulating levels of substancesincluding growth hormone (Cochran et al. 1999), dopamine(King et al. 2005) and prolactin (Thompson et al.1997).Although follicular insulin-like growth factor-1(IGF-1) in mares largely derives from the systemiccirculation, its bioactivity is regulated mostly throughlocal mechanisms (Watson et al. 2004) and the role ofIGF-1 is therefore considered in the next section.Local regulation of follicular wave development. Localregulation of follicle development in monovular specieshas been studied extensively in relation to follicleselection (reviewed in Fortune et al. 2004; Beg andGinther 2006; Knight and Glister 2006). A variety ofprotein and steroid factors including members of theIGF family, oestradiol, inhibins and activins, follistatinand vascular endothelial growth factor (VEGF) areinvolved. In general, these factors act, often in aparacrine manner, to either enhance or diminish thetrophic effects of gonadotropins on follicular cellsthrough a variety of mechanisms. Differential changesin the levels of specific factors between follicles thusensure the continuous development of the dominantfollicle and the regression of subordinate follicles duringdeviation.A differential increase in the levels of oestradiol, IGF-1,activin-A and inhibin-A in the future dominant folliclewas associated with the beginning of deviation in mares,whereas differentially elevated levels of progesteroneand inhibin-B occurred later during the development ofthe dominant follicle (Donadeu and Ginther 2002a).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Follicles and Mares 227Studies involving intrafollicular factor injection in mareshave provided insight into the complex interrelationshipsbetween these factors during follicle deviation(Ginther et al. 2005). It has been concluded thatalthough all these factors are likely involved in thedevelopment of the dominant follicle after the beginningof deviation, only IGF-1 is involved in the initiation ofdeviation by, among other actions, regulating the levelsof other growth factors in the dominant follicle (Beg andGinther 2006). This is consistent with the particularlyimportant role of the IGF-1 system in follicle selectionin other species (Mazerbourg et al. 2003). Results froma series of in vivo experiments have convincinglyconfirmed the critical role of the IGF system in follicleselection in mares. Injection of IGF-1 into the secondlargest or a smaller follicle at the beginning of deviationchanged its fate from subordinate to co-dominantresulting in the development of multiple ovulatoryfollicles (Ginther et al. 2004a; b; Gastal et al. 2007).Conversely, injection of IGF binding protein (IGFBP)-3into the largest follicle at the beginning of deviationresulted in the follicle regressing and being replaced bythe second largest follicle which became dominant(Ginther et al. 2004a). At least four types of IGFBPs(IGFBP-2, 3, 4 and 5), which negatively regulate IGFactivity, have been identified in equine follicles and theconcentrations of three of these (IGFBP-2, 4 and 5) arecorrelated negatively with those of IGF-1 during follicledevelopment (Gerard and Monget 1998; Bridges et al.2002). Consistent with the stimulatory role of IGF-1 inthe development of ovulatory follicles, reduced levels ofbioactive IGF-1 are thought to be involved in thedevelopmental deficiencies of dominant follicles duringthe spring transition that prevent them from acquiringovulatory competence (Acosta et al. 2004; Watson et al.2004).An additional factor that has begun to be explored inrelation to follicle selection in horses is VEGF. Thisangiogenic factor has been shown to be necessary forfollicle development in other species (Fraser and Wulff2001). VEGF levels differentially increase in the dominantfollicle in horses (Ginther et al. 2004b), and thisincrease is thought to be mediated, at least partly, byIGF-1. VEGF is likely involved in the reported increasein vascularization of the future dominant follicle beforethe beginning of deviation (Acosta et al. 2004) whichpresumably increases the availability of circulatinggonadotropins to the follicle. The reduced levels offollicular VEGF and low vascularity of the wall ofdominant follicles during the spring transition relative tothe ovulatory season (Watson and Al-zi’abi 2002)underscore the critical role of VEGF in the developmentof the ovulatory follicle in horses.Effects of Different Physiological Conditions onFollicle Development in MaresEffects of season. The effects of season on follicularactivity in mares have been studied in considerabledetail. Important variations in follicular activity occurnot only between the ovulatory and anovulatory seasonsbut also between different periods within each season.Studies in pony mares using follicle ablation to facilitatethe identification of individual follicular waves revealedthat, as during the ovulatory season, follicular wavesperiodically occur during the anovulatory season despitethe reduced levels of follicle development (Donadeu andGinther 2002b, 2003; Ginther et al. 2003b). Only minorwaves (largest follicle 12 mm within waves (means of 3.2 and 11.5follicles in waves developing during deep anoestrous andthe early spring transition, respectively). Althoughdominant follicles during transition may not grow tothe diameters typical of ovulatory follicles, in the samestudy transitional waves produced more follicles thanwaves developing during the ovulatory season (means of11.5 and 6.0 follicles >12 mm, respectively) attesting tothe high levels of follicular activity even in the absenceof ovulation in transitional mares (Donadeu and Ginther2003).Based on consistent temporal relationships betweenfollicles and circulating hormones, it has been concludedthat the differences in follicle development betweendifferent periods of the anovulatory season as well as thedeficient development of dominant follicles during thespring transition relative to the ovulatoy season are notattributable to deficient circulating FSH levels butrather to changes in LH and, possibly, differences infollicular sensitivity to gonadotropins (reviewed inDonadeu and Watson 2007).Season-related effects on follicular activity have alsobeen reported between the two halves of the ovulatoryseason, with higher levels of activity during the first halfof the season due to higher incidence of both secondarywaves and minor waves associated with higher gonadotropinlevels (Ginther 1992, 1993).Effects of pregnancy and parturition. Considerableresearch work is needed to better characterize folliculardynamics and associated regulatory mechanisms duringand following pregnancy in the mare. Based on combineddata from ultrasound and rectal palpation studies,follicular dynamics during the first half of pregnancyare similar to those occurring during the first half ofthe anovulatory season, with an initial period of variableactivity (between days 11 and 40 of pregnancy) characterizedby the periodic development of major waves or,more commonly, development of sporadic major wavesor only minor waves (Ginther and Bergfelt 1992),followed by a pronounced decrease in follicular activityin all mares between days 50 and 140 of pregnancy sothat the diameter of the largest follicle does not exceed15 mm by day 140 (Squires et al. 1974). A decrease infollicular activity has also been reported after the firstone-third of pregnancy in cattle (Ginther et al. 1996).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

228 FX Donadeu and HG PedersenBased on reported hormone–follicle associations, deficientFSH levels do not seem to be responsible for thereduction in follicular growth in some mares betweendays 11 and 40 of pregnancy (Ginther and Bergfelt1992), an effect that may instead be accounted for byreduced levels of LH (due to persistent progesteronenegative feedback), similar to the effects of the seasonalreduction in circulating LH on follicle growth during thefall transitional period (Ginther et al. 2003b).The mechanisms responsible for the dramatic reductionin follicular growth during mid and late pregnancyin mares have not been clarified but likely involves atemporally associated decrease in circulating FSH(reviewed in Ginther 1992). This is different from theanovulatory season during which changes in FSH levelsdo not seem to play a major role in the decrease infollicle growth during deep anoestrus (Donadeu andWatson 2007). Reduced follicle numbers during midpregnancyare also likely attributable to ovulation orluteinization of follicles into accessory corpora luteaunder the influence of chorionic gonadotropin (Ginther1992). Further complexity into the regulation of folliclegrowth during pregnancy is provided by the observationthat the effects of season on hormones and folliclescontinue to occur during pregnancy (Ginther 1992).The natural pressure to produce a foal each year in aspecies with an 11-month-long pregnancy is reflected inan early post-partum ovulation in the mare (foal heat,typically within 2 weeks of parturition). A steadyincrease in the diameter of the largest follicle andin the numbers of follicles after parturition resultingin ovulation 14 days later was recently shown inArabian mares (Gunduz et al. 2008). The post-partumincrease in follicular growth is induced by an increase ingonadotropin secretion at the time of parturition (Hineset al. 1987; Ginther et al. 1994). Studies have shown thatthe follicular response to parturition may vary inindividual horses or different types of horses, and maynot readily occur in primiparous mares (Nagy et al.1998), in the presence of a nursing foal (Ginther 1992) orduring the winter, when the negative effects of seasonmay prevail over the positive effects of parturition(Ginther et al. 1994).Follicle development before puberty. Follicle developmentin spring-born pre-puberal pony fillies has recentlybeen studied in detail (Nogueira and Ginther 2004),adding to limited information on follicle profiles fromearlier studies (reviewed in Ginther 1992). Follicularactivity during 2–10 months of life was characterized bya progressive increase in mean follicle diameter (fromapproximately 6 to 10 mm) and mean follicle numbers(from 3 to 17 follicles) between 2 and 5 months of age,a short plateau in activity coinciding with the wintermonths and a re-initiation of follicle growth after7 months of age leading to the onset of the firstovulatory season in spring (Nogueira and Ginther2004). Changes in follicle activity during the first yearof life were positively correlated with changes incirculating gonadotropins, consistent with a regulatoryrole of season on gonadotropin and follicular activitybeginning early during life in mares. Follicular growthbefore puberty was characterized by the development of(minor) follicular waves. Remarkably, these waves werenot temporally associated with statistically significantcirculating FSH surges, an observation that warrantsfurther investigation.Effects of aging on follicular activity. As highlighted in arecent review, many age-related changes in follicularactivity in the horse resemble those occurring in humans(Carnevale 2008). Follicular activity during oestrouscycles begins to decrease in mares 20 years of age orolder eventually leading to a cease in ovarian activity(Carnevale et al. 1993, 1994). Interovulatory intervalsfirst become longer in these mares due to longerfollicular phases associated with a primary follicularwave that emerges later and contains less follicles. Inaddition, the ovulatory LH surge is less pronounced inthe older mares, and there is a higher incidence ofultrasonically atypical ovulations characterized by acentral hypoechogenic area at the ovulatory site. Thereduction in follicular activity and frequency of ovulationin mares ‡20 years old is associated with an overallelevation in concentrations of FSH and LH during thefollicular phase, a phenomenon that also occurs duringthe peri-menopausal period in women. This is eventuallyassociated with persistent ovarian inactivity with follicles

Follicles and Mares 229dominant follicles and the simultaneous regression ofsubordinate follicles during follicle deviation.Follicular waves occur throughout post-natal life inthe horse until follicles irreversibly cease to grow atapproximately ‡ 20 years of age. The levels of follicularactivity are affected by factors such as stage of theoestrous cycle, season, pregnancy, age, breed andindividual. This results in variations in follicular wavepatterns that most notably include the production orabsence of a dominant follicle (major or minor waves)and the development of ovulatory-competent or ovulatory-incompetentdominant follicles. Available informationindicates that these physiological variations infollicular wave patterns are driven by changes incirculating levels of FSH and ⁄ or LH and by localchanges in responsiveness of follicles to gonadotropins.Yet, considerable work is needed to better understandthe mechanisms by which follicle development is regulatedduring different physiological conditions outsidethe oestrous cycle in mares.ReferencesAcosta TJ, Beg MA, Ginther OJ, 2004: Aberrant blood flowarea and plasma gonadotropin concentrations during thedevelopment of dominant-sized transitional anovulatoryfollicles in mares. Biol Reprod 71, 637–642.Bao B, Garverick HA, 1998: Expression of steroidogenicenzyme and gonadotropin receptor genes in bovine folliclesduring ovarian follicular waves: a review. J Anim Sci 76,1903–1921.Beg MA, Ginther OJ, 2006: Follicle selection in cattle andhorses: role of intrafollicular factors. Reproduction 132,365–377.Bergfelt DR, Ginther OJ, 1985: Delayed follicular developmentand ovulation following inhibition of FSH with equinefollicular fluid in the mare. Theriogenology 24, 99–108.Bergfelt DR, Ginther OJ, 1993: Relationships between FSHsurges and follicular waves during the estrous cycle in mares.Theriogenology 39, 781–796.Bergfelt DR, Gastal EL, Ginther OJ, 2001: Response ofestradiol and inhibin to experimentally reduced luteinizinghormone during follicle deviation in mares. Biol Reprod 65,426–432.Boerboom D, Sirois J, 2001: Equine P450 cholesterol side-chaincleavage and 3 beta-hydroxysteroid dehydrogenase ⁄ delta(5)-delta(4) isomerase: molecular cloning and regulation of theirmessenger ribonucleic acids in equine follicles during theovulatory process. Biol Reprod 64, 206–215.Bridges TS, Davidson TR, Chamberlain CS, Geisert RD,Spicer LJ, 2002: Changes in follicular fluid steroids, insulinlikegrowth factors (IGF) and IGF-binding protein concentration,and proteolytic activity during equine folliculardevelopment. J Anim Sci 80, 179–190.Carnevale EM, 2008: The mare model for follicular maturationand reproductive aging in the woman. Theriogenology69, 23–30.Carnevale EM, Bergfelt DR, Ginther OJ, 1993: Aging effectson follicular activity and concentrations of FSH, LH, andprogesterone in mares. Anim Reprod Sci 31, 287–299.Carnevale EM, Bergfelt DR, Ginther OJ, 1994: Follicularactivity and concentrations of FSH and LH associated withsenescence in mares. Anim Reprod Sci 35, 231–246.Carnevale EM, Hermenet MJ, Ginther OJ, 1997: Age andpasture effects on vernal transition in mares. Theriogenology47, 1009–1018.Cochran RA, Leonardi-Cattolica AA, Sullivan MR, KincaidLA, Leise BS, Thompson DL Jr, Godke RA, 1999: Theeffects of equine somatotropin (eST) on follicular developmentand circulating plasma hormone profiles in cyclicmares treated during different stages of the estrous cycle.Domest Anim Endocrinol 16, 57–67.Deanesly R, 1975: Germ cell development and the meioticprophase in the fetal horse ovary. J Reprod Fertil Suppl 23,547–552.Donadeu FX, Ginther OJ, 2001: Effect of number anddiameter of follicles on plasma concentrations of inhibinand FSH in mares. Reproduction 121, 897–903.Donadeu FX, Ginther OJ, 2002a: Changes in Concentrationsof follicular fluid factors during follicle selection in mares.Biol Reprod 66, 1111–1118.Donadeu FX, Ginther OJ, 2002b: Follicular waves andcirculating concentrations of gonadotrophins, inhibin andoestradiol during the anovulatory season in mares. Reproduction124, 875–885.Donadeu FX, Ginther OJ, 2003: Interactions of follicularfactors and season in the regulation of circulating concentrationsof gonadotrophins in mares. Reproduction 125,743–750.Donadeu FX, Watson ED, 2007: Seasonal changes in ovarianactivity: lessons learnt from the horse. Anim Reprod Sci 100,225–242.Driancourt MA, 1979: Follicular kinetics in the mare ovary.Ann Biol Anim Biochim Biophys 19, 1443–1453.Driancourt MA, Paris A, Roux C, Mariana JC, Palmer E,1982: Ovarian follicular populations in pony and saddletypemares. Reprod Nutr Dev 22, 1035–1047.Driancourt MA, Prunier A, Palmer E, Mariana J, 1983:Seasonal effects on ovarian follicular development in ponymares. Reprod Nutr Dev 23, 207–215.Fay JE, Douglas RH, 1987: Changes in thecal and granulosacell LH and FSH receptor content associated with follicularfluid and peripheral plasma gonadotrophin and steroidhormone concentrations in preovulatory follicles of mares. JReprod Fertil Suppl 35, 169–181.Fayad T, Levesque V, Sirois J, Silversides DW, Lussier JG, 2004:Gene expression profiling of differentially expressed genes ingranulosa cells of bovine dominant follicles using suppressionsubtractive hybridization. Biol Reprod 70, 523–533.Fortune JE, Rivera GM, Yang MY, 2004: Follicular development:the role of the follicular microenvironment in selectionof the dominant follicle. Anim Reprod Sci 82–83, 109–126.Fraser HM, Wulff C, 2001: Angiogenesis in the primate ovary.Reprod Fertil Dev 13, 557–566.Gastal EL, Gastal MO, Bergfelt DR, Ginther OJ, 1997: Roleof diameter differences among follicles in selection of afuture dominant follicle in mares. Biol Reprod 57, 1320–1327.Gastal EL, Donadeu FX, Gastal MO, Ginther OJ, 1999:Echotextural changes in the follicular wall during follicledeviation in mares. Theriogenology 52, 803–814.Gastal EL, Gastal MO, Beg MA, Ginther OJ, 2004: Interrelationshipsamong follicles during the common-growthphase of a follicular wave and capacity of individual folliclesfor dominance in mares. Reproduction 128, 417–422.Gastal EL, Gastal MO, Donadeu FX, Acosta TJ, Beg MA,Ginther OJ, 2007: Temporal relationships among LH,estradiol, and follicle vascularization preceding the firstcompared with later ovulations during the year in mares.Anim Reprod Sci 102, 314–321.Gerard N, Monget P, 1998: Intrafollicular insulin-like growthfactor-binding protein levels in equine ovarian folliclesduring preovulatory maturation and regression. BiolReprod 58, 1508–1514.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

230 FX Donadeu and HG PedersenGinther OJ, 1979: Reproductive Biology of the Mare: Basicand Applied Aspects, 1st edn. Equiservices Cross Plains,WI, USA.Ginther OJ, 1990: Folliculogenesis during the transitionalperiod and early ovulatory season in mares. J Reprod Fertil90, 311–320.Ginther OJ, 1992: Reproductive Biology of the Mare: Basicand Applied Aspects, 2nd edn. Equiservices PublishingCross Plains, WI, USA.Ginther OJ, 1993: Major and minor follicular waves during theequine estrous cycle. J Eq Vet Sci 13, 18–25.Ginther OJ, Bergfelt DR, 1992: Associations between FSHconcentrations and major and minor follicular waves inpregnant mares. Theriogenology 38, 807–821.Ginther OJ, Bergfelt DR, 1993: Growth of small follicles andconcentrations of FSH during the equine oestrous cycle. JReprod Fert 99, 105–111.Ginther OJ, Baucus KL, Bergfelt DR, 1994: Follicular andFSH responses to parturition during the anovulatory seasonin mares. Theriogenology 41, 613–627.Ginther OJ, Kot K, Kulick LJ, Martin S, Wiltbank MC, 1996:Relationships between FSH and ovarian follicular wavesduring the last six months of pregnancy in cattle. J ReprodFert 108, 271–279.Ginther OJ, Beg MA, Donadeu FX, Bergfelt DR, 2003a:Mechanism of follicle deviation in monovular farm species.Anim Reprod Sci 78, 239–257.Ginther OJ, Woods BG, Meira C, Beg MA, Bergfelt DR,2003b: Hormonal mechanism of follicle deviation as indicatedby major versus minor follicular waves during thetransition into the anovulatory season in mares. Reproduction126, 653–660.Ginther OJ, Bergfelt DR, Beg MA, Meira C, Kot K, 2004a:In vivo effects of an intrafollicular injection of insulin-likegrowth factor 1 on the mechanism of follicle deviation inheifers and mares. Biol Reprod 70, 99–105.Ginther OJ, Gastal EL, Gastal MO, Checura CM, Beg MA,2004b: Dose–response study of intrafollicular injection ofinsulin-like growth factor-I on follicular fluid factorsand follicle dominance in mares. Biol Reprod 70, 1063–1069.Ginther OJ, Gastal EL, Gastal MO, Bergfelt DR, BaerwaldAR, Pierson RA, 2004c: Comparative study of the dynamicsof follicular waves in mares and women. Biol Reprod 71,1195–1201.Ginther OJ, Gastal EL, Gastal MO, Beg MA, 2005: In vivoeffects of pregnancy-associated plasma protein-A, activin-Aand vascular endothelial growth factor on other follicularfluidfactors during follicle deviation in mares. Reproduction129, 489–496.Goudet G, Belin F, Bezard J, Gerard N, 1999: Intrafollicularcontent of luteinizing hormone receptor, {alpha}-inhibin,and aromatase in relation to follicular growth, estrous cyclestage, and oocyte competence for in vitro maturation in themare. Biol Reprod 60, 1120–1127.Gunduz MC, Kasikci G, Ekiz B, 2008: Follicular and steroidhormone changes in Arabian mares in the postpartumperiod. Anim Reprod Sci (in press).Hines KK, Fitzgerald BP, Loy RG, 1987: Effect of pulsatilegonadotrophin release on mean serum LH and FSHin peri-parturient mares. J Reprod Fertil Suppl 35, 635–640.Imboden I, Janett F, Burger D, Crowe MA, Hassig M, Thun R,2006: Influence of immunization against GnRH on reproductivecyclicity and estrous behavior in the mare. Theriogenology66, 1866–1875.King SS, Campbell AG, Dille EA, Roser JF, Murphy LL,Jones KL, 2005: Dopamine receptors in equine ovariantissues. Domest Anim Endocrinol 28, 405–415.Knight PG, Glister C, 2006: TGF-{beta} superfamily membersand ovarian follicle development. Reproduction 132, 191–206.Martoriati A, Duchamp G, Gerard N, 2003: In vivo effect ofepidermal growth factor, interleukin-1{beta}, and interleukin-1RAon equine preovulatory follicles. Biol Reprod 68,1748–1754.Mazerbourg S, Bondy CA, Zhou J, Monget P, 2003: Theinsulin-like growth factor system: a key determinant role inthe growth and selection of ovarian follicles? A comparativespecies study. Reprod Domest Anim 38, 247–258.McCue PM, Carney NJ, Hughes JP, Rivier J, Vale W, LasleyBL, 1992: Ovulation and embryo recovery rates followingimmunization of mares against an inhibin alpha-subunitfragment. Theriogenology 38, 823–831.McGee EA, Hsueh AJW, 2000: Initial and cyclic recruitmentof ovarian follicles. Endocrine Rev 21, 200–214.Mihm M, Baker PJ, Fleming LM, Monteiro AM,O’Shaughnessy PJ, 2008: Differentiation of the bovinedominant follicle from the cohort upregulates mRNAexpression for new tissue development genes. Reproduction135, 253–265.Mizunuma H, Liu X, Andoh K, Abe Y, Kobayashi J, YamadaK, Yokota H, Ibuki Y, Hasegawa Y, 1999: Activin fromsecondary follicles causes small preantral follicles toremain dormant at the resting stage. Endocrinology 140,37–42.Nagy P, Huszenicza G, Juhasz J, Kulcsar M, Solti L, Reiczigel J,Abavary K, 1998: Factors influencing ovarian activity andsexual behavior of postpartum mares under farm conditions.Theriogenology 50, 1109–1119.Nogueira GP, Ginther OJ, 2004: Dynamics of follicle populationsand gonadotropin concentrations in fillies age two toten months. Equine Vet J 32, 482–488.Palmer E, Driancourt MA, 1980: Use of ultrasonic echographyin equine gynecology. Theriogenology 13, 203–216.Pedersen HG, Berrocal B, Thomson SRM, Watson ED, 2002:Gonadotropic control of follicular growth in the mare.Theriogenology 58, 465–468.Richards JS, 2001: Perspective: the ovarian follicle –a perspective in 2001. Endocrinology 142, 2184–2193.Seaborne E, 1925: The oestrous cycle of the mare and someassociated phenomena. Anat Rec 30, 277–286.Sirois J, Ball BA, Fortune JE, 1989: Patterns of growth andregression of ovarian follicles during the oestrous cycle andafter hemiovariectomy in mares. Equine Vet J 8(Suppl.), 43–48.Sisco B, Hagemann LJ, Shelling AN, Pfeffer PL, 2003:Isolation of genes differentially expressed in dominant andsubordinate bovine follicles. Endocrinology 144, 3904–3913.Squires EL, 2006: Superovulation in mares. Veter Clin N Am:Eq Pract 22, 819–830.Squires EL, Garcia MC, Ginther OJ, 1974: Effects ofpregnancy and hysterectomy on the ovaries of pony. JAnim Sci 38, 823–830.Thompson DL Jr, Hoffman R, DePew CL, 1997: Prolactinadministration to seasonally anestrous mares: reproductive,metabolic, and hair-shedding responses. J Anim Sci 75,1092–1099.Vanderwall DKW, GL, 1990: Age-related subfertility in themare. In: Proc 35th Annual Conv American Assoc EquinePractitioners, 85–89.Watson ED, Al-zi’abi MO, 2002: Characterization of morphologyand angiogenesis in follicles of mares during springtransition and the breeding season. Reproduction 124, 227–234.Watson ED, Heald M, Tsigos A, Leask R, Steele M, GroomeNP, Riley SC, 2002: Plasma FSH, inhibin A and inhibinisoforms containing pro- and -alphaC during winter anoes-Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Follicles and Mares 231trus, spring transition and the breeding season in mares.Reproduction 123, 535–542.Watson ED, Bae SE, Thomassen R, Thomson SR, Woad K,Armstrong DG, 2004: Insulin-like growth factors-I and -IIand insulin-like growth factor-binding protein-2 in dominantequine follicles during spring transition and theovulatory season. Reproduction 128, 321–329.Author’s address (for correspondence): FX Donadeu, Roslin Institute,Royal (Dick) School of Veterinary Studies, University of Edinburgh,Edinburgh, UK. E-mail: xavier.donadeu@ed.ac.ukConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 232–237 (2008); doi: 10.1111/j.1439-0531.2008.01167.xISSN 0936-6768Proteins Associated With the Early Intrauterine Equine ConceptusMA Hayes 1 , BA Quinn 1 , ND Keirstead 1 , P Katavolos 1 , RO Waelchli 2 and KJ Betteridge 21 Departments of Pathobiology and 2 Biomedical Sciences, Ontario Veterinary College, University of Guelph, Guelph, ON, CanadaContentsA critical period of early gestation in the mare involves theimmobilization (fixation) of the encapsulated conceptus ataround days 16–17. We compared the major proteins in thenormal equine embryonic capsule and endometrial secretionsaround the period of fixation with those from pregnancies inthe process of termination induced by administration of ananalogue of prostaglandin F 2a (PGF 2a ). Uterocalin andb 2 -microglobulin (b 2 M) associated with the embryonic capsulewere proteolytically converted to smaller forms during thefixation period. These conversions were similar in conceptusesfrom control and treated mares. A 17 kDa cationic proteinidentified as a secretory phospholipase A2 (sPLA2) type IIAwas detected bound to normal capsules but increased substantiallyin response to PGF 2a . Two forms of uteroglobin weredistinguished by partial amino acid sequences of 6 kDabands in flush fluids from normal pregnant uteri. Afteradministration of PGF 2a one immunoreactive form of uteroglobinwas preferentially increased. These studies demonstratethat failure of pregnancy in this model is associated with anincrease in secretory phospholipase in the capsule and a changein the forms of uteroglobin in the uterine secretions.IntroductionEmbryonic loss before placentation accounts for asubstantial proportion of pregnancy failures in mares(Ball 1988; Baker et al. 1993; Carnevale et al. 2000;Morris and Allen 2002). In Thoroughbred studs inNewmarket, UK, 16–17% of pregnancies diagnosed byultrasound at about day 15 are subsequently lost, mostbetween days 15 and 35 (Morris and Allen 2002).During the second and third weeks of gestation, thenormal conceptus (embryo proper, the extra-embryonicmembranes and their contained fluid) expands rapidlyand migrates in the uterus until around days 16–17 whenit becomes ‘fixed’ at the site of subsequent placentation(Ginther et al. 1985). A better understanding of themolecular interactions between the conceptus andendometrium might improve methods for diagnosisand treatment of mares that are prone to losing theconceptus during this stage.The embryonic capsule, which is essential for survivalof the embryo (Stout et al. 2005), is composed largely ofacellular glycoprotein produced mainly by the trophoblast(Albihn et al. 2003). It also contains some maternally-secretedproteins, notably uterocalin-p19 which isimplicated in transport of small hydrophobic substancesacross the capsule (Crossett et al. 1998; Stewart et al.2000; Kennedy 2005). Before fixation, the yolk-sac wallcontains large amounts of GM2-activator protein(GM2AP) suggesting that this protein plays a role intransport of glycolipids or phospholipids while thecapsule is intact (Quinn et al. 2006). During the fixationperiod, sialic acid decreases in the capsule (Oriol et al.1993a,b) exposing fewer negatively charged galactoseand N-acetylgalactose residues of the major core type 1O-glycan (Arar et al. 2007). This change might beimportant in changing the permeability or ‘stickiness’ ofthe capsule. The capsule also contains b 2 -microglobulin(b 2 M) which undergoes limited proteolysis during thefixation period; the role of this conversion is unknown(Quinn et al. 2007). While such changes in the capsulemight play some functional role in the process offixation, they might also signify deterioration of thecapsule in preparation for its removal which is consideredto be completed by about day 21 (Betteridge 2007).To learn more about the molecular events associatedwith fixation during successful early pregnancy, we havecompared proteins in the capsule, yolk sac and uterusduring the fixation period in normal gestation (Quinnet al. 2007) with those in pregnancies induced to fail byadministration of a luteolytic dose of a prostaglandinF 2a analogue (Chu et al. 1997; Betteridge et al. 2006). Inthis paper, we further characterized the effects ofluteolysis on p19-uterocalin and a secretory phospholipaseassociated with the capsule, and on the forms ofuteroglobin in the uterine fluids.Materials and MethodsSeventeen mares were used according to a researchprotocol approved by the Animal Care Committee ofthe University of Guelph. They were monitored foroestrus and bred by artificial insemination with freshsemen. Ovarian follicles were monitored daily by transrectalultrasonography so that conceptuses collectedcould be aged correctly to within 0.5 day; the day thatovulation was first detected was designated as day 0.Some mares were treated on day 12 or 14 with aprostaglandin PGF2a analogue (Estrumate) to induceluteolysis (Chu et al. 1997; Betteridge et al. 2006).Between days 15 and 18 post-ovulation conceptuseswere flushed from a mare’s uterus by transcervicaluterine lavage with 1 l volumes of phosphate-bufferedsaline (PBS; pH 7.4) using the method previouslydescribed by Waelchli and Betteridge (1996). Varioussamples (capsule, yolk-sac fluid, yolk-sac wall, uterineflush fluids and endometrial biopsies) were collected forproteomic analysis as previously described (Quinn et al.2007). The endometrial samples were bisected; one halfwas flash-frozen and stored in liquid nitrogen, the otherhalf was stored in RNAlater (Ambion, Austin, TX,USA) for RT-PCR. The capsule was dissected, washedthree times in PBS to release loosely bound proteins, andcut into portions for electrophoresis. The yolk-sac wallwas dissected from the embryonic disc and separatedinto bilaminar and trilaminar portions for analysis. TheÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Proteins in Early Equine Conceptuses 233uterine flush and yolk-sac fluids were centrifuged toremove cells and particulates. Flush fluid aliquots(100 ml) were concentrated to 2 ml by centrifugalultrafiltration at a 5 kDa cutoff (Amicon Ultra, MilliporeCorporation, Bedford, MA, USA) and stored at)70°C until they were analysed. Uterine biopsies andyolk-sac wall tissues were homogenized in 0.5 ml PBScontaining proteinase inhibitors (Complete Mini, RocheDiagnostics, Mannheim, Germany).Proteins of interest were compared by one- and twodimensionalSDS–PAGE and Western blots using methodsdescribed previously (Brooks et al. 2003; Quinn et al.2007). Proteins in gels were identified by MALDI-TOFand MS ⁄ MS amino acid sequencing of in-gel trypsindigests (Brooks et al. 2003) using MS-fit (ProteinProspector,University of California, San Francisco, CA,USA) and Scaffold (Proteome Software, Portland, OR,USA). Direct amino acid sequences determined bytandem mass spectrometry at the Advanced ProteinTechnology Center, Hospital for Sick Children (Toronto,ON, Canada) (Lillie et al. 2006) were compared forsimilarity with sequences of Equus caballus via the NCBInon-redundant protein sequence database. Equalamounts of original material were loaded onto gels andcomparisons made based on band density, and the sizedetermined using pre-stained broad range molecularweight markers [Bio-Rad (Hercules, CA, USA) or NewEngland Biolabs (Ipswich, MA, USA) P7701S]. Differencesin amounts of uteroglobin in uterine flush fluidswere also compared by quantitative 2D-difference-in-gelelectrophoresis (2D-DIGE) using the methods described(Hobson et al. 2007). Gels were scanned with aTyphoon 9410 imager (excitation ⁄ emission filter of488 nm ⁄ 520 nm for Cy2, 532 nm ⁄ 580 nm for Cy3, and633 nm ⁄ 670 nm for Cy5). Statistical and quantitativeanalyses of spot changes on images were completed intriplicate samples from each group using DeCyder 6.5software (GE Healthcare, Uppsala, Sweden).Protein concentrations in samples were determined bythe Biorad protein assay (Bradford 1976). Proteins fromsome gels were transferred to nitrocellulose membranesfor immunoblot analysis, as described previously (Quinnet al. 2007). Specific binding was detected by incubationwith rabbit antiserum against recombinant equinep19 ⁄ uterocalin from which the GST fusion partnerhad been removed (Kennedy 2005), or against asynthetic peptide of uteroglobin ⁄ Clara cell secretoryprotein (Katavolos 2006), or equine secretory phopspholipaseA2 (gi: 126513263). Anti-peptide antibodieswere raised in rabbits immunized against peptidesconjugated to keyhole limpet hemocyanin (KLH)(Pacific Immunology, Ramona, CA, USA). The peptideimmunogens were KEATSSYGFYGC from the predictedsPLA2 from equine endometrium (gi: 126513263)and EPSKPDADMKAATTQLKTLV correspondingto equine secretoglobin 1A1 (gi: 126352306) (Katavolos2006).The amino acid sequence of equine sPLA2 wasdetermined from complementary DNA (cDNA) derivedfrom endometrium from pregnant mares. Tissue expressionof the putative sPLA 2 gene was detected qualitativelyby RT-PCR using individual endometrial biopsiespreviously stored in RNAlater. Total RNA was isolatedusing TRIzol Ò reagent (Ambion), and a phenol-guanidineisothiocyanate protocol (Chomczynski and Sacchi1987). Messenger RNA was copied into cDNA using theThermoscript RT-PCR system according to the manufacturer’srecommended protocol (Invitrogen, Burlington,ON, Canada). Amplification of sPLA 2 wasperformed using Platinum Taq DNA polymerase (Invitrogen)according to the manufacturer’s recommendations,using 0.5 lM of forward primer(GGTCCAGGGGCATTTGCG) and reverse primer(AACTTGGATCGGGCAGGGAG). PCR productswere electrophoresed on a 2% agarose gel and stainedwith ethidium bromide. PCR amplicons were pooledand sequenced to examine the possibility of singlenucleotide polymorphisms (SNPs) within the codingregion, including the presence of heterozygotes. DNAsequencing was performed at the Guelph MolecularSupercentre (Guelph, ON, Canada). Nucleotide sequenceswere compared against sequences at NCBIincluding Equus caballus build 1.1 (Eca1.1) genomebased on the 6.8X WGS assembly of the domestic horsegenome—EquCab1 (Broad Institute, Cambridge, MA,USA).ResultsBefore fixation, the major protein bands visible byreduced 1D SDS–PAGE in extracts of normal embryoniccapsule were uterocalin ⁄ p19 and the intact p10form of b 2 M (Fig. 1). This was consistent with previousstudies of conceptuses from normal gestation, andconfirmed by Western blot, MALDI-TOF and MS ⁄ MSsequence determination of in-gel trypsin digests. Uterocalinbands or spots in the normal capsule werevariable, in part due to degradation to smaller formsD14ND16TD16ND18 D18T Np19p17p8, p10p5, p6Fig. 1. Silver-stained 15% SDS–PAGE reducing gel of proteins elutedfrom equine embryonic capsules before and after conceptus fixation atabout day 16. Lane 1, MW markers; lanes 2, 4 and 6 are from normalpregnancies; lanes 3 and 5 are from mares treated with a luteolytic doseof a prostaglandin F 2a analogue and show increased amounts of p17,identified as secretory phospholipase A2 (sPLA2). Luteolysis had lesseffect on the proteolytic products of b 2 M (p10 to p8) and uterocalin(p19 to p5 and p6)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

234 MA Hayes, BA Quinn, ND Keirstead, P Katavolos, RO Waelchli and KJ Betteridgewith trypsin fragment masses consistent with uterocalin.By day 18, most of the p10 form of b 2 M in the capsulehad been converted to the p8 form (D9-b 2 M) that lacksnine amino acids from the N-terminus (Quinn et al.2007).A band of 17 kDa (p17) was observed around thetime of fixation in normal capsules (Fig. 1), consistentwith previous studies (Quinn et al. 2007). This wasmarkedly increased in capsules from mares that weretreated with PGF 2a (Fig. 1). The p17 band of a day 18conceptus from a treated mare was tentatively identifiedas a secretory phospholipase A2 (sPLA2) based onhomology of two partial sequences (K)LLNYKFSYRand (K)YQYYNNK determined by MS ⁄ MS. Thesesequences were most similar to a rat sPLA2 (gi: 220858),the cDNA sequence of which was used to identify asingle cDNA sequence from an equine articular cartilagecDNA library (gi: 57706753). Primers designed from thissequence were used to amplify a complete cDNAsequence from cDNA prepared by RT-PCR of RNAextracted from endometrium from a normal pregnantmare (NM_001100113). This is 100% identical with acoding sequence in chromosome 2 of the horse genome(chr2:32592684–32595740). The predicted amino acidsequence included the two partial sequences we identifiedby MS ⁄ MS of the p17 band. Also, MALDI-TOFanalysis of trypsin digested p17 matched six peptidemasses of a theoretical digest of the predicted sPLA2protein from our cDNA sequence. The predictedsequence also had homology with an N-terminalsequence (XXLXFXKMIXLMTGKQAT) reported fora 17 kDa progesterone-dependent secreted phospholipasein the equine uterus (Beier-Hellwig et al. 1995).Polyclonal antibodies generated against a syntheticpeptide (KEATSSYGFYGC) of our sequence alsorecognized p17 in immunoblots of capsules from normaland treated mares (not shown).The p17 band was not detected in 2D-PAGE ofcontrol capsules (Fig. 2a,b). However, in capsules fromtreated mares, a single 17 kDa spot was present at pI 9.8on days 16 (Fig. 2c) and 18 (not shown).By 2D-PAGE of pooled days 13–15 control capsuleextracts, a 19 kDa uterocalin spot was observed at pI 9.5(Fig. 2a). The p19 spot was less prominent in control(Fig. 2b) and treated days 16 and 18 capsules (Fig. 2c),whereas a doublet of smaller spots observed at pI 9.5were consistent with uterocalin determined by MS ⁄ MSsequences (Fig. 2b).Before fixation (Fig. 2a), there were three prominent10 kDa bands and one 8 kDa band in the pI rangeof 5.0–6.5 that were identified as forms of b 2 M byMS ⁄ MS peptide sequences of in-gel trypsin digests.After fixation, the 10 kDa forms of b 2 M decreasedwhereas the 8 kDa form became more prominent(Fig. 2b).To evaluate the changes in uterine production ofsPLA2, we examined concentrated uterine flush fluidsamples from control and treated pregnancies. At day 16of gestation, flush samples from control mares containedvariable amounts of a 19 kDa band (Fig. 3a), consistentwith uterocalin by immunoblotting (Quinn et al. 2007)and by MALDI-TOF analysis of in-gel trypsin digests.By comparison, there was a less abundant band atpI 5 7 11p19p17p10p19p17p10p19p17p10Fig. 2. Silver-stained 2D SDS–PAGE reducing gel of proteins elutedfrom equine embryonic capsules. (a) Controls pooled from days 13.5 to15.5; (b) control from day 16; (c) treated from day 166 kDa (Fig. 3a), that migrated to the position ofimmunoreactive uteroglobin also found in previousstudies of uterine flush samples (Quinn et al. 2007).Uteroglobins migrate faster than expected for theircalculated molecular masses (Mu¨ller-Scho¨ttle et al.2002; Mukherjee et al. 2007). The p6 uteroglobin bandswere moderately increased in intensity in the silverstained gels of uterine flush fluids from mares that weretreated with the PGF 2a analogue compared with normalmares (Fig. 3a). However, in Western blots with polyclonalantibodies raised against the EPSKPDADMKAATTQLKTLV, uteroglobin bands gave a markedlyÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Proteins in Early Equine Conceptuses 235pI 3 11p19p17p101 2 3 4 5 6 7 8p6pI 5.0 pI 5.8Fig. 4. Silver-stained 2D SDS–PAGE reducing gel of proteins inuterine flush collected from a control mare on day 16.5. Two bands inthe 6 kDa region were identified by MS ⁄ MS sequencing as forms ofuteroglobin. The major band at pI 4.9–5.0 and the minor band at pI5.8–6.0 both contained sequences of products of two uteroglobin genes(see Table 1)Fig. 3. Silver-stain (a) and immunoblot with antibody to a syntheticuteroglobin peptide (b) of 15% SDS-PAGE reducing gel of proteins inuterine flush samples collected on day 16.5 from control (lanes 1–4)and PGF 2a -treated (lanes 5–8) pregnant maresstronger signal in treated mares than in normal mares(Fig. 3b). This suggested that the form of uteroglobinthat increased after PGF 2a was different from theuteroglobin band in the normal pregnant uterus. By2D PAGE, the p6 bands migrated as a major form at pI4.9–5.0 and a minor form at pI 5.8–6.0 (Fig. 4), both ofwhich were identified as uteroglobin by MS ⁄ MSsequencing of in-gel trypsin digests (Table 1). Bothspots were composed of two forms, based on amino acidsequences predicted from two uteroglobin horse genomicsequences (GenBank 124072669). Two peptides ofmasses 932.547 (KTLVDFLPKN) and 1292.622(AVEPFKPDADMKA) were consistent with horsep6uteroglobin reference sequence NP_001075327 andsecretoglobin ⁄ CCSP of equine lung origin (AAW83215,AAW83216, AAW83215 and AAW83218). By comparison,two other sequenced peptides of masses 1362.712(KLMDKIAKSPLCA) and 1391.705 (KLMDKIVESPLCA) present in the major pI 4.9–5.0 spot indicatedthat two secretoglobins in the whole genome shotgunsequence of the horse (gi 124072669) were expressed.The increase in immunoreactive uteroglobin after luteolysiswas attributed to a selective increase in thesecretoglobin 1A1 form (NP_001075327) that includesthe peptide with tyrosine rather than alanine at position59 used to generate the antibody (Katavolos 2006).The minor spot (pI 5.8–6.0) was increased threefold(p = 0.025) at day 18 in the treated mares in comparisonwith control mares (N = 3 per group), as quantifiedby 2D-DIGE.DiscussionThese studies were conducted to identify potential proteinbiomarkers that might distinguish successful and unsuccessfulfixation of the equine conceptus, or that might helpto explain the molecular basis of these processes. For thisTable 1. Amino acid sequences of two forms of uteroglobin in uterine flush fluid from a day 15 normal pregnancySpot A,pISpot B,pIMeasuredmassPeptide sequence determinedby MALDI-TOF and MS ⁄ MSSimilarity with knownuteroglobin ⁄ CCSP sequences4.9–5.0 5.8–6.0 gi:20385506,gi:126352306gi:58978622, gi:58978613,gi:58978651, gi:58978641+ + 932.551 (K)TLVDFLPK(N) Yes Yes+ 1275.740 (K)TLVDFLPKNTK(D) Yes No+ + 1292.625 (A)VEPFKPDADMK(A) Yes Yes+ 1362.712 (K)LMDKIAKSPLCA(–) Yes No+ + 1391.705 (K)LMDKIVESPLCA(–) No Yes+ 1832.060 (K)TLVDFLPKNTKDSILK(L) Yes NoÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

236 MA Hayes, BA Quinn, ND Keirstead, P Katavolos, RO Waelchli and KJ Betteridgecomparison, we used an experimental model of PGF 2a -induced luteolysis that leads to failure of fixation anddelayed loss of sialic acid from the embryonic capsule(Chu et al. 1997; Betteridge et al. 2006). These studiesfollow on from a previous demonstration that b 2 Massociated with the normal capsule is truncated to an8 kDa form (D9-b 2 M) around the time of fixation (Quinnet al. 2007). The present studies show that b 2 M exists asmultiple 10 kDa forms and one 8 kDa form that can beseparated by 2D-PAGE, and that luteolysis has only aminor influence on proteolysis or binding of b 2 M associatedwith the capsule.The most obvious luteolysis-associated alteration incapsule proteins was an increase in a 17 kDa secretoryphospholipase A2 (sPLA2) classified as type IIA on thebasis of amino acid and cDNA sequence homology of thecDNA sequence we obtained from pregnant endometrium.Previous reports have noted the presence of apartially characterized PLA2 in the normal pregnantuterus (Beier-Hellwig et al. 1995; Quinn et al. 2007) butits function there is unknown. PLA2s are a complexgroups of enzymes that cleave glycerophospholipids atthe sn-2 position and release a free fatty acid, most oftenarachidonic acid which is the cyclooxygenase substrateprecursor of various eicosanoids. PLA2s can thereby havediverse signalling roles that are especially important ininflammation and haemostasis, and in the regulation ofovarian function, pregnancy and parturition (Tithof et al.2007). Secretory PLA2-IIA has a high pI and binds toanionic phospholipid interface rather than phospholipidsin intact cell membranes (Beers et al. 2003; Birts et al.2007). Human sPLA2-IIA also has an ability to bind toheparin-sulphate proteoglycans (Birts et al. 2007). Whilethe functions of sPLA2-IIA are still unclear, the evidencesupports the view that they relate to its role as an innateimmune protein involved in the catabolism of cell debris(bacteria, apoptotic cells) in extracellular locations (Birtset al. 2007). Equine uterine sPLA2 characterized in ourstudies also has a high pI of 9.8, binds to the embryoniccapsule and increases when fixation has been blocked byadministration of PGF 2a . It is therefore plausible thatsPLA2 contributes to the imminent removal and degradationof the capsule or the conceptus.Uterocalin is a well characterized lipocalin secreted byendometrial glands in the luteal stage of the ovariancycle and in the early stages of normal pregnancy inequids (Stewart et al. 2000; Suire et al. 2001; Kennedy2005). Uterocalin is also a highly cationic protein(pI 9.4 similar to that of sPLA2) (Fig. 2) and thisproperty likely explains why it also binds to theembryonic capsule. Recent evidence suggests that uterocalinis involved in the transport of small lipophilic,mainly nutrient, substances across the glycan capsuleand into the yolk sac (Suire et al. 2001; Kennedy 2005;Quinn et al. 2007). Amounts in the uterine flush werelower after treatment with PGF 2a , consistent withprogesterone dependence (Suire et al. 2001). Theamounts present in the capsule were variable and bydays 16 and 18, some had been proteolytically convertedto smaller fragments of approximately 10 kDa (Fig. 2b).It appears likely that this alteration is part of the processof degradation of the capsule rather than a processrelevant to its function.Observations in the present study indicate that thereare at least two uteroglobin genes expressed in theequine uterus during early pregnancy. The form thatwas more consistent immunoreactively with equineuteroglobin (Mu¨ller-Scho¨ttle et al. 2002) was increasedsubstantially after luteolysis. Uteroglobin was firstidentified as a low molecular weight secreted uterineprotein in rabbits and is the founding member of thelarge secretoglobin superfamily of proteins that alsoincludes Clara cell secretory proteins (CCSP) (Beier2000; Mu¨ller-Scho¨ttle et al. 2002; Mukherjee et al.2007). These proteins are expressed in various tissues,mainly in the lung and uterus. Uteroglobin (secretoglobin1A1) is expressed in various mammals including thehorse (Mu¨ller-Scho¨ttle et al. 2002) and there are threeclosely similar genes in the horse genome, located onchromosome 12. Previous studies showed that theuteroglobin that is detected in immunoblots with anantibody to a peptide sequence of uteroglobin ⁄ secretoglobin1A1 is more abundant in the non-pregnant uterusthan during the fixation period of normal pregnancy(Quinn et al. 2007). Uteroglobin has a small lipophilicbinding pocket (von der Decken et al. 2005) and canbind prostaglandins such as PGF 2a (Mukherjee et al.2007) but the significance of these properties is unclear.Secretoglobins can bind various small lipophilic moleculesincluding retinoids and polychlorinated biphenyls(von der Decken et al. 2005; Mukherjee et al. 2007).CCSP has anti-inflammatory functions in the respiratorytract, where it might bind lipid mediators andsPLA2 (Mukherjee et al. 2007). Further studies arerequired to determine whether there is a relationshipbetween the increases in sPLA2 and uteroglobin inresponse to luteolysis.These studies have employed analytical methods witha level of sensitivity most suitable for demonstration ofchanges in the most abundant proteins. It is expectedthat there are also alterations in many cytokines andenzymes that are not apparent in these approaches.Moreover, it is still unclear whether the modifications inproteins and glycans that can be observed in the capsuleduring and after the period of fixation are functionallyimportant in interactions between the endometrium andconceptus. Some of these changes might be part of thenormal removal of the capsule. However, the increase insPLA2-type IIA in the capsule and a change in theexpression of uteroglobin genes might have a role in thepending demise of the conceptus in some circumstances.Alternatively, these changes might reflect the return tothe non-pregnant post-luteolysis condition.AcknowledgementsThis research was supported by the Natural Sciences and EngineeringResearch Council of Canada (NSERC); the Grayson Jockey ClubResearch Foundation Inc.; Equine Guelph; The E.P. Taylor researchFoundation; and the Ontario Ministry of Agriculture, Food and RuralAffairs (OMAFRA). Natalie Keirstead was supported by a Fellowshipfrom the Canadian Institutes of Health Research (CIHR). We thankDorothee Bienzle, Jeff Caswell, Gordon Kirby and John Lumsden foruse of analytical equipment funded by the Canadian Foundation forInnovation (CFI). We are grateful for technical advice of DavidHobson and Paul Huber in relation to DIGE methods, and for MSanalysis by Li Zhang at the Advanced Protein Technology Centre,Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Proteins in Early Equine Conceptuses 237Hospital for Sick Children, Toronto, Ontario. The assistance of CaraLategan, Mandy Mulder and Garrett Jack and staff of the ArkellResearch Station is appreciated.ReferencesAlbihn A, Waelchli RO, Samper J, Oriol JG, Croy BA,Betteridge KJ, 2003: Production of capsular material byequine trophoblast transplanted into immunodeficient mice.Reproduction 125, 855–863.Arar S, Chan KH, Quinn BA, Waelchli RO, Hayes MA,Betteridge KJ, Monteiro MA, 2007: Desialylation of coretype 1 O-glycan in the equine embryonic capsule coincideswith immobilization of the conceptus in the uterus. CarbohydrRes 342, 1110–1115.Baker CB, Little TV, McDowell KJ, 1993: The live foaling rateper cycle in mares. Equine Vet J 15, 28–30.Ball BA, 1988: Embryonic loss in mares: incidence, possiblecauses and diagnostic considerations. Vet Clinics N AmEquine Pract 4, 263–290.Beers SA, Buckland AG, Giles N, Gelb MH, Wilton DC,2003: Effect of tryptophan insertions on the properties of thehuman group IIA phospholipase A2: mutagenesis producesan enzyme with characteristics similar to those of the humangroup V phospholipase A2. Biochemistry 42, 7326–7338.Beier HM, 2000: The discovery of uteroglobin and itssignificance for reproductive biology and endocrinology.Ann N Y Acad Sci 923, 9–24.Beier-Hellwig K, Kremer H, Bonn B, Lindner D, Beier HM,1995: Partial sequencing and identification of three proteinsfrom equine uterine secretion regulated by progesterone.Reprod Domest Anim 30, 295–298.Betteridge KJ, 2007: Equine embryology: an inventory ofunanswered questions. Theriogenology 68(Suppl 1), S9–S21.Betteridge KJ, Waelchli RO, Christie HL, Raeside JI, QuinnBA, Hayes MA, 2006: Effects of luteolysis on conceptusdevelopment during the second and third weeks of pregnancyin the mare. Anim Reprod Sci 91, 395–398.Birts CN, Barton CH, Wilton DC, 2007: A catalyticallyindependentphysiological function for human acute phaseprotein group IIA phospholipase A2: cellular uptake facilitatescell debris removal. J Biol Chem 283, 5034–5045.Bradford MM, 1976: A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Anal Biochem 72, 248–254.Brooks AS, Hammermueller J, DeLay JP, Hayes MA, 2003:Expression and secretion of ficolin beta by porcine neutrophils.Biochim Biophys Acta 1624, 36–45.Carnevale EM, Ramirez RJ, Squires EL, Alvarenga MA,Vanderwall DK, McCue PM, 2000: Factors affectingpregnancy rates and early embryonic death after equineembryo transfer. Theriogenology 54, 965–979.Chomczynski P, Sacchi N, 1987: Single-step method of RNAisolation by acid guanidinium thiocyanate-phenol-chloroformextraction. Anal Biochem 162, 156–159.Chu JW, Sharom FJ, Oriol JG, Betteridge KJ, Cleaver BD,Sharp DC, 1997: Biochemical changes in the equine capsulefollowing prostaglandin-induced pregnancy failure. MolReprod Dev 46, 286–295.Crossett B, Suire S, Herrler A, Allen WR, Stewart F, 1998:Transfer of a uterine lipocalin from the endometrium of themare to the developing equine conceptus. Biol Reprod 59,483–490.von der Decken V, Delbruck H, Herrler A, Beier HM, FischerR, Hoffmann KM, 2005: Recombinant bovine uteroglobinat 1.6 A resolution: a preliminary X-ray crystallographicanalysis. Acta Crystallogr F 61, 499–502.Ginther OJ, Bergfelt DR, Leith GS, Scraba ST, 1985:Embryonic loss in mares: incidence and ultrasonic morphology.Theriogenology 24, 73–86.Hobson DJ, Rupa P, Diaz GJ, Zhang H, Yang M, Mine Y,Turner PV, Kirby GM, 2007: Proteomic analysis ofovomucoid hypersensitivity in mice by two-dimensionaldifference gel electrophoresis (2D-DIGE). Food ChemToxicol 45, 2372–2380.Katavolos P2006: The role of Clara cells in the pathogenesis ofequine recurrent airway obstruction. PhD Thesis, Universityof Guelph.Kennedy MW, 2005: Uterocalin – provider of essential lipidsand amino acids to the pre-placentation equine conceptus.Havemeyer Found Monogr Ser 16, 53–56.Lillie BN, Hammermueller JD, Macinnes JI, Jacques M,Hayes MA, 2006: Porcine mannan-binding lectin A binds toActinobacillus suis and Haemophilus parasuis. Dev ComImmunol 30, 954–965.Morris LH, Allen WR, 2002: Reproductive efficiency ofintensively managed Thoroughbred mares in Newmarket.Equine Vet J 34, 51–60.Mukherjee AB, Zhang Z, Chilton BS, 2007: Uteroglobin: asteroid-inducible immunomodulatory protein that foundedthe secretoglobin superfamily. Endocr Rev 28, 707–725.Mu¨ller-Scho¨ttle F, Bogusz A, Grotzinger J, Herrler A,Krusche CA, Beier-Hellwig K, Beier HM, 2002: Full-lengthcomplementary DNA and the derived amino acid sequenceof horse uteroglobin. Biol Reprod 66, 1723–1728.Oriol JG, Betteridge KJ, Clarke AJ, Sharom FJ, 1993a:Mucin-like glycoproteins in the equine embryonic capsule.Mol Reprod Dev 34, 255–265.Oriol JG, Sharom FJ, Betteridge KJ, 1993b: Developmentallyregulated changes in the glycoproteins of the equineembryonic capsule. J Reprod Fertil 99, 653–664.Quinn BA, Caswell DE, Lillie BN, Waelchli RO, BetteridgeKJ, Hayes MA, 2006: The GM2-activator protein is a majorprotein expressed by the encapsulated equine trophoblast.Anim Reprod Sci 91, 391–394.Quinn BA, Hayes MA, Waelchli RO, Kennedy MW, BetteridgeKJ, 2007: Changes in major proteins in the embryonic capsuleduring immobilization (fixation) of the conceptus in the thirdweek of pregnancy in the mare. Reproduction 134, 161–170.Stewart F, Kennedy MW, Suire S, 2000: A novel uterinelipocalin supporting pregnancy in equids. Cell Mol Life Sci57, 1373–1378.Stout TA, Meadows S, Allen WR, 2005: Stage-specificformation of the equine blastocyst capsule is instrumentalto hatching and to embryonic survival in vivo. Anim ReprodSci 87, 269–281.Suire S, Stewart F, Beauchamp J, Kennedy MW, 2001:Uterocalin, a lipocalin provisioning the preattachmentequine conceptus: fatty acid and retinol binding properties,and structural characterization. Biochem J 356, 369–376.Tithof PK, Roberts MP, Guan W, Elgayyar M, Godkin JD,2007: Distinct phospholipase A2 enzymes regulate prostaglandinE2 and F2alpha production by bovine endometrialepithelial cells. Reprod Biol Endocrinol 5, 16.Waelchli RO, Betteridge KJ, 1996: Osmolality of equineblastocyst fluid from days 11 to 25 of pregnancy. ReprodFertil Dev 8, 981–988.Author’s address (for correspondence): MA Hayes, Department ofPathobiology, Ontario Veterinary College, University of Guelph,Guelph, Ontario, Canada N1G2W1. E-mail: ahayes@uoguelph.caConflict of interest: MA Hayes declares no conflict of interest; theremaining authors have not declared any conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 238–244 (2008); doi: 10.1111/j.1439-0531.2008.01168.xISSN 0936-6768Heat Stress, the Follicle, and Its Enclosed Oocyte: Mechanisms and PotentialStrategies to Improve Fertility in Dairy CowsZ RothDepartment of Animal Science, Faculty of Agricultural, Food and Environmental Quality Sciences, The Hebrew University, Rehovot, IsraelContentsReduced reproductive performance of lactating cows duringthe summer is associated with decreased thermoregulatorycompetence due to intensive genetic selection for high milkproduction. This review examines the immediate and delayedeffects of heat stress on follicular function and describes somepotential strategies for their alleviation. It focuses on how heatstress affects the follicle and its enclosed oocyte, suggestingthat perturbations in the follicular microenvironment, towhich the oocytes are exposed for long periods of development,reduce their developmental competence. Among thepotential alterations are reduction in gonadotropin secretion,alteration in follicular growth, attenuation of dominance, anddisruption of steroidogenesis. Evaporative cooling methodsare the most common strategy used to alleviate the effect ofheat stress; however, there is a compelling need to findadditional ways to improve fertility during the summer andautumn. Hormonal treatment to enhance removal of theimpaired follicles by synchronization of follicular waves withGnRH and PGF2a is suggested. An alternative method isstimulation of follicular growth by a brief treatment with bSTor FSH. Other strategies, such as timed AI and embryotransfer, have been recently used, making the optimization ofembryo cryopreservation procedures highly relevant. Protectionof the ovarian pool of oocytes from thermal stress vianutritional manipulations or administration of antioxidants orother survival factors should also be considered. A betterunderstanding of the underlying mechanisms by which heatstress impairs fertility may lead to the development ofadditional approaches to alleviate these effects.IntroductionReduced reproductive performance of lactating cowsduring the summer has been well-documented and isassociated with decreased thermoregulatory competenceof the animal due to intensive genetic selection for highmilk production. The most common strategy to alleviatethe effect of heat stress in dairy farms is to provide shadeand evaporative cooling systems, based on a combinationof sprinkling and ventilation, to enable animals tomaintain normothermia. This approach can reducebody temperature and increase milk production; however,its effect on summer fertility is limited (Hansen1997).During the autumn, although the weather is coolerand the cows are no longer subjected to thermal stress,conception rates remain low (Hansen 1997; Zeron et al.2001). Hyperthermia can directly disrupt concomitantfollicular function with adverse carry-over effects on itscompetence (Roth et al. 2000a,b). These include alterationsin follicular development, including depression ofdominance, impairment of follicular steroidogenesis andalterations in gonadotropin secretion (Wolfenson et al.2000). Given the complexity of the follicles’ activities,their interdependence and their interactions, perturbationin one component may disrupt oocyte developmentalcompetence (Webb and Campbell 2007). This reviewdescribes the immediate and delayed effects of heatstress on follicular function and some potential ways toalleviate them. It emphasizes the importance of theendocrine milieu and follicular microenvironment towhich the ovarian pool of oocytes is exposed anddescribes potential impairments in the follicle-enclosedoocyte upon heat shock, which subsequently affect itsdevelopmental competence.Immediate and Delayed Effects of Heat Stresson Follicular FunctionHeat stress and follicular dynamicsThe use of ultrasonography in recent years has demonstratedthat exposing dairy cows to elevated temperaturesalters follicular growth and function. Immediateeffects of heat stress include reduced size of the first- andsecond-wave dominant follicles (Badinga et al. 1993;Wilson et al. 1998a,b) and attenuation of dominance, asreflected by an increased number of large-sized follicles(Badinga et al. 1993, 1994; Wolfenson et al. 1995; Rothet al. 2000a) and delayed regression of subordinatefollicles (Wilson et al. 1998b; Roth et al. 2000a). Moreover,heat stress impairs the growth of medium-sizedfollicles (6–9 mm), as indicated by the earlier emergenceand delayed decrease of the second follicular wave(Roth et al. 2000a). Such alterations may lead to earlyemergence of the pre-ovulatory follicle and increasedduration of dominance (Wolfenson et al. 1995), both ofwhich are negatively associated with conception rate(Mihm et al. 1994). The picture regarding the effects ofheat stress on follicular recruitment is less clear (Wolfensonet al. 2000). A recent study in goats has indicatedthat follicles exposed to heat stress during recruitmentregress and never develop to large sizes or ovulate(Ozawa et al. 2005).Heat stress affects not only the antral folliclesemerging in the follicular wave, but also the ovarianpool of small antral follicles, resulting in carry-overeffects on follicular function. Growth of medium-sizedfollicles was attenuated in cows exposed to direct solarradiation in the previous oestrous cycle (Roth et al.2000a). The stage of follicular development that issusceptible to thermal stress has not been preciselydefined, but Roth et al. (2000a) introduced evidencesuggesting that early antral follicles of approximately0.5–1.0 mm in diameter are sensitive to heat stress.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Follicular and Oocyte Competence under Heat Stress 239The endocrine background governing alterations infollicular dynamics and depression of dominance underheat stress is not fully understood. Accumulatingevidence indicates alterations in systemic gonadotropin,inhibin and steroid concentrations, as well as in steroidogenesis(for review, see Wolfenson et al. 2000). Heatinducedincreases in the number of medium-sizedfollicles (i.e. immediate effect) were associated withincreased concentrations of follicle-stimulating hormone(FSH) and reduced concentrations of inhibin in theplasma (Wolfenson et al. 1995; Roth et al. 2000a).Inhibin and oestradiol synergistically affect FSH secretion(Kaneko et al. 1995). Thus, heat-induced impairmentof granulosa cell function, the main source ofplasma oestradiol and inhibin, can lead to increasedFSH concentrations in the plasma (Roth et al. 2000a).Follicular steroid production under heat stressSeasonal studies report that lower steroid concentrationsin the follicular fluid obtained from large folliclesduring the hot season are associated with reducedviability of granulosa cells and impaired aromataseactivity (Badinga et al. 1993; Wolfenson et al. 1995).Bridges et al. (2005) reported that follicle piecesobtained from heat-stressed cows secrete lower levelsof androstenedione and oestradiol upon gonadotropinstimulation. Similarly, thecal cells incubated at hightemperatures or collected from heat-stressed cows produceless androstenedione when stimulated by LH, butnot by forskolin, implying a disruption in LH receptorfunction upon heat stress (Wolfenson et al. 1995; Rothet al. 2001b). Nonetheless, expression of the mRNAencoding LH receptors in bovine thecal cells did notprovide evidence of such alterations (Roth et al. 2000b).In contrast, a recent study in goat reported decreasedexpression of LH receptors in heat-stressed follicles(Ozawa et al. 2005). Thus, many aspects of the molecularand cellular mechanisms underlying heat-induceddisruption of steroidogenesis remain to be elucidated.Alterations of steroidogenic capacity induced bysummer heat stress carry over to the final stage offollicle development, as evidenced by reduced androstenedioneproduction by thecal cells and low oestradiolconcentrations in follicular fluid collected from dominantfollicles in the autumn (Wolfenson et al. 1997). Inagreement with this, decreased oestradiol and androstenedioneproduction was recorded in granulosa andthecal cells obtained from follicles three to four weeksafter acute heat stress (Roth et al. 2000a). Similarly,oestradiol content in the follicular fluid aspirated fromcows was relatively low in late summer and increasedthroughout the autumn (Roth et al. 2004). Thus theextent of the heat stress effects on follicular function istransient as also reflected by the spontaneous improvementof fertility throughout autumn and early winter(Zeron et al. 2001).A strategy to enhance the removal of impairedfollicles and the reappearance of apparently normalfollicles was suggested: induction of frequent follicularwaves by hormonal treatment (i.e. GnRH followed byPGF2a (counteracted the effect of summer heat stress(Guzeloglu et al. 2001). Apparently, stimulation ofgonadotropins by GnRH has a beneficial effect ondeveloping follicles since induction of frequent follicularwaves during the autumn increased oestradiol content inpre-ovulatory follicles aspirated from previously heatstressedcows (Roth et al. 2004).Among the potential adverse effects associated withlow oestradiol levels are impairment of oestrus durationand intensity, increased incidence of anoestrus, silentovulation, and a reduced number of mounts (Gwazdauskaset al. 1981). Poor oestrus detection can beimproved by using modern aids such as the Heat-Watchsystem, a radio-telemetric pressure transducer, pedometricoestrus detection or alternatively, utilization ofovulation-synchronization protocols with fixed-timeartificial insemination (AI) procedures (Moore andThatcher 2006). Nevertheless, these procedures cannotovercome the negative effects of heat stress onconception.Reduced oestradiol concentrations in the blood duringthe follicular phase may also affect the pre-ovulatoryLH surge which, in turn, disrupts the cascade of eventsleading to oocyte ovulation. This is demonstrated by thelow amplitude of the GnRH-induced LH surge in heatstressedcows expressing low oestradiol concentration(Gilad et al. 1993). Nonetheless, studies in which GnRHwas administered at the onset of oestrus (Ullah et al.1996; Kaim et al. 2003) increased conception rates inheat-stressed primiparous cows, but this increase wasless pronounced in multiparous cows.Heat Stress and Oocyte DevelopmentalCompetenceIn-vivo and in-vitro studies support the view that bovineoocytes are susceptible to thermal stress at variousstages of follicular development. Perturbation in thephysiology of the follicle-enclosed oocyte during thelengthy period of follicular development could potentiallylead to an oocyte with reduced competence forfertilization and subsequent development.Oocytes harvested from cows during the summerexhibit a reduced ability to develop to the blastocyststage after in-vitro fertilization (Rocha et al. 1998; Al-Katanani et al. 2002) or chemical activation (Zeronet al. 2001). Cleavage to the two- and four-cell stagesfollowing chemical activation was delayed in bovineoocytes collected during the hot season (May–November)relative to oocytes collected during the cold season(December–April) (Aroyo et al. 2007b). The timing offirst cleavage (early vs delayed) is considered to havemajor long-lasting effects on subsequent embryonicdevelopmental potential (Fenwick et al. 2002), whichmay explain the inferior developmental competence ofoocytes collected during the summer.Elevated air temperatures before oestrus have beenassociated with reduced fertility (Al-Katanani et al.1999; Chebel et al. 2004). Intensive cooling from 1 daybefore oestrus to 8 days post-AI did not improve theconception rate of cows in the summer (Her et al. 1988).Similarly, using a fan-and-fogger cooling system 42 daysbefore oocyte collection did not increase the proportionof oocytes that developed to the blastocyst stage (Al-Katanani et al. 2002). Thus, either the oocytes wereÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

240 Z Rothalready compromised prior to the period of heat-stressrelief or the cooling was not efficient enough to decreaseheat stress. A study performed from late summer toearly winter indicated that a period of two to threeoestrous cycles is required for recovery from heatdamage and appearance of competent oocytes (Rothet al. 2001a). In support of this form of recovery,induction of maternal hyperthermia in mice carried overthrough three pregnancy cycles, as expressed by a lowerpregnancy rate and smaller litter sizes in the first cycleand slight increases in these parameters through thesecond and the third cycles after heat exposure (Aroyoet al. 2007a). It appears that not only the individualovulated oocyte, but also the ovarian pool of oocytescan be damaged during heat exposure. Nevertheless, theexact follicular stage at which the enclosed oocyte issusceptible to thermal stress has not been defined.Germinal-vesicle-stage oocytesMammalian oocytes are arrested at the prophase stageof the first meiotic division and acquire their meioticcompetence and fertilization potential in a stepwisemanner during follicular development. In an attempt todetermine the effect of heat stress on germinal-vesicle(GV)-stage oocytes, Payton et al. (2004) examined theeffect of heat shock on cumulus oocyte complexes(COCs) held at the GV stage using S-roscovitine, a cellcycleinhibitor. Exposure of GV-stage oocytes to 41°Cdid not impair GV breakdown but reduced the proportionof oocytes that progressed to metaphase II (MII)and was associated with further impairment of blastocystdevelopment. Similarly, in a recent study in mice,exposure of GV-stage oocytes to maternal hyperthermiadisrupted their developmental competence (Aroyo et al.2007a).The aforementioned concept of removing impairedfollicles also seems to be relevant to improving oocytequality, since frequent follicle aspirations by an ovumpick-up procedure improved the morphology and developmentalcompetence of oocytes aspirated in the fallfollowing the hot summer. Accordingly, hormonaltreatment with FSH, which is known to stimulatefollicular growth, increased the proportion of grade-1oocytes and the proportion that cleaved to the two-cellstage post-chemical activation. Similarly, short-termadministration of bovine somatotropin (bST) increasedthe proportion of grade-1 oocytes, but it did notimprove cleavage rate (Roth et al. 2002). Nevertheless,neither treatment was able to increase the rate ofblastocyst formation.MII-stage oocytesIn cattle, the process of oocyte maturation coincideswith oestrus events, which are known to be impaired byheat stress, as already described. Exposing animals toheat stress between the onset of oestrus and inseminationdisrupts subsequent embryonic development, withan increased proportion of abnormal and retardedembryos (Putney et al. 1989; Ealy et al. 1993). In-vitrostudies also indicate that exposure of cultured COCs toelevated temperatures during the first 12 h of maturationdecreases their cleavage rate (Roth et al. 2004; Rothand Hansen 2005) and the proportion of oocytes thatdevelop into blastocysts (Edwards and Hansen 1997;Roth and Hansen 2004a,b; Ju et al. 2005). Even thoughthe underlying mechanism is not entirely clear, heatstress may directly affect the oocyte or mediate negativeeffects through an impaired follicular environment (i.e.follicular fluid steroid content) and the impaired functionof surrounding cumulus cells. Lenz et al. (1983)showed that culturing COCs at 41°C for 24 h alters theirfunction, reduces cumulus-cell hyaluronic acid production,and impedes meiosis resumption. More recentstudies have reported that heat shock impairs bothnuclear and cytoplasmic maturation events, such astranslocation of cortical granules to the oolemma(Payton et al. 2004), cytoskeleton rearrangement (Rothand Hansen 2005), and spindle formation (Ju et al.2005; Roth and Hansen 2005). Similarly, heat-shockinducedperturbations of the spindle apparatus havebeen reported in mature porcine oocytes (Ju and Tseng2004) and in parthenogenetically activated bovineoocytes (Tseng et al. 2004). Such alterations maypotentially lead to incomplete nuclear maturation (Paytonet al. 2004; Roth and Hansen 2005), fertilizationfailure, and ⁄ or abnormal zygote formation. Edwardset al. (2005) reported that culturing oocytes at 41°Cdoes not compromise their nuclear and cytoplasmicmaturation but accelerates the process, and they suggestedthat fertilization performed 5 h earlier canattenuate the deleterious effect on oocyte development.Roth and Hansen (2005) reported that most heatshockedoocytes fail to undergo maturation and fertilizationand are arrested at stages MI through MII.Moreover, heat-shock-induced apoptosis has been documentedfor bovine oocytes exposed to elevated temperaturesduring maturation, in which an increasedproportion of oocytes express high caspase activity andTUNEL-positive reactions (Roth and Hansen 2004a,b).Similar findings have been recently reported for porcineoocytes (Tseng et al. 2006). Since the anti-apoptoticmolecule sphingosine 1-phosphate (S1P) blocked theeffect of heat shock on progression through meiosis(Roth and Hansen 2004b), it is reasonable to assumethat heat-shock-induced apoptosis is functionallyrelated to the inhibition of meiotic progression. Thesefindings indicate that more than one mechanism isinvolved in inducing adverse heat-stress effects in theoocyte.Heat-induced Oxidative Stress and OocyteCompetenceHeat-shock-induced oxidative stress is involved in earlyembryonic loss in mice, when stress is imposed at thezygote stage (Ozawa et al. 2002, 2004; Matsuzuka et al.2004, 2005). Whether it is also involved earlier indisrupting the bovine follicle-enclosed oocyte remainsunclear. Oxidants such as H 2 O 2 can induce meiotic cellcyclearrest as well as morphological changes characteristicof apoptosis in a stage-dependent manner,particularly in immature oocytes (Chaube et al. 2005).Reactive oxygen species (ROS) are essential, and serveas key signaling molecules in the resumption of meiosisÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Follicular and Oocyte Competence under Heat Stress 241(Takami et al. 1999), whereas excessive ROS levelswithin the follicle are associated with increased cytoplasmicdefects and abnormal chromosomal segregation(Van Blerkom et al. 1997).Given the importance of the oxidative status of theoocyte, and the fact that bovine oocytes are rich in fattyacids (McEvoy et al. 2000), membrane fatty-acid compositionis considered a major factor in determining theoocyte’s fate (Matorras et al. 1998). Seasonally increasedvariation in the fatty-acid profiles of thefollicular fluid and oocytes is associated with reduceddevelopment of the oocyte (Zeron et al. 2001). In the hotsummer, the proportions of saturated fatty acids inoocyte and granulosa cells are higher than those ofmono- and polyunsaturated fatty acids, implyingreduced oxidative status of the oocyte. Thus, administrationof antioxidants to reduce fatty-acid oxidation isone suggested strategy to stabilize the oocyte membraneupon heat stress. A recent study in mice providesevidence that administration of the antioxidant epigallocatechingallate (EGCG; 100 mg ⁄ kg body weight)before induction of hyperthermia increases the proportionof in-vivo-derived zygotes that reach the blastocyststage (Aroyo et al. 2006). Similarly, in-vitro maturationwith polyphenols improved the production of bovineembryos (Wang et al. 2007). A beneficial effect was alsoachieved by feeding lactating cows the antioxidantb-carotene (400 mg ⁄ day) for 90 days before insemination(Arechiga et al. 1998). Apparently, unlike forembryos (Hansen 2007b), antioxidant administrationcan alleviate the effect of thermal stress on the ovarianpool of oocytes.Accumulating evidence suggests that fat supplementationinfluences reproductive processes that are notdirectly related to energy balance and has beneficialeffects on the follicle, oocyte, embryo, and uterus(Mattos et al. 2000; Thatcher et al. 2003). Alterationsin fatty-acid composition of the oocyte might improveits developmental capacity. Zeron et al. (2002) reportedthat feeding ewes a diet supplemented with Ca salt offish oil increases the proportion of polyunsaturated fattyacids in the plasma and cumulus cells and improvesoocyte quality, as determined by membrane integrity.Feeding cows a diet enriched in unsaturated fatty acidsduring the summer did not have a beneficial effect onoocyte quality (Bilby et al. 2006), whereas a high level offeeding (Adamiak et al. 2005) or a high level of fat in thediet (Fouladi-Nashta et al. 2007) improved oocytedevelopmental competence. Given the potential ofnutritional manipulation, further research is warrantedto identify specific fatty acids and fat levels in diets thatmay have a beneficial effect on the ovarian pool ofoocytes.Heat Stress and Early Embryonic DevelopmentStudies indicate that embryonic loss under heat stress isdue to the sensitivity of early embryos to elevatedtemperatures. Embryos at early developmental stagesare more susceptible to thermal stress and become moreresistant during later developmental phases (Hansen2007a,b). Exposure of cows to heat stress on day 1 (butnot on days 3, 5, or 7) after oestrus decreased thedevelopment and viability of embryos on day 8 (Ealyet al. 1993). Exposure of cows to elevated temperaturesbetween onset of oestrus and insemination (Putney et al.1988) decreased subsequent embryo development. Similarly,induction of heat shock in vitro blocked thedevelopment of two-cell-stage embryos but had only amoderate effect on four- to eight-cell-stage embryos anda limited effect on the morulae (Hansen 2007a). Nevertheless,the mechanism by which the developing embryoacquires heat resistance to cellular disruption caused byelevated temperatures is not known (for reviews, see Ju2005; Hansen 2007a).Given that oocytes and early embryos are particularlysensitive to heat stress, embryo-transfer procedures thatbypass the effects of heat stress on early embryonicdevelopment have been attempted (see review, Rutledge2001; Hansen 2007b). One major limitation of thisapproach is the poor survival of embryos followingfreezing. Transferring in-vivo-derived frozen-thawedembryos increased pregnancy rates for recipient cowsrelative to traditionally inseminated cows (Putney et al.1989). Similarly, the percentage of pregnancies duringthe hot season was greater for cows receiving in-vitroderivedfresh embryos but not for those receiving frozenembryos (Ambrose et al. 1999; Drost et al. 1999).Moreover, use of vitrification did not improve thesurvival of the transferred embryos (Al-Katanani et al.2002). There is thus a compelling need to optimizeprocedures for embryo cryopreservation.Currently, fresh or in-vivo-derived frozen embryos arethe main source for embryo-production programs.Therefore, treatments that protect the ovarian pool ofoocytes can increase the number of competent oocytesavailable for embryo transfer. A protective effect of S1Phas been reported for bovine oocytes exposed to heatshock during maturation (Roth and Hansen 2004b,2005), as reflected by the reduced proportion of oocytesthat undergo apoptosis and the increased number thatreach the MII stage and develop into blastocysts. Theseembryos seem to have a high potential for development,since embryos derived from heat-shocked and S1Pprotectedoocytes were of the same quality as those fromnon-stressed oocytes, as determined by their total cellnumber, the proportion of apoptotic cells, and theintensity of caspase activity (Roth and Hansen 2004b).Moreover, pups developed from heat-stressed mice didnot differ from control pups in their learning potentialor episodic memory, as determined by behavioural andrecognition tests (Aroyo et al. 2007a).SummaryHyperthermia can impair cellular function in varioustissues of the reproductive system. However, disruptionof the follicle and its enclosed oocyte seems to be apivotal factor in the complex mechanism via which heatstress impairs fertility. This includes alterations in theendocrine milieu and follicular microenvironment towhich the ovarian pool of oocytes is exposed, leading totheir decreased developmental competence. Hyperthermiacan directly disrupt follicular function, but a carryovereffect on the follicle and its enclosed oocyte is alsoevident.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

242 Z RothEvaporative cooling systems used to maintain normothermiain high-lactating cows are likely to remainobligatory in dairy management during the hot season.However, since they only partially restore normalfertility, there is a compelling need for additionalapproaches. Beneficial strategies might involve hormonaltreatment to enhance the removal of impairedfollicles, for example: synchronization of recurringfollicular waves with the use of GnRH and PGF2a, orstimulation of follicular growth with brief treatments ofbST or small doses of FSH. More precise treatmentsdesigned to mitigate the deleterious effect on theendocrine milieu and to increase post-inseminationplasma progesterone should be further examined. Otherstrategies, such as timed embryo transfer, have beenmoderately successful. Therefore, optimizing the proceduresfor cryopreservation might allow producingembryos during the cool months of the year, when theoocytes are not subjected to heat damage. Protection ofthe ovarian oocyte pool from thermal stress throughnutritional manipulations, administration of ROSand ⁄ or other survival factors warrant further investigation.AcknowledgementsThe author thanks his colleagues, D. Wolfenson from the HebrewUniversity of Jerusalem and P.J. Hansen from the University ofFlorida, for their contribution to the study of reproduction in heatstressedcows which was discussed in this paper.ReferencesAdamiak J, Mackie K, Watt RG, Webb R, Sinclair KD, 2005:Impact of nutrition on oocyte quality: cumulative effects ofbody composition and diet leading to hyperinsulinemia incattle. Biol Reprod 73, 918–926.Al-Katanani YM, Webb DW, Hansen PJ, 1999: Factorsaffecting seasonal variation in 90-day non-return rate to firstservice in lactating Holstein cows in a hot climate. J DairySci 82, 2611–2616.Al-Katanani YM, Paula-Lopes FF, Hansen PJ, 2002: Effect ofseason and exposure to heat stress on oocyte competence inHolstein cows. J Dairy Sci 85, 390–396.Ambrose JD, Drost M, Monson RL, Rutledge JJ, Leibfried-Rutledge ML, Thatcher MJ, Kassa T, Binelli M, Hansen PJ,Chenoweth PJ, Thatcher WW, 1999: Efficacy of timedembryo transfer with fresh and frozen in vitro producedembryos to increase pregnancy rates in heat-stressed dairycattle. J Dairy Sci 82, 2369–2376.Arechiga CF, Staples CR, McDowell LR, Hansen PJ, 1998:Effects of timed insemination and supplemental beta-caroteneon reproduction and milk yield of dairy cows underheat stress. J Dairy Sci 81, 390–402.Aroyo A, Yavin S, Roth Z, Arav A, 2006: Pretreatment withantioxidant epigallocatechin gallate (EGCG) counteractsthe negative effects of maternal hyperthermia on embryonicdevelopment in mice. Reprod Fertil Dev 18, 185. Abstract156).Aroyo A, Yavin S, Arav A, Roth Z, 2007a: Maternalhyperthermia disrupts developmental competence of follicle-enclosedoocytes: in vivo and ex vivo studies in mice.Theriogenology 67, 1013–1021.Aroyo A, Yavin S, Roth Z, Arav A, 2007b: Hindering ofcleavage timing in bovine parthenotes during the hot season.Reprod Fertil Dev 19, 203. Abstract 173).Badinga L, Thatcher WW, Diaz T, Drost M, Wolfenson D,1993: Effect of environmental heat stress on folliculardevelopment and steroidogenesis in lactating Holstein cows.Theriogenology 39, 797–810.Badinga L, Thatcher WW, Wilcox CJ, Morris G, Entwistle K,Wolfenson D, 1994: Effect of season on follicular dynamicsand plasma concentrations of oestradiol-17ß, progesteroneand luteinizing hormone in lactating Holstein cows. Theriogenology42, 1263–1274.Bilby TR, Block J, do Amaral BC, Sa Filho O, Silvestre FT,Hansen PJ, Staples CR, Thatcher WW, 2006: Effects ofdietary unsaturated fatty acids on oocyte quality andfollicular development in lactating dairy cows in summer.J Dairy Sci 89, 3891–3903.Bridges PJ, Brusie MA, Fortune JE, 2005: Elevated temperature(heat stress) in vitro reduces androstenedione andestradiol and increases progesterone secretion by follicularcells from bovine dominant follicles. Domest Anim Endocrinol29, 508–522.Chaube SK, Prasad PV, Thakur SC, Shrivastav TG, 2005:Hydrogen peroxide modulates meiotic cell cycle and inducesmorphological features characteristic of apoptosis in ratoocytes cultured in vitro. Apoptosis 10, 863–874.Chebel RC, Santos JE, Reynolds JP, Cerri RL, Juchem SO,Overton M, 2004: Factors affecting conception rate afterartificial insemination and pregnancy loss in lactating dairycows. Anim Reprod Sci 84, 239–255.Drost M, Ambrose JD, Thatcher M-J, Cantrell CK, WolfsdorfKE, Hasler JF, Thatcher WW, 1999: Conception rates afterartificial insemination or embryo transfer in lactating dairycows during summer in Florida. Theriogenology 52, 1161–1167.Ealy AD, Drost M, Hansen PJ, 1993: Developmental changesin embryonic resistance to adverse effects of maternal heatstress in cows. J Dairy Sci 76, 2899–2905.Edwards JL, Hansen PJ, 1997: Differential responses of bovineoocytes and preimplantation embryos to heat shock. MolReprod Dev 46, 138–145.Edwards JL, Saxton AM, Lawrence JL, Payton RR, DunlapJR, 2005: Exposure to a physiologically relevant elevatedtemperature hastens in vitro maturation in bovine oocytes.J Dairy Sci 88, 4326–4333.Fenwick J, Platteau P, Murdoch AP, Herbert M, 2002: Timefrom insemination to first cleavage predicts developmentalcompetence of human preimplantation embryos in vitro.Hum Reprod 17, 407–412.Fouladi-Nashta AA, Gutierrez CG, Gong JG, GarnsworthyPC, Webb R, 2007: Impact of dietary fatty acids on oocytequality and development in lactating dairy cows. BiolReprod 77, 9–17.Gilad E, Meidan R, Berman A, Graber Y, Wolfenson D,1993: Effect of heat stress on tonic and GnRH-inducedgonadotrophin secretion in relation to concentration ofoestradiol in plasma of cyclic cows. J Reprod Fertil 99,315–321.Guzeloglu A, Ambrose JD, Kassa T, Diaz T, Thatcher MJ,Thatcher WW, 2001: Long-term follicular dynamics andbiochemical characteristics of dominant follicles in dairycows subjected to acute heat stress. Anim Reprod Sci 66, 15–34.Gwazdauskas FC, Thatcher WW, Kiddy CA, Paape MJ,Wilcox CZ, 1981: Hormonal patterns during heat stressfollowing PGF2a-tham salt induced luteal regression inheifers. Theriogenology 16, 271–285.Hansen PJ, 1997: Effects of environment of bovine reproduction.In: Youngquist RS (ed.), Current Therapy in LargeAnimal Theriogenology. Saunders WB, Philadelphia, pp.403–415.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Follicular and Oocyte Competence under Heat Stress 243Hansen PJ, 2007a: To be or not to be; determinants ofembryonic survival following heat shock. Theriogenology68, 40–48. (Review).Hansen PJ, 2007b: Exploitation of genetic and physiologicaldeterminants of embryonic resistance to elevated temperatureto improve embryonic survival in dairy cattle duringheat stress. Theriogenology 68, 242–249. (Review).Her E, Wolfenson D, Flamenbaum I, Folman Y, Kaim M,Berman A, 1988: Thermal productive, and reproductiveresponses of high yielding cows exposed to short termcooling in summer. J Dairy Sci 71, 1085–1092.Ju JC, 2005: Cellular responses of oocyte and embryos underthermal stress: hints to molecular signaling. Anim Reprod 2,79–90.Ju JC, Tseng JK, 2004: Nuclear and cytoskeletal alterations ofin vitro matured porcine oocytes under hyperthermia. MolReprod Dev 68, 125–133.Ju JC, Jiang S, Tseng JK, Parks JE, Yang X, 2005: Heat shockreduces developmental competence and alters spindle configurationof bovine oocytes. Theriogenology 64, 1677–1689.Kaim M, Bloch A, Wolfenson D, Braw-Tal R, Rosenberg M,Voet H, Folman Y, 2003: Effects of GnRH administered tocows at the onset of estrus on timing of ovulation, endocrineresponses, and conception. J Dairy Sci 86, 2012–2021.Kaneko H, Nakanishi Y, Akagi S, Arai K, Taya K, WatanabeG, Sasamoto S, Hasegawa Y, 1995: Immunoneutralizationof inhibin and oestradiol during the follicular phase of theoestrous cycle in cows. Biol Reprod 53, 931–939.Lenz RW, Ball GD, Leibfried ML, Ax RL, First NL, 1983: Invitro maturation and fertilization of bovine oocytes aretemperature-dependent processes. Biol Reprod 29, 173–179.Matorras R, Ruiz JI, Mendoza R, Ruiz N, Sanjurjo P,Rodriguez-Escudero FJ, 1998: Fatty acid composition offertilization-failed human oocytes. Hum Reprod 13, 2227–2230.Matsuzuka T, Ozawa M, Hirabayashi M, Ushitani A, KanaiY, 2004: Developmental competence and glutathione contentof maternally heat-stressed mouse oocytes and zygotes.Anim Sci J 75, 117–124.Matsuzuka T, Ozawa M, Nakamura A, Ushitani A, HirabayashiM, Kanai Y, 2005: Effects of heat stress on the redoxstatus in the oviduct and early embryonic development inmice. J Reprod Dev 51, 281–287.Mattos R, Staples CR, Thatcher WW, 2000: Effects of dietaryfatty acids on reproduction in ruminants. Rev Reprod 5, 38–45. (Review).McEvoy TG, Coull GD, Broadbent PJ, Hutchinson JS,Speake BK, 2000: Fatty acid composition of lipids inimmature cattle, pig and sheep oocytes with intact zonapellucida. J Reprod Fertil 118, 163–170.Mihm M, Baguisi A, Boland MP, Roche JF, 1994: Associationbetween the duration of dominance of the ovulatory follicleand pregnancy rate in beef heifers. J Reprod Fertil 102, 123–130.Moore K, Thatcher WW, 2006: Major advances associatedwith reproduction in dairy cattle. J Dairy Sci 89, 1254–1266.Ozawa M, Hirabayashi M, Kanai Y, 2002: Developmentalcompetence and oxidative state of mouse zygotes heatstressedmaternally or in vitro. Reproduction 124, 683–689.Ozawa M, Matsuzuka T, Hirabayashi M, Kanai Y, 2004:Redox status of the oviduct and Cdc2 activity in 2-cell stageembryos in heat-stressed mice. Biol Reprod 71, 291–296.Ozawa M, Tabayashi D, Latief TA, Shimizu T, Oshima I,Kanai Y, 2005: Alterations in follicular dynamics andsteroidogenic abilities induced by heat stress during follicularrecruitment in goats. Reproduction 129, 621–630.Payton RR, Romar R, Coy P, Saxton AM, Lawrence JL,Edwards JL, 2004: Susceptibility of bovine germinal vesiclestageoocytes from antral follicles to direct effects of heatstress in vitro. Biol Reprod 71, 1303–1308.Putney DJ, Drost M, Thatcher WW, 1988: Embryonicdevelopment in superovulated dairy cattle exposed toelevated ambient temperatures between days 1–7 postinsemination. Theriogenology 33, 195–209.Putney DJ, Drost M, Thatcher WW, 1989: Influence ofsummer heat stress on pregnancy rates of lactating dairycattle following embryo transfer or artificial insemination.Theriogenology 31, 765–778.Rocha A, Randel RD, Broussard JR, Lim JM, Blair RM,Roussel JD, Godke RA, Hansel W, 1998: High environmentaltemperature and humidity decrease oocyte quality inBos taurus but not in Bos indicus cows. Theriogenology 49,657–665.Roth Z, Hansen PJ, 2004a: Involvement of apoptosis in heatinduceddisruption of competence of bovine oocyte. BiolReprod 71, 1898–1906.Roth Z, Hansen PJ, 2004b: Sphingosine-1-Phosphate protectscultured bovine oocytes from physiologically-relevant thermalstress. Biol Reprod 71, 2072–2078.Roth Z, Hansen PJ, 2005: Disruption of nuclear maturationand rearrangement of cytoskeletal elements in bovineoocytes exposed to heat shock during maturation. Reproduction129, 235–244.Roth Z, Meidan R, Braw-Tal R, Wolfenson D, 2000a:Immediate and delayed effects of heat stress on folliculardevelopment and its association with plasma FSH andinhibin concentration in cows. J Reprod Fertil 120, 83–90.Roth Z, Meidan R, Preger E, Levi N, Braw-Tal R, WolfensonD 2000b: Delayed effects of heat stress on steroidogenesis inbovine preovulatory follicles. 14th International Congresson Animal Reproduction, Stockholm, Sweden, Abst. 1, 15.Roth Z, Arav A, Bor A, Zeron Y, Braw-Tal R, WolfensonD, 2001a: Improvement of quality of oocytes collected inthe autumn by enhanced removal of impaired folliclesfrom previously heat-stressed cows. Reproduction 122,737–744.Roth Z, Meidan R, Braw-Tal R, Shaham-Albalancy A,Wolfenson D, 2001b: Delayed effect of heat stress onsteroidogenesis in bovine medium-size and preovulatoryfollicles. Reproduction 121, 745–751.Roth Z, Arav A, Bor A, Zeron Y, Braw-Tal R, Wolfenson D,2002: Effect of treatment with FSH or bST on the quality ofoocytes aspirated in the autumn from previously heatstressedcows. J Dairy Sci 85, 1398–1405.Roth Z, Bor A, Braw-Tal R, Wolfenson D, 2004: Carry-overeffect of summer thermal stress on characteristics of the preovulatoryfollicle of lactating cows. J Therm Biol 29, 681–685.Rutledge JJ, 2001: Use of embryo transfer and IVF to bypasseffects of heat stress. Theriogenology 55, 105–111. (Review).Takami M, Preston SL, Toyloy VA, Behrman HR, 1999:Antioxidants reversibly inhibit the spontaneous resumptionof meiosis. Am J Physiol 276, 684–688.Thatcher WW, Guzeloglu A, Meikle A, Kamimura S, Bilby T,Kowalski AA, Badinga L, Pershing R, Bartolome J, SantosJE, 2003: Regulation of embryo survival in cattle. ReprodSuppl 61, 253–266. (Review).Tseng JK, Chen CH, Chou PC, Yeh SP, Ju JC, 2004:Influences of follicular size on parthenogenetic activationand in vitro heat shock on the cytoskeleton in cattle oocytes.Reprod Domest Anim 39, 146–153.Tseng JK, Tang PC, Ju JC, 2006: In vitro thermal stressinduces apoptosis and reduces development of porcineparthenotes. Theriogenology 15, 1073–1082.Ullah G, Fuquay JW, Keawkhong T, Clark BL, Pogue DE,Murphey EJ, 1996: Effect of gonadotropin-releasing hormoneat estrus on subsequent luteal function and fertility inÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

244 Z Rothlactating Holsteins during heat stress. J Dairy Sci 79, 1950–1953.Van Blerkom J, Antczak M, Schrader R, 1997: The developmentalpotential of the human oocyte is related to thedissolved oxygen content of follicular fluid: association withvascular endothelial growth factor levels and perifollicularblood flow characteristics. Hum Reprod 12, 1047–1055.Wang ZG, Yu SD, Xu ZR, 2007: Improvement in bovineembryo production in vitro by treatment with green teapolyphenols during in vitro maturation of oocytes. AnimReprod Sci 100, 22–31.Webb R, Campbell BK, 2007: Development of the dominantfollicle: mechanisms of selection and maintenance of oocytequality. Soc Reprod Fertil Suppl 64, 141–163.Wilson SJ, Kirby CJ, Koenigsfeld AT, Keisler DH, Lucy MC,1998a: Effects of controlled heat stress on ovarian functionof dairy cattle. 2. Heifers. J Dairy Sci 81, 2132–2138.Wilson SJ, Marion RS, Spain JN, Spiers DE, Keisler DH,Lucy MC, 1998b: Effects of controlled heat stress on ovarianfunction of dairy cattle. 1. Lactating cows. J Dairy Sci 81,2124–2131.Wolfenson D, Thatcher WW, Badinga L, Savio JD, Meidan R,Lew BJ, Braw-Tal R, Berman A, 1995: Effect of heat stresson follicular development during the oestrous cycle inlactating dairy cattle. Biol Reprod 52, 1106–1113.Wolfenson D, Lew BJ, Thatcher WW, Graber Y, Meidan R,1997: Seasonal and acute heat stress effects on steroidproduction by dominant follicles in cows. Anim Reprod Sci47, 9–19.Wolfenson D, Roth Z, Meidan R, 2000: Impaired reproductionin heat-stressed cattle: basic and applied aspects. AnimReprod Sci 61, 535–547. (Review).Zeron Y, Ocheretny A, Kedar O, Borochov A, Sklan D, AravA, 2001: Seasonal changes in bovine fertility: relation todevelopmental competence of oocytes, membrane propertiesand fatty acid composition of follicles. Reproduction 121,447–454.Zeron Y, Sklan D, Arav A, 2002: Effect of polyunsaturatedfatty acid supplementation on biophysical parameters andchilling sensitivity of ewe oocytes. Mol Reprod Dev 61, 271–278.Author’s address (for correspondence): Z Roth, Department ofAnimal Science, Faculty of Agricultural, Food and EnvironmentalQuality Sciences, The Hebrew University, PO Box 12, Rehovot 76100,Israel. E-mail: roth@agri.huji.ac.ilConflict of interest: The author declares no conflict of interest.Research was supported in part by BARD grant F1–330–2002.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 245–251 (2008); doi: 10.1111/j.1439-0531.2008.01169.xISSN 0936-6768Fertilization and Early Embryonic Development in the Porcine Fallopian TubeK-P Bru¨ssow 1 ,JRátky 2 and H Rodriguez-Martinez 31 Department of Reproductive Biology, FBN Research Institute for the Biology of Farm Animals, Dummerstorf, Germany; 2 Department of Biology ofReproduction, Research Institute for Animal Breeding and Nutrition, Herceghalom, Hungary; 3 Division of Reproduction, Faculty of VeterinaryMedicine and Animal Science, Swedish University of Agricultural Sciences (SLU), Uppsala, SwedenContentsFertilization and early embryo development relies on acomplex interplay between the Fallopian tube and thegametes before and after fertilization. Thereby the oviduct,as a dynamic reproductive organ, enables reception, transportand maturation of male and female gametes, theirfusion, and supports early embryo development. This paperreviews current knowledge regarding physiological processesbehind the transport of boar spermatozoa, their storage inand release from the functional sperm reservoir (SR), and ofthe interactions that newly ovulated oocytes play within thetube during their transport to the site of fertilization.Experimental evidence of an ovarian control on spermrelease from the SR is highlighted. Furthermore, the impactof oviductal secretion on sperm capacitation, oocyte maturation,fertilization and early embryo development isstressed.IntroductionThe oviduct (synonyms Fallopian tube, salpinx, uterinetube) firstly described by Gabriel Fallopius in 1561, wasconsidered a simple connection between the ovary andthe uterus for a long time. Today, the oviduct isregarded as one of the most dynamic reproductiveorgans, whose functional aspects are not yet fullyunderstood (Rodriguez-Martinez et al. 2001). The oviductprovides a favourable environment for spermatozoa,enabling their transport and ensuring beneficialconditions during storage while preparing them forfertilization; as well as oocytes, making ovum pick-upand transport to the site of fertilization possible, whileensuring their final maturation. Both fertilization andearly embryonic development take place within theoviduct. This paper reviews physiological processesbehind the transport of boar spermatozoa, their storagein and release from the functional sperm reservoir (SR),and of the interactions that newly ovulated oocytes playduring their transport to the site of fertilization. Someaspects are highlighted, such as experimental evidencefor an ovarian influence on sperm release from the SRand the relevance of oviductal secretion on spermcapacitation, oocyte maturation, fertilization and earlyembryo development.The OviductThe porcine oviduct is approximately 25 cm in length(Bru¨ ssow 1985) and it is anatomically divided into threemain segments, e.g. the infundibulum, the ampulla andthe isthmus. Connecting areas, i.e. the terminal section(ostium), the ampulla-isthmic-junction (AIJ) and theutero-tubal-junction (UTJ) are also distinguished. Thehistoarchitecture of the Fallopian tube is simple, witha non-glandular mucosa (endosalpinx), covered with alining epithelium with secretory and ciliated cells,a double-layered smooth muscle (myosalpinx) and acovering serosa (mesosalpinx). Towards the ostium, thethickness of the internal, circular muscle becomesthinner; the longitudinal mucosal plicae gain complexityand the number of ciliated cells considerablyincreases (Rodriguez-Martinez et al. 2001; Yaniz et al.2006). This histoarchitecture builds tubal compartmentswith apparently different roles, all with optimal environmentwith regards to pH, osmotic pressure, nutrients,specific secretory products and signal molecules(reviewed by Rodriguez-Martinez 2007). Thus, preparationof gametes, fertilization and early embryonicdevelopment are supported and regulated by theoviduct.Sperm Transport and the Tubal SpermReservoirAfter deposition of spermatozoa in the cervix by matingor AI, a sperm subpopulation is rapidly transportedthrough the uterus, whereas the majority is eliminatedfrom the uterine lumen (Rodriguez-Martinez et al.2005). Spermatozoa that ascended the uterus in the firstphase of sperm transport colonize the UTJ and thecaudal isthmus (>10 4 spermatozoa), a segment thatacts as a functional pre-ovulatory sperm reservoir,temporarily arresting spermatozoa (up to 30 h) and,presumably, activating them at a given time (Rodriguez-Martinez et al. 2005). In the SR, most spermatozoamaintain normal ultrastructure and viability (Rodriguez-Martinezet al. 1990; Mburu et al. 1997). Severalconcerted factors are thought to explain the formationof the functional SR, including the narrowed lumen(Hunter 1984), viscous mucus (Johansson et al. 2000),lower temperature (Hunter and Nichol 1986), localenzymatic and ionic milieu (Rodriguez-Martinez et al.1991), selective binding of spermatozoa to the epithelium(Fazeli et al. 1999), and specific tubal fluid components(Tienthai et al. 2004), all which primarily leadto sperm quiescence.The sequential release of a restricted number ofspermatozoa from the SR towards the AIJ ensuresfertilization of oocytes within a time window, even ifovulation lasts over a longer interval. The periovulatoryprogression of spermatozoa is suggested to be a complexand concerted process including opening of the lumenby a decrease in the hormonally-driven endosalpingealoedema, dissolution of the hyaluronan (HA)-richÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

246 K-P Bru¨ ssow, J Rátky and H Rodriguez-MartinezTable 1. Sperm subpopulations (as %; LSM ± SE) in the porcine SRat pre-, peri- and post-ovulatory oestrus (modified from Tienthai et al.2004)RetrievedSR-spermatozoa (%)Oestrous stage at sperm retrieval (accounted formspontaneous ovulation)Preovulatory()8 h)Periovulatory(±4 h)Postovulatory(+8 h)Number of sows 7 10 7Viable, non-capacitated 72.8 ± 6.9 * 69.0 ± 6.8 * 70.9 ± 7.4 *Viable, capacitated 4.9 ± 3.0 * 1.4 ± 3.0 * 13.7 ± 3.0 *Dead 15.5 ± 5.3 * 16.9 ± 5.2 * 12.8 ± 5.7 **p < 0.05.mucus, hyperactive sperm motility, increased flow oftubal fluid and re-directed oviductal contractions(Rodriguez-Martinez et al. 1998). On the other hand,there is no evidence of a massive sperm release orcapacitation during the pre- and periovulatory periods(Table 1), but rather a constant sperm release towardsthe site of fertilization in relation to the occurrence ofspontaneous ovulation (Mburu et al. 1997; Tienthaiet al. 2004).What Controls Sperm Release from the SR?It is yet to be determined which are mechanism(s) thatinduce the sperm release from the porcine SR. The mostmentioned is the proposed presence of discrete signalsacting as transduced endocrine information from thepre-ovulatory Graafian follicles (Hunter et al. 1983;Hunter 1990), although experimental evidence backingthis mechanism is yet scarce.Experimental evidence for an ovarian influence on spermrelease from the SRMicroinjection of exogenous progesterone or progesterone-richfollicular fluid (FF) under the serosal layer ofthe oviduct surrounding the SR or directly into the SRprovoked a prominent release of spermatozoa and a34% incidence of polyspermy vs 2% in controls,injected with steroid-free FF or medium (Hunter 1972;Hunter et al. 1999). This has lead to the assumptionthat progesterone could be the key effector of spermrelease from the pig SR. Progesterone concentrationsare 100 to 1000 times higher in the FF than inperipheral blood (Eiler and Nalbandov 1977; Blo¨ dowet al. 1990), and thus could stimulate sperm releasefrom the SR when entering the oviduct at ovulation.However, FF does not (or only in a negligible amount,Hansen et al. 1991) enter the porcine oviduct, asdemonstrated by our own experiments, where the entryof FF into the tube was prevented by unilateralendoscopic ligation of the oviduct on the uterine sideof the infundibulum or FF-aspiration before ovulation(Bru¨ ssow et al. 1999a, 2003). Analyses of progesteroneconcentration in FF and oviductal fluid showed thatwhile progesterone levels were similarly low prior to orafter ovulation (0.09 ± 0.13 vs 0.12 ± 0.16 ng ⁄ ml) anddid not differ from the contra-lateral control side (groupFF aspiration: 0.29 ± 0.17 vs 0.24 ± 0.35 ng ⁄ ml;Table 2. Percentages of spermatozoa detected in the ampulla andisthmus of inseminated gilts (n = 12) where FF entry into the oviductwas prevented by aspiration of the FF or ligation of the oviduct on theuterine side of the infundibulum vs sham-ligation (Bru¨ ssow et al. 2003)Group(treatment vs control)Spermatozoa (%) within oviductalsectionAmpullaIsthmusOviduct ligation 0 * 100 *Intact control 35.5 * 64.5 *FF aspiration 0 * 100 *Intact control 29.6 * 70.4 *Sham ligation 42.1 57.9Intact control 45.7 54.3*p < 0.05 vs control.group oviduct ligation: 0.22 ± 0.19 vs 0.21 ±0.22 ng ⁄ ml); those in FF were significantly higher(269.7 ± 67.9 and 389.6 ± 226.5 ng ⁄ ml) albeit notreflected in the oviductal fluid. Likewise, Hansen et al.(1991) found only 0.03 to 0.07% of the progesteroneconcentration of the FF was present within the oviductalflushing. Therefore, although a direct role ofprogesterone in eliciting sperm release from the SR isyet to be determined, a series of interesting results haveemanated from studies testing the involvement of FF inthe process of fertilization. The effect of withdrawingthe FF by aspiration of pre-ovulatory follicles or byoviduct ligation before ovulation was compared for itsrelation to the way spermatozoa were distributed withinthe oviduct right after ovulation vs in sham-manipulatedor intact oviducts (Bru¨ ssow et al. 1999b, 2003).Both a ligature of the oviduct or the FF-aspirationreduced the proportion of spermatozoa within theampullar ⁄ AIJ tubal segment (Table 2), thus suggestingthe sperm release from the SR was constrained. However,it is yet to be clarified whether an absentprogesterone-rich FF was the factor behind this difference.The FF is able to stimulate in vivo fertilization, asdemonstrated in a follow-up study (Bru¨ ssow et al. 2003)in which 6 to 10 cumulus-oocyte-complexes (COCs)recovered from donor gilts were transferred into theoviducts of inseminated recipient gilts – either togetherwith 0.6 ml of FF in one oviduct (FF group – localinfluence of FF) or together with 0.6 ml of PBS into thecontra-lateral oviduct (PBS group – without FF influence,within-animal controls). Another group of giltsserved as sham-operated control-group. The FF wascollected from both ovaries in the recipient gilts andpooled, but the COCs were rejected, to diminishconfounding effects. From this experiment, it wasevident that similarly to controls, the presence of FFcomponentswithin the oviduct increased fertilizationrate (Table 3). However, since a direct effect of FF isalready excluded in vivo, owing to the fact that the FFdoes not significantly enter the oviduct, neither it is anobligatory carrier of porcine oocytes at ovulation(Bru¨ ssow et al. 1998a, 1999a); the influence of FFcomponents on sperm migration and fertilization ispossibly generated by local counter-current transfer intothe blood and lymph circulation, as suggested by HunterÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Fertilization in the Porcine Fallopian Tube 247Table 3. In vivo fertilization and cleavage rates of porcine cumulusoocyte-complexes(COCs) transferred together with FF or PBS(n = 35 gilts; Bru¨ ssow et al. 2003)FF-treatedoviductsPBS-treatedoviductsControloviductsNumber of oviducts (n) 24 22 24Number of COCs transferred (n) 206 144 –Number of oocytes recovered (n) 138 70 235Recovery rate (%) 67.0 * 45.5 * 73.4 aNumber of oocytes fertilized (n) 78 26 119Fertilization rate (%) 56.5 * 31.7 * 50.6 *Number of cleaved embryos (n) 76 26 113Cleavage rate b (%) 97.4 100 98.3a Relative to the number of ovulated follicles.b Relative to the number of fertilized oocytes.*p < 0.05.et al. (1983), Hunter (1996). However, such hypothesisis yet to be proven.Involvement of oocytes and their cumulus cells in spermrelease form the SRThat oocytes and their cumulus vestment divert spermatozoatowards the ova has been considered previously(Austin and Walton 1960; Harper 1973; Bedfordand Kim 1985), but putative molecular mechanisms arestill speculative (Hunter 1993), and experimental evidenceof an association of oocytes with SR-spermrelease in the pig is missing. Therefore, we created amodel in which ova from donor gilts were transferred atthe periovulatory period into only one oviduct – theother serving as a control – of bilaterally ovectomized(aspiration of oocytes from the follicle) recipient giltsthat previously underwent endoscopic intrauterineinsemination (endo-IUI) with low sperm-doses intoboth uterine horns (Bru¨ ssow et al. 2006). Such a modelavoids ovulation of the recipient’s COCs, and enablesassessment of how the transfer of ova would influencesperm release from the SR. The most important findingof this study was that the presence of COCs in theoviduct significantly increased the percentages of spermatozoain the ampullar and isthmic segments, comparedto control oviducts. Both total numbers andproportions of spermatozoa were always higher, indicatingthat the presence of ova affected sperm releasefrom the SR (Table 4). However, this result does notexplain which component(s) of the oocytes promptedthe spermatozoa to leave the SR.Which components are involved in the sperm release formthe SR?One of the components of the oocytes that mayinfluence spermatozoa to leave the SR could be glycosaminoglycans(GAGs), especially the non-sulfatedGAG hyaluronan (HA) (Tienthai et al. 2004; Liberdaet al. 2006). The HA, a major component of the porcineextracellular matrix of the cumulus and zona pellucida(ZP) (Yokoo et al. 2002) is increasingly synthesized byCOCs during cumulus expansion (Kimura et al. 2002;Yokoo et al. 2007). It is thought to participate in spermcapacitation and release towards the site of fertilization(Tienthai et al. 2000a). Since the level of HA in theporcine ampullar fluid increases around ovulation(Tienthai et al. 2000b), it is possible that additionalHA enters the oviduct together with the COCs. Therefore,we experimentally analysed both the influence oftransferred ovum quality (COCs or cumulus cellremovedoocytes) and addition of HA (COCs + HA,oocytes + HA) on fertilization and the numbers ofZP-accessory spermatozoa in previously ovectomizedendo-IUI gilts (Bru¨ ssow et al. 2003, 2006). Embryodevelopment was not affected by ova quality or exogenousHA. The numbers of accessory spermatozoa werehighest in those COCs transferred together with HA(COC+HA), which did not differ from controls(Table 5). Comparable results have been achieved inprevious experiments (Bru¨ ssow et al. 2003) where additionof exogenous HA, together with cumulus-freeoocytes increased fertilization rates and the number ofaccessory spermatozoa compared to oocytes or COCs.One explanation for these results was that washingand ⁄ or manipulation of COCs in vitro, removed componentsimportant for fertilization. An alternativeinterpretation was that the proposed impact of FF viacounter-current transfer (Hunter 1972) on sperm releaseis missing in gilts from which FF was withdrawn.Summarizing the experimental data, the higher numberof accessory spermatozoa may strengthen the evidencethat HA participates in the process of sperm releasefrom the SR (Rodriguez-Martinez et al. 2005; Liberdaet al. 2006).Ovum Transport and Fertilization in theOviductOvulation, i.e. the ‘flow out’ of a mature folliclereleasing a COC into the oviduct, is a multifacetedTable 4. Distribution of boar spermatozoa (mean ± SE) withinoviductal sections in gilts (n = 10) with transferred COCs (treatedoviducts) or without COCs (control oviducts) (Bru¨ ssow et al. 2006)Treated oviducts (withCOCs)Control oviducts (withoutCOCs)n % n %Total sperm count 63 869 100 85 049 100Sperm reservoir 52 696 82.5 ± 0.5 * 80 921 95.2 ± 0.06 *Isthmus 5699 8.9 ± 0.11 * 2754 3.2 ± 0.06 *Ampulla 5474 8.6 ± 0.11 * 1374 1.6 ± 0.04 *Isthmus + Ampulla 11 173 17.5 ± 0.15 * 4128 4.9 ± 0.07 **p < 0.01 (chi-square).Table 5. Mean numbers (LSMeans ± SEM) of blastomeres and ofaccessory spermatozoa within the zona pellucida in embryos developingin vivo following oocyte or COC transfer, with or withoutexogenous HA (Bru¨ ssow et al. 2006)GroupBlastomeresAccessoryspermatozoaOocyte 2.6 ± 0.2 29.1 ± 3.1 *Oocyte + HA 2.9 ± 0.2 26.2 ± 3.9 *COC 3.1 ± 0.2 22.2 ± 3.3 *COC + HA 2.9 ± 0.2 46.2 ± 5.0 *Control 2.9 ± 0.2 45.2 ± 3.1 **p < 0.05.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

248 K-P Bru¨ ssow, J Rátky and H Rodriguez-Martinezinflammatory-like process initiated by a surge in gonadotropichormones, followed by structural changes in thefollicular wall and vascular microcirculation, consequenceof the action of proteolytic enzymes andprostaglandins (Espey and Lipner 1994). At ovulation,oocytes at the metaphase II stage are picked up by thecilia-covered fimbria and guided through the infundibulumand ampulla (Oxenreider and Day 1965; Alanko1974). Thereby oocytes with their cumulus vestmentaggregate within an ‘egg plug’ (Tanabe et al. 1949;Spalding et al. 1955). The rate of ovum transporttowards the AIJ is rapid (30–45 min, Andersen 1927;Alanko 1965), with the initial speed of ovum transportcalculated as 76 mm ⁄ h (Bru¨ ssow 1985). It is still notclear how ovum pick-up and transport are regulated, asCOCs are conveyed against a flow of tubal fluid.Possibly, this is accomplished by a concerted interactionbetween cumulus cells and the extracellular matrix theysynthesize (mainly HA) which, by dramatically increasingthe size of the egg plug, makes ciliary beating(Talbot et al. 1999), as well as myosalpingeal peristalsis(Rodriguez-Martinez et al. 1982) very effective in movingthe ova to the AIJ. Cumulus cells disperse within 6 hpost-ovulation (Hunter and Dziuk 1968; Hunter 1974;Bru¨ ssow et al. 1987); a process which is stronglyinfluenced by the presence of spermatozoa in the oviduct(Rodriguez-Martinez et al. 2001).The fertilization process within the AIJ includes thebinding of spermatozoa to the glycoprotein covering ofthe oocyte, the ZP-penetration, and finally the fusionwith the oolemma. This generates the cortical reactionand subsequent a ZP-hardening to prevent polyspermicpenetrations, a very effective system in vivo, but still aserious constrain in vitro (Funahashi et al. 2000, 2001).Gamete recognition, binding and fusion are highlyregulated processes involving a number of biochemicalevents (Jansen et al. 2001; Miller and Burkin 2001; Rathet al. 2006). The chronological and cytological details offertilization and early embryonic development in the pighave been thoroughly described by Hunter (1974). Earlyembryo development up to the 4-cell stage occurs withinthe Fallopian tube. Fertilized embryos (appearance oftwo pronuclei) can be observed approximately 5–8 hafter ovulation in the AIJ and isthmus and remain for alonger time (up to 26 h) in the one-cell stage. The 2-cellembryo stage is of short duration (6–8 h) and withinfurther 6–8 h they develop into 4-cell embryos. Earlycleavage of embryos occurs in the isthmus, and mostlyat the 4-cell stage they leave the oviduct 50–60 h postovulation(Hancock 1961; Alanko 1974; Hunter 1974;Bru¨ ssow 1985). Blastulation is impeded in the oviduct(Oxenreider and Day 1965; Rodriguez-Martinez et al.1985), for reasons still unknown.Data on the transport of fertilized and non-fertilizedova through the porcine oviduct are scarce and originateonly from experiments with either inseminated or noninseminatedsows. Some influence of fertilized embryoson oviductal transport was observed (Bru¨ ssow andRa´tky 1996), contrary to other authors (Mwanza et al.2002). However, other factors may influence ovumtransport. As such, PMSG-induced superovulation(SO) alters the distribution of embryos in the oviductwith retardation in the ampulla on d1 after hCGinduced ovulation and an accelerated transport on d3(Bru¨ ssow et al. 1987). It remains unclear if this alterationis due to the higher number of oocytes releasedand ⁄ or an altered hormonal environment in the oviductafter SO. Stress-induced alteration of cortisol levels byfood deprivation (Mburu et al. 1998) or exogenous-ACTH administration (Razdan et al. 2002; Brandt et al.2007) influenced embryo development and distributionwithin the oviduct, as well as disturbed sperm transport,estimated on accessory sperm count (Brandt et al. 2006).Contribution of Oviductal Fluid on SpermCapacitation, Fertilization and EmbryoDevelopmentOviductal fluid (ODF) is formed by selective transudationfrom the blood and specific secretion from theoviduct epithelium; it differs from blood plasma interms of ionic composition, pH, osmolarity and macromolecularcontent, and it varies with the endocrinestatus during the oestrous cycle (Leese 1988; Leese et al.2001). Two areas important to prepare sperm forfertilization, such as sperm quiescence (ensuring spermatozoaare kept alive and potentially fertile) andcapacitation (a destabilizing process, particularly relatedto the apical sperm plasma membrane; see reviewby Rodriguez-Martinez 2007). They are subscribed tobe influenced by the ODF; namely the presence ofdifferent pH environments and the presence of GAGs.For the first named, data from our laboratories indicatethe SR possesses (at least pre-ovulation) an acidic pH,with low levels of local bicarbonate (Rodriguez-Martinez2007), which might be responsible for maintainingsperm quiescence and which prevents sperm capacitationin the SR shortly before ovulation (Mburu et al.1996; Tienthai et al. 2004). Our current results suggest asignificant proportion of boar spermatozoa retrievedfrom the pre-ovulatory SR and exposed to homologousODF of either pre-, peri- or post-ovulatory oestroussows only capacitate, when exposed to post-ovulatoryODF (Tienthai et al. 2004). Capacitation is, however,triggered if these SR-spermatozoa were exposed tobicarbonate (at levels recorded to be present in theAIJ ⁄ ampulla in vivo, e.g. 33–35 mM; Tienthai et al.2004). Bicarbonate is, therefore, considered a specificeffector of the process not only in pigs but also in otherspecies, such as the bovine (Bergqvist et al. 2006). Suchexposure induces sperm membrane destabilization and,ultimately, primes them to acrosomal exocytosis, asdemonstrated in vitro, and expected in vivo. It istherefore postulated that the constant release of individualspermatozoa out of the SR is sufficient to inducetheir capacitation when adequate levels of the effectorare encountered outside of the SR area (Rodriguez-Martinez et al. 2005), thus marking capacitation as aperiovulatory process in vivo (Hunter and Rodriguez-Martinez 2004).The pig ODF contains GAGs, either non-sulphated(HA) or sulphated (S-GAGs, e.g. chondroitin sulphate,dermatan sulphate, keratan sulphate, heparan sulphateand heparin) (Tienthai et al. 2001; Buhi 2002). Themean concentrations of total S-GAGs in ODF differbetween isthmus and ampulla and vary also in relationÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Fertilization in the Porcine Fallopian Tube 249to the time of the oestrous cycle, being higher in theisthmus than in the ampulla, probably owing to thelarger secretory capacity of the latter (Tienthai et al.2001). The S-GAG-levels increase significantly in isthmusduring the pre-ovulatory oestrus, to decreasetowards metoestrus, occurring in both tubae, probablydue to the bilateral ovarian activity in this species. Thenon-sulphated GAG HA is also present in the porcineODF and without segmental differences, but with atendency to increase during standing oestrus, becominghighest around ovulation. Especially in the ampullaremains high during metoestrus (Tienthai et al. 2001).Both hyaluronan synthases, hyaluronan-binding proteinsand specific membrane receptors are present in theepithelial lining, particularly in the SR (Tienthai et al.2003a,b), where mucus accumulates before ovulation(Johansson et al. 2000). As mentioned before, GAGs,HA in particular, seem to be involved in sperm survival,capacitation and binding to and release from the SR (seeRodriguez-Martinez 2007).In-vitro development of porcine oocytes hampers stillfrom normal development as indicated by polyspermicfertilization. Apparently it is caused by insufficientmaturation with a deficient formation of cortical granulesand results in constrained ZP-reaction after firstsperm penetration. Such situation was not seen when weexamined in vivo-matured or in vivo-fertilized oocytes,suggesting that either the final maturation inside the preovulatoryfollicle (Bru¨ ssow, unpublished observations)or inside the oviduct is needed for full competence of thenewly ovulated oocytes. The latter seems most relevantconsidering the final maturation of the ZP, as establishedby Funahashi et al. (2000, 2001), implicating theODF involvement in these processes.The ODF is protein-rich and contains several oviductspecific-proteins(OSP) (Buhi et al. 1997; Buhi 2002;Buhi and Alvarez 2003). It supports fertilization andearly embryo development. For example, Wollenhauptand Bru¨ ssow (1995) described a 97 kDa OSP which waspre-dominantly present on days 1–3 after ovulation(6.8–10.3% of the total oviductal protein), and onlyattached to oviduct-derived embryos but not to intrafollicularoocytes or in vitro-derived embryos (Bru¨ ssowet al. 1998b). A contribution of this OSP on embryodevelopment was demonstrated in vitro. Supplementationof the 97 kDa protein increased de novo proteinsynthesis in in vivo matured embryos (Wollenhaupt et al.1997), reduced polyspermic penetration (Kouba et al.2000), and increased cleavage and blastocyst rate(McCauley et al. 2003).However, not only the ODF influences the fate ofgametes in the oviduct, but also oocytes and spermatozoaalter the oviductal secretory proteomic profile. Afterin vitro incubation of porcine oviducts either with boarspermatozoa or COCs, altogether 34 proteins wereregulated by the presence of spermatozoa (n = 20),oocytes (n = 5) or of both gametes (n = 9). Of theseproteins, 18 were up-regulated and 16 were downregulated(Georgiou et al. 2005). With respect to functionalcategories, gene-regulation related to proteinproduction, maintenance and repair (41%), antioxidantsand radical scavengers (18%), metabolism (15%) andmiscellaneous (25%). These changes seem to provide afavourable microenvironment for gametes and preparethe oviduct milieu for embryo arrival.Concluding RemarksAlthough numerous interactions between the Fallopiantube and gametes in pigs are recognized, our knowledgeis yet limited. Attempts have to be made to elucidate thefine tuning of intra-follicular oocyte development andovulation with sperm release and capacitation. Therelationship between gametes within the oviduct andoviductal secretion, and subsequent embryo developmentneeds further research. Increased knowledge aboutthe complex and dynamic processes within the Fallopiantube should additionally improve assisted techniquesincluding in vitro embryo production in swine.AcknowledgementsThe studies of the authors have been made possible by grants from theGerman BMELV, Hungarian OTKA, FORMAS and the SwedishFarmer’s Foundation for Agricultural Research (SLF), Stockholm.ReferencesAlanko M, 1965: The site of recovery and cleavage rate of pigova from the tuba uterina. Nord Vet Med 17, 323–327.Alanko M, 1974: Fertilization and early development of ova inAI-gilts, with special reference to role of tubal spermconcentration. A clinical and experimental study. PhDThesis, University of Helsinki.Andersen D, 1927: The rate of passage of the mammalianovum through various portions of the Fallopian tube. Am JPhysiol 82, 557–569.Austin CR, Walton A, 1960: Fertilisation. In: Parkes AS (ed.),Marshal’s Physiology of Reproduction. Longmans Green,London, 1-Part 2, pp. 310–-416.Bedford JM, Kim HH, 1985: Cumulus oophorus as a spermsequestering device, in vivo. J Exp Zool 265, 321–328.Bergqvist AS, Ballester J, Johannisson A, Herna´ndez M,Lundeheim N, Rodrı´guez-Martı´nez H, 2006: In vitrocapacitation of bull spermatozoa by oviductal fluid and itscomponents. Zygote 14, 259–273.Blo¨ dow G, Bergfeld J, Kitzig M, Bru¨ ssow K-P, 1990: Steroidhormone levels in follicular fluid of pigs with spontaneousoestrus and synchronised ovulation. Arch Exp Vet Med 44,611–620.Brandt Y, Lang A, Madej A, Rodriguez-Martinez H, EinarssonS, 2006: Impact of ACTH administration on theoviductal sperm reservoir in sows: the local endocrineenvironment and distribution of spermatozoa. AnimReprod Sci 92, 107–122.Brandt Y, Madej A, Rodriguez-Martinez H, Einarsson S,2007: Effects of exogenous ACTH during oestrus on earlyembryo development and oviductal transport in the sow.Reprod Domest Anim 42, 118–125.Bru¨ ssow K-P, 1985: Distribution of oocytes in oviduct of giltsfollowing synchronised ovulation. Mh Vet-Med 40, 264–268.Bru¨ ssow K-P, Rátky J, 1996: Distribution of ova within theoviduct of gilts after ovulation. Reprod Domest Anim 31,305–306.Bru¨ ssow K-P, Kauffold M, Bergfeld J, 1987: Effects ofdifferent PMSG doses on ovarian response as well as ondistribution and quality of oocytes in oviduct of giltsfollowing synchronization of ovulation. Mh Vet-Med 42,764–768.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

