GTMB 7 - Gene Therapy & Molecular Biology

Instructions to authors:Gene Therapy and Molecular Biology (GTMB)FREE ACCESS www.gtmb.orgScopeGene Therapy and Molecular Biology, bridging various fields is one of the most rapid with free access atgtmb.org.The scope of Gene Therapy and Molecular Biology is to promote interaction between researchers in thefields of Gene Therapy and Molecular Biology providing rapid publication of review articles and researchpapers. Articles (both invited and submitted) review or report novel findings of importance to a generalaudience in gene therapy, molecular medicine, gene discovery, and molecular biology with emphasis tomolecular mechanisms. The journal will accept papers on all aspects of gene therapy, including genedelivery systems, gene therapy of cancer and other diseases (e.g. CFTR, hemophilia, AIDS, restenosis) atthe clinical, preclinical or cell culture stage, gene discovery, cancer immunotherapy, DNA vaccines, useof DNA regulatory elements in gene transfer, cell therapy and transplantation, arraying technologies &DNA chips, peptide libraries and drug discovery related to gene therapy, cell targeting, gene targeting,therapy with oligonucleotides (antisense, ribozymes, triplex). The authors are encouraged to elaborate onthe molecular mechanisms that govern a gene therapy approach. Gene Therapy and Molecular Biologywill also publish articles on, transcription factors, DNA replication, recombination, repair, chromatin,nuclear matrix, DNA regulatory regions, locus control regions, protein phosphorylation, signaltransduction, development, and on molecular mechanism of human disease. To make the publicationattractive authors are encouraged to include color figures.Type of articlesBoth review articles and original research articles will be considered. In addition, short 1-2 page news &views will also be considered for publication. Original research articles should contain a generousintroduction in addition to experimental data. The articles contain information important to a generalaudience as the volume is also addressed to researches outside the field. There is no limit on the length ofthe articles provided that the subject is interesting to a general audience and covers exhaustively a field.The typical length of each manuscript is a approximately 4-20 printed page including Figures andTables. This is 12-60 manuscript pages.Charges, Complimentary reprints & SubscriptionsThere are no charges for color figures or page numbers. Corresponding authors get a one-year freesubscription (hard copy) plus 25 reprints free of charge. The free subscription can be renewed foradditional years by having one paper per year accepted for publication.The free electronic access to articles published in " Gene Therapy and Molecular Biology " to a biggeneral audience, the attractive journal title, the speed of the reviewing process, the no-charges for pagenumbers or color figure reproduction, the 25 complimentary reprints, the rapid electronic publication, theembracing of many fields in gene therapy (from molecular mechanisms to clinical trials), the high quality

in depth reviews and first rate research articles and most important, the eminent members of the EditorialBoard being assembled are prognostic factors of a big success for GTMB.Sections of the manuscriptEach manuscript should have a Title, Authors, Affiliation, Corresponding Author (with Tel, Fax, and E-mail), Summary, key words , running title and Introduction; review articles are subdivided intoheadings I, II, III, etc. (starting with I. Introduction) subdivided into A, B, C, and further subdivided using1, 2, 3, etc. You can further subdivide into 1, 2, 3, etc. Research articles are divided into Summary; I.Introduction; II. Materials and Methods III. Results; IV. Discussion; Acknowledgments; and References.Please include in your text citations the name of authors and year in parenthesis; for three or more authorsuse: (name of first author et al, with year); for two authors please use both names. Please delete hiddentext for references. In the reference list, please, type references with year and Journal in boldface andprovide full title of the article such as:Buschle M, Schmidt W, Berger M, Schaffner G, Kurzbauer R, Killisch I, Tiedemann J-K, Trska B,Kirlappos H, Mechtler K, Schilcher F, Gabler C, and Birnstiel ML (1998) Chemically defined, cell-freecancer vaccines: use of tumor antigen-derived peptides or polyepitope proteins for vaccination. GeneTher Mol Biol 1, 309-321.To avoid delays it is essential to submit an electronic and a hard copy version of your manuscript via e-mail and mail in a floppy, CD-ROM or ZIP, containing the manuscript that will be used to typeset thepaper. Please include in the digital media: Tables, if any, (preferably as a Microsoft Word text) and Figurelegends. Please use Microsoft Word, font “Times” (Mac users) or “Times New Roman” (PC users) andinsert Greek or other characters using the “Insert/Symbol” function in the Microsoft Word rather thansimple conversion to font “Symbol”. Please boldface Figure 1, 2, 3 etc. as well as Table 1, 2, etc.throughout the text. Please provide the highest quality of prints of your Figures; whenever possible,please provide in addition an electronic version of your figures.Article contributors are kindly requested to provide a color (or black/white) photo of themselves(preferably 4x5 cm or any size) or a group photo of the authors, as we shall include these in thepublicationSubmission and reviewingPeer reviewing is by members of the Editorial Board and external referees. Please suggest 2-3 reviewersproviding their electronic addresses, mailing addresses and telephone/fax numbers. Authors are sent pageproofs.Gene Therapy and Molecular Biology is published in on high quality paper, hardbound, and withexcellent reproduction of color figures.Reviewing is completed within 5-15 days from receiving the manuscript.Articles accepted without revisions (i.e., review articles) will be published online (www.gtmb.org) inapproximately 1 month following submission.

Please submit an electronic version of full text and figures preferably in jpeg format. The electronicversion of the figures will be used for the rapid reviewing process. High quality prints or photograph ofthe figures and the original with one copy should be sent via express mail to the Editorial Office.Editorial OfficeTeni Boulikas, Ph.D./ Maria Vougiouka, B.Sc.Gregoriou Afxentiou 7Alimos, Athens 17455GreeceTel: +30-210-985-8454Fax: +30-210-985-8453and electronically tomaria@cancer-therapy.orgThe free electronic access to articles published in "GTMB" to a big general audience, the attractivejournal title, the speed of the reviewing process, the no-charges for page numbers or color figurereproduction, the 25 complimentary reprints, the rapid electronic publication, the embracing of manyfields in cancer, the anticipated high quality in depth reviews and first rate research articles and mostimportant, the eminent members of the Editorial Board being assembled are prognostic factors of a bigsuccess for the newly established journal.

Table of contentsGene Therapy and Molecular BiologyVol 7, December 2003PagesType ofArticleArticle titleAuthors (corresponding author is inboldface)1-14 ReviewArticle15-23 ReviewArticle25-35 ReviewArticle37-42 ReviewArticle43-59 ReviewArticle61-68 ResearchArticle69-73 MiniReview75-89 ReviewArticle91-98 ReviewArticle99-102 ResearchArticleDynamic histone acetylation and itsinvolvement in transcriptionTumor therapy using radiolabelledantisense oligomers- aspects forantiangiogenetic strategy and positronemission tomographyStrategy of sensitizing tumor cells withadenovirus-p53 transfectionAntigenicity and immunogenicity of HIVenvelope gene expressed in baculovirusexpression systemCharacterization of genes transcribed inan Ixodes scapularis cell line that wereidentified by expression libraryimmunization and analysis of sequencetagsDelayed intratracheal injection ofmanganese superoxide dismutase(MnSOD)-plasmid/liposomes providessuboptimal protection against irradiationinducedpulmonary injury compared totreatment before irradiationRegulation of vascular endothelial growthfactor by hypoxiaGene therapy antiproliferative strategiesagainst cardiovascular disease.Regulation of the Sp/KLF-family oftranscription factors: focus on posttranscriptionalmodification and proteinproteininteraction in the context ofchromatinDetection of MET oncogene amplificationin hepatocellular carcinomas bycomparative genomic hybridization onmicroarraysVirginia A. Spencer and James R. DavieKalevi JA Kairemo, Mark Lubberink,Mikko Tenhunen, Antti P JekunenJekunen Antti, Miettinen Susanna,Mäenpää Johanna, Kairemo KaleviAlka Arora, Pradeep SethConsuelo Almazan, Katherine M. Kocan,Douglas K. Bergman, Jose C. Garcia-Garcia, Edmour F. Blouin and José de laFuenteMichael W. Epperly, Hongliang Guo,Michael Bernarding, Joan Gretton, MiaJefferson, Joel S. GreenbergerIlana Goldberg-Cohen, Nina S Levy,Andrew P LevyMarisol Gasc!n-Ir"n, Silvia M. Sanz-Gonz#lez and Vicente AndrésToru Suzuki, Masami Horikoshi, andRyozo NagaiW.L. Robert Li, Nagy A. Habib, SteenL. Jensen, Paul Bao, Diping Che, UweR. Müller

103-111 ResearchArticle113-133 ReviewArticle135-151 ReviewArticle153-165 ReviewArticle167-172 ResearchArticle173-179 ResearchArticle181-209 ReviewArticle211-219 ResearchArticle221-228 ReviewArticle229-238 ResearchArticle239-243 ResearchArticle245-254 ReviewArticle255-272 ReviewArticle273-289 ReviewArticleHMG-CoA-reductase inhibitiondependentand independent effects ofstatins on leukocyte adhesionCurrent progress in adenovirus mediatedgene therapy for patients with prostatecarcinomaGene therapy for vascular diseasesAngiogenic gene therapy for improvingislet graft vascularization.G-CSF Receptor-mediated STAT3activation and granulocyte differentiationin 32D cells.Calcium induces apoptosis and necrosis inhematopoetic malignant cells: Evidencefor caspase-8 dependent and FADDautonomouspathwayThe current status and future direction offetal gene therapyThe role of EBV and genomic sequencesin gene expression fromextrachromosomal gene therapy vectorsin mouse liverSite-specific kidney-targeted plasmidDNA transfer using nonviral techniquesHepatocyte-targeted delivery of SleepingBeauty mediates efficient gene transfer invivoPRL-3 as a target for cancer therapyProtective effect of heat shock proteins:potential for gene therapyLung cancer gene therapyAdvances in cationic lipid-mediated genedeliveryTriantafyllos Chavakis, ThomasSchmidt-Wöll, Peter. P. Nawroth, KlausT. Preissner, Sandip M. KanseAhter D. Sanlioglu,, Turker Koksal,Mehmet Baykara, Guven Luleci, BahriKaracay and Salih SanliogluSarah J. George, Filomena de Nigris,Andrew H. Baker, Claudio NapoliNan Zhang, Karen Anthony, KatsunoriShinozaki, Jennifer Altomonte, ZacharyBloomgarden and Hengjiang DongRuifang Xu, Akihiro Kume, YutakaHanazono, Kant M. Matsuda, YasujiUeda, Mamoru Hasegawa, FumimaroTakaku, and Keiya OzawaChristof J. Burek Malgorzata Burek,Johannes Roth, and Marek LosAnna L David, Michael Themis, SimonN Waddington, Lisa Gregory, SuzanneMK Buckley, Megha Nivsarkar, TerryCook, Donald Peebles, Charles HRodeck, Charles CoutelleStephanie M. Stoll, Leonard Meuse,Mark A. Kay, and Michele P. CalosHiroki Maruyama, Noboru Higuchi,Shigemi Kameda, Gen Nakamura, JunichiMiyazaki, and Fumitake GejyoBetsy T. Kren, Siddhartha S. Ghosh,Cheryle L. Linehan, NamitaRoyChowdhury, Perry B. Hackett,Jayanta Roy-Chowdhury, and Clifford J.SteerKoh Vicki, Fu Jianlin, Guo Ke, Lip KuoMing, Li Jie and Zeng QiDavid S. LatchmanKexia Cai, Mai Har Sham, Paul Tam,Wah Kit Lam and Ruian XuBenjamin Martin, Abderrahim Aissaoui,Matthieu Sainlos, Noufissa Oudrhiri,Michelle Hauchecorne, Jean-Pierre

Spencer and Davie: Dynamic histone acetylation and its involvement in transcription30 nm fiber is maintained by the N terminal tails (Davieand Spencer, 2001).The chromatin fiber becomes moderately folded bythe H3 and H4 N terminal tails at physiological ionicstrength. However, the N terminal tails of the four corehistones are required for the chromatin fiber to undergoextensive folding (Tse and Hansen, 1997; Logie et al,1999). At low ionic strength, the chromatin fiber assumesa three-dimensional irregular shape that is stabilized by theglobular domain of H1 and either the H1 tails or the H3 Nterminal tail (Zlatanova et al, 1998; Leuba et al, 1998a).The N terminal tails from histones H2A, H2B and H4 donot have the same effect as H3 on the chromatin fiber.However, the N terminal tail of H3 is 44 amino acids long,whereas histones H4, H2B and H2A have N terminal tailsthat are only 26, 32, and 16 amino acids long, respectively.As a result, the N terminal tail of histone H3 can extendover a significantly larger portion of linker DNAcompared to the other core histones (Leuba et al, 1998b).The H3 N terminus is also positioned close to the pointwhere linker DNA enters and exits the nucleosome, and,therefore, it can undergo extensive interactions with thelinker DNA (Zlatanova et al, 1998).The chromatin fibers within a cell interdigitate withneighboring fibers into a higher order fibrous mass thatimpedes the access of transcription factors to their targetsequences, thereby preventing transcription initiation(Schwarz et al, 1996). At physiological ionic strength, theinteraction of these neighboring fibers with one another ispartly dependent on either the H2A and H2B or the H3and H4 core histone N terminal tails (Davie and Spencer,2001). These fibrous masses are then further organizedinto compact chromosome territories within interphasenuclei (Verschure et al, 1999).In addition to binding linker DNA, the histone Nterminal tails are capable of interacting with other histonesand non-histone chromosomal proteins. The N terminus ofH4 binds to the H2A-H2B dimer of neighboringnucleosomes, and, as such, is thought to assist inchromatin folding (Luger et al, 1997). In yeast, thetranscriptional repressors Sir3, Sir4, and Ssn6/Tup1interact with the H3 and H4 N terminal domains, causingthe associated chromatin to become transcriptionallyrepressed (Grunstein, 1998). Likewise, the DrosophilaGroucho and its mammalian homologues bind to the Nterminal domain of H3 and repress transcription (Palapartiet al, 1997; Fisher and Caudy, 1998). These domains alsointeract with non-histone proteins such as HMG-14 andHMG-17 that promote the unfolding of higher orderchromatin structures (Bustin, 1999).III. Acetylation of the histone Nterminal tailsThe N terminal tails can undergo a series of posttranslationalmodifications at specific amino acidsincluding acetylation, phosphorylation, ubiquitination andmethylation (Spencer and Davie, 1999) (Figure 1). Themost extensively studied of these modifications is dynamicacetylation, a reversible process catalyzed byacetyltransferases and deacetylases which mediate thetransfer of acetyl groups on to and off of the ε-aminogroup of N terminal lysine residues, respectively (Kuo andAllis, 1998).Figure 1. General structure of the core histones and their sites of post-translational modifications. The central globular domain ofeach histone is depicted as a circle with the N and C terminal tails extending towards the left and right sides, respectively. Me, Ac, P, andUb represent methylation, acetylation, phosphorylation, and ubiquitination, respectively. HAT A (histone acetyltransferase) and HDAC(histone deacetylase) represent the enzymes that catalyze the reversible acetylation of lysine residues along the histone N terminal tails.H3 kinase and PP1 (protein phosphatase 1) represent the enzymes responsible for the reversible phosphorylation of H3 serine residue.2

Gene Therapy and Molecular Biology Vol 7, page 3This modification typically occurs on up to fivelysine residues along the H3 and H4 N terminal tails, fourresidues along H2B, and one residue along H2A (Davieand Spencer, 1999). Whether a histone is hypo- orhyperacetylated depends on the net activities ofneighboring histone acetyltransferases and deacetylases.IV. Histone acetyltransferasesThe following is only a brief summary of the histoneacetyltransferases identified to date. For a more detaileddescription of histone acetyltransferases and theirsubstrates, please refer to the following reviews (Sternerand Berger, 2000; Davie and Spencer, 2001; Marmorsteinand Roth, 2001; Bertos et al, 2001). Numeroustranscription co-activators including yGcn5, P/CAF,CBP/p300, Esa1, NuA4, and ACTR/SRC-1 have beenidentified as having intrinsic histone acetyltransferaseactivity (Sterner and Berger, 2000; Davie and Spencer,2001; Klochendler-Yeivin and Yaniv, 2001; Marmorsteinand Roth, 2001). In addition, the DNA-bindingtransactivator ATF-2, the general transcription factorsTAFII250 and Nut1, and the elongation factor Elp3 arehistone acetyltransferases (Marmorstein and Roth, 2001).Histone acetyltransferases generally exist in largecomplexes (Spencer and Davie, 1999). Each histoneacetyltransferase has a different target substrate, and thespecificity for this substrate depends on the proteinsassociated with the histone acetyltransferase (Grant et al,1999). For example, the free full-length form of yeastGcn5 preferentially acetylates H3 in vitro and H3 and H4in vivo (Zhang et al, 1998; Sterner and Berger, 2000;Davie and Spencer, 2001). However, the acetylatingefficiency of yeast Gcn5 for nucleosomal histonesincreases when assembled into high molecular weight,multi-protein complexes referred to as SAGA (Spt-Ada-Gcn5-acetyltransferase) and Ada (Grant et al, 1999). Inaddition, the pattern of histone acetylation for Gcn5assembled into the SAGA complex is distinct from thatexhibited by Gcn5 when assembled into Ada (Grant et al,1999). Similarly, the histone substrate specificity ofindividual human PCAF and yeast Esa1 acetyltransferasesbecomes altered when these enzymes are assembled intomulti-protein complexes (Davie and Spencer, 2001). Thephosphorylation of CBP by ERK1 enhances the activity ofthis acetyltransferase, suggesting that the function ofhistone acetyltransferases may be regulated byphosphorylation events (Liu et al, 1999).V. Histone deacetylasesAs many as 10 histone deacetylases have beenidentified to date (Bertos et al, 2001). Refer to thefollowing reviews (Sterner and Berger, 2000; Bertos et al,2001; Davie and Spencer, 2001; Marmorstein and Roth,2001) for a more detailed description of histonedeacetylases. These deacetylases are divided into 3 classesdefined by their size and sequence homologies to yeastdeacetylases. The class I histone deacetylases areapproximately 400-500 amino acids in length and includeHDACs 1,2,3 and 8. These class I members are nucleartranscriptional co-repressors with homology to the yeastRpd3 deacetylase. The class II histone deacetylases arelarger proteins of approximately 1000 amino acids withstructural homology to yeast Hda1 and include HDACs4,5,6,7,9 and 10 (Davie and Moniwa, 2000; Bertos et al,2001; Guardiola and Yao, 2002). Class III histonedeacetylases are encoded by genes similar to the yeastsilent information regulator (Sir 2) gene (Afshar andMurnane, 1999; Frye, 1999). These deacetylases aredependent on NAD+ and ADP-ribosylase activity (Frye,2000; Imai et al, 2000; Landry et al, 2000).Class I deacetylases are ubiquitously expressed,while class II deacetylases are tissue-, cell-anddifferentiation-specific (Davie and Moniwa, 2000). Bothclasses of deacetylases can deacetylate the four corehistones, however, each deacetylase has a site preference(Davie and Spencer, 2001). Similar to histoneacetyltransferases, the yeast Rpd3 and Hda1 deacetylasesexist in distinct multi-protein complexes, suggesting thatclass I and II deacetylases have distinct biologicalfunctions. Furthermore, the components of thesecomplexes influence the substrate specificity of theseenzymes (Davie and Moniwa, 2000). For example, the freeform of avian HDAC1 preferentially deacetylates free butnot nucleosomal H3. When assembled into a multi-proteincomplex, this deacetylase preferentially deacetylates freeH2B and histones assembled into a nucleosome (Sun et al,1999).Class I deacetylases reside in the nucleus (Davie andMoniwa, 2000). However, the sub-cellular distribution ofclass II deacetylases is not as straight forward. HDACs 4and 5 shuttle between the cytoplasm and the nucleus(Bertos et al, 2001). HDAC7 is predominantly nuclear butbinds to the membrane-associated endothelin receptor Aand most likely functions in the cytoplasm (Lee et al,2001). HDAC6 is strictly cytoplasmic, and HDAC9appears to be both nuclear and cytoplasmic (Zhou et al,2001). HDACs 4,5, and 7 are transcriptional co-repressorsthat interact with MEF2 transcription factors, as well asthe co-repressors N-CoR, BCoR, and CtBP (Bertos et al,2001; Guardiola and Yao, 2002). Similarly, HDAC9interacts with MEF-2 and represses MEF-2-mediatedtranscription (Zhou et al, 2001). HDAC10 resides in thenucleus and the cytoplasm (Guardiola and Yao, 2002). Inthe nucleus, this deacetylase functions as a transcriptionalrepressor when tethered to a promoter (Guardiola and Yao,2002). Interestingly, HDAC6 can interact with ubiquitin.As well, the mammalian homologue of UFD3, a yeastprotein involved in protein ubiquitination, is part of thecytoplasmic mammalian HDAC6 complex (Seigneurin-Berny et al, 2001).VI. The dynamics of histoneacetylationStudies of histone acetylation dynamics indicate thatboth acetylation and deacetylation occur at more than onerate (Covault and Chalkley, 1980; Zhang and Nelson,1988a). In human fibroblasts and mature avian3

Spencer and Davie: Dynamic histone acetylation and its involvement in transcriptionerythrocytes, there are two populations of acetylatedhistones. The first population, which accounts forapproximately 15% of acetylated core histones inhepatoma tissue culture cells, is rapidly hyperacetylated(t 1/2 = 7 to 15 min for monoacetylated H4) and rapidlydeacetylated (t 1/2 = 3 to 7 min). The second population,which accounts for up to 50% of acetylated histones, isslowly acetylated (t 1/2 = 140-300 min for monoacetylatedH4) and then slowly deacetylated (t 1/2 = 30 min) (Covaultand Chalkley, 1980; Zhang and Nelson, 1988a). Similarly,MCF-7 human breast cancer cells also display twopopulations of acetylated H3, H4 and H2B histones: arapidly acetylated one comprising 10% of the total nuclearacetylated histones and a slowly acetylated one thatincludes approximately 50% of acetylated histones (Sun etal, 2001).In immature chicken erythrocytes, approximately 2%of the genome is dynamically acetylated, while the rest iseither frozen in a state of mono- or di-acetylation orunacetylated (Zhang and Nelson, 1988a). The acetylatedhistones in immature avian erythrocytes are divided intotwo populations. In contrast to mature avian erythrocytes,both populations within the immature erythrocytes displaythe same rate of histone acetylation (t 1/2 =12 min formonoacetylated H4). However, in the case of H4, onepopulation is hyperacetylated to tri- or tetra-acetylatedisoforms and then rapidly deacetylated (t 1/2 = 5 min)(referred to as class I). The other population, however, isonly mono- or di-acetylated, and subsequentlydeacetylated at a slower rate (t 1/2 =90 min) (referred to asclass II)(Zhang and Nelson, 1988a; Zhang and Nelson,1988b). Histones H3 and H2B are also class I acetylatedsince butyrate-treated immature chicken erythrocytesdisplay a drastic and rapid decline in tri- and tetraacetylatedH3 and H2B within 10 minutes of incubation inthe absence of butyrate (Spencer and Davie, 2001) (Figure2).VII. The effect of histone acetylationon chromatin structureHistone acetylation affects chromatin structure inseveral ways. One theory suggests that histone acetylationalters nucleosome structure and weakens the interaction ofhistone N terminal tails with DNA (Turner, 1991; Nortonet al, 1989). Histone acetylation also maintains the openconformation of the transcriptionally active nucleosome(Walia et al, 1998). Thus, histone acetylation mayneutralize the positive charges on the N terminal lysineresidues, and loosen the contacts between histones andDNA. However, Gcn5 similarly affects transcription andcell growth whether H3 contains a lysine, arginine, orglutamine at position 14 of its N terminal tail. Similarly,replacement of lysine 8/16 residues with arginine orglutamine does not alter the affect of Gcn5 ontranscription or cell growth (Zhang et al, 1998). Thissuggests that histone acetylation may influencetranscription by mechanisms other than the neutralizationof N terminal lysine residues.Histone acetylation is also thought to disrupt thehigher order folding of chromatin fibers (Garcia-Ramirezet al, 1995; Moore and Ausio, 1997; Hansen, 1997). Atphysiological salt concentrations, acetylated chromatinfibers are salt-soluble, while unacetylated fibers areinsoluble (Ridsdale et al, 1990). However, these fibers areincapable of interacting with other fibers by the process ofoligomerization, and, therefore, are unable to form higherorder structures (Annunziato and Hansen, 2000).Figure 2. Immunoblot analyses of H2B deacetylation. Avian immature erythrocytes were incubated with sodium butyrate for 1 h, andthen incubated in the absence of butyrate for 0, 5, 10, 15 or 30 min. The total nuclear histones from erythrocytes at each time point wereextracted. Twenty µg of acid-extracted histones were electrophoresed on an Acid-Urea-Triton 15% polyacrylamide gel. The resolvedproteins were then transferred to nitrocellulose and immunostained with an antibody to hyperacetylated H2B (Serotec, UK). 0, 1, 2, 3,and 4 designate un-, mono-, di-, tri-, and tetra-acetylated histone isoforms, respectively.4

Gene Therapy and Molecular Biology Vol 7, page 5The acetylation of only 12 out of 28 lysine residues perhistone octamer promotes transcription approximately 15fold in vitro, and affects chromatin similar to theproteolytic removal of the core histone N terminal tails(Tse et al, 1998; Annunziato and Hansen, 2000). As aresult, acetylation of the histone N terminal tails is thoughtto facilitate transcription by disrupting the folding of thechromatin fiber, as well as inter-fiber interactions. Such anevent would allow transcription factors access to theirtarget DNA binding sites. In support of this, the treatmentof estrogen-responsive cells with estrogen induces H3 andH4 acetylation along the TATA sequence of the PS2promoter, subsequently, exposing the TATA binding siteand allowing the TATA binding protein to bind to this site(Sewack et al, 2001). In addition, chromatinimmunoprecipitation studies show an enrichment ofhyperacetylated H3 and H4 along the promoter regions ofseveral genes including the vitamin A and vitamin D geneswhen transcriptionally activated (Chen et al, 1999; Kadoshand Struhl, 1998; Parekh and Maniatis, 1999; Krebs et al,1999). As well, the binding of estrogen to its receptorleads to the recruitment of p300/CBP to the promoter ofestrogen-responsive genes (Chen et al, 1999).In addition to disrupting chromatin fiber-fiberinteractions, histone acetylation disrupts the interactionsbetween the histone N terminal tails and non-nucleosomalproteins or DNA. For example, H3 and H4hyperacetylation abolish Ssn6-Tup1-mediatedtranscriptional repression (Watson et al, 2000). Thehistone N terminal domains display α-helical structureswhen assembled into the nucleosome (Annunziato andHansen, 2000). This α-helical character increases uponacetylation (Wang et al, 2000). Histone acetyltransferasesmay positively influence transcription by altering thestructure of the N terminal tails and perturbing theinteractions of these tails with proteins that represstranscription. However, histone acetylation may also beassociated with transcriptional repression since theheterochromatin of several organisms contains H4acetylated at lysine 12 (Turner, 2000; Turner et al, 1992).As well, loss of the yeast RPD3 histone deacetylase causesan increase in the silencing of telomeric DNA (DeRubertis et al, 1996).It has also been suggested that histone acetylationplays a role in marking the state of genetic activity orinactivity from one cell generation to the next, therebyepigenetically determining the long-term transcriptionalcompetence of a gene (Turner, 1998). However, recentevidence shows that catalytically active histoneacetyltransferases and histone deacetylases are unable toacetylate or deacetylate chromatin in situ during mitosis(Kruhlak et al, 2001). Moreover, these enzymes becomespatially reorganized and displaced from condensingchromosomes. Instead, it appears that the spatialorganization of these enzymes relative to euchromatin andheterochromatin plays an important role in determining thepost-mitotic activation of a gene (Kruhlak et al, 2001).VIII. The effect of histone acetylationon ATP-dependent chromatin remodelingBesides playing a role in transcription factor binding,histone acetylation may also be fundamental for ATPdependentchromatin remodeling. These type ofcomplexes use ATP hydrolysis as a source of energy toalter nucleosome and chromatin structure and enhancetranscription factor binding to nucleosomal DNA-bindingsites (Davie and Moniwa, 2000). For a more detaileddescription of ATP-dependent chromatin remodelingfactors refer to the following reviews (Kingston andNarlikar, 1999; Davie and Moniwa, 2000). While thesecomplexes can alter the chromatin structure of transactivatorbinding sites, they are unable to activatetranscription alone (Gregory et al, 1999). The recruitmentof the SWI/SNF chromatin remodeling complex to nuclearreceptor and BRCA1-regulated genes is thought toincrease nucleosome fluidity, and facilitate the subsequentbinding of transcription factors to affected regions (Singhet al, 2000).In the case of the yeast HO gene, the binding of thechromatin remodeling factor, SWI/SNF leads to therecruitment of the SAGA histone acetyltransferasecomplex (Krebs et al, 1999). These two complexesfacilitate the binding of a second activator, SBF, whichmost likely recruits TBP and other components of the preinitiationcomplex. ATP-dependent chromatin remodelingare also involved in transcription repression (Davie andMoniwa, 2000). Because of this, ATP-dependentchromatin remodeling complexes may increase the rate atwhich a chromatin region fluctuates between an active andrepressed structure (Kingston and Narlikar, 1999). Iffactors are present that stabilize chromatin structure andpromote transcriptional repression, then the remodelingcomplex will drive the chromatin into a repressed state byallowing the transcriptional repressors to associate withthe chromatin. However, if transcriptional activators bindto the remodeled chromatin instead, then the remodelingcomplexes will drive the chromatin structure to atranscriptionally active state. The subsequent binding ofhistone acetyltransferases and activating complexes to thischromatin structure will then “fix” it in an active state(Kingston and Narlikar, 1999). In support of this, theelimination of SAGA acetyltransferase activity preventsproper chromatin remodeling at the PHO8 promoter invivo (Gregory et al, 1999).However, ATP-dependent chromatin remodelingcomplexes do not always bind chromatin before histoneacetyltransferases. In the case of the interferon β promoter,the enhanceosome assembles at a nucleosome-freeenhancer region of this gene and initially recruits Gcn5 toacetylate the nucleosome positioned over the TATA boxand transcription start site (Agalioti et al, 2000). This leadsto the recruitment of the CBP-PolII holoenzyme complex,and CBP subsequently recruits SWI/SNF. Therefore, insome cases, the SWI/SNF complex prefers acetylatedchromatin as a substrate (Agalioti et al, 2000). The BRG1sub-unit of the SWI/SNF complex contains abromodomain, and this type of domain can interact withacetylated histones (Winston and Allis, 1999; Cairns et al,5

Spencer and Davie: Dynamic histone acetylation and its involvement in transcription1999). The presence of acetylated histones along apromoter may increase the affinity of the SWI/SNFcomplex to this gene region. In support of this, SWI/SNFwas recruited to a promoter by a transactivator, however,its retention was enhanced when the histones along thisregion were acetylated (Hassan et al, 2001). Incubation ofthese nucleosomal arrays with SAGA and NuA4 increasedthis retention (Hassan et al, 2001). Furthermore, histoneacetyltransferases have been shown to increase the rate ofgene induction by accelerating ATP-dependent chromatinremodeling (Barbaric et al, 2001). The order ofrecruitment for chromatin-remodeling activities and thefunction of these complexes in gene activation orrepression is most likely gene-specific, and dependent onthe combination of transcription factors bound to thepromoter.IX. The effect of acetylation on nonhistoneproteinsHistone acetyltransferases can also acetylatetranscription factors (p53, ACTR, EKLF, estrogenreceptor, MyoD, GATA-1, E2F1), non-histonechromosomal proteins (HMG), components of thetranscription machinery (TFIIE, TFIIF), the nuclear importprotein importin, tubulin, and flap endonuclease-1 (Fen-1),an enzyme involved in DNA metabolism (Bannister et al,2000; Chen et al, 1999; Imhof et al, 1997; Munshi et al,1998; Hasan et al, 2001; Wang et al, 2001; Polesskaya etal, 2000; Herrera et al, 1999; Zhang and Bieker, 1998;Hung et al, 1999; L'Hernault and Rosenbaum, 1985;Martinez-Balbas et al, 2000). The acetylation of p53 andMyoD increases their binding affinity for DNA (Gu andRoeder, 1997; Polesskaya et al, 2000). As well, acetylationof E2F1 extends the half-life of this protein (Martinez-Balbas et al, 2000). Thus, along with modifying chromatinstructure, acetyltransferases may function in transcriptionby altering the DNA-binding properties of transcriptionfactors or enhancing the stability of transcription factors.The acetylation of HMGI(Y) plays an important rolein viral-induced interferon β gene activation as well as theinactivation of this event (Parekh and Maniatis, 1999).Upon infection, the enhanceosome assembles at theinterferon gene promoter with the help of HMGI(Y). Atthe same time, CBP and P/CAF are recruited to theinterferon β gene promoter where they acetylate H3 andH4 and, in combination with the enhanceosome, activatetranscription of the interferon β gene. Following induction,CBP acetylates HMGI(Y) which decreases its DNAbinding affinity and causes the disruption of theenhanceosome complex. In addition, p300 binds toestrogen receptor α in the absence of estrogen andacetylates lysine residues within the hinge/ligand bindingdomain of this receptor. This event suppresses thesensitivity of the receptor to ligand (Wang et al, 2001).The evidence from these studies suggests that the theory ofacetylation stimulating transcriptional activity is notalways true.Acetyltransferases may also function in otherbiological processes. The acetylation of flap endonuclease-1 by p300 reduces its ability to bind DNA, as well as itsnuclease activity, while acetylation of importin-alpha byCBP promotes its interaction with importin-beta in vitro(Hasan et al, 2001; Bannister et al, 2000). Furthermore, theacetylation of ACTR by another acetyltransferase suggeststhat acetylation may be a cascading event involved insignal transduction (Kouzarides, 2000; Marmorstein andRoth, 2001).X. Global versus targeted histoneacetylationNumerous studies have displayed an enrichment ofacetylated H3 and H4 along the promoter regions oftranscriptionally active genes. For example, activation ofthe human interferon gene induces H3 and H4hyperacetylation over 2-3 nucleosomes within thepromoter region (Parekh and Maniatis, 1999). Likewise,the yeast Gcn5 histone acetyltransferase complexacetylates histones only in the HO gene promoter (Krebset al, 1999). Hormone-mediated transcriptional activationalso involves the H3 and H4 hyperacetylation over thepromoter regions of hormone-responsive genes (Chen etal, 1999; Sewack G.F. et al, 2001). A similar scenariooccurs for histone deacetylation where the yeastSin3-Rpd3 histone deacetylase complex deacetylateshistones over a 1-2 nucleosome range within the promoterof a repressed gene (Kadosh and Struhl, 1998).In a recent study, the CpG island of thetranscriptionally active chicken carbonic anhydrase genewas associated with higher levels of acetylated histonescompared to the near-by promoter region (Myers et al,2001). The acetylation of H3 and H4 along this gene wasgreatest at the CpG island and showed a drastic drop atapproximately 1.5 kilobases into the transcribed region.Similarly, the chicken thymidine kinase gene displayedelevated levels of hyperacetylated histones along its CpGisland (Crane-Robinson et al, 1999). High levels ofhyperacetylated histones were also mapped to the chickenGAPDH promoter, which is located within a CpG island(Myers et al, 2001). The regions downstream of thispromoter that do not contain CpG islands displayed asharp drop in the levels of hyperacetylated H3 and H4. Aswell, chromatin fragments containing CpG islands areenriched in highly acetylated H3 and H4 isoforms (Taziand Bird, 1990). These findings suggest that histonehyperacetylation is a feature of CpG islands. In a recentstudy, acetylated histones were mapped to CpG islandslocated both within the promoter and regions downstreamfrom the transcription start site of a reporter gene (Cervoniand Szyf, 2001). The significance of histone acetylationalong CpG islands is not known. However, whenassociated with acetylated histones, a methylated DNAsequence will become demethylated (Cervoni and Szyf,2001). Because the interaction of demethylase with DNAis thought to be the limiting step in DNA demethylation,the acetylation of histones associated with CpG islandsmay increase the accessibility of demethylase to its targetDNA sequence (Cervoni and Szyf, 2001).However, histone hyperacetylation does not alwaysappear to be promoter- or CpG island-targeted. H46

Gene Therapy and Molecular Biology Vol 7, page 7acetylated at lysine 16 (H4Ac16) is distributed along theentire length of X-linked genes targeted by the malespecificlethal dosage compensation. The promoter regionsof these genes are associated with lower levels of H4Ac16compared to the middle and 3’ regions (Smith et al, 2001).Similarly, pol I- and pol II-transcribed genes containelevated levels of H4Ac16, while the levels of H4Ac12 aresignificantly elevated in yeast and Drosophilaheterochromatin (Johnson et al, 1998; Braunstein et al,1996). As well, the chicken β A -globin gene does notcontain a CpG island, but displays high levels ofwidespread H3 and H4 acetylation (Myers et al, 2001).Acetylated lysine residues are also located throughout thec-myc gene, as well as the entire adult chicken β-globindomain (Hebbes et al, 1994; Madisen et al, 1998; Myers etal, 2001).While a particular histone acetyltransferase can berecruited to and acetylate the histones along a specificgene, recent evidence suggests that some histoneacetyltransferases can also globally affect the acetylationof many genes in a non-targeted manner. Depletion ofEsa1, an acetyltransferase specifically recruited to theribosomal protein and heat shock promoters, causes adramatic decrease in H4 acetylation over many regions ofthe genome without affecting the transcription of manygenes (Reid et al, 2000). Similarly, the acetylation of theyeast PHO5 promoter by Esa1 and Gcn5, and thesubsequent deacetylation of this region by HDA1 andRpd3 also results in the widespread histoneacetylation/deacetylation of three separate chromosomalregions making to 22 kb of DNA (Vogelauer et al, 2000).Thus, the promoter-targeted acetylation activity of somehistone acetyltransferases and deacetylases may occur in abackground of non-targeted histone acetylation that ismediated by these same enzymes and not required fortranscription. However, this global acetylation can, insome cases, be targeted to particular regions of thegenome. The expression of the C/EBPα transcriptionfactor in GHFT1-5 pituitary cells causes an increase in thelevels of acetylated H3 at pericentromeric chromatindomains (Zhang et al, 2001). CBP may be the histoneacetyltransferase associated with C/EBPα, since thisenzyme concentrates at pericentromeric chromatin duringC/EBPα expression (Schaufele et al, 2001). The globalactivity of these enzymes may maintain the balance ofacetylated and deacetylated histones throughout thegenome or regions of the genome and prevent the histonesalong a gene from becoming transiently or permanentlyfully acetylated.The hyperacetylation of histones on regionsdownstream from the promoter suggests that histoneacetylation may function in transcriptional elongation. Forexample, Elp3, a 60-kilodalton subunit of theelongator/RNAPII holoenzyme has histoneacetyltransferase activity and is able to acetylate all fourcore histones in vitro (Wittschieben et al, 1999). Thishistone acetyltransferase activity is essential for theelongator function of Elp3 in vivo (Wittschieben et al,2000). Furthermore, the removal of Gcn5 and Elp3acetyltransferase activity from yeast cells causeswidespread transcription defects (Wittschieben et al,2000). Gcn5 functions in the transcription of only a subsetof genes. Therefore, Elp3 histone acetyltransferase activitymust be important for the transcription of a significantnumber of genes. Other evidence suggesting a role forhistone acetylation in transcriptional elongation comesfrom observations that transcription by T7 RNApolymerase through a nucleosome occurs at a similar rateon nucleosomal templates containing either tailless orhyperacetylated histones (Protacio et al, 2000). As well,H3 and H4 hyperacetylation is necessary to maintain thetranscriptionally active nucleosome in an openconformation for transcriptional elongation (Walia et al,1998).As a result, a cell may contain two types of histoneacetyltransferases with respect to the transcriptionalprocess: those involved in initiation, and those involved inelongation. Histone acetyltransferases required for theinitiation process would either enhance transcription factorbinding to promoter/enhancer target regions by one orseveral of the mechanisms previously described, whileacetyltransferases required for elongation would increasethe accessibility of elongation factors to the DNA withincoding regions. In support of this theory, the p300 histoneacetyltransferase interacts specifically with initiationcompetentform of RNA polymerase II, while PCAFinteracts with the elongation-competent form (Cho et al,1998). Furthermore, p300 associates with the promoterregion of an estrogen-responsive gene only duringimmediate exposure to estrogen when transcription isinitiated rather than during subsequent re-initiation stagesof transcription (Shang et al, 2000). Salt-soluble chromatinfragments enriched in active genes are associated withseveral unidentified histone acetyltransferases (Hebbesand Allen, 2000). Whether these acetyltransferasesfunction in initiation and/or elongation remains to bedetermined.Different histone acetyltransferases have differenthistone substrates along certain regions of specific targetgenes. The histone deacetylase Rpd3 preferentiallyacetylates lysine 5 of H4 at only a select number of genes(Rundlett et al, 1998). As well, the yeast histoneacetyltransferase, Esa1, interacts only with the promoterregions of ribosomal protein genes (Reid et al, 2000).Histone deacetylases along with nuclear receptor corepressorscan exist in discrete nuclear bodies (Downes etal, 2000). Similarly, nuclear matrix-associatedpromyelocytic leukemia bodies contain PML proteins thatbind and concentrate CBP into discrete domains (Boisvertet al, 2001). The differential levels of hyperacetylatedhistones observed on different regions of active genes maybe explained by the proximity of histone acetyltransferasesand deacetylases to specific regions of these genes.Regions situated close to regions of high acetyltransferaseactivity are more frequently acetylated than deacetylated,while regions close to deacetylases are deacetylated moreoften than acetylated.As well, cellular context may influence theacetylation status of histones along specific gene regions.Histone acetyltransferases and deacetylases exist in large7

Spencer and Davie: Dynamic histone acetylation and its involvement in transcriptionmulti-protein complexes, and the types of proteinsassociated with these enzymes can determine theirsubstrate specificity (Grant et al, 1999). For example, inone cell type a specific histone acetyltransferase may existin a complex capable of acetylating H4, while, in anothercell type this same enzyme may be associated withdifferent proteins and have a substrate specificity for H3.In some cases, the ability of histoneacetyltransferases and deacetylases to occupy a particulargene region may be transient (Shang et al, 2000). Within15-20 minutes following estradiol exposure, the histoneacetyltransferases AIB1 and p300 within MCF-7 humanbreast cancer cells associate with the estrogen-responsivecathepsin D promoter. RNA polymerase associates shortlyfollowing this event. This association most likely initiatestranscription since significant levels of transcription areobserved 45 min after estrogen stimulation. Theassociation of these factors then starts to decline 60 minfrom the initial time of estrogen treatment. A few minutesbefore these acetyltransferases are removed, the levels ofCBP and PCAF histone acetyltransferases associated withthe cathepsin D promoter starts to rise and peak between60 and 75 minutes. However, the levels of cathepsin Dtranscription are significantly reduced after 75 minutes.The levels of CBP and PCAF and the rate of transcriptionthen drop sharply at 90 minutes. Approximately 100minutes after estrogen stimulation, the AIB1, CBP andPCAF acetyltransferases all assemble on the promoter inthe same order as before, and the rate of transcriptionsimultaneously increases. Similar results were alsoobserved for the PS2 estrogen-responsive promoter inMCF-7 cells, and the cathepsin D promoter in ECC-1endometrial cells, showing that estrogen-inducedtranscription involves the cyclical assembly of histoneacetyltransferases along the promoters of estrogenresponsivegenes.Even though the association of histoneacetyltransferases with estrogen-responsive promoters iscyclical after estrogen stimulation, the levels of acetylatedhistones along the promoter region never drop to the levelsobserved in estrogen-deplete conditions when theacetyltransferases are displaced. Once transcription hasbeen initiated, histone acetylation may maintain the openstructure of an entire gene, and increase the accessibilityof the promoter and downstream regions to the RNApolymerase complex for subsequent rounds of initiationand elongation. Such an event may increase the rate oftranscription (Orphanides and Reinberg, 2000).Determining the structure of chromatin after initiation, butbefore and after elongation will help elucidate the functionof acetylation in elongation.XII. Transcription and the dynamicsof histone acetylationThe exact function of dynamic histone acetylation intranscription is unknown. Nuclear fractionation studiesindicate that the nuclear distribution of class I, but notclass II, acetylated histones closely follows that of thetranscriptionally active β-globin and histone H5 genes(Hendzel et al, 1991). The majority of histoneacetyltransferase and deacetylase activity, class Iacetylated histones, and transcriptionally active β-globinand histone H5 genes are located in the insoluble nuclearmaterial which contains the nuclear matrix (Hendzel et al,1991). As well, the nuclear matrix is the site oftranscription (Davie, 1995).We recently showed that intronic regions of thetranscriptionally active β-globin gene, andtranscriptionally competent, DNAse I-sensitive butinactive ε-globin genes are associated with class Iacetylated histones (Spencer and Davie, 2001). Thisassociation was shown for chromatin fragments in bothsalt-soluble and nuclear matrix-containing nuclearfractions. Of the two sequences, the β-globin intronappeared to have a higher concentration of class Iacetylated histones, while the ε-globin intron wasassociated with a mosaic of class I and class II acetylatedhistones. These findings suggest that the N terminal tailsof the core histones situated on transcriptionally activegenes contact nuclear-matrix associated histoneacetyltransferases and deacetylases in a rapid and transientmanner, while the frequency of contact between theseenzymes and the histones along transcriptionallycompetent genes is less. In support of this, the entirechicken β- A globin gene, which has a high rate oftranscription, was associated with higher levels of H3 andH4 acetylation when compared to genes transcribed atslower rates (GAPDH, carbonic anhydrase (Myers et al,2001). As well, multiple histone acetyltransferases areassociated with chromatin fragments enriched intranscriptionally active genes (Hebbes and Allen, 2000).Thus, dynamic histone acetylation may function toselectively retain transcriptionally active genes at sites oftranscription within the nuclear matrix (Spencer andDavie, 2001).In fact, evidence from a recent study on estrogenresponsivehuman breast cancer cells suggests thatexposure to estrogen changes the dynamics of histoneacetylation by altering the balance of histoneacetyltransferases and deacetylases along different regionsof estrogen-responsive genes (Sun et al, 2001). In humanbreast cancer cells, exposure to estradiol causes therecruitment of acetyltransferases and the subsequenthyperacetylation of histones at the promoter region ofestrogen-responsive genes (Chen et al, 1999). In addition,exposure of hormone-responsive human breast cancercells to estrogen reduces the rate of histone deacetylationwithout affecting the rate of histone acetylation, or thesub-nuclear location, level or activity of class I and IIhistone deacetylases (Sun et al, 2001). Instead, exposure toestrogen alters the distribution of the estrogen receptor andhistone acetyltransferases (SRC-1 and SRC-3) by causingboth types of factors to become tightly associated with thenuclear matrix (Stenoien et al, 2001; Sun et al, 2001).Thus, the binding of estrogen to the estrogen receptor maycause the estrogen receptor to recruit histoneacetyltransferases from other nuclear regions to thepromoter region of estrogen-responsive genes (Figure 3).At present, a large emphasis is placed on the role of8

Gene Therapy and Molecular Biology Vol 7, page 9histone acetyltransferases in transcriptional initiation andelongation. However, as previously mentioned, histoneacetylation is a dynamic event resulting from thecombined activities of histone acetyltransferases anddeacetylases. Thus, more attention must be given tounderstanding how acetyltransferases and deacetylasesfunction together at specific sites along transcriptionallyactive genes to fully appreciate the role of dynamic histoneacetylation in transcription.XII. The histone codeThe histone N terminal tails undergo several posttranslationalmodifications mediated by a variety ofenzymes. Research in the field of gene expression hasfocussed primarily on determining the function of eachmodification in transcription. However, a new concept hasemerged referred to as the “histone code” (Strahl andAllis, 2000; Jenuwein and Allis, 2001). This term proposesthat the different post-translational modificationsoccurring on one or more histone tails act either togetheror in sequence to form recognition sites for specificproteins involved in distinct cellular functions.Furthermore, these modifications may positively ornegatively influence the affect of one another on specificcellular functions.Evidence from several recent studies suggests thathistone phosphorylation and acetylation may functiontogether to promote gene expression. For example, thestimulation of mammalian cells by epidermal growthfactor causes the sequential phosphorylation of Ser10, andacetylation of Lys14 on H3 (Cheung et al, 2000).Moreover, Gcn5 preferentially associates with a Ser10phosphorylated form of H3 over a non-phosphorylatedform (Cheung et al, 2000). Recently, the phosphorylationof H3 Ser10 by the Snf1 kinase was shown to lead toGcn5-mediated acetylation at the INO1 promoter (Lo et al,2001).Thus, the recruitment of a kinase complex to specificpromoters may cause Ser10 phosphorylation and eitherincrease the affinity of histone acetyltransferasecomplexes for nucleosomes or increase acetyltransferasecatalytic activity (Lo et al, 2000).However, the affect of one post-translationalmodification on another may not always be positive.Heterochromatic silencing requires the methylation ofLys9 on H3 by the lysine methyltransferase Su(var)39(Rea et al, 2000). The methylation of Lys9 inhibitsphosphorylation of H3 at Ser10 possibly by hindering theaccess of kinases to this serine residue (Rea et al, 2000).Thus, methylation of Lys9 may impair transcription byinhibiting phosphorylation events required fortranscriptional stimulation (Berger, 2001).This finding, however, needs to be furtherinvestigated since immunoprecipitation studies haveidentified an association between CBP and a histonemethyltransferase that specifically targets lysines 4 and 9of H3 without significantly affecting the ability of CBP toefficiently acetylate other H3 lysine residues (Vandel andTrouche, 2001).Figure 3. Proposed model for the effect of estradiol on the distribution of histone acetyltransferases and histone deacetylases in humanbreast cancer cells. In the absence of estradiol (left), histone acetyltransferases (HAT) such as CBP, SRC-1, SRC-3, and PCAF occupythe same chromatin regions as histone deacetylases (HDAC) such as HDAC1, and HDAC2. Upon addition of estradiol (right), theestrogen receptor (ER) is recruited to nuclear matrix sites and associates with the estrogen response element of estrogen responsivegenes. When bound to estradiol, the ER recruits histone acetyltransferases from other nuclear regions, thereby altering the balance ofhistone acetyltransferases and deacetylases along specific chromatin regions.9