250 K-P Bru¨ ssow, J Rátky and H Rodriguez-MartinezBru¨ ssow K-P, Rátky J, Schneider F, Torner H, Kanitz W,Solti L, 1998a: Does follicular fluid have any importance inthe transfer of porcine oocytes into the oviduct. Theriogenology49, 340.Bru¨ ssow K-P, Alm H, Rátky J, Wollenhaupt K, Torner H1998b: The influence of gonadotropins on porcine folliculargrowth and early embryonic development. Proc 49th EAAPMeeting, Warsaw, 1998, MPhP4.4, p. 127.Bru¨ ssow K-P, Rátky J, Schneider F, Torner H, Kanitz W,Solti L, 1999a: Effects of follicular fluid on the transport ofporcine oocytes into the oviduct at ovulation. ReprodDomest Anim 34, 423–429.Bru¨ ssow K-P, Rátky J, Torner H, Sarlo´s P, Solti L, 1999b:Contribution of porcine follicular fluid in the process offertilization in vivo. Reprod Domest Anim 34, 139–145.Bru¨ ssow K-P, Rátky J, Torner H, Solti L, 2003: Possible roleof follicular fluid on porcine in vivo fertilization. In: Sato E,Miyamoto H, Manabe N (eds), Animal Frontiers. HokutoPub Co. Ltd, Kyoto, pp. 89–95.Bru¨ ssow K-P, Torner H, Rátky J, Manabe N, Tuchscherer A,2006: Experimental evidence for the influence of cumulusoocyte-complexeson sperm release from the porcine oviductalsperm reservoir. J Reprod Dev 52, 249–257.Buhi WC, 2002: Characterization and biological roles ofoviduct-specific, oestrogen-dependent glycoprotein. Reproduction123, 355–362.Buhi WC, Alvarez IM, 2003: Identification, characterizationand localization of three proteins expressed by the porcineoviduct. Therogenology 60, 225–238.Buhi WC, Alvarez IM, Kouba AJ, 1997: Oviductal regulationof fertilization and early embryonic development. J ReprodFertil 52, 285–300.Eiler H, Nalbandov AV, 1977: Sex steroids in the follicularfluid and blood plasma during the estrous cycle of pigs.Endocrinology 100, 331–338.Espey LL, Lipner H, 1994: Ovulation. In: Knobil E, Neill JD(eds), The Physiology of Reproduction. Raven Press, NewYork, pp. 725–780.Fazeli A, Duncan AE, Watson PF, Holt WV, 1999:Sperm-oviduct interaction: induction of capacitation andpreferential binding of uncapacitated spermatozoa to oviductalepithelial cells in porcine species. Biol Reprod 60,879–86.Funahashi H, Ekwall H, Rodriguez-Martinez H, 2000: Thezona reaction in porcine oocytes fertilized in vivo and in vitroas seen with scanning electron microscopy. Biol Reprod 63,1437–1442.Funahashi H, Ekwall H, Kikuchi K, Rodriguez-Martinez H,2001: Transmission electron microscopy studies of the zonareaction in pig oocytes fertilised in vivo and in vitro.Reproduction 122, 443–452.Georgiou AS, Sostaric E, Wong CH, Snijders APL, WrightPC, Moore HD, Fazeli A, 2005: Gametes alter theoviductal secretory proteome. Mol Cell Proteomics 4,1785–1796.Hancock JL, 1961: Fertilization in the pig. J Reprod Fertil 2,307–331.Hansen C, Srikandakumar A, Downey B, 1991: Presence offollicular fluid in the porcine oviduct and its contribution tothe acrosome reaction. Mol Reprod Dev 30, 148–153.Harper MJK, 1973: Stimulation of sperm movement from theisthmus to the site of fertilization in the rabbit oviduct. BiolReprod 8, 369–377.Hunter RHF, 1972: Local action of progesterone leading topolyspermic fertilization in pigs. J Reprod Fertil 31, 433–444.Hunter RHF, 1974: Chronological and cytological details offertilization and early embryonic development in thedomestic pig, Sus scrofa. Anat Rec 178, 169–185.Hunter RHF, 1984: Preovulatory arrest and periovulatoryredistribution of competent spermatozoa in the isthmus ofthe pig oviduct. J Reprod Fertil 72, 203–211.Hunter RHF, 1990: Fertilization of pig eggs in vivo and invitro. J Reprod Fertil 40, 211–226.Hunter RHF, 1993: Sperm:egg ratios and putative molecularsignals to modulate gamete interactions in polytocousmammals. Mol Reprod Dev 35, 324–327.Hunter RHF, 1996: Ovarian control of very low sperm ⁄ eggratios at the commencement of mammalian fertilization toavoid polyspermy. Mol Reprod Dev 44, 417–422.Hunter RHF, Dziuk PJ, 1968: Sperm penetration of pig eggsin relation to the timing of ovulation and insemination. JReprod Fertil 15, 199–208.Hunter RHF, Nichol R, 1986: A preovulatory temperaturegradient between the isthmus and ampulla of pig oviductsduring the phase of sperm storage. Reprod Fertil 77, 599–606.Hunter RHF, Rodriguez-Martinez H, 2004: Capacitation ofmammalian spermatozoa in vivo, with a specific focus onevents in the Fallopian tubes. Mol Reprod Dev 67, 243–250.Hunter RHF, Cook B, Poyser NL, 1983: Regulation ofoviduct function in pigs by local transfer of ovarian steroidsand prostaglandins: a mechanism to influence sperm transport.Eur J Obstet Gynecol Reprod Biol 14, 225–232.Hunter RHF, Petersen HH, Greve T, 1999: Ovarian follicularfluid, progesterone and Ca 2+ ion influences on sperm releasefrom the Fallopian tube reservoir. Mol Reprod Dev 54, 283–291.Jansen S, Ekhlasi-Hundrieser M, Toepfer-Petersen E, 2001:Sperm adhesion molecules: Structure and function. CellTissue Organs 168, 82–92.Johansson M, Tienthai P, Rodrı´guez-Martı´nez H, 2000:Histochemistry and ultrastructure of the intraluminal mucusin the sperm reservoir of the pig oviduct. J Reprod Dev 46,183–192.Kimura N, Konno Y, Miyoshi K, Matsumoto H, Sato E,2002: Expression of hyaluronan synthases and CD44 messengerRNAs in porcine cumulus-oocyte complexes duringin vitro maturation. Biol Reprod 66, 707–717.Kouba AJ, Abeydeera LR, Alvarez IM, Day BN, Buhi WC,2000: Effects of the porcine oviduct-specific glycoprotein onfertilization, polyspermy, and embryonic development invitro. Biol Reprod 63, 242–250.Leese HJ, 1988: The formation and function of oviduct fluid. JReprod Fertil 82, 843–856.Leese HJ, Tay JI, Reischl J, Downing SJ, 2001: Formation ofFallopian tubal fluid: role of a neglected epithelium.Reproduction 121, 339–346.Liberda J, Manaskova P, Prelovska L, Ticha M, Jonakova V,2006: Saccharide-mediated interactions of boar sperm surfaceproteins with components of the porcine oviduct. JReprod Immunol 71, 112–25.Mburu JN, Einarsson S, Lundeheim N, Rodriguez-MartinezH, 1996: Distribution, number and membrane integrity ofspermatozoa in the pig oviduct in relation to spontaneousovulation. Anim Reprod Sci 45, 109–121.Mburu JN, Rodriguez-Martinez H, Einarsson S, 1997:Changes in sperm ultrastructure and localization in theporcine oviduct around ovulation. Anim Reprod Sci 47,137–148.Mburu JN, Einarsson S, Kindahl H, Madej A, Rodriguez-Martinez H, 1998: Effects of postovulatory food deprivationon oviductal sperm concentration, embryo development andhormonal profiles in the pig. Anim Reprod Sci 52, 221–234.McCauley TC, Buhi WC, Wu GM, Mao J, Caamano JN,Didion BA, Day BN, 2003: Oviduct-specific glycoproteinmodulates sperm-zona binding and improves efficiency ofporcine fertilization in vitro. Biol Reprod 69, 828–834.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Fertilization in the Porcine Fallopian Tube 251Miller DJ, Burkin HR, 2001: Gamete adhesion molecules.Reproduction 58, 147–158.Mwanza M, Razdan P, Hulte´n F, Einarsson S, 2002: Transportof fertilized and unfertilized ova in sows. Anim ReprodSci 74, 69–74.Oxenreider SL, Day BN, 1965: Transport and cleavage of ovain swine. J Anim Sci 24, 413–417.Rath D, To¨ pfer-Petersen E, Michelmann H-W, Schwartz P,von Witzendorff D, Ebeling S, Ekhlasi-Hundrieser M,Piehler E, Petrunkina A, Romar R, 2006: Structural,biochemical and functional aspects of sperm–oocyte interactionsin pigs. In: Ashworth CJ, Kraeling RR (eds),Control of Pig Reproduction VII, SRF Suppl. 62. NottinghamUniversity Press, Nottingham, pp. 317–330.Razdan P, Mwanza AM, Kindahl H, Rodriguez-Martinez H,Hultén F, Einarsson S, 2002: Effect of repeated ACTHstimulationon early embryonic development and hormonalprofiles in sows. Anim Reprod Sci 70, 127–137.Rodriguez-Martinez H, 2007: Role of the oviduct in spermcapacitation. Theriogenology 68S, S138–S146.Rodriguez-Martinez H, Einarsson S, Larsson B, 1982: Spontaneousmotility of the oviduct in the anaesthetized pig. JReprod Fertil 66, 615–624.Rodriguez-Martinez H, Larsson B, Einarsson S, 1985: Ovatransport and fertility after resection of the oviductalisthmus in pigs. Acta Vet Scand 26, 218–230.Rodriguez-Martinez H, Nicander L, Viring S, Einarsson S,Larsson K, 1990: Ultrastructure of the uterotubal junctionin preovulatory pigs. Anat Histol Embryol 19, 16–36.Rodriguez-Martinez H, Ekstedt E, Ridderstra˚ le Y, 1991:Histochemical localization of carbonic anhydrase in thefemale genitalia of pigs during the oestrous cycle. Acta Anat140, 41–47.Rodriguez-Martinez H, Larsson B, Pertoft H, Kjelle´n L, 1998:GAG’s and spermatozoon competence in vivo and in vitro.In: Lauria A, Gandolfi F, Enne G, Gianaroli L (eds),Gametes: Development and Function. Serono Symposium,Serono, pp. 239–272.Rodriguez-Martinez H, Tienthai P, Suzuki K, Funahashi H,Ekwall H, Johannisson A 2001: Oviduct involvement insperm capacitation and oocyte development. Reproduction(Suppl. 58), 129–145.Rodriguez-Martinez H, Saravia F, Wallgren M, Tienthai P,Johannisson A, Vázquez JM, Martı´nez E, Roca J, Sanz L,Calvete JJ, 2005: Boar spermatozoa in the oviduct. Theriogenology63, 514–535.Spalding JF, Berry RO, Moffit JG, 1955: The maturationprocess of the ovum of swine during normal and inducedovulations. J Anim Sci 14, 609–620.Talbot P, Geiske C, Knoll M, 1999: Oocyte pickup by themammalian oviduct. Mol Biol Cell 10, 5–8.Tanabe TY, Warnick AC, Casida LE, Grummer RH, 1949:The effects of gonadotrophins administered to sows and giltsduring different stages of the estrual cycle. J Anim Sci 8,550–557.Tienthai P, Kjellen L, Pertoft H, Suzuki K, Rodriguez-Martinez H, 2000a: Localization and quantification ofhyaluronan and sulfated glycosaminoglycans in the tissueand intraluminal fluid of the pig oviduct. Reprod Fertil Dev12, 173–182.Tienthai P, Suzuki K, Pertoft H, Kjellen L, Rodriguez-Martinez H, 2000b: Production of glycosaminoglycans bythe porcine oviduct in relation to sperm storage. ReprodDomest Anim 35, 167–170.Tienthai P, Kjellen L, Pertoft H, Suzuki K, Rodriguez-Martinez H, 2001: Localization and quantitation of hyaluronanand sulfated glycosaminoglycans in the tissues andintraluminal fluid of the pig oviduct. Reprod Fertil Dev 12,173–182.Tienthai P, Kimura N, Heldin P, Sato E, Rodrı´guez-Martı´nezH, 2003a: Expression of hyaluronan synthase-3 in porcineoviductal epithelium during oestrus. Reprod Fertil Dev 15,99–105.Tienthai P, Yokoo M, Kimura N, Heldin P, Sato E,Rodriguez-Martinez H, 2003b: Immunohistochemical localizationand expression of the hyaluronan receptor CD44 inthe porcine oviductal epithelium during oestrus. Reproduction125, 119–132.Tienthai P, Johannisson A, Rodriguez-Martinez H, 2004:Sperm capacitation in the porcine oviduct. Anim Reprod Sci80, 131–146.Wollenhaupt K, Bru¨ ssow K-P, 1995: Isolation of a 97 kdporcine oviductal secretory protein using a high-performanceelectrophoresis-chromatograph (HPEC) system. ReprodDomest Anim 30, 1–7.Wollenhaupt K, Alm H, Tomek W, Bru¨ ssow K-P, 1997:Studies of the influence of a specific 97-kd porcine oviductalsecretory protein on the de novo protein synthesis of preimplantationembryos. Reprod Domest Anim 32, 213–219.Yaniz JL, Lopez-Gatius F, Hunter RHF, 2006: Scanningelectron microscopic study of the functional anatomy of theporcine oviductal mucosa. Anat Histol Embryol 35, 28–34.Yokoo M, Miyahayashi Y, Naganuma T, Kimura N, SasadaH, Sato E, 2002: Identification of hyaluronic acid-bindingproteins and their expressions in porcine cumulus-oocytecomplexes during in vitro maturation. Biol Reprod 67, 1165–1171.Yokoo M, Shimizu T, Kimura N, Tunjung WA, MatsumotoH, Abe H, Sasada H, Rodriguez-Martinez H, Sato E, 2007:Role of the Hyaluronan receptor DC44 during porcineoocyte maturation. J Reprod Dev 53, 263–270.Author’s address (for correspondence): K-P Bru¨ ssow, FBN ResearchInstitute for the Biology of Farm Animals, Wilhelm-Stahl-Allee 2,18196 Dummerstorf, Germany. E-mail: bruessow@fbn-dummerstorf.deConflict of interest: Grants for K-P Bru¨ ssow were provided byGerman Federal Ministry of Nourishment, Agriculture and ConsumersProtection (BMELV) Project 2/05; J Rátky declares thatresearch was partly supported by Hungarian OTKA; H R-M declaresno conflict of interest.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 252–259 (2008); doi: 10.1111/j.1439-0531.2008.01170.xISSN 0936-6768Mastitis in Post-Partum Dairy CowsS Pyo¨räläDepartment of Production Animal Medicine, Faculty of Veterinary Medicine, University of Helsinki, Saarentaus, FinlandContentsTransition from the dry period to lactation is a high riskperiod for the modern dairy cow. The biggest challenge atthat time is mastitis. Environmental bacteria are the mostproblematic pathogens around parturition. Coliforms areable to cause severe infections in multiparous cows, andheifers are likely to be infected with coagulase-negativestaphylococci. During the periparturient period, hormonaland other factors make the dairy cows more or lessimmunocompromised. A successful mastitis control programmeis focused on the management of dry and calvingcows and heifers. Clean and comfortable environment, properfeeding and adequate supplementation of the diet withvitamins and trace elements are essential for maintaininggood udder health. Strategies which would enhance closure ofthe teat canal in the beginning of the dry period and wouldprotect teat end from bacteria until the keratin plug hasformed decrease the risk for mastitis after calving. Dry cowtherapy has been used with considerable success. Yet, aselective approach could be recommended rather than blankettherapy. Non-antibiotic approaches can be useful tools toprevent new infections during the dry period, in herds wherethe risk for environmental mastitis is high. Vaccination hasbeen suggested as a means to support the immune defence ofthe dairy cow around parturition. In some countries, implementationof Escherichia coli core antigen vaccine hasreduced the incidence of severe coliform mastitis aftercalving.IntroductionRate of bovine intramammary infections (IMI) is at thehighest level around parturition (Fig. 1) (Smith et al.1985; Bradley and Green 2004; Valde et al. 2004;McDougall et al. 2007). The factors influencing thesusceptibility of the mammary gland to infections arepresence of bacteria at the teat end, level of efficacy ofthe protective characteristics of the teat canal anddefence mechanisms in the udder (Sordillo 2005).During the dry period, the mammary gland is consideredto be very resistant against infections. The streakcanal is closed and sealed by a keratin plug. In the fullyinvoluted gland, the concentrations of many solublefactors are at a high level and effective in preventing newinfections. Concentration of leucocytes is high in the drygland, and the environment is more favourable for theirfunction than in the lactating gland (Burvenich et al.2007). The recent decades have shown considerableprogress in the understanding of the function of thedefence system of the bovine mammary gland. Despitethis, the inefficient host defence and increased susceptibilityto mastitis during the transition period continue tobe the major problems in the dairy cows (Leslie andDingwell 2002). In this article, different factors affectingthe susceptibility of the cow to mastitis aroundparturition, as well as measures for prevention ofmastitis, are reviewed.The Effect of the Dry PeriodDuring involution and again towards the end of thedry period, the risk for mastitis is at the highest (Oliverand Sordillo 1988). After drying-off, milk is no longerremoved from the udder, and intramammary pressuremay cause leakage of milk from the teats. Leucocytesstart entering the gland within 1 week after dry-off, butdo not immediately protect the gland. The keratinplug, which also contains inhibitory substances againstbacteria, is formed within 1–2 weeks after dry-off andshould naturally seal the teat (Dingwell et al. 2004).Quarters that form a keratin plug, which completelycloses the teat soon after dry-off, have significantly lessrisk to develop an IMI. Yet, as many as 23% of thequarters have been found to be still open 6 weeks afterdrying-off (Dingwell et al. 2004). In an earlier study,5% of the teats were found to remain completely open(Williamson et al. 1995). Increasing milk yield at thedry-off has been recognized as a significant risk factor:every 5 kg increase in milk yield at dry-off above12.5 kg increased the odds of the cow by 77% to havean IMI caused by environmental bacteria at calving(Rajala-Schultz et al. 2005). The increased susceptibilityto mastitis with increasing milk yield at drying-off isprobably related to the incomplete closing of the teatcanal (Dingwell et al. 2004). Teat end condition alsoaffects mastitis susceptibility: teats with cracked endshave higher odds of developing IMI around calving(Dingwell et al. 2004).Mammary Gland Immunity Around ParturitionFrom 2 weeks prior to calving, until about 2–3 weeksafter calving, is the most critical period for the health ofthe mammary gland (Oliver and Sordillo 1988). Theinnate immune system of the periparturient cows iscompromised. During colostrogenesis, the susceptibilityof the mammary gland to infections increases as the teatcanal starts to open and leaks mammary secretion(Oliver and Sordillo 1988). At the same time, theprotective effect of dry cow therapy (DCT), if used,has disappeared (Oliver et al. 1990). Hormonal changesinclude steep rise of the concentration of 17b-oestradiolin the plasma during the last week of gestation, peakingduring the last days before parturition and a simultaneousdrop of progesterone in circulation. Blood cortisolincreases about fivefold at the day of parturition(Burton et al. 2005).Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Mastitis in Post-Partum Dairy Cows 253No of cows700600500400300200100Fig. 1. Distribution of the timefrom calving (days in milk) to thediagnosis of clinical mastitis, basedon data from 28 dairy herds inNew Zealand (McDougall et al.2007)0–7 7 21 35 49 63 77 91 105 119 133 147 161 175 189 203 217 231 246Time from calvingCalvingCellular FactorsPolymorphonuclear neutrophils (PMN) belong to theinnate immune response of the mammary gland toinfections (Paape et al. 2003; Burvenich et al. 2007).Around parturition, many functions of the PMN areimpaired. The number of immature neutrophils incirculation increases and the number of mature neutrophilsin the blood and milk are at the lowest. Theproduction of reactive oxygen species (ROS) to killbacteria is reduced from 1 week before parturition overthe first 2 weeks after calving (Hoeben et al. 2000;Mehrzad et al. 2002). The change in the respiratoryburst activity has been found to be parallel withperipartum increase of concentrations of 3b-hydroxybutyricacid (3-BHB), bovine pregnancy-associated glycoproteinand bilirubin (Hoeben et al. 2000). The rapidrise of the concentration of blood cortisol induceschanges in the function of PMN, supporting theirextended life span in the blood and increasing releasefrom the bone marrow. At that time, neutrophils favourtissue remodelling over defence against infections astheir primary task (Burton et al. 2005). At parturition,large numbers of leucocytes are recruited to the reproductivetract and placenta. The ability of the cells tomarginate on and migrate through endothelium toinfected peripheral tissue in other sites than uterusdecreases. After parturition, cortisol down-regulates itsown receptors in neutrophils and the system returns tonormal function (Burton et al. 2005). The effect ofsteroid hormones on bovine PMN function was studiedby Lamote et al. (2004), who showed that 17b-oestradioltreatment decreased the number of viable cells butprogesterone had no effect. In periparturient cows, a lossof expression of critical neutrophil adhesion moleculeshas been seen (Monfardini et al. 2002), and this loss hasbeen associated with the elevated cortisol levels (Weberet al. 2004). The same was demonstrated by externalglucocorticoid administration (Burton et al. 2005). Theproportion of PMN expressing, for example, the adhesionreceptor L-selectin, which is necessary for penetrationto the sites of infection, is diminished (Diez-Fraileet al. 2004). Neutrophil extracellular traps have recentlyshown to have a role in killing of bacteria and also to befully capable to function in the milk environment. Theimpaired efficiency of this system during the periparturientperiod may be one more explanation for theimmunosuppression of the dairy cows at that time(Lippolis et al. 2006).Lymphocytes can recognize antigens through specificreceptors and are divided to T and B lymphocytes.CD4+ T lymphocytes cells activate lymphocytes ormacrophages to secrete cytokines, which then canfacilitate either cell-mediated or humoral immuneresponse (Sordillo 2005; Rainard and Riollet 2006).The proportion of CD4+ cells in blood and mammarygland declines post-partum, and their cytokine productionis different from that in mid-lactating cows. Inperiparturient cows, the percentage of T cells has beenshown to substantially decline from that in mid-lactatingcows (Shafer-Weaver et al. 1996). Macrophages are thedominant cell type in milk of healthy, lactating gland.During infection, macrophages initiate the immuneresponse by releasing cytokines and other substancesaugmenting local inflammatory process (Rainard andRiollet 2006). Bovine macrophage numbers are highestin the mammary gland during the last week of gestation,but their phagocytic capacity is decreased (Sordillo2005).Humoral FactorsInnate and specific soluble factors represent an importantpart of the defence in the mammary gland;complement, lactoferrin, lysozyme and antimicrobialpeptides are the most common (Rainard and Riollet2006). Lactoferrin is most active during the steady stateof involution (Smith and Schanbacher 1977). Duringthat time, lactoferrin efficiently prevents growth ofbacteria with a high demand of iron such as coliforms(Todhunter et al. 1991). Complement is present in highconcentrations in colostrum, and seems not to be alimiting factor in the defence of the mammary gland atthat time (Rainard 2003). The most important factors ofthe specific immune response are opsonizing immunoglobulins(Ig) produced by antigen-activated B lymphocytes.IgG 1 is the primary isotype present in the healthymammary gland, but IgG 2 increases during inflammation.The concentration of IgGs in the bovine serum islower around parturition and in particular, the lack ofÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

254 S Pyo¨ra¨läthe IgG 2 isotype is associated with the increasedindicence of mastitis (Mallard et al. 1998).Dry Cow TherapyUse of DCT has an impact on the incidence of puerperalmastitis in two ways: first DCT should eliminateinfections present in the mammary gland at dry-offand thus prevent their flare-ups at calving; the secondaim is to protect the mammary gland for new IMIsduring the dry period (Robert et al. 2006). Antimicrobialsubstances used for DCT do not generally persist inthe udder until calving and thus do not offer protectionat that time (Oliver et al. 1990). Yet, in an UK study,reduction of clinical mastitis caused by Gram-negativeagents in the subsequent lactation was demonstratedwith a long-acting intramammary preparation withGram-negative spectrum (Bradley and Green 2001).These are the findings to date to demonstrate thatselection of dry cow treatment can influence theincidence of clinical mastitis in the subsequent lactation.During the dry period, 8–12% of previously healthyquarters develop IMI which can be detected at calving ifno DCT is administered (Leslie and Dingwell 2003).New IMIs which occur around parturition may greatlyimpact production in the subsequent lactation (Oliverand Sordillo 1988; Whist et al. 2006). Blanket DCTcontinues to be the standard procedure in most countries(Leslie and Dingwell 2003). Routine treatment ofall cows has recently been questioned, since bulk milksomatic cell counts (SCC) have markedly decreased andmastitis has changed from contagious to environmental(Leslie and Dingwell 2003). In some countries, especiallyin Scandinavia, blanket DCT has never been adaptedbut selective DCT therapy recommended (Osteras et al.1999; Robert et al. 2006). Dry cow therapy does notnecessarily protect cows from mastitis as according to aCanadian study, as much as 11% of the treated cowshad new IMI after the dry period (Dingwell et al. 2004).Pre-partum intramammary antibiotic therapy for heifershas been suggested to reduce CNS mastitis during firstlactation (Oliver et al. 2003; Middleton et al. 2005). In arecent study (Borm et al. 2006), no advantage from thispractice could be shown. Non-antibiotic internal teatsealants have become widely used, and have proven tobe effective in prevention of new infections during thedry period (Huxley et al. 2002).Effect of the Metabolic StateThe most important metabolic disturbances occurringshortly after calving are milk fever, ketosis and abomasaldisplacement. Hypocalcaemia affects the digestivesystem and pre-disposes the cow to concomitant diseases.It may affect the teat end sphincter and thusincrease the risk for mastitis. Cows with periparturienthypocalcaemia are reported to have greater chance ofdeveloping coliform mastitis (Curtis et al. 1983). Negativeenergy balance and perhaps protein imbalances inearly lactation contribute to the impaired immunedefence (Spain and Scheer 2006). Disturbance in fatmetabolism and severe negative energy balance may leadto fatty liver and ketosis. Accumulation of fat in theliver disturbs production of humoral immune factorsand is also associated with decreased functional capacityof PMN (Zerbe et al. 2000). There is evidence for adecreased capacity for phagocytosis and killing ofbacteria in cows suffering from ketosis and fatty liver(Leslie et al. 2001). High concentrations of ketonebodies such as BHB and acetoacetate found at parturitionhave been shown to inhibit the proliferation ofhaematopoietic cells (Hoeben et al. 2000). Elevatedlevels of BHB were associated with increased incidenceof clinical mastitis during early lactation (Smith et al.1985; Huszenicza et al. 2004). The course of mastitis wassevere in all ketotic cows regardless of the chemotacticresponse before infection (Kremer et al. 1993b).Role of the Causing AgentBacteriological aetiology of mastitis during the puerperalperiod differs between countries, as well asbetween heifers and older cows. The biggest bacterialchallenge for the bovine udder at parturition comesfrom the environment of the cow (Oliver and Sordillo1988). Coliform bacteria appear to be a major problemfor the periparturient cow (Burvenich et al. 2007). In theUK, a high proportion of puerperal mastitis is caused bycoliform bacteria, mainly Escherichia coli (Bradley andGreen 2004). In a Canadian study, the prevalence ofIMIs at freshening was 34% and most new IMIs werecaused by environmental streptococci and coliformbacteria (Dingwell et al. 2004). Mastitis caused byStreptococcus uberis is most common during the dryperiod and in early lactation, and in some countries thepathogen is most frequently isolated after calving (Smithet al. 1985; McDougall et al. 2007). On the contrary toenvironmental pathogens, the incidence of mastitiscaused by the contagious pathogen Staphylococcusaureus has been found to increase towards later lactation(Sol et al. 2000; McDougall et al. 2007). In Norwegiansmall herds mostly kept in tie stalls, the pathogen mostcommonly isolated at calving from cows in the firstlactation was S. aureus (Waage et al. 1999). In manycountries, the most commonly isolated organism inprimiparous cows around parturition is CNS (Matthewset al. 1992; Myllys 1995; Taponen et al. 2006, 2007;Parker et al. 2007). CNS mastitis is usually mild(Taponen et al. 2006). Inflammatory reaction in theudder in mastitis caused by CNS was found to bestronger in early lactating cows as compared with laterlactation (Pyo¨ra¨la¨ and Pyo¨ra¨la¨ 1998).The Special Problem of Puerperal E. coliMastitisMuch research has been carried out on E. coli mastitis,which in early lactating cow is often associated withsevere clinical signs (Vandeputte-Van Messom et al.1993; Burvenich et al. 2003). Pathophysiology of coliformand staphylococcal mastitis has been shown to bedifferent, and these pathogens elicit a different type ofimmune response (Bannerman et al. 2004). A strongcytokine response and acute or peracute course of thedisease is typical for E. coli mastitis in early lactation(Burvenich et al. 2007). Inflammatory reaction of theÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Mastitis in Post-Partum Dairy Cows 255mammary gland to endotoxin was shown to be significantlymore severe in cows in early lactation than in thesame cows close to drying-off (Lehtolainen et al. 2003).The greater severity of coliform mastitis after calvinghas been explained by the dysfunction of PMN (Vandeputte-VanMessom et al. 1993). The number of circulatingPMN has been shown to correlate with theseverity of E. coli mastitis (Kremer et al. 1993a). Acutephase mediators produced during E. coli mastitis cantrigger an extensive ROS production which damageshost tissues; cells from periparturient dairy cattle havebeen shown to produce significantly more TNF-a thancells from mid-lactating cows (Sordillo et al. 1995).Multiparous cows have shown to develop more severeE. coli mastitis (Vangroenweghe et al. 2004) as comparedwith young cows. Blood PMN function seems tobe more efficient in young cows than in older cows(Burvenich et al. 2003). The viability and production ofROS of the milk neutrophils to kill bacteria was foundto be depressed in multiparous cows (Mehrzad et al.2002).In the UK, studies have shown that enterobacteria areable to infect the udder during the dry period and persistthere until parturition (Bradley and Green 2000). Thishas not been confirmed in other countries. For example,in a recent US study, the proportion of coliforms asmastitis causing agents was less than 1% at drying-offand after calving (Pantoja and Ruegg 2007). In the studyby Smith et al. (1985), a relatively high proportion ofudder infections at the beginning and at the end of thedry period was found to be due to Gram-negativebacteria. Dried manure with high counts of coliformbacteria was used as bedding, which may have affectedthe results.Diagnosis of Mastitis After CalvingUdder health of a dairy cow should be assessed as soonas possible after calving. Diagnosis of clinical mastitis inearly lactation must be adjusted according to thephysical properties of milk during that period. Somaticcell count is increased at parturition, but decreases tonormal levels within 3–4 days. The difference betweeninfected and healthy quarters is significant at both times(Barkema et al. 1999a). California mastitis test (CMT)was found to have sensitivity and specificity high enoughfor detecting IMI caused by major pathogens at day 3post-partum (Sargeant et al. 2001). Regular monitoringof the udder of the cows around calving is mostimportant, and milk should be examined as soon aspossible after calving (Green et al. 2007).Effect of Dry Cow Management on theOccurrence of Mastitis Post-PartumDry cow management, such as environment, feedingand timing of transfer to the calving unit, significantlyaffect the susceptibility of heifers or dairy cows topuerperal mastitis. Generally, the incidence of clinicalmastitis after calving increases with parity (Whist et al.2006; Green et al. 2007). Dry and transition dietsshould contain the recommended levels of vitamins andtrace elements (Spain and Scheer 2006). The mostcritical are vitamin E and selenium, but also vitamin A,copper and zinc have a role in the defence of the hostagainst infections (Godden et al. 2006). Decades ago,supplementation of dairy cows with selenium andvitamin E was shown to have a positive effect onudder health (Smith et al. 1984; Hogan et al. 1993).The effect has mostly been seen in herds fed withdeficient or low levels of these elements. In herds fedwith normal diets, daily vitamin E supplementationduring the puerperal period did not affect the incidenceof clinical mastitis or other puerperal diseases (Perssonet al. 2007).Dry matter intake should not be restricted during thedry and transition period, but overfeeding of dry cowsshould be avoided. Regular body condition scoring is agood tool in monitoring the efficiency of feeding. In arecent study, it was associated with less risk of clinicalmastitis (Green et al. 2007). In a Swedish study, factorswhich increased the risk of elevated SCC after calvingwere intensive concentrate feeding to heifers, andmoving to confined housing on the day of calvinginstead of earlier (Svensson et al. 2006). The resultsfrom an Estonian study agreed with the Swedish results,as moving heifers from separate housing to the tie stallless than 2 weeks before parturition significantlyincreased the risk for clinical mastitis around parturition(Kalmus et al. 2007). An increasing risk of clinicalmastitis was found in herds which housed heifers witholder cows (Barkema et al. 1999b). In pasture-grazedconditions, risk factors for peripartum mastitis maydiffer from those in housed herds (Compton et al. 2007).Significant risk factors for clinical and subclinicalmastitis post-calving were pre-calving subclinical mastitis,low teat height above the ground, Friesian breed andudder edema.Providing a clean, dry environment, good cowcomfort and ventilation are extremely important inprevention of mastitis in dry and calving cows andheifers. A recent study from the UK demonstrated theimportance of these factors in protecting cows fromclinical mastitis after calving (Green et al. 2007). Gooddrainage of the dry cow accommodation, use ofmattresses on dry cow cubicle surfaces and disinfectionof cubicle beds were some of the factors found to beprotective against mastitis (Fig. 2). Thickness of beddingof the calving box was found to be negativelycorrelated with incidence rate of clinical mastitis(Barkema et al. 1999b). In a Belgian study (DeVliegher et al. 2004) in herds with heifers calving onslatted floors, heifers had lower SCC, probably becausethose calving places were cleaner than premises withnon-slatted floors. If possible, dry cows and pregnantheifers should not be housed in the same barn, nor thetransition cows together with the milking cows (Barkemaet al. 1999b; Green et al. 2007).Development of vaccines against mastitis has beendifficult, because even natural intramammary infectiondoes not provide protection against subsequent infections(Talbot and Lacasse 2005). The high number ofmastitis pathogens further complicates the task. Commercialcommon core antigen vaccines against coliformmastitis have been available in some countries for years.These so-called J5 vaccines are based on the wholeÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

256 S Pyo¨ra¨läFly control used in all summermonths for heifersCows of lower paritySCC in last 90 days

Mastitis in Post-Partum Dairy Cows 257associated with the incidence rate of clinical mastitis. J DairySci 82, 1643–1654.Borm AA, Fox LK, Leslie KE, Hogan JS, Andrew SM, MoyesKM, Oliver SP, Schukken YH, Hancock DD, Gaskins CT,Owens WE, Norman C, 2006: Effects of prepartum intramammaryantibiotic therapy on udder health, milk production,and reproductive performance in dairy heifers. J DairySci 89, 2090–2098.Bradley AJ, Green MJ, 2000: A study of the incidence andsignificance of intramammary enterobacterial infectionsacquired during the dry period. J Dairy Sci 83, 1957–1965.Bradley AJ, Green MJ, 2001: An investigation of the impact ofintramammary antibiotic dry cow therapy on clinicalcoliform mastitis. J Dairy Sci 84, 1632–1639.Bradley AJ, Green M, 2004: The importance of the nonlactatingperiod in the epidemiology of intramammary infectionand strategies for prevention. Vet Clin North Am FoodAnim Pract 20, 547–568.Burton JL, Madsen SA, Chang LC, Weber PS, Buckham KR,van DR, Hickey MC, Earley B, 2005: Gene expressionsignatures in neutrophils exposed to glucocorticoids: a newparadigm to help explain ‘‘neutrophil dysfunction’’ inparturient dairy cows. Vet Immunol Immunopathol 105,197–219.Burvenich C, Van MV, Mehrzad J, ez-Fraile A, Duchateau L,2003: Severity of E. coli mastitis is mainly determined bycow factors. Vet Res 34, 521–564.Burvenich C, Bannerman DD, Lippolis JD, Peelman L,Nonnecke BJ, Kehrli ME Jr, Paape MJ, 2007: Cumulativephysiological events influence the inflammatory response ofthe bovine udder to Escherichia coli infections during thetransition period. J Dairy Sci 90 (Suppl. 1), E39–E54.Compton CW, Heuer C, Parker K, McDougall S, 2007: Riskfactors for peripartum mastitis in pasture-grazed dairyheifers. J Dairy Sci 90, 4171–4180.Curtis CR, Erb HN, Sniffen CJ, Smith RD, Powers PA,Smith MC, White ME, Hillman RB, Pearson EJ, 1983:Association of parturient hypocalcemia with eight periparturientdisorders in Holstein cows. J Am Vet Med Assoc183, 559–561.De Vliegher S, Laevens H, Barkema HW, Dohoo IR, StryhnH, Opsomer G, de Kruif A., 2004: Management practicesand heifer characteristics associated with early lactationsomatic cell count of Belgian dairy heifers. J Dairy Sci 87,937–947.Diez-Fraile A, Meyer E, Duchateau L, Burvenich C, 2004:Effect of proinflammatory mediators and glucocorticoids onL-selectin expression in peripheral blood neutrophils fromdairy cows in various stages of lactation. Am J Vet Res 65,1421–1426.Dingwell RT, Leslie KE, Schukken YH, Sargeant JM, TimmsLL, Duffield TF, Keefe GP, Kelton DF, Lissemore KD,Conklin J, 2004: Association of cow and quarter-levelfactors at drying-off with new intramammary infectionsduring the dry period. Prev Vet Med 63, 75–89.Dosogne H, Vangroenweghe F, Burvenich C, 2002: Potentialmechanism of action of J5 vaccine in protection againstsevere bovine coliform mastitis. Vet Res 33, 1–12.Godden S, Leslie KE, Dingwell R, Sanford CJ, 2006: Mastitiscontrol and the dry period: what have we learned. NationalMastitis Council Regional Meeting. Charlottetown, PrinceEdward Island, Canada. Proceedings, pp. 56–70.Green MJ, Bradley AJ, Medley GF, Browne WJ, 2007: Cow,farm, and management factors during the dry period thatdetermine the rate of clinical mastitis after calving. J DairySci 90, 3764–3776.Hoeben D, Monfardini E, Opsomer G, Burvenich C,Dosogne H, De KA, Beckers JF, 2000: Chemiluminescenceof bovine polymorphonuclear leucocytes during theperiparturient period and relation with metabolic markersand bovine pregnancy-associated glycoprotein. J DairyRes 67, 249–259.Hogan JS, Weiss WP, Smith KL, 1993: Role of vitamin E andselenium in host defense against mastitis. J Dairy Sci 76,2795–2803.Huszenicza G, Janosi S, Gaspardy A, Kulcsar M, 2004:Endocrine aspects in pathogenesis of mastitis in postpartumdairy cows. Anim Reprod Sci 82–83, 389–400.Huxley JN, Greent MJ, Green LE, Bradley AJ, 2002:Evaluation of the efficacy of an internal teat sealer duringthe dry period. J Dairy Sci 85, 551–561.Kalmus P, Viltrop A, Aasma¨e B, Kask K, 2007: Occurrence ofclinical mastitis in primiparous Estonian dairy cows indifferent housing conditions. Acta Vet Scand 48, 21.Kremer WD, Noordhuizen-Stassen EN, Grommers FJ,Daemen AJ, Henricks PA, Brand A, Burvenich C, 1993a:Preinfection chemotactic response of blood polymorphonuclearleukocytes to predict severity of Escherichia colimastitis. J Dairy Sci 76, 1568–1574.Kremer WD, Noordhuizen-Stassen EN, Grommers FJ, SchukkenYH, Heeringa R, Brand A, Burvenich C, 1993b:Severity of experimental Escherichia coli mastitis in ketonemicand nonketonemic dairy cows. J Dairy Sci 76, 3428–3436.Lamote I, Meyer E, Duchateau L, Burvenich C, 2004:Influence of 17beta-estradiol, progesterone, and dexamethasoneon diapedesis and viability of bovine blood polymorphonuclearleukocytes. J Dairy Sci 87, 3340–3349.Lehtolainen T, Suominen S, Kutila T, Pyorala S, 2003: Effectof intramammary Escherichia coli endotoxin in early- vs.late-lactating dairy cows. J Dairy Sci 86, 2327–2333.Leslie KE, Dingwell R, 2002: Mastitis control: where are weand where are we going? In: Kaske M, Scholz H,Ho¨ltershinken M (eds), Recent developments and perspectivesin bovine medicine. Hildesheimer Druck und Verlag-GmbH, Hildesheim, pp. 370–382.Leslie KE, Dingwell RT, 2003: Background to dry cowtherapy: what, where, why – is it still relevant? NMCAnnual Meeting Proceedings, Forth Worth, Texas, USA,pp. 5–17.Leslie KE, Duffield TF, Sandals D, Robinson E, 2001: Theinfluence of negative energy balance on udder health.Proceedings of the 2nd International Symposium on Mastitisand Milk Quality, Vancouver, Canada, pp. 195–199.Lippolis JD, Reinhardt TA, Goff JP, Horst RL, 2006:Neutrophil extracellular trap formation by bovine neutrophilsis not inhibited by milk. Vet Immunol Immunopathol113, 248–255.Mallard BA, Dekkers JC, Ireland MJ, Leslie KE, Sharif S,Vankampen CL, Wagter L, Wilkie BN, 1998: Alteration inimmune responsiveness during the peripartum period and itsramification on dairy cow and calf health. J Dairy Sci 81,585–595.Matthews KR, Harmon RJ, Langlois BE, 1992: Prevalence ofStaphylococcus species during the periparturient period inprimiparous and multiparous cows. J Dairy Sci 75, 1835–1839.McDougall S, Arthur DG, Bryan MA, Vermunt JJ, Weir AM,2007: Clinical and bacteriological response to treatment ofclinical mastitis with one of three intramammary antibiotics.NZ Vet J 55, 161–170.Mehrzad J, Duchateau L, Pyorala S, Burvenich C, 2002: Bloodand milk neutrophil chemiluminescence and viability inprimiparous and pluriparous dairy cows during late pregnancy,around parturition and early lactation. J Dairy Sci85, 3268–3276.Middleton JR, Timms LL, Bader GR, Lakritz J, Luby CD,Steevens BJ, 2005: Effect of prepartum intramammaryÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

258 S Pyo¨ra¨lätreatment with pirlimycin hydrochloride on prevalence ofearly first-lactation mastitis in dairy heifers. J Am Vet MedAssoc 227, 1969–1974.Monfardini E, Paape MJ, Wang Y, Capuco AV, Husheem M,Wood L, Burvenich C, 2002: Evaluation of L-selectinexpression and assessment of protein tyrosinephosphorylation in bovine polymorphonuclear neutrophilleukocytes around parturition. Vet Res 33, 271–281.Myllys V, 1995: Staphylococci in heifer mastitis before andafter parturition. J Dairy Res 62, 51–60.Oliver SP, Sordillo LM, 1988: Udder health in the periparturientperiod. J Dairy Sci 71, 2584–2606.Oliver SP, Lewis TM, Lewis MJ, Dowlen HH, Maki JL, 1990:Persistence of antibiotics in bovine mammary secretionsfollowing intramammary infusion at cessation of milking.Prev Vet Med 9, 301–311.Oliver SP, Lewis MJ, Gillespie BE, Dowlen HH, Jaenicke EC,Roberts RK, 2003: Prepartum antibiotic treatment ofheifers: milk production, milk quality and economic benefit.J Dairy Sci 86, 1187–1193.Osteras O, Edge VL, Martin SW, 1999: Determinants of successor failure in the elimination of major mastitis pathogens inselective dry cow therapy. J Dairy Sci 82, 1221–1231.Paape MJ, Bannerman DD, Zhao X, Lee JW, 2003: Thebovine neutrophil: structure and function in blood and milk.Vet Res 34, 597–627.Pantoja J, Ruegg P, 2007: Somatic cell count and incidence ofnew intramammary infection across lactation. NMC AnnualMeeting, San Antonio, TX, USA, p. 226.Parker KI, Compton C, Anniss FM, Weir A, Heuer C,McDougall S, 2007: Subclinical and clinical mastitis inheifers following the use of a teat sealant precalving. J DairySci 90, 207–218.Persson WK, Hallen SC, Emanuelson U, Jensen SK, 2007:Supplementation of RRR-alpha-tocopheryl acetate to periparturientdairy cows in commercial herds with high mastitisincidence. J Dairy Sci 90, 3640–3646.Pyo¨rälä SH, Pyo¨ra¨la¨ EO, 1998: Efficacy of parenteral administrationof three antimicrobial agents in treatment ofclinical mastitis in lactating cows: 487 cases (1989–1995). JAm Vet Med Assoc 212, 407–412.Rainard P, 2003: The complement in milk and defense of thebovine mammary gland against infections. Vet Res 34, 647–670.Rainard P, Riollet C, 2006: Innate immunity of the bovinemammary gland. Vet Res 37, 369–400.Rajala-Schultz PJ, Hogan JS, Smith KL, 2005: Short communication:association between milk yield at dry-off andprobability of intramammary infections at calving. J DairySci 88, 577–579.Robert A, Seegers H, Bareille N, 2006: Incidence of intramammaryinfections during the dry period without or withantibiotic treatment in dairy cows – a quantitative analysisof published data. Vet Res 37, 25–48.Sargeant JM, Leslie KE, Shirley JE, Pulkrabek BJ, Lim GH,2001: Sensitivity and specificity of somatic cell count andCalifornia Mastitis Test for identifying intramammaryinfection in early lactation. J Dairy Sci 84, 2018–2024.Shafer-Weaver KA, Pighetti GM, Sordillo LM, 1996: Diminishedmammary gland lymphocyte functions parallel shiftsin trafficking patterns during the postpartum period. ProcSoc Exp Biol Med 212, 271–280.Smith KL, Schanbacher FL, 1977: Lactoferrin as a factor ofresistance to infection of the bovine mammary gland. J AmVet Med Assoc 170, 1224–1227.Smith KL, Harrison JH, Hancock DD, Todhunter DA,Conrad HR, 1984: Effect of vitamin E and seleniumsupplementation on incidence of clinical mastitis andduration of clinical symptoms. J Dairy Sci 67, 1293–1300.Smith KL, Todhunter DA, Schoenberger PS, 1985: Environmentalpathogens and intramammary infection during thedry period. J Dairy Sci 68, 402–417.Sol J, Sampimon OC, Barkema HW, Schukken YH, 2000:Factors associated with cure after therapy of clinical mastitiscaused by Staphylococcus aureus. J Dairy Sci 83, 278–284.Sordillo LM, 2005: Factors affecting mammary gland immunityand mastitis susceptibility. Livestock Prod Sci 98,89–99.Sordillo LM, Pighetti GM, Davis MR, 1995: Enhancedproduction of bovine tumor necrosis factor-alpha duringthe periparturient period. Vet Immunol Immunopathol 49,263–270.Spain JN, Scheer WA, 2006: Transition cow nutrition andmastitis – balancing all the factors. NMC Annual Meeting,Tampa, FL, USA, pp. 172–174.Svensson C, Nyman AK, Persson WK, Emanuelson U, 2006:Effects of housing, management, and health of dairy heiferson first-lactation udder health in southwest Sweden. J DairySci 89, 1990–1999.Talbot BG, Lacasse P, 2005: Progress in the development ofmastitis vaccines. Livestock Prod Sci 98, 101–113.Taponen S, Simojoki H, Haveri M, Larsen HD, Pyorala S,2006: Clinical characteristics and persistence of bovinemastitis caused by different species of coagulase-negativestaphylococci identified with API or AFLP. Vet Microbiol115, 199–207.Taponen S, Koort J, Bjo¨rkroth J, Saloniemi H, Pyo¨rälä S,2007: Bovine intramammary infections caused by coagulasenegativestaphylococci may persist throughout lactationaccording to amplified fragment length polymorphismbasedanalysis. J Dairy Sci 90, 3301–3307.Todhunter DA, Smith KL, Hogan JS, Schoenberger PS, 1991:Gram-negative bacterial infections of the mammary gland incows. Am J Vet Res 52, 184–188.Tomita GM, Ray CH, Nickerson SC, Owens WE, Gallo GF,2000: A comparison of two commercially available Escherichiacoli J5 vaccines against E. coli intramammary challenge.J Dairy Sci 83, 2276–2281.Valde JP, Lawson LG, Lindberg A, Agger JF, Saloniemi H,Osteras O, 2004: Cumulative risk of bovine mastitis treatmentsin Denmark, Finland, Norway and Sweden. Acta VetScand 45, 201–210.Vandeputte-Van Messom G, Burvenich C, Roets E, Massart-Leen AM, Heyneman R, Kremer WD, Brand A, 1993:Classification of newly calved cows into moderate and severeresponders to experimentally induced Escherichia coli mastitis.J Dairy Res 60, 19–29.Vangroenweghe F, Duchateau L, Burvenich C, 2004: Moderateinflammatory reaction during experimental Escherichiacoli mastitis in primiparous cows. J Dairy Sci 87, 886–895.Waage S, Mork T, Roros A, Aasland D, Hunshamar A,Odegaard SA, 1999: Bacteria associated with clinical mastitisin dairy heifers. J Dairy Sci 82, 712–719.Weber PS, Toelboell T, Chang LC, Tirrell JD, Saama PM,Smith GW, Burton JL, 2004: Mechanisms of glucocorticoid-induceddown-regulation of neutrophil L-selectin incattle: evidence for effects at the gene-expression level andprimarily on blood neutrophils. J Leukoc Biol 75, 815–827.Whist AC, Osteras O, Solverod L, 2006: Clinical mastitis inNorwegian herds after a combined selective dry-cow therapyand teat-dipping trial. J Dairy Sci 89, 4649–4659.Williamson JH, Woolford MW, Day AM, 1995: The prophylacticeffect of a dry-cow antibiotic against Streptococcusuberis. NZ Vet J 43, 228–234.Wilson DJ, Grohn YT, Bennett GJ, Gonzalez RN, SchukkenYH, Spatz J, 2007a: Comparison of J5 vaccinates andcontrols for incidence, etiologic agent, clinical severity, andÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Mastitis in Post-Partum Dairy Cows 259survival in the herd following naturally occurring cases ofclinical mastitis. J Dairy Sci 90, 4282–4288.Wilson DJ, Mallard BA, Burton JL, Schukken YH, GrohnYT, 2007b: Milk and serum J5-specific antibodyresponses, milk production change, and clinical effectsfollowing intramammary Escherichia coli challenge for J5vaccinate and control cows. Clin Vaccine Immunol 14,693–699.Zerbe H, Schneider N, Leibold W, Wensing T, Kruip TA,Schuberth HJ, 2000: Altered functional and immunophenotypicalproperties of neutrophilic granulocytes in postpartumcows associated with fatty liver. Theriogenology 54,771–786.Author’s address (for correspondence): S Pyörälä, Department ofProduction Animal Medicine, Faculty of Veterinary Medicine, Universityof Helsinki, Saarentaus, FI-07920, Finland. E-mail: satu.pyorala@helsinki.fiConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 260–267 (2008); doi: 10.1111/j.1439-0531.2008.01171.xISSN 0936-6768Embryonic and Early Foetal Losses in Cattle and Other RuminantsMG Diskin and DG MorrisTeagasc, Animal Production Research Centre, Mellows Campus, Athenry, Co. Galway, IrelandContentsEmbryo survival is a major factor affecting production andeconomic efficiency in all systems of ruminant milk and meatproduction. For heifers, beef and moderate yielding dairycows, does and camelids it appears that fertilization generallylies between 90% and 100%. In high-producing dairy cowsthere is a less substantive body of literature, but it wouldappear that it is somewhat lower and perhaps more variable.In cattle, the major component of embryo loss occurs beforeday 16 following breeding with some evidence of greater lossesbefore day 8 in high-producing dairy cows. In cattle lateembryo loss, while numerically much smaller than earlyembryo mortality loss, nevertheless, causes serious economiclosses to producers because it is often too late to rebreedfemales when they repeat. In multiple ovulating small ruminants,the loss rate is positively related to ovulation rate.Systemic concentrations of progesterone, during both the cyclepreceding and following insemination, affect embryo survivalrate with evidence that too high or indeed too low aconcentration being negatively associated with survival rate.Uterine expression of mRNA for progesterone receptor,oestradiol receptor and retinol-binding protein appears to besensitive to changes in peripheral concentrations of progesteroneduring the first week after artificial insemination. Energybalance and dry matter intake during 4 weeks after calving arecritically important in determining conception rate when cowsare inseminated at 70–100 days post-calving. Concentratesupplementation of cows at pasture during the breeding periodhas minimal effects on conception rates though suddenreductions in dietary intake should be avoided. For all systemsof milk production, more balanced breeding strategies withgreater emphasis on fertility and feed intake and ⁄ or energybalance must be developed. There is sufficient genetic variabilitywithin the Holstein breed for fertility traits. Alternativedairy breeds such as the Jersey or Norwegian Red could alsobe utilized. Genomic technology will not only provide scientistswith an improved understanding of the underlyingbiological processes involved in fertilization and the establishmentof pregnancy, but also, in the future, identify genesresponsible for improved embryo survival. Its incorporationinto breeding objectives would increase the rate of geneticprogress for embryo survival.IntroductionEmbryo mortality is a major cause of economic loss inall systems of ruminant production. Direct effects ofembryonic mortality are reflected in reduced conceptionrates to a service and reduced litter size in litter-bearingspecies. In dairying, the increase in milk yield observedover the past 40 years has been accompanied by adecline in cow fertility both in high input maize-baseddiets and in less intensive, pasture-based systems of milkproduction such as those practised in Ireland (see reviewby Diskin et al. 2006). In litter-bearing species such asthe ewe, death of one or more embryos during theimplantation stage results in the birth of smaller lambs(Rhind et al. 1980) with subsequent adverse consequencesfor their post-natal survival and growth rates.The objective of this article is to review currentinformation on embryonic and early foetal losses incattle and other ruminants and in particular on thefactors that affect their incidence and to proposepotential avenues to embryo and foetal survival rates.Fertilization Rate in CattleThere are a large number of published estimates offertilization rate in heifers and in moderate-yieldingdairy cows (see review by Sreenan and Diskin 1986).Where semen of known high fertility is used in artificialinsemination (AI), fertilization rates are of the order of90–100%. In contrast, for higher producing dairy cowsthere is little quantitative information on fertilizationrate with only three published reports (Wiebold 1988;Ryan et al. 1993; Sartori et al. 2002). Wiebold (1988),using a non-surgical embryo recovery technique on day7 following oestrus, recovered 25 ova ⁄ embryos from 23lactating cows with all recovered ova having beenfertilized. Ryan et al. (1993) in a study on the effectsof ambient temperature on fertilization rate reported noeffect of temperature and quoted fertilization rates of82.4% and 79.5% for high and low temperatures,respectively. Sartori et al. (2002) recorded a low fertilizationrate of 55.6% in lactating dairy cows comparedto 100% for heifers under high ambient temperatures,while in a subsequent study during the cool seasonfertilization rates were 87.8% and 89.5% for lactatingand non-lactating dairy cows, respectively. It appearsthat fertilization rate is similar in high- and moderateproducingdairy cows, at least during the cool season.Fertilization Rates in Small RuminantsThe available evidence is that, like cattle, a fertilizationrate of 90–95% appears to be normal in ewes (Restallet al. 1976; Mitchell et al. 1999), does (see review byNancarrow 1994) and South American camelids (Fernandez-Bacaet al. 1970) under natural mating conditions.There is some suggestion that fertilization is anall-or-none phenomenon in multiple ovulating ewes(Restall et al. 1976), though there is some evidence(Restall et al. 1976; Kleeman et al. 1990) that partialfertilization failure may occur at a low frequency. Thereis some evidence that fertilization rates may be reducedin ewes during both the early (Hulet et al. 1956) andlatter parts of the breeding season, perhaps as aconsequence of suboptimal mating activity in somerams (Colas 1983) or adverse environmental and nutritionalconditions (Mitchell et al. 1996). Yet, when ewesÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Embryo ⁄ Foetal Losses in Ruminants 261are neither environmentally nor nutritionally challengedand mated to fertile rams differences in fertilizationfailure rate are likely to be small or non-existent(Mitchell et al. 1999).Extent and Timing of Early Embryo Loss inCattleWhile fertilization rate is apparently similar in high- andmoderate-producing cows and unlikely to be affected bywhether cows are on pasture or high input total mixedration (TMR) diets, nevertheless the average calving rateto a single service is significantly lower in high- thaneither in low-producing cows or in heifers. Sreenan andDiskin (1986) calculated an embryonic and foetalmortality rate (excluding fertilization failure) of about40% for moderate-producing cows based on a fertilizationrate of 90% and an average calving rate of about55% with an estimated 70–80% of it sustained betweendays 8 and 16 after insemination. The comparativefigure for high-producing dairy cows based on afertilization rate of 90% and a calving rate of 40% is56%.There is some evidence that the pattern of earlyembryo death in the modern high-producing cow maybe different to that observed in heifers and loweryielding dairy cows. For example, in the study of Sartoriet al. (2002) conducted during the summer in Wisconsin,a significantly higher proportion of the embryos recoveredon days 6–7 were of higher quality grade fromheifers (72%) compared with lactating cows (33%).Similarly, during the winter component of the samestudy, Sartori et al. (2002) also recorded a higherproportion of higher quality grade embryos from dry(83%) compared to lactating (53%) cows. This trend ofa comparatively higher proportion of low-quality gradeor abnormal or retarded embryos from high-yieldingdairy cows is also evident in the study of Wiebold(1988), who found that 52% of the embryos from highyieldingdairy cows on days 6–7 were abnormal, basedon morphological criteria. Ryan et al. (1993) reportedthat the proportions of high-quality grade embryosrecovered under the conditions of high (58.5%) and low(51.6%) ambient temperature were similar. A feature ofthese recent studies is the consistent evidence that a high(41–67%) proportion of embryos recovered on day 7post-oestrus from moderate- or high-producing dairycows were classified as abnormal in contrast to a muchlower incidence (17–28%) in heifers and non-lactatingcows in the same studies or the 6% (average of fourstudies) in the report of Sreenan and Diskin (1986). Itwould appear that early embryo loss is greater in themodern high-producing dairy cow and that a muchhigher proportion of the embryos die before day 7following insemination. Based on the published evidence,the expected outcome of 100 inseminations ofBritish Friesian and Holstein–Friesian cows is summarizedin Fig. 1.Late Embryo and Foetal Loss in CattleOver the past 10 years, there has been significantinterest in the problem of late embryo and early foetalBritish Friesian 1980 Holstein Friesian 2006Late embryo mortality7% Early7%embryodeath 28%EarlyCalvingCalvingembryo40%55%10%death 43%Fertilisation failure10%Fig. 1. Reproductive outcomes in British–Friesian versus Holstein–Friesian cows (Source: Diskin et al. 2006)mortality, which has generally been defined as the deathof the embryo after about day 24 of gestation. With theadvent of ultrasound scanning, it has been comparativelyeasy to accurately establish the extent and timingof late embryo ⁄ foetal mortality. Silke et al. (2001)quantified the extent and pattern of embryo ⁄ foetal lossfrom days 28 to 84 of gestation in 1046 lactating dairycows and 162 dairy heifers managed on pasture-basedsystems of milk production. The overall loss ratesbetween days 28 and 84 of gestation and the pattern ofloss over this period are similar for cows (7.2%)producing on average 7247 kg of milk and heifers(6.1%). Almost half (47.5%) of the total recorded lossoccurred between days 28 and 42 of gestation. Thereare no significant association between level of milkproduction or milk energy output measured on day 120of lactation, milk fat concentration, milk proteinconcentration or milk lactose concentration andembryo ⁄ foetal loss rate. The extent and pattern ofembryo ⁄ foetal was not related to either cow or cow siregenetic merit. A more recent study by Horan et al.(2004) recorded a similar overall late embryo ⁄ foetal lossrate of 7.5% between days 30 and 67 of gestation indairy cows managed under pasture-based systems ofproduction. The extent of late embryo ⁄ foetal mortalityrecorded in these two pasture-based studies is muchlower than that reported for some US-based studies(Silke et al. 2001), though the causes for the apparentdifference are not clear. While the extent of late embryoloss is numerically much smaller than early embryoloss, it nevertheless causes serious economic losses toproducers because it is too late to rebreed cows whenthe loss occurs resulting in increased culling particularlyin seasonal calving herds.Extent and Timing of Embryo Loss in SmallRuminantsUnlike with cattle, there are a much smaller number ofreports of the serial slaughter of other ruminants atspecific time points following mating to determine thepattern of embryo loss. Most studies have measuredembryo survival at about the time of implantation andor at term. Furthermore, most studies have includedfertilization failure in their estimate of embryo loss.This is not unreasonable given that fertilization rate isusually close to 100%. In ewes and does, it is clearthat embryo survival rate is a function of ovulationÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

262 MG Diskin and DG MorrisTable 1. Probability of embryo survival as a function of ovulationrate a in naturally mated ewes and does (Source: Hanrahan 1994)No. ofcorpora luteaEwesProbability of embryo survival bDoes2 0.82 (5069) c 0.82 (306)3 0.74 (884) 0.60 d (58)4 0.65 (270) –5 0.55 (91) –6 0.45 (38) –a Based on summary of published and unpublished information on ovulation rateand the number of foetuses ⁄ off-springs.b About ‡250 and 0.06 to 0.08 otherwise.c No. of females involved (n).d Includes observations on some does with >3 corpora lutea.rate (Hanrahan 1994) and data are summarized inTable 1. While the information on goats is limitedrelative to sheep, both sets of data show that theprobability of embryo survival declines as ovulationrate increases. From a comparison of the publishedliterature on embryo survival in mated ewes or in ewesfollowing embryo transfer, Hanrahan (1994) concludedthat the change in pattern of embryo survival withincreasing number of ova shed or embryos transferredwas similar, and the clear negative relationshipbetween the number of ovulations and subsequentembryo survival is a direct consequence of the numberof embryos entering the uterus rather than anyproblems with embryo quality, the time of ovulationor uterine environment. He also concluded that therewas no evidence of an inherent physiological or geneticrelationship between ovulation rate and embryo survival.A direct consequence of these conclusions is thatinformation on embryo survival in sheep and goatsshould be examined in relation to ovulation rate andthe interpretation of differences due to management orenvironmental factors should allow for this directeffect of ovulation rate. Extrapolation of the datacontained in Table 1 provides estimates ofembryo ⁄ foetal loss of 12%, 18% and 26% for eweswith one, two or three ovulation, respectively. Incomparison with cattle, which are predominatelymono-ovular, the embryo loss rate of 12% formono-ovular sheep is much lower than comparablefigures of about 40% and 60% for heifers and highproducingdairy cows, respectively.In camelids the proportion pregnant at about60 days post-mating is about 65% of those thatovulate (Adams et al. 1991), although much lowerfigures have been recorded. Given that fertilizationrates are similar in all species, it seems that embryo lossrates are much higher in camelids than in sheep orgoats. While twin ovulations occur quite frequently incamelids, twin births are rare indicating that themechanisms that control embryo survival are probablydifferent in camelids to those in cattle, sheep and goats.One of the factors which may contribute to the higherembryo loss rate is the fact that in all camelids, over98% of the foetuses implant in the left uterine horndespite the fact both ovaries contribute almost equallyto ovulation. It appears that the right uterine horn isunable to sustain a pregnancy after about 50 days ofgestation (Arthur et al. 1982).Causes of Embryo or Foetal LossSeveral factors have been implicated in embryo andfoetal loss, usually with a lack of supporting experimentaldata. The main factors implicated are normallycategorized as those of genetic, physiological, endocrineand environmental origin (see reviews by Ashworth1994; Bazer 1994; Nancarrow 1994; Zavy 1994). Onlysome possible genetic and environmental factors areconsidered here.Genetic causesGenetic causes of embryo death include chromosomaldefects, individual genes and genetic interactions (Van-Raden and Miller 2006). The 1 ⁄ 29 Robertsonian chromosomaltranslocation (Gustavsson 1979) present inseveral beef breeds and in the Scandinavian Red breeds,but not in the Holstein breed and has been suggested asa cause of reduced fertility of males and femalesheterozygous for 1 ⁄ 29 translocation. Sweden has successfullyimplemented a testing programme to eliminatefrom AI bulls that are carriers for this conditionresulting in a significant improvement in fertility (Gustavsson1979).In the Holstein breed, two major recessive defectsaffecting embryo ⁄ foetal survival have been detected.Deficiency of uridine monophosphate synthase(DUMPS) (Robinson et al. 1984), a homozygous recessivecondition, causes foetal death a 40–50 days ofgestation (Shanks and Robinson 1989). Testing of AIsires for DUMPS has significantly reduced the frequencyof heterozygous sires and of homozygousrecessive embryos and has now almost eliminated thisas a cause of fertility losses (VanRaden and Miller2006). Complex vertebral malformation is a lethalrecessive condition that causes late foetal death in cattle.This defect was disseminated through the widespreaduse of the Holstein Carlin-M Ivanhoe Bell as a maternalsire of sires. In a Swedish study, Persson (2003) recordedthat carrier bulls had significantly lower breeding valuesfor 168-day non-return rate than non-carriers at168 days following insemination but were not differentat 28 and 56 days indicating that it primarily affects laterather than early foetal mortality. Again it is possible togreatly reduce the incidence if carriers for this conditionby testing sires.Several reproductive traits are adversely affected byinbreeding. Maternal inbreeding has been reported todecrease the 56–70-day non-return rates by between 1%(Wall et al. 2003) and 2% (Cassell et al. 2003) per 10%inbreeding of the dam. On the other hand, inbreeding ofthe embryo has bee reported to reduce the 70-day nonreturnrate by 1% for each 10% increase in the level ofinbreeding of the embryo (Cassell et al. 2003; VanRadenand Miller 2006). Genetic variation in embryo survivalmay be attributable to genetic constitution of theembryo itself and or the genetic differences among damswith respect to their ability to provide an appropriateintraovarian and uterine environment. In sheep, BodinÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Embryo ⁄ Foetal Losses in Ruminants 263et al. (1992) concluded, based on results from embryotransfer and breed-crossing studies, that genetic differencesamong embryos were not of any consequence forvariation in embryo survival rate, aside from factorssuch as chromosomal abnormalities, specific lethal genesand genetic interaction. Breed differences in embryosurvival as a trait of the ewe have been examined inmany studies based on employing embryo transferprocedures and on estimation of embryo survival usinginformation on ovulation rate and the consequent littersize. The general conclusion from the embryo transferstudies is that breed of recipient ewe did not play animportant role in determining embryo survival despitethe fact that these studies often involved breeds withvery different levels of natural prolificacy (Hanrahan1994), though there is some evidence (Hanrahan andPiper 1992) that embryo survival may be better when theFinn breed was used as an embryo recipient. There isquite a substantial literature on estimation of embryosurvival based on the information on ovulation andassociated litter size and provides a good indication ofbreed variation in embryo survival. Hanrahan (1994),following an extensive review of the literature, suggestedthat the likely breed differences for the probability ofembryo survival are 0.15, 0.13 and 0.11 for twin, tripletand quadruplet ovulations, respectively. If the essentiallylinear relationship between ovulation rate andlitter size is the same in all breeds, then a global figure of0.12 has been suggested by Hanrahan (1994) as thelikely range of breed means for embryo survivaladjusted for ovulation rate. Variation among individualswithin a population, measured as repeatability ofdifferences in embryo survival, is also quite low (»0.05)(Hanrahan 1982; Ricordeau et al. 1986), and consequentlythe heritability of these differences is alsoextremely low (Ricordeau et al. 1986).Maternal ageWhile the overall embryo survival rates would appearto be largely similar in heifers, beef cows and low–moderate-producing dairy cows, embryo survival rate islower in high-producing dairy cows. Yet, it is improbablethat this is an age- or parity-related phenomenonand is more probable due to the direct and indirectconsequences of milk production (see later). Recentdata from the US (Kuhn et al. 2006) would suggestboth that heifer conception rate is at a maximum at 15–16 month of age. Breeding heifers at 26 months of ageor older resulted in a 13% lower conception ratepresumably due to a lower embryo survival rate. Thereis evidence, at least in some breeds of sheep (Galway:Quirke and Hanrahan 1977; Romney: McMillan andMcDonald 1985), that embryo survival is lower in ewelambs than in adult ewes. These later studies and the asubsequent study of Quirke and Hanrahan (1983) allemploying embryo transfer indicate that the impairedembryo survival associated with ewe lambs is attributableto the inherent quality of embryo rather than anydeficiency of uterine environment. Ricordeau et al.(1982, 1986) found no significant differences in embryosurvival between ewe lambs and adults of the Romanovbreed. This may be dependent on the degree of sexualmaturity attained at breeding as there is evidence thatconception rate increases between first and later oestrusperiods in ewe lambs. Among age groups other thanewe lambs, there is little evidence of differences inembryo survival rate (Hanrahan 1990; Kleeman et al.1990).Progesterone during the cycle immediately prior toinsemination and embryo survival rateThere is now good evidence linking circulating concentrationof progesterone during the cycle immediatelyprior to insemination as well as during the early lutealphase of the cycle following insemination with lowembryo survival ⁄ conception rate (see review by Diskinet al. 2006). Recent data from this laboratory (Diskinet al. 2006) clearly show that there is a positive linearassociation between the concentrations of progesteroneon the day of PGF-2a-induced luteolysis and subsequentembryo survival rate. Potential mechanisms by whichlow concentrations of progesterone during the precedingoestrous cycle might reduce fertilization and or embryosurvival rates include the production of oocytes that areat a more advanced stage of maturation at time ofovulation from persistent dominant follicles; increasedfrequency of pulses of LH which in turn inducesincreased secretion of oestradiol-17b or an alterationin endometrial morphology (see review by Diskin et al.2006). The more probable effect of low concentrationsof progesterone in the cycle preceding oestrus onsubsequent embryo survival rate is to result inpre-mature oocyte maturation, which subsequentlycompromises its ability to continue normal embryodevelopment after its fertilization.Post-insemination progesterone and embryo survival rateRecent studies (Stronge et al. 2005; Diskin et al. 2006)that examined the relationship between early and midlutealphase concentrations of progesterone and subsequentembryo survival ⁄ conception rate have employedlogistic regression techniques to model the relationshipbetween the binomially distributed dependent variable(conception ⁄ embryo survival rate) and the continuouslydistributed independent variable (progesterone). In thestudy of Stronge et al. (2005) (Fig. 2), there was apositive linear and quadratic relationship between milkconcentrations of progesterone on days 5, 6, 7 postinseminationand between the rate of change in concentrationsof progesterone between days 4 and 7 andembryo survival rate. Further analysis of this data setreveals that 75%, 72% and 56% of dairy cows haveconcentrations of progesterone that are optimal forconception on days 5, 6 and 7 post-insemination,respectively. In beef heifers, Diskin et al. (2006) haveshown that a similar linear and quadratic associationbetween peripheral concentrations of progesterone andembryo survival also exists. These recent data for cattleare inline with results from earlier sheep studies of Parret al. (1987) of an inverse relationship between circulatingconcentrations of progesterone and embryo survivalin these species. Furthermore, there is evidence for sheep(Parr et al. 1987) and dairy cows (Starbuck et al. 2001)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

264 MG Diskin and DG Morris10090807060504030201001009080706050403020100Days 5 0.7 70 Days 60.6 600.5 500.4 400.3 300.2 200.1 100 00–12–34–56–78–910–1112–1314–151009080

Embryo ⁄ Foetal Losses in Ruminants 265Energy balance at around the time of insemination andsubsequent conception rateDry matter intake is lower for cows grazing pastures thanfor cows fed maize-based TMRs. High-genetic merithigh-producing cows experience greater NEB in earlylactation under grazing conditions relative to lowergenetic merit cows, notwithstanding the somewhat higherDMI capacity of the former cow type (Horan et al.2004). Meta-analysis of the combined Kennedy et al.(2003) and Horan et al. (2004) data showed that therewas no service · study · supplementation rate (p >0.10), service · study rate (p > 0.10), study · supplementationrate (p > 0.10) interaction effects on conceptionrate (see Diskin et al. 2006). Yet, there was aservice · supplementation rate effect on conception rate(p < 0.05). The rate of supplementation had no effect onfirst service conception rate (56% vs 53%) but cows onthe low level of supplementation had a lower (p < 0.05)second service conception rate (39% vs 58%) comparedwith cows on the high level of supplementation. Interestingly,the withdrawal of the supplementation commencedat about the onset of the breeding period in bothstudies. From these studies, there is no clear evidence thatconcentrate supplementation of dairy cows at pastureimproves first service conception rate, but it may bebeneficial in maintaining second service conception rates.This clearly highlights the difficulty that improving theEB of the modern dairy cows presents at this stage oflactation where grazed grass is the predominant componentof the diet. Based on the studies of Sangsritavonget al. (2002), the increased milk production as a result ofthe concentrate supplementation may well be associatedwith increased hepatic blood flow resulting in increasedmetabolism of progesterone and consequently loweringof peripheral concentrations of progesterone andpre-disposing to greater risk of embryo death.Sudden reductions in feed intake and conception rateStudies from this laboratory (Dunne et al. 1999) showthat sudden reductions in DMI at around the time ofinsemination adversely affect embryo survival in heifers.When energy intake was reduced from a high level oftwice their maintenance requirement to 0.8 times maintenancefor 2 weeks immediately after AI, embryosurvival rate in heifers was consistently less than 40%.When heifers were provided with a constant level of feedintake or when they were changed from a low to ahigher level feed intake, embryo survival was consistentlyhigh at 65–71%. In that study, where heifers wereused, there was no indication of any association betweenenergy intake and systemic progesterone concentration.Unlike the situation in sheep and pigs, there was nochange in systemic progesterone following either anincrease or reduction in energy intake. Changes inprogesterone metabolism may have been balanced bychanges in progesterone production.Protein nutrition and conception rateDairy cows at pasture frequently ingest high quantitiesof protein, often with a high proportion of theingested protein being rapidly degradable in therumen. The effects of high intakes of crude proteinon conception rate are equivocal (Diskin et al. 2006).They concluded, based on the published evidence, thatelevated concentrations of urea per se are not detrimentalto embryo survival. Yet, it needs to be clarifiedwhether the observed adverse effects of urea onembryo survival are dependent on the energy statusof the animal.Future PossibilitiesIt is well established that embryo survival is criticallyimportant in all ruminant species but particularly inseasonally bred herds. The extent of late embryomortality is numerically much smaller than earlyembryo mortality loss but nevertheless causes seriouseconomic losses to producers, because it is often too lateto rebreed females when they repeat particularly inseasonally bred herds. The major challenge will be toimprove embryonic survival rate and therefore herdreproductive efficiency particularly given the antagonisticrelationship between production and embryo survivalrate. It is clear that genetic variability exists within theHolstein breed for important fertility (Berry et al. 2003)and DMI and EB (Berry et al. 2007) traits. Because ofthe difficulty in directly measuring DMI and ⁄ or EB inlarge numbers of dairy cows, it is not currently directlyincluded in international breeding objectives. Applyinga positive weighting to traits correlated with DMIand ⁄ or EB or the identification and application ofsuitable molecular markers would be expected toimprove DMI and EB and subsequent reproductiveperformance. Alternatively, strains of cows derived frommore balanced breeding objectives, such as the NewZealand Friesian (Harris and Kolver 2001) or alternativedairy breeds such as the Jersey or Norwegian Red,could be utilized in such production systems. Advancesin genomic technology have the potential not only toprovide scientists with an improved understanding ofthe underlying biological processes involved in fertilizationand the establishment of pregnancy but also toexplore gene profiles and potentially identify genesresponsible for improved embryo survival. Markerassistedselection will increase the rate of geneticprogress for reproductive traits such as embryo survivalby increasing the accuracy of selection and permittingselection in both genders and at a young age (Rohrer2004). Reproduction-enhancing technologies such as AI,embryo transfer and perhaps cloning will have animportant role in fully exploiting their potential (Georgeand Massey 1991).ReferencesAdams GP, Sumar J, Ginther OJ, 1991: Form and function ofthe corpus luteum in llamas. Anim Reprod Sci 24, 127–138.Arthur GH, Noakes DE, Pearson H, 1982: Reproduction inthe camel. In: Arthur GH, Noakes DE, Pearson H (ed.),Veterinary Reproduction and Obstetrics. Bailli-ere Tindall,London, 482–487.Ashworth JC, 1994: Nutritional factors related to embryonicmortality in domestic species. In: Zavy MT, Geisert RDÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

266 MG Diskin and DG Morris(eds), Embryonic Mortality in Domestic Species. CRCPress, Boca Raton, FL, pp. 179–194.Ashworth CJ, Bazer FW, 1989: Changes in ovine conceptusand endometrial function following asynchronous embryotransfer or administration of progesterone. Biol Reprod 40,425–433.Bazer FW, 1994: New frontiers and approaches in the study ofembryonic mortality in domestic species. In: Geisert RD,Zavy MT (eds), Embryonic Mortality in Domestic Species.CRC Press, Boca Raton, FL, pp. 195–210.Beam SW, Butler WR, 1997: Energy balance and ovarianfollicle development prior to first ovulation postpartum indairy cows receiving three levels of dietary fat. Biol Reprod56, 133–142.Berry DP, Buckley F, Dillon P, Evans RD, Veerkamp RF,2003: Genetic parameters for body condition score, bodyweight, milk yield and fertility estimated using randomregression models. J Dairy Sci 86, 3704–3717.Berry DP, Horan B, O’Donovan M, Buckley F, Kennedy E,McEvoy M, Dillon P, 2007: Genetics of grass dry matterintake, energy balance, and digestibility in grazing Irishdairy cows. J Dairy Sci 90, 4835–4845.Bodin L, Hanrahan JP, Poivey JP, 1992: Variation in embryosurvival in sheep and goats. Paper Presented to 43rd AnnualMeeting E.A.A.P., Madrid, 24 pp.Britt JH, 1994: Follicular development and fertility, potentialimpacts of negative energy balance. In: Proceedings ofNational Reproductive Symposium, Pittsburgh, PA,pp.103–112.Cassell BG, Adamec V, Pearson RE, 2003: Maternal and fetalinbreeding depression for 70-day nonreturn and calving ratein Holsteins and Jerseys. J Dairy Sci 86, 2977–2983.Colas G, 1983: Factors affecting the quality of ram semen. In:Haresign W (ed), Sheep Production. Butterworths, London,pp. 453–465.Diskin MG, Murphy JJ, Sreenan JM, 2006: Embryo survivalin dairy cows managed under pastoral conditions. AnimReprod Sci 96, 297–311.Dunne LD, Diskin MG, Boland MP, O’Farrell KJ, SreenanJM, 1999: The effects of pre- and post-insemination plane ofnutrition on embryo survival in beef heifers. Anim Sci 69,411–417.Fernandez-Baca S, Hansel W, Novoa C, 1970: Embryonicmortality in the Alpaca. Biol Reprod 3, 243–251.Garrett JE, Geisert RD, Zavy MT, Morgan GL, 1988:Evidence for maternal regulation of early conceptus growthand development in beef cattle. J Reprod Fertil 84, 437–446.George M, Massey JM, 1991: Velogenetics, or the synergisticuse of marker assisted selection and germ-line manipulation.Theriogenology 35, 151–159.Gustavsson I, 1979: Distribution and effects of the 1.29Robertsonian translocation in cattle. J Dairy Sci 62, 825–835.Hanrahan JP, 1982: Selection for increased ovulation rate,litter size and embryo survival. In: Proceedings of 2ndWorld Congress Genetics Applied to Livestone Production,Vol. V, pp. 294–309.Hanrahan JP, 1990: Effect of selection for increased litter sizeon ovulation rate and embryo survival in sheep. In:Proceedings of British Society of Animal Production,Winter Meeting, Scarborough, UK, Paper No. 32.Hanrahan JP, 1994: Embryo survival in small ruminants:incidence and sources of variation. In: 45th AnnualMeeting European Association of Animal Production,Edinburgh, September, Sheep and Goat Commission,Session 2.Hanrahan JP, Piper LR, 1992: Nature of the genetic control ofovulation rate and its relationship with litter size. In:Proceedings of 30th Annual Meeting E.A.A.P., Leningrad,GS4.2, 12 pp.Harris BL, Kolver ES, 2001: Review of Holsteinization onintensive pastoral dairy farming in New Zealand. J Dairy Sci84E(Suppl.), E56–E61.Horan B, Mee JF, Rath M, O’Connor P, Dillon P, 2004: Theeffect of strain of Holstein–Friesian cow and feeding systemon reproductive performance in seasonal-calving milk productionsystems. Anim Sci 79, 453–467.Hulet CV, Voigtlander HP Jr, Pope AL, Casida LE, 1956: Thenature of early-season infertility in sheep. J Anim Sci 15,607–616.Kennedy J, Dillon P, O’Sullivan K, Buckley F, Rath M, 2003:The effect of genetic merit for milk production andconcentrate feeding level on reproductive performance ofHolstein–Friesian cows in a grass-based system. Anim Sci76, 297–308.Kerbler TL, Buhr MM, Jordan LT, Leslie KE, Walton JS,1997: Relationship between maternal plasma progesteroneconcentration and interferon-tau synthesis by the conceptusin cattle. Theriogenology 47, 703–714.Kleeman DO, Walker SK, Walkley JRW, Smith DH, GrimsonRJ, Seamark RF, 1990: Fertilization and embryo loss inBoorolla Merino · South Australian Merino ewes: effect ofthe F gene. Theriogenology 33, 487–498.Kuhn MT, Hutchison JL, Wiggans GR, 2006: Characterizationof Holstein Heifer fertility in the United States. J DairySci 89, 4907–4920.McMillan WH, McDonald MF, 1985: Survival of fertilizedova from ewe lambs and adult ewes in the uteri of ewelambs. Anim Reprod Sci 8, 235–240.McNeill RE, Sreenan JM, Diskin MG, Cairns MT, FitzpatrickR, Smith TJ, Morris DM, 2006: Effect of progesteroneconcentration on the expression of progesterone-responsivegenes in the bovine endometrium during the early lutealphase. Reprod Fertil Dev 18, 573–583.Mitchell LM, King ME, Aitken RP, Wallace JM, 1996: Effectof mating season and body condition on ovulation, fertilizationand pregnancy rates in crossbred ewes. Theriogenology45, 293.Mitchell LM, King ME, Aitken RP, Gebbie FE, Wallace JM,1999: Ovulation, fertilization and lambing rates, andperipheral progesterone concentrations, in ewes inseminatedat a natural oestrus during November or February.J Reprod Fertil 115, 133–140.Nancarrow CD, 1994: Embryonic mortality in the ewe anddoe. In: Zavy MT, Geisert RD (eds), Embryonic Mortalityin Domestic Species. CRC Press, London, UK, pp. 79–98.Nebel R, McGilliard M, 1993: Interactions of high milk yieldand reproductive performance in dairy cows. J Dairy Sci 76,3257–3268.Parr RA, Davis IF, Fairclough RJ, Miles MA, 1987:Overfeeding during early pregnancy reduces peripheralprogesterone concentration and pregnancy rate in sheep.J Reprod Fertil 80, 317–320.Parr RA, Davis IF, Miles MA, Squires TJ, 1993: Liver bloodflow and metabolic clearance rate of progesterone in sheep.Res Vet Sci 55, 311–316.Patton J, Kenny DA, McNamara S, Mee JF, O’Mara FP,Diskin MG, Murphy JJ, 2006: Relationships between milkproduction, energy balance, plasma analytes and reproductionin Holstein–Friesian cows. J Dairy Sci 90, 649–658.Persson A, 2003: Effect of the genetic defect complex vertebralmalformation on fertility in Swedish Holstein cattle.Undergraduate Thesis, Swedish Agricultural University,Uppsala.Quirke JF, Hanrahan JP, 1977: Comparison of the survival inthe uteri of adult ewes of cleaved ova from adult ewes andewe lambs. J Reprod Fertil 51, 487–489.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Embryo ⁄ Foetal Losses in Ruminants 267Quirke JF, Hanrahan JP, 1983: Comparison of the survival offertilized eggs from adult ewes in the uteri of adult ewes andewe lambs. J Reprod Fertil 68, 289–294.Restall BJ, Brown GH, Blickey MdeB, Cahill L, Kearns R,1976: Assessment of reproductive wastage in sheep. 1.Fertilization failure and early embryonic survival. Aust JExp Agric Anim Husb 16, 329–335.Rhind SM, Robinson JJ, McDonald I, 1980: Relationshipamong uterine and placental factors in prolific ewes andtheir relevance to variation in foetal weight. Anim Prod 30,115–124.Ricordeau G, Fazungles J, Lajous D, 1982: Heritability ofovulation rate and level of embryonic losses in Romanovbreed. 2nd Wld Congr Appl Livest Prod 7, 591–597.Ricordeau G, Poivey JP, Lajous D, Eychenne R, 1986: Geneticaspects of ovulation rate and embryo mortality in Romanovewes 3rd Wld Congr Appl Livest Prod 11, 90–95.Robinson JL, Dombrowski DB, Harpestad GW, Shanks RD,1984: Detection and prevalence of UMP synthase deficiencyamong dairy cattle. J Hered 75, 277–280.Rohrer GA, 2004: An overview of genomics research and itsimpact on livestock reproduction. Reprod Fertil Dev 16, 47–54.Ryan DP, Prichard JF, Kopel E, Godke RA, 1993: Comparingearly embryo mortality in dairy cows during hot and coolseasons of the year. Theriogenology 39, 719–737.Sangsritavong S, Combs DK, Sartori RF, Armentano LE,Wiltbank MC, 2002: High feed intake increases liver bloodflow and metabolism of progesterone and estradiol 17b indairy cattle. J Dairy Sci 85, 2831–2842.Sartori R, Sartori-Bergfelt R, Mertens SA, Guenther JN,Parish JJ, Wiltbank MC, 2002: Fertilization and earlyembryonic development in heifers and lactating cows insummer and lactating and dry cows in winter. J Dairy Sci 85,2803–2812.Shanks RD, Robinson JL, 1989: Embryonic mortality attributedto inherited deficiency of uridine monophosphatesynthase. J Dairy Sci 72, 3035–3039.Silke V, Diskin MG, Kenny DA, Boland MP, Dillon P, MeeJF, Sreenan JM, 2001: Extent, pattern and factors associatedwith late embryonic loss in dairy cows. Anim ReprodSci 15, 1–12.Sreenan JM, Diskin MG, 1986: The extent and timing ofembryonic mortality in cattle. in: Sreenan JM, Diskin MG(eds), Embryonic Mortality in Farm Animals. MartinusNijhoff, CEC, Brussels, Belgium, pp. 142–158.Sreenan JM, Diskin MG, Morris DG, 2001: Embryo survivalin cattle, a major limitation to the achievement of highfertility. BSAS Occas Publ, 26, 93–104.Starbuck GR, Darwash AO, Mann GE, Lamming GE, 2001:The detection and treatment of post insemination progesteroneinsufficiency in dairy cows. BSAS Occas Publ 26,447–450.Stronge AJH, Sreenan JM, Diskin MG, Mee JF, Kenny DA,Morris DG, 2005: Post-insemination milk progesteroneconcentration and embryo survival in dairy cows. Theriogenology64, 1212–1224.VanRaden PM, Miller RH, 2006: Effects of nonadditivegenetic interactions, inbreeding and recessive defects onembryo and fetal loss by seventy days. J Dairy Sci 89, 2716–2721.Wall E, Brotherstone S, Kearney JF, Wolliams JA, CoffeyMP, 2003: Effect of including inbreeding, heterosis andrecombination loss in prediction of breeding values forfertility traits. Interbull Bull 31, 117–121.Wiebold JL, 1988: Embryonic mortality and the uterineenvironment in first service lactating dairy cows. J ReprodFertil 84, 393–399.Zavy MT, 1994: Embryonic mortality in cattle. In: GeisertRD, Zavy MT (eds), Embryonic Mortality in DomesticSpecies. CRC Press, Boca Raton, FL, pp. 99–140.Author’s address (for correspondence): MG Diskin, Teagasc, AnimalProduction Research Centre, Mellows Campus, Athenry, Co. Galway,Ireland. E-mail: michael@diskin@teagasc.ieConflict of interest: MG Diskin is a member of a luster that hasrecently been awarded A Science Foundation Ireland Grant; DGMorris declares no conflict of interest.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 268–272 (2008); doi: 10.1111/j.1439-0531.2008.01172.xISSN 0936-6768Role of Cell Death Ligand and Receptor System on Regulation of Follicular Atresiain Pig OvariesN Manabe, F Matsuda-Minehata, Y Goto, A Maeda, Y Cheng, S Nakagawa, N Inoue, K Wongpanit, H Jin, H Gonda and J LiAnimal Resource Science Center, University of Tokyo, Kasama, JapanContentsSeveral hundred thousand primordial follicles are present inthe mammalian ovary, however, only a limited numberdevelop to the pre-ovulatory stage, and then finally ovulate.The others, more than 99%, will be eliminated through adegenerative process called ‘atresia’. The endocrinologicalregulatory mechanisms involved in follicular developmentand atresia have been characterized to a large extent, but theprecise temporal and molecular mechanisms involved in theregulation of these events have remained unknown. Frommany recent studies, it is suggested that the apoptosis inovarian granulosa cells plays a crucial role in follicular atresia.Notably, death ligand–receptor interaction and subsequentintracellular signalling have been demonstrated to be the keymechanisms regulating granulosa cell apoptosis. In this review,we provide an overview of granulosa cell apoptosis regulatedby death ligand–receptor signalling. The roles of death ligandsand receptors [Fas ligand (FasL)–Fas, tumour necrosis factor(TNF)a–TNF receptor (TNFR), and TNFa-related apoptosisinducingligand (TRAIL)–TRAIL receptor (TRAILR)] andintracellular death-signal mediators [Fas-associated deathdomain protein (FADD), TNF receptor 1-associated deathdomain protein (TRADD), caspases, apoptotic protease-activatingfactor 1 (Apaf1), TNFR-associated factor 2 (TRAF2),and cellular FLICE-like inhibitory protein (cFLIP), etc.] ingranulosa cells will be discussed.IntroductionIn cells of both vertebrate and invertebrate animalspecies, apoptosis plays a significant role in almost allphysiological functions. Apoptosis is a form of celldeath essential for the elimination of cells that aredamaged, senescent, potentially harmful, or no longeruseful. Apoptosis is characterized by internucleosomalDNA fragmentation, cell shrinkage, plasma membraneblebbing, and the formation of apoptotic bodies (Schwartzmanand Cidlowski 1993). Stimulation by deathligands or deprivation of key survival-promoting growthfactors is the main contributor to apoptosis, while stressinducers, including drugs, toxicants, oxidative stress,and radiation, are also known to cause apoptosis(Hengartner 2000). Recent studies have revealed thatapoptosis also plays a crucial role in maintainingovarian homeostasis in mammals (Hughes and Gorospe1991; Tilly et al. 1991; Manabe et al. 1996; Kaipia andHsueh 1997). During follicular growth and development,more than 99% of follicles disappear primarilydue to apoptosis of granulosa cells as the biochemicaland morphological characteristics of apoptosis havebeen observed in the granulosa cells of atretic folliclesgrowth (Grant et al. 1989; Guthrie et al. 1995). Apoptoticstimuli and intracellular signal transduction pathwaysinvolved in the apoptosis of granulosa cells remainto be determined, and investigators are studyingpotential triggers of apoptosis and how intracellularapoptotic signals are transmitted in granulosa cells(Manabe et al. 2003, 2004). Many apoptosis-relatedfactors are implicated in follicular atresia, includingdeath ligands and receptors, intracellular pro- and antiapoptoticmolecules, cytokines, and growth factors(Matsuda-Minehata et al. 2006). In particular, cell deathligand–receptor signalling has been revealed to be themajor regulatory system for apoptosis in granulosa cells(Nakayama et al. 2003; Matsuda-Minehata et al. 2008).Follicular Growth, Development and Atresia inPig OvaryAfter puberty, a number of primordial follicles start togrow during each oestrus cycle in adult females.Initiation of follicular growth involves endocrinologicalfactors, mainly FSH, and local modulating factors fromgranulosa cells, theca cells, stromal-interstitial cells, andoocytes (Hirshfield 1991; Guthrie et al. 1995). Primaryfollicles (follicles with a monolayer of follicular epithelial⁄ granulosa cells) develop into secondary follicles(follicles with stratified granulosa cells but without anantrum) and subsequently into tertiary follicles (follicleswith a follicular antrum) (Grant et al. 1989). Due to alarge increase in the proliferation of granulosa cells andan increase in the size of the antrum, tertiary folliclesshow an exponential rate of growth. In the oocyte,meiosis then restarts and the first polar body divides.Finally, selected follicles burst, and the oocytes ovulate.With increasing serum FSH concentrations at the startof the oestrous cycle, follicles produce increasingamounts of oestrogen and inhibin, produced by granulosacells. As a feedback mechanism by inhibin, FSHsecretion falls, and the remaining small follicles undergoatresia. Although atresia can occur at any time duringfollicular development, the majority of follicles becomeatretic during the early antral stage of development. Thetransition from pre-antral to antral follicular developmentoccurs after exposure of the granulosa cells togonadotropin. Then, the differentiation of granulosacells is initiated, which renders them susceptible toapoptosis. The endonuclease, DNase-I was demonstratedto exist in granulosa cells from antral, but notpre-antral, follicles (Manabe et al. 1996a,b). However,the presence of the endonuclease is not sufficient tocause apoptosis; a signal to activate DNase-I and inducecell death is required. Death ligand–receptor interactionis thought to be a major trigger of apoptosis ingranulosa cells (Manabe et al. 2003, 2004).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Death Ligand and Receptor Pig Ovarian Follicle 269Death Ligand and Receptor System RegulatesFollicular Atresia in Pig OvaryDeath ligands, which are grouped the tumour necrosisfactor (TNF) family, are synthesized as type-II membraneproteins (Ashkenazi and Dixit 1998). Most deathligands can be cleaved from the cell membrane tobecome a soluble form and act as trimers. Deathreceptors, which are classified as type-I membraneproteins, constitute a subfamily within the TNF receptorsuperfamily (TNFRsf) that has a cytoplasmic deathdomain (DD) necessary for the activation of apoptosis(Hengartner 2000). These receptors are trimerized andthen bind to death ligands, which are a trigger forapoptosis. Cell death ligand–receptor systems known inmammals include the Fas ligand (FasL) and Fas (CD95,APO-1 or TNFRsf6), TNFa and TNFa receptors(TNFRs), and TNFa-related apoptosis-inducing ligand(TRAIL, Apo2L) and TRAIL receptors (Nagata 1997;Ashkenazi and Dixit 1998; Wallach et al. 1999). In mostcases, the cell death receptor-mediated apoptotic signallingpathway is as follows: (1) Cell death ligands,binding with cell membranes or existing as a solubleform, bind to the extracellular domain of trimerized celldeath receptors, each of which contains an intracellularDD. (2) The DD of the death receptor binds with theDD of the adaptor proteins (TNF receptor 1-associateddeath domain protein: TRADD and Fas-associateddeath domain protein: FADD) through homophilicinteraction (Inoue et al. 2007). (3) An initiator caspase(procaspase-8) binds to FADD through homophilicinteraction with the death effector domain (DED; theresulting complex is called the ‘death-inducing signallingcomplex’: DISC) (Medema et al. 1997). (4) Dimerizationof procaspase-8 induces auto-proteolytic cleavageand activation (Boldin et al. 1995). (5) The activatedcaspase-8 subsequently activates downstream caspaseseither directly (‘type I’) or via mitochondrial perturbation(‘type II; mitochondrion-dependent’) (Matsui et al.2003). (6-a) In type I apoptotic cells, caspase-8 directlyactivates the effector enzyme, caspase-3. (7-a) Trancatedcaspase-3, active form, activates endogenous endonucleasesthat results in apoptosis. (6-b) In type II apoptoticcells, caspase-8’s activation leads to the release ofcytochrome c from mitochondrion, which results in theinteraction of procaspase-9 with apoptotic proteaseactivatingfactor 1 (Apaf1) [formation of cytochromec-Apaf1-caspase-9 complex (apoptosome)]. (7-b) Activatedcaspase-9 cleaves procaspase-3. (8-b) Active caspase-3activates the endonucleases that results inapoptosis (Nagata 1997; Ashkenazi and Dixit 1998;Matsui et al. 2003) (Fig. 1).FasL–Fas and other cell death ligand–receptor systems ingranulosa cellThe FasL–Fas system is one of the most studiedparadigms of instructive apoptosis, with strong apoptosis-inducingactivity (Wallach et al. 1999). The apoptosissignal triggered and mediated by FasL and Fas isas follows: (1) FasL binds to the extracellular domainof Fas. (2) The intracellular DD of Fas interacts withthe adaptor protein (FADD) through DD. (3) FADDDiscFasLFasFADDActivated caspase-8ApoptosisProcaspase-8: Death domain (DD): Death effector domain (DED): Enzymatic domainCell membraneFig. 1. Intracellular signalling in TNF family. When cell death ligandbinds to the extracellular region of cell death receptor, adoptor (Fasassociateddeath domain protein: FADD) binds to the intracellularregion of receptor through the homophilic interaction of the deathdomains (DD). Then, initiator caspase (procaspase-8) binds to FADDthrough the homophilic interaction of the death effector domain(DED). Dimerized procaspase-8 is auto-cleaved, and then truncatedcaspase-8 (active caspase-8) induces activation of down-stream caspases(ex. caspase-3), which results in apoptosisinteracts with procaspase-8 through their DED and theactive caspase-8 is formed. (4) A caspase cascade isactivated and eventually apoptosis is induced. TheFasL–Fas system is the most characterized apoptosissignalling machinery in granulosa cells (Inoue et al.2006). In many species, both FasL and Fas areexpressed in granulosa cells. In murine ovaries, FasLand Fas mRNA and protein are expressed in granulosacells of healthy and atretic follicles (Sakamaki et al.1997). Moreover, treatment of female mice with Fasactivatingantibody promoted granulosa cell apoptosisand follicular atresia, suggesting a pro-apoptotic functionof FasL–Fas signalling in vivo. In rat ovaries, FasLand Fas protein occur in granulosa cells, which tend tobe more abundant in atretic follicles than healthyfollicles (Hakuno et al. 1996). In human females, thegranulosa cells of antral follicles express Fas during theearly stages of atresia, the levels of Fas expressionincrease as atresia progresses, and granulosa FasLprotein also increases with atresia in antral follicles(Cataldo et al. 2000). In bovine ovaries, FasL mRNAlevels are higher in granulosa cells from atretic folliclesthan those from healthy follicles, and Fas mRNAexpression is stronger in granulosa cells of atretic thanhealthy follicles (Porter et al. 2000, 2001). Moreover,CD45, a pan-leucocyte marker, was not detected inbovine granulosa cells, indicating that immune cellswere not a source of FasL or Fas in the granulosa celllayer (Hu et al. 2001). Trace levels of FasL and FasmRNA and protein were detected in granulosa cells ofhealthy follicles but the levels increased during atresiain porcine ovaries (Inoue et al. 2006). These expressionpatterns of FasL and Fas strongly indicate the importantrole of FasL–Fas signalling in follicular atresia.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

270 N Manabe et al.Both FasL and Fas exist on the granulosa cellmembrane and their interaction may induce apoptosis.The pro-apoptotic effect of the FasL–Fas system ingranulosa cells was also demonstrated by experimentsin vitro. Stimulation of Fas-signalling can induceapoptosis in primary cultured granulosa cells of manyspecies including pigs (Matsuda-Minehata et al. 2006).However, pre- or co-treatment with interferon-c orcycloheximide (CHX) is necessary for inducing theapoptosis in vitro, indicating the existence of otheressential factor(s) in the FasL–Fas signalling pathwayof granulosa cell apoptosis. Moreover, we found notonly FasL–Fas system, but also TRAIL–TRAILRs(Wada et al. 2002; Inoue et al. 2003), TNFa–TNFRs(Nakayama et al. 2003), and unknown ligand–PFGreceptors (Manabe et al. 2000) systems in porcinegranulosa cells. However, we have no confirmedinformation which system plays dominant role inregulation of granulosa cell apoptosis (Matsuda-Minehataet al. 2008).Intracellular regulator (apoptosis-inhibitory factor) inporcine granulosa cellDeath ligand–receptor interaction is indeed essentialfor triggering apoptotic signalling, however, it does notnecessarily result in apoptotic cell death, indicating theimportance of intracellular inhibitors of the apoptoticsignalling pathway. Recently, we found that cellularFLICE-like inhibitory protein (cFLIP: also calledCASH, Casper, CLARP, FLAME, I-FLICE, MRIT,or usurpin) (Goltsev et al. 1997; Han et al. 1997; Huet al. 1997; Krueger et al. 2001), which is a homologueof procaspase-8 (also called FLICE) (Muzio et al.1996) and one of the intracellular proteins thatinterferes with the apoptotic effects of death ligands,plays crucial role in regulation of granulosa cellapoptosis in porcine ovaries (Goto et al. 2004).FLICE-like inhibitory protein was firstly identified inseveral viruses as viral FLIP (vFLIP), which containstwo DEDs that interact with FADD to avoid thehost’s apoptotic response (Thome et al. 1997). Inmammalian cell, homologue of vFLIP was found andnamed cFLIP (Irmler et al. 1997). There are twosplicing variants of cFLIP, short and long forms(cFLIP S and cFLIP L , respectively). cFLIP S is verysimilar in structure to vFLIP, containing two DEDs,while cFLIP L contains an additional pseudo-enzymaticdomain that is similar to the enzymatic domain ofprocaspase-8 but lacks enzymatic activity. Recently, wefound that cFLIP S and cFLIP L are expressed inporcine granulosa cells and both of them to beimportant regulators of apoptosis that block deathligand-inducible apoptosis, by competing with procaspase-8and inhibiting the activation of caspase-8(Thome and Tschopp 2001). The homology of porcinecFLIP with human and murine cFLIP is very high(more than 75% for both the mRNA and amino acidlevels) (Goto et al. 2004), and we have proposed thatcFLIP also has cell survival-promoting effects in thepig. As described above, the FasL–Fas system is wellcharacterizedas a pro-apoptotic signal in granulosacells in sows, and the expression of FasL and Fas ingranulosa cells increases during atresia, however, bothproteins are also expressed in granulosa cells of healthypre-antral and antral follicles, which rapidly grow andhave many proliferating granulosa cells. Moreover, Fascannot induce apoptosis in primarily cultured porcinegranulosa cells without either IFN-c or CHX. It hasbeen suggested that the factor(s) that blocks FasL–Fasmediatedapoptotic signalling is essential for maintaininggranulosa cells and keeping follicles healthy. Byreverse-transcription-polymerase chain reaction (RT-PCR) and Western blotting, the mRNA and protein ofcFLIP L were found to be highly expressed in thegranulosa cells of healthy follicles and decreased duringatresia (Goto et al. 2004). The mRNA levels of cFLIP Sin granulosa cells are low and showed no changesamong the stages of follicular development (Matsuda-Minehata et al. 2005, 2007). Furthermore, the proteinlevel of cFLIP S is extremely low. By in situ hybridization,cFLIP L was found to be abundant in thegranulosa cells of healthy follicles in comparison withthose of atretic follicles. Immunohistochemical analysesshowed that cFLIP protein was found to be highlyexpressed in the granulosa cells of healthy follicles butweakly expressed in those of atretic follicles. Wepresume that cFLIP, especially cFLIP L , plays ananti-apoptotic role in the granulosa cells of healthyfollicles from pig ovaries (Fig. 2). Since the antiapoptoticactivity of porcine cFLIP (pcFLIP) had notbeen confirmed in granulosa cells, we examined theeffect of pcFLIP on survival using granulosa-derivedcells (Matsuda-Minehata et al. 2007). Human ovariangranulosa tumour cell derived KGN cells (Nishi et al.2001) transfected with pcFLIP S or pcFLIP L vectorssurvive the induction of ant-Fas antibody-Fas-mediatedapoptosis, while almost all cells transfected withempty vector die, indicating the anti-apoptotic activityof pcFLIP in granulosa cells. When both cFLIP S andcFLIP L , or cFLIP L only, were suppressed by smallinterfering RNA (siRNA), the viability of KGN and Jporcine granulosa-derived JC-410 cells (Chedrese et al.1998) decreased significantly. Thus, we conclude thatporcine cFLIP functions as an anti-apoptotic factor ingranulosa-derived cells. These our findings stronglysuggest that cFLIP acts as a survival-promoting factorin granulosa cells and determines whether porcineovarian follicles survive or undergo atresia. To date,investigations have revealed that cFLIP can also inhibitTNFa- and TRAIL-signalling (Cheng et al. 2007), notonly Fas-signalling (Manabe et al. 2003). Furtherinvestigations are needed to determine whether cFLIPaffects intracellular signalling transduction in other celldeath ligand and receptor systems (e.g. TNFa–TNFRs,TRAIL–TRAILRs, etc.) in porcine granulosa cells ornot. More researches are also needed how to regulatethe expression of cFLIP in granulosa cells duringfollicular growth, development, and atresia. Our previousfindings have shown that interleukin-6 up-regulatesTNFa expression in porcine granulosa cells(Maeda et al. 2007a,b), and that TNFa up-regulatescFLIP L expression and acts a survival factor (Nakayamaet al. 2003). However, there is no confirmedinformation on what is the initial trigger to controlcFLIP expression.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Death Ligand and Receptor Pig Ovarian Follicle 271Fig. 2. Working hypothesis on regulation of apoptosis in porcine granulosa cells of healthy (left) and atretic follicles (right). Although the celldeath ligands and receptors are expressed and interact on granulosa cells of healthy follicles, subsequent apoptotic signalling is blocked bycFLIP L and ⁄ or cFLIP S , dominantly cFLIP L . As a result, the apoptosis of the granulosa cells is avoided and the follicle is kept healthy. When thecFLIP L and cFLIP S are down-regulated, cell death ligand and receptor interaction causes cleavage of procaspase-8, and subsequent apoptoticsignalling. As a result, the apoptosis of the granulosa cells is induced, and the follicle undergoes atresiaConclusionTo date, inducers of apoptosis including FasL–Fas,TNFa–TNFRs, TRAIL–TRAILRs have been the maintargets in the studies of regulation mechanism ofgranulosa cell apoptosis, and their contributions tofollicular atresia have been clarified. However, it hasbecome apparent that an intracellular anti-apoptotic⁄ inhibitory factor(s), like cFLIP, is critical for inhibitinggranulosa cell apoptosis from our recent research.Granulosa cell apoptosis is likely to be regulated by asophisticated balance of pro-apoptotic and anti-apoptoticfactors. The mechanisms of death ligand–receptorsignalling should be determined to entirely definegranulosa cell apoptosis. Solving this problem will helpto establish methods (1) of selecting healthy oocytes orto improve damaged oocytes that result in an increasedrate of gestation for in vitro fertilization in domesticanimals and humans, and (2) of rescue oocytes in theovaries of wild animals and livestock sacrificed at theslaughterhouse.AcknowledgementsOur works were supported by Grant-in-Aid for Scientific Research(13GS0008, S16108003, B18380164, E18658105, and S18108004) to N.M. and by Research Fellowship for Young Scientists to F. M-M. fromthe Japan Society for the Promotion of Science.ReferencesAshkenazi A, Dixit VM, 1998: Death receptors: signaling andmodulation. Science 281, 1305–1308.Boldin MP, Mett IL, Varfolomeev EE, Chumakov I, Shemer-Avni Y, Camonis JH, Wallach D, 1995: Self-association ofthe ‘‘death domains’’ of the p55 tumor necrosis factor(TNF) receptor and Fas ⁄ APO1 prompts signaling for TNFand Fas ⁄ APO1 effects. J Biol Chem 270, 387–391.Cataldo NA, Dumesic DA, Goldsmith PC, Jaffe RB, 2000:Immunolocalization of Fas and Fas ligand in the ovaries ofwomen with polycystic ovary syndrome: relationship toapoptosis. Hum Reprod 15, 1889–1897.Chedrese PJ, Rodway MR, Swan CL, Gillio-Meina C, 1998:Establishment of a stable steroidogenic porcine granulosacell line. J Mol Endocrinol 20, 287–292.Cheng Y, Maeda A, Goto Y, Matsuda-Minehata F, ManabeN, 2007: Molecular cloning of porcine (Sus scrofa) tumornecrosis factor receptor 2. J Reprod Dev 53, 1291–1297.Goltsev YV, Kovalenko AV, Arnold E, Varfolomeev EE,Brodianskii VM, Wallach D, 1997: CASH, a novel caspasehomologue with death effector domains. J Biol Chem 272,19641–19644.Goto Y, Matsuda-Minehata F, Inoue N, Matsui T, Maeda A,Manabe N, 2004: Porcine (Sus scrofa) cellular FLICE-likeinhibitory protein (cFLIP): molecular cloning and comparisonwith the human and murine cFLIP. J Reprod Fertil 50,549–555.Grant SA, Hunter MG, Foxcroft GR, 1989: Morphologicaland biochemical characteristics during ovarian folliculardevelopment in the pig. J Reprod Fertil 86, 171–183.Guthrie HD, Cooper BS, Welch GR, Zakaria AD, JohnsonLA, 1995: Atresia in follicle grown after ovulation in the pig:Measurement of increased apoptosis in granulosa cells andreduced follicular fluid estradiol-17b. Biol Reprod 52, 920–927.Hakuno N, Koji T, Yano T, Kobayashi N, Tsutsumi O,Taketani Y, Nakane PK, 1996: Fas ⁄ APO-1 ⁄ CD95 system asa mediator of granulosa cell apoptosis in ovarian follicleatresia. Endocrinology 137, 1938–1948.Han DK, Chaudhary PM, Wright ME, Friedman C, Trask BJ,Riedel RT, Baskin DG, Schwartz SM, Hood L, 1997:MRIT, a novel death-effector domain-containing protein,interacts with caspases and BclXL and initiates cell death.Proc Natl Acad Sci U S A 94, 11333–11338.Hengartner MO, 2000: The biochemistry of apoptosis. Nature407, 770–776.Hirshfield AN, 1991: Development of follicles in the mammalianovary. Int Rev Cytol 124, 43–101.Hu S, Vincenz C, Ni J, Gentz R, Dixit VM, 1997: I-FLICE, anovel inhibitor of tumor necrosis factor receptor-1- and CD-95-induced apoptosis. J Biol Chem 272, 17255–17257.Hu CL, Cowan RG, Harman RM, Porter DA, Quirk SM,2001: Apoptosis of bovine granulosa cells after serumwithdrawal is mediated by Fas antigen (CD95) and Fasligand. Biol Reprod 64, 518–526.Hughes FM Jr, Gorospe WC, 1991: Biochemical identificationof apoptosis (programmed cell death) in granulosa cells:evidence for a potential mechanism underlying follicularatresia. Endocrinology 129, 2415–2422.Inoue N, Manabe N, Matsui T, Maeda A, Nakagawa S, WadaS, Miyamoto H, 2003: Roles of tumor necrosisÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

272 N Manabe et al.factor-related ligand signaling pathway in granulosa cellapoptosis during atresia in pig ovaries. J Reprod Dev 49,313–321.Inoue N, Maeda A, Matsuda-Minehata F, Fukuta K, ManabeN, 2006: Expression and localization of Fas ligand and Fasduring atresia in porcine ovarian follicles. J Reprod Dev 52,723–730.Inoue N, Matsuda-Minehata F, Manabe N, 2007: Molecularcharacteristics of porcine Fas-associated death domain(FADD) and procaspase-8. J Reprod Dev 53, 427–436.Irmler M, Thome M, Hahne M, Schneider P, Hofmann K,Steiner V, Bodmer JL, Schroter M, Burns K, Mattmann C,Rimoldi D, French LE, Tschopp J, 1997: Inhibition of deathreceptor signals by cellular FLIP. Nature 388, 190–195.Kaipia A, Hsueh AJ, 1997: Regulation of ovarian follicleatresia. Ann Rev Physiol 59, 349–363.Krueger A, Baumann S, Krammer PH, Kirchhoff S, 2001:FLICE-inhibitory proteins: regulators of death receptormediatedapoptosis. Mol Cell Biol 21, 8247–8254.Maeda A, Inoue N, Matsuda-Minehata F, Cheng Y, ManabeN, 2007a: The role of interleukin-6 in the regulation ofgranulosa cell apoptosis during follicular atresia in pigovaries. J Reprod Dev 53, 481–490.Maeda A, Goto Y, Matsuda-Minehata F, Cheng Y, Inoue N,Manabe N, 2007b: Changes in expression of interleukin-6receptors in granulosa cells during follicular atresia in pigovaries. J Reprod Dev 53, 727–736.Manabe N, Imai Y, Ohno H, Takahagi Y, Sugimoto M,Miyamoto H, 1996a: Apoptosis occurs in granulosa cells butnot cumulus cells in the atretic antral follicles in pig ovaries.Experientia 52, 647–651.Manabe N, Imai Y, Kimura Y, Myoumoto A, Sugimoto M,Miyamoto H, 1996b: Ca 2+ ⁄ Mg 2+ -dependent endonucleasebut not Ca 2+ -dependent, Mg 2+ -dependent or cation-independentendonuclease is involved in granulosa cell apoptosisof pig atretic follicles. J Reprod Dev 42, 247–253.Manabe N, Myoumoto A, Tajima C, Fukumoto M, NakayamaM, Uchio K, Yamaguchi M, Miyamoto H, 2000:Immunochemical characteristics of a novel cell deathreceptor and a decoy receptor on granulosa cells of porcineovarian follicles. Cytotechnology 33, 189–201.Manabe N, Inoue N, Miyano T, Sakamaki K, Sugimoto M,Myamoto H, 2003: Ovarian follicle selection in mammalianovaries: regulatory mechanisms of granulosa cell apoptosisduring follicular atresia. In: Leung PK, Adashi E (eds), TheOvary, 2nd edn. Academic Press and Elsevier, Amsterdam,pp. 369–385.Manabe N, Goto Y, Matsuda-Minehata F, Inoue N, MaedaA, Sugimoto M, Sakamaki K, Miyano T, 2004: Regulationmechanism of selective atresia in porcine follicles: regulationof granulosa cell apoptosis during atresia. J Reprod Dev 50,493–514.Matsuda-Minehata F, Goto Y, Inoue N, Manabe N, 2005:Changes in expression of anti-apoptotic protein, cFLIP, ingranulosa cell during follicular atresia in porcine ovaries.Mol Reprod Dev 72, 145–151.Matsuda-Minehata F, Inoue N, Goto Y, Manabe N, 2006:The regulation of ovarian granulosa cell death by pro- andanti-apoptotic molecules. J Reprod Dev 52, 695–705.Matsuda-Minehata F, Goto Y, Inoue N, Sakamaki K, PChedrese PJ, Manabe N, 2007: Anti-apoptotic activity ofporcine cFLIP in ovarian granulosa cell lines. Mol ReprodDev 74, 1165–1170.Matsuda-Minehata F, Maeda A, Cheng Y, Sai T, Gonda H,Manabe N, 2008: Regulation of granulosa cell apoptosis bydeath ligand-receptor signaling. Anim Sci J 79, 1–10.Matsui T, Manabe N, Goto Y, Inoue N, Nishihara S,Miyamoto H, 2003: Expression and activity of Apaf1 andcaspase-9 in granulosa cells during follicular atresia in pigovaries. Reproduction 126, 113–120.Medema JP, Scaffidi C, Kischkel FC, Shevchenko A, MannM, Krammer PH, Peter ME, 1997: FLICE is activated byassociation with the CD95 death-inducing signaling complex(DISC). EMBO J 16, 2794–2804.Muzio M, Chinnaiyan AM, Kischkel FC, O’Rourke K,Shevchenko A, Ni J, Scaffidi C, Bretz JD, Zhang M, GentzR, Mann M, Krammer PH, Peter ME, Dixit VM, 1996:FLICE, a novel FADD-homologous ICE ⁄ CED-3-like protease,is recruited to the CD95 (Fas ⁄ APO-1) death-inducingsignaling complex. Cell 85, 817–827.Nagata S, 1997: Apoptosis by death factor. Cell 88, 355–365.Nakayama M, Manabe N, Inoue N, Matsui T, Miyamoto H,2003: Changes in the expression of tumor necrosis factor(TNF)a, TNFa receptor (TNFR) 2, TNFR-associatedfactor 2 in granulosa cells during atresia in pig ovaries. BiolReprod 68, 530–535.Nishi Y, Yanase T, Mu Y, Oba K, Ichino I, Saito M, NomuraM, Mukasa C, Okabe T, Goto K, Takayanagi R, KashimuraY, Haji M, Nawata H, 2001: Establishment andcharacterization of a steroidogenic human granulosa-liketumor cell line, KGN, that expresses functional folliclestimulatinghormone receptor. Endocrinology 142, 437–445.Porter DA, Vickers SL, Cowan RG, Huber SC, Quirk SM,2000: Expression and function of Fas antigen vary in bovinegranulosa and theca cells during ovarian follicular developmentand atresia. Biol Reprod 62, 62–66.Porter DA, Harman RM, Cowan RG, Quirk SM, 2001:Relationship of Fas ligand expression and atresia duringbovine follicle development. Reproduction 121, 561–566.Sakamaki K, Yoshida H, Nishimura Y, Nishikawa S, ManabeN, Yonehara S, 1997: Involvement of Fas antigen in ovarianfollicular atresia and luteolysis. Mol Reprod Dev 47, 11–18.Schwartzman RA, Cidlowski JA, 1993: Apoptosis: the biochemistryand molecular biology of programmed cell death.Endocr Rev 14, 133–151.Thome M, Tschopp J, 2001: Regulation of lymphocyteproliferation and death by FLIP. Nat Rev Immunol 1, 50–58.Thome M, Schneider P, Hofmann K, Fickenscher H, Meinl E,Neipel F, Mattmann C, Burns K, Bodmer JL, Schroter M,Scaffidi C, Krammer PH, Peter ME, Tschopp J, 1997: ViralFLICE-inhibitory proteins (FLIPs) prevent apoptosisinduced by death receptors. Nature 386, 517–521.Tilly JL, Kowalski KI, Johnson AL, Hsueh AJ, 1991:Involvement of apoptosis in ovarian follicular atresia andpostovulatory regression. Endocrinology 129, 2799–2801.Wada S, Manabe N, Inoue N, Nakayama M, Matsui T,Miyamoto H, 2002: TRAIL-decoy receptor-1 disappears ingranulosa cells of atretic follicles in porcine ovaries. JReprod Dev 48, 167–173.Wallach D, Varfolomeev EE, Malinin NL, Goltsev YV,Kovalenko AV, Boldin MP, 1999: Tumor necrosis factorreceptor and Fas signaling mechanisms. Ann Rev Immunol17, 331–367.Author’s address (for correspondence): N Manabe, Animal ResourceScience Center, University of Tokyo, Kasama 319-0206, Japan.E-mail: amanabe@mail.ecc.u-tokyo.ac.jpConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 273–279 (2008); doi: 10.1111/j.1439-0531.2008.01174.xISSN 0936-6768Epigenetic Programming of Porcine Endometrial Function and the LactocrineHypothesisFF Bartol 1 , AA Wiley 1 and CA Bagnell 21 Departments of Animal Sciences and Anatomy, Physiology and Pharmacology, Cellular and Molecular Biosciences Program, Auburn University,Auburn, AL, USA; 2 Department of Animal Sciences, Endocrinology and Animal Biosciences Program, Rutgers University, New Brunswick, NJ, USAContentsEpigenetic programs controlling development of the femalereproductive tract (FRT) are influenced by the effects ofnaturally occurring bioactive agents on patterns of geneexpression in FRT tissues during organizationally criticalperiods of foetal and perinatal life. Aberrations in suchimportant cellular and molecular events, as may occur withexposure to natural or manmade steroid or peptide receptormodulatingagents, disrupt the developmental program andcan change the developmental trajectory of FRT tissues,including the endometrium, with lasting consequences. In thepig, as in other mammals, maternal programming of FRTdevelopment begins pre-natally and is completed postnatally,when maternal effects on development can becommunicated via signals transmitted in milk. Studiesinvolving relaxin (RLX), a prototypic milk-borne morphoregulatoryfactor (MbF), serve as the basis for ongoingefforts to identify maternal programming events that affectuterine and cervical tissues in the neonatal pig. Data supportthe lactocrine hypothesis for delivery of MbFs to neonatesas a specific consequence of nursing. Components of amaternally driven lactocrine mechanism for RLX-mediatedsignalling in neonatal FRT tissues, including evidence thatmilk-borne RLX is delivered into the neonatal circulationwhere it can act on RLX receptor (RXFP1) -positiveneonatal tissues to affect their development, are in place inthe pig. The fact that all newborn mammals drink milkextends the timeframe of maternal influence on neonataldevelopment across many species. Thus, lactocrine transmissionof milk-borne developmental signals is an element ofthe maternal epigenetic programming equation that deservesfurther study.IntroductionIn the pig (Sus scrofa domesticus), as in other mammals,development of the female reproductive tract (FRT)begins pre-natally but is completed post-natally (Yinand Ma 2005; Bartol et al. 2006). Structural patterning(morphogenesis) and functional programming (cytodifferentiation)of epithelial-mesenchymal tissues derivedfrom the Mu¨llerian ducts, including the oviducts, uterus,cervix and anterior vagina, are coupled processessupported by the progressive generation of increasinglycomplex and specific cellular relationships and interactions(Gray et al. 2001; Yin and Ma 2005). Over time,these interactions drive the evolution of organizationallycritical, temporally and spatially unique morphoregulatorygene expression domains that define micro-environmentalconditions which, in turn, direct and specifycell fate, dictate patterns of development, and determinecell and tissue identity and functionality. For a giventissue, this complex sequence of events defines thedevelopmental program and, in so doing, establishes adevelopmental trajectory for cells and tissues that willeventually dictate phenotype (Burggren 1999).Genetic potential for developmental success is definedat conception. Thereafter, if development proceedsalong a normal course through embryonic, foetal andperinatal life, a normal phenotypic trajectory is establishedand an optimal phenotypic outcome is realized(Fig. 1). However, developmental programs can bedisrupted by a host of epigenetic factors of both biotic(physiological, endocrinological, metabolic) and abiotic(anthropogenic ⁄ macro-environmental) origin (Burggren1999; Bartol 2002). Disruption of the developmentalprogram during organizationally critical periods, whencells and tissues are uniquely sensitive to aberrantstimuli, alters micro-environmental conditions requiredfor normal development. Divergence from the normaldevelopmental program can be sufficient to alter thedevelopmental trajectory and, ultimately, the phenotypeof a cell, tissue or organ (Fig. 1) (Bartol et al. 1999;Bartol 2002; Nathanielsz 2006). A wealth of evidenceindicates that exposure to organizationally disruptiveconditions during perinatal life can have lasting effectson both the form and function of FRT tissues (Yin andMa 2005; Bartol et al. 2006). Factors required toestablish a normal developmental program and toinsure an optimal developmental trajectory for FRTtissues remain incompletely defined.The term epigenetics, coined by Waddington in the1940s, was originally defined to describe the interactionsof genes with their environment which give rise tophenotype (Waddington 1940). This definition hasevolved to describe the study of changes in geneexpression that occur without a change in DNAsequence (Jirtle and Skinner 2007). The term is nowused broadly in reference to the ‘inheritance of informationbased on gene expression levels rather than ongene sequence’ (Junien 2006). Thus, factors that affectdevelopmentally critical gene expression events, particularlythose with lasting and potentially heritableconsequences, constitute epigenetic elements of thedevelopmental program (Jirtle and Skinner 2007).‘Maternal programming’ refers to maternally drivenepigenetic events with the potential to affect both thedevelopmental program and trajectory of embryonic,foetal and ⁄ or perinatal tissues (Szyf et al. 2005; Nathanielsz2006; Wells 2007). It is likely that maternaleffects on development do not end with parturition, butextend into the early neonatal period. During this time,colostrum (first milk) serves as the conduit for communicationof organizationally important developmentalsignals from mother to offspring.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

274 FF Bartol, AA Wiley and CA BagnellCourse of developmentDevelopmentalprogram affectedDCriticalPeriodTimeDevelopmentaltrajectory affectedFig. 1. Potential effects of an organizationally disruptive event on thedevelopmental program and trajectory. Depicted is a scheme in whicha disruptive event (D) is encountered during a critical organizationalperiod early in development. The normal developmental program isdepicted by the solid line, with a range of developmental eventsassociated with normal outcomes indicated by the hatched area.Consequences of disruption may include: (1) alteration of the normaldevelopmental program without effects on the developmental trajectory(dashed line); or (2) alteration of both the developmental programand trajectory with long-term consequences for adult phenotype(dotted line)Here, data are reviewed to show that the basic elementsof a signalling system activated by a milk-borne morphoregulatorygrowth factor (MbF) exemplified byrelaxin (RLX) are present in the pig at birth. Collectively,data support the ‘lactocrine hypothesis’ for maternalprogramming of FRT tissues whereby milk-borne RLX,absorbed into the neonatal circulation during the firstdays of post-natal life, acts directly through its ownreceptor to induce and ⁄ or support oestrogen receptor-a(ER) expression and activation in a manner necessary toinsure normal FRT programming and an optimal developmentaltrajectory for FRT tissues in the pig.Reproductive Tract Development andConsequences of Developmental DisruptionClassic data indicating that pre-natal exposure ofhuman foetuses to the synthetic oestrogen diethylstilbestrolalters the developmental program of FRT tissuesand sets the stage for cervicovaginal cancer and othercomplications (Herbst et al. 1979; Iguchi and Sato 2000)drew attention to the fact that disruption of hormonesensitivedevelopmental events can have serious consequencesfor reproductive performance and health.Studies catalysed by these observations provided importantinsights into the roles of the steroid hormonesuperfamily of receptors and related ligands in normaland aberrant FRT programming. For the uterus, lossof-functionstudies showed that ER expression isrequired for normal growth, while both the progesteronereceptor (PR) and ER are required for normal uterinefunction (Conneely et al. 2001; Emmen and Korach2003). Neither ER nor PR expression is necessary tosupport primary uterine patterning events in pre- orperinatal life. However, aberrant activation of these andrelated receptor systems during critical organizationalperiods can affect the developmental program withsignificant functional consequences in both rodents andungulates (Bartol et al. 1999; Gray et al. 2000; Iguchiand Sato 2000; Huang et al. 2005). In laboratoryanimals, perinatal oestrogen exposure produced lesionsin adult uteri that included altered steroid receptorconcentrations and responsiveness; changes in oestrogenmetabolism and protein synthesis; persistent inductionor deregulation of gene expression; deregulation ofprotooncogene expression affecting epithelial proliferationand apoptosis; a myriad of structural lesions and,classically, general hypoplasia of FRT tissues andorgans (Iguchi and Sato 2000; Huang et al. 2005).Complementary data involving ungulate models indicateclearly that adult uterine phenotype can beprogrammed by targeted disruption of both steroidand peptide hormone-sensitive post-natal organizationalevents (Bartol et al. 1999, 2006; Carpenter et al. 2003;Tarleton et al. 2003).Events associated with normal and, particularly,oestrogen-induced disruption of neonatal uterine developmentin the pig have been reviewed (Bartol et al.1993; Yan et al. 2006b) and concepts related to maternaland ⁄ or environmental programming of reproductive⁄ somatic tissues in the pig are illustrated in Fig. 2.Data for the porcine uterus show that disruption ofFRT development during the early post-natal period hasboth morphological and functional consequences. Radialpatterning of the porcine uterine wall is incompleteat birth (PND 0), when tissues are ER-negative (Bartolet al. 1993; Tarleton et al. 1998). Early post-natal eventsassociated with development of these tissues includedifferentiation of glandular epithelium (GE) from luminalepithelium (LE), proliferation of nascent endometrialglands, and onset of regular ER expression instroma and GE, but not in LE (Spencer et al. 1993a;Yan et al. 2006a; b; Masters et al. 2007). When administeredfor 14 days from birth, the ER antagonist ICI182,780 retards development of the uterine wall andinhibits gland genesis in the neonatal endometrium(Tarleton et al. 1999). Thus, the ER is both a marker ofGE differentiation and a mediator of radial patterningin the neonatal porcine uterus.Aberrant activation of the neonatal ER system byadministration of oestradiol valerate (EV) to gilts fromPND 0–13, while acutely uterotrophic (Spencer et al.1993b), alters patterns of morphoregulatory geneexpression in the developing endometrium (Bartol et al.2006) and is ultimately both antiuterotrophic andantiembryotrophic (Bartol et al. 1993; Tarleton et al.2001, 2003). The hypoplastic, neonatally EV-exposedadult porcine uterus does not respond normally tosignals associated with the periattachment stage of earlypregnancy. Compared to unexposed controls on gestationalday 12 (GD 12), uterine luminal fluid proteincontent was reduced, uterine growth responses to earlypregnancy were inhibited, and endometrial gene expressionpatterns were altered in neonatally EV-exposed,pregnant adult gilts (Tarleton et al. 2003; Chen et al.2007). Treatment effects on the adult endometrialproteome for GD 12 were also pronounced (Bartolet al. 2006), and embryo survival in neonatallyEV-exposed gilts was reduced by 22% when assessedon GD 45 (Bartol et al. 1993). Thus, disruption ofoestrogen-sensitive, ER-dependent developmentalevents between birth and PND 14 has marked andlasting effects on uterine form and function in the pig. ItÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Lactocrine Programming of Uterine Development 275Fig. 2. Maternal and environmental programming of development.Natural and anthropogenic macroenvironmental factors (oval) impingeupon the maternal system throughout pregnancy and lactation(triangle). Pre-natally, maternal effects on foetal development areprofound and factors of macroenvironmental origin with the potentialto affect development are communicated via the maternal system (left).During this period the uterus serves as the conduit for communicationbetween mother and (potential) offspring. With parturition, communicationbetween mother and offspring continues via the mammarygland and milk (right). Post-natally, macroenvironmental factors mayaffect neonatal development directly and may also be conductedthrough (and modified in) the maternal system. The term lactocrinewas coined to describe a mechanism through which bioactive milkbornefactors are delivered from mother to offspring as a specificconsequence of nursingis clear that factors with the potential to affect uterineER expression and activation, either directly or indirectly,should be expected to have significant effects ondevelopmental programming of these tissues.Relaxin Receptor Expression and Functionalityin the Neonatal Porcine UterusWhile RLX has been studied for over 80 years (Hisaw1926), the cognate receptor for this 6000 Da peptidehormone was identified in 2002 (Hsu et al. 2002). A typeC, leucine-rich, G-protein coupled receptor (LGR)originally designated LGR7, the RLX receptor is nowrecognized as a member of the RLX family of peptide(RXFP) receptors and was recently re-designatedRXFP1 (Bathgate et al. 2006).LGR7 ⁄ RXFP1 is expressed by porcine uterine (Yanet al. 2006b) and cervical tissues from birth (Yan et al.2005). Uterine RXFP1 expression in the pig, which ispredominantly if not exclusively stromal, increases fromPND 0 through PND 14. Immunohistochemical evidencesupports in situ hybridization (ISH) data showingstromal RXFP1 staining similar to that reported foradult primate and human endometrium (Ivell et al.2003). Stromal expression of RXFP1 suggests that directeffects of RLX on epithelial growth and developmentare mediated through a stromally-driven paracrinemechanism (Bagnell et al. 2005). However, the fact thattrophic effects of RLX administered for 2 days fromPND 12, after onset of uterine ER expression, wereinhibited by pre-treatment with ICI 182,780 suggeststhat RLX may also act indirectly via crosstalk with theER system (Pillai et al. 1999; Yan et al. 2006a). Cervicalexpression of RXFP1 in the porcine neonate is higherthan that observed for the uterus (Yan et al. 2005).However, RLX administration for 2 days from birthdecreased cervical RXFP1 expression on PND 2 (Yanet al. 2005), suggesting elements of a negative autoregulatorymechanism governing cervical RXFP1 expression.Preliminary data (Yan and Bagnell 2003; Yanet al. 2008) also indicate that both uterine and cervicalER expression increased on PND 2 following administrationof RLX for 2 days from birth. Since RXFP1expression precedes ER expression in the neonataluterus, which is stimulated by exogenous RLX, subtlebut developmentally critical uterotrophic effects of RLXexposure from birth are likely to be RXFP1-specific.The more pronounced effects of RLX exposure observedduring the second week of neonatal life, after the onsetof ER expression, could reflect amplification of the RLXsignal through cross-talk with the ER (Pillai et al. 1999;Yan et al. 2006a). The fact that RLX can increaseoestrogen-stimulated uterine Hoxa10 expression (Guiet al. 1999) indicates that RLX signalling may also affectER-dependent events driving morphoregulatory geneexpression in the neonatal uterus (Bartol et al. 2006) andother RXFP1-positive, E-sensitive FRT tissues.Like oestrogen, uterotrophic effects of RLX in neonatalpigs are age-specific (Spencer et al. 1993b; Bagnellet al. 2005). When given for 2 days from birth, prior toonset of endometrial ER expression, RLX increaseduterine LE height, but not uterine weight on PND 2(Yan et al. 2006a). In striking contrast, RLX increasedcervical weight by PND 2 when compared with E-treated or control gilts (Yan et al. 2005). In gilts treatedfor 2 days from PND 12, after onset of uterine ERexpression, RLX increased both uterine LE height anduterine weight on PND 14 (Yan et al. 2006a). Data areconsistent with other studies showing that RLXincreases uterine LE height in pre-pubertal gilts (Ryanet al. 2001) and is important for maintenance ofreproductive epithelia in RLX-null mice (Zhao et al.2000). In RLX-deficient rats, proliferation of vaginalepithelial and stromal cells decreased while rates ofapoptosis in these cell types increased more than 10- and3-fold relative to controls (Lee et al. 2005).Image analysis based evaluation of cell proliferation,in which proliferating cell nuclear antigen (PCNA)positive cells were identified in situ and PCNA labellingindices were determined, was used to assess effects of ageat 3-day intervals from birth through PND 15 onpatterns of epithelial proliferation in the neonatalporcine endometrium (Masters et al. 2007). Overall,PCNA labelling index was greater in GE than in LE andwas affected by neonatal age in both cell types. Resultsindicated that the developmental transition from amorphogenetically inactive ‘infantile’ endometrial conditionat birth to a morphogenetically active ‘proliferative’state (Spencer et al. 1993a) has occurred by PND3 in the pig as marked by a dramatic increase inepithelial PCNA labelling index (Masters et al. 2007).This labelling index remained high through PND 6 inboth LE and GE, declined thereafter in both cellcompartments, and increased to PND 15 only in GE.Furthermore, administration of oestrogen or RLX inweek two induced LE and GE proliferation on PND 14Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Lactocrine Programming of Uterine Development 277Fig. 3. Schematic of porcine endometrial development between birth (neonatal day 0) and post-natal day 15. At birth (top left) endometrialglands are absent and the porcine endometrium is ERa negative and RXFP1 positive. Stromal RXFP1 expression suggests that effects of RLX ondeveloping epithelium may involve unidentified RLX-induced factors of stromal origin (relaximedins). Stromal–epithelial interactions (doubleheadedarrows) and the local microenvironmental conditions they affect are essential for uterine developmental success. Milk RLX concentrationsare highest within 24h of birth and decline linearly thereafter. RLX is present (200 pg ⁄ ml) in the peripheral circulation of nursing piglets andremains so through the second day of neonatal life. Other bioactive milk-borne factors (MbFs) may also be presented to these tissues. Onset ofuterine gland genesis is marked by ERa expression in nascent glandular epithelium (GE) and increased proliferative activity in GE, evident byPND 3. Data suggest a maternally lactocrine-driven system in which milk-borne RLX delivered into the neonatal circulation as a consequence ofnursing, acts through its cognate receptor system (RXFP1) to stimulate endometrial ERa expression, an event required for normal uterine glandgenesis and endometrial development in the neonatal pig. The potential influence of lactocrine-acting factors wanes as the influence of ERmediatedevents increases during the first weeks of neonatal lifemediated programming of the porcine FRT during theearly post-natal period are in place. Included is evidencethat: (1) the window of opportunity for post-natalmaternal programming of FRT tissues is open duringthe first 3 days of neonatal life as a critical morphoregulatorygene expression axis is developing; (2) neonataluterine and cervical tissues are RXFP1-positive at birth,prior to onset of uterine ER expression; (3) exogenousporcine RLX has trophic effects on uterine and cervicaltissues that are evident by PND 2 when exposure beginsat birth; (4) RLX administered from birth induces ERexpression in developing uterine and cervical tissues byPND 2; (5) a natural source of RLX is found in sow’smilk, where RLX concentrations are highest within 48 hof parturition; and (6) RLX is not detectable in thesystemic circulation of newborn pigs prior to nursing orin neonatal pigs fed hormone-free milk replacer, but isdetectable at serum concentrations approaching200 pg ⁄ ml on PND 0–1 in neonatal pigs allowed tonurse. Using uterine data as a reference, observationssuggest that RLX, delivered via a lactocrine mechanism,is available to support levels of ER expression requiredto insure that critical oestrogen-sensitive, ER-dependentelements of the post-natal organizational program areoptimized (Bartol et al. 2006). These relationships aresummarized in Fig. 3.Relaxin also has diverse actions in non-reproductivetissues (Sherwood 2004). RXFP1 transcripts have beenidentified in immature rat (Hsu et al. 2000), adult mouse(Kamat et al. 2004) and human (Sudo et al. 2003)tissues in males and females including the brain, heart,kidney, lung and liver. Thus, milk-borne RLX maysupport development of other vital tissues and organsduring neonatal life in animals of both sexes. In thislight, the lactocrine-driven, RLX-dependent neonataltissue programming mechanism envisioned here for theporcine FRT may have broader developmental andsomatotrophic implications. Given that RLX is not theonly bioactive peptide found in milk (Donovan andOdle 1994; Meisel 2005), it is important to consider thatother MbFs may be delivered to the neonatal system viaa lactocrine route. Nevertheless, the extent to whichmaternally-derived MbFs contribute to establishment ofan optimal organizational program and affect thedevelopmental trajectory of porcine FRT tissues remainsto be determined. This is a central topic ofongoing research in our laboratories.AcknowledgementsContributions of Dr Wenbo Yan, Ms Bethany Crean-Harris and MrJoseph Chen are gratefully acknowledged. This work was supported byNational Research Initiative Competitive Grants 2003-35203-13572and 2007-35203-18098 from the USDA-CSREES to FFB and CAB,NSF EPS-0447675, and the NJ and AL Agricultural ExperimentStations.ReferencesArmitage JA, Taylor PD, Poston L, 2005: Experimentalmodels of developmental programming: consequences ofexposure to an energy rich diet during development. JPhysiol 565, 3–8.Baer C, Bilkei G, 2005: Ultrasonographic and gross pathologicalfindings in the mammary glands of weaned sowshaving suffered recidiving mastitis metritis agalactia. ReprodDomest Anim 40, 544–547.Bagnell CA, Yan W, Wiley AA, Bartol FF, 2005: Effects ofrelaxin on neonatal porcine uterine growth and development.Ann N Y Acad Sci 1041, 248–255.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

278 FF Bartol, AA Wiley and CA BagnellBartol FF, 2002: The female reproductive system: targets fortoxicants. Proc Soc Theriol Ann Conf Sympos 427–436.Bartol FF, Wiley AA, Spencer TE, Vallet JL, Christenson RK,1993: Early uterine development in pigs. J Reprod Fertil48(Suppl), 99–116.Bartol FF, Wiley AA, Floyd JG, Ott TL, Bazer FW, GrayCA, Spencer TE, 1999: Uterine differentiation as afoundation for subsequent fertility. J Reprod Fertil 54(Suppl),287–302.Bartol FF, Wiley AA, Bagnell CA2006: Uterine developmentand endometrial programming. In: Asworth CJ, KraelingRR(eds), Control of Pig Reproduction VII. NottinghamUniversity Press, UK, pp. 113–130.Bathgate RA, Ivell R, Sanborn BM, Sherwood OD, SummersRJ, 2006: International union of pharmacology vii: recommendationsfor the nomenclature of receptors for relaxinfamily peptides. Pharmacol Rev 58, 7–31.Ben-Shlomo Y, Holly J, McCarthy A, Savage P, Davies D,Davey Smith G, 2005: Prenatal and postnatal milk supplementationand adult insulin-like growth factor I: Long-termfollow-up of a randomized controlled trial. Cancer EpidemiolBiomarkers Prev 14, 1336–1339.Burggren W, 1999: Genetic, environmental and maternalinfluences on embryonic cardiac rhythms. Comp BiochemPhys A 124, 423–427.Burrin DG, Davis TA, Ebner S, Schoknecht PA, Fiorotto ML,Reeds PJ, 1997: Colostrum enhances the nutritional stimulationof vital organ protein synthesis in neonatal pigs.J Nutr 127, 1284–1289.Carpenter KD, Gray CA, Noel S, Gertler A, Bazer FW,Spencer TE, 2003: Prolactin regulation of neonatal ovineuterine gland morphogenesis. Endocrinology 144, 110–120.Chen J, Frankshun A-L, Wiley A, Ho TY, Bartol FF, BagnellCA2007: Estrogen exposure from birth affects uterine geneexpression in neonatal gilts that persists in pregnant adults.Biol Reprod Special Issue, 152 .Coalson JA, Lecce JG, 1973: Influence of nursing intervals onchanges in serum proteins (immunoglobulins) in neonatalpigs. J Anim Sci 36, 381–385.Conneely OM, Mulac-Jericevic B, Lydon JP, De Mayo FJ,2001: Reproductive functions of the progesterone receptorisoforms: lessons from knock-out mice. Mol Cell Endocrinol179, 97–103.Donovan SM, Odle J, 1994: Growth factors in milk asmediators of infant development. Annu Rev Nutr 14, 147–167.Eddie LW, Sutton B, Fitzgerald S, Bell RJ, Johnston PD,Tregear GW, 1989: Relaxin in paired samples of serum andmilk from women after term and preterm delivery. Am JObstet Gynecol 161, 970–973.Emmen JM, Korach KS, 2003: Estrogen receptor knockoutmice: phenotypes in the female reproductive tract. GynecolEndocrinol 17, 169–176.Goldsmith LT, Lust G, Steinetz BG, 1994: Transmission ofrelaxin from lactating bitches to their offspring via suckling.Biol Reprod 50, 258–265.Gray CA, Taylor KM, Bazer FW, Spencer TE, 2000:Mechanisms regulating norgestomet inhibition of endometrialgland morphogenesis in the neonatal ovine uterus. MolReprod Devel 57, 67–78.Gray CA, Bartol FF, Tarleton BJ, Wiley AA, Johnson GA,Bazer FW, Spencer TE, 2001: Developmental biology ofuterine glands. Biol Reprod 65, 1311–1323.Grosvenor CE, Picciano MF, Baumrucker CR, 1993: Hormonesand growth factors in milk. Endocrinol Rev 14, 710–728.Gui Y, Zhang J, Yuan L, Lessey BA, 1999: Regulation ofhoxa-10 and its expression in normal and abnormal endometrium.Mol Hum Reprod 5, 866–873.Herbst AL, Scully RE, Robboy SJ1979: Prenatal diethylstilbestrolexposure and human genital tract abnormalities.Natl Cancer I Monogr 25–35.Hisaw FL, 1926: Experimental relaxation of the pubicligament of the guinea pig. PSEBM 23, 661–663.Hsu SY, Kudo M, Chen T, Nakabayashi K, Bhalla A, van derSpek PJ, van Duin M, Hsueh AJ, 2000: The three subfamiliesof leucine-rich repeat-containing g protein-coupledreceptors (lgr): identification of LGR6 and LGR7 and thesignaling mechanism for LGR7. Mol Endocrinol 14, 1257–1271.Hsu SY, Nakabayashi K, Nishi S, Kumagai J, Kudo M,Sherwood OD, Hsueh AJW, 2002: Activation of orphanreceptors by the hormone relaxin. Science 295, 671–674.Huang W-W, Yin Y, Bi Q, Chiang T-C, Garner N, Vuoristo J,McLachlan JA, Ma L, 2005: Developmental diethylstilbestrolexposure alters genetic pathways of uterine cytodifferentiation.Mol Endocrinol 19, 669–682.Iguchi T, Sato T, 2000: Endocrine disruption and developmentalabnormalities of female reproduction. Am Zool 40,402–411.Ivell R, Balvers M, Pohnke Y, Telgmann R, Bartsch O, Milde-Langosch K, Bamberger AM, Einspanier A, 2003: Immunoexpressionof the relaxin receptor LGR7 in breast anduterine tissues of humans and primates. Reprod BiolEndocrinol 1, 114.Jirtle RL, Skinner MK, 2007: Environmental epigenomics anddisease susceptibility. Nat Rev Genet 8, 253–262.Junien C, 2006: Impact of diets and nutrients ⁄ drugs on earlyepigenetic programming. J Inherit Metab Dis 29, 359–365.Kamat AA, Feng S, Bogatcheva NV, Truong A, Bishop CE,Agoulnik AI, 2004: Genetic targeting of relaxin and insulinlikefactor 3 receptors in mice. Endocrinology 145, 4712–4720.Kling PJ, 2002: Roles of erythropoietin in human milk. ActaPaediatr Suppl 91, 31–35.Koldovsky O, 1996: The potential physiological significance ofmilk-borne hormonally active substances for the neonate.J Mammary Gland Biol 1, 317–323.Lee H-Y, Zhao S, Fields PA, Sherwood OD, 2005: The extentto which relaxin promotes proliferation and inhibits apoptosisof cervical epithelial and stromal cells is greatest duringlate pregnancy in rats. Endocrinology 146, 511–518.Lecce JG, 1973: Effect of dietary regimen on cessation ofuptake of macromolecules by piglet intestinal epithelium(closure) and transport to the blood. J Nutr 103, 751–756.Masters RA, Crean BD, Yan W, Moss AG, Ryan PL, WileyAA, Bagnell CA, Bartol FF, 2007: Neonatal porcineendometrial development and epithelial proliferation affectedby age and exposure to estrogen and relaxin. DomestAnim Endocrinol 33, 335–346.Meisel H, 2005: Biochemical properties of peptides encryptedin bovine milk proteins. Curr Med Chem 12, 1905–1919.Nathanielsz PW, 2006: Animal models that elucidate basicprinciples of the developmental origins of adult diseases.ILAR J 47, 73–82.Nusser KD, Frawley S, 1997: Depriving neonatal rats of milkfrom early lactation has long-term consequences on mammotropedevelopment. Endocr 7, 319–323.O’Byrne EM, Steinetz BG, 1976: Radioimmunoassay (RIA) ofrelaxin in sera of various species using an antiserum toporcine relaxin. PSEBM 152, 272–276.Pacha J, 2000: Development of intestinal transport function inmammals. Physiol Rev 80, 1633–1667.Pierce AE, Risdall PC, Shaw B, 1964: Absorption of orallyadministered insulin by the newly born calf. J Physiol 171,203–215.Pillai SB, Rockwell LC, Sherwood OD, Koos RD, 1999: Relaxinstimulates uterine edema via activation of estrogen receptors:Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Lactocrine Programming of Uterine Development 279blockade of its effects using ici 182,780, a specific estrogenreceptor antagonist. Endocrinology 140, 2426–2429.Ryan PL, Baum DL, Lenhart JA, Ohleth KM, Bagnell CA,2001: Expression of uterine and cervical epithelial cadherinduring relaxin-induced growth in pigs. Reproduction 122,929–937.Sherwood OD, 2004: Relaxin’s physiological roles and otherdiverse actions. Endocrinol Rev 25, 205–234.Spencer TE, Bartol FF, Wiley AA, Coleman DA, Wolfe DF,1993a: Neonatal porcine endometrial development involvescoordinated changes in DNA synthesis, glycosaminoglycandistribution, and 3h-glucosamine labeling. Biol Reprod 48,729–740.Spencer TE, Wiley AA, Bartol FF, 1993b: Neonatal age andperiod of estrogen exposure affect porcine uterine growth,morphogenesis, and protein-synthesis. Biol Reprod 48, 741–751.Sudo S, Kumagai J, Nishi S, Layfield S, Ferraro T, BathgateRA, Hsueh AJ, 2003: H3 relaxin is a specific ligand for LGR7and activates the receptor by interacting with both theectodomain and the exoloop 2. J Biol Chem 278, 7855–7862.Szyf M, Weaver ICG, Champagne FA, Diorio J, Meaney MJ,2005: Maternal programming of steroid receptor expressionand phenotype through DNA methylation in the rat. FrontNeuroendocrinol 26, 139–162.Tarleton BJ, Wiley AA, Spencer TE, Moss AG, Bartol FF,1998: Ovary-independent estrogen receptor expression inneonatal porcine endometrium. Biol Reprod 58, 1009–1019.Tarleton BJ, Wiley AA, Bartol FF, 1999: Endometrialdevelopment and adenogenesis in the neonatal pig: effectsof estradiol valerate and the antiestrogen ici 182,780. BiolReprod 61, 253–263.Tarleton BJ, Wiley AA, Bartol FF, 2001: Neonatal estradiolexposure alters uterine morphology and endometrial transcriptionalactivity in prepubertal gilts. Domest AnimEndocrinol 21, 111–125.Tarleton BJ, Braden TD, Wiley AA, Bartol FF, 2003:Estrogen-induced disruption of neonatal porcine uterinedevelopment alters adult uterine function. Biol Reprod 68,1387–1393.Tashima LS, Mazoujian G, Bryant-Greenwood GD, 1994:Human relaxins in normal, benign and neoplastic breasttissue. J Mol Endocrinol 12, 351–364.Waddington CH1940: Organisers and Genes. CambridgeUniversity Press, Cambridge, UK.Wan Y, Saghatelian A, Chong L-W, Zhang C-L, Cravatt BF,Evans RM, 2007: Maternal PPAR gamma protects nursingneonates by suppressing the production of inflammatorymilk. Gene Dev 21, 1895–1908.Wells JCK, 2007: The thrifty phenotype as an adaptivematernal effect. Biol Rev Camb Phil Soc 82, 143–172.Yan W, Bagnell CA, 2003: Effects of relaxin on estrogenreceptor and vascular endothelial growth factor expressionin the cervix and vagina of neonatal pigs. Proc Soc Theriol23.Yan W, Wiley AA, Bartol FF, Bagnell CA, 2005: Tissuespecificeffects of relaxin on the reproductive tract ofneonatal gilts. Ann N Y Acad Sci 1041, 132–135.Yan W, Ryan PL, Bartol FF, Bagnell CA, 2006a: Uterotrophiceffects of relaxin related to age and estrogen receptoractivation in neonatal pigs. Reproduction 131, 943–950.Yan W, Wiley AA, Bathgate RAD, Frankshun A-L, Lasano S,Crean BD, Steinetz BG, Bagnell CA, Bartol FF, 2006b:Expression of LGR7 and LGR8 by neonatal porcine uterinetissues and transmission of milk-borne relaxin into theneonatal circulation by suckling. Endocrinology 147, 4303–4310.Yan W, Chen J, Wiley AA, Crean-Harris BD, Bartol FF,Bagnell CA, 2008: Relaxin (RLX) and estrogen affectestrogen receptor a, vascular endothelial growth factor,and RLX receptor expression in the neonatal porcine uterusand cervix. Reproduction 135, 705–712.Yin Y, Ma L, 2005: Development of the mammalian femalereproductive tract. J Biochem 137, 677–683.Zhao L, Samuel CS, Tregear GW, Beck F, Wintour EM, 2000:Collagen studies in late pregnant relaxin null mice. BiolReprod 63, 697–703.Author’s address (for correspondence): FF Bartol, Departments ofAnimal Sciences and Anatomy, Physiology and Pharmacology,Cellular and Molecular Biosciences Program, Auburn University,Auburn, AL 36849, USA. E-mail: bartoff@auburn.eduConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 280–287 (2008); doi: 10.1111/j.1439-0531.2008.01175.xISSN 0936-6768Endocrine Regulation of the Establishment of Spermatogenesis in PigsKC Caires, JA Schmidt, AP Oliver, J de Avila and DJ McLeanWashington State University, Pullman, Washington, USAContentsSomatic and germ cell maturation precedes the start ofspermatogenesis and is coordinated, so efficient spermatogenesiswill occur in the adults. The present study was conductedto evaluate endocrine regulation of germ and somatic cellhomeostasis in the neonatal boar testis associated with theestablishment of spermatogenesis. Testis tissue obtained from3-, 5-, 7- and 14-day-old piglets were ectopically xenograftedonto castrated, immunodeficient nude mice. Grafts wereremoved 22 weeks later and evaluated for growth and theestablishment of spermatogenesis. Recipient mouse testosteronebiosynthesis and follicle-stimulating hormone (FSH)concentrations were also assayed. Testis tissue graft growthwas significantly greater in testis grafts from 3-day donortissue when compared to all other ages; 5-, 7- and 14-day-olddonor tissue weights were not significantly different atremoval. Follicle-stimulating hormone concentrations inrecipient mice supporting testis grafts from 5-, 7- and14-day-old donor tissues did not differ and were similar tonormal physiological levels in age-matched, intact nude mice.Serum FSH levels were significantly lower in recipient micesupporting testis grafts from 3-day-old donor tissue. Radioimmunoassayand biological assay indicated no differences intestosterone production by testis tissue grafts of varyingdonor age. Porcine testis tissue obtained from 3-, 5-, 7- and14-day-old neonatal boars were all capable of producinground and elongate spermatids after 22 weeks of grafting, buttestis grafts from 14-day-old donors had a significantlygreater (eightfold) percentage of seminiferous tubules withspermatids compared to all other donor ages (p < 0.05).Cryopreservation did not affect the ability of testis tissuegrafts to grow, produce testosterone or establish spermatogenesiswhen compared to controls (p < 0.05). Collectively,these data demonstrate intrinsic differences in the biologicalactivity of germ and somatic cell populations during neonatalboar testis development associated with the establishment ofspermatogenesis.IntroductionSpermatogenesis is an organized process requiring manysystemic and local factors (Zirkin 1998; Parvinen andVentela 1999) and includes all cellular transformationsthat occur within the seminiferous epithelium leading tothe production of mature spermatozoa in the seminiferoustubule of the testis. Yet, germ cell interactionswith nearby Sertoli cells are necessary for germ cells toundergo mitosis, meiosis and spermiogenesis (Russellet al. 1990). Furthermore, it is believed that Sertoli cellscontribute to the process of spermatogonial stem cell(SSC) renewal (McLean 2005), allowing for continualproduction of sperm during a male’s adult life span.Leydig and Sertoli cells also account for the majority oftestosterone and oestrogen synthesis in the male (Waiteset al. 1985), and secrete factors exerting negative feedbackon the pituitary gland; both processes promotingnormal male fertility (Padmanabhan and Sharma 2001;Plant and Marshall 2001).The majority of information about the cell biologyassociated with spermatogenesis has been gained withthe use of rodent models and as a result, there remains ageneral lack of knowledge regarding testis developmentand sperm production in livestock. Further, due to thecomplexity involved with germ cell differentiation, mostresearch investigating testis biology is often restricted toin vitro studies. Although insightful, the intimate relationshipbetween Sertoli cells and germ cells are oftenlost, and cell lines can mutate during culture. Ectopictestis tissue xenografting is a procedure whereby smallpieces of testicular parenchyma from a donor animal aregrafted under the skin of an immunodeficient nudemouse. This technique maintains interactions betweengerm and somatic cells in the grafted tissue and thusprovides a biological assay for the establishment ofspermatogenesis in a variety of species (Honaramoozet al. 2002; Oatley et al. 2004). Additionally, multiplestudies have demonstrated that testis tissue xenograftsare representative of in vivo testis development (Schmidtet al. 2006; Zeng et al. 2006).The onset of puberty is commonly associated with thecompletion of the first wave of spermatogenesis in avariety of species (Amann 1970). It is known that Sertolicell number established before puberty determines adulttestis size (Russell 1993; Sharpe et al. 2000) and maturespermatogenic capacity (Orth et al. 1988) in variousmammalian species. Puberty in most domestic pigbreeds occurs around 20 weeks of age (Lunstra et al.1986) and researchers have investigated the establishmentof the spermatogenesis in the post-natal boar testisfrom 1 month of age to adulthood (Lunstra et al. 2003;McCoard et al. 2003; Almeida et al. 2006), but little isknown about relationships between somatic and germcells in the neonatal testis. Although porcine Sertoli cellsstill express markers of proliferating cells around4 months of age (Klobucar et al. 2003), several reportssuggest the importance of Sertoli cell proliferation in theboar testis during the first 2 weeks of life and neonatally(McCoard et al. 2001, 2003). This represents severalcandidate physiological stages in development that maygovern the spermatogenic capacity of the mature boartestis.Porcine Sertoli cell numbers increase fourfold duringthe first 2 weeks of neonatal life, and this represents animportant time for Sertoli cell homeostasis, thus wehypothesized that testicular tissue from 14-day-oldboars would have the greatest ability to establish andsupport spermatogenesis following grafting when comparedto other donor ages. We also predicted that nodifferences in testis graft androgen biosynthesis orÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Endocrine Regulation of Porcine Spermatogenesis 281recipient mouse serum follicle-stimulating hormone(FSH) would exist between the donor ages. Completionof these objectives will provide a better understanding ofthe mechanisms regulating neonatal germ and somaticcell development associated with establishing the firstwave of spermatogenesis and future spermatogeniccapacity in the boar testis.Materials and MethodsChemicals and reagents were all purchased from Sigma(http://www.sigmaaldrich.com) unless otherwise stated.Donor animals and tissue collectionThe Washington State University Animal Care and UseCommittee approved all animal procedures. Neonataltestis was obtained from 3-, 5-, 7- and 14-day-old whiteline composite boars at the Washington State UniversitySwine Facility using standard castration protocols. Aminimum of three boars were randomly chosen torepresent each donor age. To assess age-related differencesin testicular characteristics of porcine donormaterial prior to grafting, individual testis weights,Sertoli cell number and markers of Sertoli cell maturitystatus were measured (n = 6 neonatal boars per age).We also evaluated these traits in 21-day-old boars.Ectopic testis xenograftingTesticular parenchyma tissue was cut into 5 mg piecesand ectopically xenografted onto castrated, recipientimmunodeficient NCr nude mice (Taconic, http://www.taconic.com; CrTac:NCR-Fox1 ) or placedin Bouin’s fixative for 4 h at 4°C followed by dehydrationand storage in 70% ethanol for eventual morphologicalevaluation. Briefly, mice were anesthetized withketamine (0.1 mg ⁄ kg body weight, BW) and xylazine(0.5 mg ⁄ kg BW) castrated using surgical procedures,and managed as previously described (Oatley et al.2004).Analysis of donor grafts and recipient miceRecipient mice were sacrificed 22 weeks after grafting.Testis grafts were removed, weighed and processed forhistological analysis of establishment of spermatogenesisas previously described (Schmidt et al. 2006). Atsacrifice, mouse vesicular gland weights were recordedas a bioassay for androgen biosynthesis, and bloodwas also collected from recipient mice by cardiacpuncture and serum evaluated for FSH and testosteroneby radioimmunoassay (RIA). The extent of germcell differentiation in testis tissue grafts was visuallyevaluated by comparing the average number ofseminiferous tubules cross-sections containing spermatogonia,meiotic germ cells, differentiating spermatidsor only Sertoli cells within the largest centre section ofeach tissue sample. Digital images were captured usinga Leica DFC 280 camera and a Leica DMEcompound microscope (Leica Microsystems ImagingSolutions Ltd, http://www.leica-microsystems.com) at400 · magnification.ImmunohistochemistryAll antisera were obtained from Santa Cruz Biotechnology(Santa Cruz, CA, USA). Tissues were fixed inBouin’s solution and embedded in paraffin accordingto standard procedures. The tissues were sectioned(thickness, 5 lm), deparaffinized, rehydrated and microwavedon high for 15 min while submerged in0.01 M of sodium citrate (pH 6.0) to retrieve antigens.Tissues were submerged for 20 min in 3% hydrogenperoxide (in methanol) to quench endogenous peroxidaseactivity. Sections were then blocked with 10%normal serum for 30 min at room temperature, andincubated overnight at 4°C with an affinity-purified,rabbit anti-mouse GATA4 polyclonal antiserum(C-20), mouse cytokeratin 18 monoclonal antiserum(C-04), rabbit anti-human GATA1 polyclonal antiserum(H-200), each diluted 1 : 200; or rabbit antihumanandrogen receptor (AR) polyclonal antiserum(1 : 100). As a negative control, serial sections wereprocessed without any primary antibody. At roomtemperature, sections were then washed in phosphatebufferedsaline (PBS) (2 · 5 min) and incubated withbiotinylated secondary antibody diluted 1 : 300 for 1 h,rinsed in PBS (2 · 5 min) and exposed to streptavidin⁄ horseradish peroxidase (HRP) for 30 min. Immunoreactivitywas detected following a 5-min incubationin diaminobenzidine (DAB) and sections were counterstainedlightly with haematoxylin. Slides were storedat room temperature until morphological analysis.Photomicrograph cross-sections representing immunolocalizationof GATA-4, cytokeratin 18, GATA-1 andAR can be found in Fig. 3.CryopreservationOn the day of collection, four pieces of testis tissue from5-day-old boars (n = 3) were ectopically grafted ontocastrated NCr nude mice as a pre-cryopreservationcontrol. The remainder of the samples that was notgrafted immediately was frozen in a dimethyl sulfoxide(DMSO)-based cryopreservation medium using a programmableembryo freezer. Briefly, freezing media wasprepared by mixing 0.68 g sucrose (0.1 M) and 0.2 gbovine serum albumin (1%) into Dulbecco’s modifiedeagle medium (DMEM) to a final volume of 17.87 ml.Freezing media was then filtered with a 0.4 mm syringefilter, and 2.13 ml of DMSO (1.5 M) was added. About4–8 pieces of testis tissue were added to 1 ml of freezingmedia in cryovials. The cryovials were placed in aCrylogic (Model CL856) freezer at 20°C. The vials werecooled at 2°C ⁄ min to )7°C and seeded with liquidnitrogen cooled tongs. The freezing procedure continuedat 0.3°C ⁄ min to )30°C and then at 0.1°C ⁄ min to )35°C.Once the tissue reached )30°C, it was transferred to an)80°C freezer for 1 h and then submersed in liquidnitrogen. Tissue was stored for 1 week in liquid nitrogen,removed and subsequently thawed. After removalfrom liquid nitrogen, the cryovials were placed at roomtemperature for 1 min and then incubated for 1 min in a37°C water bath. Samples were washed in mediacontaining decreasing concentrations of DMSO (1.0and 0.5 M) in DMEM for 3 min, each followed byÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

282 KC Caires, JA Schmidt, AP Oliver, J de Avila and DJ McLeanwashing in Hank’s balanced salt solutions (HBSS)(0.0 M DMSO). After the final wash, donor tissues(n = 3) were grafted onto nine mice (three mice perdonor) in the same manner as the controls. To determinethe immediate affects of cryopreservation on tissuemorphology, samples of control and cryopreserved pregrafttestis tissue were also processed for evaluation bylight microscopy.Statistical analysisGerm cell proportions were averaged per mouse toprevent unequal weighting due to variation in thesetraits. All statistical comparisons were performed usingthe Proc GLM function of Statistical Analysis System(SAS) software, version 9.13 (SAS Institute, Cary, NC,USA), and differences were considered significant at thep £ 0.05 level. Pair-wise comparisons were evaluatedbetween donor ages using the Duncan’s multiple rangetest for significance. Data are presented as the mean ± -SEM and differences between donor ages were consideredsignificant at the p £ 0.05 level.ResultsNeonatal donor testicular characteristicsTesticular growth and Sertoli cell numberTestis weight increased with advancing age(p < 0.001). The average increase in testis weight after3 days post-partum (dpp) in neonatal boars was about1.5-, 2-, 6- and 12-fold at ages 5, 7, 14 and 21 dpp,respectively (Fig. 1). The number of Gata4-positiveSertoli cells per seminiferous cord cross-section also(a)(b)Fig. 1. Porcine testis tissue weight and Sertoli cell number from prepubertalboars at 3, 5, 7, 14 and 21 days post-partum (dpp). (a)Testicular growth and (b) number of Gata4-positive Sertoli cell nucleiper seminiferous tubule (cross-section). Bars with different lettersindicate significant differences between means (p < 0.05)increased with advancing age (p < 0.05, Fig. 1b).Sertoli cell nuclei reacted strongly with the Gata4antibody, but a low level of expression was alsolocalized to Leydig cell nuclei in the interstitial compartment(Fig. 2b), as previously described (McCoardet al. 2003).Protein markers of Sertoli cell maturationTo determine the maturation status of Sertoli cells theexpression of cytokeratin 18 and Gata1, proteinmarkers for immaturity and differentiation wereimmunolocalized. Strong cytoplasmic staining forcytokeratin 18, a foetal Sertoli cell marker, wasvisualized in Sertoli cells within 3, 5, 7, 14 and 21dpp boar testis tissues (Fig. 2c,f). Immunolocalizationof GATA1, a lineage specific differentiation marker,revealed weak but specific cytoplasmic staining nearthe apical region of Sertoli cell nuclei of all neonatalages evaluated (Fig. 2d) as is consistent with theexpression pattern of precursor Sertoli cells in the prenataland early neonatal mouse testis (Yomogida et al.1994). Nuclear immunoreactivity for AR was visualizedin Sertoli, peritubular myoid, Leydig cells andgonocytes within 3, 5, 7, 14 and 21 dpp boar testistissues (Fig. 2e).Porcine testis tissue graft growth, androgen biosynthesisand recipient mouse serum follicle-stimulating hormoneconcentrationTo assess age-related differences in growth potential ofporcine donor tissue at removal, testis graft weightswere measured and the numbers of seminiferoustubule cross-sections were determined. Analysis ofgrowth potential indicated donor-age-related variationin testis tissue graft weight and tubule number. Bothparameters for graft growth were significantly higher(p < 0.05) in testis grafts from 3-day donor tissue,when compared to all other ages (Fig. 3). In contrast,donor testis tissue from 5-, 7- and 14-day-old neonatalboars had similar growth rates as no differences weredetected in graft weight or seminiferous tubule number.To assess the ability for testis tissue grafts tosynthesize biologically active androgen, recipient mouseserum testosterone levels and vesicular gland weightswere compared between donor age groups. Despitevariation in donor age, RIA for serum testosterone andbiological assay of vesicular gland weights indicated nodifferences in the ability of testis tissue grafts to producetestosterone (Table 1). We also investigated the abilityof porcine testis tissue xenografts to produce factors thatexert negative feedback on pituitary FSH synthesiswhen compared to normal, intact mice of similar ageand strain by RIA. Follicle-stimulating hormone concentrationsin recipient mice supporting testis graftsfrom 5-, 7- and 14-day-old donor tissue did not differand were similar to normal physiological levels in agematched,intact nude mice (Table 1). Yet, serum FSHlevels were significantly lower (p < 0.05) than normal inrecipient mice supporting testis grafts from 3-day-olddonor tissue.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Endocrine Regulation of Porcine Spermatogenesis 283(a)(b)(c)(d)(e)(f)Fig. 2. Immunolocalization ofSertoli cell protein markers inpre-pubertal porcine testis tissue:(a) negative control (no primaryIgG), (b) GATA4, (c, f) cytokeratin18, (d) GATA1 and (e) androgenreceptor. Scale bars = 50 lmDegree of germ cell differentiation in porcine testis tissuegraftsThe extent of spermatogenesis in testis tissue fromneonatal boars of different ages was determined after22 weeks of grafting. Porcine testis tissue obtained from3-, 5-, 7- and 14-day-old neonatal boars were all capableof producing round and elongate spermatids aftergrafting (Fig. 4). Testis grafts from 14-day-old donorscontained a significantly greater (p < 0.05) percentageof seminiferous tubules with spermatids compared to allother donor ages (Fig. 4). In contrast, spermatogenesiswas established at similar rates in donor testis tissuefrom 3-, 5- and 7-day-old neonatal boars (Fig. 4). Theaverage percentage of tubules containing spermatids intestis tissue from 14-day-old donors was about eightfoldhigher (45.4%) when compared to other donor agesaveraging 5.4%. Photomicrograph cross-sections representingthe degree of germ cell differentiation withinporcine testis grafts can be found in Fig. 5.Effect of cryopreservation on porcine testis tissuedevelopment post-graftingWe evaluated the affect of pre-graft cryopreservationon the ability of porcine testis tissue grafts to grow,produce testosterone and support complete germ celldifferentiation by utilizing 5-day-old donor testis tissue(n = 3). Morphological evaluation prior to graftingrevealed that cryopreservation adversely affected seminiferouscord organization when compared to controls(Fig. 6a,b). Following the grafting period, no differenceswere observed between control and cryopreserved testistissue weight, seminiferous tubule number and diameteror testosterone production (data not shown). Establishmentof spermatogenesis in cryopreserved and controlporcine testis tissue grafts was similar as no differenceswere observed in the distribution of seminiferous tubulecross-sections containing spermatogonia, meiotic germcells and spermatids, respectively (Fig. 6c). Photomicrographcross-sections representing complete germ celldifferentiation in control and cryopreserved porcinetestis tissue grafts can be found in Fig. 6d,e.DiscussionIn the present study, porcine testis tissue grafts from alldonor ages initiated the first round of spermatogenesisassociated with pubertal age in a similar developmentaltimeline to the boar testis in situ, and were capable ofcomplete germ cell differentiation after 22 weeks ofgrafting. Yet, we observed a significant increase (eightfold)in the extent of germ cell differentiation in testisÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

284 KC Caires, JA Schmidt, AP Oliver, J de Avila and DJ McLean(a)(b)Fig. 4. Percentage of seminiferous tubules containing round andelongating spermatids within porcine testis tissue grafts of varyingdonor age. Different letters show significant differences (p < 0.05)between meansFig. 3. Weight of porcine testis tissue grafts obtained from prepubertalboars of 3, 5, 7 and 14 days of age following the ectopicgrafting period. (a) Testis tissue graft weight and (b) number ofseminiferous tubule cross-sections per testis graft. Different lettersindicate significant differences (p < 0.05) between meanstissue grafts from 14-day-old donors, as demonstratedby about 40% more seminiferous tubules containingspermatids, when compared to other donor ages. Thus,intrinsic differences exist in neonatal boar testis developmentassociated with the supporting the first wave ofspermatogenesis in porcine testis tissue xenografts.Differences also existed in the ability for testis graftsto grow and exert negative feedback on pituitary FSHproduction. Testis tissue grafts from 3-day donor tissuewere on average twofold larger when compared to testistissue obtained from other donor ages, while serum FSHlevels were significantly lower than controls (p < 0.05)in recipient mice supporting testis grafts from 3-day-olddonor tissue. In contrast, testis grafts from 5-, 7- and14-day-old donor tissue were of similar size andrecipient mice supporting these testis grafts containedserum concentrations of FSH in similar to normalphysiological levels in age-matched, intact nude mice.This finding is interesting, as no differences wereobserved in testosterone production by RIA or biologicalassay, and suggests increased potential for testisgrafts from 3-day donors to negatively regulate pituitaryproduction of FSH, likely due to increased inhibinproduction from a larger population of Sertoli cells.This effect may also be due to germinal aplasia (Zarateet al. 1974) or failure of germ cells to develop, butthis appears unlikely based on the data demonstratingthe establishment of spermatogenesis was similar intestis tissue grafts from 3-, 5- and 7-day neonataldonors.The timeline for Sertoli cell proliferation and differentiationevents are highly species-dependent, occurbefore puberty and are essential for spermatogenesis inthe adult. It is well established that during pre-natal andpost-natal testis development, FSH stimulates Sertolicells to proliferate and in response to thyroid hormones,retinoic acid and testosterone, Sertoli cells cease proliferationand initiate terminal differentiation (Orth 1984;Buzzard et al. 2003; Holsberger and Cooke 2005).Although androgens may limit Sertoli cell expansionin rodents (Buzzard et al. 2003), elevated testosteronelevels during pre-pubertal development in the rhesusmonkey testis increases Sertoli proliferation and thepopulation of germ cells independent of FSH (Arslanet al. 1993; Ramaswamy et al. 2000).The first 3 weeks of neonatal life in boars correspondswith significant increases in testis weight and Sertoli cellnumber (Fig. 1), as previously reported (McCoard et al.2003). Our morphologic evaluation of testicular characteristicsindicates that Sertoli cells in the post-natalboar testis maintain a primitive phenotype until at least21 days of age, as defined by the expression patterns oftwo well-characterized protein markers of Sertoli cellmaturation: cytokeratin 18 (Stosiek et al. 1990) andGATA1 (Yomogida et al. 1994; Bartu˚ nek et al. 2003).During this same timeline, the germ cell populationconsists of mitotically dormant gonocytes, locatedpericentrally in the seminiferous tubule, that do notmigrate to the basement membrane of until after14 days of age.Non-graftedGrafted: listed by donor age (dpp)Intact Castrate 3 5 7 14VG wt. (mg) 220.4 ± 14.8* 14.7 ± 2.5* 139.0 ± 21.3 84.0 ± 29.3 63.5 ± 36.4 92.5 ± 22.3Test. (ng ⁄ ml) 0.88 ± 0.4 0.09 ± 0.4* 1.26 ± 0.2 0.94 ± 0.2 0.54 ± 0.2 0.65 ± 0.3FSH (ng ⁄ ml) 36.7 ± 4.5 NA 22.4 ± 2.8* 32.4 ± 4.3 36.7 ± 5.2 29.5 ± 3.8Table 1. Vesicular gland weights,serum concentrations of testosterone(test.) and follicle stimulatinghormone (FSH) of intact (positivecontrol), castrated (negativecontrol) and castrated micesupporting testis tissue grafts*Significant differences (p < 0.05) between means.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Endocrine Regulation of Porcine Spermatogenesis 285(a)(b)(c)(d)Fig. 5. Morphological analysis ofgerm cell populations in porcinetestis tissue grafts. Tubule crosssectionsrepresenting mostadvanced germ cell type: (a) spermatocytesa, (b) round spermatidsb, (c, d) elongating spermatids c.Scale bars = 50 lm(a) (b) (c)Fig. 6. Morphological analysis of(a) non-frozen control and (b)cryopreserved porcine testis tissueprior to grafting. (c) Morphometricevaluation of germ cell differentiationin control and frozen porcinetestis tissue after grafting(20 weeks) represented by percentageof seminiferous tubule crosssectionscontaining spermatogonia[Gonia], spermatocytes [Cytes]and spermatids [Tids] as mostadvanced cell type. There were nosignificant differences betweenmeans (p > 0.05). Micrographs of(d) control and (e). Cryopreservedporcine testis tissue following thegrafting period Bar = 50 lm(d)(e)In vivo studies have suggested the potential tomanipulate establishment of Sertoli cell populations innon-human primates (Sharpe et al. 2000), rodents (Orthet al. 1988), bulls (Majdic et al. 1998) and boars(Lunstra et al. 2003; Nonneman et al. 2005). In anattempt to increase Sertoli cell number and pubertaltestis size similar to rodents, researchers used theapproach of inducing post-natal hypothyroidism in 21-day-old, pre-pubertal boars (Klobucar et al. 2003). Yet,it was concluded that post-natal hypothyroidism doesnot affect testis development, Sertoli cell number orsperm production in boars, despite observations ofporcine Sertoli cells expressing markers of proliferatingcells until 4 months of age (Klobucar et al. 2003).Interestingly, Sertoli cells and gonocytes were immunopositivefor nuclear AR protein in the testes of allÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

286 KC Caires, JA Schmidt, AP Oliver, J de Avila and DJ McLeanneonatal boar ages evaluated (Fig. 2e). This suggeststhat testosterone may directly or indirectly regulate thehomeostasis of these cell types in the neonatal boartestis, potentially contributing to the observed donor agevariation in the establishment of spermatogenesis in thegrafting study, and it is known that post-natal testisdevelopment is conserved between grafted and in vivotestis tissue (Oatley et al. 2004; Zeng et al. 2006).Our results demonstrate age-related differences in thepotential of neonatal testis tissue to grow and supportcomplete germ cell differentiation when grafted onimmunodeficient mice. More importantly, these findingsindicate important physiological changes in germ andSertoli cell homeostasis during neonatal testis developmentin the boar, and may explain why previousattempts to increase testis size and sperm productionin boars have been unsuccessful. This conclusion issupported by in vivo evidence (At-Taras et al. 2006,2008), and as a result we suggest future efforts to impactlifetime fertility in boars should occur during the first2 weeks of neonatal life.From an application standpoint, tissue originatingfrom 14-day boar testis resulted in the greatest degree ofgraft germ cell differentiation; this stage in developmentmay be ideally suited for germ line genetic manipulationof cells, potentially facilitating a novel means to producetransgenic boar spermatozoa. Further, as Sertoli cellsprovide an immunoprotective barrier for developinggerm cells, neonatal porcine Sertoli cells have seenincreasing use to encapsulate porcine islet cells prior totransplantation in human patients with diabetes mellitus(Dufour et al. 2003; Valdes-Gonzalez et al. 2007), andthus a better understanding of the basic mechanismsregulating their homeostasis, and survival after transplantationis necessitated. We also demonstrated nodifferences in the ability of cryopreserved and freshlygrafted donor porcine testis tissue to grow, producetestosterone and establish spermatogenesis. Thus, assomatic cell function is restored after freezing, cryopreservationof pre-pubertal testis tissue prior to graftingpresents a novel means for male germ linepreservation. The factors regulating the biologicalactivity of Sertoli and germ cell homeostasis in theneonatal boar testis will be the focus of future experimentsand may provide insights for both agriculturaland biomedical applications.AcknowledgementThis research was supported in part by a graduate research fellowshipprovided by the Seattle Chapter of the Achievement Rewards forCollege Scientists Foundation on behalf of K.C.C.ReferencesAlmeida FF, Leal MC, Franca LR, 2006: Testis morphometry,duration of spermatogenesis, and spermatogenic efficiencyin the wild boar (Sus scrofa). Biol Reprod 75, 792–799.Amann RP, 1970: Sperm production rates. In: Johnson AD,Gomes WR, VanDemark NL (eds), The Testis, Vol. 1.Academic Press, New York, pp. 433–482.Arslan M, Weinbauer GF, Schlatt S, Shahab M, Nieschlag E,1993: FSH and testosterone, alone or in combination,initiate testicular growth and increase the number ofspermatogonia and Sertoli cells in a juvenile non-humanprimate (Macaca mulatta). J Endocrinol 136, 235–243.At-Taras EE, Berger T, McCarthy MJ, Conley AJ, Nitta-OdaBJ, Roser JF, 2006: Reducing estrogen synthesis in developingboars increases testis size and total sperm production.J Androl 27, 552–559.At-Taras EE, Kim IC, Berger T, Conley A, Roser JF, 2008:Reducing endogenous estrogen during development altershormone production by porcine Leydig cells and seminiferoustubules. Domest Anim Endocrinol 34, 100–108.Bartu˚ nek P, Králova´ J, Blendinger G, Dvora´k M, Zenke M,2003: GATA-1 and c-myb crosstalk during red blood celldifferentiation through GATA-1 binding sites in the c-mybpromoter. Oncogene 22, 1927–1935.Buzzard JJ, Wreford NG, Morrison JR, 2003: Thyroidhormone, retinoic acid, and testosterone suppress proliferationand induce markers of differentiation in cultured ratsertoli cells. Endocrinology 144, 3722–3731.Dufour JM, Rajotte RV, Korbutt GS, Emerich DF, 2003:Harnessing the immunomodulatory properties of Sertolicells to enable xenotransplantation in type I diabetes.Immunol Invest 32, 275–297.Holsberger DR, Cooke PS, 2005: Understanding the role ofthyroid hormone in Sertoli cell development: a mechanistichypothesis. Cell Tissue Res 322, 133–140.Honaramooz A, Snedaker A, Boiani M, Scholer H, DobrinskiI, Schlatt S, 2002: Sperm from neonatal mammalian testesgrafted into mice. Nature 418, 778–781.Klobucar I, Kosec M, Cebulj-Kadunc N, Majdic G, 2003:Postnatal hypothyroidism does not affect prepubertaltestis development in boars. Reprod Domest Anim 38,193–198.Lunstra DD, Ford JJ, Christenson RK, Allrich RD, 1986:Changes in Leydig cell ultrastructure and function duringpubertal development in the boar. Biol Reprod 34, 145–158.Lunstra DD, Wise TH, Ford JJ, 2003: Sertoli cells in the boartestis: changes during development and compensatoryhypertrophy after hemicastration at different ages. BiolReprod 68, 140–150.Majdic G, Snoj T, Horvat A, Mrkun J, Kosec M, Cestnik V,1998: Higher thyroid hormone levels in neonatal life result inreduced testis volume in postpubertal bulls. Int J Androl 21,352–357.McCoard SA, Lunstra DD, Wise TH, Ford JJ, 2001: Specificstaining of Sertoli cell nuclei and evaluation of Sertoli cellnumber and proliferative activity in Meishan and WhiteComposite boars during the neonatal period. Biol Reprod64, 689–695.McCoard SA, Wise TH, Ford JJ, 2003: Endocrine andmolecular influences on testicular development in Meishanand White Composite boars. J Endocrinol 178, 405–416.McLean DJ, 2005: Spermatogonial stem cell transplantationand testicular function. Cell Tissue Res 322, 21–31.Nonneman D, Rohrer GA, Wise TH, Lunstra DD, Ford JJ,2005: A variant of porcine thyroxine-binding globulin hasreduced affinity for thyroxine and is associated with testissize. Biol Reprod 72, 214–220.Oatley JM, de Avila DM, Reeves JJ, McLean DJ, 2004:Establishment of spermatogenesis and the seminiferousepithelium in ectopically xenografted bovine testiculartissue. Biol Reprod 71, 494–501.Orth JM, 1984: The role of follicle-stimulating hormone incontrolling Sertoli cell proliferation in testes of fetal rats.Endocrinology 115, 1248–1255.Orth JM, Gunsalus GL, Lamperti AA, 1988: Evidence fromSertoli cell-depleted rats indicates that spermatid numberin adults depends on numbers of Sertoli cells producedduring perinatal development. Endocrinology 122, 787–794.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Endocrine Regulation of Porcine Spermatogenesis 287Padmanabhan V, Sharma TP, 2001: Neuroendocrine vs.paracrine control of follicle-stimulating hormone. ArchMed Res 32, 533–543.Parvinen M, Ventela S, 1999: Local regulation of spermatogenesis:a living cell approach. Hum Fertil (Camb) 2, 138–142.Plant TM, Marshall GR, 2001: The functional significance ofFSH in spermatogenesis and the control of its secretion inmale primates. Endocr Rev 22, 764–786.Ramaswamy S, Plant TM, Marshall GR, 2000: Pulsatilestimulation with recombinant single chain human luteinizinghormone elicits precocious sertoli cell proliferation in thejuvenile male rhesus monkey (Macaca mulatta). Biol Reprod63, 82–88.Russell LD, 1993: Form, dimensions and cytology of mammalianSertoli cells. In: Russell LD, Griswold MD (eds),The Sertoli Cell. Cache River Press, Clearwater, FL, pp.1–37.Russell LD, Ettlin RA, Sinha Hikim AP, Clegg ED, 1990:Mammalian spermatogenesis. in: Russell LD, Ettlin RA,Sinha Hikim AP, Clegg ED (eds), Histological and HistopathologicalEvaluation of the Testis. Cache River Press,Clearwater, FL, pp. 1–40.Schmidt JA, de Avila JM, McLean DJ, 2006: Graftingperiod and donor age affect the potential for spermatogenesisin bovine ectopic testis xenografts. Biol Reprod 75,160–166.Sharpe RM, Walker M, Millar MR, Atanassova N, Morris K,McKinnell C, Saunders PT, Fraser HM, 2000: Effect ofneonatal gonadotropin-releasing hormone antagonistadministration on Sertoli cell number and testicular developmentin the marmoset: comparison with the rat. BiolReprod 62, 1685–1693.Stosiek P, Kasper M, Karsten U, 1990: Expression of cytokeratin8 and 18 in human Sertoli cells of immature and atrophicseminiferous tubules. Differentiation 43, 66–70.Valdes-Gonzalez RA, White DJ, Dorantes LM, Teran L,Garibay-Nieto GN, Bracho-Blanchet E, Davila-Perez R,Evia-Viscarra L, Ormsby CE, Ayala-Sumuano JT, Silva-Torres ML, Ramirez-Gonzalez B, 2007: Three-yr follow-upof a type 1 diabetes mellitus patient with an islet xenotransplant.Clin Transplant 21, 352–357.Waites GM, Speight AC, Jenkins N, 1985: The functionalmaturation of the Sertoli cell and Leydig cell in themammalian testis. J Reprod Fertil 75, 317–326.Yomogida K, Ohtani H, Harigae H, Ito E, Nishimune Y,Engel JD, Yamamoto M, 1994: Developmental stage- andspermatogenic cycle-specific expression of transcriptionfactor GATA-1 in mouse Sertoli cells. Development 120,1759–1766.Zarate A, Garrido J, Canales ES, Soria J, Schally AV, 1974:Disparity in the negative gonadal feedback control for LHand FSH secretion in cases of germinal aplasia or Sertolicell-onlysyndrome. J Clin Endocrinol Metab 38, 1125–1127.Zeng W, Avelar GF, Rathi R, Franca LR, Dobrinski I, 2006:The length of the spermatogenic cycle is conserved inporcine and ovine testis xenografts. J Androl 27, 527–533.Zirkin BR, 1998: Spermatogenesis: its regulation by testosteroneand FSH. Semin Cell Dev Biol 9, 417–421.Author’s address (for correspondence): DJ McLean, Washington StateUniversity, Pullman, Washington, 99164–6353, USA. E-mail:dmclean@wsu.eduConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 288–294 (2008); doi: 10.1111/j.1439-0531.2008.01176.xISSN 0936-6768Male Germ Cell TransplantationI DobrinskiSchool of Veterinary Medicine, Center for Animal Transgenesis and Germ Cell Research, University of Pennsylvania, Kennett Square, PA, USAContentsTransplantation of male germ line stem cells from a donoranimal to the testes of an infertile recipient was first describedin 1994. Donor germ cells colonize the recipient’s testis andproduce donor-derived sperm, such that the recipient male candistribute the genetic material of the germ cell donor. Germcell transplantation represents a functional reconstitutionassay for male germ line stem cells and as such has vastlyincreased our ability to study the biology of stem cells in thetestis and define phenotypes of infertility. First developed inrodents, the technique has now been used in a number ofanimal species, including domestic mammals, chicken and fish.There are three major applications for this technology inanimals: first, to study fundamental aspects of male germ linestem cell biology and male fertility; second, to preserve thereproductive potential of genetically valuable individuals bymale germ cell transplantation within or between species;third, to produce transgenic sperm by genetic manipulation ofisolated germ line stem cells and subsequent transplantation.Transgenesis through the male germ line has tremendouspotential in species in which embryonic stem cells are notavailable and somatic cell nuclear transfer has limited success.Therefore, transplantation of male germ cells is a uniquelyvaluable approach for the study, preservation and manipulationof male fertility in animals.IntroductionMale fertility requires efficient production of spermatozoathroughout the adult life of the male. Spermatogenesisis characterized by sequential steps of cellproliferation and differentiation resulting in the productionof virtually unlimited numbers of spermatozoa(Russell et al. 1990). The foundation of this system is themale germ line stem cell, which has the unique potentialfor both self-renewal and production of differentiateddaughter cells that ultimately form spermatozoa (Huckins1971; Clermont 1972; Meistrich and van Beek1993a). The male germ line stem cell is the only cell in anadult body that divides and can contribute genes tosubsequent generations, making it an obvious target forgenetic manipulations (see below). Because stem cellsare ultimately defined by function, unequivocal identificationdepends on an assay to demonstrate thepotential to reconstitute the appropriate body system.For male germ line stem cells, this assay wasestablished in 1994, when it was demonstrated thattransplantation of germ cells from fertile donor mice tothe testes of infertile recipient mice resulted in donorderivedspermatogenesis and sperm production by therecipient animal (Brinster and Zimmermann 1994). Theuse of donor males carrying a marker gene allowed foridentification of donor-derived spermatogenesis in therecipient mouse testis and proved that the donorhaplotype is passed on to the offspring by recipientanimals (Brinster and Avarbock 1994). In 13 years sincethe initial report, the technique has found widespreaduse in rodents, and more recently, also in non-rodentanimals (see below). Some of the most importantmilestones are summarized in Table 1.Cross-Species Transplantation of Male GermCellsIn 1996, production of rat sperm in mouse testes wasachieved following cross-species (xenogeneic) spermatogonialtransplantation from rats to mice (Clouthieret al. 1996) and was subsequently successful from miceto rats (Ogawa et al. 1999b; Zhang et al. 2003). Thiswork illustrated that the cell cycle during spermatogenesisis controlled by the germ cell, not the Sertolicell (Franca et al. 1998). Recently, it was confirmedthat rat sperm produced in a host mouse testis arecapable of supporting normal development whenintroduced into rat oocytes by ICSI (Shinohara et al.2006). Hamster spermatogenesis also occurred successfullyin the mouse host (Ogawa et al. 1999a); yet, withincreasing phylogenetic distance between donor andrecipient species, meiotic differentiation could no longerbe achieved in the mouse testis. Transplantation ofgerm cells from donors ranging from rabbits and dogs,to pigs and bulls, and ultimately non-human primatesand humans, resulted in colonization of the mousetestis, but spermatogenesis became arrested at the stageof spermatogonial expansion (Dobrinski et al. 1999,2000; Nagano et al. 2001b, 2002a). It appears that theinitial steps of germ cell recognition by the Sertoli cells,localization to the basement membrane and initiationof spermatogonial proliferation are conserved betweenevolutionary divergent species. Yet, the testicularenvironment (Sertoli cells and paracrine factors) ofthe recipient mouse appears to be unable to supportspermatogenic differentiation and meiosis from donorspecies other than rodents. This incompatibility couldtheoretically be addressed by co-transplantation ofgerm cells and donor Sertoli cells to the mouse testis(Shinohara et al. 2003), and complete spermatogenesisfrom different mammalian species in a mouse host wasachieved by testis tissue transplantation (Honaramoozet al. 2002b). Although xenogeneic spermatogonialtransplantation to rodent testes did not result inspermatogenesis from donor species other thanrodents, it nonetheless provides a bioassay for stemcell potential of germ cells isolated from other species(Dobrinski et al. 1999, 2000; Izadyar et al. 2003a).Recently, an elegant study demonstrated spermatogenesisand fertility after transplantation of primordialgerm cells between rainbow trout and salmon,Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Male Germ Cell Transplantation 289Table 1. Male germ cell transplantation – a time line of progressYear of first report Major finding References1994 Spermatogenesis and transmission of donor haplotype after germ cell transplantation in mice Brinster and Avarbock (1994),Brinster and Zimmermann (1994)1995 Male germ cell transplantation in rats Jiang and Short (1995)1996 Reconstitution of spermatogenesis after transplantation of frozen spermatogonial stem cells Avarbock et al. (1996)Rat spermatogenesis in mouse testis Clouthier et al. (1996)1998 Culture of mouse spermatogonial stem cells Nagano et al. (1998)1999 Transplantation of germ cells from domestic animals into mouse testis Dobrinski et al. (1999)2000 Restoration of fertility by germ cell transplantation in mice Ogawa et al. (2000)2001 Transgenic mice produced by retroviral transduction of male germ line stem cells Nagano et al. (2001a)2002 Production of transgenic rats by lentiviral transduction of male germ line stem cells Hamra et al. (2002)Germ cell transplantation into X-irradiated monkey testes Schlatt et al. (2002)Germ cell transplantation in pigsHonaramooz et al. (2002a)Generation of a spermatogonial cell line in mice Feng et al. (2002)2003 Fertility and transmission of donor haplotype after germ cell transplantation in goats Honaramooz et al. (2003b)Transplantation of bovine spermatogonial stem cellsIzadyar et al. (2003b)2004 Generation of pluripotent stem cells from neonatal mouse testis Kanatsu-Shinohara et al. (2004a)2006 Testicular germ cells can colonize sexually undifferentiated embryonic gonad and produceOkutsu et al. (2006)functional eggs in fishKnock-out mice produced after transplantation of cultured germ cells Kanatsu-Shinohara et al. (2006)2008 Adeno-associated virus-mediated transgenesis after germ cell transplantation in mice and goats Honaramooz et al. (2008)highlighting the plasticity of the approach in fish(Okutsu et al. 2007).Male Germ Cell Transplantation in Non-RodentSpeciesWhile the majority of studies are performed in rodentmodels, germ cell transplantation is also applied to nonrodentspecies and has so far been reported in pigs,goats, cattle, monkeys and recently fish and chickens(Honaramooz et al. 2002a, 2003a; b; Schlatt et al. 2002;Izadyar et al. 2003b; Takeuchi et al. 2003; Yoshizakiet al. 2005; Lee et al. 2006; Mikkola et al. 2006; Okutsuet al. 2006; Trefil et al. 2006).The application of germ cell transplantation to nonrodentmammalian species was initially difficult due todifferences in testicular anatomy and physiology. Whiledirect injection of donor cells into rodent seminiferoustubules is possible via the efferent ducts, this is notfeasible in larger mammalian species. Instead, a combinationof ultrasound-guided cannulation of the centrallylocated rete testis with delivery of germ cells by gravityflow (Honaramooz et al. 2002a, 2003a) was shown to bea successful approach in large animals. When germ cellsfrom transgenic donor goats were transplanted into thetestes of immunocompetent, pre-pubertal recipient animals,the recipients produced sperm carrying the donorhaplotype and transmitted the donor genetic makeup tothe offspring. This provided proof-of-principle thatgerm cell transplantation results in donor-derived spermproduction and fertility also in a non-rodent species(Honaramooz et al. 2003b).In cattle, Izadyar et al. (2003b) first reported thetechnique of germ cell transfer and showed that transplantedautologous germ cells can initiate spermatogenesisin the recipient testis. Subsequently, we have shownthat heterologous transplantation of bovine germ cellscan succeed between breeds of cattle (Herrid et al. 2006).The success of germ cell transplantation requires theavailability of a stem cell niche in the recipient testis. Inrodents, the use of young mice and treating recipientswith GnRH agonists to suppress high intratesticulartestosterone levels improved donor cell colonization(Ogawa et al. 1998, 1999b; Dobrinski et al. 2001;Shinohara et al. 2001). Similarly, using pre-pubertalmales as recipients has proven a successful strategy ingerm cell transplantation in pigs, goats and cattle(Honaramooz et al. 2003a; b; Herrid et al. 2006). Theefficiency of colonization of seminiferous tubules by thetransplanted germ cells can be further improved ifthe recipient testes have little or no endogenousspermatogonia. Busulfan, a DNA-alkylating agent thatdestroys proliferating cells, is frequently used in rodentsto deplete recipient germ cells prior to germ celltransplantation. However, the sterilizing dose of busulfanis species- and strain-specific and treatment can belethal due to severe bone marrow depression (Ogawaet al. 1999b; Brinster et al. 2003). While its use in adultboars has been described in combination with methylprednisoloneto prevent thrombocytopenia (Mikkolaet al. 2006), in utero treatment of pigs during a period ofhigh proliferation of foetal gonocytes provides a morepractical approach that resulted in depletion of germcells with no observed adverse effects on the piglets oron sow health or fertility (Honaramooz et al. 2005). Asan alternative to cytotoxic treatment, irradiation of thetestes will also result in depletion of endogenous germcells (Creemers et al. 2002; Schlatt et al. 2002). Provideda suitable radiation source is available, local testicularirradiation appears to be the method of choice in specieswhere the anatomical position of the testes facilitatesshielding of the body from irradiation, thereby minimizingsystemic effects. Fractionated testicular irradiationresulted in depletion of germ cells in goats, ramsand cats (Honaramooz et al. 2005; Oatley et al. 2005;Kim et al. 2006). Interestingly, germ cell transplantationin rodents and perhaps also in older pigs and cattlerequires that donor and recipients are closely related orthat recipient animals are immunosuppressed (Izadyaret al. 2003b; Kanatsu-Shinohara et al. 2003b; Zhanget al. 2003; Mikkola et al. 2006), whereas germ celltransplantation in young pigs, goats and bulls wasÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

290 I Dobrinskisuccessful also between unrelated individuals (Honaramoozet al. 2002a, 2003a; b; Herrid et al. 2006). Thetestis is considered to be an immune-privileged site, butit is unclear why transplantation between unrelated,immunocompetent animals is possible in domesticanimal species but not in rodents. Nonetheless, thismakes the technique infinitely more applicable in nonrodentspecies.Applications of Male Germ CellTransplantationTo study fundamental aspects of male germ line stem cellbiology and male fertilityGerm cell transplantation in rodents made it possible tostudy the stem cell niche in the testis and to characterizeputative spermatogonial stem cells (Parreira et al. 1998;Nagano et al. 1999; Ventela et al. 2002; Kubota et al.2003; Nagano 2003; Hamra et al. 2004). It was alsoconfirmed that sperm arising from transplanted donorgerm cells are capable of fertilization in vivo and in vitro(Brinster and Avarbock 1994; Goossens et al. 2003,2006; Honaramooz et al. 2003b).As there are only few stem cells in the testis (Meistrichand van Beek 1993b; Tegelenbosch and de Rooij 1993),a great deal of work has been directed towards enrichmentof stem cells from the testis and their expansion inculture. Initially it was shown that selection of mousegerm cells for expression of a 6 - and b 1 -integrin in theabsence of c-kit receptor, as well as collection of cellsfrom experimentally induced cryptorchid testes resultedin a significant enrichment for spermatogonial stem cells(Shinohara et al. 1999, 2000a; b). Expression of Thy-1,CD9 or Egr3 was also utilized as markers for enrichmentof mouse germ line stem cell populations (Kubotaet al. 2003; Hamra et al. 2004; Kanatsu-Shinohara et al.2004b). In contrast, germ cell isolation and enrichmentin pigs and cattle relies on expression of different markerproteins (Luo et al. 2006; Herrid et al. 2007).Nagano et al. (1998) showed first that stem cells couldbe maintained in culture for a long period of time.Co-culture with embryonic fibroblast or bone marrowstromal cells, but not Sertoli cell lines, and addition ofseveral growth factors known to be beneficial for cultureof other stem cell types or primordial germ cells, such asglial cell line-derived neurotrophic factor (GDNF),leukemia inhibitory factor (LIF), epidermal growthfactor (EGF) and basic fibroblast growth factor(bFGF), successfully maintained mouse germ line stemcells in culture for varying periods of time (Naganoet al. 1998, 2003; Kanatsu-Shinohara et al. 2003a; Yehet al. 2007). Efficient long-term culture systems formouse and rat spermatogonial stem cells have now beendescribed (Kubota et al. 2004; Hamra et al. 2005;Kanatsu-Shinohara et al. 2005; Tenenhaus Dann et al.2006). The majority of the work to date was performedwith primary cultures of putative male germ line stemcells; yet, immortalized germ cell lines have beendescribed in rat and mouse (Feng et al. 2000; van Peltet al. 2002; Hofmann et al. 2005).Recently, the developmental plasticity of culturedmale germ cells was highlighted by reports that pluripotentstem cells could be isolated from cultures ofneonatal and adult mouse testis cells (Kanatsu-Shinoharaet al. 2004a; Guan et al. 2006; Seandel et al. 2007)and by the demonstration that adult male germ line stemcells can even give rise to fertilization competent eggswhen transplanted into an undifferentiated gonad in fish(Okutsu et al. 2006, 2007).When presented with a phenotype of male infertilitywith a defect in spermatogenesis, it is often difficult todetermine whether the defect originates in the germ cells,or in the somatic components of the testis (Ogawa et al.2000). To characterize an unknown defect, standardexperimental design now includes reciprocal transplantationof germ cells from affected donors to wild-typetestes and vice versa. Using this approach, male germcell transplantation was successfully applied to characterizethe role of the c-kit receptor and its ligand stemcell factor in regulation of germ cell proliferation (Ohtaet al. 2000), to show that the defect associated with thejuvenile spermatogonial depletion (jsd) mutation wasinherent to the germ cells (Boettger-Tong et al. 2000;Ohta et al. 2001), that germ cells do not requireoestrogen receptors (Mahato et al. 2000) or androgenreceptors (Johnston et al. 2001) for development, andthat germ cell differentiation is regulated by GDNF(Creemers et al. 2002; Yomogida et al. 2003), plzf(Buaas et al. 2004; Costoya et al. 2004) and CREMfunction (Wistuba et al. 2002). Transplantation of wildtypegerm cells into the testes of Dazl null miceestablished that the somatic compartment of the Dazlnull testes remains functional (Rilianawati et al. 2003).In the rat, germ cell transplantation experiments elucidatedthe defect underlying the as-mutation (Noguchiet al. 2002).To preserve the reproductive potential of geneticallyvaluable individuals by male germ cell transplantationMouse spermatogonial stem cells can be cryopreservedfor prolonged periods of time before transplantation,can still establish spermatogenesis in the recipient testis(Avarbock et al. 1996) and live off spring resulted fromspermatozoa produced after transplantation of frozenthawedmouse germ cells (Kanatsu-Shinohara et al.2003c; Frederickx et al. 2004). Therefore, germ celltransplantation could serve to restore male fertilityafter an insult to the testis, such as irradiation orcytotoxic treatment in cancer patients (Orwig andSchlatt 2005). Germ cell transplantation has an advantageover the cryopreservation of sperm prior totreatment, in that it could be applied to pre-pubertalmales where sperm cannot be obtained or to adultmales rendered azoospermic or teratozoospermic by thedisease. The technique is also of great interest indomestic or endangered animals for its potential topreserve genetic material from immature males that arelost before they reach puberty (Pukazhenthi et al. 2006;Dobrinski and Travis 2007). Even when cryopreservationof sperm is possible from adult individuals, thepreserved sperm will provide a finite resource. Incontrast, cryopreserved germ cells will undergo geneticrecombination after transplantation, thereby virtuallypreserving the entire genetic potential of the donormale. This will provide an invaluable advantage forÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Male Germ Cell Transplantation 291application of germ cell transplantation to the conservationof genetic diversity.To produce transgenic sperm by genetic manipulation ofthe male germ lineTransgenic mice and rats have been generated by viraltransduction of germ cells prior to transplantation(Nagano et al. 2001a, 2002b; Hamra et al. 2002; Orwiget al. 2002; Honaramooz et al. 2008). Recently, theadvances in mouse germ cell culture even allowed forgene targeting and production of knock-out mice usinggerm cell transplantation (Kanatsu-Shinohara et al.2006). This approach has tremendous potential inspecies where embryonic stem cell technology is notavailable and options to generate genetically modifiedanimals are inefficient. Current strategies to generategenetically modified large animals include pronuclearmicroinjection of DNA (Hammer et al. 1985) andnuclear transfer technology using modified donor cells(Schnieke et al. 1997; Cibelli et al. 1998; Chen et al.2002; Lai et al. 2002; Park et al. 2002), lentiviraltransduction of oocytes (Hofmann et al. 2003, 2004;McGrew et al. 2004; Whitelaw et al. 2004), as well assperm-mediated DNA transfer (Lazzereschi et al. 2000;Lavitrano et al. 2002). Yet, except for lentiviral transductionof oocytes, currently available technology isfrequently fraught with low efficiency and developmentalabnormalities in the few resulting off-spring, makingthe approach of using germ cell transplantation a veryvaluable alternative. The approach is schematicallyillustrated in Fig. 1. Introduction of a genetic modificationprior to fertilization will circumvent problemsassociated with manipulation of gametes and earlyembryos and developmental abnormalities associatedwith nuclear reprogramming. We recently providedproof-of-principle for transgenesis through the malegerm line in a non-rodent animal by demonstratingtransgene transmission after transduction of goat germcells prior to transplantation with a GFP marker genecarried by an adeno-associated viral vector (Honaramoozet al. 2008). While still experimental, transgenesisthrough the male germ line is expected to provide anefficient alternative to existing technology for the introductionof genetic modifications in domestic animals.Fig. 1. Schematic representation of transgene introduction throughmale germ cell transplantationConclusionsGerm cell transplantation was initially developed inrodents in 1994. Application to a domestic animalspecies was first reported in the pig and subsequently ingoats and cattle, chickens and fish. Male germ celltransplantation allows us to explore basic biologicalaspects of male germ line stem cells and testis function aswell as potential causes of male infertility. Practicalapplications include the introduction of genetic modificationsinto the germ line of domestic animals, and thepreservation of fertility in rare and endangered animalsand in patients undergoing potentially sterilizing treatmentsfor cancer therapy. Transplantation of germ cellswill continue to significantly enhance our understandingof testis function and our ability to control and preservemale fertility.AcknowledgementWork from the author’s laboratory presented here was supported by5R01RR017359-05 (NCRR ⁄ NIH) and 2R42-HD044780-02 (NIH ⁄ -NICHD).ReferencesAvarbock MR, Brinster CJ, Brinster RL, 1996: Reconstitutionof spermatogenesis from frozen spermatogonial stem cells.Nat Med 2, 693–696.Boettger-Tong H, Johnston DS, Russell LD, Griswold MD,Bishop CE, 2000: Seminiferous tubules from mutant jsd(juvenile spermatogonial depletion) mice are capable ofsupporting transplanted spermatogenesis. Abstract. BiolReprod 62(Suppl. 1), 108.Brinster RL, Avarbock MR, 1994: Germline transmission ofdonor haplotype following spermatogonial transplantation.Proc Natl Acad Sci U S A 91, 11303–11307.Brinster RL, Zimmermann JW, 1994: Spermatogenesis followingmale germ-cell transplantation. Proc Natl Acad SciUSA91, 11298–11302.Brinster CJ, Ryu B-Y, Avarbock MR, Karagenc L, BrinsterRL, Orwig KE, 2003: Restoration of fertility by germ celltransplantation requires effective recipient preparation. BiolReprod 69, 412–420.Buaas FW, Kirsh AL, Sharma M, McLean DJ, Morris JL,Griswold MD, de Rooij DG, Braun RE, 2004: Plzf isrequired in adult male germ cells for stem cell self-renewal.Nat Genet 36, 647–652.Chen S-H, Vaught TD, Monahan JA, Boone J, Emslie E,Jobst PM, Lamborn AE, Schnieke A, Robertson L, ColmanA, Dai Y, Polejaeva IA, Ayares DL, 2002: Efficientproduction of transgenic cloned calves using preimplantationscreening. Biol Reprod 67, 1488–1492.Cibelli JB, Stice SL, Golueke PG, Kane JJ, Jerry J, BlackwellC, Ponce de Leon FA, Robl JM, 1998: Cloned transgenicclaves produced from nonquiescent fetal fibroblasts. Science280, 1256–1258.Clermont Y, 1972: Kinetics of spermatogenesis in mammals:seminiferous epithelium cycle and spermatogonial renewal.Physiol Rev 52, 198–236.Clouthier DE, Avarbock MR, Maika SD, Hammer RE,Brinster RL, 1996: Rat spermatogenesis in mouse testis.Nature 381, 418–421.Costoya JA, Hobbs RM, Barna M, Cattoretti G, Manova K,Sukhwani M, Orwig KE, Wolgemuth DJ, Pandolfi PP, 2004:Essential role of Plzf in maintenance of spermatogonial stemcells. Nat Genet 36, 653–659.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

292 I DobrinskiCreemers LB, Meng X, den Ouden K, van Pelt AMM, IzadyarF, Santoro M, de Rooij DG, Sariola H, 2002: Transplantationof germ cells from glial cell line-derived neurotrophicfactor-overexpressing mice to host testes depleted of endogenousspermatogenesis by fractionated irradiation. BiolReprod 66, 1579–1584.Dobrinski I, Travis AJ, 2007: Germ cell transplantation for thepropagation of companion animals, non-domestic andendangered species. Reprod Fertil Dev 19, 732–739.Dobrinski I, Avarbock MR, Brinster RL, 1999: Transplantationof germ cells from rabbits and dogs into mouse testes.Biol Reprod 61, 1331–1339.Dobrinski I, Avarbock MR, Brinster RL, 2000: Germ celltransplantation from large domestic animals into mousetestes. Mol Rep Dev 57, 270–279.Dobrinski I, Ogawa T, Avarbock MR, Brinster RL, 2001:Effect of the GnRH-agonist leuprolide on colonization ofrecipient testes by donor spermatogonial stem cells aftertransplantation in mice. Tissue Cell 33, 200–207.Feng B, Inaba M, Lian Z, Cui Y, Toki J, Ito T, Jin T, Fan T,Yang G, Yu C, Kushida T, Ikehara S, 2000: Development ofmouse dendritic cells from lineage-negative c-kit

Male Germ Cell Transplantation 293Kanatsu-Shinohara M, Inoue K, Lee J, Yoshimoto M,Ogonuki N, Miki H, Baba S, Kato T, Kazuki Y, ToyokuniS, Toyoshima M, Niwa O, Oshimura M, Heike T, NakahataT, Ishino F, Ogura A, Shinohara T, 2004a: Generation ofpluripotent stem cells from neonatal mouse testis. Cell 119,1001–1012.Kanatsu-Shinohara M, Toyokuni S, Shinohara T, 2004b: CD9is a surface marker on mouse and rat male germline stemcells. Biol Reprod 70, 70–75.Kanatsu-Shinohara M, Miki H, Inoue K, Ogonuki N,Toyokuni S, Ogura A, Shinohara T, 2005: Long-termculture of mouse male germline stem cells under serum- orfeeder-free conditions. Biol Reprod 72, 985–991.Kanatsu-Shinohara M, Ikawa M, Takehashi M, Ogonuki N,Miki H, Inoue K, Kazuki Y, Lee J, Toyokuni S, OshimuraM, Ogura A, Shinohara T, 2006: Production of knockoutmice by random or targeted mutagenesis in spermatogonialstem cells. Proc Natl Acad Sci U S A 103, 8018–8023.Kim Y, Selvaraj V, Dobrinski I, Lee H, McEntee MC, TravisAJ, 2006: Recipient preparation and mixed germ cellisolation for spermatogonial stem cell transplantation indomestic cats. J Androl 27, 248–256.Kubota H, Avarbock MR, Brinster RL, 2003: Spermatogonialstem cells share some, but not all, phenotypic and functionalcharacteristics with other stem cells. Proc Natl Acad Sci U SA 100, 6487–6492.Kubota H, Avarbock MR, Brinster RL, 2004: Cultureconditions and single growth factors affect fate determinationof mouse spermatogonial stem cells. Biol Reprod 71,722–731.Lai L, Kolber-Simonds D, Park K-W, Cheong H-T, GreensteinJL, Im G-S, Samuel M, Bonk A, Rieke A, Day BN,Murphy CN, Carter DB, Hawley RJ, Prather RS, 2002:Production of a-1,3-galactosyltransferase knockout pigs bynuclear transfer cloning. Science 295, 1089–1092.Lavitrano M, Bacci ML, Forni M, Lazzereschi D, Di StefanoC, Fioretti D, Giancotti P, Marfe G, Pucci L, Renzi L,Wang H, Stoppacciaro A, Stassi G, Sargiacomo M,Sinibaldi P, Turchi V, Giovannoni R, Della Casa G, SerenE, Rossi G, 2002: Efficient production by sperm-mediatedgene transfer of human decay accelerating factor (hDAF)transgenic pigs for xenotransplantation. Proc Natl Acad SciUSA99, 14230–14235.Lazzereschi D, Forni M, Cappello F, Bacci ML, Di Stefano C,Marfe G, Giancotti P, Renzi L, Wang HJ, Rossi M, DellaCasa G, Pretagostini R, Frati G, Bruzzone P, Stassi G,Stoppacciaro A, Turchi V, Cortesini R, Sinibaldi P, Frati L,Lavitrano M, 2000: Efficiency of transgenesis using spermmediatedgene transfer: generation of hDAF transgenic pigs.Transplant Proc 32, 892–894.Lee YM, Jung JG, Kim JN, Park TS, Kim TM, Shin SS, KangDK, Lim JM, Han JY, 2006: A testis-mediated germlinechimera production based on transfer of chicken testicularcells directly into heterologous testes. Biol Reprod 75, 380–386.Luo J, Megee SO, Rathi R, Dobrinski I, 2006: Protein geneproduct 9.5 is a spermatogonia-specific marker in the pigtestis: application to enrichment and culture of porcinespermatogonia. Mol Reprod Dev 73, 1531–1540.Mahato D, Goulding EH, Korach KS, Eddy EM, 2000:Spermatogenic cells do not require estrogen receptor-a fordevelopment or function. Endocrinology 141, 1273–1276.McGrew MJ, Sherman A, Ellard FM, Lillico SG, GilhooleyHJ, Kingsman AJ, Mitrophanous KA, Sang H, 2004:Efficient production of germline transgenic chickens usinglentiviral vectors. EMBO Rep 5, 728–733.Meistrich ML, van Beek MEAB, 1993a: Spermatogonial stemcells. In: Desjardins C, Ewing LL (eds), Cell and MolecularBiology of the Testis. Oxford University Press, New York,pp. 266–295.Meistrich ML, van Beek MEAB, 1993b: Spermatogonial stemcells: assessing their survival and ability to produce differentiatedcells. Meth Toxicol 3A, 106–123.Mikkola M, Sironen A, Kopp C, Taponen J, Sukura A, VilkkiJ, Katila T, Andersson M, 2006: Transplantation of normalboar testicular cells resulted in complete focal spermatogenesisin a boar affected by the immotile short-tail spermdefect. Reprod Domest Anim 41, 124–128.Nagano M, 2003: Homing efficiency and proliferation kineticsof male germ line stem cells following transplantation inmice. Biol Reprod 69, 701–707.Nagano M, Avarbock MR, Leonida EB, Brinster CJ, BrinsterRL, 1998: Culture of mouse spermatogonial stem cells.Tissue Cell 30, 389–397.Nagano M, Avarbock MR, Brinster RL, 1999: Pattern andkinetics of mouse donor spermatogonial stem cell colonizationin recipient testes. Biol Reprod 60, 1429–1436.Nagano M, Brinster CJ, Orwig KE, Ryu B-Y, Avarbock MR,Brinster RL, 2001a: Transgenic mice produced by retroviraltransduction of male germ-line stem cells. Proc Natl AcadSci U S A 98, 13090–13095.Nagano M, McCarrey JR, Brinster RL, 2001b: Primatespermatogonial stem cells colonize mouse testes. BiolReprod 64, 1409–1416.Nagano M, Patrizio P, Brinster RL, 2002a: Long-term survivalof human spermatogonial stem cells in mouse testes. FertilSteril 78, 1225–1233.Nagano M, Watson DJ, Ryu B-Y, Wolfe JH, Brinster RL,2002b: Lentiviral vector transduction of male germ line stemcells in mice. FEBS Lett 524, 111–115.Nagano M, Ryu B-Y, Brinster CJ, Avarbock MR, BrinsterRL, 2003: Maintenance of mouse male germ line stem cellsin vitro. Biol Reprod 68, 2207–2214.Noguchi J, Toyama Y, Yuasa S, Kikuchi K, Kaneko M, 2002:Hereditary defects in both germ cells and the blood–testisbarrier system in as-mutant rats: evidence from spermatogonialtransplantation and tracer-permeability analysis. BiolReprod 67, 880–888.Oatley JM, Tibary A, de Avila DM, Wheaton JE, McLean DJ,Reeves JJ, 2005: Changes in spermatogenesis and endocrinefunction in the ram testis due to irradiation and activeimmunization against luteinizing hormone-releasing hormone.J Anim Sci 83, 604–612.Ogawa T, Dobrinski I, Avarbock MR, Brinster RL, 1998:Leuprolide, a gonadotropin-releasing hormone agonist,enhances colonization after spermatogonial transplantationinto mouse testes. Tissue Cell 30, 583–588.Ogawa T, Dobrinski I, Avarbock MR, Brinster RL, 1999a:Xenogeneic spermatogenesis following transplantation ofhamster germ cells to mouse testes. Biol Reprod 60, 515–521.Ogawa T, Dobrinski I, Brinster RL, 1999b: Recipient preparationis critical for spermatogonial transplantation in therat. Tissue Cell 31, 461–472.Ogawa T, Dobrinski I, Avarbock MR, Brinster RL, 2000:Transplantation of male germ line stem cells restores fertilityin infertile mice. Nat Med 6, 29–34.Ohta H, Yomogida K, Dohmae K, Nishimune Y, 2000:Regulation of proliferation and differentiation in spermatogonialstem cells: the role of c-kit and its ligand SCF.Development 127, 2125–2131.Ohta H, Yomogida K, Tadokoro Y, Tohda A, Dohmae K,Nishimune Y, 2001: Defect in germ cells, not supportingcells, is the cause of male infertility in the jsd mutant mouse:proliferation of spermatogonial stem cells without differentiation.Int J Androl 24, 15–23.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

294 I DobrinskiOkutsu T, Suzuki K, Takeuchi Y, Takeuchi T, Yoshizaki G,2006: Testicular germ cells can colonize sexually undifferentiatedembryonic gonad and produce functional eggs infish. Proc Natl Acad Sci U S A 103, 2725–2729.Okutsu T, Shikina S, Kanno M, Takeuchi Y, Yoshizaki G,2007: Production of trout offspring from triploid salmonparents. Science 317, 1517.Orwig KE, Schlatt S, 2005: Cryopreservation and transplantationof spermatogonia and testis tissue for preservation ofmale fertility. J Natl Cancer Inst Monogr 34, 56.Orwig KE, Avarbock MR, Brinster RL, 2002: Retrovirusmediatedmodification of male germline stem cells in rats.Biol Reprod 67, 874–879.Park K-W, Lai L, Cheong H-T, Cabot R, Sun Q-Y, Wu G,Rucker EB, Durtschi D, Bonk A, Samuel M, Rieke A, DayBN, Murphy CN, Carter DB, Prather RS, 2002: Mosaicgene expression in nuclear transfer-derived embryos and theproduction of cloned transgenic pigs from ear-derivedfibroblasts. Biol Reprod 66, 1001–1005.Parreira GC, Ogawa T, Avarbock MR, Franca LR, BrinsterRL, Russell LD, 1998: Development of germ cell transplantsin mice. Biol Reprod 59, 1360–1370.van Pelt aMM, Roepers-Gajadien HL, Gademan IS, CreemersLB, de Rooij DG, van Dissel-Emiliani FMF, 2002: Establishmentof cell lines with rat spermatogonial stem cellcharacteristics. Endocrinology 143, 1845–1850.Pukazhenthi BS, Comizzoli P, Travis AJ, Wildt DE, 2006:Applications of emerging technologies to the study andconservation of threatened and endangered species. ReprodFertil Dev 18, 77–90.Rilianawati R, Speed R, Taggart M, Cooke HJ, 2003:Spermatogenesis in testes of Dazl null mice after transplantationof wild-type germ cells. Reproduction 126, 599–604.Russell LD, Ettlin RA, SinhaHikim AP, Clegg ED, 1990:Mammalian spermatogenesis. In: Russell LD, Ettlin RA,SinhaHikim AP, Clegg ED (eds), Histological and HistopathologicalEvaluation of the Testis. Cache River Press,Clearwater, pp. 1–40.Schlatt S, Foppiani L, Rolf C, Weinbauer GF, Nieschlag E,2002: Germ cell transplantation into X-irradiated monkeytestes. Hum Reprod 17, 55–62.Schnieke AE, Kind AJ, Ritchie WA, Mycock K, Scott AR,Ritchie M, Wilmut I, Colman A, Campbell KHS, 1997:Human factor IX transgenic sheep produced by transfer ofnuclei from transfected fetal fibroblasts. Science 278, 2130–2133.Seandel M, James D, Shmelkov SV, Falciatori I, Kim J,Chavala S, Scherr DS, Zhang F, Torres R, Gale NW,Yancopoulos GD, Murphy A, Valenzuela DM, Hobbs RM,Pandolfi PP, Rafi S, 2007: Generation of functional multipotentadult stem cells from GPR125+ germline progenitors.Nature 449, 346–352.Shinohara T, Avarbock MR, Brinster RL, 1999: b 1 - and a 6 -integrin are surface markers on mouse spermatogonial stemcells. Proc Natl Acad Sci U S A 96, 5504–5509.Shinohara T, Avarbock MR, Brinster RL, 2000a: Functionalanalysis of spermatogonial stem cells in Steel and cryptorchidinfertile mouse models. Dev Biol 220, 401–411.Shinohara T, Orwig KE, Avarbock MR, Brinster RL, 2000b:Spermatogonial stem cell enrichment by multiparameterselection of mouse testis cells. Proc Natl Acad Sci U S A 97,8346–8351.Shinohara T, Orwig KE, Avarbock MR, Brinster RL, 2001:Remodeling of the postnatal mouse testis is accompanied bydramatic changes in stem cell number and niche accessibility.Proc Natl Acad Sci U S A 98, 6186–6191.Shinohara T, Orwig KE, Avarbock MR, Brinster RL, 2003:Restoration of spermatogenesis in infertile mice by Sertolicell transplantation. Biol Reprod 68, 1064–1071.Shinohara T, Kato M, Takehashi M, Lee J, Chuma S,Nakatsuji N, Kanatsu-Shinohara M, Hirabayashi M, 2006:Rats produced by intraspecies spermatogonial transplantationin mice and in vitro microinsemination. Proc Natl AcadSci U S A 103, 13624–13628.Takeuchi Y, Yoshizaki G, Takeuchi T, 2003: Generation oflive fry from intraperitoneally transplanted primordial germcells in rainbow trout. Biol Reprod 69, 1142–1149.Tegelenbosch RAJ, de Rooij DG, 1993: A quantitative studyof spermatogonial multiplication and stem cell renewal inthe C3H ⁄ 101 F 1 hybrid mouse. Mut Res 290, 193–200.Tenenhaus Dann C, Alvadro AL, Hammer RE, Garbers DL,2006: Heritable and stable gene knockdown in rats. ProcNatl Acad Sci U S A 103, 11246–11251.Trefil P, Micakova A, Mucksova J, Hejnar J, Poplstein M,Bakst MR, Kalina J, Brillard J-P, 2006: Restoration ofspermatogenesis and male fertility by transplantation ofdispersed testicular cells in the chicken. Biol Reprod 75,575–581.Ventela S, Ohta H, Parvinen M, Nishimune Y, 2002: Developmentof stages of the cycle in mouse seminiferousepithelium after transplantation of green fluorescent protein-labeledspermatogonial stem cells. Biol Reprod 66,1422–1429.Whitelaw CBA, Radcliffe PA, Ritchie WA, Carlisle A, EllardFM, Pena RN, Rowe J, Clark AJ, King TJ, MitrophanousKA, 2004: Efficient generation of transgenic pigs usingequine infectious anaemia virus (EIAV) derived vector.FEBS Lett 571, 233–236.Wistuba J, Schlatt S, Cantauw C, von Schonfeldt V, NieschlagE, Behr R, 2002: Transplantation of wild-type spermatogonialeads to complete spermatogenesis in testes of cyclic3’,5’-adenosine monophosphate response element modulator-deficientmice. Biol Reprod 67, 1052–1057.Yeh JR, Zhang X, Nagano M, 2007: Establishment of a shorttermin vitro assay for mouse spermatogonial stem cells. BiolReprod 77, 904.Yomogida K, Yagura Y, Tadokoro Y, Nishimune Y, 2003:Dramatic expansion of germinal stem cells by ectopicallyexpressed human glial cell line-derived neurotrophic factorin mouse Sertoli cells. Biol Reprod 69, 1303–1307.Yoshizaki G, Tago Y, Takeuchi Y, Sawatari E, Kobayashi T,Takeuchi T, 2005: Green fluorescent protein labeling ofprimordial germ cells using a nontransgenic method and itsapplication for germ cell transplantation in salmonidae. BiolReprod 73, 88–93.Zhang Z, Renfree MB, Short RV, 2003: Successful intra- andinterspecific male germ cell transplantation in the rat. BiolReprod 68, 961–967.Author’s address (for correspondence): I Dobrinski, School ofVeterinary Medicine, Center for Animal Transgenesis and Germ CellResearch, University of Pennsylvania, Kennett Square, PA 19348,USA. E-mail: dobrinsk@vet.upenn.eduConflict of interest: The author received grants from NIH whosupported the work described in this paper.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 295–301 (2008); doi: 10.1111/j.1439-0531.2008.01177.xISSN 0936-6768Sexual Maturation in the BullN Rawlings 1 , ACO Evans 2 , RK Chandolia 3 and ET Bagu 41 Department of Veterinary Biomedical Sciences, University of Saskatchewan, Saskatoon, Saskatchewan, Canada; 2 School of Agriculture FoodSciences and Veterinary Medicine, University College Dublin, Dublin, Ireland; 3 Institute fu¨r Reproducktions Medizin, Universitat Mu¨nster, Mu¨nster,Germany; 4 Centre de recherche´ en biologie de la Reproduction, Laurier, Quebec, CanadaContentsIn this review, we describe the process of sexual maturationin the bull calf. The testes of the bull grow relatively slowlyuntil approximately 25 weeks of age and then a rapid phaseof growth occurs until puberty, at 37–50 weeks of age.During the early postnatal phase of slower growth of thetestis pre-spermatogonia and some spermatogonia are established,adult Leydig cells appear and undifferentiated Sertolicells are produced. The rapid testicular growth, after25 weeks of age, consists of marked increases in thediameter and length of the seminiferous tubules, dramaticproliferation and differentiation of germ cells, with maturespermatozoa occurring between 32 and 40 weeks of age.The adult Leydig cell population is largely in place by30 weeks of age and that of Sertoli cells by 30–40 weeksof age. Serum concentrations of LH increase from 4 to5 weeks of age, to an early postnatal peak at 12–16 weeksof age, followed by a decline to 25 weeks of age.Serum FSH concentrations are high postnatally, decliningto approximately 25 weeks of age. Serum testosteroneconcentrations increase during the phase of rapid testiculargrowth. Hypothalamic opioidergic inhibition may abatetransiently to allow the early postnatal increase in LHsecretion, while testicular androgenic negative feedbackprobably contributes to the decline in gonadotropinsecretion to 25 weeks of age. Several lines of study haveled us to suggest that early postnatal gonadotropin secretionis pivotal in initiating the process of sexual maturation inthe bull calf.IntroductionA fairly widely accepted end point for pubertal developmentin the bull was defined by Wolf et al. (1965) asan ejaculate that contained at least 50 million spermwith no less than 10% progressive motility. In bullsfrom European breeds puberty is reached over a rangeof approximately 37–50 weeks of age, with dairy breedsmaturing somewhat earlier than beef breeds (Lunstraet al. 1978; Amann 1983; Evans et al. 1995). Althoughscrotal circumference at puberty, as defined by Wolfet al. (1965), varies somewhat between breeds, a scrotalcircumference of 28 cm is often used as the point of, orage at puberty (Wolf et al. 1965; Lunstra et al. 1978).However, sperm output and quality increase for sometime after the age of puberty (Almquist and Cunningham1967; Lunstra and Echternkamp 1982). Thispresent review will focus on development of the reproductivesystem in the bull, with particular emphasis onregulatory mechanisms. As much as possible we havetried to describe trends and mechanisms for bulls ingeneral, regardless of breed, season of birth andnutrition.Testicular Growth and Development of theReproductive TractTesticular growth follows a sigmoid pattern in the bullcalf, with more rapid growth after approximately25 weeks of age through puberty, slowing as the bullachieves adult sperm output (Fig. 1; Abdel-Raouf 1960;Macmillan and Hafs 1968; Amann 1983). Seminiferoustubule diameter increases gradually until 20–25 weeks ofage and then more rapidly (Abdel-Raouf 1960; Macmillanand Hafs 1969; Amann 1983; Evans et al. 1996).Increased diameter and length of the tubules accountsfor most of the increase in testis size up to 32 weeks ofage, but tubule length predominates during later stagesof development (Curtis and Amann 1981). Luminationof the tubules occurs at approximately 25 weeks of age(Evans et al. 1996). Fetal Leydig cells degenerate byapproximately 8 weeks after birth and adult Leydig cellnumbers increase rapidly from 4 to 30 weeks of age(Wrobel 1990). Undifferentiated Sertoli cells proliferatefrom 4 weeks of age up until 13–20 weeks of age andthen the number decreases as these cells differentiateinto mature, adult Sertoli cells; differentiation of matureSertoli cells appears to be complete somewhere between30 and 40 weeks of age (Abdel-Raouf 1960; Curtis andAmann 1981; Sinowatz and Amselgruber 1986; Wrobel2000; Bagu et al. 2006a). The number of Sertoli cells is acritical determinant of daily sperm production in thebull (Berndtson et al. 1987). At birth the solid seminiferoustubules or chords contain the primordial germcells or gonocytes; these largely disappear by 30 weeksof age (Curtis and Amann 1981; Wrobel 2000; Baguet al. 2006a). Proliferation of pre-spermatogonia andsome spermatogonia occurs from 4 to 5 weeks of ageonwards (Abdel-Raouf 1960; Curtis and Amann 1981;Wrobel 1990; Evans et al. 1996; Bagu et al. 2006a). Prespermatogonialcell numbers decrease after 24 weeksof age (Curtis and Amann 1981), but numbers ofspermatogonia increase rapidly until 44 weeks of age(Abdel-Raouf 1960). Primary spermatocytes appear atapproximately 20 weeks of age, secondary spermatocytesat 20–30 weeks of age, round spermatids between25 and 30 weeks of age, long spermatids at 25–35 weeksof age and finally mature spermatozoa are clearly seenbetween 32 and 40 weeks of age (Abdel-Raouf 1960;Macmillan and Hafs 1968; Curtis and Amann 1981;Evans et al. 1996; Bagu et al. 2006a). It is interestingthat rapid testicular growth, differentiation and development,occur after 20–25 weeks of age.As with testicular development, epididymal growth isslow initially, but increases rapidly after 28 weeks of ageÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

296 N Rawlings, ACO Evans, RK Chandolia and ET BaguFig. 1. Growth and developmentof some segments of the reproductivesystem of the bull calf frombirth to puberty based on Abdel-Raouf (1960), Macmillan and Hafs(1969), Curtis and Amann (1981),Sinowatz and Amselgruber (1986),Evans et al. (1996), Wrobel (2000)and Bagu et al. (2006a). Horizontallines indicate approximate periodsfor the activities shown; lines witharrow heads indicate start pointsfor continued activities(Fig. 1; Abdel-Raouf 1960; Macmillan and Hafs 1969).Rapid development of the vesicular glands occurs afterapproximately 2 months of age (Lindner and Mann1960; Macmillan and Hafs 1969; Chandolia et al. 1997c)in the bull. Content of the two principle secretoryproducts of the vesicular glands, fructose and citric acid,increase markedly in bulls after 20–24 weeks of age(Abdel-Raouf 1960; Macmillan and Hafs 1969) but inone report (Lindner and Mann 1960) this increase wasseen as early as 8 weeks of age.Endocrine Changes During Sexual Maturation;Alignment with Testicular DevelopmentSerum concentrations of LH are low in the first 4–5 weeks after birth but increase thereafter to a peak atapproximately 12–16 weeks of age and then concentrationsdecrease to approximately 25 weeks of age,remaining low but variable through the time of puberty(Fig. 2; Rawlings et al. 1978; Lacroix and Pelletier1979; McCarthy et al. 1979; Amann and Walker 1983;Rodriguez and Wise 1989; Rawlings and Evans 1995;Fig. 2. The temporal patterns of serum concentrations of reproductivehormones from birth to puberty in the bull calf based on Evans et al.(1996), Chandolia et al. (1997c), Aravindakshan et al. (2000a) andWrobel (2000). Horizontal lines indicate approximate periods for theactivities shownAravindakshan et al. 2000a; Bagu et al. 2006a). Thisearly, postnatal increase in LH secretion is caused by atransient increase in the frequency of peaks or episodesin LH secretion (Rawlings and Evans 1995). The earlypostnatal increase in LH pulse frequency is clearlydriven by an increase in the frequency of pulses ofGnRH secretion (Rodriguez and Wise 1989) and isaccompanied by an increase in pituitary LH stores(Rodriguez and Wise 1991), pituitary GnRH receptors(Amann et al. 1986; Rodriguez and Wise 1991) and LHrelease in vitro (McAndrews et al. 1994). The pattern ofserum concentrations of FSH during development in thebull is less clear ranging from no real trend (McCarthyet al. 1979; Amann and Walker 1983) to elevatedconcentrations postnatally, decreasing to 25 weeks ofage; FSH secretion in the bull does not appear to bepulsatile (Miyamoto et al. 1989; Evans et al. 1996; Baguet al. 2006a).Interestingly, serum concentrations of inhibin arehigh postnatally declining to 7 or 8 months of age(Miyamoto et al. 1989; Matsuzaki et al. 2001). Serumconcentrations of testosterone increase slowly after birthand up to approximately 20 weeks of age; subsequently,concentrations increase rapidly, starting anywhere fromapproximately 20–35 weeks of age (Secchiara et al.1976; Lacroix et al. 1977; Sundby and Velle 1980;Miyamoto et al. 1989; Rawlings and Evans 1995; Evanset al. 1996). During the early postnatal period, up to20 weeks of age, the testis also secretes significantamounts of androstenedione and dihydrotestosterone,albeit at lower levels than testosterone (Rawlings et al.1972; McCarthy et al. 1979; Sundby et al. 1984; Rawlingsand Cook 1986).The slow increase in testosterone secretion andproduction of androstenedione and dihydrotestosterone,seen prior to 20 weeks of age, occur during thephase of the rapid increase in number of adult Leydigcells in the testes and the early postnatal transientÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Puberty in the Bull 297increase in LH secretion. The later, rapid increase inserum concentrations of testosterone occurs after theearly rise in LH secretion and once the production ofadult Leydig cells is virtually complete, but importantly,over a period of time when gonadotropin secretion islow. Rapid growth of the testes and the androgendependentsexual accessory structures occurs after20 weeks of age during the rapid increase in testosteronesecretion. When LH and FSH receptor concentrationsper milligram of protein, in the testes, were studied every4 weeks from 5 to 33 weeks of age and again at56 weeks of age, the major finding was a transientdecrease in concentrations from 13 to 25 weeks of age(Bagu et al. 2006a). For these bulls (Hereford · Charolais)this transient decrease in gonadotropin receptorconcentrations occurred as serum gonadotropin concentrationsdeclined from their high postnatal levels andas numbers of adult Leydig and Sertoli cells increasedquite rapidly at the onset of rapid testicular growth.Proliferation of immature Sertoli cells occurs duringthe early postnatal period in the bull calf while serumFSH concentrations are high. Maturation of immatureto mature Sertoli cells starts around the time of the earlypostnatal rise in LH secretion and continues as gonadotropinsecretion declines. It is interesting that thephase of most active production of spermatogonia andforward progression of spermatogenesis, in the developingbull, occurs at the end of the early postnatalincrease in LH secretion and as serum concentrations ofFSH decline from their postnatal zenith. Again, theperiod of rapid testicular growth and development ofspermatogenesis, occur after 20–25 weeks of age whenserum concentrations of LH and FSH are lower than inthe early postnatal period. However, pulsatile secretionof LH is still important for testis growth in this period ofrapid testicular growth, as suppression of LH secretionwith oestradiol slows testicular growth, but treatmentwith LHRH restores it (Schanbacher 1981; Schanbacheret al. 1982).The Critical Role of the Early PostnatalIncrease in Gonadotropin SecretionWhen 48 Hereford · Charolais bull calves were splitinto two groups, based on age at puberty as defined byWolf et al. (1965), serum LH concentrations were higherduring the postnatal increase in LH secretion, in earlymaturing (42 weeks age) compared to late maturingbulls (48 weeks of age; Evans et al. 1995; Aravindakshanet al. 2000b). Serum FSH concentrations did notdiffer between groups. We were surprised to see subsequently,that late maturing bull calves had higher serumconcentrations of LH in response to LHRH at 4 and20 months of age (at the onset and end of the earlypostnatal increase in LH secretion) than early maturingbull calves (Bagu et al. 2006b). This was perhapsbecause early maturing bull calves experienced an earlierand greater increase in GnRH secretion postnatally, thatcaused depletion of LH stores in the early developingpituitary, giving a lower response to exogenous LHRH,compared to late maturing bull calves. Treatment ofHereford · Charolais bull calves with a GnRH agonistdesigned to lower LH secretion, from 6 to 18 weeks ofage (Chandolia et al. 1997a), delayed the peak in theearly postnatal rise in LH secretion from 20 (controlcalves) to 25 weeks of age. Serum concentrations ofFSH were lower in treated calves at 14, 16, 18 and26 weeks of age compared to control calves. Serumtestosterone concentrations were significantly suppressedat 14, 16 and 18 weeks of age. Scrotal circumferencewas smaller in treated calves from birth to50 weeks of age. At 50 weeks of age spermatid andspermatocyte numbers, but not Sertoli cell numbers,were lower in tubule cross sections for tubules in stageVI of spermatogenesis in treated compared to controlbulls. In another study, similar bull calves were givenLHRH intravenously, every 2 h for 14 days, between 4and 6 weeks of age (Chandolia et al. 1997b). Thistreatment increased the frequency of secretory pulsesof LH during the period of treatment and resulted in agreater scrotal circumference. Spermatogenesis wasenhanced and numbers of Sertoli cells were increasedbased on histology of testicular tubular cross sections instage VI of spermatogenesis, from testes collected at54 weeks of age. Serum concentrations of FSH were notaffected by treatment. In subsequent studies, in Hereford· Charolais bull calves, treatments were given lessfrequently; GnRH was given (i.m.) twice every day from4 to 8 weeks of age (Madgwick et al. 2007) and inseparate sets of bull calves, either bovine LH or bovineFSH were given every 2 days from 4 to 8 weeks of age(Bagu et al. 2004). Based on achievement of a scrotalcircumference of 28 cm (Wolf et al. 1965; Lunstra et al.1978) GnRH treatment advanced age at puberty(42 weeks of age) compared to control bulls (47 weeksof age). The FSH treatment also advanced puberty(39 weeks of age) compared to control bulls (45 weeksof age). Schuenemann et al. (2007) gave FSH at 27 daysof age, when bull calves were immunized against inhibinand saw increased germ cell numbers at 17 weeks of age.It appears that early postnatal secretion of FSH andparticularly LH, prior to 25 weeks of age, are critical toinitiate testicular differentiation and development in thebull calf, with the process continuing after this timeagainst a background of lower serum gonadotropinconcentrations, but increasing serum concentrations oftestosterone.Regulation of Gonadotropin Secretion DuringDevelopmentThe regulation of the relatively high levels of FSHsecretions seen postnatally in the bull calf and the early,transient, postnatal rise in LH secretion are critical asthey appear to be pivotal signals for the onset of sexualmaturation (Fig. 3). When bull calves were castrated atdifferent ages there were only limited increases in FSHsecretion but LH secretion rose markedly in calves24 weeks old and older (Amann and Walker 1983;Deaver and Peters 1988). In another study, castrationdid not result in an increase in LH secretion in bullcalves at 7, 8 or 9 weeks of age, but a response was seenat 10, 11, and 13 weeks of age (Wise et al. 1987).However, Bass et al. (1978) concluded that castrationresulted in increased LH secretion at 1, 7 and 17 weeksof age. Testosterone and dihydrotestosterone suppressÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

298 N Rawlings, ACO Evans, RK Chandolia and ET BaguFig. 3. Regulation of LH and FSH secretion during sexual maturationin bull calves based on Amann and Walker (1983), Wise et al. (1987),Deaver and Peters (1988), Rodriguez and Wise (1989), Macdonaldet al. (1990), Evans et al. (1993), Shahab et al. (1993, 1995), Evanset al. (1996), Chandolia et al. (1997c) and Aravindakshan et al.(2000a). Horizontal lines and grey boxes indicate approximate periodsfor the activities shown; lines with arrow heads indicate start points forcontinued activities. NMA, N-methyl-D,L-aspartic acidserum concentrations of LH and FSH in bull calvesbetween 4 and 20 weeks of age (Kennedy et al. 1985;Godfrey et al. 1992). However, when Hereford · Charolaiscalves were treated with three injections, 12 hoursapart, of the androgen receptor blocker Flutamide, at 8,16 and 24 weeks of age, the authors concluded thatreduced suppression of LH secretion by androgens isnot critical for the onset of the early postnatal increasein LH secretion but increased suppression could be afactor in the subsequent decrease in both LH and FSHsecretion (Rawlings and Evans 1995). There has beenconsiderable interest in the role of oestradiol as afeedback regulator of gonadotropin secretion in the bullcalf. Oestradiol is produced by the bovine testes but anadrenal source is also likely (Henricks et al. 1988).Serum concentrations of oestradiol appear to increasefrom 1 to 6 weeks of age (Amann et al. 1986; Evanset al. 1993) and again after 24 weeks of age, duringrapid testicular growth (Evans et al. 1993). The concentrationsof oestradiol receptors decrease in the hypothalamus,but increase in the pituitary, from 6 to10 weeks of age (Amann et al. 1986). Treatment of bullcalves with oestradiol as early as 5 weeks of age willsuppress LH and FSH secretion (Schanbacher et al.1982; Wise et al. 1987; Deaver et al. 1988; Wildeus et al.1988; Macdonald et al. 1990; Godfrey et al. 1992).Interestingly, in the study of Wildeus et al. (1988),treatment with oestradiol suppressed mean LH secretionat 4, 6, 8 and 10 weeks of age but LH pulse frequencywas suppressed only at 4 and 8 weeks of age.It would appear that negative feedback suppression ofgonadotropin secretion, by products of the testes, mayterminate the early postnatal increase in LH secretionand contribute to decreased FSH secretion at that time(circa 20 to 25 weeks of age). This negative feedbackstrengthens as the calf gets older and probably, principally,involves androgens, but oestradiol is also involved.The role of inhibin in regulating FSH secretionduring development is unclear; although immunizationagainst inhibin in the bull calf does cause an increase inFSH secretion (Kaneko et al. 2001). Some attemptshave been made to further study the neuroendrocrineregulation of gonadotropin secretion in the postnatalbull calf between birth and 25 weeks of age. WhenNaloxone, a l opiate receptor antagonist, was administeredto bull calves, some evidence was produced thatsupported opioidergic suppression of LH secretionduring the early postnatal period (Macdonald et al.1990; Evans et al. 1993). The authors of these studiessuggested that a transient decrease in opioidergic suppressionof LH secretion, particularly LH pulse frequency,may allow the early postnatal increase in LHsecretion (Macdonald et al. 1990; Evans et al. 1993;Chandolia et al. 1997d). Some evidence for opioidergicsuppression of FSH secretion was seen in 24-week-oldbull calves (Evans et al. 1993; Chandolia et al. 1997d).In the study of Chandolia et al. (1997d), the dopamineantagonist pimozide was also used and the authorsconcluded that a dopaminergic drive for LH and FSHsecretion was established in 24-week-old bull calves. At24 weeks of age, but not earlier, opioidergic inhibitionof gonadotropin secretion largely involved inhibition ofthe dopaminergic drive. Administration of MK801, anNMDA (N-methyl-D,L-aspartic acid) receptor antagonist,did not affect LH secretion at 1 and 12 weeks ofage, but at 24 weeks of age pulsed secretion of LH wasdecreased (Shahab et al. 1995); NMA caused LH releaseat 24 weeks of age (Shahab et al. 1993).The early postnatal increase in LH secretion consistsof an increase in the frequency of pulses of LH secretion,driven by increased GnRH pulse frequency (Rodriguezand Wise 1989). Diminished opioidergic tone may allowthis increased GnRH secretion and the termination ofthe early postnatal increase in LH secretion is perhapspartly due to negative feedback by testicular androgensas well as opioidergic tone. Similar regulatory mechanismsmay cause the decrease in FSH secretion fromapproximately 20 to 30 weeks of age. Central drives forgonadotropin secretion, involving NMA and dopamine,appear to be established after the early postnatalincrease in LH secretion, but androgenic negativefeedback of gonadotropin secretion also strengthens atand after this time. The role of inhibin in the regulationof the temporal patterns of FSH secretion duringdevelopment in the bull calf is unclear.Effects of Breed, Season and NutritionIn the introduction to this review the effect of breed onage at puberty was briefly alluded to; nutrition (Brattonet al. 1959) and season of birth (Aravindakshan et al.2000a; Tatman et al. 2004) can also influence reproductivedevelopment. In general, it appears that lownutritional intake after weaning hinders sexual development,whereas, the effects of enhanced nutrition,although less clear, appear positive (Mann et al. 1967;Pruitt et al. 1986). It has been suggested that variousmetabolic hormones may provide a reflection of nutritionalintake and influence patterns of secretion ofreproductive hormones during development (I’Ansonet al. 1991). In the bull calf, serum concentrations ofinsulin, IGF-1 and leptin increase with age, but growthhormone concentrations decrease (Breier et al. 1988;Ronge and Blum 1989; McAndrews et al. 1993; Renavilleet al. 1996; Brito et al. 2007a; b). Treatment of bullcalves with recombinant bovine somatotrophin from 4to 32 weeks of age, did not affect sexual developmentÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Puberty in the Bull 299(MacDonald and Deaver 1993). When nutritionalrestriction was applied pre-weaning and through theperiod of the early postnatal increase in LH secretion,LH secretion was reduced and puberty delayed; testicularweight was reduced (Brito et al. 2007a). Enhancednutrition during the same period of time positivelyaffected the early postnatal increase in LH secretion andincreased testicular growth (Brito et al. 2007b). Duringthe treatment period, nutritional restriction caused adecrease in serum concentrations of IGF-1 and improvednutrition increased serum concentrations ofIGF-1 and insulin (Brito et al. 2007a,b). It is intriguingthat nutritional manipulation of the early increase in LHsecretion affects testicular growth and age at puberty.Fall born bull calves reached puberty later (Tatmanet al. 2004) or the timing of puberty was more variable(Aravindakshan et al. 2000a) compared to spring borncalves. Fall born bull calves experienced a prolongedearly postnatal increase in LH secretion, with greater LHpulse amplitude, compared to spring born calves (Aravindakshanet al. 2000a). Although photoperiod wouldbe the obvious basis for these seasonal differences,treatment of spring born bulls with implants releasingmelatonin at birth, 6 and 11 weeks of age, did not disruptdevelopment in any way (Aravindakshan et al. 2000a).ConclusionsThe testes of the bull grow relatively slowly untilapproximately 25 weeks of age and then a rapid phaseof growth occurs until puberty at 37–50 weeks of age.During the early postnatal phase of slower growth of thetestis, pre-spermatogonia and some spermatogonia areestablished, adult Leydig cells appear and undifferentiatedSertoli cells are produced. The rapid testiculargrowth, after 25 weeks of age, consists of markedincreases in the diameter and length of the seminiferoustubules, dramatic proliferation and differentiation ofgerm cells, with mature spermatozoa occurring between32 and 40 weeks of age. The adult Leydig cell populationis largely in place by 30 weeks of age and that ofSertoli cells by 30–40 weeks of age. Serum concentrationsof LH increase from 4 to 5 weeks of age, to anearly postnatal peak at 12–16 weeks of age, followed bya decline to 25 weeks of age. Serum FSH concentrationsare high postnatally, declining to approximately25 weeks of age. Serum testosterone concentrationsincrease during the phase of rapid testicular growth.Hypothalamic opioidergic inhibition may abate transientlyto allow the early postnatal increase in LHsecretion, while testicular androgenic negative feedbackprobably contributes to the decline in gonadotropinsecretion to 25 weeks of age. Several lines of study haveled us to suggest that early postnatal gonadotropinsecretion is pivotal to initiate the process of sexualmaturation in the bull calf.AcknowledgementsWe thank Dr R. Kennedy, Ms S. Cook, Dr J.P. Aravindakshan, Dr S.Madgwick and Mr W. Kerr for help with our studies. Financialsupport was from the Canadian Natural Sciences and EngineeringResearch Council and the Saskatchewan Agricultural Research Fund.ReferencesAbdel-Raouf M, 1960: The postnatal development of thereproductive organs in bullswith special reference topuberty. (Including growth of the hypophysis and theadrenals). Acta Endocrinol 34(Suppl. 49), 1–109.Almquist J, Cunningham D, 1967: Reproductive capacity ofbeef bulls. I. Postpuberal changes in semen production atdifferent ejaculation frequencies. J Anim Sci 26, 174–181.Amann RP, 1983: Endocrine changes associated with onset ofspermatogenesis in Holstein bulls. J Dairy Sci 66, 2606–2622.Amann R, Walker O, 1983: Changes in pituitary-gonadal axisassociated with puberty in Holstein bulls. J Anim Sci 57,422–433.Amann R, Wise M, Glass J, Nett T, 1986: Prepubertal changesin the hypothalamic–pituitary axis of Holstein bulls. BiolReprod 34, 71–80.Aravindakshan JP, Honaramooz A, Bartlewski PM, BeardAP, Pierson RR, Rawlings NC, 2000a: Gonadotrophinsecretion in prepubertal bull calves born in spring andautumn. J Reprod Fertil 120, 159–167.Aravindakshan JP, Honaramooz A, Bartlewski PM, BeardAP, Pierson RA, Rawlings NC, 2000b: Pattern of gonadotropinsecretion and ultrasonographic evaluation of developmentalchanges in the testis of early and late maturingbull calves. Theriogenology 54, 339–354.Bagu ET, Madgwick S, Duggavathi R, Bartlewski PM, BarrettDM, Huchkowsky S, Cook SJ, Rawlings NC, 2004: Effectsof treatment with LH or FSH from 4 to 8 weeks of age onthe attainment of puberty in bull calves. Theriogenology 62,861–873.Bagu ET, Cook SJ, Gratton C, Rawlings NC, 2006a: Postnatalchanges in testicular gonadotropin receptors, serum gonadotropin,and testosterone concentrations and functionaldevelopment of the testes in bulls. Reproduction 132, 403–411.Bagu ET, Cook SJ, Honaramooz A, Aravindakshan JP,Huchkowsky S, Rawlings NC, 2006b: Changes in serumluteinizing hormone (LH) concentrations in response toluteinizing hormone releasing hormone (LHRH) in bullcalves that attained puberty early or late. Theriogenology66, 937–944.Bass JJ, Peterson AJ, Payne E, 1978: Plasma concentrations ofluteinizing hormone, testosterone and androstenedione incastrated and adrenalectomized bull calves. J Endocrinol 79,137–138.Berndtson WE, Igboeli G, Parker WG, 1987: The numbers ofSertoli cells in mature Holstein bulls and their relationshipto quantitative aspects of spermatogenesis. Biol Reprod 37,60–67.Bratton R, Musgrave S, Dunn H, Foote R, 1959: Causes andprevention of reproductive failures in dairy cattle. II.Influence of underfeeding and overfeeding from birth to80 weeks of age on growth, sexual development, and semenproduction of Holstein bulls. Bull Cornell Agric Exp Sta 94,45.Breier BH, Gluckman PD, Bass JJ, 1988: Influence ofnutritional status and oestradiol-17 beta on plasma growthhormone, insulin-like growth factors-I and -II and theresponse to exogenous growth hormone in young steers.J Endocrinol 118, 243–250.Brito LF, Barth AD, Rawlings NC, Wilde RE, Crews DH Jr,Boisclair YR, Ehrhardt RA, Kastelic JP, 2007a: Effect offeed restriction during calfhood on serum concentrations ofmetabolic hormones, gonadotropins, testosterone, and onsexual development in bulls. Reproduction 134, 171–181.Brito LF, Barth AD, Rawlings NC, Wilde RE, Crews DH Jr,Mir PS, Kastelic JP, 2007b: Effect of improved nutritionÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

300 N Rawlings, ACO Evans, RK Chandolia and ET Baguduring calfhood on serum metabolic hormones, gonadotropins,and testosterone concentrations, and on testiculardevelopment in bulls. Domest Anim Endocrinol 33, 460–469.Chandolia RK, Evans ACO, Rawlings NC, 1997a: Effect ofinhibition of increased gonadotrophin secretion before20 weeks of age in bull calves on testicular development.J Reprod Fertil 109, 65–71.Chandolia RK, Honaramooz A, Bartlewski PM, Beard AP,Rawlings NC, 1997b: Effects of treatment with LH releasinghormone before the early increase in LH secretion onendocrine and reproductive development in bull calves.J Reprod Fertil 111, 41–50.Chandolia RK, Honaramooz A, Omeke BC, Pierson R, BeardAP, Rawlings NC, 1997c: Assessment of development of thetestes and accessory glands by ultrasonography in bullcalves and associated endocrine changes. Theriogenology48, 119–132.Chandolia RK, Evans ACO, Rawlings NC, 1997d: Theinvolvement of dopaminergic and opioidergic neuronalsystems in the control of the early rise in LH secretion inbull calves. J Neuroendocrinol 9, 121–127.Curtis S, Amann R, 1981: Testicular development andestablishment of spermatogenesis in Holstein bulls. J AnimSci 53, 1645–1647.Deaver DR, Glass GD, Grieger DM, Reeves JJ, 1998: Effectsof estradid on secretion of LH, hypothalamic functionand testical development in bull calves. Domest AnimEndocrinal 5, 307–316.Evans A-CO, Currie WD, Rawlings NC, 1993: Opioidergicregulation of gonadotrophin secretion in the early prepubertalbull calf. J Reprod Fertil 99, 45–51.Evans A-CO, Davies FJ, Nasser LF, Bowman P, RawlingsNC, 1995: Differences in early patterns of gonadotrophinsecretion between early and late maturing bulls, and changesin semen characteristics at puberty. Theriogenology 43, 569–578.Evans A-CO, Pierson RA, Garcia A, McDougall LM, HrudkaF, Rawlings NC, 1996: Changes in circulating hormoneconcentrations, testes histology and testes ultrasonographyduring sexual maturation in beef bulls. Theriogenology 46,345–357.Godfrey RW, Lunstra DD, Schanbacher BD, 1992: Effect ofimplanting bull calves with testosterone propionate, dihydrotestosteronepropionate or oestradiol-17 beta prepubertallyon the pituitary–testicular axis and onpostpubertal social and sexual behaviour. J Reprod Fertil94, 57–69.Henricks D, Hoover J, Gimenez T, Grimes L, 1988: A study ofthe source of estradiol-17 beta in the bull. Horm Met Res 20,494–497.I’Anson H, Foster DL, Foxcroft GR, Booth PJ, 1991:Nutrition and reproduction. Oxf Rev Reprod Biol 13,239–311.Kaneko H, Noguchi J, Kikuchi K, Akagi S, Shimada A, TayaK, Watanabe G, Hasegawa Y, 2001: Production andendocrine role of inhibin during the early development ofbull calves. Biol Reprod 65, 209–215.Kennedy R I, Rawlings N, Cook S, 1985: Effects ofdihydrotestosterone and 5 alpha-androstane-3 alpha,17beta-diol on serum concentrations of LH and FSH in theprepubertal bull. Can J Anim Sci 65, 95–99.Lacroix A, Pelletier J, 1979: Short-term variations in plasmaLH and testosterone in bull calves from birth to 1 year ofage. J Reprod Fertil 55, 81–85.Lacroix A, Garnier DH, Pelletier J, 1977: Temporal fluctuationsof plasma LH and testosterone in Charolais bull calvesduring the first year of life. Ann Biol Anim Biochim Biophys17, 1013–1019.Lindner H, Mann T, 1960: Relationship between the contentof androgenic steroids in the testes and the secretory activityof the seminal vesicles in the bull. J Endocrinol 21, 341–361.Lunstra D, Echternkamp S, 1982: Puberty in beef bulls:acrosome morphology and semen quality in bulls ofdifferent breeds. J Anim Sci 55, 638–648.Lunstra DD, Ford JJ, Echterkamp SE, 1978: Puberty in beefbulls: hormone concentrations, growth, testicular development,sperm production and sexual aggressiveness in bulls ofdifferent breeds. J Anim Sci 46, 1054–1062.MacDonald R, Deaver D, 1993: Testicular development inbulls treated with recombinant bovine somatotropin.J Anim Sci 71, 1540–1545.Macdonald RD, Peters JL, Deaver DR, 1990: Effect ofnaloxone on the secretion of LH in infantile and prepubertalHolstein bull calves. J Reprod Fertil 89, 51–58.Macmillan K, Hafs H, 1968: Gonadal and extra gonadalsperm numbers during reproductive development of Holsteinbulls. J Anim Sci 27, 1614–1620.Macmillan K, Hafs H, 1969: Reproductive tract of Holsteinbulls from birth through puberty. J Anim Sci 28, 233–239.Madgwick S, Bagu ET, Duggavathi R, Bartlewski PM, BarrettDM, Huchkowsky S, Cook SJ, Beard AP, Rawlings NC,2007: Effects of treatment with GnRH from 4 to 8 weeks ofage on the attainment of sexual maturity in bull calves.Anim Reprod Sci, in press.Mann T, Rowson L, Short RV, Skinner J, 1967: Therelationship between nutrition and androgenic activity inpubescent twin calves, and the effect of orchitis. Endocrinology38, 455–468.Matsuzaki S, Uenoyama Y, Okuda K, Watanabe G, KitamuraN, Taya K, Cruzana MB, Yamada J, 2001: Prepubertalchanges in immunoreactive inhibin concentration in bloodserum and testicular tissue in Holstein bull calves. J Vet MedSci 63, 1303–1307.McAndrews JM, Stroud CM, Macdonald RD, Hymer WC,Deaver DR, 1993: Age-related changes in the secretion ofgrowth hormone in vivo and in vitro in infantile andprepubertal Holstein bull calves. J Endocrinol 139, 307–315.McAndrews JM, Peters JL, Deaver DR, 1994: Age-relatedchanges in the secretion of LH in vivo and in vitro ininfantile and prepubertal Holstein bull calves. J ReprodFertil 101, 453–458.McCarthy MS, Hafs HD, Convey EM, 1979: Serum hormonepatterns associated with growth and sexual development inbulls. J Anim Sci 49, 1012–1020.Miyamoto A, Umezu M, Ishii S, Furusawa T, Masaki J,Hasegawa Y, Ohta M, 1989: Serum inhibin, FSH, LH andtestosterone levels and testicular inhibin content in beef bullsfrom birth to puberty. Anim Reprod Sci 20, 165–178.Pruitt RJ, Corah LR, Stevenson JS, Kiracofe GH, 1986: Effectof energy intake after weaning on the sexual development ofbeef bulls. II. Age at first mating, age at puberty, testosteroneand scrotal circumference. J Anim Sci 63, 579–585.Rawlings NC, Cook SJ, 1986: Plasma concentrations oftestosterone, androstenedione, dihydrotestosterone, 5alpha-androstane-3alpha,17beta-diol,5alpha-androstane-3beta,17beta-diol,and androsterone in bull calves: response tohCG. Can J Anim Sci 66, 975–982.Rawlings NC, Evans ACO, 1995: Androgen negative feedbackduring the early rise in LH secretion in bull calves.J Endocrinol 145, 243–249.Rawlings N, Hafs H, Swanson LV, 1972: Testicular and bloodplasma androgens in Holstein bulls from birth throughpuberty. J Anim Sci 34, 435–440.Rawlings NC, Fletcher PW, Henricks DM, Hill JR, 1978:Plasma luteinizing hormone (LH) and testosterone levelsduring sexual maturation in beef bull calves. Biol Reprod19, 1108–1112.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Puberty in the Bull 301Renaville R, Massart S, Sneyers M, Falaki M, Gengler N,Burny A, Portetelle D, 1996: Dissociation of increases inplasma insulin-like growth factor I and testosterone duringthe onset of puberty in bulls. J Reprod Fertil 106, 79–86.Rodriguez RE, Wise ME, 1989: Ontogeny of pulsatile secretionof gonadotropin-releasing hormone in the bull calfduring infantile and pubertal development. Endocrinology124, 248–256.Rodriguez RE, Wise ME, 1991: Advancement of postnatalpulsatile luteinizing hormone secretion in the bull calf bypulsatile administration of gonadotropin-releasing hormoneduring infantile development. Biol Reprod 44, 432–439.Ronge H, Blum J, 1989: Insulin-like growth factor I duringgrowth in bulls. Reprod Nutr Dev 29, 105–111.Schanbacher BD, 1981: Importance of the episodic nature ofluteinizing hormone secretion for normal development ofthe bovine testis during puberty: interference with oestradiol-17beta. J Endocrinol 88, 393–400.Schanbacher BD, D’Occhio MJ, Kinder JE, 1982: Initiation ofspermatogenesis and testicular growth in oestradiol-17 beta-implanted bull calves with pulsatile infusion of luteinizinghormone releasing hormone. J Endocrinol 93, 183–192.Schuenemann GM, Mendis-Handagama SM, Hopkins FM,Kania SA, Schrick FN, 2007: Changes in the testis seminiferoustubules and interstitium in prepubertal bull calvesimmunised against inhibin at the time of gonadotropinadministration. Reprod Fertil Dev 19, 840–849.Secchiara P, Martorana F, Pellegrini S, Luisi M, 1976:Variation of plasma testosterone in developing Friesianbulls. J Anim Sci 42, 405–409.Shahab M, Nusser KD, Griel LC, Deaver DR, 1993: Effect ofa single intravenous injection of N-methyl-D,L-aspartic acidon secretion of luteinizing hormone and growth hormone inHolstein bull calves. J Neuroendocrinol 5, 469–473.Shahab M, Nusser KD, Peters JL, Deaver DR, 1995:Involvement of N-methyl-D-aspartate receptor using excitatoryamino acid neurotransmitters in control of pulsatilesecretion of LH during sexual development in Holstein bullcalves. J Reprod Fertil 105, 77–78.Sinowatz F, Amselgruber W, 1986: Postnatal development ofbovine Sertoli cells. Anat Embryol 174, 413–423.Sundby A, Velle W, 1980: Plasma concentration of testosteronein young bulls in relation to age, rate of weight gain andstimulation with human chorionic gonadotrophin. J Endocrinol86, 465–469.Sundby A, Andersen D, Purvis K, Hansson V, 1984: Testiculargonadotropin receptors, testicular testosterone, dihydrotestosteroneand androstenedione in the developing bull.Archiv Androl 12, 59–64.Tatman SR, Neuendorff DA, Wilson TW, Randel RD,2004: Influence of season of birth on growth andreproductive development of Brahman bulls. Theriogenology62, 93–102.Wildeus S, Randel R, Humphrey W, 1988: Age-relatedchanges in the suppression of serum luteinizing hormoneby estradiol-17b in developing steers. Theriogenology 29,1065–1073.Wise ME, Rodriguez RE, Kelly CM, 1987: Gonadal regulationof LH secretion in prepubertal bull calves. DomestAnim Endocrinol 4, 175–181.Wolf F, Almquist J, Hale E, 1965: Prepuberal behavior andpuberal characteristics of beef bulls on high nutrientallowance. J Anim Sci 24, 761–765.Wrobel K, 1990: The postnatal development of the bovineLeydig cell population. Reprod Dom Anim 25, 51–60.Wrobel KH, 2000: Prespermatogenesis and spermatogoniogenesisin the bovine testis. Anat Embryol 202, 209–222.Author’s address (for correspondence): NC Rawlings, The DeansOffice, WCVM, University of Saskatchewan, Saskatoon, SaskatchewanS7N 5B4, Canada. E-mail: norman.rawlings@usask.caConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 302–309 (2008); doi: 10.1111/j.1439-0531.2008.01178.xISSN 0936-6768Epigenetic Regulation of Foetal Development in Nuclear Transfer Animal ModelsA Dinnyes 1,2 , XC Tian 3 and X Yang 31 Genetic Reprogramming Group, Agricultural Biotechnology Centre, Godollo, Hungary; 2 Molecular Animal Biotechnology Laboratory, Szent IstvanUniversity, Godollo, Hungary; 3 Department of Animal Science ⁄ Center for Regenerative Biology, University of Connecticut, Storrs, CT, USAContentsSomatic cell nuclear transfer (SCNT, ‘cloning’) holds greatpotential for agricultural applications, generation of medicalmodel animals, transgenic farm animals or by ‘therapeuticcloning’ for generating human embryonic stem cells for thetreatment of human diseases. However, the low survival rate ofSCNT-derived pregnancies represents a serious limitation ofthe current technology. In order to overcome this hurdle, adeeper understanding of the epigenetic reprogramming of thesomatic cell nuclei and its effect on the pregnancy is needed.Here we review the literature on nuclear reprogramming bySCNT, including studies of gene expression, DNA methylation,chromatin remodelling, genomic imprinting and Xchromosome inactivation. Reprogramming of genes expressedin the inner cell mass, from which the body of the foetus isformed, seems to be highly efficient. Defects in the extraembryonictissues are probably the major cause of the lowsuccess rate of reproductive cloning. Methods to partiallyovercome such problems exist, yet more future research isneeded to find practical and efficient methods to remedy thisproblem. Improvement of the survival of foetuses is a centralissue for the future of agricultural SCNT not only for itseconomic viability, but also because in lack of improvementsin animal welfare current regulations can block the use of themethod in the EU and several other countries.Importance of Somatic Cell Nuclear TransferSomatic cell nuclear transfer is a promising technologyfor agricultural applications on its own or combinedwith transgenesis for the generation of transgenic farmanimals or novel medical animal models. Somatic cellnuclear transfer offsprings have been produced in 18mammalian species, among these the largest number ofsomatic cell nuclear transfer (SCNT) progeny have beenborn in cattle and pig, with an estimated number of 3000and 600, respectively. Although, most of the ‘clones’(F 0 ) are not alive by now, by adding the number (severalthousands) of healthy progeny derived by naturalbreeding from such ‘clone’ F 0 founders, the impact ofthe technology is much bigger. Even though the numberof F 0 ‘clones’ produced for food purposes is still small,the potential of subsequent generations developed fromthose ‘clones’ by sexual reproduction is already high.The recently published risk assessment report of FDA(Animal Cloning 2006) on the food safety of the milkand meat of SCNT-derived animals concludes that thereare no food safety concerns with such products, thusincreases the future chances for more animals producedfor human consumption.Despite all these advances, the technology is still veryinefficient, mostly due to the major (80–99%) loss ofembryos transferred during the pregnancy and withfurther losses in the perinatal period and during the firstyear of life. Such inefficiency makes the technologycommercially less valuable and socially hardly acceptable.This later issue is particularly important in the EUwhere the legal framework for animal welfare regulationsand the negative public perception can block theuse of such a wasteful method, therefore improvementof the survival of foetuses is a central issue for the futureof agricultural SCNT.Pre- and Post-implantation Development ofSCNT EmbryosBasic steps of nuclear reprogrammingSomatic cell nuclear reprogramming involves two majorsteps: (i) dedifferentiation of the already differentiateddonor cell to a totipotent ‘embryonic’ state, followed by(ii) redifferentiation of SCNT embryos to differentprecursor and somatic cell types during later development.The first step of nuclear reprogramming involvesthe erasure of the donor cell epigenetic pattern after NTand the re-establishment of embryonic epigenetic characteristicsand gene expression in the SCNT embryo.The second step of nuclear reprogramming refers toredifferentiation of SCNT embryos from a totipotentstatus to various differentiated stages of organogenesisduring post-implantation development. During thesesteps there is evidence for normal and abnormalreprogramming, as well. Recent results indicate thatmany of the defects in SCNT embryos occur in the laterreprogramming stage, which may be attributable toproblems with placental development and function.Pre- and post-implantation developmentMammalian embryos derived by SCNT are capable ofdevelopment to the blastocyst stage with an efficiency of30–50%, comparable to that of embryos produced byin vitro fertilization (IVF) in species such as cattle andpigs (30–40%) (Cibelli et al. 2002; Zhu et al. 2002).Unlike embryos derived from fertilization, however,most SCNT embryos die during post-implantationdevelopment, and those that survive to term arefrequently defective.In cattle Panarace et al. (2007) reported efficiency ofSCNT up to 20%, but analysing data in three countries,Brazil, Argentine and USA over 5 years shows thatamong the 3374 cloned embryos transferred into surrogatefemales, 9% live calves were born, 8% were alive24 h after birth, and 7% survived for 150 days or more.Within a given species, success rates can vary extensivelyreflecting a lack of full understanding of the roleof various factors involved in the SCNT process such asÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Epigenetics of Foetal Development in Clones 303donor cell selection, cell cycle stage, culture conditions.For unknown reasons, approximately one-third of thedonor cell lines lead to a success rate of live calvesobtained from initiated pregnancy as high as 40%whereas approximately 25% of donor cell lines totallyfailed (Panarace et al. 2007). Those differences in therate of live calves even occur when donor cell linecultures are run simultaneously within the same experimentalprogram. The different cell lines with noevidence of abnormal chromosomal constitution gavethe same high number of blastocysts in vitro afterSCNT, irrespectively of the following success rate ofdevelopment (Renard et al. 2007).Higher success rates of SCNT in cattle are largely dueto the extensive knowledge on the female (and male)reproductive physiology in this economically importantspecies. In most mammalian species studied thus far, thesurvival rate to birth for cloned blastocysts is onlyapproximately 1–5%, compared with a 30–60% birthrate for IVF blastocysts. Many pathologies have beendescribed in conceptuses and neonates derived fromSCNT (Cibelli et al. 2002). Large offspring syndrome(LOS), which also occurs in cattle and sheep derivedfrom IVF, is common in cloned cattle, sheep and mice.LOS refers to a heterogeneous group of symptoms:notably, large size at birth and severe placental deficiency(Young et al.1998). Other symptoms are morevariable but include prolonged gestation, dystocia(strenuous labour), foetal and placental oedema, abnormalsize of organs, hydrallantois and hydramnios,respiratory problems and perinatal death (Young et al.1998; Farin et al. 2006). Most of the observed defectsinvolve abnormal placentation (Cibelli et al. 2002).Placentas of cloned calves have fewer but much largerplacentomes, presumably to compensate for the reducednumber of sites of foeto-maternal exchange (Chavatte-Palmer et al. 2002). In SCNT mice placentas are alwaysenlarged (Suemizu et al. 2003). It is possible that manyof the observed abnormalities are consequences ofprimary defects in placental function (Constant et al.2006). SCNT embryos from mice and cattle develop tothe blastocyst stage with a high success rate, and NTembryonicstem cells (NT-ESCs) can be derived fromsuch embryos with high efficiency (Wang et al. 2005;Wakayama et al. 2006) . Indeed, we have producednumerous NT-ESC lines in mouse from donor cellsincluding cumulus, fibroblast and ESCs. Gene expressionanalyses of such cell lines have shown almost nodifferences compared to the ESCs with the same geneticbackground or from each others, regardless of the celltype of donor cells (Dinnyes et al., unpublished results).The molecular, epigenetic mechanisms behind the placentalabnormalities are particularly interesting as aunique model for epigenetic regulation of the pregnancy,as discussed in details below.Molecular Level Changes in SCNT EmbryosGlobal gene expressionMost SCNT embryos have abnormalities at the molecularlevel. Early reports on reprogramming of variouscandidate genes in SCNT embryos are inconsistent(Daniels et al. 2000; Wrenzycki et al. 2001; Inoue et al.2006). For example, expression of the pluripotencymarker gene Pou5f1 (formerly known as Oct4) afterSCNT has been reported to be both normal (Danielset al. 2000; Jouneau et al. 2006), or abnormal (Bortvinet al. 2003). The global gene expression pattern inSCNT bovine blastocysts reflects well the extent ofnuclear reprogramming. In our studies the gene expressionprofiles of SCNT embryos were very different fromthose of somatic donor cells, and they resembled thoseof naturally fertilized embryos with