Spencer and Davie: Dynamic histone acetylation and its involvement in transcriptionA recent study mapping the distribution of di-methylatedlysine 9 on H3 across the chicken β-globin domain duringerythropoiesis showed that regions enriched in methylatedlysine 9 were depleted of di-acetylated H3 (K9 and K14).However, H3 acetylation correlated with lysine 4methylation, suggesting that transcriptional activation isassociated with H3 methylated at K4, as well as withacetylated H3 and H4 isoforms (Litt et al, 2001).Likewise, in Tetrahymena, methylated Lys4 of H3 isfound only in transcriptionally active macronuclei (Strahlet al, 1999).AcknowledgmentsResearch supported by grants from the CanadianInstitutes of Health Research (CIHR) (MT-9186,RO-15183), CancerCare Manitoba, and the U.S. ArmyMedical and Materiel Command Breast Cancer ResearchProgram (#DAM17-00-1-0319), and the National CancerInstitute of Canada with funds from the Canadian CancerSociety. A CIHR Senior Scientist Award to J.R.D. and aU.S. Army Medical and Materiel Command Fellowship toV.A.S. are gratefully acknowledged.ReferencesAfshar G, and Murnane JP (1999) Characterization of a humangene with sequence homology to Saccharomyces cerevisiaeSIR2. Gene 234, 161-168.Agalioti T, Lomvardas S, Parekh B, Yie J, Maniatis T, andThanos D (2000) Ordered recruitment of chromatinmodifying and general transcription factors to the IFN-βpromoter. Cell 103, 667-678.Annunziato AT, and Hansen J C (2000) Role of histoneacetylation in the assembly and modulation of chromatinstructures. Gene Expr 9, 37-61.Bannister AJ, Miska EA, Gorlich D, and Kouzarides T (2000)Acetylation of importin-alpha nuclear import factors byCBP/p300. Curr Biol 10, 467-470.Barbaric S, Walker J, Schmid A, Svejstrup JQ, and Horz W(2001) Increasing the rate of chromatin remodeling and geneactivation--a novel role for the histone acetyltransferaseGcn5. EMBO J 20, 4944-4951.Berger SL (2001) An embarrassment of niches: the manycovalent modifications of histones in transcriptionalregulation. Oncogene 20, 3007-3013.Bertos NR, Wang AH, and Yang XJ (2001) Class II histonedeacetylases: structure, function, and regulation. BiochemCell Biol 79, 243-252.Boisvert FM, Kruhlak MJ, Box AK, Hendzel MJ, and Bazett-Jones DP (2001) The transcription coactivator CBP is adynamic component of the promyelocytic leukemia nuclearbody. J Cell Biol 152, 1099-1106.Braunstein M, Sobel RE, Allis CD, Turner BM, and Broach JR(1996) Efficient transcriptional silencing in Saccharomycescerevisiae requires a heterochromatin histone acetylationpattern. Mol Cell Biol 16, 4349-4356.Bustin M (1999) Regulation of DNA-dependent activities by thefunctional motifs of the high-mobility-group chromosomalproteins. Mol Cell Biol 19, 5237-5246.Cairns BR, Schlichter A, Erdjument-Bromage H, Tempst P,Kornberg RD, and Winston F (1999) Two functionallydistinct forms of the RSC nucleosome-remodeling complex,containing essential AT hook, BAH, and bromodomains.Mol Cell 4, 715-723.Cervoni N and Szyf M (2001) Demethylase activity is directedby histone acetylation. J Biol Chem 276, 40778-40787.Chen H, Lin RJ, Xie W, Wilpitz D, and Evans RM (1999)Regulation of hormone-induced histone hyperacetylation andgene activation via acetylation of an acetylase. Cell 98, 675-686.Cheung P, Tanner KG, Cheung WL, Sassone-Corsi P, Denu JM,and Allis CD (2000) Synergistic coupling of histone H3phosphorylation and acetylation in response to epidermalgrowth factor stimulation. Mol Cell 5, 905-915.Cho H, Orphanides G, Sun X, Yang XJ, Ogryzko V, Lees E,Nakatani Y, and Reinberg D (1998). A human RNApolymerase II complex containing factors that modifychromatin structure. Mol Cell Biol 18, 5355-5363.Covault J and Chalkley R (1980) The identification of distinctpopulations of acetylated histone. J Biol Chem 255, 9110-9116.Crane-Robinson C, Myers FA, Hebbes TR, Clayton AL, andThorne AW (1999). Chromatin immunoprecipitation assaysin acetylation mapping of higher eukaryotes. MethodsEnzymol 304, 533-547.Davie JR (1995) The nuclear matrix and the regulation ofchromatin organization and function. Int Rev Cytol 162A,191-250.Davie JR and Moniwa M (2000) Control of chromatinremodeling. Crit Rev Eukaryot Gene Expr 10, 303-325.Davie JR and Spencer VA (1999) Control of histonemodifications. J Cell Biochem Suppl 32-33, 141-148.Davie JR and Spencer VA (2001) Signal transduction pathwaysand the modification of chromatin structure. Prog NucleicAcid Res Mol Biol 65, 299-340.De Rubertis F, Kadosh D, Henchoz S, Pauli D, Reuter G, StruhlK, and Spierer P (1996) The histone deacetylase RPD3counteracts genomic silencing in Drosophila and yeast.Nature 384, 589-591.Downes M, Ordentlich P, Kao HY, Alvarez JG, and Evans RM(2000) Identification of a nuclear domain with deacetylaseactivity. Proc Natl Acad Sci USA 97, 10330-10335.Fisher AL and Caudy M (1998) Groucho proteins: transcriptionalcorepressors for specific subsets of DNA-bindingtranscription factors in vertebrates and invertebrates. GenesDev 12, 1931-1940.Frye RA (1999) Characterization of five human cDNAs withhomology to the yeast SIR2 gene: Sir2-like proteins (sirtuins)metabolize NAD and may have protein ADPribosyltransferaseactivity. Biochem Biophys Res Commun260, 273-279.Frye RA (2000) Phylogenetic classification of prokaryotic andeukaryotic Sir2-like proteins. Biochem Biophys ResCommun 273, 793-798.Garcia-Ramirez M, Rocchini C, and Ausio J (1995) Modulationof chromatin folding by histone acetylation. J Biol Chem270, 17923-17928.Grant PA, Eberharter A, John S, Cook RG, Turner BM, andWorkman JL (1999) Expanded lysine acetylation specificityof Gcn5 in native complexes. J Biol Chem 274, 5895-5900.Gregory PD, Schmid A, Zavari M, Munsterkotter M, and HorzW (1999) Chromatin remodeling at the PHO8 promoterrequires SWI-SNF and SAGA at a step subsequent toactivator binding. EMBO J 18, 6407-6414.10

Gene Therapy and Molecular Biology Vol 7, page 11Grunstein M (1998) Yeast heterochromatin: regulation of itsassembly and inheritance by histones. Cell 93, 325-328.Gu W and Roeder RG (1997) Activation of p53 sequencespecificDNA binding by acetylation of the p53 C-terminaldomain. Cell 90, 595-606.Guardiola AR and Yao TP (2002) Molecular cloning andcharacterization of a novel histone deacetylase HDAC10. JBiol Chem 277, 3350-3356.Hansen JC (1997) The core histone amino-termini: combinatorialinteraction domains that link chromatin structure withfunction. Chemtracts Biochem Mol Biol 10, 56-69.Hansen JC, Tse C, and Wolffe AP (1998) Structure and functionof the core histone N-termini: more than meets the eye.Biochemistry 37, 17637-17641.Hasan S, Stucki M, Hassa PO, Imhof R, Gehrig P, Hunziker P,Hubscher U, and Hottiger MO (2001) Regulation of humanflap endonuclease-1 activity by acetylation through thetranscriptional coactivator p300. Mol Cell 7, 1221-1231.Hassan AH, Neely KE, and Workman JL (2001) Histoneacetyltransferase complexes stabilize swi/snf binding topromoter nucleosomes. Cell 104, 817-827.Hebbes TR and Allen SC (2000) Multiple histoneacetyltransferases are associated with a chicken erythrocytechromatin fraction enriched in active genes. J Biol Chem275, 31347-31352.Hebbes TR, Clayton AL, Thorne AW, and Crane-Robinson C(1994) Core histone hyperacetylation co-maps withgeneralized DNase I sensitivity in the chicken β-globinchromosomal domain. EMBO J 13, 1823-1830.Hendzel MJ, Delcuve GP, and Davie JR (1991) Histonedeacetylase is a component of the internal nuclear matrix. JBiol Chem 266, 21936-21942.Herrera JE, Sakaguchi K, Bergel M, Trieschmann L, Nakatani Y,and Bustin M (1999) Specific acetylation of chromosomalprotein HMG-17 by PCAF alters its interaction withnucleosomes. Mol Cell Biol 19, 3466-3473.Hung HL, Lau J, Kim AY, Weiss MJ, and Blobel GA (1999)CREB-Binding protein acetylates hematopoietic transcriptionfactor GATA-1 at functionally important sites. Mol Cell Biol19, 3496-3505.Imai S, Armstrong CM, Kaeberlein M, and Guarente L (2000)Transcriptional silencing and longevity protein Sir2 is anNAD-dependent histone deacetylase. Nature 403, 795-800.Imhof A, Yang XJ, Ogryzko VV, Nakatani Y, Wolffe AP, andGe H (1997) Acetylation of general transcription factors byhistone acetyltransferases. Curr Biol 7, 689-692.Jenuwein T and Allis CD (2001) Translating the histone code.Science 293, 1074-1080.Johnson CA, O'Neill LP, Mitchell A, and Turner BM (1998)Distinctive patterns of histone H4 acetylation are associatedwith defined sequence elements within both heterochromaticand euchromatic regions of the human genome. NucleicAcids Res 26, 994-1001.Kadosh D and Struhl K (1998) Targeted recruitment of the Sin3-Rpd3 histone deacetylase complex generates a highlylocalized domain of repressed chromatin in vivo. Mol CellBiol 18, 5121-5127.Kimura H and Cook PR (2001) Kinetics of core histones inliving human cells: little exchange of H3 and H4 and somerapid exchange of H2B. J Cell Biol 153, 1341-1353.Kingston RE and Narlikar GJ (1999) ATP-dependent remodelingand acetylation as regulators of chromatin fluidity. GenesDev 13, 2339-2352.Klochendler-Yeivin A and Yaniv M (2001). Chromatin modifiersand tumor suppression. Biochim Biophys Acta 1551, M1-10.Kouzarides T (2000) Acetylation: a regulatory modification torival phosphorylation? EMBO J 19, 1176-1179.Krebs JE, Kuo MH, Allis CD, and Peterson CL (1999) Cellcycle-regulated histone acetylation required for expression ofthe yeast HO gene. Genes Dev 13, 1412-1421.Kruhlak MJ, Hendzel MJ, Fischle,W, Bertos NR, Hameed S,Yang XJ, Verdin E, and Bazett-Jones DP (2001) Regulationof global acetylation in mitosis through loss of histoneacetyltransferases and deacetylases from chromatin. J BiolChem 276, 38307-38319.Kuo MH and Allis CD (1998) Roles of histone acetyltransferasesand deacetylases in gene regulation. Bioessays 20, 615-626.L'Hernault SW and Rosenbaum JL (1985) Chlamydomonasalpha-tubulin is posttranslationally modified by acetylationon the epsilon-amino group of a lysine. Biochemistry 24,473-478.Litt MD, Simpson M, Gaszner M, Allis CD, and Felsenfeld G(2001) Correlation between histone lysine methylation anddevelopmental changes at the chicken β-globin locus.Science 293, 2453-2455.Landry J, Slama JT, and Sternglanz R (2000) Role of NAD(+) inthe deacetylase activity of the SIR2-like proteins. BiochemBiophys Res Commun 278, 685-690.Lee,HJ, Chun M, and Kandror KV (2001) Tip60 and HDAC7interact with the endothelin receptor a and may be involvedin downstream signaling. J Biol Chem 276, 16597-16600.Leuba SH, Bustamante C, van Holde K, and Zlatanova J (1998a)Linker histone tails and N-tails of histone H3 are redundant:scanning force microscopy studies of reconstituted fibers.Biophys J 74, 2830-2839.Leuba SH, Bustamante C, Zlatanova J, and van Holde K (1998b)Contributions of linker histones and histone H3 to chromatinstructure: scanning force microscopy studies on trypsinizedfibers. Biophys J 74, 2823-2829.Liu YZ, Thomas NS, and Latchman DS (1999) CBP associateswith the p42/p44 MAPK enzymes and is phosphorylatedfollowing NGF treatment. Neuroreport 10, 1239-1243.Lo WS, Duggan L, Tolga NC, Emre, Belotserkovskya R, LaneWS, Shiekhattar R, and Berger SL. (2001) Snf1--a histonekinase that works in concert with the histoneacetyltransferase Gcn5 to regulate transcription. Science 293,1142-1146.Lo WS, Trievel RC, Rojas JR, Duggan L, Hsu JY, Allis CD,Marmorstein R, and Berger SL (2000) Phosphorylation ofserine 10 in histone H3 is functionally linked in vitro and invivo to Gcn5-mediated acetylation at lysine 14. Mol Cell 5,917-926.Logie C, Tse C, Hansen JC, and Peterson CL (1999) The corehistone N-terminal domains are required for multiple roundsof catalytic chromatin remodeling by the SWI/SNF and RSCcomplexes. Biochemistry 38, 2514-2522.Luger K, Mader AW, Richmond RK, Sargent DF, and RichmondTJ (1997) Crystal structure of the nucleosome core particle at2.8 A resolution. Nature 389, 251-260.Madisen L, Krumm A, Hebbes TR, and Groudine M (1998) Theimmunoglobulin heavy chain locus control region increaseshistone acetylation along linked c-myc genes. Mol Cell Biol18, 6281-6292.11

Spencer and Davie: Dynamic histone acetylation and its involvement in transcriptionMarmorstein R and Roth SY (2001) Histone acetyltransferases:function, structure, and catalysis. Curr Opin Genet Dev 11,155-161.Martinez-Balbas MA, Bauer UM, Nielsen SJ, Brehm A, andKouzarides T (2000) Regulation of E2F1 activity byacetylation. EMBO J 19, 662-671.Moore SC and Ausio J (1997) Major role of the histones H3-H4in the folding of the chromatin fiber. Biochem Biophys ResCommun 230, 136-139.Munshi N, Merika M, Yie J, Senger K, Chen G, and Thanos D(1998) Acetylation of HMG I(Y) by CBP turns off IFN-βexpression by disrupting the enhanceosome. Mol Cell 2,457-467.Myers FA, Evans DR, Clayton AL, Thorne AW, and Crane-Robinson C (2001) Targeted and extended acetylation ofhistones H4 and H3 at active and inactive genes in chickenembryo erythrocytes. J Biol Chem 276, 20197-20205.Norton VG, Imai BS, Yau P, and Bradbury EM (1989) Histoneacetylation reduces nucleosome core particle linking numberchange. Cell 57, 449-457.Orphanides G and Reinberg D (2000) RNA polymerase IIelongation through chromatin. Nature 407, 471-475.Palaparti A, Baratz A, and Stifani S (1997) TheGroucho/transducin-like enhancer of split transcriptionalrepressors interact with the genetically defined aminoterminalsilencing domain of histone H3. J Biol Chem 272,26604-26610.Parekh BS and Maniatis T (1999) Virus infection leads tolocalized hyperacetylation of histones H3 and H4 at the IFNβpromoter. Mol Cell 3, 125-129.Polesskaya A, Duquet A, Naguibneva I, Weise C, Vervisch A,Bengal E, Hucho F, Robin P, and Harel-Bellan A (2000)CREB-binding protein/p300 activates MyoD by acetylation.J Biol Chem 275, 34359-34364.Protacio RU, Li G, Lowary PT, and Widom J (2000) Effects ofhistone tail domains on the rate of transcriptional elongationthrough a nucleosome. Mol Cell Biol 20, 8866-8878.Rea S, Eisenhaber F, O'Carroll D, Strahl BD, Sun ZW, SchmidM, Opravil S, Mechtler K, Ponting CP, Allis CD, andJenuwein T (2000) Regulation of chromatin structure by sitespecifichistone H3 methyltransferases. Nature 406, 593-599.Reid JL., Iyer VR, Brown PO, and Struhl K (2000) Coordinateregulation of yeast ribosomal protein genes is associated withtargeted recruitment of Esa1 histone acetylase. Mol Cell 6,1297-1307.Ridsdale JA, Hendzel MJ, Delcuve GP, and Davie JR (1990)Histone acetylation alters the capacity of the H1 histones tocondense transcriptionally active/competent chromatin. JBiol Chem 265, 5150-5156.Rundlett SE, Carmen AA, Suka N, Turner BM, and Grunstein M(1998) Transcriptional repression by UME6 involvesdeacetylation of lysine 5 of histone H4 by RPD3. Nature392, 831-835.Schaufele F, Enwright JF, Wang X, Teoh C, Srihari R, EricksonR, MacDougald OA, and Day RN (2001) CCAAT/EnhancerBinding Protein alpha Assembles Essential CooperatingFactors in Common Subnuclear Domains. Mol Endocrinol15, 1665-1676.Schwarz PM, Felthauser A, Fletcher TM, and Hansen JC (1996)Reversible oligonucleosome self-association: dependence ondivalent cations and core histone tail domains. Biochemistry35, 4009-4015.Seigneurin-Berny D, Verdel A, Curtet S, Lemercier C, Garin J,Rousseaux S, and Khochbin S (2001) Identification ofComponents of the Murine Histone Deacetylase 6 Complex:Link between Acetylation and Ubiquitination SignalingPathways. Mol Cell Biol 21, 8035-8044.Sewack GF, Ellis TW, and Hansen U (2001) Binding of TATAbinding protein to a naturally positioned nucleosome isfacilitated by histone acetylation. Mol Cell Biol 21, 1404-1415.Shang Y, Hu X, DiRenzo J, Lazar MA, and Brown M ( 2000)Cofactor dynamics and sufficiency in estrogen receptorregulatedtranscription. Cell 103, 843-852.Singh H, Sekinger EA, and Gross DS (2000) Chromatin andcancer: causes and consequences. J Cell Biochem Suppl 35,61-68.Smith ER, Allis CD, and Lucchesi JC (2001) Linking globalhistone acetylation to the transcription enhancement of X-chromosomal genes in Drosophila males. J Biol Chem 276,31483-31486.Spencer VA and Davie JR (1999) Role of covalent modificationsof histones in regulating gene expression. Gene 240, 1-12.Spencer VA and Davie JR (2001) Dynamically acetylatedhistone association with transcriptionally active andcompetent genes in the avian adult β-globin gene domain. JBiol Chem 276, 34810-34815.Stenoien DL, Patel K, Mancini MG, Dutertre M, Smith CL,O'Malley BW, and Mancini MA (2001) FRAP reveals thatmobility of oestrogen receptor-α is ligand- and proteasomedependent.Nat Cell Biol 3, 15-23.Sterner DE and Berger SL (2000) Acetylation of histones andtranscription-related factors. Microbiol Mol Biol Rev 64,435-459.Strahl BD and Allis CD (2000) The language of covalent histonemodifications. Nature 403, 41-45.Strahl BD, Ohba R, Cook RG, and Allis CD (1999) Methylationof histone H3 at lysine 4 is highly conserved and correlateswith transcriptionally active nuclei in Tetrahymena. ProcNatl Acad Sci USA 96, 14967-14972.Sun JM, Chen HY, and Davie JR (2001) Effect of Estradiol onhistone acetylation dynamics in human breast cancer cells. JBiol Chem 276, 49435-49442.Sun JM, Chen HY, Moniwa M, Samuel S, and Davie JR (1999)Purification and characterization of chicken erythrocytehistone deacetylase 1. Biochemistry 38, 5939-5947.Tazi J and Bird A (1990) Alternative chromatin structure at CpGislands. Cell 60, 909-920.Tse C and Hansen JC (1997) Hybrid trypsinized nucleosomalarrays: identification of multiple functional roles of theH2A/H2B and H3/H4 N-termini in chromatin fibercompaction. Biochemistry 36, 11381-11388.Tse C, Sera T, Wolffe AP, and Hansen JC (1998) Disruption ofhigher-order folding by core histone acetylation dramaticallyenhances transcription of nucleosomal arrays by RNApolymerase III. Mol Cell Biol 18, 4629-4638.Turner BM (1991) Histone acetylation and control of geneexpression. J Cell Sci 99 ( Pt 1), 13-20.Turner BM (1998) Histone acetylation as an epigeneticdeterminant of long-term transcriptional competence. CellMol Life Sci 54, 21-31.Turner BM (2000) Histone acetylation and an epigenetic code.Bioessays 22, 836-845.12

Gene Therapy and Molecular Biology Vol 7, page 13Turner BM, Birley AJ, and Lavender J (1992) Histone H4isoforms acetylated at specific lysine residues defineindividual chromosomes and chromatin domains inDrosophila polytene nuclei. Cell 69, 375-384.Vandel L and Trouche D (2001) Physical association betweenthe histone acetyl transferase CBP and a histone methyltransferase. EMBO Rep 2, 21-26.Verschure PJ, van Der Kraan, I, Manders EM, and van Driel R(1999) Spatial relationship between transcription sites andchromosome territories. J Cell Biol 147, 13-24.Vogelauer M, Wu J, Suka N, and Grunstein M (2000) Globalhistone acetylation and deacetylation in yeast. Nature 408,495-498.Walia H, Chen HY, Sun JM, Holth LT, and Davie JR (1998)Histone acetylation is required to maintain the unfoldednucleosome structure associated with transcribing DNA. JBiol Chem 273, 14516-14522.Wang C, Fu M, Angeletti RH, Siconolfi-Baez L, Reutens AT,Albanese C, Lisanti MP, Katzenellenbogen BS, Kato S,Hopp T, Fuqua SA, Lopez GN, Kushner PJ, and Pestell RG(2001). Direct acetylation of the estrogen receptor α hingeregion by p300 regulates transactivation and hormonesensitivity. J Biol Chem 276, 18375-18383.Wang X, Moore SC, Laszckzak M, and Ausio J (2000)Acetylation increases the α-helical content of the histonetails of the nucleosome. J Biol Chem 275, 35013-35020.Watson AD, Edmondson DG, Bone JR, Mukai Y, Yu Y, Du W,Stillman DJ, and Roth SY (2000) Ssn6-Tup1 interacts withclass I histone deacetylases required for repression. GenesDev 14, 2737-2744.Winston F and Allis CD (1999) The bromodomain: a chromatintargetingmodule? Nat Struct Biol 6, 601-604.Wittschieben BO, Fellows J, Du W, Stillman DJ, and SvejstrupJQ (2000) Overlapping roles for the histone acetyltransferaseactivities of SAGA and elongator in vivo. EMBO J 19,3060-3068.Wittschieben BO, Otero G, de Bizemont T, Fellows J,Erdjument-Bromage H, Ohba R, Li Y, Allis CD, Tempst P,and Svejstrup JQ (1999) A novel histone acetyltransferase isan integral subunit of elongating RNA polymerase IIholoenzyme. Mol Cell 4, 123-128.Zhang DE and Nelson DA (1988b) Histone acetylation inchicken erythrocytes. Rates of acetylation and evidence thathistones in both active and potentially active chromatin arerapidly modified. Biochem J 250, 233-240.Zhang DE and Nelson DA (1988a) Histone acetylation inchicken erythrocytes. Rates of deacetylation in immature andmature red blood cells. Biochem J 250, 241-245.Zhang W and Bieker JJ (1998) Acetylation and modulation oferythroid Kruppel-like factor (EKLF) activity by interactionwith histone acetyltransferases. Proc Natl Acad Sci USA 95,9855-9860.Zhang W, Bone JR, Edmondson DG, Turner BM, and Roth SY(1998) Essential and redundant functions of histoneacetylation revealed by mutation of target lysines and loss ofthe Gcn5p acetyltransferase. EMBO J 17, 3155-3167.Zhang WH, Srihari R, Day RN, and Schaufele F (2001)CCAAT/enhancer-binding protein alpha alters histone H3acetylation at large subnuclear domains. J Biol Chem 276,40373-40376.Zhou X, Marks PA, Rifkind RA, and Richon VM (2001) Cloningand characterization of a histone deacetylase, HDAC9. ProcNatl Acad Sci USA 98, 10572-10577.Zlatanova J, Leuba SH, and van Holde K (1998) Chromatin fiberstructure: morphology, molecular determinants, structuraltransitions. Biophys J 74, 2554-2566.Virginia A. Spencer and James R. Davie13

Spencer and Davie: Dynamic histone acetylation and its involvement in transcription14

Gene Therapy and Molecular Biology Vol 7, page 15Gene Ther Mol Biol Vol 7, 15-23, 2002Tumor therapy using radiolabelled antisenseoligomers- aspects for antiangiogenetic strategy andpositron emission tomographyReview ArticleKalevi JA Kairemo 1* , Mark Lubberink 2 , Mikko Tenhunen 3 , Antti P Jekunen 4Department of Nuclear Medicine 1 and Hospital Physics 2 Uppsala University, Uppsala SwedenDepartment of Oncology 3 and Department of Clinical Pharmacology 4 , Helsinki University Central Hospital, Helsinki,Finland__________________________________________________________________________________*Correspondence: Kalevi J A Kairemo, MD, PhD, MSc (Eng); Professor, Department of Nuclear Medicine, Uppsala UniversityHospital, Sweden; Tel. +46-18-611 1006; Fax. +46-18-611 4124; e-mail: kalevi.kairemo@onkologi.uas.lul.seKey words: antisense therapy, oligonucleotides, phosphorus radioisotopes, sulphur radioisotopes, AIDS, cancer, dosimetry, positronemission tomographyReceived: 17 January 2002; accepted: 29 January, 2002; electronically published: July 2003SummaryAngiogenesis provides a putative target for radiochemotherapy as endothelial cells on vascular wall are sensitive forradiation and by destructing of one endothelial cell may lead to death hundred of tumor cells. Endothelial cells inthe angiogenic vessels within solid tumors express several proteins that are absent or faintly expressing inestablished blood vessels, including α v integrins (Hammes, 1996) and receptors for certain angiogenic growthfactors (Hanahan, 1997) (Risau, 1997). Recently, vascular endothelial cell growth factor (VEGF)-inducedinvasiveness has been inhibited specifically by ETS-1 antisense oligonucleotide. ETS-1 gene expression can beinduced, while there are several other systems with constant expression. In this paper, we extent use of oligos fromconventional biokinetic studies to therapeutic use by comparing radioactive oligos to peptide counterparts.Radiolabelled oligos have a potential of having both direct antisense inhibition and radiation effects. Previously wehave shown theoretically that oligonucleotide therapy may be effective with internally labelled (P-32, P-33 and S-35)oligodeoxynucleotide phosphorothioates. This has also been demonstrated in vitro using P-33 (Kairemo et al, 1999).We investigate also the possibility of using 15-mer oligodeoxynucleotide phosphorothioates (oligos) or oligomers inwhich the phosphate-ribose backbone has been replaced with polyamide backbone (peptide nucleic acids). Theabsorbed organ doses of these radiolabelled compounds were estimated from biodistribution data. Subcellularbiodistribution was used in evaluation of the best targeting inside the cell with one oligomer. Our results indicatethat oligos can give significantly up to 130-fold higher absorbed organ doses in oligos than in peptides. Mainly this isdue to slower biokinetics of oligos (35-fold slower half-lives). For imaging, positron emitters such as F-18 and Br-76,offer an advantage for radiopharmacokinetic studies (Wu at al., 2000). We have therefore calculated the subcellulardosimetry for these isotopes in different cell dimensions (nuclear diameter 6-16µm, cellular diameter 12-20µm).I. IntroductionAngiogenesis is a cascade of processes involvingboth soluble angiogenic factors and insoluble extracellularmatrix factors (Jekunen and Kairemo, 1997). Solublemultiple molecules, that induce angiogenesis, are releasedby both tumor cells and host cells, including endothelialcells, epithelial cells, mesothelial cells, and leukocytes.These processes provide several targets for developmentof angiogenesis inhibitors. We have used twoangiogenetic factors in our model; tie tyrosine kinasereceptor and ets, representing a factor participating andinducing angiogenesisOn the basis of amino acid sequence and structuralsimilarities, receptor tyrosine kinases can be divided intoseveral families (Ullrich and Schlessinger, 1990). Tie isthe protein product of a recently described receptortyrosine kinase cDNA, which together with tek defines anew subfamily. The tie gene is mandatory for the normalgrowth and differentiation of endothelial cells during fetal15

Kairemo et al: Oligonucleotide radiotherapydevelopment (Korhonen, 1992). It is abundantly expressedin vascular endothelia during development, and in somemegakaryoblastic and erythroleukemia cell lines; as wellas tieRNA accumulates in the epithelium of local vesselsduring ovulation and wound healing (Korhonen, 1992).Tie receptor has an important role in the angiogenesisassociated with melanoma metastasis (Kaipainen, 1994).Radioantibodies against tie receptor have been used intargeting studies in vivo with success (Kairemo et al,1996). As the location of tie receptor is at the outer cellmembrane, receptor is easily reachable and effects ofradiation and receptor blocking should occur immediately,which may be beneficiary for the radioantibody treatment.For further development of these receptors the crucialpoint is to find inducers for normally low levels. Ligandsfor endothelial cell receptors tyrosine kinases, Tie-1 andTie-2 are not known. Ligands with agonistic andantagonistic activities for Tie-2 have now been identified:angiopoetin 1 is an activating ligand for Tie 2 andregulates blood vessel maturation (Suri, 1996), whileangiopoetin 2 serves as antagonist (Maisonpierre, 1997).The ETS family proteins are transcription factors thatbind to the regulatory control region of certain genes viaETS binding motif, which has been found in numerousgenes including proteases and receptor tyrosine kinases(Wasylyk, 1993). ETS regulates the expression ofproteases and migration of endothelial cells, and in fact,the induction of ETS-1 mRNA is a mutual phenomenon inendothelial cells stimulated with angiogenic growthfactors (Iwasaka, 1996). It has also been shown that ETS -1 antisense oligo markedly reduced the DNA- ETScomplex diminishing the responsiveness to the stimulus ofangiogenic factor (Iwasaka, 1996). Induction of expressionof ETS gene is faster and more prominent than proteinexpression providing better although transient target fortherapy.The specificity resides in the sequence of oligo,which interacts with its complementary mRNA, but onlyminimally with noncomplementary structures. Theantisense oligo, through the formation of a mRNA-DNAduplex, specifically prevents the translation of that mRNAinto protein (Figure 1). For oligos to be effective antisenseagents, they first must enter the cells and achieveappropriate concentration in the correct intracellularcompartment. Cellular nucleases are highly potent indigesting phosphodiester oligos. Thus several nucleaseresistant oligos have been developed. Phosphorothioateoligo has a non-bridging oxygen atom replacing a sulphuratom. Peptide nucleic acid (PNA) is an oligomer in whichthe charged phosphate-ribose backbone has beeneliminated and replaced with an uncharged backbone(Egholm 1992) and PNAs have been reported to resistnuclease and protease degradation (Egholm 1993).Oligos bind to serum albumin and other proteins withlow affinity and distribute to all peripheral tissues with thekidneys and liver accumulating most of the drug. They arecleared by slow metabolism with an elimination half-lifeup to 50 hrs. The biokinetics of GEM 91 phosphorothioateoligodeoxynucleotide has been evaluated in six AIDSpatients, where the plasma mean residence time variedfrom 24.7 to 49.6 hrs, the mean being 41.7 ± 3.6 hrs(Zhang, 1995a).Figure 1. Schematic presentation of radionanotargeting16

Gene Therapy and Molecular Biology Vol 7, page 17Phosphorothioate oligodeoxynucleotides have severaladvantages: they are relatively resistant to destruction bynucleases; they have good aqueous solubility; theyhybridize efficiently with target RNA with relatively highspecificity; they are relatively efficiently taken up by cells;and they are widely used in automated oligonucleotidesynthesizers (Zhang, 1995b). Phosphorothioateoligodeoxynucleotides labelled internally either withsulphur or phosphorus do not require any extra couplingtechniques as in the case with transition metals. Thetherapeutic possibilities of radiolabelled antisenseoligodeoxynucleotides or peptides are still unknown, andone of the basic questions in radiotherapy is the optimalsource of radiation. Here we have estimated dosimetricproperties of different radiolabels on oligonucleotides andpeptides at cellular level, that could be predicted fromexisting data. The aim of this study was to calculateinternal radiation dose from the known data and assess thesuitability of different isotopes for the labels. Macroscopicdoses were calculated for oligonucleotides labelled with76 Br, 111 In, 90 Y and 211 At, as examples of positron emitters,Auger-electron emitters, high-energy beta radiationemitters, and alpha emitting nuclides.We have previously shown by using calculationsfrom the biodistribution data of oligonucleotidephosphorothioates in a xenograft model thatoligonucleotide radiotherapy can optimally be given withP-32 and P-33 (Kairemo et al, 1996). Calculations cansuggest recommendable source of radiation, and thusallow a proper selection of the optimal label. By selectinga radiation source the penetrability of radiation can becontrolled and severe side effects may be avoidedefficiently.II. Dosimetric calculationsThe accumulated dose from radionuclides usedinternal labelling of oligos, phosphorus-32 (P-32),phosphorus-33 (P-33) and sulphur-35 (S-35) wasestimated using the MIRD (Medical Internal RadiationDose) formalism, the basic equations of which areD = Ã × S (1)andà =x∫0A(t)dt= A 0T effln2where D, A, S and T eff refer to absorbed doses, activities,geometric factors and effective half-lives.The effective half-life can be calculated usingmonoexponential kinetics byTeffTbTf=T + Tbfwhere T b is the biological half-life of the oligomer andT f physical half-life of the specific radionuclide.Two different situations were investigated to calculate therelative dose.(2)(3)Case 1: rapid kinetics compared with physical decay:T T , Tf>>b 1 b 2D 1= à 1= A 01×T T f b1× T f+ T b2=D 2à 2A 02T f+ T b1T fT b2= A 01A 02× T b1T b2(4)Case 2: rapid kinetics compared with very slow kinetics:T >> T >> Tb2 f b1D 1= à 1= A 01×T fT b1× T f+ T b2=D 2à 2A 02T f+ T b1T fT b2= A 01A 02× T b1T f(5)The absorbed dose of P-32, P-33 and S-35 labelledoligonucleotides were estimated using publishedbiodistribution data with several oligonucleotides andmouse models. (Crooke et al, 1996) have investigatedpharmacokinetics of a 20-mer oligodeoxynucleotidephosphorothioate (ISIS 3082) and its 2_-propoxyphosphorothioate (ISIS 9045) in mice. Thisoligodeoxynucleotide inhibits the expression of mouseintercellular adhesion molecule (Crooke et al, 1996).(Dewanjee et al, 1994a) have published the data in mousefor 15-mer oligonucleotide sequence coupled withdiethylenetriamine pentaacetate (DTPA)-isothiocyanate.(Mardirossian et al, 1997) have published thepharmacokinetic and stability data for radiolabeled aminederivatized15-base DNA oligomer in mice. Thepharmacokinetics of the compounds were expected not tochange depending on P-32, P-33 or S-35 labelling. Herewe also studied positron emitters F-18 and Br-76, betaemitterY-90, Auger-emitter In-111 and alpha-emitter At-211. The whole organ uptakes as percent of the injectedactivity were used.III. Dosimetric dataTable I summarizes actual delivered doses in liver,kidney and tumor. Data was collected from differentpublished reports on pharmacokinetic data with differentS-35 labelled oligonucleotides and mouse models. Theliver doses in mouse models varied from 0.003 to 30Gy/MBq. The kidney doses in the same animal modelsvaried from from 0.01 to 35 Gy/MBq. The values in thesemodelswere all within tolerance limits of radiotoxicityexcept those for ISIS 9045. In the mammary tumor modelthe observed kidney dose of 9.1 Gy/MBq for P-32 (notshown) is close to the maximum tolerated dose, whereasfor S-35 the absorbed radiation dose in kidneys wasacceptable 1.3 Gy/MBq. The tumor dose was 1.0 Gy peradministered MBq.Table I shows that oligos deliver up to 130-foldhigher organ doses (including tumors) than peptide nucleic17

Kairemo et al: Oligonucleotide radiotherapyacids of the same size. The PNAs have rapid biokinetics;the half-lives are approximately 35-fold faster than thoseof oligo phosphorothiates. The lipophilic oligophosphorothiate 9045 with is 2´-propoxy modificationgives very high organ doses. All other 15-21-mer oligosgive identical liver doses. The smallest kidney dose wascalculated for the 15-mer oligo, and both ISIS 3082 and2105 had 3.2-fold higher kidney dose. Despite theheterogeneity of the origin of the input data and usedapproximations of the time-activity distribution, consistentresults were obtained. Subcellular dosimetry was appliedin situations as described in Figure 2. The followingresults were obtained as shown in Figure 3. Itdemonstrates subcellular dosimetric data in different celldimensions (nuclear diameter 3-8 µm,cellular diameter 6-10 µm) for positron emitters F-18 and Br-76 in fourdifferent oligodeoxynucleotide target systems. If highnuclear DNA target is used,large variation especially inBr-76 dose can be observed. This means that the cellnuclear dose is very much dependent on cell dimensions.If highly inductable RNA target is used, variation is muchsmaller as as in less extreme subcellular concentrations ofoligodeoxynucleotide.Kinetics of oligonucleotides are highly dependent onthe chemistry of the sugar-phosphate backbone of themolecules, and of the length of the molecules. Here, the20h SUVs and cellular distribution reported by (Wu et al,2000) for antisense 76 Br-phosphorothioate oligonucleotidesof length 20 mer was used, combined with octreotidekinetics. For tumour, a SUV of 17.5 was used, as likely foroctreoscan, since no oligonucleotide data was found.Cellular uptake values in tumour are assumptions. Theonly data on oligonucleotide kinetics found was made(Tavitian et al, 1998), describing only the first 90 min afteradministration of three different oligonucleotides inbaboons as measured by PET with 18 F.Macroscopic doses were calculated foroligonucleotides labelled with 76 Br, 111 In, 90 Y and 211 At, asexamples of positron emitters, Auger-electron emitters,high-energy beta radiation emitters, and alpha emittingnuclides (Table III). Absorbed doses were calculatedusing the Mirdose 3.1 program by Stabin (Stabin, 1996),except for 211 At where gamma radiation was ignored andlocal absorbtion of all alpha and beta radiation energy wasassumed. Kidney, liver, spleen and remainder of the bodywere used as source organs.Using cellular S-value data (Bolch, 1999), nucleus tonucleus absorbed doses were calculated for the subcellulardistributions (Table II, IV), and compared to macroscopicdoses. The mean number of decays in each cell wascalculated assuming a uniform distribution of the activitywithin each organ, and assuming spherical cells with adiameter of 14 µm and a nucleus diameter of 10 µm.IV. DiscussionHere, we have emphasized the possible role ofradiolabelled antisense oligos in the anti-angiogenetictherapy. It is known that new tumor vessels due toangiogenesis differ from capillaries in normal tissues dueto properties of regulation of blood flow and alsointerstitial fluid pressure in tumors is elevated. Molecules,related to angiogenesis in tumors may retain longer intumors and thus give for a longer effect for therapeuticagents. The ETS1 gene has a direct role in angiogenesis:the antisense oligonucleotides directed against the ETS1gene thus altered a cellular property of endothelial cellsthat is correlated with the ability of the cells to migratethrough basement membranes (Chen 1997). While ETS1regulates the expression of various proteins by endothelialcells related their growth, it is also regulating variousproteins affecting coagulation and other factors whichperform important endothelial functions.Table I. The calculated organ doses for different oligomers in mouse modelsOligomerPeptide nucleic acid,15-merc-myc, antisense, 15-merISIS 308220-merISIS 9045, 20-merISIS 2105, 21-merInitial activity (% ofinjected dose)0.19% (liver)1.45 % (kidney)6.95 % (liver)5.15 % (kidney)18 % (liver)25 % (kidney)45 % (liver)12 % (kidney)18 % (liver)25 % (kidney)Biologic halflife,T b (hours)5.1% (liver)4.8 (kidney)178.2 (liver)170.7 (kidney)62 (liver)112 (kidney)∞ (liver)∞ (kidney)62 (liver)112 (kidney)Liver dose (S-35)Gy/ MBq0.078 %0.003 Gy/ MBq100 %0.4 Gy/ MBq90%0.4 Gy/ MBq7620 % (S-35)30 Gy/ MBq90%0.4 Gy/ MBqc-myc, antisense, 15-mer 11.0 % (tumor) 194 (tumor) 100 % (tumor)1.0 Gy/ MBqKidney dose (S-35) Gy/ MBq0.79%0.01 Gy/ MBq100 %1.3 Gy/ MBq320 %4.0 Gy/ MBq2710 % (S-35)35 Gy/ MBq320 %4.0 Gy/ MBqReferenceMardirossianet al, 1997Dewanjeeet al, 1994Crooke et al,1996Crooke et al,1996Crooke et al,1996Dewanjeeet al, 199418

Gene Therapy and Molecular Biology Vol 7, page 19Figure 2: Schematic model for cellular calculations in real andextreme situations. Subcellular dosimetry was applied in thesesituationsDose calculationsFigure 3. It demonstrates subcellular dosimetric data in different cell dimensions (nuclear diameter 3-8 µm, cellular diameter 6-10 µm)for positron emitters F-18 and Br-76 in four different oligodeoxynucleotide target systems.19

Kairemo et al: Oligonucleotide radiotherapyTable II. Shows subcellular distributions calculated by the nucleus to nucleus absorbed dosesThe following SUVs at 20h after injection were given by Wu et al, 1999:6 mer 12 mer 20 mer 30 merKidney 53.1 13.3 17.8 1.9Liver 0.5 0.5 8.6 12.3Spleen 0.5 0.5 3.4 5.1The following subcellular distribution was assumed for 20 mer, approximately as in Wu et al, 1999:NucleusRestKidney 30% 70%Liver 30% 70%Tumour 80%, 50% 20%, 50%Table III shows the calculated absorbed doses for a number of organs and tumours.Macroscopic absorbed doses (mGy/MBq)Organ 111 In 90 Y 76 Br 211 AtLiver 0.63 4.41 1.29 4.90Spleen 0.43 3.59 0.96 1.91Kidney 0.95 10.1 2.35 8.74Whole body (mGy/MBq) 0.13 0.57 0.23 0.54Tumour, 100g 1.03 12.0 2.65 10.9Tumour, 0.01g 0.54 3.29 0.59 10.9Organ 111 In 90 Y 76 Br 211 AtLiver 0.63 4.41 1.29 4.90Table IV shows the cellular dosesAverage nucleus self-dose (mGy/MBq), and percentage of average nucleus absorbed doseOrgan 111 In 90 Y 76 Br 211 AtLiver 0.03 (4.0%) 0.004 (0.09%) 0.006 (0.5%) 0.21 (4.4%)Kidney 0.05 (4.8%) 0.007 (0.04%) 0.011 (0.5%) 0.38 (4.4%)Tumour, 100g 0.15 (14.5%), 0.09 0.023 (0.19%), 0.015 0.036 (1.4%), 0.023 1.26 (11.6%), 0.79Tumour, 20g 0.15 (27.8%), 0.09 0.023 (0.71%), 0.015 0.036 (6.1%), 0.023 1.26 (11.6%), 0.79Furthermore, ETS1 has expression in B and Tlymphocytes and thymus. Vascular endothelial growthfactor (VEGF) is an endothelial cell-specific mitogen thatpromotes angiogenesis in solid tumors. The VEGFinducedinvasiveness was inhibited by ETS1 antisenseoligonucleotides but not by a sense control (Chen 1997).Antisense-VEGF has been successfully used to controltumor growth and it may provide another basis for thedevelopment of antiangiogenic gene therapy (Saleh 1996).Rat glioma cells were transfected with a eukaryoticexpression vector bearing an antisense-VEGF cDNA andtransplanted into nude mice: growth of the antisense-VEGF cell lines was inhibited compared to control cells,despite the fact that they have a faster division time invitro. These tumors had fewer blood vessels and a higherdegree of necrosis explaining the reduced tumor size(Saleh 1996). Also, human melanoma cells transfectedwith sense vascular permeability factor (VPF)/VEGFexpressed and secreted large amounts of mouseVPF/VEGF and formed well-vascularized tumors withhyperpermeable blood vessels and minimal necrosis innude/SCID mice (Claffey, 1996).20

Gene Therapy and Molecular Biology Vol 7, page 21VPF/VEGF promoted melanoma growth bystimulating angiogenesis and constitutive VPF/VEGFexpression dramatically promoted tumor colonization inthe lung up to 50-fold of that of controls (Claffey, 1996).Minimal sequence information required for high-affinitybinding to VEGF is contained in 29-36-nucleotide motifsfor the development of potent and specific VEGFantagonists (Jellinek, 1994). Transforming growth factoralpha (TGF-alpha) has been shown to induce VEGF/VPFin normal human epidermal keratinocytes in vitro (Smyth,1997). By using a 19-mer antisense phosphorothioateoligodeoxynucleotide complementary to bases 6-24relative to the translational start site of the VEGF/VPFmRNA, modulation of VEGF/VPF induction by TGFalphawas examined in vitro. The anti-sense oligo wascapable of inhibiting VEGF/VPF RNA and protein tonear-basal levels providing an antiangiogenetic strategy(Smyth, 1997).Previously, it was shown that phosphorothioateantisense oligonucleotides directed against basic fibroblastgrowth factor (bFGF) mRNA inhibited both the growth ofKaposi's sarcoma (KS) cells derived from differentpatients and the angiogenic activity associated with thesecells, including the induction of KS-like lesions in nudemice (Ensoli, 1994). These effects were due to the blockof the production of bFGF which is required by AIDS-KScells to enter the cell cycle and which, after release,mediates angiogenesis (Ensoli, 1994).We describe oligos to be superior to peptide oligos invehicle characteristics of radiation. Whilephosphorothioate oligos have rapid disappearance fromplasma within an hour, and a biexponential elimination,their half lives apparently longer than in the peptideoligos. Although a phosphorothioate oligo leaves plasmarapidly, it requires days to leave the whole body. There isalso significant extravascular accumulation of greater than50 % of the injected dose over a period of 3 to 12 hr.Furthermore, uptake into tissues is not saturated, as someuptake is happening even at 28 days during continuosinfusion (Iversen et al, 1994). The oligos are extensivelyeliminated in the urine over first 3 days after bolusinjection. Distribution to, and tissue accumulation anddistribution is tissue-specific (Iversen et al, 1992, 1994).It can be addressed that the behavior of the radiationat small distancies is crucial. This would be crucial inoligoradiotherapy with highest possible uptake in thetarget cell and minimal radiation toxicity to surroundingnormal cells. Here, oligos are transferring radioactivesource inside the cell and finally to close contact withtarget RNA macromolecule.We have shown earlier that for subcellular targetinginternal labels give the lowest variation in estimatedabsorbed nuclear doses in our cell model with givendimensions (nuclear diameter 6-16 µm, cellular diameter12-20 µm) (Kairemo et al, 1996). From the published data(Crooke et al, 1995) for ISIS 2105,21-mer oligonucleotidethe following subcellular distribution was obtained: thenuclearuptake 0.2 %, cytoplasmic uptake 1.3 %, and cellsurface uptake 0.3 % of injected dose. In this anti-humanpapilloma virus (HPV) model these uptakes as % cellvolume are 11 % for nucleus, 72% for cytoplasm and 17% for cell surface. We calculated concentrationdistributions including the uniform distribution andpublished biodistribution. We normalized the resultsrelative to the uniform distribution and the effect of theactivity outside the cell was not taken into account, whichassumption lead to the maximal possible inhomogeneity inabsorbed dose distribution within a single cell.We have also calculated in vivo subcellular tissuedistribution for oligodeoxynucleotide phosphorothioateswith some Auger emitting radionuclides. Augeremittersare low-range electrons with high biologicalefficiency with a tendency of becoming more and morefrequently used, at least theoretically. The doses varyconsiderably depending on cellular dimensions whenusing Auger-emitting isotopes; however, in small cellsthey may give a high dose. In tumors cell dimensions mayvary and therefore these Auger-emitting isotopes shouldbe applied only when nuclear target circumstances arewell characterized. High energy β-emitter P-32 gives thenuclear dose closest to uniform distribution in cell sizes,but this is due to high energy. We have previously shown(Kairemo) that when using P-32 labelled oligos other thantarget cells will be destroyed because of long range. Thisis not the case when using β-emitters, P-33 and S-35,which are optimal when targets are smaller than 300 µm indiameter. P-33 was not studied here separately because itscharacteristics are very close to those of S-35. Now wealso demonstrate that calculations related to positronemitters F-18 and Br-76, beta-emitter Y-90, Auger-emitterIn-111 and alpha-emitter At-211 add substantialinformation to radionanotargeting dosimetry. Calculationsusing Br-76 demonstrate up to 5-fold differences in cellnuclear dose only in different cellular dimensions. Thisindicates the importanc of careful selection of a properradionuclide.It is possible to use a mixture of radioisotopes toensure a complete coverage of targets in more than onelocations, e.g. targeting nuclear related and cellular RNAat the same time. In addition, modern imaging techniqueallows visual control over kinetic events. Dual labellingmay provide therapeutic benefits when treating smallerand larger targets simultaneously. Further in vivodevelopment, especially with various labels for oligos ishighly indicated.ReferencesAgrawal S, Temsamani J, Galbraith W, Tang J. (1995)Pharmacokinetics of antisense oligonucleotides. ClinPharmacokinet 28, 7-16.Agrawal S, Temsamani J, Tang JY. (1991) Pharmacokinetics,biodistribution and stability of oligodeoxynucleotidephosphorothioates in mice. Proc Natl Acad Sci USA 88,7595-7599.Bolch WE, Bouchet LG, Robertson JS, Wessels BW, Siegel JA,Howell RW, Erdi AK, Aydogan B, Costes S, Watson EE,Brill AB, Charkes ND, Fisher DR, Hays MT, Thomas SR.(1999) MIRD pamphlet No. 17, the dosimetry of nonuniformactivity distributions--radionuclide S values at the voxel21

Kairemo et al: Oligonucleotide radiotherapylevel. Medical Internal Radiation Dose Committee. J NuclMed 40, 11S-36S.Chen Z, Fisher RJ, Riggs CW, Rhim JS, Lautenberger JA. (1997)Inhibition of vascular endothelial growth factor-inducedendothelial cell migration by ETS1 antisenseoligonucleotides. Cancer Res 57, 2013-9Claffey KP, Brown LF, del Aguila LF, Tognazzi K, Yeo KT,Manseau EJ, Dvorak HF. (1996) Expression of vascularpermeability factor/vascular endothelial growth factor bymelanoma cells increases tumor growth, angiogenesis, andexperimental metastasis. Cancer Res 56, 172-81Crooke RM, Graham MJ, Cooke ME, Crooke ST. (1995) In vitropharmacokinetics of phosphorothioate antisenseoligonucleotides. J Pharmacol Exp Ther 275, 462-473.Crooke ST, Graham MJ, Zuckerman JE, Brooks D, Conklin BS,Cummins LL, Greig MJ, Guinosso CJ, Kornburst D,Manorahan M, Sasmor HM, Schleich T, Tivel KL, GriffeyRH. (1996) Pharmacokinetic properties of several noveloligonucleotide analogs in mice. J Pharmacol Exp Ther277, 923-937Dewanjee M.K, Ghafouripour A.K, Kapadvanjwala M,Dewanjee S, Serafini AN, Lopez DM, and Sfakianakis GN.(1994a) Noninvasive imaging of c-myc oncogene messengerRNA with indium-111-antisense probes in a mammarytumor-bearing mouse model. J. Nucl. Med. 35, 1054-1063.Egholm M, Buchardt O, Christensen L, Behrens C, Freier SM,Driver DA, Berg RH, Kim SK, Norden B, Nielsen PE.(1993) PNA hybridizes to complementary oligonucleotidesobeying the Watson-Crick hydrogen-bonding rules. Nature365, 566-568.Egholm M, Burchardt O, Nielsen PE, Berg RH. (1992) Peptidenucleic acids (PNA), oligonucleotide analogs with an achiralpeptide backbone. J Am Chem Soc 114, 1895-1897.Ensoli B, Markham P, Kao V, Barillari G, Fiorelli V, GendelmanR, Raffeld M, Zon G, Gallo RC. (1994) Block of AIDS-Kaposi's sarcoma (KS) cell growth, angiogenesis, and lesionformation in nude mice by antisense oligonucleotidetargeting basic fibroblast growth factor. A novel strategy forthe therapy of KS. J Clin Invest 94, 1736-46Geselowitz DA, Neckers LM. (1992) Analysis of oligonucleotidebinding, internalization and intracellular trafficking utilizinga novel radiolabeled crosslinker. Antisense Res Dev 2, 17-25.Hammes HP, Brownlee M, Jonczyk A, Sutter A, Preissner KT.(1996) Subcutaneous injection of a cyclic peptide antagonistof vitronectin receptor-type integrins inhibits retinalneovascularization. Nat Med. 2, 529-33.Hanahan D. (1997) Signaling vascular morphogenesis andmaintenance. Science 277, 48-50.Iversen PL, Mata J, Tracewell WG, and Zon G. (1994)Pharmacokinetics of an antisense phosphorothioateoligodeoxynucleotide against rev from humanimmunodeficiency virus type 1 in the adult male ratfollowing single injections and continuos infusion. AntisenseRes. Dev 4, 43-52.Iversen PL, Shu S, Meter A, and Zon G. (1992) Cellular uptakeand subcellular distribution of phosphorothioateoligonucleotides into cultured cells. Antisense Res. Dev 2,211-222.Iwasaka C, Tanaka K, Abe M, Sato Y. (1996) Ets-1 regulatesangiogenesis by inducing the expression of urokinase-typeplasminogen activator and matrix metalloproteinase-1 andmigration of vascular endothelial cells. J Cell Physiol 169,522-531Jekunen AP,Kairemo KJA. (1997) Inhibition of malignantangiogenesis. Cancer Treat Rev 23, 263-86.Jellinek D, Green LS, Bell C, Janjic N. (1994) Inhibition ofreceptor binding by high-affinity RNA ligands to vascularendothelial growth factor. Biochemistry 33, 10450-6Kaipainen A, Vlaykova T, Hatva E, Böhling T, Jekunen A,Pyrhönen S, Alitalo K. (1994) Enhanced expression of the tiereceptor tyrosine kinase messenger RNA in the vascularendothelium of metastatic melanomas. Cancer Res 54,6571-6577Kairemo KJA, Jekunen A, Karnani P. (1996) Modulation ofantibody kinetics by the cell membrane active agent Tween80 in vivo.Anticancer Res 16, 3542-3550Kairemo KJA, Jekunen AP, Tenhunen M. (1999) Essentials ofradionanotargeting using oligodeoxynucleotides. Gene TherMol Biol4, 171-176Kairemo KJA, Tenhunen M, Jekunen AP. (1996)Oligoradionuclidetherapy using radiolabelled antisenseoligodeoxynucleotide phosphorothioates. Anti-Cancer DrugDesign 11, 439-449Kairemo KJA,Tenhunen M, Jekunen AP. (1996) Dosimetry ofradionuclide therapy using radiophosphonated antisenseoligodeoxynucleotide phosphorothioates based on animalpharmacokinetic and tissue distribution data. Antisense NuclAcid Drug Dev 6, 215-220.Kairemo KJA, Thorstensen K,Mack M, Tenhunen M, JekunenAP. (1999) Ets-1 mRNA as target for antisense radiooligonucleotidetherapy in melanoma cells. Gene Ther MolBiol 4, 177-182Korhonen J, Partanen J, Armstrong E, Vaahtokari A, Elenius K,Jalkanen M, Alitalo K. (1992) Enhanced expression of the tiereceptor tyrosine kinase in cells during neovascularization.Blood 20, 2548-2555Maisonpierre PC, Suri C, Jones PF, Bartunkova S, Wiegand SJ,Radziejewski C, Compton D, McClain J, Aldrich TH,Papadopoulos N, Daly TJ, Davis S, Sato TN, YancopoulosGD. (1997) Angiopoietin-2, a natural antagonist for Tie2 thatdisrupts in vivo angiogenesis. Science 277, 55-60.Mardirossian G, Lei K, Rusckowski M, Chang F, Qu T, EgholmM, Hnatowich DJ. (1997) In vivo hybridization oftechnetium-99m-labeled peptide nucleic acid (PNA). J NuclMed 38, 907-913Masood R, Cai J, Zheng T, Smith DL, Naidu Y, Gill PS. (1997)Vascular endothelial growth factor/vascular permeabilityfactor is an autocrine growth factor for AIDS-Kaposisarcoma. Proc Natl Acad Sci U S A. 94, 979-84Risau W. (1997) Mechanisms of angiogenesis Nature 386, 671-4.Saleh M, Stacker SA, Wilks AF. (1996) Inhibition of growth ofC6 glioma cells in vivo by expression of antisense vascularendothelial growth factor sequence. Cancer Res 56, 393-401Sands H, Gorey-Feret LJ, Cocuzza AJ, Hobbs FW, Chidester D,Trainor GL. (1994) Biodistribution and metabolism ofinternally 3 H-labeled oligonucleotides. I. Comparison of aphosphodiester and a phosphorothioate. Mol Pharmacol 45,932-943.Shoji Y, Akhtar S, Periasamy A. (1991) Mechanism of cellularuptake of modified oligonucleotides methylphosphonatelinkage. Nucl Acid Res 19, 5543-5550.Smyth AP, Rook SL, Detmar M, Robinson GS. (1997) Antisenseoligonucleotides inhibit vascular endothelial growthfactor/vascular permeability factor expression in normal22

Gene Therapy and Molecular Biology Vol 7, page 23human epidermal keratinocytes. J Invest Dermatol 108,523-6Stabin MG. (1996) MIRDOSE, personal computer software forinternal dose assessment in nuclear medicine. J Nucl Med37,538-46.Suri C, Jones PF, Patan S, Bartunkova S, Maisonpierre PC,Davis S, Sato TN, Yancopoulos GD. (1996) Requisite role ofangiopoietin-1, a ligand for the TIE2 receptor, duringembryonic angiogenesis. Cell 87, 1171-80.T Wu J, Zhou L, Tonissen K, Tee R, Artzt K. (1999) Thequaking I-5 protein (QKI-5) has a novel nuclear localizationsignal and shuttles between the nucleus and the cytoplasm. JBiol Chem 274,29202-10.Tavitian B, Terrazzino S, Kuhnast B, Marzabal S, Stettler O,Dolle F, Deverre JR, Jobert A, Hinnen F, Bendriem B,Crouzel C, Di Giamberardino L. (1998) In vivo imaging ofoligonucleotides with positron emission tomography. NatMed 4,467-71.Ullrich A, Schlessinger J. (1990) Signal transduction byreceptors with tyrosine kinase activity. Cell 61, 203-212Wasylyk B, Hahn SL, Giovane A. (1993) The ets family oftranscription factors. Eur J Biochem 211, 7-18.Wu F, Yngve U, Hedberg E, Honda M, Lu L, ErikssonB,Watanabe Y, Bergström M, L_ngström B. (2000)Distribution of 76Br-labelled antisense oligonucleotides ofdifferent lengthdetermined ex vivo in rats. Eur J Pharm Sci10, 179-186Zhang R, Diasio RB, Lu Z, Liu T, Jiang Z, Galbraith WM, andAgrawal S. (1995) Pharmacokinetics and tissue distributionin rats of an oligodeoxynucleotide phosphorothioate (GEM91) developed as a therapeutic agent for humanimmunodeficiency virus type-1. Biochem Pharmacol 49,929-939.Zhang R, Yan J, Shahinian H, Amin G, Lu Z, Liu T, Saag MS,Jiang Z, Temsamani J, Martin RR, et al (1995)Pharmacokinetics of an anti-human immunodeficiency virusantisense oligodeoxynucleotide phosphorothioate (GEM 91)in HIV-infected subjects. Clin Pharmacol Ther 58, 44-53.23

Kairemo et al: Oligonucleotide radiotherapy24

Gene Therapy and Molecular Biology Vol 7, page 25Gene Ther Mol Biol Vol 7, 25-35, 2003Strategy of sensitizing tumor cells with adenovirusp53transfectionReview ArticleJekunen Antti 1* , Miettinen Susanna 2 , Mäenpää Johanna 3 , Kairemo Kalevi 41 Department of Clinical Pharmacology, Helsinki University, and Department of Oncology, Turku University, and AventisPharma Finland, Finland. 2 Department of Anatomy, Tampere University, Finland. 3 Department of Obstetrics andGynecology, Division of Gynecologic Oncology, Tampere University Hospital and Tampere University, Finland.4 Department of Nuclear Medicine, Uppsala University Hospital, Sweden__________________________________________________________________________________*Correspondence: Antti Jekunen, MD, PhD, PL96, 00241 Helsinki, Finland; Tel. +358400 755208; Fax. +3589 47638140; e-mail:antti.jekunen@aventis.comReceived: 29 January 2002; accepted: 06 March 2002; electronically published: July 2003SummaryLoss or malfunction of the p53-mediated apoptotic pathway has been proposed as one mechanism by which tumorsbecome resistant to chemotherapy. While it may be the most frequently mutated gene in human tumor samples, thefunction of p53 is critical for maintaining the integrity of the cellular genome in its responses to treatment withcytotoxic agents. Intact p53 protein in nuclei of normal cells acts as a transcriptional activator for a group of genesinvolved in cell cycle arrest, DNA repair and apoptosis. The transfection of adenovirus p53 (adeno-p53) alone hasbeen shown in ovarian cancer cell culture models to inhibit cell growth and to promote apoptosis regardless of theendogenous p53 status of the cells. Both mutant p53 in the tumor cells and the loss of p53 function were associatedwith resistance to chemotherapeutic agents. There are various reports of at least additive interactions betweenadeno-p53 and several chemotherapeutic agents in a number of cancers, e.g. bladder cancer, NSCLC, prostatecancer, breast cancer, and ovarian cancer both in vitro and in vivo. The mechanisms of these interactions areunknown, but they may depend on the chemotherapeutic agents used, the targets and critical tissues, and theintracellular signal transduction pathways affected.Results obtained with a speculative treatment regimenconsisting of oligonucleotide therapy and p53 transfection suggest that p53 expression in tumor cells may improvetheir sensitivity to routine chemotherapy, e.g. docetaxel and irinotecan, which are efficacious drugs possessingdifferent modes of action: prevention of depolymerization of tubulin and specific DNA topoisomerase I inhibition,respectively. It is known, however, that even these new agents cannot achieve responses in all tumors, and that insome tumors the efficacy, once established, diminishes along with the treatment. In these cases of resistant tumorsor recurrences and relapses, combined treatment with adeno-p53 and chemotherapeutic agents may be anattractive strategy for inhibiting the progression of local cancers. In fact, the ground is ready for a rapid practicaldevelopment of adeno-p53, which itself causes only minimal side-effects after administration, e.g. injection siterashes and fever, and an immunostimulation that seems to be quite mild and transient in nature. Future cancertherapy strategies may consist of effective chemotherapy coupled to molecular medicine specifically targeting tumorcells. So far, we do not have proper means in molecular medicine for achieving high enough tumor access with anyof the current systemic virus vectors having the proper level of selectivity between tumor and normal cells. We havealready some clinical experience, however, with intratumoral approaches that ensure the highest possibleconcentrations inside NSCLC, ovarian cancer and head and neck cancer tumors. It seems that there is clearevidence of good tolerability at non-maximal doses, but unfortunately, only modest activity when the construct isused alone. We review here the published data on the use of adenovirus p53 for sensitizing tumors tochemotherapeutic agents and outline perspectives for the future.I. IntroductionA. Function of p53The p53 protein, a nuclear phosphoprotein, isindispensable for genomic integrity and cell cycle control.Its basic function is to control the entry of the cell into theS phase of the cell cycle. p53 extends the time availablefor DNA repair before S phase entry (Fan et al, 1995). Thewild- type gene product regulates cell growth and divisionnegatively. Although not essential for progression of the25

Jekunen et al: Strategy of sensitizing tumor cells with adenovirus-p53 transfectioncell cycle, it is critical as a checkpoint that blocksuncontrolled cell division (Levine, 1992). In the nuclei ofnormal cells, the intact p53 protein acts as a transcriptionalactivator for a group of genes involved in cell cycle arrest(p21 cip1/waf1 ), DNA repair (GADD45), and apoptosis (Bax)(O'Connor et al, 1997; Sugrue et al, 1997; Yin et al, 1997;Carrier et al, 1999). In addition to this, p53 is a potentinducer of programmed cell death (apoptosis) within a cellin which the DNA has been damaged. Normally, the p53gene is inactive. When, after DNA damage, the normalp53 is activated, the levels of p21, p27, and GADD 45may become very high (Sherr, 1994). DNA damage incells induces expression of p53 and interruption of the cellcycle in both G1 and G2 (Chu and DeVita, 2001). If DNArepair is successful, the cell continues its cycle. If repairdoes not succeed, the cell undergoes apoptosis.B. Mutation of p53Mutations in the p53 gene are among the mostcommon genetic alterations observed in human tumorsamples (Oren, 1992). The specific cytotoxic treatment,the conditions of treatment, the p53 status, and otherelements of cell-cycle regulation may all contribute to theoutcome of exposure of a cell to DNA-damaging agents(Chu and DeVita, 2001). p53 can activate an apoptoticresponse to DNA damage, especially in hematopoietic andlymphoid cells, which often overrides the G1 checkpointresponse (Fan et al, 1995). In cell types programmed forapoptosis, loss of p53 function decreases their sensitivityto a wide variety of DNA-damaging agents, while in cellapoptosis, it has been more difficult to establish a clearrelationship between p53 gene status and chemosensitivitytypes of some solid tumors not inherently programmed for(Fan et al, 1995). If the DNA is damaged, the cell withintact p53 function will undergo p53-dependent apoptosis(Chu and DeVita, 2001). In tumor cells with mutated p53,the loss of p53 function, is thought to result in resistanceto chemotherapeutic agents (Lowe et al, 1994; Righetti etal, 1996; Blandino et al, 1999). A recent study of ovariancancer shows that women with tumors having the p53 nullmutation have a survival disadvantage over those with p53missense mutations (Shahin et al, 2000).II. Evidence of the role of p53 inchemosensitizingA. p53 and chemotherapeutic agentsDysregulation of the p53 pathway may lead to drugresistance due to overproduction of the gene productsresponsible for entry into the S phase and rapid cellgrowth (Figure 1).Activation of these genes could theoretically increasethe resistance of cells to the following chemotherapeuticagents: methotrexate, 2-chlorodeoxyadenosine,hydroxyurea, fludarabine, cytosine arabinoside, and 5-fluorouracil. Under some experimental circumstances, celldeath in response to exposure to DNA-damaging agentsmay require an intact p53-dependent apoptoticmechanism. Some of the genes that are transcriptionallyactivated by p53 belong to a class of proteins known toinhibit cyclin-dependent kinases (cdk). p21 forms acomplex with proliferating cell nuclear antigen or inhibitscdk’s, e.g. cdk4 (Polyak et al, 1997). Activated p53 cancause a G1 cell cycle arrest by increasing the transcriptionof the cdk inhibitor p21 (Figure 2), which block cdk4activity, preventing reitinoblastoma gene product (RB)phosphorylation (Sherr, 1994) and release of E2F blockingthe transcription of a number of genes, and inhibiting entryinto S phase (Kirsch, 1998). The E2F family oftranscription factors bind to the regulatory regions of anumber of genes that participate in the synthesis of DNA(Figure 2).Figure 1. Effect of chemotherapy via p53 pathway. After chemotherapy has induced DNA damage, p53 protein is activated andtranscription of many genes is increased, resulting in cell cycle arrest and apoptosis. For apoptotically sensitive cells, genotoxic damagecan signal an immediate apoptotic response, while for apoptotically insensitive cells, the primary apoptotic decision point is disabled.Cells that avoid apoptotic or necrotic death after DNA repair can survive and grow. (Kirsch 1998; Brown and Wouters 1999)26

Gene Therapy and Molecular Biology Vol 7, page 27These genes include ribonucleotide reductase,dihydrofolate reductase, DNA-dependent RNApolymerase, thymidylate synthase, c-myc, c-fos, and c-myb. Activation of these gene products facilitates the entryof the cell into the S phase.There is much evidence in support of the idea that amutation in p53 may lead to resistance to cytotoxic agents.In premenopausal women with node negative breastcancer, it has been shown by immunohistochemistry thatp53(+) tumors are less sensitive to treatment with aregimen including 5-fluorouracil, doxorubicin, andcyclophosphamide than p53 (-) tumors. (Clahsen et al,1998). Under in vitro conditions Koechli et al, have shownthat mutant p53 can increase chemoresistance to 5-fluorouracil, cyclophosphamide, and methotrexate(Koechli, 1994). Cisplatin resistance seems to beconnected with p53 mutations, and in advanced ovariancancer, the p53 mutational status is a predictor of theresponsiveness to platinum-based chemotherapy (Calvert,1999). However, there are also reports that apparentlydisagree with the chemoresistance effect of p53 (Fan et al,1995; Stal 1995; Hawkins et al, 1996).Human fibroblasts lacking functional p53 were moresensitive to cisplatin, carboplatin, paclitaxel, nitrogenmustard or melphalan than cells with functional p53(Hawkins et al, 1996). Similar results, loss of p53 functionand the sensitizing effect of cisplatin, have beendemonstrated in MCF-7 breast cancer cells and RKOmethotrexate, and 5-fluorouracil have been reported incolon cancer cell lines with or without disruption of p53function by a dominant negative p53 transgene (Fan et al,1995).Increased rates of response to cyclophosphamide,patients with breast cancer who were determined to beimmunohistochemically p53(+) (Stal,1995).B. In vitro interactionsSynergy between two chemical agents in vitro is anempirical phenomenon, in which the observed effect of thecombination is greater than would be predicted from theeffect of each agent working alone. While synergy is notdirectly measurable in clinical practice, it may predict afavorable outcome when two treatments are combined invivo and may strongly suggest the presence of synergy invivo. Nielsen et al, used three-dimensional statisticalmodeling to evaluate the presence of synergistic, additive,or antagonistic efficacy between adenovirus-mediated p53gene transfer and paclitaxel in a panel of human tumor celllines, including those for ovarian, head and neck, prostate,and breast cancer (Nielsen et al, 1998). Cells were eitherpretreated with paclitaxel 24 h or not, before proliferationwas measured 3 days later. Paclitaxel had synergistic oradditive efficacy with p53 transfer, independently ofwhether the cells expressed mutant p53 protein or no p53protein at all. Cell cycle analysis demonstrated that, priorto apoptotic cell death, p52 transfection arrested cells inthe G0/G1 stage, whereas paclitaxel arrested cells in theG2-M stage. When combined, the relative concentrationsof the two agents determined the dominant cellularresponse. The observed synergy remained unexplained;however, some speculations were offered. P53 has beenshown to down regulate the expression of the antiapoptoticbcl-2 gene and up regulate the expression of the proapoptoticbax gene in other tumor cells (Selter andMontenarh 1994). Thus, p53 and paclitaxel may potentiateeach other in stimulating the apoptotic pathway inneoplastic cells (Nielsen et al, 1998). It may also be thatpaclitaxel increased the number of cells transfected by theadenovirus. Particularly, the concentrations of paclitaxelresponsible for increased adenovirus transduction arelower than the concentrations required for microtubulecondensation. Moreover, the rate of change in the numberof cells transduced by adenovirus appears to beindependent of paclitaxel-induced cell death. The authorsalso determined the efficacy of the combination therapy invivo. In some instances, it seems that loss of p53 mayincrease resistance to one agent, while simultaneouslyincreasing sensitivity to another. Bunz et al, (1999) havereported that deletion of p53 in colorectal cancer cell linesmaintained the cells that were resistant to 5-fluorouracil,but increased the sensitivity to doxorubicin and radiationin vitro. If the compound exerts it effects by apoptosis, asdoes 5-fluorouracil, loss of the apoptotic pathway maylead to resistance.Figure 2. Two examples of cell cycle arrest via p53 activation. P53 mediated cell-cycle arrest is demonstrated with two examples: A)inhibition of cdk4 and cdk2 resulting G1-S and G2-M arrest, respectively. B) p53 activation increases the transcription of the cyclindependentkinase (cdk) inhibitor p21. Increase levels of p21 protein prevent cdk’s from phosphorylating their substrates, such as theretinoblastoma protein (RB) and thus block cell-cycle progression from G1 into S phase. (Kirsch 1998; Brown and Wouters 1999)27

Jekunen et al: Strategy of sensitizing tumor cells with adenovirus-p53 transfectionRecently, a report using isobologram modelling haveshowed that the combination of adeno-p53 + radiationproduced significantly synergistic effects in NSCL celllines, whereas the combination of docetaxel + adeno-p53and docetaxel + radiation produced mixed effects rangingbetween additive and synergistic (Nguyen al., 1996). Thethree-agent combination also produced significantlysynergistic effects.Brown and Wouters have criticized the sensitizingresults obtained in cell cultures. They have pointed out theneed for further evidence in relating p53 to the sensitivityof anticancer agents (Brown and Wouters, 1999). Becauseapoptosis, particularly p53–dependent apoptosis, can occurrapidly after drug exposure, short-term growth rate assaystend to underestimate overall death of cells with mutantp53 or of cells not undergoing apoptosis. This may resultin a situation where short-term assays may incorrectlyassess overall cell death in tumor cells with differentprobabilities of undergoing early apoptosis. Thus, resultsmay have a bias toward increased cell death in wild-typep53 cells and decreased cell kill in mutant p53 cells.Results of experiments with normal cells transformed withdominant oncogenes have often been extrapolated totumor cells, instead of initially using cancer cell models.Transformed normal cells are usually apoptotically moresensitive than cancer cells. Therefore, in sensitizingexperiments, both long term clonogenic assays and tumorcell models with solid tumors should be used rather thangrowth rate assays and transformed normal cells.However, the more widely accepted conclusion drawnfrom studies conducted in cancer cell lines and tumors ofdifferent origin is still that restoration of normal p53function in tumors restores the apoptotic pathway andleads to an increased response to chemotherapy (Peller,1998; Ferreira, 1999; Chang, 2000).C. Transfection of cell cultures with theadenovirus p53 gene constructAdenovirus vectors have many advantages over otherviral and non-viral vectors. Their transfection efficacy ishigh, in both dividing and resting cells, and they showhigh expression levels (Hwu, 2001). As adenoviral DNAis not incorporated into the cell genome, expression of thetransgene is transient, but adenoviral vectors can beproduced at high titers. Introduction of wild-type p53 intotumors with non functional p53 offers a novel strategy fortreating cancer, by inducing apoptotic death in neoplasticcells.Genomic instability accompanied by loss of p53-mediated apoptosis can also lead to therapy resistance. Thesupport for this rationale is that loss of p53 coulddesensitize cells to the damaging effects of drugs. Normaltransgenic hematopoetic cells (Lotem and Sachs, 1993),E1A-expressing transgenic fibroblasts (Lowe et al, 1993),and transformed transgenic fibroblasts (Lowe et al, 1994)were all more resistant to apoptosis following treatmentwith any of a wide variety of anticancer agents, than werecomparable cells from the parental strain of mice, whichexpressed wild-type p53. Apoptosis seemed to beenhanced in cells that expressed wild-type p53 and wereable to trigger their own cell death program.In cell culture models, adenovirus-mediated p53 genetransfer alone inhibits cell growth and promotes apoptosis,regardless of the endogenous p53 status of the ovariancancer cells (Santoso et al, 1995). In tumor cells, mutatedp53 and also loss of p53 function were associated withresistance to chemotherapeutic agents. There are severalreports of at least an additive interaction between adenop53and cisplatin in bladder cancer (Miyake et al, 2000),between adeno-p53 and cisplatin, SN-38 (a metabolite ofirinotecan), 5-fluorouracil, taxanes, bleomycin, andcyclophosphamide in NSCLC (Fujiwara et al, 1994)(Horio et al, 2000), and between adeno-p53 and paclitaxelin ovarian cancer (Nielsen et al, 1998). In the ovariancancer model, enhanced efficacy has been reported in athree-drug combination of adeno-p53, cisplatin, andpaclitaxel (Gurnani et al, 1999).There is some evidence that chemosensitivity can beincreased by replacement of the p53 gene. Roth (Roth,1996) reported that recombinant-adenovirus-mediatedtransfer of the wild-type p53 gene into several human cellswith homozygous deletions of p53 markedly increasedcellular chemosensitivity to the major chemotherapeuticdrugs. An additive antiproliferative effect was reported inp53null H358 lung cancer cells when cultured withcisplatin for 24 h before transduction with adeno-p53(Fujiwara et al, 1994). Enhanced apoptosis, detected byDNA fragmentation, was reported for the combinationcompared with each agent alone.A viability assay demonstrated that a replicationdefectiveadenovirus encoding the wild-type p53 gene(INGN 201, Introgen Therapeutics, Inc.) suppressesgrowth and enhances sensitivity to DNA-damagingchemotherapeutic drugs (5-fluorouracil, doxorubicin,cisplatin) in p53-mutant-expressing cell lines (Gjerset andMercola, 2000). These cells lines represent DLD-1 coloncancer, T47D breast cancer, PC-3 prostate cancer, andT98G glioblastoma. Transfection efficiencies were 60-70%. It seems that restoration of the wild-type p53 tomutant p53-expressing or p53null cells results in markedenhancement of sensitivity to several DNA damagingagents. This enhancement of sensitivity was not observedin two wild-type p53-expressing cell lines, MCF7 andLS174T, suggesting that, in this model, wild-type p53gene transfer is effective as therapy sensitization only intumors that have lost wild-type p53 function.1. Glioma and pancreatic cancerSomatic gene therapy based on the reintroduction ofp53 limits the proliferation of human malignant gliomacells, but is unlikely to induce clinically relevantsensitization to chemotherapy in these tumors. Wild-typep53 failed to sensitize glioma cells to cytotoxic drugsincluding BCNU, cytarabine, doxorubicin, teniposide, andvincristine. The combined effects of the wild-type p53gene transfer and drug treatment were less than additiverather than synergistic, suggesting that the intracellularcascades activated by p53 and chemotherapy wereredundant. Unexpectedly, forced expression of mutant-28

Gene Therapy and Molecular Biology Vol 7, page 29p53-modulated drug sensitivity enhanced the toxicity ofsome drugs but attenuated the effects of others (Trepel etal, 1998). Likewise, in p53-null pancreatic carcinomacells, wild-type p53 gene transduction had no effect on invitro chemosensitivity to cisplatin, etoposide, 5-fluorouracil and paclitaxel (Kimura et al, 1997).Moreover, in anaplastic thyroid cancer cells, adeno-p53increased the sensitivity to doxorubicin with a 10-folddecrease in IC 50 values.2. Hepatocellular cancerOne of the goals of gene therapy for treating canceris selective expression of cytotoxic gene products in tumorcells. When replication-defective retroviruses wereconstructed containing p53 cDNA that wastranscriptionally regulated by the human hepatocellularcarcinoma-associatedalpha-fetoprotein genetranscriptional control elements, the expression ofexogenous wild-type p53 from this retroviral vector waslimited to the cells producing alpha-fetoprotein.Introduction of wild-type p53 into alpha-fetoproteinpositive human hepatocellular carcinoma cells byretroviral infection markedly inhibited their clonal growthin a monolayer and increased the sensitivity of these cellsto the chemotherapeutic drug cisplatin (Xu et al, 1996).3. Ovarian cancerIn cell culture models adenovirus-mediated p53 genetherapy is one way to inhibit cell growth and promotesapoptosis, regardless of the endogenous p53 status of theovarian cancer cells (Santoso et al, 1995) (Wolf et al,1999). Adeno-p53 gene transfer, combined with cisplatin,doxorubicin, 5-fluorouracil, methotrexate, or etoposide,inhibited cell proliferation more effectively thanchemotherapy alone in head and neck, ovarian, prostateand breast tumor cell lines. Of particular significance, inan ovarian cancer model enhanced efficacy was notedwhen using the three-drug combination of adeno-p53,cisplatin, and paclitaxel (Gurnani et al, 1999). In humanhead and neck, ovarian, prostate, and breast cancer cells,low concentrations of paclitaxel also increase the numberof cells transduced by recombinant adeno-p53 in a dosedependentmanner (Nielsen et al, 1998). The concentrationof paclitaxel responsible for increased adenovirustransduction is lower than that required for microtubulecondensation.4. Breast cancerTransduction of cells using replication-deficientadenovirus vectors can induce endogenous p53 expressionin cells containing the wild-type p53 gene and thisresponse is different from the p53 induction observed afterDNA damage (McPake et al, 1999). Lebedeva et al, haveexamined the effects of a replication-defective adenovirusencoding p53 (INGN 201, Ad5CMV-p53), alone or incombination with the breast cancer therapeuticdoxorubicin, in suppressing growth and inducing apoptosisin breast cancer cells in vitro (Lebedeva et al, 2001). Theyfound that whereas in vitro treatment of cells with adenop53reduced 3 H-thymidine incorporation by about 90% at48 hr, cell viability at 6 days was reduced by only about50% relative to controls. Although apoptosis is detectablein the adeno-p53-treated cultures, these results suggest thata large fraction of adeno-p53-treated cells merely undergoreversible cell cycle arrest. Combined treatment withadeno-p53 and doxorubicin results in a greater thanadditive loss of viability in vitro and increased apoptosis.These data indicate an additive to synergistic effect ofadeno-p53 and doxorubicin for the treatment of primaryand metastatic breast cancer.However, in breast cancer cell lines results withoutany clear cut link between transfection of p53 and asensitizing effect have been reported. Two human breastcancer cell lines, MDA-MB-231 and MDA-MB-435, bothwith p53 mutations, were transduced with adenoviralvectors containing wild-type p53 and the effects on growthwere determined by clonogenic assays (Parker et al, 2000).Combining VP-16 and paclitaxel with Ad5CMV-p53 didnot consistently or significantly decrease clonogenicsurvival.5. Bladder cancerCombined treatment with Ad5CMV-p53 andcisplatin could be an attractive strategy for inhibitingprogression of bladder cancer. In human bladder cancerKoTCC-1 cells, transfer of an adenovirus-mediated p53gene enhances cisplatin cytotoxicity in vitro, andAd5CMV-p53 and cisplatin synergistically inhibit growthand metastasis in vivo. Ad5CMV-p53 substantiallyenhances cisplatin chemosensitivity in a dose-dependentmanner, reducing the median IC 50 by more than 50%.Furthermore, orthotopic injection of adeno-p53 combinedwith cisplatin therapy synergistically inhibits growth ofsubcutaneous KoTCC-1 tumors and the incidence ofmetastasis (Miyake et al, 2000). In contrast, p21 cip1/waf1gene therapy had no effect on in vitro or in vivochemosensitivity to cisplatin (Miyake et al, 1998).6. Lung cancerRecombinant adenovirus-mediated transfer of thewild-type p53 gene into monolayer cultures ormulticellular tumor spheroids of the human NSCLC cellline H358, in which there is homozygous deletion of p53,markedly increased the cellular sensitivity of these cells tocisplatin (Fujiwara et al, 1994). In a study made by Osakiet al,(Osaki et al, 2000), an alteration in drugchemosensitivity caused by the adenovirus-mediatedtransfer of the wild-type p53 gene in human lung cancercells was tested on a human pulmonary squamous cellcarcinoma cell line, NCI-H157, and a human pulmonarylarge-cell carcinoma cell line, NCI-H1299. Based onisobologram data, a supra-additive effect was observed for5-fluorouracil and SN-38 on NCI-H157 cells. An additiveeffect was also observed for cisplatin, paclitaxel,bleomycin, and cyclophosphamide on NCI-H157 cells.Cisplatin, paclitaxel, 5-fluorouracil, and SN-38 had anadditive effect on NCI-H1299 cells. No drug showed anysubadditive or protective effects. These findings suggest29

Jekunen et al: Strategy of sensitizing tumor cells with adenovirus-p53 transfectionthat CPT-11 and 5-fluorouracil may be useful asanticancer agents for use in a combination therapyregimen, using wild-type p53 gene transfer. These resultsindicate that CPT-11, as well as cisplatin, is a candidatefor the combination of chemotherapy and gene therapy forNSCLC. Adeno-p53 and DNA-damaging agents, cisplatin,etoposide and CPT-11 showed synergistic effects inNSCLC, but, in contrast had additive effects withantitubulin agents such as paclitaxel and docetaxel (Horio,Hasegawa et al, 2000). Perdomo et al, (Perdomo et al,1998) have demonstrated that human NSCLC cells havinga mutant form of p53 grow faster in vivo than wild-typep53 cell lines and the treatment with cisplatin or radiationdoes not reduce the size of mutant p53 tumors, althoughwild-type p53 tumors regress markedly. Apoptosisoccurred in mutant p53 cell types only at high cisplatindoses and not at the magnitude detected in wild-typetumors.III. In vivo evidence ofchemosensitization by adenovirus p53These observations have been extended to in vivomodels. Tumors have been treated in vivo withreplication-defective p53 adenovirus and chemotherapy.Nguyen et al, have reported convincing in vivo studies, inwhich p53null H1299 lung tumor xenografts were giveni.p. cisplatin before, concurrently with, or afterintratumoral adenovirus p53 (Nguyen et al, 1996). Themost effective dosing regimen was cisplatin given twodays before p53 therapy. Cisplatin and CPT-11 had asignificant antitumoral effect on lung cancer H157 cellxenografts of nude mice in vivo. Human head and neckcancer and colon cancer (Gjerset et al, 1997) and prostatecancer (Gjerset and Mercola 2000) in nude mice models invivo have been found to exhibit a similar sensitizationeffect with adenovirus plus cisplatin as in studies in vitro.Gjerset et al, demonstrated increased sensitivity tocisplatin cytotoxicity in p53mut T98G glioblastoma andp53 mut H23 small cell lung carcinoma cells transducedwith p53 expression vectors one or two days beforeexposure to cisplatin (Gjerset et al, 1995). These resultsare consistent with other in vivo studies in animal modelsshowing a combined benefit of p53 and chemotherapy(Badie et al, 1998), (Fujiwara et al, 1994), (Miyake et al,1998), (Nielsen et al, 1998), (Nguyen et al, 1996). Gjersetand Mercola are convinced that these results support theclinical application of adenovirus p53 combinationapproaches to tumors expressing mutant p53 (Gjerset andMercola 2000). Chemosensitization by p53 has also beenstudied using ex vivo modified cells in an orthotopicmodel of glioblastoma in Fisher rats (Dorigo et al, 1998).The combination of p53 with 5-fluorouracil andtopotecan has been studied in p53mut SW480 colorectaltumor cells transfected with an inducible p53 construct(Yang et al, 1996). Dose-dependent enhancement ofcytotoxicity was observed with these drugs by theconcurrent expression of wild-type p53. Increasedcytotoxicity has been reported in p53mut SkBr3 mammarytumor cells when transduction with p53 was followed 8 hrlater by doxorubicin or mitomycin-C, but not byvincristine (Blagosklonny and El-Deiry 1996).In the p53 null SK-OV-2 xenograft model of ovariancancer, a dosing schedule of the p53 therapy that, by itself,had a relatively minimal effect on the tumor burden (16%)caused a much greater decrease in tumor burden (55%)when combined with paclitaxel (Nielsen et al, 1998).Further, in nude mice implanted intraperitoneally with2774 human ovarian cancer cells (mutated p53), theresponse to adeno-p53 gene therapy showed significantsurvival duration, with a survival time greater than that ofuntreated animals. However, no statistically significantsurvival advantage was observed between adeno-p53- andadenovirus-βgal-treated mice (von Gruenigen et al, 1998).In another ovarian cancer study using nude mice, theadeno-p53 treatment effectively suppressed the growth ofperitoneal tumors and prolonged the survival of the treatedgroup, especially when the tumor burden was small (Kimet al, 1999). Greater combined efficacy was observed inthe p53null DU-145 prostate, p53Mut MDA-MB-468breast, and p53met MDA-MB-231 breast cancer xenograftmodels in vivo. The authors concluded that their data,taken together, offer the possibility of enhanced antitumoractivity with lower than normal doses of paclitaxel andadenovirus p53, when the two drugs are administered incombination (Nielsen et al, 1998). They noted that thiscould potentially decrease the chemotherapy-induced sideeffects, increasing the quality of life of the patients and,perhaps, reducing the overall expense of a complete courseof cancer treatment.IV. Clinical results of adenovirus p53transfection with chemotherapyThe first evidence of the efficacy of p53 gene therapyfor cancer was given by a pilot study in which retroviralp53 expression vectors were directly injected into smallendobronchial lesions of NSCLC patients (Roth et al,1996). Tumor regression was noted in three patients out ofnine, and tumor growth stabilized in three other patients.The safety and feasibility of the intratumoral injection ofadenoviral wild-type p53 expression vectors have beenestablished in NSCLC patients, with clear evidence fortransgenic expression, and possibly induction of apoptosis(Swisher et al, 1999; see Table 1). The antitumor activityin this trial was consistent with the activity of retroviralp53 injection in NSCLC patients. Twenty-four patientsreceived intratumor injections of adenovirus p53 and twopatients achieved a partial response, while 17 patientsachieved stable disease as the best clinical response.A nonrandomized, phase I, dose-escalating study byClayman et al expanded these findings into head and necksquamous cell carcinoma (Clayman et al, 1998). Patientswith incurable recurrent local or regionally metastaticHNSCC received multiple intratumoral injections ofadeno-p53, either with or without tumor resection. P53expression was detected in tumor biopsies despiteantibody responses after injections. prevent the appearanceof adeno-p53 in blood and urine. were seen in the study Asexpected, almost Neither dose-limiting effects nor serious30

Gene Therapy and Molecular Biology Vol 7, page 31adverse events all the patients developed anti-adenovirusantibodies in the course of treatment, but this immuneresponse did not treatment. The most common treatmentrelatedadverse event was pain at injection site. Otherreported adverse events were transient fever, headache,pain, and edema. No evidence of systemic hypersensitivityor allergic reactions was seen, despite the fact that patientsreceived many repeated courses of treatment. In somepatients, adenovirus p53 administration led to objectiveantitumor activity. Two out of 17 patients showedobjective tumor regressions greater than 50% and sixpatients showed stable disease for up to 3.5 months. Inaddition, one patient showed a complete pathologicresponse. The median survival for responding patients was13.6 months, and the overall median survival was 267days, which is about 60% longer than that reported inchemotherapy trials with a similar patient profile(Schornagel et al, 1995). Of course, it is impossible, for aphase-one study with limited numbers of patients to stateanything more than that these results are promising andthat further studies are needed, and are underway, todetermine the actual role of adenovirus-mediated p53intratumoral injections as a treatment option for HNSCC.The next step in the development of p53 treament is toinclude combination therapy with cytotoxic agents.There is also a negative trial published by Schullerand coworkers (2001). Twenty-five patients with nonresectableNSCLC were enrolled in an open-label,multicenter, phase II study of three cycles ofchemotherapeutics with intratumoral injection ofrecombinant adenovirus p53. The main idea of this smallstudy was to compare the isolated responses of a tumorlesion treated by transfer of the adenoviral wild-type p53gene with a comparable lesion not receiving any injectionsin patients undergoing first-line chemotherapy forNSCLC. In the 13 patients receiving carboplatin andpaclitaxel, there was no obvious difference between themean response of gene-therapy-treated and the referencelesions. In contrast, the mean regression of the referencelesions in patients treated with cisplatin and vinorelbinewas 15%, whereas it amounted to 55% in lesions that wereadditionally injected with the gene construct.There was no difference between the responses of lesionstreated with p53 gene therapy in addition to chemotherapy(52%) and those of lesions treated with chemotherapyalone (48%). The authors concluded that, in these patientsthe therapy appears to provide no additional benefit.However, there were several possible shortcomings in theclinical set-up: no injections to the reference lesions,highly restrictive inclusion criteria may result in selectionbias, a higher response rate (50%) than is normallyachieved in this disease, a chance of having a biologicallyinactive virus construct, and insufficient spreading of thereplication-defective adenoviral vectors within the tumorsafter only one central intralesional injection.Recently, attemps have been made to overcome theproblem of ineffective vector spreading by administrationof replication-competent adenoviruses (Heise, Sampson etal, 1997) and encouraging clinical results have beenreported (Khuri et al, 2000). There were concerns aboutthe safety, which, however, turned out to be exaggerated.Khuri et al, (2000) demonstrated an acceptable safetypattern with no sign of any dissemination to theenvironment. A Phase II trial of a combination ofintratumoral ONYX-015 injection with cisplatin and 5-fluorouracil was carried out with patients having recurrentsquamous cell cancer of the head and neck. Only pain atthe injection site (45%), mucous membrane disorder(21%), syncope (5%), kidney failure (5%), and anorexia(3%) could not be ruled out as attributable to Onyx-015.In addition, the injected tumors achieved objectiveresponses at a substantially higher rate (9 of the 11) thanthe non-injected tumors (3 of the 11) within the samepatients. In six patients, the injected tumor responded andthe non injected tumor did not respond. The time to tumorprogression was also longer for the injected tumors thanfor the non-injected tumors. There was no correlationbetween the response and the baseline tumor size, baselineneutralizing antibody titer, p53 gene status, or priortreatment. It was also clear that the efficacy of theintratumoral injection was not prevented by neutralizingantibodies. There has been discussion about whether or notenough evidence about viral replication of ONYX-015 inpatients, as along experience based on 190 patients treatedby a replication-defective adenovirus demonstratingsimilar biodistribution (Clayman et al, 1998; Constenla-Figueiras et al, 1999). It may simple be that Taqman realtimepolymerase chain reaction technology is not sufficientto prove that viral reproduction is taking place (Yver et al,2001).Table 1. Sensitising effect of adenovirus-p53 on chemotherapeutic agents, major clinical treatment resultsDisease Phase Combination n Treatment responses Reference (first author year)NSCLC II no 24 2 PR, 17 SD_ (Swisher et al, 1999)Head & neck II no 17 1CR, 2 PR, 6 SD_ (Clayman et al, 1998)NSCLC II Cisplatin + vinorelbin 25 13 PR* (Schuler et al, 2001)Heach & neck II Cisplatin +5-FU 11 9 PR* (Khuri et al, 2000)(_) on patients(*) on measurable lesions31

Jekunen et al: Strategy of sensitizing tumor cells with adenovirus-p53 transfectionV. ConclusionSeveral subsequent studies have confirmed thatvarious malignant cell lines and tumors expressing mutantor deleted p53 are chemoresistant to a wide range ofanticancer agents. However, other studies disagreesuggesting that cells with impaired p53 function canbecome sensitized to various anticancer agents. Thus, therelationship between p53 status and chemosensitivity iscomplex and presumably depends on a number of factors,including the specific cytotoxic stimuli, tissue-specificdifferences, and the specific cellular context thatincorporates the overall genetic machinery and the variousintracellular signaling pathways (Chu and DeVita 2001).The relationship between p53 and chemotherapy dependson the chemotherapeutic agents used, the target and thecritical tissues, and the intracellular signal transductionpathways affected.The theoretical basis of the sensitizing effect ofchemotherapeutic agents in combination with adenovirusp53 has been presented and so have a number ofsupportive data. As adenovirus p53 has its own activity,there seems to be a possibility that the cytotoxicity may beenhanced at least in some cell lines by transfer of the geneinto the tumor cells. This concept has reached the level ofproof in some, although not all, experimental conditions.This leaves a room for doubt, as all spontaneous solidtumors are heterogeneous and there may always remaincell clones that fail to obey the sensitizing principle. It isclear that more evidence is needed to support thisprinciple, especially clonogenic assays and classicalinteraction studies. Although the in vivo experiments areconvincing and strongly positive, it may not be altogethercorrect to extrapolate these results into clinical practice.There is a relative lack of pharmacokinetic studies andpharmacokinetic interaction studies in adenovirus p53gene therapy.Several strategies may be used to develop p53-basedanticancer therapies, with the goal of resensitizing tumorcells to conventional chemotherapy (Chang 2000). Theseinclude reintroduction of the gene encoding wild-type p53and methods for restoring normal p53 function to mutantp53. In addition, methods are being developed that targetthe p53-mdm-2 interaction of using lack of wild-type p53in tumors to protect normal tissue from the adverse effectsof chemotherapy. Replacement of the wild-type p53 byintratumoral transfection has already reached the phase IIIstage of clinical trials. Transfection of p53 can becombined with radioimmunotherapy as part of a tumormanipulation scheme (Kairemo, Jekunen et al, 1999).Increasing suppressor gene p53 expression in tumor cellsimproves the sensitivity of the tumor cells to routinechemotherapy. In a variety of tumor types, docetaxel andirinotecan are efficacious drugs with a new mode ofaction: prevention of depolymerization of tubulin andinhibition of specific DNA topoisomerase I, respectively.But we cannot obtain responses from all tumors, and insome tumors the efficacy, although established,diminished with time. In these cases of resistant tumors orrecurrences and relapses, combined treatment with adenop53and chemotherapeutic agents may be an attractivestrategy for inhibiting progression of local cancers. It isclear that even a modest change in drug sensitivity maybring some refractory tumors within a range that istreatable with conventional chemotherapy. Future therapymight couple standard cytotoxic agents with new biologicagents that attack specific molecular targets to reregulatethe cell-cycle checkpoint.Human data supporting the effect of sensitizingchemotherapy with adenovirus p53 is still maturing,although we have not found a way to use systemicadministration. We know that is s safe to performintratumoral gene therapy with adenovirus either with areplication non-competent or replication competent vector.As yet, there is no clinical evidence to support a definiteconclusion that adenovirus p53 provides a clinicallymeaningful improvement on conventional chemotherapy.However, it is clear that in some trial set ups it has beenpossible to demonstrate encouraging results and thepossibility of a clinical sensitizing effect of p53 genetherapy on the chemotherapy used when specificallyindicated. Intratumoral expression of transgenes andtumor-selective tissue destruction have been documentedin phase I and phase II clinical trials of adenovirus p53mediated gene therapy. However, durable responses andthe clinical benefit seen have been limited, with of 10-15%response rates.The rationale of combining p53 gene therapy with achemotherapeutic agent in the clinical setting has beennoted to be as follows: combinations of agents withdifferent toxicologic profiles can result in increasedefficacy without increased overall toxicity, they maythwart the development of resistance to the single agents,they may offer a solution to the problem of heterogeneoustumor cell populations with different drug sensitivityprofiles and they allow the physician to take advantage ofpossible synergies between drugs, resulting in increasedanticancer efficacy in patients (Nielsen, Lipari et al, 1998).Several phase III clinical trials with adenovirus p53therapy in head and neck cancer, NSCLC, and ovariancancer, will be completed in the near future, and the roleof gene therapy may become routine a part of treatmentregimens.AcknowledgmentsWe would like to thank Aventis Pharma Finland forsupporting this work.ReferencesBadie B, Kramar MH, Lau R, Boothman DA, Economou JS,Black KL. (1998). “Adenovirusmediated p53 gene deliverypotentiates the radiation-induced growth inhibition ofexperimental brain tumors.” J Neurooncol 37: 217-222.Blagosklonny, M. V. and W. S. El-Deiry (1996). “In vitroevaluation of a p53-expressing adenovirus as an anti-cancerdrug.” Int J Cancer 67: 386-392.Blandino G, Levine AJ, Oren M. (1999). “Mutant p53 gain offunction: differential effects of different p53 mutants onresistance of cultured cells to chemotherapy.” Oncogene18(2): 477-85.32

Gene Therapy and Molecular Biology Vol 7, page 33Brown, M. J. and B. G. Wouters (1999). “Apoptosis, p53, andtumor cell sensitivity to anticancer agents.” Cancer Res 59:1391-1399.Bunz, F. (1999). “Disruption of p53 in human cancer cells altersthe responses of chemotherapeutic agents.” J Clin Invest104: 263-269.Calvert, A. H. (1999). “Carboplatin and paclitaxel, alone and incombination: dose escalation, measurement of renal function,and role of the p53 tumor supressor gene.” Semin. Oncol.29: 90-94.Carrier, F., Georgel, P.T., Pourquier, P., Blake, M., Kontny,H.U., Antinore, M.J., Gariboldi, M., Myers, T. G, Weinstein,J.N., Pommier, Y,and Fornace, A.J., Jr. (1999). “Gadd45, ap53-responsive stress protein, modifies DNA accessibility ondamaged chromatin.” Mol Cell Biol 19(3): 1673-85.Chang, E. H. (2000). “Tp53 gene therapy: akey to modulatingresistance to anticancer therapies?” Molecular MedicineToday 6: 358-364.Chu, E. and V. T. J. DeVita (2001). Apoptosis, cell-cycle control,and resistance to chemotherapy. Cancer. Principles &Practice of Oncolgy. V. T. J. DeVita, S. Hellman and S. A.Rosenberg. Philadelphia, Lippincott Williams & Wilkins. 6:289-306.Clahsen PC, van de Velde CJ, Duval C, Pallud C, Mandard AM,Delobelle-Deroide A, van den Broek L, Sahmoud TM, vande Vijver MJ. (1998). “p53 protein accumulation andresponse to adjuvant chemotherapy in premenopausal womenwith node-negative early breast cancer.” J Clin Oncol 16(2):470-3945.Clayman GL, el-Naggar AK, Lippman SM, Henderson YC,Frederick M, Merritt JA, Zumstein LA, Timmons TM, LiuTJ, Ginsberg L, Roth JA, Hong WK, Bruso P, Goepfert H.(1998). “Adenovirus-mediated p53 gene transfer in patientswith advanced recurrent head and neck squamous cellcarcionoma.” JCO 16: 2221-2232.Constenla-Figueiras M, Betticher DC, DelCampo JM, Hitt R,Rochlitz C, Dhondt V, Gautier E, Saulnier P, Yver A, BadriN. (1999). “A phase II trial with Ad5CMV-p53 as a singleagent in recurrent/refractory SCCHN looking at vectorbiodistribution and horizontal trnasmission under normal lifeconditions.” Proc Am Soc Clin Oncol 18: 444a.Dorigo O, Turla ST, Lebedeva S, Gjerset RA. (1998).“Sensitization of rat gliblastoma multiforme to cisplatin invivo following restoration of wild-type p53 function.” JNeurosurg 88: 535-540.Fan S, Smith ML, Rivet DJ 2nd, Duba D, Zhan Q, Kohn KW,Fornace AJ Jr, O'Connor PM. (1995). “Disruption of p53function sensitizes breast cancer MCF-7 cells to cisplatin andpentoxifylline.” Cancer Res 55(8): 1649-907.Ferreira, C. G. (1999). “p53 and chemosensitivity.” Ann Oncol9: 1011-1021.Fujiwara T, Grimm EA, Mukhopadhyay T, Zhang WW, Owen-Schaub LB, Roth JA. (1994). “Induction of chemosensitivityin human lung cancer cells in vivo by adenovirus-mediatedtransfer of the wild-type p53 gene.” Cancer Res 54(9):2287-91.Gjerset, R. A., O. Dorigo, et al, (1997). “Tumor regression invivo following p53 combination therapy.” Cancer GeneTherapy 4: 0-2.Gjerset, R. A. and D. Mercola (2000). “Sensitizing of tumors tochemotherapy through gene therapy.” Cancer GeneTherapy 465: 273-291.Gjerset RA, Turla ST, Sobol RE, Scalise JJ, Mercola D, CollinsH, Hopkins PJ. (1995). “Use of wild-type p53 to achievecomplete treatment sensitization of tumor cells expressingendogenous mutant p53.” Mol Carcinog 14: 275-285.Gurnani M, Lipari P, Dell J, Shi B, Nielsen LL. (1999).“Adenovirus-mediated p53 gene therapy has greater efficacywhen combined with chemotherapy against human head andneck, ovarian, prostate, and breast cancer.” CancerChemother Pharmacol 44(2): 143-51.Gurnani M, Lipari P, Dell J, Shi B, Nielsen LL. (1998).“Adenovirus-mediated p53 gene therapy and paclitaxel havesynergistic efficacy in models of human head and neck,ovarian, prostate, and breast cancer.” Clin Cancer Res 4:835-846.Hawkins DS, Demers GW, Galloway DA. (1996). “Inactivationof p53 enhances sensitivity to multiple chemotherapeuticagents.” Cancer Res 56: 892-898.Heise C, Sampson-Johannes A, Williams A, McCormick F, VonHoff DD, Kirn DH. (1997). “ONYX-015, an E1B geneattenuatedadenovirus, causes tumor-specific cytolysis andantitumoral efficaacy that can be augmented by standardchemotherapuetic agents.” Nature Medicine 3: 639-645.Horio Y, Hasegawa Y, Sekido Y, Takahashi M, Roth JA,Shimokata K. (2000). “Synergistic effects of adenovirusexpressing wild-type p53 on chemosensitivity of non-smallcell lung cancer cells.” Cancer Gene Ther 7(4): 537-44.Hwu, P. (2001). Gene therapy. Cancer. Principles & Practiceof Oncology. V. T. J. DeVita, S. Hellman and S. A.Rosenberg. Philadelphia, Lippincott Williams & Wilkins. 6:3161-3180.Kairemo KJ, Jekunen AP, Tenhunen M. (1999). Dosimetry andoptimization of in vivo targeting with radiolabeled antisenseoligonucleotides: oligonucleotide radiotherapy. Methods ofEnzymology. New York, Academic Press. 314: 506-525.Khuri FR, Nemunaitis J, Ganly I, Arseneau J, Tannock IF,Romel L, Gore M, Ironside J, MacDougall RH, Heise C,Randlev B, Gillenwater AM, Bruso P, Kaye SB, Hong WK,Kirn DH. (2000). “A controlled trial of intraumoral ONYX-015, a selectively-replicating adenovirus, in combinationwith cisplatin and 5-fluorouracil in patients with recurrenthead and neck cancer.” Nature Medicine 6: 879-885.Kim J, Hwang ES, Kim JS, You EH, Lee SH, Lee JH. (1999).“Intraperitoneal gene therapy with adenoviral-mediated p53tumor suppressor gene for ovarian cancer model in nudemouse.” Cancer Gene Ther 6(2): 172-8.Kimura M, Tagawa M, Takenaga K, Yamaguchi T, Saisho H,Nakagawara A, Sakiyama S. (1997). “Inability to induce thealteration of tumorigenicity and chemosensitivity of p53-nullhuman pancreatic carcinoma cells after the transduction ofwild-type p53 gene.” Anticancer Res 17(2A): 879-83.Kirsch, D., Kastan, M.B. (1998). “Tumor-suppressor p53:implications for tumor development and prognosis.” JCO16: 3158-3168.Koechli, O. (1994). “Mutant p53 protein associated withchemosensitivity in breast cancer specimens.” Lancet 344:1647-1648.Lebedeva S, Bagdasarova S, Tyler T, Mu X, Wilson DR, GjersetRA. (2001). “Tumor suppression and therapy sensitization oflocalized and metastatic breast cancer by adenovirus p53.”Hum Gene Ther 12(7): 763-772.Levine, A. J. (1992). “The p53 tumour suppressor gene andproduct.” Cancer Surveys 12: 59-79.Lotem, J. and L. Sachs (1993). “Hematopoietic cells from micedeficient in wild-type p53 are more resistant to induction ofapoptosis by some agents.” Blood 82: 1092-1096.33