304 A Dinnyes, XC Tian and X Yanggrowth and development. Some genes are onlyimprinted in the placenta (Wagschal and Feil 2006).In the mouse distinct allelic expression patterns ofimprinted genes can be established as early as the twocellstage, and by the blastocyst stage, monoallelicexpression of many imprinted genes is observed, whichclosely follows the re-establishment of genome-wideDNA methylation in early development (Latham1999).Epigenetic perturbations in SCNT embryosFor cloning to succeed, the adult pattern of epigeneticmodifications must be erased, and the patternsdescribed above must be established in the SCNTembryos. In the development of naturally fertilizedembryos, epigenetic marks are established over longperiods of time during gametogenesis and fertilization(Morgan et al. 2005). In SCNT, however, an adultsomatic pattern of epigenetic modification that isnormally very stable must be reversed within a shortperiod of time before zygotic genome activation(Zuccotti et al. 2000). Although no study has beenable to measure the epigenetic status of SCNT embryosand then follow their development after transfer torecipients, several studies have shown aberrant methylationpatterns in SCNT bovine embryos whencompared with in vivo- or in vitro-fertilized controls(Bourc’his et al. 2001; Dean et al. 2001; Kang et al.2002). Interestingly, at the blastocyst stage, the methylationstatus of the ICM is relatively normal, whereasthe trophoblast cells show abnormal hypermethylation(Dean et al. 2001). Methylation of histone H3K9 alsoshows a similar pattern, with abnormally high levels ofmethylation in the trophectoderm of cloned bovineblastocysts (Santos et al. 2003). Studies on the acetylationof histones in cloned embryos have foundaberrancies, including hypoacetylation (Santos et al.2003; Enright et al. 2005; Suteevun et al. 2006). Thehistone deacetylase (HDAC) inhibitor trichostatin A(TSA) increases cloning efficiency including live births(Kishigami et al. 2006). The reprogramming of imprintedgenes in SCNT can be altered and someabnormalities in cloned animals resemble those seen inhuman imprinting diseases and in mutant mice withexperimentally disrupted imprinting (Debaun et al.2003). SCNT animals with abnormal phenotypes canbe naturally bred, and all their offspring show normalphenotypes, suggesting that the abnormalities of theclones are due to epigenetic errors rather than geneticmutations (Tamashiro et al. 2002). There is evidencethat the placenta is especially vulnerable to problems ofthis type. In cloned mice several imprinted genes showabnormally low expression in the placenta, whereas nodifferences are seen in foetal expression (Inoue et al.2002). We also found abnormal allelic expressionpattern of the imprinted IGF2R gene in placentasbut not in the organs of cloned bovine calves (Yanget al. 2005). In our microarray study (Smith et al.2005), we found that Cd81, a gene that in mice showsimprinting only in the placenta, was downregulated inSCNT bovine blastocysts compared with in vivoembryos.X chromosome inactivationIn all eutherian mammalian species XCI is used toachieve an equality of expression of X-linked genesbetween males and females (Lyon 1961). Inactivationof the X chromosome is achieved through epigeneticmechanisms. The Xist gene encodes a non-codingRNA that is expressed in cis from the chromosomethat is to be inactivated (Borsani et al. 1991). Xisttranscripts coat the X chromosome and recruit chromatin-modifyingproteins that convert the X chromosomeinto a heterochromatic, silenced state (Okamotoet al. 2004). X chromosome inactivation is also subjectto imprinting. In mice imprinted XCI occurs duringpre-implantation development, with the paternal chromosomebeing preferentially silenced in the placenta(Takagi and Sasaki 1975). By preferential paternalXist expression at the time of zygotic genome activation,leading to Xist transcript accumulation andsilencing of the paternal X chromosome at the fourcellstage (Okamoto et al. 2005). This pattern persistsin the trophectoderm lineage, such that only thematernal X chromosome is expressed in the placenta.In the ICM, however, the paternal X chromosome isreactivated, after which the paternal and maternalchromosomes are subject to random inactivation inthe developing embryo (Okamoto et al. 2005). There isalso evidence of imprinted XCI in cattle placentas(Xue et al. 2002). Although a systematic analysis ofXCI in cattle has yet to be done, Xist transcripts havebeen detected as early as the two-cell stage (De LaFuente et al. 1999). The late-replicating (and presumptiveinactive) X chromosome is first observed at theearly blastocyst stage (day 8), and XCI is complete byday 14.X chromosome inactivation perturbations during SCNTIn SCNT animals that are generated from femaledonor nuclei, XCI appears to be random, which isnormal. This means that the inactive X chromosomefrom the donor nucleus must have been reactivated,after which the embryo must have reestablishedrandom activation of maternal and paternal chromosomes.This pattern of XCI in the ICM of clonedmouse blastocysts was confirmed to be normal. Otherstudies have confirmed that the inactive X chromosomeis reactivated in cloned mouse blastocysts (Eggan et al.2000) but also observed aberrant expression of X-linked genes at later developmental stages (particularlyin the mid- to late-gestation placenta) that failed toshow the normal pattern of imprinted (paternal-inactive)XCI (Nolen et al. 2005). Placental XCI is usuallyskewed in these clones, with the original inactive Xchromosome inactivated in the (female) donor nucleus.This suggests that no reprogramming occurs or thataberrant reprogramming of XCI occurs later in theextraembryonic lineages after NT. We have performedsimilar studies in cattle, with similar results. At theblastocyst stage, we did not find any evidence forabnormal expression of X-linked genes in clonedembryos (Smith et al. 2005). In later development,however, we observed biallelic expression of X-linkedÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Epigenetics of Foetal Development in Clones 305genes in the placentas of dead clones rather than thematernal-specific expression that is expected if placentalXCI is normal. Interestingly, the placentas ofsurviving clones, unlike those of dead clones, had onlyone active X chromosome, suggesting that aberrantpatterns of XCI might have contributed to foetal death(Xue et al. 2002). Aberrations in X-linked gene expressioncould have severe consequences for placentaldevelopment (Hemberger 2002).Trophectoderm and placental defects in SCNT embryosRecent molecular evidence supports the hypothesis thatthe placental lineage is especially vulnerable to problemsarising from reprogramming of the somatic nucleus afterNT. Anomalies in the trophectoderm of cloned embryoshave been described as early as the blastocyst stage.SCNT bovine blastocysts have been found to have alower ratio of trophectoderm-to-ICM cells comparedwith fertilized ones (Koo et al. 2002). Furthermore, TP-1 (also known as IFN-tau), the gene specificallyexpressed by the trophoblast for maternal recognitionof pregnancy, was abnormally expressed in SCNTbovine blastocysts (Wrenzycki et al. 2001). In theplacentas or placentomes of SCNT bovine foetusesaberrant gene expression has been detected as early asday 25 and throughout gestation to term (Hashizumeet al. 2002; Hill et al. 2002). Furthermore, 60 proteinswere differentially expressed in term placentas of clonedcalves compared with fertilized controls (Kim et al.2005). The majority of the cloned mouse embryos diebefore a functional placenta can develop (Jouneau et al.2006) showing phenotypic defects during gastrulation(E7–E8) characterized by an abnormal ratio between theembryonic and extraembryonic portions of the embryo.The abnormal phenotypes were not fully rescued withthe addition of ESCs or ICM cells from normal embryosbut were rescued by the addition of tetraploid cells.Gene expression in cloned mice derived from nuclei ofdifferent cell types shown some cell type-specific effects,but most of the abnormalities were independent ofdonor cell type and seemed to be a consequence of theNT procedure, with at least 4% of genes expressed in theplacenta showing dysregulation (Humpherys et al.2002).Strategies to Increase the Pregnancy Rates andReduce Foetal Abnormalities After SCNTFrom a practical point of view SCNT pregnanciespresent different level of problems: (i) negligible reprogrammingerrors, without any obvious physiologicalconsequences; (ii) small errors, where favourable environmentalconditions or veterinary care can compensatefor negative effects; (iii) serious reprogramming errorsresulting in placental or foetal failures and loss ofpregnancy.Currently there is a lack of diagnostic methods topredict the expected level of problems in individualembryos and for their pre-selection prior to ET.However, there are several different practical strategieswhich might increase the live birth rates of SCNTpregnancies:Modification of reprogramming success of both foetal andplacental genesThis strategy includes efforts to find the best combinationof many variables during SCNT. The main effortsare focused on the core elements of the SCNT procedure:the donor cells, the recipient cytoplast, and certainmodifiers of the reprogramming.The ‘best’ donor cellFinding of cell types more amenable for reprogrammingwould improve the SCNT, although the existing data isstill controversial concerning the effect of cell type onthe reprogramming – as approximately 200 cell typescould be tested, the task is rather big.Improving donor cell culture methods and using lowpassage number of such cells could help to avoid geneticaberrations and mutations accumulated during cellculture – although long term cultured cells can be usedas well for SCNT (Kubota et al. 2000) and telomereshortening seems not to be a limiting factor in SCNT(Lanza et al. 2000). The genetic origin of the cells is animportant factor, as it was studied in mouse in details(Wakayama et al. 1999; 2001; Rideout et al. 2000;Eggan et al. 2001). However, there is almost no dataon the effect of genetic component in domestic animals.The cell cycle status of the donor cells plays animportant role, and several studies reported on thepositive effect of serum starvation induced G0 (Wilmutet al. 1997) or cell culture confluency induced G1 stages(Cibelli et al. 1998a, 1998b, 2002; Wells et al. 2003). Thecell cycle must be in harmony with the state of therecipient cytoplast (Campbell et al. 1996; 2002; Wilmutet al. 2002) (see below).The ‘best’ recipient cytoplastThe origin of the recipient oocyte can be in vivo derivedor in vitro matured. The later ones in farm animals aremuch less expensive if slaughterhouse ovaries areavailable. In mouse there is evidence, that naturallyovulated oocytes are better quality than superovulatedones (Hiiragi and Solter 2005), but in farm animals thischoice would increase the cost substantially. The geneticorigin of the recipient oocyte also plays a role, accordingto mouse data (Wakayama et al. 1999; Gao et al. 2004),but there is very limited information on this issue inother species. Genotype in that case is probably importantdue to variations in the amount of maternal-originreprogramming factors available in the cytoplast. Furthermore,individual differences among oocyte donorsprobably exist, but scientific examination of such factorsis rather difficult, due to the low number of observationspossible when the oocyte donors are not used repetitively.Usually cytoplasts for SCNT are generated byenucleation of non-activated Metaphase II stage oocytes(Wilmut et al. 1997, 2002; Cibelli et al. 2002), butTelophase II stage can be also used as recipient withsimilar efficacy (Bordignon and Smith 1998; Baguisiet al. 1999), and recently it was demonstrated, that theuse of mitotic zygotes is also a viable option (Gredaet al. 2006; Schurmann et al. 2006; Egli et al. 2007).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

306 A Dinnyes, XC Tian and X YangIncreasing the volume of the recipient cytoplast can alsoprovide a better reprogramming environment (Peuraet al. 1998).In vitro modifiers of reprogrammingA number of in vitro approaches have been devised toincrease SCNT success. These include treating donorcells with pharmacological agents to modify theirepigenetic marks (Enright et al. 2003, 2005; Shi et al.2003) or cell cycle stage (Wells et al. 2003). In mice theHDAC inhibitor TSA can significantly improve theefficiency of both reproductive cloning (Enright et al.2005; Kishigami et al. 2006) and NT-ESC derivation(Kishigami et al. 2006). The efficiency of TSA treatmentvaries among species, but preliminary data in rabbit(Dinnyes et al., unpublished results) suggest that chemicaltreatments may be a promising alternative. Anotherapproach is fusing transiently permeabilized cells containingartificially condensed chromatin (Sullivan et al.2004). Activating with sperm rather than by artificialmethods (Schurmann et al. 2006) can have a positiveinfluence on SCNT by providing a better Ca-signallingimpulse and contributing some factors (mRNAs, microRNAs) to the cytoplast. Serial NT (Ono et al. 2001) hasbeen shown to improve the reprogramming process, aswell.Favourable environmental conditionsThe environmental conditions to which the SCNTembryos and foetuses are exposed influence the epigenomeand can compensate for a certain level of geneticreprogramming errors. Epigenetics also depend on thatcomplex ‘environment’ of the genes in question and IVCconditions can play a major role in species, where theIVC period is fairly long prior to embryo transfer.Early-stage embryo transfer can reduce aberrationsaccumulated due to IVC, and strong empirical evidencein rabbit support the importance of asynchrony betweenthe SCNT embryo age and the post-ovulatory age of therecipient oviduct or uterus (Chesne et al. 2002).Co-transfer of parthenogenetically activated or in vivoembryos together with the SCNT ones can overcomesome of the signalling and ‘communicational’ problems ofthe SCNT embryos with the uterine environment. Thisapproach was beneficial in pig (De Sousa et al. 2002) andin mouse (Dinnyes et al., unpublished data).Husbandry, nutritional and veterinary managementof pregnancy and delivery can be particularly importantin SCNT cases and must be aimed at improved birthrates and animal welfare for both the recipient femalesand the newborns. Measures include precise monitoringof foetal and placental development using tools such asrepeated ultrasonography and, if needed, treating hydropsand other in-utero complications, providinginduced deliveries, timed caesarean sections, and intensivecare options in the perinatal period.Tetraploid complementationThe method of tetraploid placenta complementation isoften used for producing ESC-derived mice (Nagyet al. 1990) when the ESCs are mixed with atetraploid embryo, which will contribute only to theplacenta and not the foetus itself. In case of SCNTembryos the extra trophectoderm cells contributed bythe tetraploid embryo can rescue the foetus (Jouneauet al. 2006). In mouse NT-ESCs can develop to termwith a much higher efficiency when combined withfertilized tetraploid placenta complement, comparedwith direct transfer of SCNT embryos (Eggan et al.2001; Hochedlinger and Jaenisch 2002). These studiesshowed supporting evidence that poor cloningefficiency is largely due to placenta abnormality. Incattle chimeric calves have been generated with NT-ESC-like cells (Cibelli et al. 1998a, 1998b; Iwasakiet al. 2000).Embryo aggregationCompleting SCNT embryos with each other by embryoaggregation has been shown to be beneficial as presumablyeach embryo will contain a different set ofreprogramming errors, potentially compensating forsome of the individual errors. Indeed, in mouse,aggregation of two or three SCNT embryos led toincreased cell numbers, normalized gene expressions andeightfold higher in vivo development of the resultingsingle blastocysts (Boiani et al. 2003). In bovine, similaraggregation of three one-cell NT embryos during in vitroculture affected development of embryonic and somaticcloned embryos differently. Aggregates of embryonicclones showed a higher cloning efficiency but there wasno evidence for complementary interactions. In contrast,SCNT aggregates developed normally in vitro, butthe in vivo development was not improved (Misica-Turner et al. 2007).ConclusionsAfter the first decade of mammalian SCNT it is notclear yet how to overcome the current serious limitationsof the technology, but the overall progress ispromising, and there are no reasons to doubt thatwithin another decade a much more efficient SCNTmethodology will be available for the breeders. Asdescribed above, several lines of evidence point towardto the hypothesis of relatively normal reprogrammingin the embryonic lineage of SCNT embryos butaberrant reprogramming in their trophectoderm (Cibelliet al. 1998a, 1998b; Iwasaki et al. 2000; Jouneauet al. 2006; Wakayama et al. 2006). A better reprogrammingis the key for safety, reproducibility andanimal welfare. Fundamental knowledge on reprogrammingmechanisms, epigenetics and developmentalpathways will provide the base for practical resolutions.Replacement or rescue of faulty SCNT placentaSomatic cell nuclear transfer placental deficiencies couldbe rescued by providing a well functioning placenta.AcknowledgementsThe work was supported by EU FP6 (MEXT-CT-2003-509582,LSHG-CT-2006-518240, MRTN-CT-2006-035468).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Epigenetics of Foetal Development in Clones 307ReferencesAnimal Cloning, 2006. A Draft Risk Assessment. FDA,Rockville, MD, USA. Available at: http://www.fda.gov/cvm/Documents/Cloning_Risk_Assessment.pdf.Baguisi A, Behboodi E, Melican DT, Pollock JS, DestrempesMM, Cammuso C, Williams JL, Nims SD, Porter CA,Midura P, Palacios MJ, Ayres SL, Denniston RS, HayesML, Ziomek CA, Meade HM, Godke RA, Gavin WG,Overstrom EW, Echelard Y, 1999: Production of goats bysomatic cell nuclear transfer. Nat Biotechnol 17, 456–461.Boiani M, Eckardt S, Leu NA, Scholer HR, McLaughlin KJ,2003: Pluripotency deficit in clones overcome by clone-cloneaggregation: epigenetic complementation? EMBO J 22,5304–5312.Bordignon V, Smith LC, 1998: Telophase enucleation: animproved method to prepare recipient cytoplasts for use inbovine nuclear transfer. Mol Reprod Dev 49, 29–36.Borsani G, Tonlorenzi R, Simmler Mc, Dandolo L, Arnaud D,Capra V, Grompe M, Pizzuti A, Muzny D, Lawrence C,Willard Hf, Avner P, Ballabio A, 1991: Characterization ofa murine gene expressed from the inactive X chromosome.Nature 351, 325–329.Bortvin A, Eggan K, Skaletsky H, Akutsu H, Berry DL,Yanagimachi R, Page DC, Jaenisch R, 2003: Incompletereactivation of Oct4-related genes in mouse embryos clonedfrom somatic nuclei. Development 130, 1673–1680.Bourc’his D, Le Bourhis D, Patin D, Niveleau A, Comizzoli P,Renard JP, Viegas-Pequignot E, 2001: Delayed and incompletereprogramming of chromosome methylation patternsin bovine cloned embryos. Curr Biol 11, 1542–1546.Campbell KH, Loi P, Otaegui PJ, Wilmut I., 1996: Cell cycleco-ordination in embryo cloning by nuclear transfer. RevReprod 1, 40–46.Cardoso MC, Leonhardt H, 1999: DNA methyltransferase isactively retained in the cytoplasm during early development.J Cell Biol 147, 25–32.Chavatte-Palmer P, Heyman Y, Richard C, Monget P,LeBourhis D, Kann G, Chilliard Y, Vignon X, Renard JP,2002: Clinical, hormonal, and hematologic characteristics ofbovine calves derived from nuclei from somatic cells. BiolReprod 66, 1596–1603.Chesne P, Adenot PG, Viglietta C, Baratte M, Boulanger L,Renard JP, 2002: Cloned rabbits produced by nucleartransfer from adult somatic cells. Nat Biotechnol 20, 366–369.Cibelli JB, Stice SL, Golueke PJ, Kane JJ, Jerry J, Blackwell C,de Leon FAP, Robl JM, 1998a: Cloned transgenic calvesproduced from nonquiescent fetal fibroblasts. Science 280,1256–1258.Cibelli JB, Stice SL, Golueke PJ, Kane JJ, Jerry J, Blackwell C,de Leon FAP, Robl JM, 1998b: Transgenic bovine chimericoffspring produced from somatic cell-derived stem-like cells.Nat Biotechnol 16, 642–646.Cibelli JB, Campbell KH, Seidel GE, West MD, Lanza RP,2002: The health profile of cloned animals. Nat Biotechnol20, 13–14.Constant F, Guillomot M, Heyman Y, Vignon X, Laigre P,Servely JL, Renard JP, Chavatte-Palmer P, 2006: Largeoffspring or large placenta syndrome? Morphometric analysisof late gestation bovine placentomes from somaticnuclear transfer pregnancies complicated by hydrallantois.Biol Reprod 75, 122–130.Daniels R, Hall V, Trounson AO, 2000: Analysis of genetranscription in bovine nuclear transfer embryos reconstructedwith granulosa cell nuclei. Biol Reprod 63, 1034–1040.De La Fuente R, Hahnel A, Basrur PK, King WA, 1999: Xinactive-specific tran-script (Xist) expression and X chromosomeinactivation in the preattachment bovine embryo.Biol Reprod 60, 769–775.De Sousa PA, Dobrinsky JR, Zhu J, Archibald AL, Ainslie A,Bosma W, Bowering J, Bracken J, Ferrier PM, Fletcher J,Gasparrini B, Harkness L, Johnston P, Ritchie M, RitchieWA, Travers A, Albertini D, Dinnyes A, King TJ, Wilmut I,2002: Somatic cell nuclear transfer in the pig: control ofpronuclear formation and integration with improved methodsfor activation and maintenance of pregnancy. BiolReprod 66, 642–650.Dean W, Santos F, Stojkovic M, Zakhartchenko V, Walter J,Wolf E, Reik W, 2001: Conservation of methylationreprogramming in mammalian development: aberrantreprogramming in cloned embryos. Proc Natl Acad Sci USA98, 13734–13738.Dean W, Santos F, Reik W, 2003: Epigenetic reprogrammingin early mammalian development and following somaticnuclear transfer. Semin Cell Dev Biol 14, 93–100.Debaun MR, Niemitz EL, Feinberg AP, 2003: Association ofin vitro fertilization with Beckwith-Wiedemann syndromeand epigenetic alterations of Lit1 and H19. Am J HumGenet 72, 156–160.Eggan K, Akutsu H, Hochedlinger K, Rideout W III,Yanagimachi R, Jaenisch R, 2000: X-chromosome inactivationin cloned mouse embryos. Science 290, 1578–1581.Eggan K, Akutsu H, Loring J, Jackson-Grusby L, Klemm M,Rideout WM, Yanagimachi R, Jaenisch R, 2001: Hybridvigor, fetal overgrowth, and viability of mice derived bynuclear cloning and tetraploid embryo complementation.Proc Natl Acad Sci U S A 98, 6209–6214.Egli D, Rosains J, Birkhoff G, Eggan K, 2007: Developmentalreprogramming after chromosome transfer into mitoticmouse zygotes. Nature 447, 679–685.Enright BP, Kubota C, Yang X, Tian XC, 2003: Epigeneticcharacteristics and development of embryos cloned fromdonor cells treated by trichostatin A or 5-aza-2’-deoxycytidine.Biol Reprod 69, 896–901.Enright BP, Sung LY, Chang CC, Yang X, Tian XC, 2005:Methylation and acetylation characteristics of cloned bovineembryos from donor cells treated with 5-aza-2¢-deoxycytidine.Biol Reprod 72, 944–948.Farin PW, Piedrahita JA, Farin CE, 2006: Errors in developmentof fetuses and placentas from in vitro-produced bovineembryos. Theriogenology 65, 178–191.Gao S, Czirr E, Chung YG, Han Z, Latham KE, 2004:Genetic variation in oocyte phenotype revealed throughparthenogenesis and cloning: correlation with differencesin pronuclear epigenetic modification. Biol Reprod 70,1162–1170.Greda P, Karasiewicz J, Modlinski JA, 2006: Mouse zygotesas recipients in embryo cloning. Reproduction 132, 741–748.Hashizume K, Ishiwata H, Kizaki K, Yamada O, TakahashiT, Imai K, Patel OV, Akagi S, Shimizu M, Takahashi S,Katsuma S, Shiojima S, Hirasawa A, Tsujimoto G, TodorokiJ, Izaike Y, 2002: Implantation and placental developmentin somatic cell clone recipient cows. Cloning StemCells 4, 197–209.Hemberger M, 2002: The role of the X chromosome inmammalian extra embryonic development. Cytogenet GenomeRes 99, 210–217.Hiiragi T, Solter D, 2005: Reprogramming is essential innuclear transfer. Mol Reprod Dev 70, 417–421.Hill JR, Schlafer DH, Fisher PJ, Davies CJ, 2002: Abnormalexpression of trophoblast major histocompatibility complexclass I antigens in cloned bovine pregnancies is associatedwith a pronounced endometrial lymphocytic response. BiolReprod 67, 55–63.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

308 A Dinnyes, XC Tian and X YangHochedlinger K, Jaenisch R, 2002: Monoclonal mice generatedby nuclear transfer from mature B and T donor cells.Nature 415, 1035–1038.Humpherys D, Eggan K, Akutsu H, Friedman A, HochedlingerK, Yanagimachi R, Lander ES, Golub TR, Jaenisch R,2002: Abnormal gene expression in cloned mice derivedfrom embryonic stem cell and cumulus cell nuclei. Proc NatlAcad Sci U S A 99, 12889–12894.Inoue K, Kohda T, Lee J, Ogonuki N, Mochida K, NoguchiY, Tanemura K, Kaneko-Ishino T, Ishino F, Ogura A,2002: Faithful expression of imprinted genes in cloned mice.Science 295, 297.Inoue K, Ogonuki N, Miki H, Hirose M, Noda S, Kim JM,Aoki F, Miyoshi H, Ogura A, 2006: Inefficient reprogrammingof the hematopoietic stem cell genome followingnuclear transfer. J Cell Sci 119, 1985–1991.Iwasaki S, Campbell KH, Galli C, Akiyama K, 2000:Production of live calves derived from embryonic stem-likecells aggregated with tetraploid embryos. Biol Reprod 62,470–475.Jouneau A, Zhou Q, Camus A, Brochard V, Maulny L,Collignon J, Renard JP, 2006: Developmental abnormalitiesof NT mouse embryos appear early after implantation.Development 133, 1597–1607.Kang YK, Park JS, Koo DB, Choi YH, Kim SU, Lee KK,Han YM, 2002: Limited demethylation leaves mosaic-typemethylation states in cloned bovine pre-implantationembryos. EMBO J 21, 1092–1100.Kim HR, Kang JK, Yoon JT, Seong HH, Jung JK, Lee HM,Park CS, Jin DI, 2005: Protein profiles of bovine placentaderived from somatic cell nuclear transfer. Proteomics 5,4264–4273.Kishigami S, Mizutani E, Ohta H, Hikichi T, Van Thuan N,Wakayama S, Bui HT, Wakayama T, 2006: Significantimprovement of mouse cloning technique by treatment withtrichostatin A after somatic nuclear transfer. BiochemBiophys Res Commun 340, 183–189.Koo DB, Kang YK, Choi YH, Park JS, Kim HN, Oh KB, SonDS, Park H, Lee KK, Han YM, 2002: Aberrant allocationsof inner cell mass and trophectoderm cells in bovine nucleartransfer blastocysts. Biol Reprod 67, 487–492.Kubota C, Yamakuchi H, Todoroki J, Mizoshita K, TabaraN, Barber M, Yang XZ, 2000: Six cloned calves producedfrom adult fibroblast cells after long-term culture. Proc NatlAcad Sci U S A 97, 990–995.Lanza RP, Cibelli JB, Blackwell C, Cristofalo VJ, FrancisMK, Baerlocher GM, Mak J, Schertzer M, Chavez EA,Sawyer N, Lansdorp PM, West MD, 2000: Extension of celllife-span and telomere length in animals cloned fromsenescent somatic cells. Science 288, 665–669.Latham KE, 1999: Epigenetic modification and imprinting ofthe mammalian genome during development. Curr Top DevBiol 43, 1–49.Li E, 2002: Chromatin modification and epigenetic reprogrammingin mammalian development. Nat Rev Genet 3, 662–673.Lyon MF, 1961: Gene action in the X-chromosome of themouse (Mus musculus L.). Nature 190, 372–373.Mayer W, Niveleau A, Walter J, Fundele R, Haaf T, 2000:Demethylation of the zygotic paternal genome. Nature 403,501–502.Misica-Turner PM, Oback FC, Eichenlaub M, Wells DN,Oback B, 2007: Aggregating embryonic but not somaticnuclear transfer embryos increases cloning efficiency incattle. Biol Reprod 76, 268–278.Morgan HD, Santos F, Green K, Dean W, Reik W, 2005:Epigenetic reprogramming in mammals. Hum Mol Genet14, R47–R58.Nagy A, Gocza E, Diaz Em, Prideaux Vr, Ivanyi E, MarkkulaM, Rossant J, 1990: Embryonic stem cells alone are able tosupport fetal development in the mouse. Development 110,815–821.Nolen LD, Gao SR, Han ZM, Mann MRW, Chung YG, OtteAP, Bartolomei MS, Latham KE, 2005: X chromosomereactivation and regulation in cloned embryos. Dev Biol279, 525–540.Okamoto I, Otte A P, Allis C D, Reinberg D, Heard E, 2004:Epigenetic dynamics of imprinted X inactivation duringearly mouse development. Science 303, 644–649.Okamoto I, Arnaud D, Le Baccon P, Otte AP, Disteche CM,Avner P, Heard E, 2005: Evidence for de novo imprinted X-chromosome inactivation independent of meiotic inactivationin mice. Nature 438, 369–373.Ono Y, Shimozawa N, Ito M, Kono T, 2001: Cloned micefrom fetal fibroblast cells arrested at metaphase by a serialnuclear transfer. Biol Reprod 64, 44–50.Panarace M, Agu¨ero JI, Garrote M, Jauregui G, Segovia A,Canè L, Gutie´rrez J, Marfil M, Rigali F, Pugliese M, YoungS, Lagioia J, Garmil C, Forte Pontes JE, Ereno Junio JC,Mower S, Medina M, 2007: How healthy are clones andtheir progeny: 5 years of field experience. Theriogenology67, 142–151.Peura TT, Lewis IM, Trounson AO., 1998: The effect ofrecipient oocyte volume on nuclear transfer in cattle. MolReprod Dev 50, 185–191.Reik W, Santos F, Mitsuya K, Morgan H, Dean W, 2003:Epigenetic Asymmetry In The Mammalian Zygote AndEarly Embryo: Relationship to lineage commitment? PhilosTrans R Soc Lond B Biol Sci 358, 1403–1409.Renard J P, Maruotti J, Jouneau A, Vignon X, 2007: Nuclearreprogramming and pluripotency of embryonic cells: applicationto the isolation of embryonic stem cells in farmanimals. Theriogenology 68, S196–S205.Rideout WM, Wakayama T, Wutz A, Eggan K, Jackson-Grusby L, Dausman J, Yanagimachi R, Jaenisch R, 2000:Generation of mice from wild-type and targeted ES cells bynuclear cloning. Nat Genet 24, 109–110.Santos F, Zakhartchenko V, Stojkovic M, Peters A, JenuweinT, Wolf E, Reik W, Dean W, 2003: Epigenetic markingcorrelates with developmental potential in cloned bovinepreimplantation embryos. Curr Biol 13, 1116–1121.Schurmann A, Wells DN, Oback B, 2006: Early zygotes aresuitable recipients for bovine somatic nuclear transfer andresult in cloned offspring. Reproduction 132, 839–848.Shi W, Hoeflich A, Flaswinkel H, Stojkovic M, Wolf E,Zakhartchenko V, 2003: Induction of a senescent-likephenotype does not confer the ability of bovine immortalcells to support the development of nuclear transferembryos. Biol Reprod 69, 301–309.Smith SL, Everts RE, Tian XC, Du FL, Sung LY, Rodriguez-Zas SL, Jeong BS, Renard JP, Lewin HA, Yang XZ, 2005:Global gene expression profiles reveal significant nuclearreprogramming by the blastocyst stage after cloning. ProcNatl Acad Sci U S A 102, 17582–17587.Somers J, Smith C, Donnison M, Wells DN, Henderson H,McLeay L, Pfeffer PL, 2006: Gene expression profiling ofindividual bovine nuclear transfer blastocysts. Reproduction131, 1073–1084.Suemizu H, Aiba K, Yoshikawa T, Sharov AA, Shimozawa N,Tamaoki N, Ko MSH, 2003: Expression profiling ofplacentomegaly associated with nuclear transplantation ofmouse es cells. Dev Biol 253, 36–53.Sullivan EJ, Kasinathan S, Kasinathan P, Robl JM, Collas P,2004: Cloned calves from chromatin remodeled in vitro. BiolReprod 70, 146–153.Suteevun T, Parnpai R, Smith SL, Chang CC, MuenthaisongS, Tian XC, 2006: Epigenetic characteristics of cloned and invitro-fertilized swamp buffalo (Bubalus bubalis) embryos.J Anim Sci 84, 2065–2071.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Epigenetics of Foetal Development in Clones 309Takagi N, Sasaki M, 1975: Preferential inactivation of thepaternally derived X chromosome in the extraembryonicmembranes of the mouse. Nature 256, 640–642.Tamashiro KLK, Wakayama T, Akutsu H, Yamazaki Y,Lachey JL, Wortman MD, Seeley RJ, D’Alessio DA,Woods SC, Yanagimachi R, Sakai RR, 2002: Cloned micehave an obese phenotype not transmitted to their offspring.Nat Med 8, 262–267.Wagschal A, Feil R, 2006: Genomic imprinting in the placenta.Cytogenet Genome Res 113, 90–98.Wakayama T, Yanagimachi R, 2001: Mouse cloning withnucleus donor cells of different age and type. Mol ReprodDev 58, 376–383.Wakayama T, Rodriguez I, Perry AC, Yanagimachi R,Mombaerts P, 1999: Mice cloned from embryonic stemcells. Proc Natl Acad Sci U S A 96, 14984–14989.Wakayama S, Jakt ML, Suzuki M, Araki R, Hikichi T,Kishigami S, Ohta H, Van Thuan N, Mizutani E, SakaideY, Senda S, Tanaka S, Okada M, Miyake M, Abe M,Nishikawa SI, Shiota K, Wakayama T, 2006: Equivalencyof nuclear transfer-derived embryonic stem cells to thosederived from fertilized mouse blastocyst. Stem Cells 24,2023–2033.Wang L, Duan E, Sung LY, Jeong BS, Yang X, Tian XC,2005: Generation and characterization of pluripotentstem cells from cloned bovine embryos. Biol Reprod 73,149–155.Wells DN, Laible G, Tucker FC, Miller AL, Oliver JE, XiangT, Forsyth JT, Berg MC, Cockrem K, L’Huillier PJ, TervitHR, Oback B, 2003a: Coordination between donor cell typeand cell cycle stage improves nuclear cloning efficiency incattle. Theriogenology 59, 45–59.Wells DN, Laible G, Tucker FC, Miller AL, Oliver JE, XiangT, Forsyth JT, Berg MC, Cockrem K, L’Huillier PJ, TervitHR, Oback B, 2003b: Coordination between donor cell typeand cell cycle stage improves nuclear cloning efficiency incattle. Theriogenology 59, 45–59.Wilkins JF, Haig D, 2003: What good is genomic imprinting:the function of parent-specific gene expression. Nat RevGenet 4, 359–368.Wilmut I, Schnieke AE, McWhir J, Kind AJ, Campbell KH,1997: Viable offspring derived from fetal and adult mammaliancells. Nature 385, 810–813.Wilmut I, Beaujean N, de Sousa PA, Dinnyes A, King TJ,Paterson LA, Wells DN, Young LE, 2002: Somatic cellnuclear transfer. Nature 419, 583–587.Wrenzycki C, Wells D, Herrmann D, Miller A, Oliver J, TervitR, Niemann H, 2001: Nuclear transfer protocol affectsmessenger RNA expression patterns in cloned bovineblastocysts. Biol Reprod 65, 309–317.Xue F, Tian XC, Du FL, Kubota C, Taneja M, Dinnyes A,Dai YP, Levine H, Pereira LV, Yang XZ, 2002: Aberrantpatterns of X chromosome inactivation in bovine clones.Nat Genet 31, 216–220.Yang L, Chavatte-Palmer P, Kubota C, O’Neill M, HoaglandT, Renard JP, Taneja M, Yang XZ, Tian XC, 2005:Expression of imprinted genes is aberrant in deceasednewborn cloned calves and relatively normal in survivingadult clones. Mol Reprod Dev 71, 431–438.Young LE, Sinclair KD, Wilmut I, 1998: Large offspringsyndrome in cattle and sheep. Rev Reprod 3, 155–163.Zhu J, Telfer EE, Fletcher J, Springbett A, Dobrinsky JR, DeSousa PA, Wilmut I, 2002: Improvement of an electrical activationprotocol for porcine oocytes. Biol Reprod 66, 635–641.Zuccotti M, Garagna S, Redi CA, 2000: Nuclear transfer,genome reprogramming and novel opportunities in celltherapy. J Endocrinol Invest 23, 623–629.Author’s address (for correspondence): A Dinnyes, Genetic ReprogrammingGroup, Agricultural Biotechnology Centre and MolecularAnimal Biotechnology Laboratory, Szent Istvan University, Godollo,H–2100, Hungary. E-mail: dinnyes@abc.huConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 310–316 (2008); doi: 10.1111/j.1439-0531.2008.01179.xISSN 0936-6768A Proposed Role for VEGF Isoforms in Sex-Specific Vasculature Development in theGonadRC Bott, DT Clopton and AS CuppAnimal Science Department, University of Nebraska-Lincoln, Lincoln, NE, USAContentsMany scientists have expended efforts to determine whatregulates development of an indifferent gonad into either atestis or ovary. Expression of Sry and upregulation of Sox9 arefactors that initiate formation of the testis-specific pathway toallow for both sex-specific vasculature and seminiferous cordformation. Migration of mesonephric precursors of peritubularmyoid cells and endothelial cells into the differentiatingtestis is a critical step in formation of both of these structures.Furthermore, these events appear to be initiated downstreamfrom Sry expression. Sertoli cell secretion of growth factorsacts to attract these mesonephric cells. One hypothesis is that agrowth factor specific for these cell linages act in concert tocoordinate migration of both peritubular and endothelial cells.A second hypothesis is that several growth factors stimulatemigration and differentiation of mesonephric ‘stem-like’ cellsto result in migration and differentiation into several differentcell lineages. While the specific mechanism is unclear, severalgrowth factors have been implicated in the initiation ofmesonephric cell migration. This review will focus on theproposed mechanisms of a growth factor, Vascular EndothelialGrowth Factor, and how different angiogenic and inhibitoryisoforms from this single gene may aid in development oftestis-specific vascular development.IntroductionWhy study gonadal differentiation and sex-specificvascular development?Infertility affects 40–70 million couples; of thoseapproximately 50% of the infertility problems areattributed to male-related factors which include: lowsperm count, abnormal spermatogenesis, and reducedandrogen production (Skakkebaek 2004). In the past12 years, the incidence of male infertility cases hasincreased at an alarming rate resulting in the developmentof a new syndrome – testicular dysgenesis syndrome.The reasons for the increase in testicularabnormalities are not understood but may be a resultof genetic or environmental disruption of cell differentiationduring testis morphogenesis (Skakkebaek 2004).Formation of testicular cords, and sex-specific vasculature,are the two morphological hallmarks that distinguisha testis from an ovary. Testis development isinitiated by Sertoli cell differentiation and expression ofSry, causing mesonephric cell migration into the differentiatingtestis to form seminiferous cords. Removal ofthe mesonephros or blockage of mesonephric cellmigration prevents cord formation. Thus, the migrationof mesonephric derived cells, which include pre-endothelialand pre-peritubular cells, are critical to testisdevelopment.What is known about embryonic testis morphogenesis?In the mouse, the Sry gene is expressed in embryos at10.5 days post-coitus (dpc) and ceases after 12.5 dpc.Sry induces expression of genes which cause differentiationof the Sertoli cell lineage from precursorsomatic cells in the coelomic epithelium (Fig. 1) (Cuppand Skinner 2005). So even though Sry function isimportant, expression in the rodent is brief. The Sertolicell is the first cell of the testis to differentiate (Magreand Jost 1991). After differentiation, Sertoli cells beginto proliferate and simultaneously move into the gonadproper, forming aggregates with primordial germ cells.Proliferation of Sertoli cells increases the size of thetestis, and this proliferation appears to be solelydependent on the expression of Sry (Schmahl et al.2000). Without differentiation and proliferation of theSertoli cells, the indifferent gonad would not developinto a testis.After Sertoli cell differentiation, testicular cords andsex-specific vasculature develop to establish adult testismorphology. In the mouse, these events occur during11.5–12 dpc and are complete by 12.5 dpc while in therat these events occur between embryonic day 13.5–14(Martineau et al. 1997). Induction of cord formation isinitiated by cell migration from the adjacent mesonephrosinto the developing testis to surround theprimordial germ and Sertoli cell aggregates (Buehret al. 1993; Merchant-Larios et al. 1993). Mesonephriccell migration is the result of Sry regulated expressionof paracrine growth factors secreted by Sertoli cells(Martineau et al. 1997). Paracrine growth factors act aschemo-attractants inducing mesonephric cell migrationand cord formation (Ricci et al. 1999; Cupp et al. 2000,2002, 2003; Colvin et al. 2001). If mesonephric cellmigration is blocked, or if the mesonephros is removed,no cords will form. The population of cells that migratefrom the mesonephros is proposed to be pre-endothelialand ⁄ or pre-peritubular cells (Buehr et al. 1993; Merchant-Larioset al. 1993). Therefore, in addition tosurrounding the Sertoli-primordial germ cell aggregatesto form testis cords, the migrating mesonephric cellsmay also initiate the formation of vasculature withinthe developing testis. Platelet-derived growth factor(Uzumcu et al. 2002a) and VEGF (Bott et al. 2006)have been demonstrated to be necessary for cordformation; however, only VEGF has been demonstratedto be critical for both vascular development andcord formation.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

VEGFA Isoforms as Regulators of Testis Vasculature 311Sex determinationTestis morphogenesis# Tail somites (ts)Da yspostcoitum(dpc)8 10–15 16–18 19–23 24–3010. 5 10.7–11.2 11.3–11.5 11.6–11.9 12–12.5Fig. 1. Sry-regulated events occurringduring sex determination andtestis morphogenesis using dayspost-coitum and tail somites asdevelopmental markers (modifiedfrom Cupp and Skinner 2005)MesonephrosGenital RidgeSRY mRN APrim ordialGe rm CellSertoli celldifferentiationSertoli cellproliferationMesonephric ce llmigrationCord formatio nWhat is known about formation of sex-specific vasculardevelopment?Very little is known about what regulates sex specificvascular development in the gonads. Pre-endothelialcells in the indifferent gonad express markers for bothveins (EphB4) and arteries (ephrin B2); however thischanges as each gonad develops sex-specific structures.In the testis, endothelial cells migrate from the mesonephrosand form vascular networks with arterialmarkers between cords and the coelomic vessel. Thecoelomic vessel is the major testicular artery that isformed when mesonephric endothelial cells migratefrom the mesonephros into the testis and is unique tothe testis. As endothelial cells differentiate a greaterpercentage develop into arteries and contribute to thecoleomic vessel than those which contribute to veinswithin the testis. The development of arteries may benecessary to aid movement of testosterone to thedeveloping male reproductive tract which allows for itsmaintenance and differentiation (Brennan et al. 2002).In contrast, the ovary develops similar amounts ofarterial and venous networks from endothelial cells.The origin of vasculature is different in the testis andovary with vascular networks developing throughangiogenesis (branching from existing vasculature inthe mesonephros) in the testis and through vasculogenesis(de novo or neovascularization) in the ovary(Brennan et al. 2002). It is likely that the testis andovary develop vascular patterns from different endothelialcell origins due to expression of sex-specifictranscription factors. Furthermore, sex-specific transcriptionfactors stimulate different growth factors tonot only initate blood vessel formation but organspecificsupport structures such as seminiferous cordsor oogonial cysts.Experiments using knockout mice demonstrate thatdisruption of vascular development also causes sexreversaland abnormalities in germ cell development.Mice that are null for Wnt4 (Jeays-Ward et al. 2003)and Follistatin (Fst) (Yao et al. 2004) have sex-reversedXX gonads with ectopic expression of a coelomic vesseland vasculature surrounding seminiferous cord-likestructures. The Wnt4 gene regulates expression of Fstso it is not surprising that mice null for both of thesegenes have the same phenotype. Follistatin is onlyexpressed in the XX gonad and in general acts to inhibitactivins. Therefore, there is potential for activins to beinvolved in formation of the coelomic vessel. In contrastInhibin beta b (Inhbb) XY null mice do not formcoelomic vessels 50% of the time. Thus, Inhbb null XYgonads are sometimes sex-reversed to XX vascularphenotypes. Activin A is composed of two inhibin betaa, Inhba, subunits while Activin B is composed of twoInhbb subunits. In the ovary, only Inhbb is expressedand not Inhba. Furthermore, expression of Inhbb isfourfold higher in the testis compared to the ovary at12.5 dpc. Mating of Wnt4 and Inhbb null mice recapitulatethe normal phenotype in XY gonads to form acoelomic vessel. Yet, matings between Inhbb and Fstnull mice did not recapitulate the XY vascular phenotype,thus Wnt4 suppression of Inhbb appears to beindependent of Fst inhibitory actions on activins (Jeays-Ward et al. 2003; Yao et al. 2006).During normal testis development, Sry antagonizesexpression of Wnt4 to suppress Fst and relieve inhibitionof Inhbb (Fig. 2). We can only speculate that Inhbbtranscriptional regulation is due to expression of Sox-9,or other growth factors downstream of Sry expression.Inhbb stimulates endothelial cell migration and formationof a coelomic vessel. In XX gonads with no Wnt4repression, Fst expression is increased and Inhbbexpression is down regulated to prevent endothelial cellmigration and formation of ovarian-specific vascularpatterns (Yao et al. 2004).AMHXYSrySox9FGF-9?InhbbActivin ABMPsTestis vascular patternsXXWnt4FstOvarian vascular patternsFig. 2. Genes demonstrated to be involved in sex-specific vascularpatterns through development of null mice or treatment of indifferentgonads in organ cultureÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

312 RC Bott, DT Clopton and AS CuppIn addition to the potential actions of Activin B, XXgonads treated with bone morphogenic proteins(BMP), anti-mullerian hormone (AMH) and ActivinA in vitro developed a sex-reversed phenotype withformation of an ectopic coelomic vessel; however,AMH deficient mice did not develop this same phenotype(Ross et al. 2003; Yao et al. 2004). Bone morphogenicprotein, AMH, Activin A and Inhbb aremembers of the transforming growth factor-beta(TGF-b) superfamily and directly regulate cell functionthrough inhibition or activation of the SMAD signaltransduction pathway. In the zebrafish, SMAD responseelements in the promoter region of Vegfa genealso appear to regulate expression of different isoforms(He and Chen 2005). Thus, the Vegfa gene may be adownstream gene regulated by members of the TGFfamily to regulate sex-specific vasculature in the developinggonad (Fig. 2).What is known about the vascular endothelial growthfactor family?The VEGF family is composed of five ligands: VEGF(VEGF-A), VEGF-B, VEGF-C, VEGF-D and placentagrowth factor. VEGF (VEGF-A) is the bestcharacterized and most potent VEGF molecule.VEGF works through both Fms-like tyrosine kinase1 (FLT1) and Kinase domain region receptor (KDR),to elicit its effects on endothelial cell migration,differentiation, proliferation and survival and apoptosis.The primary receptor involved in neovascularizationof organs with VEGF is KDR which contains atyrosine kinase signal transduction domain (Waltenbergeret al. 1994). There are conflicting reports aboutthe role of FLT1 in regulation of VEGF’s actions. InFlt1 knockouts, the mice die due to vasculatureovergrowth suggesting that FLT1 acts as a decoyreceptor to inhibit VEGF-dependent vascular development.Furthermore, FLT1 can also inhibit VEGF’sactions by dimerizing with KDR to inhibit signaltransduction. Thus, FLT1 appears to act as a negativeregulator of VEGF’s actions on endothelial cellmigration, survival and proliferation (Ferrara 2000;Olsson et al. 2006).Both mRNA and protein for VEGF and KDR arepresent during testis morphogenesis and are expressedin cells, like the Sertoli cell, that direct testis development.However, Flt1 is not expressed until after cordformation in the testis (Bott et al. 2006). The absenceof FLT1 during mesonephric endothelial cell migrationand establishment of vasculature in the testis mayallow for endothelial cell migration through VEGF. Incontrast, the ovary, which does not have mesonephriccell migration, expresses both Flt1 and Kdr duringdevelopment (Pohlmann et al. 2004). Thus, VEGFangiogenic isoforms may be bound by FLT1 receptorin the ovary inhibiting the ability of mesonephricendothelial cells to be recruited into the developingovary; thus providing a different endothelial cell originfor ovarian vasculature.Vascular endothelial growth factor is transcribedfrom a single gene that has eight exons and is alternativelyspliced into different isoforms each containing aVEGF 205VEGF 188VEGF 164VEGF 144VEGF 120VEGF 110VEGF 164bExons 1–5 6A 6B 7 8aExons 1–5 6A 7 8aExons 1–5 7 8aExons 1–5 6A 8aExons 1–5Exons 1–37Bdifferent number of amino acids. The most commonangiogenic isoforms are VEGF205, 188, 164, 144 and120 (in humans the angiogenic isoforms are one aminoacid longer; Fig. 3). The predominant isoforms expressedin tissues are VEGF188, 164 and 120 (Veikkolaand Alitalo 1999).In 2002, an additional isoform, VEGF165b, wasidentified which contained part of the 3’ UTR that isnow determined to be exon 8b (Fig. 3). Furthermore,recent studies have demonstrated that the humanVEGF165b isoform is anti-angiogenic in function andinhibits signal transduction through KDR (Bates et al.2002; Woolard et al. 2004). Thus, this isoform isinhibitory to the actions of VEGF. Recently severalother inhibitory isoforms have been identified in thehuman and we have cloned several in both the rat andbovine (VEGF165b accession number EU040284;VEGF189b accession number EU040285). Therefore,it appears that for every angiogenic isoform there is asister inhibitory isoform that is formed when exon 8ais replaced with exon 8b. (In rodents and cattle thesister inhibitory isoform is one amino acid longer.)The primary functions of angiogenic VEGF isoformsare to induce endothelial cell migration, survival,proliferation and differentiation to initiate angiogenesiswithin developing organs and tumors. The inhibitoryVEGF isoforms appear to modulate thesefunctions.In addition to the multiple VEGF isoforms, two coreceptors,neuropilin1 (NRP1) and neuropilin2 (NRP2),that bind to specific VEGF isoforms (i.e. VEGF164,VEGF188) have been identified. The predominantfunction of NRP1 appears to be stabilization ofVEGF164 binding to KDR which augments signaltransduction. The two co-receptors can also stabilizebinding of VEGF isoforms to FLT1. In fact, FLT1 mayregulate VEGF actions by tying up NRP boundisoforms. While both NRP co-receptors can bind toFLT1 it appears that only NRP1 can stabilize signaltransduction through KDR. Interestingly, KDR, butnot FLT1 is expressed during seminiferous cord formation(Bott et al. 2006) thus the ability of NRP1 toenhance signal transduction upon binding of VEGFisoforms to KDR may be critical during testis differentiation.8a8aExons 1–5 7 8bFig. 3. Exons in VEGF gene that compose each VEGF isoformÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

VEGFA Isoforms as Regulators of Testis Vasculature 313What Signal Transduction Pathways DoesVEGF Activate Through KDR and Which areCritical for Testis Cord Formation?Vascular endothelial growth factor induces the activationof several signal transduction pathways uponbinding to KDR. Of these pathways, Phospholipase Cgamma activates protein kinase C causing activation ofRas or Raf and ultimately MAPK and ERK phosphorylation.The result of ERK phosphorylation is endothelialcell proliferation (Shu et al. 2002). The PI3-kinasepathway has been demonstrated to affect migration andpermeability through Rac and permeability and survivalthrough Akt (Gerber et al. 1998). The p38MAPK andFAK pathways are also postulated to affect actinreorganization and focal adhesion turnover to influencemigration (Goligorsky et al. 1999). Several of theseVEGF signal transduction pathways may interact toinitiate endothelial cell differentiation, migration, proliferationand survival. Inhibition of the MAPK pathwayaffects cord formation (Uzumcu et al. 2002b) intestis organ cultures; however, PI3-kinase inhibits bothseminiferous cord formation (Uzumcu et al. 2002b; Bottet al. 2006) and vascular development (Bott et al. 2006).Thus, both of these sex-specific events, vascular developmentand cord formation may be regulated bymultiple growth factors through a similar signal transductionpathway.What specific roles do the VEGF isoforms have inendothelial cell migration?Endothelial cell recruitment and chemoattractionrequires several different VEGF isoforms to establish aVEGF concentration gradient. At least three of theangiogenic VEGF isoforms are associated with specificendothelial cell functions (Veikkola and Alitalo 1999).VEGF120, a highly diffusible isoform, recruits endothelialcells into tissue to initiate the development ofvasculogenesis while VEGF164 can recruit endothelialcells and establish large blood vessels. VEGF188, due toits heparin binding domain, acts locally to inducebranching of large blood vessels into smaller capillariesand the development of capillary beds (Grunstein et al.2000) (Fig. 3). Furthermore, these three isoforms arethought to establish a VEGF chemoattractant gradientto allow for endothelial cell migration. In the testis,VEGF120, 188 and 164 are present during seminiferouscord formation and can develop a viable chemoattractantgradient (Bott et al. 2006). In contrast, in the ovaryonly VEGF164 and 120 are present (Pohlmann et al.2004). The absence of the larger VEGF isoforms toanchor the chemoattractant gradient inhibits endothelialcell migration and weakens the attraction to VEGFsecreting cells (Ruhrberg 2003). Therefore, in the ovarya viable chemoattractant gradient may not be presentcontributing to the lack of mesonephric cell migration.VEGF165b is anti-angiogenic and appears to inhibitthe actions of VEGF164 and to a lesser extent otherVEGF isoforms. The inhibitory isoform, VEGF165b isexpressed in all tissues but is down-regulated in prostatetumors (Woolard et al. 2004), renal tumors (Bates et al.2002) and in disorders of the eye such as maculardegeneration (Perrin et al. 2005). The mechanism ofdifferential VEGF isoform expression is not known butis likely related to transcriptional and ⁄ or post-transcriptionalregulation of the VEGF gene which results indifferent ratios of angiogenic to inhibitory VEGFisoforms expressed.One role of the inhibitory VEGF isoforms may be toinhibit the chemoattractant gradient and inhibit endothelialcell migration (Bates et al. 2002; Woolard et al.2004). Our laboratory has subcloned and sequenced therat VEGF165b and determined that there is a sexspecificexpression of VEGF165b (fourfold higher inovary than testis; unpublished data) during the timeendothelial cell migration occurs. Therefore, theVEGF165b isoform may be expressed in the ovary toprevent establishment of a chemoattractant gradientand migration of endothelial cells from the mesonephrosinto the developing ovary.How are specific VEGF isoform actions regulated?Neuropilin1, a VEGFA co-receptor, appears to modulateVEGFA signal transduction (Klagsbrun and Eichmann2005). Neuropilins were first identified asreceptors for axon guidance forming complexes withthe plexin-A subfamily to modulate semaphorin signals(Takahashi et al. 1998). However, recent studies haveidentified a role for neuropilins in VEGF-dependentangiogenic processes. Specifically, NRP1 knockout micehave vasculature, heart and neural defects and die atE10.5–12.5 suggesting that NRP1 is necessary fornormal vascular development (Kawasaki et al. 1999).If NRP1 binds to semaphorin instead of VEGF thenendothelial cell migration is inhibited (Miao et al. 1999).Likewise, if NRP1 ⁄ VEGF complexes bind to FLT1 thisalso may inhibit endothelial cell migration and thedevelopment of vasculature.The predominant function of NRP1 appears to bestabilization of VEGFA164 binding to KDR whichaugments signal transduction. The two co-receptors canalso stabilize binding of angiogenic VEGFA isoforms toFLT1. Infact, FLT1 may regulate angiogenic VEGFactions by enticing NRP bound isoforms to bind to thisdecoy receptor. While both NRP co-receptors can bindto FLT1 it appears that only NRP1 can stabilize signaltransduction through KDR.Initially, VEGFA isoforms containing exon 7 or aheparin binding domain (also in exon 6) were thought tobind to NRP1. However, there have been recent studiesthat suggest NRPs (in addition to binding to VEGF164and 188 that contain heparin binding domains) alsobind to VEGFA120 (which does not have a heparinbinding domain). Although they bind VEGFA120, theinteraction does not allow for the bridging of the tworeceptors (NRP1 and KDR) that occur with VEGF164or 188. Therefore, the interaction with VEGF120 doesnot appear to stabilize and amplify signal transductionwhich occurs with VEGF164 (Pan et al. 2007).Exon 8a also appears to be important in binding ofNRPs to VEGF angiogenic isoforms (Jia et al. 2006). Ithas been demonstrated that inhibitory VEGF isoformssuch as 165b cannot bind NRPs since exon 8b replacesÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

314 RC Bott, DT Clopton and AS Cupp8a. Therefore, NRPs can convey angiogenic vs inhibitoryisoform specificity to cells and if NRPs areeliminated from cells then the actions of angiogenicisoforms may be impaired.What regulates VEGF gene expression?Hypoxia (Schweda et al. 2000), oestradiol (Muelleret al. 2000), progesterone (Mueller et al. 2003), andfactors signalling through the SMAD signal transductionpathway (He and Chen 2005) have all beenimplicated in increased or decreased expression ofVEGF isoforms. Results from in vitro studies (ActivinA ⁄ B, BMPs, AMH) (Yao et al. 2004) and null mice(Inhbb, Follistatin and Wnt4) (Yao et al. 2004, 2006)suggest that TGF-b family members or proteins thatinhibit their actions are involved in modulation ofvascular development in the developing gonad. Interestingly,in the zebrafish SMAD response elementsregulate VEGF gene expression (He and Chen 2005).Deletion of either SMAD1 or SMAD5 response elementson the zebrafish VEGF promoter resulted ineither inhibition or enhanced transcription of the VEGFgene (He and Chen 2005). Within the murine VEGFpromoter there are multiple consensus sequences forSMAD response elements (AS Cupp, unpublished data).Therefore, it is highly possible that growth factorssignalling through the SMAD signal transduction pathwaysalter the expression of VEGF and VEGF-signaltransduction mediators or potentially affect post-translationalmodification of VEGF to allow for differentratios of angiogenic vs inhibitory isoforms to beexpressed (Fig. 2).What Mouse Models Can Be Used toUnderstand the Function of VEGF DuringTestis Morphogenesis?Null mutations of VEGF are embryonic lethal with eventhe loss of one allele resulting in embryonic death at E11dpc (Carmeliet et al. 1996; Ferrara et al. 1996). Similarly,homozygous knockouts for either receptor, FLT1or KDR, die on 9–10 dpc (Millauer et al. 1994; Hiratsukaet al. 1998). Thus, these ubiquitous knockoutmodels prevent us from determining the role of VEGFin testis morphogenesis and ⁄ or vascular patterning ofthe gonad. Overexpression of VEGF is equally detrimentalto fertility and embryonic viability (Miquerolet al. 2000; Huminiecki et al. 2001).Many conditional mutants have been developedwhich knock out VEGF ubiquitously later in development(in the neonate; Gerber et al. 1999) or in specifictissues (i.e. nerves in skin (Mukouyama et al. 2005). Inthese studies VEGF was determined to be required forproliferation and survival of endothelial cells. Thedeletion of VEGF in newborn mice in every cell of thebody resulted in smaller mice that did not survive toadulthood. However, if VEGF was knocked out later inthe fully developed animal there was less of an effect ongrowth.Overexpression of VEGF in transgenic mice results ininfertility (Korpelainen et al. 1998). Furthermore, overexpressionof human165b through the MMLV promoterresults in disruption of mammary gland development,and death postnatally in offspring from over-expressingfemales due to malnutrition (Qiu et al. 2007). However,until a testis-specific knockout mouse is developed it isunclear how removal of Sertoli-cell secretion of VEGFwill affect testis development and function. Thus, furtherexperiments need to be conducted in vivo to demonstratethe role of VEGF on sex-specific vascular development,seminiferous cord formation and testis function.ConclusionApproximately 2 million couples seek treatment forinfertility every year and less than half find successfultreatments (Carlsen et al. 2005). Infertility problems inat least half of these couples are a result of male-relatedfactors that are created by testicular dysgenesis. Manyof the problems associated with testicular dysgenesis areproposed to involve a disruption in embryonic differentiationof cells within the indifferent gonad resulting inaltered testicular development. Elucidating the factorsinvolved in sex-specific vascular development will allowfor a better understanding of how transcription factorscoordinate regulation of growth factors to result in atestis-specific vascular system. Furthermore, delineatingthe interaction of VEGF angiogenic and inhibitoryisoforms in sex-specific vascular development promisesto be an interesting piece in the puzzle of gonadaldevelopment.AcknowledgementThis research was funded by NIH grants (R03HD41546-01;R03HD045350-01).ReferencesBates D, Cui T-G, Doughty J, Winkler M, Sugiono M, ShieldsJ, Peat D, Gillatt D, Harper S., 2002: VEGF 165 b, aninhibitory splice variant of vascular endothelial growthfactor, is down-regulated in renal cell carcinoma. CancerRes 62, 4123–4131.Bott RC, McFee RM, Clopton DT, Toombs C, Cupp AS,2006: Vascular endothelial growth factor and kinase domainregion receptor are involved in both seminiferous cordformation and vascular development during testis morphogenesisin the rat. Biol Reprod 75, 56–67.Brennan J, Karl J, Capel B, 2002: Divergent vascularmechanisms downstream of Sry establish the arterial systemin the XY gonad. Dev Biol 244, 418–428.Buehr M, Gu S, McLaren A, 1993: Mesonephric contributionto testis differentiation in the fetal mouse. Development 117,273–281.Carlsen E, Swan SH, Petersen JH, Skakkebaek NE, 2005:Longitudinal changes in semen parameters in young Danishmen from the Copenhagen area. Hum Reprod 20, 942–949.Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L,Gertsenstein M, Fahrig M, Vandenhoeck A, HarpalK, Eberhardt C, Declercq C, Pawling J, Moons L, CollenD, Risau W, Nagy A, 1996: Abnormal blood vesseldevelopment and lethality in embryos lacking a singleVEGF allele. Nature 380, 435–439.Colvin JS, Green RP, Schmahl J, Capel B, Ornitz DM, 2001:Male-to-female sex reversal in mice lacking fibroblastgrowth factor 9. Cell 104, 875–889.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

VEGFA Isoforms as Regulators of Testis Vasculature 315Cupp A, Skinner M, 2005: Embryonic sertoli cell differentiation.in: Griswold M, Skinner M (eds), Sertoli Cell Biology,Vol. 1. Elsevier Academic Press, San Diego, pp. 43–70.Cupp AS, Kim GH, Skinner MK, 2000: Expression and actionof neurotropin-3 and nerve growth factor in embryonic andearly postnatal rat testis development. Biol Reprod 63,1617–1628.Cupp AS, Tessarollo L, Skinner MK, 2002: Testis developmentalphenotypes in neurotropin receptor trkA and trkCnull mutations: role in formation of seminiferous cords andgerm cell survival. Biol Reprod 66, 1838–1845.Cupp AS, Uzumcu M, Skinner MK, 2003: Chemotactic role ofneurotropin 3 in the embryonic testis that facilitates malesex determination. Biol Reprod 68, 2033–2037.Ferrara N, 2000: VEGF: an update on biological andtherapeutic aspects. Curr Opin Biotechnol 11, 617–624.Ferrara N, Carver-Moore K, Chen H, Dowd M, Lu L, O’SheaKS, Powell-Braxton L, Hillan KJ, Moore MW, 1996:Heterozygous embryonic lethality induced by targetedinactivation of the VEGF gene. Nature 380, 439–442.Gerber HP, McMurtrey A, Kowalski J, Yan M, Keyt BA,Dixit V, Ferrara N, 1998: Vascular endothelial growthfactor regulates endothelial cell survival through the phosphatidylinositol3’-kinase ⁄ Akt signal transduction pathway.Requirement for Flk-1 ⁄ KDR activation. J Biol Chem 273,30336–30343.Gerber H.-P, Hilan K, Ryan A, Kowalski J, Keller G.A,Rangell L, Wright B, Radtke F, Aguet M, Ferrara N, 1999:VEGF is required for growth and survival in neonatal mice.Development 126, 1149–1159.Goligorsky MS, Abedi H, Noiri E, Takhtajan A, Lense S,Romanov V, Zachary I, 1999: Nitric oxide modulation offocal adhesions in endothelial cells. Am J Physiol 276,C1271–C1281.Grunstein J, Masbad JJ, Hickey R, Giordano F, Johnson RS,2000: Isoforms of vascular endothelial growth factor act in acoordinate fashion to recruit and expand tumor vasculature.Mol Cell Biol 20, 7282–7291.He C, Chen X, 2005: Transcription regulation of the vegf geneby the BMP ⁄ Smad pathway in the angioblast of zebrafishembryos. Biochem Biophys Res Commun 329, 324–330.Hiratsuka S, Minowa O, Kuno J, Noda T, Shibuya M, 1998:Flt-1 lacking the tyrosine kinase domain is sufficient fornormal development and angiogenesis in mice. Proc NatlAcad Sci U S A 95, 9349–9354.Huminiecki L, Chan H, Lui S, Poulsom R, Stamp G, Harris A,Bicknell R, 2001: Vascular endothelial growth factor transgenicmice exhibit reduced male fertility and placentalrejection. Mol Human Reprod 7, 255–264.Jeays-Ward K, Hoyle C, Brennan J, Dandonneau M, AlldusG, Capel B, Swain A, 2003: Endothelial and steroidogeniccell migration are regulated by WNT4 in the developingmammalian gonad. Development 130, 3663–3670.Jia H, Bagherzadeh A, Hartzoulakis B, Jarvis A, Lohr M,Shaikh S, Aqil R, Cheng L, Tickner M, Esposito D, HarrisR, Driscoll PC, Selwood DL, Zachary IC, 2006: Characterizationof a bicyclic peptide neuropilin-1 (NP-1) antagonist(EG3287) reveals importance of vascular endothelial growthfactor exon 8 for NP-1 binding and role of NP-1 in KDRsignaling. J Biol Chem 281, 13493–13502.Kawasaki T, Kitsukawa T, Bekku Y, Matsuda Y, Sanbo M,Yagi T, Fujisawa H, 1999: A requirement for neuropilin-1 inembryonic vessel formation. Development 126, 4895–4902.Klagsbrun M, Eichmann A, 2005: A role for axon guidancereceptors and ligands in blood vessel development andtumor angiogenesis. Cytokine Growth Factor Rev 16, 535–548.Korpelainen E, Karkkainen M, Tenhunen A, Lakso M,Rauvala H, Vierula M, Parvinen M, Alitalo K, 1998:Overexpression of VEGF in testis and epididymis causesinfertility in transgenic mice: evidence for nonendothelialtargets for VEGF. J Cell Biol 143, 1705–1712.Magre S, Jost A, 1991: Sertoli cells and testicular differentiationin the rat fetus. J Electron Microsc Tech 19, 172–188.Martineau J, Nordqvist K, Tilmann C, Lovell-Badge R, CapelB, 1997: Male-specific cell migration into the developinggonad. Curr Biol 7, 958–968.Merchant-Larios H, Moreno-Mendoza N, Buehr M, 1993:The role of the mesonephros in cell differentiation andmorphogenesis of the mouse fetal testis. Int J Dev Biol 37,407–415.Miao HQ, Soker S, Feiner L, Alonso JL, Raper JA, KlagsbrunM, 1999: Neuropilin-1 mediates collapsin-1 ⁄ semaphorin IIIinhibition of endothelial cell motility: functional competitionof collapsin-1 and vascular endothelial growth factor-165. J Cell Biol 146, 233–242.Millauer B, Shauver L, Plate K, Risau W, Ullrich A, 1994:Glioblastoma growth inhibited in vivo by a dominantnegativeFlk-1 mutant. Nature 367, 576–579.Miquerol L, Langille BL, Nagy A, 2000: Embryonic developmentis disrupted by modest increases in vascular endothelialgrowth factor gene expression. Development 127, 3941–3946.Mueller MD, Vigne JL, Minchenko A, Lebovic DI, LeitmanDC, Taylor RN, 2000: Regulation of vascular endothelialgrowth factor (VEGF) gene transcription by estrogenreceptors alpha and beta. Proc Natl Acad Sci U S A 97,10972–10977.Mueller MD, Vigne JL, Pritts EA, Chao V, Dreher E, TaylorRN, 2003: Progestins activate vascular endothelial growthfactor gene transcription in endometrial adenocarcinomacells. Fertil Steril 79, 386–392.Mukouyama YS, Gerber HP, Ferrara N, Gu C, Anderson DJ,2005: Peripheral nerve-derived VEGF promotes arterialdifferentiation via neuropilin 1-mediated positive feedback.Development 132, 941–952.Olsson AK, Dimberg A, Kreuger J, Claesson-Welsh L, 2006:VEGF receptor signalling – in control of vascular function.Nat Rev 7, 359–371.Pan Q, Chathery Y, Wu Y, Rathore N, Tong RK, Peale F,Bagri A, Tessier-Lavigne M, Koch AW, Watts R. J, 2007:Neuropilin-1 binds to VEGF121 and regulates endothelialcell migration and sprouting. J Biol Chem 282, 24049–24056.Perrin RM, Konopatskaya O, Qiu Y, Harper S, Bates DO,Churchill AJ, 2005: Diabetic retinopathy is associated with aswitch in splicing from anti- to pro-angiogenic isoforms ofvascular endothelial growth factor. Diabetologia 48, 2422–2427.Pohlmann R, Clopton D, McFee R, Rozell T, Cupp A, 2004:Inhibition of Vascular Endothelial Growth Factor (VEGF)Signaling blocks follicle progression in the perinatal ratovary. J Anim Sci 82, 233.Qiu YBH, Weeraperuma S, Wratting D, Murphy D, Neal CR,Bates DO, Harper SJ, 2008: Mammary alveolar developmentduring lactation is inhibited by the endogenousantiangiogenic growth factor isoform, VEGF165b. FASEBJ 22, 1104–1112.Ricci G, Catizone A, Innocenzi A, Galdieri M, 1999: Hepatocytegrowth factor (HGF) receptor expression and role ofHGF during embryonic mouse testis development. Dev Biol216, 340–347.Ross AJ, Tilman C, Yao H, MacLaughlin D, Capel B, 2003:AMH induces mesonephric cell migration in XX gonads.Mol Cell Endocrinol 211, 1–7.Ruhrberg C, 2003: Growing and shaping the vascular tree:multiple roles for VEGF. BioEssays 25, 1052–1060.Schmahl J, Eicher EM, Washburn LL, Capel B, 2000: Sryinduces cell proliferation in the mouse gonad. Development127, 65–73.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

316 RC Bott, DT Clopton and AS CuppSchweda F, Blumberg FC, Schweda A, Nabel C, Holmer SR,Riegger GA, Pfeifer M, Kramer BK, 2000: Effects of chronichypoxia on renal PDGF-A, PDGF-B, and VEGF geneexpression in rats. Nephron 86, 161–166.Shu X, Wu W, Mosteller RD, Broek D, 2002: Sphingosinekinase mediates vascular endothelial growth factor–inducedactivation of ras and mitogen–activated protein kinases.Mol Cell Biol 22, 7758–7768.Skakkebaek NE, 2004: Testicular dysgenesis syndrome: newepidemiological evidence. Int J Androl 27, 189–191.Takahashi T, Nakamura F, Jin Z, Kalb RG, Strittmatter SM,1998: Semaphorins A and E act as antagonists of neuropilin-1 and agonists of neuropilin-2 receptors. Nat Neurosci 1,487–493.Uzumcu M, Dirks KA, Skinner MK, 2002a: Inhibition ofplatelet-derived growth factor actions in the embryonic testisinfluences normal cord development and morphology. BiolReprod 66, 745–753.Uzumcu M, Westfall SD, Dirks KA, Skinner MK, 2002b:Embryonic testis cord formation and mesonephric cellmigration requires the phosphotidylinositol 3-kinase signalingpathway. Biol Reprod 67, 1927–1935.Veikkola T, Alitalo K, 1999: VEGF’s receptors and angiogenesis.Semin Cancer Biol 9, 211–220.Waltenberger J, Claesson-Welsh L, Siegbahn A, Shibuya M,Heldin CH, 1994: Different signal transduction properties ofKDR and Flt1, two receptors for vascular endothelialgrowth factor. J Biol Chem 269, 26988–26995.Woolard J, Wang W-Y, Bevan H, Qui Y, MorbidelliL, Pritchard-Jones R, Cui T-G, Sugiono M, WaineE, Perrin R, Foster R, Digby-Bell J, Shields J,Whittles C, Muchens R, Gillatt D, Ziche M,Harper S, Bates D, 2004: VEGF 165 b, an inhibitoryvascular endothelial growth factor splice variant:mechanism of action, in vivo effect on angiogenesisand endogenous protein expression. Cancer Res 64,7822–7835.Yao HH, Matzuk MM, Jorgez CJ, Menke DB, Page DC,Swain A, Capel B, 2004: Follistatin operates downstream ofWnt4 in mammalian ovary organogenesis. Dev Dyn 230,210–215.Yao HH, Aardema J, Holthusen K, 2006: Sexually dimorphicregulation of inhibin beta B in establishing gonadal vasculaturein mice. Biol Reprod 74, 978–983.Author’s address (for correspondence): AS Cupp, A224i AnimalScience Department, 38th and Fair Street, University of Nebraska-Lincoln, Lincoln, NE 68583-0908, USA. E-mail: acupp2@unl.eduConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 317–323 (2008); doi: 10.1111/j.1439-0531.2008.01180.xISSN 0936-6768Comparative Aspects of the Endotoxin- and Cytokine-Induced Endocrine CascadeInfluencing Neuroendocrine Control of Growth and Reproduction in Farm AnimalsBK Whitlock 1 , JA Daniel 2 , RR Wilborn 3 , TH Elsasser 4 , JA Carroll 5 and JL Sartin 11 Department of Anatomy, Physiology and Pharmacology, College of Veterinary Medicine, Auburn University, Auburn, AL, USA; 2 Department ofAnimal Science, Berry College, Mt. Berry, GA, USA; 3 Department of Clinical Sciences, College of Veterinary Medicine, Auburn University, Auburn,AL, USA; 4 Growth Biology Laboratory, USDA ⁄ ARS, Beltsville, MD, USA; 5 Livestock Issues Research Unit, USDA ⁄ ARS, Lubbock, TX, USAContentsDisease in animals is a well-known inhibitor of growth andreproduction. Earlier studies were initiated to determine theeffects of endotoxin on pituitary hormone secretion. Thesestudies found that in sheep, growth hormone (GH) concentrationwas elevated, whereas insulin-like growth factor-I(IGF-I) was inhibited, as was luteinizing hormone (LH).Examination of the site of action of endotoxin in sheepdetermined that somatotropes expressed the endotoxin receptor(CD14) and that both endotoxin and interleukin-Ibactivated GH secretion directly from the pituitary. In the faceof elevated GH, there is a reduction of IGF-I in all speciesexamined. As GH cannot activate IGF-I release duringdisease, there appears to be a downregulation of GH signallingat the liver, perhaps related to altered nitration of Janus kinase(JAK). In contrast to GH downregulation, LH release isinhibited at the level of the hypothalamus. New insights havebeen gained in determining the mechanisms by which diseaseperturbs growth and reproduction, particularly with regard tonitration of critical control pathways, with this perhaps servingas a novel mechanism central to lipopolysaccharide suppressionof all signalling pathways. This pathway-based analysis iscritical to the developing novel strategies to reverse thedetrimental effect of disease on animal production.IntroductionGram-negative bacterial infections such as Escherichiacoli (E. coli), Salmonella, Pseudomonas and Proteus tendto occur frequently in farm animals. Common portals ofentry for these Gram-negative (endotoxin-containing)bacteria include, but are not limited to, the gastrointestinalsystem, the reproductive tract (especially uteri ofpostpartum animals), and the mammary gland. Infectionsof these systems or organs result in some of themost common and costly diseases of production ⁄ farmanimals such as enteritis, endometritis ⁄ metritis andmastitis. Regardless of the chronicity, infections in farmanimals with bacteria containing endotoxin, and thecytokine and endocrine changes that result, inevitablyimpact growth, metabolism and reproduction in anegative manner.Acute diarrhoea, often caused by Gram-negativebacteria such as enterotoxigenic E. coli and Salmonella,is a common disease in newborn calves and accounts formore than 50% of pre-weaning deaths in intensivelyraised calves (USDA 1996). However, pre-weaningmortality of farm animals may be just the beginning ofeconomic losses secondary to enteritis caused by Gramnegativebacteria. Other consequences of neonataldiarrhoea are as follows: (1) greater morbidity secondaryto increased susceptibility to other pathogens, (2)impaired growth rates due to reduced food intake andmetabolic disturbances and (3) delayed puberty anddecreased reproductive productivity (production ofoffspring and milk) as a result of poor growth ratesand inadequate energy stores.The outcomes of bacterial infection associated withthe postpartum uterus include puerperal metritis, clinicalendometritis, pyometra and subclinical endometritis(Sheldon et al. 2006). Arcanobacterium pyogenes(A. pyogenes), E. coli and other Gram-negative bacteria,namely Fusobacterium necrophorum (F. necrophorum)and Bacteroides spp., are predominant in the uterus ofclinically diseased animals (Hirvonen et al. 1999). Thesecommon forms of reproductive tract diseases in farmanimals (especially dairy cows) may delay the completeregeneration of endometrium, disrupt the resumption ofcyclic ovarian function resulting in postponement of thefirst insemination, increase the numbers of inseminationsper conception and thus prolong the calvinginterval and decrease the calving rate (Hussain andDaniel 1991). It is clear that uterine infections andconsequential diseases have detrimental effects onreproductive performance of dairy cows. As mostclinical and reproductive consequences are attributedto the presence of A. pyogenes in combination withorganisms like E. coli and other Gram-negative bacteria,a better understanding in pathogenesis and the mechanismsinvolved is of great practical and economicimportance.Mastitis is one of the major bacterial diseases inpostpartum farm animals, especially dairy cows. In theearly weeks of lactation, Gram-negative bacteria may bethe predominant mastitis pathogen. Clinical cases ofGram-negative mastitis load the hosts with endotoxin.In lactating cows, marked changes in plasma levels ofcertain energy-related metabolites were reported simultaneouswith the endotoxin-induced endocrine alterations.Concentrations of glucose tended to increaseinitially then subsequently declined and there was atendency for increased non-esterified fatty acid values,whereas plasma b-hydroxybutyrate (BHB) decreasedlinearly in a dose-dependent manner after lipopolysaccharide(LPS) infusion (Waldron et al. 2003).In addition to endotoxin from Gram-negative bacteriacausing metabolic perturbations, abnormal metabolicstatus of farm animals can influence their responseto infection. Epidemiological studies have demonstratedinterrelations among negative energy balance-relatedmetabolic disorders (hepatic lipidosis and ketosis), theincreased incidence of clinical mastitis and theÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

318 BK Whitlock, JA Daniel, RR Wilborn, TH Elsasser, JA Carroll and JL Sartinsubsequent decrease in reproductive performance inhigh-producing postpartum dairy cows (Valde et al.1997; Washburn et al. 2002). Some trials confirmed theindirect negative impact of clinical and subclinicalmastitis on reproductive performance (Barker et al.1998), whereas others revealed direct mastitis-inducedabnormalities in ovarian function (Moore et al. 1991;Hockett et al. 2000; Huszenicza et al. 2005). Understandingthe mechanisms involved in how endotoxininteracts with reproduction, metabolism and endocrinologymay lead to unique strategies to reverse thenegative effects of infectious disease on farm animalsand humans.Understanding Endotoxin and NeuroendocrineRegulationWhile the term ‘endotoxin’ is often mistakenly appliedto any toxin derived from the microbes, the technicaldefinition is specific for the group of LPS complexesextractable from or released from the outer membraneof Gram-negative pathogenic and non-pathogenic speciesof bacteria such as E. coli, Salmonella, Neisseria,Mannheimia, Pseudomonas and others. For the mostpart, LPS complexes are rather stabile and confinedwithin the membrane of these bacteria under states ofwhat might be called ‘bacterial good health’. Forexample, upon infection with a Gram-negative pathogen,several interactions between the invading bacteriaand cellular and biochemical factors in the immunesystem lead to the degeneration of the outer bacterialmembrane with the consequential release of LPS intothe host’s internal environment. The most commoncauses for this release of LPS are bacterial autolysis,phagocytosis and digestion of bacteria by prowlingactivated immune cells such as macrophages and neutrophils,and exogenous lysis facilitated by LPS activationof the complement cascade and lysosome activity.Of concern in human clinical medicine is the recentconfirmation that massive septic crisis can be furthercomplicated acutely with the administration of certainantibiotics because of the mode of action of these drugsto facilitate cell wall breakdown with the resultingrelease of LPS. In this situation, untimely administrationof drugs with modes of action like that of thepenicillins might precipitate an acute crisis through theinitiation of multiple organ failure. Though not acommon occurrence, the scenario should be recognizedas a potential confounding factor in the treatment ofsepsis.Endotoxin challenges metabolism through direct andindirect mechanisms. The direct effects of LPS, in theabsence of cytokine production, on the release ofpituitary hormones critical to the maintenance ofmetabolism as well as reproduction were demonstratedby Coleman et al. (1993). In this regard, some nonimmunecells such as hepatocytes (Liu et al. 1998),adipocytes (Daniel et al. 2003) and pituitary cells(Daniel et al. 2005) differentially express the LPSbinding receptor CD14 on the plasma membrane,providing a mechanistic explanation as to how cellsmight respond directly to LPS. Moreover, we havepreliminary immunohistochemical evidence that theToll-like Receptor 4 is also expressed on somatotropes(Elsasser and Sartin, unpublished). While hepatocyteCD14 was shown to upregulate following LPS challenge(Liu et al. 1998), somatotrope expression decreased andconstitutive presence on gonadotropes, lactotropes andcorticotropes was unchanged (Daniel et al. 2005). Thisrecent discovery of the downregulation of CD14 onsomatotropes after LPS release is consistent with anactivation of the endotoxin receptor on the pituitary. Inregard to the so-called indirect mechanisms of actions,the classical pattern of proinflammatory cytokinesreleased from a challenged immune system functions inboth an endocrine and paracrine manner to interactwith specific receptors for these cytokines. This affectsnot only the metabolic character of target cells (liver,muscle and adipose), but also metabolic regulatoryorgans, the pituitary and pancreas in particular.As we have reviewed earlier (Elsasser et al. 2000;Daniel et al. 2002; Elsasser and Kahl 2002; Carroll2008), LPS elicits a well-timed elaboration of proinflammatorycytokines, prostaglandin derivatives, catecholaminesand free radicals from neutrophils,monocytes and macrophages which, depending on theseverity of the response, largely halt anabolic processesand initiate catabolic breakdown of tissue reserves.Most influential on metabolic processes and prototypicalof the anti-anabolic character of the endotoxinproinflammatory response is the release of tumournecrosis factor-a (TNF-a). This relatively short-termeffector has a unique influence on metabolism, but canalso be considered as an initiator of further inflammatoryresponse. Data indicate that the TNF-a response toinfused or bolus-administered LPS is largely turned offafter approximately 4–6 h. Regardless of the administrationmodel, the TNF-a event initiates progression ofadditional proinflammatory cytokines such as IL-6, andalso elaborates anti-inflammatory cytokines such as IL-4, c-interferon and IL-10 (Fig. 1). Significant in thisprocess are the homeostatic survival mechanisms associatedwith the development of what is characterized asearly and late endotoxin tolerance (West and Heagy2002; Elsasser et al. 2004). The response is dependent onPlasma cytokine (pg/ml)1e+51e+41e+31e+21e+11e+01e–1IL-6IL-10 1 2 3 4 6Time (h)Fig. 1. Effect of LPS on plasma TNF-a, IL-6, IL-1 and INF-c incalves. Calves were treated with endotoxin (arrow; 0.6 lg ⁄ kg BW),n = 5. Data are redrawn from unpublished data by the authors forillustrative purposesÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Endotoxin and Neuroendocrinology 319the timely generation of nitric oxide (NO) as driven byTNF-a and accompanied by translocation of NFjb tothe nucleus which in turn attenuates transcription of theTNF-a gene. The result is a necessary turndown orsuppression of proinflammatory signalling elementsthat, if left unchecked, leads to free radical tissuedamage (Zeisberger and Roth 1998).Sites of Endotoxin Action on Growth Hormone(GH) and Luteinizing Hormone (LH)Both GH and LH are produced and secreted by thepituitary under the influence of releasing hormones fromthe hypothalamus. Thus, LPS may influence circulatingconcentrations of GH and LH directly by alteringproduction and secretion at the pituitary, or indirectlyby altering production and secretion of releasinghormones at the level of the hypothalamus.Endotoxin was reported to impair adenohypophysialLH release in rats (Rettori et al. 1994) and sheep(Coleman et al. 1993; Fig. 2). In cycling heifers receivingan experimental challenge 42 h after the PGF 2a (dinoprost)-inducedluteolysis (Suzuki et al. 2001), LPSreduced the pulse frequency of LH for 6 h, andincreased the mean concentration and pulse amplitudeof LH. Plasma concentrations of cortisol andLH (ng/ml)GH (ng/ml)87654321–1 0 1 2 3 4 5 6 7 8 9 10Hours after treatment18161412108642ControlEndotoxinEndotoxin (0.4 µg/kg BW)Control0–1 0 1 2 3 4 5 6 7 8 9Hours after treatmentFig. 2. Effects of LPS on pulsatile patterns of LH and GH in sheeptreated with LPS (Coleman et al. 1993)10progesterone were simultaneously and transientlyincreased due to the adrenocortical over-production ofthese hormones. Plasma oestradiol concentrations weredecreased and the preovulatory LH pulse was delayed orcompletely blocked. A similar disruption was demonstratedin the preovulatory rise of oestradiol and in thesecretory pattern of LH in ewes following endotoxinchallenge (Battaglia et al. 2000). Endotoxin absorbedfrom the uterine lumen was reported to suppress theformation of the preovulatory LH peak and to inducethe cystic degeneration of dominant follicles in postpartumcows (Peter et al. 1989). Also of particularimportance in farm animals is the finding that neonatalexposure to LPS actually programs long-term sensitivityof the GnRH regulatory system, such that post-natalresponses to LPS produce a greater inhibition of GnRHand LH (Li et al. 2007).Endotoxin suppresses circulating concentrations ofLH at the level of the hypothalamus (Coleman et al.1993). This is supported as follows by three primarypieces of evidence. Firstly, portal vein cannulationindicates that LPS results in reduced secretion of GnRHfrom the hypothalamus, thus suggesting endotoxininhibits circulating concentrations of LH by inhibitinghypothalamic stimulation of LH secretion (Battagliaet al. 1997). Secondly, the increased secretion of LH inresponse to LPS by dispersed pituitary cells (Colemanet al. 1993) further suggests that LPS acts at the level ofthe hypothalamus to inhibit circulating concentrationsof LH. Finally, LPS challenge reduces electrical activityin areas of the hypothalamus associated with thegeneration of GnRH and thus LH pulses (Takeuchiet al. 1997; Yoo et al. 1997).In contrast to the effects on LH, endotoxin’s effect oncirculating concentration of GH is more complicated. Inspecies where LPS decreases, the circulating concentrationof GH (cattle and rat), the effect is primarily at thelevel of the hypothalamus. However, in species whereLPS increases circulating concentrations of GH (sheepand human), the effect appears to occur primarily at thepituitary. In pigs, an acute increase in circulatingconcentrations of GH following an LPS challenge hasbeen reported (Parrott et al. 1995; Hevener et al. 1997).However, this effect was only short-lived, and subsequentLPS-induced uncoupling of the GH ⁄ IGF-I axispersisted. Pituitary production and secretion of GH isprimarily under the regulation of GH releasing hormone(GHRH) and somatostatin from the hypothalamus. Inspecies where GH concentration is decreased by LPS,the effect is likely mediated through cytokine-inducedstimulation of somatostatin production (Scarborough1990). Sheep injected with LPS secrete GH during thesame period when LH secretion is reduced (Fig. 2;Coleman et al. 1993). Challenge with LPS also results inan increase in somatostatin concentrations in hypophysialportal blood with no change in the concentrations ofGHRH in sheep (Briard et al. 1998). Under normalphysiological conditions, an increase in somatostatinaccompanied with no change in GHRH would beexpected to result in decreased circulating concentrationsof GH. However, Briard et al. (1998) observedincreased circulating concentrations of GH associatedwith increased hypophysial portal blood concentrationsÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

320 BK Whitlock, JA Daniel, RR Wilborn, TH Elsasser, JA Carroll and JL Sartinof somatostatin in sheep challenged with LPS (Briardet al. 1998). Thus, the effect of LPS to increase circulatingconcentrations of GH was not explained byaltered hypothalamic function, and the effects of LPS toincrease circulating concentrations of GH appear to bemediated at the pituitary. Further supporting a pituitarysite of action, LPS challenge also results in increasedsecretion of GH from dispersed pituitary cells (Colemanet al. 1993).Possible Mechanisms by Which EndotoxinAlters LH and GHMultiple changes occur in response to endotoxin challengewhich could potentially affect the decrease incirculating concentrations of LH. One key response to adisease challenge and stress in general is activation ofthe hypothalamic–pituitary–adrenal axis with an associatedincrease in glucocorticoids, particularly cortisol.Indeed, administration of cortisol will reduce circulatingconcentrations of LH (Debus et al. 2002) and GH(Thompson et al. 1995). However, inhibition of cortisolsynthesis during the LPS challenge does not preventLPS suppression of pulsatile GnRH and LH secretion(Debus et al. 2002). Thus, while cortisol may play a rolein LPS-induced suppression of circulating concentrationsof LH, other factors are also likely involved.The endogenous opioid system may also be involvedin LPS suppression of circulating concentrations of LH.Central administration of the opiate antagonist naloxoneblocked LPS suppression of circulating concentrationsof LH in monkeys (Xiao et al. 2000). In addition,administration of naloxone to heifers following LPStreatment resulted in increased circulating concentrationsof LH (Kujjo et al. 1995). However, naloxone didnot prevent the decrease in circulating concentrations ofLH observed following challenge with E. coli in sheep(Leshin and Malven 1984). The inability of an opiateantagonist to block E. coli-induced LH suppression maybe associated with the differential proinflammatorycytokine profiles observed in LPS vs E. coli challengestudies. For instance, in pigs, LPS is commonly knownto induce the primary proinflammatory cytokines TNFa,IL-Ib and IL-6 (Carroll et al. 2003). However, in pigschallenged with live E. coli, TNF-a concentrations arenot elevated (Strauch et al. 2004) indicating that theactivation of the acute phase immune response isdepending upon the immunological challenge. Thus,while the endogenous opioid system may play a role inLPS suppression of circulating concentrations of LH,other systems are involved in the effect of Gramnegativebacterial challenge to suppress LH.As discussed earlier, challenge with LPS and Gramnegativebacteria results in activation of the immunesystem. Components of the innate immune system maybe responsible for the suppression of circulating concentrationsof LH. Indeed, investigators have examinedthe role of the cytokines TNF-a and IL-1b as well asprostaglandins in LPS suppression of LH. Centraladministration of both TNF-a and IL-1b suppressedcirculating concentrations of LH (Daniel et al. 2005).Central administration of IL-1b to monkeys also suppressedcirculating concentrations of LH (Xiao et al.2000). However, neither central nor peripheral administrationof the cytokine antagonists TNF-R1 norIL-1RA prevented the LPS-induced suppression ofcirculating concentrations of LH (Xiao et al. 2000;Daniel et al. 2005). Thus, while increased concentrationsof inflammatory cytokines may play a role in LPSsuppression of circulating concentrations of LH, themechanism for LH inhibition is clearly multifactorial.In response to LPS challenge, prostaglandin productionis also enhanced. Treatment with a prostaglandinsynthesis inhibitor (flurbiprofen) prevented endotoxinsuppression of LH and GnRH (Harris et al. 2000).Thus, prostaglandin formation is a crucial step in LPSsuppression of pulsatile secretion of GnRH and LH.In contrast, the oestradiol-induced surge of LH wasblocked by LPS, but the inhibition was found to functionvia prostaglandin independent pathways (Breen et al.2004).The mechanisms by which LPS alters circulatingconcentrations of GH are clearer than the means bywhich LPS suppresses LH. Early data suggested that insome species (rat), LPS administration results in suppressedcirculating concentrations of GH. The reducedGH is due to increased somatostatin release in responseto inflammatory cytokines, specifically IL-1, TNF andIL-6 (Scarborough 1990). In contrast, other speciesincrease circulating concentrations of GH in response toinflammatory cytokines and to LPS administration. Inthis case, the site of action is the pituitary as opposed tothe hypothalamus. More recently, this disparate effect ofLPS between different species was evaluated by Priegoet al. (2003). Rats administered low doses of LPS hadincreased plasma GH, whereas high doses were inhibitoryto GH, indicating that differences observed weredue to dose of LPS and not species differences. Thus, thetypical model in rats is a model of endotoxic shock,whereas the model in sheep and humans tends to modela less severe disease.Studies with dispersed ovine pituitary cells have foundthat treatment with IL-1 stimulated GH synthesis andsecretion while TNF-a reduced GRH-stimulated GHrelease (Fry et al. 1998). Moreover, TNF-a will inhibitGH release from cultured bovine pituitary cells (Elsasseret al. 1991). In addition, peripheral administration ofTNF-a and IL-1b resulted in increased circulatingconcentrations of GH (Daniel et al. 2005). Intravenous(IV) but not intracerebroventricular administration ofthe cytokine antagonists, sTNF-R1 or IL-1RA, preventedthe LPS-induced increase in circulating concentrationsof GH (Daniel et al. 2005). The differing effectsof TNF-a between in vivo and in vitro models suggestthat in vivo, TNF-a activates IL-1b release which in turnis a stimulus to GH. This increase in inflammatorycytokines in response to LPS plays a critical role inendotoxin-induced alterations in circulating concentrationsof GH. Finally, Daniel et al. (2005) discoveredthat CD14 is expressed on somatotropes and is downregulatedin response to LPS injection, whereas thetypical dogma suggests that the major effects on therodent pituitary are via folliculostellate cells. In agreementwith previous research (Coleman et al. 1993), thesedata also provide evidence that GH can be released fromthe somatotrope in response to direct exposure to LPS.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Endotoxin and Neuroendocrinology 321Another mediator which may be involved in the GHand LH response to LPS is ghrelin. Ghrelin is anendogenous ligand for the growth hormone secretagoguereceptor which stimulates GH secretion (Kojimaet al. 1999). In pigs, ghrelin has been reported toincrease body weight gain and circulating concentrationsof GH and insulin (Salfen et al. 2004). Recent dataindicate that LPS challenge results in increased circulatingconcentrations of ghrelin in humans, a specieswith increased GH in response to LPS (Vila et al. 2007).Perhaps an increase in ghrelin stimulates the increase inGH observed following an LPS challenge. In addition,ghrelin has been demonstrated to reduce circulatingconcentrations of LH in rats (Furuta et al. 2001),monkeys (Vulliemoz et al. 2004), sheep (Iqbal et al.2006) and humans (Kluge et al. 2007). The possiblerelationship of ghrelin to the LPS inhibition of LH andstimulation of GH should be a focus of future researchinvolving the mechanisms of LPS action in ruminants.Endotoxin Disrupts GH SignallingInitial studies demonstrated that the proinflammatorymediators of LPS action within the immune systemmediated a reduction in circulating plasma IGF-1concentrations and this reduction was independent ofthe decrease in voluntary food intake observed in theLPS-treated calves (Elsasser et al. 1995). The chronicdecline of this key anabolic hormone during catabolicdisease suggested that calves suffering weight loss frominfectious diseases might benefit from the use ofanabolic hormones to decrease the catabolic milieu. Assuch, studies were conducted to evaluate the actions ofanabolic agents such as oestradiol and GH (Elsasseret al. 1998; Sartin et al. 1998). Elsasser et al. (1998)examined the possibility that treatment with exogenousGH could be used to overcome the effects of sarcocystisinfection in cattle by increasing plasma IGF-I concentrationsand normalizing metabolism. However, GHhad no ability to normalize plasma concentrations ofIGF-I or other indices of metabolism. Since no problemswere found with liver GH receptor functions, thislack of effect of GH was hypothesized to relate toaltered GH signalling. In a series of follow-up studiesusing the LPS model, Elsasser et al. (2004, 2007a,b)determined that the activation of the major signaltransduction regulator of GH action, protein tyrosinekinase JAK-2, by GH was reduced by infection, thusexplaining the GH resistance observed during disease.Moreover, the specific locus of the resistance was relatedto a reduced capacity for JAK-2 to be activated at itskinase epitope by phosphorylation (Fig. 3). Moredetailed studies have determined that a major targetwhere JAK-2 is nitrated is the position normallyphosphorylated (… 1007 tyrosine- 1008 tyrosine…) in theGH activation of the signal transduction cascade, thusproviding a novel mechanism to explain some verylocalized aspects of GH resistance in disease (Elsasseret al. 2007a,b). In this instance, increased nitration ofJAK-2 is associated with decreased phosphorylation,dimerization and translocation to the nucleus of the keynuclear gene transcription factor for GH activation ofIGF-1 production, STAT5b. The interesting feature ofFig. 3. Representative Western blot of liver homogenate proteins fromcontrol (before LPS, lane 1) and after LPS (24 h, 2.5 lg ⁄ kg E. coli055:B5 LPS, lanes 2 and 3) biopsy samples. (Elsasser et al. 2007a)this nitration response is that while it appears to be atransient phenomenon in states of normal health, theformation of nitrated JAK-2 is an overt and prolongedoccurrence in chronic disease states. The demonstratedability to specifically modulate this nitration response byaltering the generation of nitric oxide and superoxidemay prove useful to further pharmacological strategiesthat might be used to re-establish stabile metabolismand mitigate the impacts of disease stress on animalhealth.ConclusionMany diseases common to farm and production animalsare caused by infections with Gram-negative bacteria.The consequences of these infections are a result of theliberation of endotoxin ⁄ LPS from the bacteria and thereaction of the immune system ⁄ inflammatory cells.Cytokine (e.g. IL-I and TNF-a) production by inflammatorycells in response to LPS is a normal andnecessary function of the immune system in animals toprevent and alleviate infections. However, the inflammatorycytokines can also initiate a cascade of eventsthat impair hormonal and metabolic homeostatic processesregulating growth, metabolism and reproduction.The purpose for impairing these functions is most likelya consequence of the lofty nutrient ⁄ metabolic demandsfor the immune system’s response to infections. In anattempt to redirect or conserve nutrients for immunefunctions, growth is impaired by direct or indirectactions of cytokines on the somatotropic axis. In adultanimals, nutrients may be conserved by inhibitingenergetically risky behaviour [i.e. reproduction (oestrus,pregnancy and lactation)] through manipulation of thehypothalamic–pituitary–gonadal axis at any point.More research is needed to completely understand themechanisms behind the effects of disease stress ongrowth, metabolism and reproduction. While previousdata on severe proinflammatory-mediated dysfunctionwere associated with stark pathology, the nitrationconcept follows closely with perturbations associatedwith low level responses to immune challenge that donot culminate in death. For example, the implications ofthe recent data by Elsasser et al. (2007a,b) suggest anovel mechanism for inhibition of endocrine signallingwhich should be examined as a possible unifyingmechanism by which LPS actions suppress key functionsin cells. By understanding the intricacies of LPS-inducedcytokine release and how cytokines affect farm animalproduction, we can target more precisely the strategiesneeded to combat infectious agents as well as stabilizeÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

322 BK Whitlock, JA Daniel, RR Wilborn, TH Elsasser, JA Carroll and JL Sartinhost responses to infection in order to speed recoveryand improve animal welfare.ReferencesBarker AR, Schrick FN, Lewis MJ, Dowlen HH, Oliver SP,1998: Influence of clinical mastitis during early lactation onreproductive performance of Jersey cows. J Dairy Sci 81,1285–1290.Battaglia DF, Bowen JM, Krasa HB, Thrun LA, Viguie C,Karsch FJ, 1997: Endotoxin inhibits the reproductiveneuroendocrine axis while stimulating adrenal steroids: asimultaneous view from hypophyseal portal and peripheralblood. Endocrinology 138, 4273–4281.Battaglia DF, Krasa HB, Padmanabhan V, Viguie C, KarschFJ, 2000: Endocrine alterations that underlie endotoxininduceddisruption of the follicular phase in ewes. BiolReprod 62, 45–53.Breen KM, Billings HJ, Debus N, Karsch FJ, 2004: Endotoxininhibits the surge secretion of gonadotropin-releasing hormonevia a prostaglandin-independent pathway. Endocrinology145, 221–227.Briard N, Guillaume V, Frachebois C, Rico-Gomez M, SauzeN, Oliver C, Dutour A, 1998: Endotoxin injection increasesgrowth hormone and somatostatin secretion in sheep.Endocrinology 139, 2662–2669.Carroll JA 2008: Bi-directional communication: growth andimmunity in domestic livestock. J Anim Sci 86, (suppl. 14)E126–E137.Carroll JA, Gaines AM, Spencer JD, Allee GL, Kattesh HG,Roberts MP, Zannelli ME, 2003: Effect of menhaden fish oilsupplementation and lipopolysaccharide exposure on nurserypigs. I. Effects on the immune axis when fed dietscontaining spray-dried plasma. Domest Anim Endocrinol24, 341–351.Coleman ES, Elsasser TH, Kemppainen RJ, Coleman DA,Sartin JL, 1993: Effect of endotoxin on pituitary hormonesecretion in sheep. Neuroendocrinology 58, 111–122.Daniel JA, Whitlock BK, Wagner CG, Sartin JL, 2002:Regulation of the growth hormone and luteinizing hormoneresponse to endotoxin in sheep. Domest Anim Endocrinol23, 361–370.Daniel JA, Elsasser TH, Morrison CD, Keisler DH, WhitlockBK, Steele B, Pugh D, Sartin JL, 2003: Leptin, tumornecrosis factor-alpha (TNF), and CD14 in ovine adiposetissue and changes in circulating TNF in lean and fat sheep.J Anim Sci 81, 2590–2599.Daniel JA, Elsasser TH, Martinez A, Steele B, Whitlock BK,Sartin JL, 2005: Interleukin-1beta and tumor necrosisfactor-alpha mediation of endotoxin action on growthhormone. Am J Physiol Endocrinol Metab 289, E650–E657.Debus N, Breen KM, Barrell GK, Billings HJ, Brown M,Young EA, Karsch FJ, 2002: Does cortisol mediate endotoxin-inducedinhibition of pulsatile luteinizing hormoneand gonadotropin-releasing hormone secretion? Endocrinology143, 3748–3758.Elsasser TH, Kahl S, 2002: Adrenomedullin has multiple rolesin disease stress: development and remission of the inflammatoryresponse. Microsc Res Tech 57, 120–129.Elsasser TH, Caperna TJ, Fayer R, 1991: Tumor necrosisfactor-alpha affects growth hormone secretion by a directpituitary interaction. Proc Soc Exp Biol Med 198, 547–554.Elsasser TH, Caperna TJ, Rumsey TS, 1995: Endotoxinadministration decreases plasma insulin-like growth factor(IGF)-I and IGF-binding protein-2 in Angus x Herefordsteers independent of changes in nutritional intake. JEndocrinol 144, 109–117.Elsasser TH, Sartin JL, McMahon C, Romo G, Fayer R, KahlS, Blagburn B, 1998: Changes in somatotropic axis responseand body composition during growth hormone administrationin progressive cachectic parasitism. Domest AnimEndocrinol 15, 239–255.Elsasser TH, Kahl S, Rumsey TS, Blum JW, 2000: Modulationof growth performance in disease: reactive nitrogen compoundsand their impact on cell proteins. Domest AnimEndocrinol 19, 75–84.Elsasser TH, Kahl S, MacLeod C, Nicholson B, Sartin JL, LiC, 2004: Mechanisms underlying growth hormone effects inaugmenting nitric oxide production and protein tyrosinenitration during endotoxin challenge. Endocrinology 145,3413–3423.Elsasser TH, Kahl S, Li C-J, Sartin JL, Garrett WM, RodrigoJ, 2007a: Caveolae nitration of JAK-2 at the 1007Y-1008Ysite: coordinating inflammatory response and metabolichormone readjustment within the somatotropic axis. Endocrinology148, 3792–3802.Elsasser TH, Li CJ, Caperna TJ, Kahl S, Schmidt WF, 2007b:Growth hormone (GH)-associated nitration of Januskinase-2 at the 1007Y-1008Y epitope impedes phosphorylationat this site: mechanism for and impact of a GH, AKT,and nitric oxide synthase axis on GH signal transduction.Endocrinology 148, 3792–3802.Fry C, Gunter DR, McMahon CD, Steele B, Sartin JL, 1998:Cytokine-mediated growth hormone release from culturedovine pituitary cells. Neuroendocrinology 68, 192–200.Furuta M, Funabashi T, Kimura F, 2001: Intracerebroventricularadministration of ghrelin rapidly suppresses pulsatileluteinizing hormone secretion in ovariectomized rats.Biochem Biophys Res Commun 288, 780–785.Harris TG, Battaglia DF, Brown ME, Brown MB, CarlsonNE, Viguie C, Williams CY, Karsch FJ, 2000: Prostaglandinsmediate the endotoxin-induced suppression ofpulsatile gonadotropin-releasing hormone and luteinizinghormone secretion in the ewe. Endocrinology 141, 1050–1058.Hevener W, Almond GW, Armstrong JD, Richards RG, 1997:Effects of acute endotoxemia on serum somatotropin andinsulin-like growth factor I concentrations in prepubertalgilts. Am J Vet Res 58, 1010–1013.Hirvonen J, Huszenicza G, Kulcsar M, Pyorala S, 1999:Acute-phase response in dairy cows with acute postpartummetritis. Theriogenology 51, 1071–1083.Hockett ME, Hopkins FM, Lewis MJ, Saxton AM, DowlenHH, Oliver SP, Schrick FN, 2000: Endocrine profiles ofdairy cows following experimentally induced clinical mastitisduring early lactation. Anim Reprod Sci 58, 241–251.Hussain AM, Daniel RC, 1991: Bovine endometritis: currentand future alternative therapy. Zentralbl Veterinarmed A38, 641–651.Huszenicza Gy, Ja´nosi Sz, Kulcsa´r M,Ko´ro´di P, Reiczigel J,Kátai L, Peters AR, De Rensis F, 2005: Effects of clinicalmastitis on ovarian function in postpartum dairy cows.Reprod Domest Anim 40, 199–204.Iqbal J, Kurose Y, Canny B, Clarke IJ, 2006: Effects of centralinfusion of ghrelin on food intake and plasma levels ofgrowth hormone, luteinizing hormone, prolactin, and cortisolsecretion in sheep. Endocrinology 147, 510–519.Kluge M, Schussler P, Uhr M, Yassouridis A, Steiger A, 2007:Ghrelin suppresses secretion of luteinizing hormone inhumans. J Clin Endocrinol Metab 92, 3202–3205.Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H,Kangawa K, 1999: Ghrelin is a growth-hormone-releasingacylated peptide from stomach. Nature 402, 656–660.Kujjo LL, Bosu WT, Perez GI, 1995: Opioid peptidesinvolvement in endotoxin-induced suppression of LH secre-Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Endotoxin and Neuroendocrinology 323tion in ovariectomized Holstein heifers. Reprod Toxicol 9,169–174.Leshin LS, Malven PV 1984: Bacteremia-induced changes inpituitary hormone release and effect of naloxone. Am JPhysiol 247 (5 Pt 1) E585–E591.Li XF, Kinsey-Jones JS, Knox AM, Wu XQ, Tahsinsoy D,Brain SD, Lightman SL, O’Byrne KT, 2007: Neonatallipopolysaccharide exposure exacerbates stress-induced suppressionof luteinizing hormone pulse frequency in adulthood.Endocrinology 148, 5984–5990.Liu S, Khemlani LS, Shapiro RA, Johnson ML, Liu K, GellerDA, Watkins SC, Goyert SM, Billiar TR, 1998: Expressionof CD14 by hepatocytes: upregulation by cytokines duringendotoxemia. Infect Immun 66, 5089–5098.Moore DA, Cullor JS, Bondurant RH, Sischo WM, 1991:Preliminary field evidence for the association of clinicalmastitis with altered interestrus intervals in dairy cattle.Theriogenology 36, 257–265.Parrott RF, Vellucci SV, Goode JA, Lloyd DM, Forsling ML,1995: Cyclo-oxygenase mediation of endotoxin-inducedfever, anterior and posterior pituitary hormone release,and hypothalamic c-Fos expression in the prepubertal pig.Exp Physiol 80, 663–674.Peter AT, Bosu WT, DeDecker RJ, 1989: Suppression ofpreovulatory luteinizing hormone surges in heifers afterintrauterine infusions of Escherichia coli endotoxin. Am JVet Res 50, 368–373.Priego T, Granado M, Iba´n˜ ez de Cáceres I, Martı´n AI, Villanu´aMA, Lo´pez-Caldero´n A., 2003: Endotoxin at low dosesstimulates pituitary GH whereas it decreases IGF-I and IGFbindingprotein-3 in rats. J Endocrinol 179, 107–117.Rettori V, Belova N, Kamat A, Lyson K, Gimeno M, McCannSM, 1994: Blockade by interleukin-1-alpha of nitricoxidergiccontrol of luteinizing hormone-releasing hormone releasein vivo and in vitro. Neuroimmunomodulation 1, 86–91.Salfen BE, Carroll JA, Keisler DH, Strauch TA, 2004: Effectsof exogenous ghrelin on feed intake, weight gain, behavior,and endocrine responses in weanling pigs. J Anim Sci 82,1957–1966.Sartin JL, Elsasser TH, Gunter DR, McMahon CD, 1998:Endocrine modulation of physiological responses to catabolicdisease. Domest Anim Endocrinol 15, 423–429.Scarborough DE 1990: Somatostatin regulation by cytokines.Metabolism 39 (9 Suppl. 2) 108–111.Sheldon IM, Lewis GS, LeBlanc S, Gilbert RO, 2006: Definingpostpartum uterine disease in cattle. Theriogenology 65,1516–1530.Strauch TA, Carroll JA, Fangman TJ, Wiedmeyer CE,Hamback AK, 2004: The acute phase response of Escherichiacoli challenged pigs exhibiting a febrile response in theabsence of elevated TNF-alpha. J Anim Vet Adv 3, 227–238.Suzuki C, Yoshioka K, Iwamura S, Hirose H, 2001: Endotoxininduces delayed ovulation following endocrine aberrationduring the proestrous phase in Holstein heifers.Domest Anim Endocrinol 20, 267–278.Takeuchi Y, Nagabukuro H, Kizumi O, Mori Y, 1997:Lipopolysaccharide-induced suppression of the hypothalamicgonadotropin-releasing hormone pulse generator inovariectomized goats. J Vet Med Sci 59, 93–96.Thompson K, Coleman ES, Hudmon A, Kemppainen RJ,Soyoola EO, Sartin JL, 1995: Effects of short-term cortisolinfusion on growth hormone-releasing hormone stimulationof growth hormone release in sheep. Am J Vet Res 56, 1228–1231.USDA 1996: Dairy ‘96. Animal Plant Health InspectionService. National Animal Health Monitoring Service. Ft.Collins, CO, USA.Valde JP, Hird DW, Thurmond MC, Osteras O, 1997:Comparison of ketosis, clinical mastitis, somatic cell count,and reproductive performance between free stall and tie stallbarns in Norwegian dairy herds with automatic feeding.Acta Vet Scand 38, 181–192.Vila G, Maier C, Riedl M, Nowotny P, Ludvik B, Luger A,Clodi M, 2007: Bacterial endotoxin induces biphasicchanges in plasma ghrelin in healthy humans. J ClinEndocrinol Metab 92, 3930–3934.Vulliemoz NR, Xiao E, Xia-Zhang L, Germond M, Rivier J,Ferin M, 2004: Decrease in luteinizing hormone pulsefrequency during a five-hour peripheral ghrelin infusion inthe ovariectomized rhesus monkey. J Clin EndocrinolMetab 89, 5718–5723.Waldron MR, Nishida T, Nonnecke BJ, Overton TR, 2003:Effect of lipopolysaccharide on indices of peripheral andhepatic metabolism in lactating cows. J Dairy Sci 86, 3447–3459.Washburn SP, White SL, Green JT Jr, Benson GA, 2002:Reproduction, mastitis, and body condition of seasonallycalved Holstein and Jersey cows in confinement or pasturesystems. J Dairy Sci 85, 105–111.West MA, Heagy W, 2002: Endotoxin tolerance: a review. CritCare Med 30(1 Suppl.), S64–S73.Xiao E, Xia-Zhang L, Ferin M, 2000: Inhibitory effects ofendotoxin on LH secretion in the ovariectomized monkeyare prevented by naloxone but not by an interleukin-1receptor antagonist. Neuroimmunomodulation 7, 6–15.Yoo MJ, Nishihara M, Takahashi M, 1997: Tumor necrosisfactor-alpha mediates endotoxin induced suppression ofgonadotropin-releasing hormone pulse generator activity inthe rat. Endocr J 44, 141–148.Zeisberger E, Roth J, 1998: Tolerance to pyrogens. Ann N YAcad Sci 856, 116–131.Author’s address (for correspondence): JL Sartin, Department ofAnatomy, Physiology and Pharmacology, College of VeterinaryMedicine, Auburn University, Auburn, AL 36849, USA. E-mail:sartijl@vetmed.auburn.eduConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 324–330 (2008); doi: 10.1111/j.1439-0531.2008.01173.xISSN 0936-6768Energy Metabolism and Leptin: Effects on Neuroendocrine Regulation ofReproduction in the Gilt and SowCR Barb 1 , GJ Hausman 1 and CA Lents 21 USDA ⁄ ARS, Richard B. Russell Agriculture Research Center; 2 Animal and Dairy Science Department, University of Georgia, Athens, GA, USAContentsIt is well established that reproductive function is metabolicallygated. However, the mechanisms whereby energy storesand metabolic cues influence appetite, energy homeostasis andfertility are yet to be completely understood. Adipose tissue isno longer considered as only a depot to store excess energy.Recent findings have identified numerous genes, severalneurotrophic factors, interleukins, insulin-like growth factorbinding protein-5, ciliary neurotrophic factor and neuropeptideY (NPY) as being expressed by adipose tissue duringpubertal development. These studies demonstrated for the firsttime the expression of several major adipokines or cytokines inpig adipose tissue which may influence local and centralmetabolism and growth. Leptin appears to be the primarymetabolic signal and is part of the adipose tissue-hypothalamicregulatory loop in the control of appetite, energy homeostasisand luteinizing hormone (LH) secretion. Leptin’s actions onappetite regulation are mediated by inhibition of hypothalamicNPY and stimulation of proopiomelanocortin. Its effects ongonadotropin-releasing hormone (GnRH) ⁄ LH secretion aremediated by NPY and kisspeptin. Thus, leptin appears to bean important link between metabolic status, the neuroendocrineaxis and subsequent fertility in the gilt and sow.IntroductionIt is generally accepted that there are two modes ofluteinizing hormone (LH) secretion in the pig (Kraelingand Barb 1990), pulsatile secretion and surge secretion.These patterns of LH secretion reflect the pattern ofgonadotropin-releasing hormone (GnRH) released fromneurosecretory neurones within the hypothalamus intothe hypothalamic-hypophysial portal system (Williams1989). The importance of pulsatile GnRH ⁄ LH secretionwas demonstrated in studies that induced precociousoestrus and ovulation in intact pre-pubertal gilts withhourly intravenous (i.v.) injections of GnRH (Lutz et al.1984; Pressing et al. 1992). In addition, hourly administrationof GnRH to anestrous post-partum sowsinduced oestrus and ovulation (Cox and Britt 1988).Thus, GnRH neurones secrete their product in anepisodic manner, but interoceptive and exteroceptivefactors detected by the central nervous system aretranslated by neuroendocrine mechanisms into signalswhich alter the pattern of GnRH and subsequent LHsecretion. For example, interoceptive signals, such asgonadal and adrenal steroids, metabolites, and otherneuronal signals act to modulate frequency and amplitudeof GnRH pulses.The importance of nutrition and metabolic state ininitiating and maintaining reproductive function andgrowth is well established (reviewed by Prunier et al.1993; Prunier and Quesnel 2000; Barb et al. 2002).Dietary nutrients influence expression of metabolicpathways that allow animals to achieve their full geneticpotential for reproduction and growth. These pathwaysare complex and involve appetite regulation along withregulation of the reproductive and growth axis, as wellas gonadal function (Wade et al. 1996; Barb et al. 1999;Muller et al. 1999). Identification of blood-borne metabolicsignals that activate the GnRH ⁄ LH pulse generator,alter the pattern of growth hormone (GH)secretion, and regulate appetite remains elusive. Recentreports have identified adipose tissue as a source ofputative metabolic signals that regulate the neuroendocrineaxis (Schwartz et al. 1996; Barb and Kraeling2004; Barb et al. 2006a).Nutrition and ReproductionResearch has examined backfat levels in gilts relative tofuture reproductive performances. Several reports demonstratedthat gilts with higher backfat subsequentlyhad a shorter weaning to first oestrus interval, largerlitter size, and higher farrowing rate as second paritysows compared to gilts with lower amounts of backfat(Tummaruk et al. 2001). Whittemore (1996) suggestedthat primiparous sows should not be allowed to havebackfat depth fall below 14 mm or rise above 25 mm(lipid ⁄ protein >1:1 and

Energy and Reproduction in the Pig 325lactation (Weldon et al. 1994; Whittemore 1996).Almeida et al. (2001) showed that moderate feedrestriction during the luteal phase of the oestrous cyclein the gilt affects ovulation rate and the progesteronerise after the LH surge.The role of insulin on LH secretion in the postpartumsow is somewhat controversial. Tokach et al.(1992) reported that insulin levels during early lactationwere correlated with LH peak amplitude and thusappear to be associated with reproductive function.Koketsu et al. (1998) reported greater lactation feedintake was associated with greater concentrations ofinsulin and glucose, greater LH pulse frequency prior toweaning and shorter farrowing-to-oestrus interval. Inaddition, exogenous insulin administered on the day ofweaning for 4 consecutive days in primiparous sowsdecreased the average interval from weaning to oestrusand increased the percentage of sows in oestrus (Whitleyet al. 1998). One model used to study the effect of insulinon the hypothalamic-pituitary axis is the diabetesinducedanimal. In the diabetic ovariectomized gilt,withdrawal of insulin therapy for 4 days prevented theoestradiol-induced preovulatory-like LH surge, but didnot affect pulsatile LH secretion (Angell et al. 1996).This finding indicates that diabetes mellitus alters thesensitivity of the hypothalamic-pituitary axis to oestradioland pituitary responsiveness to GnRH. Pituitarycell culture experiments confirmed that the sensitivity ofthe pituitary gland to GnRH decreased after removal ofinsulin therapy for 7 days in diabetic pigs (Angell et al.1996), suggesting that insulin may play a role inmaintaining pituitary responsiveness to GnRH. Incontrast, insulin administration on 5 consecutive daysprior to weaning in feed-restricted sows did not affectweaning to oestrus interval or ovulation rate (Quesneland Prunier 1998). In primiparous sows, neither glucosenor insulin concentrations were correlated with LHsecretion during lactation, after weaning or with theweaning to oestrus interval (van den Brand et al. 2000).Moreover, van den Brand et al. (2001) found inprimiparous sows that plasma IGF-I concentrationson day 22 (weaning) were positively correlated with LHpulse frequency while Quesnel et al. (1998) reported norelationship between IGF-I concentrations and LHpulse frequency before or after weaning in primiparoussows. Thus, the definitive role of IGF-I and insulin onLH secretion and subsequent reproductive performancein the post-partum sow has yet to be determined.Serum and milk leptin concentrations in the primaparousand multiparous lactating sow were positivelycorrelated with backfat thickness and level of dietaryenergy fed during gestation, as well as feed consumption(Estienne et al. 2000, 2003). A positive correlation wasobserved among plasma insulin, leptin and LH concentrationsin lactating sows fed ad libitum but not in feedrestrictedsows. Moreover, the weaning to oestrusinterval was greater in the feed-restricted sows comparedto controls (Mao et al. 1999). De Rensis et al. (2005)reported that in sows classified as fat, medium or thinbased on backfat thickness at farrowing, serum leptinconcentrations were greater in fat sows compared tomedium and thin animals at weaning. In addition, therewas no relationship between leptin concentrations andreproductive performance after weaning. However,plasma leptin concentrations were associated withbackfat depth, and loss of backfat was associated withreproductive performance. These findings provide evidencethat circulating leptin, LH concentrations andfeed consumption during lactation are influenced bydietary energy intake during pregnancy or lactation inthe sow, suggesting that leptin may serve as a permissivemetabolic signal that may be necessary for activation ofthe reproductive axis in the post-partum sow.Growth and Metabolic SignalsIn addition to developmentally related maturation of theneuroendocrine axis, permissive peripheral signals areassociated with attainment of a minimum percentage ofbody fat (Frisch 1984); one example is leptin which mayplay a role in the timing of puberty (Barb et al. 2001a;Barb and Kraeling 2004). During the pre-pubertalperiod, expression of a number of hypothalamic genesassociated with appetite and growth regulation appearto be developmentally regulated. For example, hypothalamicexpression of the biological form of the leptinreceptor (OB-rb), adipose tissue leptin expression andserum leptin concentrations increased by 3.5 months ofage (Qian et al. 1999; Lin et al. 2001). In addition,Kojima et al. (2007) demonstrated a positive relationshipbetween pre-weaning weight and post-weanimghypothalamic agouti-related protein, orexin, type-2orexin receptor and NPY gene expression. These genesmay play an important role in post-weaning growth anddevelopment in the gilt. Furthermore, their encodedproteins are well positioned anatomically to interactwith GnRH (Kraeling and Barb 1990) and GHRH(Leshin et al. 1994) neurones and appetite regulatingneurones (Lawrence et al. 1999; Matteri 2001). Consistentwith this idea, central administration of leptinincreased GH secretion and suppressed feed intake inthe pre-pubertal gilt (Barb et al. 1998) and in vitro leptinstimulated GnRH release from porcine hypothalamicexplants (Barb et al. 2004). These changes in thegrowth ⁄ reproductive axis appear to be in concert withthe timing of puberty.We previously reported that metabolic response toacute feed deprivation occurred more rapidly in prepubertalgilts compared to mature gilts, likely becausepre-pubertal gilts have a higher metabolic rate, smallerenergy reserves and thus a greater nutrient intakerequirement for growth (Barb et al. 1997). An acute24 h fast increased serum free fatty acid concentrations,and decreased leptin pulse frequency but not meanserum leptin concentrations in the ovariectomized prepubertalgilt (Barb et al. 2001a). Furthermore, shortterm feed restriction decreased leptin secretion and LHpulse frequency in the mature ovariectomized gilt(Whisnant and Harrell 2002) and decreased LH secretionin the intact pre-pubertal gilt (Booth et al. 1996),while a 3-day fast reduced adipose tissue leptin mRNAin castrate male pigs (Spurlock et al. 1998). These resultssupport the idea that leptin may serve as a metabolicsignal in the activation of the reproductive axis.In pre-pubertal gilts short term feed restriction to33% of control diet for 8 days failed to affect LH orÓ 2008 No claim to original government works

326 CR Barb, GJ Hausman and CA LentsTable 1. Microarray analysis of several fat depots demonstrated that thyroid hormone receptors were collectively upregulated while several otherreceptors (MCIR, FGF4) and lipogenic (LPL, SCD) enzymes were downregulated with fasting aFat depot Gene Estimate SE T-value p-valuePerirenal Thyroid hormone receptor alpha 2 (Sus scrofa) )0.20083 0.051199 )3.92248 0.000254Mesenteric Thyroid hormone receptor, alpha )0.16131 0.047101 )3.42473 0.001195Leaf Thyroid hormone receptor, alpha )0.15323 0.030834 )4.96954 0.000007Leaf Thyroid hormone receptor alpha 2 (Sus scrofa) )0.31586 0.089853 )3.51524 0.000909Leaf Melanocortin 1 receptor 0.357846 0.175065 2.044075 0.045929Perirenal Melanocortin 1 receptor 0.396369 0.117805 3.364614 0.001431Leaf Fibroblast growth factor receptor 4 0.40214 0.121846 3.300404 0.001731Perirenal Fibroblast growth factor receptor 4 0.44749 0.139023 3.218823 0.002199Perirenal Stearoyl-CoA desaturase 0.914822 0.323623 2.826815 0.006618Lipoprotein lipase 0.43321 0.125486 3.452266 0.0011Mesenteric Stearoyl-CoA desaturase 1.40823 0.304109 4.630675 0.000024Lipoprotein lipase 0.14577 0.069703 2.091302 0.04131Leaf Stearoyl-CoA desaturase 0.170878 0.071794 2.380111 0.020937Lipoprotein lipase 0.380183 0.079236 4.798133 0.000013a Analysis of the microarray data was conducted with either LOWESS normalization based or an ANOVA normalization based method to determine the influence offasting on gene expression.leptin secretion or to affect backfat leptin, Ob-rb,adipocyte fatty acid binding protein, or the transcriptionfactors peroxisome proliferator-activated receptor-c2and CCAAT-enhancer-binding protein-a expression(Hart et al. 2007). However, based on maintenancerequirements for the pre-pubertal gilt (NRC 1998) usedin that study, the feed restricted gilts were actually fed124% of the maintenance requirement. This may explainthe failure of feed restriction to affect LH or leptinconcentrations. Although feed restriction did preventbody weight and backfat gain adipocyte function wasaltered as evidenced by upregulation of thyroid hormonereceptor-a in perirenal, leaf and mesenteric fatdepots in feed restricted animals; this coincided withelevated thyroxin concentration (GJ Hausman, CRBarb, HA Hart, unpublished data; Table 1). Thesechanges may in part represent an attempt to maintainthermogenesis, as well as immune and neruoendocrinehomeostasis.Adipose Tissue as an Endocrine OrganAdipose tissue plays a more dynamic role thanpreviously thought in physiological mechanisms andwhole-body homeostasis. Studies now show that therole of adipose tissue includes responding to nutrient,neural and hormonal signals, and secreting factors or‘adipokines’ that control feeding, thermogenesis, immunity,and neuroendocrine function (review by Ahimaet al. 2006). The current evidence indicates that of allthe adipose tissue secreted factors, or ‘adipokines’,interleukins (IL)-6, IL-8, plasminogen activator inhibitor-1,leptin and adiponectin can be considered trueendocrine factors (review by Hauner 2005). Furthermore,the secretory function of adipose tissue isadversely influenced by both obesity (Hauner 2005)and impaired adipose tissue accretion (review byGuerre-Millo 2004). Therefore, body condition oradiposity by virtue of secreted adipokines could impactor regulate, in part, physiological states such asreproductive condition or status. Adipose tissue constitutesthe largest amount of stored energy in the body(Loftus 1999), and regulation of energy expenditureinvolves a balance of several factors, such as feedingbehaviour, adipose tissue mass and activation ofcatabolic processes (e.g. lactation). Changes in bodyweight or nutritional status are characterized byalterations in serum concentrations of many hormonesand growth factors that regulate adipocyte functionand leptin secretion (Barb et al. 2001b). Recent microarrayanalyses by our laboratory revealed that 21 genesencoding secreted proteins were expressed 40-fold overbackground in neonatal porcine adipose tissue andporcine pre-adipocyte cultures (Hausman et al. 2006).Additionally, agouti gene expression was detected byRT-PCR in pig adipose tissue. Proteomic analysis ofadipocyte culture conditioned media identified severalsecreted proteins, including several neurotrophic factors,IL-1A, IL-1B, IL-8, IL-6, IL-15, and IGF bindingprotein-5. Through microarray and RT-PCR analyses,we recently demonstrated that in addition to these,several other cytokines, e.g., ciliary neurotrophic factorand NPY, are expressed by adipose tissue duringpubertal development (Hausman et al. 2007). Thesestudies demonstrate for the first time the expressionof several major adipokines or cytokines in pigadipose tissue which may influence local and centralmetabolism and growth. Thus, these reports supportthe idea that adipose tissue functions as an endocrineorgan.Morphological studies have revealed that adiposetissue is innervated by adrenergic nerve fibres (Hausmanand Richardson 1987). Further, immunocytochemicaldata revealed that most of the subpopulations of theadrenergic leptin receptor immuno-reactive (OBR-IR)neurones supplying fat tissue in the pig were positive forNPY and tyrosine hydroxylase immunoactivity (Czajaet al. 2002). Moreover, immuno-positive neurones forOBR were located in the paraventricular nucleus,ventromedial nucleus, anterior hypothalamic area, preopticarea, arcuate nucleus and supraoptic nucleus(Czaja et al. 2003). Collectively, these studies providemorphological data demonstrating that hypothalamicOBR containing neurones are transsynaptically connectedto the perirenal fat depot. Therefore, the aboveevidence supports a direct link between hypothalamicÓ 2008 No claim to original government works

Energy and Reproduction in the Pig 327Fig. 1. Schematic illustration of leptin coordination of energy homeostasisand neuroendocrine function: Leptin is secreted in response tochanges in energy balance and acts on the hypothalamus to controlfood intake, reproduction and adipocyte function. Positive energybalance increases blood leptin concentrations, suppresses expression ofneuropeptide Y (NPY) and agouti related peptide (AgRP), andupregulates proopiomelanocortin (POMC) and a-melanocyte stimulatinghormone (a-MSH). Negative energy balance decreases bloodleptin concentrations, stimulates expression of NPY and AgRP anddownregulates POMC and a-MSH. Thick black lines represent strongregulator tone and thin black lines weak regulator tone. Modified fromLoftus (1999)neurones in the regulation of fat metabolism andreproduction.Under changing metabolic states, which may increaseadipose tissue mass and circulating leptin concentrationsand stimulate catabolic pathways within the hypothalamus,conversely decreased adipose mass and reducedblood leptin concentrations stimulate anabolic pathwaysand inhibit catabolic pathways (Fig. 1). These twopathways comprise a number of genes that not onlyregulate appetite and energy balance but also impact thereproductive axis either directly in the case of NPY, orindirectly by altering adipocyte function and hence theexpression of genes and their protein products. Forexample, hypothalamic NPY and proopiomelanocortin(POMC) associated products a-melanocyte stimulatinghormone (a-MSH) and beta endorphin can affecthypothalamic GnRH release and subsequent feedingbehaviour (Barb et al. 1994, 1998, 2006b; Crown et al.2007). Peripherally, adipocyte leptin expression andsecretion can suppress appetite (Barb et al. 1998),stimulate hypothalamic GnRH release (Barb et al.2004) and reverse the inhibitory effect of energy deprivationon LH secretion (Whisnant and Harrell 2002).Thus, as previously stated, reproductive function ismetabolically gated, but mechanisms interfacing energystores or metabolic cues and fertility are not completelyunderstood.Identification of targets of leptin in the hypothalamusIt is well established that reproductive function ismetabolically gated. However, the mechanisms wherebyenergy stores and metabolic cues influence fertility areyet to be completely understood. The effects of leptinappear to be mediated through modulation of hypothalamicNPY expression (Campfield et al. 1996). In thepig, the presence of biologically-active OBR in thehypothalamus and pituitary (Lin et al. 2000) and thefact that leptin increased LH secretion from pig pituitarycells and GnRH release from hypothalamic tissue invitro (Barb et al. 2004) suggest that leptin acts throughthe hypothalamus. There is strong evidence from colocalizationof leptin receptor mRNA with NPY geneexpression that hypothalamic NPY is the primarypotential target for leptin in the pig (Czaja et al. 2002).Moreover, central administration of NPY suppressedLH secretion and stimulated feed intake and reversedthe inhibitory action of leptin on feed intake (Barb et al.2006b). However, NPY alone may not mediate theaction of leptin, since leptin failed to effect NPY releasefrom pig hypothalamic-preoptic area tissue fragments(Barb et al. 2004). Furthermore, metabolic signals mayin part be communicated to GnRH neurones via otherneuropeptides such as galanin-like peptide (Rich et al.2007), a-MSH (Crown et al. 2007), b-endorphin (Barbet al. 1994), or kisspeptin (Arreguin-Arevalo et al. 2007;Luque et al. 2007).The kisspeptins are potent stimulators of theGnRH ⁄ LH axis (Castellano et al. 2006; Arreguin-Arevalo et al. 2007; Luque et al. 2007). The kisspeptinsare a group of structurally related peptides that areproducts of the kisspeptin-1 (KiSS-1) gene (Ohtaki et al.2001; Kotani et al. 2001). Synthesized as a pre-prohormone,it is cleaved to liberate a 54 amino acid peptidewhich can be proteolytically processed (Takino et al.2003) to shorter variants, all of which share the sameamidated C-terminus and retain full biological activity.Kisspeptins acting through their cognate receptor,GPR54, are thought to be an important determinate inthe onset of puberty (Shahab et al. 2005; Smith andClarke 2007). Thus, hypothalamic kisspeptin and itsreceptor, GPR54, may serve as an essential gatekeeperof GnRH neurones and hence of reproductive function.Kisspeptins play a role in the timing of puberty onset(Tena-Sempere 2006c) and act on the gonadotropic axisvia the release of hypothalamic GnRH (Tena-Sempere2006a; Dungan et al. 2006). The hypothalamic expressionof KiSS-1 gene is under the control of sex steroids,and KiSS-1 neurones are involved in mediating thenegative and positive feedback effects of oestradiol ongonadotropin secretion (Smith et al. 2005, 2006). Moreover,the hypothalamic KiSS-1 system may also conveythe modulatory action of metabolic signals to theGnRH neurones (Tena-Sempere 2006b; Luque et al.2007). A recent report demonstrated that short-termfasting resulted in a decline in hypothalamic KiSS-1 andGPR54 mRNA levels at 12 and 24 h which preceded thereduction in GnRH gene expression at 48 h (Luqueet al. 2007). Moreover, leptin and not IGF-I or insulinstimulated KiSS-1 expression in mouse hypothalamiccell line N6. Furthermore, hypothalamic KiSS-1 expressionwas decreased in NPY null mice and this wasreversed by NPY administration (Luque et al. 2007).These reports support the idea that leptin and NPY arekey mediators of metabolic regulation of the hypothalamicKiSS-1 system and subsequent GnRH release.Lents et al. (2008) recently demonstrated that administrationof kisspeptin into the lateral ventricle of thebrain stimulated LH and follicle stimulating hormone(FSH) secretion but not GH in the pre-pubertal gilt,Ó 2008 No claim to original government works