Jekunen et al: Strategy of sensitizing tumor cells with adenovirus-p53 transfectionLowe SW, Bodis S, McClatchey A, Remington L, Ruley HE,Fisher DE, Housman DE, Jacks T. (1994). “p53 status andthe efficacy of cancer therapy in vivo.” Science 266(5186):807-10.Lowe SW, Ruley HE, Jacks T, and Housman DE. (1993). “p53-mediated apoptosis modulates the cytotoxicity of anticanceragents.” Cell 74: 957-967.McPake CR, Shetty S, Kitchingman GR, Harris LC.. (1999).“Wild-Type p53 Induction Mediated by Replication-deficientAdenoviral Vectors.” Cancer Res 59(17): 4247-907.Miyake H, Hara I, Gohji K, Yamanaka K, Arakawa S, KamidonoS. (1998). “Enhancement of chemosensitivity in humanbladder cancer cells by adenoviral-mediated p53 genetransfer.” Anticancer Res 18(4C): 3087-92.Miyake H, Hara I, Hara S, Arakawa S, Kamidono S. (2000).“Synergistic chemosensitization and inhibition of tumorgrowth and metastasis by adenovirus-mediated P53 genetransfer in human bladder cancer model.” Urology 56(2):332-6.Nguyen DM, Spitz FR, Yen N, Cristiano RJ, Roth JA (1996).“Gene therapy for lung cancer: enhancement of tumorsuppression by a combination of sequential systemiccisplatin and adenovirus-mediated p53 transfer.” J ThoracCardiovasc Surg 112: 1372-1376.O'Connor PM, Jackman J, Bae I, Myers TG, Fan S, Mutoh M,Scudiero DA, Monks A, Sausville EA, Weinstein JN, FriendS, Fornace AJ Jr, Kohn KW. (1997). “Characterization of thep53 tumor suppressor pathway in cell lines of the NationalCancer Institute anticancer drug screen and correlations withthe growth-inhibitory potency of 123 anticancer agents.”Cancer Res 57(19): 4285-300.Oren, M. (1992). “p53: the ultimate tumor suppressor gene?”Faseb J 6(13): 3169-76.Osaki S, Nakanishi Y, Takayama K, Pei XH, Ueno H, Hara N.(2000). “Alteration of drug chemosensitivity caused by theadenovirus-mediated transfer of the wild-type p53 gene inhuman lung cancer cells.” Cancer Gene Ther 7(2): 300-7.Parker LP, Wolf JK, Price JE. (2000). “Adenoviral-mediatedgene therapy with Ad5CMVp53 and Ad5CMVp21 incombination with standard therapies in human breast cancercell lines.” Ann Clin Lab Sci 30(4): 395-405.Peller, S. (1998). “Clinical implications of p53: effect onprognosis, tumor progression and chemotherapy response.”Semin. Cancer Biol. 8: 379-387.Perdomo JA, Naomoto Y, Haisa M, Fujiwara T, Hamada M,Yasuoka Y, Tanaka N. (1998). “In vivo influence of p53status on proliferation and chemoradiosensitivity in nonsmall-celllung cancer.” J Cancer Res Clin Oncol 124(1):10-8.Polyak K, Xia Y, Zweier JL, Kinzler KW, Vogelstein B. (1997).“A model for p53-induced apoptosis.” Nature 389: 300.Righetti SC, Della Torre G, Pilotti S, Menard S, Ottone F,Colnaghi MI, Pierotti MA, Lavarino C, Cornarotti M, OrianaS, Bohm S, Bresciani GL, Spatti G, Zunino F. (1996). “Acomparative study of p53 gene mutations, proteinaccumulation, and response to cisplatin-based chemotherapyin advanced ovarian carcinoma.” Cancer Res 56(4): 689-93.Roth, J. A. (1996). “Modification of tumor-suppressor-geneexpression and induction of apoptosis in non-small-cell lungcancer (NSCLC) with an adenovirus vector expressing wildtypep53 and cisplatin.” Hum Gene Ther 7: 1013-1030.Roth JA, Nguyen D, Lawrence DD, Kemp BL, Carrasco CH,Ferson DZ, Hong WK, Komaki R, Lee JJ, Nesbitt JC, PistersKM, Putnam JB, Schea R, Shin DM, Walsh GL, DolormenteMM, Han CI, Martin FD, Yen N, Xu K, Stephens LC,McDonnell TJ, Mukhopadhyay T, Cai D. (1996).“Retrovirus-mediated wild-type p53 gene transfer to tumorsof patients with lung cancer.” Nature Medicine 2(9): 985-991.Santoso JT, Tang DC, Lane SB, Hung J, Reed DJ, Muller CY,Carbone DP, Lucci JA 3rd, Miller DS, Mathis JM. (1995).“Adenovirus-based p53 gene therapy in ovarian cancer.”Gynecol Oncol 59(2): 171-8.Schornagel JH, Verweij J, de Mulder PH, Cognetti F, VermorkenJB, Cappelaere P, Armand JP, Wildiers J, de Graeff A,Clavel M, et al, (1995). “Randomized phase III trial ofadatrexate versus methotrexate in patients with metastaticand/or recurrent squamous cell carcinoma of the head andneck: A European Organization for Research and Treatmentof Cancer Head and Neck Cancer Cooperative Group study.”JCO 13: 1649-1655.Schuler M, Herrmann R, De Greve JL, Stewart AK, GatzemeierU, Stewart DJ et al (2001). “Adenovirus-Mediated Wild-Type p53 Gene Transfer in Patients ReceivingChemotherapy for Advanced Non-Small-Cell Lung Cancer:Results of a Multicenter Phase II Study.” J Clin Oncol19(6): 1750-1758.Selter, H. and M. Montenarh (1994). “The emerging picture ofp53.” Int J Biochem 26: 154-154.Shahin MS, Hughes JH, Sood AK, Buller RE. (2000). “Theprognostic significance of p53 tumor suppressor genealterations in ovarian carcinoma.” Cancer 89(9): 2006-17.Sherr, C. J. (1994). “G1 phase progression: cycling on cue.” Cell79: 55-62.Stal, O. (1995). “p53 expression and the result of adjuvanttherapy of breast cancer.” Acta Oncologica 34: 767-770.Sugrue MM, Shin DY, Lee SW, Aaronson SA. (1997). “Wildtypep53 triggers a rapid senescence program in humantumor cells lacking functional p53.” Proc Natl Acad Sci U SA 94(18): 9648-53.Swisher SG, Roth JA, Nemunaitis J, Lawrence DD, Kemp BL,Carrasco CH, Connors DG, El-Naggar AK, Fossella F,Glisson BS, Hong WK, Khuri FR, Kurie JM, Lee JJ, Lee JS,Mack M, Merritt JA, Nguyen DM, Nesbitt JC, Perez-Soler R,Pisters KM, Putnam JB Jr, Richli WR, Savin M, Waugh MK,et al, (1999). “Adenovirus mediated p53 gene tranfer inadvanced non-small-cell lung cancer.” J Natl Cancer Inst91: 763-771.Trepel M, Groscurth P, Malipiero U, Gulbins E, Dichgans J,Weller M. (1998). “Chemosensitivity of human malignantglioma: modulation by p53 gene transfer.” J Neurooncol39(1): 19-32.Wolf JK, Mills GB, Bazzet L, Bast RC Jr, Roth JA, GershensonDM. (1999). “Adenovirus-mediated p53 growth inhibition ofovarian cancer cells is independent of endogenous p53status.” Gynecol Oncol 75(2): 261-6.von Gruenigen VE, Santoso JT, Coleman RL, Muller CY, MillerDS, Mathis JM. (1998). “In vivo studies of adenovirus-basedp53 gene therapy for ovarian cancer.” Gynecol Oncol 69(3):197-204.Xu GW, Sun ZT, Forrester K, Wang XW, Coursen J, Harris CC.(1996). “Tissue-specific growth suppression andchemosensitivity promotion in human hepatocellularcarcinoma cells by retroviral-mediated transfer of the wildtypep53 gene.” Hepatology 24(5): 1264-8.Yang B, Eshleman JR, Berger NA, Markowitz SD. (1996).“Wild-type p53 protein potentiates cytotoxicity of34

Gene Therapy and Molecular Biology Vol 7, page 35therapeutic agents in human colon cancer cells.” ClinCancer Res 1996: 1649-1657.Knudson CM, Tung KSK, Tourtellotte WG, Brown, GAJ,Korsmeyer SJ. (1997). “Bax suppresses tumorigenesis andstimulates apoptosis in vivo.” Nature 385(6617): 637-40.Yver, A., J. J. Nemunaitis, et al, (2001). “Does detection ofcirculating ONYX-015 genome by polymerase chain reactionindicate vector replication.” JCO 19(12): 3155-3157.35

Jekunen et al: Strategy of sensitizing tumor cells with adenovirus-p53 transfection36

Gene Therapy and Molecular Biology Vol 7, page 37Gene Ther Mol Biol Vol 7, 37-42, 2003Antigenicity and immunogenicity of HIV envelopegene expressed in baculovirus expression systemResearch ArticleAlka Arora 1 , Pradeep Seth 2*1Post Doctoral Fellow, Department of Medical Genetics and Microbiology, University of Toronto, Canada.2Professor and Head Department of Microbiology, All India Institute of Medical Sciences, India.__________________________________________________________________________________*Correspondence: Dr. Pradeep Seth, Professor and Head, Department of Microbiology, All India Institute of Medical Sciences, AnsariNagar, New Delhi, India -110029.Tel: 91-11-652 6814; Fax: 91-11-686 2663; E mail: pseth@aiims.aiims.ac.inReceived: 28 December 2002; Accepted: 5 February 2003; electronically published: July 2003SummaryHuman immunodeficiency virus type I (HIV-1) envelope gene was expressed in Spodoptera frugiperda (Sf21) cells.DNA constructs encoding env-tat-rev genes were cloned into the baculovirus expression vector pBacPAK9.Recombinant baculovirus was prepared by cotransfection with linearized wild type virus DNA. Western blotting ofcell extracts containing recombinant HIV-1 proteins demonstrated expression of HIV-1 gp160 and its completecleavage products gp120 and gp41. A time course experiment suggested that the maximum expression was observedat 48-hrs post infection. In order to measure the biological activity recombinant HIV envelope proteins were usedfor lymphocyte proliferation assay. The results demonstrated that recombinant gp160 and its cleavage productswere antigenically and functionally authentic.I. IntroductionHIV genome, like other retroviruses encode for Gag,Pol and Env. In addition, it also encodes for 6 regulatoryand accessory proteins Tat, Rev, Nef, Vif, Vpr and Vpu.The major structural protein encoded by env gene of HIV-1 consists of a protein of 850-880 amino acids. Extensiveglycosylation of this precursor protein results in theproduction of Gp160 monomers, which then assemble intooligomers for transport from ER to the plasma membrane(Earl et al, 1991). During transport from Golgi,intracellular cleavage of Gp160 yields an outer envelopeglycoprotein Gp120 and trans-membrane glycoproteinGp41 (Kozarsky et al, 1989). Specifically, the HIV viralenvelope protein Gp120 is important for virus-receptorinteraction and virus entry (Kowalski et al, 1987, Hill et al,1997). Gp41 is known to play a central role in theenvelope glycoprotein oligomerization and fusion function(Poumbourios et al, 1997). HIV infection results in theproduction of HIV specific antibodies, therefore detectionof these antibodies by ELISA and Western blot assayremains the basis of blood donor and patient screening.Serum specimen from HIV infected people regardless oftheir clinical stage react efficiently with precursorglycoprotein Gp160 or its cleavage product Gp120 andGp41 (Lange et al, 1986; Goudsmit et al, 1987).Antibodies to gag protein p24 are the earliest proteindetectable by Western blot after infection, however, thesetend to decrease with progression of clinical symptoms(Lange et al, 1986; Goudsmit et al, 1987). Recombinantantigen based EIAs have been shown to be more sensitive,especially in detecting early seroconverters and specificthan peptide or virus lysate based EIAs (Johnson 1992;Galli et al, 1996). The main objective of this study was toobtain large quantities of purified recombinant protein,suitable to be used as an immunogen and for developmentof HIV-1 detection kit. We used Baculovirus expressionvector system for expressing HIV-1 Gp160 as this systemresults in efficient processing of the protein, posttranslationalmodifications and is known to give highyields of expressed protein.II. Materials and MethodsA. Plasmids, cells, reagents and peptidespCR-Script SK (+) cloning vector was purchased fromStratagene, LaJolla, CA, USA. pBRU plasmid containingcomplete genome of BRU strain of HIV-1 cloned in pUC18 wasobtained through the AIDS Research and Reference ReagentProgram, Division of AIDS, NIAID, NIH, Bethesda, MD, USA.BacPAK, Baculovirus expression system was purchased fromClontech (BD Biosciences Clontech, Palo Alto, CA). Plasmidswere grown in DH5α strains of Escherichia coli (LifeTechnologies, Gaithesburg, MD, USA), and purified usingWizard miniprep columns (Promega Corp, Madison, WI). TNM-FH media for insect cell culture was obtained from HyClone(Genetix, New Delhi, India). TNM-FH medium contains Grace's37

Arora and Seth: Antigenicity and immunogenicity of HIV envelope genemedium, lactalbumin hydrolysate and yeast extract. Sf 21 cellswere cultured at 27 o C in TNM-FH medium supplemented with10% FBS (TNM-FH/FBS). Vaccinia expressed recombinantgp120 and gp160 (vPE8 and vPE16) were obtained through NIHAIDS Research and Reference Reagent Program.B. DNA constructs3kb env-tat-rev gene segment (nt 5352- nt 8354) of HIV-1subtype B strain, BRU, was PCR amplified using primers API(5352-5390) TTATTCTAGAGAGAAGAGCAAGAAATGGATCCAGTAGAT and APII (8316-8354)TTTTTGAGCTCTTGCCACCCATTTTAAAGTAAAGACCTTand cloned into pCR-Script (SK+) cloning vector to producepSBRU-TRE as described earlier (Arora and Seth, 2001). The3kb HIV-1 env-tat-rev gene segment was released by restrictiondigestion of pSBRU-TRE with Xba I, Not I and Bgl I. The env,tat and rev gene fragment was then purified from low meltingpoint agarose gel and subcloned into baculovirus transfer vector,pBacPAK9 predigested with Xba I and Not I to generatepBacBRU-TRE. Recombinant clone was screened by colonyhybridization followed by restriction enzyme analysis. pCIBRU-TRE, mammalian expression vector expressing 3kb HIV-1 envtat-revgene under the control of Immediate-EarlyPromoter/Enhancer of CMV, used in this study for immunizingBalb/c mice has been described earlier (Arora and Seth, 2001).C. Generating a recombinant virusRecombinant virus was prepared as per manufacturer'sinstructions. Briefly, 35mm tissue culture dishes were seededwith 1x10 6 Spodoptera frugiperda cells (Sf21) (Vaughn et al,1977) in 1.5 ml of complete TNM-FH/FBS medium andincubated overnight at 27 o C in a humid chamber. 500ng ofplasmid pBacBRU-TRE DNA, along with Bsu 361 digestedBacPAK6 viral DNA was mixed with 5µg of lipofectin andincubated at room temperature for 15 min. Culture medium in thetissue culture dishes containing Sf21 cells was replaced with 1.5ml of serum free TNM-FH. Lipofectin-DNA complex was thengently added to Sf21 cells. Plates were incubated at 27 o C for 5hrs. Thereafter, serum free TNM-FH medium was replaced withTNM-FH/FBS medium and the plates were returned forincubation at 27 o C for 4 days.D. Isolation of recombinant virusPlaque assay was performed using co-transfectionsupernatant to generate a pure clone of recombinant virus. 1x10 6Sf21 cells were seeded in 35mm tissue culture dishes andincubated overnight at 27 o C. These cells were then infected with100µl of neat or 10 -1 dilution of co-transfection supernatant. Onehour later, the virus inoculum was removed and infected cellswere overlaid with 1.5ml of agarose (1.5% in TNM-FH/FBS).After agarose was set 1.5 ml of TNM-FH/FBS medium wasadded to each dish and incubated for 4 days at 27 o C. Plaqueswere stained with .03% of neutral red solution. 4 plaques werepicked up and transferred into an eppendorf tube containing500µl of TNM-FH/FBS and stored at 4 o C overnight.E. Virus propagation and evaluationThe plaque picks were used as a source of virus to infectcells in a 96 well plate. Infections were performed in duplicate.Cells were harvested 4 days following infection and cell lysatewas used to perform dot blot analysis to detect the recombinantvirus. Each sample was suspended in 200µl of 0.5N NaOH and20µl of 10M-ammonium acetate. Samples were then spotted onto the nitrocellulose membrane by loading on the wells of the dotblot manifold apparatus (Bio Rad Laboratory, Richmond, CA).Vacuum suction was applied to drain off the entire solution.Membrane was dried at room temperature for 5-10 min and thenbaked for 2 hrs at 80 o C. Hybridization was performed usingα 32 P-dCTP labeled envelope probe prepared by random primerlabeling using Klenow fragment of DNA polymerase 1(Amersham Biosciences, Piscataway, NJ). The membrane wasthen washed and exposed to a Kodak-X film overnight at -70 o C.F. In vitro expressionA time course experiment was performed to examine theexpression of HIV-1 env gene in Sf 21 cells infected with therecombinant virus. Cells were harvested at various time intervalspost infection. SDS PAGE, immunofluorescence and WesternBlot analysis of cell lysate were conducted to study expression ofproteins. SDS-PAGE was performed according to Laemmli. ForWestern Blot analysis proteins were resolved by SDS-PAGE andtransferred onto a nitrocellulose membrane using Trans-blot SDsemi-dry electrophoretic transfer Cell (Bio Rad Laboratories)The membrane was treated with non-fat powdered milk in TTBS(Tween 20- Tris buffer Saline) for 1 hr at room temp. and reactedwith HIV-1 positive human polyclonal serum (at a dilution of1:200) in TBS for 1h at room temperature. After washing thricewith TTBS, the membrane was incubated at room temperaturefor 1 hr. with anti-human IgG conjugated with alkalinephosphatase (1:10,000). Membrane was then washed thrice withTTBS and incubated in the substrate solution (Sigma fastBCIP/NBT tablet dissolved in 10ml of deionized water, SigmaChemicals Co., St. Louis). For Immunofluorescence, P4(recombinant baculovirus) infected cells, uninfected cells(control) and AcNPv (wild type virus) infected cells wereharvested at different time points and washed thrice with PBS.1x10 4 cells were spotted onto the wells of a teflon-coated slideand fixed with acetone: methanol (1:1) at -20 o C for 30 min. Forstaining, cells were allowed to react with HIV-1 positive humanpolyclonal serum (1:50) for 1h at 37 o C. Cells were then washedwith PBS and incubated with FITC conjugated anti-human IgG(Sigma) and incubated for 1hr at 37 o C. Thereafter, the cells werewashed and mounted with glycerol buffer and visualized underfluorescent microscope.G. T cell proliferation assay3 H thymidine uptake assay was used to measure theproliferation of splenocytes after antigenic stimulation. Balb/cmice were immunized intramuscularly with pCIBRU-TRE orpCI (control vector) DNA as described earlier (Arora and Seth,2001). Six groups of Balb/c mice were taken (each groupcomprising 5 mice) (Table 1). In-group D3 three doses of 100 µgDNA each were given at bi-weekly intervals. In D0P2 groupanimals were immunized with 2 doses of P4 with no DNApriming. In-group D3P2 animals were immunized with 3 dosesof pCIBRU-TRE DNA followed by 2 doses of P4. Group D3V2consisted of mice immunized with 3 doses of pCIBRU-TREfollowed by 2 doses of recombinant vaccinia virus expressedgp120 and gp160 (vPE8 and vPE16). D0V2 group consisted ofmice immunized with 2 doses of vPE8 and vPE16 with nopriming with DNA construct and control group. Stimulatingantigens included vaccinia expressed recombinant gp160/gp120(vPE16/ vPE8) and baculovirus expressed gp160 (P4).Splenocytes from various groups of mice were harvested and resuspendedat a concentration of 2x10 6 cells/ml in RPMI 1640medium supplemented with 10% FCS. Cells were stimulated in38

Gene Therapy and Molecular Biology Vol 7, page 39triplicate. Five µg/ml of vPE16/vPE8 infected vero cell lysates/P4 infected Sf21 cell lysates was used in cell proliferation assay.Lysates of wild type vaccinia virus (WR) infected Vero cells/wild type baculovirus (AcNPv) infected Sf21 cell lysate wasused as control to study the non specific 3 H-thymidine uptakedue to wild type vaccinia/vero cell proteins or wild typebaculovirus/Sf21 cell protein in the cell lysates. Stimulationindex was calculated by the following formula.SI = Mean cpm of 3 H thymidine incorporated in the presence of stimulating antigen (vPE 16, vPES or P4)Mean cpm of 3 H thymidine incorporated in wild type virus (VacWR or AcNPv) controlII. ResultsA. Generating a Recombinant Baculovirus:Complete HIV-1 envelope glycoprotein along with theregulatory protein Tat and Rev were PCR amplified fromsubtype B, BRU strain of HIV-1 and cloned intopBacPAK9, baculovirus transfer vector, downstream tothe baculovirus polyhedrin gene promoter (Figure 1).Recombinant baculovirus transfer vector was screened bycolony hybridization followed by restriction enzymeanalysis and was termed as pBacBRU-TRE (Figure 2).Following co-transfection, recombinant baculovirus wasformed by the homologous recombination betweenpBacBRU-TRE and Bsu361 digested BacPAK6 viralDNA in the region flanking the chimeric gene, whichallows its insertion into the genome of the wild type virus.The BacPAK6 DNA is missing an essential portion of thebaculovirus genome, ORF1629, that is essential for viralreplication (Possee et al, 1991) When the DNArecombines with the vector (the transfer vector carries themissing ORF1629 sequence), the essential element isrestored and the target gene is transferred to thebaculovirus genome. Recombinant viruses were collectedand selected by plaque purification. Recombinantphenotype of the plaques is verified by Dot-Blot analysis.Two of the plaques were found to be positive by Dot-Blotanalysis and were termed as P4 and P5 (Figure 3). PlaqueP4 gave the stronger signal and was therefore amplifiedand used for further infections.Figure 1. a) pBacPAK9 baculovirus transfer vector, b)Recombinant plasmid pBacBRU-TRE. HIV-1 env, tat and revgene released on digestion of pSBRU-TRE was gel purified andsubcloned into baculovirus transfer vector pBacPAK9predigested with restriction enzymes Xba 1 and Not 1.Figure 2 a) Autoradiograph showing recombinant colonies asdetected by colony hybridization, b) Restriction enzyme analysisof the recombinant plasmid pBacBRU-TRE with differentenzymes. Lanes M: Lambda DNA digested with Hind IIIenzyme. Positions of the molecular weight markers are indicated,1: uncut; 2: pBacBRU-TRE digested with Bam H1; 3:pBacBRU-TRE digested with Hind III; 4: pBacBRU-TREdigested with Pvu II39

Arora and Seth: Antigenicity and immunogenicity of HIV envelope geneFigure 3. Autoradiograph showing dot blot analysis of celllysates from plaque picks infected Sf21 cells. 2 plaques labeledas P4 and P5 were found to be positive. Cells infected with wildtype baculovirus AcNPv served as the negative control. pBRUplasmid DNA served as the positive control.Figure 4. The photograph showing Immunofluorescencemicroscopy of the recombinant baculovirus infected Sf21 cells at48h-post infection. HIV-1 positive human polyclonal serumserved as the source of primary antibody.B. Expression of HIV-1 Envelopeglycoprotein by Recombinant BaculovirusExpre s sion of gp160 in Sf21 ce lls w a s exa mine d byindire ct immunofluores ce nce and w es te rn blot ana lysis ofinfec ted c e lls using HIV-1 pos itive huma n polyc lonal se ra .A 3+ fluore sc enc e wa s obs e rved at 48-hrs post infe c tion ona s ca le of 0 to 4+ that is from no fluore sc enc e to inte nsefluore sc enc e (Figur e 4). These results were supported bywestern blot analysis of the infected cells at 48hrs-postinfection. Gp160 and its cleavage products, Gp120 andGp41, could be detected after immunostaining. Since thetotal carbohydrate load added to the insect cell expressedglycoprotein is marginally less than that added duringsecretion from a mammalian cell, the baculovirusexpressed glycoprotein are correspondingly smaller (105kDa) than their mammalian counterparts (120 kDa) Nocorresponding protein bands were detected on from wildtype baculovirus (AcNPv) infected cells and uninfectedcells (Figure 5).Figure 5. Western blot analysis of recombinant baculovirusexpressed gp160. Lanes M: protein high range molecular weightmarker; 1: uninfected cell lysate; 2: cell lysate from AcNPvinfected cells; 3 & 4: cell lysate from recombinant baculovirusinfected cells.C. Lymphocyte Proliferation AssayIn vitro T cell proliferative activity of splenocytes fromanimals immunized with DNA vaccine pCIBRU-TREalone (group D3), boosted with P4 or vPE8/vPE16 (groupsD3P2, D3V2) or P4 and vPE8/vPE16 alone (groups D0P2,D0V2) was studied. (Table 1). Splenocytes from all theanimal groups showed positive proliferative response on invitro stimulation (Figure 6). Splenocytes from groupD0P2 mice demonstrated proliferation in response to P4cell lysate (SI-8.16), as well as to vPE8 and vPE16antigens (SI of 4 and 5.6). Splenocytes from DNA vaccineimmunized mice group D3 and D3P2 proliferated with SIof 8.8 on stimulation with vPE8 and with SI of 3.8 and 4.4respectively on stimulation with P4. Splenocytes frommice immunized with 2 doses of vaccinia expressedrecombinant Gp120/Gp160 with no DNA priming (GroupD0V2) showed better proliferation with vPE8, ascompared with vPE16 and P4. However, splenocytes frommice immunized with 3 doses of DNA followed by 2doses of vaccinia expressed recombinant Gp120/Gp160(Group D3V2) gave almost equal proliferation with P4,vPE8 and vPE16 respectively (Figure 6).Figure 6. In vitro T cell proliferative response to P4, vPE8 &vPE16 (recombinant baculovirus expressed gp160) ofsplenocytes from Balb/c mice immunized with pCIBRU-TRE (3doses at biweekly intervals) and boosted with 2 doses of eitherrecombinant baculovirus (P4) or recombinant vaccinia virus(vPE8 & vPE16). These groups of mice were marked as D3P2 orD3V2 respectively. Animals from groups D0P2 and D0V2 wereinjected only with recombinant baculovirus or recombinantvaccinia virus (no DNA priming).40

Gene Therapy and Molecular Biology Vol 7, page 41Table 1. Different groups of mice primed with pCIBRU-TRE DNA and boosted with baculovirus expressed (P4) orvaccinia expressed (vPE8 and vPE16) recombinant gp160.Group pCIBRU-TRE P4 vPE8 and vPE16D3D0P23 doses2 dosesD3P2 3 doses 2 dosesD0V22 dosesD3V2 3 doses 2 dosesIV. DiscussionThe main objective of this study was to prepare largeamounts of HIV-1 envelope protein, which may be used asa source of antigen for studying immune response againstHIV-1. HIV-1 gp160 with its signal sequence along withthe regulatory genes tat and rev was used to producerecombinant baculovirus (Malim et al, 1989; Ruben et al,1989 Rosen and Pavlakis; 1990, Roy et al, 1990). Thissystem has several advantages over other systemsincluding high level of protein production and posttranslationalmodification, which cannot be achieved inbacterial system (Luckow and Summers 1988, 1989). Weobserved poor expression of envelope proteins followinginfection of Sf21 cells as no protein was observed afterSDS-PAGE of the P4 infected Sf21 cell lysate followed bycoommassie blue staining. Several other studies haveindicated that env protein is refractory to efficientrecombinant expression (Lasky et al, 1986 Hu et al, 1987;Hu et al, 1987). Replacement of the signal sequence of theHIV-1 envelope protein with those of herpes simplex virusglycoprotein or human tPA results in efficient expression(Lasky et al, 1986; Berman et al, 1988). These studiestherefore suggest that the signal sequence of HIV-1envelope gene, which consists of 5 positively chargedamino acids, may be responsible for the poor expression.Li et al, (1994), showed that substitution of the gp120natural signal sequences with the signal sequences fromhoneybee mellitin or murine interleukin 3 promotes a highlevel of expression of a glycosylated form of gp120 andefficient secretion. These heterologous signal sequencescontain one (mellitin) or no (IL-3) positively chargedamino acids. These workers also demonstrated that onstepwise substitution of positively charged amino acidswith neutral amino acids resulted in enhanced expressionof HIV-1 gp120. Similarly, Golden et al, 1998, comparedthree different signal sequences [human tissueplasminogen activator (tPA), human placental alkalinephosphatase (pap), or baculovirus envelope glycoprotein(gp67)] and found that the tPA leader yielded the highestlevel of secreted protein, followed by the gp67 and papsequences.In this study, however, HIV-1 gp160 and its completecleavage products were observed on Western Blot analysisusing HIV-1 positive human polyclonal sera. Suggestingthereby that the envelope protein retained its antigenicityand may be used as a source of antigen for Western Blotanalysis. Immunogenicity as well as antigenicity of thisbaculovirus expressed envelope protein was alsodemonstrated by lymphocyte proliferation assays. Largescaleprotein purification is being pursued for furtherstudies.AcknowledgmentsThe Department of Biotechnology, Ministry ofScience and Technology, Government of India hasprovided financial support for this research. Ms AlkaArora received Research Fellowship from CSIR duringthis study.ReferencesArora A and Seth P (2001). Immuniz ation w ith H IV- 1 Subtype Bgp160-D NA Induces Spec if ic as we ll as c ross- rea ctive I mmuneRe sponses in Mice . Indian J Med Res 114, 1-9.Berman PW, Nunes WM and Haffar OK (1988) Expression ofmembrane-associated and secreted variants of gp160 ofhuman immunodeficiency virus type 1 in vitro and incontinuous cell lines. J Virol 62, 3135-42.Earl PL, Moss B and Doms RW (1991) Folding, interaction withGRP78-BiP, assembly, and transport of the humanimmunodeficiency virus type 1 envelope protein. J Virol 65,2047-55.Galli RA, Castriciano S, Fearon M, Major C, Choi KW, MahonyJ and Chernesky M (1996) Performance Characteristics ofRecombinant Enzyme Immunoassay To Detect Antibodies toHuman Immunodeficiency Virus Type 1 (HIV-1) and HIV-2and To Measure Early Antibody Responses inSeroconverting Patients. J Clin Microbiol 34, 999–1002.Golden A, Austen DA, van Schravendijk MR, Sullivan BJ,Kawasaki ES, Osburne MS (1998) Effect of promoters andsignal sequences on the production of secreted HIV-1 gp120protein in the baculovirus system. Protein Expr Purif 14, 8-12.Goudsmit J, Lange JMA, Paul DA, Dawson GJ (1987)Antigenemia and antibody titers to core and envelopeantigens in AIDS, AIDS-related complex, and subclinicalhuman immunodeficiency virus infection. J Infect Dis 155,558-60.Hill CM, Deng H, Unutmaz D, Kewalramani VN, Bastiani L,Gorny MK, Zolla-Pazner S, Littman DR (1997) Envelopeglycoproteins from human immunodeficiency virus types 1and 2 and simian immunodeficiency virus can use humanCCR5 as a co-receptor for viral entry and make direct CD4-dependent interactions with this chemokine receptor. J Virol71, 6296-304.Hu SI, Kosowski SG and Schaaf KF (1987) Expression ofenvelope glycoproteins of human immunodeficiency virus byan insect virus vector. J Virol 61, 3617-20.Johnson JE (1992) Detection of human immunodeficiency virustype 1 antibody by using commercially available whole-cellviral lysate, synthetic peptide, and recombinant proteinenzyme immunoassay systems. J Clin Microbiol 30,216–218.Kozarsky K, Penman M, Basiripour L, Haseltine W, Sodroski Jand Krieger M (1989) Glycosylation and processing of thehuman immunodeficiency virus type 1 envelope protein. JAcquir Immune Defic Syndr 2, 163-9.Kowalski M, Potz J, Basiripour L, Dorfman T, Goh WC,Terwilliger E, Dayton A, Rosen C, Haseltine W, Sodroski J41

Arora and Seth: Antigenicity and immunogenicity of HIV envelope gene(1987) Functional regions of the envelope glycoprotein ofhuman immunodeficiency virus type 1. Science 237, 1351-5.Lange JM, Paul DA, Huisman HG, de Wolf F, van den Berg H,Coutinho RA, Danner SA, van der Noordaa J, Goudsmit J(1986) Persistent HIV antigenemia and decline of HIV coreantibodies associated with transition to AIDS. Brit Med J293, 1459-62.Lasky LA, Groopman JE, Fennie CW, Benz PM, Capon DJ,Dowbenko DJ, Nakamura GR, Nunes WM, Renz ME,Berman PW (1986) Neutralization of the AIDS retrovirus byantibodies to a recombinant envelope glycoprotein. Science233, 209-12.Li Y, Luo L, Thomas DY, Kang CY (1994) Control ofexpression, glycosylation, and secretion of HIV-1 gp120 byhomologous and heterologous signal sequences Virology204, 266-78.Luckow VA and Summers MD (1988) Signals important forhigh-level expression of foreign genes in Autographacalifornica nuclear polyhedrosis virus expression vectors.Virology 167, 56-71Luckow VA and Summers MD (1989) High level expression ofnonfused foreign genes with Autographa californica nuclearpolyhedrosis virus expression vectors. Virology 170, 31-9.Malim MH, Hauber J, Le SY, Maizel JV and Cullen BR (1989)The HIV-1 rev trans-activator acts through a structured targetsequence to activate nuclear export of unspliced viralmRNA. Nature 338, 254-257.Possee RD, Sun TP, Howard SC, Ayres MD, Hill-Perkins M,Gearing KL (1991) Nucleotide sequence of the Autographacalifornica nuclear polyhedrosis 9.4 kbp EcoRI-I and -R(polyhedrin gene) region. Virology. 185, 229-41.Poumbourios P, Wilson KA, Center RJ, El Ahmar W and KempBE (1997) Human immunodeficiency virus type 1 envelopeglycoprotein oligomerization requires the gp41 amphipathicalpha-helical/leucine zipper-like sequence. J Virol 71, 2041-9.Rosen CA and Pavlakis GN (1990) Tat and Rev: positiveregulators of HIV gene expression. AIDS 4, A51Roy S, Delling U, Chen CH, Rosen CA, Sonenberg N (1990) Abulge structure in HIV-1 TAR RNA is required for Tatbinding and Tat-mediated trans-activation. Genes Dev 4,1365-1373.Ruben S, Perkins A, Purcell R, Joung K, Sia R, Burghoff R,Haseltine WA, Rosen CA (1989) Structural and functionalcharacterization of human immunodeficiency virus tatprotein. J Virol 63, 1-8.Vaughn JL, Goodwin RH, Tompkins GJ, McCawley P (1977)The establishment of two cell lines from the insectSpodoptera frugiperda (Lepidoptera; Noctuidae).In Vitro.13, 213-7.Pradeep Seth42

Gene Therapy and Molecular Biology Vol 7, page 43Gene Ther Mol Biol Vol 7, 43-59, 2003Characterization of genes transcribed in an Ixodesscapularis cell line that were identified by expressionlibrary immunization and analysis of expressedsequence tagsResearch ArticleConsuelo Almazán, Katherine M. Kocan, Douglas K. Bergman, Jose C. Garcia-Garcia, Edmour F. Blouin and José de la Fuente*Department of Veterinary Pathobiology, College of Veterinary Medicine, Oklahoma State University, Stillwater, OK74078.__________________________________________________________________________________*Correspondence: José de la Fuente, Department of Veterinary Pathobiology, College of Veterinary Medicine, Oklahoma StateUniversity, Stillwater, OK 74078; Phone: (405) 744-0372; Fax: (405) 744-5275; e-mail: jose_delafuente@yahoo.comKey words: tick, vaccine, tick cell culture, cDNA library immunization, EST, expression library immunizationReceived: 23 May 2003; Accepted: 06 June 2003; electronically published: June 2003SummaryExpression library immunization (ELI) combined with analysis of expressed sequence tags (ESTs) were used toidentify genes transcribed in a cell line (IDE8) that was originally derived from embryos of Ixodes scapularis. AcDNA expression library was constructed from the IDE8 cells and cDNA clones were screened by ELI. Miceinjected with cDNA clones were then infested with I. scapularis larvae. cDNA clones affecting larval feeding ordevelopment were subjected to single pass 5’ sequence analysis and the non-redundant sequences were putativelyidentified by sequence identity using the protein Basic Local Alignment Search Tool (BLAST) algorithm.Sequences of the clones were grouped according to the predicted function of the encoded proteins. 351 cDNAs thataffected larval feeding and/or development were identified, of which 316 cDNA clones contained non-redundantsequences and 101 produced a significant identity to sequences reported previously. Gene ontologies could beassigned to 87 clones. Vaccination of mice with plasmid DNA followed by tick infestation resulted in identificationof cDNA clones that inhibited tick infestation or promoted tick feeding. cDNAs that inhibited tick infestation wereidentical to nucleotidase, heat shock proteins, beta-adaptin, chloride channel, ribosomal proteins, and proteinswith unknown function. cDNA clones that promoted tick feeding were identical to beta-amyloid precursor, block ofproliferation, mannose-binding lectin, RNA polymerase III, ATPases and a protein of unknown function. Herein,we describe the sequence analysis of I. scapularis ESTs selected by ELI that affected larval tick feeding and/ordevelopment. These proteins may be useful for incorporation into vaccine preparations designed to interrupt thelife cycle of I. scapularis and/or interfere with transmission of pathogens.I. IntroductionTicks are ectoparasites of wild and domestic animalsand humans, and are considered to be the most importantvector of pathogens in North America (Parola and Raoult,2001). Ixodes spp. (Acari: Ixodidae) are distributedworldwide and are vectors of human pathogens, includingBorrelia burgdorferi (Lyme disease), Anaplasmaphagocytophilum (human granulocytic ehrlichiosis),Coxiella burnetti (Q fever), Francisella tularensis(tularemia), B. afzelii, B. lusitaniae, B. valaisiana and B.garinii, Rickettsia helvetica, R. japonica and R. australis,Babesia divergens, as well as tick-borne encephalitis(TBE) and Omsk Hemorrhagic fever viruses (Estrada-Peña and Jongejan, 1999; Parola and Raoult, 2001).Throughout eastern and southeastern United States andCanada, I. scapularis (the black legged tick) is the mainvector of B. burgdorferi sensu stricto and A.phagocytophilum (Estrada-Peña and Jongejan, 1999;Parola and Raoult, 2001).Control of tick infestations is difficult, particularlyfor multi-host ticks such as Ixodes spp. Presently, tick43

Almazán et al: Expressed sequence tags in Ixodes scapulariscontrol is effected by integrated pest management inwhich different control methods are adapted in ageographic area against one tick species with dueconsideration to their environmental effects. Recently,development of vaccines against one-host Boophilus spp.has provided new possibilities for identification ofprotective antigens for use in vaccines for control of tickinfestations (Willadsen, 1997; Willadsen and Jongejan,1999; de la Fuente et al, 1999, 2000a; de Vos et al, 2001).Control of ticks by vaccination would avoid environmentalcontamination and selection of drug resistant ticks that canresult from repeated acaricide application (de la Fuente etal, 1998; Garcia-Garcia et al, 1999). Anti-tick vaccinesalso allow for inclusion of multiple antigens in order totarget a broad range of tick species, as well as pathogenblockingantigens.Development of high throughput DNA sequencingtechnologies and bioinformatic tools facilitate assignmentof provisional function to expressed sequence tags (ESTs;Boguski et al, 1993). This approach has resulted invaluable information for the study of biological systemsand for the identification of potential vaccine candidates(Lizotte-Waniewski et al, 2000; Knox et al, 2001; Tarletonand Kissinger, 2001; Touloukian et al, 2001; Kessler et al,2002). In ticks, construction of EST databases has beenreported for B. microplus (Crampton et al, 1998),Amblyomma americanum (Hill and Gutierrez, 2000) andA. variegatum (Nene et al, 2002). The application of ESTtechnology has been used for characterization of geneexpression in salivary glands of I. scapularis (Valenzuelaet al, 2002), I. ricinus (Valenzuela, 2002), A. americanumand Dermacentor andersoni (Bior et al, 2002), foridentification of genes differentially expressed in D.variabilis ovaries in response to rickettsial infection(Mulenga et al, 2003) and in I. ricinus salivary glands inresponse to blood feeding (Leboulle et al, 2002).A new technique, expression library immunization(ELI), in combination with sequence analysis of ESTs,provides an alternative approach for identification ofpotential vaccine antigens that is based on rapid screeningof the expressed genes without prior knowledge of theantigens encoded by the cDNAs. ELI was first reported forMycoplasma pulmonis (Barry et al, 1995) and since thenhas been used for unicellular and multicellular pathogensand viruses (Manoutcharian et al, 1998; Alberti et al,1998; Brayton et al, 1998; Melby et al, 2000; Smooker etal, 2000; Moore et al, 2001; Singh et al, 2002; Leclercq etal, 2003). Recently, we reported the first application ofELI to arthropods, specifically to I. scapularis (Almazánet al, 2003) in a mouse model system. A combination ofcDNA ELI and EST analysis resulted in the selection of351 cDNA clones affecting tick larval development(Almazán et al, 2003). After grouping the clonesaccording to the putative function of predicted proteins,some cDNA pools resulted in the inhibition of tickinfestation and others promoted tick feeding after ELI(Almazán et al, 2003).Herein we describe the sequence analysis andcharacterization of I. scapularis ESTs that were identifiedby Almazán et al. (2003) using cDNA ELI and a mousemodel for tick infestation.II. Materials and methodsA. Construction of the I. scapularisexpression cDNA libraryThe cDNA library was constructed from I. scapulariscultured embryonic IDE8 cells (Munderloh et al, 1994) asreported previously (Almazán et al, 2003). The expression librarywas constructed in the vector pEXP1 containing the stronghuman cytomegalovirus major immediate earlypromoter/enhancer (CMV IE ) (Clontech, Palo Alto, CA). ThecDNA library contained 4.4 x 10 6 independent clones and a titerof approximately 10 10 cfu/ml with more than 93% of the cloneswith cDNA inserts. The average cDNA size was 1.7 kb (0.5-4.0kb).B. DNA vaccination and tick infestationVaccinations with plasmid DNA and tick infestations weredone as reported previously for the screening of the expressioncDNA library by ELI using the mouse model of I. scapualrisinfestations (Almazán et al, 2003). Briefly, plasmid DNA waspurified (Wizard SV 96 plasmid DNA purification system,Promega, Madison, WI) and used to inject CD-1 female mice, 5-6 weeks of age at the time of first vaccination. Mice were caredfor in accordance with standards specified in the Guide for Careand Use of Laboratory Animals. Mice were injected using a 1 mltuberculin syringe and a 27αG needle at days 0 and 14. Three to6 mice per group were each immunized IM in the thigh with 1 µgtotal DNA/dose in 50 µl PBS. Control mice were injected with 1µg vector DNA alone. Two weeks after the last immunization,mice were infested with 100 I. scapularis larvae per mouse. Fortick infestations, mice were retrained in a small wire cage in acardboard carton. One hundred larvae were counted and appliedto the mice with a brush. Ticks were reared at the OklahomaState University Tick Rearing Facility by feeding larvae on mice,nymphs on rabbits and adults on sheep. For these experiments,larvae were obtained from the eggs oviposited by sister females.Twelve hours after tick infestation, larvae in the bottom of thecage that did not attach were counted in order to calculate thenumber of attached larvae per mouse. Mice were then transferredto individual cages in which they were placed on an elevated1/4” mesh wire platform over water (1/2” deep). Replete larvaedropping from each mouse were collected daily from the waterand counted during 7 days. Time for larval development wasevaluated from the day of tick infestation to the day in which themaximum number of replete larvae was collected. The inhibitionof tick infestation (I) for each test group was calculated withrespect to vector-immunized controls as [1-(RLn/RLc xRLic/RLin)] x 100, where RLn is the average number of repletelarvae recovered per mouse for each test group, RLc is theaverage number of replete larvae recovered per mouse for controlgroup, RLic is the average number of larvae attached per mousefor control group, and RLin is the average number of larvaeattached per mouse for each test group. Engorged larvae wereheld in a 95% humidity chamber and allowed to molt. Molting ofengorged larvae was evaluated 34 days after the last larvalcollection by visual examination of ticks under a dissecting lightmicroscope. The inhibition of molting (M) for each test groupwas calculated with respect to controls as [1-(MLn/MLc xRLc/RLn)] x 100, where MLn is the average number of nymphsfor each test group, MLc is the average number of nymphs forthe control group, RLc is the average number of larvae recoveredfor the control group, and RLn is the average number of larvaerecovered for each test group.44

Gene Therapy and Molecular Biology Vol 7, page 45C. Plasmid DNA preparation and sequencingBacterial colonies were inoculated in Luria-Bertani with 50µg/ml ampicillin, grown for 16 hr in a 96-well plate and plasmidDNA purified (Wizard SV 96 plasmid DNA purification system,Promega, Madison, WI) and partially sequenced with a 5’ vectorspecificprimer (5’-CGACTCACTATAGGGAG-3’) at the CoreSequencing Facility, Department of Biochemistry and MolecularBiology, Noble Research Center, Oklahoma State University,using ABI Prism dye terminator cycle sequencing protocolsdeveloped by Applied Biosystems (Perkin-Elmer Corp., FosterCity, CA). In most cases a sequence larger than 700 nucleotideswas obtained.D. Data analysisNucleotide sequences were analyzed using the programAlignX (Vector NTI Suite V 5.5, InforMax, North Bethesda,MD). Multiple sequence alignment was performed using anengine based on the Clustal W algorithm (Thompson et al, 1994).Nucleotides were coded as unordered, discrete characters withfive possible character-states; A, C, G, T, or N (missing) andgaps were coded as missing data. Phylogenetic trees wereconstructed based on a sequence distance method utilizing theNeighbor Joining algorithm of Saitou and Nei (1987). BLAST(Altschul et al, 1990) was used to search the NCBI databases toidentify previously reported sequences with identity to those thatwe sequenced. Gene ontology assignments were made accordingto Ashburner et al. (2000) for non-redundant EST sequence datawith the help of GoFish v.1.0 (Berriz et al, 2003).III. ResultsThe screening of the I. scapularis expression cDNAlibrary by ELI and EST analysis resulted in 351 cDNAsaffecting larval development in the mouse model of tickinfestation (Almazán et al, 2003). Of them, 316 cDNAclones contained non-redundant sequences and 101 (32%)produced a significant identity to previously reportedsequences by BLAST analysis of NCBI nucleotide andprotein databases (Table 1). Gene ontologies could beassigned to 87 clones (27.5% of non-redundant sequencesand 86.1% of clones with identity to sequences reportedpreviously) (Table 2).Table 1. cDNA clones with identity to previously reported sequences.EST clone Predicted protein GenBankaccessionnumber1C111E6Translation initiation factor 5A(eIF5A)Translation initiation factor 5C(eIF-5C)CD052489CD0524902D2 Initiate factor 5 (if5) CD0524911A10 Elongation factor 2 CD0524924F7 Elongation factor 1alpha CD0524931F6 Ribosomal protein S4 (RpS4) CD0524942B8 Ribosomal protein S11 (RpS11) NR2F8Laminin receptor 1 (ribosomalprotein SA)CD0524962F10 Ribosomal protein L3 (RpL3) NR3A10 Ribosomal protein L7A (RpL7A) CD0524973G9 Ribosomal protein S8 (RpS8) CD0524953G103C3Ribosomal protein L27A(RpL27A)QM homolog (DQM) ribosomalproteinCD052498CD0524994D12 Proteasome/Signalosome subunit CD0525004E7 Proteasome subunit CD0525014D11 Proteasome subunit CD0525023D10 Ribophorin I CD0525031B12 Ubiquitin-conjugating enzyme CD0525041D10 Ubiquitin CD0525051A9V-ATPase D subunitContains microsatellite sequenceCD0525061B2 V-ATPase C subunit CD052507EST clone Predicted protein GenBankaccessionnumber4A4 V-ATPase E subunit CD0525081C5 Na+/K+ ATPase, alpha subunit CD0525092A9 NADH dehydrogenase CD0525101D6 NADH dehydrogenase subunit 5(nad5)CD0525111A4 Aldehyde dehydrogenase CD0525121C8 Virilizer (vir) CD0525131C10 Hsp70 CD0525143F6 Hsp60 CD0525151D1 Nucleotide binding protein 1(Nubp1)1D8Identity to D. melanogasterGH03607 full length cDNAcoding for a putative membraneproteinCD052516CD0525171D11 Putative membrane protein CD0525181E7 Sterol carrier protein CD0525191F3 Cyclin C (CycC) CD0525203D9 Alpha tubulin CD0525212A7 Beta tubulin CD0525222A11 Notchless (Nle) CD0525232B2 Export factor binding protein 2(Refbp2)CD0525242B7 G protein-coupled receptor CD0525252B92C122D1Succinate dehydrogenase B(SdhB)Beta-amyloid precursor protein(APP)Fructose-1,6-bisphosphatase (fbpgene)CD052526CD052527CD0525282D5 DNA repair protein Rad1 (Rad1) CD0525292D6Identity to S. pombe dim1+,helicase protein 1CD05253045