328 CR Barb, GJ Hausman and CA LentsFig. 2. Schematic illustration of leptin coordination of energy homeostasisand neuroendocrine function: Leptin is secreted in response tochanges in energy balance. Leptin acts on the hypothalamus to controlfood intake, reproduction and, by suppressing expression of neuropeptideY (NPY) and agouti related peptide (AGRP) and upregulationof proopiomelanocortin (POMC), a-melanocyte stimulating hormone(a-MSH) and kisspeptin (KiSS)while peripheral administration of kisspeptin increasedserum concentrations of LH but not FSH or GH. Thesedata illustrate that kisspeptin can activate the brainpituitaryaxis and may be part of an importantmechanism regulating activation of the GnRH ⁄ LHrelease and initiating onset of puberty in swine. Moreoverit is tempting to speculate that the kisspeptinneuronal pathway may play a role in interfacingmetabolic status of the post-partum sow with theGnRH ⁄ gonadotropic axis and hence may affect thepost-weaning fertility of the sow.In conclusion evidence presented supports the conceptthat metabolic signals, which reflect changes in energybalance, affect both the hypothalamic GnRH ⁄ LH pulsegenerator and appetite. Changes in fat metabolism inresponse to changes in feed intake and energy balancealter adipocyte function and circulating concentrationsof IGF-I and leptin. A preponderance of evidence citedabove suggests that leptin may serve as the primarymetabolic signal interacting with neuropeptides such askisspeptin and NPY that link energy status with theneuroendocrine axis and subsequent reproduction(Fig. 2).ReferencesAherne FX, Kirkwood RN, 1985: Nutrition and sow prolificacy.J Reprod Fertil 33, 169–183.Ahima RS, Qi Y, Singhal NS, Jackson MB, Scherer PE, 2006:Brain adipocytokine action and metabolic regulation. Diabetes55, S145–S154.Almeida FR, Mao J, Novak S, Cosgrove JR, Foxcroft GR,2001: Effects of different patterns of feed restriction andinsulin treatment during the luteal phase on reproductive,metabolic, and endocrine parameters in cyclic gilts. J AnimSci 79, 200–212.Angell CA, Tubbs RC, Moore AB, Barb CR, Cox NM, 1996:Depressed luteinizing hormone response to estradiol in vivoand gonadotropin-releasing hormone in vitro in experimentallydiabetic swine. Domest Anim Endocrinol 13, 453–463.Arreguin-Arevalo JA, Lents CA, Farmerie TA, Nett TM, ClayCM, 2007: KiSS-1 peptide induces release of LH by a directeffect on the hypothalamus of ovariectomized ewes. AnimReprod Sci 101, 265–275.Barb CR, Kraeling RR, 2004: Role of leptin in the regulationof gonadotropin secretion in farm animals. Anim ReprodSci 82–83, 155–167.Barb CR, Chang W-C, Leshin LS, Rampacek GB, KraelingRR, 1994: Opioid modulation of gonadotropin releasinghormone release from the hypothalamic preoptic area in thepig. Domest Anim Endocrinol 11, 375–382.Barb CR, Kraeling RR, Rampacek GB, Dove CR, 1997:Metabolic changes during the transition from the fed to theacute feed-deprived state in prepuberal and mature gilts. JAnim Sci 75, 781–789.Barb CR, Yan X, Azain MJ, Kraeling RR, Rampacek GB,Ramsay TG, 1998: Recombinant porcine leptin reduces feedintake and stimulates growth hormone secretion in swine.Domest Anim Endocrinol 15, 77–86.Barb CR, Barrett JB, Kraeling RR, Rampacek GB, 1999: Roleof leptin in modulating neuroendocrine function: a metaboliclink between the brain-pituitary and adipose tissue.Reprod Domest Anim 34, 111–125.Barb CR, Barrett JB, Kraeling RR, Rampacek GB, 2001a:Serum leptin concentrations, luteinizing hormone andgrowth hormone secretion during feed and metabolic fuelrestriction in the prepuberal gilt. Domest Anim Endocrinol20, 47–63.Barb CR, Hausman GJ, Houseknecht KL, 2001b: Biology ofleptin in the pig. Domest Anim Endocrinol 21, 297–317.Barb CR, Kraeling RR, Rampacek GB, 2002: Metabolicregulation of the neuroendocrine axis in pigs. ReproductionSuppl 59, 203–217.Barb CR, Barrett JB, Kraeling RR, 2004: Role of leptin inmodulating the hypothalamic-pituitary axis in the pig.Domest Anim Endocrinol 26, 201–214.Barb CR, Hausman GJ, Ramsay TG, 2006a: Leptin in farmanimals. in: Castracane VD, Henson MC (eds), Leptin.Endocrinology Update Series. Springer Publishing Group,New York, NY, pp. 263–308.Barb CR, Kraeling RR, Rampacek GB, Hausman GJ, 2006b:The role of neuropeptide Y and interaction with leptin inregulating feed intake and luteinizing hormone and growthhormone secretion in the pig. Reproduction 131, 1127–1135.Booth PJ, Cosgrove JR, Foxcroft GR, 1996: Endocrine andmetabolic responses to realimentation in feed-restrictedprepuberal gilts: associations among gonadotropins, metabolichormones, glucose, and uteroovarian development. JAnim Sci 74, 840–848.van den Brand H, Dieleman SJ, Soede NM, Kemp B, 2000:Dietary energy source at two feeding levels during lactationof primiparous sows: I. Effects on glucose, insulin, andluteinizing hormone and on follicle development, weaningto-estrusinterval, and ovulation rate. J Anim Sci 78, 396–404.van den Brand H, Prunier A, Soede NM, Kemp B, 2001: Inprimiparous sows, plasma insulin-like growth factor-I canbe affected by lactational feed intake and dietary energysource and is associated with luteinizing hormone. ReprodNutr Dev 41, 27–39.Brendemuhl JH, Lewis AJ, Poe ER Jr, 1989: Influence ofenergy and protein intake during lactation on body compositionof primiparous sows. J Anim Sci 67, 1478–1488.Campfield LA, Smith FJ, Burn P, 1996: The OB protein(leptin) pathway – a link between adipose tissue mass andcentral neural networks. Horm Metab Res 28, 619–632.Castellano JM, Navarro VM, Fernandez-Fernandez R,Castano JP, Malagon MM, Aguilar E, Dieguez C, MagniP, Pinilla L, Tena-Sempere M, 2006: Ontogeny andÓ 2008 No claim to original government works

Energy and Reproduction in the Pig 329mechanisms of action for the stimulatory effect of kisspeptinon gonadotropin-releasing hormone system of the rat. MolCell Endocrinol 257–258, 75–83.Clowes EJ, Aherne FX, Foxcroft GR, Baracos VE, 2003:Selective protein loss in lactating sows is associated withreduced litter growth and ovarian function. J Anim Sci 81,753–764.Cox NM, Britt JH, 1988: Pulsatile administration of gonadotropinreleasing hormone to lactating sows: endocrinechanges associated with induction of fertile estrus. BiolReprod 27, 1126–1137.Crown A, Clifton DK, Steiner RA, 2007: Neuropeptidesignaling in the integration of metabolism and reproduction.Neuroendocrinology 86, 175–182.Czaja K, Lakomy M, Sienkiewicz W, Kaleczyc J, Pidsudko Z,Barb CR, Rampacek GB, Kraeling RR, 2002: Distributionof neurons containing leptin receptors in the hypothalamusof the pig. Biochem Biophys Res Commun 298, 333–337.Czaja K, Kraeling RR, Barb CR, 2003: Are hypothalamicneurons transsynaptically connected to porcine adiposetissue? Biochem Biophys Res Commun 311, 482–485.De Rensis F, Gherpelli M, Superchi P, Kirkwood RN, 2005:Relationship between backfat depth and plasma leptinduring lactation and sow reproductive performance afterweaning. Anim Reprod Sci 90, 95–100.Dungan HM, Clifton DK, Steiner RA, 2006: Minireview:kisspeptin neurons as central processors in the regulation ofgonadotropin-releasing hormone secretion. Endocrinology147, 1154–1158.Estienne MJ, Harper AF, Barb CR, Azain MJ, 2000:Concentrations of leptin in serum and milk collected fromlactating sows differing in body condition [In ProcessCitation]. Domest Anim Endocrinol 19, 275–280.Estienne MJ, Harper AF, Kozink DM, Knight JW, 2003:Serum and milk concentrations of leptin in gilts fed a highorlow-energy diet during gestation. Anim Reprod Sci 75,95–105.Frisch RE, 1984: Body fat, puberty and fertility. Biol Rev 59,161–188.Guerre-Millo M, 2004: Adipose tissue and adipokines: forbetter or worse. Diabetes Metab 30, 13–19.Hart HA, Azain MJ, Hausman GJ, Reeves DE, Barb CR,2007: Failure of short term feed restriction to affectluteinizing hormone and leptin secretion or subcutaneousadipose tissue expression of leptin in the prepuberal gilt. CanJ Anim Sci 87, 191–197.Hauner H, 2005: Secretory factors from human adipose tissueand their functional role. Proc Nutr Soc 64, 163–169.Hausman GJ, Richardson RL, 1987: Adrenergic innervationof fetal pig adipose tissue. Histochemical and ultrastructuralstudies. Acta Anat (Basel) 130, 291–297.Hausman GJ, Poulos SP, Richardson RL, Barb CR, Andacht T,Mynatt R, Kirk H, 2006: Secreted proteins and genes in fetaland neonatal pig adipose tissue and stromal-vascular cells.J Anim Sci 84, 1666–1681.Hausman GJ, Barb CR, Dean RG, 2007: Patterns of geneexpression in pig adipose tissue: transforming growthfactors, interferons, interleukins and other cytokines. JAnim Sci 85, 2445–2456.King RH, Martin GB, 1989: Relationships between proteinintake during lactation, LH levels and oestrous activity infirst-litter sows. Anim Reprod Sci 19, 283–292.King RH, Williams IH, 1984: The effect of nutrition on thereproductive performance of first litter sows 2. Protein andenergy intakes during lactation. Anim Prod 38, 249–256.Kojima CJ, Carroll JA, Matteri RL, Touchette KJ, Allee GL,2007: Effects of weaning and weaning weight on neuroendocrineregulators of feed intake in pigs. J Anim Sci 85,2133–2139.Koketsu Y, Dial GD, Pettigrew JE, Xue J, Yang H, Lucia T,1998: Influence of lactation length and feed intake onreproductive performance and blood concentrations ofglucose, insulin and luteinizing hormone in primiparoussows. Anim Reprod Sci 52, 153–163.Kotani M, Detheux M, Vandenbogaerde A, Communi D,Vanderwinden JM, Le Poul E, Brezillon S, Tyldesley R,Suarez-Huerta N, Vandeput F, Blanpain C, Schiffmann SN,Vassart G, Parmentier M, 2001: The metastasis suppressorgene KiSS-1 encodes kisspeptins, the natural ligands of theorphan G protein-coupled receptor GPR54. J Biol Chem276, 34631–34636.Kraeling RR, Barb CR, 1990: Hypothalamic control ofgonadotrophin and prolactin secretion in pigs. J ReprodFertil Suppl 40, 3–17.Lawrence CB, Turnbull AV, Rothwell NJ, 1999: Hypothalamiccontrol of feeding. Curr Opin Neurobiol 9, 778–783.Lents CA, Heidorn NL, Barb CR, Ford JJ, 2008: Central andperipheral administration of kisspeptin activates gonadotropinbut not somatotropin secretion in prepubertal gilts.Reproduction 135, 879–887.Leshin LS, Barb CR, Kiser TE, Rampacek GB, Kraeling RR,1994: Growth hormone-releasing hormone and somatostatinneurons within the porcine and bovine hypothalamus.Neuroendocrinology 59, 251–264.Lin J, Barb CR, Matteri RL, Kraeling RR, Chen X,Meinersmann RJ, Rampacek GB, 2000: Long form leptinreceptor mRNA expression in the brain, pituitary, andother tissues in the pig. Domest Anim Endocrinol 19,53–61.Lin J, Barb CR, Kraeling RR, Rampacek GB, 2001: Developmentalchanges in the long form leptin receptor andrelated neuropeptide gene expression in the pig brain. BiolReprod 64, 1614–1618.Loftus TM, 1999: An adipocyte-central nervous systemregulatory loop in the control of adipose homeostasis. CellDev Biol 10, 11–18.Luque RM, Kinemam RD, Tena-Sempere M, 2007: Regulationof hypothalamic expression of KiSS-1 and GPR54genes by metabolic factors: analyses using mouse modelsand a cell line. Endocrinology 148, 4601–4611.Lutz JB, Rampacek GB, Kraeling RR, Pinkert CA, 1984:Serum luteinizing hormone and estrogen profiles beforepuberty in the gilt. J Anim Sci 58, 686–691.Mao J, Zak LJ, Cosgrove JR, Shostak S, Foxcroft GR, 1999:Reproductive, metabolic, and endocrine responses to feedrestriction and GnRH treatment in primiparous, lactatingsows. J Anim Sci 77, 724–735.Matteri RL, 2001: Overview of central targets for appetiteregulation. J Anim Sci 79, E148–E158.Muller EE, Locatelli V, Cocchi D, 1999: Neuroendocrinecontrol of growth hormone secretion. Physiol Rev 79, 511–607.NRC, 1998: Nutrient Requirements in Swine. National AcademyPress, Washington, DC.Ohtaki T, Shintani Y, Honda S, Matsumoto H, Hori A,Kanehashi K, Terao Y, Kumano S, Takatsu Y, Masuda Y,Ishibashi Y, Watanabe T, Asada M, Yamada T, Suenaga M,Kitada C, Usuki S, Kurokawa T, Onda H, Nishimura O,Fujino M, 2001: Metastasis suppressor gene KiSS-1 encodespeptide ligand of a G-protein-coupled receptor. Nature 411,613–617.Pressing A, Dial GD, Esbenshade KL, Stroud CM, 1992:Hourly administration of GnRH to prepubertal gilts:endocrine and ovulatory responses from 70 to 190 days ofage. J Anim Sci 70, 232–242.Prunier A, Quesnel H, 2000: Nutritional influences on thehormonal control of reproduction in female pigs. LivestProd Sci 63, 1–16.Ó 2008 No claim to original government works

330 CR Barb, GJ Hausman and CA LentsPrunier A, Dourmad JY, Etienne M, 1993: Feeding level,metabolic parameters and reproductive performance ofprimiparous sows. Livest Prod Sci 37, 185–196.Qian H, Barb CR, Compton MM, Hausman GJ, Azain MJ,Kraeling RR, Baile CA, 1999: Leptin mRNA expression andserum leptin concentrations as influenced by age, weight andestradiol in pigs. Domest Anim Endocrinol 16, 135–143.Quesnel H, Prunier A, 1998: Effect of insulin administrationbefore weaning on reproductive performance in feedrestricted primiparous sows. Anim Reprod Sci 51, 119–129.Quesnel H, Pasquier A, Mounier AM, Louveau I, Prunier A,1998: Influence of feed restriction in primiparous lactatingsows on body condition and metabolic parameters. ReprodNutr Dev 38, 261–274.Quesnel H, Mejia-Guadarrama CA, Dourmad JY, Farmer C,Prunier A, 2005: Dietary protein restriction during lactationin primiparous sows with different live weights at farrowing:I. Consequences on sow metabolic status and litter growth.Reprod Nutr Dev 45, 39–56.Rich N, Reyes P, Reap L, Goswami R, Fraley GS, 2007: Sexdifferences in the effect of prepubertal GALP infusion ongrowth, metabolism and LH secretion. Physiol Behav 92,814–823.Schwartz MW, Seeley RJ, Campfield LA, Burn P, Baskin DG,1996: Identification of targets of leptin action in rathypothalamus. J Clin Invest 98, 1101–1106.Shahab M, Mastronardi C, Seminara SB, Crowley WF, OjedaSR, Plant TM, 2005: Increased hypothalamic GPR54signaling: a potential mechanism for initiation of pubertyin primates. Proc Natl Acad Sci U S A 102, 2129–2134.Shields RG, Mahan DC, Maxson PF, 1985: Effect of dietarygestation and lactation protein levels on reproductionperformance and body composition of first litter femaleswine. J Anim Sci 60, 179–189.Smith JT, Clarke IJ, 2007: Kisspeptin expression in the brain:catalyst for the initiation of puberty. Rev Endocr MetabDisord 8, 1–9.Smith JT, Dungan HM, Stoll EA, Gottsch ML, Braun RE,Eacker SM, Clifton DK, Steiner RA, 2005: Differentialregulation of KiSS-1 mRNA expression by sex steroidsin the brain of the male mouse. Endocrinology 146, 2976–2984.Smith JT, Clifton DK, Steiner RA, 2006: Regulation of theneuroendocrine reproductive axis by kisspeptin-GPR54signaling. Reproduction 131, 623–630.Spurlock ME, Frank GR, Cornelius SG, Ji S, Willis GM,Bidwell CA, 1998: Obese gene expression in porcine adiposetissue is reduced by food deprivation but not by maintenanceor submaintenance intake. J Nutr 128, 677–682.Takino T, Koshikawa N, Miyamori H, Tanaka M, Sasaki T,Okada Y, Seiki M, Sato H, 2003: Cleavage of metastasissuppressor gene product KiSS-1 protein ⁄ metastin by matrixmetalloproteinases. Oncogene 22, 4617–4626.Tena-Sempere M, 2006a: GPR54 and kisspeptin in reproduction.Hum Reprod Update 12, 631–639.Tena-Sempere M, 2006b: KiSS-1 and reproduction: focus onits role in the metabolic regulation of fertility. Neuroendocrinology83, 275–281.Tena-Sempere M, 2006c: The roles of kisspeptins andG protein-coupled receptor-54 in pubertal development.Curr Opin Pediatr 18, 442–447.Tokach MD, Pettigrew JE, Dial GD, Wheaton JE, CrookerBA, Johnston LJ, 1992: Characterization of luteinizinghormone secretion in the primiparous, lactating sow: relationshipto blood metabolites and return-to-estrus interval.J Anim Sci 70, 2195–2201.Tummaruk P, Lundeheim N, Einarsson S, Dalin AM, 2001:Effect of birth litter size, birth parity number, growth rate,backfat thickness and age at first mating of gilts on theirreproductive performance as sows. Anim Reprod Sci 66,225–237.Wade GN, Schneider JE, Li HY, 1996: Control of fertility bymetabolic cues. Am J Physiol 270, E1–E19.Weldon WC, Lewis AJ, Louis GF, Kovar JL, Giesemann MA,Miller PS, 1994: Postpartum hypophagia in primiparoussows: I. Effects of gestation feeding level on feed intake,feeding behavior, and plasma metabolite concentrationsduring lactation. J Anim Sci 72, 387–394.Whisnant CS, Harrell RJ, 2002: Effect of short-term feedrestriction and refeeding on serum concentrations of leptin,luteinizing hormone and insulin in ovariectomized gilts.Domest Anim Endocrinol 22, 73–80.Whitley NC, Payne DB, Zhang H, Cox NM, 1998: Theinfluence of insulin administration after weaning the firstlitter on ovulation rate and embryo survival in sows.Theriogenology 50, 479–485.Whittemore CT, 1996: Nutrition reproduction interactions inprimiparous sows. Livest Prod Sci 46, 65–83.Williams LT, 1989: Signal transduction by the platelet-derivedgrowth factor receptor. Science 243, 1564–1570.Author’s address (for correspondence): C Richard Barb, USDA ⁄ ARS,Richard B, Russell Agriculture Research Center, PO Box 5677,Athens, GA 30604, USA. E-mail: richard.barb@ars.usda.govConflict of interest: This research was supported by USDA funds(C.R.B.; G.J.H.) and State funds allocated to the Georgia AgriculturalExperiment Station (C.A.L.) and in part by a National ResearchInitiative Competitive Grant (no. 2005-35023-15949 and 2005-35203-16852) from the USDA Cooperative State Research, Education, andExtension Service awarded to C.A.L. Mention of trade name,proprietary product, or specific equipment does not constitute aguarantee or warranty by the US Department of Agriculture or TheUniversity of Georgia and does not imply its approval to the exclusionof other products which may be suitable. The authors declare thatthere is no conflict of interest that would prejudice their impartiality ofthis scientific work.Ó 2008 No claim to original government works

Reprod Dom Anim 43 (Suppl. 2), 331–337 (2008); doi: 10.1111/j.1439-0531.2008.01181.xISSN 0936-6768Somatic Cell Nuclear Transfer in HorsesCesare Galli 1,2 , Irina Lagutina 1 , Roberto Duchi 1 , Silvia Colleoni 1 and Giovanna Lazzari 11 Laboratorio di Tecnologie della Riproduzione, Istituto Sperimentale Italiano Lazzaro Spallanzani, CIZ srl, Cremona, Italy; 2 Dip. ClinicoVeterinario, Universita` di Bologna, Cremona, ItalyContentsThe cloning of equids was achieved in 2003, several years afterthe birth of Dolly the sheep and also after the cloning ofnumerous other laboratory and farm animal species. The delaywas because of the limited development in the horse of moreclassical-assisted reproductive techniques required for successfulcloning, such as oocyte maturation and in vitro embryoproduction. When these technologies were developed, theapplication of cloning also became possible and cloned horseoffspring were obtained. This review summarizes the maintechnical procedures that are required for cloning equids andthe present status of this technique. The first step is competentoocyte maturation, this is followed by oocyte enucleation andreconstruction, using either zona-enclosed or zona-freeoocytes, by efficient activation to allow high cleavage ratesand finally by a suitable in vitro embryo culture technique.Cloning of the first equid, a mule, was achieved using anin vivo-matured oocytes and immediate transfer of the reconstructedembryo, i.e. at the one cell stage, to the recipientoviduct. In contrast, the first horse offspring was obtainedusing a complete in vitro procedure from oocyte maturation toembryo culture to the blastocyst stage, followed by nonsurgicaltransfer. Later studies on equine cloning report highefficiency relative to that for other species. Cloned equidoffspring reported to date appear to be normal and those thathave reached puberty have been confirmed to be fertile. Insummary, horse cloning is now a reproducible technique thatoffers the opportunity to preserve valuable genetics andnotably to generate copies of castrated champions andtherefore, offspring from those champions that would beimpossible to obtain otherwise.IntroductionCloning mammals by somatic cell nuclear transfer(SCNT) has become a common technology in recentyears, following the work of Wilmut and co-workers(Wilmut et al. 1997). Today, a decade later, mostdomestic and laboratory species have been cloned,including the horse (Campbell et al. 2007). Initially, infarm animals, the application of SCNT was focussed onthe multiplication of superior genotypes of high geneticvalue. Very soon, yet, it became evident that thetechnique could offer, mainly because of its peculiarfailures, an important biological model for basicresearch as well as a wealth of opportunities forbiomedical research (Hochedlinger and Jaenisch 2006).It is in this latter field that most of the potentialapplications lie today. They range from the creation ofgenetically modified large animals to the derivation ofgenetically matched embryonic stem cells derived fromcloned embryos (Rideout et al. 2002).With regard to Equids, the birth of three mule foalscloned from foetal cells (Woods et al. 2003) and onehorse foal cloned from adult somatic cells (Galli et al.2003a) were reported in 2003 and additional cloned foalshave also been documented (Lagutina et al. 2005;Hinrichs et al. 2006, 2007). The relatively late applicationof SCNT to horse reproduction is a consequence ofthe limited information available in the publishedstudies on assisted reproductive techniques in Equids,in particular oocyte maturation, activation and in vitroculture of early pre-implantation embryos. The cloningof the mule by Woods and co-workers (Woods et al.2003) relied on the use of in vivo matured oocytes thatwere transferred to the oviducts of recipient maresimmediately after nuclear transfer and activation. Incontrast, Galli (Galli et al. 2003a), Lagutina (Lagutinaet al. 2005) and Hinrichs (Hinrichs et al. 2006, 2007)have all reported cloning of horse embryos up to theblastocyst stage carried out completely in vitro.In general, few studies are available in the publishedstudies on equine in vitro embryo production and it isonly recently that reports have been published oncomplete in vitro production of equine pre-implantationembryos by means of in vitro oocyte maturation,fertilization by intracytoplasmic sperm injection (ICSI)and in vitro culture (Galli et al. 2002a; Lazzari et al.2002a; Choi et al. 2004; Hinrichs et al. 2005). Theselatter reports demonstrate the feasibility of applying acompletely in vitro procedure to obtain horse ICSIblastocysts able to establish a pregnancy (Galli et al.2002a,b; Lazzari et al. 2002a) and develop into live termoffspring (Galli et al. 2007).One additional reason for the limited development ofassisted reproduction techniques in Equids is the lack ofinterest demonstrated by the horse industry. SCNT, inparticular, has received a sceptical welcome in mostequestrian disciplines. Some of the mayor players in thefield (such as the American Quarter Horse Associationand the Jockey Club) are opposed to the use of cloningand will not register offspring produced by this technique(Church 2006).The refinement of the in vitro culture conditionssuitable for equine oocyte maturation and embryogrowth (Hinrichs et al. 2005; Galli et al. 2007), thedevelopment of adequate horse oocyte activation protocols(Lazzari et al. 2002b) and the application of apiezoelectric device (Westhusin et al. 2003) or zona-freemanipulation for embryo reconstruction (Booth et al.2001; Oback et al. 2003; Lagutina et al. 2007), have allbeen the fundamental steps for defining a successfulin vitro procedure for SCNT in the horse.Although nuclear transfer, in general, and SCNT, inparticular, are routine techniques in several laboratoriesworldwide, both in farm animals and in laboratoryspecies, they are very demanding. The major hurdle inÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

332 C Galli, I Lagutina, R Duchi, S Colleoni and G Lazzaricloning, and not only for beginners is to master all thelengthy and complicated procedures involved in SCNTto a high level of efficiency and reproducibility in orderto obtain scientifically meaningful results (Galli et al.2003b; Ritchie 2006).In principle, all techniques used today in mammalsare based on those described by Willadsen (Willadsen1986). Basically, there is a requirement for maturedoocytes, perhaps in large numbers and of good quality,the oocytes then need to be enucleated before the donorcell is fused or injected in the enucleated oocyte andfinally the reconstructed embryo must be activated. Atthis point, embryos can be transferred to the oviduct ofrecipients for further development. Yet, in large animals,it is preferable to culture in vitro the clonedembryos up to the blastocyst stage, so that a nonsurgicaluterine transfer can be performed (Galli et al.2003b).Source of OocytesOocytes can be harvested from the ovaries of live donorsby ovum pick up (OPU) or from the ovaries ofslaughtered mares (Zhang et al. 1989), the latter beingthe only economically sustainable source of oocytes forSCNT. On average, it is possible to recover three to fouroocytes per abattoir ovary when compared with oocyteretrieval from live donors by OPU, in which, therecovery is slightly more variable and can be from threeto six oocytes per session of OPU (Galli et al. 2007). Thematuration rate of horse oocytes is also quite variable,averaging between 25% and 70% in published studies(Carneiro et al. 2001; Dell’Aquila et al. 2001, 2003;Bogh et al. 2002; Choi et al. 2002; Lorenzo et al. 2002).The recovery of oocytes from abattoir horse ovariesrequires incision and scraping of the follicle wall with acurette and extensive washing to detach the cumulusoocytecomplexes (COCs). Another interesting aspect,peculiar to the horse, is the frequent collection ofoocytes with expanded cumulus; in our laboratory thisaccounts for approximately a third of the recoveredCOCs. In other species such as ruminants and pigs, anexpanded cumulus is linked to collection from atreticfollicles and these oocytes are generally discardedimmediately because of their extremely low developmentalcapacity (de Loos et al. 1989). In the horse, yet,oocytes with an expanded cumulus mature normally andhave normal developmental competence. Table 1 presentsthe data from a large study conducted in ourlaboratory to measure the efficiency of oocyte collectionfrom abattoir ovaries and the expected maturation ratefrom compact and expanded COCs. The data indicatethat each ovary contains 5.3 follicles on an average witha diameter above 5 mm that are suitable for oocytecollection. If these follicles are scraped and washed, therecovery is 3.8 oocytes corresponding to approximately70% efficiency. The need to scrape the follicles derivesfrom the tight connections between the cumulus and themembrana granulosa and between the latter and thefollicle wall. Once the oocytes are collected and selectedon the basis of cumulus morphology, they are transferredto maturation medium and allowed to mature for24 h. In our culture conditions, as shown in Table 1(Galli et al. 2007), the maturation rate ranges from51.1% to 60% for compact and expanded COCs,respectively. This difference is statistically significantand in agreement with other studies (Hinrichs andWilliams 1997; Hinrichs and Schmidt 2000), in which,oocytes with expanded cumulus were found to be morecapable to complete maturation than were oocytes witha compact cumulus. Yet, the ability of oocytes withexpanded cumuli to develop to the blastocyst stage wasnot statistically different from oocytes with a compactcumulus (see Table 2). In calculating the maturationrate, the authors included the large number of degeneratingoocytes that are found after IVM. In fact, at thetime of collection, the oocytes that will degenerateduring maturation cannot be identified; this is anotherpeculiarity of the equine species when oocytes arecollected from abattoir ovaries. At the end of maturation,after the removal of the surrounding somatic cells,degenerated oocytes are easily detected. If these oocytesthat degenerate in culture are not included in thecalculation, the percentage of mature oocytes increasesconsiderably into the range of 70–85% and is comparableto that obtained from OPU oocytes from livemares. As no lysis is observed with OPU oocytes, thisfinding implies that the postmortem modificationsoccurring in the large equine ovaries are responsiblefor the degeneration of abattoir oocytes. Indeed, whenequine oocytes were placed into maturation immediatelyafter slaughter, the rate of maturation was higher andthe rate of degeneration lower than those for oocytesrecovered after transport of ovaries to the laboratoryTable 2. Effect of cumulus morphology on the developmental competenceof equine oocytes matured in vitro, fertilized by ICSI andcultured in vivo in the sheep oviduct (Galli et al. 2007)CumulusmorphologyNo. injected(MII)No.cleavedCleavagerate (%)No. ofblastocysts% Blastocyst ⁄MIIinjectedCompact 150 110 73.3 a 43 28.6 aExpanded 73 43 58.9 b 13 17.8 aChi square test: values within columns with different letters differ (p < 0.05).Table 1. Maturation competence of horse oocytes derived from expanded or compact COCs (Galli et al. 2007)No. ofovariesNo. of follicles(per ovary)No. of COCsexpanded(per ovary)No. of COCscompact(per ovary)No. of COCs matured after IVMNo. of COCs degenerated after IVMExpanded (%) Compact (%) Expanded (%) Compact (%)603 3204 (5.3) 590 (1) 1672 (2.8) 354 (60.0) a 855 (51.1) b 177 (30.0) 558 (33.4)Chi square test: values with different letters differ (p < 0.05).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Cloning in Horses 333Table 3. Effect of maturation media on maturation, cleavage rate andICSI embryo development (Galli et al. 2007)MaturationmediumNo. of No. ofoocytes degeneratedNo. ofmetaphaseII (%)No. ofinjectedNo. ofcleaved(% ofinjected)No. ofblastocysts(% ofinjected)TCM 199 434 105 205 (47.2 a) 191 111 (58.1 a) 23 (12.0 a)DMEM ⁄ F12 338 71 159 (45.6 a) 140 108 (77.1 b) 37 (26.4 b)Replicates = 6.Chi square test. Numbers within columns with different letters differ (p < 0.05).Table 4. Effect of maturation media on maturation, cleavage rate andNT embryo development (Galli et al. 2007)MaturationmediumNo. ofoocytesNo. oflysedNo. ofmetaphaseII (%)No. ofNTembryosNo. ofcleaved(% of NT)No. ofblastocysts(% of NT)TCM 199 164 35 64 (39.0 a) 41 38 (92.7 a) 4 (9.77 a)DMEM ⁄ F12 166 35 67 (40.4 a) 47 46 (97.9 a) 13 (27.7 b)Chi square test. Numbers within columns with different letters differ (p < 0.05).(Hinrichs et al. 2005). In that study, high rates ofdegeneration in culture were seen only in oocytes with acompact cumulus; evaluation of oocyte chromatinconfigurations led to the conclusion that these oocyteswere juvenile and that their chromatin was damagedduring storage within the ovary.It is known that in vitro maturation conditionsdramatically influence the development of fertilizedoocytes to the blastocyst stage. Different media are usedfor maturation of equine oocytes, although the mostcommon remains medium TCM199, probably becauseof its widespread use for bovine oocyte maturation. Inthe last 2 years, the authors have tested other mediaincluding DMEM-F12 supplemented with 15% serumreplacement (Invitrogen sri, San Giuliano Milanese,Italy) instead of TCM 199 supplemented with 10%foetal calf serum (FCS). Hormones and growth factorswere added to both media as reported earlier (Lagutinaet al. 2005). We found a clear advantage in the use of aDMEM-F12 based medium not only on cleavage ratebut also on embryo development to the blastocyst stage(Table 3). Preliminary experiments have confirmed similarfindings in horse SCNT (Table 4). This positiveeffect is evident also on blastocyst development, which issignificantly higher in the DMEM-F12 group than in theTCM199 group. This remarkable improvement increasesthe efficiency of equine embryo production andmakes it possible to use a totally in vitro systemfollowing OPU in the horse. The time required forcompleting IVM ranges from 22 to 24 h for expandedoocytes and 26–36 h for compacted onesSource of Donor CellsA variety of donor cells from different origin (foetal andadult) and various tissues have been used for SCNTwithout observing significant differences in the overallefficiency (Oback and Wells 2002). In the horse, foetalfibroblasts (Li et al. 2003; Lagutina et al. 2005; WoodsTable 5. Cleavage and blastocyst development after nuclear transferwith confluent or roscovitine-treated fibroblasts (Hinrichs et al. 2006)TreatmentNo. ofinjected(MII)No. of Cleavage No. of D +cleaved rate 7 blastocystsNo. of D +8 blastocystsTotalblastocyst(%)Confluent 55 40 73 1 1 2 (3.6)Roscovitine 56 44 79 0 2 2 (3.6)et al. 2003) as well as adult fibroblasts and granulosacells have been used (Vanderwall et al. 2004; Lagutinaet al. 2005). For recovering adult fibroblasts, skinbiopsies are taken under local anaesthesia on the neckor the chest of the donor horse. The recovered tissue isthen minced into small pieces that are plated inTCM199 ⁄ DMEM (1:1) culture medium supplementedwith 10% FCS. In 2–3 weeks, the fibroblasts outgrowthe tissue and form a confluent monolayer. Uponsubsequent sub-culture and expansion, the cells arefrozen for later use or prepared for the SCNT procedureat early passages (Li et al. 2003). There is a considerablevariation in the success rate of horse SCNT because ofthe specific cell line source of donor nuclei (Galli et al.2003a; Lagutina et al. 2005) as reported for other species(Oback and Wells 2002). The cell cycle of the donor cellis crucial for success and G0 or G1 stages have beenused in horses. This stages can be achieved throughserum starvation (reduction of FCS to 0.5%), growth toconfluency (contact inhibition) or the use of kinaseinhibitors like Roscovitine (Table 5) (Oback and Wells2002; Hinrichs et al. 2006, 2007). Given the relativelylow number of experiments performed, it cannot beascertained if any of these approaches is better than theother.Embryo Reconstruction and ActivationIn our laboratory, the oocytes are stripped of theircumulus after 22–24 h of maturation. First, by pipettingduring exposure to hyaluronidase; second, in 0.25%trypsin for 1.5 min and finally, in SOF-Hepes containing10% FCS. For reconstruction of NT-embryos, theauthors use both zona-enclosed and zona-free methods(Oback et al. 2003). For the zona-enclosed method,oocytes with an extruded polar body are stained withHoechst 33342 in the presence of cytochalasin B(5 lg ⁄ ml). Enucleation is performed by the aspirationof the polar body and metaphase II plate in a minimalvolume of ooplasm under UV light. All manipulationsare carried out in SOF-Hepes with 6 mg ⁄ ml BSA(Gardner et al. 1994), except fusion. Individual nucleardonor cells are then transferred into the perivitellinespace of zona-enclosed cytoplasts using the enucleationpipette. The cytoplast–karyoplast couplets are transferredinto 0.3 M mannitol (50 lM Ca and 100 lM Mg)solution and fused one or two times within 15–30 mininterval by a double DC-pulse of 1.8–2.4 kV ⁄ cm appliedfor 30 ls at 26–27 h of maturation. Because of theheterogeneous characteristics of the horse zona pellucida,the fusion rate is often inconsistent or low. Li et al.(2002) have found that even high voltage DC pulses of2.2–2.5 kV ⁄ cm were able to fuse less than 60% ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

334 C Galli, I Lagutina, R Duchi, S Colleoni and G Lazzaricouplets, the cleavage rate was 35–50% and onlyapproximately 2.5% of NT-embryos reached the blastocyststage. Hinrichs (Hinrichs et al. 2006, 2007) usedpiezoelectric micromanipulation both to enucleate themetaphase oocyte and to inject the somatic cell directlyinto the enucleated oocyte, thus overcoming the limitationof cell fusion. The horse oocyte tolerates themicroinjection procedure very well as demonstrated alsoby the success of the ICSI procedure (Galli et al. 2007).For the zona-free method, the zona pellucida ofoocytes with extruded polar bodies is digested with0.5% pronase in PBS. Zona-free oocytes are enucleatedunder UV light with a blunt micropipette (with perpendicularbreak). All manipulations are performed inSOF-Hepes containing 10% FCS. Subsequently, zonafreecytoplasts are individually washed for few secondsin 300 lg ⁄ ml phytohemagglutinin P in TCM 199-Hepesand then quickly dropped over a single donor cell (Vajtaet al. 2003) settled to the bottom of a drop of TCM 199with 0.5% of FCS. Formed cell couplets are washed in0.3 M mannitol (50 lM Ca and 100 lM Mg) solutionand fused one or two times at 15–30 min intervals by asingle DC-pulse of 1.2 kV ⁄ cm applied for 30 ls at 26–27 h of maturation. With such an approach, the fusionrate approximates 100% and because of the lowerelectric field required, both cleavage rate and postcleavagedevelopment to the blastocyst stage are higher(Lagutina et al. 2005, 2007). Activation is usuallyperformed 2–4 h after fusion. In our experience, horseoocytes are difficult to activate and the standardprotocols used in ruminants do not yield a high rate ofcleavage. The authors found that the use of ionomycin(a calcium ionophore) followed by the synergistic effectof 6-DMAP (1 mM) and Cycloheximide (5 lg ⁄ ml)resulted in over 90% of parthenogenetic activation(Lazzari et al. 2002b; Galli et al. 2007) and cleavage ofcloned embryos (Lagutina et al. 2005). Woods (Woodset al. 2003) increased the calcium level during activationof in vivo matured oocytes. Hinrichs (Hinrichs et al.2006, 2007) extensively compared different activationprotocols including the injection of sperm extract.Sperm extract in combination with ionomycin, whilenot statistically different from other treatments, providedthe highest blastocyst development rate althoughthis was not superior to that reported by Lagutina(Lagutina et al. 2005) using the association of ionomycinwith both 6-DMAP and cycloheximide.Embryo CultureThe progress in in vitro maturation and ICSI technologyhas increased efforts to design suitable culture systemsfor early cleavage-stage embryos. Very little is knownabout the specific in vitro culture requirements of equineembryos. The assessment of culture conditions has beencarried out with embryos produced in vitro by ICSI andusing various media. Many different culture conditionshave been reported for pre-implantation development ofICSI fertilized horse oocytes, including defined mediasuch as G1.2 (Choi et al. 2002), DMEM-F12 and CZB(Choi et al. 2004) and modified SOF (Galli et al. 2002a).Earlier work evaluated co-culture with somatic cellsincluding Vero cells (Dell’Aquila et al. 1997), oviductepithelial cells (Battut et al. 1991), cumulus cells (Liet al. 2001), granulosa cells (Rosati et al. 2002) orculture in conditioned media (Choi et al. 2001). In mostof these systems, yet, the blastocyst rates remained low,ranging from 4% to 16% of injected oocytes. Incontrast, the culture of presumptive zygotes followingICSI, in vivo either in the mare oviduct or in thesurrogate sheep oviduct, allowed much higher development,being approximately 30% (Galli et al. 2002b,2007; Lazzari et al. 2002b; Tremoleda et al. 2003; Choiet al. 2004) (Table 6). When cell number counts werecompared among in vivo produced embryos and thoseproduced by in vitro culture in a modified SOF medium,both on day 7 of development, in vitro producedembryos had significantly lower cell numbers, resemblinga day 5 embryo rather than a day 7 (Tremoledaet al. 2003). This difference is now taken into accountwhen embryos are transferred to synchronized recipients(see Section ‘Embryo Transfer and Offspring’).The results reported above where obtained using aTCM 199-based medium for oocyte maturation. Interestingly,the authors found that maturing the oocytes ina DMEM-F12-based medium allows to increase blastocystdevelopment to a much closer rate to in vivo culture,being in the range of 26% when compared with 27–50%for the sheep oviduct system, both for ICSI and forSCNT embryos (Galli et al. 2007). Yet, other laboratorieshave achieved equivalent blastocyst development(25–35%) with oocytes matured in M199, usingDMEM-F12 medium for embryo culture under a mixedgas atmosphere (Hinrichs et al. 2005; Choi et al.2006a,b).In our laboratory, equine embryos are cultured in20 ll microdrops under oil with up to 20 embryos perdrop of mSOF. The culture of zona-free embryos isperformed individually in 5 ll drops or in the WOWsystem (Vajta et al. 2000). Fifty per cent of the media isreplaced on D 4 with mSOF and on D6 withDMEM ⁄ F12 supplemented with 5%FCS and 5% SR(serum replacement).The residual difference in embryo development ratesbetween the in vitro and in vivo culture environmentsindicates the need for further improvement of theTable 6. Oocyte recovery rate by OPU and effect of in vitro or in vivo sheep oviduct culture on embryos development (Galli et al 2007)No. ofOPUsNo. of oocytes(no. per OPU)No. metaphase II(% of oocytes)(no. per OPU)No. of cleaved(% of injected)(no. per OPU)No. of comp.morulae ⁄ blastocysts(% of injected) (no. per OPU)Sheep oviduct culture 20 60 (3.0) 46 (76.7) (2.3) 41 (89.1% a) (2.1) 23 (50.0 a) (1.2)In vitro culture 12 46 (4.1) 36 (73.5) (3.0) 25 (69.4% a) (2.1) 5 (13.9 b) (0.4)Student’s t-test. Numbers within columns with different letters are significantly different (p < 0.05).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Cloning in Horses 335embryo culture system. In our experience, the viabilityof cloned embryos is somewhat lower than their IVFcounterpart and for this reason more than one SCNTembryo is transferred into each recipient in our laboratory.Embryo Transfer and OffspringTo date, a few groups have reported the birth of foalsfollowing ICSI of in vitro matured oocytes, after eithersurgical transfer of early embryos to the oviduct(Cochran et al. 1998) or in vitro embryo culture to theblastocyst stage and transcervical transfer to the uterus(Li et al. 2001; Hinrichs et al. 2005; Galli et al. 2007).This has been true also for SCNT embryos where to datethree labs have reported in the scientific literature, thebirth of a dozen equids following oviduct transfer ofearly cleaving embryos (Woods et al. 2003) or uterinetransfer of blastocyst stage embryos (Galli et al. 2003a;Lagutina et al. 2005; Hinrichs et al. 2006, 2007). Ashorse cloning is entering the commercial arena, additionalanimals are currently being cloned or have beencloned by commercial companies. While one suchcompany has announced the birth of cloned foals tothe popular press, information on the methods used andtheir efficiency is not in the public domain.In our laboratory, recipient mares are examined twoto three times a week by ultrasound to determine the dayof ovulation. A total of 3–7 (mostly 5) days afterovulation, each mare receives up to four nuclear transferembryos at the blastocyst stage on day 7 or day 8 ofdevelopment by non-surgical transfer. On day 17, afterovulation, the animals are examined for pregnancydiagnosis and thereafter, are scanned weekly throughoutthe first trimester of pregnancy and later at monthlyintervals until foaling.Table 7 summarizes the results reported in thepublished studies for the transfer of SCNT equidembryos. The work of Woods and Vanderwall for muleand horse SCNT, respectively, refer to the surgicaltransfer to the recipient mare oviduct of one-cellnuclear transfer embryos. Results indicate a total ofthree mule foals born from 305 zygotes transferred, outof 21 pregnancies diagnosed (Woods et al. 2003) andno live foals from 63 zygotes transferred and 7pregnancies diagnosed (Vanderwall et al. 2004). Inour laboratory, the authors have transferred 118embryos (morulae or blastocysts), fresh or frozen, into46 recipients, obtaining 13 pregnancies with oneembryonic vesicle each and 3 live foals, i.e. an embryosurvival rate of 11% and a foaling rate of 23% (3 ⁄ 13)(Galli et al. 2003a; Lagutina et al. 2005). Hinrichs et al.(2006, 2007) reported the transfer of a total of 42blastocyst-stage embryos for 20 pregnancies and 11 livefoals. In our experience, the viability of clonedembryos is lower than for ICSI embryos. In fact,during the same period, in which the SCNT embryoswere transferred a parallel study was conducted transferringICSI embryos both from abattoir and OPUoocytes cultured under the same conditions (Galli et al.2007). Overall, from 34 embryos produced by ICSItransferred into 21 recipients, the authors obtained 13pregnancies at day 17, indicating a survival rate of38% within this study. The best result involved oocytescollected by OPU, with embryos cultured in vitro andcryopreserved, in which 9 vesicles were observed onday 17 out of 13 embryos transferred into 11 recipients.In this case, the survival rate after transfer was 69%and the subsequent foaling rate was 88% (7 ⁄ 8) ofestablished pregnancies. Therefore, results confirm thelower viability of SCNT embryos when compared withfertilized embryos as already described in other species.Pregnancy losses with equine SCNT embryos occurthroughout gestation (Table 7) as described for ruminantsbut at a much lower rate. No large offspringsyndrome has been described in equids. The femalehorse produced by Galli (Galli et al. 2003a) (Fig. 1) iscurrently 4 years old and its foal was born on the 17 thMarch 2008 and it is normal and in good health. Themale produced by Lagutina (Lagutina et al. 2005) hasproduced semen for artificial insemination and it hasconfirmed to be fertile. Altogether, these results suggestthat cloned horses are reproductively normal, as hasbeen shown in other species.Table 7. Pre- and post-implantation development of SCNT embryosreported in the published studiesAuthorNo. ofoocytesfusedNo. ofoocytescleavedNo. ofblastocystsNo. oftransferredNo. of No. ofpregnancies foalsWoods et al.(2003)Galli et al.(2003)Vanderwall et al.(2004)Lagutina et al.(2005)Hinrichs et al.(2006)Hinrichs et al.(2007)307 NA NA 305 a 21 3841 753 21 17 4 172 NA NA 62 a 7 01508 1339 101 101 9 2567 457 14 10 4 2589 480 35 26 16 7a Transferred immediately after fusion to the oviduct.Fig. 1. Prometea, the world’s first cloned horse, born in 2003, in frontof her nuclear donor who was also her surrogate motherÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

336 C Galli, I Lagutina, R Duchi, S Colleoni and G LazzariConcluding RemarksAssisted-reproductive techniques have been developedin recent years and efficient techniques for oocytematuration, ICSI, embryo culture and SCNT are nowavailable. Their application for breeding purposes hasonly started to be explored because of the limitednumber of laboratories currently able to master them.Interestingly, cloned equine offspring reported so farappear not to exhibit the placental, foetal and neonatalpathologies described in other species. Furthermore, thefirst two cloned horses that have reached puberty haveshown normal reproductive performance and offspringfrom both are expected soon. For these reasons, thehorse breeding industry is watching with some interest,although often mixed with suspicion, the performance ofthe first cloned horses because cloning offers theunprecedented opportunity of generating copies ofcastrated champions and, therefore, offspring thatwould be impossible to obtain otherwise (Church2006). This possibility promises to dramatically increasethe impact of superior male genotypes in those horsebreeds that will accept cloning in their breeding schemes.AcknowledgementsThis work has been supported in part by grants from MUR (TECLA),Cariplo Foundation (NOBEL), the European Commission (Xenome,LSHB-CT-2006-037377) and the European Science Foundation (EuroStellsERAS-CT-2003-980409).ReferencesBattut I, Bezard J, Palmer E, 1991: Establishment of equineoviduct cell monolayers for co-culture with early equineembryos. J Reprod Fertil Suppl 44, 393–403.Bogh IB, Bezard J, Duchamp G, Baltsen M, Gerard N, DaelsP, Greve T, 2002: Pure preovulatory follicular fluid promotesin vitro maturation of in vivo aspirated equineoocytes. Theriogenology 57, 1765–1779.Booth PJ, Tan SJ, Reipurth R, Holm P, Callesen H, 2001:Simplification of bovine somatic cell nuclear transfer byapplication of a zona-free manipulation technique. CloningStem Cells 3, 139–150.Campbell KH, Fisher P, Chen WC, Choi I, Kelly RD, Lee JH,Xhu J, 2007: Somatic cell nuclear transfer: past, present andfuture perspectives. Theriogenology 68(Suppl. 1), S214–S231.Carneiro G, Lorenzo P, Pimentel C, Pegoraro L, Bertolini M,Ball B, Anderson G, Liu I, 2001: Influence of insulin-likegrowth factor-I and its interaction with gonadotropins,estradiol, and fetal calf serum on in vitro maturation andparthenogenic development in equine oocytes. Biol Reprod65, 899–905.Choi YH, Chung YG, Seidel GE Jr, Squires EL, 2001:Developmental capacity of equine oocytes matured andcultured in equine trophoblast-conditioned media. Theriogenology56, 329–339.Choi YH, Love CC, Chung YG, Varner DD, Westhusin ME,Burghardt RC, Hinrichs K, 2002: Production of nucleartransfer horse embryos by Piezo-driven injection of somaticcell nuclei and activation with stallion sperm cytosolicextract. Biol Reprod 67, 561–567.Choi YH, Love LB, Westhusin ME, Hinrichs K, 2004:Activation of equine nuclear transfer oocytes: methodsand timing of treatment in relation to nuclear remodeling.Biol Reprod 70, 46–53.Choi YH, Love LB, Varner DD, Hinrichs K, 2006a: Blastocystdevelopment in equine oocytes with low meiotic competenceafter suppression of meiosis with roscovitine prior to in vitromaturation. Zygote 14, 1–8.Choi YH, Love LB, Varner DD, Hinrichs K, 2006b: Holdingimmature equine oocytes in the absence of meiotic inhibitors:effect on germinal vesicle chromatin and blastocystdevelopment after intracytoplasmic sperm injection. Theriogenology66, 955–963.Church SL, 2006: Nuclear transfer saddles up. Nat Biotechnol24, 605–607.Cochran R, Meintjes M, Reggio B, Hylan D, Carter J, PintoC, Paccamonti D, Godke RA, 1998: Live foals producedfrom sperm-injected oocytes derived from pregnant mares.J Equine Vet Sci 18, 736–740.Dell’Aquila ME, Cho YS, Minoia P, Traina V, LacalandraGM, Maritato F, 1997: Effects of follicular fluid supplementationof in-vitro maturation medium on the fertilizationand development of equine oocytes after in-vitrofertilization or intracytoplasmic sperm injection. HumReprod 12, 2766–2772.Dell’Aquila ME, Masterson M, Maritato F, Hinrichs K, 2001:Influence of oocyte collection technique on initial chromatinconfiguration, meiotic competence, and male pronucleusformation after intracytoplasmic sperm injection (ICSI) ofequine oocytes. Mol Reprod Dev 60, 79–88.Dell’Aquila ME, Albrizio M, Maritato F, Minoia P, HinrichsK, 2003: Meiotic competence of equine oocytes and pronucleusformation after intracytoplasmic sperm injection(ICSI) as related to granulosa cell apoptosis. Biol Reprod68, 2065–2072.Galli C, Crotti G, Turini P, Duchi R, Mari G, Zavaglia G,Duchamp G, Daels P, Lazzari G, 2002a: Frozen-thawedembryos produced by ovum pick up of immature oocytesand ICSI are capable to establish pregnancies in the horse.Theriogenology 58, 705–708.Galli C, Maclellan LJ, Crotti G, Turini P, Ponderato N, DuchiR, Merlo B, Mari G, Barbacini S, Lazzari G, 2002b:Development of equine oocytes matured in vitro in differentmedia and fertilised by ICSI. Theriogenology 57, 718.Galli C, Lagutina I, Crotti G, Colleoni S, Turini P, PonderatoN, Duchi R, Lazzari G, 2003a: Pregnancy: a cloned horseborn to its dam twin. Nature 424, 635.Galli C, Lagutina I, Lazzari G, 2003b: Introduction to cloningby nuclear transplantation. Cloning Stem Cells 5, 223–232.Galli C, Colleoni S, Duchi R, Lagutina I, Lazzari G, 2007:Developmental competence of equine oocytes and embryosobtained by in vitro procedures ranging from in vitromaturation and ICSI to embryo culture, cryopreservationand somatic cell nuclear transfer. Anim Reprod Sci 98, 39–55.Gardner DK, Lane M, Spitzer A, Batt PA, 1994: Enhancedrates of cleavage and development for sheep zygotescultured to the blastocyst stage in vitro in the absence ofserum and somatic cells: amino acids, vitamins, and culturingembryos in groups stimulate development. Biol Reprod50, 390–400.Hinrichs K, Schmidt AL, 2000: Meiotic competence in horseoocytes: interactions among chromatin configuration, folliclesize, cumulus morphology, and season. Biol Reprod 62,1402–1408.Hinrichs K, Williams KA, 1997: Relationships among oocytecumulusmorphology, follicular atresia, initial chromatinconfiguration, and oocyte meiotic competence in the horse.Biol Reprod 57, 377–384.Hinrichs K, Choi YH, Love LB, Varner DD, Love CC,Walckenaer BE, 2005: Chromatin configuration within thegerminal vesicle of horse oocytes: changes post mortem andÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Cloning in Horses 337relationship to meiotic and developmental competence. BiolReprod 72, 1142–1150.Hinrichs K, Choi YH, Love CC, Chung YG, Varner DD,2006: Production of horse foals via direct injection ofroscovitine-treated donor cells and activation by injection ofsperm extract. Reproduction 131, 1063–1072.Hinrichs K, Choi YH, Varner DD, Hartman DL, 2007:Production of cloned horse foals using roscovitine-treateddonor cells and activation with sperm extract and ⁄ orionomycin. Reproduction 134, 319–325.Hochedlinger K, Jaenisch R, 2006: Nuclear reprogrammingand pluripotency. Nature 441, 1061–1067.Lagutina I, Lazzari G, Duchi R, Colleoni S, Ponderato N,Turini P, Crotti G, Galli C, 2005: Somatic cell nucleartransfer in horses: effect of oocyte morphology, embryoreconstruction method and donor cell type. Reproduction130, 559–567.Lagutina I, Lazzari G, Duchi R, Turini P, Tessaro I, BrunettiD, Colleoni S, Crotti G, Galli C, 2007: Comparative aspectsof somatic cell nuclear transfer with conventional and zonafreemethod in cattle, horse, pig and sheep. Theriogenology67, 90–98.Lazzari G, Crotti G, Turini P, Duchi R, Mari G, Zavaglia G,Barbacini S, Galli C, 2002a: Equine embryos at thecompacted morula and blastocyst stage can be obtainedby intracytoplasmic sperm injection (ICSI) of in vitromatured oocytes with frozen-thawed spermatozoafrom semen of different fertilities. Theriogenology 58, 709–712.Lazzari G, Mari G, Galli C, 2002b: Synergistic effect ofcycloheximide and 6-DMAP on activation of equine andbovine oocytes. J Reprod Fertil Abstr Ser 28, 73.Li X, Morris LH, Allen WR, 2001: Influence of co-cultureduring maturation on the developmental potential of equineoocytes fertilized by intracytoplasmic sperm injection(ICSI). Reproduction 121, 925–932.Li X, Morris LH, Allen WR, 2002: In vitro development ofhorse oocytes reconstructed with the nuclei of fetal andadult cells. Biol Reprod 66, 1288–1292.Li X, Tremoleda JL, Allen WR, 2003: Effect of the numberof passages of fetal and adult fibroblasts on nuclearremodelling and first embryonic division in reconstructedhorse oocytes after nuclear transfer. Reproduction 125, 535–542.de Loos F, van Vliet C, van Maurik P, Kruip TA, 1989:Morphology of immature bovine oocytes. Gamete Res 24,197–204.Lorenzo PL, Liu IK, Carneiro GF, Conley AJ, Enders AC,2002: Equine oocyte maturation with epidermal growthfactor. Equine Vet J 34, 378–382.Oback B, Wells D, 2002: Donor cells for nuclear cloning: manyare called, but few are chosen. Cloning Stem Cells 4, 147–168.Oback B, Wiersema AT, Gaynor P, Laible G, Tucker FC,Oliver JE, Miller AL, Troskie HE, Wilson KL, Forsyth JT,Berg MC, Cockrem K, McMillan V, Tervit HR, Wells DN,2003: Cloned cattle derived from a novel zona-free embryoreconstruction system. Cloning Stem Cells 5, 3–12.Rideout WM III, Hochedlinger K, Kyba M, Daley GQ,Jaenisch R, 2002: Correction of a genetic defect by nucleartransplantation and combined cell and gene therapy. Cell109, 17–27.Ritchie WA, 2006: Nuclear transfer in sheep. Methods MolBiol 325, 11–23.Rosati I, Berlinguer F, Bogliolo L, Leoni G, Ledda S, NaitanaS, 2002: The effect of co-culture on the development of invitro matured equine oocytes after intracytoplastic sperminjection. Equine Vet J 34, 673–678.Tremoleda JL, Stout TA, Lagutina I, Lazzari G, Bevers MM,Colenbrander B, Galli C, 2003: Effects of in vitro productionon horse embryo morphology, cytoskeletal characteristics,and blastocyst capsule formation. Biol Reprod 69,1895–1906.Vajta G, Peura TT, Holm P, Paldi A, Greve T, Trounson AO,Callesen H, 2000: New method for culture of zona-includedor zona-free embryos: the Well of the Well (WOW) system.Mol Reprod Dev 55, 256–264.Vajta G, Lewis IM, Trounson AO, Purup S, Maddox-HyttelP, Schmidt M, Pedersen HG, Greve T, Callesen H, 2003:Handmade somatic cell cloning in cattle: analysis of factorscontributing to high efficiency in vitro. Biol Reprod 68, 571–578.Vanderwall DK, Woods GL, Aston KI, Bunch TD, Li G,Meerdo LN, White KL, 2004: Cloned horse pregnanciesproduced using adult cumulus cells. Reprod Fertil Dev 16,675–679.Westhusin M, Hinrichs K, Choi YH, Shin T, Liu L, KraemerD, 2003: Cloning companion animals (horses, cats, anddogs). Cloning Stem Cells 5, 301–317.Willadsen SM, 1986: Nuclear transplantation in sheepembryos. Nature 320, 63–65.Wilmut I, Schnieke AE, McWhir J, Kind AJ, CampbellKH1997: Viable offspring derived from fetal and adultmammalian cells. Nature 385, 810–813.Woods GL, White KL, Vanderwall DK, Li GP, Aston KI,Bunch TD, Meerdo LN, Pate BJ, 2003: A mule cloned fromfetal cells by nuclear transfer. Science 301, 1063.Zhang JJ, Boyle MS, Allen WR, Galli C, 1989: Recent studieson in vivo fertilization of in vitro matured horse oocytes.Equine Vet J 8, 101–104.Author’s address (for correspondence): Cesare Galli, Laboratorio diTecnologie della Riproduzione, CIZ srl, Via Porcellasco 7 ⁄ f, 26100Cremona, Italy. E-mail: cesaregalli@ltrciz.itConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 338–346 (2008); doi: 10.1111/j.1439-0531.2008.01182.xISSN 0936-6768Application and Commercialization of Flow Cytometrically Sex-Sorted SemenD Rath 1 and LA Johnson 21 Institute of Farm Animal Genetics, Mariensee (FLI), Neustadt, Germany; 2 Mount Airy, MD, USAContentsThe current technology to sort X and Y chromosomebearing sperm population requires individual identificationand selection of spermatozoa in a modified high-speed flowcytometer. For farm animal species, the technology iscapable of producing sexed sperm at greater than 90%purity. However, only in the bovine, the technology hasreached a developmental level that allows its commercialapplication. Meanwhile, the demand for female calves hasgrown rapidly, which encourages the demand for sex-sortedsemen from high genetic value bulls. The success of thetechnology will depend mainly on the fertilizing capacity ofthe sorted spermatozoa, as this is the most affecting andeconomically relevant factor. To date, fertility is still variableand is quite dependant on post-sort processing. Newprocessing techniques are under investigation and will likelybe able to improve the fertility rates after AI with sex-sortedsemen. It is of great importance to select the right bulls andto test the sorted samples on a routine basis. In addition tothe demand for sex-sorted semen by the cattle industry, thereis also a significant demand expressed by pig farmers.However, it is still unknown if the use of sex-sorted sementhrough commercial pig AI will be economically feasible. Forthe pig, the combination of in vitro fertilization with sexedsemen and non-surgical embryo transfer is an alternative thatmerits further scientific attention. Recent developments inovine AI and ET will make it very likely that commercialsheep industry will adopt the sexing technology in theirbreeding concepts.IntroductionControlling the sex of offspring prior to conceptionpermits the livestock industry to produce the optimalproportion of males and females to take advantage ofsex-limited and sex-influenced traits, thus providingeconomically flexible management practices for theproducer. Controlling sex ratio permits faster geneticprogress, higher productivity, improves animal welfareby decreasing obstetric difficulties in cattle, avoidingcastration in pigs, and producing less environmentalimpact due to the elimination of the unwanted sexbefore they grow to adulthood. Sex pre-selection inanimals must be effective and efficient, must result infertility equal to or better than with unsexed semen, andmust be reasonably inexpensive and convenient to bewidely applied (Foote and Miller 1971). Currently, theonly known means of effectively producing separatepopulations of X and Y sperm in mammals is throughthe use of DNA differentiation by high-speed flowcytometry (Beltsville Sperm Sexing Technology; Johnsonet al. 1989). Applicable to virtually all mammals,this method can produce populations of X or Y spermwith greater than 90% purity and subsequent offspringwhose phenotypic sex is consistent with initial purityof the sorted sperm population. Limitations of thetechnology are entwined with the physiology andanatomy of the specific species in terms of the numberof sperm required for fertilization and subsequentproduction of offspring. The current technology requiresthat each sperm be separately interrogated for DNAcontent, thus limiting the number of sorted X or Ysperm in cattle, sheep, swine and horses to approximately12–20 million sperm per hour (Johnson andWelch 1999; de Graaf et al. 2007c; Heer 2007).As flow cytometers and computers were developed andimproved, it soon became apparent that flow cytometryheld the key to separation of viable X and Y chromosomebearing sperm based on their DNA content, which couldthen be used for fertilization. Successful development of aviable sperm sexing procedure encompassed four steps toachieve a verifiable method of controlling sex ratio inmammals. These steps were:(1) modification of a commercially available flowcytometer ⁄ cell sorter into a sperm cell sorter byadding a forward fluorescence detector and a bevelledsample injection needle to accommodate sperm orientationand minimize DNA variability (Johnson andPinkel 1986),(2) development of a method to stain living spermwith a vital fluorescent dye (Hoechst 33342, Sigma-Aldrich, Mu¨ nchen, Germany) in order to maintainviability through the sorting process and to the timeof fertilization (Johnson et al. 1987a),(3) merging the analytical and sorting capability of thesperm sorter to produce separate populations of livingX and Y sperm based on differential DNA content(Johnson et al. 1989) and(4) development of a method to re-analyse separatesorted populations of X and Y living sperm for theirDNA content to verify the purity of X or Y sperm inthe laboratory (Johnson et al. 1987b; Welch andJohnson 1999) to be a predictor of ultimate phenotypicsex of the live offspring (Johnson et al. 1989;Johnson 1991) and prove the efficacy of using DNA todifferentiate X and Y sperm that would be used toachieve fertilization.The limitations of single-cell measurement technologyhave been followed throughout the development of thetechnology and continued even to this day in commercialapplication. Consistent with these limitations, themethod is applied most easily in the bovine (2 millionsperm per AI; Seidel et al. 1997), ovine (1–5 millionsperm per AI; de Graaf et al. 2007c) but hardly at all inswine (50–100 million sperm per AI; Rath et al. 2003;Vazquez et al. 2003; Grossfeld et al. 2005). Even withthe obvious limitations, the technology has begun toflourish in commercial bovine practice. It is estimatedÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Commercialization of Sex-Sorted Semen 339that more than 2 million sexed calves have beenproduced since the technology was first applied tocommercial practice in 2000. Although the license tobegin commercialization using AI was granted by theUS Department of Agriculture in 1996 (Patent#5,135,759) to XY-Inc. Fort Collins, CO, USA, commercialproduction of bovine sexed sperm did not beginuntil 2000 (Cogent Breeding Ltd., Chester, UK). However,insufficient development of the technology by theparent company hampered faster and wider commercialapplication. Within the past year, XY-Inc. was sold to asub-licensee (Sexing Technologies, Navasota, TX,USA). Prior to this sale, Sexing Technologies had some20 sorters operating in North America, which has nowbeen expanded to more than 50 sorters in numerouslocations in North and South America, Europe andother parts of the world. These recent developmentssuggest that sexed bovine sperm will soon be available tomost cattle producers around the world. Intensiveresearch is still required to further improve the existingtechnology or find alternatives and ⁄ or combinationswith other biotechniques. Increased technology developmentis critical in order to develop the largecommercial opportunity that exists for sexed semencattle and in other livestock species, especially sheep,swine and horses. In this review, the current status of thesex-sorting technology associated with commercialapplications and the consequences for farm animalagriculture are summarized.The Limitations of Single-Cell MeasurementTechnologyThe current technology requires single sperm identification,individual recognition of the orienting position infront of the laser beam and single droplet charging.Although there has been significant progress in thenumber of sperms sorted per unit time in the last12 years (1–2 million sperm per hour to the current levelof approximately 20 million sperm per hour), theprocess remains inefficient in the production of spermfor AI. Standard AI dose numbers of fresh or frozenunsexed sperm are out of reach for this technology.Efficient utilization of sex-sorted semen requires asignificant reduction in spermatozoa per AI dose orfor use in in vitro fertilization (IVF) and other biotechniques.For example, 2 million live spermatozoa frommany bulls or for some bulls even below seem to besufficient for AI. The usability of these bulls for sexsortingdepends only partly on the strength of theirspermatozoa to survive the sorting process. It is wellknown that the fertilizing ability with unsortedspermatozoa varies among individual bulls when lowsperm concentrations are used for AI (Den Daas et al.1998).In addition to the effects of reduced sperm number,high dilution effects diminish the fertilizing potentialof spermatozoa as the protecting and regulatingsubstances of seminal plasma are also diluted oreliminated (Maxwell and Johnson 1999; Centurionet al. 2003).The differential sperm DNA content is identified withthe Bis-benzimide Hoechst 33342 that emits a bluefluorescence when excited with UV light. Hoechst 33342was selected based on the fact that it was a vital stainand one that effectively was able to penetrate the livingsperm membrane and bind to the DNA of the highlycondensed chromatin of the sperm nucleus (Johnsonet al. 1987a). In several experiments, the influence of theHoechst 33342 on DNA integrity was tested. Johnsonet al. (1989) postulated that fluorochrome dyes reduceembryonic viability by mid-gestation. This coincideswith the findings of Spinaci et al. (2005) with boarspermatozoa. Co-incubation with Hoechst dye as well asthe sorting process itself diminished the percentage oflive spermatozoa. The damaged ability to fertilize andcarry the embryo to term may be the result of thecombined effects of the dye and UV laser or of eitherindividually. Higher laser intensity is more damagingthan the lower laser intensity as shown for rabbitspermatozoa by Johnson et al. (1996). Recently, Schenkand Seidel (2007) reported a similar finding for bovinesemen. However, less laser intensity diminishes theresolution and indirectly the sort rates. Guthrie et al.(2002) saw no differences on embryo development whenpig spermatozoa were illuminated with 125 or 25 mWlaser power, but optimum resolution during sortingbetween X and Y intact sperm required at least 125 mWlaser power.Catt et al. (1997) labelled human spermatozoa onmicroscope slides with the Hoechst dye and exposedthem to UV laser light. No changes were found in thefrequency of endogenous DNA nicks. This is inagreement with a recent study from Parrilla et al.(2004), indicating no genotoxic effects of Hoechst33342 in porcine spermatozoa. Boe-Hansen et al.(2005) used the neutral Comet assay and the sperm condensationstructure assay (SCSA) to evaluatesperm chromosome integrity. Both tests showed thatsperm integrity is improved in the sorted populationwhen compared with unsorted semen. This finding isno doubt due to the presence of FD#40 red food dye,which is added to the pre-sort sample to eliminatemembrane damaged sperm from the sorting process(Johnson et al. 1999). Similar results were obtained byDe Ambrogi et al. (2006), who found no effect on thedefragmentation index after sorting, and de Graafet al. (2007a,b) getting even better fertilization and ETresults with sorted semen when compared with highdiluted controls.A significant factor during sorting is also therepeated electrical charging and electrostatic deviation.Membranes of the mid-piece of the sperm tail aresensitive to the electric field and may undergo depolarization.Furthermore, we believe that mitochondrialactivity is reduced due to the presence of reactiveoxygen species (ROS) produced by electric forces(Klinc and Rath 2007; Klinc et al. 2007). Similarpositive results were reported for sex-sorted boarspermatozoa when media were supplemented withPSP I ⁄ II heterodimers (Garcı´ a et al. 2007), whichincreased motility and mitochondrial activity(Centurion et al. 2003). However, in a study reportedby de Graaf et al. (2007b), they were unable to seesimilar benefits to ram semen antioxidants or seminalplasma.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

340 D Rath and LA JohnsonCommercial Application of Sex-SortedSpermatozoa in Livestock ProductionExcellent reviews on the commercialization, especially ofbovine sexed semen, have been published by Amann(1999), Seidel (2003a,b), Garner (2006) and Garner andSeidel (2008). Although their calculation models aremainly directed to the American farm situation, they arehelpful guidelines for other countries too. The situationin other species had been summarized by Maxwell et al.(2004) and for pigs by Johnson et al. (2005). Spermsorting based on the Beltsville Sperm Sexing Technologyhas reached a certain technical standard that makes itrobust enough for use under practical conditions, even ifit has clear limitations as described above. Its utilizationdiffers significantly among species mainly because of thebiological differences related to site of semen deposition,length of cycle, demand on sperm numbers, sortabilityand freezability of spermatozoa as well as commercialdemands and management conditions.The Commercial Use of Bovine Sex-SortedSemenStage of applicabilityDue to the high sorting index (131), an approximation ofthe ability to flow cytometrically sort sperm consisting ofthe head profile area (lm 2 ) · X-Y sperm DNA difference(%; Garner 2006), bovine spermatozoa are more suitablefor high-speed sorting than other species. The methodhas reached a standard that is likely to make it profitablein dairy herds soon. However, impairments to thefertilizing capacity of sorted bull sperm are obviousand impairment of fertilizing capacity because of theimpact of sorting and high dilution on capacitation andseminal plasma components continue to be limitations tobe dealt with. Further improvement of the pre- and postsorthandling of the semen is critical and requiresintensive research to minimize negative effects on thepost-thaw lifespan of sexed spermatozoa as the fertilephase has direct implications on AI regimens. In a recentstudy employing a modified sperm handling protocolduring sorting and freezing and in combination with anew extender for both sorting and freezing (Sexcess Ò ,Masterrind GmbH, Verden, Germany), we were able toextend the post-thaw viability and progressive motilityfor several hours as shown by thermo tolerance test andcalving rates (Klinc 2005; Klinc et al. 2007). In this AIstudy, sorted semen (2 million live sperm ⁄ straw) wasthawed and inseminated in accordance with the existingprotocols for conventional semen (thawing at 37°C for20 s; AI 12–24 h after onset of heat; semen deposition inthe uterine body or if possible without force into theipsilateral horn). Pregnancy and calving rates were equalto controls. Similar data were also obtained for sortedliquid semen stored for up to 72 h (Klinc and Rath 2007).Significant technical improvements in sperm quality aftersorting were developed using the latest laser technologyand replacing the water-cooled Argon gas laser with apulsed solid-state laser. Shorter exposure to the laserlight seems to further reduce sperm stress, and observationswith ram spermatozoa indicate increased fertilityrates (de Graaf et al. 2007c). Another stress factor is thehydrodynamic pressure. The percentage of live bull andstallion spermatozoa increased significantly when thefluid pressure was lowered from 2.07 to 2.59 mm Hg andincreased developmental rates of bovine IVF embryos(Campos-Chillon and de la Torre 2003; Suh et al. 2005).For many reasons, it may be helpful to sort conventionallyfrozen semen after thawing and refreeze thesorted samples. Such methods are under investigation.Hollinshead et al. (2004) showed for the first time thatbull semen survives an adapted protocol from ram(Hollinshead et al. 2003) employing a gradient centrifugationto separate egg yolk from the thawed samples andto enrich intact sperm prior to sorting. Maxwell et al.(2007) tested PureSperm Ò (Labotect, Go¨ ttingen, Germany)as effective gradient for high diluted, unsorted bullsperm, and in a recent study, Parrilla et al. (unpublished)showed that PureSperm and Bovipure Ò (Labotect,Go¨ ttingen, Germany) are suitable for refreezing sexedbull sperm and to produce embryos to the same extent aswith unsorted sperm from the same ejaculates.Altogether, the technical improvements made in therecent years allow for the maintenance of an acceptablestandard of bovine semen quality. Nevertheless, it is selfevidentto control the sorted sperm quality on a dailybasis. Some suggested prerequisites for sex-sorted bullsemen of optimal quality are as follows:(1) Bulls must be pre-tested and those failing withunsorted low-dose insemination should be excludedfrom their use to produce sex-sorted sperm.(2) A complete spermiogram should be made fromeach ejaculate before sorting. The test should include:macroscopical evaluation, mass movement categorization,single sperm motility estimation, evaluation ofsperm concentration and a complete morphologicalevaluation test, especially as a faster acrosome reactionis commonplace in sex-sorted sperm (Mocé et al.2006).(3) After sorting, aliquots of each batch should betested for single sperm motility, motility pattern in athermal tolerance test preferably for 6 h at 38°C,precise morphological evaluation for acrosome integrityand total morphology, FACS Analysis (FITC-PNA ⁄ Syto17 ⁄ PI; SCSA), and a general test ofmicrobiological contamination. These tests may beperformed with sorted sperm of the unwanted sex.Additionally, one straw from each batch of thewanted sex needs to be reanalysed for sorting purityand the total number of spermatozoa per straw.Ballester et al. (2007) confirmed that it is not sufficientto test semen quality directly after thawing. Incubationof the samples at 38°C revealed a reduction oflinearity, sperm viability (Seminaphtharhodafluor)and acrosome integrity.It is recommended to put at least 2 million livespermatozoa in a straw as the absolute minimum.Andersson et al. (2004, 2006) clearly showed thatmore than 2 million spermatozoa are necessary toobtain satisfactory pregnancy and calving rates.Caution must also be exercised to account for highindividual bull effects. In order to avoid losses of thehighly diluted semen, the liquid column in the strawshould be divided into a first segment containingÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Commercialization of Sex-Sorted Semen 341extender only, which will close the cotton plug, and asecond segment containing the sexed spermatozoa.Both segments need to be divided by an air bubble.Automatic systems to fill the straws in this manner arecommercially available.(4) It is recommended that only heifers are inseminatedwith sex-sorted semen at the present time. Fromseveral studies on our research farm, we have noindication that neither timing nor variability ofovulation nor site of insemination is responsible forthe large differences in non-return rates betweenheifers and cows.On the female side, it is a prerequisite to provideoptimal herd management to attain high female fertility(Seidel 2003a,b). Virtually, all fertility results to datefrom the AI of sex-sorted semen in cattle have been fromheifers. Using AI with sorted semen in cows frequentlyproduces unacceptable non-return rates, possibly becauseof the inability to consistently pinpoint ovulationfor optimum AI.Several groups (Cran et al. 1994; Merton et al. 1997;Lu et al. 1999) reported similar cleavage rates butreduced blastocyst development following fertilizationin vitro with sorted bovine sperm cells. In a recentstudy, we demonstrated for the first time that bovineIVF-embryos derived from sex-sorted spermatozoadisplay a reduction in the relative abundance ofdevelopmentally important genes like Gluc-3 andG6PD, compared with their counterparts derived fromunsorted semen (Morton et al. 2007). It requiresfurther in-depth study as epigenetic changes alreadyoccurring in the early embryo are thought to beinvolved in postnatal abnormalities. These are deleteriouseffects of sperm sexing on spermatozoa beyondthose previously recognized and reflected in the developmentalcompetence of embryos. Similarly, cleavagerates after IVF with sex-sorted spermatozoa were 30%below those of unsorted spermatozoa of the sameejaculate, blastocyst formation on day 8 was 30–40%lower than for the controls (Bermejo-Alvarez et al.2007) and cell cycles were reduced (Beyhan et al. 1999)or disturbed in timing (Cran et al. 1993; Lu et al. 1999;Morton et al. 2005). On the contrary, Seidel et al.(1999) found no excess embryonic loss between 1 and2 months of gestation in heifers inseminated withsorted sperm.Commercial demandsThe demand for sex-sorted semen may vary betweencountries. Based on the American market situation,advantages are foreseeable for the decoupling ofproduction of replacement heifers and cows from thenumber of culled cows. More heifers are available forherd replacement and by this a stronger selection willbe possible. A higher degree of female selection willhave an impact on the genetic development of thepopulation as females will contribute with up to 15%on genetic selection, which was based so far on sires(Weigel and Barlass 2003). As the number of replacementanimals is satisfied, herd size may grow andmore females are sold. In parallel, the milk productioncan rise and the market price may decrease as benefitfor consumers. In addition, the costs for progenytesting and embryo transfer will decrease and eventuallythe economic benefit for genetic markers will beenhanced as more siblings of the same sex can beproduced of the best bulls and cows. The more thetechnology develops and more reasonable pricing isable to be applied the more the prices for replacementand export heifers will decrease. Therefore, only theearly users will benefit most by selling surplus heifersat substantially higher prices than the cost to raise orreplace (De Vries et al. 2008). It is essential that ascommercialization increases, more data on fertility,embryonic losses, foetal deformations, abortions andheifer health can be collected. Such information willhelp to evaluate the technology correctly for itspractical use.Status of commercial implementationThe technology is the most highly developed for bovinesemen and was introduced into commercial applicationin the United Kingdom in 2000. Recently, severalAmerican AI centres have contracted for the technologyand offer sexed semen from their bulls worldwide.As this manuscript is written, there are about 50sorters in the USA and three new centres in Europestarting production of sexed bull semen. At the currentsituation, the high price for sexed semen may beadequate as long as the demand is high for heifers butits value will decrease with decreasing milk, heifer andcull cow prices, with increasing feed costs and lowerpricing of conventional semen. If the sex-sorted semenis highly fertile and female prices are $200 (approximately130€) above males, Seidel (2003a) considered anadditional price of $21.56 per AI dose as affordable formarketing.In a recent study, Ettema et al. (2007) calculated thesituation in Denmark as an example for the Europeanmarket. They developed a spreadsheet model includingprice for springing heifers, replacement costs, price perbeef calf, price of sexed semen, conception rates withsexed semen, replacement rates, sex ratio of sexedsemen, incidence of dystocia and stillbirth. The mainfactors affecting price are lowered fertility, the high costof equipment, necessity for skilled personnel and thecost for intellectual property. From this data, it isobvious that the net return to assets (NRA) will benegative for the first year and it will take 3–4 yearsbefore herds have reached a steady state with respect toraising and selling replacements.In summary, sex-sorted bull semen is in high demand.The success of a widespread application depends on itsprice. It has to be moderate enough to allow areasonable profit for farmers, and it has to be highenough to make its production lucrative (Seidel et al.2003). A prerequisite for integration in genetic programmesis the quality of the sorted semen. Underoptimal conditions, fertility is beginning to reach satisfactorylevels though semen variability and managementfactors continue to be critically important to success incommercial practice.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

342 D Rath and LA JohnsonThe Commercial Use of Ovine Sex-SortedSemenStage of applicabilityIn sheep, insemination with less than 1 million motileunsexed sperm has been demonstrated to result inacceptable fertility (Walker et al. 1984; 5 million motile,Eppleston et al. 1986) but the results vary considerably(Salamon and Maxwell 2000). Lambs from sex-sortedsemen were produced via laparoscopy by Cran et al.(1997). Subsequent experiments with 1–16 million sexsortedfrozen-thawed spermatozoa produced disappointinglevels of fertility (Hollinshead et al. 2002,2003). Even inseminations with up to 40 million totalsorted, cryopreserved spermatozoa were inconclusive(Hollinshead et al. 2003). However, in a very recentstudy using an adapted protocol, de Graaf et al. (2007c)demonstrated that dosages of 1 or 5 million sortedmotile ram sperm gave superior fertility to non-sortedsperm when laparoscopically inseminated. These resultshave significance for the future commercialization of sexpre-selection technology in sheep as a reduction in theminimum effective sperm number will allow a correspondingdecrease in the associated cost per dose and setaside a larger percentage of the sheep industry to utilizesex pre-selection. In another trial, the same groupshowed for the first time for any species that frozenthawedspermatozoa, after sex-sorting and a secondcryopreservation step are capable of producing offspringof the predicted sex following AI (de Graaf et al. 2006).Sorting of frozen semen and subsequent refreezing withthe birth of lambs had only been described so far afterIVF and ET (Hollinshead et al. 2004). Sex-sorted ramsperm can be used effectively with intra-cytoplasmicsingle sperm injection (ICSI; Catt et al. 1996; Hollinsheadet al. 2002) and IVF (Rhodes et al. 1994; Maxwellet al. 2004; Morton et al. 2004). de Graaf et al. (2007a)also showed that the potential of sex-sorted frozenthawedram semen is equal to unsorted semen whenused for laparoscopic insemination, or intrauterineinsemination in superovulated ewes and subsequentembryo transfer of morula and blastocysts.Status of commercial implementationCommercialization of sex-sorted ram spermatozoa forlaparoscopical insemination may play a role in the nearfuture, though it is questionable, whether its use will besufficiently profitable to make it routinely available. Thelatest results of the University of Sydney group (deGraaf et al. 2006, 2007c) are very promising and theimplementation of sex-sorted ram sperm in the commercialpractice may have a significant impact on theAustralian and New Zealand markets.The Commercial Use of Porcine SexSortedSemenStage of applicabilityIn principle, porcine sperm are as easy to sort as otherlivestock species. Its sorting index (115) ranges betweenbull and ram sperm (Garner 2006). The primarylimitations are the large amount of sperm necessaryfor AI and the sensitivity of sorted sperm against highdilution and cryopreservation. Factors associated withprocessing boar sperm for sexing lead to capacitationlike membrane changes (Ashworth et al. 1994; Maxwellet al. 1997, 1998; Barrios et al. 2000). Seminal plasma asa constituent of the sample medium and of the catchmedium has been shown to be beneficial to decapacitatesorted spermatozoa (Caballero et al. 2004).Embryos from IVF with sexed boar semen were firstproduced 15 years ago (Rath et al. 1993). Subsequently,offspring were produced after sex-sorting and IVF ofnon-frozen spermatozoa (Rath and Niemann 1996,1997; Abeydeera et al. 1998). In addition, ICSI hasbeen used successfully to produce male offspring (Probstand Rath 2003).A major step forward was made with the idea of lowdoseinsemination (for review see Rath 2002) and withthe development of a flexible catheter (Firflex Ò , Minitu¨b, Tiefenbach, Germany) for deep intrauterine insemination(DIU) in Spain (Martinez et al. 2001, 2002;Cuello et al. 2005). Progeny have been produced afterDIU of non-frozen bulk-sorted (Vazquez et al. 2003)and sex-sorted non-frozen sperm (Rath et al. 2003;Grossfeld et al. 2005), using as few as 50 · 10 6 spermatozoain a volume of 2 ml (Rath et al. 2003). However,the method is only available for sows and not forprimiparous gilts. Placing the semen directly in front ofthe uterotubal junction may avoid major sperm lossesduring uterine transportation. Unilateral insemination issufficient to find adequate number of embryos in bothhorns (Tummaruk et al. 2007).Boar semen has been sex-sorted, frozen, thawed andsurgically inseminated into the oviduct to produce liveoffspring (Johnson et al. 2000). However, a repeatablemethod for freezing sex-sorted boar sperm remains to bedeveloped. Although attempts to freeze sex-sorted boarsemen resulted in sufficient post-thaw sperm quality, nopregnancies were able to go to term after non-surgicalDUI (Bathgate et al. 2008). Similarly, embryos derivedfrom IVF with sex-sorted frozen sperm and transferrednon-surgically did not develop to term (Bathgate et al.2007).Commercial demandsIn swine, the greatest general demand for sex-sortedsemen is to produce female piglets for meat production.Several European countries will soon begin to forbidcastration of male piglets; therefore a shift in sex ratioswill be a necessity for pig producers under suchregulations.Status of commercial implementationThe Beltsville Sperm Sexing Technology, as it iscurrently used in research and commercial practice, isunable to produce sufficient numbers of sperm to makeswine AI feasible at the present time. In its current stageof development, sex-sorted pig sperm could be used inspecialized commercial practice, e.g. nuclear herds, ifcombined with IVF ⁄ ET and ⁄ or laparoscopic insemination(Rath et al. 1999; Vazquez et al. 2006; Garcı´ a et al.2007). Optimum utilization of sex-sorted boar spermÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Commercialization of Sex-Sorted Semen 343would suggest that it would be advantageous andessential to freeze sex-sorted boar spermatozoa so thatthe greatest flexibility for insemination and ovulationcan be exercised.The Commercial Use of Equine Sex-SortedSemenSemen from horses can be sorted even though theefficiency is lower than in other livestock species.However, the value of a single dose may cover allproduction costs. Buchanan et al. (2000) reported a40% pregnancy rate in mares after AI of 25 · 10 6 sexsortednon-frozen spermatozoa and pregnancies haveeven been obtained with as few as 5 · 10 6 liquid orfrozen-thawed spermatozoa using hysteroscopic insemination(Lindsey et al. 2002a, b). Recently, a filly wasborn after hysteroscopic insemination with low numbersof frozen-thawed sexed spermatozoa into the uterotubaljunction. Highly diluted unsorted semen from the sameejaculates did no better. Undiluted semen inseminated ata dose of 5 · 10 8 spermatozoa yielded satisfactoryconception rates (Clulow et al. 2007). Therefore, thepoor fertility after hysteroscopic insemination with lowdoses of sex-sorted or non-sorted spermatozoa may bedirectly attributable to the low-dose insemination conditionswith frozen-thawed rather than sex-sorted spermatozoa.In order to improve the post-thaw quality ofsex-sorted stallion semen, different approaches werefocused on cushion centrifugation before and aftersorting as well as pre-selection of a stable spermpopulation using a PureSperm gradient (Knop et al.2005; Heer 2007) and different automated freezingprotocols (Buss 2006; Clulow et al. 2007). The qualityof the sorted frozen semen was improved stepwise andinsemination trials have yet to be performed. The majordifficulty with equine semen is its variability amongstallions and ejaculates. The commercialization of sexedstallion semen will depend on the consistency of thefertility after sorting and freezing, a broader number ofstallions with sortable ejaculates and improvements ofthe current hysteroscopic insemination technique.ReferencesAbeydeera LR, Johnson LA, Welch GR, Wang WH, BoquestAC, Cantley TC, Rieke A, Day BN, 1998: Birth of pigletspreselected for gender following in vitro fertilization of invitro matured pig oocytes by X and Y chromosome bearingspermatozoa sorted by high speed flow cytometry. Theriogenology50, 981–988.Amann RP, 1999: Issues affecting commercialization of sexedsperm. Theriogenology 52, 1441–1457.Andersson M, Taponen J, Koskinen E, Dahlbom M, 2004:Effect of insemination with doses of 2 or 15 million frozenthawedspermatozoa and semen deposition site on pregnancyrate in dairy cows. Theriogenology 61, 1583–1588.Andersson M, Taponen J, Kommeri M, Dahlbom M, 2006:Pregnancy rates in lactating Holstein-Friesian cows afterartificial insemination with sexed sperm. Reprod DomestAnim 41, 95–97.Ashworth PJC, Harrison RAP, Miller NGA, Plummer JM,Watson PF, 1994: Survival of ram spermatozoa at highdilution: protective effect of simple constituents of culturemedia as compared with seminal plasma. Reprod Fertil Dev6, 173–180.Ballester J, Johannisson A, Saravia F, Håård M, GustafssonH, Bajramovic D, Rodriguez-Martinez H, 2007: Post-thawviability of bull AI-doses with low-sperm numbers. Theriogenology68, 934–943.Barrios B, Pérez-Pe´ R, Gallego M, Tato A, Osada J, Muin˜ o-Blanco T, Cebrián-Pérez JA, 2000: Seminal plasma proteinsrevert the cold-shock damage on ram sperm membrane. BiolReprod 63, 1531–1537.Bathgate R, Morton KM, Eriksson BM, Rath D, Sieg B,Maxwell WM, Evans G, 2007: Non-surgical deep intrauterinetransfer of in vitro produced porcine embryosderived from sex-sorted frozen-thawed boar sperm. AnimReprod Sci 99, 82–92.Bathgate R, Grossfeld R, Susetio D, Ruckholdt M, HeasmanK, Rath D, Evans G, Maxwell WM, 2008: Early pregnancyloss in sows after low dose, deep uterine artificial inseminationwith sex-sorted, frozen-thawed sperm. Anim ReprodSci 104, 440–444.Bermejo-Alvarez P, Rizos D, Rath D, Lonergan P, Gutierrez-Adan A, 2007: Epigenetic differences between male andfemale bovine blastocysts produced in vitro. Physiol Genomics17, 264–272.Beyhan Z, Johnson LA, First NL, 1999: Sexual dimorphism inIVM-IVF bovine embryos produced from X and Y chromosome-bearingspermatozoa sorted by high speed flowcytometry. Theriogenology 52, 35–48.Boe-Hansen GB, Morris ID, Annette B, Ersbølla K, Greve T,Christensen P, 2005: DNA integrity in sexed bull spermassessed by neutral Comet assay and sperm chromatinstructure assay. Theriogenology 63, 1789–1802.Buchanan BR, Seidel GE, McCue PM, Schenk JL, HerickhoffLA, Squires EL, 2000: Insemination of mares with lownumbers of either unsexed or sexed spermatozoa. Theriogenology53, 1309–1321.Buss H, 2006: Improvement of the freezability of sex-sortedstallion spermatozoa. Thesis, University of VeterinaryMedicine Hannover, Hannover, Germany. (in German).Caballero I, Vazquez JM, Centurio´ n F, Rodrı´ guez-MartinezH, Parrilla I, Roca J, Cuello C, Martinez EA, 2004:Comparative effects of autologous and homologous seminalplasma on the viability of largely extended boar spermatozoa.Reprod Domest Anim 39, 370–375.Campos-Chillon LF, de la Torre JF, 2003: Effect on concentrationof sexed bovine sperm sorted at 40 and 50 psi ondevelopmental capacity of in vitro produced embryos.Theriogenology 59, 506. (Abstract).Catt SL, Catt JW, Gomez MC, Maxwell WMC, Evans G,1996: Birth of a male lamb derived from in vitro maturedoocyte fertilized by intracytoplasmic injection of a single‘male’ sperm. Vet Rec 139, 494–495.Catt SL, Sakkas D, Bizzaro D, Bianchi PG, Maxwell WMC,Evans G, 1997: Hoechst staining and exposure to UV laserduring flow cytometric sorting does not affect the frequencyof detected endogenous DNA nicks in abnormal and normalhuman sperm. Mol Hum Reprod 3, 821–825.Centurion F, Vazquez JM, Calvete JJ, Roca J, Sanz L, ParrillaI, Garcia EM, Martinez EA, 2003: Influence of porcinespermadhesins on the susceptibility of boar spermatozoa tohigh dilution. Biol Reprod 69, 640–646.Clulow JR, Buss H, Sieme H, Rodger JA, Cawdell-Smith AJ,Evans G, Rath D, Morris LH, Maxwell WM2007: Fieldfertility of sex-sorted and non-sorted frozen-thawed stallionspermatozoa. Anim Reprod Sci. doi:10.1016/j.anireprosci.2007.08.015.(Epub ahead of print)Cran DG, Johnson LA, Miller NGA, Cochrane D, Polge CE,1993: Production of bovine calves following separation ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

344 D Rath and LA JohnsonX- and Y-bearing sperm and in vitro fertilization. Vet Rec132, 40–41.Cran DG, Cochrane DJ, Johnson LA, Wei H, Lu KH, PolgeC, 1994: Separation of X- and Y-chromosome bearingbovine sperm by flow cytometry for use in IVF. Theriogenology41, 183. (Abstract).Cran DG, McKelvey WAC, King ME, Dolman DF, McEvoyTG, Broadbent PJ, Robinson JJ, 1997: Production of lambsby low dose intrauterine insemination with flow cytometricallysorted and unsorted sperm. Theriogenology 47, 267.(Abstract).Cuello C, Berthelot F, Martinat-Botte´ F, Venturi E, GuillouetP, Vázquez JM, Roca J, Martı´ nez EA, 2005: Piglets bornafter non-surgical deep intrauterine transfer of vitrifiedblastocysts in gilts. Anim Reprod Sci 85, 275–286.De Ambrogi M, Spinaci M, Galeati G, Tamanini C, 2006:Viability and DNA fragmentation in differently sorted boarspermatozoa. Theriogenology 66, 1994–2000.De Vries A, Overton M, Fetrow J, Leslie K, Eicker S, RogersG, 2008: Exploring the impact of sexed semen on thestructure of the dairy industry. J Dairy Sci 91, 847–856.Den Daas JH, De Jong G, Lansbergen LM, Van Wagtendonk-De Leeuw AM, 1998: The relationship between thenumber of spermatozoa inseminated and the reproductiveefficiency of individual dairy bulls. J Dairy Sci 81, 1714–1723.Eppleston J, Maxwell WMC, Battye KM, Roberts EM,1986: Effect of thawed motility and intrauterine dose ofmotile sperm on fertility in ewes. Proc Aust Soc ReprodBiol 18, 19.Ettema JF, Hoag DL, Seidel DE Jr, 2007: Economic opportunitiesfor sexed semen on commercial dairies. West DairyNews 7, 67–68.Foote RH, Miller P, 1971: What might sex ratio control meanin the animal world? In: Kiddy CA, Hafs HD(eds), SexRatio at Birth Prospects for Control. American Society ofAnimal Science, Champaign, pp. 1–9.Garcı´ a EM, Vázquez JM, Parrilla I, Calvete JJ, Sanz L,Caballero I, Roca J, Vazquez JL, Martı´ nez EA, 2007:Improving the fertilizing ability of sex sorted boar spermatozoa.Theriogenology 68, 771–778.Garner DL, 2006: Flow cytometric sexing of mammaliansperm. Theriogenology 65, 943–957.Garner DL, Seidel GE Jr, 2008: History of commercializingsexed semen for cattle. Theriogenology 69, 886–895.de Graaf SP, Evans G, Maxwell WM, O’Brien JK, 2006: Invitro characteristics of fresh and frozen-thawed ram spermatozoaafter sex-sorting and re-freezing. Reprod FertilDev 18, 867–874.de Graaf SP, Beilby K, O’Brien JK, Osborn D, Downing JA,Maxwell WM, Evans G, 2007a: Embryo production fromsuperovulated sheep inseminated with sex-sorted ram spermatozoa.Theriogenology 67, 550–555.de Graaf SP, Evans G, Gillan L, Guerra MMP, Maxwell WM,O’Brien JK, 2007b: The influence of antioxidant, cholesteroland seminal plasma on the in vitro quality of sortedand non-sorted ram spermatozoa. Theriogenology 67, 217–227.de Graaf SP, Evans G, Maxwell WM, Downing JA, O’BrienJK, 2007c: Successful low dose insemination of flowcytometrically sorted ram spermatozoa in sheep. ReprodDomest Anim 42, 648–653.Grossfeld R, Klinc P, Sieg B, Rath D, 2005: Production ofpiglets with sexed semen employing a non-surgical inseminationtechnique. Theriogenology 63, 2269–2277.Guthrie HD, Johnson LA, Garrett WM, Welch GR, DobrinskyJR, 2002: Flow cytometric sperm sorting: effects ofvarying laser power on embryo development in swine. MolReprod Dev 61, 87–92.Heer P, 2007: Adjustment of cryopreservation of stallionspermatozoa to the Beltsville Sperm Sexing Technology.Doctorial Thesis, University of Veterinary Medicine Hannover,Hannover, Germany (in German).Hollinshead FK, O’Brien JK, Maxwell WMC, Evans G, 2002:Production of lambs of predetermined sex after the inseminationof ewes with low numbers of frozen-thawed sortedX- or Y-chromosome-bearing spermatozoa. Reprod FertilDev 14, 503–508.Hollinshead FK, Gillian L, O’Brien JK, Evans G, MaxwellWMC, 2003: In vitro and in vivo assessment of functionalcapacity of flow cytometrically sorted ram spermatozoaafter freezing and thawing. Reprod Fertil Dev 15, 351–359.Hollinshead FK, Evans G, Evans KM, Catt SL, MaxwellWMC, O’Brien JK, 2004: Birth of lambs of a predeterminedsex after in vitro production of embryos usingfrozen-thawed sex-sorted and re-frozen-thawed ram spermatozoa.Reproduction 127, 557–568.Johnson LA, 1991: Sex preselection in swine: altered sex ratiosin offspring following surgical insemination of flow sortedX- and Y-bearing sperm. Reprod Domest Anim 26, 309–314.Johnson LA, Pinkel D, 1986: Modification of a laser-basedflow cytometer for high-resolution DNA analysis of mammalianspermatozoa. Cytometry 7, 268–273.Johnson LA, Welch GR, 1999: Sex preselection: high-speedflow cytometric sorting of X- and Y-sperm for maximumefficiency. Theriogenology 52, 1134–1323.Johnson LA, Flook JP, Look MV, 1987a: Flow cytometry ofX and Y chromosome-bearing sperm for DNA using animproved preparation method and staining with Hoechst33342. Gamete Res 17, 203–212.Johnson LA, Flook JP, Look MV, Pinkel D, 1987b: Flowsorting of X and Y chromosome-bearing spermatozoa intotwo populations. Gamete Res 16, 1–9.Johnson LA, Flook JP, Hawk HW, 1989: Sex preselection inrabbits: live births from X and Y sperm separated by DNAand cell sorting. Biol Reprod 41, 199–203.Johnson LA, Cran DG, Welch GR, Polge C, 1996: Genderpreselection in mammals. In: Miller RH, Pursel VG,Norman HD(eds), Beltsville Symposium XX: Biotechnology’sRole in the Genetic Improvement of Farm Animals.American Society of Animal Science, Savoy, pp. 151–164.Johnson LA, Welch GR, Rens W, 1999: The Beltsville spermsexing technology: high-speed sperm sorting gives improvedsperm output for in vitro fertilization and AI. J Anim Sci77(Suppl. 2), 213–220.Johnson LA, Guthrie HD, Fiser P, Maxwell WMC, WelchGR, Garrett WM2000: Cryopreservation of flow cytometricallysorted boar sperm: effects on in vivo embryodevelopment. J Anim Sci 78 (Suppl. 1), 198. (Abstract).Johnson LA, Rath D, Vazquez JM, Maxwell WMC, DobrinskyJR, 2005: Preselection of sex of offspring in swine forproduction: current status of the process and its application.Theriogenology 63, 615–624.Klinc P, 2005: Improved fertility of flow cytometrically sexselected bull spermatozoa. Doctoral thesis, University ofVeterinary Medicine Hannover, Hannover, Germany.Klinc P, Rath D, 2007: Reduction of oxidative stress in bovinespermatozoa during flow cytometric sorting. Reprod DomestAnim 42, 63–67.Klinc P, Frese D, Osmers H, Rath D, 2007: Insemination withsex sorted fresh bovine spermatozoa processed in thepresence of antioxidative substances. Reprod Domest Anim42, 58–62.Knop K, Hoffmann N, Rath D, Sieme H, 2005: Effects ofcushioned centrifugation technique on sperm recovery andsperm quality in stallions with good and poor semenfreezability. Anim Reprod Sci 89, 294–297.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Commercialization of Sex-Sorted Semen 345Lindsey AC, Morris LHA, Allen WR, Schenk JL, Squires EL,Bruemmer JE, 2002a: Hysteroscopic insemination of mareswith low numbers of nonsorted or flow sorted spermatozoa.Equine Vet J 34, 128–132.Lindsey AC, Schenk JL, Graham JK, Bruemmer JE, SquiresEL, 2002b: Hysteroscopic insemination of low numbers offlow sorted fresh and frozen ⁄ thawed stallion spermatozoa.Equine Vet J 34, 121–127.Lu KH, Cran DG, Seidel GE, 1999: In vitro fertilization withflow-cytometrically sorted bovine sperm. Theriogenology52, 1393–1405.Martinez EA, Vazquez JM, Roca J, Lucas X, Gil MA, ParrillaI, 2001: Successful non-surgical deep intrauterine inseminationwith small numbers of spermatozoa in sows. Reproduction122, 289–296.Martinez EA, Vazquez JM, Roca J, Lucas X, Gil MA, ParrillaI, Vazquez JL, Day BN, 2002: Minimum number ofspermatozoa required for normal fertility after deep intrauterineinsemination in non-sedated sows. Reproduction123, 163–170.Maxwell WMC, Johnson LA, 1999: Physiology of spermatozoaat high dilution rates: the influence of seminal plasma.Theriogenology 52, 1353–1362.Maxwell WMC, Welch GR, Johnson LA, 1997: Viability andmembrane integrity of spermatozoa after dilution and flowcytometric sorting in the presence or absence of seminalplasma. Reprod Fertil Dev 8, 1165–1178.Maxwell WMC, Long CR, Johnson LA, Dobrinsky JR, WelchGR, 1998: The relationship between membrane status andfertility of boar spermatozoa after flow cytometric sorting inthe presence or absence of seminal plasma. Reprod FertilDev 10, 433–440.Maxwell WMC, Evans G, Hollinshead FK, Bathgate R, deGraaf SP, Eriksson BM, Gillan L, Morton KM, O’BrienJK, 2004: Integration of sperm sexing technology into theART toolbox. Anim Reprod Sci 82, 79–95.Maxwell WM, Parrilla I, Caballero I, Garcia E, Roca J,Martinez EA, Vazquez JM, Rath D, 2007: Retainedfunctional integrity of bull spermatozoa after double freezingand thawing using PureSperm density gradient centrifugation.Reprod Domest Anim 42, 489–494.Merton JS, Haring RM, Stap J, Hoebe RA, Aten JA, 1997:Effect of flow cytometrically sorted frozen ⁄ thawed semen onsuccess rate of in vitro bovine embryo production. Theriogenology47, 295. (Abstract).Moce´ E, Graham JK, Schenk JL, 2006: Effect of sex-sortingon the ability of fresh and cryopreserved bull sperm toundergo an acrosome reaction. Theriogenology 66, 929–936.Morton KM, Catt SL, Hollinshead FK, Maxwell WMC,Evans G2004: Lambs born after in vitro embryo productionfrom prepubertal lamb oocytes and frozen-thawed unsortedand sex-sorted spermatozoa. Reprod Fertil Dev 16, 260.(Abstract).Morton KM, Catt SL, Maxwell WM, Evans G, 2005: Anefficient method of ovarian stimulation and in vitro embryoproduction from prepubertal lambs. Reprod Fertil Dev 17,701–706.Morton KM, Herrmann D, Sieg B, Struckmann C, MaxwellWM, Rath D, Evans G, Lucas-Hahn A, Niemann H,Wrenzycki C, 2007: Altered mRNA expression patterns inbovine blastocysts after fertilisation in vitro using flowcytometricallysex-sorted sperm. Mol Reprod Dev 74, 931–940.Parrilla I, Vazquez JM, Cuello C, Gil MA, Roca J, DiBerardino D, Martınez EM, 2004: Hoechst 33342 stainand u.v. laser exposure do not induce genotoxic effectsin flow-sorted boar spermatozoa. Reproduction 128, 615–621.Probst S, Rath D, 2003: Production of piglets using intracytoplasmicsperm injection (ICSI) with flowcytometricallysorted boar semen and artificially activated oocytes. Theriogenology59, 961–973.Rath D, 2002: Low dose insemination in the sow. A review.Reprod Domest Anim 37, 1291–1304.Rath D, Niemann H, 1996: In vitro production of swineembryos. Dtsch Tierarztl Wochenschr 103, 331–336.Rath D, Niemann H, 1997: In vitro fertilization of porcineoocytes with fresh and frozen-thawed ejaculated or frozenthawedepididymal semen obtained from identical boars.Theriogenology 47, 785–793.Rath D, Johnson LA, Welch GR, 1993: In vitro culture ofporcine embryos: development to blastocysts after in vitrofertilization (IVF) with flow cytometrically sorted andunsorted semen. Theriogenology 39, 293. (Abstract).Rath D, Long CR, Dobrinsky JR, Welch GR, Schreier LL,Johnson LA, 1999: In vitro production of sexed embryos forgender preselection: high-speed sorting of X-chromosomebearingsperm to produce pigs after embryo transfer. J AnimSci 77, 3346–3352.Rath D, Ruiz S, Sieg B, 2003: Birth of female piglets followingintrauterine insemination of a sow using flow cytometricallysexed boar semen. Vet Rec 152, 400–401.Rhodes SL, Maxwell WMC, Evans G1994: DNA analysis andsorting of ram spermatozoa by flow cytometry andsubsequent fertilization of in vitro matured sheep oocytes.In: Proceedings of the 7th International Symposium ofSpermatology, Cairns, Australia, pp. 9.26. (Abstract).Salamon S, Maxwell WMC, 2000: Storage of ram semen.Anim Reprod Sci 62, 77–111.Schenk JL, Seidel GE Jr, 2007: Pregnancy rates in cattle withcryopreserved sexed spermatozoa: effects of laser intensity,staining conditions and catalase. Soc Reprod Fertil Suppl64, 165–177.Seidel GE Jr, 2003a: Economics of selecting for sex: the mostimportant genetic trait. Theriogenology 59, 585–598.Seidel GE Jr, 2003b: Sexing mammalian sperm-intertwining ofcommerce, technology, and biology. Anim Reprod Sci 79,145–156.Seidel GE Jr, Allen CH, Johnson LA, Holland MD, Brink Z,Welch GR, Graham JK, Cattell MB, 1997: Uterine horninsemination of heifers with very low numbers of non-frozenand sexed spermatozoa. Theriogenology 48, 1255–1264.Seidel GE Jr, Schenk JL, Herickhoff LA, Doyle SP, Brink Z,Green RD, Cran DG, 1999: Insemination of heifers withsexed sperm. Theriogenology 52, 1407–1420.Seidel GE Jr, Brink Z, Schenk JL, 2003: Use of heterospermicinsemination with fetal sex as the genetic marker to studyfertility of sexed sperm. Theriogenology 59, 515.Spinaci M, De Ambrogi M, Volpe S, Galeati G, Tamanini C,Seren E, 2005: Effect of staining and sorting on boar spermmembrane integrity, mitochondrial activity and in vitroblastocyst development. Theriogenology 64, 191–201.Suh TK, Schenk JL, Seidel GE Jr, 2005: High pressure flowcytometric sorting damages sperm. Theriogenology 64,1035–1048.Tummaruk P, Sumransap P, Techakumphu M, KunavongkritA, 2007: Distribution of spermatozoa and embryos in thefemale reproductive tract after unilateral deep intra uterineinsemination in the pig. Reprod Domest Anim 42, 603–609.Vazquez JM, Martinez EA, Parrilla I, Roca J, Gil MA,Vazquez JL, 2003: Birth of piglets after deep intrauterineinsemination with flow cytometrically sorted boar spermatozoa.Theriogenology 59, 1605–1614.Vazquez JM, Martinez EA, Parrilla I, Cuello C, Gil MA,Garcia E, 2006: Improving the efficiency of laparoscopicintraoviductal insemination with sex sorted boar spermatozoa.Reprod Fertil Dev 18, 283. (Abstract).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

346 D Rath and LA JohnsonWalker SK, Smith DH, Little DL, Warnes GM, Quinn P,Seamark RF, 1984: Artificial insemination and transfer ofembryos by laparoscopy. In: Lindsay DR, Pearce DT(eds),Reproduction in Sheep. Australian Academy of Science andAustralian Wool Corporation, Canberra, pp. 306–309.Weigel KA, Barlass KA, 2003: Results of a producer surveyregarding crossbreeding on US dairy farms. J Dairy Sci 86,4148–4154.Welch GR, Johnson LA, 1999: Sex preselection: laboratoryvalidation of the sperm sex ratio of flow sorted X- and Y-sperm by sort reanalysis for DNA. Theriogenology 52,1343–1352.Author’s address (for correspondence): D. Rath, Institute of FarmAnimal Genetics, Mariensee (FLI), 31535 Neustadt, Germany.E-mail: detlef.rath@fli.bund.deConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 347–354 (2008); doi: 10.1111/j.1439-0531.2008.01183.xISSN 0936-6768Low-Dose Insemination in Pigs: Problems and PossibilitiesJM Vazquez, J Roca, MA Gil, C Cuello, I Parrilla, I Caballero, JL Vazquez and EA Martı´nezDepartment of Animal Medicine and Surgery, Faculty of Veterinary Medicine, Campus de Espinardo, University of Murcia, Murcia, SpainContentsLow-dose AI procedures are required by the pig industry toefficiently utilize emerging sperm technologies, such ascryopreservation and sex-sorting. Currently, several differentprocedures for inseminating with a low or very low numberof spermatozoa have been described. Deep intrauterineinsemination allows the deposition of the spermatozoa inthe depth of the uterine horn, allowing a significantreduction in the number of spermatozoa inseminated withmaintenance of optimal reproductive performance. Intraoviductallaparoscopic insemination has been recentlyapplied in pigs. This technique has proved to be applicablewith diluted and sex-sorted spermatozoa. This reviewdiscusses several problems encountered during the developmentof deep intrauterine insemination and intra-oviductallaparoscopic insemination of pigs and provides potentialsolutions for the practical application of both the technologies.IntroductionLow-dose insemination techniques are recommended inpigs when the available number of spermatozoa islimited, since standard pig artificial insemination (AI)protocols employ 3 · 10 9 spermatozoa per dose depositedintracervically two or three times during oestrus.Under these standard conditions, one ejaculate can beused to inseminate only a limited number of sows, thusconstraining the efficient use of the boars. The standardAI technique is either unsuitable or inefficient whenapplied to the emerging sperm technologies, such asfrozen-thawed spermatozoa and sperm sexing. For thesereasons, a new procedure has been developed fordepositing spermatozoa deep into the uterine horn(DUI), which allows a reduction of the number ofspermatozoa per dose (Martinez et al. 2001a,b, 2002).When fresh semen is used under field conditions, thesperm dose can be reduced to 150 · 10 6 spermatozoawith acceptable fertility results (Martinez et al. 2002,2006). This practical procedure should offer a greatbenefit for the optimization of the use of fresh semenfrom superior boars or in sanitary contingencies whenthe number of doses to be used is decreased. Currently,the DUI technique has the potential to achieve highfertilityresults using as few as 1–2 · 10 9 total frozenthawedspermatozoa (Roca et al. 2003, 2006). The DUItechnology has the potential to counteract factorslimiting routine application of frozen-thawed spermatozoa,such as the normally high number of spermatozoarequired per dose and the low fertility achieved.The most sought after reproductive technology is preconceptionsex pre-selection. The only accurate andpotentially cost-effective approach for achieving sex preselectioninvolves separating the X- from the Y-chromosomebearing spermatozoa using flow cytometry andsperm sorting (Johnson et al. 2005). However, thenumber of available flow-sorted spermatozoa is toolow for an extended use of the technology in pigproduction, even using DUI methodology (Vazquezet al. 2003, 2005). Laparoscopic insemination into theoviduct (ILI) might be an alternative method, at leastwhen applied under specialized production situations(Vazquez et al. 2006).The following review identifies objections and problems,which appeared during the development of theDUI and ILI procedures, and describes potentialsolutions for practical application of both technologies.Low Dose: Deep Intrauterine InseminationTwo main objections have been raised to the procedureof DUI in sows: (1) putative damage to the cervix andthe uterine wall by the insemination device during itsadvancement along the lumen of the cervical canal andthe uterus and (2) the incidence of the unilateralfertilization when 150 · 10 6 fresh spermatozoa are used.Putative damage of the cervix and uterine wall by the DUIcatheterIt has been suggested that the DUI catheter may causedamage to the cervix and the uterus mucosa during itsinsertion in the sow reproductive tract, potentiallycompromising subsequent fertility. Using a fibre opticendoscope procedure developed in our laboratories fornon-surgical DUI in non-sedated sows (Martinez et al.2001a), slight bleeding of the cervical canal wasobserved in 3 out of 33 sows (9.1%) during the insertionof the endoscope. In addition, during removal of theendoscope from the genital tract, a visible mark on theendometrium of the first uterine curvature was found inseven animals (21.2%). The endometrial damage wasnot accompanied by internal or external bleeding in anysow and did not have a detrimental effect on fertility ofhormonally treated oestrous sows used in that study.Recently, results from Australia indicated that a highproportion of sows (27%) bled during or after DUI andthat 22% of the inseminations resulted in some form ofbleeding from the reproductive tract (Bathgate et al.2007). In that paper, the type of bleeding was dividedsubjectively into three categories: cervical (blood wasseen on the catheter or at anytime during insemination),uterine (blood not noticed during insemination, butbecame evident 12 or more hours after insemination)and both (both types of blood were observed). Theauthors did not find significant differences in non-returnrates, farrowing rates and litter sizes between sows withor without bleeding. The high proportion of sows withÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

348 JM Vazquez, J Roca, MA Gil, C Cuello, I Parrilla, I Caballero, JL Vazquez and EA Martı´nezbleeding as a result of the insertion of the DUI catheterobtained by the Australian group is difficult to explain ifwe take into account other researches where theincidence of bleeding during or after DUI was low ornon-existent (Martinez et al. 2001a, 2002, 2006; Dayet al. 2003; Rath et al. 2003; Roca et al. 2003; Vazquezet al. 2003; Bolarin et al. 2005a; Grossfeld et al. 2005;Wongtawan et al. 2006). In a recent trial performed on acommercial farm in Spain, we could detect blood on theDUI catheter after insemination in only 2 out of 95 sowsinseminated which was similar to that observed in sowsinseminated with the traditional AI method (1 out of95). Each of those three sows farrowed a high number ofpiglets (Martinez et al. 2006). More recently, similardata on bleeding have been obtained by our group usingDUI methodology (6 sows with blood out of 108) andstandard AI (4 sows with blood out of 114) with noinfluence on performance reproductive of the sows(unpublished data).While it seems clear that the presence of blood duringor after insemination does not exert a detrimental effecton fertility results after DUI, the contradictory results inrelation to the incidence of bleeding as a consequence ofthe insertion of a DUI catheter remain to be explained.Several factors could be implicated in the differencesobserved in relation to the bleeding phenomenon. First,it is evident that during the development of any newtechnology, problems during the period of training areexpected. Deep into the uterine horn technologyrequires more patience and skill when compared tostandard AI procedures. In addition, improper manipulationof the device when placing the catheter into theuterus could result in damage of the cervix and ⁄ orthe uterus. The device should always be inserted carefullyand not forced when some resistance is encountered.Although the DUI procedure can be performed safelyand quickly, a minimum period of training is necessaryin order to achieve optimal results. Second, the differencesmay be due to disparity between the commercialcatheters and the experimental prototype used in thepublished reports. While the experimental prototypeswere carefully hand-made, some abnormalities wereobserved in several parts of the DUI catheter in somebatches of the first commercial catheters constructed.This limitation has been solved by a later refinement ofthe DUI catheters carried out by a German company(Minitube, Tiefenbach, Germany). Currently, the incidenceof bleeding after the DUI procedure using thecurrent commercial DUI catheter (Minitube) is low andsimilar to that observed after the standard AI technique.In our first studies, using a fibre optic endoscope, avisible mark on the endometrium of the first uterinecurvature was visible in a high proportion of sows(Martinez et al. 2001a). Recently, Bathgate and coworkersperformed an experiment in order to examineendometrial trauma resulting from the DUI technique(Bathgate et al. 2007). In this experiment, five sows duefor slaughter were weaned at a time to match the onsetof oestrus with their arrival at the abattoir. Less than1 h prior to slaughter, the DUI catheter was inserted asif the sow was to be inseminated, without deposition ofthe semen. Immediately after slaughter, the entirereproductive tract was removed from each sow andtransported to the laboratory where the tracts weredissected and inspected for damage to the uterineendometrium. The authors observed slight damage tothe endometrial lining in all sows and that none of thesesows bled during or after the insemination. These resultsare in agreement with our previous observations usingthe endoscope (Martinez et al. 2001a). While in ourstudy the endometrial damage did not affect the fertilityof hormonally treated oestrous sows, the implications ofany damage to the reproductive tract of the sowsinduced by insertion of the DUI catheter needs to beconsidered, as suggested by Bathgate et al. (2007).Consequently, two experiments have been conductedby our group. In the first experiment (Bolarin et al.2005b), the effect on future fertility of DUI catheterinsertion into the uterine horn during deep intrauterineinsemination was evaluated. A total of 159 weaned sows(parity 2–8) were twice DUI-inseminated in a commercialfarm during 10 months with 1–2 · 10 9 frozenthawedspermatozoa. Ninety-nine sows (62.7%) werepregnant and 97 (61.0%) farrowed with a mean littersize of 9.6 ± 0.28. Reproductive data on pigs inseminatedby standard AI before and after DUI wererecorded and compared (Table 1). No differences(p > 0.05) were found in pregnancy and farrowingrates nor litter size for AI before and after DUI.Moreover, ‘not in pig’ rates were also similar. Fromthese results, we concluded that the insertion of the DUIcatheter into the uterine horn did not affect subsequentfertility of the sows.In the second experiment (Rodriguez-Vilar et al.2006), the effect of insertion of the DUI device into oneuterine horn on the reproductive performance ofmultiparous sows was evaluated. A total of 51 sowswere conventionally inseminated twice with cooledsemen (3 · 10 9 sperm ⁄ 100 ml doses) at 0 and 24 hafter onset of oestrus. At the time of the firstinsemination, the sows were randomly divided intotwo groups. In one group, a DUI device was insertedat the same time that conventional insemination wascarried out (DUI group). The remaining 23 sows wereonly conventionally inseminated (control group). Therewere no differences (p > 0.05) between the DUI andcontrol groups for pregnancy and farrowing rates orlitter size (Table 2). A supplementary study with ahigher number of sows (n = 186) offered similarresults (unpublished data). These results show thatthe insertion of the DUI catheter into the uterus doesnot affect the reproductive performance of multiparoussows.Table 1. Reproductive performance of sows traditionally inseminatedwith 3 · 10 9 fresh spermatozoa ⁄ 100 ml of extender before and afterthe use of the DUI procedure (from Bolarin et al. 2005b)ParameterStandard AIbefore DUI insertion(n = 159)Standard AIafter DUI insertion(n = 159)Pregnancy rate (%) 87.4 86.2Farrowing rate (%) 84.9 84.3Total piglets born11.1 ± 0.17 11.5 ± 0.15(mean ± SEM)Not in pig (%) 2.9 2.2Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Low-Dose Insemination in Pigs 349Table 2. Effect of the insertion of the DUI catheter in sowstraditionally inseminated with 3 · 10 9 fresh spermatozoa ⁄ 100 ml ofextender on fertility parameters (from Rodriguez-Vilar et al. 2006)ParameterStandard AI plusDUI catheter insertion(n = 28)Standard AIwithout DUI catheterinsertion (n = 23)Pregnancy rate (%) 89.3 91.3Farrowing rate (%) 85.7 89.3Total piglets born(mean ± SEM)11.0 ± 0.54 11.1 ± 0.53It may be concluded that: (1) the incidence of bleedingafter the DUI procedure is low and similar to thatobserved after standard AI when practiced by suitablytrained personnel and with appropriate DUI cathetersand (2) the insertion of the DUI catheter through thecervix into the uterus does not produce significantdamage to the cervix and ⁄ or uterine wall and does notadversely affect the present and ⁄ or future reproductiveperformance of the sows.Incidence of the unilateral fertilization following DUIwith 150 · 106 spermatozoaThe earliest experiments carried out in our laboratory,comparing DUI with standard AI (3 · 10 9 spermatozoain 80–100 ml), indicated that DUI allows a 20-foldreduction in the number of fresh spermatozoa inseminated,and at least a 16–20-fold reduction in the dosevolume, without affecting farrowing rate or litter size inweaned sows, hormonally induced to ovulate (Martinezet al. 2002). However, preliminary studies using theDUI procedure on weaned, untreated sows under farmconditions indicated that although pregnancy and farrowingrates were not affected, litter sizes after DUI(150 · 10 6 fresh spermatozoa ⁄ 5 ml) could be up to 1.0piglet less (p < 0.01) than after standard AI (3 · 10 9spermatozoa ⁄ 95 ml) (Vazquez et al. 2001; Day et al.2003). More recent data using spontaneously ovulatingweaned sows confirmed these results (Martinez et al.2006).Various experiments have been conducted to determinethe possible factors implicated in such a reductionof prolificacy. In one of them (Martinez et al. 2006), 42weaned sows were inseminated at 12, 24 and 36 h afterthe onset of spontaneous oestrus using one of thefollowing two regimes: (1) DUI (treatment) with150 · 10 6 fresh spermatozoa in 5 ml of BTS extenderor (2) standard cervical AI (control) with 2.85 · 10 9fresh spermatozoa in 95 ml of BTS extender. On day 6,after onset of oestrus, the proximal segment of theindividual uterine horns of the sows were surgicallyflushed under anaesthesia to retrieve ova ⁄ embryos andevaluate the success of fertilization per horn (e.g.occurrence of effective uni- vs bilateral sperm transportrendering uni- or bilateral, complete or partial fertilization).Retrieved embryos were assessed for cleavage andnumber of accessory spermatozoa. Although identicaloverall fertilization rates were achieved in both inseminationgroups, the percentage of sows with partialbilateral fertilization and unilateral fertilization wasmarkedly higher (p < 0.05) in the DUI group (35%)compared with the control (standard AI) group (5%),with a consequent lower (p < 0.001) percentage ofviable early embryos after DUI. The number of spermbound to the zona pellucida (ZP) was higher in embryosfrom control sows than in DUI sows, indicating that thenumber of spermatozoa reaching the sperm reservoirslargely depends on the number of spermatozoa inseminated,as reported in other species (Morton and Glover1974; Nadir et al. 1993).Our results also indicate a high individual variabilityin the number of accessory spermatozoa in the ZP ofembryos from sows inseminated on the same day andwith seminal doses originating from the same ejaculates.This variability was also observed within sow betweenuterine horns, regardless of the AI method used (Fig. 1).In addition, no accessory spermatozoa were found inoocytes retrieved from sows with unilateral fertilization(Fig. 2). This indicates that spermatozoa were not ableto reach the contralateral oviduct in these sows andsuggests that intrinsic characteristics of the sow (i.e.myometrial contractility and ⁄ or local phagocytosis byinvading leucocytes) could be implicated in the incidenceof unilateral fertilizations. It is known that once thesperm dose is deposited deeply into a uterine horn,spermatozoa are able to reach the contralateral oviductand to fertilize a high proportion of oocytes (Martinezet al. 2002; Tummaruk et al. 2007) although spermatozoacan be recovered from only one oviduct via theflushing technique (Tummaruk et al. 2007). It could bespeculated that in some sows, the number of spermatozoareaching the contralateral oviduct might be too lowto establish an efficient sperm population in the reservoirto fertilize the oocytes, and as a consequence, theFig. 1. Individual variations in the number of spermatozoa bound tothe ZP of normal embryos in sows inseminated three times duringoestrus either by deep intrauterine (DUI; 150 · 10 6 spermatozoa⁄ 5 ml) or in contemporaries inseminated cervically (AI;2.85 · 10 9 spermatozoa ⁄ 95 ml). The sows were inseminated on thesame day using doses from the same batch of semen. Each dose ofsemen used for DUI was taken directly from the same bottle used toinseminate the contemporary control sow. *Differences betweenuterine horns of the same sow (p £ 0.002). (a) Differences comparedto the same uterine horn of the contemporary sow (at least p < 0.01).Unilateral fertilization was observed in sow DUI 3. In that sow, allstructures recovered from uterine horn 2 were oocytes with nospermatozoa bound to the ZP and the differences in sperm countbetween her uterine horns or in relation to uterine horn 2 of hercontemporary control AI sow were not analysed (from Martinez et al.2006)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

350 JM Vazquez, J Roca, MA Gil, C Cuello, I Parrilla, I Caballero, JL Vazquez and EA Martı´nezFig. 2. Number of spermatozoa bound to the ZP of normal embryosin deep intrauterine inseminated sows (DUI; 0.15 · 10 9 spermatozoa⁄ 5 ml) with unilateral fertilization and in contemporary controlsows (AI; 2.85 · 10 9 spermatozoa ⁄ 95 ml). Each dose of semen usedfor DUI was taken directly from the same bottle used to inseminate thecontemporary control sow. Sows from both groups were inseminatedthree times during oestrus. *Differences between uterine horns of thesame sow (at least p < 0.002). (a) Differences compared to the sameuterine horn of the contemporary sow (at least p < 0.01). Allstructures recovered from uterine horn 2 in DUI sows were oocyteswith no spermatozoa bound to the ZP and differences between uterinehorns of the same sow or in relation to the uterine horn 2 of AI sowswere not analysed (from Martinez et al. 2006)oocytes of that oviduct are not fertilized. Although twodifferent pathways (transperitoneal and intrauterine)have been confirmed for the transport of spermatozoainto the reproductive tract after the DUI procedure(Martinez et al. 2005), more investigations are needed tounderstand the mechanisms by which spermatozoacolonize the oviducts (reviewed by Rodriguez-Martinezet al. 2005) after DUI with low numbers of spermatozoa.From the above results, it can be concluded that thelower litter size after DUI with 150 · 10 6 fresh spermatozoain 5 ml of BTS extender is related to a decreaseddistribution of spermatozoa, leading to a higher incidenceof partial bilateral and unilateral fertilization thanwith standard AI.The discrepancies existing between our first experimentsusing hormonally induced oestrous sows (Martinezet al. 2001a, 2002) and other studies using sows inspontaneous oestrus (Vazquez et al. 2001; Day et al.2003; Martinez et al. 2006), in relation to the differencesobserved in the litter size between DUI and standard AIgroups, could be related to the ovulatory response to thehormonal treatment. It is possible that the hormonaltreatment of sows inseminated in the depth of oneuterine horn with 150 million spermatozoa increased thenumber of ovulations from each ovary and, as a result, agreater number of oocytes could have been fertilized inrelation to standard AI sows (spontaneous oestroussows), even with unilateral or partial bilateral fertilization.Consequently, the litter sizes were similar betweenboth groups of sows.There are several possibilities for solving the problemof unilateral fertilization after DUI. It is evident that thenumbers of spermatozoa, volume of inseminates and ⁄ orthe place of deposition of semen plays a major role.Thus, we have observed that fertilization rates and thepercentage of bilateral fertilization after DUI with600 · 10 6 spermatozoa in 20 ml of BTS extender didnot differ from those of the standard AI group (97.8%and 100% vs 98.5% and 100%, respectively) (Martinezet al. 2005). Other factors, such as the insemination–ovulation interval, uterine contractility and phagocytosis,could also be implicated in the incidence ofunilateral fertilization when DUI is used under fieldconditions. Therefore, hormonal treatments to controlovulation, stimulants of myometrial contractility andthe addition of substances for decreasing local phagocytosiscould be ways to increase the litter size whensows in spontaneous oestrus are inseminated using DUIwith 150 · 10 6 fresh spermatozoa.In conclusion, the incidence of unilateral fertilizationin sows in spontaneous oestrus after DUI with 150 · 10 6fresh spermatozoa in 5 ml of extender can be overcomeby increasing the number of spermatozoa and volume ofthe dose to 600 · 10 6 and 20 ml, respectively. Otherpotential remedies that influence uterine contractibilityand local phagocytosis remain to be explored.Very Low Dose: Laparoscopic InseminationThe deposition of the spermatozoa higher in thereproductive tract than the uterus (uterotubal junctionor oviduct) allows a greater proportion of spermatozoato survive and colonize the oviduct; and therefore, fewerwould be necessary to achieve the same probability offertility than with standard AI (Watson 2000). Techniquesto deposit a very low number of spermatozoa in asmall volume of inseminate would increase the efficiencyof insemination when using spermatozoa with a highadded value (e.g. sex-sorted semen or sperm-mediatedgene transfer) (Rath 2002; Johnson et al. 2005). Inseminationinto the uterotubal junction or the oviductwould reduce the number of spermatozoa required tothe minimum possible, decreasing the loss of spermatozoaby phagocytosis or backflow.Successful fertility results have been obtained whenfrozen-thawed (Maxwell and Salamon 1977) or sexsortedspermatozoa (Johnson 1991) were depositeddirectly into the oviduct by laparotomy. Moreover, itis well known that optimal fertility can be achieved withonly 1 · 10 7 spermatozoa after surgical inseminationnear the utero-tubal junction, instead of the 3 · 10 9spermatozoa normally inseminated in the cervix (Kruegeret al. 1999). The number of spermatozoa requiredfor insemination is directly dependant on the place ofdeposition of these spermatozoa (Vazquez et al. 2005).The disadvantage of a protocol based on laparotomy isthat the invasive procedure cannot be applied on farms.Surgical laparotomy results in distress for the sows andin post-operative adhesions that complicate the subsequentinseminations.Laparoscopy is a less-invasive technique than laparotomyfor depositing semen directly into the uterus oroviduct and could be performed on farm by technicallytrained personnel. This procedure provides sharp andmagnified images of the genital tract. Although laparoscopyhas been routinely applied to commercialinseminations of sheep with a low number of spermatozoa(reviewed by Maxwell and Watson 1996), theÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Low-Dose Insemination in Pigs 351application in pigs should be further evaluated before itcan be applied on farm. The effectiveness (safety andsuccess) of this insemination procedure as a solution forthe application of spermatozoa with a high added value(very low dose) is discussed.Laparoscopic procedure in pigsLaparoscopy is an alternative procedure that allowsaccess into the abdominal cavity. This methodology isless invasive than laparotomy since a large incision isreplaced by several smaller incisions. The laparoscopicprocedure reduces the exposure of the peritoneal cavity,and the handling of the bowel is avoided (Garry andFrancog 2006).However, laparoscopic insemination hasthe disadvantage that, although minimally invasive, it isalso a surgical procedure. Moreover, it is important totake into consideration that laparoscopy deprives theoperator of three-dimensional vision when comparedwith full laparotomy procedures.While laparoscopic insemination has been successfullyused in millions of humans and animals worldwide, theexperience with pigs is limited. We first developed amodel in sows to carry out laparoscopic insemination.The sows, under general anaesthesia, were placed in aTrendelenburg position at an angle of approximately20–30° above horizontal. A 2-cm mid-line incision wasperformed at the umbilical level, where a Veres needle(Veres Insufflation Needle, Aesculap Inc., Center Valley,USA) was inserted into the peritoneal cavity to insufflatecarbon dioxide to establish the pneumoperitoneum andto distend the abdominal cavity to ensure the visualizationof the genital structures.Subsequently, a trocar-canula was inserted in the sameincision to allow the passage of the laparoscope, maintainingthe gas pressure by a device attached to the trocarcanula.The need to insert the trocar-canula as well as thedistension of the cavity are weaknesses of laparoscopy.Initially, in our experimental trials, approximately 200laparoscopic inseminations were conducted using thisprocedure. The two main disadvantages observed were:(1) the time taken to insufflate carbon dioxide into theperitoneal cavity and (2) the risk of gastrointestinal(bowel or vessel) injury during the laparoscopy.Gas insufflation provides a safe space to insert theadditional trocars as well as allowing the clear visualizationof the genital structures. Our experience indicatesthat the gas pressure required for insemination inpigs should be 13–15 mmHg. The time needed to reachthis pressure in pigs was 8–10 min due to the diameter ofthe Veres needle and the pressure output of thelaparoscopic insufflator. Consequently, increasing thecarbon dioxide flow rate should decrease the timerequired to reach the correct intra-abdominal pressure.The incidence of the gastrointestinal injury wasapproximately 0.5–1% based on the diagnosis madeduring exploration of the abdominal cavity. The blindinsertion of the Veres needle and the first trocar maycause complications, especially in large and fat sows, asobserved with obese human patients. It is also importantto take into consideration that this technique requiresspecific instrumentation and, especially, a skilled operator.A number of modifications to the laparoscopicprocedure have been developed based on these limitations.As an alternative to the Veres needle, opticaltrocars can be safely positioned under direct visualization.These trocars have been shown in humans to besafe and easy to handle, offering several advantages overthe use of the Veres needle and the mini-laparotomy(String et al. 2001). In our laboratory, we carried out anexperiment using a 12-mm Optiview trocar (EthiconEndo-Surgery, Cincinnati, OH, USA) instead of theVeres needle. This trocar and a 0° laparoscope areinserted and advanced into the wound, allowing the tipof the trocar to be viewed while it is entering theabdominal cavity. After the cavity is entered, the handpiece of the Optiview (Ethicon Endo-Surgery) isremoved and replaced by the 0° laparoscope, and theabdominal cavity directly inflated to 13–15 mmHg withcarbon dioxide. Using this procedure, less than 3 minare required to insufflate the abdominal cavity, and nogastrointestinal injuries have been observed after theinsemination of nearly 500 sows.Once the reproductive tract has been located, thelaparoscope allows the in situ observation of ovarianstatus. Two accessory posts are placed in the right andleft part of the hemi-abdomen, which provides accessfor laparoscopic Duval forceps for manipulation ofthe uterine horn and grasping of the uterus or theoviduct for the insemination needle. To carry out theinsemination, the needle is inserted through the UTJor oviduct wall with a fast, precise and controlledmovement and spermatozoa are injected into thelumen. In some cases (low percentage,

352 JM Vazquez, J Roca, MA Gil, C Cuello, I Parrilla, I Caballero, JL Vazquez and EA Martı´nezAlthough the number of spermatozoa deposited bylaparotomy is very low when compared with the numberrequired for standard AI, 10 · 10 6 could still be a highnumber when valuable spermatozoa (e.g. sex-sorted) areused. Furthermore, failures of the insemination havebeen observed in 29% of the animals and no fertilizationwas observed in one uterine horn (Fantinati et al. 2005).In our own experience, once the UTJ is identified, theneedle is advanced to touch the uterine part and theninserted into the uterine lumen. However, it is easy toaccidently introduce the tip of the needle into the uterinewall, rather than into the lumen, resulting in inseminationfailure. The difficulty in feeling the location of thetip of the needle makes it very hard to ensure its correctpositioning in the reproductive tract, even for welltrainedoperators. Only after expulsion of the semen,does the absence of progressive swelling of uterine tip, aswell as the resistance encountered while attempting toeject the semen, indicate that semen has been depositedincorrectly.A variation of this technique has been developed inour laboratory by deposition of semen directly into theoviduct. This laparoscopic insemination method is easyand fast in pigs when done by technically trainedpersonnel. The main advantage is that the number ofspermatozoa deposited can be much lower than thatrequired in the UTJ as demonstrated by the goodfertility when frozen-thawed (Maxwell and Salamon1977) or sex-sorted spermatozoa (Johnson 1991) weredeposited directly into the oviduct by laparotomy.Moreover, the correct positioning of the tip of theneedle in the lumen of the oviduct is much easier toachieve than for UTJ inseminations. Nevertheless, it isimportant to take into consideration that entry ofspermatozoa into the oviduct is normally a highlyselective process, being dependent mainly on the intrinsicmotility of each individual spermatozoan. Thisselectivity is abolished when the spermatozoa aredeposited directly into the oviduct.Efficiency of the intra-oviductal insemination techniqueIt is well known that the polyspermy block from theporcine oocyte is low when compared with other species(Day 2000) and, in vitro, the sperm : oocyte ratio affectsthe percentage of polyspermic oocytes (Gil et al. 2004,2007). From more than 50 · 10 9 spermatozoa depositedinto the sows after natural mating, only tens to hundredsof sperm reached the oviduct (Hunter 2002). Consequently,a high percentage of polyspermic oocytes couldbe expected after intra-oviductal laparoscopic inseminationwhen compared with conventional insemination,since a higher than physiological number of spermatozoaare deposited in the oviduct. As expected, a highpercentage of oocytes were penetrated after intraoviductallaparoscopic insemination with 0.3, 0.5 or1 · 10 6 spermatozoa per oviduct in our laboratory.Surprisingly, however, regardless of the high number ofspermatozoa deposited in the oviductal ampulla, theincidence of polyspermic penetration was very low when0.3 or 0.5 · 10 6 spermatozoa were inseminated beforeovulation (Table 3). These results show the potentialeffectiveness of laparoscopic insemination as a methodTable 3. Percentages of penetrated and monospermic oocytes in sowsat an induced ovulation following intra-oviductal laparoscopic inseminationof 0.3, 0.5 or 1 million spermatozoa per oviductInseminationDose (·10 6 )% Oocytespenetrated% Monospermy0.3 91.8 ± 1.5 88.1 ± 2.1 a0.5 92.2 ± 1.8 86.1 ± 2.1 a1.0 95.0 ± 2.1 68.6 ± 29 bDifferent letters within the same column represent a significant difference(p < 0.05).for obtaining pregnancies with very low numbers ofspermatozoa. There was evidence of polyspermy onlywhen 1 · 10 6 spermatozoa were inseminated.It is well known that the presence of the gametes inthe oviduct modifies the amount of protein in theoviductal fluid. Nineteen proteins have been identifiedin vitro that are specifically regulated by spermatozoa,four proteins regulated specifically by oocytes, and oneprotein that is commonly regulated by both spermatozoaand oocytes (Georgiou et al. 2005). The oviductalenvironment, and consequently the fertilization process,also depends on the hormones secreted by the follicles(Hunter 2002). We have recently identified a number ofproteins that are regulated by the presence of spermatozoaor oocytes in the oviduct in vivo (Georgiou et al.2007). Therefore, the oviductal environment may modulatethe penetration of oocytes by spermatozoa as aconsequence of their deposition directly into the oviduct.All of the proteins regulated by the gametes arewell known to influence gamete maturation, viabilityand function. Consequently, the deposition of thespermatozoa before, during or after the arrival of theoocyte in the oviductal ampulla should be an importantvariable to be considered in the analysis of the efficiencyof intra-oviductal laparoscopic insemination.Recently, we have evaluated the effect of the timing ofinsemination in relation to ovulation (pre-ovulatory,ovulating or ovulated follicles) on the penetration andpolyspermic fertilization of oocytes (Table 4). The rateof polyspermy was very high when sows with ovulatedoocytes were inseminated directly into the oviduct.While polyspermy has been associated with aged oocytes(Hunter 2002), in this experiment, it seemed to be morerelated to the oviductal environment than the age ofthe oocytes, since all of the sows ovulated only 1–3 hbefore insemination. When the spermatozoa were presentin the oviduct before ovulation (pre-ovulatoryinseminations), the rate of polyspermy was very low.Table 4. Percentages of penetrated and monospermic oocytes in sowsat an induced ovulation following intra-oviductal laparoscopic insemination(0.3 spermatozoa per oviduct) before, during or after ovulationInsemination timerelative to ovulation% Oocytespenetrated% MonospermyBefore 98.1 ± 1.3 97.5 ± 1.1 aDuring 89.3 ± 2.2 87.8 ± 2.1 bAfter 94.3 ± 2.7 49.8 ± 2.9 cDifferent letters within the same column represent a significant difference(p < 0.05).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Low-Dose Insemination in Pigs 353However, if the oocytes were present at the time ofinsemination (peri-ovulatory insemination), polyspermywas significantly higher (p < 0.05).Using the intra-oviductal laparoscopic inseminationprocedure, pregnancy rates higher than 80% have beenobtained with only 0.3 · 10 6 diluted or sex-sortedspermatozoa per oviduct (Vazquez et al. 2006). Highfertility has also been reported after intra-oviductallaparoscopic insemination with sex-sorted spermatozoainseminated in presence of the PSP-I ⁄ PSP-II heterodimerextracted from boar seminal plasma (Garcia et al.2007).ConclusionsUsing appropriate insemination procedures, it is nowfeasible to achieve high fertility rates in pigs with low orvery low doses of spermatozoa. When compared withstandard AI, post-cervical insemination allows a threefoldreduction in the number of cooled spermatozoa anddeep intrauterine insemination allows a substantialreduction in the number of cooled (5–20-fold) orfrozen-thawed (sixfold) spermatozoa. Intra-oviductalinsemination by laparoscopy successfully allows inseminationswith only 0.3 · 10 6 sex-sorted spermatozoa. Allthese insemination procedures may be useful tools forthe pig industry.AcknowledgementsThe authors wish to thank Professor W.M.C. Maxwell for helpfuldiscussion and critical reading of the manuscript. Financial supportfrom CICYT (AGF98-0533; AGL2001-0471; AGL2004-07546 andAGF2005-00760), CDTI (B288 ⁄ 98; 03.180; 03-402, 04-0231), INIA(RZ01-019) and Fundacion Séneca de Murcia (PB ⁄ 74 ⁄ FS ⁄ 02,03002 ⁄ PI ⁄ 05 and 04543 ⁄ GERM ⁄ 07).ReferencesBathgate R, Eriksson BM, Thomson PC, Maxwell WMC,Evans G, 2007: Field fertility of frozen-thawed boar spermat low doses using non-surgical, deep uterine insemination.Anim Reprod Sci. doi:10.1016/j.anireprosci.2007.01.008.Bolarin A, Roca J, Rodriguez-Martinez H, Hernandez M,Vazquez JM, Martinez EA, 2005a: Dissimilarities in sows’ovarian status at the insemination time could explaindifferences in fertility between farms when frozen thawedsemen is used. Theriogenology 65, 669–680.Bolarin A, Hernandez M, Vazquez JM, Martinez EA, Roca J,2005b: Catheter insertion into uterus horn during deepintrauterine insemination does not compromise futurefertility potential of sows. Reprod Domest Anim 40, 340(abstract).Brussow KP, Torner H, Ratky J, Manabe N, Tuchscherer A,2006: Experimental evidence for the influence of cumulus–oocyte-complexes on sperm release from the porcine oviductalsperm reservoir. J Reprod Dev 52, 249–257.Day BN, 2000: Reproductive biotechnologies: current status inporcine reproduction. Anim Reprod Sci 60–61, 161–172.Day BN, Mathias K, Didion BA, Martinez EA, Caamano JN,2003: Deep intrauterine insemination in sows: first field trialin USA commercial farm with a newly developed device.Theriogenology 59, 213 (Abstract).Fantinati P, Zannoni A, Bernardini C, Webster N, LavitranoM, Forni M, Seren E, Bacci ML, 2005: Laparoscopicinsemination technique with low numbers of spermatozoa insuperovulated prepuberal gilts for biotechnological application.Theriogenology 63, 806–817.Garcia EM, Vazquez JM, Parrilla I, Calvete JJ, Sanz L,Caballero I, Roca J, Vazquez JL, Martinez EA, 2007:Improving the fertilizing ability of sex sorted boar spermatozoa.Theriogenology 68, 771–778.Garry R, Francog MDF, 2006: Laparoscopic surgery. BestPract Clin Obstet Gynaecol 20, 89–104.Georgiou AS, Sostaric E, Wong CH, Snijders AP, Wright PC,Moore HD, Fazeli A, 2005: Gametes alter the oviductalsecretory proteome. Mol Cell Proteomics 4, 1785–1796.Georgiou AS, Snijders APL, Sostaric E, Afaltoonian R,Vazquez JL, Vazquez JM, Roca J, Martinez EA, WrightPC, Fazeli A, 2007: Modulation of the oviductal environmentby gametes. J Proteome Res 6, 4656–4666.Gil MA, Ruiz M, Cuello C, Vazquez JM, Roca J, MartinezEA, 2004: Influence of sperm: oocyte ratio during in vitrofertilization of in vitro matured cumulus-intact pig oocyteson fertilization parameters and embryo development. Theriogenology61, 551–560.Gil MA, Alminana C, Cuello C, Parrilla I, Roca J, VazquezJM, Martinez EA, 2007: Brief coincubation of gametes inporcine in vitro fertilization: Role of sperm: oocyte ratio andpost-coincubation medium. Theriogenology 67, 620–626.Grossfeld R, Klinc P, Sieg B, Rath D, 2005: Production ofpiglets with sexed semen employing a non-surgical inseminationtechnique. Theriogenology 63, 2269–2277.Hunter RHF, 2002: Vital aspects of Fallopian tube physiologyin pigs. Reprod Domest Anim 37, 186–190.Johnson LA, 1991: Sex preselection in swine: altered sex ratioin offspring following surgical insemination of flow sortedX- and Y-bearing sperm. Reprod Domest Anim 26, 309–314.Johnson LA, Rath D, Vazquez JM, Maxwell WMC, DobrinskyJR, 2005: Preselection of sex of offspring in swine forproduction: current status of the process and its application.Theriogenology 63, 615–624.Krueger C, Rath D, Johnson LA, 1999: Low dose inseminationin synchronized gilts. Theriogenology 52, 1363–1373.Martinez EA, Vazquez JM, Roca J, Lucas X, Gil MA, ParrillaI, Vazquez JL, Day BN, 2001a: Successful non-surgical deepintrauterine insemination with low number of spermatozoain sows by a fiberoptic endoscope technique. Reproduction122, 289–296.Martinez EA, Vazquez JM, Roca J, Lucas X, Gil MA,Vazquez JL, 2001b: Deep intrauterine insemination andembryo transfer in pigs. Reproduction 58(Suppl.), 301–311.Martinez EA, Vazquez JM, Roca J, Lucas X, Gil MA, ParrillaI, Vazquez JL, Day BN, 2002: Minimal number of spermatozoarequired for normal fertility after deep intrauterineinsemination in non-sedated sows. Reproduction 123, 163–170.Martinez EA, Vazquez JM, Roca J, Cuello C, Gil MA, ParrillaI, Vazquez JL, 2005: An update on reproductive technologieswith potential short-term application in pig production.Reprod Domest Anim 40, 300–309.Martinez EA, Vazquez JM, Parrilla I, Cuello C, Gil MA,Rodriguez-Martinez H, Roca J, Vazquez JL, 2006: Incidenceof unilateral fertilizations after low dose deep intrauterineinsemination in spontaneously ovulating sows underfield conditions. Reprod Dom Anim 41, 41–47.Maxwell WMC, Salamon S, 1977: Fertility of boar semen afterlong term frozen storage. Theriogenology 8, 206 (abstract).Maxwell WMC, Watson PF, 1996: Recent progress in thepreservation of ram semen. Anim Reprod Sci 42, 55–65.Morton DB, Glover JD, 1974: Sperm transport in the femalerabbit: the effect of inseminate volume and sperm density.J Reprod Fertil 38, 139–142.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

354 JM Vazquez, J Roca, MA Gil, C Cuello, I Parrilla, I Caballero, JL Vazquez and EA Martı´nezNadir S, Saacke RG, Bame JH, Mullins J, Degelos S, 1993:Effect of freezing semen and dosage of sperm on number ofaccessory sperm, fertility and embryo quality in artificiallyinseminated cattle. J Anim Sci 71, 199–204.Rath D, 2002: Low Dose insemination in the Sow-A review.Rep Dom Anim 37, 201–205.Rath D, Ruiz S, Sieg B, 2003: Birth of female piglets followingintrauterine insemination of a sow using flow cytometricallysexed boar semen. Vet Rec 152, 400–401.Roca J, Carvajal G, Lucas X, Vazquez JM, Martinez EA,2003: Fertility of weaned sows after deep intrauterineinsemination with a reduced number of frozen-thawedspermatozoa. Theriogenology 60, 77–87.Roca J, Vazquez JM, Gil MA, Cuello C, Parrilla I, MartinezEA, 2006: Challenges in pig artificial insemination. ReprodDom Anim 41(Suppl. 2), 43–53.Rodriguez-Martinez H, Saravia F, Wallgren M, Tienthai P,Johannisson A, Vazquez JM, Martinez EA, Roca J, Sanz L,Calvete JJ, 2005: Boar spermatozoa in the oviduct. Theriogenology63, 514–535.Rodriguez-Vilar L, Lo´pez-Sánchez C, Boları´n A, Herna´ndezM, Vazquez JM, Martinez EA, Roca J, 2006: Effects ofinsertion of deep intrauterine insemination device into oneuterine horn on the reproductive performance of multiparoussows. Reprod Dom Anim 41 (Suppl. 2), 124 (abstract).String A, Berber E, Foroutani A, Macho JR, Pearl JM,Siperstein AE, 2001: Use of optical access trocar for safe andrapid entry in various laparoscopic procedures. Surg Endosc15, 570–573.Tummaruk P, Sumransap P, Techakumphu M, KunavongkritA, 2007: Distribution of spermatozoa and embryos in thefemale reproductive tract after unilateral deep intra uterineinsemination in the pig. Reprod Domest Anim 42, 603–609.Vazquez JM, Martı´nez EA, Parrilla I, Cuello C, Gil MA,Lucas X, Roca J, Vazquez JL, Didion BA, Day BN2001:Deep intrauterine insemination in natural post-weaningoestrus sows. In: Sixth International Conference on PigReproduction, Columbia, MO, USA, pp. 132.Vazquez JM, Martinez EA, Parrilla I, Roca J, Gil MA,Vazquez JL, 2003: Birth of piglets after deep intrauterineinsemination with flow cytometrically sorted spermatozoa.Theriogenology 59, 1605–1614.Vazquez JM, Martinez EA, Roca J, Gil MA, Parrilla I, CuelloC, Carvajal G, Lucas X, Vazquez JL, 2005: Improving theefficiency of sperm technologies in pigs: the value of deepintrauterine insemination. Theriogenology 63, 536–547.Vazquez JM, Martinez EA, Parrilla I, Cuello C, Gil MA,Garcia E, Caballero I, Alminana C, Roca J, Vazquez JL,2006: Laparoscopic intraoviductal insemination in pigs: Anew tool for improving the efficiency of sex sorted spermatozoa.Reprod Domest Anim 41, 298 (abstract).Watson PF2000: The causes of reduced fertility with cryopreservedsemen. Anim Reprod Sci 60–61, 481–492.Wongtawan T, Saravia F, Wallgren M, Caballero I, Rodriguez-MartinezH, 2006: Fertility after deep intra-uterineartificial insemination of concentrated low-volume boarsemen doses. Theriogenology 65, 773–787.Author’s address (for correspondence): JM Vazquez, Department ofAnimal Medicine and Surgery, Faculty of Veterinary Medicine,Campus de Espinardo, University of Murcia, E-30071, Murcia, Spain.E-mail: vazquez@um.esConflict of interest: All authors declare that the financial supportwithin the past two years has been obtained from: CICYT (AGF98-0533; AGL2001-0471; AGL2004-07546 and AGF2005-00760), CDTI(B288/98; 03.180; 03-402, 04-0231), INIA (RZ01-019), FundacionSéneca de Murcia (PB/74/FS/02, 03002/PI/05 and (04543/GERM/07).The authors have no further conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 355–358 (2008); doi: 10.1111/j.1439-0531.2008.01184.xISSN 0936-6768Production of Transgenic Farm Animals by Viral Vector-Mediated Gene TransferCBA Whitelaw, SG Lillico and T KingThe Roslin Institute and Royal (Dick) School of Veterinary Studies, University of Edinburgh, EdinburghUKContentsTransgenic technology holds considerable promise to advanceunderstanding in biomedical and agricultural systems withsome believing that one day transgenic animals may directlycontribute to farming and breeding practice. Nevertheless,applications in livestock have been restricted in part by theinefficiency of the technology. The recent development oflentivirus vectors for transgene delivery may overcome some ofthis limitation. This presentation describes these vectors, theiradvantages and limitations.History of Transgenesis in LivestockThe first transgenic livestock were generated more thantwo decades ago (Hammer et al. 1985). As then the useof the pronuclear microinjection technique, althoughinefficient, has been the most commonly used methodfor generating non-rodent transgenic animals. Pronuclearinjection involves the direct injection of thetransgene DNA into one of the two pronuclei in thezygote (Simons et al. 1988; Nottle et al. 2001). Althoughinefficient, with usually only 1% of injected eggsresulting in a transgenic founder animal, the reliabilityof pronuclear injection has ensured that it remains themost used method.The inefficiency of pronuclear injection has been themain driver in the search for more efficient methods.These investigations have resulted in a range of techniqueseach displaying different benefits and limitations(Clark and Whitelaw 2003). The use of sperm as adelivery route to generate transgenic livestock is attractivebecause of its simplicity but suffers from apparentvariability, while nuclear transfer using transgenic cellshas the advantage that all founder animals had to betransgenic but is currently restricted because of lowfoetal survival and high neonatal mortality.Five years after, pronuclear injection was first demonstratedin livestock, an exciting method was developedthat involved sperm-mediated gene transfer(Lavitrano et al. 1989). The concept of simply mixingthe transgene with sperm was very appealing because ofits simplicity. Unfortunately, this method has not seenwide uptake because of apparent variability in success(Brinster et al. 1989; Noseno 2003). Recently, modificationsincluding the use of intracytoplasmic sperminjection (Moreira et al. 2007) or the direct injection ofthe transgene into testes (Coward et al. 2007) may yetrekindle interest in this approach.Then came a huge technological advance; nucleartransfer (Wilmut et al. 1997; Campbell et al. 2006) –commonly called cloning – offers the precise modificationto the genome. In this method, the donor cell usedfor reconstituting the embryo is first made transgenic.Both random (Schnieke et al. 1997) and sequencetargeted (McCreath et al. 2000) transgene integrationis possible. This approach to transgenesis is therefore,somewhat analogous to the long established use ofembryonic stem cells in mice (Capecchi 1989). Incontrast, pronuclear microinjection only offers randomintegration of the transgene, resulting in the transgenesuffering from position-effects, which make expressionof the transgene unpredictable (al-Shawi et al. 1990;Wilson et al. 1990). Therefore, the precise nature of thegenetic modification possible using nuclear transfermakes this approach to altering the activity of a specificgene very attractive. Yet, the method remains inefficientand technically demanding (Campbell et al. 2007); andfor many hypothesis driven studies, this inefficiencyrenders this method too expensive both in cost andanimal numbers.Recently, livestock have been produced using anothermethod, one that has the benefit of being very efficient.This method utilizes a viral vector and builds on theprogress achieved in the development of gene therapystrategies. The use of viral vectors in transgenesis is notnew, with the first ever transgenic mice carrying aretroviral vector transgene (Jaenisch 1976). Yet, thesefirst generation vectors were often transcriptionallysilenced (Pannell and Ellis 2001). The new vectors arebased on a specific type of retrovirus called a lentivirusand are both robustly expressed and offer spectaculartransgenesis efficiencies (Pfeifer 2004; Whitelaw 2004).This presentation describes these vectors, their advantagesand limitations. It does not address what transgenicanimals might be used for: this is left to theimagination of the reader.Lentivirus Vectors – Pros and ConsLentiviruses are infectious retroviral pathogens thatcause a number of diseases in a range of speciesincluding man. Therefore, a lentivirus had to be disabledfor it to be used as a vector (Amado and Chen 1999;Naldinin and Verma 2000; Loewen and Poeschla 2005).Lentivirus vectors are generated by deletion of key genesfrom the viral genome involved in packaging andreplication of the virus from the viral genome. In thismethod, the replication-defective vector is only able togo through the first phase of the infectious cycle once,and not able to produce infectious virus. Vector particlesare produced by introducing the vector DNA into apackaging cell line in trans with the missing proteins bytransfection strategies (Pfeifer 2004).An additional key feature is that exogenous transcriptionelements can be inserted into the vector (Pfeifer2004). This allows control of the spatial and temporaltransgene expression profile. Apart from the introducedÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

356 CBA Whitelaw, SG Lillico and T Kingpromoter element, the vector is otherwise transcriptionallysilent. This is achieved by introducing mutationsinto the viral genome transcription control sequences(the long-terminal repeats) to generate a self-inactivatingvector (Pfeifer 2004). This distinguishes the lentivirusvector from other retrovirus vectors, which rely on viraltranscriptional elements for transgene activity (Pfeifer2004).With regard to transgenesis, lentivirus vectors haveseveral advantages (Pfeifer 2004; Whitelaw 2004).A distinguishing property of lentivirus vectors is theirability to infect both dividing and non-dividing cells. Itis this property that has promoted their development asgene delivery vectors since they can be delivered to eitherthe egg or zygote. Second, delivery of the vector,although still requiring the use of microinjection equipment,is less damaging to the egg ⁄ zygote when comparedwith pronuclear microinjection resulting in aconsiderable proportion of injected embryos developingto term. Exploiting the natural efficiency of infectionafforded to virus by evolution, the vector fuses with thecell (egg or zygote) and is internalized. This means thatthe lentivirus vector can be introduced (Fig. 1), byinjection into the perivitelline space of the zygote (Loiset al. 2002; Ritchie et al. 2007) or by co-culture with azona-free (or laser microdrilled) zygote (Ewerling et al.2006). This delivery method results in a very highproportion of injected embryos surviving. Recent datasuggests that many embryo manipulation methods causeDNA damage and thus limited embryo development(Yamauchi et al. 2007). Lentivirus vector-mediatedtransgene delivery appears not to suffer from thisproblem, perhaps because of the ‘natural’ route ofinfection that they use to deliver the transgene.A further and significant property of lentivirus vectorsis the spectacular efficiency of transgenesis that can beachieved. Transgenic mice (Lois et al. 2002; Pfeifer et al.LTRPromoter - transgeneFig. 1. Diagram of lentiviral vector delivery methods. A prototypicalvector is shown long-terminal repeats encompassing the promotertransgeneof choice. Packaged viral vectors can then be either coculturedwith zona denuded (or microdrilled) zygotes, which requiresindividual culturing of each embryo or injected into the perivitellinespace of an oocyte or zygote. Both methods have been demonstrated toproduce live transgenic animalsLTR2002), rats (Lois et al. 2002; Dann 2007; Michalkiewiczet al. 2007), poultry (McGrew et al. 2004; Lillico et al.2007), pigs (Hofmann et al. 2003; Whitelaw et al. 2004),cattle (Hofmann et al. 2004), sheep (Whitelaw, unpublished)and recently rabbits (Zsuzsa Bosze, personalcommunication) have been made using this method. Formany of these species, transgenesis rates of up to 100%of injected embryos carrying an integrated transgenehave been achieved. This had to be compared with thestandard pronuclear injection method, where 5% ofmouse zygotes and 1% of livestock zygotes result in atransgenic founder animal. Furthermore, unlike pronuclearinjection, most of these integration events carrytranscriptionally active transgenes. Taken together, theadvantageous properties of lentivirus vectors make thema very attractive transgene delivery route, especially forspecies, where the more standard methods are verycostly in terms of animal numbers and in the financialsupport required.Unfortunately like all good things, lentivirus vectorshave properties, which can constrain their application(Pfeifer 2004; Whitelaw et al. 2004). Perhaps, the mostsignificant concern is the reduced ‘cargo’ capacity. Thesevectors can only carry small transgenes (up to 8 kb),although in practice, transgenes bigger than 6 kb oftendisplay poor packaging efficiency and low vector titres.Second, the vector integrates as a single copy butmultiple integrations can occur. This leads to segregationissues in subsequent generations. Yet, since founderanimals carry different copies of the transgene, this canbe an advantage for founder population studies, offeringa dose–response to transgene activity.Integration is random, as in the pronuclear injectionmethod. This can result in position-effect phenomenonthat is often associated with pronuclear injection.Position-effects can result in aberrant or ectropic transgeneactivity, and poorly expressing or silent transgeneloci (Wilson et al. 1990). In our experience, the authorshave rarely seen these inappropriate transgene expressionprofiles, although others have (Hofmann et al.2006). The authors have seen expression in 4-year-oldpigs at the same level they displayed as a young pig; insheep, pigs and chickens the vast majority of transgenicanimals that they produced using lentiviral vectorsexpress the transgene; although the authors haveobserved transgenerational reduction in expression(after the 6th generation) in a few lines of transgenemice (the majority of lines do not show silencing), theauthors have yet to see any transgenerational silencingin chickens and pigs. Random integration also raises thepossibility of functional impairment of the host genomeat the site of integration. Lentiviral transgene integrationsites have yet to be mapped in animals but do lieclose to genes in studies on human cells and, although atvery low frequency, can cause gene activation (Themiset al. 2005; Montini et al. 2006). To overcome thislimitation, there is a considerable effort being directed todesign lentivirus vectors that can target transgeneintegration to specific sequences within the genome(Balciunas et al. 2006; Philippe et al. 2006; Clark et al.2007; Lombardo et al. 2007; Shinohara et al. 2007).Finally, it must be remembered that VSV-G pseudotyped(Pfeifer 2004) lentivirus vectors have the potentialÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Lentiviral Transgenesis 357to infect the handler. This function is lost uponintegration since these vectors are replication-defective.Nevertheless this method, at least the initial laboratorybasedsteps of virus preparation and handling, requiresgreater focus on biosafety than other method oftransgenesis.Livestock Transgenesis – The Best has Yet toComeLentivirus vectors represent a useful new method fortransgene delivery, especially to non-rodent species,offering unprecedented numbers of transgenic founderanimals. In essence, regardless of what species, eachlitter will carry transgenic founder animals. This meansthat for many studies, evaluation of transgene functioncan be carried out in the founder population. Overall,the efficiency and founder study strategy results in theuse of significantly fewer animals per study anddramatically reduces the timeline of a study. Therefore,there are substantial cost savings making it feasible todesign studies that were previously restricted to mousebasedexperiments. Perhaps, the possibility of combininglentivirus vector efficiency with the exciting potentialof RNA-interference offers an exciting future for transgeniclarge animals (Clark and Whitelaw 2003; :Tiscornia et al. 2003; Shin et al. 2006).Already this method of transgene delivery hasyielded successful applications, exemplified by thetargeted production of two therapeutic proteins intransgenic hens eggs (Lillico et al. 2007); and more areunderway. The authors are aware of studies lookingto deliver RNAi-based systems to combat viral infectionin chickens and sheep, and attempts to produceanimal models of human disease focussing on diseasesof the eye, lung and throat. In addition, studiesutilizing lentiviral vectors are now allowing mechanisticstudies of basic biological traits, for examplephotoperiodism in sheep, to be investigated in largeanimals, experiments that were simply not feasibleusing standard transgene delivery methods in thesespecies.Lentivirus vectors suffer from two drawbacks; limitedcargo capacity and random integration. It is not easy tosee how the carrying capacity of viral vectors can besignificantly improved; this aspect is likely to be overcomethrough the generation of robust methods forartificial chromosome transfer (Kuroiwa et al. 2002).Yet, to prevent insertional mutagenesis or positioneffects,the targeted integration of viral vectors is underintense investigation and is an achievable goal. As theauthors learnt more about how to harness sequencebinding activities of factors such zinc-finger bindingproteins, they might able to overcome this limitation.Alternatively, manipulating the binding specificity affordedby recombinase enzymatic activities is likely toresult in sequence-directed transgene integration strategies(Balciunas et al. 2006; Clark et al. 2007; Lombardoet al. 2007; Shinohara et al. 2007). These technologiesare just around the corner but perhaps the best is yet tocome. With the research activity in this area, in concertto the energy being applied to the development of genetherapy delivery methods, there is the real prospect thatefficient, sequence-specific transgene integration in theegg or zygote will be available soon.AcknowledgementsThe authors recognize the support and discussion with both presentand past members of this group. All animal work at The RoslinInstitute is performed under licence from the UK Home Office. Workin the authors’ group is supported by a number of sources,in particular, a SFHEFC Strategic Research Development Grant‘Development of Novel Technologies to Fight Viral Diseases in FarmAnimals’, the European Commission NEST029025 project INTEGRAand the BBSRC.ReferencesAmado RG, Chen IS, 1999: Lentiviral vectors – the promise ofgene therapy in reach. Science 285, 674–676.Balciunas D, Wangensteen KJ, Wilber A, Bell J, Geurts A,Sivasubbu S, Wang X, Hackett PB, Largaespada DA,McIvor RS, Ekker SC, 2006: Harnessing a high cargocapacitytransposon for genetic applications in vertebrates.PLoS Genet 2, e169.Brinster RL, Sandgren EP, Behringer RR, Palmiter RD, 1989:No simple solution for making transgenic mice. Cell 59,239–241.Campbell KH, McWhir J, Ritchie WA, Wilmut I, 2006: Sheepcloned by nuclear transfer from a cultured cell line. Nature380, 64–68.Campbell KH, Fisher P, Chen WC, Choi I, Kelly RD, Lee JH,Xhu J, 2007: Somatic cell nuclear transfer: past, present andfuture perspectives. Theriogenology 68(Suppl. 1), S214–S231.Capecchi MR, 1989: Altering the genome by homologousrecombination. Science 244, 1288–1292.Clark AJ, Whitelaw CBA, 2003: A future for transgeniclivestock. Nat Rev Genet 4, 825–833.Clark KJ, Carlson DF, Foster LK, Kong BW, Foster DN,Fahrenkrug SC, 2007: Enzymatic engineering of the porcinegenome with transposons and recombinases. BMC Biotechnol17, 42.Coward K, Kubota H, Parrington J, 2007: In vivo genetransfer into testis and sperm: developments and futureapplication. Arch Androl 53, 187–197.Dann CT, 2007: New technology for an old favorite: lentiviraltransgenesis and RNAi in rats. Transgenic Res 16, 571–580.Ewerling S, Hofmann A, Klose R, Weppert M, Brem G, RinkK, Pfeifer A, Wolf E, 2006: Evaluation of laser-assistedlentiviral transgenesis in bovine. Transgenic Res 15, 447–454.Hammer RE, Pursel VG, Rexroad CE, Wall RJ, Bolt DJ,Ebert KM, Palmiter RD, Brinster RL, 1985: Production oftransgenic rabbits, sheep and pigs by microinjection. Nature315, 680–683.Hofmann A, Kessler B, Ewerling S, Weppert W, Vogg B,Ludwig H, Stojkovic M, Boelhauve M, Brem G, Wolf E,Pfeifer A, 2003: Efficient transgenesis in farm animals bylentiviral vectors. EMBO Rep 4, 1054–1060.Hofmann A, Zakhartchenko V, Weppert M, Sebald H,Wenigerkind H, Brem G, Wolf E, Pfeifer A, 2004: Generationof transgenic cattle by lentiviral gene transfer intooocytes. Biol Reprod 71, 405–409.Hofmann A, Kessler B, Ewerling S, Kabermann A, Brem G,Wolf E, Pfeifer A, 2006: Epigenetic regulation of lentiviraltransgene vectors in a large animal model. Mol Ther 13, 59–66.Jaenisch R, 1976: Germ line integration and Mendeliantransmission of the exogenous Moloney leukemia virus.Proc Natl Acad Sci USA 73, 1260–1264.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

358 CBA Whitelaw, SG Lillico and T KingKuroiwa Y, Kasinathan P, Choi YJ, Naeem R, Tomizuka K,Sullivan EJ, Knott JG, Duteau A, Goldsby RA, OsborneBA, Ishida I, Robl JM, 2002: Cloned transchromosomiccalves producing human immunoglobulin. Nat Biotechnol20, 889–894.Lavitrano M, Camaioni A, Fazio VM, Dolci S, Farace MG,Spadafora C, 1989: Sperm cells as vectors for introducingforeign DNA into eggs: genetic transformation of mice. Cell57, 717–723.Lillico SG, Sherman A, McGrew MJ, Roberston CD, Smith J,Haslam C, Barnard P, Radcliffe PA, Mitrophanous KA,Elliot EA, Sang H, 2007: Oviduct-specific expression of twotherapeutic proteins in transgenic hens. Proc Natl Acad SciUSA 104, 1771–1776.Loewen N, Poeschla EM, 2005: Lentiviral vectors. AdvBiochem Eng Biotechnol 99, 169–191.Lois C, Hong EJ, Pease S, Brown EJ, Baltimore D, 2002:Germline transmission and tissue-specific expression oftransgenes delivered by lentiviral vectors. Science 295,868–872.Lombardo A, Genovese P, Peausejour CM, Colleoni S, leeYL, Kim KA, Ando D, Urnow FD, Galli C, Gregory PD,Holmes MC, Naldini L, 2007: Gene editing in human stemcells using zinc finger nucleases and integrase-defectivelentiviral vector delivery. Nat Biotechnol 25, 1298–1306.McCreath KJ, Howcroft J, Campbell KH, Colman A,Schneike AE, Kind AJ, 2000: Production of gene-targetedsheep by nuclear transfer from cultured somatic cells.Nature 405, 1066–1069.McGrew MJ, Sherman A, Ellard FM, Lillico SG, GilhooleyHJ, Kingsman AJ, Mitrophanous KA, Sang H, 2004:Efficient production of germline transgenic chickens usinglentiviral vectors. EMBO Rep 5, 728–733.Michalkiewicz M, Michalkiewicz T, Geurts AM, Roman RJ,Slocum GR, Singer O, Wihrauch D, Greene AS, KaldunskiM, Verma IM, Jacob HJ, Cowley AW Jr, 2007: Efficienttransgenic rat production by a lentiviral vector. Am JPhysiol Heart Circ Physiol 293, H881–H894.Montini E, Cesana D, Shcnidt M, Sanvito F, Ponzoni M,Bartholomae C, Sergi Sergi L, Benedicenti F, Ambrosi A, DiSerio C, Doglioni C, von Kalle C, Naldini L, 2006:Hematopoietic stem cell gene transfer in a tumor-pronemouse model uncovers low genotoxicity of lentiviral vectorintegration. Nat Biotechnol 24, 687–696.Moreira PN, Pozueta J, Perez-Crespo M, Valdivieso F,Gutierrez-Adan A, Montoliu L, 2007: Improving the generationof genomic-type transgenic mice by ICSI. TransgenicRes 16, 163–168.Naldinin L, Verma IM, 2000: Lentiviral vectors. Adv VirusRes 55, 599–609.Noseno N, 2003: Piglets add some colour to transgenic story.Nature 424, 604.Nottle MB, Haskard KA, Verma PJ, Du ZT, Grupen CG,McIlfatrick SM, Ashman RJ, Harrison SJ, Barlow H,Wigley PL, Lyons IG, Cowan PJ, Crawford RJ, TolstoshevPL, Pearse MJ, Robins AJ, d’Apice AJ, 2001: Effect ofDNA concentration on transgenesis rates in mice and pigs.Transgenic Res 10, 523–531.Pannell D, Ellis J, 2001: Silencing of gene expression: implicationsfor design of retrovirus vectors. Rev Med Virol 11, 205–217.Pfeifer A, 2004: Lentiviral transgenesis. Transgenic Res 13, 513–522.Pfeifer A, Ikawa M, Dyn Y, Verma IM, 2002: Transgenesis bylentiviral vectors: lack of gene silencing in mammalianembryonic stem cells and preimplantation embryos. ProcNatl Acad Sci USA 99, 2140–2145.Philippe S, Sarkis C, Barkats M, Mammeri H, Ladroue C,Petit C, Mallet J, Serguera C, 2006: Lentiviral vectors with adefective integrase allow efficient and sustained transgeneexpression in vitro and in vivo. Proc Natl Acad Sci USA103, 17684–17689.Ritchie WA, Neil C, King T, Whitelaw CBA, 2007: Transgenicembryos and mice produced from low titre lentiviral vectors.Transgenic Res 16, 661–664.Schnieke AE, Kind AJ, Ritchie WA, Mycock K, Scott AR,Ritchie M, Wilmut I, Colman A, Campbell KH, 1997:Human factor IX transgenic sheep produced by transfer ofnuclei from transfected fetal fibroblasts. Science 278, 2130–2133.al-Shawi R, Kinnaird J, Burke J, Bishop JO, 1990: Expressionof a foreign gene in a line of transgenic mice is modulated bya chromosomal position effect. Mol Cell Biol 10, 1192–1198.Shin KJ, Wall EA, Zavzavadjian JR, Santat LA, Liu J, HwangJI, Rebres R, Roacj T, Seaman W, Simon MI, Fraser ID,2006: A single lentiviral vector platform for microRNAbasedconditional RNA interference and coordinated transgeneexpression. Proc Natl Acad Sci USA 103, 13759–13764.Shinohara ET, Kaminski JM, Segal DJ, Pelczar P, Kolhe R,Ryan T, Costes CJ, Fraser MJ, Handler AM, YamagimachiR, Moisyadi S, 2007: Active integration: new strategies fortransgenesis. Transgenic Res 16, 333–339.Simons JP, Wilmut I, Clark AJ, Archibald AL, Bishop JO,Lathe R, 1988: Gene transfer into sheep. Biotechnology 6,179–183.Themis M, Waddington SN, Schmidt M, von Kalle C, WangY, Al-Allaf F, Gregory LG, Novsarkar M, Holder MV,Buckley SM, Dighe N, Ruthe AT, Mistry A, Bigger B,Rahim A, Nguyen TH, Trono D, Thrasher AJ, Coutelle C,2005: Oncogenesis following delivery of a nonprimatelentiviral gene therapy vector to fetal and neonatal mice.Mol Ther 12, 763–771.Tiscornia G, Singer O, Ikawa M, Verma IM, 2003: A generalmethod for gene knockdown in mice by using lentiviralvectors expressing small interfering RNA. Proc Natl AcadSci USA 100, 1844–1848.Whitelaw CBA, 2004: Transgenic livestock made easy. TrendsBiotechnol 22, 157–159.Whitelaw CBA, Radcliffe PA, Ritchie WA, Carlisle A, EllardFM, Pena RN, Rowe J, Clark AJ, King TJ, MitrophanousKA, 2004: Efficient generation of transgenic pigs usingequine infectious anaemia virus (EIAV) derived vector.FEBS Lett 571, 233–236.Wilmut I, Schnieke AE, McWhir J, Kind AJ, Campbell KH,1997: Viable offspring derived from fetal and adult mammaliancells. Nature, 385, 810–813.Wilson C, Bellen HJ, Gehring WJ, 1990: Position effects oneukaryotic gene expression. Annu Rev Cell Biol 6, 679–714.Yamauchi Y, Doe B, Ajduk A, Ward MA, 2007: GenomicDNA damage in mouse transgenesis. Biol Reprod 77, 803–812.Author’s address (for correspondence): C Bruce A Whitelaw, TheRoslin Institute and Royal (Dick) School of Veterinary Studies,University of Edinburgh, Roslin, Midlothian, Edinburgh EH25 9PS,UK. E-mail: bruce.whitelaw@roslin.ed.ac.ukConflict of interest: CBA Whitelaw has received both National andInternational research funding from various public bodies to pursuethis research; the remaining authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 359–367 (2008); doi: 10.1111/j.1439-0531.2008.01185.xISSN 0936-6768Use of Microarray Technology to Profile Gene Expression Patterns Important forReproduction in CattleACO Evans 1 , N Forde 1 , GM O’Gorman 1 , AE Zielak 2 , P Lonergan 1 and T Fair 11 School of Agriculture, Food Science and Veterinary Medicine, and the Conway Institute, University College Dublin, Dublin, Ireland; 2 Institute ofAnimal Breeding, Faculty of Biology and Animal Science, Wroclaw University of Environmental and Life Sciences, Wroclaw, PolandContentsFertility in cattle is a major component of many agriculturalenterprises and there is pressure to devise methods to improvethis. A number of approaches are ongoing, one of which is tobetter understand the cellular and molecular events of thedevelopment of reproductive tissues and to use these as targetsfor developing new strategies. Microarray technologies nowallow us the potential to determine the transcriptional profileof expressed genes in a given tissue. This review focuses on thetypes of microarrays available for studies in cattle andconcludes that genes associated with one or more of thecellular processes of cell survival ⁄ death, intracellular signalling,transcription and translation, cell division and proliferationand cellular metabolism are the main transcriptionalpathways that control the development of ovarian follicles,oocytes, early embryos and the uterine endometrium about thetime of the establishment of pregnancy.IntroductionFertility rates in cattle (pregnancies per insemination)are depending on the types of animals being bred andthe production system that they are in; however, ratesof more than 50% in multiparous animals (i.e. notheifers) are not often achieved (Diskin et al. 2006).Strategies that aim to improve fertility rates depend onimproving our understanding of organ, cellular andmolecular events that control reproduction and it is forthis reason that there is a great interest in determiningthe relationship between the expression of genes intissues and their consequences for fertility. A numberof recent publications have reviewed the potential formicroarrays and bovine reproduction (Niemann et al.2007; Smith and Rosa 2007; Wrenzycki et al. 2007;Bonnet et al. 2008; Mitko et al. 2008). In this review,we focus on the different microarray approaches andtechnologies that are in use and then go on to reviewthe genomics of ovarian follicle development, oocytedevelopment, embryo development and uterine functionabout the time of the establishment of pregnancyin cattle.Microarrays and Functional GenomicsThe term genomics was proposed in 1987 as thebringing together of molecular biology, cell biology andgenetics (McKusick and Ruddle 1987) and refers to thestudy of whole sets of genes and their interactions. Thestudy of the expression of particular genes at specifictimes in cells, that leads to a description of theirfunction, is termed functional genomics. Advances inthe scale and sophistication of DNA microarrays haveheralded a new era away from the traditional gene-bygeneapproach and towards the study of complexinterrelated processes that frequently occur in biologicalsystems. Although many different microarraysystems have been developed, the most commonly usedsystems are complementary DNA (cDNA) microarrays,oligonucleotide microarrays and high-density(Affymetrix-style) fabricated short oligonucleotidemicroarrays.Complementary DNA (cDNA) microarraysThe earliest and still most widely used form of geneexpression arrays are DNA microarrays where a presynthesizedprobe (cDNA or oligonucleotide) is spottedonto a support (Hager 2006). The first generation ofmicroarrays used cDNA as probes on glass slides(Schena et al. 1995). These early microarrays weregenerally available for human, rodent and other majormodel organisms, while those for cattle were developedlater and continue to be developed. Many of thesebovine cDNA microarrays are customized or specificfor particular research areas of interest, which has anadvantage of producing ‘bespoke’ in-house microarrays.Probes for cDNA microarrays are usuallyproducts of the polymerase chain reaction (PCR)generated from cDNA libraries or clone collections,using either vector-specific or gene-specific primers, andare printed onto glass slides or nylon membranes asspots at defined locations (Schulze and Downward2001). On most slides, genes are spotted in duplicate ortriplicate, where the cDNA in each spot is usually awhole gene or expressed sequence tag (EST) and isoptimally approximately 300–800 nucleotides long(Bryant et al. 2004). The availability of cDNA clonesets for arraying has depended on large-scale ESTsequencing projects that have resulted in millions ofEST sequences being deposited in GenBank; subsequentlythe complexity of these collections is reducedby the assembly of related sequences into clusters(Lyons 2003). Overall, cDNA microarrays have beenvital for the initial development and dissemination ofmicroarray technology. However, a range of technicalissues arise during their production; these include phagecontamination, incorrect annotation and errors in highthroughputproduction of probe sets generated fromtens of thousands of bacterial clones. These problemshave prompted a search for an alternative type of probeand spotted oligonucleotide microarrays have been onepopular alternative.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

360 ACO Evans, N Forde, GM O’Gorman, AE Zielak, P Lonergan and T FairOligonucleotide microarraysThe reliability of synthesis and falling production costsof pre-synthesized long (50–70 mer) oligonucleotideshave made them a viable alternative probe choice inmicroarray production (Lyons 2003). Improvements inslide surface chemistry and, in particular, the ability toattach oligonucleotide probes to slides at a sufficientconcentration has resulted in the successful developmentof the oligonucleotide microarray platform (Hollowayet al. 2002). In addition, genes that are expressed in lowabundance and are not represented in available cDNAlibraries can be included provided that sequence informationis available, as is the case for many organismsnow. The design of oligonucleotides is highly critical andalgorithms have been developed to increase theirperformance (Kreil et al. 2006). A study by Hugheset al. (2001) examined oligonucleotide sensitivity andspecificity, and found that 60-mer oligonucleotides werereliable when complex biological samples were used andsuggested that microarrays with one single carefullyselected oligonucleotide per gene correlated closely withcDNA microarrays (Hughes et al. 2001). Overall, synthesizedoligonucleotide probes have become the mostpopular format in many DNA microarray facilities.When good oligonucleotide probe design practices areemployed they can potentially discriminate betweenhighly similar targets, such as splice variants or genefamilies (Kreil et al. 2006).The processing steps for both types of spottedmicroarrays, cDNA and oligonucleotide, are the samein many respects. Two samples (mRNA to convertedcDNA), a ‘test’ and a ‘control’ are differentiallyfluorescently labelled with fluorochrome dyes. Dependingon the experimental design selected, one sample maybe a reference that is used in all hybridizations and actsas an internal control. The labelling of the samples canoccur during the reverse transcription (direct method)or afterwards (indirect method). The two samples arethen combined and allowed to hybridize to the immobilizedtargets on the microarray under sensitive andspecific conditions. Competitive hybridization yields aquantitative ratio of the two fluorescent dyes, determinedafter the slide is scanned and gives informationabout the comparative over- or under-expression ofgenes in that particular experimental sample (Bryantet al. 2004).Affymetrix GeneChips ÒThe high-density, synthesized in situ or fabricatedoligonucleotide platform employs the most sophisticatedmicroarray technology available to date. Inparticular, the Affymetrix GeneChip Ò has been widelyaccepted and is considered the gold standard in globalgene expression analysis (http://www.affymetrix.com).The GeneChip Ò technology platform consists of highdensitymicroarrays, which are used with standardizedassays, reagents, instrumentation, data managementand analysis tools (McGall and Christians 2002). TheGeneChip Ò bovine genome microarray was designedbased on content from bovine UniGene Build 57(March 24, 2004) and bovine mRNAs deposited inthe GenBank database. Specifically, the microarraycontains 24 027 (Bos taurus) probe sets that representmore than 23 000 (Bos taurus) transcripts, includingassemblies from approximately 19 000 UniGene Clusters.The bovine GeneChip Ò also has probe features formultiple hybridization controls, poly-A controls andhousekeeping (control or reference) genes. GeneChip Òmicroarrays consist of 25-mer oligonucleotides (probes)that are synthesized at specific locations on a coatedquartz surface. The 25-mer probe length confers highspecificity, while a number of probes are used to spaneach gene conferring high sensitivity and reproducibility.For each probe on the microarray that perfectlymatches its target sequence (perfect match or PM),there is also a paired ‘mismatch’ probe (MM). The MMprobe contains a single mismatch located directly in themiddle of the 25-base probe sequence. Informationfrom the PM and MM probe sets may be summarizedusing either the MAS5 Statistical algorithm, or alternativelyRobust Multichip Analysis (RMA) can beperformed which ignores MM values within theAffymetrix Ò Expression ConsoleÔ Software package.A recent review describes the GeneChip Ò technologyplatform including its individual components andapplications in more detail (Dalma-Weiszhausz et al.2006). The preparation of RNA samples for geneexpression profiling on Affymetrix GeneChips Ò requiresa number of sequential steps that results in thesynthesis of a biotin-labelled cRNA sample. AftercRNA fragmentation, the sample is ready to behybridized to the Affymetrix GeneChip Ò . Followinghybridization, probe arrays are washed, stained andscanned.BioinformaticsCollated raw microarray expression data must besubjected to a computational pipeline comprising initiallyof quality control validation and filtering. Anappropriate normalization methodology, for the experimentand data set in question, is then applied to removesystemic bias. These biases can be introduced byunequal quantities of starting RNA, differences inlabelling or detection efficiencies between the fluorescentdyes used, and systematic biases in the measuredexpression levels (Quackenbush 2002). Downstreamstatistical analysis can subsequently be performed togenerate meaningful lists of differentially expressedgenes. Free open source software packages are available,including, e.g. TM4 (http://www.tm4.org/); which includesa collection of software solutions (MicroarrayData Manager (MADAM), TIGR_Spotfinder, MicroarrayData Analysis System (MIDAS), and MultiexperimentViewer (MeV)), to capture, manage and analyzedata from DNA microarray experiments. Alternatively,the Bioconductor software project (http://www.bioconductor.org/;based on the R programming language) hasa range of powerful statistical and graphical methods forthe analysis of genomic data (see also De Bruyne et al.2007; Grant et al. 2007) for further information on theanalysis and management of microarray expressiondata).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Use of Microarray Technology for Reproduction in Cattle 361Global Gene Expression During OvarianFollicle DevelopmentFolliculogenesis starts during foetal life when millions ofprimordial germ cells surrounded by pregranulosa cellsdevelop and form a life-time supply of follicles. At as yetpoorly understood intervals, these follicles develop fromthe resting pool to grow and mature. During thisdevelopment, most follicles undergo atresia and die, witha small proportion developing to the later stages beyond5 mm in diameter and few developing to ovulatory sizeand fewer still going on to ovulate. These later stages offollicle development are the most susceptible to manipulationand have been the focus of many studies over theyears (reviewed in Mihm and Evans 2008).It is clear that antral follicle growth occurs insuccessive waves of development in cattle that are eachcharacterized by the growth of a cohort of follicles thatdevelop in parallel until the growth of one folliclerapidly increases and becomes dominant (this follicle isselected to become dominant) with the other folliclescease growing, begin to regress and undergo atresia(subordinate follicles). The hormonal changes andassociated regulation of growth, selection and atresiaof follicles has been well studied. After selection, theparameters that differentiate the dominant follicle fromsubordinate follicles are larger diameter, enhancedoestradiol production, more LH receptors in granulosacells, and an increase in IGF binding protein (IGFBP)protease activity in follicular fluid, low concentrations ofIGFBPs and increased concentrations of free IGF infollicular fluid (Fortune et al. 2004; Webb and Campbell2007; Mihm and Evans 2008). As with other areas ofbiomedical science, the importance of the expression ofgenes in controlling follicle development in cattle hasbeen studied one gene at a time (the candidate geneapproach) and has been highly revealing [reviewed by(Mihm and Bleach 2003)]. Alternatively, studying theexpression of many genes simultaneously has beenachieved using suppressive subtraction hybridization(Sisco et al. 2003), serial analysis of gene expression(SAGE) (Mihm et al. 2008) or microarrays (see next).Microarrays have been used to study the progressionof ovarian follicle development and have shown changesin gene expression associated with a decrease in FSHdependence and an increase in LH dependence asdominant follicles continue to develop after selection(Mihm et al. 2006). Continued development of thedominant follicle was associated with decreased mRNAexpression for genes known to be induced by FSH ingranulosa cells (FSHR, ESR2, INHA, ACVR1 andCCND2), increased mRNA expression for the LHreceptor (LHCGR) in granulosa cells and alterationsin mRNA expression of numerous genes potentiallyinvolved in survival and apoptosis of granulosa andtheca cells (SIVA, FADD, TIAF1, LASS4 andTNFSF8) (Mihm et al. 2006). In another study, microarrayswere used to characterize the global pattern ofgene expression in aberrant persistent dominant follicles(low fertility) and concluded that alteration of theexpression of genes involved in energy and proteinmetabolism (e.g. phosphofructokinase and fructose-1,6-bisphosphatase), amino acid transport (e.g. amino acidtransporter A2) and apoptosis (e.g. caspase-3, Bcl-2 andBcl-x) in granulosa cells may negatively impact on theirability to support a healthy oocyte and thereby contributeto decreased fertility (Lingenfelter et al. 2008).In a number of experiments, we have used bovinecDNA microarrays to ask what genes in granulosa andtheca cells contribute to selection of the dominant fromsubordinate follicles. We identified 261 genes ⁄ ESTs thatwere differentially expressed between dominant andsubordinate follicles and focused our attentions offamilies (ontologies) of genes. First, we examined genesinvolved in apoptosis and showed that dominant folliclesurvival was associated with genes involved in theinhibition of apoptosis (aromatase, LH receptor, oestradiolreceptor b, DICE-1 tumour suppressor protein andMCL-1) and the regression of subordinate follicles wasassociated with enhanced expression of genes involved inthe apoptotic pathways (Betaglycan, COX1, TNFalpha,caspase-activated DNase, DRAK-2, caspase 13,P58(IPK), Apaf-1, BTG-3 and TS-BCLL) (Evans et al.2004). We also identified 83 genes involved in signaltransduction (largely hormone receptors and intracellularsignalling molecules) that were differentially expressedbetween dominant and subordinate folliclesand after a number of additional experiments hypothesizedimportant roles for CAMkinase1 and EphA4 intheca cells and BCAR1 in granulosa cells for thedevelopment of dominant follicles and for betaglycanand FIBP in granulosa cells of regressing subordinatefollicles (Forde et al. 2008a). The fate of cells is regulatedin part by the transcription of new genes and we alsoexamined the relative expression of genes coding fortranscription factors in follicles during the development.We identified 34 genes that code for transcription factorsin granulosa and theca cells with good evidence tosuggest that genes for CEBP-b, SRF, FKHRL1,NCOR1 and Midnolin may be important in determiningthe fate of follicle cells (Zielak et al. 2007a). In summary,these studies used a microarray approach to identify alarge number of genes which indicates that ovarianfollicle development involves the concerted actions ofgroups of molecules with diverse functions. However, themicroarray approach also lead to the discovery of genesthat have not previously been thought to be involved infollicle development, for example, the prion protein(Forde et al. 2007) and completely novel (some as yetunidentified) genes (Zielak et al. 2007b).Global Gene Expression During OocyteMaturationPrior to embryonic genome activation, embryo developmentin cattle is driven by maternal mRNAs andproteins produced during the oocyte growth phase.Intrinsic oocyte developmental competence, as assessedby development to the blastocyst stage has beenpositively associated with a number of factors, includingfollicular environment (Lonergan et al. 2004) and thesite of oocyte maturation (Rizos et al. 2002). Numerousstudies have been carried out to investigate the affect ofthese factors on the oocyte transcriptome using a widerange of gene expression analysis techniques such asÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

362 ACO Evans, N Forde, GM O’Gorman, AE Zielak, P Lonergan and T Faircandidate gene PCR, global gene expression microarrayanalysis and cDNA subtraction.The first global analysis of the bovine oocyte transcriptomepre- and post-meiotic maturation wasreported in 2003 (Dalbies-Tran and Mermillod 2003).Using the Atlas Human 1.2 cDNA expression array(Clontech Laboratories, Mountain View, CA, USA),the authors identified approximately 300 genes withknown identities that were expressed in the bovineoocyte, of which 70 were differentially expressed duringmeiotic maturation, and these genes were associatedwith cell cycle regulation, DNA transcription andapoptosis regulation. More recently, using the samecross-species hybridization approach, Adjaye et al.(2007) compared mRNA transcript profiles of bovineoocytes and blastocysts using a human cDNA arrayconsisting of over 15 000 probe sets. A total of 164 geneswere exclusively expressed in the oocyte. Among thegenes that were either exclusive to, or enriched inoocytes, the authors identified 31 genes that wereoocyte-conserved across mouse and human oocyte datasets. The genes included germ cell-specific transcriptssuch as DAZL and NASP and a panel of genesimplicated in cell proliferation such as FER, KIT andMAPRE1 (Adjaye et al. 2007). Since then with access tothe Affymetrix Bovine GeneChip, we recently carriedout a global mRNA analysis of bovine immature andin vitro matured oocytes (Carter et al. 2007; Fair et al.2007). We have identified at least 8000 genes that areexpressed in bovine oocytes, over 800 of which aredifferentially expressed (‡2-fold) during meiotic maturation(Fair et al. 2007). The resulting differentiallyexpressed and present ⁄ absent gene lists were classifiedaccording to their gene ontology (GO); cellular location,biological process and molecular function. In general,the same GO categories were represented in the twooocytestages. However, 25% of the genes were upregulatedand 75% were downregulated in the in vitromatured oocytes when compared with their immaturecounterparts. Similarly, Misirlioglu et al. (2006) alsoused the Affymetrix Bovine GeneChip to compare geneexpression of bovine matured oocytes (MII) and eightcell-stage embryos. Genes controlling DNA methylationand metabolism were upregulated in MII oocytes(Misirlioglu et al. 2006). When the two MII oocyte genelists were analyzed, they not only appeared to confirmeach other but also the findings of Dalbies-Tran andMermillod (2003), as a number of common genes thatwere primarily involved in metabolism, transport andcell death regulation were identified.To date many models of bovine oocyte developmentalcompetence have been described (Lonergan et al. 2003;Fair et al. 2004; Sirard et al. 2006). Recently, a broadranging study employed a combination of SSH, cDNAmicroarray analysis and quantitative PCR to interrogatemRNA transcript profiles in oocytes from several suchmodels, including; oocytes cultured with or withoutrFSH, early- vs late-cleaving embryos, and oocytes fromdifferent follicle sizes. The gene candidates CCNB2,PTTG1, H2A, CKS1, PSMB2, SKIIP, CDC5L, RGS16and PRDX1, representing cell cycle, cell metabolism,cell signalling and gene expression functional genecategories were identified as significantly associated withdevelopmental competence (Mourot et al. 2006). Focusingjust on follicle status, Ghanem et al. (2007) comparedtranscript abundance of bovine oocytes retrievedfrom small growing follicles vs dominant folliclesrecovered during the first follicular wave using theBlueChip cDNA microarray (Sirard et al. 2006; Ghanemet al. 2007). Fifty-one differentially regulated geneswere identified, of which genes regulating proteinbiosynthesis (RPLP0, RPL8, RPL24, ARL6IP, RpS14,RpS15, RpS4x and RPS3A) were enriched in the morecompetent oocytes, i.e. oocytes recovered from growingfollicles, other over expressed genes fell into a number offunctional classifications, including the following: translationelongation, ATP binding, NADH dehydrogenaseactivity, cytoskeleton, calcium ion binding. In addition,using the BlueChip cDNA microarray, a recent reportdescribes a novel non-invasive approach to identifyingdevelopmentally competent and incompetent bovineoocytes and zygotes (Dessie et al. 2007). In this study,bovine oocytes and zygotes were dielectrophoreticallyseparated according to their speed of migration to theopposite electrode. The authors reported that themajority of differentially regulated genes were moreabundant in the very fast moving oocytes and zygotes,and fell mainly under functional categories associatedwith regulation of protein biosynthesis and metabolicregulation. Among these transcripts, RPL2, RPL8,RPL35 and RPLP0 were more abundant in the veryfast moving developmentally superior oocytes, confirmingthe findings of Mourot et al. (2006) described earlier.Finally, comparing RNA isolated from germinal vesiclestage oocytes collected from adult vs pre-pubertalanimals, as model of high and low oocyte competence,on a bovine cDNA array representing over 15 000unique genes (Suchyta et al. 2003), Patel et al. (2007)described significant over-representation of transcriptsencoding for genes in hormone secretion classificationwithin adult oocytes. Further analysis revealed follistatinas a maternal mRNA marker of high developmentalcompetence in two models (Patel et al. 2007).Global Gene Expression During EmbryoDevelopmentMammalian pre-implantation embryos are very sensitiveto the environment in which they develop. In vitro,the period of post-fertilization embryo culture is themost critical period affecting blastocyst quality assessedin terms of cryotolerance, gene expression pattern andability to establish a pregnancy (Corcoran et al. 2006;Lonergan et al. 2006). Precise control of gene expressionduring this phase of development is particularly importantas several critical developmental events occurduring this period including: (i) the first cleavagedivision, the timing of which has been associated withthe developmental competence in a variety of mammalianspecies, (ii) embryonic genome activation, when theembryo transfers from a reliance on maternal RNAderived from the oocyte to expression of its owngenome, (iii) morula compaction, which involves theestablishment of the first intimate cell-to-cell contacts inthe embryo and (iv) blastocyst formation, involving thedifferentiation of two cell types, the trophectoderm andÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Use of Microarray Technology for Reproduction in Cattle 363the inner cell mass and subsequent elongation. Analysisof expression patterns of developmentally importantgenes provides a useful tool to assess the normality ofin vitro produced embryos alongside their in vitroderivedcounterparts and may provide a blueprintagainst which to manipulate culture conditions in orderto improve embryo quality.Gene expression studies in cattle embryos have beenconducted employing qualitative, semi-quantitative andquantitative approaches. Most studies have focused on asmall set of ‘candidate’ genes putatively important forembryogenesis including growth factors, metabolicenzymes, transcription factors involved in the regulationof early developmental events and imprinted geneproducts (see Wrenzycki et al. 2005 for review). A muchmore limited, but increasing, number of studies haveused cDNA microarrays to examine global gene expressionchanges in cattle embryos (Pfister-Genskow et al.2005; Smith et al. 2005; Corcoran et al. 2006; Somerset al. 2006; Adjaye et al. 2007).Ushizawa et al. (2004) used a custom-designed uteroplacentalcDNA microarray to study gene expression inbovine embryos on days 7, 14 and 21, extra-embryonicmembranes on day 28 and foetuses on day 28 ofpregnancy. Comparison of day 7 and day 14 revealedmost genes increased during this period, and a smallnumber of genes exhibiting altered expression decreasedas gestation progressed (Ushizawa et al. 2004). Theexpression of trophoblast cell-specific molecules such asplacental lactogens, prolactin-related proteins, interferon-tauand adhesion molecules was enhanced duringthe pre-implantation period suggesting a pivotal role inthe preparation needed for implantation.We recently compared mRNA expression across awide range of biological processes in bovine blastocystsderived from either culture in vivo or in vitro usingbovine cDNA microarrays (BOTL). The vast majority(85%) of the differentially expressed transcripts betweenthe two embryo populations were downregulated inin vitro cultured blastocysts, suggesting that the primaryreason why in vitro embryos are of inferior developmentalcompetence compared to in vivo cultured may bebecause of a deficiency of the machinery associated withtranscription and translation (Corcoran et al. 2006).Based on previous data showing the feasibility andhigh reproducibility of cross species hybridizationapproach using a human cDNA array to study geneexpression in bovine tissues (Adjaye et al. 2004), Adjayeet al. (2007) used a human 15 529 chip (the ENSEMBLchip) to compare gene expression in bovine oocytes andblastocysts; 164 and 1324 genes were expressed exclusivelyat the oocyte or blastocyst stage, respectively,while 419 genes expressed in both stages. Few studieshave attempted to correlate the differences in mRNAabundance observed at the blastocyst stage, such asthose outlined earlier, with the ability of the embryo toestablish a pregnancy. One such recent study (El-Sayedet al. 2006) addressed the relationship between transcriptionalprofile of embryos and the pregnancy successbased on gene expression analysis of biopsies fromblastocysts taken prior to transfer to recipients. Microarraydata (BlueChip) analysis revealed a total of 52differentially regulated genes between embryos resultingin a calf delivery vs those not resulting in a pregnancy(El-Sayed et al. 2006). Biopsies from embryos that weretransferred to recipients and resulted in calf deliverywere enriched with genes necessary for implantation(COX2 and CDX2), carbohydrate metabolism(ALOX15), signal transduction (PLAU) and placentaldevelopment (PLAC8), while those failing to establish apregnancy were enriched with transcripts for an inflammatorycytokine (TNF), protein amino acid binding(EEF1A1), transcription factors (MSX1 and PTTG1),glucose metabolism (PGK1 and AKR1B1) and CD9which is an inhibitor of implantation (El-Sayed et al.2006).Bovine embryos produced by nuclear transfer (NT)are capable of developing to the blastocyst stage with anefficiency of 30–50% which is comparable to thatachieved with routine IVF (Yang et al. 2007). However,following somatic cell cloning and transfer to surrogaterecipients a very limited percentage (0.5–5%) of theembryos completes full-term development. This ismainly due to a high frequency of post-implantationdevelopmental arrest. Such losses are predominantlyduring the first trimester of pregnancy but can occurmuch later (Heyman et al. 2002) and are often associatedwith aberrant placental development (Hill et al.2000). The high incidence of pregnancy loss andneonatal death following somatic cell NT (SCNT) arehypothesized to result from incomplete nuclear reprogramming;several authors have convincingly demonstratedthat the donor somatic cell is reprogrammedsuch that expression pattern at the blastocyst stage issubstantially different from that of the somatic cell priorto NT (Smith et al. 2005; Beyhan et al. 2007). Therehave been several recent cDNA microarray studiesexamining differential gene expression in NT-derivedembryos (Pfister-Genskow et al. 2005; Smith et al. 2005;Somers et al. 2006) in an effort to explain the poorgestational success.In one of the first reports comparing individualembryos produced by NT and IVF using cDNAmicroarray technology for any species, Pfister-Genskowet al. (2005) found 18 genes differentially expressedbetween NT- and IVF-produced embryos includingthree intermediate-filament protein genes (cytokeratin8, cytokeratin 19 and vimentin), three metabolic genes(phosphoribosyl pyrophosphate synthetase 1, mitochondrialacetoacetyl-coenzyme A thiolase and alpha-glucosidase),two lysosomal-related genes (prosaposin andlysosomal-associated membrane protein 2), and a geneassociated with stress responses (heat shock protein 27)along with major histocompatibility complex class I,nidogen 2, a putative transport protein, heterogeneousnuclear ribonuclear protein K, mitochondrial 16S rRNA,and ES1 (a zebrafish orthologue of unknown function).Smith et al. (2005) used microarray technology toanalyze global differences in gene expression betweenblastocysts derived from (i) natural fertilization in vivo,(ii) in vitro fertilization or (iii) NT. Perhaps surprisingly,the NT embryos most closely resembled true in vivoembryos in terms of numbers of differentially expressedgenes. They reported a low variability in the pattern ofmRNA expression among individual in vivo-derivedblastocysts and among NT blastocysts when comparedÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

364 ACO Evans, N Forde, GM O’Gorman, AE Zielak, P Lonergan and T Fairwith in vitro produced blastocysts suggesting a relativelyuniform reprogramming mechanism in the NT embryosby the blastocyst stage. Despite this apparent similarityin gene expression pattern between NT- and AI-derivedblastocysts, the low pregnancy rate and developmentalanomalies associated with NT suggest that reprogrammingof the donor genome is incomplete. To investigatethe possible causes of these losses, transcript profiling oftrophoblast and embryonic disc tissue collected fromday 25 conceptuses derived by NT or AI was conductedusing a microarray consisting of 13 257 oligonucleotides-derivedfrom cattle gene sequences. Approximately9000 genes were differentially expressed between trophoblastand disc demonstrating major gene expressiondifferences in embryonic and extra-embryonic tissues.Comparing NT with AI, 188 genes were differentiallyexpressed in trophoblast and 10 in the embryonic disc,providing evidence that nuclear reprogramming isdefective in a large proportion of NT-derived clonesand that aberrant gene expression in the trophoblastcontributes to pregnancy failure at day 25 and beyond(Everts et al. 2005).Using Affymetrix GeneChip bovine arrays, Beyhanet al. (2007) compared global gene expression of twodifferent somatic cell lines whose capacity to generatehealth NT-derived calves is significantly different, thecloned blastocysts derived from the cell lines and IVFblastocysts in order to elucidate some of the key genesresponsible for successful reprogramming. The majorityof genes in SCNT blastocysts had comparable expressionpatterns to IVF blastocysts, with the exception of asmall number of genes whose relative expression valueswere similar to their corresponding donor cells, indicatinga failure of reprogramming in SCNT blastocysts(Beyhan et al. 2007).Global Gene Expression in the UterineEndometriumThe study of endometrial function is necessary tounderstand the factors associated with the establishmentof pregnancy and to get an insight into the causes ofearly embryo loss. The endometrium is responsible forthe secretion of the numerous cytokines, growth factorsand proteins that together make are collectively termedthe histotroph. The histotroph is secreted from theglandular epithelium in the endometrium into the lumenof the uterus and it is in this environment that theembryo develops. Studies involving endometrial geneexpression in cattle have mainly focused on the changesthat occur at different stages of the oestrous cycle. Somestudies have examined the mRNA expression profiles ofbovine epithelial cells of the ipsilateral vs the contralateraloviduct (Bauersachs et al. 2003) and cells from theipsilateral oviduct (Bauersachs et al. 2004) or endometrium(Bauersachs et al. 2005) at oestrus (low progesterone)and dioestrus (high progesterone), to identifychanges in gene expression at different stages of theoestrous cycle in cattle.Microarray studies that give an insight into thechanges that occur in gene expression in the endometrium,which have a functional consequence for embryosurvival, have been performed in humans (Horcajadaset al. 2006), mice (Kashiwagi et al. 2007), sheep (Chenet al. 2007) and pigs (Ross et al. 2007). Although there isscant literature on the use of microarrays to studyendometrial gene expression in cattle, two papers(Bauersachs et al. 2006; Klein et al. 2006) have takentwo different animal model approaches to address thedifferences in endometrial gene expression betweenpregnant and cycling animals on day 18. In one study,monozygotic twin recipients were used along within vitro-derived embryos and suppression subtractivehybridization to enrich the cDNA for transcripts thatwere upregulated in the endometrium of pregnantanimals before microarray hybridization (Klein et al.2006). They identified 87 different genes ⁄ mRNAs, ofwhich 80 were known bovine genes or were inferredfrom human orthologues. Ontological classification ofthese genes showed that the largest class was type Iinterferon-stimulated genes. Of these differentially expressedgenes, several have already been described asbeing involved in changes in endometrial gene expressionin other species. For example, interferon-stimulatedgene-15 has shown to be expressed in the endometriumof cows (Austin et al. 2004), sheep (Johnson et al. 2000),pigs (Joyce et al. 2003), mice (Austin et al. 2003) andhumans and baboons (Bebington et al. 1999). However,the power of the microarray technology allowed for theidentification of genes that are involved in cell adhesion(connective tissue growth factor – CTGF, glycosylphosphatidylinositolspecific phospholipase D1 – GPLD1,milk fat globule-EGF factor 8 protein – MFGE8) aswell as genes involved in endometrial remodelling(matrix metalloproteinase 19 – MMP19, tissue inhibitorof metalloproteinase 2 – TIMP2) that have not previouslybeen described as being differentially regulated inthe endometrium (Klein et al. 2006). The second modelutilized SSH technology also, but used in vivo-derivedembryos (Bauersachs et al. 2006). This study identified109 transcripts that were differentially regulated in day18 pregnant animals when compared with day 18 cyclinganimals. Of these 109 transcripts, 34, 28 and 3 geneswere enriched in the GO terms for immune responsegenes, response to stimulus and antigen presentation,respectively.The earlier studies looked at differences in geneexpression in the endometrium associated with pregnancyon a day when maternal recognition had alreadyoccurred. However, in cattle, the majority of embryonicloss occurs prior to maternal recognition of pregnancy(Dunne et al. 2000; Humblot 2001). Retrospectivestudies have shown a correlation between concentrationsof progesterone and pregnancy rates both in beefand dairy cattle (Diskin et al. 2006). Other studies incattle have shown that progesterone supplementationincreases embryo size on day 14 of pregnancy (Garrettet al. 1988). Recent studies in our laboratory (Carteret al. 2008), Forde et al. (2008b) have tried to addressthe issue of early embryo mortality by examining thedifferential effects of elevated progesterone in theimmediate post-conception period and pregnancy onendometrial gene expression in cycling and pregnantanimals at different physiological stages of the earlydevelopmental axis (i.e. days 5, 7, 13 and 16 postconception,corresponding to the 16-cell ⁄ early morulaÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Use of Microarray Technology for Reproduction in Cattle 365stage, the blastocyst stage, the beginning of elongationand the time of maternal recognition of pregnancy,respectively). In our model, no genes were differentiallyexpressed in pregnant compared with cycling animalsbefore day 16. Of the genes that were differentiallyexpressed between pregnant and cycling animals (on day16), GO analyses reveals that genes involved in immuneresponse were significantly over-represented in pregnantvs cycling animals. Comparisons between our lists ofdifferentially expressed genes on day 16 of pregnancyand those from (Bauersachs et al. 2006) (day 18) showthat 32 genes are differentially regulated on both days inboth the models. This suggests that these genes areinvolved in very early maternal recognition of pregnancyand may be critical for the establishment of pregnancy.In the same study, we identified large numbers of genesthat were differentially expressed in the endometria ofanimals with high vs low concentrations of progesteroneon days 5, 7 and 13. GO analysis of differentiallyexpressed genes on day 5 indicate that genes involved inthe positive regulation of transcription are progesteroneregulated and on day 7 there is a shift towards genesinvolved in cell division. This indicates a potentiallycritical role of progesterone in regulating endometrialfunction early in pregnancy (days 5–13) when theembryo is developing prior to its signalling for thematernal recognition of pregnancy. Further work needsto be done to ascertain a functional role for these newlydescribed genes involved in differences in endometrialgene expression associated with pregnancy.ConclusionsKnowledge of the global pattern of gene expression isimportant for understanding critical regulatory pathwaysthat are necessary for successful tissue developmentand in recent years there has been a big increase inthe number of publications reporting the use of microarraysto study reproduction in cattle. These publicationsconclude that genes involved with cellsurvival ⁄ death, intracellular signalling, transcriptionand translation, cell division and proliferation andcellular metabolism are the main cellular processes thatcontrol the development of reproductive tissues. In thiscontext, microarrays have proved themselves to be apowerful tool. However, a note of caution reminds usthat our understanding of these cellular events will likelyimprove further in the future with the development ofbetter tools to analyse these data (bioinformatics). Inaddition, information provided using microarrays islimited by the number and nature of the spots on eacharray and new sequencing technologies now have thepromise to overcome this problem and have thepotential to replace microarrays as the method of choicefor global gene expression studies (Fields 2007).AcknowledgementsThis research, as well as similar ongoing studies, is funded by ScienceFoundation Ireland (grant numbers: 02 ⁄ IN1 ⁄ B78, PICA award,06 ⁄ INI ⁄ B62, 07 ⁄ SRC ⁄ B1156). The opinions, findings and conclusionsor recommendations expressed in this material are those of the authorsand do not necessarily reflect the views of the Science FoundationIreland.ReferencesAdjaye J, Herwig R, Herrmann D, Wruck W, Benkahla A,Brink TC, Nowak M, Carnwath JW, Hultschig C, NiemannH, Lehrach H, 2004: Cross-species hybridisation of humanand bovine orthologous genes on high density cDNAmicroarrays. BMC Genomics 5, 83.Adjaye J, Herwig R, Brink TC, Herrmann D, Greber B,Sudheer S, Groth D, Carnwath JW, Lehrach H, NiemannH, 2007: Conserved molecular portraits of bovine andhuman blastocysts as a consequence of the transition frommaternal to embryonic control of gene expression. PhysiolGenomics 31, 315–327.Austin KJ, Bany BM, Belden EL, Rempel LA, Cross JC,Hansen TR, 2003: Interferon-stimulated gene-15 (Isg15)expression is up-regulated in the mouse uterus in response tothe implanting conceptus. Endocrinology 144, 3107–3113.Austin KJ, Carr AL, Pru JK, Hearne CE, George EL, BeldenEL, Hansen TR, 2004: Localization of ISG15 and conjugatedproteins in bovine endometrium using immunohistochemistryand electron microscopy. Endocrinology 145, 967–975.Bauersachs S, Blum H, Mallok S, Wenigerkind H, Rief S, Prelle K,Wolf E, 2003: Regulation of ipsilateral and contralateral bovineoviduct epithelial cell function in the postovulation period:a transcriptomics approach. Biol Reprod 68, 1170–1177.Bauersachs S, Rehfeld S, Ulbrich SE, Mallok S, Prelle K,Wenigerkind H, Einspanier R, Blum H, Wolf E, 2004:Monitoring gene expression changes in bovine oviductepithelial cells during the oestrous cycle. J Mol Endocrinol32, 449–466.Bauersachs S, Ulbrich SE, Gross K, Schmidt SE, Meyer HH,Einspanier R, Wenigerkind H, Vermehren M, Blum H,Sinowatz F, Wolf E, 2005: Gene expression profiling ofbovine endometrium during the oestrous cycle: detection ofmolecular pathways involved in functional changes. J MolEndocrinol 34, 889–908.Bauersachs S, Ulbrich SE, Gross K, Schmidt SE, Meyer HH,Wenigerkind H, Vermehren M, Sinowatz F, Blum H, WolfE, 2006: Embryo-induced transcriptome changes in bovineendometrium reveal species-specific and common molecularmarkers of uterine receptivity. Reproduction 132, 319–331.Bebington C, Bell SC, Doherty FJ, Fazleabas AT, FlemingSD, 1999: Localization of ubiquitin and ubiquitin crossreactiveprotein in human and baboon endometrium anddecidua during the menstrual cycle and early pregnancy.Biol Reprod 60, 920–928.Beyhan Z, Ross PJ, Iager AE, Kocabas AM, Cunniff K, RosaGJ, Cibelli JB, 2007: Transcriptional reprogramming ofsomatic cell nuclei during preimplantation development ofcloned bovine embryos. Dev Biol 305, 637–649.Bonnet A, Dalbies-Tran R, Sirard MA, 2008: Opportunitiesand challenges in applying genomics to the study ofoogenesis and folliculogenesis in farm animals. Reproduction135, 119–128.Bryant PA, Venter D, Robins-Browne R, Curtis N, 2004:Chips with everything: DNA microarrays in infectiousdiseases. Lancet Infect Dis 4, 100–111.Carter F, Fair T, Park S, Wade M, Evans ACO, Lonergan P,2007: Gene expression profiling of immature and in vitromatured bovine oocytes using Affymetrix Genechip technology.Reprod Fertil Dev 19, Abstract 263, 2480.Carter F, Forde N, Duffy P, Wade M, Fair T, Crowe MA,Evans ACO, Kenny DA, Roche JF, Lonergan P, 2008:Effect of increasing progesterone concentration from Day 3of pregnancy on subsequent embryo survival and developmentin beef heifers. Reprod Fertil Dev 20, 368–375.Chen Y, Antoniou E, Liu Z, Hearne LB, Roberts RM, 2007:A microarray analysis for genes regulated by interferon-tauin ovine luminal epithelial cells. Reproduction 134, 123–135.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

366 ACO Evans, N Forde, GM O’Gorman, AE Zielak, P Lonergan and T FairCorcoran D, Fair T, Park S, Rizos D, Patel OV, Smith GW,Coussens PM, Ireland JJ, Boland MP, Evans ACO, LonerganP, 2006: Suppressed expression of genes involved intranscription and translation in in vitro compared with invivo cultured bovine embryos. Reproduction 131, 651–660.Dalbies-Tran R, Mermillod P, 2003: Use of heterologouscomplementary DNA array screening to analyze bovineoocyte transcriptome and its evolution during in vitromaturation. Biol Reprod 68, 252–261.Dalma-Weiszhausz DD, Warrington J, Tanimoto EY, MiyadaCG, 2006: The affymetrix GeneChip platform: an overview.Methods Enzymol 410, 3–28.De Bruyne V, Al-Mulla F, Pot B, 2007: Methods formicroarray data analysis. Methods Mol Biol 382, 373–391.Dessie SW, Rings F, Holker M, Gilles M, Jennen D, Tholen E,Havelicek V, Besenfelder U, Sukhorukov VL, ZimmermanU, Endter JM, Sirad MA, Schellander K, Tesfaye D, 2007:Dielectrophoretic behavior of in vitro-derived bovine metaphaseII oocytes and zygotes and its relation to in vitroembryonic developmental competence and mRNA expressionpattern. Reproduction 133, 931–946.Diskin MG, Murphy JJ, Sreenan JM, 2006: Embryo survivalin dairy cows managed under pastoral conditions. AnimReprod Sci 96, 297–311.Dunne LD, Diskin MG, Sreenan JM, 2000: Embryo and foetalloss in beef heifers between day 14 of gestation and full term.Anim Reprod Sci 58, 39–44.El-Sayed A, Hoelker M, Rings F, Salilew D, Jennen D, TholenE, Sirard MA, Schellander K, Tesfaye D, 2006: Large-scaletranscriptional analysis of bovine embryo biopsies in relationto pregnancy success after transfer to recipients. PhysiolGenomics 28, 84–96.Evans ACO, Ireland JLH, Winn ME, Lonergan P, Smith GW,Coussens PM, Ireland JJ, 2004: Identification of genesinvolved in apoptosis and dominant follicle developmentduring follicular waves in cattle. Biol Reprod 70, 1475–1484.Everts RE, Band MR, Liu ZL, Kumar CG, Liu L, Loor JJ,Oliveira R, Lewin HA, 2005: A 7872 cDNA microarray andits use in bovine functional genomics. Vet Immunol Immunopathol105, 235–245.Fair T, Murphy M, Rizos D, Moss C, Martin F, Boland MP,Lonergan P, 2004: Analysis of differential maternal mRNAexpression in developmentally competent and incompetentbovine two-cell embryos. Mol Reprod Dev 67, 136–144.Fair T, Carter F, Park S, Evans ACO, Lonergan P, 2007:Global gene expression analysis during bovine oocyte invitro maturation. Theriogenology 68(Suppl. 1), S91–S97.Fields S, 2007: Molecular biology. Site-seeing by sequencing.Science 316, 1441–1442.Forde N, Rogers M, Canty MJ, Lonergan P, Smith GW,Coussens PM, Ireland JJ, Evans ACO, 2007: Association ofthe prion protein and its expression with ovarian follicledevelopment in cattle. Mol Reprod Dev 75, 248–249.Forde N, Mihm M, Canty M, Zielak A, Baker P, Park S,Lonergan P, Smith G, Coussens P, Ireland J, Evans A, 2008a:Differential expression of signal transduction factors inovarian follicle development; a role for betaglycan andFIBP in granulosa cells in cattle. Physiol Genomics 33,193–204.Forde N, Carter F, Fair T, Crowe MA, Evans ACO, SpencerTE, Bazer FW, O’Gaora P, McBride R, Boland MP,Lonergan P, Roche JF, 2008b: Effect of pregnancy andprogesterone on gene expression in the uterine endometriumof cattle. Annual Meeting of the Society for the Study ofReproduction, Hawaii, USA (27 to 30 May 2008). Biologyof Reproduction, Special Issue, Abstract #37.Fortune JE, Rivera GM, Yang MY, 2004: Follicular development:the role of the follicular microenvironment in selectionof the dominant follicle. Anim Reprod Sci 82–83, 109–126.Garrett JE, Geisert RD, Zavy MT, Morgan GL, 1988:Evidence for maternal regulation of early conceptus growthand development in beef cattle. J Reprod Fertil 84, 437–446.Ghanem N, Holker M, Rings F, Jennen D, Tholen E, SirardMA, Torner H, Kanitz W, Schellander K, Tesfaye D, 2007:Alterations in transcript abundance of bovine oocytesrecovered at growth and dominance phases of the firstfollicular wave. BMC Dev Biol 7, 90.Grant GR, Manduchi E, Stoeckert CJ Jr, 2007: Analysis andmanagement of microarray gene expression data. CurrProtoc Mol Biol (Suppl. 77) Chapter 19.6, 1–30.Hager J, 2006: Making and using spotted DNA microarrays inan academic core laboratory. Methods Enzymol 410, 135–168.Heyman Y, Chavatte-Palmer P, LeBourhis D, Camous S,Vignon X, Renard JP, 2002: Frequency and occurrence oflate-gestation losses from cattle cloned embryos. BiolReprod 66, 6–13.Hill JR, Burghardt RC, Jones K, Long CR, Looney CR, ShinT, Spencer TE, Thompson JA, Winger QA, Westhusin ME,2000: Evidence for placental abnormality as the major causeof mortality in first-trimester somatic cell cloned bovinefetuses. Biol Reprod 63, 1787–1794.Holloway AJ, van Laar RK, Tothill RW, Bowtell DD, 2002:Options available – from start to finish – for obtaining datafrom DNA microarrays II. Nat Genet 32(Suppl.), 481–489.Horcajadas JA, Sharkey AM, Catalano RD, Sherwin JR,Dominguez F, Burgos LA, Castro A, Peraza MR, PellicerA, Simon C, 2006: Effect of an intrauterine device on thegene expression profile of the endometrium. J Clin EndocrinolMetab 91, 3199–3207.Hughes TR, Mao M, Jones AR, Burchard J, Marton MJ,Shannon KW, Lefkowitz SM, Ziman M, Schelter JM, MeyerMB, Kobayashi S, Davis C, Dai H, He YD, Stephaniants SB,Cavet G, Walker WL, West A, Coffey E, Shoemaker DD,Stoughton R, Blanchard AP, Friend SH, Linsley PS, 2001:Expression profiling using microarrays fabricated by an inkjetoligonucleotide synthesizer. Nat Biotechnol 19, 342–347.Humblot P, 2001: Use of pregnancy specific proteins andprogesterone assays to monitor pregnancy and determinethe timing, frequencies and sources of embryonic mortalityin ruminants. Theriogenology 56, 1417–1433.Johnson GA, Spencer TE, Burghardt RC, Joyce MM, BazerFW, 2000: Interferon-tau and progesterone regulate ubiquitincross-reactive protein expression in the ovine uterus.Biol Reprod 62, 622–627.Joyce M, Burghardt R, Bazer F, Zaunbrecher G, Johnson G,2003: Interferon-stimulated genes (ISGs) are induced in theendometrium of pregnant but not pseudopregnant pigs. BiolReprod 68 (Suppl. 1), Abstract 230, 206.Kashiwagi A, DiGirolamo CM, Kanda Y, Niikura Y, EsmonCT, Hansen TR, Shioda T, Pru JK, 2007: The postimplantationembryo differentially regulates endometrial gene expressionand decidualization. Endocrinology 148, 4173–4184.Klein C, Bauersachs S, Ulbrich SE, Einspanier R, Meyer HH,Schmidt SE, Reichenbach HD, Vermehren M, Sinowatz F,Blum H, Wolf E, 2006: Monozygotic twin model reveals novelembryo-induced transcriptome changes of bovine endometriumin the preattachment period. Biol Reprod 74, 253–264.Kreil DP, Russell RR, Russell S, 2006: Microarray oligonucleotideprobes. Methods Enzymol 410, 73–98.Lingenfelter BM, Dailey RA, Inskeep EK, Vernon MW, PooleDH, Rhinehart JD, Yao J, 2008: Microarray analysis ofgene expression in granulosal cells from persistent follicles incattle. Anim Reprod Sci 104, 405–413.Lonergan P, Rizos D, Gutierrez-Adan A, Fair T, Boland MP,2003: Oocyte and embryo quality: effect of origin, cultureconditions and gene expression patterns. Reprod DomestAnim 38, 259–267.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Use of Microarray Technology for Reproduction in Cattle 367Lonergan P, Pedersen HG, Rizos D, Greve T, Thomsen PD,Fair T, Evans A, Boland MP, 2004: Effect of the postfertilizationculture environment on the incidence of chromosomeaberrations in bovine blastocysts. Biol Reprod 71,1096–1100.Lonergan P, Fair T, Corcoran D, Evans ACO, 2006: Effect ofculture environment on gene expression and developmentalcharacteristics in IVF-derived embryos. Theriogenology 65,137–152.Lyons P, 2003: Advances in spotted microarray resources forexpression profiling. Brief Funct Genomic Proteomic 2, 21–30.McGall GH, Christians FC, 2002: High-density geneChipoligonucleotide probe arrays. Adv Biochem Eng Biotechnol77, 21–42.McKusick VA, Ruddle FH, 1987: A new discipline, a newname, a new journal. Genomics 1, 1–2.Mihm M, Bleach EC, 2003: Endocrine regulation of ovarianantral follicle development in cattle. Anim Reprod Sci 78,217–237.Mihm M, Evans ACO, 2008: Mechanisms for dominant follicleselection in monovulatory species: a comparison of morphological,endocrine and intraovarian events in cows, maresand women. Reprod Domest Anim 43(Suppl. 2), 48–56.Mihm M, Baker PJ, Ireland JL, Smith GW, Coussens PM,Evans ACO, Ireland JJ, 2006: Molecular evidence that growthof dominant follicles involves a reduction in follicle-stimulatinghormone dependence and an increase in luteinizinghormone dependence in cattle. Biol Reprod 74, 1051–1059.Mihm M, Baker PJ, Fleming LM, Monteiro AM, O’ShaughnessyPJ, 2008: Differentiation of the bovine dominantfollicle from the cohort upregulates mRNA expression fornew tissue development genes. Reproduction 135, 253–265.Misirlioglu M, Page GP, Sagirkaya H, Kaya A, Parrish JJ,First NL, Memili E, 2006: Dynamics of global transcriptomein bovine matured oocytes and preimplantationembryos. Proc Natl Acad Sci U S A 103, 18905–18910.Mitko K, Ulbrich SE, Wenigerkind H, Sinowatz F, Blum H,Wolf E, Bauersachs S, 2008: Dynamic changes in messengerRNA profiles of bovine endometrium during the oestrouscycle. Reproduction 135, 225–240.Mourot M, Dufort I, Gravel C, Algriany O, Dieleman S, SirardMA, 2006: The influence of follicle size, FSH-enrichedmaturation medium, and early cleavage on bovine oocytematernal mRNA levels. Mol Reprod Dev 73, 1367–1379.Niemann H, Carnwath JW, Kues W, 2007: Application ofDNA array technology to mammalian embryos. Theriogenology68(Suppl. 1), S165–S177.Patel OV, Bettegowda A, Ireland JJ, Coussens PM, LonerganP, Smith GW, 2007: Functional genomics studies of oocytecompetence: evidence that reduced transcript abundance forfollistatin is associated with poor developmental competenceof bovine oocytes. Reproduction 133, 95–106.Pfister-Genskow M, Myers C, Childs LA, Lacson JC, PattersonT, Betthauser JM, Goueleke PJ, Koppang RW, LangeG, Fisher P, Watt SR, Forsberg EJ, Zheng Y, Leno GH,Schultz RM, Liu B, Chetia C, Yang X, Hoeschele I,Eilertsen KJ, 2005: Identification of differentially expressedgenes in individual bovine preimplantation embryos producedby nuclear transfer: improper reprogramming ofgenes required for development. Biol Reprod 72, 546–555.Quackenbush J, 2002: Microarray data normalization andtransformation. Nat Genet 32(Suppl.), 496–501.Rizos D, Ward F, Duffy P, Boland MP, Lonergan P, 2002:Consequences of bovine oocyte maturation, fertilization orearly embryo development in vitro versus in vivo: implicationsfor blastocyst yield and blastocyst quality. MolReprod Dev 61, 234–248.Ross JW, Ashworth MD, White FJ, Johnson GA, Ayoubi PJ,DeSilva U, Whitworth KM, Prather RS, Geisert RD, 2007:Premature estrogen exposure alters endometrial gene expressionto disrupt pregnancy in the pig. Endocrinology 148,4761–4773.Schena M, Shalon D, Davis RW, Brown PO, 1995: Quantitativemonitoring of gene expression patterns with acomplementary DNA microarray. Science 270, 467–470.Schulze A, Downward J, 2001: Navigating gene expressionusing microarrays – a technology review. Nat Cell Biol 3,E190–E195.Sirard MA, Richard F, Blondin P, Robert C, 2006: Contributionof the oocyte to embryo quality. Theriogenology 65,126–136.Sisco B, Hagemann LJ, Shelling AN, Pfeffer PL, 2003:Isolation of genes differentially expressed in dominant andsubordinate bovine follicles. Endocrinology 144, 3904–3913.Smith GW, Rosa GJ, 2007: Interpretation of microarray data:trudging out of the abyss towards elucidation of biologicalsignificance. J Anim Sci 85, E20–E23.Smith SL, Everts RE, Tian XC, Du F, Sung LY, Rodriguez-Zas SL, Jeong BS, Renard JP, Lewin HA, Yang X, 2005:Global gene expression profiles reveal significant nuclearreprogramming by the blastocyst stage after cloning. ProcNatl Acad Sci U S A 102, 17582–17587.Somers J, Smith C, Donnison M, Wells DN, Henderson H,McLeay L, Pfeffer PL, 2006: Gene expression profiling ofindividual bovine nuclear transfer blastocysts. Reproduction131, 1073–1084.Suchyta SP, Sipkovsky S, Kruska R, Jeffers A, McNulty A,Coussens MJ, Tempelman RJ, Halgren RG, Saama PM,Bauman DE, Boisclair YR, Burton JL, Collier RJ, DePetersEJ, Ferris TA, Lucy MC, McGuire MA, Medrano JF,Overton TR, Smith TP, Smith GW, Sonstegard TS, SpainJN, Spiers DE, Yao J, Coussens PM, 2003: Developmentand testing of a high-density cDNA microarray resource forcattle. Physiol. Genomics 15, 158–164.Ushizawa K, Herath CB, Kaneyama K, Shiojima S, HirasawaA, Takahashi T, Imai K, Ochiai K, Tokunaga T, TsunodaY, Tsujimoto G, Hashizume K, 2004: cDNA microarrayanalysis of bovine embryo gene expression profiles duringthe pre-implantation period. Reprod Biol Endocrinol 2, 77.Webb R, Campbell BK, 2007: Development of the dominantfollicle: mechanisms of selection and maintenance of oocytequality. Soc Reprod Fertil Suppl 64, 141–163.Wrenzycki C, Herrmann D, Lucas-Hahn A, Korsawe K,Lemme E, Niemann H, 2005: Messenger RNA expressionpatterns in bovine embryos derived from in vitro proceduresand their implications for development. Reprod Fertil Dev17, 23–35.Wrenzycki C, Herrmann D, Niemann H, 2007: MessengerRNA in oocytes and embryos in relation to embryoviability. Theriogenology 68(Suppl. 1), S77–S83.Yang X, Smith SL, Tian XC, Lewin HA, Renard JP,Wakayama T, 2007: Nuclear reprogramming of clonedembryos and its implications for therapeutic cloning. NatGenet 39, 295–302.Zielak AE, Canty MJ, Forde N, Coussens PM, Smith GW,Lonergan P, Ireland JJ, Evans ACO, 2007a: Differentialexpression of genes for transcription factors in theca andgranulosa cells following selection of a dominant follicle incattle. Mol Reprod Dev 19, 967–975.Zielak AE, Forde N, Park SDE, Doohan F, Coussens PM,Smith GW, Ireland JJ, Lonergan P, Evans ACO, 2007b:Identification of novel genes associated with dominantfollicle development in cattle. Reprod Fertil Dev 19, 967–975.Author’s address (for correspondence): ACO Evans, UCD VeterinarySciences Centre, Belfield, Dublin 4, Ireland. E-mail: alex.evans@ucd.ieConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 368–373 (2008); doi: 10.1111/j.1439-0531.2008.01186.xISSN 0936-6768Breeding Soundness Evaluation and Semen Analysis for Predicting Bull FertilityJP Kastelic 1,2 and JC Thundathil 21 Agriculture and Agri-Food Canada, Lethbridge Research Centre, Lethbridge; 2 Faculty of Veterinary Medicine, University of Calgary, Calgary, AB,CanadaContentsBull fertility is influenced by numerous factors. Although 20–40% of bulls in an unselected population may have reducedfertility, few are completely sterile. Breeding soundness refersto a bull’s ability to get cows pregnant. A standard breedingsoundness evaluation identifies bulls with substantial deficits infertility, but does not consistently identify sub-fertile bulls.In this regard, the use of frozen-thawed semen (from bulls incommercial AI centres) that meets minimum quality standardscan result in pregnancy rates that differ by 20–25 percentagepoints. Although no single diagnostic test can accuratelypredict variations in fertility among bulls that are producingapparently normal semen, recent studies suggested that acombination of laboratory tests were predictive of fertility.This review is focused on recent developments in prediction ofbull fertility, based on assessments at the molecular, cellularand whole-animal levels.IntroductionBreeding soundness refers to a bull’s ability to get cowspregnant. Although 20–40% of bulls may have reducedfertility, few are completely sterile (Kastelic et al. 2000).Sub-fertile bulls delay conception, prolong the calvingseason, reduce calf weaning weights and increase femaleculls. Multiple sire breeding groups and low breedingpressure may mask sub-fertile bulls, but single-siremating groups and AI increase the importance of bullfertility.Traditional Breeding Soundness EvaluationIn general, bull breeding soundness potential is eitherdetermined by breeding many normal, fertile females(and assessing pregnancy rates) or conducting a breedingsoundness evaluation. No single measurement orcriterion reliably predicts fertility and therefore, severalcriteria are usually evaluated. The standards of theSociety for Theriogenology (http://www.therio.org) areintended to assess the likelihood of a bull establishingpregnancy in ‡25 healthy, cycling females in a 65–70 daybreeding season. The classification is based on a physicalevaluation and acceptable thresholds for testiculardevelopment, sperm motility and normal sperm morphology(libido and tests for venereal diseases are notroutinely evaluated). A breeding soundness evaluationincludes assessment of size, muscling, body conditionand freedom from disease and physical defects. Scrotalcircumference (SC) is highly correlated with pairedtestes weight, which is correlated with daily spermproduction and semen quality (reviewed by Barth 2007).Bulls with large SC have half-sib heifers and daughterswith earlier puberty and greater fertility (reviewed byBarth 2007). As the heritability of SC in young bulls(1–2 years) is 0.5, it responds well to selection(reviewed by Barth 2007). Coe and Gibson (1993)evaluated 264 beef bulls (13 breeds), and at 200 daysof age, calves with SC >23 cm had a 95% probability ofSC >34 cm by 365 days of age, whereas calves with SC34 cm by365 days.Scrotal circumference is measured by forcing the testesto the bottom of the scrotum and using a flexible tape toapply moderate tension at the largest circumference. Thescrotum should have a distinct neck, testes should befreely moveable, similar in size, firm and resilient, withnormal epididymides and spermatic cords. The sheathand penis should be of appropriate size and free ofdefects. The most common abnormality of the accessorysex glands is vesicular adenitis (usually unilateral), and itoccurs in 2–4% of bulls 1–2 years of age, but is uncommon(30% morphologicallyabnormal sperm or >20% head defects (http://www.therio.org; reviewed by Barth 2007).A bull that is healthy and sound, with adequate SC,‡70% morphologically normal sperm, and ‡30% progressivemotility, is designated Satisfactory PotentialBreeder (http://www.therio.org; Barth 2007). A bullwith temporary conditions (likely to resolve) is designatedClassification Deferred. This typically includesrecent puberty, an injury or lameness that is likely toresolve, or temporary testicular degeneration (e.g.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Breeding Soundness and Semen Analysis in Bulls 369because of hot weather). A bull with undesirableheritable defects, small SC, debilitating injury or disease,or permanent testicular degeneration, is designatedUnsatisfactory Potential Breeder.Bulls in AI CentresBulls in commercial AI centres often have substantialvariations in fertility, and frozen semen that meetsminimum quality standards can yield pregnancy ratesthat differ by 20–25% (Larson and Miller 2000).Standard semen end points identify grossly abnormalsemen, but do not consistently identify sub-fertile bullswith apparently normal semen (Gadea et al. 2004). Asbull fertility is influenced by a wide range of factors, nosingle diagnostic test can accurately predict fertility(Petrunkina et al. 2007; Rodriguez-Martinez 2007).Sperm Plasma Membrane ViabilityAs a functional sperm plasma membrane is essential forfertilization, its viability (integrity) should be assessed.Eosin permeates non-viable sperm (pink to red spermheads), whereas viable sperm appear white. Combinationsof various fluorescent probes, specific for viable ornon-viable sperm, are also used, and the most commonis Syber-14 and propidium iodide (Gillan et al. 2005).Syber-14 penetrates ‘viable’ sperm (with an intactplasma membrane), making the DNA bright green(Gillan et al. 2005), whereas propidium iodide penetratesdead and non-viable sperm and renders them red.The proportion of live and dead sperm (green and red,respectively) is determined with fluorescent microscopyor flow cytometry.Brito et al. (2003) studied several methods for evaluatingsperm membrane viability (eosin ⁄ nigrosin, Trypanblue, fluorescent probes and the response of sperm to theexposure of a hypoosmotic solution) and their relationshipto cleavage (IVF). Although these stains evaluate thephysical integrity of sperm membrane, the hypoosmoticswelling test (HOST; Jeyendran et al. 1984) evaluates itsfunctional competence. This was the only plasmalemmaevaluation method that contributed to conventionalsperm quality tests in predicting the success of in vitrofertilization (based on cleavage rate; Brito et al. 2003).An Annexin-V binding assay using fluorescence-activatedcell sorting (FACS) assesses externalization ofphosphatidyl serine of cryopreserved bovine semen (Anzaret al. 2002; Januskauskas et al. 2003). This test detectstranslocation of membrane phosphatidyl serine, whichprecedes loss of membrane integrity during cryopreservation-induced,apoptosis-like cell degradation. However,its correlation with fertility of frozen-thawed bullsemen varied among studies (Rodriguez-Martinez 2003).Sperm MotilitySperm motility (total or progressive) was traditionallyevaluated subjectively with light microscopy. Thismethod detects gross abnormalities, but not subtlevariations in motility that can affect fertility. However,computer-assisted sperm analysis (CASA), a much moreobjective method, measures specific motion characteristicsassociated with functional status. For example,capacitated sperm are hyperactivated (high amplitudeand low frequency; Yanagimachi 1994). Furthermore,combinations of sperm motion characteristics assessedby CASA (beat cross frequency, linearity, average pathvelocity, straightness and curvilinear velocity) weresignificantly correlated with bull fertility (Farrell et al.1998), as were average path velocity, total motility,linearly motile sperm and total number of motile spermin swim-up preparations (Hallap et al. 2006). However,comparisons of CASA results are only reliable whenidentical sperm evaluation settings are used.Sperm–Oviduct InteractionBillions of sperm are ejaculated, but

370 JP Kastelic and JC Thundathilsperm–oocyte interactions. Following capacitation,sperm interact with the zona pellucida (ZP). For thisinteraction, galactosyl transferase (Larson and Miller2000), p47 (Ensslin et al. 1998), sp56 (Cheng et al. 1994)and zonadhesin (Hardy and Garbers 1995) are apparentlysperm receptors, whereas zona glycoprotein (ZP3)is the ligand (Yanagimachi 1994). This interaction leadsto an acrosome reaction (Yanagimachi 1994); the inneracrosomal membrane interacts with a second glycoprotein,ZP2, facilitating the secondary binding of sperm tothe zona matrix during penetration (Yanagimachi 1994;Aitken 2006).There are several reports regarding the associationbetween sperm–zona binding and fertility. There weredifferences among bulls (with known fertility) in therelative number of sperm bound to bovine oocytes(Fazeli et al. 1993). However, large numbers of oocytesare needed and differences among bulls are detectableonly if they exceed differences among oocytes (Zhanget al. 1995). Alternatively, a hemizona assay (Fazeliet al. 1997) uses both halves of a ZP to compare bindingability of control sperm (known fertility) and test sperm(unknown fertility). Zona binding and hemizona bindingassays and non-return rates were significantlycorrelated. In vitro zona penetration assays have beenused to predict in vivo fertility (Puglisi et al. 2004).However, sperm penetration varies according to sperm–oocyte ratio, duration of incubation and heparin concentration,limiting the value of this test.Sperm Oolemma Fusion and Sperm DNADecondensationHenault et al. (1995) used a homologous zona-freeoocyte penetration assay to demonstrate the effects ofaccessory gland fluid from low- vs high-fertility bulls.Moreover, competitive penetration of zona-free bovineoocytes by fluorochrome-labelled bull sperm was relatedto in vivo fertility (Henault and Killian 1995) andevaluated the ability of sperm to undergo chromatindecondensation and pronucleus formation. In theabsence of a ZP, there is an equal opportunity for allsperm to fuse with the oocyte membrane, undergo DNAdecondensation and pronuclei formation (Thundathilet al. 2001a).Sperm chromatin integrity is an important determinantof the ability of sperm to form normal pronuclei.Sperm DNA is associated with histone nucleoproteinsand organized into classical nucleosome core particlesduring early spermatogenesis. However, these histonenucleoproteins are replaced by transition proteins,which are subsequently replaced by protamines (Yanagimachi1994). Ultimately, chromatin of mature spermhas a compact toroidal structure that resists denaturation.Lewis and Aitken (2005) suggested oxidativestress was a major cause of sperm DNA damage, withreduced pre-implantation embryo development andpregnancy rates. Increased testicular temperature reducedthe stability of sperm DNA (Karabinus et al.1997) and impaired the ability of these sperm to undergoDNA decondensation and pronuclei formation (Walterset al. 2006). However, links among elevated testiculartemperature, oxidative stress and sperm DNA damagein bulls remain unclear. The sperm cell structure assay(SCSA) uses flow cytometry to determine chromatinintegrity, based on resistance to acid denaturation.Sperm are exposed to low pH and stained with acridineorange, which emits green or red fluorescence when itbinds to double- (intact) or single-stranded DNA(denatured), respectively. The ratio of red to (red + -green) fluorescence measures chromatin denaturation,which is significantly correlated with fertility (Ballacheyet al. 1987, 1988; Januskauskas et al. 2001; Waterhouseet al. 2006). In brief, flow cytometry-based approachesprovide a quantitative measure of the structural integrityof sperm chromatin, based on a large number ofsperm, whereas IVF-based tests evaluate the ability ofsperm to undergo DNA decondensation and pronuclearformation during fertilization.Association Between In Vitro Fertilization andFertilityMarquant-Le Guienne et al. (1990) reported that pronuclearformation can be used to predict field fertility ofbulls. Zhang et al. (1997) reported that fertilizationbased on cleavage rate was highly correlated with nonreturnrates, but blastocyst production varied amongtest dates. However, fertility predictions were moreaccurate when based on several laboratory assays vs asingle assay (Truelson et al. 1996). In this regard, aseven-variable model (post-thaw total motility, postthawsperm with a linear motile pattern, sperm concentration,concentration of motile sperm after swim-up,sperm ZP-binding, cleavage rate of total oocytes andblastocyst rate of total oocytes) accounted for 84.6% ofthe variation in non-return rates (Zhang et al. 1999).However, this approach may not be sensitive enough todiscriminate among highly fertile bulls. Similarly, amodel with 30 post-thaw sperm characteristics (includingcleavage rate) accurately predicted fertility (based onconception and non-return rates) of both high- and lowfertilitybulls (Phillips et al. 2004).Future DirectionsThe cell biology approaches described earlier may serveas supplementary tests to a standard breeding soundnessevaluation and improve the reliability of fertility predictions.However, a better understanding of the regulationof sperm function and its contributions to earlyembryo development at the molecular level may lead toreliable molecular markers of fertility, with implicationsfor predicting fertility variations in bulls used for AI.As DNA is transcriptionally inactive in ejaculatedsperm, physiological functions are regulated by structuraland functional proteins and identifying theseproteins may have implications for predicting fertility.The role of heat-shock proteins in sperm capacitation(Kamaruddin et al. 2004) and their potential role insperm–oocyte binding have been reported (Matweeet al. 2001). Specific sperm proteins involved in theinteraction between sperm and ZP (Larson and Miller2000; Inoue et al. 2005) and sperm–oocyte adhesion andegg activation (Evans and Florman 2002) have beenidentified, and the role of angiotensin converting enzymeÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Breeding Soundness and Semen Analysis in Bulls 371(ACE) in fertilization (Kondoh et al. 2005) has beendescribed. In addition, PAWP, a sperm-specific WWdomain-binding protein, promotes meiotic resumptionand pronuclear development during fertilization (Wuet al. 2007).We recently used scrotal insulation to induce abnormalspermatogenesis (unpublished) and identified spermproteins associated with abnormal spermatogenesis as aresult of elevated testicular temperature. There wasdifferential expression (between normal and abnormalsperm of the same bull) of the alpha 4 subunit ofNa + ⁄ K + ATPase, tissue inhibitor of metalloproteinase-2 (TIMP-2), ACE and hexokinase-1. Further studies areneeded to elucidate the specific role of these spermproteins in fertilization and early development.Although the sperm proteins described earlier areassociated with the regulation of specific sperm functions,proteins with changes in expression patterns inresponse to variations in fertility remain unknown, buthave great potential as fertility markers. In this regard,heparin-binding proteins (24–31 kDa) were proposed asa genetic marker for fertility differences in bullsproducing normal semen (Bellin et al. 1998; McCauleyet al. 1999). Furthermore, there are associations betweenspecific seminal plasma proteins and fertility(Killian et al. 1993; Moura et al. 2006). Sperm proteinsfrom low- vs high-fertility Nelore bulls with acceptablesemen had differential expression (two-dimensional gelelectrophoresis) of sperm membrane proteins (Roncolettaet al. 2006). Similarly, accessory gland fluids fromhigh-fertility Holstein bulls had more bovine seminalplasma protein (BSP) 30 kDa and phospholipase A2,whereas osteopontin appeared to improve the ability ofepididymal sperm (from low-fertility bulls) to penetrateoocytes in vitro (Moura et al. 2007). Therefore, identifyingfertility-associated sperm or seminal plasma proteinsby comparing low- vs high-fertility bulls, andinvestigating the role of these proteins in the regulationof sperm function, fertilization or embryo development,may identify markers that predict fertility.Morphologically abnormal sperm failed during gameteinteraction or pre-implantation development (Thundathilet al. 2000, 2001b; Walters et al. 2005). Inaddition, embryos resulting from the fertilization ofoocytes by morphologically normal sperm, coexisting inthe ejaculate along with abnormal sperm, had reduceddevelopmental competence, suggesting that these spermwere functionally impaired. Therefore, increasing theinsemination dose to compensate for infertility as aresult of compensable factors (Saacke et al. 2000)requires further investigation. Data from commercialembryo production units suggested that bulls differ intheir ability to produce pre-implantation embryos in vivo(Thundathil and Mapletoft, unpublished data). In thisregard, damage to sperm DNA because of oxidativestress, chromosome anomalies and environmental effects,including elevated testicular temperature, time ofAI relative to oestrus, duration of semen storage,duration of sperm–oocyte interaction, age of malesand infectious agents in semen influence quality ofembryos (reviewed by Chenoweth 2007). However, morestudies are needed to elucidate the effects of paternalgenes (Browing and Strome 1996), sperm RNA(Krawetz 2005; Miller et al. 2005; Miller and Ostermeier2006; Boerke et al. 2007), specific sperm and seminalplasma proteins (Cancel et al. 1997; McCauley et al.2001; Roncoletta et al. 2006) and evolutionarily conservedfactors associated with sperm DNA (Wu andChu 2008) on fertilization and early embryo development.As fertility is affected by numerous factors, thesearch for fertility markers should include the wholeanimallevel. Bovine testes must be 4–5°C below bodycoretemperature (38°C) for normal spermatogenesis(Setchell 1978). Assessment of scrotal surface temperatureswith infrared thermography (scrotal thermogram)provided detailed information regarding a bull’sability to regulate testicular temperature (Coulter 1988;Kastelic et al. 2000). Beef bulls with abnormal scrotalthermograms had lower fertility to natural service(Lunstra and Coulter 1997). Furthermore, ultrasonographicevaluations of the testicular vascular cone andits fat cover were indicative of thermoregulatorycapability, and associated with semen productionpotential and semen quality (Kastelic et al. 2001;Arteaga et al. 2005).ConclusionA traditional breeding soundness examination willusually identify bulls that are grossly abnormal. However,a comprehensive approach, including assessingsperm function and fertility at the molecular, cellularand whole-animal levels, is needed to predict fertility ofbulls that are producing apparently normal sperm.ReferencesAitken RJ, 2006: Sperm function tests and fertility. Int JAndrol 29, 69–75.Anzar M, He L, Buhr MM, Kroetsch TG, Pauls KP, 2002:Sperm apoptosis in fresh and cryopreserved bull semendetected by flow cytometry and its relationship with fertility.Biol Reprod 66, 354–360.Arteaga AA, Barth AD, Brito LF, 2005: Relationship betweensemen quality and pixel-intensity of testicular ultrasonogramsafter scrotal insulation in beef bulls. Theriogenology64, 408–415.Ballachey BE, Hohenboken WD, Evenson DP, 1987: Heterogeneityof sperm nuclear chromatin structure and itsrelationship to bull fertility. Biol Reprod 36, 915–925.Ballachey BE, Evenson DP, Saacke RG, 1988: The spermchromatin structure assay. Relationship with alternate testsof semen quality and heterospermic performance of bulls.J Androl 9, 109–115.Barth AD, 2007: Evaluation of potential breeding soundnessof the bull. in: Youngquist RS, Threlfall WR (eds), CurrentTherapy in Large Animal Theriogenology 2. SaundersElsevier, Philadelphia, pp. 228–240.Bellin ME, Oyarzo JN, Hawkins HE, Zhang H, Smith RG,Forrest DW, Sprott LR, Ax RL, 1998: Fertility-associatedantigen on bull sperm indicates fertility potential. J Anim Sci76, 2032–2039.Boerke A, Dieleman SJ, Gadella BM, 2007: A possible role forsperm RNA in early embryo development. Theriogenology68(Suppl. 1), S147–S155.Boilard M, Bailey J, Collin S, Dufour M, Sirard M-A, 2002:Effect of bovine oviduct epithelial cell apical plasmaÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

372 JP Kastelic and JC Thundathilmembranes on sperm function assessed by a novel flowcytometric approach. Biol Reprod 67, 1125–1132.Brito LFC, Barth AD, Bilodeau-Goeseels S, Panich PL,Kastelic JP, 2003: Comparison of methods to evaluate theplasmalemma of bovine sperm and their relationship with invitro fertilization rate. Theriogenology 60, 1539–1551.Browing H, Strome S, 1996: A sperm-supplied factor requiredfor embryogenesis in C. elegans. Development 122, 391–404.Cancel AM, Chapman DA, Killian GJ, 1997: Osteopontin isthe 55-kilodalton fertility-associated protein in Holstein bullseminal plasma. Biol Reprod 57, 1293–1301.Cheng A, Le T, Palacios M, Bookbinder LH, Wassarman PM,Suzuki F, Bleil JD, 1994: Sperm-egg recognition in themouse: characterization of sp56, a sperm protein havingspecific affinity for ZP3. J Cell Biol 125, 867–878.Chenoweth PJ, 2007: Influence of the male on embryo quality.Theriogenology 68, 308–315.Coe PH, Gibson CD, 1993: Adjusted 200-day scrotal size as apredictor of 365-day scrotal circumference. Theriogenology40, 1065–1072.Cormier N, Sirard MA, Bailey JL, 1997: Premature capacitationof bovine spermatozoa is initiated by cryopreservation.J Androl 18, 461–468.Coulter GH, 1988: Thermography of bull testes. In: Proceedingsof the 12th Technical Conference on Artificial Inseminationand Reproduction, National Association of AnimalBreeders, pp. 58–63.Ensslin M, Vogel T, Calvete JJ, Thole HH, Schmidtke J,Matsuda T, To¨pfer-Petersen E, 1998: Molecular cloningand characterization of P47, a novel boar sperm-associatedzona pellucida-binding protein homologous to afamily of mammalian secretory proteins. Biol Reprod 58,1057–1064.Evans JP, Florman HM, 2002: The state of the union: the cellbiology of fertilization. Nat Cell Biol 4, S57–S63.Farrell PB, Presicce DA, Brockett CC, Foote RH, 1998:Quantification of bull sperm characteristics measured bycomputer assisted sperm analysis (CASA) and the relationshipto fertility. Theriogenology 49, 871–879.Fazeli AR, Steenweg W, Bevers MM, de Loos FA, van denBroek J, Colenbrander B, 1993: Development of a spermzona pellucida binding assay for bull semen. Vet Rec 2, 14–16.Fazeli AR, Zhang BR, Steenweg W, Larsson B, Bevers MM,van den Broek J, Rodriguez-Martinez H, Colenbrander B,1997: Relationship between sperm-zona pellucida bindingassays and the 56-day nonreturn rate of cattle inseminatedwith frozen-thawed bull semen. Theriogenology 48, 853–863.Gadea J, Selles E, Marco MA, 2004: The predictive value ofporcine seminal parameters on fertility outcome undercommercial conditions. Reprod Dom Anim 39, 303–308.Gillan L, Evans G, Maxwell WMC, 2005: Flow cytometricevaluation of sperm parameters in relation to fertilitypotential. Theriogenology 63, 445–457.Gwathmey TM, Ignotz GG, Mueller JL, Manjunath P, SuarezSS, 2006: Bovine seminal plasma proteins PDC-109, BSP-A3, and BSP-30-kDa share functional roles in storing spermin the oviduct. Biol Reprod 75, 501–507.Hallap T, Jaakma U, Rodriguez-Martinez H, 2006: Changes insemen quality in Estonian Holstein AI bulls at 3, 5 and7 years of age. Reprod Dom Anim 41, 214–218.Hardy DM, Garbers DL, 1995: A sperm membrane proteinthat binds in a species-specific manner to the egg extracellularmatrix is homologous to von Willebrand factor. J BiolChem 270, 26025–26028.Henault MA, Killian GJ, 1995: Effects of sperm preparationand bull fertility on in vitro penetration of zona-free bovineoocytes. Theriogenology 43, 739–749.Henault MA, Killian GJ, Kavanaugh JF, Griel LC Jr, 1995:Effect of accessory sex gland fluid from bulls of differingfertilities on the ability of cauda epididymal sperm topenetrate zona-free bovine oocytes. Biol Reprod 52, 390–397.Hunter RHF, 2008: Sperm release from oviduct epithelialbinding is controlled hormonally by peri-ovulatory graafianfollicles. Mol Reprod Dev 75, 167–174.Hunter RHF, Fléchon B, Fléchon JE, 1987: Pre- and periovulatorydistribution of viable spermatozoa in the pigoviduct: Scanning electron microscope study. Tissue Cell 19,423–436.Hunter RHF, Fle´chon B, Fle´chon JE, 1991: Distribution,morphology and epithelial interactions of bovine spermatozoain the oviduct before and after ovulation: A scanningelectron microscope study. Tissue Cell 23, 641–656.Ignotz GG, Cho MY, Suarez SS, 2007: Annexins are candidateoviductal receptors for bovine sperm surface proteins andthus may serve to hold bovine sperm in the oviductalreservoir. Biol Reprod 77, 906–913.Inoue N, Ikawa M, Isotani A, Okabe M, 2005: The immunoglobulinsuperfamily protein Izumo is required for sperm tofuse with eggs. Nature 434, 152–153.Januskauskas A, Johannisson A, So¨derquist L, Rodriguez-Martinez H, 2000: Assessment of the fertilizing potential offrozen-thawed bovine spermatozoa by calcium ionophoreA23187-induced acrosome reaction in vitro. Theriogenology53, 859–875.Januskauskas A, Johanmsson A, Rodriguez-Martinez H,2001: Assessment of sperm quality through fluorometryand sperm chromatin structure assay in relation to fieldfertility of frozen thawed semen from Swedish AI bulls.Theriogenology 55, 947–961.Januskauskas A, Johannisson A, Rodriguez-Martinez H,2003: Subtle membrane changes in cryopreserved bull semenin relation with sperm viability, chromatin structure, andfield fertility. Theriogenology 60, 743–758.Jeyendran RS, Van der Ven HH, Perez-Pelaez M, Crabo BG,Zaneveld LJD, 1984: Development of an assay to assess thefunctional integrity of the human sperm membrane and itsrelationship to other semen characteristics. J Reprod Fertil70, 219–228.Kamaruddin M, Kroetsch T, Basrur PK, Hansen PJ, KingWA, 2004: Immunolocalization of heat shock protein 70 inbovine spermatozoa. Andrologia 36, 327–334.Karabinus DS, Vogler CJ, Saacke RG, Evenson DP, 1997:Chromatin structural changes in sperm after scrotal insulationof Holstein bulls. J Androl 18, 549–555.Kastelic JP, Cook RB, Coulter GH, 2000: Scrotal ⁄ testicularthermoregulation in bulls. in: Chenoweth PJ (ed.), Topics inBull Fertility. International Veterinary Information Service.Available from http://www.ivis.org.Kastelic JP, Cook RB, Pierson RA, Coulter GH, 2001:Relationships among scrotal and testicular characteristics,sperm production, and seminal quality in 129 beef bulls.Can J Vet Res 65, 111–115.Killian GJ, Chapman DA, Rogowski LA, 1993: Fertilityassociatedproteins in Holstein bull seminal plasma. BiolReprod 49, 1202–1207.Kondoh G, Tojo H, Nakatani Y, Komazawa N, Murata C,Yamagata K, Maeda Y, Kinoshita T, Okabe M, Taguchi R,Takeda J, 2005: Angiotensin-converting enzyme is a GPIanchoredprotein releasing factor crucial for fertilization.Nat Med 11, 160–166.Krawetz SA, 2005: Paternal contribution: new insights andfuture challenges. Nat Rev Genet 6, 633–642.Larson JL, Miller DJ, 2000: Can relative spermatozoalgalactosyltransferase activity be predictive of dairy bullfertility? J Dairy Sci 83, 2473–2479.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Breeding Soundness and Semen Analysis in Bulls 373Lewis SE, Aitken RJ, 2005: DNA damage to spermatozoa hasimpacts on fertilization and pregnancy. Cell Tissue Res 322,33–41.Lunstra DD, Coulter GH, 1997: Relationship between scrotalinfrared temperature patterns and natural-mating fertility inbeef bulls. J Anim Sci 75, 767–774.Marquant-Le Guienne B, Humblot P, Thibier M, Thibault C,1990: Evaluation of bull semen fertility by homologous invitro fertilization tests. Reprod Nutr Dev 30, 259–266.Matwee C, Kamaruddin M, Betts DH, Basrur PK, King WA,2001: The effects of antibodies to heat shock protein 70 infertilization and embryo development. Mol Hum Reprod 7,829–837.McCauley TC, Zhang H, Bellin ME, Ax RL, 1999: Purificationand characterization of fertility-associated antigen(FAA) in bovine seminal fluid. Mol Reprod Dev 54, 145–153.McCauley TC, Zhang HM, Bellin ME, Ax RL, 2001:Identification of a heparin-binding protein in bovine seminalfluid as tissue inhibitor of metalloproteinases-2. Mol ReprodDev 58, 336–341.Miller D, Ostermeier GC, 2006: Spermatozoal RNA: Why is itthere and what does it do? Gynecol Obstet Fertil 34, 840–846.Miller D, Ostermeier GC, Krawetz SA, 2005: The controversy,potential and roles of spermatozoal RNA. Trends Mol Med11, 156–163.Moura AA, Koc H, Chapman DA, Killian GJ, 2006:Identification of proteins in the accessory sex gland fluidassociated with fertility indexes of dairy bulls: a proteomicapproach. J Androl 27, 201–211.Moura AA, Chapman DA, Killian GJ, 2007: Proteins of theaccessory sex glands associated with the oocyte-penetratingcapacity of cauda epididymal sperm from Holstein bulls ofdocumented fertility. Mol Reprod Dev 74, 214–222.Petrunkina AM, Waberski D, Gunzel-Apel AR, Topfer-Petersen E, 2007: Determinants of sperm quality and fertilityin domestic species. Reproduction 134, 3–17.Phillips NJ, Mcgowan MR, Johnston SD, Mayer DG, 2004:Relationship between thirty post-thaw spermatozoal characteristicsand the field fertility of 11 high-use Australiandairy AI sires. Anim Reprod Sci 81, 47–61.Puglisi R, Balduzzi D, Galli A, 2004: In vitro sperm penetrationspeed and its relationship with in vivo bull fertility.Reprod Dom Anim 39, 424–428.Rodriguez-Martinez H, 2003: Laboratory semen assessmentand prediction of fertility: still utopia? Reprod Dom Anim38, 312–318.Rodriguez-Martinez H, 2007: State of the art in farm animalsperm evaluation. Reprod Fertil Dev 19, 91–101.Roncoletta M, Morani Eda S, Esper CR, Barnabe VH,Franceschini PH, 2006: Fertility-associated proteins inNelore bull sperm membranes. Anim Reprod Sci 91, 77–87.Saacke RG, Dalton JC, Nadir S, Nebel RL, Bame JH, 2000:Relationship of seminal traits and insemination time tofertilization rate and embryo quality. Anim Reprod Sci 2,663–677.Setchell BP, 1978: The scrotum and thermoregulation. In:Setchell BP (ed), The Mammalian Testis. Cornell UniversityPress, Ithaca, pp. 90–108.Suarez SS, 2002: Formation of a reservoir of sperm in theoviduct. Reprod Domest Anim 37, 140–143.Thundathil J, Gil J, Januskauskas A, Larsson B, So¨derquist L,Mapletoft R, Rodrıguez-Martınez H, 1999: Relationshipbetween the proportion of capacitated spermatozoa presentin frozen-thawed semen and fertility with artificial insemination.Int J Androl 22, 366–373.Thundathil J, Meyer R, Palasz AT, Barth AD, Mapletoft RJ,2000: Effect of the knobbed acrosome defect in bovinesperm on IVF and embryo production. Theriogenology 54,921–934.Thundathil J, Palomino J, Barth A, Mapletoft R, Barros C,2001a: Fertilizing characteristics of bovine sperm with flattenedor indented acrosomes. Anim Reprod Sci 67, 231–243.Thundathil J, Palasz AT, Barth AD, Mapletoft RJ, 2001b: Theuse of in vitro fertilization techniques to investigate thefertilizing ability of bovine sperm with proximal cytoplasmicdroplets. Anim Reprod Sci 65, 181–192.Thundathil JC, Anzar M, Buhr MM, 2006: Na + ⁄ K + ATPaseas a signaling molecule during bovine sperm capacitation.Biol Reprod 75, 308–317.Truelson SL, Graham JK, Mortimer RG, Field TG, 1996: Invitro penetration into bovine oocytes and zona-free hamsteroocytes by bull spermatozoa treated with liposomes. J DairySci 79, 991–999.Walters AH, Eyestone WE, Saacke RG, Pearson RE, GwazdauskasFC, 2005: Bovine embryo development after IVFwith spermatozoa having abnormal morphology. Theriogenology63, 1925–1937.Walters AH, Saacke RG, Pearson RE, Gwazdauskas FC,2006: Assessment of pronuclear formation following in vitrofertilization with bovine spermatozoa obtained after thermalinsulation of the testis. Theriogenology 65, 1016–1028.Waterhouse KE, Haugan T, Kommisrud E, Tverdal A,Flatberg G, Farstad W, Evenson DP, De Angelis PM,2006: Sperm DNA damage is related to field fertility ofsemen from young Norwegian Red bulls. Reprod Fertil Dev18, 781–788.Watson PF, 2000: The causes of reduced fertility withcryopreserved semen. Anim Reprod Sci 60–61, 481–492.Wu TF, Chu DS, 2008: Epigenetic processes implementedduring spermatogenesis distinguish the paternal pronucleusin the embryo. Reprod Biomed Online 16, 13–22.Wu AT, Sutovsky P, Manandhar G, Xu W, Katayama M,Day BN, Park KW, Yi YJ, Xi YW, Prather RS, Oko R,2007: PAWP, a sperm-specific WW domain-binding protein,promotes meiotic resumption and pronuclear developmentduring fertilization. J Biol Chem 282, 12164–12175.Yanagimachi R, 1994: Mammalian fertilization. In: Knobil E,Neill JD (eds), The Physiology of Reproduction. RavenPress, New York, pp. 189–317.Zhang BR, Larsson B, Rodriguez-Martinez H, 1995: Influenceof batches of bovine oocytes on the outcome of an intactzona pellucida binding assay and in vitro fertilization. Int JAndrol 18, 213–220.Zhang BR, Larsson B, Lundeheim N, Rodriguez-Martinez H,1997: Relationship between embryo development in vitroand 56-day non return rates of cows inseminated withfrozen-thawed semen from dairy bulls. Theriogenology 48,221–231.Zhang BR, Larsson B, Lundeheim N, Håård MG, Rodriguez-Martinez H, 1999: Prediction of bull fertility by combined invitro assessments of frozen-thawed semen from young dairybulls entering an AI-programme. Int J Androl 22, 253–260.Zhao Y, Buhr MM, 1996: Localization of various ATPases infresh and cryopreserved bovine spermatozoa. Anim ReprodSci 44, 139–148.Author’s address (for correspondence): John Kastelic, LethbridgeResearch Centre, 5403-1st Ave South, P.O. Box 3000, Lethbridge, AB,Canada T1J 4B1. E-mail: kastelicj@agr.gc.caConflict of interest: J Kastelic is the Co-Editor-in-Chief of the journalTheriogenology, published by Elsevier Science, Inc; J Thundthil hasreceived funding support from the University of Calgary; Agricultureand Food Council of Alberta and Alberta Livestock IndustryDevelopment Fund; Westgen Endowment Fund; and L’AllianceBoviteq Inc.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 374–378 (2008); doi: 10.1111/j.1439-0531.2008.01187.xISSN 0936-6768Sanitary Procedures for the Production of Extended SemenGC AlthouseDepartment of Clinical Studies, New Bolton Center, University of Pennsylvania, Kennett Square, PA, USAContentsSemen is collected and processed from a variety of animalspecies for use in artificial insemination breeding programmes.Because of the inherent nature of the semen collection process,bacterial contamination of the ejaculate is a common occurrence.Additionally, manipulation of the ejaculate duringprocessing in the laboratory can expose the sample to possibleintroduction of bacterial contamination. If preventative measuresat the stud fail to adequately control these risks,decreases in semen quality, dose longevity and fertility mayoccur. Multiple mammalian and non-mammalian sources havebeen identified as origins of contamination in the stud.Knowledge of these sources has aided the industries indeveloping strategies that help in controlling the introductionof contaminant bacteria in extended semen. A primary step inminimizing contamination is in the practice of good hygiene bystud personnel. Prudent general sanitation protocols shouldalso be followed in the laboratory, animal housing and semencollection areas. Cleanliness and attention to the actual semencollection process can also aid in reducing bacterial loadoriginating from the stud semen donor. Attentiveness to all ofthese steps significantly contributes to an overall reduction inthe type and amount of bacterial contamination. However,their complete elimination stills remains unavoidable. Toaddress residual bacteria load in the sample, antimicrobialsare commonly used in semen extenders intended to promotein vitro sperm longevity beyond that of a few hours. Currentresearch by the animal industries continues in the selection andprudent use of antimicrobials that will lead to the successand sustainability of this modality in controlling bacterialcontamination.IntroductionMany animal industries have come to embrace certainassisted reproductive technologies (ART). Of particularinterest is the technique of artificial insemination (AI).Not only was it one of the earliest ART’s applied at acommercial level, but it remains the most common andwidespread ART used today (Althouse 2007). Bothdomestic food (i.e. beef, dairy, poultry, swine) andcompanion (i.e. dogs, horses) animal industries havereadily incorporated AI as a viable option in theirbreeding management programmes. Underlying thereproductive success of any of these programmes,however, is the need for and availability of a goodquality extended semen product.Driven in part by scales of efficiency and everprevalent biosecurity concerns, many animal industrieshave evolved to the point of having dedicated farms (i.e.studs) which house the stud animals from which semenis collected, analysed and processed. The end result ofthis processing is the development of semen doses, whichare then fresh-extended ⁄ chilled or frozen and subsequentlydelivered by ground or mail courier to thebreeding farm. In those industries which utilizefresh-extended ⁄ chilled semen, an evolution in extendertechnology has and continues to occur, which lengthensthe number of days that sperm viability can be maintainedin this liquid, chilled state. An undesirable sideeffect of this process, however, has been a concomitantincrease in the incidence of extended semen doses whichcontain spermicidal levels of bacterial contamination(Althouse et al. 2000). These spermicidal effectsare direct and are bacterial concentration-dependent(Teague et al. 1971; Auroux et al. 1991; Wolff et al.1993; Monga and Roberts 1994; Diemer et al. 1996).The purpose of this paper is to present the generalconcepts and current understanding of bacterial dynamicsin extended semen, and to discuss methods thereofto reduce and control bacterial contamination.Bacterial Growth Dynamics in Semen ExtendersGiven the inherent nature and impact of those bacteriawhich exhibit pathogenesis, much work has beenperformed over the years investigating bacterial growthand the identification of factors which regulate suchgrowth. This available body of work in addition to morerecent works investigating bacteria growth dynamics inselect semen extenders (Varner et al. 1998; Althouse andLu 2005; Althouse et al. 2008) provides us with a goodunderstanding of bacterial behaviour under variousconditions, and of which much of the following discussionis based.Most bacteria genera today exhibit the qualities ofbeing phenomenally hardy organisms with a survivalistability to adapt to a wide variation of conditions. Withextenders used for storing semen, a primary purpose ofthese diluents is to provide the necessary nutrients andmilieu to prolong in vitro sperm cell viability. Unfortunately,these same attributes make for a potential mediathat contaminant bacteria can flourish in if thesecontaminants exhibit resistance to the inclusion antibiotics.Growth patterns of bacteria are similar acrossmany genera. A typical bacterial growth curve ispresented in Fig. 1.After introduction of bacteria to a novel environment(e.g. semen extender), a period of negligible or laggrowth occurs. The length of time for this lag phase isdependent upon such things as the genera inoculated,level of inoculums, time needed for recovery from thephysical damage or shock because of the transfer,environmental conditions, and the innate temporalrequirements of the bacterium to synthesize essentialcomponents necessary to metabolize the available substrate(s).Once adaptation is met, a period of rapidgrowth ensues. During this log or exponential phase,cells are doubling on a regular basis, leading to aÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Sanitary Procedures for Extended Semen 375Fig. 1. Phases of bacterial growth (adapted from Banwart 1979)geometric progression in cell number. Essential componentsof this growth process may include an increase incell mass and ribosome number, duplication of thebacterial chromosome, cell wall and plasma membranesynthesis, chromosomal partitioning with septum formation,and finally cell division. This rapid growthphase eventually reaches a point at which it plateaus,transitioning into the stationary phase. The bacterialconcentration and length of time of the stationary phasehas traditionally been thought to be controlled bynutrient exhaustion, physical space limitation, and ⁄ oraccumulation of inhibitory levels of metabolites, includingend products. Recent work has also alluded to thesecretion of quorum-sensing compounds, which allowfor interbacterial communication in the gauging ofpopulation density (Carbonell et al. 2002; Gonzalez andKeshavan 2006; Fulghesu et al. 2007). If no change ismade in the present conditions, a death phase eventuallycommences, which is driven by the decline in the viablecell population.As with spermatozoa, bacteria are unable to regulatetheir temperature. Spermatozoa are dependent upon theenvironment to provide this critical component toprolonged viability. Likewise, this component is ofequal importance to the viability and growth of bacteria.All bacteria have an optimum temperature (T o ) – thetemperature at which maximum growth rate is attained.If the bacteria are exposed to environmental temperaturesabove their T o , plasma membrane fluidityincreases, with concomitant alteration of cell functionand decreased growth. At some temperature point, nofurther compensation can occur, leading to perturbationof the cell and death. Spermatozoa are much moresensitive to increased temperature change than manyother cell types, including bacteria. Unlike bacteria,sperm do not have intracellular organelles to aid inmodifying plasma membrane fluidity for survivability.Consequently, exposure to a matter of a few degreesabove body temperature is all that is necessary for spermdeath to occur.As environmental temperature decreases (e.g. chillingfresh-extended semen), cell plasma membrane fluiditychanges in conjunction with decreased growth rate andmetabolism. At some point, growth and metabolismcease, and the cells become dormant. This phenomenonis employed to reduce metabolism and induce dormancyof sperm in extended semen. Although this is beneficialto sperm longevity, it can benefit contaminant bacteriaeven more, as they have the intracellular machinery toadapt to lower environmental temperatures to the pointthat some genera may again thrive giving way to theirprofuse overgrowth in the chilled extender.From prior studies in swine, the majority of contaminantsidentified in extended semen are Gram-negativebacteria, with a majority of the contaminant bacteriabelonging to the Family Enterobacteriaceae (Althouseet al. 2000, 2008; Althouse and Lu 2005). Our laboratoryhas seen similar types of contaminants in extendedsemen from other species as well. In these and otherstudies, bacterial contaminant overgrowth of certaingenera has been associated with deleterious effects onsemen quality and longevity (Sone et al. 1989; Althouseet al. 2000; Akhter et al. 2008; Aurich and Spergser2007). Spermatozoal agglutination, decreased spermatozoalmotility and viability are potential end effects tobacteriospermic overgrowth. When such doses are usedin an AI programme, increased regular returns tooestrus, post-insemination vulvar discharges andreduced herd reproductive performance have beenreported in some species (Sone et al. 1989; Althouseet al. 2000; Bukharin et al. 2000). Depending uponbacterial type and species along with species origin ofthe sperm, an actual sperm to bacteria (CFU) ratio ofapproximately 1:1 up to 100:1 in a sample appearssufficient to generate the undesired effects on the semendose and fertility (Tamuli et al. 1984; Auroux et al.1991; Althouse et al. 2000), with these effects being of atime-dependent nature. Hence, extenders which providethe ability to extend storage life of sperm also providethe opportunity for certain contaminant bacteria toreach their necessary threshold for eliciting a negativeeffect on fertility potential.Sources and Control of ContaminationUsing pigs as a model, multiple contamination pointshave been identified in the production of extendedsemen (Althouse et al. 2000; Althouse and Lu 2005).From this work, contamination sources have beenclassified into either being of mammalian or nonmammalianorigin, and are usually found to originatefrom a single source on a stud ⁄ farm. These singlesources have been found to yield the same or differentbacterial strains at different times at the stud ⁄ farm.Likewise, different sources of contamination have beenfound at different times on the same stud ⁄ farm. Commonsources of bacterial contamination of extendedsemen are outlined in Table 1. Many of these sources,once contaminated, act as seeding points for furtherejaculates which come into contact with this source.From our laboratories work performed to date,a multitude of contamination sources appear to exist.The first step in minimizing bacterial contaminationof extended semen is the practice of good hygiene andgeneral sanitation by personnel. Personnel which comein contact with any materials which are subsequentlyused in the collection and processing of semen, or whoÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