Almazán et al: Expressed sequence tags in Ixodes scapularis2E8 Esterase CD0525312F92F12Identity to AvGI TC255 (A.variegatum) & hypotheticalprotein FLJ12475 (H. sapiens)Transmembrane G-proteinresponsiveadenylyl cyclaseCD052532CD0525332G8 Lysyl-tRNA synthetase CD0525342H11Sodium- and chloride-dependenttaurine transporterCD0525353C12 RNA polymerase III CD0525363E1 Beta-adaptin CD0525373E2Microtubule-associated protein,RP/EB familyCD0525383E4 Myosin II regulatory light chain CD0525393E6UnknownZinc finger like proteinCD0525403E10 Mannose binding lectin (rhea) CD0525413E12 Clathrin heavy chain (Chc) CD0525423F4Identity to M. musculus adultmale testis cDNA3F10 Identity to D. melanogaster P-element somatic inhibitor (Psi)3G114A84A104A12Identity to D. melanogaster BM-40 extracellular basementmembrane proteinIdentity to D. melanogasterregulator of gene transcription(Chi)Identity to D. melanogasterhomeoprotein phtfAmino acid transporter system A(ATA2)CD052543CD052544CD052545CD052546CD052547CD0525484B2 Calmodulin CD0525494B7 Alpha-tubulin CD0525504C94C114D64D74E64D8Identity to D. melanogastertransducin (G protein)-likeenhancer of split 3, homolog ofE(spl)Intracellular receptor of activatedprotein kinase C1 (Rack1)Identity to D. melanogasterCG10395 cDNAIdentity to D. melanogasterLD23959 cDNAIdentity to D. melanogasterCG13597 cDNAIdentity to H. sapienshypothetical protein FLJ10342CD052551CD052552CD052553CD052554CD052555CD0525564E1 Pre-mRNA splicing factor CD0525574E3Receptor signaling proteinserine/threonine kinaseCD0525584F8 Nucleotidase CD0525594F1 Block of proliferation 1 (Bop1) CD0525604G14G2Identity to H. sapienshypothetical protein MGC2404LRP/alpha-2-macroglobulinreceptorCD052561CD0525624G5 Disulfide isomerase CD0525634G8 Fumarate hydratase CD0525644G10Rab3D (member of the Rassuperfamily of small GTPases)CD0525654G11 Chloride channel CD0525664H4 Solute carrier protein CD0525671B7 Mitochondrion NR1B8 Mitochondrion NR2E9 Mitochondrion NR2G11 Mitochondrion NR3C6 Mitochondrion NR3G4 Mitochondrion NR4A2 Mitochondrion NR4E9 Mitochondrion NR2A6 Mitochondrion NR4G7NAD-dependent malatedehydrogenaseNR3D4 Cytochrome c oxidase I (COI) NR1C2 Cytochrome c oxidase II (COII) NR4D2Cytochrome c oxidase III(COIII)NR1G4 Cytochrome b (cytb) NR2G9 16S ribosomal RNA NR1F42C73B64G124H2UnknownIdentity to I. scapularis cloneAC22 microsatellite sequence(AF331735)UnknownContains microsatellite sequenceUnknownContains a microsatellitesequenceUnknownContains microsatellite sequenceUnknownContains microsatellite sequenceCD052568CD052569CD052570CD052571CD052572NR, Not reported to the EST database for being identicalto mitochondrial sequencesThe majority of clones with gene ontology assignedcorresponded to non-nuclear gene products involved incell growth and maintenance, including genes with ligandbinding, carrier or enzymatic activities (Table 2).Seventeen clones contained sequences corresponding totick mitochondrion and were not submitted to the ESTdatabase. Other clones such as 2A9 and 1D6, althoughprobably coding for mitochondrial proteins, were analyzedand submitted to the EST database. Interestingly, 11clones encoded gene products localized in the cell nucleus(Table 2).The average G + C content of the EST dataset(47,503 bases excluding the poly-A tails with 171 (0.4%)undetermined nucleotide positions) was 54%, but somesequences, such as clone 2A9 which probably codes for amitochondrial protein, had only a 25% G + C content.46

Gene Therapy and Molecular Biology Vol 7, page 47Some short ESTs in clones 1D1 and 2D5 contained a longstretch of T.Vaccination of mice with plasmid DNA followed bytick infestation resulted in some cDNA clones that had aninhibitory effect on tick infestations, while others appearedto promote tick feeding (Table 3). The cDNAs inhibitingtick infestation were identical to nucleotidase, heat shockproteins, beta-adaptin, chloride channel, ribosomalproteins and proteins with unknown function. cDNAclones identical to beta-amyloid precursor, block ofproliferation, mannose-binding lectin, RNA polymeraseIII, ATPases and a protein of unknown function enhancedtick feeding.Further characterization of cDNAs that affectedlarval development (Table 3) was conducted for allclones except for 4D8, 4F8, 4D6 and 4E6, which producedhigh inhibition of tick infestation and are currently beingstudied separately as recombinant proteins expressed inEscherichia coli.The pool of heat shock proteins hsp70 and hsp60cDNAs conferred partial protection against tickinfestations and did not affect molting (Table 3). ThecDNA sequences for hsp70 and hsp60 in clones 1C10 and3F6, respectively, were partial and contained the regioncoding for the C-terminal of the protein, and were highlyidentical to other hsp70 sequences (data not shown).The sequence of hsp70 contained a 3’ untranslatedregion (UTR) of 299 bp before the poly-A tail. The clone3E1 contained a cDNA identical to the beta-adaptin thatproduced a 27% inhibition of tick infestation and a 5%inhibition of molting to the nymphal stage aftervaccination and tick challenge (Table 3). The completesequence was determined for the clone 3E1 (Figure 1A),and contained an insert of 1,942 bp encoding for apredicted protein of 191 amino acids. The sequence of thisprotein was shorter than that for other beta-adaptins(Figure 1B), suggesting that it could encode for a betaadaptinappendage or it may be a partial cDNA sequencebecause of a long 3’ UTR of 1,334 bp located before thepoly-A tail.Table 2. I. scapularis gene ontology assignments.Category Number of clones % of 87 clones with geneontology assignmentsCellular component% of 101 clones with identity toreported sequencesCell 32 36.78 31.88Mitochondria 17 15.54 16.83Cell membrane 14 16.09 13.86Nucleus 11 12.64 10.89Extracellular 2 2.30 1.98Unlocalized 2 2.30 1.98Unknown 9 10.34 8.91Biological processCell growth or maintenance 61 70.11 60.40Physiological process 8 9.20 7.92Developmental process 5 5.75 4.95Cell communication 2 2.30 1.98Unknown 11 12.64 10.89Molecular functionLigand binding or carrier 30 34.48 29.70Enzyme 29 33.33 28.71Transporter 9 10.34 8.91Chaperone 2 2.30 1.98Structural molecule 7 8.05 6.93Unknown 10 11.49 9.90Gene ontology assignments were made according to Ashburner et al. (2000) for non-redundant EST sequence data with the help ofGoFish v.1.0 (Berriz et al, 2003). The number of clone sequences falling into each category are listed and then calculated as a percent ofclones for which gene ontology was assigned and the total number of clones for which identity was found to previously publishedsequences.47

Almazán et al: Expressed sequence tags in Ixodes scapularisTable 3. Summary of results of DNA vaccination and challenge with I. scapularis larvae in the mouse model of tickinfestations.EST cDNA clone Predicted protein Inhibition of tick infestation4D8Identity to H. sapiens hypothetical proteinFLJ10342 with unknown functionI (%)Inhibition of moltingM (%)40 a 7 a4F8 Nucleotidase 50 a 17 a1C10 b Hsp70 17 a 0 a3F6 b4D64E6Hsp60Identity to D. melanogaster CG10395 cDNAwith unknown functionIdentity to D. melanogaster CG13597 cDNAwith unknown function61 1120 ND3E1 Beta-adaptin 27 54G11 Chloride channel 38 3017 clones b Ribosomal proteins 15 a 0 a2C12 Beta-amyloid precursor protein (APP) -8 c ND4F1 Block of proliferation Bop1 -39 c ND3E10 Mannose binding lectin -48 a, c ND3C12 b RNA polymerase III -104 a, c ND2F9 bIdentity to A. variegatum AvGI TC255 &Homo sapiens hypothetical proteinFLJ12475 with unknown functions1A9, 1B2, 4A4 b ATPase -57 a, c NDa Data reported by Almazán et al. (2003). For all other experiments, mice were immunized with cDNA-containing expression plasmidDNA as described above. I and M were calculated as described in Materials and Methods section. ND, not determined.b Pooled together for vaccination experiments by ELI (Almazán et al, 2003) (1C10 and 3F6, cDNA pool “Heat shock”; 3C12 and 2F9,cDNA pool “Secreted protein”; ribosomal clones, cDNA pool “Ribosomal”; 1A9, 1B2 and 4A4, cDNA pool “ATPase”).c Resulted in enhanced tick feeding after mouse vaccination and tick challenge.The cDNA in clone 4G11 was identical to a chloridechannel but it contained only a partial sequence (Figure2A). This sequence protected against tick infestations andinhibited larval molting (Table 3). Chloride channels havebeen found in living organisms from bacteria to mammals,with some amino acid positions being conserved in allsequences (Figure 2A). As expected, phylogeneticanalysis of chloride channel sequences demonstrated thatthe I. scapularis sequence comprised a sister group toother insect sequences that have been reported (Figure2B).Vaccination with ribosomal sequences had someinhibitory effect on tick infestations but did not affectmolting (Table 3). The pool of ribosomal cDNAs includedEST sequences coding for cellular and mitochondrialribosomal proteins and translation factors (Table 4), andthese genes are highly conserved across species. However,proteins encoded by I. scapularis ESTs were 43% to 95%identical to arachnida or insect sequences and 36% to 85%identical to mouse sequences (Table 4). The cDNA inclone 2C12 that was found to be identical to the betaamyloidprecursor protein (APP) contained a fragmentencoding for the C-terminal of the protein (Figure 3),suggesting that it contains a partial cDNA with a long(1,400 bp) 3’ UTR. Nonetheless, the C-terminal sequenceof the I. scapularis APP contained regions of amino acidsidentical to fly and mosquito sequences (Figure 3).Vaccination with this cDNA resulted in 8% enhancementof larval feeding (Table 3). Vaccination with cDNA clone4F1 resulted in enhanced larval feeding (Table 3). Thecomplete sequence of clone 4F1 cDNA was determinedand contained an insert of 2,475 bp with 30 bp and 66 bpof 5’ and 3’ UTR, respectively and a poly-A tail of 114bases.An open reading frame of 2,265 bp encoded for aprotein of 754 amino acids that was identical to mouseblock of proliferation (Bop 1) (Figure 4).Similar proteins have been identified in otherorganisms including Drosophila melanogaster, Anophelesgambiae and humans (Figure 4), suggesting that thisprotein has been highly conserved during evolution. Theclone 3E10 had a pronounced stimulatory effect on larvalfeeding (Table 3). This clone was completely sequencedand contained an insert of 1,848 bp with 50 bp and 279 bpof 5’ and 3’ UTR, respectively and a short poly-A tail of24 bases. An open reading frame of 1,494 bp encoded fora protein of 497 amino acids that was identical tomannose-binding lectins found in many eukaryotes(Figure 5). A similar sequence was described in A.variegatum ESTs, which clustered together with the I.scapularis sequence (Figure 5).48

Gene Therapy and Molecular Biology Vol 7, page 49ABcgATGCAGGCGATGACGGGCTTTGCGGTGCAGTTCAACAAAAACAGTTTCGGGCTGACTCCAGCTCAGCCGCTGCAGTTGCAGATTCCCCTGCAGCCCAACTTCCCAGCTGATGCGAGCTTGCAGCTGGGAACCAACGGTCCCGTGCAGAAGATGGACCCCCTCACCAACCTTCAGGTGGCCATCAAGAACAATGTGGACGTGTTCTACTTCAGCTGCCTGGTGCCCATGCACGTGCTGAGCACGGAGGACGGCCTGATGGACAAGCGGGTGTTCCTGGCCACCTGGAAAGACATCCCCGCCCAAAACGAGGTCCAGTACACCCTCGACAACGTCAACCTCACTGCAGACCAAGTTTCCCAGAAGCTGCAGAACAACAACATTTTCACGATAGCCAAGAGGAACGTGGACGGCCAGGACATGCTGTACCAGTCCCTGAAGCTCACCAACGGCATTTGGGTGTTGGCGGAGCTCAAGATACAGCCCGGCAATCCAAGGATCACGTTGTCTTTGAAGACAAGAGCACCTGAAGTGGCAGCAGGTGTACAACAAACTTACGAACTCATTCTACACAGCTGAggctgctgtgaatgaaactcttctcccacccccttcttttgatggcagtcaatgtctcgtttcattttcttgttttcttttgcggcgtgctacggaacaaggtcctacattcccaagttatatggtgttgtcgcgtagggggcagagtgccgctgagcccgcgacagccttgtttctgaggagagccgaacgcaccacttcgaaaaagaaaaagtgaaaacggaaaaatgaaaaattttccagttgcttcaaattaacattcctcgtagtcagtctgtggccgttgagtttggtgtaaagaagaaaaaggtgtctcttttagtgaaaatggttgctttttattggtatcccctatcacaccgagcacgaacataagaaatcctgacaaggattctcctttagttgtattatggtggctggagcacacgaggcacctgttgccaattcgacccagcaaatgcccaattctcaagatttgagttcattgaggttgttttgctcctccccccccaccccccaactttgtcgttggattgtctaacagtgtaaatgggcgacgactcgttattctttttttcttcattctttctttttgttgtcacgcgccccgggggacgcgacacaacttatgtgcataattgattttcacaggctgcgacgcagtctgtaaaagaaggggaagtgaaactctgctccgccgctgctagtgtcatcacgggacgaccatcgcgttttctctgactatttaaacaaaactgcatagcttagggggcagtctgtgcaaagtggaacaaccaaactgagccctgccctttcggtgtgtgtacaagcatctctgtgtaacatgaactactttacatgaactacattgcatgaacgggagaagtttagttgtttttttgttttttttttcaggtgactatgtcaacagattagaaccattttttggaacggctggaaagataaccgctcattttgtttctactaaaagactacgaaaagtgttgactttttgcatcggtttggcaacgtttgtttggcatgcatgtagttgagcgtaatggtatcacccctcgtaaacaataacagtgcaatggagcagtactgtagtgtccattaaagagcgagagtttggttaaaggttgttaattgaggtccgtgttatcctttgagtaggagagcggcactttttgcaaatagcgctgctgggggcgtcatatctgccctccaaaacatgcacattttaagtgtgaattgttgcggcggcttgtacaagtatgtgtgttatgtgtagaaaaagaactcttaattaaaatatttgtggccaaaacgtcaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaM. musculus (747) LQHMTDFAIQFNKNSFGVIPSTPLAIHTPLMPNQSIDVSLPLNTLGPVMKD. melanogaster (731) MQPMTNFAIQLNKNSFGLVPASPMQ-AAPLPPNQSIEVSMALGTNGPIQRH. sapiens (68) LQHMTDFAIQFNKNSFGVIPSTPLAIHTPLMPNQSIDVSLPLNTLGPVMKI. scapularis (1) MQAMTGFAVQFNKNSFGLTPAQPLQLQIPLQPNFPADASLQLGTNGPVQKConsensus (748) LQHMTDFAIQFNKNSFGLIPATPLQIHTPLMPNQSIDVSLPLNTNGPVQKM. musculus (797) MEPLNNLQVAVKNNIDVFYFSCLIPLNVLFVEDGKMERQVFLATWKDIPND. melanogaster (780) MEPLNNLQVAVKNNIDIFYFACLVHGNVLFAEDGQLDKRVFLNTWKEIPAH. sapiens (118) MEPLNNLQVAVKNNIDVFYFSCLIPLNVLFVEDGKMERQVFLATWKDIPNI. scapularis (51) MDPLTNLQVAIKNNVDVFYFSCLVPMHVLSTEDGLMDKRVFLATWKDIPAConsensus (798) MEPLNNLQVAVKNNIDVFYFSCLIPLNVLFVEDGKMDKRVFLATWKDIPNM. musculus (847) ENELQFQIKECHLNADTVSSKLQNNNVYTIAKRNVEGQDMLYQSLKLTNGD. melanogaster (830) ANELQYTLSGVIGTTDGIASKMTTNNIFTIAKRNVEGQDMLYQSLKLTNNH. sapiens (168) ENELQFQIKECHLNADTVSSKLQNNNVYTIAKRNVEGQDMLYQSLKLTNGI. scapularis (101) QNEVQYTLDNVNLTADQVSQKLQNNNIFTIAKRNVDGQDMLYQSLKLTNGConsensus (848) ENELQFTIKEVHLTADTVSSKLQNNNIFTIAKRNVEGQDMLYQSLKLTNGM. musculus (897) IWILAELRIQPGNPNYTLSLKCRAPEVSQYIYQVYDSILKN-D. melanogaster (880) IWVLLELKLQPGNPEATLSLKSRSVEVANIIFAAYEAIIRSPH. sapiens (218) IWILAELRIQPGNPNYTLSLKCRAPEVSQYIYQVYDSILKN-I. scapularis (151) IWVLAELKIQPGNPRITLSLKTRAPEVAAGVQQTYELILHS-Consensus (898) IWILAELKIQPGNPNYTLSLKCRAPEVAQYIYQVYDSILKSFigure 1. Analysis of clone 3E1 identical to beta-adaptin. (A) Nucleotide sequence of complete cDNA. Non-coding sequence is shownin lower case letters and coding sequence is shown in capital letters with translation initiation and termination codons in bold letters. (B)Alignment of M. musculus (GenBank accession number XP_109938), D. melanogaster (CAA53509) and Homo sapiens (AAA35583)protein sequences and the translation product of clone 3E1 identified as I. scapularis beta-adaptin appendage (AY296113). Proteinsequences are shown in the single letter amino acid code. Identical amino acids are shown in red and amino acids conserved in 3 of 4sequences are shown in blue.AE. coli (4) DTPSLETPQAARLRRRQLIRQLLERDKTPLAILFMAAVVGTLVGLAA-VAO. mossambicus (98) DLKEGVCLSALWFNH--------EQ----------CCWTSNETTFAERDKX. laevis (146) DLKEGICLPWFWFNH--------EQ----------CCWQSNNVTFEDRNNI. scapularis (1) DLKEGICPQAFWLNK--------EQ----------CCWASNDTFFKG-DDC. elegans (141) DLKTGVCADRFWLDH--------EH----------CCWSSNDTFYKD-DDD. melanogaster (223) DLKHGICPPAFWFNR--------EQ----------CCYPAKQSVFEE-GNL. major (114) AFRSGICANFFWLGR-------------------------N-MCCVDCREA. gambiae (272) DLKFGICPQAFWLNR--------EQ----------CCWSSNETSFDS-GN49

Almazán et al: Expressed sequence tags in Ixodes scapularisM. musculus (155) DLKEGICLSALWYNH--------EQ----------CCWGSNETTFEERDKS. tuberosum (108) GFKLLLTSNLMLDGK-----------------------------------S. cerevisiae (102) NWKTGHCQRNWLLNKS-------------------FCCNGVVNEVTSTSNConsensus (272) DLK GIC AFWLNR EQ CCW SN T F DE. coli (53) FDKGVAWLQNQRMGALVHTADNYPLLLTVAFLCSAVLAMFGYFLVRKYAPO. mossambicus (130) CPQWKSWAELILGQ--AEGPGSYIMNYFMYIYWALSFAFLAVCLVKVFAPX. laevis (178) CPEWRSWSQLVLGR--SEGAFPYILNYFMYVMWALLFSLLAVLLVRNFAPI. scapularis (32) CKQWYRWPEMFDSGMDKDGAGFYLLSYLLYVMWSVLFATLAVMLVRTFAPC. elegans (172) CKAWTKWPWMLNYYN-SSSFLFLFLEWIFYIGWAVAMSTLAVLFVKIFAPD. melanogaster (254) CSTWKTWPEIFGLD--RNGTGPYIVAYIWYVLWALLFASLSASLVRMFAPL. major (138) CGEYYSWGEFFLGR---DNHVVAFVDFVMYVSFSTMAAVTAAYLCKTYAPA. gambiae (303) CSQWYAWSEIFTSS--REGFGAYVISYFFYIMWAMLFALLAASLVRMFAPM. musculus (187) CPQWKTWAELIIGQ--AEGPGSYIMNYIMYIFWALSFAFLAVSLVKVFAPS. tuberosum (123) ----------------------YFQAFAAFAGCNVFFATCAAALCAFIAPS. cerevisiae (133) LLLKRQEFECEAQG-LWIAWKGHVSPFIIFMLLSVLFALISTLLVKYVAPConsensus (322) C W W EL EG YIL YIMYILWALLFA LA LVK FAPE. coli (103) EAGGSGIPEIEGALE---DQRPVRWWRVLPVKFFGGLGTLGGGMVLGREGO. mossambicus (178) YACGSGIPEIKTILSGF-IIRGYLGKWTLMIKTITLVLAVASGLSLGKEGX. laevis (226) YACGSGIPEIKTILSGF-IIRGYLGKWTLIIKTMTLVLAVSSGLSLGKEGI. scapularis (82) YACGSGIPEIKTILSGF-IIRGYLGKWTLTIKSVCLVLAVGAGLSLGKEGC. elegans (221) YACGSGIPEIKCILSGF-VIRGYLGKWTFIIKSVGLILSSASGLSLGKEGD. melanogaster (302) YACGSGIPEIKTILSGF-IIRGYLGKWTLLIKSVGLMLSVSAGLTLGKEGL. major (185) YASGGGIAEVKTIVSGH-HVKRYLGGWTLITKVVGMCFSTGSGLTVGKEGA. gambiae (351) YACGSGIPEIKTILSGF-IIRSYLGKWTLIIKSVGIMLSVSAGLSLGKEGM. musculus (235) YACGSGIPEIKTILSGF-IIRGYLGKWTLMIKTITLVLAVASGLSLGKEGS. tuberosum (151) AAAGSGIPEVKAYLNG-IDAHSILAPSTLLVKIFGSILGVSAGFVVGKEGS. cerevisiae (182) MATGSGISEIKVWVSGFEYNKEFLGLLTLVIKSVALPLAISSGLSVGKEGConsensus (372) YACGSGIPEIKTILSGF IIRGYLGKWTLIIKSVGLVLAVSSGLSLGKEGE. coli (150) PTVQIGGNIGRMV----------LDIFRLKG--DEARHTLLATGAAAGLAO. mossambicus (227) PLVHVACCCGNIF----------SYLFPKYSKNEAKKREVLSAASAAGVSX. laevis (275) PLIHVACCCGNIL----------CHLFTKYRKNEAKRREVLSAAAAAGVSI. scapularis (131) PLVHVACCIGNIF----------SYLFPKYGKNEAKKREILSAAAAAGVSC. elegans (270) PMVHLACCIGNIF----------SYLFPKYGLNEAKKREILSASAAAGVSD. melanogaster (351) PMVHIASCIGNIF----------SHVFPKYGRNEAKKREILSAAAAAGVSL. major (234) PFVHIGACVGGII----------SGALPSYQQ-EAKERELITAGAGGGMAA. gambiae (400) PMVHIASCIGNIL----------SYLFPKYGRNEAKKREILSAAAAAGVSM. musculus (284) PLVHVACCCGNIF----------SYLFPKYSTNEAKKREVLSAASAAGVSS. tuberosum (200) PMVHTGACIANLLGQGGSRKYHLTWKWLKYFKNDRDRRDLITCGAAAGVAS. cerevisiae (232) PSVHYATCCGYLL----------TKWLLRDTLTYSTQYEYLTAASGAGVAConsensus (422) PLVHIA CIGNILSYLFPKY KNEAKKREILSAAAAAGVSE. coli (188) AAFNAPLAGILFIIEEMRPQ--FRYTLISIKAVFIGVIMSTIMYRIFNHEO. mossambicus (267) VAFGAPIGGVLFSLEEVSYY--FPLKTLWRSFFAALVAAFVLRSINPFGNX. laevis (315) VAFGAPIGGVLFSLEEVSYY--FPLKTLWRSFFAALVAAFTLRSINPFGNI. scapularis (171) VAFGAPIGGVLFSLEEVSYY--XPLKTLWRSFFCALVAASVLRSINPFGNC. elegans (310) VAFGAPIGGVLFSLEEASYY--FPLKTMWRSFFCALVAGIILRFVNPFGSD. melanogaster (391) VAFGAPIGGVLFSLEEVSYY--FPLKTLWRSFFCALIAAFVLRSLTPFGNL. major (273) VAFGAPVGGVIFALEDVSTS--YNFKALMAALICGVTAVLLQSRVDLWHTA. gambiae (440) VAFGAPIGGVLFSLEEVSYY--FPLKTLWRSFFCALIAAFILRSINPFGNM. musculus (324) VAFGAPIGGVLFSLEEVSYY--FPLKTLWRSFFAALVAAFVLRSINPFGNS. tuberosum (250) AAFRAPVGGVLFALEEIASW--WRSALLWRTFFTTAIVAMVLRSLIQFCRS. cerevisiae (272) VAFGAPIGGVLFGLEEIASANRFNSSTLWKSYYVALVAITTLKYIDPFRNConsensus (472) VAFGAPIGGVLFSLEEVSYY FPLKTLWRSFF ALVAA VLRSINPFGNE. coli (236) VA----------LIDVGKLSDAPLO. mossambicus (315) SR----------LVLFYVEYHTPWX. laevis (363) SR----------LVLFYVEFHAPWI. scapularis (219) DH----------LVMFYVEYDFPWC. elegans (358) NQ----------TSLFHVDYMMKWD. melanogaster (439) EH----------SVLFFVEYNKPWL. major (321) GR----------IVQFSVNYQHNWA. gambiae (488) EH----------SVLFYVEYNKPWM. musculus (372) SR----------LVLFYVEYHTPWS. tuberosum (298) GGNCGLFGQGGLIMFDVNSGVSNYS. cerevisiae (322) GR----------VILFNVTYDRDWConsensus (522)LVLFYVEY PW50

Gene Therapy and Molecular Biology Vol 7, page 51BFigure 2. Analysis of clone 4G11 identical to chloride channel. (A) Alignment of M. musculus (XP_134186), D. melanogaster(AAM76180), Solanum tuberosum (T07608), Oreochromis mossambicus (AAD56388), A. gambiae (EAA11899), C. elegans(NP_495940), Leishmania major (strain Friedlin) (T02805), Saccharomyces cerevisiae (P37020), Escherichia coli K12 (AAC73266),and Xenopus laevis (CAA71071) protein sequences and the translation product of clone 4G11 identified as a fragment of I. scapularischloride channel (AY296114). Protein sequences are shown in the single letter amino acid code. Identical amino acids are shown in redand amino acids conserved in 6-10 of 11 sequences are shown in blue. (B) Phylogenetic tree constructed from analysis of chloridechannel protein sequences based on a sequence distance method utilizing the Neighbor Joining algorithm of Saitou and Nei (1987).D. melanogaster PHAQGFIEVDQNVTTHHPIVREEKIVPNMQINGYENPTYKYFEI. scapularis PQAQGFVQVDQGALPASPEER---HLASMQVNGYENPTYKYFEA. gambiae PHAQGFVEVDQAVGAPVTPEE--RHVANMQINGYENPTYKYFEConsensus PHAQGFVEVDQ V P ER HVANMQINGYENPTYKYFEFigure 3. Analysis of clone 2C12 identical to beta-amyloid precursor protein. Alignment of D. melanogaster (AF181628) and A.gambiae (EAA07868) protein sequences and the translation product of clone 2C12 identified as I. scapularis beta-amyloid peptide (ß-AP) (AY296115). Protein sequences are shown in the single letter amino acid code. Identical amino acids are shown in red and aminoacids conserved in 2 of 3 sequences are shown in blue.Table 4. Characterization of I. scapularis ESTs encoding for ribosomal proteinsEST clone Predicted protein Identical aminoacids4F71A2Elongation factor 1-alpha 95%85%SpeciesNeacarus texanusMus musculusGenBank accessionnumberAAK12660NP_0319321A10 Elongation factor-2 88%80%Mastigoproctus giganteusMus musculusAAK12348BAC262031C11 eIF-5A 65%59%Drosophila melanogasterMus musculusAAM68297XP_2033361F62C3RpS4 79%75%Spodoptera frugiperdaMus musculusAAL26580AAH091002B8 RpS11 92%80%Dermacentor variabilisMus musculusAAO92287XP_1334772F8 Laminin receptor 1(RpSA)66%73%Anopheles gambiaeMus musculusEAA00413NP_0351592F10 RpL3 70%68%Spodoptera frugiperdaMus musculusAAL62468AAH096553A10 RpL7A 55%60%Drosophila melanogasterMus musculusNP_511063A302413D10 Ribophorin I 57%50%Drosophila melanogasterMus musculusAAN71150BAC266793G9 RpS8 70%71%Spodoptera frugiperdaMus musculusAAL62472XP_1349043G10 RpL27A 42% Spodoptera frugiperda AAK9215851

Almazán et al: Expressed sequence tags in Ixodes scapularis4D11 Proteasome subunit 60%4D12Proteasome/Signalosomesubunit36% Mus musculus XP_13711855%43%56%4E7 Proteasome subunit 84%85%Drosophila melanogasterMus musculusAnopheles gambiaeMus musculusAnopheles gambiaeMus musculusNP_524115NP_035315EAA11895AAC33900EAA10351NP_036096The sequences of I. scapularis ESTs identical to ribosomal proteins pooled for DNA vaccination as described in Almazán et al. (2003),were compared to all non-redundant sequences in GenBank DNA and protein databases (1,419,727 sequences total; Apr-09-2003) usingBLASTX 2.2.6 (Altschul et al, 1997). The percent of identical amino acids to arachnida or insect and mouse sequences are showntogether with their corresponding GenBank accession number. The GenBank accession numbers for I. scapualris sequences are shownon Table 1.1 50M. musculus (1) ------------------------MAGACGKPHMSPASLPGKRRLEPDQED. melanogaster (1) MTKKLALKRRGKDSEPTNEVVASSEASENEEEEEDLLQAVKDPGEDSTDDH. sapiens (1) ----------------------------SVRPEKRRSEPELEPEPEPEPPA. gambiae (1) ---------------------QENLLGSIENEGEDSSDSDGEYATDDDEDI. scapularis (1) ----------------------MGPKTLSKQPAKASSSTSKRTAGPTISKConsensus (1) P S E A D D D51 100M. musculus (27) LQIQEPPLLSD-PDSSLSDSEESVFSGLEDSGSDSSEEDTEGVA----GSD. melanogaster (51) EGIDQEYHSDSSEELQFESDEEGNYLGRKQSSSAEEDEESSDEEDN---EH. sapiens (23) LLCTSPLSHSTGSDSGVSDSEESVFSGLEDSGSDSSEDDDEGDEEGEDGAA. gambiae (30) DVLSFESLNSDGEE---EDEEEDAGTTLEEVEREAEEDDDEEDAERKQREI. scapularis (29) QTEDSDDEGSSSAYSDLEDSEGADSSDSNDLSDTEASEDDYDDSQDEENTConsensus (51) I E SS DS LEDSEES FSGLEDS SDSSEEDDEDDAE101 150M. musculus (72) SGDEDNHRAEETSEELAQAAPLCSRTEE--------------AGALAQDED. melanogaster (98) EEESTDGEEVEDEEKDSKSKQTDDKPSGSGAASKKALTAELPKRDSSKPEH. sapiens (73) LDDEGHSGIKKTTEEQVQASTPCPRTEM--------------ASARIGDEA. gambiae (77) EQFESDDEPLPDDLKLGRIEDVLGTGEKKTRGLGVFPPVPKRKGKAAQDEI. scapularis (79) KITLTGVEGKDLELRGKDQEAPVESGKRSAWHRQQEDAKEDRRTQVVEDEConsensus (101) DET E E EEK A R E K A DE151 200M. musculus (108) YEE-DSSDEEDIRNTVGNVPLAWYDEFPHVGYDLDGKRIYKPLRTRDELDD. melanogaster (148) YQDSDTSDEEDIRNTVGNIPMHWYDEYKHIGYDWDAKKIIKPPQG-DQIDH. sapiens (109) YAE-DSSDEEDIRNTVGNVPLEWYDDFPHVGYDLDGRRIYKPLRTRDELDA. gambiae (127) YAAGDTSDEEDIRNTVGNIPMHWYDEYKHVGYDWDAKKIIKAKKG-DAIDI. scapularis (129) YAF-DSSDEEDVRNTVGNIPLEWYEHYPHIGYDLEGKPILKPPRV-SDLDConsensus (151) YAE DSSDEEDIRNTVGNIPL WYDEYPHVGYDLDGKKIIKP R DELD201 250M. musculus (157) QFLDKMDDPDFWRTVQDKMTGRDLRLTDEQVALVHRLQRGQFGDSGFNPYD. melanogaster (197) EFLRKIEDPDFWRTVKDPLTGQDVRLTDEDIALIKRIVSGRIPNKDHEEYH. sapiens (158) QFLDKMDDPDYWRTVQDPMTGRDLRLTDEQVALVRRLQSGQFGDVGFNPYA. gambiae (176) DFLQRMEDPNFWRTVTDPQTGQKVVLSDEDIGLIKRIMSGRNPDAEYDDYI. scapularis (177) DFLRKMDDPNYWRTVKDKSTGQDVVLTDEDVDLIQRLQKGQFPSSTTDPYConsensus (201) DFL KMDDPDFWRTV DPMTGQDVRLTDEDVALIKRLQSGQFPDS FDPY251 300M. musculus (207) EPAVDFFSGDIMIHPVTNRPADKRSFIPSLVEKEKVSRMVHAIKMGWIKPD. melanogaster (247) EPWIEWFTSEVEKMPIKNVPDHKRSFLPSVSEKKRVSRMVHALKMGWMKTH. sapiens (208) EPAVDFFSGDVMIHPVTNRPADKRSFIPSLVEKEKVSRMVHAIKMGWIQPA. gambiae (226) EPFIEWFTSEVEKMPIRNIPESKRSFLPSKAEKHKIGRYVHALKMGWMKTI. scapularis (227) EPFEDIFSHETMIHPVTRHPPQKRSFVPSRIEKAMVSKMVHAIKMGWIKPConsensus (251) EPFIDFFS EVMIHPVTN P KRSFIPSLVEK KVSRMVHAIKMGWIKP301 350M. musculus (257) RRPHD------PTPSFYDLWAQEDPNAVLG-RHKMHVPAPKLALPGHAESD. melanogaster (297) TEEVEREKQAKRGPKFYMLWETDTSREHMR-RIHDPVSAPKRDLPGHAESH. sapiens (258) RRPRD------PTPSFYDLWAQEDPNAVLG-RHKMHVPAPKLALPGHAESA. gambiae (276) MAEKRRLEAIRRQPKFYMLWTTDHGKEEMR-RIHDHVAAPKRMLPGHAESI. scapularis (277) RVKKH------DPERFSLLWDKDDSTAGSNERMQRHIPAPKMKLPGHEESConsensus (301) R KD PKFYMLW DD A L RI HVPAPKL LPGHAES351 400M. musculus (300) YNPPPEYLPTEEERSAW--MQQEPVERKLNFLPQKFPSLRTVPAYSRFIQD. melanogaster (346) YNPPPEYLFDAKETKEWLKLKDEPHKRKLHFMPQKFKSLREVPAYSRYLRH. sapiens (301) YNPPPEYLLSEEERLAW--EQQEPGERKLSFLPRKFPSLRAVPAYGRFIQ52

Gene Therapy and Molecular Biology Vol 7, page 53A. gambiae (325) YNPPPEYLFDEKELEEWNKLANQPWKRKRAYVPQKYNSLREVPGYTRYVKI. scapularis (321) YNPPAEYLFTEEEEAKWR--EQEPEERRINFLPAKYPCLRAVPAYERFIEConsensus (351) YNPPPEYLFTEEE W L QEP ERKL FLPQKFPSLR VPAYSRFI401 450M. musculus (348) ERFERCLDLYLCPRQRKMRVNVDPEDLIPKLPRPRDLQPFPVCQALVYRGD. melanogaster (396) ERFLRCLDLYLCPRAKRVKLNIDAEYLIPKLPSPRDLQPFPTVESMVYRGH. sapiens (349) ERFERCLDLYLCPRQRKMRVNVDPEDLIPKLPRPRDLQPFPTCQALVYRGA. gambiae (375) ERFLRCLDLYLAPRMRRSRVAVGAEYLIPKLPSPRDLQPFPTLQNLIYTGI. scapularis (369) ERFERCLDLYLCPRQRKMRVNVDAEDLIPQLPKPKDLQPFPSIQSIVYEGConsensus (401) ERFERCLDLYLCPRQRKMRVNVDAEDLIPKLPRPRDLQPFPTIQALVYRG451 500M. musculus (398) HSDLVRCLSVSPGGQWLASGSDDGTLKLWEVATARCMKTVHVGGVVRSIAD. melanogaster (446) HTDLVRSVSVEPKGEYLVSGSDDKTVKIWEIATGRCIRTIETDEVVRCVAH. sapiens (399) HSDLVRCLSVSPGGQWLVSGSDDGSLRLWEVATARCVRTVPVGGVVKSVAA. gambiae (425) HTSLIRCISVEPKGEYIVTGSDDMTVKIWEISTARCIRTIPTGDIVRSVAI. scapularis (419) HTDCVLCLSLEPAGQFFASXSEDGTVRIWELLTGXCLKKFQFEAPVKSVAConsensus (451) HTDLVRCLSVEPGGQWLVSGSDDGTVKIWEIATARCIRTI GGVVRSVA501 550M. musculus (448) WNPNPTICLVAAAMDDAVLLLNPALGDRLLVGSTDQLLEAF----TPPEED. melanogaster (496) WCPNPKLSIIAVATGNRLLLVNPKVGDKVLVKKTDDLLAEAPSQDVIESEH. sapiens (449) WNPSPAVCLVAAAVEDSVLLLNPALGDRLVAGSTDQLLSAF----VPPEEA. gambiae (475) WCPNSKISLVAAASGKRVLLINPKVGDYMLVKKTDDLLTEAPRSDTVDSEI. scapularis (469) WCP--VVVPMKLCVDKTVSMLDAGVTDKLLPFTTGHRVVCPPRRVLGPGGConsensus (501) WCPNP I LVAAAVD VLLLNPAVGDKLLV STD LL P V P E551 600M. musculus (494) PALQPARWLEVSEEEHQRGLRLRICHSKPVTQVTWHGRGDYLAVVLSSQED. melanogaster (546) RIKTAVQWSNAEADEQEKGVRVVITHFKPIRQVTWHGRGDYLATVMPEGAH. sapiens (495) PPLQPARWLEASEEERQVGLRLRICHGKPVTQVTWHGRGDYLAVVLATQGA. gambiae (525) RIRSAVQWGEVTEEEKKLGVRIVITHFREVRQVTWHGRGDYFATVMPDGAI. scapularis (517) GSGVGADVGLLSRVPLPGGASAGRSPPR-CGAGDVALEGRLLCHCHGRGTConsensus (551) AA W EVSEEE GLRL ITH KPV QVTWHGRGDYLA VL GA601 650M. musculus (544) HTQVLLHQVSRRRSQSPFRRSHGQVQCVAFHPSRPFLLVASQRSIRIYHLD. melanogaster (596) NRSALIHQLSKRRSQIPFSKSKGLIQFVLFHPVKPCFFVATQHNIRIYDLH. sapiens (545) HTQVLIHQLSRRRSQSPFRRSHGQVQRVAFHPARPFLLVASQRSVRLYHLA. gambiae (575) YRSVMIHQLSKRRSQVPFSKSKGLIQCVLFHPIKPCLFVATQRHIRVYDLI. scapularis (566) GHRACPSVVHAAVRRLPFSKAKGGVSRVLFHPLRPFLLVACQRTVRVYHLConsensus (601) H VLIHQLSKRRSQIPFSKSKG VQ VLFHPIRPFLLVASQRSIRIYHL651 700M. musculus (594) LRQELTKKLMPNCKWVSSMAVHPAGDNIICGSYDSKLVWFDLDLSTKPYKD. melanogaster (646) VKQELVKKLLTNSKWISGMSIHPKGDNLLVSTYDKKMLWFDLDLSTKPYQH. sapiens (595) LRQELTKKLMPNCKWVSSLAVHPAGDNVICGSYDSKLVWFDLDLSTKPYRA. gambiae (625) VKQLMMKKLYPGCKWISSMAIHPKGDNLLIGTYEKRLMWFDLDLSTKPYQI. scapularis (616) LKQELAKRLTSNCKWISCMGRPPPGDNLLIGTYEKRLMWFDLDLSTKPYQConsensus (651) LKQEL KKLMPNCKWISSMAIHP GDNLLIGTYDKKLMWFDLDLSTKPYQ701 750M. musculus (644) VLRHHKKALRAVAFHPRYPLFASGSDDGSVIVCHGMVYNDLLQNPLLVPVD. melanogaster (696) TMRLHRNAVRSVAFHLRYPLFASGSDDQAVIVSHGMVYNDLLQNPLIVPLH. sapiens (645) MLRHHKKALRAVAFHPRYPLFASGSDDGSVIVCHGMVYNDLLQNPLLVPVA. gambiae (675) QLRIHNAAIRSVAFHPRYPLFASAGDDRSVIVSHGMVYNDLLQNPLIVPLI. scapularis (666) QLRIHNAAIRSVAFHPRYPLFASAGDDRSVIVSHGMVYNDLLQNPLIVPLConsensus (701) LRIHK AIRSVAFHPRYPLFASGSDD SVIVSHGMVYNDLLQNPLIVPL751 790M. musculus (694) KVLKGHTLTRDLGVLDVAFHPTQPWVFSSGADGTIRLFS-D. melanogaster (746) KKLQTHEKRDEFGVLDVNWHPVQPWVFSTGADSTIRLYT-H. sapiens (695) KVLKGHVLTRDLGVLDVIFHPTQPWVFSSGADGTVRLFT-A. gambiae (725) RRLKNHAVVNDFSVFDVVFHPTQPWVFSSGADNTVRLYT-I. scapularis (716) RRLKNHAISKGMGVLDCAFHPHQPWIVTAGADSTLRLFT-Consensus (751) KRLK H LTRDLGVLDV FHPTQPWVFSSGAD TIRLFTFigure 4. Analysis of clone 4F1 identical to block of proliferation (Bop1). (A) Alignment of M. musculus (AAH12693), D.melanogaster (NP_611270), A. gambiae (EAA04116), and H. sapiens (AAH07274) protein sequences and the translation product ofclone 4F1 identified as I. scapularis Bop (AY296116). Protein sequences are shown in the single letter amino acid code. Identical aminoacids are shown in red and amino acids conserved in 3-4 of 5 sequences are shown in blue.53

Almazán et al: Expressed sequence tags in Ixodes scapularisThe clone 3C12, together with clone 2F9, produced thegreatest enhancement of tick feeding after vaccination andtick challenge (Table 3). The clone 3C12 was completelysequenced and contained an insert of 447 bp with 5 bp and86 bp of 5’ and 3’ UTR, respectively and a short poly-Atail of 29 bases. An open reading frame of 327 bp encodedfor a protein of 108 amino acids that was identical to RNApolymerase III, and had a high degree of identity withhuman and insect sequences (Figure 6A). The EST inclone 2F9 was identical to human and A. variegatumsequences coding for proteins of unknown function(Figure 6B).Vaccination with the pool of ESTs identical toATPases resulted in a 57% increase in larval feeding(Table 3). This pool originally contained 6 sequences(Almazán et al, 2003) but only 3 were non-redundant(clones 1A9, 1B2 and 4A4). All sequences were identicalto vacuolar proton pump ATPases (EC 3.6.1.34). Thesequence of 1A9 was identical to D. melanogaster(TC112371) V-ATPase subunit D, 1B2 was identical to A.americanum (AAU03374) V-ATPase subunit C and 4A4was identical to D. melanogaster (TC112172) V-ATPasesubunit E.Six clones of the I. scapularis ESTs contained shorttandem repeat (STR) microsatellite sequences. STRs werefound in 5 clones (1F4, 2C7, 3B6, 4G12 and 4H2)containing sequences of unknown function and in oneclone (1A9) that was identical to the D. melanogaster V-ATPase subunit D (Table 1). Microsatellite sequencescontained perfect and imperfect STRs (Table 5). Clones1A9, 4G12 and 3B6 contained 9, 6 and 12 TA repeats,respectively. Clone 1F4 contained an imperfect repeat of15 GC/T and the clone 2C7 contained 9 GT repeats. Theclone 4G12 contained a second STR of 10 CA/GA/CTrepeats.IV. DiscussionThe feasibility of controlling tick infestations throughimmunization of hosts with tick antigens has beendemonstrated previously for Boophilus spp. (reviewed byWilladsen, 1997; Willadsen and Jongejan, 1999; de laFuente et al, 1999, 2000a). However, a limiting step fordevelopment of effective anti-tick vaccines is theidentification of tick protective antigens. In the past, tickprotective antigens were identified by (a) evaluatingproteins after host immunization and tick challenge thatwere derived from progressive fractionation of crude tickextracts, (b) immunomapping of tick antigens which elicitan antibody response in the infested host, and (c) testingtick proteins in vaccination experiments that wereconsidered to be important for the parasite function and/orsurvival.However, construction of cDNA libraries and ESTdatabases from different tick tissues, developmental stagesand from genes expressed in response to various stimuli(i.e., tick feeding or infection of cDNAs encoding for tickimmunosuppressants, anticoagulants and other proteinswith low antigenicity that may enhance tick feeding.Alternatively, they may encode for proteins homologousto host proteins associated with anti-tick or growthsuppression activity which neutralization results in a tickpro-feeding effect. The former could be the case forATPases. These proteins are highly conserved acrossspecies and, therefore, could elicit a poor immuneresponse. However, ATPases are expressed in tickembryos and salivary glands of unfed adults and adultfemales at all stages of feeding and some evidencessuggest that these proteins may participate in salivary fluidsecretion in A. americanum (McSwain et al, 1997).Therefore, although the mechanism is not known,DNA vaccination with ATPase-coding cDNAs couldproduce enhanced larval feeding. Although we presentlydo not have evidence to support the latter hypothesis,proteins of unknown function, such as the one encoded byclone 2F9 that is identical to host proteins of unidentifiedfunction, and Bop 1, a nonribosomal protein that is highlyconserved from yeast to human with a growth suppressorfunction that plays a key role in the formation of mature28S and 5.8S rRNAs and in the biogenesis of the 60Sribosomal subunit (Pestov et al, 1998; Strezoska et al,2000), are examples that may enhance tick feeding.Figure 5. Analysis of clone 3E10 identical to mannose-binding lectin. Phylogenetic tree constructed from analysis of C. elegans(NP_492548), A. gambiae (EAA11908), D. melanogaster (NP_524776), M. musculus (XP_128952), R. norvegicus (NP_446338),Cercopithecus aethiops (Q9TU32), H. sapiens (NP_005561), Polyandrocarpa misakiensis (BAB20045), X. laevis (AAC59755),Dictyostelium discoideum (AAL92589), A. variegatum (BM290898) and I. scapularis (AY296117) protein sequences based on asequence distance method utilizing the Neighbor Joining algorithm of Saitou and Nei (1987).54

Gene Therapy and Molecular Biology Vol 7, page 55A1 50D. melanogaster (1) MLFFCPSCGNILIIEEDTNCHRFTCNTCPYISKIRRKISTKTFPRLKEVDH. sapiens (1) MLLFCPGCGNGLIVEEGQRCHRFSCNTCPYVHNITRKVTNRKYPKLKEVDA. gambiae (1) MLMFCPTCGNLLLVEESTDSLRFSCNTCPYICKIRRTISSRIYPTLKEVDI. scapularis (1) MLLFCPTCANILIVEQGLECFRFACNTCPYVHNIKAKMSNRKYPRLKDVDConsensus(1) MLLFCPTCGNILIVEEGTDCHRFSCNTCPYIHNIRRKISNRKYPRLKEVD51 100D. melanogaster (51) HVLGGKAAWENVDSTDAECPTCGHKRAYFMQIQTRSADEPMTTFYKCCNHH. sapiens (51) DVLGGAAAWENVDSTAESCPKCEHPRAYFMQLQTRSADEPMTTFYKCCNAA. gambiae (51) HVMGGSAAWENVDSTDAVCPSCSHNRAYFMQMQTRSADEPMTTFYKCCNQI. scapularis (51) DVLGGAAAWENVDSTEEKCPKCGHERAYFMQIQTRSADEPMTTFYKCCNQConsensus(51) HVLGGAAAWENVDSTDE CPKCGH RAYFMQIQTRSADEPMTTFYKCCNQ101D. melanogaster (101) ECNHTWRDH. sapiens (101) QCGHRWRDA. gambiae (101) TCGHNWRDI. scapularis (101) LCGHQWRDConsensus (101) CGHNWRDBI. scapularis (78) MVDPEDEEVQLDEAMDEMAAYFRKEYTPKLLITTSDNPHRRTIKFCRELKA. variegatum (1) MVQADDEEVQLDEAMDEMAAYFRKEYIPKLLITTSDNPHTRTIRFCRELKH. sapiens (115) TVDPNDEEVAYDEATDEFASYFNKQTSPKILITTSDRPHGRTVRLCEQLSConsensus (115) MVDP DEEVQLDEAMDEMAAYFRKEY PKLLITTSDNPH RTIRFCRELKI. scapularis (128) QSIPDAEFRWRNRSRIKKTVEQAVERGYSDIAIINEDRRHPSKFVVQFLA. variegatum (51) QSIPNADFRWRNRSRIKKTVEQAIERGYSDIAVINEDRRHPNGLLLTHLH. sapiens (165) TVIPNSHVYYRRGLALKKIIPQCIARDFTDLIVINEDRKTPNGLILSHLConsensus (165) QSIPNA FRWRNRSRIKKTVEQAIERGYSDIAVINEDRRHPNGL L HLFigure 6. Analysis of clones 3C12 and 2F9 identical to RNA polymerase III and a hypothetical protein of unknown function,respectively. (A) Alignment of D. melanogaster (AAF57437), A. gambiae (TC6088), and H. sapiens (AAK61210) RNA polymerase IIIprotein sequences and the translation product of clone 3C12 identified as I. scapularis RNA polymerase III (AY296118). (B) Alignmentof A. variegatum (TC255), H. sapiens (FLJ12475) and I. scapularis clone 2F9 (AY296119) partial protein sequences. Protein sequencesare shown in the single letter amino acid code. Identical amino acids are shown in red and amino acids conserved in 2-3 of 4 (A) and 2 of3 (B) sequences are shown in blue.Table 5. Microsatellite STR sequences in I. scapularis ESTs.cDNA clone1A94G121F42C73B64H2Microsatellite sequenceTATATATATATATATATACACACACAGACACACTCACAATATATATATATAGCGCGCGCGTGTGCGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGTTATATATATATATATATATATATATGAAATGAAATGAAATGAAANonetheless, cDNAs associated with enhanced tickfeeding could be made as recombinant proteins to modifytheir immunogenicity and then be evaluated as candidateprotective antigens. Additionally, these antigens may alsobe good candidates for blocking the transmission of tickbornepathogens (Wikel et al, 1997; Labuda et al, 2002).The enhanced feeding effect of cDNA clones withidentity to App (2C12), mannose-binding lectin (3E10)and RNA polymerase III (3C12) is difficult to explain.The beta-amyloid protein precursor is involved in differentphysiological processes, including development of theembryonic nervous system in D. melanogaster (Rosen etal, 1989) and pharyngeal pumping in Caenorhabditiselegans (Zambreano et al, 2002). The sequence containedin clone 2C12 corresponded to the beta-amyloid peptide(ß-AP), a ≈40 amino acids peptide derived from the APPprotein found as the major component of dense plaques inbrains of Alzheimer disease patients (reviewed byCummings, 2003). Vaccination with ß-AP prevented theformation of ß-AP plaques in transgenic mice, opening anew possible approach for treatment of Alzheimer disease(McGeer and McGeer, 2003). However, we do notunderstand the apparent enhanced feeding effect of thetick ß-AP in cDNA-vaccinated mice. The lectin in clone3E10 was identical to mannose-binding endoplasmicreticulum-Golgi intermediate compartment protein (Araret al, 1995; Lahtinen et al, 1996). However, thecarbohydrate-binding domain is shared by other lectinsfound in different cell compartments. The clone 3C12encoded for an RNA polymerase III. Enhanced tick55