376 GC AlthouseTable 1. Potential sources of bacterial contamination of extendedsemenMammalianFaecalPreputial cavity fluidsSkin ⁄ hairRespiratory secretionsPersonnelNon-mammalianWater (e.g. tap, purified)Plant matter (i.e. feed, bedding)Sinks ⁄ drainsAir ⁄ ventilation systemsInanimate objectsare involved in the actual semen collection process, needto be aware that their person can be a source ofcontamination or act as a fomite in the transfer ofcontamination. In these situations, prudent hand washingand ⁄ or use of protective gloves should be encouraged.Personnel should avoid any contact of their barehands with materials which can later come into contactwith the semen or extender. Personnel with upperrespiratory infections should be cognizant of and avoidcontamination of materials, semen or extender throughaerosolization of the contaminant during sneezing orcoughing. Caps and hair nets can be of value if worn bypersonnel performing the semen collection process andby laboratory personnel as an aid in minimizing hairand dander as a contamination source. Clean protectivegarments and shoes ⁄ boots, provided on site by the stud,should be available for use by all stud personnel.Prudent general sanitation protocols should be inplace in the laboratory, animal housing and semencollection areas. In the barn, regular removal of organicmaterial, a byproduct of normal husbandry, followed bya thorough cleaning with a broad-spectrum disinfectantis recommended. Use of sanitizers and disinfectantsshould be used prudently in the laboratory, but productsand protocols need to be specific to avoid residues,which may be detrimental to semen quality upon latercontact.To minimize bacterial load originating from the studsemen donor, the ventral abdomen may need to becleaned and dried prior to commencing with semencollection. Trimming of hair surrounding the preputialorifice should occur on an as needed basis to eliminatethe accumulation of organic matter at this site and itsinadvertent introduction into the ejaculate during semencollection. Cleaning of the preputial opening andsurrounding area with a single-use disposable wipeshould be considered if the area is wet and ⁄ or hasorganic material present. In some species, preputialfluids can contain high numbers of bacteria (Aamdalet al. 1958); therefore, these fluids should be evacuatedimmediately prior to the semen collection process. Whencollecting semen using an artificial vagina or glovedhand, the collector should position the penis in a mannerconsistent with minimizing gravitational contaminationof the semen collection vessel with preputial fluids.Lastly, if performing gloved hand semen collection,diverting the pre-sperm fraction from the semen collectionvessel may help in reducing ejaculate bacterial load.A synopsis outlining general stud hygiene and sanitationprocedures is provided in Table 2.Table 2. General stud hygiene and sanitation recommendationsStud personnel1. Application of good hand hygiene, including appropriate washing and use of protective gloves, should be practiced throughout all areas of the stud2. Personnel should avoid any contact of bare hands with materials which can later come into contact with the semen or extender3. Personnel with upper respiratory infections should be cognizant of and avoid contamination of materials, semen or extender through aerosolization duringsneezing or coughing4. Caps and hair nets can be of value if worn by personnel performing the semen collection process and by laboratory personnel as an aid in minimizing hairand dander as a contamination source5. Clean protective garments and shoes ⁄ boots, provided on site by the stud, should be available for use by all stud personnelAnimal housing ⁄ handling1. Animal housing should be put on a regular sanitary maintenance schedule, including removal of organic material and application of a broad-spectrumdisinfectant2. Trimming of hair surrounding the preputial orifice should occur on an as needed basis to eliminate the accumulation of organic matter at this site and itsinadvertent introduction into the ejaculate during semen collection3. The ventral abdomen should be clean and dry prior to commencing with semen collection4. Cleaning of the preputial opening and surrounding area with a single-use disposable wipe should be considered if the area is wet and ⁄ or has organic materialpresent5. In some species, preputial fluids can contain high numbers of bacteria, therefore, these fluids should be evacuated immediately prior to the semen collectionprocess6. When collecting semen using an artificial vagina or gloved hand, the collector should position the penis in such a way as to minimize gravitationalcontamination of the semen collection vessel with preputial fluids7. If performing gloved hand semen collection, diverting the pre-sperm fraction from the semen collection vessel may aid in reducing ejaculate bacterial load8. The semen collection area and any collection equipment should be thoroughly cleaned and disinfected at the end of each collection dayLaboratory1. Encourage single-use disposable products when economically feasible to minimize cross-contamination2. When using reusable laboratory materials (i.e. glassware, plasticware, plastic tubing, containers, etc.) which cannot be heat ⁄ gas sterilized or boiled, cleanthese reusable’s initially using a laboratory-grade detergent (residue-free) with water, followed by a distilled water rinse, and lastly through a 70% alcohol(non-denatured) rinse. Allow sufficient time and proper ventilation for complete evaporation of residual alcohol on the reusable. Rinse reusable with semenextender prior to their first use of the day3. Laboratory purified water should be checked on a minimum quarterly basis if in-house, and by lot if outsourced. Any bacterial growth should be consideredsignificant and appropriate action taken to identify and eliminate the contaminant source4. Disinfect countertops and contaminated lab equipment at end of processing day with a residue-free detergent and rinse5. Floor should be mopped at end of day with a disinfectant6. Break down bulk products into smaller, daily use quantities immediately after opening7. Ultraviolet lighting can be installed to aid in sanitizing reusable’s and laboratory surfaces; however, safety precautions should be integrated to preventexposure to personnelÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Sanitary Procedures for Extended Semen 377Because of the level of frequency in which watersources have been identified as a source of contamination,regular monitoring of these sources is an importantcomponent in a control programme. Well or municipalwater should be checked on a minimum quarterly basisfor coliform load, which should meet or exceed localpublic health ordinances. Purified water systems shouldbe checked on a minimum quarterly basis; any bacterialgrowth should be considered significant and appropriateaction taken to eliminate the contaminant from thesystem. Facilities using purchased purified water shouldhave each lot checked for bacterial growth; once again,any bacterial growth is significant, and another watersource should be considered.Final quality assurance for bacterial control should beascertained by regular monitoring of the extended semensamples. Depending upon the desired level of qualitycontrol, whole system or point dose screening formicrobial contamination can be performed on randomlyselected individual or pooled semen which are at least48 h of age post-processing (Althouse et al. 2003). Ourown preference is to perform this monitoring on theextended semen product on its day of expiration. As ageneral rule, recommended number of doses screenedper month should be 1% of total monthly collections orfour samples ⁄ week, whichever is greater (Althouse et al.2003). Any positive contaminant growth should becritically assessed and a follow-up performed.In Vitro Control of Contamination in ExtendedSemenBecause of the unavoidable presence of bacteria in theejaculate, even under the best hygiene and sanitaryconditions, antimicrobials are a necessary component ofsemen extenders which are intended to promote in vitrolongevity of sperm beyond that of a few hours. Whenlooking across our common domestic animal species, themost common classes of antimicrobials used today insemen extenders are the aminoglycosides, aminocyclitols,b-lactams, lincosamides, macrolides, polypeptides andquinolones (Table 3). In general, these antimicrobialsTable 3. Common antimicrobials currently used in semen extendersBovine Equine Porcine PoultryAminocyclitolsSpectinomycin d dAminoglycosidesAmakacindGentamicin d d d dNeomycindStreptomycin d d dBeta-lactamsAmoxicillindPenicillin d d dTicarcillindLincosamidesLincomycin d dMacrolidesTylosin d d dPolypeptidesPolymixin d dQuinolonesEnrofloxacin d dwork on susceptible bacteria by blocking or interferingwith a biochemical reaction essential to the life cycle.Several thresholds must be obtained for antimicrobialsto work in bacterially contaminated extended semenin vitro. The selected antimicrobial(s) must be at sufficientconcentrations to provide adequate readily-availableactive product, the compound must be in a state tobe able to permeate (active and ⁄ or passive mechanisms)the bacteria in sufficient quantities, and the compoundmust occupy a sufficient number of active sites over aperiod of time within the bacteria to elicit their detrimentaleffect. Interference in this process at any point canlead to antimicrobial resistance of the bacteria (forreview, see Althouse and Lu 2005). Recent work hasbegun to elucidate the thermotemporal dynamics ofcontaminant bacteria and antimicrobials in semenextenders (Althouse et al. 2008). Through this typework, it is hoped that prudent antimicrobial selectionand use by the industries will lead to the success andsustainability of this modality in our efforts to controlcontamination.ConclusionsThe current body of research substantiates the risk thatexists of contaminating semen with bacteria during thecollection and processing of an ejaculate. If preventativemeasures at the stud fail to adequately control this risk,decreases in semen quality, dose longevity and fertilitymay occur. To date, multiple mammalian and nonmammaliansources have been identified as origins ofthis contamination at a stud. In addition to generalcleanliness, targeted hygiene and sanitation procedurescan greatly aid in controlling the introduction ofcontaminant bacteria in the extended semen product.Even with these measures in place to reduce incomingbacterial load, it is unrealistic to expect that an extendedsemen product will be devoid of bacterial contamination.Hence, prudent use of antimicrobials in the semendiluents’ is an essential component to the overall controlof contaminant bacteria in extended semen usedin programmes incorporating assisted reproductivetechnologies.ReferencesAamdal J, Hogset I, Filseth O, 1958: Extirpation of thepreputial diverticulum of boars used for artificial insemination.J Am Vet Med Assoc 132, 522–524.Akhter S, Ansari MS, Andrabi SM, Ullah N, Qayyum M,2008: Effects of antibiotics in extender on bacterial andspermatozoal quality of cooled buffalo (Bubalus bubalis)bull semen. Reprod Domest Anim 43, 272–278.Althouse GC, 2007: Artificial insemination. In: Schatten H,Constantinescu GM (eds), Comparative Reproductive Biology.Blackwell Publishing, Ames, IA, USA, pp. 159–169.Althouse GC, Lu KG, 2005: Bacteriospermia in extendedporcine semen. Theriogenology 63, 573–584.Althouse GC, Kuster CE, Clark SG, Weisiger RM, 2000:Field investigations of bacterial contaminants and theireffects on extended porcine semen. Theriogenology 53,1167–1176.Althouse GC, Reicks D, Spronk GD, Trayer TP, 2003: Health,hygiene, and sanitation guidelines for boar studs providingÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

378 GC Althousesemen to the domestic market. J Swine Health Prod 11, 204–206.Althouse GC, Pierdon MS, Lu KG, 2008: Thermotemporaldynamics of contaminant bacteria and antimicrobials inextended porcine semen. Theriogenology, in press.Aurich C, Spergser J, 2007: Influence of bacteria and gentamicinon cold-stored stallion spermatozoa. Theriogenology67, 912–918.Auroux MR, Jacques L, Mathieu D, Auer J, 1991: Is thesperm bacterial ratio a determining factor in impairment ofsperm motility: an in-vitro study in man with Escherichiacoli. Int J Androl 14, 264–270.Banwart GJ, 1979: Basic Food Microbiology. AVI PublishingCo., Westport, i pp.Bukharin OV, Kuz’min MD, Ivanov IuB, 2000: [The role ofthe microbial factor in the pathogenesis of male infertility].Zh Mikrobiol Epidemiol Immunobiol 2, 106–110.Carbonell X, Chorchero JL, Cubarsi R, Vila P, Villaverde A,2002: Control of Escherichia coli growth rate through celldensity. Microbiol Res 157, 257–265.Diemer T, Weidner W, Michelmann HW, Schiefer HG, RovanE, Mayer F, 1996: Influence of Escherichia coli on motilityparameters of human spermatozoa in vitro. Int J Androl 19,271–277.Fulghesu L, Giallorenzo C, Savoia D, 2007: Evaluation ofdifferent compounds as quorum sensing inhibitors in Pseudomonasaeruginosa. J Chemother 19, 388–391.Gonzalez JE, Keshavan ND, 2006: Messing with bacterialquorum sensing. Microbiol Mol Biol Rev 70, 859–875.Monga M, Roberts JA, 1994: Sperm agglutination by bacteria:receptor-specific interactions. J Androl 15, 151–156.Sone M, Kawarasaki T, Ogasa A, Nakahara T, 1989: Effectsof bacteria-contaminated boar semen on the reproductiveperformance. Jpn J Anim Reprod 35, 159–164.Tamuli MK, Sharma DK, Rajkonwar CK, 1984: Studies onthe microbial flora of boar semen. Indian Vet J 61, 858–861.Teague NS, Bogarsky S, Glenn JF, 1971: Interference ofhuman spermatozoa motility by Escherichia coli. Fertil Steril22, 281–285.Varner DD, Scanlan CM, Thompson JA, Brumbaugh GW,Blanchard TL, Carlton CM, Johnson L, 1998: Bacteriologyof preserved stallion semen and antibiotics in semenextenders. Theriogenology 50, 559–573.Wolff H, Panhans A, Stolz W, Meurer M, 1993: Adherence ofEscherichia coli to sperm: a mannose mediated phenomenonleading to agglutination of sperm and E. coli. Fertil Steril 60,154–158.Author’s address (for correspondence): GC Althouse, Department ofClinical Studies, New Bolton Center, University of Pennsylvania, 382West Street Road, Kennett Square, Pennsylvania 19348-1692, USA.E-mail: gca@vet.upenn.eduConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 379–385 (2008); doi: 10.1111/j.1439-0531.2008.01188.xISSN 0936-6768Management of Goat Reproduction and Insemination for Genetic Improvement inFranceB Leboeuf 1 , JA Delgadillo 2 , E Manfredi 3 , A Piace` re 4 , V Cle´ment 4 , P. Martin 5 , M Pellicer 6 , P Boue´5 and R de Cremoux 41 INRA-UEICP, Venours, Rouille´, France; 2 CIRCA, Universidad Autonoma Agraria Antonio Narro, Torreon, Coahuila, Mexico; 3 INRA SAGA;4 Institut de l’Elevage, Castanet-Tolosan cedex; 5 Caprige`ne-France, Agropole, Mignaloux-Beauvoir; 6 Physiologie de la Reproduction et desComportements, UMR 6175 INRA-CNRS-Universite´ de Tours-Haras Nationaux, Nouzilly, FranceContentsReproductive seasonality observed in all breeds of goatsoriginating from temperate latitudes and in some breeds fromsubtropical latitudes can now be controlled by artificialchanges in photoperiod. Short days stimulate sexual activity,while long days inhibit it. This knowledge has allowed thedevelopment of photoperiodic treatments to control sexualactivity in goats, for both the buck and doe. In the Frenchintensive milk production system, goat AI plays an importantrole to control reproduction and, in conjunction with progenytesting, to improve milk production. Most dairy goats areinseminated out of the breeding season with deep frozensemen, after induction of oestrus and ovulation by hormonaltreatments. This protocol provides a kidding rate of approximately65%. New breeding strategies have been developed,based on the buck effect associated with AI, to reduce the useof hormones. With the development of insemination withfrozen semen, a classical selection programme was set up,including planned mating, progeny testing and the diffusion ofproved sires by inseminations in herds. Functional traits havebecome important for efficient breeding schemes in the dairygoat industries. Based on knowledge gained over the pastdecade, the emphasis in selective breeding has been placed onfunctional traits related to udder morphology and health. Newwindows have been opened based on new molecular tools,allowing the detection and mapping of genes of economicimportance.IntroductionGoats and sheep exhibit seasonal changes in reproductiveactivity, accompanied by variations in the availabilityof products over the year. Out-of-breeding-seasonreproduction allows milk and meat production throughoutthe year in accordance with commercial requirementsand consumer expectations. Artificialinsemination has an important role in goat breeding,especially in intensive systems of production, to controlreproduction and, in conjunction with accurate progenytesting, to improve the production of milk, hair andmeat. At the farm level, the control of reproductionallows kidding at a precise season of the year,a synchronization of kidding over a limited period oftime, and facilitates supplementary feeding to meet thedemands of lactation. Most goat milk is processed intohigh-quality cheese. As a result of the evolution ofdemand by consumers for typically regional cheese andfor more quality and safety, dairy goats have to beimproved for production (milk yield) and milk composition(fat and proteins) to remain competitive. Newfunctional traits must also be considered so as to reduceproduction costs and increase product quality and safety.Reproductive Seasonality and Control of theTiming of ReproductionSeasonal variations in female and male sexual activityA reproductive seasonality is observed in most breeds ofgoat originating from high latitudes (>35°) and in somelocal breeds from subtropical latitudes (25°–35°). In theAlpine breed, as in the local does from subtropicalMexico, the breeding season begins in the early autumnand ends in the late winter (Chemineau et al. 1992;Delgadillo et al. 2004a). Males of these breeds alsodisplay wide changes in their sexual activity. In Alpinebucks, the breeding season begins in the early autumnand ends in late winter, while in local bucks fromsubtropical Mexico, it begins in the late spring and endsin the early winter (Delgadillo et al. 1991, 1999). In theseseasonal breeds, sexual behaviour, testicular size, anindex of the spermatogenetic activity, and the quantitativeand qualitative sperm production of bucks decreasesdramatically during the non-breeding season (Delgadilloet al. 1991, 1999).Treatments to control the timing of reproductionPhotoperiodic control of reproductionThe annual reproductive rhythm in breeds of goats andsheep from temperate latitudes and in some local breedsfrom the subtropics is controlled by photoperiod(Delgadillo et al. 2004b; Chemineau et al. 2006). Inartificial conditions, plasma testosterone levels increaseduring short days and decrease during long days(Delgadillo et al. 2004b). In females, ovulations startduring short days and end during long days (Gebbieet al. 1999). In order to manipulate sexual activity, theanimals must perceive alternations between long andshort days to prevent the establishment of sexualrefractoriness.Control of male sexual activity by abolition of seasonalvariations or induction of sexual activity in thenon-breeding seasonIn Alpine and Saanen male goats subjected to alternationsof 1 or 2 months of long days (16 h of light ⁄ day:LD) and 1 or 2 months of short days (8 h of light ⁄ day:SD), marked seasonality of sexual behaviour, testis sizeand quantitative and qualitative sperm production wereabolished or reduced (Delgadillo et al. 1991, 1992).When bucks were collected twice a week during twoconsecutive years, these light-treated males producedÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

380 B Leboeuf, JA Delgadillo, E Manfredi, A Piace` re, V Cle´ment, P Martin, M Pellicer, P Boue´, and R de Cremoux50% more AI doses than a control group under naturallight (Delgadillo et al. 1991). Fertility of deep-frozensemen was in the same range as that obtained in thecontrol animals during the natural sexual season. Theseresults were improved using alternations of 1.5 monthsof LD and 1.5 months of SD (Leboeuf et al. 2004a).This photoperiodic treatment is now used in the Frenchnational selection programme.In male goats, sexual activity can be stimulated duringthe non-breeding season using artificial long daysfollowed by melatonin or natural photoperiod(Delgadillo et al. 2004a; Pellicer-Rubio et al. 2007). Insubtropical Mexican male goats, 2.5 months of longdays (16 h of light ⁄ day) from 1st of November followedby natural photoperiod, increased testosterone secretionfrom February to April, during the non-breedingseason. These long-day treated animals displayed anintense sexual behaviour when exposed to anoestrusdoes. In the control groups, testosterone secretion andsexual behaviour were lower than in the animals treatedwith artificial photoperiod (Delgadillo et al. 2004a).Control of female sexual activity in the non-breedingseason by the male effectThe so-called male effect is a technique to stimulate thesexual activity in seasonally anovulatory goats (Pellicer-Rubio et al. 2007). Most female goats have a shortovarian cycle of 5–7 days, followed by a second ovulationassociated with oestrous behaviour and a normalluteal phase (Chemineau et al. 2006). The response ofdoes to the male effect can be improved by the use ofmales rendered sexually active during the non-breedingseason by photoperiodic treatments. In subtropicalMexican goats, males treated with 2.5 months of longdays were very effective at inducing fertile ovulatoryactivity in anovulatory goats kept under natural photoperiodbut not untreated bucks (80% vs 10%, respectively;Delgadillo et al. 2004a). In more seasonal breeds(Alpine and Saanen), the treatment of females withartificial photoperiod is also necessary to improve theresponse to the male effect. Under these conditions,most does exposed to males ovulated (99%) and 81%kidded (Pellicer-Rubio et al. 2007).Production and Storage of Semen for ArtificialInseminationManagement of semen production and storageThe major problem for goat semen preservation is thebulbo-urethral gland secretion (BUS) responsible fordeterioration of spermatozoa diluted in skimmed milkor in media containing egg yolk when stored at 37°C ordeep frozen. The component from BUS responsible forthis effect was identified as a 55- to 60-kDa glycoproteinlipase (Busgp60) belonging to the pancreatic lipaserelatedprotein 2 family (GoPLRP2; Sias et al. 2005).This enzyme was able to hydrolyse both triolein andmilk triglycerides into free fatty acids, which stronglyinhibit the motility and damage the membranes of buckspermatozoa (Pellicer-Rubio and Combarnous 1998).Techniques of production and storage of semen basedon cryopreservation have been proposed by Corteel(1974). After separation of seminal plasma from spermcells to avoid deterioration of spermatozoa, sperm cellsare diluted in a skimmed milk-based glucose (0.5 M) andglycerol 7%; Corteel 1974). Semen is then stored in0.2 ml straws containing 1 · 10 8 sperm cells and deepfrozen progressively in three steps into liquid nitrogen()196°C).For liquid preservation, goat semen is usually stockedat 4°C. Many extenders were tested for liquid preservationbut the most efficient one is a skimmed milk-basedmedia (Dauzier 1966), or Native Phosphocaseinate(PPCN) (Leboeuf et al. 2004b). However, sperm fertilityafter AI is retained for only 12–24 h and decreasesthereafter with the duration of liquid preservation.Furthermore, there was no significant difference offertility after AI between semen ‘washed’ before dilutionand semen ‘unwashed’. So, it seems that the deleteriouseffect of seminal lipase is minimized at 4°C whencompared with its effect at 37°C or after freezing ⁄ thawing.Use of stored semen by artificial inseminationInsemination of cycling goats in the breeding season afternatural oestrusDifferent practical methods of oestrous detection areused, including entire males fitted with an apron,vasectomized males or androgenized castrated malesor females fitted with a harness and marking crayon.Fertility of goats inseminated once in the 24 h followingthe beginning of oestrus with deep-frozen-thawed spermatozoaranges from 60% to 65% (Corteel 1977).Insemination after synchronized and induced oestrus byhormonal treatmentIn France, the synchronization treatment most commonlyused for dairy goats consists of the insertion of avaginal sponge impregnated with a synthetic analogue ofprogesterone (45 mg fluorogestone acetate) for 11 days,together with an intramuscular injection of pregnantmore serum gonadotrophin (PMSG) (400–600 IUaccording to the milk production and season of treatment)and 50 lg of cloprostenol, both administered 48 hbefore sponge removal. Treated goats are inseminatedonce with a frozen-thawed dose of 1 · 10 8 spermatozoa,at 43–45 h after sponge removal (Corteel et al. 1988). Thefertility rate obtained in dairy goats given one inseminationis 60–65% (Boué and Sigwald 2001).Insemination after male effect in anoestrusThe new societal trends and European legislation, whichare opposed to the use of hormones and syntheticsubstances in the animal industries, have encouragedproducers to adopt practices that minimize or completelyavoid the use of synthetic chemicals andhormone treatments (Martin et al. 2004). In this regard,the male effect may be an efficient non-pharmacologicalalternative to hormones for inducing and synchronizingoestrus in AI programmes. Artificial insemination protocolsincluding one or two inseminations over a 24-hperiod determined by the occurrence of oestrus or by theintroduction of the buck have been validated. In theÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Management of Goat Reproduction and Insemination 381middle of seasonal anoestrus (May), a high level offertility (71–78% of kidding) was obtained using frozensemen on goats subjected to treatment with artificiallong days and progestagen followed by one AI 52 hafter introduction of the buck equipped with an apron(Pellicer-Rubio et al. 2008).Genetic Variability of Selected TraitsProduction traitsGenetic parametersGenetic parameters of milk production traits [milk, fatand protein yields (PYs), fat and PCs] have beenestimated for Alpine and Sannen goats from differentcountries (Iloeje et al. 1981; Boichard et al. 1989;Belichon et al. 1998; Muller et al. 2002) but also onother breeds (Mavrogenis et al. 1984; Analla et al.1996). The same trends have been found for theseestimates: for yields, reported values of heritability varyfrom 0.30 to 0.40. Heritability of contents is higher,from 0.50 to 0.60, but with a genetic coefficient ofvariation lower than that for yields. The correlationbetween fat and PC is approximately 0.50 and thesetraits are in opposition with milk yield, in particular PC.A strong positive correlation (approximately 0.80–0.90)is usually found between milk, fat and PYs.Casein polymorphismCompared with the sheep and cow, the types of casein ingoat milk are extremely variable, each having severalvariants, and casein polymorphisms are associated withthe rate of casein synthesis. Variants can be classifiedaccording to their synthesis rate of as1-cas (Grosclaudeand Martin 1997): A, B and C (strong alleles associatedwith high synthesis rates), E (intermediate rate), G andF (weak rates) and O (absence or only traces of as1-cas).Mahé et al. (1994) and Barbieri et al. (1995) found thatthe difference between extreme genotypes (AA and FF)for PC is approximately 4.5 g ⁄ kg which representsapproximately three genetic standard deviations, asestimated under the classical polygenic model for fielddata (r = 1.7 g ⁄ kg; Belichon et al. 1998).Allele frequencies at the as1-cas locus vary accordingto breed (Jordana et al. 1995). Variation also existswithin-breed according to the subpopulations sampled.Manfredi et al. (1998) showed, for the Alpine breedwhich is selected for PY and PC, that the frequency ofstrong alleles in the dams-of-AI bucks subpopulationwas almost twice the corresponding frequency in theunselected female subpopulation (0.46 vs 0.25). Therefore,allele frequencies at the as1-cas locus changed overtime in this population under selection pressure.Functional traitsType appraisalApart from influencing longevity by reducing thefrequency of culling for type traits, selection onmorphology, in particular udder morphology, willfacilitate the labour of the breeder by reducing milkingtime. Another advantage is the benefit to udder health,for example a reduction of clinical and subclinicalmastitis. Genetic variability has been found for morphologytraits (Luo et al. 1997; Piace` re et al. 1998;Manfredi et al. 2001; Cle´ment et al. 2006). Heritabilityestimates for traits related to the udder vary from 0.20 to0.35. Genetic correlations with dairy traits can be lowfor some traits, whereas others, in particular udder floor(UF) morphology, show a significant genetic oppositionwith milk, fat and PYs.Resistance to diseaseUdder health, namely clinical and subclinical mastitis,influences animal welfare and the quantity and quality ofmilk. Bacterial infections in goats are measured by directbacterial counting or by indirect measurements such asthe California mastitis test (CMT) and somatic cell counts(SCC). The reasons for the high SCC usually found ingoat milk are unclear. Current research is focused on theadequate interpretation of SCC (Sierra et al. 1998) andthe definition of SCC thresholds for detection of mammaryinfections in goats (Baudry et al. 1998). Geneticparameters for SCC have been estimated (Rupp et al.2004). The heritability was approximately 0.20 and thegenetic correlation with milk production close to 0.Besides mastitis, other diseases have an impact onproduction costs, quality of products and animalwelfare. The caprine arthritis-encephalitis virus (CAEV)provokes arthritis but it may also affect the mammarygland and aggravate bacterial mastitis. Susceptibility toCAEV infection might be associated to polymorphismsat the MHC (Ruff and Lazary 1988; Amills et al. 1999).Among goat diseases, scrapie also has been considered.The incubation period of this lethal spongiform encephalopathyis associated with a polymorphism of the PrPgene (Goldmann et al. 1996). As far as resistance toparasitism is concerned, genetic variability also existsand moderate heritabilities are exhibited (Baker et al.1999; Mandonnet et al. 1999).Milking abilityEfficient milking results from the interaction betweenthe milking conditions, including machine parameters(vacuum level, pulsation rate, pulsation ratio; Lu et al.1991), and the udder anatomy and physiology(Bruckmaier et al. 1994). The existence of a major geneinfluencing milking ability (volume collected after 1 minof milking) was postulated and confirmed by segregationanalyses (Ricordeau et al. 1990; Ilahi et al. 2000),indicating that the major gene explains approximatelyhalf of the total genetic variation (total heritability of 0.5and polygenic heritability of 0.3). Phenotypic associationsbetween udder characteristics and milking abilityhave been reported (Montaldo and Martinez-Lozano1993; Bruckmaier et al. 1994) but genetic correlationsare unknown. Genetic correlations between milkingspeed (volume collected after 1 min of milking) andlactation yield were found to be low (rg = 0.10) by Ilahiet al. (2000) while phenotypic correlations between dailyyield and milking speed were over 0.25 (Peris et al. 1996;Ilahi et al. 1999). Phenotypic correlations betweenÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

382 B Leboeuf, JA Delgadillo, E Manfredi, A Piace` re, V Cle´ment, P Martin, M Pellicer, P Boue´, and R de Cremouxmilking ability and mastitis, measured as CMT scores orSCC, are low according to Montaldo and Martinez-Lozano (1993) and Ilahi (1999). Genetic correlationsbetween milking ability and udder health remainunknown in goats.Contribution of AI to Genetic ImprovementProgrammes: The French Selection ProgrammeProduction from French goats, as measured by milkrecording on more than 100 000 animals in each breed,is on average 800 kg of milk for a lactation length of270 days. Fat and PC are higher in the Alpine than inthe Saanen breed: 3.7% and 3.2%, respectively, vs 3.5%and 3.1%. These results have been obtained afterapproximately 20 years of genetic improvement, with aselection scheme based on the use of AI.Determination of the selection objectivesThe objectives of the selection programme are revisedperiodically in order to adapt the genetic goals tomarket demands, and to take into account the genetictrends of the populations under selection (Piace` re et al.2002). Since 1970, the objective of the French programmewas the improvement of PY and PC withselection thresholds defined for each trait. However, asproduction traits are either negatively or positivelycorrelated, threshold selection on several traits is notefficient. Thus, in 1995, the former independent levels ofselection were replaced by a single selection thresholddetermined for a combined index ICC = PY + 0.4 PC(Piace` re et al. 1997). Selection pressure on proteinresulted in positive trends for protein and for fat content(FC) because of the positive correlation between bothtraits. However, gains on PC were higher than those forFC, thus inducing a slight negative trend in the ratioFC ⁄ PC. In 1999, the ICC was modified to stabilize theFC ⁄ PC ratio, by giving a slight positive weighting to fatyield and FC. As Alpine and Saanen goats have similarmatter yields, the production selection objective was thesame for both breeds.The same procedure was undertaken for type indicesto determine a single type objective. The type combinedindex, named IMC (Index Morphologique Caprin, goattype index) (Cle´ment et al. 2006) combines most relevanttype indices for breeders: udder profile (UP), udderfloor (UF) position, rear udder (RU) attachment andrear udder shape (RS). These four traits are responsiblefor 80% of the whole udder variability. However, typephenotypic and genetic differences between Alpine andSaanen goats were observed. So the IMC was not thesame for both breeds:Alpine IMC ¼ 1:5 UP þ UF þ RU þ RS;Saanen IMC ¼ UP þ UF þ RU þ 0:5 RSAs breeders wanted to obtain genetic progress for uddertype, as well as improving the milk production, they hadto define the respective weightings for both criteria in aglobal combined index. According to the geneticparameters of different traits, which were slightlydifferent according to the breed, they finally decided togive a weight of 44% and 33% on type in the Saanenand Alpine breeds, respectively.Technical structures and general organizationSelection programmes for dairy ruminants benefit fromthe cooperative action of farmers, milk plants, AIcentres and milk recording organizations, with thetechnical and scientific support of specialized institutions.The breeding law specifies the missions of eachorganization participating to the selection programme.There are seven organizations concerned with thebreeding of dairy goats:1. The EDE (Regional Breeding Bureaux) monitorsanimal identification. According to the Europeanrules, all French goats are officially identified with eartags and registered in a national database;2. France Controˆ le Laitier is responsible for milkrecording and providing technical advice to breederson breeding and feeding. Selection on productiontraits needs exhaustive records. In France, 340 000goats in 2200 herds are officially milk recorded andany breeder can access the milk recording services.Records are registered in a national database;3. The Centres Re´gionaux Informatiques (CRI,Regional Computing Centers) are responsible forthe regional databases and all regional data are thencentralized in a national database: Le Centre deTraitement de l’Information Ge´ne´tique (CTIG,National Computing Center for Genetics);4. Capri-IA is the national semen production centreand the national union of AI cooperatives for goats.Capri-IA produces frozen semen, which is sent to AIcooperatives responsible for on-farm inseminations;5. The Institut de l’Elevage (IE, Breeding Institute)gives technical support and coordinates data collection,validation and storage in the national database;6. The Institut National de la Recherche Agronomique(INRA) gives scientific support for the definitionof selection objectives, the genetic evaluation (indexcalculation) and the design of the selection programme;7. Caprige` ne France is the goat breeders association.It groups 800 breeders and 160 000 Alpine andSaanen goats. The Caprige` ne committee gathersrepresentatives of breeders, cheese producers, milkplants, Ministry of Agriculture and the technicalorganizations previously described. It is in charge ofthe general management of the selection programmeand the committee determines the selection objectives.Furthermore, Caprige` ne is responsible for the typerecording and the pedigree file management.Genetic evaluation of animalsThe statistical method used for genetic evaluation is aBLUP applied to an animal model. The advantage ofthis approach is that it provides unbiased breedingvalues: first, all available information is taken intoaccount and second, it allows the dissociation of geneticand environmental effects. However, to separate part ofthe variance as a result of genetic effects from those ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Management of Goat Reproduction and Insemination 383the husbandry system, the method assumes that thereare some genetic links between herds, all these linksbeing called genetic connectedness. The main source ofconnectedness is AI.Genetic evaluation of French dairy goats for productiontraits is computed for all flocks recorded for milkproduction, but breeding values are published only forconnected herds, that is, for herds which participate inthe selection programme. Methods used to evaluateconnectedness are based on the percentage of goats withAI sires, but also on other source of connectedness, inparticular, the use of goats with an AI paternal grandsire(Fouilloux et al. 2007), as the majority of naturalmatingbucks are sired by AI.For genetic evaluation of type traits, connectedness isnot a problem, because approximately two-thirds ofscored goats are daughters of AI bucks and type scoringis performed only in flocks using AI. Thus, breedingvalues are calculated and published for a subpopulationof connected herds.Creation of genetic improvementA major step in the selection programme conducted in1992 with the creation of a structured selection nucleusformed with goat breeders highly interested in selectionand genetic improvement. These breeders must inseminate30% of their goats, at least, and use 30% ofprogeny tested semen for their inseminations. In return,breeders enjoy technical advice for selection and get thelargest access to the best AI semen. In 2006, the selectionprogramme involved 160 000 goats owned by 800breeders.Planned matingThe most important step in the genetic improvementprogramme is to breed AI males of high genetic level.They will come from planned matings between the bestfemales and the best males in the population. Therefore,elite females selected to be dams of the next AI buckgeneration are inseminated with frozen semen from thebest AI males, named Sires. Dams and Sires are selectedaccording to their breeding values on production andtype traits, combined in a global index. Planned matingis determined to avoid inbreeding and to preservegenetic variability.Sanitary and individual controlEvery year, 350 males born from planned mating areinspected on farm by Capri-IA in order to select the bestfor the semen production centre. They are examined forpedigree (DNA validation), breed standard and sanitarystandard. The sanitary standard of each dam and eachherd is also tested. According to the tests’ results, 200males are selected for the quarantine of the semenproduction centre, where they are controlled on thebasis of sanitary standards and growth. At the end ofthe quarantine, only 120 males enter the semen productioncentre, where they are controlled on the basis ofsexual behaviour (libido and ability to give semen in anartificial vagina), semen production (semen quantity andquality) and sperm survival after deep freezing-thawing.After these controls, remaining 70 males are progenytested to determine their genetic level. All eliminatedmales are slaughtered.Progeny testingFor progeny testing, 200 inseminations are performedper male across milk-recorded herds. The goal is toobtain a minimum of 30 milk-recorded daughters pertested male (the average number is 80 daughters permale). Milk records and type appraisal of the daughtersare used to compute the breeding value of each male.The 35 best males are selected and the other ones areslaughtered.ConclusionThis review has mainly addressed the methods and thetechniques that have provided improvements to routinepractical AI programmes and genetic improvementprogrammes. An important consideration is the timingof the AI, which should be performed close to the timeof ovulation in the females. Accurate and carefuldetection of oestrus or control of oestrus and ovulationare necessary to reach a satisfactory fertility level.Validation of new procedures such as the use of themale effect in highly seasonal goats will be required tomeet to the new expectations of the animal industriesand consumers which are opposed to the use of syntheticchemicals and hormonal treatments. The efficiency ofthe French selection programme is now widely provedand recognized. It fits with most dairy goat breeders’objectives, but further studies are still in progress tointegrate new selection criteria and to adapt selectionobjectives to new needs and diversified market demands.The goat genetic map is being constructed, thus givingthe basis for QTL (Quantitative Trait Loci) detection.Molecular information will be particularly useful forgenetic improvement of dairy traits.ReferencesAmills M, Sanchez A, Manfredi E, 1999: Allelic frequencies ofthe Mhc class II DRB gene in caprine arthritis-encephalitisvirus infected goats. 2nd Annual Conference on new andRe-Emerging Infectious Diseases, University of Illinois,Urbana-Champaign, pp. 23.Analla M, Jime´nez-Gamero I, Mun˜ oz Serrano A, SerradillaJM, Falagan A, 1996: Estimation of genetic parameters formilk yield and fat and protein contents of milk fromMurciano-Granadina goats. J Dairy Sci 79, 1895–1898.Baker RL, Audho JO, Aduda EO, Thorpe W, 1999: GeneticResistance to Gastrointestinal Nematode Parasites in someindigenous breeds of sheep and goats in Kenya. In: FromJay L. Lush to Genomics: Visions for Animal Breeding andGenetics. University Press, Ames, p. 180.Barbieri ME, Manfredi E, Elsen JM, Ricordeau G, Bouillon J,Grosclaude F, Mahé MF, Bibe´ B, 1995: Influence du locusde la case´ine as1 sur les performances laitie` res et lesparame` tres géne´tiques des che` vres de race Alpine. GenetSel Evol 27, 437–450.Baudry C, Mercier P, Mallereau MP, Lenfant D, 1998:Utilisation des numérations cellulaires individuelles pour laÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

384 B Leboeuf, JA Delgadillo, E Manfredi, A Piace` re, V Cle´ment, P Martin, M Pellicer, P Boue´, and R de Cremouxde´tection des infections mammaires subcliniques de laChe` vre: de´finition de seuils. In: Milking and Milk Productionof Dairy Sheep and Goats, EAAP Publication No. 95,Proceedings of the 6th International Symposium on themilking of Small Ruminants, Athens, Greece, 26th September–1stOctober, pp. 119–123.Belichon S, Manfredi M, Piacère A, 1998: Genetic parametersof dairy traits in the Alpine and Saanen goat breeds. GenetSel Evol 30, 529–534.Boichard D, Bouloc N, Ricordeau G, Piacere A, Barillet F,1989: Genetic parameters for first lactation dairy traits inthe Alpine and Saanen goat breeds. Genet Sel Evol 21,201–215.Boue´ P, Sigwald JP, 2001: Statistiques et bilan ge´ne´tique del’IA en espe` ce caprine. CR No. 3309, November 2002, 34pp.Capri-IA et I.E. (Ed), 86800 Mignaloux Beauvoir, France.Bruckmaier R, Ritter C, Schams D, Blun J, 1994: Machinemilking of dairy goats during lactation: udder anatomy,milking characteristics, and blood concentrations of oxytocinand prolactin. J Dairy Res 61, 457–466.Chemineau P, Daveau A, Maurice F, Delgadillo JA, 1992:Seasonality of oestrus and ovulation is not modified bysubjecting female Alpine goats to a tropical photoperiod.Small Rumin Res 8, 299–312.Chemineau P, Pellicer-Rubio MT, Lassoued N, Khaldi G,Monniaux D, 2006: Male induced short oestrous andovarian cycles in sheep and goats: a working hypothesis.Reprod Nutr Dev 46, 417–429.Cle´ment V, Martin P, Barillet F, 2006: Elaboration d’un indexsynthe´tique caprin combinant les caracte` res laitiers et descaracte` res de morphologie mammaire. Renc Rech Rumin13, 209–212.Corteel JM, 1974: Viabilite´ des spermatozoı¨des de boucconserve´s et congelés avec ou sans leur plasma seminal:Effet du glucose. Ann Biol Anim Biochim Biophys 14, 741–745.Corteel JM, 1977: Production, storage and artificial inseminationof goat semen. In: Proceeding of the Symposium‘‘Management of Reproduction in Sheep and Goats’’,Madison, WI, USA, 24–25 July 1977, pp. 41–57.Corteel JM, Leboeuf B, Baril G, 1988: Artificial breeding ofadult goats and kids induced with hormones to ovulateoutside the breeding season. Small Rumin Res 1, 19–33.Dauzier L, 1966: Artificial insemination in the goat. In:Dalling (ed), Encyclopedia of Veterinary Medicine. W. Greenand Son, pp. 269–271.Delgadillo JA, Leboeuf B, Chemineau P, 1991: Decrease in theseasonality of sexual behaviour and sperm production inbucks by exposure to short photoperiodic cycles. Theriogenology36, 755–770.Delgadillo JA, Leboeuf B, Chemineau P, 1992: Abolition ofseasonal variations in semen quality and maintenance ofsperm fertilizing ability by short photoperiodic cycles in goatbucks. Small Rumin Res 9, 47–59.Delgadillo JA, Canedo GA, Chemineau P, Guillaume D,Malpaux B, 1999: Evidence for an annual reproductiverhythm independent of food availability in male Creolegoats in subtropical northern Mexico. Theriogenology 52,727–737.Delgadillo JA, Fitz-Gonza´lez G, Duarte G, Véliz FG, Carrillo E,Flores JA, Vielma J, Herna´ndez H, Malpaux B, 2004a:Management of photoperiod to control caprine reproductionin the subtropics. Reprod Fertil Dev 16, 471–478.Delgadillo JA, Cortez ME, Duarte G, Chemineau P, Malpaux B,2004b: Evidence that photoperiod controls annual changes intestosterone secretion and body weight in subtropical malegoats. Reprod Nutr Dev 44, 183–193.Fouilloux MN, Cle´ment V, Laloe¨ D, 2008: Measuringconnectedness among herds in mixed linear models: fromtheory to practice in large-size genetic evaluations. Acceptedfor publication in Genet Sel Evol.Gebbie FE, Forsyth IA, Arendt J, 1999: Effects of maintainingsolstice light and temperature on reproductive activity, coatgrowth, plasma prolactin and melatonin in goats. J ReprodFertil 116, 25–35.Goldmann W, Martin T, Foster J, Hughes S, Smith G,Hughes K, Dawson M, Hunter N, 1996: Novel polymorphismsin the caprine PrP gene: a codon 142 mutation associatedwith scrapie incubation period. J Gen Virol 77, 2885–2891.Grosclaude F, Martin P, 1997: Milk protein polymorphism.In: Proceedings of the International Dairy FederationSeminar, Palmerston North, New Zealand, pp. 241–253.Ilahi H1999: Variabilite´ ge´nétique du de´bit de traite chez lescaprins laitiers. Ph.D. thesis, ENSAR, Rennes, France.Ilahi H, Chastin P, Bouvier F, Arhainx J, Ricard E, ManfrediE, 1999: Milking characteristics of dairy goats. Small RuminRes 34, 97–102.Ilahi H, Manfredi E, Chastin P, Monod F, Elsen JM, Le RoyP, 2000: Genetic variability in milking speed of dairy goats.Genet Res 75, 315–319.Iloeje MU, Van Vleck LD, Wiggans GR, 1981: Components ofvariance for milk and fat yields in dairy goats. J Dairy Sci64, 2290–2293.Jordana J, Amills M, Diaz E, Angulo C, Serradilla JM,Sanchez A, 1995: Gene frequencies of caprine as1-caseinpolymorphism in Spanish goat breeds. Small Rumin Res 20,215–221.Leboeuf B, Furstoss V, Guillouet P, Boue´ P, 2004a: Productionof artificial insemination doses from Alpine and Saanenbucks under various photoperiodic cycles. S Afr J Anim Sci34, 230–232.Leboeuf B, Guillouet P, Bonne´ JL, Forgerit Y, Magistrini M,2004b: Goat semen preserved at 4°C until 76 hours beforeartificial insemination: different attempts to maintain thefertility. S Afr J Anim Sci 34, 233–235.Lu CD, Potchoiba MJ, Loetz ER, 1991: Influence ofvacuum level, pulsation ratio and rate on milking performanceand udder health in dairy goats. Small Rumin Res5, 1–8.Luo MF, Wiggans GR, Hubbard SM, 1997: Variance componentestimation and multitrait genetic evaluation for typetraits of dairy goats. J Dairy Sci 80, 594–600.Mahe´ MF, Manfredi E, Ricordeau G, Piacere A, Grosclaude F,1994: Effets du polymorphisme de la case´ine as1 sur lesperformances laitie` res: analyse intra-descendance de boucsde race Alpine. Genet Sel Evol 26, 151–157.Mandonnet N, Aumont G, Gruner L, Bouix J, Vu Tien Khang J,Arquet R, Varo H, 1999: Direct and maternal genetic effectsfor resistance to strongyles in Creole goats. In: 17th InternationalConference of the WAAP, Copenhagen, 15–19 August1999. Abst.Manfredi E, Leroux C, Piace` re A, Martin P, Elsen JM,Grosclaude F, 1998: Use of a major gene in a dairy goatselection scheme. In: 49th Annual Meeting of the EAAP,Warsaw, Poland, 24–27 August, pp. 1–4.Manfredi E, Piace` re A, Lahaye P, Ducrocq V, 2001: Geneticparameters of type appraisal in Saanen and Alpine goats.Livest Prod Sci 70, 183–189.Martin GB, Milton JTB, Davison RH, Hunzicker GEB,Lindsay DR, Blache D, 2004: Natural methods for increasingreproductive efficiency in small ruminants. Anim ReprodSci 82–83, 231–246.Mavrogenis AP, Constantinuou A, Louca A, 1984: Environmentaland genetic causes of variation in production traitsof Damascus goats. II. Goat productivity. Anim Prod 38,99–104.Montaldo H, Martinez-Lozano FJ, 1993: Phenotypic relationshipsbetween udder and milking characteristics, milkÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Management of Goat Reproduction and Insemination 385production and California mastitis test in goats. SmallRumin Res 12, 329–337.Muller CJC, Cloet SWP, Schoeman SJ, 2002: Estimation ofgenetic parameters for milk yield and milk composition ofSouth African Saanen goats. In: Proceedings of the 7thWorld Congress on Genetics Applied to Livestock Production,Montpellier, France, pp. 259–262.Pellicer-Rubio M, Combarnous Y, 1998: Deterioration of goatspermatozoa in skimmed milk-based extenders as a result ofoleic acid released by the bulbourethral lipase BUSgp60. JReprod Fertil 112, 95–105.Pellicer-Rubio MT, Leboeuf B, Bernelas D, Forgerit Y,Pougnard JL, Bonne´ JL, Senty E, Chemineau P, 2007:Highly synchronous and fertile reproductive activityinduced by the male effect during deep anoestrus in lactatinggoats subjected to treatment with artificially long daysfollowed by a natural photoperiod. Anim Reprod Sci 98,241–258.Pellicer-Rubio MT, Leboeuf B, Bernelas D, Forgerit Y,Pougnard JL, Bonne´ JL, Senty E, Breton S, Brun F,Chemineau P, 2008: High fertility using artificial inseminationduring deep anoestrus after induction and synchronisationof ovulatory activity by the ‘‘male effect’’ in lactatinggoats subjected to treatment with artificial long days andprogestagens. Accepted for publication in Anim Reprod Sci.Peris S, Such X, Caja G, 1996: Milkability of Murcinao-Granadina dairy goats: milk partitioning and flow rateduring milking according to parity, prolificacy and mode ofsuckling. J Dairy Res 63, 1–9.Piace` re A, Bouloc-Duval N, Sigwald JP, Larzul C, Manfredi E,1997: Utilisation de l’index combine´ caprin et du polymorphismede la case´ine alpha s1 dans le sche´ma de sélectioncaprin. Renc Rech Rumin 4, 187–190.Piace` re A, Manfredi E, Lahaye P, 1998: Analyse ge´nétique dela morphologie des che` vres Saanen et Alpine françaises. In:Proceedings of the 6th International Symposium on themilking of Small Ruminants, Athens, Greece, 26th September–1stOctober, pp. 178–180.Ricordeau G, Bouillon J, Le Roy P, Elsen JM, 1990:De´terminisme géne´tique du de´bit de traite au cours de latraite des che` vres. INRA Prod Anim 3, 121–126.Ruff G, Lazary S, 1988: Evidence for linkage between theCaprine leucocyte antigen (CLA) system and susceptibilityto CAE virus-induced arthritis in goats. Immunogenetics 28,303–309.Rupp R, Cle´ment V, Piace` re A, Manfredi E, 2004: Goat milksomatic cell count is a heritable trait. In: 55th AnnualMeeting of the EAAP, Bled, Slovenia, 3–8 September 2004.Sias B, Ferrato F, Pellicer-Rubio M, Forgerit Y, Guillouet P,Leboeuf B, Carrie` re F, 2005: Cloning and seasonal secretionof the pancreatic lipase-related protein2 present in goatseminal plasma, 2004. Biochim Biophys Acta 1686, 169–180.Sierra D, Sanchez A, Corrales JC, Contreras A, 1998: In:Proceedings of the 6th International Symposium on themilking of Small Ruminants, Athens, Greece, 26th September–1stOctober, pp. 178–180.Author’s address (for correspondence): B Leboeuf, INRA-UEICP,Venours, 86480 Rouillé, France. E-mail: leboeuf@lusignan.inra.frConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 386–392 (2008); doi: 10.1111/j.1439-0531.2008.01189.xISSN 0936-6768Regulation of the Spermatogonial Stem Cell NicheN Kostereva 1 and M-C Hofmann 1,21 Department of Veterinary Biosciences; 2 Institute for Genomic Biology, University of Illinois at Urbana-Champaign, Urbana, IL, USAContentsSpermatogonial stem cells (SSCs) reside within specializedmicroenvironments called ‘niches’, which are essential for theirmaintenance and self-renewal. In the mammalian testis, themain components of the niche include the Sertoli cell, thegrowth factors that this nursing cell produces, the basementmembrane, and stimuli from the vascular network between theseminiferous tubules. This review focuses on signalling pathwaysmaintaining SSCs self-renewal and differentiation anddescribes potential mechanisms of regulation of the spermatogonialstem cell niche.Stem Cells and their NicheStem cells are required for the growth, maintenanceand repair of many tissues. They are characterized bytheir abilities to self-renew and to produce numerousdifferentiated daughter cells, enabling them to play acentral role in tissue homeostasis. In order for theirmaintenance and self-renewal to be ensured, theyreceive essential signals from their microenvironment,which is called the stem cell ‘niche’. The concept ofstem cell niche was first suggested by Schofield in 1978,to describe a microenvironment that supports stemcells in the mammalian hematopoietic system (Schofield1978). Other stem cell niches have now beenidentified in most tissues of model organisms, includingthe intestine, skin, brain and testis. The niche regulatesspecific properties of the stem cell, including selfrenewal,pluripotency, quiescence and the ability todifferentiate into single or multiple lineages (Adamsand Scadden 2006). To this effect, the niche can bedefined as a complex interplay of short- and long-rangestimuli between the stem cells, their differentiatingdaughters, neighbouring cells, and the extracellularmatrix, collectively making up a microenvironmentthat controls stem cell behaviour. Ultimately, thisbehaviour will depend on cellular intrinsic factors thatare modulated by these signals (Watt and Hogan2000).The Spermatogonial Stem Cell NicheIn the mammalian testis, the stem cells that are at theorigin of spermatogenesis are called spermatogonialstem cells (SSCs). Spermatogonial stem cells reside inthe basal part of the seminiferous epithelium. These cellsare the only stem cells in the body that undergo selfrenewalthroughout life and transmit genetic informationto the offspring (De Rooij and Russell 2000).Spermatogonial stem cells are morphologically undifferentiatedsingle cells that are not connected byintercellular bridges like the more advanced germ cells(Dym and Fawcett 1971; Huckins 1971; Oakberg 1971).They reside on the basement membrane and are also inintimate contact with the Sertoli cells, the only somaticcells present within the seminiferous epithelium. Spermatogonialstem cells self-renew in order to maintainspermatogenesis throughout the life of adult maleanimals. These cells also differentiate to produceApaired, Aaligned, A1-A4 spermatogonia and type Bspermatogonia, through a series of steps that willamplify the number of germ cells. Finally, type Bspermatogonia will differentiate into spermatocytes thatwill translocate to the inner layer of the seminiferousepithelium, undergo meiosis and give rise to haploidspermatids that will differentiate into spermatozoa.Although some advances have recently been made, themolecular mechanisms underlying the maintenance andself-renewal of SSCs only begin to be elucidated. Inaddition, the signal that mediates the decision of SSCsto differentiate rather than self-renew is still unknown.This slow progress is due to the fact that these cells existin low numbers (0.03% of the total number of germcells in an adult testis; Meistrich and Van Beek 1993)and that no specific membrane marker is available toisolate them unequivocally. However, enrichment techniquesare now available, that allow significant advancesin our understanding of the behaviour of SSCsin vitro (Kubota et al. 2003; Buageaw et al. 2005;Hofmann et al. 2005b; Braydich-Stolle et al. 2007). Inaddition, the transplantation technique established bythe group of R. Brinster a decade ago providedresearchers with an in vivo functional assay to assessself-renewal and germline transmission after geneticmanipulations of these cells (Nagano et al. 2001;Brinster 2002).In the mammalian testis, the somatic Sertoli cell, thebasement membrane and the cellular components of theinterstitial space between the seminiferous tubules(Shetty and Meistrich 2007) are crucial components ofthe niche. The Sertoli cell provides growth factorsnecessary for self-renewal such as glial cell line-derivedneurotrophic factor (GDNF) and basic fibroblastgrowth factor (bFGF) (Meng et al. 2000; Kubota et al.2004; Hofmann et al. 2005b), the basement membraneand integrins provide for anchorage (Shinohara et al.1999), and stimuli from the vascular network andinterstitial cells are crucial for the localization ofundifferentiated spermatogonia along specific portionsof the basement membrane (Yoshida et al. 2007). Theintegration of these signals provides the cues necessaryfor self-renewal and retention of the SSCs in theirundifferentiated state. These extrinsic signals will modulateSSC intrinsic signals such as kinases, secondmessengers and transcription factors to ensure homeostasis.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Regulation of the Spermatogonial Stem Cell Niche 387Cues for Self-RenewalOne factor that is essential for SSC self-renewal isGDNF. It is a protein secreted by Sertoli cells afterbirth, and is specifically responsible for the maintenanceand proliferation of SSCs in vivo and in vitro (Menget al. 2000; Kubota et al. 2004; Hofmann et al. 2005b).Glial cell line-derived neurotrophic factor belongs to thetransforming growth factor-beta superfamily, and signalsthrough a multi-component receptor formed by theRet tyrosine kinase transmembrane receptor and its coreceptorglial cell line-derived neurotrophic factor familyreceptor alpha-1 (GFRa-1). The binding of GDNFtriggers the activation of multiple signalling pathways inresponsive cells (Airaksinen and Saarma 2002).Mice lacking GDNF or its receptors die within thefirst day of birth with renal and neuronal abnormalities(Schuchardt et al. 1994; Moore et al. 1996; Pichel et al.1996; Sanchez et al. 1996). The problem of neonatalmortality can be overcome by grafting GDNF, GFRa-1or Ret deficient neonatal testes under the back skin ofimmunodeficient mice, and subsequently monitoring thedevelopment of the grafted testes (Naughton et al.2006). This strategy revealed that any disruption ofGDNF-mediated Ret signalling results in a lack of SSCself-renewal and induces the progressive loss of spermatogenesisby germ cell depletion. In comparison,normal spermatogenesis and maintenance of SSC populationswas observed in the grafted wild type (WT)testes. Thus, GDNF, Ret and GFRa-1 are all crucial forSSC maintenance, emphasizing the essential role of theGDNF ⁄ Ret ⁄ GFRa1 signalling pathway in SSCs.Spermatogonial stem cells have been difficult toisolate because of their low numbers and the lack ofsurface markers specific enough for immunosorting.Some laboratories, including ours, have isolated SSCsusing antibodies against the GFRa-1 receptor (Buageawet al. 2005; Hofmann et al. 2005b). Although expressedby both the stem cell and its direct progeny, the Apairedspermatogonia, GFRa-1 is an adequate marker forpurifying SSCs from testes using antibody selection.Using gravity sedimentation on a 2–4% bovine serumalbumin (BSA) gradient followed by magnetic beadsisolation with a GFRa-1 antibody, we obtained a cellpurity of up to 98%. However, while the purity is high,the number of cells recovered is low. Other investigatorshave been successful at enriching SSCs using antibodiesto thymus cell antigen 1 (Thy-1) and fluorescenceactivatedcell sorting, producing an enrichment that wassufficient for most of their studies (Kubota et al. 2003).Glial cell line-derived neurotrophic factor, together withother growth factors, has recently been used to expandthe number of SSCs in long-term cultures provided thatthey are grown onto feeder layers (Kanatsu-Shinoharaet al. 2003; Ogawa et al. 2003; Kubota et al. 2004).We recently elucidated some of the pathways inducedby GDNF in SSCs by using serum-free short-termcultures of SSCs and a spermatogonial stem cell line thatwe established (Hofmann et al. 2005a). After stimulationby GDNF, we observed that several kinases fromthe Src family co-precipitate with the Ret receptorin SSCs (Braydich-Stolle et al. 2007). Further, wedemonstrated that Src activates a PI3K ⁄ Akt signallingFig. 1. Glial cell line-derived neurotrophic factor can signal throughactivation of Src kinases in SSCs. The GDNF can promote cell cycleprogression via activation of Src family kinases (SFKs) such as Src,Yes, Fyn and Lyn. The SFKs further activate a PI3K ⁄ Akt pathway,which ultimately leads to an increase in N-myc gene expression (fromBraydich-Stolle et al. 2007)pathway, which ultimately leads to N-Myc expressionand promotes SSC proliferation (Fig. 1). Subsequently,the groups of T. Shinohara and R. Brinster establishedthe in vivo relevance of this signalling axis for SSC selfrenewalby using germ cell transplantations after downregulatingAkt and Src expression by RNA interferenceor pharmacological inhibitors (Lee et al. 2007; Oatleyet al. 2007).Ret activation by the binding of GDNF also inducesthe anchoring of the protein adaptors Shc and Grb2 andthe activation of Ras in SSCs (He et al. 2008). Ras thenactivates ERK1 ⁄ 2, which ultimately leads to the phosphorylationand activation of transcription factors suchas CREB-1, ATF-1 and CREM-1. Finally, theGDNF ⁄ Ret ⁄ Ras axis up-regulates the transcription ofthe immediate-early gene c-fos, the cell cycle activatorcyclin A, as well as CDK2 (Fig. 2). This scenario issimilar to what is observed in other cell types, whereCREB and c-fos enhance the expression of cyclin A,favouring the G1 ⁄ S cell cycle transition (Desdouetset al. 1995). Therefore, our studies suggest that GDNFinduces SSC self-renewal through multiple signallingpathways.Other transcription factors have been recently identified,that are crucial for SSC maintenance orself-renewal. These include B cell CLL ⁄ lymphoma 6,member b (Bcl6b) (Oatley et al. 2006), TATA boxbinding protein (TBP)-associated factor 4b (Taf4b)(Falender et al. 2005) and promyelocytic leukaemiaÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

388 N Kostereva and M-C HofmannFig. 2. Glial cell line-derived neurotrophic factor can signal through aRas-induced pathway in SSCs. The GDNF can promote cell cycleprogression via activation of the Ras ⁄ ERK1 ⁄ 2 pathway. Downstreamevents are also depicted. ‘P’ indicates ‘phosphorylate’, and ‘A’ denotes‘activate’ (from He et al. 2008)zinc-finger (Plzf) (Buaas et al. 2004; Costoya et al.2004). The exact function of these molecules is notelucidated yet. However, there is evidence that Bcl6b is atarget of GDNF and plays a role in stem cell maintenance,as mice with a targeted disruption of Bcl6b havean increased incidence of Sertoli cell-only seminiferoustubules (Oatley et al. 2006). In addition, it seems clearthat Plzf in WT animals represses SSC differentiation,while its loss in mutant or knockout mice shifts thebalance towards differentiation at the cost of selfrenewal.In addition, Plzf seems to directly repress thetranscription of the receptor c-kit, which is importantfor spermatogonial differentiation (Filipponi et al.2007), and undifferentiated spermatogonia isolated fromPlzf - ⁄ - mice exhibit a marked increase in c-kit expression.Therefore, Plzf might maintain the pool of SSCsthrough direct repression of c-kit expression.Regulation of DifferentiationWhile the exact determinant of the self-renewal ⁄ differentiationswitch has not been elucidated yet, progresshas also been made in our understanding of the controland maintenance of SSCs differentiation. As mentionedabove, differentiating spermatogonia express the c-kitreceptor, and Kit ligand, produced by Sertoli cells,stimulates their DNA synthesis (Dolci et al. 2001). Inaddition, a recent study demonstrated that Kit ligandup-regulates the expression of early meiotic genes inthese cells (Rossi et al. 2008), emphasizing the importanceof this growth factor for germ cell differentiation.In 2001, two groups of researchers described theexpression of Notch receptors and their ligands, Jagged-1and Jagged-2, in the mammalian testis (Diramiet al. 2001; Hayashi et al. 2001). In mammals, the Notchsignalling pathway is involved in cell fate decisionsduring various cellular and developmental processesFig. 3. The Notch signalling pathway. After the binding of Jagged 1 ⁄ 2to the Notch receptor, the intracellular portion of the Notch receptor iscleaved and becomes Notch intracellular domain (NICD). The NICDis a transcription factor that translocates into the nucleus to activatethe expression of target genes (adapted from E. Lai, Memorial Sloan-Kettering Cancer Center, Sloan-Kettering Institute, New York, NY,USA)(Weinmaster 1997). Notch is a transmembrane receptorof approximately 300 kDa in size. Upon binding to itsligand, Notch is proteolytically processed to generate a180 kDa fragment containing most of the extracellulardomain and a 120 kDa fragment containing the transmembraneand cytoplasmic domains. This latter fragmentis in turn cleaved, releasing a shorter 80 kDAfragment, called Notch intracytoplamic domain(NICD), which migrates into the cell nucleus andfunctions as a transcription factor (Fig. 3). In thenucleus, activated Notch (NICD) interacts with othertranscription factors to regulate cell fate decision.Because both ligand and receptor are localized in theplasma membrane of the effector and the target cell,respectively, a close cell–cell contact is necessary for theactivation of this pathway. In mammals, four Notchgenes (Notch1–4) have been isolated (Weinmaster et al.1991; Lardelli et al. 1994; Uyttendaele et al. 1996). AllNotch receptors show complementary and combinatorialexpression patterns during the development ofvarious tissues, including the testis (Williams et al.1995; Mori et al. 2003). Immunocytochemistry revealedthat the Notch family proteins are activated in specificgerm cell types in the seminiferous tubules during germcell development, and that their ligands Jagged-1 andJagged-2 are expressed by Sertoli cells (Dirami et al.2001). We observed that the expression of the Notch-1intracytoplasmic domain (N1-ICD) starts before birth ingonocytes, increases as the germ cells proliferate andÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Regulation of the Spermatogonial Stem Cell Niche 389(a)(b)Fig. 4. Influence of GDNF andJagged-1 on SSCs in vitro. (a)Example of a colony of SSCsgrown in vitro (from Braydich-Stolle et al. 2007). (b) Example of achain of differentiating spermatogoniagrown in vitro (from Hofmannet al. 2006). (c) Number ofcolonies and chains of spermatogoniaobtained after culturingSSCs with 100 ng ⁄ ml GDNF for 3and 6 days. (d) Number of coloniesand chains of spermatogonia obtainedafter culturing SSCs with10 lg ⁄ ml Jagged-1 for 3 and6 days. A significant increase in thenumber of chains can be observed,indicating that Jagged-1 mediatesSSC differentiation (*p < 0.05,Student’s t-test)(c)(d)differentiate into type A and B spermatogonia, andpeaks in spermatocytes indicating that activated Notch-1 might function to maintain the proliferation of germcells as they differentiate. Further, addition of Jagged-1in the culture media of freshly isolated SSCs increasedtheir differentiation, as the number of Aaligned cellssignificantly increased in these cultures in comparisonwith the controls (Fig. 4). Suppression of Notch-1signalling by using antibodies against Notch-1 orJagged-2 led to maturation arrest in the rat testis(Hayashi et al. 2001). To evaluate a possible relationshipbetween an alteration of Notch signalling and thepathogenesis of maturation arrest, Hayashi and colleaguesexamined the expression of Notch-1 and itsligand Jagged-2 in the testis of non-obstructive infertilepatients (Hayashi et al. 2004). Patients showed maturationarrest at the spermatogonia, spermatocytes orspermatids stage. Patients with only spermatogonia intheir seminiferous tubules lacked expression of bothNotch-1 and Jagged-2 surface proteins. Patients exhibitinggerm cells that reached the spermatocyte andspermatid stage lacked either Notch-1 or Jagged-2.None of the patients affected with maturation arrestexpressed both Notch-1 and Jagged-2. These resultsindicate that Notch-1 is specifically involved in germ celldifferentiation in humans, and that some compensationmechanisms by other Notch family members or Jagged-1 can occur.An interesting possibility is that factors induced byGDNF in SSCs would down-regulate Notch-1, and thusfavour self-renewal at the expense of differentiation. Forexample, microarray analysis of GDNF-regulated genesindicated that Numb, an antagonist of Notch signalling,is up-regulated (Braydich-Stolle et al. 2005; Hofmannet al. 2005b). In vitro, we observed that Numb promotesthe degradation of Notch-1 at the protein level in germlinestem cells. Immunocytochemistry of Numb expressionin sections of seminiferous tubules parallels theexpression of Notch; it is found in type A spermatogoniaand peaks in spermatocytes, indicating that Notch-1is tightly regulated during the first step of spermatogenesis(Corallini et al. 2006).Regulation of the Spermatogonial Stem CellNicheRecent studies have demonstrated that GDNF productionby Sertoli cells is regulated by FGF2 (Simon et al.2007). However, FGF2 knockout mice are viable andfertile, indicating that other factors are necessary. Thesefactors include interleukin-1 beta (IL-1b) and tumournecrosis factor alpha (TNFa), which are producedessentially by monocytes, macrophages and dendriticcells (Simon et al. 2007). It is known that macrophagesconstitute approximately 20% of the interstitial tissue ofthe testis, where they are a major source of cytokines(Kern et al. 1995). In addition, GDNF production bySertoli cells is also regulated by FSH (Tadokoro et al.2002). Therefore, these data confirm that the stem cellniche is regulated at least in part by the vasculature andcellular components of the interstitium between theseminiferous tubules. The regulation of Jagged-1 ⁄ 2inSertoli cells is not well understood, but studies havesuggested that EGF might up-regulate Jagged-2 throughinteraction with the Sonic Hedgehog ⁄ Gli signallingpathway (Kasper et al. 2006).Another molecule essential for the maintenance of theSSC niche and spermatogenesis is the Ets-related moleculeEtv5 (Chen et al. 2005; Hess et al. 2006). The Etv5is a transcription factor expressed by Sertoli cells andgerm cells. While Etv5 expression is under control ofGDNF in germ cells, it is driven by FGF2 and EGF inSertoli cells (Oatley et al. 2006; Simon et al. 2007). TheEtv5 binds to other transcription factors to up-regulatethe expression of target genes (Gutierrez-Hartmannet al. 2007), and it is likely that its binding partners andfunction in Sertoli cells and germ cells are different. Inmice with a targeted deletion of etv5 (etv5- ⁄ -), SSCs failto renew, and these cells are lost through differentiation.Microarray analysis revealed that Etv-5 expression wasessential for the production of several chemokines,including Stromal cell-derived factor (SDF-1, CXCL-12), CXCL5 and CCL7 (Chen et al. 2005). In addition,Etv5 seems important for the expression of the matrixmetalloproteinase-12 (MMP-12). As chemokines andÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

390 N Kostereva and M-C HofmannMMPs are well known for their involvement in stem cellrecruitment, migration and homing (Heissig et al. 2002;Von Luttichau et al. 2005), it appears that in Sertolicells, Etv5 is responsible for the production of factorsthat retain stem cells inside their niches. Additionalstudies also suggest that the blood–testis barrier (Sertoli–Sertolitight junctional complex) is abnormal in theEtv5(- ⁄ -) mice, and several chemokines and MMPs havebeen shown to be critical in modulating various componentsthat contribute to blood-testes barrier function(Morrow et al. 2008).Alterations of the spermatogonial stem cell niche willoccur with aging, and these changes contribute todeficient stem cell number and activity. This wasrecently demonstrated by transplanting SSCs fromyoung, fertile male mice into the 1-year and 2-yearatrophied testes of old males (Zhang et al. 2006). Theresults showed that 1-year-old testes are permissive forregeneration of spermatogenesis, while the 2-year-oldtestes are not, suggesting a gradual change in the SSCmicroenvironment. Further studies demonstrated asignificant decrease of the production of GDNF bySertoli cells with age, which could in part explain thedecline of stem cell numbers (Ryu et al. 2006). However,in reciprocal transplantation experiments, colonizationof young testes by 2-year-old SSCs was not as efficient ascolonization by 1-year-old SSCs, indicating that stemcell intrinsic factors are also altered as the animal ages(Zhang et al. 2006). Age-related decline of the spermatogonialstem cell niche occurs as well in theDrosophila testis, where the decrease of expression of akey self-renewal signal, Unpaired (Upd), correlates witha decrease in SSCs with aging (Boyle et al. 2007).Conversely, forced expression of Upd by the somaticcells of the niche maintains SSCs in older males.Therefore, similar molecular mechanisms within thetesticular niche maintain self-renewal across species, andtheir alterations contribute to a decline of the stem cellpool and spermatogenesis.ConclusionSpermatogonial stem cells are at the origin of spermatogenesis.Their fate is regulated by a complex interplay ofgrowth factors produced by Sertoli cells and other germcells, the extracellular matrix and vasculature components.The ensemble of these factors is called thespermatogonial stem cell niche. The behaviour of theSSCs is the result of interactions between the signalsgiven from the niche and stem cell intrinsic factors suchas kinases, phosphatases and transcription factors.Growth factors provided by the niche that regulate thefate of SSCs include GDNF, stem cell factor (SCF) andJagged-1 ⁄ 2. There is little known about the mechanismsthat regulate the production of these factors by Sertolicells, but FGF2, FSH and cytokines produced byinterstitial cells seem involved. Another molecule thatregulates the function of the niche is the Sertoli celltranscription factor Etv5. By maintaining the blood–testis barrier, and by regulating the production ofchemokines, and metalloproteinases by Sertoli cells,Etv5 contributes to the homing and physical maintenanceof SSCs within the niche.AcknowledgementsFunded by NIH grants R01-HD044543 and K02-HD054607.ReferencesAdams G, Scadden D, 2006: The hematopoietic stem cell in itsplace. Nat Immunol 7, 333–337.Airaksinen M, Saarma M, 2002: The GDNF family: signalling,biological functions and therapeutic value. Nat RevNeurosci 3, 383–394.Boyle M, Wong C, Rocha M, Jones D, 2007: Decline in selfrenewalfactors contributes to aging of the stem cell niche inthe Drosophila testis. Cell Stem Cell 1, 470–478.Braydich-Stolle L, Nolan C, Dym M, Hofmann M, 2005: Roleof glial cell line-derived neurotrophic factor in germ-linestem cell fate. Ann N Y Acad Sci 1061, 94–99.Braydich-Stolle L, Kostereva N, Dym M, Hofmann M, 2007:Role of Src family kinases and N-Myc in spermatogonialstem cell proliferation. Dev Biol 304, 34–45.Brinster R, 2002: Germline stem cell transplantation andtransgenesis. Science 296, 2174–2176.Buaas F, Kirsh A, Sharma M, Mclean DJ, Morris J, GriswoldM, De Rooij DG, Braun R, 2004: Plzf is required in adultmale germ cells for stem cell self-renewal. Nat Genet 36,647–652.Buageaw A, Sukhwani M, Ben-Yehudah A, Ehmcke J, RaweV, Pholpramool C, Orwig K, Schlatt S, 2005: GDNFfamily receptor alpha1 phenotype of spermatogonial stemcells in immature mouse testes. Biol Reprod 73, 1011–1016.Chen C, Ouyang W, Grigura V, Zhou Q, Carnes K, LimH, Zhao G, Arber S, Kurpios N, Murphy TL, Cheng A,Hassell J, Chandrashekar V, Hofmann MC, Hess R, MurphyKM, 2005: ERM is required for transcriptionalcontrol of the spermatogonial stem cell niche. Nature436, 1030–1034.Corallini S, Fera S, Grisanti L, Falciatori I, Muciaccia B,Stefanini M, Vicini E, 2006: Expression of the adaptorprotein m-Numb in mouse male germ cells. Reproduction132, 887–897.Costoya JA, Hobbs R, Barna M, Cattoretti G, Manova K,Sukhwani M, Orwig K, Wolgemuth D, Pandolfi P, 2004:Essential role of Plzf in maintenance of spermatogonial stemcells. Nat Genet 36, 653–659.De Rooij DG, Russell L, 2000: All you wanted to knowabout spermatogonia but were afraid to ask. J Androl 21,776–798.Desdouets C, Matesic G, Molina CA, Foulkes N, Sassone-Corsi P, Brechot C, Sobczak-Thepot J, 1995: Cell cycleregulation of cyclin A gene expression by the cyclic AMPresponsivetranscription factors CREB and CREM. MolCell Biol 15, 3301–3309.Dirami G, Ravindranath N, Achi M, Dym M, 2001:Expression of Notch pathway components in spermatogoniaand Sertoli cells of neonatal mice. J Androl 22, 944–952.Dolci S, Pellegrini M, Di AS, Geremia R, Rossi P, 2001:Signaling through extracellular signal-regulated kinase isrequired for spermatogonial proliferative response to stemcell factor. J Biol Chem 276, 40225–40233.Dym M, Fawcett D, 1971: Further observations on thenumbers of spermatogonia, spermatocytes, and spermatidsconnected by intercellular bridges in the mammalian testis.Biol Reprod 4, 195–215.Falender A, Freiman R, Geles K, Lo K, Hwang K, Lamb D,Morris PL, Tjian R, Richards J, 2005: Maintenance ofspermatogenesis requires TAF4b, a gonad-specific subunitof TFIID. Genes Dev 19, 794–803.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Regulation of the Spermatogonial Stem Cell Niche 391Filipponi D, Hobbs R, Ottolenghi S, Rossi P, Jannini E,Pandolfi P, Dolci S, 2007: Repression of kit expression byPlzf in germ cells. Mol Cell Biol 27, 6770–6781.Gutierrez-Hartmann A, Duval D, Bradford A, 2007: ETStranscription factors in endocrine systems. Trends EndocrinolMetab 18, 150–158.Hayashi T, Kageyama Y, Ishizaka K, Xia G, Kihara K,Oshima H, 2001: Requirement of Notch 1 and its ligandjagged 2 expressions for spermatogenesis in rat and humantestes. J Androl 22, 999–1011.Hayashi T, Yamada T, Kageyama Y, Negishi T, Kihara K,2004: Expression failure of the Notch signaling system isassociated with the pathogenesis of maturation arrest inmale infertility patients. Fertil Steril 81, 697–699.He Z, Jiang J, Kokkinaki M, Golestaneh N, Hofmann M,Dym M, 2008: GDNF up-regulates c-fos transcription viathe Ras ⁄ ERK1 ⁄ 2 pathway to promote mouse spermatogonialstem cell proliferation. Stem Cells 26, 266–278.Heissig B, Hattori K, Dias S, Friedrich M, Ferris B, HackettN, Crystal RG, Besmer P, Lyden D, Moore M, Werb Z,Rafii S, 2002: Recruitment of stem and progenitor cells fromthe bone marrow niche requires MMP-9 mediated release ofkit-ligand. Cell 109, 625–637.Hess R, Cooke P, Hofmann M, Murphy K, 2006: Mechanisticinsights into the regulation of the spermatogonial stem cellniche. Cell Cycle 5, 1164–1170.Hofmann M, Braydich-Stolle L, Dettin L, Johnson E, Dym M,2005a: Immortalization of mouse germ line stem cells. StemCells 23, 200–210.Hofmann M, Braydich-Stolle L, Dym M, 2005b: Isolation ofmale germ-line stem cells; influence of GDNF. Dev Biol 279,114–124.Huckins C, 1971: The spermatogonial stem cell population inadult rats. I. Their morphology, proliferation and maturation.Anat Rec 169, 533–557.Kanatsu-Shinohara M, Ogonuki N, Inoue K, Miki H, OguraA, Toyokuni S, Shinohara T, 2003: Long-term proliferationin culture and germline transmission of mouse male germlinestem cells. Biol Reprod 69, 612–616.Kasper M, Schnidar H, Neill G, Hanneder M, Klingler S, BlaasL, Schmid C, Hauser-Kronberger C, Regl G, Philpott M,Aberger F, 2006: Selective modulation of Hedgehog ⁄ GLItarget gene expression by epidermal growth factor signalingin human keratinocytes. Mol Cell Biol 26, 6283–6298.Kern S, Robertson S, Mau V, Maddocks S, 1995: Cytokinesecretion by macrophages in the rat testis. Biol Reprod 53,1407–1416.Kubota H, Avarbock M, Brinster R, 2003: Spermatogonialstem cells share some, but not all, phenotypic and functionalcharacteristics with other stem cells. Proc Natl Acad SciUSA 100, 6487–6492.Kubota H, Avarbock M, Brinster R, 2004: Growth factorsessential for self-renewal and expansion of mouse spermatogonialstem cells. Proc Natl Acad Sci USA 101, 16489–16494.Lardelli M, Dahlstrand J, Lendahl U, 1994: The novel Notchhomologue mouse Notch 3 lacks specific epidermal growthfactor-repeats and is expressed in proliferating neuroepithelium.Mech Dev 46, 123–136.Lee J, Kanatsu-Shinohara M, Inoue K, Ogonuki N, Miki H,Toyokuni S, Kimura T, Nakano T, Ogura A, Shinohara T,2007: Akt mediates self-renewal division of mouse spermatogonialstem cells. Development 134, 1853–1859.Meistrich M, Van Beek M, 1993: Spermatogonial stem cells.In: Desjardins C, Ewing LL (eds), Spermatogonial StemCells. Oxford University Press, New York, pp. 266–295.Meng X, Lindahl M, Hyvonen M, Parvinen M, De Rooij DG,Hess M, Raatikainen-Ahokas A, Sainio K, Rauvala H,Lakso M, Pichel JG, Westphal H, Saarma M, Sariola H,2000: Regulation of cell fate decision of undifferentiatedspermatogonia by GDNF. Science 287, 1489–1493.Moore M, Klein R, Farinas I, Sauer H, Armanini M, PhillipsH, Reichardt L, Ryan A, Carver-Moore K, Rosenthal A,1996: Renal and neuronal abnormalities in mice lackingGDNF. Nature 382, 76–79.Mori S, Kadokawa Y, Hoshinaga K, Marunouchi T, 2003:Sequential activation of Notch family receptors duringmouse spermatogenesis. Dev Growth Differ 45, 7–13.Morrow C, Hostetler C, Griswold M, Hofmann MC, MurphyK, Cooke P, Hess R, 2008: ETV5 is required for continuousspermatogenesis in adult mice and may mediate blood testesbarrier function and testicular immune privilege. Ann N YAcad Sci 1120, 144–151.Nagano M, Brinster C, Orwig K, Ryu B, Avarbock M,Brinster R, 2001: Transgenic mice produced by retroviraltransduction of male germ-line stem cells. Proc Natl AcadSci USA 98, 13090–13095.Naughton C, Jain S, Strickland A, Gupta A, Milbrandt J,2006: Glial cell-line derived neurotrophic factor (GDNF)-mediated RET signaling regulates spermatogonial stem cellfate. Biol Reprod 74, 314–321.Oakberg E, 1971: Spermatogonial stem-cell renewal in themouse. Anat Rec 169, 515–531.Oatley JM, Avarbock M, Telaranta A, Fearon DT, Brinster R,2006: Identifying genes important for spermatogonial stemcell self-renewal and survival. Proc Natl Acad Sci USA 103,9524–9529.Oatley JM, Avarbock M, Brinster R, 2007: Glial cell linederivedneurotrophic factor regulation of genes essential forself-renewal of mouse spermatogonial stem cells is dependenton Src family kinase signaling. J Biol Chem 282, 25842–25851.Ogawa T, Ohmura M, Yumura Y, Sawada H, Kubota Y,2003: Expansion of murine spermatogonial stem cellsthrough serial transplantation. Biol Reprod 68, 316–322.Pichel J, Shen L, Sheng H, Granholm A, Drago J, Grinberg A,Lee E, Huang S, Saarma M, Hoffer B, Sariola H, WestphalH, 1996: Defects in enteric innervation and kidney developmentin mice lacking GDNF. Nature 382, 73–76.Rossi P, Lolicato F, Grimaldi P, Dolci S, Di Sauro A,Filipponi D, Geremia R, 2008: Transcriptome analysis ofdifferentiating spermatogonia stimulated with kit ligand.Gene Expr Patterns 8, 58–70.Ryu B, Orwig K, Oatley JM, Avarbock M, Brinster R, 2006:Effects of aging and niche microenvironment on spermatogonialstem cell self-renewal. Stem Cells 24, 1505–1511.Sanchez MP, Silos-Santiago I, Frisen J, He B, Lira S, BarbacidM, 1996: Renal agenesis and the absence of enteric neuronsin mice lacking GDNF. Nature 382, 70–73.Schofield R, 1978: The relationship between the spleen colonyformingcell and the haemopoietic stem cell. Blood Cells 4,7–25.Schuchardt A, D’agati V, Larsson-Blomberg L, Costantini F,Pachnis V, 1994: Defects in the kidney and enteric nervoussystem of mice lacking the tyrosine kinase receptor Ret.Nature 367, 380–383.Shetty G, Meistrich M, 2007: The missing niche for spermatogonialstem cells: do blood vessels point the way? Cell StemCell 1, 361–363.Shinohara T, Avarbock M, Brinster R, 1999: Beta1- andalpha6-integrin are surface markers on mouse spermatogonialstem cells. Proc Natl Acad Sci USA 96, 5504–5509.Simon L, Ekman G, Tyagi G, Hess R, Murphy KM, Cooke P,2007: Common and distinct factors regulate expression ofmRNA for ETV5 and GDNF, Sertoli cell proteins essentialfor spermatogonial stem cell maintenance. Exp Cell Res 313,3090–3099.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

392 N Kostereva and M-C HofmannTadokoro Y, Yomogida K, Ohta H, Tohda A, Nishimune Y,2002: Homeostatic regulation of germinal stem cell proliferationby the GDNF ⁄ FSH pathway. Mech Dev 113, 29–39.Uyttendaele H, Marazzi G, Wu G, Yan Q, Sassoon D,Kitajewski J, 1996: Notch4 ⁄ int-3, a mammary proto-oncogene,is an endothelial cell-specific mammalian Notch gene.Development 122, 2251–2259.Von Luttichau I, Notohamiprodjo M, Wechselberger A,Peters C, Henger A, Seliger C, Djafarzadeh R, Huss R,Nelson PJ, 2005: Human adult CD34-progenitor cellsfunctionally express the chemokine receptors CCR1,CCR4, CCR7, CXCR5, and CCR10 but not CXCR4. StemCells Dev 14, 329–336.Watt F, Hogan B, 2000: Out of Eden: stem cells and theirniches. Science 287, 1427–1430.Weinmaster G, 1997: The ins and outs of notch signaling. MolCell Neurosci 9, 91–102.Weinmaster G, Roberts VJ, Lemke G, 1991: A homolog ofDrosophila Notch expressed during mammalian development.Development 113, 199–205.Williams R, Lendahl U, Lardelli M, 1995: Complementaryand combinatorial patterns of Notch gene family expressionduring early mouse development. Mech Dev 53, 357–368.Yoshida S, Sukeno M, Nabeshima Y, 2007: A vasculatureassociatedniche for undifferentiated spermatogonia in themouse testis. Science 317, 1722–1726.Zhang X, Ebata K, Robaire B, Nagano M, 2006: Agingof male germ line stem cells in mice. Biol Reprod 74, 119–124.Author’s address (for correspondence): Marie-Claude Hofmann,Department of Veterinary Biosciences and Institute for GenomicBiology, University of Illinois at Urbana-Champaign, 2001 SouthLincoln Avenue, Urbana, IL 61802, USA. E-mail: mhofmann@uiuc.eduConflict of interest: The authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 393–400 (2008); doi: 10.1111/j.1439-0531.2008.01190.xISSN 0936-6768Factors Affecting Oocyte Quality: Who is Driving the Follicle?P Mermillod, R Dalbie` s-Tran, S Uzbekova, A The´lie, J-M Traverso, C Perreau, P Papillier and P MongetPhysiologie de la Reproduction et des Comportements, UMR 6175 INRA, CNRS, Universite´ de Tours, Haras Nationaux, Nouzilly, FranceContentsMammalian ovaries contain a large stock of oocytes enclosedin primordial follicles. Ovarian cyclic activity induces some ofthese follicles to initiate growth towards a possible ovulation.However, most of these follicles terminate their growth at anymoment and degenerate through atresia. In growing follicles,only a subset of oocytes are capable to support meiosis,fertilization and early embryo development to the blastocyststage, as shown through embryo in vitro production experiments.This proportion of competent oocytes is increasingalong with follicular size. Growing lines of evidence suggestthat oocyte competence relies on the storage of gene products(messenger RNA or protein) that will be determinant tosupport early stages of embryo development, before fullactivation of embryonic genome. Given these facts, thequestion is: are these gene products stored in oocytes duringlate folliculogenesis, allowing an increasing proportion of themto become competent? Alternatively, these transcripts may bestored during early folliculogenesis as the oocyte grows anddisplays high transcription activity. Several arguments supportthis latter hypothesis and are discussed in this review: (i) manyattempts at prolonged culture of oocytes from antral follicleshave failed to increase developmental competence, suggestingthat developmental competence may be acquired before antralformation; (ii) the recent discovery of oocyte secreted factorsand of their ability to regulate many parameters of surroundingsomatic cells, possibly influencing the fate of folliclesbetween ovulation or atresia, suggests a central role of oocytequality in the success of folliculogenesis. Finally, in addition totheir role in interfollicular regulation of ovulation rate, latefolliculogenesis regulation and atresia could also be seen as aselective process aimed at the elimination through follicularatresia of oocytes that did not succeed to store proper geneproducts set during their growth.IntroductionIn mammals, the ovulation delivers a very special cell:the oocyte. This cell has encountered a long andcomplex process of differentiation, leading to thereduction of DNA complement from 2n to n and tothe preparation of the cytoplasm to orchestrate thefusion of the male and female genome complements andthe remodelling of the resulting complete genomecapable to ensure the early development and cell lineagedifferentiation.These particularities place the mammalian oocyte atthe centre of the procreation process. In addition, therecent emergence of embryo-based technologies [in vitrofertilization (IVF), transgenesis and cloning] has reinforcedthe interest of the scientific community into theoocyte. Beyond these applications, the study of oocytefunctions and specificities provides interesting clues forthe better comprehension of several basic biologicalprocesses (cell cycle regulation, genome silencing, posttranscriptionalregulation of gene expression, etc.).The oocyte enters the meiotic division cycle in thefoetal gonad and stops the meiotic progression at thelate prophase stage (diplotene stage, germinal vesicle)around the time of birth, depending on the species. Itremains at this meiotic stage for the complete durationof oocyte growth in the ovarian follicle. During folliculogenesis,the oocyte grows, undergoes modifications ofits ultrastructure and stores the RNA and proteinmaterials necessary to finally become competent toresume and complete maturation, support fertilizationand initiate chromatin remodelling and embryo development.These different competencies are sequentialacquisitions during oocyte differentiation (Sirard et al.2006). The ability to resume meiosis [germinal vesiclebreakdown (GVBD)], to progress to metaphase I and tometaphase II may appear sequentially in some specieslike sheep and goats (Mermillod et al. 1999). In cattle,one can consider that all oocytes are meioticallycompetent soon after antrum formation. Indeed,oocytes from antral follicles spontaneously resumemeiosis when placed in culture (Edwards 1965).At this stage of meiotic competence, not all oocytesare able to be fertilized and to initiate early embryodevelopment. These acquisitions seem to be progressivein the course of follicular growth. It means that theproportion of competent oocytes increases during folliculogenesis,but at the end very few oocytes reach thefinal stage of full competence because of the loss of mostfollicles by physiological atresia occurring at any time oftheir growth. Therefore, the progression of the proportionof competent oocytes within the population ofgrowing follicles may be the result of late oocytedifferentiation through accumulation of factors (messengers,proteins), which will be involved in the successof early developmental steps or the result of the selectionof follicles containing the more fit oocytes, in anincreasingly selective hormonal environment and challenginginterfollicular regulation.The scope of this short review will be to define oocytecompetence, to describe experimental models designedin domestic species for the study of this competence andto discuss respective inputs of oocyte and surroundingsomatic cells in driving an ovarian follicle to ovulationand delivery of a fully competent oocyte.Oocyte Competence Definition and EvaluationDevelopmental competence of the oocyte (or oocytequality) may be defined as its ability to mature, befertilized and give rise to normal and fertile offspringafter normal gestation (Duranthon and Renard 2001).Although the ability of oocytes to reach theblastocyst stage in culture is not a perfect reflection ofÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

394 P Mermillod, R Dalbie` s-Tran, S Uzbekova, A The´lie, J-M Traverso, C Perreau, P Papillier and P Mongetdevelopmental competence, it could be used as a filter todiscriminate between oocyte populations, which aremore or less advanced on the way to competence. In thisview, in vitro production (IVP) of embryos is commonlyused as a means to diagnose the competence of oocytebatches representing different physio-pathological situations.Additional criteria may be used to investigate thequality of resulting embryos such as the evaluation ofthe level of expression of specific gene sets (Corcoranet al. 2005; Knijn et al. 2005), embryo metabolism, orin vitro survival of embryos after freezing or vitrification(Rizos et al. 2001). Nevertheless, embryo transfer,pregnancy results and offspring viability remain theultimate way to finally conclude about oocyte quality.In vitro production experiments provide informationabout the proportion of competent oocytes in a batch.In addition to this global evaluation, experimentalprotocols have been developed to analyze the competenceof individual oocytes (Carolan et al. 1996; Fenget al. 2007). In these conditions, it becomes possible tocorrelate oocyte competence to some environmentalparameters (i.e. follicular characteristics). This approachallows for a more precise evaluation of the links betweencumulus oocyte complex (COC) morphology or otherfollicular parameters and quality of the enclosed oocyte.Several morphological or molecular markers ofoocyte quality have been proposed (Wang and Sun2007). Because the presence of cumulus cells is requiredfor the success of in vitro maturation (IVM) and IVF,morphological evaluation is usually applied to the wholeCOC. Compact, continuous multilayer cumulus investmentand bright, homogeneous ooplasm are consideredas a sign of immature oocyte quality, whereas incomplete,expanded, granulated or too thin (less than threelayers) cumulus or dark, heterogeneous oocyte cytoplasmare related to lower quality immature oocytes(Blondin and Sirard 1995).Different intrinsic oocyte parameters may be measuredsuch as mitochondrial activity, calcium, ATP orglutathione contents, phosphodiesterase activity, baxtranscript expression, etc. (Wang and Sun 2007). Despitethe fact that the evaluation of these parameters isinvasive and requires oocyte destruction, they could beused to evaluate global quality of oocytes from differentin vivo or in vitro treatments, but they cannot beconsidered as predictive parameters. Brilliant cresyl blue(BCB) staining has been proposed as an easy and vitalway to evaluate glucose 6-phosphate dehydrogenase(G6PDH) activity in oocytes (Alm et al. 2005). The blueBCB stain is converted to colourless by G6PDHenzymatic activity. BCB stained immature oocytes (i.e.oocytes with low G6PDH activity, remaining blue afterBCB treatment) provided higher blastocyst rate ascompared with BCB negative oocytes or control ones,indicating a negative correlation between G6PDHactivity level and oocyte quality. In addition, BCBstaining before IVM did not affect IVP results, makingthis method an appropriate way of predictive oocytequality evaluation.Molecular markers of oocyte quality could also beidentified in surrounding tissues, like cumulus cells, or infollicular fluid. For example, the rate of apoptosis inbovine cumulus cells before (Zeuner et al. 2003; Fenget al. 2007) or during (Ikeda et al. 2003) IVM isnegatively correlated to oocyte quality. This is ofparticular interest because, whereas apoptosis evaluationrequires the destruction of the cumulus cells, thismeasure could be performed on cumulus biopsy in thecourse of an IVF process or on cells removed frommature oocytes in an intracytoplasmic sperm injectionprocess.Oocyte Differentiation In vivoIt has been well established in several domestic andmodel species that the proportion of competent oocytesincreases with follicular size (cattle: Lonergan et al.1994; Lequarre et al. 2005; pig: Marchal et al. 2002;mouse: Eppig et al. 1992). This is represented in Fig. 1for cattle oocytes from follicles of less vs more than4 mm in diameter. The current hypothesis to explain thisobservation is that germinal vesicle arrested oocytescontinue their functional differentiation during thewhole course of follicular growth to finally reach fullcompetence (Mermillod et al. 1999). However, this alsoindicates that even small classes of antral folliclesalready contain a small proportion of fully competentoocytes. Therefore, the kinetics of oocyte differentiationalong the course of follicular growth is different fromone follicle to another.Several physiological factors are known to interferewith the kinetics of evolution of the ovarian oocytepopulation towards competence. For example, in sheep,it has been shown that oocytes from ewes heterozygousfor the booroola fecundity gene were larger in diameterand more competent for development as compared withoocytes collected from follicles of the same size class inwild genotype ewes (Cognie et al. 1998). Interestingly,the booroola mutation has been identified as a singlenucleotide inactivating mutation of the BMPRIb receptorof TGF-b (Transforming Growth Factor-beta)superfamily members (Mulsant et al. 2001), whichincludes several oocyte secreted factors (OSF).In cattle, the age of the donor is influencing oocytequality. Oocyte collected from pre-pubertal calves havebeen shown to be less competent as compared withtheir adult counterparts (Khatir et al. 1996). TheFig. 1. Rate of cleavage 48 h post-insemination (open bars) anddevelopment to the blastocyst stage on Day 7 post-insemination (blackbars) for cattle oocytes collected from follicles of less than 4 mm indiameter (n = 203), from 4 to 5 mm (n = 260) and more than 6 mm(n = 109). Percentage of total oocytes, mean from five replicates± SEM (a, b: p < 0.05). Adapted from (Lequarre et al. 2005)Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Factors Affecting Oocyte Quality 395physiological status of the donor cow also affects oocytequality. For example, oocytes from small follicles aremore competent during follicular growth phase ascompared with oocyte collected during the dominantphase (Machatkova et al. 2004). In addition, individualdifferences between donor cows in the developmentalcompetence of oocytes collected by ovum pick up mayindicate genetic control of oocyte competence (Tamassiaet al. 2003).In mammals, atresia is observed at any stage offollicular growth (Cahill et al. 1979). One of the earliestsigns of follicular atresia is the drop of estradiolproduction by granulosa cells. We measured aromataseactivity in granulosa cells of dissected bovine follicleswhile the corresponding oocytes were submitted individuallyto IVP and we observed that oocytes able todevelop to the blastocyst stage were originating fromfollicles displaying high aromatase activity (Mermillodet al. 1999). Interestingly, when atresia was induced byprolonged follicular phase under gonadotropin suppressionin superovulated heifers, oocyte quality was notaffected (Oussaid et al. 2000). This contrast may indicatethat growing follicles containing low-quality oocyteare induced into atresia, whereas artificially inducingatresia of large follicles does not affect oocyte quality.Therefore, atresia appears a consequence of oocyte lowquality rather than a cause of oocyte degradation.Modifying hormonal environment by exogenous FSHinjection allows more follicles to reach ovulation. Thisprocedure is widely used in multiple ovulation andembryo transfer schemes in cattle (Lonergan 2007) andsmall ruminant (Cognie et al. 2003) species. Optimizedtreatment involving FSH stimulation followed by a 48-hcoasting period even allowed to produce quite homogeneouspopulations of competent oocytes, reaching up to80% of development to the blastocyst stage (Blondinet al. 2002). Modifying the hormonal environmentreduces the selection pressure on follicles, resulting ina higher proportion of follicles progressing to the preovulatorystage. However, some ultrastructure abnormalitieshave been reported in sheep superovulatedoocytes (O’Callaghan et al. 2000). This may be the resultof artificially driving to ovulation follicles containingpoorly fitted oocytes.Oocyte Differentiation In vitroBased on the hypothesis that developmental competenceis progressively acquired by an increasing proportion ofoocytes during follicular growth, it would appearinteresting to mimic this differentiation in vitro to allowmore oocytes to become competent. In vitro continuationof oocyte differentiation implies that culturedoocytes are maintained at the germinal vesicle stage toallow ongoing regulation of expression of relevant genesduring a prematuration period. Several methods havebeen proposed to maintain oocytes in meiotic arrestin vitro, including the use of chemicals aimed atmaintaining high cAMP level in oocytes such asforskolin, isobutyl methylxanthine or cAMP stableanalogues (Sirard et al. 1998). Other drugs inhibitingprotein synthesis (cycloheximide) or phosphorylation (6-dimethyl aminopurine) were also tested (Lonergan et al.1997). All of these drugs were able to maintain a more orless reversible inhibition of meiotic resumption incultured cattle oocytes. However, no improvement ofoocyte quality was obtained after such inhibitions.More recently, drugs directly acting by inhibition ofenzymatic activity of M-phase promoting factor (MPF),the key element of G2 ⁄ M transition in eukaryotic cellcycle, were used to maintain oocytes at the germinalvesicle stage. MPF is a heterodimer of CDK1 catalyticsubunit and cyclin B regulatory subunit. Interestingresults have been obtained by using puric moleculescompeting with ATP for binding the catalytic pocket ofMPF heterodimer. Butyrolactone I (Lonergan et al.2000) and roscovitine (Mermillod et al. 2000) weresuccessfully used in this view and provided an efficientand reversible way to maintain cattle oocytes at thegerminal vesicle stage in culture for 24–48 h. However,as shown in Fig. 2, 24 h culture of germinal vesicleoocyte did not result in improvement of developmentpotential even under different stimulation (Ponderatoet al. 2002). However, although these molecules allowedan efficient block of meiotic progression, several meiosisrelated cytoplasmic events did occur under inhibition,such as modifications of protein synthesis andphosphorylation patterns, indicating the initiation ofcytoplasmic maturation (Vigneron et al. 2004a). Interestingly,this approach allowed to point out severalMPF-independent signalling pathways (Akt, JNK andAurora A) activated during meiotic resumption (Vigneronet al. 2004b) and probably involved in the regulationof some aspects of cytoplasmic maturation.Amongst these pathways, Aurora A, a serine ⁄ threonineprotein kinase, seems to be a major player in thephosphorylation cascade leading to MPF activation,meiotic resumption and post-transcriptional control ofgene expression (Uzbekova et al. 2008). Failing toincrease the quality of immature oocytes by allowingthem to continue their late differentiation during aprematuration culture under chemical meiotic inhibitionmay be attributed to the diversity of the signallingmechanisms involved in the regulation of differentmaturation aspects. Alternatively, it may be due to anearlier determination of oocyte quality, by the time ofhigher transcriptional activity in growing oocytes.Oocyte Gene ExpressionThe current hypothesis to explain differential ability ofoocytes to develop after fertilization is that the oocyteshould store the appropriate set of mRNA and proteinsthat will be required during the period of genomicsilencing from GVBD to major zygotic transition(MZT) and for the onset of MZT. Transcriptionalactivity is high in growing oocytes up to early antralstages of follicular growth and decreases to a basallevel thereafter, as follicular diameter reaches 3 mmand the diameter of the oocyte becomes 110 lm inbovine (Fair et al. 1997). A transcription burst issuspected just before GVBD (Rodriguez and Farin2004a,b). Genes expressed at that time are supposed tobe involved in meiotic resumption and progression butmay also have a role during early development(Rodriguez et al. 2006).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

396 P Mermillod, R Dalbie` s-Tran, S Uzbekova, A The´lie, J-M Traverso, C Perreau, P Papillier and P MongetEGF (n = 195)Rescovitine (n = 202)Rescovitine = EGF (n = 208)Fig. 2. Cleavage and rate of blastocystsat 7, 8 and 9 days post-inseminationfor oocytes matured in199 medium supplemented with10 ng ⁄ ml of epidermal growthfactor (EGF) either immediatelyafter collection (EGF) or after aprematuration treatment underroscovitine inhibition in mediumalone (roscovitine) or in the presenceof EGF. Means of three replicates.Inhibition of meioticresumption did not affect embryodevelopment, even under EGF stimulationNevertheless, if oocyte quality relies on mRNAstorage, these mRNAs could be stored during maintranscriptional activity in oocytes from pre-antral folliclesand kept in a stable form for future use duringmaturation or after fertilization. Indeed, whereas insomatic tissues, mRNA content directly reflects contemporaryprotein expression pattern, post-transcriptionalgene expression regulation is slightly different inoocyte where messengers could be masked and maintainedfor long periods of time and then be processed fortranslation or destruction through polyadenylation regulation(Bettegowda and Smith 2007) or RNA interferencemechanisms (Murchison et al. 2007).However, if it is admitted that oocyte developmentalcompetence may be acquired during late folliculargrowth, one should admit that determinant messengersmay be produced in oocytes from late antral follicles.This transcription activity could be quantitatively lowbut functionally determinant. Much research has beendevoted towards identification of these critical mRNAduring the past 10 years.The development of highly sensitive molecular biologymethods has allowed to identify a panel of genesthat are preferentially expressed in oocytes of mammals(Cui and Kim 2007). The list of these genes is currentlyactively expanding and will probably continue to growduring the forthcoming years, because of active researchwork and increasing sensitivity of molecular techniques.In particular, subtractive suppressive hybridization(SSH) approach allowed to evidence hundreds of genesoverexpressed in cattle oocytes as compared withsomatic tissues (Pennetier et al. 2005; Vallee et al. 2005).In mouse oocytes, knockout experiments allowed topoint out several genes, which should be expressed inoocyte to allow embryo development after fertilization.These genes are referred to as ‘maternal effect genes’such as maternal antigen that embryo require (Mater) orzygotic arrest-1 (Zar-1). These genes may be involved inthe onset of transcriptional activity in zygote nucleusduring MZT occurring at the two-cell stage in mice(Minami et al. 2007) and at the eight-cell stage inruminant species.Some orthologous genes of mouse maternal-effectgenes have been cloned in domestic species. For example,we cloned bovine Mater and Zar-1 orthologousgenes in cattle (Pennetier et al. 2004; Uzbekova et al.2006). However, the study of bovine Mater expressionprofile (Pennetier et al. 2006) indicated that this genewas transcribed and translated at early stages offollicular growth, long before the suspected time ofdevelopmental competence acquisition. Although nomaternal effect gene has been shown to be translatedspecifically during late folliculogenesis, it could beproposed that some determinant genes are progressivelyaccumulated during this period and that full developmentalcompetence is acquired as a threshold of mRNAor protein from these genes are reached in the oocyte.However, given that a significant population offollicles degenerate by atresia at different stages of theirgrowth, another explanation may be that folliclescontaining competent oocytes are more able to survivethe increasingly challenging hormonal environmentthrough a better coordination of follicular somatic cellsproliferation and differentiation.Oocyte Regulation of Follicular GrowthThe role of somatic follicular cells in the establishmentof oocyte competence has been well established. Cumuluscells are participating to oocyte meiotic arrest andresumption and oocyte metabolism (glutathione storage,glucose metabolism), they also participate to oocyteÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Factors Affecting Oocyte Quality 397capture by infundibulum after ovulation and to fertilizationregulation (Van Soom et al. 2002).During the last few years, increasing lines of evidenceshave suggested that the oocyte itself also regulates manyparameters of follicular somatic cells proliferation anddifferentiation. The TGF-b superfamily gathers morethan 35 members involved in the regulation of severalphysiological processes in mammals. In particular,TGF-b, members are strongly involved in many aspectsof ovarian function regulation (Knight and Glister2006). This superfamily includes different families ofgrowth factors such as TGF-b family, bone morphogeneticproteins (BMP), growth and differentiation factors(GDF), activin ⁄ inhibin and anti-mullerian hormone. Italso includes type I and type II receptors acting throughSmad signalling to modulate target genes expression, aswell as non-signalling secreted or membrane-boundbinding proteins that may regulate TGF-b familymembers’ biologic activity. TGF-b family membersoriginated from somatic follicular cells are involved inthe regulation of several aspects of follicular growthinitiation and progression (Knight and Glister 2006). Inaddition, GDF-9, BMP-15 (also called GDF-9B) andBMP-6 have been shown to be produced by oocytesfrom the primordial follicle stage (McNatty et al. 2001).In mouse, GDF-9 knockout induces arrest of folliculardevelopment at the primary stage and sterility (Donget al. 1996) and GDF-9 may also be involved inprimordial to primary follicle transition in the rat ovary(Vitt et al. 2000). Oocyte secreted factors are necessaryto induce the differentiation of pre-antral mouse granulosacells to cumulus cells [ability to expand uponepidermal growth factor (EGF) stimulation (Diaz et al.2007)]. In the sheep, naturally occurring inactivatingmutations of GDF-9 or BMP-15 increase ovulation ratein heterozygous ewes while homozygous females areinfertile and active immunization against GDF-9 andBMP-15 blocks folliculogenesis at the primary stage inewe (McNatty et al. 2005). During antral folliculargrowth, GDF-9, BMP-15 and BMP-6 promote granulosacells proliferation and modulate their sensitivity toFSH, thus participating to the selection of the dominantfollicle (Knight and Glister 2006). In addition, it hasbeen shown recently in bovine that oocyte-secretedBMP-6 and BMP-15 are protecting cumulus cellsagainst apoptosis occurring during culture of cattleCOC (Hussein et al. 2005) and GDF-9 preventspremature differentiation (luteinization) of bovine granulosacells (Spicer et al. 2006). Recently, JY-1 has beendiscovered as another oocyte-specific secreted factormodulating granulosa cell functions (Bettegowda et al.2007), highlighting again the central role of oocyte andOSF in the regulation of follicular growth.In addition, the oocyte is able to regulate surroundingsomatic cells function through the modulation of theirmetabolism. Subtractive suppressive hybridizationallowed to identify six genes involved in glycolysis thatwere overexpressed in mouse cumulus cells as comparedwith mural granulosa cells of antral follicles and thisoverexpression was dependant on the presence of theoocyte or of OSF (Sugiura et al. 2005). More recently, ithas been established that this oocyte regulation ofcumulus cells glycolysis is mediated by GDF-9 andFGF-8 OSF cooperation (Sugiura et al. 2007). In thesame way, GDF-9 and BMP-15 promote cholesterolsynthesis by cumulus cells (Su et al. 2008). This cholesterolmay be useful to oocyte that is unable to produceit. Interesting oocyte grafting experiments in mice(Eppig et al. 2002) showed that, placed in similarenvironment, oocytes collected from secondary follicleswere more able to promote follicular growth with appropriatemural granulosa ⁄ cumulus cell differentiation (LHFig. 3. Two hypotheses to explain the increasing ratio of competent oocytes in growing follicular populations. (a) Primordial follicles entergrowth phase, oocytes are not yet competent (open circles). By the time of antrum formation, all oocytes become meiotically competent (grey),few are already developmentally competent (black). During further growth, some follicles are lost in atresia, whatever the status of their oocyte.During further follicular growth, the proportion of competent oocytes increases because of ongoing oocyte differentiation. More follicles are lostduring terminal growth. (b) Early follicular growth is similar. Because their oocytes are less able to drive follicular growth, follicles enclosing lessdevelopmentally competent oocytes (grey) are lost in atresia, whereas follicles containing developmentally competent oocytes (black) continuetheir growth. Finally, only follicles containing fully competent oocytes reach ovulationÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

398 P Mermillod, R Dalbie` s-Tran, S Uzbekova, A The´lie, J-M Traverso, C Perreau, P Papillier and P Mongetreceptor expression) as compared with oocytes fromprimary follicles.It appears finally that oocyte is able to regulate manyaspects of follicle somatic cells proliferation, differentiationand metabolic activities involved in late folliculargrowth regulation. This opens the way to a newhypothesis, i.e. that oocytes in early antral folliclesalready display intrinsic developmental competence.This competence may translate into differential abilityto drive proper follicular differentiation in a decreasingFSH support context and increasing interfollicularcompetition. Poorly differentiated oocytes (oocytesbearing some transcriptome abnormalities) fail in maintainingproper somatic cells differentiation status or toprotect these cells against apoptosis induced by increasinglychallenging hormonal environment and finallyleave their follicle degenerate on the way to atresia(Fig. 3). Hormonal stimulation of follicular growthpartly alleviates the selection pressure imposed onfollicles, and allows lower quality oocytes to drive theirfollicles to ovulation in these permissive conditions.The expression of many oocyte-specific genes duringoocyte growth in pre-antral follicles is coordinated bygermline-specific transcription factors such as FIGLAor NOBOX (Pangas and Rajkovic 2006). Interestingly,NOBOX regulates oocyte genes involved both in developmentalcompetence (Mater, Zar-1 and Mos) and incontrol of follicular function by oocyte (GDF-9 andBMP-15). This observation indicates that oocytes deficientin terms of competence may also be deficient intheir ability to properly drive their follicle to ovulation.ConclusionsDuring its early growth in pre-antral follicles, the oocytestores specific gene products that will be determinant forits ability to coordinate follicular growth and that willbe required during early embryo development afterfertilization. Increasing lines of evidence suggest that bythe time of antrum formation, the quality of oocyte isalready determined. This quality will first be expressedthrough the ability of the oocyte to drive folliculargrowth under an increasingly challenging hormonalenvironment up to ovulation, avoiding atresia. Thisquality will be then expressed after fertilization throughthe ability to support early embryo development and toinitiate embryonic genome transcriptional activity. Ofcourse, further experimental data are required to supportthis hypothesis.ReferencesAlm H, Torner H, Lohrke B, Viergutz T, Ghoneim IM, KanitzW, 2005: Bovine blastocyst development rate in vitro isinfluenced by selection of oocytes by brilliant cresyl bluestaining before IVM as indicator for glucose-6-phosphatedehydrogenase activity. Theriogenology 63, 2194–2205.Bettegowda A, Smith GW, 2007: Mechanisms of maternalmRNA regulation: implications for mammalian earlyembryonic development. Front Biosci 12, 3713–3726.Bettegowda A, Yao J, Sen A, Li Q, Lee K-B, Kobayashi Y,Patel OV, Coussens PM, Ireland JJ, Smith GW, 2007: JY-1,an oocyte-specific gene, regulates granulosa cell functionand early embryonic development in cattle. In: Proceedingsof the National Academy of Sciences, %R 10.1073/pnas.0706383104, 0706383104.Blondin P, Sirard MA, 1995: Oocyte and follicularmorphology as determining characteristics for developmentalcompetence in bovine oocytes. Mol Reprod Dev41, 54–62.Blondin P, Bousquet D, Twagiramungu H, Barnes F, SirardMA, 2002: Manipulation of follicular development toproduce developmentally competent bovine oocytes. BiolReprod 66, 38–43.Cahill LP, Mariana JC, Mauleon P, 1979: Total follicularpopulations in ewes of high and low ovulation rates.J Reprod Fertil 55, 27–36.Carolan C, Lonergan P, Khatir H, Mermillod P, 1996: In vitroproduction of bovine embryos using individual oocytes. MolReprod Dev 45, 145–150.Cognie Y, Benoit F, Poulin N, Khatir H, Driancourt MA,1998: Effect of follicle size and of the FecB Booroola gene onoocyte function in sheep. J Reprod Fertil 112, 379–386.Cognie Y, Baril G, Poulin N, Mermillod P, 2003: Currentstatus of embryo technologies in sheep and goat. Theriogenology59, 171–188.Corcoran D, Fair T, Lonergan P, 2005: Predicting embryoquality: mRNA expression and the preimplantation embryo.Reprod Biomed Online 11, 340–348.Cui XS, Kim NH, 2007: Maternally derived transcripts:identification and characterisation during oocyte maturationand early cleavage. Reprod Fertil Dev 19, 25–34.Diaz FJ, Wigglesworth K, Eppig JJ, 2007: Oocytes determinecumulus cell lineage in mouse ovarian follicles. J Cell Sci120, 1330–1340.Dong J, Albertini DF, Nishimori K, Kumar TR, Lu N,Matzuk MM, 1996: Growth differentiation factor-9 isrequired during early ovarian folliculogenesis. Nature 383,531–535.Duranthon V, Renard JP, 2001: The developmental competenceof mammalian oocytes: a convenient but biologicallyfuzzy concept. Theriogenology 55, 1277–1289.Edwards RG, 1965: Maturation in vitro of mouse, sheep, cow,pig, rhesus monkey and human ovarian oocytes. Nature208, 349–351.Eppig JJ, Schroeder AC, O’Brien MJ, 1992: Developmentalcapacity of mouse oocytes matured in vitro: effects ofgonadotrophic stimulation, follicular origin and oocyte size.J Reprod Fertil 95, 119–127.Eppig JJ, Wigglesworth K, Pendola FL, 2002: The mammalianoocyte orchestrates the rate of ovarian follicular development.Proc Natl Acad Sci U S A 99, 2890–2894.Fair T, Hulshof SC, Hyttel P, Greve T, Boland M, 1997:Nucleus ultrastructure and transcriptional activity of bovineoocytes in preantral and early antral follicles. Mol ReprodDev 46, 208–215.Feng WG, Sui HS, Han ZB, Chang ZL, Zhou P, Liu DJ, BaoS, Tan JH, 2007: Effects of follicular atresia and size on thedevelopmental competence of bovine oocytes: a study usingthe well-in-drop culture system. Theriogenology 67, 1339–1350.Hussein TS, Froiland DA, Amato F, Thompson JG,Gilchrist RB, 2005: Oocytes prevent cumulus cellapoptosis by maintaining a morphogenic paracrine gradientof bone morphogenetic proteins. J Cell Sci 118, 5257–5268.Ikeda S, Imai H, Yamada M, 2003: Apoptosis in cumulus cellsduring in vitro maturation of bovine cumulus-enclosedoocytes. Reproduction 125, 369–376.Khatir H, Lonergan P, Carolan C, Mermillod P, 1996:Prepubertal bovine oocyte: a negative model for studyingoocyte developmental competence. Mol Reprod Dev 45,231–239.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Factors Affecting Oocyte Quality 399Knight PG, Glister C, 2006: TGF-beta superfamily membersand ovarian follicle development. Reproduction 132, 191–206.Knijn HM, Wrenzycki C, Hendriksen PJ, Vos PL, ZeinstraEC, van der Weijden GC, Niemann H, Dieleman SJ, 2005:In vitro and in vivo culture effects on mRNA expression ofgenes involved in metabolism and apoptosis in bovineembryos. Reprod Fertil Dev 17, 775–784.Lequarre AS, Vigneron C, Ribaucour F, Holm P, Donnay I,Dalbies-Tran R, Callesen H, Mermillod P, 2005: Influenceof antral follicle size on oocyte characteristics and embryodevelopment in the bovine. Theriogenology 63, 841–859.Lonergan P, 2007: State-of-the-art embryo technologies incattle. Soc Reprod Fertil Suppl 64, 315–325.Lonergan P, Monaghan P, Rizos D, Boland MP, Gordon I,1994: Effect of follicle size on bovine oocyte quality anddevelopmental competence following maturation, fertilization,and culture in vitro. Mol Reprod Dev 37, 48–53.Lonergan P, Khatir H, Carolan C, Mermillod P, 1997: Bovineblastocyst production in vitro after inhibition of oocytemeiotic resumption for 24 h. J Reprod Fertil 109, 355–365.Lonergan P, Dinnyes A, Fair T, Yang X, Boland M, 2000:Bovine oocyte and embryo development following meioticinhibition with butyrolactone I. Mol Reprod Dev 57, 204–209.Machatkova M, Krausova K, Jokesova E, Tomanek M, 2004:Developmental competence of bovine oocytes: effects offollicle size and the phase of follicular wave on in vitroembryo production. Theriogenology 61, 329–335.Marchal R, Vigneron C, Perreau C, Bali-Papp A, MermillodP, 2002: Effect of follicular size on meiotic and developmentalcompetence of porcine oocytes. Theriogenology 57,1523–1532.McNatty KP, Juengel JL, Wilson T, Galloway SM, Davis GH,2001: Genetic mutations influencing ovulation rate in sheep.Reprod Fertil Dev 13, 549–555.McNatty KP, Smith P, Moore LG, Reader K, Lun S,Hanrahan JP, Groome NP, Laitinen M, Ritvos O, JuengelJL, 2005: Oocyte-expressed genes affecting ovulation rate.Mol Cell Endocrinol 234, 57–66.Mermillod P, Oussaid B, Cognie Y, 1999: Aspects of follicularand oocyte maturation that affect the developmentalpotential of embryos. J Reprod Fertil Suppl 54, 449–460.Mermillod P, Tomanek M, Marchal R, Meijer L, 2000: Highdevelopmental competence of cattle oocytes maintained atthe germinal vesicle stage for 24 hours in culture by specificinhibition of MPF kinase activity. Mol Reprod Dev 55, 89–95.Minami N, Suzuki T, Tsukamoto S, 2007: Zygotic geneactivation and maternal factors in mammals. J Reprod Dev53, 707–715.Mulsant P, Lecerf F, Fabre S, Schibler L, Monget P, LannelucI, Pisselet C, Riquet J, Monniaux D, Callebaut I, Cribiu E,Thimonier J, Teyssier J, Bodin L, Cognie Y, Chitour N,Elsen JM, 2001: Mutation in bone morphogenetic proteinreceptor-IB is associated with increased ovulation rate inBooroola Merino ewes. Proc Natl Acad Sci U S A 98, 5104–5109.Murchison EP, Stein P, Xuan Z, Pan H, Zhang MQ, SchultzRM, Hannon GJ, 2007: Critical roles for Dicer in the femalegermline. Genes Dev 21, 682–693.O’Callaghan D, Yaakub H, Hyttel P, Spicer LJ, Boland MP,2000: Effect of nutrition and superovulation on oocytemorphology, follicular fluid composition and systemichormone concentrations in ewes. J Reprod Fertil 118,303–313.Oussaid B, Lonergan P, Khatir H, Guler A, Monniaux D,Touze JL, Beckers JF, Cognie Y, Mermillod P, 2000: Effectof GnRH antagonist-induced prolonged follicular phase onfollicular atresia and oocyte developmental competencein vitro in superovulated heifers. J Reprod Fertil 118, 137–144.Pangas SA, Rajkovic A, 2006: Transcriptional regulation ofearly oogenesis: in search of masters. Hum Reprod Update,12, 65–76.Pennetier S, Uzbekova S, Perreau C, Papillier P, Mermillod P,Dalbies-Tran R, 2004: Spatio-temporal expression of thegerm cell marker genes MATER, ZAR1, GDF9, BMP15,andVASA in adult bovine tissues, oocytes, and preimplantationembryos. Biol Reprod 71, 1359–1366.Pennetier S, Uzbekova S, Guyader-Joly C, Humblot P,Mermillod P, Dalbies-Tran R, 2005: Genes preferentiallyexpressed in bovine oocytes revealed by subtractiveand suppressive hybridization. Biol Reprod 73, 713–720.Pennetier S, Perreau C, Uzbekova S, Thelie A, Delaleu B,Mermillod P, Dalbies-Tran R, 2006: MATER proteinexpression and intracellular localization throughout folliculogenesisand preimplantation embryo development in thebovine. BMC Dev Biol 6, 26.Ponderato N, Crotti G, Turini P, Duchi R, Galli C, Lazzari G,2002: Embryonic and foetal development of bovine oocytestreated with a combination of butyrolactone I and roscovitinein an enriched medium prior to IVM and IVF. MolReprod Dev 62, 513–518.Rizos D, Ward F, Boland MP, Lonergan P, 2001: Effect ofculture system on the yield and quality of bovine blastocystsas assessed by survival after vitrification. Theriogenology 56,1–16.Rodriguez KF, Farin CE, 2004a: Developmental capacity ofbovine cumulus oocyte complexes after transcriptionalinhibition of germinal vesicle breakdown. Theriogenology61, 1499–1511.Rodriguez KF, Farin CE, 2004b: Gene transcription andregulation of oocyte maturation. Reprod Fertil Dev 16, 55–67.Rodriguez KF, Blomberg LA, Zuelke KA, Miles JR, AlexanderJE, Farin CE, 2006: Identification of candidate mRNAsassociated with gonadotropin-induced maturation of murinecumulus oocyte complexes using serial analysis of geneexpression. Physiol Genomics 27, 318–327.Sirard MA, Richard F, Mayes M, 1998: Controlling meioticresumption in bovine oocytes: a review. Theriogenology 49,483–497.Sirard MA, Richard F, Blondin P, Robert C, 2006: Contributionof the oocyte to embryo quality. Theriogenology 65,126–136.Spicer LJ, Aad PY, Allen D, Mazerbourg S, Hsueh AJ, 2006:Growth differentiation factor-9 has divergent effects onproliferation and steroidogenesis of bovine granulosa cells.J Endocrinol 189, 329–339.Su YQ, Sugiura K, Wigglesworth K, O’Brien MJ, Affourtit JP,Pangas SA, Matzuk MM, Eppig JJ, 2008: Oocyte regulationof metabolic cooperativity between mouse cumulus cells andoocytes: BMP15 and GDF9 control cholesterol biosynthesisin cumulus cells. Development 135, 111–121.Sugiura K, Pendola FL, Eppig JJ, 2005: Oocyte control ofmetabolic cooperativity between oocytes and companiongranulosa cells: energy metabolism. Dev Biol279, 20–30.Sugiura K, Su YQ, Diaz FJ, Pangas SA, Sharma S, WigglesworthK, O’Brien MJ, Matzuk MM, Shimasaki S, Eppig JJ,2007: Oocyte-derived BMP15 and FGFs cooperate topromote glycolysis in cumulus cells. Development 134,2593–2603.Tamassia M, Heyman Y, Lavergne Y, Richard C, Gelin V,Renard JP, Chastant-Maillard S, 2003: Evidence of oocytedonor cow effect over oocyte production and embryodevelopment in vitro. Reproduction 126, 629–637.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