Almazán et al: Expressed sequence tags in Ixodes scapularisfeeding was produced in mice vaccinated with a DNApool containing this clone and clone 2F9 of unknownfunction. It is therefore possible that the enhanced feedingeffect on tick larvae was due to clone 2F9 with little or nocontribution of clone 3C12.Microsatellites are a class of genetic markers that arecomposed of STR sequences flanked by unique DNAsequences (Hearne et al, 1992). STRs are highlypolymorphic and widely distributed through the genome.The analysis of tick STRs has been used for identificationof strains of B. microplus (de la Fuente et al, 2000b) andfor the development of a preliminary genetic linkage mapof I. scapularis (Ullman et al, 2003). The STR sequencesdescribed in this study could be used for completion of thegenetic map of I. scapularis as the first step toward thesequencing of this tick genome.Most sequences in the I. scapularis EST data setwere relatively G + C rich, with an average G + C contentof 54%, similar to the 52% reported by Nene et al. (2002)for A. variegatum. The few sequences with a high A + Tcontent probably corresponded to mitochondrial genes,with pathogens) provides new exciting possibilities forscreening and identifying antigens protective against tickinfestations. This approach may also allow foridentification of antigens that interfere with pathogendevelopment and transmission.Recently, Almazán et al. (2003) used cDNA ELIcombined with EST analysis as a rapid method for theidentification of protective antigens against I. scapularisinfestations, demonstrating the role of sequenceinformation in conjunction with new technologies such asbioinformatics and ELI for a systematic andcomprehensive approach to vaccine discovery.One of the advantages of ELI for identification ofprotective antigens is that a priori criteria are notintroduced to direct the selection of candidate genes. Thisapproach, as shown in this study, resulted in potentialvaccine antigens otherwise not predicted, such as clone4F8 that was found to be identical to a nucleotidase.However, nucleotidases are essential for cell growth andthe inhibition of its enzymatic activity would be cytotoxic(Spiegelberg et al, 1999), providing a possible explanationfor their protective properties against tick infestations. TheI. scapularis sequence in clone 4F8 was different from the5’-nucleotidase that was identified and characterizedpreviously by Liyou et al. (1999, 2000) in B. microplus.However, the protective capacity of this protein has notbeen evaluated.As discussed previously by Almazán et al, (2003), apossible explanation for the inhibitory effect on larval tickdevelopment of other vaccine candidates that wereidentified in this study is based on the role that they playin cell growth and maintenance, which is evident forclones identical to beta-adaptin (3E1) and chloride channel(4G11). Beta adaptins are adaptor components required inthe assembly of clathrin-coated plasma membrane pits thatfunction in cell vesicular transport mechanisms includingendocytosis (Camidge and Pearse, 1994; Boehm andBonifacino, 2002), a process actively involved in blooddigestion by ticks and other hematophagous arthropods(Akov, 1982). Chloride channels are also involved in vitalcell functions including the catalysis of counter ioncurrents that accompany primary proton fluxes inendosomal and lysosomal acidification (Koprowski andKubalski, 2001; Iyer et al, 2002). Therefore, interferencewith the process of endocytosis may impair acquisitionand digestion of the tick bloodmeal and result in inhibitionof tick infestations. Another I. scapularis EST (clone3E12) encoded for a protein identical to D. melanogasterclathrin heavy chain, a protein involved in synaptic vesicleendocytosis (Chang et al, 2002). This cDNA is also acandidate protective antigen because it interfers withendocytosis in feeding larvae.The protection capacity of ribosomal and heat shockprotein preparations has been documented previously inother organisms (Elad and Segal, 1995; Silva, 1999;Melby et al, 2000; Cassataro et al, 2002). Recently, Hsp70was demonstrated to be induced in I. ricinus salivaryglands during blood feeding (Leboulle et al, 2002),documenting the role of heat shock proteins inphysiological responses in ticks. Even in the case wheresubstantial homology exists between tick proteins and host(mouse) proteins, analysis of ribosomal proteins suggeststhat differences in the amino acid sequence could directthe host immune response against distinctive, non-selfepitopes, which could be sufficient to induce a protectiveresponse.The results of vaccination and tick infestationdemonstrated that some cDNAs enhance tick feeding. Thiseffect could be due to the expression corroborating thehypothesis that there is a marked difference in codonusage between mitochondrial and nuclear protein codinggenes in the Ixodidae (Nene et al, 2002).Most of the ESTs in our database, although initiallyidentified by ELI of cDNA pools that produced inhibitionof tick infestation, were not characterized further andremain potential candidate antigens for vaccinedevelopment against I. scapularis infestations. Particularlyinteresting were cDNAs that may be involved indevelopmental processes. Clone 4B2, identical to D.melanogaster sequence NP_523710, encoded forcalmodulin, a Ca ++ -binding protein of 149 amino acids thatis involved in fly development. This protein was found tobe expressed in several larval and adult tissues, includingthe larval midgut (Takamatsu et al, 2002). Clone 1C8 hada low degree of identity to D. melanogaster virilizer, agene involved in Sex-lethal (Sxl) splicing and essential forfly male and female viability and embryonic development(Niessen et al, 2001). Clone 2A11 also had a low degree ofidentity to D. melanogaster developmental regulator,Notchless, a key player in the signaling by Notch familyreceptors that are involved in many cell-fate decisionsduring development (Royet et al, 1998). Similarly, clone4A10 had partial identity to the putative homeodomaintranscriptional factor, phtf, a member of a gene family thatplays an important role during development and isconserved between fly, mouse and human (Manuel et al,2000). Other clones with special interest as vaccinecandidates may include those identical to membraneproteins (1D8, 1D11, 3G11) and those putatively involved56

Gene Therapy and Molecular Biology Vol 7, page 57in G-protein-coupled signaling (2B7, 2F12, 4C9). In fact,the clone 3G11 was identical to D. melanogaster BM-40, aprotein of the group of extracellular basement membraneproteins which includes the protective antigen p29 fromHaemaphysalis longicornis (Mulenga et al, 1999).In summary, we have characterized I. scapularis ESTsequences that were selected by cDNA ELI in themouse/tick challenge model because they affected tickdevelopment. Characterization of these ESTs provides abasis for future research on ticks and is a source ofcandidate antigens for use in vaccine developmentdesigned to control tick infestations and/or reducetransmission of pathogens. The combination of ELI withEST appears to be a productive systematic andcomprehensive approach to vaccine discovery.AcknowledgmentsThis research was supported by the project No. 1669of the Oklahoma Agricultural Experiment Station, theEndowed Chair for Food Animal Research (K. M. Kocan,College of Veterinary Medicine, Oklahoma StateUniversity), NIH Centers for Biomedical ResearchExcellence through a subcontract to J. de la Fuente fromthe Oklahoma Medical Research Foundation, and theOklahoma Center for the Advancement of Science andTechnology, Applied Research Grant, AR00(1)-001 andAR02(1)-037. Consuelo Almazán is supported by a grantin-aidfrom the CONACYT, Mexico and an assistantshipfrom the College of Veterinary Medicine, Oklahoma StateUniversity. J. C. Garcia-Garcia is supported by a HowardHughes Medical Institute Predoctoral Fellowship inBiological Sciences. Jerry Bowman is acknowledged forproviding tick larvae. Janet J. Rogers and Sue AnnHudiburg (Core Sequencing Facility, Department ofBiochemistry and Molecular Biology, Noble ResearchCenter, Oklahoma State University) are acknowledged forDNA sequencing and oligonucleotide synthesis,respectively. We thank Joy Yoshioka for editorialassistance.ReferencesAlberti E, Acosta A, Sarmiento ME, Hidalgo C, Vidal T,Fachado A, Fonte L, Izquierdo L, Infante JF, Finlay CM,Sierra G (1998) Specific cellular and humoral immuneresponse in Balb/c mice immunised with an expressiongenomic library of Trypanosoma cruzi. Vaccine 16, 608-612.Almazán C, Kocan KM, Bergman DK, Garcia-Garcia JC, BlouinEF, de la Fuente J (2003) Identification of protective antigensfor the control of Ixodes scapularis infestations using cDNAexpression library immunization. Vaccine 21, 1492-1501.Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990)Basic local alignment search tool. J Mol Biol 215, 403-410.Altschul, SF, Madden TL, Schäffer AA, Zhang J, Zhang Z,Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database searchprograms. Nuc Acids Res 25, 3389-3402.Arar C, Carpentier V, Le Caer JP, Monsigny M, Legrand A,Roche AC (1995) ERGIC-53, a membrane protein of theendoplasmic reticulum-Golgi intermediate compartment, isidentical to MR60, an intracellular mannose-specific lectin ofmyelomonocytic cells. J Biol Chem 270, 3551-3553.Ashburner M, Ball CA, Blake JA, et al (2000) Gene ontology:tool for the unification of biology. Nature Genet 25, 25-29.Akov S. Blood digestion in ticks. In: Obenchain FD, Galun R,editors (1982) Physiology of ticks. Current themes intropical science (vol. 1). Oxford, Pergamon, pp. 197-211.Barry MA, Lai WC, Johnston SA (1995) Protection againstmycoplasma infection using expression-libraryimmunization. Nature 377, 632-635.Berriz GF, White JV, King OD, Roth FP (2003) GoFish findsgenes with combinations of Gene Ontology attributes.Bioinformatics 19, 788-789.Bior AD, Essenberg RC, Sauer JR (2002) Comparison ofdifferentially expressed genes in the salivary glands of maleticks, Amblyomma americanum and Dermacentor andersoni.Insect Biochem Mol Biol 32, 645-655.Boguski MS, Lowe TM, Tolstoshev CM (1993) dbEST-databasefor "expressed sequence tags". Nat Genet 4, 332-333.Boehm M, Bonifacino JS (2002) Genetic analyses of adaptinfunction from yeast to mammals. Gene 286, 175-186.Brayton KA, Vogel SW, Allsopp BA (1998) Expression libraryimmunization to identify protective antigens from Cowdriaruminantium. Ann N Y Acad Sci 849, 369-371.Camidge DR, Pearse BM (1994) Cloning of Drosophila betaadaptinand its localization on expression in mammaliancells. J Cell Sci 107, 709-718.Cassataro J, Velikovsky CA, Giambartolomei GH, Estein S,Bruno L, Cloeckaert A, Bowden RA, Spitz M, Fossati CA(2002) Immunogenicity of the Brucella melitensisrecombinant ribosome recycling factor-homologous proteinand its cDNA. Vaccine 20, 1660-1669.Chang HC, Newmyer SL, Hull MJ, Ebersold M, Schmid SL,Mellman I (2002) Hsc70 is required for endocytosis andclathrin function in Drosophila. J Cell Biol 159, 477-487.Crampton AL, Miller C, Baxter GD, Barker SC (1998)Expressed sequenced tags and new genes from the cattle tick,Boophilus microplus. Exp Appl Acarol 22, 177-186.Cummings JL (2003) Alzheimer's disease: from molecularbiology to neuropsychiatry. Semin Clin Neuropsychiatry 8,31-36.de la Fuente J, Rodriguez M, Redondo M, Montero C, Garcia-Garcia JC, Mendez L, Serrano E, Valdes M, Enriquez A,Canales M, Ramos E, Boue O, Machado H, Lleonart R, deArmas CA, Rey S, Rodriguez JL, Artiles M, Garcia L (1998)Field studies and cost-effectiveness analysis of vaccinationwith Gavac against the cattle tick Boophilus microplus.Vaccine 16, 366-373.de la Fuente J, Rodriguez M, Montero C, Redondo M, Garcia-Garcia JC, Mendez L, Serrano E, Valdes M, Enriquez A,Canales M, Ramos E, Boue O, Machado H, Lleonart R(1999) Vaccination against ticks (Boophilus spp.): theexperience with the Bm86-based vaccine Gavac. Genet Anal15, 143-148.de la Fuente J, Rodriguez M, Garcia-Garcia JC (2000a)Immunological control of ticks through vaccination withBoophilus microplus gut antigens. Ann N Y Acad Sci 916,617-621.de la Fuente J, Garc_a-Garc_a JC, Gonz_lez DM, Izquierdo G,Ochagavia ME (2000b) Molecular analysis of Boophilus spp.(Acari: Ixodidae) tick strains. Vet Parasitol 92, 209-222.de Vos S, Zeinstra L, Taoufik O, Willadsen P, Jongejan F (2001)Evidence for the utility of the Bm86 antigen from Boophilus57

Almazán et al: Expressed sequence tags in Ixodes scapularismicroplus in vaccination against other tick species. ExpAppl Acarol 25, 245-261.Elad D, Segal E (1995) Immunogenicity in calves of a cruderibosomal fraction of Trichophyton verrucosum: a field trial.Vaccine 13: 83-87.Estrada-Peña A, Jongejan F (1999) Ticks feeding on humans: areview of records on human-biting Ixodoidea with specialreference to pathogen transmission. Exp Appl Acarol 23,685-715.Garcia-Garcia JC, Gonzalez IL, Gonzalez DM, Valdes M,Mendez L, Lamberti J, D'Agostino B, Citroni D, Fragoso H,Ortiz M, Rodriguez M, de la Fuente J (1999) Sequencevariations in the Boophilus microplus Bm86 locus andimplications for immunoprotection in cattle vaccinated withthis antigen. Exp Appl Acarol 23, 883-395.Hearne CM, Ghosh S Todd J (1992) Microsatellites for linkageanalysis of genetic traits. Trends Genet 8, 288-294.Hill CA, Gutierrez JA (2000) Analysis of the expressed genomeof the lone star tick, Amblyomma americanum (Acari:Ixodidae) using an expressed sequence tag approach. MicrobComp Genomics 5, 89-101.Iyer R, Iverson TM, Accardi A, Miller C (2002) A biological rolefor prokaryotic ClC chloride channels. Nature 419, 715-718.Kessler MM, Willins DA, Zeng Q, Del Mastro RG, Cook R,Doucette-Stamm L, Lee H, Caron A, McClanahan TK, WangL, Greene J, Hare RS, Cottarel G, Shimer GH Jr (2002) Theuse of direct cDNA selection to rapidly and effectivelyidentify genes in the fungus Aspergillus fumigatus. FungalGenet Biol 36, 59-70.Knox DP, Redmond DL, Skuce PJ, Newlands GF (2001) Thecontribution of molecular biology to the development ofvaccines against nematode and trematode parasites ofdomestic ruminants. Vet Parasitol 101, 311-335.Koprowski P, Kubalski A (2001) Bacterial ion channels and theireukaryotic homologues. Bioessays 23, 1148-1158.Labuda M, Trimnell AR, Lickova M, Kazimirova M, Slovak M,Nuttall PA (2002) Recombinant tick salivary antigens(64TRP) as a TRANSBLOK vaccine against tick-borneencephalitis virus. Abstracts 4 th International Conferenceon Ticks and Tick-Borne Pathogens, Banff, Canada, p. 51.Lahtinen U, Hellman U, Wernstedt C, Saraste J, Pettersson RF(1996) Molecular cloning and expression of a 58-kDa cis-Golgi and intermediate compartment protein. J Biol Chem271, 4031-4037.Leboulle G, Rochez C, Louahed J, Ruti B, Brossard M, BollenA, Godfroid E (2002) Isolation of Ixodes ricinus salivarygland mRNA encoding factors induced during blood feeding.Am J Trop Med Hyg 66, 225-233.Leclercq S, Harms JS, Oliveira SC (2003) Enhanced efficacy ofDNA vaccines against an intracellular bacterial pathogen bygenetic adjuvants. Curr Pharm Biotechnol 4, 99-107.Liyou N, Hamilton S, Elvin C, Willadsen P (1999) Cloning andexpression of ecto 5’-nucleotidase from the cattle tickBoophilus microplus. Insect Mol Biol 8, 257-266.Liyou N, Hamilton S, Mckenna R, Elvin C, Willadsen P (2000)Localization and functional studies on the 5’-nucleotidase ofthe cattle tick Boophilus microplus. Exp Appl Acarol 24,235-246.Lizotte-Waniewski M, Tawe W, Guiliano DB, Lu W, Liu J,Williams SA, Lustigman S (2000) Identification of potentialvaccine and drug target candidates by expressed sequence taganalysis and immunoscreening of Onchocerca volvulus larvalcDNA libraries. Infect Immun 68, 3491-3501.Manoutcharian K, Terrazas LI, Gevorkian G, Govezensky T(1998) Protection against murine cysticercosis using cDNAexpression library immunization. Immunol Lett 62, 131-136.Manuel A, Beaupain D, Romeo PH, Raich N (2000) Molecularcharacterization of a novel gene family (PHTF) conservedfrom Drosophila to mammals. Genomics 64, 216-220.McGeer PL, McGeer E (2003) Is there a future for vaccination asa treatment for Alzheimer's disease? Neurobiol Aging 24,391-395.McSwain JL, Luo C, deSilva GA, Palmer MJ, Tucker JS, SauerJR, Essenberg RC (1997) Cloning and sequence of a gene fora homologue of the C subunit of the V-ATPase from thesalivary gland of the tick Amblyomma americanum (L).Insect Mol. Biol 6, 67-76.Melby PC, Ogden GB, Flores HA, Zhao W, Geldmacher C,Biediger NM, Ahuja SK, Uranga J, Melendez M (2000)Identification of vaccine candidates for experimental visceralleishmaniasis by immunization with sequential fractions of acDNA expression library. Infect Immun 68, 5595-5602.Moore RJ, Lenghaus C, Sheedy SA, Doran TJ (2001) Improvedvectors for expression library immunization-application toMycoplasma hyopneumoniae infection in pigs. Vaccine 20,115-120.Mulenga A, Sugimoto C, Sako Y, Ohashi K, Musoke A, ShubashM, Onuma M (1999) Molecular characterization of aHaemaphysalis longicornis tick salivary gland-associated 29-kilodalton protein and its effect as a vaccine against tickinfestation in rabbits. Infect Immun 67, 1652-1658.Mulenga A, Macaluso KR, Simser JA, Azad AF (2003)Dynamics of Rickettsia-tick interactions: identification andcharacterization of differentially expressed mRNAs inuninfected and infected Dermacentor variabilis. Insect MolBiol 12, 185-193.Munderloh UG, Wang YLM, Chen C, Kurtti TJ (1994)Establishment, maintenance and description of cell linesfrom the tick Ixodes scapularis. J Parasitol 80, 533-543.Nene V, Lee D, Quackenbush J, Skilton R, Mwaura S, GardnerMJ, Bishop R (2002) AvGI, an index of genes transcribed inthe salivary glands of the ixodid tick Amblyommavariegatum. Int J Parasitol 32, 1447-1456.Niessen M, Schneiter R, Nothiger R (2001) Molecularidentification of virilizer, a gene required for the expressionof the sex-determining gene Sex-lethal in Drosophilamelanogaster. Genetics 157, 679-688.Parola P, Raoult D (2001) Tick-borne bacterial diseasesemerging in Europe. Clin Microbiol Infect 7, 80-83.Pestov DG, Grzeszkiewicz TM, Lau LF (1998) Isolation ofgrowth suppressors from a cDNA expression library.Oncogene 17, 3187-3197.Rosen DR, Martin-Morris L, Luo LQ, White K.A (1989)Drosophila gene encoding a protein resembling the humanbeta-amyloid protein precursor. Proc Natl Acad Sci U S A86, 2478-2482.Royet J, Bouwmeester T, Cohen SM (1998) Notchless encodes anovel WD40-repeat-containing protein that modulates Notchsignaling activity. EMBO J 17, 7351-7360.Saitou N, Nei M (1987). The neighbor-joining method: a newmethod for reconstructing phylogenetic trees. Mol Biol Evol4, 406-425.Silva CL (1999) The potential use of heat-shock proteins tovaccinate against mycobacterial infections. Microbes Infect1, 429-435.58

Gene Therapy and Molecular Biology Vol 7, page 59Singh RA, Wu L, Barry MA (2002) Generation of genome-wideCD8 T cell responses in HLA-A*0201 transgenic mice by anHIV-1 ubiquitin expression library immunization vaccine. JImmunol 168, 379-391.Smooker PM, Setiady YY, Rainczuk A, Spithill TW (2000)Expression library immunization protects mice against achallenge with virulent rodent malaria. Vaccine 18, 2533-2540.Spiegelberg BD, Xiong JP, Smith JJ, Gu RF, York JD (1999)Cloning and characterization of a mammalian lithiumsensitivebisphosphate 3'-nucleotidase inhibited by inositol1,4-bisphosphate. J Biol Chem 274, 13619-13628.Strezoska Z, Pestov DG, Lau LF (2000) Bop1 is a mouse WD40repeat nucleolar protein involved in 28S and 5. 8S RRNAprocessing and 60S ribosome biogenesis. Mol Cell Biol 20,5516-5528.Takamatsu Y, Nakagoshi H, Rachidi M, Lopes C, Nishida Y,Ohsako S (2002) Characterization of the dCaMKII-GAL4driver line whose expression is controlled by the DrosophilaCa(2+)/calmodulin-dependent protein kinase II promoter.Cell Tissue Res 310, 237-252.Tarleton RL, Kissinger J (2001) Parasite genomics: current statusand future prospects. Curr Opin Immunol 13, 395-402.Thompson JD, Higgins DG, Gibson TJ (1994). CLUSTAL W:improving the sensitivity of progressive multiple sequencealignment through sequence weighting, positions-specificgap penalties and weight matrix choice. Nuc Acids Res 22,4673-4680.Touloukian CE, Leitner WW, Robbins PF, Rosenberg SA,Restifo NP (2001) Mining the melanosome for tumor vaccinetargets: P.polypeptide is a novel tumor-associated antigen.Cancer Res 61, 8100-8104.Ullmann AJ, Piesman J, Dolan MC, Iv WC (2003) A preliminarylinkage map of the hard tick, Ixodes scapularis. Insect MolBiol 12, 201-210.Valenzuela JG (2002) Exploring the messages of the salivaryglands of Ixodes ricinus. Am J Trop Med Hyg 66, 223-224.Valenzuela JG, Francischetti IM, Pham VM, Garfield MK,Mather TN, Ribeiro JM (2002) Exploring the sialome of thetick Ixodes scapularis. J Exp Biol 205, 2843-2864.Willadsen P (1997) Novel vaccines for ectoparasites. VetParasitol 71, 209-222.Willadsen P, Jongejan F (1999) Immunology of the tick-hostinteraction and the control of ticks and tick-borne diseases.Parasitol Today 15, 258-262.Wikel SK, Ramachandra RN, Bergman DK, Burkot TR, PiesmanJ (1997) Infestation with pathogen-free nymphs of the tickIxodes scapularis induces host resistance to transmission ofBorrelia burgdorferi by ticks. Infect Immun 65, 335-338.Zambrano N, Bimonte M, Arbucci S, Gianni D, Russo T,Bazzicalupo P (2002) feh-1 and apl-1, the Caenorhabditiselegans orthologues of mammalian Fe65 and beta-amyloidprecursor protein genes, are involved in the same pathwaythat controls nematode pharyngeal pumping. J Cell Sci 115,1411-1422Back row from left to right: Jose C. Garcia-Garcia, KatherineM. Kocan, Jose de la Fuente;Front row: Consuelo Almazán and Edmour F. Blouin59

Almazán et al: Expressed sequence tags in Ixodes scapularis60

Gene Therapy and Molecular Biology Vol 7, page 61Gene Ther Mol Biol Vol 7, 61-68, 2003Delayed intratracheal injection of manganesesuperoxide dismutase (MnSOD)-plasmid/liposomesprovides suboptimal protection against irradiationinducedpulmonary injury compared to treatmentbefore irradiationResearch ArticleMichael W. Epperly, Hongliang Guo, Michael Bernarding, Joan Gretton, MiaJefferson, Joel S. Greenberger*Department of Radiation Oncology, University of Pittsburgh Cancer Institute, Pittsburgh, PA 15213__________________________________________________________________________________*Correspondence: Joel S. Greenberger, M.D., Professor and Chairman, Department of Radiation Oncology, University of PittsburghCancer Institute, B346-PUH 200 Lothrop Street, Pittsburgh, PA 15213; Telephone: 412-647-3607; Fax: 412-647-6029; Email:greenbergerjs@msx.upmc.eduKey words: MnSOD, reactive oxygen species, pulmonary fibrosisAbbreviations: OCT Optimum Cutting Temperature, ROS reactive oxygen speciesReceived: 10 May 2003; Accepted: 10 June 2003; electronically published: June 2003SummaryIonizing irradiation results in cellular production of reactive oxygen species (ROS), which cause DNA strandbreaks, lipid peroxidation or other cellular damage leading to cell death. Antioxidant enzymes neutralize these ROSand provide cellular protection against sources of oxidative stress including ionizing irradiation. Intratrachealinjection of the transgene for antioxidant protein MnSOD in plasmid/liposome (PL) complex 24 hours beforeirradiation has been shown to protect the murine lung from irradiation-induced organizing alveolitis/fibrosis. Todetermine whether intratracheal injection of MnSOD-PL at later times of macrophage infiltration andinflammation following irradiation had a detectable protective effect against irradiation fibrosis, control noninjectedor MnSOD-PL complex injected C57BL/6J mice were irradiated to 20 Gy. Subgroups received a delayedinjection of MnSOD-PL at day 1, 80, 90 or 100 after irradiation and all were followed for the development oforganizing alveolitis/fibrosis. While mice injected with MnSOD-PL prior to irradiation demonstrated the best levelof protection, we observed that mice injected with MnSOD-PL at 80 or 100 days after irradiation also showedsignificant protection of the lung compared to irradiated, control mice. Thus, delayed administration of MnSOD-PLhas detectable radioprotective effects on C57BL/6J mouse lung but pre-irradiation injection remains the optimaltreatment paradigm.I. IntroductionMnSOD is a mitochondrial localized enzyme whichreduces superoxides produced during respiration (Quinlanet al, 1994; Fridovich, 1995). Therapeutic increase inexpression of MnSOD by transgene administrationprotects tissues and organs from irradiation damageincluding lung (Epperly et al, 1998; 1999b; 2000a)esophagus, (Stickle et al, 1999; Epperly et al, 2001a;Epperly et al, 2000b) oral cavity (Guo et al, 2003) andbladder (Kanai et al, 2002). Increased expression ofMnSOD at the time of irradiation also decreases theirradiation induction of inflammatory cytokines includingtumor necrosis factor-alpha (TNF-α), interleukin (IL)-1,and transforming growth factor-beta (TGF-β) (Epperly etal, 1999c). Approximately 80 days after total lungirradiation C57BL/6J mice show increased TNF-α mRNAand this level decreases to background levels by 120 daysfollowing irradiation (Epperly et al, 1999c). As TNF-αmRNA expression decreases, that for TGF-β increases at100 days after irradiation and continues to elevate duringthe development of the pathologic changes of organizingalveolitis/fibrosis (Epperly et al, 1999c). At the initiation61

Epperly et al: Late injection of MnSOD-PL protects against pulmonary fibrosisof organizing alveolitis/fibrosis, an increase in TGF-β1 isalso detected (Epperly et al, 1999c). This late increase inTGF-β1 persists to day 120, the time at which TGF-β2expression also increases (Epperly et al, 1999c). Levels ofTGF-β2 remain elevated throughout the development oforganizing alveolitis/fibrosis (Epperly et al, 1999c).We have previously demonstrated that intratrachealinjections of MnSOD-PL complex or adenoviruscontaining the human MnSOD transgene 24 hours beforeirradiation protects the murine lung from irradiationinduceddamage (Epperly et al, 1998; 1999b; 2000a,2001b). Protection of the murine lung was measured as:(a) increased survival (Epperly et al, 1998, 1999b), (b)decreased pathologically quantifiable percent of lungshowing organizing alveolitis/fibrosis, (Epperly et al,1998; 1999b; 2000a, 2001b) and (c) decreased productionof inflammatory cytokine mRNA for IL-1, TNF-α, andTGF-β (Epperly et al, 1998, Epperly et al, 2001b). Theoptimal schedule for administration of MnSOD-PL is notknown. Injection prior to irradiation might be effective bypreventing ROS mediated DNA damage or protectingagainst mitochondrial mediated apoptosis (Epperly et al,1999a, 2002). However, injection following irradiation orat delayed time points when increases in TNF-α and TGFβmRNA are detected may reduce cytokine mediatedproduction of ROS and also protect against tissue injury.To determine the optimal time of MnSOD-PLadministration in the C57BL/6J mouse model, mice wereinjected with MnSOD-PL at 1, 80, 90, or 100 days after 20Gy whole lung irradiation and data were compared to thatwith mice treated before irradiation. The mice werefollowed for development of organizing alveolitis/fibrosisand the percent of lung displaying organizingalveolitis/fibrosis was determined. Since MnSOD is amitochondrial enzyme that dismutates superoxides only(Quinlan et al, 1994; Fridovich, 1995) the detection ofincreased survival in delayed injection groups of micemight indicate the presence of delayed increases insuperoxide production, and thus be interpreted to play arole in the development of pulmonary fibrosis. In thepresent studies, we sought to determine whether delayedelevation of MnSOD by transgene therapy protects lungsfrom irradiation-induced pulmonary fibrosis.II. Materials and methodsA. Injection of MnSOD-PLC57BL/6J were anesthetized using Nembutal and injectedintratracheally with MnSOD-PL complexes (500 µg plasmidDNA in a volume of 50 µl plus 28 µl of lipofectant) (Epperly etal, 1998; 1999b) Twenty-four hours later the MnSOD-PLinjectedmice plus control non-injected mice were irradiated to20 Gy to the pulmonary cavity. The mice were shielded so thatonly the pulmonary cavity was irradiated. A subgroup of thecontrol, irradiated mice was injected with MnSOD-PL 24 hoursafter irradiation. Other subgroups of each control irradiated orMnSOD-PL pre-irradiation injected mice were injectedintratracheally a second time at day 80, 90 or 100 followingirradiation. All mice were followed for development oforganizing alveolitis/fibrosis, at which time the mice weresacrificed.B. Determination of organizingalveolitis/fibrosisWhen 80% of the control, irradiated mice had beensacrificed due to moribund condition as indicator of pulmonaryorganizing alveolitis/fibrosis, a subgroup of mice from eachgroup was also sacrificed. The lungs were expanded withOptimum Cutting Temperature (OCT), removed, frozen in OCT,sectioned, and hematoxylin and eosin (H&E)-stained (Epperly etal, 1998; Epperly et al, 1999b). The sections were examinedmicroscopically and the percent of organizing alveolitis/fibrosiswas determined using an Optimus Image Analysis System(Epperly et al, 1998; Epperly et al, 1999b). In this system, thearea of organizing alveolitis/fibrosis was compared to the area ofthe entire lobe, and the percent of lung developing organizingalveolitis/fibrosis calculated.C. StatisticsThe irradiation survival curves of the different subgroupswere compared with control irradiated mice using a Log RankTest (Epperly et al, 1998; 1999b). The percent organizingalveolitis/fibrosis for the different subgroups of mice werecompared using a Student’s t-Test (Epperly et al, 1998; Epperlyet al, 1999b).D. Animal protocolsProtocols for animal usage were approved by theInstitutional Animal care and Use Committee of the Universityof Pittsburgh. Veterinary support was provided by the Divisionof Laboratory Animal Research of the University of Pittsburgh.III. ResultsA. Delayed injection of MnSOD-PL afterlung irradiation improves survivalTo determine whether intratracheal injection ofMnSOD-PL at delayed intervals following irradiationprotected the murine lung from irradiation-induceddamage, C57BL/6J mice were injected intracheally with500 µg of plasmid DNA containing the MnSOD transgeneat 1, 80, 90 or 100 days following 20 Gy irradiation to thepulmonary cavity. The mice were then followed for thedevelopment of organizing alveolitis/fibrosis and weresacrificed when moribund. Mice injected with MnSOD-PLat 80 or 100 days after irradiation showed a significantincrease in survival compared to 20 Gy irradiated noninjectedcontrol mice while mice injected with MnSOD-PL at day 1 or 90 after irradiation showed a detectable butnot significant increase in survival (Figure 1).B. Pre-irradiation injection of MnSOD-PL affords optimal protection and is notfurther enhanced by a second delayedtreatmentGroups of mice were next injected with MnSOD-PL24 hours before 20 Gy irradiation to the pulmonary cavityand then evaluated to determine whether a secondinjection of MnSOD-PL at 80, 90 or 100 days afterirradiation resulted in an additional increase in survivalcompared to single pre-irradiation therapy. In this study,62

Gene Therapy and Molecular Biology Vol 7, page 63subgroups of mice received a second injection of MnSOD-PL at 80, 90 or 100 days later. The mice were thenfollowed for the development of organizingalveolitis/fibrosis at which time they were sacrificed. Asshown in Figure 2, there was no significant improvementin survival following a second injection of MnSOD-PLcompared to the improvement seen with one preirradiationinjection. A second injection at day 90 resultedin a significantly decreased survival compared to preinjectiononly.A comparison of these injection groups is shown inFigure 3. All subgroups of mice injected with MnSOD-PL24 hours before irradiation had increased survivalcompared to mice that received no injection and only 20Gy irradiation. Furthermore, mice injected with MnSOD-PL 24 hours prior to irradiation showed the best survivalcompared to all other groups of mice including those thatreceived a second delayed injection.C. Decreased lung irradiation damagehistopathologically correlates to MnSOD-PLmediated increased survivalTo determine whether the differences in survival ofmice between groups correlated with histopathologicchanges in the lung, specifically the development oforganizing alveolitis/fibrosis, representatives of eachsubgroup of mice were euthanized at the time point when80% of the 20 Gy irradiated, control mice were sacrificeddue to moribund condition from developing organizingalveolitis/fibrosis. The lungs were expanded in OCT,removed, frozen in OCT, sectioned, and H&E-stained.The percent of lung displaying organizingalveolitis/fibrosis was calculated using an Optimus ImageAnalysis system as described in the Methods.Figure 2: Improved survival of mice injected with MnSOD-PL24 hours before pulmonary irradiation is not further enhanced bya second delayed injection. C57BL/6J mice were injected withMnSOD-PL 24 hours before 20 Gy irradiation to the pulmonarycavity. Subgroups were injected with a second dose of MnSOD-PL at 80, 90 or 100 days after the initial irradiation. There was nosignificant improvement in the overall survival by a secondinjection 80, 90 or 100 days after irradiation (p=0.547, 0.039, and0.309 respectively) compared to pre-irradiation administrationabove. A second injection at day 90 resulted in significantlydecreased survival compared to pre-injection only. Groupscontained ≥10 mice/group.Figure 1: Improved survival of pulmonary irradiated C57BL/6Jmice injected with MnSOD-PL at day 1, 80, 90 or 100 followingirradiation. C57BL/6J mice were irradiated to 20 Gy to thepulmonary cavity. Subgroups were subsequently injected withMnSOD-PL on day 1, 80, 90, or 100 following irradiation. Themice were followed for the development of organizingalveolitis/fibrosis, at which time they were sacrificed. Theseresults demonstrated that injection of MnSOD-PL at day 80 or100 following irradiation (or times when TNF-α and TGF-βproduction are increased) increases survival compared toirradiated, control mice (p = 0.0015 or 0.0005, respectively).Groups contained ≥10 mice/group.Figure 3: Pre-irradiation injection of MnSOD-PL providesoptimal protection from lung irradiation damage. C57BL/6J micewere injected with MnSOD-PL and irradiated 24 hours later to20 Gy to the lung, as were non-injected control mice. Subgroupsof mice were subsequently injected with MnSOD-PL at day 1(control, irradiated mice only), 80, 90 or 100 after irradiation.The mice were then followed for development of organizingalveolitis/fibrosis, and were sacrificed when moribund. All miceinjected with MnSOD-PL 24 hours before irradiation had asignificantly increased life span compared to control, irradiatedmice (p ≤ 0.0066). Groups contained ≥10 mice/group.63

Gene Therapy and Molecular Biology Vol 7, page 65Figure 5: Delayed injection of MnSOD-PL provides detectable protection from irradiation-induced organizing alveolitis/fibrosis.C57BL/6J mice were injected with MnSOD-PL 24 hours before 20 Gy irradiation to the pulmonary cavity. Subgroups of the MnSOD-PL-injected mice were given a second injection of MnSOD-PL at day 80, 90 or 100. Subgroups of non-injected but 20 Gy irradiatedcontrol mice were injected with MnSOD-PL only at day 1, 80, 90 or 100 following irradiation. Once 80% of the non-injected control,irradiated mice had been sacrificed due to moribund condition resulting from organizing alveolitis/fibrosis, representative mice in eachother group were sacrificed. The lungs were expanded in OCT, excised, frozen in OCT, sectioned, and H&E-stained. Representativephotographs of the lungs at the time of sacrifice are shown for: non-irradiated mice (A); 20 Gy non-injected control, irradiated mice, (B);20 Gy irradiated mice injected with MnSOD-PL at 80 days, (C); MnSOD-PL-injected mice 24 hours before 20 Gy, (D); or mice injectedwith MnSOD-PL both 24 hours before irradiation and again at day 80 (E).65

Epperly et al: Late injection of MnSOD-PL protects against pulmonary fibrosisTherefore, MnSOD-PL action on ROS produced byinflammatory cells such as macrophages at 80 days doesnot appear to explain the present data implyingsuperoxides might have been produced by macrophagesand neutralized by injections of MnSOD-PL at 80 or 100days after irradiation (Epperly et al, 2003).We previously demonstrated that at 80 days afterirradiation of the mouse lung there is an increase in TNF-αmRNA expression which decreases to background level byday 120 (Epperly et al, 1999b). This increase isaccompanied by increased expression of mRNA for TGFβat day 100. Initially, there is an increase in TGF-β1isoform until day 120 at which time TGF-β1 expressiondecreases, and TGF-β2 expression increases and stayselevated until development of organizing alveolitis/fibrosis(Epperly et al, 1999b). The detectable protection byMnSOD-PL injection at day 80 might have been attributedto an effect on the TNF-α elevation at that time point.ROS production might increase TNF-α expression at day80 leading to further increased ROS production (Haddad,2002). Treatment of alveolar epithelial cells with a ROSgenerating system results in increased TNF-α expressionand a depletion of glutathione (Haddad, 2002). TNF-αtreatment inhibits myogenesis by causing a decrease inglutathione levels and elevation of ROS (Langen et al,2002). Pre-treatment with the anti-oxidant N-acetyl-1-cysteine (NAC) restored the formation of multi-nucleatedmyotubes, indicating that myogenesis inhibition wasattributable to ROS expression (Langen et al, 2002). Pretreatmentof HELA cells with gammaglutamylcysteinylglycineinhibits TRAIL-inducedapoptosis (Lee et al, 2002). TNF-α expression mayincrease the production of ROS and result in a furtherincrease in TNF-α expression. ROS response to andinduction of TNF-α expression may be a cyclicmechanism in the lung at day 80, and MnSOD-PLtreatment at this time point may have interrupted the cycle.Further studies will be required to explain the protectionby injections of MnSOD-PL at 80 days after irradiation.Pulmonary increases in TGF-β1 and TGF-β2 mRNAat 100 to 120 days after irradiation have been detected(Epperly et al, 1999b). It has been demonstrated that ROScan also increase TGF-β expression (Bellocq et al, 1999)The treatment of human alveolar lung cell line A549 withxanthine and xanthine oxidase or nitric oxide generator S-nitroso-N-acetyl-penicillamine (SNAP) leads to release ofTGF-β1 (Bellocq et al, 1999). The xanthine-xanthineoxidase induced release of TGF-β1 can be inhibited by theaddition of catalase but not superoxide dismutase,implicating the involvement of hydrogen peroxide(Bellocq et al, 1999) TGF-β1 has been demonstrated toinduce production of extracellular hydrogen peroxide inhuman fibroblasts that mediate oxidative dityrosinedependentcross-linking of ECM (Larios et al, 2001).TGF-β and hydrogen peroxide have been observed toinduce connective tissue factor (CTGF) that then inducescollagen type 1 and fibronectin, a deposition leading tofibrosis (Park et al, 2001).The mechanism of action of TGF-β in cells of thelung may involve the mitochondria since TGF-β1 can leadto downregulation of Bcl-2 and Bcl-xl, which normallyprevent apoptosis (Lafon et al, 1996; Herrera et al, 2001a).Overexpression of Bcl-2 suppresses the effects of TGF-β(Huang and Chou, 1998). Following exposure to TGF-βthere is also a loss of mitochondrial membrane potential,release of cytochrome-C, and activation of caspase-3(Herrera et al, 2001a). TGF-β1 activates caspase- 3, 8 and9, which precede the loss of mitochondrial membranepotential (Herrera et al, 2001b). Activation of caspase-8results in cleavage of Bid and Bcl-xl, which may lead toan amplification loop resulting in the mitochondrialmediated apoptosis (Zha et al, 2000). Irradiation of murinebone marrow stromal cell line D2XRII in vitro inducesrelease of TGF-β into the culture medium (Greenberger etal, 1996). Co-cultivation of 32D cl 3 cells or subclones1F2 or 2C6 overexpressing MnSOD with irradiated bonemarrow stromal cells resulted in higher levels ofintracellular ROS in the non-irradiated 32D cl 3, 1F2 or2C6 cells compared to cells co-cultivated with nonirradiatedstromal cell lines (Greenberger et al, 1996). TheMnSOD overexpressing subclonal line 1F2 or 2C6 formedmore cobblestone islands on the irradiated stromal cellsthan 32D cl 3 cells (Greenberger et al, 1996). IncreasedMnSOD activity in 1F2 or 2C6 cells may have resulted ina decrease in ROS, thus allowing for greater attachment ofthe MnSOD overexpressing cell lines to the irradiatedstromal cells. Therefore, injections of MnSOD-PL into thelung at day 100 when TGF-β levels are beginning toincrease may inhibit ROS production, and/or stabilize thebronchoalveolar cell or endothelial cell mitochondria,preventing some (but not all) of the late effects ofirradiation damage to the lung. We are currently exploringthis possible mechanism.The present report indicates that a singleadministration of MnSOD-PL 24 hours prior to 20 Gylung irradiation is significantly more effective thanadministration at any of four post-irradiation time pointsranging from 1-100 days after irradiation. We did notevaluate time points between 1 and 80 days as there wasno histopathologic or other evidence to suggest thatinitiation steps in the late organizing alveolitis/fibrosisresponse began prior to 80 days. Our results may helpexplain the dynamics of late irradiation pulmonary injury.One interpretation of the results is that it representsevidence of a pleiotropic effect of ionizing irradiation onseveral cellular and physiologic targets within the lung.Initiation events at the time of irradiation may lead to amultiplicity of effectuating events beginning at around day100 and leading to rapid organizing alveolitis/fibrosis.Prevention of some of the initiating events by MnSOD-PLadministration prior to irradiation may have a significantlygreater effect at reducing the overall outcome compared tomodulation of some of the late effectuating events byMnSOD-PL administration at that time. For example,neutralization of free radical moieties induced byirradiation at day 0 by overexpression of MnSOD at thattime may be a significant early event which impacts onmultiple downstream/delayed effectuating targets(macrophage migration, fibroblast migration into the lung,66

Gene Therapy and Molecular Biology Vol 7, page 67endothelial upregulation of adhesion molecules, and othercomponents of the fibrosis response not yet elucidated). Incontrast, modulation of some of the effectuating events byMnSOD-PL administration at the late time points mayhave a significantly decreased effect in preventing latelesion simply due to the multiplicity of events already inprogress, and that many of these may be unrelated to thefree radical neutralization capacity of MnSOD at that latetime point. The same mechanism explaining a greatereffect of treatment prior to irradiation might also hold truefor anti-apoptotic effects of MnSOD overexpression in themitochondria.The present data also indicate that significantprotective effects afforded by MnSOD-PL administrationprior to irradiation were not significantly further improvedby a second delayed administration. This result maysimply be attributable to the dominant mechanism ofprevention of initiating events compared to effectuatingevents. The present results add support to utilization offractionated inhalation of freeze-dried MnSOD-PL duringcourses of fractionated radiotherapy which would beappropriate to the clinical translational model of normallung irradiation protection in lung cancer patientsreceiving chemoradiotherapy over a 60-day time course.Fractionation experiments currently in progressincorporate twice weekly inhalation of freeze-driedMnSOD-PL by mice receiving 24 fractions of irradiationduring 35 days. The present observation of a lack oftoxicity of a second delayed administration of MnSOD-PLin the present data supports the concept that multi-fractionadministration of this gene therapy technique should notexacerbate and may decrease pulmonary irradiationdamage.AcknowledgmentsThis research has been supported by the NationalInstitutes of Health, Grant #R01-HL-60132ReferencesBellocq A, Azoulay E, Marullo S, Flahault A, Fouqueray B,Philippe C, Cadranel J, Baud L (1999) Reactive oxygen andnitrogen intermediates increase transforming growth factorbeta1 release from human epithelial alveolar cells throughtwo different mechanisms. AJRCMB 21, 128-136.Bowler RP, Crapo JD (2002) Oxidative stress in airways, is therea role for extracellular superoxide dismutase? AJRCCM166, 38-43.Epperly MW, Bray JA, Kraeger S, Swacka R, Engelhardt JF,Travis E, and Greenberger (1998) Prevention of late effectsof irradiation lung damage by manganese superoxidedismutase gene therapy. Gene Ther 5, 196-208.Epperly MW, Bray JA, Esocobar P, Bigbee WL, Watkins S,Greenberger JS (1999a) Overexpression of the humanMnSOD transgene in subclones of murine hematopoieticprogenitor cell line 32D cl 3 decreases irradiation-inducedapoptosis but does not alter G2/M or G1/S phase cell cyclearrest. Radiat Oncol Invest 7, 331-342.Epperly, MW, Bray, JA, Krager, S, Berry, LM, Gooding, W,Engelhardt, JF, Zwacka, R, Travis,EL, and Greenberger, JS(1999b) Intratracheal injection of adenovirus containing thehuman MnSOD transgene protects athymic nude mice fromirradiation-induced organizing alveolitis. Int J Radiat OncolPhys 43, 169-181.Epperly MW, Travis EL, Sikora C, Greenberger JS (1999c)Magnesium superoxide dismutase (MnSOD)plasmid/liposome pulmonary radioprotective gene therapy,Modulation of irradiation-induced mRNA for IL-1, TNF-α,and TGF-β correlates with delay of organizingalveolitis/fibrosis. Biol Blood Bone Marrow Transplant 5,204-214.Epperly MW, Epstein CJ, Travis EL, Greenberger JS (2000a)Decreased pulmonary radiation resistance of manganesesuperoxide dismutase (MnSOD)-deficient mice is correctedby human manganese superoxide dismutaseplasmid/liposome(SOD2-PL) intratracheal gene therapy.Radiat Res 154, 365-374.Epperly MW, Sikora C, Defilippi S, Bray J, Koe G, Liggitt D,Luketich JD, Greenberger JS (2000b) Plasmid/liposometransfer of the human manganese superoxide dismutase(MnSOD) transgene prevents ionizing irradiation-inducedapoptosis in human esophagus organ explant culture. RadiatOncol Invest 90, 128-137.Epperly MW, Gretton JA, DeFilippi SJ, Greenberger JS, SikoraCA, Liggitt D, Koe G (2001a) Modulation of radiationinducedcytokine elevation associated with esophagitis andesophageal stricture by manganese superoxide dismutaseplasmid/liposome(SOD-PL) gene therapy. Radiat Res 155,2-14.Epperly MW, Travis EL, Whitsett JA, Raineri I, Epstein CJ,Greenberger JS (2001b) Overexpression of manganesesuperoxide dismutase (MnSOD) in whole lung or alveolartype II (AT-II) cells of MnSOD transgenic mice does notprovide intrinsic lung irradiation protection. Radiat OncolInvest 96, 11-21.Epperly MW, Guo HL, Gretton JE, Greenberger JS. (2003) Bonemarrow origin of myofibroblasts in irradiation pulmonaryfibrosis. AJRCMB, 29, in press.Epperly MW, Sikora CA, DeFilippi SJ, Gretton JA, Zhan Q,Kufe DW, Greenberger JS (2002) MnSOD inhibitsirradiation-induced apoptosis by stabilization of themitochondrial membrane against the effects of SAP kinasesp38 and Jnk1 translocation. Radiat Res 157, 568-577.Fridovich I (1995) Superoxide radical and superoxidedismutases. Annu Rev Biochem 64, 97-112.Gorbunov NV, Pogue-Geile KL, Epperly MW, Bigbee WL,Draviam R, Day BW, Wald N, Watkins SC, Greenberger JS(2000) Activation of the nitric oxide synthase 2 pathway inthe response of bone marrow stromal cells to high doses ofionizing radiation. Radiat Res 154, 73-86.Gossart S, Cambon C, Orfila C, Seguelas MH, Lepert JC, RamiJ, Carre P, Pipy B (1996) Reactive oxygen intermediates aregulators of TNF-alpha production in rat lung inflammationinduced by silica. J Immun 156, 1540-1548.Greenberger JS, Epperly MW, Zeevi A, Brunson KW, GoltryKL, Pogue-Geile KL, Bray J, Berry L (1996) Stromal cellinvolvement in leukemogenesis and carcinogenesis. In Vivo10, 1-18.Guo HL, et al. (2003) Prevention of irradiation-induced oralcavity mucositis by plasmid/liposome delivery of the humanmanganese superoxide dismutase (MnSOD) transgene.Radiat Res 159, 361-370.Haddad JJ (2002) Redox regulation of pro-inflammatorycytokines and IkappaB-alpha/NF-kappaB nuclear67

Epperly et al: Late injection of MnSOD-PL protects against pulmonary fibrosistranslocation and activation. Biochem Biophys ResCommun 296, 847-856.Herrera B, Alvarez AM, Sanchez A, Fernandez M, Roncero C,Benito M, Fabregat I (2001) Reactive oxygen species (ROS)mediates the mitochondrial-dependent apoptosis induced bytransforming growth factor (beta) in fetal hepatocytes.FASEB J 15, 741-751.Herrera B, Fernandez M, Alvarez AM, Roncero C, Benito M, GilJ, Fabregat I (2001) Activation of caspases occursdownstream from radical oxygen species production, Bcl-xldown-regulation and early cytochrome C release in apoptosisinduced by transforming growth factor beta in rat fetalhepatocytes. Hepatology 34, 548-556.Hirose K, Longo DL, Oppenheim JJ, Matsushima K (1993)Overexpression of mitochondrial manganese superoxidedismutase promotes the survival of tumor cells exposed toIL-1, TNF, selected anticancer drugs and ionizing irradiation.FASEB J 7, 361-368.Huang YL, Chou CK (1998) Bcl-2 blocks apoptotic signal oftransforming growth factor-beta in human hepatoma cells. JBiomed Sci 5, 185-191.Kanai AJ, Zeidel ML, Lavelle JP, Greenberger JS, Birder LA, deGroat WC, Apodaca GL, Meyers SA, Ramage R, EpperlyMW (2002) Manganese superoxide dismutase gene therapyprotects against irradiation-induced cystitis. Am J PhysiolRenal Physiol. 283, 1304-1312.Lafon C, Mathieu C, Guerrin M, Pierre O, Vidal S, Valette A(1996) Transforming growth factor beta 1-induced apoptosisin human ovarian carcinoma cells, protection by theantioxidant N-acetylcysteine and Bcl-2. Cell Growth Diff 7,1095-1104.Langen RC, Schols AM, Kelders MC, Van Der Velden JL,Wouters EF, Janssen-Heininger YM (2002) Tumor necrosisfactor-alpha inhibits myogenesis through redox-dependentand independent pathways. Am J Physiol Cell Physiol 283,714-721.Larios JM, Budhiraja R, Fanburg BL, Thannickal VJ (2001)Oxidative protein cross-linking reactions involving L-tyrosine in transforming growth factor-beta1-stimulatedfibroblasts. J Biol Chem 276, 17437-17441.Lee MW, Park SC, Kim JH, Kim IK, Han KS, Kim KY, LeeWB, Jung YK, Kim SS (2002) The involvement of oxidativestress in tumor necrosis factor (TNF)-related apoptosisinducingligand (TRAIL)-induced apoptosis in HeLa cells.Cancer Letters 182, 75-82.Li JJ, Oberley LW (1997) Overexpression of manganesecontainingsuperoxide dismutase confers resistance to thecytotoxicity of TNF-α and/or hyperthermia. Cancer Res 57,1991-1998.Park SK, Kim J, Seomun Y, Choi J, Kim DH, Han IO, Lee EH,Chung SK, Joo CK (2001) Hydrogen peroxide is a novelinducer of connective tissue growth factor. BiochemBiophys Res Commun 284, 966-971.Quinlan T, Spivack S, Mossman BT (1994) Regulation ofantioxidant enzymes in lung after oxidant injury. EnvironHealth Perspect 102, 79-87.Stickle RL, Epperly MW, Klein E, Bray JA, Greenberger JS(1999) Prevention of irradiation-induced esophagitis byintraesophageal plasmid/liposome delivery of the humanmanganese superoxide dismutase (MnSOD) transgene.Radiat Oncol Invest 7, 204-217.Urano M, Kuroda M, Reynolds R, Oberley TD, St Clair DK(1995) Expression of manganese superoxide dismutasereduces tumor control radiation dose, gene radiotherapy.Cancer Res 55, 2490-2493.Vujaskovic Z, Feng QF, Rabbani ZN, Anscher MS, SamulskiTV, Brizel DM (2002) Radioprotection of lungs byamifostine is associated with reduction in profibrogeniccytokine activity. Radiat Res 157, 656-660.Wong GHW, Elwell JH, Oberley LW, Goeddel DV (1989)Manganese superoxide dismutase is essential for cellularresistance to cytotoxicity of tumor necrosis factor. Cell 58,923-931.Zha J, Weiler S, Oh KJ, Wei MC, Korsmeyer SJ (2000)Posttranslational N-myristoylation of BID as a molecularswitch for targeting mitochondria and apoptosis. Science290, 1761-1765.68

Gene Therapy and Molecular Biology Vol 7, page 69Gene Ther Mol Biol Vol 7, 69-73, 2003Regulation of vascular endothelial growth factor byhypoxiaMini ReviewIlana Goldberg-Cohen*, Nina S Levy, Andrew P LevyTechnion Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel__________________________________________________________________________________*Correspondence: Ilana Goldberg-Cohen, Technion Faculty of Medicine, Haifa, Israel; Tel 011-972-4-8295202; Fax 011-972-4-8514103; email: gilana@tx.technion.ac.ilKey words: VEGF (vascular endothelial growth factor), hypoxia, HuRReceived: 04 June 2003; Accepted: 27 June 2003; electronically published: July 2003SummaryThe past few decades have singled out the growth of new blood vessels, termed angiogenesis, as a key process in thecourse of normal development as well as in pathological disease processes. VEGF, an endothelial cell specificmitogen, is now accepted as a key mediator of angiogenic events and as such may be a powerful tool in manipulatingthe growth of new blood vessels. VEGF expression is regulated to a great extent by hypoxia. The lack of oxygen tosupply a tissue triggers several molecular mechanisms that increase VEGF mRNA transcription, stability andtranslation, and thus upregulate the expression of VEGF protein. This review focuses on the increase in VEGFmRNA stability through its recognition by the RNA binding protein HuR. Binding of HuR to its cognate site on the3´UTR of VEGF mRNA results in a several fold increase in VEGF mRNA stability, possibly due to the masking of anearby binding site for ribonucleases. Mastering the regulatory mechanisms of VEGF expression is of greatimportance for the future manipulation of VEGF and angiogenesis in the disease setting.I. IntroductionThe ability to grow new blood vessels to supply theneeds of a growing tissue is critical in both physiologicalprocesses such as embryogenesis and in pathologicalprocesses that include tumor growth and metastasis.Vascular Endothelial Growth Factor (VEGF), anendothelial cell specific mitogen, (Ferrara and Henzel,1989; Plouet et al, 1989) is a critical mediator in theestablishment of new blood vessels in bothvasculogenesis, the de novo foundation of vascularsystems (Risau, 1997), and angiogenesis, the developmentof new blood vessels from a pre existing network (Risau,1997). The VEGF gene, found on chromosome 6p21(Vincenti et al, 1996), consists of eight exons separated byseven introns and is alternatively spliced to form fivedifferent VEGF isoforms, the most prominent beingVEGF 165 , that differ in length and ability to bind heparin(Houck et al, 1991).Two tyrosine kinase family receptors flt-1(VEFGR1) and flk-1 (VEGFR2) were identified as VEGFreceptors (de Vries et al, 1992; Terman et al, 1992). Theyhave a similar structure of seven immunoglobulin-likeloops in their extracellular domain, a transmembraneregion and a tyrosine kinase consensus sequence (Shibuyaet al, 1990; Terman et al, 1991). The two receptors inducedifferent signal transduction cascades when activated andthus mediate separate responses to VEGF (Waltenberger etal, 1994; Yoshida et al, 1996). A third receptor familyunrelated to the receptor families described above, theneuropillin receptor family, binds mainly to VEGF 165 andits members are thought to act as coreceptors (Soker et al,1996).II. Regulation of VEGF geneexpressionIn light of its potency and importance in vasculaturedevelopment, VEGF itself is carefully regulated to providefor the appropriate amount of VEGF at the appropriatetime. Growth factors, cytokines and other extracellularmolecules such as PDGF, TNFα and others influenceangiogenesis by governing VEGF expression (Deroanne etal, 1997; Finkenzeller et al, 1997; Frank et al, 1995;Pertovaara et al, 1994; Ryuto et al, 1996). Oncogenes andtumor suppressor genes also play a role in VEGFmodulation as in the case of the von Hipple Lindau tumorsuppressor gene whose absence or inactivationdramatically increases VEGF expression (Iliopoulos et al,1996; Maher and Kaelin, 1997; Mukhopadhyay et al,1997).69

Goldberg-Cohen et al: Regulation of vascular endothelial growth factor by hypoxiaOne of the key factors, which controls VEGFexpression, is oxygen tension. A growing mass such as anembryo or a tumor is in need of oxygen when it can nolonger rely on diffusion to sustain itself. The lack ofoxygen, termed hypoxia, induces a cascade of events,which increase VEGF expression and ultimately thegrowth of new blood vessels.III. Hypoxic regulation of VEGFHypoxia increases VEGF expression by severalmechanisms which act at the level of mRNA transcription,stabilization and translation.A. Upregulation of VEGF mRNAtranscriptionVEGF transcription, as well as that of several otherhypoxia inducible genes such as the glycolytic enzymesand erythropoietin, is increased with hypoxia. Most ofthese genes have Hypoxia Response Elements (HREs) thatbind a heterodimeric helix-loop-helix transcription factorcalled Hypoxia Inducible Factor 1 (HIF-1) (Wang andSemenza, 1995; Semenza et al, 1996). HIF-1 binds to itsrecognition site on VEGF 5´ promoter and together withother trans acting factors mediates the increase in VEGFtranscription with hypoxia. Several other transcriptionfactors such as AP-1 and CREB also appear to influencethe hypoxic induction of VEGF transcription most likelyvia direct interaction with HIF-1 (Abate et al, 1990).B. Hypoxic regulation of VEGF mRNAtranslationVEGF mRNA has an unusually long 5´ untranslatedregion (5´ UTR) containing stable secondary structuresand a short in-frame initiation and termination codons.This significantly inhibits initiation of protein synthesis bythe classical model of the cap-dependant ribosomescanning. VEGF mRNA can also be translated in a capindependentmanner through an Internal Ribosome EntrySite (IRES). Under hypoxic conditions, and otherconditions of stress, cap dependant translation is reduced.The presence of an IRES site allows the translation ofVEGF and other IRES containing mRNAs to continue(Akiri et al, 1998; Stein et al, 1998).C. Hypoxic stabilization of VEGF mRNAThe half life of VEGF mRNA, like that of severalother cytokine and oncogene mRNAs, is very short.Increased stability of a mRNA renders it more accessibleto the translational machinery and thus increases theamount of its gene product. Shaw and Kamen (1986)reported a considerable decrease in the stability of β-globin mRNA when an AU-rich element (ARE) from the3´UTR of GM-CSF was introduced 3´ to the β-globin gene(Shaw and Kamen, 1986). Further studies indicated thatthe pentameric sequence AUUUA is necessary but notsufficient to induce degradation of mRNAs and mutationsthat specifically interrupted this pentameric sequenceabolished the destabilizing properties of the entire AU richelement (Akashi et al, 1994; Chen et al, 1994).The degradation of mRNAs containing AU richelements in their 3´ UTR is facilitated by the binding oftrans-acting factors which may promote exonuclease aswell as site-specific endonucleolytic events.Tristetraproline (TTP) and AUF1 are two such trans-actingRNA binding proteins that bind AU rich elements anddestabilize the mRNAs carrying these sequences (Brewer,1991; Carballo et al, 1998; Lai and Blackshear, 2001).While AU rich elements allow for the rapiddegradation of mRNAs they also appear to be able to bindtrans-acting factors that act to increase mRNA stabilityunder certain circumstances as discussed below for VEGFmRNA. Like GM-CSF, the 3´UTR of VEGF mRNAconsists of multiple AU rich elements that render itvulnerable to rapid degradation. However, under hypoxicconditions, RNA binding proteins recognize and bind totheir cognate AU rich sites on the 3´UTR of VEGFmRNA, increasing its stability and thus its expressionseveral fold.IV. HuRA prominent member of the ARE binding proteinfamily that acts to increase mRNA stability with hypoxiais HuR. This RNA binding protein belongs to theEmbryonic Letal Abnormal Visual (ELAV) protein familyfirst described in Drosophila (Robinow et al, 1988). Thefounding member, ELAV, is expressed immediatelyfollowing neuroblast differentiation into neurons and isinvolved in the subsequent neuronal differentiation andmaintenance (Robinow and White, 1991; Campos et al,1985). Further studies identified four human homologuesthat were characterized as tumor antigens (Szabo et al,1991). Three of the human ELAV-like proteins areexpressed solely in terminal differentiation of neurons andneuroendocrine tumors (King et al, 1994; Barami et al,1995; Jain et al, 1997) while the fourth, termed HuR, isfound in proliferating cells and in tumors throughout thebody (Ma et al, 1996). Classification as tumor antigensgave rise to extensive research into the essence of theirRNA binding properties and resulted in the identificationof three highly conserved RNA recognition motifs. Two ofthe RNA recognition motifs are in tandem separated fromthe third by a basic segment (Kenan et al, 1991).Subsequent studies confirmed that the ELAV-like proteinsare prone to bind AU rich elements present in the 3´UTRsof mRNAs as well as to their polyA tails, which maycontribute to their ability to protect mRNAs fromribonuclease degradation (Ma et al, 1997).As discussed above, HuR, the only ELAV familymember not restricted to the nervous system but ratherexpressed throughout the body, is involved in increasingVEGF mRNA stability with hypoxia by binding to an AUrich recognition site on the VEGF mRNA 3´UTR. A studyinvestigating the binding of HuR to c-fos mRNAidentified a high affinity site containing three AU richmotifs AUUUA, AUUUUA, and AUUUUUA, all ofwhich are critical for maximal binding (Ma et al, 1996).The requirement for a nonspecific number of U residues in70

Gene Therapy and Molecular Biology Vol 7, page 71the target sequence points to an inclination towardsbinding a particular structure rather than a primarysequence (Kim et al, 1974).V. HuR binding site on VEGF mRNA3´UTRThe 3´UTR of VEGF mRNA contains long stretchesof AU residues, which confer rapid mRNA degradationthrough the binding of ribonucleases to the AU richelements. However, under hypoxic conditions, these AUrich elements allow the binding of RNA binding proteinssuch as HuR, which block binding of ribonucleases andthus increase the stability of the VEGF mRNA and VEGFexpression under hypoxia (Stein et al, 1995;; Damert et al,1997; Claffey et al, 1998).Attempts to characterize the minimal binding site ofHuR on the 3´UTR of VEGF mRNA that is still able toconfer increased stability with hypoxia were carried out inour lab and resulted in the identification of a 40 base pairelement at position 1285 of the 3´UTR of VEGF mRNA(nucleotides 1285-1325 of the VEGF 3´UTR, GenBankaccession number U22372)(Goldberg-Cohen et al, 2002).Transient cotransfection of a vector carrying the 40 basepair element positioned 3´ to the luciferase reporter geneand a plasmid overexpressing HuR showed an increase inreporter activity that correlated with an increase incotransfected HuR. Furthermore, when incubatedovernight under hypoxic conditions, cells transfected withthe reporter vector containing the 40 base pair element hadgreater reporter activity than cells transfected with reportervector alone. These observations were confirmed in an invitro model where the stability of an RNA containing the40 base pair element was shown to be increased in anRNA degradation assay in the presence of HuR.RNase T1 and lead protection assays mapped the HuRbinding site to nucleotides 23-39 of the 40 base pairelement. Deletion of the HuR specific binding sitedramatically reduced reporter activity in the transienttransfection assay (Goldberg-Cohen et al, 2002).In view of the ability of HuR to bind and stabilizeVEGF mRNA with hypoxia, a model was constructed. Inthis model, under normoxic conditions VEGF mRNA isextremely unstable by virtue of its recognition byribonucleases that bind the VEGF mRNA 3´UTR andcause its rapid degradation. This labile character of VEGFmRNA can be overcome under hypoxic conditionsthrough the binding of HuR to its recognition site on the3´UTR of VEGF mRNA rendering it less vulnerable toribonuclease digestion.The hypoxic regulation of HuR is not completelyunderstood. Under normoxic conditions the bulk of HuR issequestered in the nucleus. Under hypoxia, cytoplasmicHuR levels increase with no apparent increase in totalHuR levels (Levy et al, 1998). This would suggestnucleocytoplasmic transport of HuR and indeed it wasreported that HuR possess a shuttling signal termed HuRNucleocytoplasmic Shuttling sequence (HNS) that may besignificant to the process (Fan and Steitz, 1998). It remainsto be investigated whether HuR binds VEGF mRNA in thenucleus and is transported to the cytoplasm as a complexor whether HuR is first transported to the cytoplasm whereit binds and stabilizes VEGF.VI. PerspectivesThe manipulation of angiogenic events as atherapeutic tool marks the dawn of a new era in treatingnumerous afflictions in medicine today. Whetherangiogenesis is stimulated to feed the ischemic heart oranti angiogenic agents are applied to prevent tumor growthand metastases, understanding and controlling angiogeneicfactors will be critical to the outcome. As a majorparticipant in the angiogenic process, VEGF has been thetarget of a worldwide research effort in the past twodecades. Cancer research has focused on the importance ofVEGF in tumor growth and metastases and the findingssupport this notion. VEGF is now recognized as a keyregulator in tumor induced angiogenesis and several antiVEGF treatments that utilize antibodies against VEGF andVEGF receptors have successfully blocked tumor growthin mouse models (Kim et al, 1993; Ferrara, 1999; Gerberet al, 2000; Lee et al, 2000). Cardiovascular research hasalso benefited from the advances in VEGF research. Genetherapy techniques that augment VEGF expression in thediseased heart or ischemic leg are currently under way.The success of these VEGF treatments will depend to alarge extent on our understanding the complex regulationof VEGF. In addition, deciphering the molecularmechanisms that govern the expression of VEGF andother angiogenic factors will give rise to new possibilitiesnot only in the ever growing field of vascular biology butalso in the vast area of embryonic development and tissuetransplantation.ReferencesAbate C, Luk D, Gagne E, Roeder RG, Curran T (1990) Fos andjun cooperate in transcriptional regulation via heterologousactivation domains. Mol Cell Biol 10, 5532-5535.Akashi M, Shaw G, Hachiya M, Elstner E, Suzuki G, Koeffler P(1994) Number and location of AUUUA motifs: role inregulating transiently expressed RNAs. Blood 83, 3182-3187.Akiri G, Nahari D, Finkelstein Y, Le SY, Elroy-Stein O, Levi BZ(1998) Regulation of vascular endothelial growth factor(VEGF) expression is mediated by internal initiation oftranslation and alternative initiation of transcription.Oncogene 17, 227-236.Barami K, Iversen K, Furneaux H, Goldman SA (1995) Huprotein as an early marker of neuronal phenotypicdifferentiation by subependymal zone cells of the adultsongbird forebrain. J Neurobiol 28, 82-101.Brewer G (1991) An A + U-rich element RNA-binding factorregulates c-myc mRNA stability in vitro. Mol Cell Biol 11,2460-2466.Campos AR, Grossman D, White K (1985) Mutant alleles at thelocus elav in Drosophila melanogaster lead to nervoussystem defects. A developmental-genetic analysis. JNeurogenet 2, 197-218.Carballo E, Lai WS, Blackshear PJ (1998) Feedback inhibition ofmacrophage tumor necrosis factor-alpha production bytristetraprolin. Science 281, 1001-1005.71

Goldberg-Cohen et al: Regulation of vascular endothelial growth factor by hypoxiaChen CY, Chen TM, Shyu AB (1994) Interplay of twofunctionally and structurally distinct domains of the c-fosAU-rich element specifies its mRNA-destabilizing function.Mol Cell Biol 14, 416-426.Claffey KP, Shih SC, Mullen A, Dziennis S, Cusick JL, AbramsKR, et al (1998) Identification of a human VPF/VEGF 3'untranslated region mediating hypoxia-induced mRNAstability. Mol Biol Cell 9, 469-481.Damert A, Machein M, Breier G, Fujita MQ, Hanahan D, RisauW, et al (1997) Up-regulation of vascular endothelial growthfactor expression in a rat glioma is conferred by two distincthypoxia-driven mechanisms. Cancer Res 57, 3860-3864.De Vries C, Escobedo JA, Ueno H, Houck K, Ferrara N,Williams LT (1992) The fms-like tyrosine kinase, a receptorfor vascular endothelial growth factor. Science 255, 989-991.Deroanne CF, Hajitou A, Calberg-Bacq CM, Nusgens BV,Lapiere CM (1997) Angiogenesis by fibroblast growth factor4 is mediated through an autocrine up-regulation of vascularendothelial growth factor expression. Cancer Res 57, 5590-5597.Fan XC, Steitz JA (1998) HNS, a nuclear-cytoplasmic shuttlingsequence in HuR. Proc Natl Acad Sci U.S.A 95, 15293-15298.Ferrara N (1999) Role of vascular endothelial growth factor inthe regulation of angiogenesis. Kidney Int 56, 794-814.Ferrara N, Henzel WJ (1989) Pituitary follicular cells secrete anovel heparin-binding growth factor specific for vascularendothelial cells. Biochem Biophys Res Commun 161, 851-858.Finkenzeller G, Sparacio A, Technau A, Marme D, Siemeister G(1997) Sp1 recognition sites in the proximal promoter of thehuman vascular endothelial growth factor gene are essentialfor platelet-derived growth factor-induced gene expression.Oncogene 15, 669-676.Frank S, Hubner G, Breier G, Longaker MT, Greenhalgh DG,Werner S (1995) Regulation of vascular endothelial growthfactor expression in cultured keratinocytes. Implications fornormal and impaired wound healing. J Biol Chem 270,12607-12613.Gerber HP, Kowalski J, Sherman D, Eberhard DA, Ferrara N.(2000) Complete inhibition of rhabdomyosarcoma xenograftgrowth and neovascularization requires blockade of bothtumor and host vascular endothelial growth factor. CancerRes. 60, 6253-6258.Goldberg-Cohen I, Furneauxb H, Levy AP (2002) A 40-bp RNAelement that mediates stabilization of vascular endothelialgrowth factor mRNA by HuR. J Biol Chem 277, 13635-13640.Houck KA, Ferrara N, Winer J, Cachianes G, LI B, Leung DW(1991) The vascular endothelial growth factor family:identification of a fourth molecular species andcharacterization of alternative splicing of RNA. MolEndocrinol 5, 1806-1814.Iliopoulos O, Levy AP, Jiang C, Kaelin WGJ, Goldberg MA(1996) Negative regulation of hypoxia-inducible genes bythe von Hippel-Lindau protein. Proc Natl Acad Sci U.S.A93, 10595-10599.Jain RG, Andrews LG, Mcgowan KM, Pekala PH, Keene JD(1997) Ectopic expression of Hel-N1, an RNA-bindingprotein, increases glucose transporter (GLUT1) expression in3T3-L1 adipocytes. Mol Cell Biol 17, 954-962.Kenan DJ, Query CC, Keene JD (1991) RNA recognition:towards identifying determinants of specificity. TrendsBiochem Sci 16, 214-220.Kim KJ, Li B, Winer J, Armanini M, Gillett N, Phillips HS, et al(1993) Inhibition of vascular endothelial growth factorinducedangiogenesis suppresses tumour growth in vivo.Nature 362, 841-844.Kim SH, Suddath FL, Quigley GJ, Mcpherson A, Sussman JL,Wang AH, et al (1974) Three-dimensional tertiary structureof yeast phenylalanine transfer RNA. Science 185, 435-440.King PH, Levine TD, Fremeau RTJ, Keene JD (1994)Mammalian homologs of Drosophila ELAV localized to aneuronal subset can bind in vitro to the 3' UTR of mRNAencoding the Id transcriptional repressor. J Neurosci 14,1943-1952.Lai WS, Blackshear PJ. (2001) Interactions of CCCH zinc fingerproteins with mRNA: tristetraprolin-mediated AU-richelement-dependent mRNA degradation can occur in theabsence of a poly(A) tail. J Biol Chem 276, 23144-23154.Lee CG, Heijn M, Di Tomaso E, Griffon-Etienne G,Ancukiewicz M, Koike C, et al (2000) Anti-Vascularendothelial growth factor treatment augments tumor radiationresponse under normoxic or hypoxic conditions. Cancer Res60, 5565-5570.Levy NS, Chung S, Furneaux H, Levy AP (1998) Hypoxicstabilization of vascular endothelial growth factor mRNA bythe RNA-binding protein HuR. J Biol Chem 273, 6417-6423.Ma WJ, Cheng S, Campbell C, Wright A, Furneaux H (1996)Cloning and characterization of HuR, a ubiquitouslyexpressed Elav-like protein. J Biol Chem 271, 8144-8151.Ma WJ, Chung S, Furneaux H (1997) The Elav-like proteins bindto AU-rich elements and to the poly(A) tail of mRNA.Nucleic Acids Res 25, 3564-3569.Maher ER, Kaelin WGJ (1997) von Hippel-Lindau disease.Medicine 76, 381-391.Mukhopadhyay D, Knebelmann B, Cohen HT, Ananth S,Sukhatme VP (1997) The von Hippel-Lindau tumorsuppressor gene product interacts with Sp1 to repressvascular endothelial growth factor promoter activity. MolCell Biol 17, 5629-5639.Pertovaara L, Kaipainen A, Mustonen T, Orpana A, Ferrara N,Saksela O, et al (1994) Vascular endothelial growth factor isinduced in response to transforming growth factor-beta infibroblastic and epithelial cells. J Biol Chem 269, 6271-6274.Plouet J, Schilling J, Gospodarowicz D (1989) Isolation andcharacterization of a newly identified endothelial cellmitogen produced by AtT-20 cells. EMBO J 8, 3801-3806.Risau W (1997) Mechanisms of angiogenesis. Nature 386, 671-674.Robinow S, Campos AR, YAO KM, White K (1988) The elavgene product of Drosophila, required in neurons, has threeRNP consensus motifs. Science 242, 1570-1572.Robinow S, White K (1991) Characterization and spatialdistribution of the ELAV protein during Drosophilamelanogaster development. J Neurobiol 22, 443-461.Ryuto M, Ono M, Izumi H, Yoshida S, Weich HA, Kohno K, etal (1996) Induction of vascular endothelial growth factor bytumor necrosis factor alpha in human glioma cells. Possibleroles of SP-1. J Biol Chem 271, 28220-28228.Semenza GL, Jiang BH, Leung SW, Passantino R, Concordet JP,Maire P, et al (1996) Hypoxia response elements in thealdolase A, enolase 1, and lactate dehydrogenase A genepromoters contain essential binding sites for hypoxiainduciblefactor 1. J Biol Chem 271, 32529-32537.72

Gene Therapy and Molecular Biology Vol 7, page 73Shaw G, Kamen R (1986) A conserved AU sequence from the 3'untranslated region of GM-CSF mRNA mediates selectivemRNA degradation. Cell 46, 659-667.Shibuya M, Yamaguchi S, Yamane A, Ikeda T, Tojo A,Matsushime H, et al (1990) Nucleotide sequence andexpression of a novel human receptor-type tyrosine kinasegene (flt) closely related to the fms family. Oncogene 5, 519-524.Soker S, Fidder H, Neufeld G, Klagsbrun M (1996)Characterization of novel vascular endothelial growth factor(VEGF) receptors on tumor cells that bind VEGF165 via itsexon 7-encoded domain. J Biol Chem 271, 5761-5767.Stein I, Itin A, EinaT P, Skaliter R, Grossman Z, Keshet E (1998)Translation of vascular endothelial growth factor mRNA byinternal ribosome entry: implications for translation underhypoxia. Mol Cell Biol 18, 3112-3119.Stein I, Neeman M, Shweiki D, Itin A, Keshet E (1995)Stabilization of vascular endothelial growth factor mRNA byhypoxia and hypoglycemia and coregulation with otherischemia-induced genes. Mol Cell Biol 15, 5363-5368.Szabo A, Dalmau J, Manley G, Rosenfeld M, Wong E, Henson J,et al (1991) HuD, a paraneoplastic encephalomyelitisantigen, contains RNA-binding domains and is homologousto Elav and Sex-lethal. Cell 67, 325-333.Terman BI, Carrion ME, Kovacs E, Rasmussen BA, Eddy RL,Shows TB (1991) Identification of a new endothelial cellgrowth factor receptor tyrosine kinase. Oncogene 6, 1677-1683.Terman BI, Dougher-Vermazen M, Carrion ME, Dimitrov D,Armellino DC, Gospodarowicz D, et al (1992) Identificationof the KDR tyrosine kinase as a receptor for vascularendothelial cell growth factor. Biochem Biophys ResCommun 187, 1579-1586.Vincenti V, Cassano C, Rocchi M, Persico G (1996) Assignmentof the vascular endothelial growth factor gene to humanchromosome 6p21.3. Circulation 93, 1493-1495.Waltenberger J, Claesson-Welsh L, Siegbahn A, Shibuya M,Heldin CH (1994) Different signal transduction properties ofKDR and Flt1, two receptors for vascular endothelial growthfactor. J Biol Chem 269, 26988-26995.Wang GL, Semenza GL (1995) Purification and characterizationof hypoxia-inducible factor 1. J Biol Chem 270, 1230-1237.Yoshida A, Anand-Apte B, Zetter BR (1996) Differentialendothelial migration and proliferation to basic fibroblastgrowth factor and vascular endothelial growth factor.Growth Factors 13, 57-64.73

Goldberg-Cohen et al: Regulation of vascular endothelial growth factor by hypoxia74

Gene Therapy and Molecular Biology Vol 7, page 75Gene Ther Mol Biol Vol 7, 75-89, 2003Gene therapy antiproliferative strategies againstcardiovascular diseaseReview ArticleMarisol Gascón-Irún, Silvia M. Sanz-González and Vicente Andrés*Laboratory of Vascular Biology, Department of Molecular and Cellular Pathology and Therapy, Instituto de Biomedicinade Valencia, Spanish Council for Scientific Research (CSIC), Valencia, Spain__________________________________________________________________________________*Correspondence: Vicente Andrés, Ph.D; Laboratory of Vascular Biology, Department of Molecular and Cellular Pathology andTherapy, Instituto de Biomedicina de Valencia, Spanish Council for Scientific Research (CSIC), C/ Jaime Roig, 11 46010 Valencia(SPAIN); Tel.: +34-963391752 (office), +34-963391751 (lab), Fax: +34-963690800; e-mail: vandres@ibv.csic.esKey words: atherosclerosis, restenosis, bypass graft failure, cell cycle, gene therapyList of abbreviations: apoE, apolipoprotein E; AP-1, activator protein-1; BrdU, 5-bromodeoxyuridine; CDK, cyclin-dependent kinase;CKI, CDK inhibitory protein; EC, endothelial cell; ERK, extracellular signal-regulated kinase; IVUS, intravascular ultrasound; JNK, c-jun NH 2 -terminal protein kinase; MAPK, mitogen-activated protein kinase; ODN, oligodeoxynucleotide; PCNA, proliferating cellnuclear antigen; PDGF, platelet-derived growth factor; pRb, retinoblastoma protein; PTCA, percutaneous transluminal angioplasty;SAPK, stress-activated protein kinase; TGF-β, transforming growth factor-β; VSMC, vascular smooth muscle cell.Received: 17 June 2003; Accepted: 27 June 2003; electronically published: July 2003SummaryExcessive cellular proliferation is thought to contribute to the pathogenesis of several forms of cardiovasculardisease (e. g., atherosclerosis, restenosis after angioplasty, and vessel bypass graft failure). Therefore, candidatetargets for the treatment of these disorders include cell cycle regulatory factors, such as cyclin-dependent kinases(CDKs), cyclins, CDK inhibitory proteins (CKIs), tumor suppressors, growth factors and their receptors, andtranscription factors. Importantly, animal models of atherosclerosis have demonstrated an inverse correlationbetween neointimal cell proliferation and atheroma size, suggesting that excessive cell growth prevails at the onsetof atherogenesis. Cell growth may also predominate at the onset of human atherosclerosis. Thus, given that affectedhumans often exhibit advanced atherosclerotic plaques when first diagnosed, the potential benefit ofantiproliferative strategies for the treatment of atherosclerosis in clinic is doubtful. The antiproliferativeapproaches used so far in the setting of vascular obstructive disease have focused on restenosis and graftatherosclerosis, during which neointimal hyperplasia is spatially localized and develops over a short period of time(typically 2-12 months). Vascular interventions, both endovascular and open surgical, allow minimally invasive,easily monitored gene delivery. Thus, gene therapy strategies are emerging as an attractive approach for thetreatment of vascular proliferative disease. In this review, we will discuss the use of gene therapy strategies againstcellular proliferation in animal models and clinical trials of cardiovascular disease.I. IntroductionLarge-scale clinical trials conducted over the lastdecades have allowed the identification of independentrisk factors that increase the prevalence and severity ofatherosclerosis (e. g., hypercholesterolemia, hypertension,smoking). Cardiovascular risk factors initiate andperpetuate an inflammatory response within the injuredarterial wall that promotes the development ofatherosclerotic plaques (Ross, 1999; Lusis, 2000; Dzau etal, 2002; Steinberg, 2002) (Figure 1). Chemokines andcytokines secreted by leukocytes that accumulate withinthe injured arterial wall promote their own proliferation, aswell as the growth and migration of the underlyingvascular smooth muscle cells (VSMCs) (Figure 2). Thisinflammatory response also plays a critical role duringrestenosis after angioplasty and graft atherosclerosis.Thus, understanding the molecular mechanisms thatcontrol hyperplastic growth of vascular cells should helpdevelop novel therapeutic strategies for the treatment ofvascular obstructive disease.Although arterial cell proliferation occurs in animalmodels during all phases of atherogenesis (Ross, 1999;Díez-Juan and Andrés, 2001; Cortés et al, 2002), studieswith hyperlipidemic rabbits have shown an inversecorrelation between atheroma size and cellularproliferation within the atheromatous plaque (Spraragen etal, 1962; McMillan and Stary, 1968; Rosenfeld and Ross,1990). Experimental angioplasty is also characterized by75

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular diseaseabundant proliferation of VSMCs, followed by thereestablishment of the quiescent phenotype, typicallywithin 2-4 weeks (Bauters and Isner, 1997; Libby andTanaka, 1997; Andrés, 1998). These animal studiessuggest that vascular cell proliferation prevails at the onsetof atherogenesis and restenosis.Figure 1. Neointimal lesion development in response to cardiovascular risk factors and mechanical injury. Exposure of the arterialwall to cardiovascular risk factors and mechanical injury leads to endothelial damage. Recruitment of circulating leukocytes is promotedby the expression of adhesion molecules by the injured endothelial cells. Neointimal leukocites release a plethora of cytokines andchemokines that initiate and perpetuate an inflammatory response, which activates signal transduction pathways and transcription factorsthat promote the hyperplastic growth of the lesion. Accumulation of noncellular material also contributes to atheroma development.Figure 2. Early atherogenesis is associated with abundant cell proliferation within the arterial wall. Immunohistochemical analysis ofaortic arch cross-section of male New Zealand rabbits fed control chow or a cholesterol-rich diet for 2 months. Animals were injectedwith 5-bromodeoxyuridine (BrdU) prior to sacrifice. Specimens were incubated with anti-BrdU and anti-RAM11 antibodies to monitorcell proliferation and to identify macrophages, respectively (Cortés et al, 2002). Arrowheads indicate the internal elastic lamina. Notelack of atherosclerosis and undetectable immunoreactivity for BrdU and RAM11 within the aortic arch of control rabbits. In contrast,prominent fatty streaks enriched in lipid-laden macrophages are seen in cholesterol-fed animals. Some macrophages are also detectedwithin the media. Abundant BrdU immunoreactivity demonstrates a high proliferative activity, particularly within the atheroscleroticlesion. All photomicrographs are at the same magnification.76

Gene Therapy and Molecular Biology Vol 7, page 77Expression of proliferation markers in humanprimary atheromatous plaques and restenotic lesions hasbeen well documented (Essed et al, 1983; Gordon et al,1990; Burrig, 1991; Nobuyoshi et al, 1991; Katsuda et al,1993; Kearney et al, 1997; O'Brien et al, 1993, 2000;Rekhter and Gordon, 1995; Wei et al, 1997; Orekhov et al,1998; Tanner et al, 1998; Veinot et al, 1998). However,controversy exists regarding the magnitude of theproliferative response, ranging from a very low index ofcell proliferation (Gordon et al, 1990; Katsuda et al, 1993;O'Brien et al, 1993; 2000; Rekhter and Gordon, 1995;Veinot et al, 1998) to abundance of dividing cells (Essedet al, 1983; Nobuyoshi et al, 1991; Pickering et al, 1993;Kearney et al, 1997). Aside from methodological issues (e.g., differences in the fixatives used for tissue preservation,antigen accessibility, diversity of proliferation markersanalyzed in these studies), some of the reported variancewith regard to the issue of cell proliferation might relate todifferences in the arteries being analyzed (i. e., peripheral,coronary and carotid arteries) and variance in the stage ofatherogenesis at the time of tissue harvesting (Isner, 1994).The cell types that undergo cell proliferation withinhuman atherosclerotic tissue include VSMCs, leukocytesand endothelial cells (ECs) (Gordon et al, 1990; Burrig,1991; Katsuda et al, 1993; O'Brien et al, 1993; Rekhterand Gordon, 1995; Orekhov et al, 1998; Veinot et al,1998). Histological examination in 20 patients undergoingantemortem coronary angioplasty revealed that the extentof intimal proliferation was significantly greater in lesionswith evidence of medial or adventitial tears than in lesionswith no or only intimal tears (Nobuyoshi et al, 1991).Human carotid artery primary atherosclerotic tissueretrieved by endarterectomy surgery displayed greaterproliferative activity in the intimal lesion versus theunderlying media (Rekhter and Gordon, 1995). Moreover,monocyte/macrophage proliferation predominated in theintima (46% versus 9.7% α-actin immunoreactiveVSMCs, 14.3% ECs, 13.1% T lymphocytes), whereasVSMC proliferation prevailed in the media (44.4% versus20% ECs, 13.0% monocyte/macrophages, and 14.3% Tlymphocytes). It is also noteworthy that cell proliferationin human peripheral and coronary ateries is greater inrestenotic versus primary lesions (O'Brien et al, 1993;2000; Pickering et al, 1993). Furthermore, culturedVSMCs from human advanced primary stenosingdisclosed lower proliferative capacity than cells from freshrestenosing lesions (Dartsch et al, 1990). Thus, similar tothe situation in animal models, proliferation during humanatherosclerosis and restenosis might peak at the onset ofthese pathologies and then progressively decline.Cell cycle progression is controlled by several cyclindependentkinases (CDKs) that associate with regulatorycyclins (Morgan, 1995) (Figure 3). Active CDK/cyclinholoenzymes hyperphosphorylate the retinoblastomaprotein (pRb) and the related pocket proteins p107 andp130 from mid G1 to mitosis. Phosphorylation of pRb andrelated pocket proteins contributes to the transactivation ofgenes with functional E2F-binding sites, including severalgrowth and cell-cycle regulators (i.e., c-myc, pRb, cdc2,cyclin E, cyclin A), and genes encoding proteins that arerequired for nucleotide and DNA biosynthesis (i. e., DNApolymerase α, histone H2A, proliferating cell nuclearantigen, thymidine kinase) (Dyson, 1998; Lavia andJansen-Durr, 1999; Stevaux and Dyson, 2002). Interactionof CDK/cyclins with CDK inhibitory proteins (CKIs)attenuates CDK activity and promotes growth arrest(Philipp-Staheli et al, 2001). CKIs of the Cip/Kip family(p21 Cip1 , p27 Kip1 and p57 Kip2 ) bind to and inhibit a widespectrum of CDK/cyclin holoenzymes, while members ofthe Ink4 family (p16 Ink4a , p15 Ink4b , p18 Ink4c , p19 Ink4d ) arespecific for cyclin D-associated CDKs.Figure 3. Control of mammalian cell cycle by CDK/cyclin holoenzyme and growth suppresssors of the CKI family. Sequentialactivation of specific CDK/cyclin complexes leads to progression through the different phases of the cell cycle. Inhibitory proteins of theCKI family (Cip/Kip and Ink4) inhibit CDK/cyclin activity.77

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular diseaseMitogenic and antimitogenic stimuli affect the rates ofsynthesis and degradation of CKIs, as well as theirredistribution among different CDK/cyclin pairs (Philipp-Staheli et al, 2001). For example, p27 Kip1 promotes theassembly of CDK4/cyclin D complexes by binding tothem, thus facilitating CDK2/cyclin E activation throughG1/S phase.VSMC proliferation in the balloon-injured rat carotidartery is associated with a temporally and spatiallycoordinated expression of CDKs and cyclins (Wei et al,1997; Braun-Dullaeus et al, 2001). Importantly,augmented expression of these factors is associated withan increase in their kinase activity (Abe et al, 1994; Wei etal, 1997), demonstrating the assembly of functionalCDK/cyclin holoenzymes in the injured arterial wall.Expression of CDK2 and cyclin E was also detected inhuman VSMCs within atherosclerotic and restenotic tissue(Kearney et al, 1997; Wei et al, 1997; Ihling et al, 1999),suggesting that induction of positive cell-cycle controlgenes is a hallmark of vascular proliferative disease inhuman patients.In the following sections, we will discuss the use ofgene therapy strategies targeting cellular proliferation inpreclinical (Table 1) and clinical studies (Table 2) relatedto cardiovascular disease.Table 1: Attenuation of neointimal thickening by antiproliferative gene therapy approaches in animal models of vascularproliferative disease.Strategy Target gene Animal model Ref.Antisense (ODN)CDK2 Balloon angioplasty (rat) Abe et al, 1994; Morishita et al, 1994aCDC2 Balloon angioplasty (rat) Abe et al, 1994; Morishita et al, 1994bCyclin B1 Balloon angioplasty (rat) Morishita et al, 1994bCDC2/PCNA Graft arteriosclerosis (rabbit, rat) Mann et al, 1995; Miniati et al, 2000CDC2/PCNA Balloon angioplasty (rat) Morishita et al, 1993CDK2 Graft arteriosclerosis (mouse) Suzuki et al, 1997c-myb * Balloon angioplasty (pig, rat) Simons et al, 1992; Gunn et al, 1997c-myc * Balloon angioplasty (rat, pig, rabbit) Bennett et al, 1994a; Shi et al, 1994b;Kipshidze et al, 2001, 2002c-myc * Graft arteriosclerosis (pig) Mannion et al, 1998PDGFβ receptor Balloon angioplasty (rat) Cohen-Sacks et al, 2002Antisense (retrovirus) Cyclin G1 Balloon angioplasty (rat) Zhu et al, 1997Ribozyme‘Decoy’ ODNOveexpression ofgrowth suppressorsOverexpression ofdominant-negativePCNA Stent (pig) Frimerman et al, 1999TGF-β1 Balloon angioplasty (rat) Yamamoto et al, 2000PDGF-A Balloon angioplasty (rat) Kotani et al, 200312-lipoxygenase Balloon angioplasty (rat) Gu et al, 2001E2F Balloon angioplasty (rat, pig) Morishita et al, 1995; Ahn et al, 2002a;Nakamura et al, 2002E2FAP-1Graft arteriosclerosis (rabbit, mouse,monkey)Balloon angioplasty (rat, rabbit,minipig)Mann et al, 1997; Kawauchi et al, 2000;Ehsan et al, 2001Ahn et al, 2002b; Buchwald et al, 2002; Kumeet al, 2002p21 Cip1 Balloon angioplasty (rat, mouse, pig) Chang et al, 1995a; Yang et al, 1996; Ueno etal, 1997a; Condorelli et al, 2001;p21 Cip1 Graft arteriosclerosis (rabbit) Bai et al, 1998p27 Kip1 Balloon angioplasty (rat, pig) Chen et al, 1997; Tanner et al, 2000pRb Balloon angioplasty (rat, pig) Chang et al, 1995b; Smith et al, 1997bRB2/p130 Balloon angioplasty (rat) Claudio et al, 1999p53 Balloon angioplasty (rabbit, rat) Yonemitsu et al, 1998; Scheinman et al, 1999;Matsushita et al, 2000GAX Balloon angioplasty (rat, rabbit) Maillard et al, 1997; Smith et al, 1997a;Perlman et al, 1999GATA-6 Balloon angioplasty (rat) Mano et al, 1999RAS Balloon angioplasty (rat) Indolfi et al, 1995; Ueno et al, 1997bERK Balloon angioplasty (rat) Izumi et al, 2001mutants JNK Balloon angioplasty (rat) Izumi et al, 2001* These inhibitory effects might be caused by a nonantisense mechanism (Burgess et al, 1995; Chavany et al, 1995; Guvakova et al,1995; Villa et al, 1995; Wang et al, 1996).78

Gene Therapy and Molecular Biology Vol 7, page 79II. Preclinical studiesAntiproliferative gene therapy strategies designed forthe treatment of experimental cardiovascular diseaseinclude the following: 1) inactivation of positive cell cycleregulators (e. g., CDK/cyclins, protooncogenes, E2F,growth factors) by antisense approaches, ribozymes, andtranscription factor ‘decoy’ strategies (Figure 4), 2)overexpression of negative regulators of cell growth (e. g.,CKIs, p53, pRb, GAX, and GATA-6), and 3)overexpression of transdominant negative mutants ofpositive cell cycle regulators (e. g., Ras, mitogen-activatedprotein kinases).Table 2: Gene therapy clinical trials for vascular proliferative disease based on cytostatic strategies.Trial Design Strategy Disease Outcome Refs.PREVENT IPREVENT IIITALICSPREVENT:ITALICS:Randomized,double-blinded,single centerE2F decoy ODNex vivotransfectionof vein graftAutologousvein graftfailure afterperipheralartery bypass70-74% decreases inthe level of positivecell cycle regulatorsexpressed by VSMCsin the vein, andreduction in primarygraft failureMann etal, 1999RandomizedE2F decoy ODN Autologous vein Larger patency and Dzau et al,multicenter,ex vivograft failure after inhibition neointimal 2002double-blinded, transfection coronary artery thickeningplacebo-controlled of vein graft bypassRandomized, c-myc antisense In-stentNo reduction inKutryk etplacebo-controlled ODN delivery coronaryangiographical, 2002after stent restenosisrestenosis rateimplantationProject of ex-vivo vein graft engineering via transfectionInvestigation by the thoraxcenter of antisense DNA using local delivery and IVUS after coronary stentingFigure 4. Targeted gene inactivation by means of gene therapy strategies. Decoy approach by delivering a double-stranded ODNcorresponding to the optimum DNA recognition sequence of the transcription factor of interest (TF) leads to attenuation of its interactionwith the authentic cis-elements in cellular target genes, thus resulting in reduced gene transcription. Ribozymes inactivate the gene ofinterest by degrading their transcript. Antisense ODNs hybridize in a complementary fashion and stoicheometrically with the targetmRNA, thus causing blockade of translation or synthesis of a truncated (inactive) protein.79

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular diseaseA. Antisense approachThe gene of interest is inactivated by using asynthetic antisense oligodeoxynucleotide (ODN) thathybridizes in a complementary fashion andstoicheometrically with the target mRNA.1. CDKs and cyclinsThe efficacy of antisense ODN strategies targetingCDKs and cyclins to reduce neointimal lesion formationhas been demonstrated in several animal models ofballoon angioplasty. These studies include antisenseoligodeoxynucleotides against CDK2 (Abe et al, 1994;Morishita et al, 1994a), CDC2 (Morishita et al, 1993;1994b; Abe et al, 1994) and cyclin B1 (Morishita et al,1994b). Interestingly, cotransfection of antisense ODNagainst CDC2 kinase and cyclin B1 resulted in furtherinhibition of neointima formation, as compared toblockade of either gene target alone (Morishita et al,1994b). Of note, Morishita et al. (1993) reported sustainedinhibition of neointima formation in the rat carotidballoon-injury model after a single intraluminal moleculardelivery of combined CDC2 and proliferating cell nuclearantigen (PCNA) antisense ODNs, whereas this approachhad no effect in the coronary arteries of pigs after balloonangioplasty (Robinson et al, 1997). Downregulation ofcyclin G1 expression by retrovirus-mediated antisensegene transfer inhibited VSMC proliferation and neointimaformation after balloon angioplasty (Zhu et al, 1997).Attenuated graft atherosclerosis has been also observedupon inactivation of CDC2/PCNA (Mann et al, 1995;Miniati et al, 2000) and CDK2 (Suzuki et al, 1997) withantisense ODN.2. Mitogen-responsive nuclear factors thatpromote cell growthSeveral “immediate-early” genes (e. g., c-fos, c-jun,c-myc, c-myb, egr-1) are induced in serum-stimulatedVSMCs, and their overexpression can promote VSMCproliferation in vitro (Castellot et al, 1985; Kindy andSonenshein, 1986; Reilly et al, 1989; Brown et al, 1992;Campan et al, 1992; Rothman et al, 1994; Bennett et al,1994b; Gorski and Walsh, 1995). VSMCs cultured fromatheromatous plaques present higher levels of c-mycmRNA than in VSMCs from normal arteries (Parkes et al,1991), and arterial injury induced the expression of several“immediate-early” gene (Lambert et al, 2001; Miano et al,1990; 1993; Sylvester et al, 1998). Antisense ODNsagainst c-myc and c-myb reportedly inhibited in asequence-specific manner both VSMC proliferation invitro (Pukac et al, 1990; Brown et al, 1992; Ebbecke et al,1992; Simons and Rosenberg, 1992; Biro et al, 1993; Shiet al, 1993; Bennett et al, 1994a; Shi et al, 1994a; Gunn etal, 1997), and neointima formation after angioplasty(Simons et al, 1992; Bennett et al, 1994a; Shi et al, 1994b;Gunn et al, 1997; Kipshidze et al, 2001, 2002) and veingrafting (Mannion et al, 1998) in vivo. However, theseinhibitory effects may be mediated by a nonantisensemechanism (Burgess et al, 1995; Chavany et al, 1995;Guvakova et al, 1995; Villa et al, 1995; Wang et al, 1996).It has been recently shown that nanospherescontaining antisense ODN against PDGFβ receptor inhibitneointimal thickening in the rat carotid model of balloonangioplasty (Cohen-Sacks et al, 2002).B. RibozymesRibozymes represent a unique class of RNAmolecules that catalytically cleave the specific targetRNA, thus resulting in targeted gene inactivation. Su et al.(2000) designed a DNA-RNA chimeric hammerheadribozyme targeted to human transforming growth factorβ1(TGF-β1) that significantly inhibited angiotensin IIstimulatedTGF-β1 mRNA and protein expression inhuman VSMCs, and efficiently inhibited the growth ofthese cells. Likewise, cleavage of the platelet-derivedgrowth factor (PDGF) A-chain mRNA by hammerheadribozyme attenuated human and rat VSMC growth in vitro(Hu et al, 2001a,b) and inhibited neointima formation inthe rat carotid artery model of balloon injury (Kotani et al,2003).Studies using experimental models of angioplastyprovided the first evidence that ribozymes might representuseful tools in cardiovascular therapy. Frimerman et al.(1999) reported the efficacy of chimeric hammerheadribozyme to PCNA in reducing stent-induced stenosis in aporcine coronary model, and ribozyme strategy againstTGF-β1 inhibited neointimal formation after ballooninjury in the rat carotid artery model (Yamamoto et al,2000). 12-Lipoxygenase products of arachidonatemetabolism have growth and chemotactic effects invascular smooth muscle cells, and ribozyme against thisenzyme prevents intimal hyperplasia in balloon-injured ratcarotid arteries (Gu et al, 2001).C. Transcription factor ‘decoy’ strategiesThis approach consists of delivering a doublestrandedODN corresponding to the optimum DNA targetsequence of the transcription factor of interest, thusleading to the sequestration of the specific trans-actingfactor and attenuation of its interaction with the authenticcis-elements in cellular target genes.1. E2FE2F participates in the transcriptional activation ofgenes encoding proteins that are required for nucleotideand DNA biosynthesis (e. g., DNA polymerase α, histoneH2A, pcna, thymidine kinase) (Dyson, 1998; Lavia andJansen-Durr, 1999) and in several growth and cell-cycleregulators (e. g., c-myc, pRb, cdc2, cyclin E, cyclin A).Experimental neointimal thickening in ballooninjuredarteries (Morishita et al, 1995; Nakamura et al,2002), vein grafts (Mann et al, 1997; Ehsan et al, 2001),and cardiac allografts (Kawauchi et al, 2000) is preventedby the use of a synthetic ‘decoy’ ODN containing an E2Fconsensus binding site that inactivates the transcriptionfactor E2F. Ahn et al. (2002a) developed a novel E2F80