400 P Mermillod, R Dalbie` s-Tran, S Uzbekova, A The´lie, J-M Traverso, C Perreau, P Papillier and P MongetUzbekova S, Roy-Sabau M, Dalbies-Tran R, Perreau C,Papillier P, Mompart F, Thelie A, Pennetier S, Cognie J,Cadoret V, Royere D, Monget P, Mermillod P, 2006:Zygote arrest 1 gene in pig, cattle and human: evidence ofdifferent transcript variants in male and female germ cells.Reprod Biol Endocrinol 4, 12.Uzbekova S, Arlot-Bonnemains Y, Dupont J, Dalbies-Tran R,Papillier P, Pennetier S, Thelie A, Perreau C, Mermillod P,Prigent C, Uzbekov R, 2008: Spatio-temporal expressionpatterns of Aurora kinases A, B, and C and cytoplasmicpolyadenylation-element-binding protein in bovine oocytesduring meiotic maturation. Biol Reprod 78, 218–233.Vallee M, Gravel C, Palin MF, Reghenas H, Stothard P,Wishart DS, Sirard MA, 2005: Identification of novel andknown oocyte-specific genes using complementary DNAsubtraction and microarray analysis in three differentspecies. Biol Reprod 73, 63–71.Van Soom A, Tanghe S, De Pauw I, Maes D, de Kruif A,2002: Function of the cumulus oophorus before and duringmammalian fertilization. Reprod Domest Anim 37, 144–151.Vigneron C, Perreau C, Dalbies-Tran R, Joly C, Humblot P,Uzbekova S, Mermillod P, 2004a: Protein synthesis andmRNA storage in cattle oocytes maintained under meioticblock by roscovitine inhibition of MPF activity. MolReprod Dev 69, 457–465.Vigneron C, Perreau C, Dupont J, Uzbekova S, Prigent C,Mermillod P, 2004b: Several signaling pathways areinvolved in the control of cattle oocyte maturation. MolReprod Dev 69, 466–474.Vitt UA, McGee EA, Hayashi M, Hsueh AJ, 2000: In vivotreatment with GDF-9 stimulates primordial and primaryfollicle progression and theca cell marker CYP17 in ovariesof immature rats. Endocrinology 141, 3814–3820.Wang Q, Sun QY, 2007: Evaluation of oocyte quality:morphological, cellular and molecular predictors. ReprodFertil Dev 19, 1–12.Zeuner A, Muller K, Reguszynski K, Jewgenow K, 2003:Apoptosis within bovine follicular cells and its effect onoocyte development during in vitro maturation. Theriogenology59, 1421–1433.Author’s address (for correspondence): P Mermillod, INRA, Physiologiede la Reproduction et des Comportements, UMR6175 INRA,CNRS, Universite´ de Tours, Haras Nationaux, 37380 Nouzilly,France. E-mail: mermillo@tours.inra.frConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 401–406 (2008); doi: 10.1111/j.1439-0531.2008.01191.xISSN 0936-6768Selected Aspects of Advanced Porcine Reproductive TechnologyK Kikuchi 1 , N Kashiwazaki 2 , T Nagai 3 , M Nakai 1 , T Somfai 1 , J Noguchi 1 and H Kaneko 11 Division of Animal Sciences, National Institute of Agrobiological Sciences, Tsukuba, Ibaraki; 2 Laboratory of Animal Reproduction, School ofVeterinary Medicine, Azabu University, Sagamihara, Kanagawa; 3 National Institute of Livestock and Grassland Science, Tsukuba, Ibaraki, JapanContentsIn vitro fertilization (IVF) of in vitro matured (IVM) oocytesin pigs has become the most popular method of studyinggametogenesis and embryogenesis in this species. Furthermore,because of recent advances in in vitro culture (IVC) ofIVM–IVF embryos, in vitro production (IVP) of embryosnow enables us to generate viable embryos as successfully asfor in vivo-derived embryos and with less cost and in lesstime. These technologies contribute not only to developmentsin reproductive physiology and agriculture but also tothe conservation of porcine genetic resources and theproduction of cloned or genetically modified pigs. However,in IVP, there still remains the problem of abnormal ploidy,which is caused by performing procedures under nonphysiologicalconditions. In recent years, unique technologiessuch as intracytoplasmic sperm injection (ICSI) or xenograftingof gonadal tissue into immunodeficient experimentalanimals have been developed to help conserve gameteresources. These technologies combined with IVP areexpected to be useful for the conservation of gametes fromimportant genetic resources. Here, we discuss the developmentalability and normality of porcine IVP embryos andalso the utilization of ICSI and xenografting in advancingbiotechnology in pigs.IVF and the Normality of Porcine ZygotesThe in vitro developmental competence or viability ofporcine in vitro matured (IVM)–in vitro fertilized (IVF)oocytes to the blastocyst stage was first confirmed andreported by Mattioli et al. (1989). Furthermore, pigletshave been born from IVM–IVF embryos after in vitroculture (IVC) to the 2- to 4-cell stages (Mattioli et al.1989; Yoshida et al. 1993). Since then, some laboratorieshave succeeded in producing piglets from embryoscleaved at the 2- to 4-cell stages after IVM–IVF andIVC for 24–36 h (Funahashi et al. 1996, 1997). Viablepiglets were also generated after transfer of in vitroproduced (IVP) embryos at the blastocyst stage (Marchalet al. 2001; Kikuchi et al. 2002). Over several recentyears, IVC procedures have been improved, but IVMand IVF systems still have unsolved problems, including(1) imbalance of nuclear and cytoplasmic maturationand (2) polyspermy. Both these phenomena causeabnormal ploidy in IVP embryos, potentially resultingin loss of embryos after their transfer to recipients. Weneed to make every effort to achieve normality in IVPembryos after IVM and IVF.Fertilization at the immature stage, before nuclearmaturationImmature oocytes before nuclear maturation, such asthose at the metaphase-I stage, acquire fertilization oractivation potential (Kikuchi et al. 1999) and developmentalpotential, at least to the blastocyst stage (Somfaiet al. 2005). Our recent data suggest that treatment ofimmature oocytes with cytochalasin B, a drug thatinhibits actin filament polymerization, results in thefailure of extrusion of the first polar body during thefirst meiosis (Somfai et al. 2006). This failure, whichmay also be caused under physiological conditions,leads to lack of completion of meiotic maturation inoocytes (Somfai et al., unpublished). The developmentalcompetence of these immature oocytes is poor, resultingin low rates of blastocyst formation and low cellnumbers in the blastocysts. Chromosomal analysissuggests that these immature porcine oocytes fertilizedin vitro develop abnormally in terms of ploidy andespecially show increased rates of triploidy (Somfaiet al. 2005).Polyspermic fertilization of matured oocytesThe best-known problem that occurs during porcineIVF is polyspermy (Nagai 1996). Many researchershave been trying to solve this problem by differentapproaches, but so far no one has succeeded.Although it has been suggested that polyspermicoocytes have the ability to extrude the extra spermchromosomes in the early embryonic stage, beforeblastulation (Funahashi and Day 1997), the abnormalploidies are reportedly maintained until the blastocyststage (Han et al. 1999; McCauley et al. 2003; Somfaiet al. in press).Utilization of Intracytoplasmic Sperm Injectionfor PigsIntracytoplasmic sperm injection (ICSI) was first performedto study the early events of fertilization inhamsters (Uehara and Yanagimachi 1976, 1977). Fromthe 1990s, ICSI has been used for assisted reproductivetechnology in humans. Also in animals, viable offspringcan be generated by using sperm with poor motility orviability (Table 1). These successful reports confirm theooplasmic capacity for fertilization and development toterm after ICSI in both in vivo matured and IVMoocytes in mammals. If the spermatozoa have lost theirmotility, ICSI is indispensable for fertilization to producethe next generation. In pigs, the use of in vivomatured oocytes led to the first successful production ofpiglets after ICSI (Kolbe and Holtz 2000; Martin 2000);the use of ICSI in porcine IVM oocytes for successfulpiglet production was first confirmed in our laboratory(Nakai et al. 2003).Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

402 K Kikuchi, N Kashiwazaki, T Nagai, M Nakai, T Somfai, J Noguchi and H KanekoTable 1. First successful productions of offspring after ICSI indifferent speciesSpeciesIn vivo maturedOocyte sourceIn vitro maturedHuman Palermo et al. (1992) Nagy et al. (1996)Mouse Ahmadi et al. (1995)Kimura and Yanagimachi (1995a)Lacham-Kaplan and Trounson (1995)Cow Goto et al. (1991)Sheep Catt et al. (1996)Rabbit Li et al. (2001)Monkey Hewitson et al. (1999)Cat Gomez et al. (2000)Horse Cochran et al. (1998)RatHirabayashi et al. (2002a)Hamster Yamauchi et al. (2002)Pig Kolbe and Holtz (2000) Nakai et al. (2003)Martin (2000)Generation of zygotes from immotile spermatozoa orimmature sperm cellsOffspring have also been obtained by using immaturesperm cells such as round spermatids in mice (Oguraet al. 1994) and rats (Hirabayashi et al. 2002b), orelongated spermatids in monkeys (Hewitson et al. 2002).In mice, both primary (Kimura et al. 1998; Ogura et al.1998; Sasagawa et al. 1998) and secondary (Kimura andYanagimachi 1995b) spermatocytes could be injected,and the procedure yielded living pups. In pigs, however,when round spermatids were injected into IVM oocytes(Kim et al. 1999), no piglets were obtained aftertransfer. No other immature sperm cells are useful forembryonic development in pigs.Avoiding polyspermyConsidering the fact that the ultimate procedure toblock polyspermy has not yet been established inporcine IVF, single sperm injection can be a mostwelcome procedure to generate normal (monospermic)embryos and piglets. However, the efficacy of ICSI inpigs remains low (Nakai et al. 2003). In this report, wehave transferred 598 oocytes after ICSI into total sevenrecipients; however, only three live piglets obtained froma single recipient. To accelerate the acrosome reaction,we have pre-treated spermatozoa with calcium ionophorebefore sperm injection (Nakai et al. 2003), or withdithiothreitol to induce in vitro decondensation of thesperm nuclei (Nakai et al. 2006). However, no increasedefficacy of embryo production was confirmed in ourlaboratory.Freeze-drying of spermatozoaNew technologies for the conservation of spermatozoaother than by cryopreservation in liquid nitrogen havebeen anticipated. One of the most welcome procedures isfreeze-drying of spermatozoa, because the spermatozoacan be stored at less cost, and without the need for liquidnitrogen, at 5°C or room temperature. After freezedrying,the spermatozoa lose their motility and shouldbe fertilized by ICSI. Successful offspring productionhas been reported in mice (Wakayama and Yanagimachi1998; Kaneko et al. 2003a; Ward et al. 2003; Kanekoand Nakagata 2006; Kawase et al. 2007), rabbits (Liuet al. 2004), and rats (Hirabayashi et al. 2005a; Hochiet al., 2008). In mice, DNA damage was induced duringthe holding period after rehydration before sperminjection (Wakayama and Yanagimachi 1998). Therelationship between the duration of rehydration offreeze-dried sperm and the in vitro development offreeze-dried sperm-injected oocytes was examined inpigs (Nakai et al. 2007). The results suggested thatembryos obtained after ICSI with a single freeze-driedsperm head have the competence to be implanted and todevelop to the early fetal stage, and also that theduration of rehydration does not influence the in vitrodevelopmental ability of the embryos in pigs.Xenografting of testicular tissues into immunodeficientmiceIn mice, germ cells used for transplantation into thetestis can be homolog taken from an unrelated male(Brinster and Avarbock 1994; Brinster and Zimmermann1994). The possibility of spermatogenesis by thismethod has been also suggested in pigs (Honaramoozet al. 2002a). In these cases, germ cells, includingspermatogonial stem cells, were injected directly intothe seminiferous tubules, and it was expected thatspermatozoa from the introduced cells would be producedin the ejaculate. However, this technique requiresspecial skills for the injection and also accurate separationprocedures to select the spermatozoa derived fromthe introduced cells. Germ cell transplantation into micehas produced complete donor-derived spermatogenesisin rodents, but unfortunately it did not succeed in othermammalian species.Xenografting of testicular tissues can be performedinto immunodeficient mice heterotopically, such asunder the skin of the back. The report by Honaramoozet al. (2002b) suggested first the possibility of productionof mammalian sperm in testicular tissues graftedinto nude mice. However, successful embryo productionby using the sperm cells from xenografted testiculartissues has been limited to rhesus monkeys (Honaramoozet al. 2004). In our study (Kikuchi et al. 2006a),testicular tissues from male piglets 6–15 days old weretransplanted under the skin of the backs of castratednude mice 5–8 weeks old. When porcine testes werexenografted into mice, spermatids and spermatozoawere obtained in 70.4% (19 ⁄ 27) of the mice but showedonly faint motility, suggesting the need for ICSI to givetheir ability to fertilize oocytes. When a total of 253oocytes (four replications) were injected with sperm, 63oocytes (24.9% ± 7.1%) developed to the blastocyststage. Their total mean number of cells was 41.9 ± 3.9.These values were comparable to those from in vitrofertilization with frozen-thawed spermatozoa [25.3%and 48.7 cells, respectively (Kikuchi et al. 2002)], showingthe promising ability of the embryos to develop intopiglets after embryo transfer. The results suggest thepossibility of producing embryos with developmentalpotential by using porcine spermatozoa differentiatedfrom the gonocytes within the xenografts.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Advanced Porcine Reproductive Technology 403Other technologies related to ICSIThe establishment of a porcine IVM–ICSI proceduremay help us to generate transgenic pigs by spermmediatedgene transfer. The possibility of mammalianspermatozoa being able to act as vectors for foreignDNA has been reported by Brackett et al. (1971), andsuccessful fertilization by gene-mediated spermatozoahas been reported in mice (Lavitrano et al. 1989) andpigs (Lavitrano et al. 1997, 1999; Lazzereschi et al.2000). This method combined with ICSI has succeededin mice (Perry et al. 1999) and rats (Hirabayashi et al.2005b), and is now expected to be used in othermammalian species, including pigs. In general, theefficacy of the sperm-mediated gene transfer method,as compared with microinjection method into thepronuclei, was reviewed as excellent in the productionof transgenic animals (Smith 1999; Wall 2002). Theefficacy of transgenic pig production may be improvedby using a combination of IVM oocytes and spermmediatedgene transfer by ICSI, as has been suggestedby Kurome et al. (2006).Conservation of Ovarian Resources byXenografting into MicePrimordial follicles act as stores of ovarian follicles andare potential resources of oocytes for medical, agriculturaland zoological purposes. Ovarian grafting is apotential method of maturing the oocytes in theprimordial follicles (primordial oocytes) of large mammals.Grafting of ovarian tissues to another site in thebody (autografting) has been successful in producingviable offspring in humans (Donnez et al. 2004) andother primates (Lee et al. 2004). Cross-species ovariangrafting (xenografting) seems to be more advantageousfor the multiplication and conservation of domestic orendangered animals. Mouse oocytes that grow withinovarian tissue xenografted to nude rats acquire theability to generate pups (Snow et al. 2002). To date,ovarian tissues have been prepared from species phylogeneticallydistant from mice, including humans (Oktayet al. 1998; Weissman et al. 1999; Kim et al. 2002; Gooket al. 2003), dogs (Metcalfe et al. 2001), monkeys(Candy et al. 1995), sheep (Gosden et al. 1994), cows(Senbon et al. 2003), pigs (Kaneko et al. 2003b; Kagawaet al. 2005), tammar wallabies (Mattiske et al. 2002) andcommon wombats (Cleary et al. 2003, 2004), and thenxenografted into immunodeficient mice. To our knowledge,only our previous study (Kaneko et al. 2003b), inwhich neonatal pig ovarian tissues were xenografted,had proven that primordial oocytes can develop in thehost mice and acquire fertilizing ability.In our series of studies (Kaneko et al. 2003b, 2006;Kikuchi et al. 2006b), we have generated viable embryosfrom porcine primordial oocytes xenografted into nudemice. Here, we summarize the recent data and discuss thepossibility of further improvement of this technology.IVM–IVF of porcine oocytes grown in host miceOvarian tissues from 20-day-old piglets, in which mostof the follicles were primordial upon histological evaluation,were transplanted under the capsules of bothkidneys of ovariectomized nude mice (Kaneko et al.2003b). Forty-five to 70 days after grafting, the hostmice in all groups for the first time showed vaginalcornification. They were treated with 5 IU of equinechorionic gonadotropin (eCG) 10 days (eCG-10),30 days (eCG-30), or 60 days (eCG-60) after the detectionof cornified epithelial cells in their vaginal smears.Ovarian grafts and blood samples were obtained 48 hafter eCG treatment. The mice in all groups formedsmall numbers of antral follicles in the grafts. However,we recovered large numbers of full-sized oocytes(‡115 lm in diameter) only from mice in the eCG-60group; the numbers of full-sized oocytes in the othergroups were low. Peripheral levels of total inhibin werehighest in the eCG-60 group; this supports our findingthat the most enhanced growth of antral folliclesoccurred in this eCG-60 group. When the oocytesobtained from the eCG-60 group were matured in vitro,17% were found to be at metaphase II. Moreover, 55%of IVM oocytes with a first polar body were fertilized invitro, resulting in both male and female pronuclearformation. These results clearly demonstrate that fertilizationof oocytes from porcine primordial follicles isachievable by a combination of xenografting and in vitroculture.Developmental ability in vitro and in vivo in recipient miceafter IVF of porcine oocytes grown in most miceIn the next step, we evaluated the developmental abilityof the in vitro fertilized oocytes after transfer torecipients or in vitro culture (Kikuchi et al. 2006b). Asdetermined from the incidence of the first polar body,the maturation rates of the oocytes collected from thehost mice 48 h after treatment with eCG ranged from25.1% to 42.5%. The oocytes were fertilized in vitro andcultured in vitro for 6 days, or transferred into oestrussynchronizedrecipients and recovered after 6 days. Onday 6 of culture, 15.4% of the matured oocytes hadcleaved to the 2- to 8-cell stage. However, neither theembryos cultured in vitro nor those transferred andrecovered developed to advanced embryonic stages,such as morulae or blastocysts. These results suggestthat the developmental ability of xenografted oocytes isinsufficient, even after in vitro maturation.Improvement of developmental ability after treatment ofhost mice with gonadotropinsTo improve the developmental ability of primordialoocytes xenografted to nude mice, we treated the hostmice with gonadotrophins to accelerate folliculargrowth (Kaneko et al. 2006). Gonadotrophin treatmentswere commenced around 60 days after vaginalcornification. Ovarian grafts were obtained 2 or 3 daysafter treatment with eCG (eCG-2 and eCG-3 groups,respectively), after porcine FSH infusion by osmoticpump for 7 or 14 days, or after infusion of porcineFSH for 14 days with a single injection of estradiolantiserum (FSH-7, FSH-14 and FSH-14EA groups,respectively). Gonadotropin treatments accelerated folliculargrowth within the xenografts compared withÓ 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

404 K Kikuchi, N Kashiwazaki, T Nagai, M Nakai, T Somfai, J Noguchi and H Kanekothat in control mice given no gonadotropins, consistentwith markedly higher circulating inhibin levels in thegonadotropin-treated mice. In contrast, circulatingmouse FSH levels were depressed. We recovered largenumbers of full-sized oocytes with meiotic competenceto the mature stage from the eCG-3, FSH-7 and FSH-14EA groups, unlike in the control group. Moreover,56% of IVM oocytes with a first polar body werefertilized in vitro in the FSH-14EA group. After IVFand subsequent IVC for 7 days, one blastocyst wasobtained from each of the eCG-3, FSH-7 and FSH-14EA groups, whereas no blastocysts appeared in theother groups. These results suggest that exogenousgonadotropins – not mouse FSH – stimulated thegrowing follicles that had developed from the primordialfollicles in the xenografts: the effects were incompletebut improved the meiotic and developmentalabilities of the oocytes to some extent.Further improvementOur studies suggest that, in pigs, oocytes derived fromprimordial follicles, even after in vivo growth and IVM,seem to have difficulty achieving cytoplasmic maturation.One of the procedures to enhance embryonicdevelopment is the transfer of metaphase-II chromosomesto an enucleated cytoplast with full developmentalability; this has been successfully reported in mice(Wang et al. 2001). On the contrary, we recentlyreported the ‘Centri-Fusion’ method of nuclear transfer(Fahrudin et al. 2007; Nagai et al. 2007). This mayenable the modification of cytoplasmic maturationalability, even in oocytes from xenografted ovaries.ConclusionPorcine IVP and related technologies are expected to beuseful for the conservation of genetic resources andother related technologies. To increase the reliability ofthese technologies, improvements in the procedures usedare still required. We believe that more basic research toclarify the mechanisms of oocyte maturation, fertilizationand embryonic development will improve thesetechnologies in pigs.AcknowledgementsThis study was supported in part by a Grant-in-Aid for ScientificResearch (17380170) to H.K. from the Japanese Society for Promotionof Science.ReferencesAhmadi A, Ng SC, Liow SL, Ali J, Bongso A, Ratnam SS,1995: Intracytoplasmic sperm injection of mouse oocyteswith 5 mM Ca 2+ at different intervals. Hum Reprod 10,431–435.Brackett BG, Baranska W, Sawicki W, Koprowski H, 1971:Uptake of heterologous genome by mammalian spermatozoaand its transfer to ova through fertilization. Proc NatlAcad Sci U S A 68, 353–357.Brinster RL, Avarbock MR, 1994: Germline transmission ofdonor haplotype following spermatogonial transplantation.Proc Natl Acad Sci U S A 91, 11303–11307.Brinster RL, Zimmermann JW, 1994: Spermatogenesis followingmale germ-cell transplantation. Proc Natl Acad SciUSA91, 11298–11302.Candy CJ, Wood MJ, Whittingham DG, 1995: Folliculardevelopment in cryopreserved marmoset ovarian tissue aftertransplantation. Hum Reprod 10, 2334–2338.Catt SL, Catt JW, Gomez MC, Maxwell WM, Evans G, 1996:Birth of a male lamb derived from an in vitro maturedoocyte fertilised by intracytoplasmic injection of a singlepresumptive male sperm. Vet Rec 139, 494–495.Cleary M, Paris MC, Shaw J, Jenkin G, Trounson A, 2003:Effect of ovariectomy and graft position on cryopreservedcommon wombat (Vombatus ursinus) ovarian tissue followingxenografting to nude mice. Reprod Fertil Dev 15, 333–342.Cleary M, Shaw JM, Jenkin G, Trounson AO, 2004: Influenceof hormone environment and donor age on cryopreservedcommon wombat (Vombatus ursinus) ovarian tissue xenograftedinto nude mice. Reprod Fertil Dev 16, 699–707.Cochran R, Meintjes M, Reggio B, Hylan D, Carter J, PintoC, Paccamonti D, Godke RA, 1998: Live foals producedfrom sperm-injected oocytes derived from pregnant mares.J Equine Vet Sci 18, 736–740.Donnez J, Dolmans MM, Demylle D, Jadoul P, Pirard C,Squifflet J, Martinez-Madrid B, van Langendonckt A, 2004:Live birth after orthotopic transplantation of cryopreservedovarian tissue. Lancet 364, 1405–1410.Fahrudin M, Kikuchi K, Karja NWK, Ozawa M, Somfai T,Ohnuma K, Noguchi J, Kaneko H, Nagai T, 2007:Development to the blastocyst stage of porcine somatic cellnuclear transfer embryos reconstructed by the fusion ofcumulus cells and cytoplasts prepared by gradient centrifugation.Cloning Stem Cells 9, 216–228.Funahashi H, Day BN, 1997: Advances in in vitro productionof pig embryos. J Reprod Fertil Suppl 52, 271–283.Funahashi H, Kim NH, Stumpf TT, Cantley TC, Day BN,1996: Presence of organic osmolytes in maturation mediumenhances cytoplasmic maturation of porcine oocytes. BiolReprod 54, 1412–1419.Funahashi H, Cantley TC, Day BN, 1997: Synchronizationof meiosis in porcine oocytes by exposure to dibutyrylcyclic adenosine monophosphate improves developmentalcompetence following in vitro fertilization. Biol Reprod 57,49–53.Gomez MC, Pope CE, Harris, Davis A, Mikota S, Dresser BL,2000: Births of kittens produced by intracytoplasmic sperminjection of domestic cat oocytes matured in vitro. ReprodFertil Dev 12, 423–433.Gook DA, Edgar DH, Borg J, Archer J, Lutjen PJ, McBainJC, 2003: Oocyte maturation, follicle rupture and luteinizationin human cryopreserved ovarian tissue followingxenografting. Hum Reprod 18, 1772–1781.Gosden RG, Boulton MI, Grant K, Webb R, 1994: Folliculardevelopment from ovarian xenografts in SCID mice.J Reprod Fertil 101, 619–623.Goto K, Kinoshita A, Takuma Y, Ogawa K, 1991: Birth ofcalves after the transfer of oocytes fertilized by sperminjection. Theriogenology 35, 205.Han YM, Abeydeera LR, Kim JH, Moon HB, Cabot RA, DayBN, Prather RS, 1999: Growth retardation of inner cellmass cells in polyspermic porcine embryos produced invitro. Biol Reprod 60, 1110–1113.Hewitson L, Dominko T, Takahashi D, Martinovich C,Ramalho-Santos J, Sutovsky P, Fanton J, Jacob D, MonteithD, Neuringer M, Battaglia D, Simerly C, Schatten G,1999: Unique checkpoints during the first cell cycle offertilization after intracytoplasmic sperm injection in rhesusmonkeys. Nat Med 5, 431–433.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Advanced Porcine Reproductive Technology 405Hewitson L, Martinovich C, Simerly C, Takahashi D, SchattenG, 2002: Rhesus offspring produced by intracytoplasmicinjection of testicular sperm and elongated spermatids.Fertil Steril 77, 794–801.Hirabayashi M, Koto M, Aoto T, Sekimoto A, Ueda M,Miyoshi I, Kasai N, Hochi S, 2002a: Offspring derived fromintracytoplasmic injection of transgenic rat sperm. TransgenicRes 11, 221–228.Hirabayashi M, Kato M, Aoto T, Ueda M, Hochi S, 2002b:Rescue of infertile transgenic rat lines by intracytoplasmicinjection of cryopreserved round spermatids. Mol ReprodDev 62, 295–299.Hirabayashi M, Kato M, Ito J, Hochi S, 2005a: Viable ratoffspring derived from oocytes intracytoplasmically injectedwith freeze-dried sperm heads. Zygote 13, 79–85.Hirabayashi M, Kato M, Ishikawa A, Kaneko R, Yagi T,Hochi S, 2005b: Factors affecting production of transgenicrats by ICSI-mediated DNA transfer: effects of sonicationand freeze-thawing of spermatozoa, rat strains for spermand oocyte donors, and different constructs of exogenousDNA. Mol Reprod Dev 70, 422–428.Hochi S, Watanabe K, Kato M, Hirabayashi M, 2008: Liverats resulting from injection of oocytes with spermatozoadried and stored for one year. Mol Reprod Dev 75, 890–894.Honaramooz A, Megee SO, Dobrinski I, 2002a: Germ celltransplantation in pigs. Biol Reprod 66, 21–28.Honaramooz A, Snedaker A, Boiani M, Scholer H, DobrinskiI, Schlatt S, 2002b: Sperm from neonatal mammalian testesgrafted in mice. Nature 418, 778–781.Honaramooz A, Li MW, Penedo MC, Meyers S, Dobrinski I,2004: Accelerated maturation of primate testis by xenograftinginto mice. Biol Reprod 70, 1500–1503.Kagawa N, Sakurai Y, Miyano T, Manabe N, 2005: Effects oflong-term grafting on follicular growth in porcine ovariancortical grafts xenoplanted to severe combined immunodeficient(SCID) mice. J Reprod Dev 51, 77–85.Kaneko T, Nakagata N, 2006: Improvement in the long-termstability of freeze-dried mouse spermatozoa by adding of achelating agent. Cryobiology 53, 279–282.Kaneko T, Whittingham DG, Overstreet JW, Yanagimachi R,2003a: Tolerance of the mouse sperm nuclei to freeze-dryingdepends on their disulphide status. Biol Reprod 69, 1859–1862.Kaneko H, Kikuchi K, Noguchi J, Hosoe M, Akita T, 2003b:Maturation and fertilization of porcine oocytes fromprimordial follicles by a combination of xenografting andin vitro culture. Biol Reprod 69, 1488–1493.Kaneko H, Kikuchi K, Noguchi J, Ozawa M, Ohnuma K,Maedomari N, Kashiwazaki N, 2006: Effects of gonadotropintreatments on meiotic and developmental competence ofoocytes in porcine primordial follicles following xenograftingto nude mice. Reproduction 131, 279–288.Kawase Y, Hani T, Kamada N, Jishage K, Suzuki H, 2007:Effects of pressure at primary drying of freeze-drying mousereproduction ability and preservation potential. Reproduction133, 841–846.Kikuchi K, Nagai T, Ding J, Yamauchi N, Noguchi J, IzaikeY, 1999: Cytoplasmic maturation for activation of pigfollicular oocytes cultured and arrested at metaphase-I. JReprod Fertil 116, 143–156.Kikuchi K, Onishi A, Kashiwazaki N, Iwamoto M, Noguchi J,Kaneko H, Akita T, Nagai T, 2002: Successful pigletproduction after transfer of blastocysts produced by amodified in vitro system. Biol Reprod 66, 1033–1041.Kikuchi K, Nakai M, Kashiwazaki N, Ozawa M, MaedomariN, Noguchi J, Ohnuma K, Kaneko H, 2006a: Porcineblastocysts derived from in vitro matured oocytes injectedintracytoplasmically with a sperm from testicular tissuexenografted into nude mice. Reprod Fertil Dev 18, 247.Kikuchi K, Kaneko H, Nakai M, Noguchi J, Ozawa M,Ohnuma K, Kashiwazaki N, 2006b: In vitro and in vivodevelopmental ability of oocytes derived from porcineprimordial follicles xenografted into nude mice. J ReprodDev 52, 51–57.Kim NH, Shin JS, Kim C, Jun SH, Lee HT, Chung KS, 1999:Fertilization and in vitro development of porcine oocytesfollowing intracytoplasmic injection of round spermatid orround spermatid nuclei. Theriogenology 51, 1441–1449.Kim SS, Soules MR, Battaglia DE, 2002: Follicular development,ovulation, and corpus luteum formation in cryopreservedhuman ovarian tissue after xenotransplantation.Fertil Steril 78, 77–82.Kimura Y, Yanagimachi R, 1995a: Intracytoplasmic sperminjection in the mouse. Biol Reprod 52, 709–720.Kimura Y, Yanagimachi R, 1995b: Development of normalmice from oocytes injected with secondary spermatocytenuclei. Biol Reprod 53, 855–862.Kimura Y, Tateno H, Handel MA, Yanagimachi R, 1998:Factors affecting meiotic and developmental competence ofprimary spermatocyte nuclei injected into mouse oocytes.Biol Reprod 59, 871–877.Kolbe T, Holtz W, 2000: Birth of a piglet derived from anoocyte fertilized by intracytoplasmic sperm injection (ICSI).Anim Reprod Sci 64, 97–101.Kurome M, Ueda H, Tomii R, Naruse K, Nagashima H, 2006:Production of transgenic-clone pigs by the combination ofICSI-mediated gene transfer with somatic cell nucleartransfer. Transgenic Res 15, 229–240.Lacham-Kaplan O, Trounson A, 1995: Intracytoplasmicsperm injection in mice: increased fertilization and developmentto term after induction of the acrosome reaction. HumReprod 10, 2462–2469.Lavitrano M, Camaioni A, Fazio VM, Dolci S, Farace MG,Spadafora C, 1989: Sperm cells as vectors for introducingforeign DNA into eggs: genetic transformation of mice. Cell57, 717–723.Lavitrano M, Forni M, Varzi V, Pucci L, Bacci ML, DiStefano C, Fioretti D, Zoraqi G, Moioli B, Rossi M,Lazzereschi D, Stoppacciaro A, Seren E, Alfani D, CortesiniR, Frati L, 1997: Sperm mediated gene transfer: productionof pigs transgenic for a human regulator of complementactivation. Transplant Res 29, 3508–3509.Lavitrano M, Stoppacciaro A, Bacci ML, Forni M, Fioretti D,Pucci L, Di Stefano C, Lazzereschi D, Rughetti A, CerettaS, Zannoni A, Rahimi H, Moioli B, Rossi M, Nuti M, RossiG, Seren E, Alfani D, Cortesini R, Frati L, 1999: Humandecay accelerating factor transgenic pigs for xenotransplantationobtained by sperm-mediated gene transfer. TransplantRes 31, 927–974.Lazzereschi D, Forni M, Cappello F, Bacci ML, Di Stefano C,Marfe G, Giancotti P, Renzi L, Wang HJ, Rossi M, DellaCasa G, Pretagostini R, Frati G, Bruzzone P, Stassi G,Stoppacciaro A, Turchi V, Cortesini R, Sinibaldi P, Frati L,Lavitrano M, 2000: Efficiency of transgenesis using spermmediatedgene transfer: generation of hDAF transgenic pigs.Transplant Res 32, 892–894.Lee DM, Yeoman RR, Battaglia DE, Stouffer RL, Zelinski-Wooten MB, Fanton JW, Wolf DP, 2004: Live birth afterovarian tissue transplant. Nature 428, 137–138.Li GP, Chen DY, Lian L, Sun QY, Wang MK, Liu JL, Li JS,Han ZM, 2001: Viable rabbits derived from reconstructedoocytes by germinal vesicle transfer after intracytoplasmicsperm injection (ICSI). Mol Reprod Dev 58, 180–185.Liu JL, Kusakabe H, Chang CC, Suzuki H, Schmidt DW,Julian M, Pfeffer R, Bormann CL, Tian XC, YanagimachiR, Yang X, 2004: Freeze-dried sperm fertilization leads tofull-term development in rabbits. Biol Reprod 70, 1776–1781.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

406 K Kikuchi, N Kashiwazaki, T Nagai, M Nakai, T Somfai, J Noguchi and H KanekoMarchal R, Feugang JM, Perreau C, Venturi E, Terqui M,Miemillod P, 2001: Meiotic and developmental competenceof prepubertal and adult swine oocytes. Theriogenology 56,17–29.Martin MJ, 2000: Development of in vivo-matured porcineoocytes following intracytoplasmic sperm injection. BiolReprod 63, 109–112.Mattioli M, Bacci ML, Galeati G, Seren E, 1989: Developmentalcompetence of pig oocytes matured and fertilized invitro. Theriogenology 39, 1201–1207.Mattiske D, Shaw G, Shaw JM, 2002: Influence of donor ageon development of gonadal tissue from pouch young of thetammar wallaby, Macropus eugenii, after cryopreservationand xenografting into mice. Reproduction 123, 143–153.McCauley TC, Mazza MR, Didion BA, Mao J, Wu G,Coppola G, Coppola GF, Di Berardino D, Day BN, 2003:Chromosomal abnormalities in Day-6, in vitro-produced pigembryos. Theriogenology 60, 1569–1580.Metcalfe SS, Shaw JM, Gunn IM, 2001: Xenografting ofcanine ovarian tissue to ovariectomized severe combinedimmunodeficient (SCID) mice. J Reprod Fertil Suppl 57,323–329.Nagai T, 1996: In vitro maturation and fertilization of pigoocytes. Anim Reprod Sci 42, 153–163.Nagai T, Kikuchi K, Imai K, Fahrudin M, 2007: Handmadesomatic cell cloning and its related studies in farm animals. JMammal Ova Res 24, 99–106.Nagy ZP, Cecile J, Liu J, Loccufier A, Devroey P, VanSteirteghem A, 1996: Pregnancy and birth after intracytoplasmicsperm injection of in vitro matured germinal-vesiclestage oocytes: case report. Fertil Steril 65, 1047–1050.Nakai M, Kashiwazaski N, Takizawa A, Hayashi Y, NakatsukasaE, Fuchimoto D, Noguchi J, Kaneko H, Shino M,Kikuchi K, 2003: Viable piglets generated from porcineoocytes matured in vitro and fertilized by intracytoplasmicsperm head injection. Biol Reprod 68, 1003–1008.Nakai M, Kashiwazaki N, Takizawa A, Maedomari N, OzawaM, Noguchi J, Kaneko H, Shino M, Kikuchi K, 2006:Morphological changes in boar sperm nuclei with reduceddisulfide bonds in electrostimulated porcine oocytes. Reproduction131, 603–611.Nakai M, Kashiwazaki N, Takizawa A, Maedomari N, OzawaO, Noguchi J, Kaneko H, Shino M, Kikuchi K, 2007:Effects of chelating agents during freeze-drying of boarspermatozoa on DNA fragmentation and on developmentalability in vitro and in vivo after intracytoplasmic sperm headinjection. Zygote 15, 15–24.Ogura A, Matsuda J, Yanagimachi R, 1994: Birth of normalyoung after electrofusion of mouse oocytes with roundspermatids. Proc Natl Acad Sci U S A 91, 7460–7462.Ogura A, Suzuki O, Tanemura K, Mochida K, Kobayashi Y,Matsuda J, 1998: Development of normal mice frommetaphase I oocytes fertilized with primary spermatocytes.Proc Natl Acad Sci U S A 95, 5611–5615.Oktay K, Newton H, Mullan J, Gosden RG, 1998: Developmentof human primordial follicles to antral stages inSCID ⁄ hpg mice stimulated with follicle stimulating hormone.Hum Reprod 13, 1133–1138.Palermo G, Joris H, Devroey P, Van Steirteghem AC, 1992:Pregnancies after intracytoplasmic injection of single spermatozooninto an oocyte. Lancet 340, 17–18.Perry AC, Wakayama T, Kishikawa H, Kasai T, Okabe M,Toyoda Y, Yanagimachi R, 1999: Mammalian transgenesisby intracytoplasmic sperm injection. Science 284, 1180–1183.Sasagawa I, Kuretake S, Eppig JJ, Yanagimachi R, 1998:Mouse primary spermatocytes can complete two meioticdivisions within the oocyte cytoplasm. Biol Reprod 58, 248–254.Senbon S, Ota A, Tachibana M, Miyano T, 2003: Bovineoocytes in secondary follicles grow and acquire meioticcompetence in severe combined immunodeficient mice.Zygote 11, 139–149.Smith KR, 1999: Sperm cell mediated transgenesis: a review.Anim Biotechnol 10, 1–13.Snow M, Cox S-L, Jenkin G, Trounson A, Shaw J, 2002:Generation of live young from xenografted mouse ovaries.Science 297, 2227.Somfai T, Kikuchi K, Medvedev SU, Onishi A, Iwamoto M,Fuchimoto D, Ozawa M, Noguchi N, Kaneko H, OhnumaK, Sato E, Nagai T, 2005: Development to the blastocyststage of immature pig oocytes arrested before the metaphase-IIstage and fertilized in vitro. Anim Reprod Sci 90,307–328.Somfai T, Ozawa M, Noguchi J, Kaneko H, Ohnuma K,Karja NWK, Farhudin M, Maedomari N, Dinnye´s A,Nagai T, Kikuchi K, 2006: Diploid porcine parthenotesproduced by the inhibition of first polar body extrusionduring in vitro maturation of follicular oocytes. Reproduction132, 559–570.Somfai T, Ozawa M, Noguchi J, Kaneko H, Karja NWK,Fahrudin M, Nakai M, Maedomari N, Dinnye´s A, Nagai T,Kikuchi K(in press): In vitro development of polyspermicporcine oocytes: relationship between early fragmentationand excessive number of penetrating spermatozoa. AnimReprod Sci 117, 131–147.Uehara T, Yanagimachi R, 1976: Microsurgical injection ofspermatozoa into hamster eggs with subsequent transformationof sperm nuclei into male pronuclei. Biol Reprod 15,467–470.Uehara T, Yanagimachi R, 1977: Behavior of nuclei oftesticular, caput and cauda epididymal spermatozoa injectedinto hamster eggs. Biol Reprod 16, 315–321.Wakayama T, Yanagimachi R, 1998: Development of normalmice from oocytes injected with freeze-dried spermatozoa.Nat Biotechnol 16, 639–641.Wall RJ, 2002: New gene transfer methods. Theriogenology57, 189–201.Wang MK, Chen DY, Lui JL, Li GP, Sun QY, 2001: In vitrofertilisation of mouse oocytes reconstructed by transfer ofmetaphase-II chromosomes results in live births. Zygote 9,9–14.Ward MA, Kaneko T, Kusakabe H, Biggers JD, WhittinghamDG, Yanagimachi R, 2003: Long-term preservation ofmouse spermatozoa after freeze-drying and freezing withoutcryoprotection. Biol Reprod 69, 2100–2108.Weissman A, Gotlieb L, Colgan T, Jurisicova A, GreenblattEM, Casper RF, 1999: Preliminary experience with subcutaneoushuman ovarian cortex transplantation in the NOD-SCID mouse. Biol Reprod 60, 1462–1467.Yamauchi Y, Yanagimachi R, Horiuchi T, 2002: Full-termdevelopment of golden hamster oocytes following intracytoplasmicsperm head injection. Biol Reprod 67, 534–553.Yoshida M, Mizoguchi Y, Ishiaki K, Kojima T, Nagai T,1993: Birth of piglets derived from in vitro fertilization ofpig oocytes matured in vitro. Theriogenology 39, 1303–1311.Author’s address (for correspondence): Kazuhiro Kikuchi, Division ofAnimal Sciences, National Institute of Agrobiological Sciences,Kannondai 2-1-2, Tsukuba, Ibaraki 305-8602, Japan. E-mail:kiku@affrc.go.jpConflict of interest: All authors declare no conflict of interests.Ó 2008 The Authors. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 407–416 (2008); doi: 10.1111/j.1439-0531.2008.01192.xISSN 0936-6768Climbing Mount Efficiency – Small Steps, Not Giant Leaps Towards Higher CloningSuccess in Farm AnimalsBjo¨rn ObackAgResearch Ltd., Ruakura Research Centre, Hamilton, New ZealandContentsDespite more than a decade of research efforts, farm animalcloning by somatic cell nuclear transfer (SCNT) is stillfrustratingly inefficient. Inefficiency manifests itself at differentlevels, which are currently not well integrated. At the molecularlevel, it leads to widespread genetic, epigenetic andtranscriptional aberrations in cloned embryos. At the organismallevel, these genome-wide abnormalities compromisedevelopment of cloned foetuses and offspring. Specific moleculardefects need to be causally linked to specific clonedphenotypes, in order to design specific treatments to correctthem. Cloning efficiency depends on the ability of the nucleardonor cell to be fully reprogrammed into an embryonic stateand the ability of the enucleated recipient cell to carry out thereprogramming reactions. It has been postulated that reprogrammabilityof the somatic donor cell epigenome is influencedby its differentiation status. However, directcomparisons between cells of divergent differentiation statuswithin several somatic lineages have found no conclusiveevidence for this. Choosing somatic stem cells as donors hasnot improved cloning efficiency, indicating that donor cell typemay be less critical for cloning success. Different recipient cells,on the otherhand, vary in their reprogramming ability. Inbovine, using zygotes instead of oocytes has increased cloningsuccess. Other improvements in livestock cloning efficiencyinclude better coordinating donor cell type with cell cycle stageand aggregating cloned embryos. In the future, it will beimportant to demonstrate if these small increases at every stepare cumulative, adding up to an integrated cloning protocolwith greatly improved efficiency.The Importance of Farm Animal CloningIn more than a decade since the birth of Dolly the sheep,cloned offspring have been produced by somatic cellnuclear transfer (SCNT) in 18 mammalian species.Despite this ever growing list, SCNT remains veryinefficient compared with other assisted reproductivetechnologies such as in vitro fertilization (IVF) orartificial insemination. Typically, cloning efficiency,quantified as the proportion of all embryos transferredinto surrogate mothers that develop into viable offspring,is about 1%–5% (Oback and Wells 2007a). Overthree-quarters of all cloning laboratories are working onfarm animals (cattle, pig, goat, sheep, buffalo and deer),illustrating that the main objective behind SCNT is stillcommercially driven – namely to multiply elite animalswith desired phenotypic traits and to produce geneticallymodified animals (Oback and Wells 2007a). As aconsequence of increased research effort and fundingin the area, the total number of cloning publications hasincreased by an order of magnitude in the past decade,and still continues to grow (Fig. 1). Cattle SCNT haslong dominated the NT publication record, accountingfor an annual average of about 25% of PubMed-listedpapers since 1994. Pig is the second most importantcloned farm animal by this measure (13% of NTpublications), followed by goat, sheep, buffalo and reddeer (altogether 6%). Overall, farm animal cloning thusaccounts for 44% of cloning publications, laboratoryanimals (mouse, rabbit, monkey and rat) for 22%, otherspecies (including human) for 16% and general reviewarticles, which are not species-specific, for the remaining18%. Based solely on past research investment andoutput, i.e. the number of labs involved and theirpublications, cattle is still the most important clonedlivestock species (Oback and Wells 2007a).Nuclear ReprogrammingAfter NT of a fully differentiated donor cell into acytoplast, the resulting reconstruct can develop into anembryo and even a viable animal. The logical alternative,i.e. that the NT reconstruct cleaves into fullydifferentiated donor cells, has never been observed.Erasing transcriptional programme and epigenetic identityof the donor cell is referred to as nuclear reprogramming.The molecular dominance of the oocyte overany somatic cell type tested may simply be due to itbeing a 1000-fold larger in volume and thus containing a1000-fold excess of oocyte-specific factors, in which casethe reprogramming dominance should disappear oncecell size differences are experimentally adjusted. This issupported by the observation that nuclear reprogrammingalso occurs in differentiated cells fused to nondividingmultinucleate heterokaryons (Blau et al. 1983;Terranova et al. 2006) with the direction of reprogrammingbeing dictated by the ratio of the nuclei derivedfrom each cell type (Pavlath and Blau 1986). Thecapacity to reverse stable heritable epigenetic modifications,such as DNA-methylation, is not particular tooocytes, but also occurs in embryonic stem (ES) cells(Tada et al. 2003) and even fully differentiated skeletalmuscle cells (Zhang et al. 2007). However, so far onlyoocytes have been capable of reprogramming somaticcells to the extent of giving rise to a completely newcloned organism. Clone survival into adulthood is thusthe most informative and meaningful measure of extensivedonor cell reprogramming. Reprogramming efficiencyafter NT critically depends on two processes: theability of the nuclear donor cell to be fully reprogrammedand the ability of the oocyte to carry out thereprogramming reactions. As it is currently unclearwhich process is more important for reprogrammingsuccess, both will be discussed in this review.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

408 B ObackNumber of publications200180160140120100806040200sheep cattle goatpigred deer,buffalo1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Publication yearFig. 1. Increase in cloning-relatedpublications over time. Arrowspoint to the year when the firstclone of the indicated farm animalwas published. The number ofpeer-reviewed publications wasbased on searching PubMed, Webof Science Ò [v3.0], and conferenceproceedingsReprogrammability of the Donor CellA hypothesis has been emerging that epigenetic reprogrammingafter NT and cloning efficiency are inverselycorrelated with donor cell differentiation (Oback andWells 2007b). This hypothesis is mainly supported bythree lines of evidence from comparative mouse cloningexperiments using: (i) progressively advanced blastomeredonor nuclei from early-cleavage stages, (ii) EScells and (iii) terminally differentiated lymphocytes andneurons. These comparisons demonstrated that earlyblastomeres result in much higher cloning efficiency thansomatic cells (Hiiragi and Solter 2005). However, theyfailed to conclusively determine whether differentiationstatus significantly affects cloning efficiency withinsomatic donor cell lineages (Oback and Wells 2007b).Terminally differentiated cells vs terminally differentiatedcellsThe first study that compared NT cloning efficiency oftwo epigenetically distinct donor cell subpopulationswithin the same somatic lineage was performed inlymphocytes. It compared two fully differentiated T celltypes, namely natural killer T (NKT) lymphocytes andhelper T cells. The two lymphocyte populations wereisolated from the same strain of male mice, purified to>98% using fluorescence-activated cell sorting and usedfreshly for NT. As shown before (Hochedlinger andJaenisch 2002), cloning from T lymphocytes wasextremely inefficient. NKT cells supported early postimplantationdevelopment (60% vs 7% to implantation)much better than helper T cells; however, cloningefficiency was not significantly different (1.5% vs 0%;Fig. 2). The lymphocyte data confirmed that terminallydifferentiated cells can be reprogrammed, but providedno comparison for the efficiency of this process with lessdifferentiated cells from the same lineage, e.g. haematopoieticstem or precursor cells.Progenitor cells vs terminally differentiated cellsCells of divergent differentiation status from the samesomatic lineage were first compared in mouse, butagain the results were not conclusive (Yamazaki et al.2001). Undifferentiated neural progenitors and moredifferentiated immature young neurons from the foetalcerebral cortex resulted in live offspring with differentefficiencies (12% vs 2%; Fig. 2); however, numberswere too small to be significant. In order to directly testthe hypothesis in farm animals, we used increasingly% cloning efficiency20 antlerblood18muscle16neural14skin121086420ASCadipocyteHSCgranulocyteNKT cellT cellmuscle fibroblastMPCmyotubeNSCNPCimmatureneuronmatureneuron 1KSC (male)KPC (male)Antler Blood Muscle Neural SkinIncreasing differentiationKSC (female)KPC (female)Fig. 2. Somatic cloning efficiencyand donor cell type. Cell typesfrom different lineages are listed bydegree of differentiation. Cloningefficiency is measured as theproportion of surviving offspringper number of embryos transferred;1 neuronal clones werescored for normal development atmid-gestation. ASC, antler stemcell; HSC, haematopoietic stemcell; NKT, natural killer T cell;MPC, muscle precursor cell; NSC,neural stem cell; KSC, keratinocytestem cell; KPC, keratinocyteprecursor cell. For mouse NT(blood, neural and skin cells), datafrom different research groupsusing the same cell type and cellcycle stage but different genotypeswere pooledÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Steps to Improve Farm Animal Cloning Efficiency 409differentiated donors from the bovine skeletal musclelineage, namely foetal myogenic precursors and theirin vitro differentiated progeny, post-mitotic mononucleatedmyotubes (Green et al. 2007). Several factorsserved as early (PAX7 and MYOD1) and late (MYOG,MYF6 and MHY) markers of myogenic determinationand differentiation, respectively. Rather than relying onantigen profiling of bulk populations, we determinedthe purity of individually size-selected cells actuallyprepared for NT. After selection, single precursors wereessentially all MYOD1- and PAX7-positive as well asMYOG- and MYH-negative, whereas myotubes werePAX7-negative, weakly MYOD1- and stronglyMYOG-positive and already expressed the late differentiationmarker MHY. Following NT, muscle-specificgenes PAX7, MYOD1, MYOG and MYF6, weresilenced and remained undetectable up to the blastocyststage. Despite significant differences in development toblastocysts, in vivo viability of those cloned embryoswas not dependant on donor cell type or differentiationstatus.Adult stem cells vs terminally differentiated cellsWe then compared cells from either end of the somaticdifferentiation continuum, namely undifferentiated antlerstem cells (ASCs) and their terminally differentiatedprogeny, adipocytes (Berg et al. 2007). Antler stemcells are anatomically and histologically well definedand capable of proliferation, lifelong self-renewal andproduction of differentiating daughter cells. In vivo,they give rise to all different antler lineages (e.g.skin, blood, nerve, cartilage, bone and connectivetissue), even after ectopic transplantation into mice(Li et al. 2001). In vitro, ASCs are highly proliferative,giving rise to cell lines that can be differentiated intoseveral mesodermal cell lineages, such as bone, cartilageand even adipocytes, a cell type which they wouldnot normally form in vivo (Berg et al. 2007). FollowingNT, antlerogenic (POU5F1 and PTHLH) as well asadipogenic markers (COL1A2, PPARG and GAPDH)were successfully reprogrammed in cloned blastocysts.Despite differences prior to NT, transcript abundanceof donor-specific markers did not differ significantly incloned blastocysts from both cell types, indicating thattranscriptional reprogramming was independent of thedonor cell source. In keeping with the similar degree ofgene expression reprogramming for ASCs vs adipocytes,there were also no differences in development tothe blastocyst stage. Importantly, survival to weaningwas not significantly different between ASCs andadipocytes. This finding was consistent with previousresults in bovine using mesenchymal stem cells withoutany obvious increase in cloning efficiency (Kato et al.2004). The farm animal data are also supported by anumber of comparative studies in mouse using stemcells isolated from haematopoietic (Inoue et al. 2005,2006), mesenchymal (Inoue et al. 2007), neuronal(Yamazaki et al. 2001) and skin (Li et al. 2007)lineages. None of these studies have found anysignificant evidence that adult stem cells result inhigher cloning efficiency than differentiated cells fromeither the same or some unrelated lineage (Fig. 2). Infact, at least for haematopoietic and mesenchymal stemcells, cloning efficiencies were among the lowestreported for any somatic donor cell type (Inoue et al.2006, 2007).Differentiated vs embryonic stem cellsEmbryonic stem cells are immortal pluripotent (able togive rise to all cell types of the embryo proper) cellsderived from placing blastomeres into in vitro culture(Evans and Kaufman 1981; Martin 1981). In mouse,high ES-cell cloning efficiencies have been reported(Rideout et al. 2000); however, these are matched bycertain somatic cell types, provided they share the samegenotype, sex and cell cycle stage (Oback and Wells2007b). All attempts to derive livestock bona fide EScells have failed so far. However, long-term ‘ES-like cell’cultures in farm animal species resemble murine ES cellsin morphology, antigen profile, extensive in vitro differentiationability even after long-term culture andsomatic cell chimerism (Renard et al. 2007). Cloningefficiency with ES-like cells appears to be relatively high(Saito et al. 2003). Because ES cells can be differentiatedinto a variety of cell types, it would be instructive tocompare their cloning efficiency with that of theirdifferentiated progeny. Given the limited availabilityof ES cells in livestock, this experiment would perhapsbest be conducted in mouse. Initial experiments alongthose lines demonstrated that embryos cloned from EScellderived neural stem cells (NSCs) showed muchhigher cloning efficiency than differentiated neurons asdonors (Blelloch et al. 2006). However, cloning efficiencywas quantified by the frequency of deriving EScell lines from NT blastocysts. Repeating such anexperiment and measuring survival into adulthoodmight clarify whether two epigenotypes from each endof the differentiation spectrum show any differences inreprogrammability. Early embryos are not the onlysource for derivation of pluripotent cells. Multipotentialspermatogonial stem cells, isolated from mouse testis,also result in germline transmission (Kanatsu-Shinoharaet al. 2004) and would make interesting donor candidates.Induced pluripotent stem (iPS) cells, obtainedafter transduction of somatic cells with a core set oftranscription factors (Takahashi and Yamanaka 2006),may be even more similar to ES cells and theirapplicability for livestock cloning will be very importantto evaluate.Cell cycle coordinationCloned calves have been generated from serum-starved(Kato et al. 1998; Wells et al. 1999), proliferating(Cibelli et al. 1998; Kasinathan et al. 2001) and mitoticallyarrested somatic donor cells (Tani et al. 2001;Heyman et al. 2002b). For non-transgenic fibroblasts,the overall output of viable calves at weaning wassignificantly higher with serum-starved than with G 1cells (Wells et al. 2003). For transgenic fibroblasts,however, the results were reversed (Wells et al. 2003).This suggests that it may be necessary to coordinatedonor cell type and cell cycle stage in order to maximizeoverall cloning efficiency.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

410 B ObackReprogramming Ability of the Recipient CellWhile there are over 200 donor cell types to choosefrom, choices for the recipient cell for NT are muchmore limited. The most commonly used type is harvestedas non-activated MI oocytes from follicles ofslaughtered adult animals. The maturation into MIIoocytes can occur in vivo or in vitro and both types havebeen successfully used for cloning. In mice, naturallyovulated oocytes improve cloning efficiency comparedwith superovulated ones (Hiiragi and Solter 2005);however, using them in mono-ovulatory species, suchas cattle, is impractical. In contrast to the donor, geneticfactors should be less important for choosing therecipient cell because its nuclear genome is removedprior to NT. However, genotype affects transcriptomeand proteome of the oocyte. The fact that hybrid vigouris beneficial for recipient cells (Gao et al. 2004) and thatearly transcriptional activation varies in clones reconstructedwith oocytes from different mouse strainsemphasizes the importance of the recipient cell’s genotypeduring cloning (Vassena et al. 2008). Furthercomparative studies are needed to fully evaluate theinfluence of oocyte source and maturation method onlivestock cloning efficiency. Cytoplasts prepared fromartificially activated MII oocytes (early telophase or TII)have also sometimes been used to clone goats (Baguisiet al. 1999) and cattle (Bordignon and Smith 1998;Kurosaka et al. 2002; Bordignon et al. 2003). Presently,there is no evidence that MII and TII cytoplasm differ intheir reprogramming ability.In early mammalian cloning experiments, fertilizedoocytes (zygotes) were successfully used as recipientswith two-cell stage donor cells in mouse (McGrath andSolter 1984; Robl et al. 1986), cattle (Prather and First1990) and pig (Prather et al. 1989). Once more advancedembryonic or even somatic donors were used, enucleatedzygotes repeatedly failed to support early developmentof the reconstituted NT embryo (Wakayama et al.2000). This 20-year-old perception that SCNT successcritically depended on the use of unfertilized oocyteschanged radically when the first cloned cattle (Schurmannet al. 2006) and mice (Greda et al. 2006; Egli et al.2007) were born from NT with somatic and embryonicdonors, respectively, into zygotes.Recipient zygotes can be either in interphase or inmitosis. Interphase zygotes were first reproducibly usedfor cloning mice in 2006 (Greda et al. 2006), perhapspartially confirming much earlier attempts (Illmenseeand Hoppe 1981). They succeeded by developing a newmicromanipulation technique for enucleation. The classicalway of enucleating (McGrath and Solter 1983),removes intact pronuclei including all nuclear componentsand some surrounding cytoplasm. In contrast, thenew technique appears to more selectively remove onlythe pronuclear envelope and its attached chromatin,while leaving other pronuclear components, such asnucleoli, behind. These zygotic cytoplasts supportedfull-term development after NT of eight-cell donor cells.It would be informative to determine which factorsremain in the zygotic cytoplasm after selective enucleationand if these factors would be sufficient to alsoreprogram somatic donor nuclei. However, the latterexperiment is technically challenging because of the cellcycle incompatibility between somatic donor andrecipient cell. Donor–recipient cell cycle coordinationis important to maintain normal ploidy and promotereprogramming, both of which are linked to the level ofcyclin B ⁄ Cdk 1 complex activity (M-Cdk, formerlymaturation-promoting factor) in the recipient (Obackand Wells 2002). Interphase zygotes contain lowM-Cdk-activity, thus an introduced donor nucleus ofany cell cycle stage no longer undergoes efficientchromatin remodelling, completely abolishing developmentpast the eight-cell stage (Tani et al. 2001). In orderto circumvent this problem for SCNT, zygotes have tobe at a stage when they contain enough M-Cdk-activityto disassemble the donor nuclear envelope and allowearly embryo development to proceed. This wasachieved by using zygotes which are either just exiting(i.e. early telophase) or entering mitosis (i.e. metaphase),and thus contain enough residual or newly synthesizedM-Cdk-activity, respectively. With early bovine zygotes,enucleated just 4 h after insemination, post-implantationdevelopment of SCNT embryos was improvedcompared with using chemically activated MII cytoplasts(Schurmann et al. 2006). Pregnancy establishmentwas significantly higher and this initial difference translatedinto about twofold higher development into calvesat weaning (Schurmann et al. 2006). There are severalreasons why sperm-activated oocytes might have beenbeneficial. Firstly, sperm is the natural agent of activation,inducing long-lasting Ca 2+ -oscillations that moreclosely resemble natural fertilization than the singlelarge Ca 2+ -rise triggered by artificial activation (Fissoreet al. 1992). Secondly, sperm delivers factors to the eggthat may improve early embryonic development (Krawetz2005). These include: (i) the centrosome, which isusually removed together with its associated componentsduring enucleation; (ii) some 3000 differentmRNAs and (iii) micro-RNAs, which do not code forproteins but play a role in controlling gene activity.Thirdly, there may be early sperm-derived transcriptsarising in the time interval between sperm–egg fusionand enucleation. Although there is no direct evidence forsuch early paternal transcription, it has been shown thatsome embryonic genome activation occurs shortly afterfertilization in cattle (Memili et al. 1998). The nature ofthese early transcripts and whether they come from thematernal and ⁄ or paternal genome is not known. Allthese molecules may participate in chromatin remodelling,pronuclear formation, establishment of imprints orembryonic genome activation (Krawetz 2005). Theincreased reprogramming ability of early zygotes clearlywarrants a more detailed analysis into their RNA andprotein composition.Soon after the cattle study, the birth of cloned miceafter NT into mitotically arrested late zygotes confirmedtheir suitability as NT recipients (Egli et al. 2007). Usingmetaphase-arrested zygotes before their first cleavage,cloned offspring were obtained after NT with mitoticallyarrested zygotes, 2-, 8- and ES-cells, but not somaticdonors. Because a direct comparison between metaphase-arrestedoocyte and zygote recipients, using donorcells of the same type and cell cycle stage, has not yetÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Steps to Improve Farm Animal Cloning Efficiency 411been reported in mouse, it is currently unknown ifzygotes can also improve cloning efficiency in thisspecies. Taken together, these studies unequivocallydemonstrated that zygotes retain the factors necessaryto completely reprogram embryonic and somatic genomes.These are likely to reside in the pronuclei duringinterphase, being redistributed throughout the cytoplasmduring mitosis.Improving Reprogramming In vitroA number of in vitro approaches have been devised toincrease somatic cloning success. These include treatingdonor cells with pharmacological agents to alter theirepigenetic marks (Enright et al. 2003; Shi et al. 2003),fusing transiently permeabilized cells containing artificiallycondensed chromatin (Sullivan et al. 2004) orusing serial NT (Ono et al. 2001). One of the mostpromising methods has been embryo aggregation. Inmouse, aggregation of two or three NT embryos led tonormalized gene expression and higher cloning efficiencyin some studies (Boiani et al. 2003) but not in others(Yabuuchi et al. 2002). In bovine, we aggregated threeone-cell NT embryos during in vitro culture (Obacket al. 2003; Misica-Turner et al. 2007). Aggregationaffected development of embryonic and somatic clonedembryos differently. In aggregates of embryonic clones,in vitro development was impaired, but the fewdeveloping blastocysts appeared molecularly normal.Theoretically, one would expect an n-fold increase inpost-blastocyst survival using n numbers of aggregationpartners, provided there were no positive or negativeinteractions within the composite blastocyst. Embryoclones were in agreement with this prediction, showingon average 2.5-fold higher cloning efficiency. In otherwords, there was no evidence for complementary interactionsresulting in increased survival beyond whatwould be expected as a direct numerical consequence ofaggregation. A similar increase in term survival ofaggregated embryonic clones has been described before(Peura et al. 1998). In contrast, SCNT aggregatesdeveloped normally in vitro, but the resulting blastocystsshowed reduced POU5F1 expression and no effecton vivo survival, indicating that they were in factcompromised compared with embryo clone aggregates.These differences in timing of developmental failurereveal striking biological differences between embryonicand somatic clones in response to aggregation.Measuring Reprogramming Before EmbryoTransferCloned embryos before transfer display a number ofepigenetic abnormalities such as aberrant DNA-methylation(Kang et al. 2003) and post-translational histonemodifications (Santos et al. 2003; Wang et al. 2007;Yang et al. 2007a,b). It is likely that abnormal epigenotypewill result in aberrant gene expression. At leastfor one gene, a mechanistic link between DNA-methylationand transcriptional errors has indeed been established.This gene encodes the transcription factorPOU5F1, which is essential for regulating embryoniccell pluripotency and has often been used to monitorreprogramming success after NT (Boiani et al. 2002;Munsie et al. 2002; Bortvin et al. 2003; Misica-Turneret al. 2007; Wuensch et al. 2007). In interspecies NTexperiments using Xenopus recipient oocytes and mousedonor thymocytes, selective POU5F1 promoter DNAdemethylationwas causally linked to incorrectlyre-activating POU5F1 transcription (Simonsson andGurdon 2004). Because discrepancies between RNAand protein levels exist (Tian et al. 2004), it remains tobe formally demonstrated that POU5F1 protein expressionwas also aberrant and that this subsequentlycontributed to some of the aberrant organismal phenotypesobserved in clones. Proteins ultimately carry outcellular functions, thus their expression profiling representscellular phenotype better than RNA profiling.However, the proteome with its many low-abundanceproteins, often carrying post-translational modifications,is more difficult to analyze and has consequentlynot been comprehensively profiled in clones (Fig. 3).RNA levels, on the contrary, can be amplified, makingthem easier and cheaper to quantify globally. NumerousNT ES cell-like cellsFig. 3. Genome-wide profiling ofclones at different levels. Shown isnuclear transfer (NT) embryodevelopment from one-cell to blastocyst,which can either be used togenerate ES cell-like cells or clonedoffspring. Bars indicate differentgenome-wide profiling approachesto examine different levels ofcellular information (DNA, RNA,protein, etc.) at different developmentaltime windows. Shadingwithin bars indicates to what degreeeach approach has been used for aparticular developmental interval,ranging from little use (white) toheavy use (dark shading). Arrowsbetween bars indicate the flow ofinformation from genotype tophenotype+NTrecipient donor NT 1-cell NT blastocystGenomeEpigenomeTranscriptomeProteomePhenomeNT animalÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

412 B Obackstudies have shown that the spatiotemporal regulationof gene expression is abnormal in cloned pre-implantationembryos. This applies to both termination of thesomatic and initiation of the embryonic gene expressionprogramme. Analysis of transcriptional reprogramminghas focused on two developmental stages: early-cleavagestages at the time of embryonic genome activation andblastocysts.Repressing donor genesTransient and reversible changes in donor cell-specificgene expression can occur simply as an artefact oftrypsinization (Green et al. 2007), emphasizing theimportance of analyzing cells in the state they are inwhen used for NT. However, even if the initial silencingof some genes reflects their manipulation for NT, manydonor cell-specific mRNAs remain repressed until theblastocyst stage. Microarray data have shown that therecipient ooplasm, which is transcriptionally inactive atthe time of NT (Latham et al. 1992), establishes atranscriptionally repressive state within the donor cellgenome of the early NT reconstruct (Vassena et al.2007). As a result, the transcriptome of cloned murineone-cell embryos was very similar to that of controls.Only 259 (=1.6%) of total mRNAs were differentiallyexpressed between SCNT and fertilized embryos at theone-cell stage, with 53 (=20%) of those being newtranscripts and 206 (=80%) being maternally suppliedmRNAs, which were either stabilized or precociouslydegraded in clones. The few newly transcribed genesthat were overexpressed in SCNT embryos, comparedwith both fertilized and parthenogenetic controls,encoded mostly transcription factors that were alsohighly expressed in the donor cells.Reprogramming is a process that occurs over aprotracted period of time, and the next developmentalstage that was analyzed in great detail is the blastocyst.Expression of several donor-specific candidate genesfrom different lineages, e.g. cumulus (Bortvin et al.2003; Inoue et al. 2006) and haematopoietic in mouse(Inoue et al. 2006), skeletal muscle in bovine (Greenet al. 2007) and antler vs adipocyte in cervine NT (Berget al. 2007) was extinguished by that stage. Likewise,global gene expression comparisons demonstrated thatthe majority of genes was differentially expressedbetween donors and their NT blastocyst derivatives,indicating extensive donor transcriptome reprogramming(Smith et al. 2005). Even two female adult ear skinfibroblast lines that differed in the expression of over3000 transcripts prior to NT resulted in indistinguishablegene expression profiles at the blastocyst stage(Smith et al. 2005). Despite their similar degree of geneexpression reprogramming, these two lines still showedsignificantly different cloning efficiency, suggesting thatglobal transcriptional profiling may be of limited valuewhen screening for ‘good’ vs ‘bad’ donor lines. However,not every gene was properly repressed, leaving asmall set of differentially expressed genes that maintainedtheir expression levels in both donors and theircorresponding blastocysts. These fibroblast genes wereinvolved in various biological functions and it would beinformative to determine if the same genes, or theirassociated chromatin modifications, are generally resistantto silencing, independent of cell type. Other genesthat maintain epigenetic donor cell memory have beendescribed in endoderm and neuroectoderm donors infrog (Ng and Gurdon 2005) and myogenic donors aftermouse (Gao et al. 2003) and frog NT (Ng and Gurdon2008). Apart from altered metabolic demands (Gaoet al. 2003), functional consequences of continued donorcell gene expression in cloned embryos are not known.In summary, the donor cell genome was largelyrepressed by the late one-cell stage, but still detectablewith a small array of transcripts at both early and latepreimplantation stages. This raises the question ifinduced donor genome silencing would improve development.Activating embryo genesFollowing fertilization, the embryo undergoes drasticmorphological changes that culminate in the differentiationof two separate lineages, trophectoderm andinner cell mass. During this process, failure to correctlyactivate embryonic genes at the right time, place andabundance will have detrimental effects on development,as illustrated by a large number of single genemutations that interfere with normal embryogenesis inthe mouse. Cloned embryos are clearly defective inrecapitulating this correct stage-specific gene expression.In mouse, already the first transiently inducedgenes (TIG) transcribed from the embryonic genomewere absent or greatly reduced in cloned two-cellembryos, especially if they were also absent from thedonor cell (Vassena et al. 2007, 2008). Even expressionof those few TIG transcripts already present in thedonor cell was reduced in clones, probably becausethey became silenced after NT and only partially reactivatedat the two-cell stage. Other studies, usingdifferent sets of early markers, also observed suppressionof embryonically activated mRNAs in clonedtwo-cells; however, the donor cell expression was notanalyzed (Suzuki et al. 2006; Inoue et al. 2007). Microarrayanalysis confirmed substantial differences betweenthe transcriptomes of NT vs fertilized or artificiallyactivated embryos at the one-cell and, at an order ofmagnitude greater, the two-cell stages (Vassena et al.2007). At the one-cell stage, 53 (=0.3% of total) newtranscripts were at least twofold differentially expressed(8 down- and 45 upregulated), at the two-cell stage thatnumber increased to 1539 (452 down- and 1087upregulated) of different transcripts relative to controlembryos. In addition, a large number of maternalmRNAs underwent either precocious or delayed degradation.The combined effects of aberrant transcriptionand mRNA handling prominently affected genesinvolved in transcriptional regulation. SCNT embryosmay be predisposed to aberrantly express genes encodingproteins that are responsible for establishing andmaintaining a stable donor cell differentiation status.Their widespread dysregulation likely caused a ‘rippleeffect’ that alters the transcriptome of many otherfunctions including oxidative phosphorylation, membranetransport, and mRNA transport and processing(Vassena et al. 2007).Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Steps to Improve Farm Animal Cloning Efficiency 413In most farm animals, the transfer of early-cleavagestage embryos into oviducts is technically difficult andrecipient animals are expensive. Because high embryowastage during pre-implantation development wouldfurther increase recipient cost, cloned embryos areusually transferred at the blastocyst stage. Being ableto select blastocysts with a high chance of developinginto viable offspring is thus a very important goal incloning research. This has been addressed by a numberof microarray studies comparing bovine NT and IVFblastocysts (Pfister-Genskow et al. 2005; Smith et al.2005; Somers et al. 2006; Beyhan et al. 2007; Long et al.2007; Zhou et al. 2007). Comparing these studies, therewere three striking observations. Firstly, transcriptabundance for over 95% of genes analyzed was similarbetween NT and IVF embryos, suggesting that mostgenes were reprogrammed. Secondly, no distinct set ofconsistently misexpressed genes was found in clonedembryos. Altogether, 281 genes were classified as differentiallyexpressed (169 down- and 112 upregulated).Only four of these genes, mostly housekeeping ribosomaltranscripts, were found in two independentstudies and none of them were consistently up- ordownregulated. Thirdly, average changes in transcriptabundance were only twofold, with less than 3%showing more than fourfold changes. The lack of acommon transcriptional signature can have either technicalor biological reasons. Technically, the studiesdiffered in microarrays and analytical algorithms used,adding cross-platform to the inevitable cross-laboratoryvariation (Bammler et al. 2005). Moreover, transcriptomeprofiling can only be as powerful as the biologicaldefinition of the starting cell material. In that respect,comparative analysis of the data sets was clearlyconfounded by multiple parameter changes, e.g. ingenotype, donor and recipient cell source, activationmethod, culture protocol, morphological grade, developmentalstage and state of blastocysts (fresh vscryopreserved), to name just a few. Despite thesedifferences in experimental nuances, analysis of a totalof 130 NT and 121 IVF embryos across the six studies,either individually or in pools, should have revealedrobust ‘hotspots’ of preferentially affected transcripts ifthey indeed existed. Why have they not been found? Amajor problem is that the intracellular range of transcriptabundance in blastocyst cell types is largelyunknown. In yeast, transcript abundance varies oversix orders of magnitude, ranging from a few hundredcopies per cell for glycolytic mRNAs to one-thousandthper cell for transcripts encoding transcription factorswith critical regulatory roles (Holland 2002). It isentirely plausible that ‘hotspot’ transcripts are expressedat low levels or in a small fraction of the cells studiedand microarrays are simply not sensitive enough todetect them (Evans et al. 2002). Alternatively, thesenegative results may argue for a model of randomizedreprogramming. At the single-cell level, gene expressionlevels already fluctuate randomly under normal conditions(Kaern et al. 2005; Raser and O’Shea 2005).Nuclear reprogramming will add to this intrinsicstochastic variation, increasing the noise and maskingexpression differences at individual gene level. A stochasticmodel is supported by the finding that transcriptlevels of small sets of candidate genes, representing 0.1%of the about 10 000 genes expressed in blastocysts (Zenget al. 2004), vary greatly between individual NTembryos (Bortvin et al. 2003) and cannot distinguishbetween NT and IVF embryos (deCamargo et al. 2005;Smith et al. 2007). Carefully quantifying abundance ofseveral functionally unrelated transcripts in singleembryos revealed that they are independently reprogrammed:one gene can be close to ‘normal’, i.e. close tothe average expression level for this gene across manyembryos, whereas another gene in the same embryowould be aberrantly expressed. These patterns of normalvs aberrantly expressed genes would be different fordifferent embryos, arguing against a deterministic modelof reprogramming (C. Smith, personal communication).Further evidence comes from induced reprogramming iniPS cells. Ectopic expression of key transcription factorsinitiates a slow reprogramming process that depends onstochastic epigenetic events, which sequentially activateES-cell markers, eventually resulting in stable pluripotency(Meissner et al. 2007). However, if reprogrammingwas completely random, it should also switch on(or off) genes that were silent (or active) in both thedonor cell and during normal embryogenesis. Genesfalling in these two categories have not yet beenanalyzed comprehensively.ConclusionsDespite more than a decade of research efforts, routinefarm animal cloning efficiency is still frustratingly low.The molecular cause for this low efficiency has not beenelucidated. There is a wealth of information describingwidespread genetic, epigenetic and transcriptional aberrationsin cloned embryos; however, these genome-wideanalyses at different levels are currently not wellintegrated. Importantly, these observations have yet tobe causally connected with the observed phenotypes incloned foetuses and offspring. Because some clonedphenotypes are highly reproducible in farm animals, e.g.hydroallantois in cattle (Heyman et al. 2002a), it shouldbe possible to establish their underlying molecularcauses and design specific treatments to overcome them.Current methods to improve cloning efficiency are stillfairly unspecific and have at best led to a slight increasein success rates. Choosing undifferentiated somatic stemcells as donors has not improved cloning efficiency infive different lineages, indicating that donor cell typemay not be critical for cloning success. It remains to beseen if the limited number of NT experiments reportedwas simply not sufficient to detect a hierarchical relationbetween cell differentiation and cloning efficiency or ifsuch a relation is not universally true. In any case,varying just one parameter alone, e.g. somatic donor celldifferentiation status, is unlikely to lift cloning success tothe efficiency level of other assisted reproductive technologies,such as IVF or artificial insemination. So far,improvements resulted from better coordination betweendonor cell type and cell cycle stage, the use ofmitotic zygotes as NT recipients and, at least forembryonic clones, the aggregation of cloned embryos(Fig. 4). Each of these improvements optimized one stepof the cloning procedure at a time. In the future, it willÓ 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

414 B ObackFig. 4. Climbing mount efficiency: the path to cloning success. Shownare key steps that improve cloning efficiency, resulting in healthyclones. 1. Coordinating donor cell type and cell cycle stage. 2. Usingzygotes rather than unfertilized oocytes as reprogramming reactors.3. Aggregating cloned embryos during in vitro culturebe important to demonstrate if the small sequentialincreases at every step are cumulative, adding up to oneintegrated cloning protocol with greatly improved efficiency.AcknowledgementsI would like to thank Dr. David Wells for critical reading of themanuscript and all members of our cloning team for excellent technicalsupport in the field and laboratory. This work was supported by theNew Zealand Foundation for Research, Science and Technology andAgResearch.ReferencesBaguisi A, Behboodi E, Melican DT, Pollock JS, DestrempesMM, Cammuso C, Williams JL, Nims SD, Porter CA,Midura P, Palacios MJ, Ayres SL, Denniston RS, HayesML, Ziomek CA, Meade HM, Godke RA, Gavin WG,Overstrom EW, Echelard Y, 1999: Production of goats bysomatic cell nuclear transfer. Nat Biotechnol 17, 456–461.Bammler T, Beyer RP, Bhattacharya S, Boorman GA, BoylesA, Bradford BU, Bumgarner RE, Bushel PR, Chaturvedi K,Choi D, Cunningham ML, Deng S, Dressman HK, FanninRD, Farin FM, Freedman JH, Fry RC, Harper A, HumbleMC, Hurban P, Kavanagh TJ, Kaufmann WK, Kerr KF,Jing L, Lapidus JA, Lasarev MR, Li J, Li YJ, LobenhoferEK, Lu X, Malek RL, Milton S, Nagalla SR, O’Malley JP,Palmer VS, Pattee P, Paules RS, Perou CM, Phillips K, QinLX, Qiu Y, Quigley SD, Rodland M, Rusyn I, Samson LD,Schwartz DA, Shi Y, Shin JL, Sieber SO, Slifer S, SpeerMC, Spencer PS, Sproles DI, Swenberg JA, Suk WA,Sullivan RC, Tian R, Tennant RW, Todd SA, Tucker CJ,Van Houten B, Weis BK, Xuan S, Zarbl H, 2005:Standardizing global gene expression analysis betweenlaboratories and across platforms. Nat Methods 2, 351–356.Berg DK, Li C, Asher G, Wells DN, Oback B, 2007: Red deercloned from antler stem cells and their differentiatedprogeny. Biol Reprod 77, 384–394.Beyhan Z, Ross PJ, Iager AE, Kocabas AM, Cunniff K, RosaGJ, Cibelli JB, 2007: Transcriptional reprogramming ofsomatic cell nuclei during preimplantation development ofcloned bovine embryos. Dev Biol 305, 637–649.Blau HM, Chiu CP, Webster C, 1983: Cytoplasmic activationof human nuclear genes in stable heterocaryons. Cell 32,1171–1180.Blelloch R, Wang Z, Meissner A, Pollard S, Smith A, JaenischR, 2006: Reprogramming efficiency following somatic cellnuclear transfer is influenced by the differentiation andmethylation state of the donor nucleus. Stem Cells 24, 2007–2013.Boiani M, Eckardt S, Leu NA, Scholer HR, McLaughlin KJ,2003: Pluripotency deficit in clones overcome by clone-cloneaggregation: epigenetic complementation? EMBO J 22,5304–5312.Boiani M, Eckardt S, Scholer HR, McLaughlin KJ, 2002: Oct4distribution and level in mouse clones: consequences forpluripotency. Genes Dev 16, 1209–1219.Bordignon V, Keyston R, Lazaris A, Bilodeau AS, Pontes JH,Arnold D, Fecteau G, Keefer C, Smith LC, 2003: Transgeneexpression of green fluorescent protein and germ linetransmission in cloned calves derived from in vitrotransfectedsomatic cells. Biol Reprod 68, 2013–2023.Bordignon V, Smith LC, 1998: Telophase enucleation: animproved method to prepare recipient cytoplasts for use inbovine nuclear transfer. Mol Reprod Dev 49, 29–36.Bortvin A, Eggan K, Skaletsky H, Akutsu H, Berry DL,Yanagimachi R, Page DC, Jaenisch R, 2003: Incompletereactivation of Oct4-related genes in mouse embryos clonedfrom somatic nuclei. Development 130, 1673–1680.Cibelli JB, Stice SL, Golueke PJ, Kane JJ, Jerry J, Blackwell C,Ponce de Leon FA, Robl JM, 1998: Cloned transgeniccalves produced from nonquiescent fetal fibroblasts. Science280, 1256–1258.deCamargo ALS, Powell AM, do Vale Filho VR, Wall RJ,2005: Comparison of gene expression in individual preimplantationbovine embryos produced by in vitro fertilisationor somatic cell nuclear transfer. Reprod Fertil Dev 17,487–496.Egli D, Rosains J, Birkhoff G, Eggan K, 2007: Developmentalreprogramming after chromosome transfer into mitoticmouse zygotes. Nature 447, 679–685.Enright BP, Kubota C, Yang X, Tian XC, 2003: Epigeneticcharacteristics and development of embryos cloned fromdonor cells treated by trichostatin A or 5-aza-2¢-deoxycytidine.Biol Reprod 69, 896–901.Evans MJ, Kaufman MH, 1981: Establishment in culture ofpluripotential cells from mouse embryos. Nature 292, 154–156.Evans SJ, Datson NA, Kabbaj M, Thompson RC, VreugdenhilE, De Kloet ER, Watson SJ, Akil H, 2002: Evaluation ofAffymetrix Gene Chip sensitivity in rat hippocampal tissueusing SAGE analysis. Serial analysis of gene expression. EurJ Neurosci 16, 409–413.Fissore RA, Dobrinsky JR, Balise JJ, Duby RT, Robl JM,1992: Patterns of intracellular Ca2 + concentrations infertilized bovine eggs. Biol Reprod 47, 960–969.Gao S, Chung YG, Williams JW, Riley J, Moley K, LathamKE, 2003: Somatic cell-like features of cloned mouseembryos prepared with cultured myoblast nuclei. BiolReprod 69, 48–56.Gao S, Czirr E, Chung YG, Han Z, Latham KE, 2004: Geneticvariation in oocyte phenotype revealed through parthenogenesisand cloning: correlation with differences in pronuclearepigenetic modification. Biol Reprod 70, 1162–1170.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Steps to Improve Farm Animal Cloning Efficiency 415Greda P, Karasiewicz J, Modlinski JA, 2006: Mouse zygotesas recipients in embryo cloning. Reproduction 132, 741–748.Green AL, Wells DN, Oback B, 2007: Cattle cloned fromincreasingly differentiated muscle cells. Biol Reprod 77,395–406.Heyman Y, Chavatte-Palmer P, LeBourhis D, Camous S,Vignon X, Renard JP, 2002: Frequency and occurrence oflate-gestation losses from cattle cloned embryos. BiolReprod 66, 6–13.Heyman Y, Zhou Q, Lebourhis D, Chavatte-Palmer P, RenardJP, Vignon X, 2002: Novel approaches and hurdles tosomatic cloning in cattle. Cloning Stem Cells 4, 47–55.Hiiragi T, Solter D, 2005: Reprogramming is essential innuclear transfer. Mol Reprod Dev 70, 417–421.Hochedlinger K, Jaenisch R, 2002: Monoclonal mice generatedby nuclear transfer from mature B and T donor cells.Nature 415, 1035–1038.Holland MJ, 2002: Transcript abundance in yeast varies oversix orders of magnitude. J Biol Chem 277, 14363–14366.Illmensee K, Hoppe PC, 1981: Nuclear transplantation in Musmusculus: developmental potential of nuclei from preimplantationembryos. Cell 23, 9–18.Inoue K, Noda S, Ogonuki N, Miki H, Inoue S, Katayama K,Mekada K, Miyoshi H, Ogura A, 2007: Differentialdevelopmental ability of embryos cloned from tissue-specificstem cells. Stem Cells 25, 1279–1285.Inoue K, Ogonuki N, Miki H, Hirose M, Noda S, Kim JM,Aoki F, Miyoshi H, Ogura A, 2006: Inefficient reprogrammingof the hematopoietic stem cell genome followingnuclear transfer. J Cell Sci 119, 1985–1991.Inoue K, Wakao H, Ogonuki N, Miki H, Seino K, Nambu-Wakao R, Noda S, Miyoshi H, Koseki H, Taniguchi M,Ogura A, 2005: Generation of cloned mice by directnuclear transfer from natural killer T cells. Curr Biol 15,1114–1118.Kaern M, Elston TC, Blake WJ, Collins JJ, 2005: Stochasticityin gene expression: from theories to phenotypes. Nat RevGenet 6, 451–464.Kanatsu-Shinohara M, Inoue K, Lee J, Yoshimoto M,Ogonuki N, Miki H, Baba S, Kato T, Kazuki Y, ToyokuniS, Toyoshima M, Niwa O, Oshimura M, Heike T, NakahataT, Ishino F, Ogura A, Shinohara T, 2004: Generation ofpluripotent stem cells from neonatal mouse testis. Cell 119,1001–1012.Kang YK, Lee KK, Han YM, 2003: Reprogramming DNAmethylation in the preimplantation stage: peeping withDolly’s eyes. Curr Opin Cell Biol 15, 290–295.Kasinathan P, Knott JG, Wang Z, Jerry DJ, Robl JM, 2001:Production of calves from G1 fibroblasts. Nat Biotechnol19, 1176–1178.Kato Y, Imabayashi H, Mori T, Tani T, Taniguchi M, HigashiM, Matsumoto M, Umezawa A, Tsunoda Y, 2004: Nucleartransfer of adult bone marrow mesenchymal stem cells:developmental totipotency of tissue-specific stem cells froman adult mammal. Biol Reprod 70, 415–418.Kato Y, Tani T, Sotomaru Y, Kurokawa K, Kato J, DoguchiH, Yasue H, Tsunoda Y, 1998: Eight calves cloned fromsomatic cells of a single adult. Science 282, 2095–2098.Krawetz SA, 2005: Paternal contribution: new insights andfuture challenges. Nat Rev Genet 6, 633–642.Kurosaka S, Nagao Y, Minami N, Yamada M, Imai H, 2002:Dependence of DNA synthesis and in vitro development ofbovine nuclear transfer embryos on the stage of the cell cycleof donor cells and recipient cytoplasts. Biol Reprod 67, 643–647.Latham KE, Solter D, Schultz RM, 1992: Acquisition of atranscriptionally permissive state during the 1-cell stage ofmouse embryogenesis. Dev Biol 149, 457–462.Li C, Harris AJ, Suttie JM, 2001: Tissue interactions andantlerogenesis: new findings revealed by a xenograftapproach. J Exp Zool 290, 18–30.Li J, Greco V, Guasch G, Fuchs E, Mombaerts P, 2007: Micecloned from skin cells. Proc Natl Acad Sci U S A. 104, 2738–2743.Long JE, Cai X, He LQ, 2007: Gene profiling of cattleblastocysts derived from nuclear transfer, in vitro fertilizationand in vivo development based on cDNA library. AnimReprod Sci 100, 243–256.Martin GR, 1981: Isolation of a pluripotent cell line from earlymouse embryos cultured in medium conditioned by teratocarcinomastem cells. Proc Natl Acad Sci U S A 78, 7634–7638.McGrath J, Solter D, 1983: Nuclear transplantation in themouse embryo by microsurgery and cell fusion. Science 220,1300–1302.McGrath J, Solter D, 1984: Inability of mouse blastomerenuclei transferred to enucleated zygotes to support developmentin vitro. Science 226, 1317–1319.Meissner A, Wernig M, Jaenisch R, 2007: Direct reprogrammingof genetically unmodified fibroblasts into pluripotentstem cells. Nat Biotechnol 25, 1177–1181.Memili E, Dominko T, First NL, 1998: Onset of transcriptionin bovine oocytes and preimplantation embryos. MolReprod Dev 51, 36–41.Misica-Turner PM, Oback FC, Eichenlaub M, Wells DN,Oback B, 2007: Aggregating embryonic but not somaticnuclear transfer embryos increases cloning efficiency incattle. Biol Reprod 76, 268–278.Munsie M, O’Brien C, Mountford P, 2002: Transgenic strategyfor demonstrating nuclear reprogramming in the mouse.Cloning Stem Cells 4, 121–130.Ng RK, Gurdon JB, 2005: Epigenetic memory of active genetranscription is inherited through somatic cell nucleartransfer. Proc Natl Acad Sci U S A 102, 1957–1962.Ng RK, Gurdon JB, 2008: Epigenetic memory of an active genestate depends on histone H3.3 incorporation into chromatinin the absence of transcription. Nat Cell Biol 10, 102–109.Oback B, Wells D, 2002: Donor cells for nuclear cloning: manyare called, but few are chosen. Cloning Stem Cells 4, 147–168.Oback B, Wells DN, 2007a: Cloning cattle: the methods in themadness. Adv Exp Med Biol 591, 30–57.Oback B, Wells DN, 2007b: Donor cell differentiation,reprogramming, and cloning efficiency: elusive or illusivecorrelation? Mol Reprod Dev 74, 646–654.Ono Y, Shimozawa N, Ito M, Kono T, 2001: Cloned micefrom fetal fibroblast cells arrested at metaphase by a serialnuclear transfer. Biol Reprod 64, 44–50.Pavlath GK, Blau HM, 1986: Expression of muscle genes inheterokaryons depends on gene dosage. J Cell Biol 102, 124–130.Peura TT, Lewis IM, Trounson AO, 1998: The effect ofrecipient oocyte volume on nuclear transfer in cattle. MolReprod Dev 50, 185–191.Pfister-Genskow M, Myers C, Childs LA, Lacson JC,Patterson T, Betthauser JM, Goueleke PJ, Koppang RW,Lange G, Fisher P, Watt SR, Forsberg EJ, Zheng Y, LenoGH, Schultz RM, Liu B, Chetia C, Yang X, Hoeschele I,Eilertsen KJ, 2005: Identification of differentially expressedgenes in individual bovine preimplantation embryos producedby nuclear transfer: improper reprogramming ofgenes required for development. Biol Reprod 72, 546–555.Prather RS, First NL, 1990: Cloning embryos by nucleartransfer. J Reprod Fertil Suppl 41, 125–134.Prather RS, Sims MM, First NL, 1989: Nuclear transplantationin early pig embryos. Biol Reprod 41, 414–418.Raser JM, O’Shea EK, 2005: Noise in gene expression: origins,consequences, and control. Science 309, 2010–2013.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

416 B ObackRenard JP, Maruotti J, Jouneau A, Vignon X, 2007: Nuclearreprogramming and pluripotency of embryonic cells: Applicationto the isolation of embryonic stem cells in farmanimals. Theriogenology. 68, Suppl 1 S196–S205.Rideout WM, 3rd, Wakayama T, Wutz A, Eggan K, Jackson-Grusby L, Dausman J, Yanagimachi R, Jaenisch R, 2000:Generation of mice from wild-type and targeted ES cells bynuclear cloning. Nat Genet 24, 109–110.Robl JM, Gilligan B, Critser ES, First NL, 1986: Nucleartransplantation in mouse embryos: assessment of recipientcell stage. Biol Reprod 34, 733–739.Saito S, Sawai K, Ugai H, Moriyasu S, Minamihashi A,Yamamoto Y, Hirayama H, Kageyama S, Pan J, Murata T,Kobayashi Y, Obata Y, Yokoyama KK, 2003: Generationof cloned calves and transgenic chimeric embryos frombovine embryonic stem-like cells. Biochem Biophys ResCommun 309, 104–113.Santos F, Zakhartchenko V, Stojkovic M, Peters A, JenuweinT, Wolf E, Reik W, Dean W, 2003: Epigenetic markingcorrelates with developmental potential in cloned bovinepreimplantation embryos. Curr Biol 13, 1116–1121.Schurmann A, Wells DN, Oback B, 2006: Early zygotes aresuitable recipients for bovine somatic nuclear transfer andresult in cloned offspring. Reproduction 132, 839–848.Shi W, Hoeflich A, Flaswinkel H, Stojkovic M, Wolf E,Zakhartchenko V, 2003: Induction of a senescent-likephenotype does not confer the ability of bovine immortalcells to support the development of nuclear transferembryos. Biol Reprod 69, 301–309.Simonsson S, Gurdon J, 2004: DNA demethylation isnecessary for the epigenetic reprogramming of somatic cellnuclei. Nat Cell Biol 6, 984–990.Smith C, Berg D, Beaumont S, Standley NT, Wells DN,Pfeffer PL, 2007: Simultaneous gene quantitation of multiplegenes in individual bovine nuclear transfer blastocysts.Reproduction 133, 231–242.Smith SL, Everts RE, Tian XC, Du F, Sung LY, Rodriguez-Zas SL, Jeong BS, Renard JP, Lewin HA, Yang X, 2005:Global gene expression profiles reveal significant nuclearreprogramming by the blastocyst stage after cloning. ProcNatl Acad Sci U S A 102, 17582–17587.Somers J, Smith C, Donnison M, Wells DN, Henderson H,McLeay L, Pfeffer PL, 2006: Gene expression profiling ofindividual bovine nuclear transfer blastocysts. Reproduction131, 1073–1084.Sullivan EJ, Kasinathan S, Kasinathan P, Robl JM, Collas P,2004: Cloned calves from chromatin remodeled in vitro. BiolReprod 70, 146–153.Suzuki T, Minami N, Kono T, Imai H, 2006: Zygoticallyactivated genes are suppressed in mouse nuclear transferredembryos. Cloning Stem Cells 8, 295–304.Tada M, Morizane A, Kimura H, Kawasaki H, Ainscough JF,Sasai Y, Nakatsuji N, Tada T, 2003: Pluripotency ofreprogrammed somatic genomes in embryonic stem hybridcells. Dev Dyn 227, 504–510.Takahashi K, Yamanaka S, 2006: Induction of pluripotentstem cells from mouse embryonic and adult fibroblastcultures by defined factors. Cell 126, 663–676.Tani T, Kato Y, Tsunoda Y, 2001: Direct exposure ofchromosomes to nonactivated ovum cytoplasm is effectivefor bovine somatic cell nucleus reprogramming. Biol Reprod64, 324–330.Terranova R, Pereira CF, Du Roure C, Merkenschlager M,Fisher AG, 2006: Acquisition and extinction of geneexpression programs are separable events in heterokaryonreprogramming. J Cell Sci 119, 2065–2072.Tian Q, Stepaniants SB, Mao M, Weng L, Feetham MC, DoyleMJ, Yi EC, Dai H, Thorsson V, Eng J, Goodlett D, BergerJP, Gunter B, Linseley PS, Stoughton RB, Aebersold R,Collins SJ, Hanlon WA, Hood LE, 2004: Integrated genomicand proteomic analyses of gene expression in mammaliancells. Mol Cell Proteomics 3, 960–969.Vassena R, Han Z, Gao S, Baldwin DA, Schultz RM, LathamKE, 2007: Tough beginnings: alterations in the transcriptomeof cloned embryos during the first two cell cycles. DevBiol 304, 75–89.Vassena R, Han Z, Gao S, Latham KE, 2007: Deficiency inrecapitulation of stage-specific embryonic gene transcriptionin two-cell stage cloned mouse embryos. Mol Reprod Dev74, 1548–1556.Vassena R, Han Z, Gao S, Latham KE, 2008: Erratum:Deficiency in recapitulation of stage-specific embryonic genetranscription in two-cell stage cloned mouse embryos. MolReprod Dev. 75, 217.Wakayama T, Tateno H, Mombaerts P, Yanagimachi R, 2000:Nuclear transfer into mouse zygotes. Nat Genet 24, 108–109.Wang F, Kou Z, Zhang Y, Gao S, 2007: Dynamic reprogrammingof histone acetylation and methylation in the first cellcycle of cloned mouse embryos. Biol Reprod 77, 1007–1016.Wells DN, Laible G, Tucker FC, Miller AL, Oliver JE, XiangT, Forsyth JT, Berg MC, Cockrem K, L’Huillier PJ, TervitHR, Oback B, 2003: Coordination between donor cell typeand cell cycle stage improves nuclear cloning efficiency incattle. Theriogenology 59, 45–59.Wells DN, Misica PM, Tervit HR, 1999: Production of clonedcalves following nuclear transfer with cultured adult muralgranulosa cells. Biol Reprod 60, 996–1005.Wuensch A, Habermann FA, Kurosaka S, Klose R,Zakhartchenko V, Reichenbach HD, Sinowatz F,McLaughlin KJ, Wolf E, 2007: Quantitative monitoring ofpluripotency gene activation after somatic cloning in cattle.Biol Reprod 76, 983–991.Yabuuchi A, Kato Y, Tsunoda Y, 2002: Effects of aggregationof nuclear-transferred mouse embryos developed fromenucleated eggs receiving ES cells on in vitro and in vivodevelopment. J Reprod Dev 48, 393–397.Yamazaki Y, Makino H, Hamaguchi-Hamada K, Hamada S,Sugino H, Kawase E, Miyata T, Ogawa M, Yanagimachi R,Yagi T, 2001: Assessment of the developmental totipotencyof neural cells in the cerebral cortex of mouse embryo bynuclear transfer. Proc Natl Acad Sci U S A 98, 14022–14026.Yang F, Hao R, Kessler B, Brem G, Wolf E, ZakhartchenkoV, 2007: Rabbit somatic cell cloning: effects of donor celltype, histone acetylation status and chimeric embryocomplementation. Reproduction 133, 219–230.Yang J, Yang S, Beaujean N, Niu Y, He X, Xie Y, Tang X,Wang L, Zhou Q, Ji W, 2007: Epigenetic marks in clonedrhesus monkey embryos: comparison with counterpartsproduced in vitro. Biol Reprod 76, 36–42.Zeng F, Baldwin DA, Schultz RM, 2004: Transcript profilingduring preimplantation mouse development. Dev Biol 272,483–496.Zhang F, Pomerantz JH, Sen G, Palermo AT, Blau HM, 2007:Active tissue-specific DNA demethylation conferred bysomatic cell nuclei in stable heterokaryons. Proc Natl AcadSci U S A 104, 4395–4400.Zhou W, Xiang T, Walker S, Farrar V, Hwang E, Findeisen B,Sadeghieh S, Arenivas F, Abruzzese RV, Polejaeva I, 2007:Global gene expression analysis of bovine blastocystsproduced by multiple methods. Mol Reprod Dev 75, 744–758.Author’s address (for correspondence): B Oback, AgResearch Ltd.,Ruakura Research Centre, East Street, Private Bag 3123, Hamilton,New Zealand. E-mail: bjorn.oback@agresearch.co.nzConflict of interest: The author declares no conflict of interest.Ó 2008 The Author. Journal compilation Ó 2008 Blackwell Verlag

Reprod Dom Anim 43 (Suppl. 2), 417–422 (2008); doi: 10.1111/j.1439-0531.2008.01193.xISSN 0936-6768Nuclear Transfer of Freeze-Dried Somatic Cells into Enucleated Sheep OocytesP Loi 1 , K Matzukawa 2 , G Ptak 1 , Y Natan 3 , J Fulka Jr 4 and A Arav 31 Department of Comparative Biomedical Sciences, Teramo University, Teramo, Italy; 2 National Institute of Livestock and Grassland Science,Tsukuba, Japan; 3 Institute of Animal Science, Agricultural Research Organization, The Volcani Center, Bet Dagan, Israel; 4 Institute of AnimalProduction, Prague, Czech RepublicContentsLyophilization has been used since long time to preserveyeast and bacteria strains. Subsequently, a great deal ofefforts has been dedicated to the preservation in a dry state ofred blood cells and platelets. However, despite more than30 years passed by, no significant progress has been achieved.Recently, it has been reported that freeze-dried mice spermatozoawere able to generate normal offspring followinginjection into the mature mice oocytes. In this work, weprompted to apply the lyophilization protocol developed formice spermatozoa to sheep somatic cells (lymphocytes andgranulosa cells). More than 350 enucleated sheep oocyteswere injected with granulosa cells, and freeze dried using theprotocol developed for mice sperm cells. Transplanted nucleiorganized large pronuclei with fragmented DNA, but none ofthem entered the first mitosis. In the second part of theexperiments, trehalose and EGTA were found to reducesignificantly the extent of nuclear damage (65% and 55%intact nuclei in lymphocyte and granulosa cells, respectively)following freeze drying. Granulosa cells lyophilized withEGTA ⁄ trehalose and stored at room temperature for 3 yearswere used for nuclear transfer, and the injected oocytes werecultured in vitro for 7 days. Approximately 16% of the oocyteinjected with freeze-dried cells developed into blastocysts. Toconclude, we demonstrated for the first time that nucleatedcells maintain genomic integrity after prolonged storage in adry state, and we were able to achieve early embryonicdevelopment following injection of these cells into enucleatedsheep oocytes.IntroductionLyophilization has been used for the preservation offowl spermatozoa already in the 1950s (Polge et al.1949). The original protocol was applied later to otherspecies (Sherman 1954), but the results in terms ofoffspring production were contradictory (Saacke andMalquist 1961). The definitive proof that dry spermatozoaretain genetic integrity was established only whenmicrosurgical procedures were developed for bypassingthe lack of mobility of freeze-dried spermatozoa, andnormal mice were produced by intracytoplasmic sperminjection (ICSI) of freeze-dried sperm (Wakayama andYanagimachi 1998). A following paper from the samegroup demonstrated the preservation of genomic integrityin freeze-dried spermatozoa (Kusakabe et al. 2001),and more recently these results have been demonstratedin other species (Keskintepe et al. 2002; Liu et al. 2004).The possibility to store male gametes in a dry staterepresents a major breakthrough for storing and shippingmale gametes from mutants, transgenic and otherstrains of laboratory mouse. Lyophilization of spermcould also be an attractive way to store spermatozoa infarm animal species, and especially in human, whereICSI has become a routine procedure in assistedreproduction.The progressive reduction of large animals worldwide(Hilton-Taylor 2000; Margules and Pressey 2000) hassuggested to establish gene banks from species threatenedby extinction (Myers et al. 2000), with the aim touse the cells for somatic cell nuclear transfer (SCNT,Wilmut et al. 1997).Somatic cell nuclear transfer has indeed an obviouspotential for the multiplication of rare genotypes (Corley-Smithand Brandhorst 1999; Loi et al. 2001), but itswide application is prevented by the low efficiency interms of offspring outcomes. Furthermore, the majorityof the endangered mammals are practically unknownfrom a reproductive point of view. Therefore, thestorage of somatic cells for future use, once theprocedure of somatic cell cloning will be reliable, iscertainly a wise step to be undertaken. However, theestablishment of gene banks in the form of cell linesencounters several problems, represented by the costs ofliquid nitrogen. Recently, our group demonstrated thatsomatic cells rendered unviable by heat treatmentretained their potential to generate a normal lamb afternuclear transplantation (Loi et al. 2002). The mainscope of this work was to demonstrate the feasibility ofusing radical approaches for the nuclear reprogrammingof somatic cells, but, as in case of freeze-dried sperm, wealso established that lack of cell viability, as indicated bymassive membrane and cytoplasmic damage, is not anabsolute prerequisite for SCNT.In this work, we developed robust procedures for thepreservation of sheep somatic cells in a freeze-driedstate, and we tested their ability to direct early embryonicdevelopment of enucleated oocytes. The ensuingresults demonstrate for the first time the production ofnormal embryos from nuclear transfer of somatic cellsstored freeze dried for more than 3 years at roomtemperature.Materials and MethodsCell collection and freeze dryingIn the first experiment, granulosa cells dissociated fromin vitro-matured Cumulus Oocyte Complexes (COCs)from Sarda breed ewes were dispersed in 1 ml of HepesbufferedDMEM medium supplemented with 10% FCS;then 100 ll aliquots of the cell suspension were loadedinto 5 ml ampoules and plunged directly into liquidnitrogen. Frozen cells were then passed on pre-cooledstage of a freeze-drying apparatus (Freezone 4.5; LabconcoCorporation, Kansas City, MO, USA) andÓ 2008 Teramo University

418 P Loi, K Matzukawa, G Ptak, Y Natan, J Fulka Jr and A Aravlyophilized for 24 h under an inlet pressure of 1 mtorr at)50°C, following the protocol used for sperm celllyophilization (Wakayama and Yanagimachi 1998). Atthe end of the process, each ampoule was flame sealed,placed into cardboard and stored at room temperature(18–23°C) until use.On the light of the massive nuclear damage in rehydratedcells lyophilized using the sperm cell protocol,different solutions were tested for lyophilization. Briefly,peripheral blood lymphocytes were isolated fromperipheral blood (from Assaf ewes) through a Ficoll–Paque density gradient; the purity of the cells wasassessed by an automatic cell counter (Pentra 60; ABX,Montpellier, France), more than 80% of the cells werelymphocytes.Granulosa cells were obtained by COCs collectedfrom the ovaries of slaughtered Assaf ewes. CumulusOocyte Complexes were matured for 24 h in the mediumTCM 199 plus 10% FCS, FSH, LH and oestradiol inincubator at 38.5°C with 5% CO 2 . At the end ofmaturation, expanded COCs were shortly incubatedin hyaluronidase (300 USP units ⁄ ml; Sigma-Aldrich,Milano, Italy) and dissociated into single-cell populationby repeated pipetting.The freezing solution (2000 ll) was added to the cells(lymphocytes or granulosa) to be lyophilized. Thefreezing solution was composed of 50% FCS and0.1 M trehalose in Hepes Talp buffer. Samples at avolume of 2 ml were frozen using the MTG freezingapparatus (IMT, Nes Ziona, Israel) at a cooling rate of5.1°C ⁄ min. The concentration of the cells ranged from1–6 · 10 6 cells ⁄ ml. After freezing, the samples werestored in liquid nitrogen until transfer into the lyophilizer(Freezone Plus 6; Labconco). Samples werelyophilized for 72 h, after which each ampoule wasflame sealed, placed into cardboard and stored at roomtemperature (18–23°C) until use. Aliquots of the sampleswere dispatched by air mail to Teramo, Italy, fornuclear transfer studies.RehydrationImmediately before use for nuclear transfer, the ampouleswere opened and 100 ll or 2000 ll of milliQ water(according to the freeze-drying protocol) were added.Rehydrated cells were washed twice with medium 199plus Hepes, antibiotics and BSA before use for nucleartransfer. For each ampoule, viability was assessed onsmall aliquots cells by propidium iodine staining (200cells counted). DNA fragmentation was assessed withcomet essay according the manufacturer instruction(R&D Systems, Milano, Italy) in every replicate.DNA integrityDNA integrity was evaluated using the single cell gelelectrophoresis assay (aka comet assay) (R&D Systems)as described by the manufacturer. Cells were diluted to aconcentration of 105 cells ⁄ ml in PBS. The cells werecombined with molten LM agarose at a ratio of:1 : 10(v ⁄ v), then 75 ll were placed on comet slides.The slides were put in the dark at 4°C for 10 min. Slideswere then immersed in a pre-cooled (4°C) lysis solutionfor 1 h at 4°C. Afterward, the slides were immersed in afreshly prepared alkali solution, which consisted of 0.6 gNaOH pellets, 250 ll of 250 mM EDTA, pH 10.0 and49.75 ml deionized water, for 60 min in the dark atroom temperature. Finally, the slides were washed in aTBE buffer (Tris base, Boric acid and EDTA) for 5 minand then the slides were submerged in TBE buffer in ahorizontal electrophoresis apparatus; and 1 V ⁄ cm wasapplied for 10 min. In addition, each experiment wasconducted with two controls; the first one was of cellsthat were previously treated with 100 lM of hydrogenperoxide for 10 min at 2–8°C as described in themanufacturer kit (this served as a positive control forthe assay showing damaged DNA) and the secondcontrol was of fresh cells, indicating endogenous levelsof damage within the cells. After the samples were dried,they were stained with SYBR green. Scoring was carriedout using a fluorescent microscope (Zeiss, Vertrieb,Germany) connect to a digital camera (Sony, Tokyo,Japan) and analysed using the Image J free software(NIH, Baltimore, MD, USA).SEMFor evaluation of the morphology of dry samples,Scanning Electron Microscopy (Philips Inc., Milano,Italy) was used. The samples were gold plated beforebeing placed in the SEM, and observed with a voltage ofthe electron scatter of 25 kV.Oocyte maturationMethods of in vitro embryo production were as previouslydescribed (Ptak et al. 1999a,b). Briefly, followingcollection of ovaries from adult sheep (Comisana andBergamasca breeds) at slaughter, oocytes were aspiratedand evaluated under a dissecting microscope. Onlyoocytes surrounded by at least two layers of granulosacells and with evenly granulated cytoplasm were selectedfor in vitro maturation (IVM). Oocytes were maturedin vitro in bicarbonate-buffered TCM-199 (Gibco) (275mOsm) containing 2 mM glutamine, 100 lM cysteamine,0.3 mM sodium pyruvate, 10% foetal bovine serum(FBS) (Gibco Milan, Italy), 5 lg ⁄ ml FSH (OvagenAuckland, New Zealand), 5 lg ⁄ ml LH, 1 lg ⁄ ml oestradiolin a humidified atmosphere of 5% CO 2 in air at39°C for 24 h.Oocyte enucleation and nuclear transferMatured oocytes were enucleated generally after 22 h,after the start of the maturation. Cumulus OocyteComplexes were shortly incubated with hyaluronidase,and the granulosa cells removed by repeated pipetting.Oocytes were incubated in Hepes-buffered 199 mediumcontaining 4 mg ⁄ ml of BSA and 7.5 lg ⁄ ml of cytochalasinB and Hoechst 33342, 10 mg ⁄ ml for 15 min inincubator.Enucleation was carried out in Hepes-buffered 199medium plus BSA and cytochalasin B using the routinemanipulation procedures.Enucleated oocytes were allowed to recover fromcytochalasin B treatment for 10 min in the incubator,Ó 2008 Teramo University

Dry Cells for Nuclear Transfer 419then directly injected with freeze-dried cell. Reconstructedoocytes were activated in Hepes-bufferedmedium 199 containing 5 lg ⁄ ml Ionomycin for 5 min,then incubated in SOF medium plus antibiotics andBSA containing 10 mM dimethylaminopurine and7.5 lg ⁄ ml cytochalasin B for 3–5 h. Subsamples ofreconstructed embryos were processed for comet assayas described earlier.Embryo cultureReconstructed embryos were transferred into 20-lldrops of SOF enriched with 1% (v : v) basal mediumEagle (BME)-essential amino acids, 1% (v : v) minimumessential medium (MEM)-non-essential aminoacids (Gibco), 1 mM glutamine, and 8 mg ⁄ ml fattyacid-free BSA (SOFaaBSA). Zygote cultures weremaintained in humidified atmosphere of 5% CO 2 ,7%O 2 , 88% N 2 at 39°C, and the medium renewed on day 3and day 5 of culture (day 0 = day of activation, 16).Embryonic development was monitored every 24 h, andarrested embryos were fixed in acetic acid–methanol3 : 1 for 24 h, then stained with 2% aceto-orcein forchecking number and gross morphology of nuclei. Inthree replicates, aliquots of reconstructed embryos werefixed at late nuclear swelling stage (13–16 h postactivation)and processed for Comet analysis to investigatethe presence and extent of DNA damage.DNA microsatellite analysisGenomic DNA from the pooled blastocysts derivedfrom batches of freeze-dried cells (one ampoule of cellwas used for each replicate), and from the re-hydratedcell used as nuclei donor was prepared following thestandard procedures. Microsatellite analyses were performedwith multiplex polymerase chain reaction usingnine microsatellite markers (CP49, FCB11, AE129,FCB304, INRA063, MAF214, PZ963, CSRD247 andHSC in table 1). Based on the degree of polymorphismof these microsatellites in the sheep population, theprobability that cell line and the cloned embryos wouldhave the same genotype was determined to be smallerthan 2 · 10 )11 . Genotypes were determined by polyacrylamidegel electrophoresis using an automatedDNA sequencer (Perkin-Elmer, Bucks, UK) and wereanalysed by Genscan and Genotyper software (GenscanDetection System, Woburn, MA, USA).Stastical analysisThe experimental observations, expressed as percentageof Comet assay positive cells and embryonic developmentto the blastocyst stage, were processed by a onewayanalysis of variance using the statistical packageSPSS (SPSS, Inc., Chicago, IL, USA) following themodel:y ij ¼ l þ a i þ e ij ;where y ij is the experimental observations, l, the generalmean, a i , effect caused by the treatment (i = 1 forcontrol groups (cells and embryos), i = 2 for Cometpositive cells ⁄ freeze-dried cells nuclear transfer embryos,and ij = casual effect of the error (0, r).ResultsThe first embryo reconstructions were made usinggranulosa cells freeze-dried according the methods publishedby Wakayama and Yanagimachi 1998 (4). Over350 oocytes were injected in these preliminary experiments.The majority of them (90%) were activated anddeveloped a pronuclear-like structure, but none of themcompleted the first mitosis. Nuclear morphology revealedhighly decondensed pronucleus-like structure,with massive DNA fragmentation (Fig. 2e). Lymphocytesand granulosa cells were unviable following rehydration,for all of them stained red with propidiumiodine. However, the inclusion of trehalose in thefreezing medium dramatically reduced DNA damageassessed by Comet, with 65% of lymphocytes and 55%of granulosa negative to the Comet assay (Fig. 2a).Despite the higher proportion of damaged DNA, granulosacells were used for nuclear transfer, for we have anestablished and robust cloning protocol based on the useof granulosa cells as nuclear donor.The positive effect exerted by the inclusion of threalosein the freezing media was also evident from the SEManalysis of cells (lymphocytes) freeze dried with orwithout threalose. In the presence of threalose, the driedmixture displayed a glassy, smooth appearance (Fig. 2b),on the contrary of the control samples, where the powderhas rather a splinter-like appearance (Fig. 2c).Embryo development following nuclear transfer ofcells frozen with trehalose is indicated in Table 2.Cleavage rate was similar to both the control and thefreeze-dried groups, although the latter reached theblastocyst stage (Fig. 2d) at lower extent (15.6% vs21%, p = 0.56). Microsatellite analysis performed withmultiplex polymerase chain reaction using nine microsatellitemarkers demonstrated early left any doubts thatthe blastocysts obtained were isogenic with the freezedrieddonor granulosa cells (data not shown).DiscussionWe have convincingly shown that the procedure whichhas been developed by Wakayama and Yanagimachi1998 for mouse spermatozoa (4) was not suitable tofreeze-dried granulosa cells which needed to serve asTable 1. Microsatellite analysis of cell lines and corresponding clonedblastocysts (three replicates)LocusCell lineClonedblast.Cell lineClonedblast.Cell lineClonedblast.PF PFC PE1 PE1C PE2 PE2COAR CP 49 90 98 90 98 90 98 90 98 90 98 90 98FCB 11 118 122 124 134 124 134 124 134 124 134 118 122OAR AE 129 145 149 145 149 145 149 145 149 145 149 145 149FCB 304 166 170 170 170 170 170 170 170 170 170 166 170INRA 063 175 179 175 179 175 179 175 179 175 179 175 179MAF 214 189 261 189 189 189 189 189 189 189 189 189 261CSRD247 223 227 213 227 213 227 213 227 213 227 223 227HSC 269 269 269 293 – – 269 293 269 293 269 269Ó 2008 Teramo University

420 P Loi, K Matzukawa, G Ptak, Y Natan, J Fulka Jr and A Arav% DNA Intact100806040200aFreshbFreeze-driedLymphocytesFreeze-DriedGranulosa cellsa-b = p < 0.05; b-c = p < 0.05; a,b,c-d = p < 0.001;H 2 O 2 treatedFig. 1. Proportion of nuclei with undamaged DNA (assessed byComet assay; 200 nuclei counted per group, three replicates each).a–b = p < 0.05; b–c = p < 0.05; a,b,c–d = p < 0.001nucleus donor in somatic cell cloning. Although apronuclear-like structure was found in more than 90%of the reconstructed embryos 7–9 h after activation,cdnone of these embryos entered the first mitosis, andcytological analysis revealed in all embryos extensiveDNA damage, like the one observed in occurrence ofasynchronous transfer between s-phase nuclei andmitotic cytoplasts (Johnson and Rao 1970).Comet analysis confirmed the presence of DNAfragmentation in all the cells after rehydration (Figs 1and 2a). Interestingly, all cloned embryos arrested atbefore the first mitosis, confirming the presence of a veryrobust DNA damage checkpoint in zygotes (Kalogeropouloset al. 2004).These results were in part expected. Spermatozoa aresmall cells with a low level of hydration (Ward andZalensky 1996), and with transcriptionally inactiveDNA tightly packed around the basic protamine core;altogether, this nuclear structure is very suitable forfreeze drying. On the contrary, somatic cells with largerdiameter, higher water content, and upper most withtheir nuclear organization render them extremely vul-(a)(b)(c)(d)(e)(g)(h)(f)Fig. 2. (a) Comet analysis ofrehydrated lymphocytes lyophilizedin 0.1 M threalose, 40·. (b)SEM appearance of dry powderlyophilized with 0.1 M threalose,350·. (c) SEM appearance of drypowder lyophilized without threalose,350·. (d) Fragmented DNAin a pronuclear-like structureorganized after 7 h post-activation,100·. (e) Normal pronucleus with aregular nuclear membrane in controlreconstructed embryos, 100·.(f) Control embryo reconstructedwith fresh granulosa cells; nocomet tail visible. (g) Embryoreconstructed with a freeze-driedcell: note the massive DNA damageindicated by the migration offragmented DNA outside theembryo. (h) Blastocyst producedseven after culture of enucleatedoocytes injected with freeze-driedgranulosa cells, 40·Ó 2008 Teramo University

Dry Cells for Nuclear Transfer 421Table 2. In vitro development of enucleated oocytes injected withfreeze-dried and fresh, control granulosa cellsSource of cells Cultured 2–8 Cells Morula BlastocystGranulosa control 129 43 31 27 (20.9%) aGranulosa freeze-dried 160 52 28 25 (15.6%) bThe cultures were maintained for 7–8 days in the medium SOF plus amino-acidsand BSA, with FCS added on day 4. Reconstructed embryos were checked every24 h for development.a–b p = 0.56.nerable to the combined osmotic ⁄ dehydration stressimposed the freeze-drying process.These preliminary observations, and the followingreports published on the topics (Kaneko et al. 2003),prompted us to formulate alternative test medium forfreeze-drying somatic cells. Our first target was to avoidthe massive DNA damage evident in rehydrated cellsdried with the sperm protocol.Trehalose produced by some plants, yeasts, sporesand a range of unicellular organism confers them theability to survive conditions of almost absence of water(Crowe et al. 1992). The mechanism by which trehaloseconfers desiccation tolerance is not clear. Trehalose perse is thought to replace the shell of water on the surfaceof macromolecules, proteins in particular (Prestrelskiet al. 1993) while on lipids it seems that its protectiveeffect is due to the inhibition of phase transitiontemperature of membranes (Leslie et al. 1994). Remarkably,the induced threalose expression in fibroblastsconferred desiccation tolerance for 5 days (Guo et al.2000). Based on these indications, trehalose wasincluded in the formulation of the freeze-dryingmedium.As reported in results, freeze-dried medium containing0.1 M trehalose was highly effective in protecting thenuclear compartment, for about half of the cells hadintact DNA after rehydration. The best results observedin lymphocytes, comparing to granulosa cells (65% vs55%, respectively), p < 0.05) was probably due to thesmaller size of the former cell.Enucleated oocytes reconstructed with freeze-driedgranulosa cells developed a pronucleus-like structureand started to cleave an equal rates compared withcontrol ones (Table 2), and 16% of them reached theblastocyst stage (Fig. 2). This result was quite surprising,because if we consider that about half of thegranulosa cells displayed damaged DNA followingrehydration, we would have expected a much lowerproportion of oocytes injected with dried cells developingto blastocyst stage. Exonuclease and recombinationactivity both increase during oogenesis, and fully growneggs have found to catalyse homologous recombination,ligation and illegitimate recombination of exogenousDNA (Lehman et al. 1993), suggesting that multiplepathways are available in the oocyte’s cytoplasm for therepair of double strand break. Therefore, it is possiblethat DNA damage induced by the dehydration arerepaired by the oocyte and do not affect the development,at least to blastocyst stage.Our results showed for the first time the developmentalpotential of freeze-dried cells after the nucleartransplantation. However, while in the original reportsperm cells were kept lyophilized for 4 months prior touse for ICSI, the cells used in our study were storedfreeze dried for more than 3 years at room temperature.Therefore, we believe that our method is a majorcontribution towards alternative ways or the establishmentand maintaining cell lines for nuclear transfer.Cloning by nuclear transfer is a very dynamic field,where joint efforts of embryologist and molecularbiologist are producing knowledge of nuclear reprogrammingmechanism at very rapid pace (Gurdon et al.2003; Meissner and Jaenisch 2006), letting foreseeing asuccessful application of cloning procedures in themedium term. Unfortunately, the list on endangeredspecies is not limited to mammals, but a wide range ofminor vertebrates are also disappearing quickly. However,cloning procedures have been recently adapted tovery unconventional animal models with two papersreporting about successful cloning of zebra fish (Ju et al.2004) and Drosophila (Haigh et al. 2005) these accomplishmentsbroad the range of species that can potentiallybenefit from cloning technology.To conclude, we demonstrated for the first time thatsomatic cells stored in a freeze-dried state are able todirect embryonic development of enucleated oocytes tillblastocyst stage. These results represent a substantialbreakthrough innovative for the establishment ofgenetic banks from endangered animals.ReferencesCorley-Smith GE, Brandhorst BP, 1999: Preservation ofendangered species and populations: a role for genomebanking, somatic cell cloning, and androgenesis? MolReprod Dev 53, 363–367.Crowe JH, Hoekstra FA, Crowe LM, 1992: Anhydrobiosis.Annu Rev Physiol 54, 579–599.Guo N, Puhlev I, Brown DR, Mansbridge J, Levine F, 2000:Trehalose expression confers desiccation tolerance on humancells. Nat Biotechnol 18, 168–171.Gurdon JB, Byrne JA, Simonsson S, 2003: Nuclear reprogrammingand stem cell creation. Proc Natl Acad Sci U S A30, 11819–11822.Haigh AJ, Macdonald WA, Lloyd VK, 2005: The generationof cloned Drosophila melanogaster. Genetics 169, 1165–1167.Hilton-Taylor C, 2000: (compiler): IUCN Red List ofThreatened Species. IUCN ⁄ SSC, Gland, Switzerland andCambridge, UK.Johnson RT, Rao PN, 1970: Mammalian cell fusion: inductionof premature chromosome condensation in interphasenuclei. Nature 226, 717–722.Ju B, Huang H, Lee KY, Lin S, 2004: Cloning zebrafish bynuclear transfer. Methods Cell Biol 77, 403–411.Kalogeropoulos N, Christoforou C, Green AJ, Gill S,Ashcroft NR, 2004: chk-1 is an essential gene and isrequired for an S-M checkpoint during early embryogenesis.Cell Cycle 3, 1196–1200.Kaneko T, Whittingham DG, Yanagimachi R, 2003: Effect ofpH value of freeze-drying solution on the chromosomeintegrity and developmental ability of mouse spermatozoa.Biol Reprod 68, 136–139.Keskintepe L, Pacholczyk G, Machnicka A, Norris K, CurukMA, Khan I, Brackett BG, 2002: Bovine blastocyst developmentfrom oocytes injected with freeze-dried spermatozoa.Biol Reprod 67, 409–415.Ó 2008 Teramo University

422 P Loi, K Matzukawa, G Ptak, Y Natan, J Fulka Jr and A AravKusakabe H, Szczygiel MA, Whittingham DG, YanagimachiR, 2001: Maintenance of genetic integrity in frozen andfreeze-dried mouse spermatozoa. Proc Natl Acad Sci U S A98, 13501–13506.Lehman CW, Clemens M, Worthylake DK, Trautman JK,Carroll D, 1993: Homologous and illegitimate recombinationin developing Xenopus oocytes and eggs. Mol Cell Biol13, 6897–6906.Leslie SB, Teter SA, Crowe LM, Crowe JH, 1994: Trehaloselowers membrane phase transitions in dry yeast cells.Biochim Biophys Acta 1192, 7–13.Liu JL, Kusakabe H, Chang CC, Suzuki H, Schmidt DW,Julian M, Pfeffer R, Bormann CL, Tian XC, YanagimachiR, Yang X, 2004: Freeze-dried sperm fertilization leads tofull-term development in rabbits. Biol Reprod 70, 1776–1781.Loi P, Ptak G, Barboni B, Fulka J Jr, Cappai P, Clinton M,2001: Genetic rescue of an endangered mammal by crossspeciesnuclear transfer using post-mortem somatic cells.Nat Biotechnol 19, 962–964.Loi P, Clinton M, Barboni B, Fulka J Jr, Cappai P, Feil R,Moor RM, Ptak G, 2002: Nuclei of nonviable ovine somaticcells develop into lambs after nuclear transplantation. BiolReprod 67, 126–132.Margules CR, Pressey RL, 2000: Systematic conservationplanning. Nature 405, 243–253.Meissner A, Jaenisch R, 2006: Mammalian nuclear transfer.Dev Dyn 235, 2460–2469.Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB,Kent J, 2000: Biodiversity hotspots for conservation priorities.Nature 403, 853–858.Polge C, Smith A, Parkes AS, 1949: Revival of Spermatozoaafter vitrification and dehydration at low temperature.Nature 164, 666–667.Prestrelski SJ, Arakawa T, Carpenter JF, 1993: Separation offreezing- and drying-induced denaturation of lyophilizedproteins using stress-specific stabilization. II. Structuralstudies using infrared spectroscopy. Arch Biochem Biophys303, 465–473.Ptak G, Dattena M, Loi P, Tischner M, Cappai P, 1999a:Ovum pick-up in sheep: efficiency of in vitro embryoproduction, vitrification and birth of offspring. Theriogenology52, 1105–1114.Ptak G, Loi P, Dattena M, Tischner M, Cappai P, 1999b:Offspring from one-month-old lambs: studies on the developmentalcapability of prepubertal oocytes. Biol Reprod 61,1568–1574.Saacke RG, Malquist JO, 1961: Freeze-drying of bovinespermatozoa. Nature 192, 995–996.Sherman JK, 1954: Freezing and freeze-drying of humanspermatozoa. Fertil Steril 5, 357–371.Wakayama T, Yanagimachi R., 1998: Development of normalmice from oocytes injected with freeze-dried spermatozoa.Nat Biotechnol 16, 639–644.Ward SW, Zalensky AO, 1996: The unique, complex organizationof the transcriptionally silent sperm chromatin. CritRev Eukaryot Gene Expr 6, 139–147.Wilmut I, Schnieke AE, McWhir J, Kind AJ, Campbell KHS,1997: Viable offspring derived from fetal and adult mammaliancells. Nature 385, 810–813.Author’s address (for correspondence): Lino Loi, Department ofComparative Biomedical Sciences, Teramo University, Piazza AldoMoro 45, 64100, Teramo, Italy. E-mail: ploi@unite.itConflict of interest: All authors declare no conflict of interests.Ó 2008 Teramo University

The largest on-line library includes textbooks, proceedings, journals,manuals, a calendar, and much more. Free unlimited access forveterinarians, veterinary technicians and students.Books in Veterinary Medicine:o A Concise Review of Veterinary Virology byCarter, Wise and Floreso Topics in Bull Fertility by Chenowetho Recent Advances in Small AnimalReproduction by Concannon et al.o Recent Advances in Equine Reproduction byBallo Recent Advances in Yak Reproduction by Zhaoand Zhango Recent Advances in Anesthetic Managementof Large Domestic Animals by Steffeyo Recent Advances in Goat Diseases byTempestao Recent Advances in Laparoscopy andThoracoscopy by Wilsono Recent Advances in Veterinary Anesthesiaand Analgesia: Companion Animals by Gleedand Ludderso Clinical Avian Medicine by Harrison andLightfooto Equine Respiratory Diseases by Lekeuxo Care of Donkeys by Matthews and Tayloro A Guide to Hematology in Dogs and Cats byRebar, MacWilliams, Feldman, et al.o Dermatology for the Small Animal Practitionerby Muellero Veterinary Toxicology by Beasleyo Animal Disease Factsheets The Center for FoodSecurity & Public Health - Iowa State UniversityProceedings in Veterinary Medicine:o NAVC - North American Veterinary Conferenceo IPVS - International Pig Veterinary SocietyCongresso WBC - World Buiatrics Congresso AAEP - American Association of EquinePractitionerso SIVE - Italian Association of EquineVeterinarians Annual Congresso BEPS - Belgian Equine Practitioners SocietyStudy Dayso Geneva Congress on Equine Medicine andSurgeryo SCIVAC - Società Culturale Italiana Veterinariper Animali da Compagnia Congresso EENHC - European Equine Health andNutrition Biannual Congresso AAVPT – American Academy of VeterinaryPharmacology and Therapeuticso WSAVA - World Small Animal VeterinaryAssociation Annual Congresso Canine Cancer Conference: Genes, Dogs,and Cancero Cross-Species Approach to Pain andAnalgesiao EAVDI - European Association of VeterinaryDiagnostic Imagingo EAZWV - European Association of Zoo andWildlife Veterinarians Biannual Meetingso Lameness in Ruminants SymposiumVeterinarians, students and Technicians have free access to all books andproceedings. Visit the IVIS Website at www.ivis.org and register today. For free!!

Table of Contents Volume 43 · Supplement 2 · July 2008 · 1-422CONTENTS continued...B. M. A. O. PERERAReproduction in Domestic Buffalo 200-206P. S. BRAR, A. S. NANDAPostpartum Ovarian Activity in South Asian Zebu Cattle 207-212M. J. R. PARANHOS DA COSTA, A. SCHMIDEK, L. M. TOLEDOMother-Offspring Interactions in Zebu Cattle 213-216B. S. PRAKASH, M. SARKAR, M. MONDALAn Update on Reproduction in Yak and Mithun 217-223F. X. DONADEU, H. G. PEDERSENFollicle Development in Mares 224-231M. A. HAYES, B. A. QUINN, N. D. KEIRSTEAD, P. KATAVOLOS, R. O. WAELCHLI, K. J. BETTERIDGEProteins Associated With the Early Intrauterine Equine Conceptus 232-237Z. ROTHHeat Stress, the Follicle, and Its Enclosed Oocyte: Mechanisms and Potential Strategies to ImproveFertility in Dairy Cows 238-244K. - P. BRÜSSOW, J. RÁTKY, H. RODRIGUEZ-MARTINEZFertilization and Early Embryonic Development in the Porcine Fallopian Tube 245-251S. PYÖRÄLÄMastitis in Post-Partum Dairy Cows 252-259M. G. DISKIN, D. G. MORRISEmbryonic and Early Foetal Losses in Cattle and Other Ruminants 260-267N. MANABE, F. MATSUDA-MINEHATA, Y. GOTO, A. MAEDA, Y. CHENG, S. NAKAGAWA, N. INOUE,K. WONGPANIT, H. JIN, H. GONDA, J. LIRole of Cell Death Ligand and Receptor System on Regulation of Follicular Atresia in Pig Ovaries 268-272F. F. BARTOL, A. A. WILEY, C. A. BAGNELLEpigenetic Programming of Porcine Endometrial Function and the Lactocrine Hypothesis 273-279K. C. CAIRES, J. A. SCHMIDT, A. P. OLIVER, J. DE AVILA, D. J. MCLEANEndocrine Regulation of the Establishment of Spermatogenesis in Pigs 280-287I. DOBRINSKIMale Germ Cell Transplantation 288-294N. RAWLINGS, A. C. O. EVANS, R. K. CHANDOLIA, E. T. BAGUSexual Maturation in the Bull 295-301A. DINNYES, X. C. TIAN, X. YANGEpigenetic Regulation of Foetal Development in Nuclear Transfer Animal Models 302-309R. C. BOTT, D. T. CLOPTON, A. S. CUPPA Proposed Role for VEGF Isoforms in Sex-Specific Vasculature Development in the Gonad 310-316B. K. WHITLOCK, J. A. DANIEL, R. R. WILBORN, T. H. ELSASSER, J. A. CARROLL, J. L. SARTINComparative Aspects of the Endotoxin- and Cytokine-Induced Endocrine Cascade InfluencingNeuroendocrine Control of Growth and Reproduction in Farm Animals 317-323C. R. BARB, G. J. HAUSMAN, C. A. LENTSEnergy Metabolism and Leptin: Effects on Neuroendocrine Regulation of Reproduction in the Gilt and Sao 324-330C. GALLI, I. LAGUTINA, R. DUCHI, S. COLLEONI, G. LAZZARISomatic Cell Nuclear Transfer in Horses 331-337D. RATH, L. A. JOHNSONApplication and Commercialization of Flow Cytometrically Sex-Sorted Semen 338-346J. M. VAZQUEZ, J. ROCA, M. A. GIL, C. CUELLO, I. PARRILLA, I. CABALLERO, J. L. VAZQUEZ, E. A. MARTÍNEZLow-Dose Insemination in Pigs: Problems and Possibilities 347-354C. B. A. WHITELAW, S. G. LILLICO, T. KINGProduction of Transgenic Farm Animals by Viral Vector-Mediated Gene Transfer 355-358A. C. O. EVANS, N. FORDE, G. M. O'GORMAN, A. E. ZIELAK, P. LONERGAN, T. FAIRUse of Microarray Technology to Profile Gene Expression Patterns Important for Reproduction in Cattle 359-367J. P. KASTELIC, J. C. THUNDATHILBreeding Soundness Evaluation and Semen Analysis for Predicting Bull Fertility 368-373G. C. ALTHOUSESanitary Procedures for the Production of Extended Semen 374-378B. LEBOEUF, J. A. DELGADILLO, E. MANFREDI, A. PIACÈRE, V. CLÉMENT, P. MARTIN, M. PELLICER,P. BOUÉ, R. DE CREMOUXManagement of Goat Reproduction and Insemination for Genetic Improvement in France 379-385N. KOSTEREVA, M. - C. HOFMANNRegulation of the Spermatogonial Stem Cell Niche 386-392P. MERMILLOD, R. DALBIÈS-TRAN, S. UZBEKOVA, A. THÉLIE, J. - M. TRAVERSO, C. PERREAU,P. PAPILLIER, P. MONGETFactors Affecting Oocyte Quality: Who is Driving the Follicle? 393-400K. KIKUCHI, N. KASHIWAZAKI, T. NAGAI, M. NAKAI, T. SOMFAI, J. NOGUCHI, H. KANEKOSelected Aspects of Advanced Porcine Reproductive Technology 401-406B. OBACKClimbing Mount Efficiency – Small Steps, Not Giant Leaps Towards Higher Cloning Success in Farm Animals 407-416P. LOI, K. MATZUKAWA, G. PTAK, Y. NATAN, J. FULKA JR, A. ARAVNuclear Transfer of Freeze-Dried Somatic Cells into Enucleated Sheep Oocytes 417-422