Gene Therapy and Molecular Biology Vol 7, page 81‘decoy’ ODN with a circular dumbbell structure (CD-E2F)and compared its properties with those of conventionalphosphorothioated E2F ‘decoy’ ODN (PS-E2F). CD-E2Fdisplayed more stability and stronger antiproliferativeactivity than PS-E2F when assayed in cultured VSMCs,and was more effective in inhibiting neointimal formationin vivo.2. Activator protein-1 (AP-1)Cell proliferation in the rat carotid artery model ofangioplasty correlated with elevated expression and highDNA-binding activity of transcription factors of the AP-1family (Miano et al, 1990; Miano et al, 1993; Hu et al,1997; Sylvester et al, 1998; Andrés et al, 2001). Underconditions of PDGF stimulation, AP-1 ‘decoy’ ODNdelivery into cultured human VSMCs significantlyreduced cell number and TGF-β1 production (Kume et al,2002), and attenuated neointimal thickening when appliedat the site of balloon angioplasty in rabbit carotid artery(Kume et al, 2002) and minipig coronary arteries(Buchwald et al, 2002). Circular dumbbell AP-1 ‘decoy’ODN was more effective in inhibiting the proliferation ofVSMCs in vitro and neointimal hyperplasia in vivocompared to conventional phosphorothioated AP-1 decoyODN, (Ahn et al, 2002b).D. Overexpression of growth suppressors1. CKIsThe efficacy of CKIs in inhibiting CDK activity andcell cycle progression has been widely documented in avariety of normal and tumour cells in vitro. The firstevidence that p21 Cip1 and p27 Kip1 may function as negativeregulators of neointimal hyperplasia was suggested inanimal studies showing the upregulation of these CKIs atlate time points following balloon angioplasty, coincidingwith the restoration of the quiescent phenotype after theinitial proliferative wave (Chen et al, 1997; Tanner et al,1998). The protective role of p27 Kip1 against neointimalthickening has been rigorously demonstrated inhypercholesterolemic apolipoprotein E (apoE)-deficientmice, in which genetic inactivation of p27 Kip1 acceleratedatherogenesis in a dose-dependent manner (Díez-Juan andAndrés, 2001). However, neointimal hyperplasia aftermechanical damage of the arterial wall was similar inwild-type and p27 Kip1 -null mice (Roque et al, 2001b).Redundant roles between p21 Cip1 and p27 Kip1 , orcompensatory increase in p21 Cip1 expression (or otherCKIs) might account for the lack of phenotype of p27 Kip1 -null mice in the setting of mechanical arterial injury.Several studies have suggested a role of CKIs inestablishing regional phenotypic variance in VSMCs fromdifferent vascular beds. Using human VSMCs isolatedfrom internal mammary artery and saphenous vein, Yanget al. (1998) suggested that sustained p27 Kip1 expression inspite of growth stimuli may contribute to the resistance togrowth of VSMCs from internal mammary artery and tothe longer patency of arterial versus venous grafts (Yanget al, 1998). Likewise, different expression of p15 Ink4b andp27 Kip1 has been correlated with distinct proliferativeresponse of intimal and medial VSMCs towards basicfibroblast growth factor (bFGF or FGF2) (Olson et al,2000). Intrinsic differences in the regulation of p27 Kip1might also play an important role in creating variance inthe proliferative and migratory capacity of VSMCsisolated from different vascular beds, which might in turncontribute to establishing regional variability inatherogenicity (Castro et al, 2003).Tanner et al (1998) have reported more frequentexpression of p27 Kip1 and p21 Cip1 within regions of humancoronary atheromas not undergoing proliferation.Concordant expression of TGF-β receptors I and II invirtually all cells positive for p27 Kip1 within humanatherosclerotic plaques indicates that TGF-β1 present inthese lesions may contribute to p27 Kip1 upregulation (Ihlinget al, 1999). Moreover, coexpression of p53 and p21 Cip1 inhuman carotid atheromatous plaque cells that revealedlack of proliferation markers suggests that induction ofp21 Cip1 may occur via transcriptional activation by p53(Ihling et al, 1997).Ectopic expression of p21 Cip1 and p27 Kip1 , but notp16 Ink4a , significantly reduced neointimal thickening inseveral animal models of angioplasty (Chang et al, 1995a;Yang et al, 1996; Chen et al, 1997; Ueno et al, 1997a;Tanner et al, 2000; Condorelli et al, 2001). Overexpressionof p21 Cip1 also attenuated neointimal lesion formation in arabbit model of vein grafting (Bai et al, 1998).2. p53p53 is a transcription factor that functions as a tumorsuppressor displaying both antiproliferative andproapoptotic actions. These effects result from complexregulatory networks, including transcriptional activation ofantiproliferative and proapoptotic genes (e. g., p21 Cip1 andBax, respectively), transcriptional repression ofproproliferative and antiapoptotic genes (e. g., IGF-II andbcl-2, respectively), and direct protein-protein interactions(e. g., with helicases and caspases). Increased VSMCproliferation has been shown as a result of antisense p53ODN transfection (Aoki et al, 1999; Matsushita et al,2000), and p53 gene transfer has the opposite effect(Yonemitsu et al, 1998). Mayr et al (2002) showed ahigher rate of proliferation and migration of VSMCsisolated from p53-deficient mice than its wild-typecounterparts. Consistent with these findings, earlymigration and proliferation of VSMCs happened inexplanted porcine tunica media tissue after mitogeninduceddownregulation of p53 (Rodriguez-Campos et al,2001).p53 deficiency has been demonstrated to have aproatherogenic effect in studies of genetic inactivation inhypercholesterolemic apoE and apoE*3-Leiden mice,although the relative contribution of increased cellularproliferation and decreased apoptosis in these animalmodels remains obscure (Guevara et al, 1999; van Vlijmenet al, 2001). Mice deficient for p53 also disclosedaccelerated vein graft atherosclerosis (Mayr et al, 2002).Regarding human atherosclerosis, p53 is overexpressedbut not mutated in human atherosclerotic tissue (Iacopetta81

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular diseaseet al, 1995), and lack of proliferation markers in vascular expression and G1 cell cycle arrest (Perlman et al, 1998).The GATA transcription factors play a critical role in ERKs disclosed persistent hyperexpression and activationthe establishment of hematopoietic cell lineages and in atherosclerotic lesions of cholesterol-fed rabbits,during the development of the cardiovascular system suggesting that these factors play critical roles in initiating(Simon, 1995). GATA-6 is rapidly downregulated upon and perpetuating cell proliferation during the developmentmitogen stimulation of quiescent VSMCs (Suzuki et al, of atherosclerosis (Hu et al, 2000; Metzler et al, 2000).1996), and overexpression of GATA-6 induced p21 Cip1 Likewise, angioplasty in porcine and rat arteries led to thecells coexpressing p53 and p21 Cip1 within advanced human Importantly, p21 Cip1 -null mouse embryonic fibroblastsatherosclerotic lesions suggests that transcriptional were refractory to the GATA-6-induced growth inhibitionactivation of the p21 Cip1 gene by p53 may be a protective (Perlman et al, 1998). The level of GATA-6 mRNA,mechanism against excessive vascular cell growth (Ihling protein, and DNA-binding activity is transientlyet al, 1997).downregulated at early time points after balloonp53 appears to play an important role in the angioplasty in the rat carotid artery, and reversal ofpathogenesis of restenosis, as suggested by both animal GATA-6 downregulation by adenovirus-mediated GATAandhuman studies. Transfection of antisense p53 ODN 6 gene transfer to the vessel wall inhibited intimalinto rat intact carotid artery decreased p53 protein hyperplasia in this animal model (Mano et al, 1999).expression and resulted in a significant increase inneointimal lesion growth at 2 and 4 weeks after balloonangioplasty(Matsushita et al, 2000). Evidence suggests5. GAXthat human cytomegalovirus (HCMV) infection Gax is a homeobox gene highly expressed in culturescontributes to the development of atherosclerosis and of quiescent VSMCS, which is rapidly downregulated inrestenosis, and part of this effect may be due to increased vitro upon growth factor stimulation of VSMCs, and afterVSMC proliferation and migration by inactivation of p53 balloon angioplasty in vivo (Gorski et al, 1993; Weir et al,(Speir et al, 1994; Zhou et al, 1996; 1999; Tanaka et al, 1995). Overexpression of GAX inhibited VSMC1999). It is also noteworthy that human VSMCs from proliferation in vitro and attenuated neointimal thickeningrestenosis or in-stent stenosis sites demonstrate normal or in balloon-injured rat carotid arteries in a p21 Cip1 -enhanced responses to p53 when compared to VSMCs dependent manner (Smith et al, 1997a; Perlman et al,from normal vessels (Scott et al, 2002). Moreover, p53 1999). Percutaneous delivery of the Gax gene alsogene transfer effectively inhibited neointimal hyperplasia inhibited vessel stenosis in a rabbit model of balloonafter experimental angioplasty (Yonemitsu et al, 1998; angioplasty (Maillard et al, 1997).Scheinman et al, 1999; Matsushita et al, 2000), and inhuman saphenous vein (George et al, 2001).E. Overexpression of transdominantnegative mutants of positive cell cycle3. pRbregulators.The complex interplay between pRb and 1. Rastranscription factors of the E2F family plays a critical rolein the control of cell growth (Stevaux and Dyson, 2002).Ras-dependent signaling plays an important role inmitogen-stimulated cell growth (Pronk and Bos, 1994).E2F-dependent transactivation of genes required for cell Ras is implicated in the activation of the G1cycle progression is prevented in quiescent cells due to the CDK/cyclin/E2F pathway (Winston et al, 1996;Aktas et al,accumulation of hypophosphorylated pRb. 1997; Kerkhoff and Rapp, 1997; Leone et al, 1997; LloydHyperphorylation of pRb by mitogenic stimuli leads toE2F activation and cell growth. Transfer of antisense pRbODN into human VSMCs resulted in the induction of theproapoptotic factors bax and p53, and this was associatedwith increased number of apoptotic cells and a higher rateet al, 1997; Peeper et al, 1997; Zou et al, 1997) and iscritical for the normal induction of cyclin A promoteractivity and DNA synthesis in mitogen-stimulated VSMCs(Sylvester et al, 1998). Consistent with these findings,local delivery of transdominant negative mutants of Rasof DNA synthesis (Aoki et al, 1999). Inhibition of VSMC attenuated neointimal thickening after experimentalproliferation in vitro and attenuation of neointimaformation after balloon angioplasty can be achieved byballoon angioplasty (Indolfi et al, 1995; Ueno et al,1997b).adenovirus-mediated transfer of several forms of pRb,including full-length constitutively active(nonphosphorylatable) and phosphorylation-competent 2. Mitogen-activatedpRb, and truncated versions of pRb (Chang et al, 1995b; (MAPKs)protein kinasesSmith et al, 1997b). Similarly, adenoviral transfer of the The MAPK pathway is critical in the transducction ofpRb related protein RB2/p130 inhibited VSMC proliferative signals in many mammalian tissues, includingproliferation in vitro and prevented neointimal hyperplasia the cardiovascular system (Zou et al, 1998; Bogoyevitch,after experimental angioplasty (Claudio et al, 1999). 2000). Several families of MAPKs have been described,including the stress-activated protein kinases/c-jun NH 2 -terminal protein kinases (SAPKs/JNKs), extracellular4. GATA-6signal-regulated kinases (ERKs), and p38. JNKs and82

Gene Therapy and Molecular Biology Vol 7, page 83rapid activation of ERKs and JNKs (Lai et al, 1996; Lilleet al, 1997; Pyles et al, 1997; Koyama et al, 1998).Consistent with this notion, gene transfer of dominantnegativemutants of ERK or JNK prevented neointimalformation in balloon-injured rat artery (Izumi et al, 2001).III. Clinical studiesThe antiproliferative approaches used so far for thetreatment of cardiovascular disease have focused onrestenosis and graft atherosclerosis, during whichneointimal hyperplasia is rapid and localized. Thesedisorders remain the major limitation of revascularizationby percutaneous transluminal angioplasty (PTCA) andartery bypass surgery.A. E2F ‘decoy’Encouraging results of the E2F ‘decoy’ strategy inanimal models of balloon angioplasty and graftatherosclerosis (see above) led to the initiation of the firstProject of Ex-vivo Vein graft Engineering via Transfection(PREVENT I) (Mann et al, 1999). In this single-centre,randomized, controlled gene therapy trial, 41 patientsundergoing bypass for the treatment of peripheral arterialocclusions were randomly assigned untreated (n=16), E2F-‘decoy’-ODN-treated (n=17), or scrambled-ODN-treated(n=8) human infrainguinal vein grafts. Ex vivo delivery ofODNs was achieved intraoperatively via pressuremediatedtransfection. This procedure was associated witha 70-74% decrease in the level of PCNA and c-mycmRNA expressed by the VSMCs in the vein, and astatistically significant reduction in primary graft failurecompared to control groups. Following to this pilot trial, arandomized, double-blinded, placebo controlled Phase IIbtrial (PREVENT II) was carried out in patients undergoingcoronary artery bypass surgery. The results of quantitativecoronary angiography and intravascular ultrasound (IVUS)showed larger patency and inhibition of neointimalthickening in treated patients at 12 months afterintervention (Dzau et al, 2002).B. c-myc antisense ODNPharmacokinetics and clinical safety of ascendingdoses of c-myc antisense ODN (LR-3280) administeredafter PTCA was assessed by Roque et al. (2001a). Seventyeight patients were randomized to receive either standardcare (n = 26) or standard care and escalating doses (1 to 24mg) of LR-3280 (n = 52), administered into target vesselthrough a guiding catheter. The peak plasmaconcentrations of LR-3280 occurred at 1 minute anddecreasing rapidly after approximately 1 hour, with littleLR-3280 detected in the urine between 0-6 hours and 12-24 hours. The intracoronary administration of LR-3280was well tolerated at doses up to 24 mg and produced noadverse effects in dilated coronary arteries, thus providingthe basis for the evaluation of local delivery of c-mycantisense ODN for the prevention of humanvasculoproliferative disease.Kutryk et al. (2002) recently reported the results ofthe Investigation by the Thoraxcenter of Antisense DNAusing Local delivery and IVUS after Coronary Stenting(ITALICS) trial. This randomized, placebo controlledstudy was designed to determine the efficacy of antisenseODN against c-myc in inhibiting in-stent restenosis.Eighty-five patients were randomly assigned to receiveeither c-myc antisense ODN or saline vehicle byintracoronary local delivery after coronary stentimplantation. Follow-up included the percent neointimalvolume obstruction measured by IVUS, clinical outcomeand quantitative coronary angiography. There was noreduction in either the neointimal volume obstruction orthe angiographic restenosis rate after treatment with 10 mgof phosphorothioate-modified ODN directed against c-myc as demonstrated by the analysis of 77 patients.IV. ConclusionsExcessive cell proliferation within the arterial wall isthought to contribute to neointimal thickening during thepathogenesis of atherosclerosis, in-stent restenosis, andvessel bypass graft failure. Animal models ofatherosclerosis have demonstrated an inverse correlationbetween neointimal cell proliferation and atheroma size,suggesting that excessive cell growth prevails at the onsetof atherogenesis. Cell proliferation may also predominateat the early stages of human atheroma development. Thus,given that patients frequently exhibit advancedatherosclerotic plaques when first diagnosed, the potentialbenefit of antiproliferative strategies for the treatment ofhuman atherosclerosis is uncertain. The antiproliferativeapproaches used so far in the setting of vascularobstructive disease have focused on restenosis and graftatherosclerosis, during which neointimal hyperplasia isspatially localized and develops over a short period of time(typically 2-12 months). Gene therapy is emerging as anattractive strategy in the treatment of vascular proliferativedisease due to minimally invasive and easily monitoredgene delivery in vascular interventions. Antiproliferativegene therapy strategies that have proven efficient ininhibiting neointimal thickening in animal models ofvascular obstructive disease include the use of antisenseandribozyme-mediated inactivation of positive cell cycleregulators, overexpression of negative regulators of cellgrowth, and ‘decoy’ strategies to inactivate transcriptionfactors that promote cell cycle progression. Althoughsome of these strategies have shown encouraging results inhumans, further studies are required to override the currentpractical barriers and limitations placed on most clinicaltrials before gene therapy strategies exhibit wideapplication in clinic. These should include the clarificationof safety issues, development of better gene deliveryvectors, and improvement of transgene expression. Asidefrom these technical improvements, significant effort inbasic research is warranted to identify more effective andsafer treatment genes.83

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular diseaseAcknowledgmentsWork in the laboratory of V. Andrés is partiallysupported by the Ministerio de Ciencia y Tecnología ofSpain (MCyT) and Fondo Europeo de Desarrollo Regional(grants SAF2001-2358 and SAF2002-1143), and fromInstituto de Salud Carlos III (ISCIII) (Red de CentrosC03/01). S. M. Sanz and M. Gascón are predoctoralfellows of the ISCIII and MCyT, respectively.ReferencesAbe, J, Zhou, W, Taguchi, J, Takuwa, N, Miki, K, Okazaki, H,Kurokawa, K, Kumada, M, and Takuwa, Y (1994).Suppression of neointimal smooth muscle cell accumulationin vivo by antisense cdc2 and cdk2 oligonucleotides in ratcarotid artery. Biochem Biophys Res Commun 198, 16-24.Ahn, JD, Morishita, R, Kaneda, Y, Kim, HS, Chang, YC, Lee,KU, Park, JY, Lee, HW, Kim, YH, and Lee, IK (2002a).Novel E2F decoy oligodeoxynucleotides inhibit in vitrovascular smooth muscle cell proliferation and in vivoneointimal hyperplasia. Gene Ther 9, 1682-1692.Ahn, JD, Morishita, R, Kaneda, Y, Lee, SJ, Kwon, KY, Choi,SY, Lee, KU, Park, JY, Moon, IJ, Park, JG, Yoshizumi, M,Ouchi, Y, and Lee, IK (2002b). Inhibitory effects of novelAP-1 decoy oligodeoxynucleotides on vascular smoothmuscle cell proliferation in vitro and neointimal formation invivo. Circ Res 90, 1325-1332.Aktas, H, Cai, H, and Cooper, GM (1997). Ras links growthfactor signaling to the cell cyle machinery via regulation ofcyclin D1 and the cdk inhibitor p27 KIP1 . Mol Cell Biol 17,3850-3857.Andrés, V (1998). Control of vascular smooth muscle cellgrowth and its implication in atherosclerosis and restenosis.Int J Molec Med 2, 81-89.Andrés, V, Ure_a, J, Poch, E, Chen, D, and Goukassian, D(2001). The role of Sp1 in the induction of p27 geneexpression in vascular smooth muscle cells in vitro and afterballoon angioplasty. Arterioscl Thromb Vasc Biol 21, 342-347.Aoki, M, Morishita, R, Matsushita, H, Hayashi, S, Nakagami, H,Yamamoto, K, Moriguchi, A, Kaneda, Y, Higaki, J, andOgihara, T (1999). Inhibition of the p53 tumor suppressorgene results in growth of human aortic vascular smoothmuscle cells. Potential role of p53 in regulation of vascularsmooth muscle cell growth. Hypertension 34, 192-200.Bai, H, Morishita, R, Kida, I, Yamakawa, T, Zhang, W, Aoki, M,Matsushita, H, Noda, A, Nagai, R, Kaneda, I, Higaki, J,Ogihara, T, Sawa, Y, and Matsuda, H (1998). Inhibition ofintimal hyperplasia after vein grafting by in vivo transfer ofhuman senescent cell-derived inhibitor-1 gene. Gene Ther 5,761-769.Bauters, C, and Isner, JM (1997). The biology of restenosis.Prog Cardiovasc Dis 40, 107-116.Bennett, MR, Anglin, S, McEwan, JR, Jagoe, R, Newby, AC,and Evan, GI (1994a). Inhibition of vascular smooth musclecell proliferation in vitro and in vivo by c-myc antisenseoligodeoxynucleotides. J Clin Invest 93, 820-828.Bennett, MR, Evan, GI, and Newby, AC (1994b). Deregulatedexpression of the c-myc oncogene abolishes inhibition ofproliferation of rat vascular smooth muscle cells by serumreduction, interferon-g, heparin, and cyclic nucleotideanalogues and induces apoptosis. Circ Res 74, 525-536.Biro, S, Fu, Y-M, Yu, Z-X, and Epstein, SE (1993). Inhibitoryeffects of antisense oligodeoxynucleotides targeting c-mycmRNA on smooth muscle cell proliferation and migration.Proc Natl Acad Sci U S A 90, 654-658.Bogoyevitch, MA (2000). Signalling via stress-activatedmitogen-activated protein kinases in the cardiovascularsystem. Cardiovasc Res 45, 826-842.Braun-Dullaeus, RC, Mann, MJ, Seay, U, Zhang, L, von DerLeyen, HE, Morris, RE, and Dzau, VJ (2001). Cell cycleprotein expression in vascular smooth muscle cells in vitroand in vivo is regulated through phosphatidylinositol 3-kinase and mammalian target of rapamycin. ArteriosclerThromb Vasc Biol 21, 1152-1158.Brown, KE, Kindy, MS, and Sonenshein, GE (1992). Expressionof the c-myb proto-oncogene in bovine vascular smoothmuscle cells. J Biol Chem 267, 4625-4630.Buchwald, AB, Wagner, AH, Webel, C, and Hecker, M (2002).Decoy oligodeoxynucleotide against activator protein-1reduces neointimal proliferation after coronary angioplasty inhypercholesterolemic minipigs. J Am Coll Cardiol 39, 732-738.Burgess, TL, Fisher, EF, Ross, SL, Bready, JV, Qian, YX,Bayewitch, LA, Cohen, AM, Herrera, CJ, Hu, SS, Kramer,TB, and et al. (1995). The antiproliferative activity of c-myband c-myc antisense oligonucleotides in smooth muscle cellsis caused by a nonantisense mechanism. Proc Natl Acad SciU S A 92, 4051-4055.Burrig, KF (1991). The endothelium of advanced arterioscleroticplaques in humans. Arterioscler Thromb 11, 1678-1689.Campan, M, Desgranges, C, Gadeau, A-P, Millet, D, and Belloc,F (1992). Cell cycle dependent gene expression in quiescent,stimulated, and asynchronously cycling arterial smoothmuscle cells in culture. J Cell Physiol 150, 493-500.Castellot, JJ, Cochran, DL, and Karnovsky, MJ (1985). Effect ofheparin on vascular smooth muscle cells. I. Cell metabolism.J Cell Physiol 124, 21-28.Castro, C, Díez-Juan, A, Cortés, MJ, and Andrés, V (2003).Distinct regulation of mitogen-activated protein kinases andp27 Kip1 in smooth muscle cells from different vascular beds.A potential role in establishing regional phenotypic variance.J Biol Chem 278, 4482-4490.Chang, MW, Barr, E, Lu, MM, Barton, K, and Leiden, JM(1995a). Adenovirus-mediated over-expression of thecyclin/cyclin-dependent kinase inhibitor, p21 inhibitsvascular smooth muscle cell proliferation and neointimaformation in the rat carotid artery model of balloonangioplasty. J Clin Invest 96, 2260-2268.Chang, MW, Barr, E, Seltzer, J, Jiang, Y, Nabel, GJ, Nabel, EG,Parmacek, MS, and Leiden, JM (1995b). Cytostatic genetherapy for vascular proliferative disorders with aconstitutively active form of the retinoblastoma geneproduct. Science 267, 518-522.Chavany, C, Connell, Y, and Neckers, L (1995). Contribution ofsequence and phosphorothioate content to inhibition of cellgrowth and adhesion caused by c-myc antisense oligomers.Mol Pharmacol 48, 738-746.Chen, D, Krasinski, K, Chen, D, Sylvester, A, Chen, J, Nisen,PD, and Andrés, V (1997). Downregulation of cyclindependentkinase 2 activity and cyclin A promoter activity invascular smooth muscle cells by p27 Kip1 , an inhibitor ofneointima formation in the rat carotid artery. J Clin Invest99, 2334-2341.Claudio, PP, Fratta, L, Farina, F, Howard, CM, Stassi, G,Numata, S, Pacilio, C, Davis, A, Lavitrano, M, Volpe, M,Wilson, JM, Trimarco, B, Giordano, A, and Condorelli, G(1999). Adenoviral RB2/p130 gene transfer inhibits smoothmuscle cell proliferation and prevents restenosis afterangioplasty. Circ Res 85, 1032-1039.Cohen-Sacks, H, Najajreh, Y, Tchaikovski, V, Gao, G, Elazer, V,Dahan, R, Gati, I, Kanaan, M, Waltenberger, J, and Golomb,G (2002). Novel PDGFbetaR antisense encapsulated in84

Gene Therapy and Molecular Biology Vol 7, page 85polymeric nanospheres for the treatment of restenosis. GeneTher 9, 1607-1616.Condorelli, G, Aycock, JK, Frati, G, and Napoli, C (2001).Mutated p21/WAF/CIP transgene overexpression reducessmooth muscle cell proliferation, macrophage deposition,oxidation-sensitive mechanisms, and restenosis inhypercholesterolemic apolipoprotein E knockout mice.FASEB J 15, 2162-2170.Cortés, MJ, Díez-Juan, A, Pérez, P, Pérez-Roger, I, Arroyo-Pellicer, R, and Andrés, V (2002). Increased earlyatherogenesis in young versus old hypercholesterolemicrabbits by a mechanism independent of arterial cellproliferation. FEBS Letters 522, 99-103.Dartsch, PC, Voisard, R, Bauriedel, G, Hofling, B, and Betz, E(1990). Growth characteristics and cytoskeletal organizationof cultured smooth muscle cells from human primarystenosing and restenosing lesions. Arteriosclerosis 10, 62-75.Díez-Juan, A, and Andrés, V (2001). The growth suppressorp27 Kip1 protects against diet-induced atherosclerosis. FASEBJ 15, 1989-1995.Dyson, N (1998). The regulation of E2F by pRB-family proteins.Genes Dev 12, 2245-2262.Dzau, VJ, Braun-Dullaeus, RC, and Sedding, DG (2002).Vascular proliferation and atherosclerosis: new perspectivesand therapeutic strategies. Nat Med 8, 1249-1256.Ebbecke, M, Unterberg, C, Buchwald, A, Stohr, S, and Wiegand,V (1992). Antiproliferative effects of a c-myc antisenseoligonucleotide on human arterial smooth muscle cells. BasicRes Cardiol 87, 585-591.Ehsan, A, Mann, MJ, Dell'Acqua, G, and Dzau, VJ (2001).Long-term stabilization of vein graft wall architecture andprolonged resistance to experimental atherosclerosis afterE2F decoy oligonucleotide gene therapy. J ThoracCardiovasc Surg 121, 714-722.Essed, CE, Van den Brand, M, and Becker, AE (1983).Transluminal coronary angioplasty and early restenosis.Fibrocellular occlusion after wall laceration. Br Heart J 49,393-396.Frimerman, A, Welch, PJ, Jin, X, Eigler, N, Yei, S, Forrester, J,Honda, H, Makkar, R, Barber, J, and Litvack, F (1999).Chimeric DNA-RNA hammerhead ribozyme to proliferatingcell nuclear antigen reduces stent-induced stenosis in aporcine coronary model. Circulation 99, 697-703.George, SJ, Angelini, GD, Capogrossi, MC, and Baker, AH(2001). Wild-type p53 gene transfer inhibits neointimaformation in human saphenous vein by modulation of smoothmuscle cell migration and induction of apoptosis. Gene Ther8, 668-676.Gordon, D, Reidy, MA, Benditt, EP, and Schwartz, SM (1990).Cell proliferation in human coronary arteries. Proc NatlAcad Sci U S A 87, 4600-4604.Gorski, DH, LePage, DF, Patel, CV, Copeland, NG, Jenkins,NA, and Walsh, K (1993). Molecular cloning of a divergedhomeobox gene that is rapidly down-regulated during theG0/G1 transition in vascular smooth muscle cells. Mol CellBiol 13, 3722-3733.Gorski, DH, and Walsh, K (1995). Mitogen-responsive nuclearfactors that mediate growth control signals in vascularmyocytes. Cardiovasc Res 30, 585-592.Gu, JL, Pei, H, Thomas, L, Nadler, JL, Rossi, JJ, Lanting, L, andNatarajan, R (2001). Ribozyme-mediated inhibition of ratleukocyte-type 12-lipoxygenase prevents intimal hyperplasiain balloon-injured rat carotid arteries. Circulation 103,1446-1452.Guevara, NV, Kim, HS, Antonova, EI, and Chan, L (1999). Theabsence of p53 accelerates atherosclerosis by increasing cellproliferation in vivo. Nat Med 5, 335-339.Gunn, J, Holt, CM, Francis, SE, Shepherd, L, Grohmann, M,Newman, CM, Crossman, DC, and Cumberland, DC (1997).The effect of oligonucleotides to c-myb on vascular smoothmuscle cell proliferation and neointima formation afterporcine coronary angioplasty. Circ Res 80, 520-531.Guvakova, MA, Yakubov, LA, Vlodavsky, I, Tonkinson, JL, andStein, CA (1995). Phosphorothioate oligodeoxynucleotidesbind to basic fibroblast growth factor, inhibit its binding tocell surface receptors, and remove it from low affinitybinding sites on extracellular matrix. J Biol Chem 270,2620-2627.Hu, WY, Fukuda, N, Kishioka, H, Nakayama, M, Satoh, C, andKanmatsuse, K (2001a). Hammerhead ribozyme targetinghuman platelet-derived growth factor A- chain mRNAinhibited the proliferation of human vascular smooth musclecells. Atherosclerosis 158, 321-329.Hu, WY, Fukuda, N, Nakayama, M, Kishioka, H, andKanmatsuse, K (2001b). Inhibition of vascular smoothmuscle cell proliferation by DNA-RNA chimerichammerhead ribozyme targeting to rat platelet-derivedgrowth factor A-chain mRNA. J Hypertens 19, 203-212.Hu, Y, Cheng, L, Hochleitner, BW, and Xu, Q (1997). Activationof mitogen-activated protein kinases (ERK/JNK) and AP-1transcription factor in rat carotid arteries after balloon injury.Arterioscler Thromb Vasc Biol 17, 2808-2816.Hu, Y, Dietrich, H, Metzler, B, Wick, G, and Xu, Q (2000).Hyperexpression and activation of extracellular signalregulatedkinases (ERK1/2) in atherosclerotic lesions ofcholesterol-fed rabbits. Arterioscler Thromb Vasc Biol 20,18-26.Iacopetta, B, Wysocki, S, Norman, P, and House, A (1995). Thep53 tumor suppressor gene is overexpressed but not mutatedin human atherosclerotic tissue. Int J Oncol 7, 399-402.Ihling, C, Menzel, G, Wellens, E, Monting, JS, Schaefer, HE,and Zeiher, AM (1997). Topographical association betweenthe cyclin-dependent kinases inhibitor p21, p53accumulation, and cellular proliferation in humanatherosclerotic tissue. Arterioscler Thromb Vasc Biol 17,2218-2224.Ihling, C, Technau, K, Gross, V, Schulte-Monting, J, Zeiher,AM, and Schaefer, HE (1999). Concordant upregulation oftype II-TGF-beta-receptor, the cyclin- dependent kinasesinhibitor p27 Kip1 and cyclin E in human atherosclerotic tissue:implications for lesion cellularity. Atherosclerosis 144, 7-14.Indolfi, C, Avvedimento, EV, Rapacciuolo, A, Di Lorenzo, E,Esposito, G, Stabile, E, Feliciello, A, Mele, E, Giuliano, P,Condorelli, G, and Chiariello, M (1995). Inhibition ofcellular ras prevents smooth muscle cell proliferation aftervascular injury in vivo. Nature Med 1, 541-545.Isner, JM (1994). Vascular remodeling. Honey, I think I shrunkthe artery. Circulation 89, 2937-2941.Izumi, Y, Kim, S, Namba, M, Yasumoto, H, Miyazaki, H,Hoshiga, M, Kaneda, Y, Morishita, R, Zhan, Y, and Iwao, H(2001). Gene transfer of dominant-negative mutants ofextracellular signal- regulated kinase and c-Jun NH2-terminal kinase prevents neointimal formation in ballooninjuredrat artery. Circ Res 88, 1120-1126.Katsuda, S, Coltrera, MD, Ross, R, and Gown, AM (1993).Human atherosclerosis. IV. Immunocytochemical analysis ofcell activation and proliferation in lesions of young adults.Am J Pathol 142, 1787-1793.Kawauchi, M, Suzuki, J, Morishita, R, Wada, Y, Izawa, A,Tomita, N, Amano, J, Kaneda, Y, Ogihara, T, Takamoto, S,and Isobe, M (2000). Gene therapy for attenuating cardiacallograft arteriopathy using ex vivo E2F decoy transfectionby HVJ-AVE-liposome method in mice and nonhumanprimates. Circ Res 87, 1063-1068.85

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular diseaseKearney, M, Pieczek, A, Haley, L, Losordo, DW, Andrés, V,Schainfield, R, Rosenfield, R, and Isner, JM (1997).Histopathology of in-stent restenosis in patients withperipheral artery disease. Circulation 95, 1998-2002.Kerkhoff, E, and Rapp, UR (1997). Induction of cell proliferationin quiescent NIH 3T3 cells by oncogenic c-Raf-1. Mol CellBiol 17, 2576-2586.Kindy, MS, and Sonenshein, GE (1986). Regulation of oncogeneexpression in cultured aortic smooth muscle cells: posttranscriptionalcontrol of c-myc mRNA. J Biol Chem 261,12865-12868.Kipshidze, N, Keane, E, Stein, D, Chawla, P, Skrinska, V,Shankar, LR, Khanna, A, Komorowski, R, Haudenschild, C,Iversen, P, Leon, MB, Keelan, MH, and Moses, J (2001).Local delivery of c-myc neutrally charged antisenseoligonucleotides with transport catheter inhibits myointimalhyperplasia and positively affects vascular remodeling in therabbit balloon injury model. Catheter Cardiovasc Interv54, 247-256.Kipshidze, NN, Kim, HS, Iversen, P, Yazdi, HA, Bhargava, B,New, G, Mehran, R, Tio, F, Haudenschild, C, Dangas, G,Stone, GW, Iyer, S, Roubin, GS, Leon, MB, and Moses, JW(2002). Intramural coronary delivery of advanced antisenseoligonucleotides reduces neointimal formation in the porcinestent restenosis model. J Am Coll Cardiol 39, 1686-1691.Kotani, M, Fukuda, N, Ando, H, Hu, WY, Kunimoto, S, Saito, S,and Kanmatsuse, K (2003). Chimeric DNA-RNAhammerhead ribozyme targeting PDGF A-chain mRNAspecifically inhibits neointima formation in rat carotid arteryafter balloon injury. Cardiovasc Res 57, 265-276.Koyama, H, Olson, NE, Dastvan, FF, and Reidy, MA (1998).Cell replication in the arterial wall: activation of signalingpathway following in vivo injury. Circ Res 82, 713-721.Kume, M, Komori, K, Matsumoto, T, Onohara, T, Takeuchi, K,Yonemitsu, Y, and Sugimachi, K (2002). Administration of adecoy against the activator protein-1 binding site suppressesneointimal thickening in rabbit balloon-injured arteries.Circulation 105, 1226-1232.Kutryk, MJ, Foley, DP, van den Brand, M, Hamburger, JN, vander Giessen, WJ, deFeyter, PJ, Bruining, N, Sabate, M, andSerruys, PW (2002). Local intracoronary administration ofantisense oligonucleotide against c-myc for the prevention ofin-stent restenosis: results of the randomized investigation bythe Thoraxcenter of antisense DNA using local delivery andIVUS after coronary stenting (ITALICS) trial. J Am CollCardiol 39, 281-287.Lai, K, Lee, W-S, Jain, MK, Lee, M-E, and Haber, E (1996).Mitogen-activated protein kinase phosphatase-1 in rat arterialsmooth muscle cell proliferation. J Clin Invest 98, 1560-1567.Lambert, DL, Malik, N, Shepherd, L, Gunn, J, Francis, SE, King,A, Crossman, DC, Cumberland, DC, and Holt, CM (2001).Localization of c-Myb and induction of apoptosis byantisense oligonucleotide c-Myb after angioplasty of porcinecoronary arteries. Arterioscler Thromb Vasc Biol 21, 1727-1732.Lavia, P, and Jansen-Durr, P (1999). E2F target genes and cellcyclecheckpoint control. Bioessays 21, 221-230.Leone, G, DeGregori, J, Sears, R, Jakoi, L, and Nevins, JR(1997). Myc and Ras collaborate in inducing accumulation ofactive cyclin E/cdk2 and E2F. Nature 387, 422-426.Libby, P, and Tanaka, H (1997). The molecular basis ofrestenosis. Prog Cardiovasc Dis 40, 97-106.Lille, S, Daum, G, Clowes, MM, and Clowes, AW (1997). Theregulation of p42/p44 mitogen-activated protein kinases inthe injured rat carotid artery. J Surg Res 70, 178-186.Lloyd, AC, Obermüller, F, Staddon, S, Barth, CF, McMahon, M,and Land, H (1997). Cooperating oncogenes converge toregulate cyclin/cdk complexes. Genes Dev 11, 663-677.Lusis, AJ (2000). Atherosclerosis. Nature 407, 233-241.Maillard, L, Van Belle, E, Smith, RC, Le Roux, A, Denèfle, P,Steg, G, Barry, JJ, Branellec, D, Isner, JM, and Walsh, K(1997). Percutaneous delivery of the gax gene inhibits vesselstenosis in a rabbit model of balloon angioplasty.Cardiovasc Res 35, 536-546.Mann, M, Gibbons, GH, Kernoff, RS, Diet, FP, Tsao, PS, Cooke,JP, Kaneda, Y, and Dzau, VJ (1995). Genetic engineering ofvein grafts resistant to atherosclerosis. Proc Natl Acad SciUSA 92, 4502-4506.Mann, MJ, Gibbons, GH, Tsao, PS, von der Leyen, HE, Cooke,JP, Buitrago, R, Kernoff, R, and Dzau, VJ (1997). Cell cycleinhibition preserves endothelial function in geneticallyengineered rabbit vein grafts. J Clin Invest 99, 1295-1301.Mann, MJ, Whittemore, AD, Donaldson, MC, Belkin, M, Conte,MS, Polak, JF, Orav, EJ, Ehsan, A, Dell'Acqua, G, and Dzau,VJ (1999). Ex-vivo gene therapy of human vascular bypassgrafts with E2F decoy: the PREVENT single-centre,randomised, controlled trial. Lancet 354, 1493-1498.Mannion, JD, Ormont, ML, Magno, MG, O'Brien, JE, Shi, Y,and Zalewski, A (1998). Sustained reduction of neointimawith c-myc antisense oligonucleotides in saphenous veingrafts. Ann Thorac Surg 66, 1948-1952.Mano, T, Luo, ZY, Malendowicz, SL, Evans, T, and Walsh, K(1999). Reversal of GATA-6 downregulation promotessmooth muscle differentiation and inhibits intimalhyperplasia in balloon-injured rat carotid artery. Circ Res 84,647-654.Matsushita, H, Morishita, R, Aoki, M, Tomita, N, Taniyama, Y,Nakagami, H, Shimozato, T, Higaki, J, Kaneda, Y, andOgihara, T (2000). Transfection of antisense p53 tumorsuppressor gene oligodeoxynucleotides into rat carotid arteryresults in abnormal growth of vascular smooth muscle cells.Circulation 101, 1447-1452.Mayr, U, Mayr, M, Li, C, Wernig, F, Dietrich, H, Hu, Y, and Xu,Q (2002). Loss of p53 accelerates neointimal lesions of veinbypass grafts in mice. Circ Res 90, 197-204.McMillan, GC, and Stary, HC (1968). Preliminary experiencewith mitotic activity of cellular elements in theatherosclerotic plaques of cholesterol-fed rabbits studied bylabeling with tritiated thymidine. Ann N Y Acad Sci 149,699-709.Metzler, B, Hu, Y, Dietrich, H, and Xu, Q (2000). Increasedexpression and activation of stress-activated proteinkinases/c-Jun NH(2)-terminal protein kinases inatherosclerotic lesions coincide with p53. Am J Pathol 156,1875-1886.Miano, JM, Tota, RR, Vlasic, N, Danishefsky, KJ, andStemerman, MB (1990). Early proto-oncogene expression inrat aortic smooth muscle cells following endothelial removal.Am J Pathol 137, 761-765.Miano, JM, Vlasic, N, Tota, RR, and Stemerman, MB (1993).Localization of Fos and Jun proteins in rat aortic smoothmuscle cells after vascular injury. Am J Pathol 142, 715-724.Miniati, DN, Hoyt, EG, Feeley, BT, Poston, RS, and Robbins,RC (2000). Ex vivo antisense oligonucleotides toproliferating cell nuclear antigen and Cdc2 kinase inhibitgraft coronary artery disease. Circulation 102, III237-242.Morgan, DO (1995). Principles of CDK regulation. Nature 374,131-134.Morishita, R, Gibbons, GH, Ellison, KE, Nakajima, M, von derLeyen, H, Zhang, L, Kaneda, Y, Ogihara, T, and Dzau, VJ(1994a). Intimal hyperplasia after vascular injury is inhibited86

Gene Therapy and Molecular Biology Vol 7, page 87by antisense cdk2 kinase oligonucleotides. J Clin Invest 93,1458-1464.Morishita, R, Gibbons, GH, Ellison, KE, Nakajima, M, Zhang, L,Kaneda, Y, Ogihara, T, and Dzau, VJ (1993). Singleintraluminal delivery of antisense cdc2 kinase andproliferating-cell nuclear antigen oligonucleotides results inchronic inhibition of neointimal hyperplasia. Proc Natl AcadSci USA 90, 8474-8478.Morishita, R, Gibbons, GH, Horiuchi, M, Ellison, KE, Nakajima,M, Zhang, L, Kaneda, Y, Ogihara, T, and Dzau, VJ (1995).A gene therapy strategy using a transcription factor decoy ofthe E2F binding site inhibits smooth muscle cell proliferationin vivo. Proc Natl Acad Sci USA 92, 5855-5859.Morishita, R, Gibbons, GH, Kaneda, Y, Ogihara, T, and Dzau,VJ (1994b). Pharmacokinetics of antisenseoligodeoxyribonucleotides (cyclin B1 and CDC 2 kinase) inthe vessel wall in vivo: enhanced therapeutic utility forrestenosis by HVJ-liposome delivery. Gene 149, 13-19.Nakamura, T, Morishita, R, Asai, T, Tsuboniwa, N, Aoki, M,Sakonjo, H, Yamasaki, K, Hashiya, N, Kaneda, Y, andOgihara, T (2002). Molecular strategy using cis-element'decoy' of E2F binding site inhibits neointimal formation inporcine balloon-injured coronary artery model. Gene Ther 9,488-494.Nobuyoshi, M, Kimura, T, Ohishi, H, Horiuchi, H, Nosaka, H,Hamasaki, N, Yokoi, H, and Kim, K (1991). Restenosis afterpercutaneous transluminal coronary angioplasty: pathologicobservations in 20 patients. J Am Coll Cardiol 17, 433-439.O'Brien, ER, Alpers, CE, Stewart, DK, Ferguson, M, Tran, N,Gordon, D, Benditt, EP, Hinohara, T, Simpson, JB, andSchwartz, SM (1993). Proliferation in primary and restenoticcoronary atherectomy tissue. Implications forantiproliferative therapy. Circ Res 73, 223-231.O'Brien, ER, Urieli-Shoval, S, Garvin, MR, Stewart, DK,Hinohara, T, Simpson, JB, Benditt, EP, and Schwartz, SM(2000). Replication in restenotic atherectomy tissue.Atherosclerosis 152, 117-126.Olson, NE, Kozlowski, J, and Reidy, MA (2000). Proliferation ofintimal smooth muscle cells. Attenuation of basic fibroblastgrowth factor 2-stimulated proliferation is associated withincreased expression of cell cycle inhibitors. J Biol Chem275, 11270-11277.Orekhov, AN, Andreeva, ER, Mikhailova, IA, and Gordon, D(1998). Cell proliferation in normal and atherosclerotichuman aorta: proliferative splash in lipid-rich lesions.Atherosclerosis 139, 41-48.Parkes, JL, Cardell, RR, Hubbard, FC, Hubbard, D, Meltzer, A,and Penn, A (1991). Cultured human atherosclerotic plaquesmooth muscle cells retain transforming potential and displayenhanced expression of the myc protooncogene. Am JPathol 138, 765-775.Peeper, DS, Upton, TM, Ladha, MH, Neuman, E, Zalvide, J,Bernards, R, DeCaprio, JA, and Ewen, ME (1997). Rassignalling linked to the cell-cycle machinery by theretinoblastoma protein. Nature 386, 177-181.Perlman, H, Luo, Z, Krasinski, K, Le Roux, A, Mahfoudi, A,Smith, RC, Branellec, D, and Walsh, K (1999). Adenovirusmediateddelivery of the Gax transcription factor to ratcarotid arteries inhibits smooth muscle proliferation andinduces apoptosis. Gene Ther 6, 758-763.Perlman, H, Suzuki, E, Simonson, M, Smith, RC, and Walsh, K(1998). GATA-6 induces p21 Cip1 expression and G1 cellcycle arrest. J Biol Chem 273, 13713-13718.Philipp-Staheli, J, Payne, SR, and Kemp, CJ (2001). p27 Kip1 :regulation and function of a haploinsufficient tumorsuppressor and its misregulation in cancer. Exp Cell Res264, 148-168.Pickering, JG, Weir, L, Jekanowski, J, Kearney, MA, and Isner,JM (1993). Proliferative activity in peripheral and coronaryatherosclerotic plaque among patients undergoingpercutaneous revascularization. J Clin Invest 91, 1469-1480.Pronk, GJ, and Bos, JL (1994). The role of p21ras in receptortyrosine kinase signalling. Biochim Biophys Acta 1198,131-147.Pukac, LA, Castellot, JJ, Jr., Wright, TC, Jr., Caleb, BL, andKarnovsky, MJ (1990). Heparin inhibits c-fos and c-mycmRNA expression in vascular smooth muscle cells. CellRegul 1, 435-443.Pyles, JM, March, KL, Mehdi, K, Wilenski, RL, and Adam, LP(1997). Activation of MAP Kinase in vivo follows balloonoverstretch injury of porcine coronary and carotid arteries.Circ Res 81, 904-910.Reilly, CF, Kindy, MS, Brown, KE, Rosenberg, RD, andSonenshein, GE (1989). Heparin prevents vascular smoothmuscle cell progression through the G1 phase of the cellcycle. J Biol Chem 264, 6990-6995.Rekhter, MD, and Gordon, D (1995). Active proliferation ofdifferent cell types, including lymphocytes, in humanatherosclerotic plaques. Am J Pathol 147, 668-677.Robinson, KA, Chronos, NA, Schieffer, E, Palmer, SJ, Cipolla,GD, Milner, PG, and King, SB, 3rd (1997). Endoluminallocal delivery of PCNA/cdc2 antisense oligonucleotides byporous balloon catheter does not affect neointima formationor vessel size in the pig coronary artery model ofpostangioplasty restenosis. Cathet Cardiovasc Diagn 41,348-353.Rodriguez-Campos, A, Ruiz-Enriquez, P, Faraudo, S, andBadimon, L (2001). Mitogen-induced p53 downregulationprecedes vascular smooth muscle cell migration from healthytunica media and proliferation. Arterioscler Thromb VascBiol 21, 214-219.Roque, F, Mon, G, Belardi, J, Rodriguez, A, Grinfeld, L, Long,R, Grossman, S, Malcolm, A, Zon, G, Ormont, ML,Fischman, DL, Shi, Y, and Zalewski, A (2001a). Safety ofintracoronary administration of c-myc antisense oligomersafter percutaneous transluminal coronary angioplasty(PTCA). Antisense Nucleic Acid Drug Dev 11, 99-106.Roque, M, Reis, ED, Cordon-Cardo, C, Taubman, MB, Fallon,JT, Fuster, V, and Badimon, JJ (2001b). Effect of p27deficiency and rapamycin on intimal hyperplasia: in vivo andin vitro studies using a p27 knockout mouse model. LabInvest 81, 895-903.Rosenfeld, ME, and Ross, R (1990). Macrophage and smoothmuscle cell proliferation in atherosclerotic lesions of WHHLand comparably hypercholesterolemic fat-fed rabbits.Arteriosclerosis 10, 680-687.Ross, R (1999). Atherosclerosis: an inflammatory disease. NEngl J Med 340, 115-126.Rothman, A, Woler, B, Button, D, and Taylor, P (1994).Immediate-early gene expression in respose to hypertrophicand proliferative stimuli in pulmonary arterial smooth musclecells. J Biol Chem 269, 6399-6404.Scheinman, M, Ascher, E, Levi, GS, Hingorani, A, Shirazian, D,and Seth, P (1999). p53 gene transfer to the injured ratcarotid artery decreases neointimal formation. J Vasc Surg29, 360-369.Scott, S, O'Sullivan, M, Hafizi, S, Shapiro, LM, and Bennett, MR(2002). Human vascular smooth muscle cells from restenosisor in-stent stenosis sites demonstrate enhanced responses top53. Implications for brachytherapy and drug treatment forrestenosis. Circ Res 90, 398-404.Shi, Y, Dodge, GR, Hall, DJ, Desrochers, PE, Fard, A, Shaheen,F, Talbot, C, Yurgenev, L, and Zalewski, A (1994a).Inhibition of type 1 collagen synthesis in vascular smooth87

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular diseasemuscle cells by c-myc antisense oligomers. Circulation 90,I-147.Shi, Y, Fard, A, Galeo, A, Hutchinson, HG, Vermani, P, Dodge,GR, Hall, DJ, Shaheen, F, and Zalewski, A (1994b).Transcatheter delivery of c-myc antisense oligomers reducesneointimal formation in a porcine model of coronary arteryballoon injury. Circulation 90, 944-951.Shi, Y, Hutchinson, HG, Hall, DJ, and Zalewski, A (1993).Downregulation of c-myc expression by antisenseoligonucleotides inhibits proliferation of human smoothmuscle cells. Circulation 88, 1190-1195.Simon, MC (1995). Gotta have GATA. Nat Genetics 11, 9-11.Simons, M, Edelman, ER, DeKeyser, J-L, Langer, R, andRosenberg, RD (1992). Antisense c-myb oligonucleotidesinhibit intimal arterial smooth muscle cell accumulation invivo. Nature 359, 67-70.Simons, M, and Rosenberg, RD (1992). Antisense nonmusclemyosin heavy chain and c-myb oligonucleotides supresssmooth muscle cell proliferation in vitro. Circ Res 70, 835-843.Smith, RC, Branellec, D, Gorski, DH, Guo, K, Perlman, H,Dedieu, J-F, Pastore, C, Mahfoudi, A, Denèfle, P, Isner, JM,and Walsh, K (1997a). p21 CIP1 -mediated inhibition of cellproliferation by overexpression of the Gax homeodomaingene. Genes Dev 11, 1674-1689.Smith, RC, Wills, KN, Antelman, D, Perlman, H, Truong, LN,Krasinski, K, and Walsh, K (1997b). Adenoviral constructsencoding phosphorylation-competent full-length andtruncated forms of the human retinoblastoma protein inhibitmyocyte proliferation and neointima formation. Circulation96, 1899-1905.Speir, E, Modali, R, and Huang, E-S (1994). Potential role ofhuman cytomegalovirus and p53 interaction in coronaryrestenosis. Science 265, 391-394.Spraragen, SC, Bond, VP, and Dahl, LK (1962). Role ofhyperplasia in vascular lesions of cholesterol-fed rabbitsstudied with thymidine-3H autoradiography. Circ Res 11,329-336.Steinberg, D (2002). Atherogenesis in perspective:hypercholesterolemia and inflammation as partners in crime.Nat Med 8, 1211-1217.Stevaux, O, and Dyson, NJ (2002). A revised picture of the E2Ftranscriptional network and RB function. Curr Opin CellBiol 14, 684-691.Su, JZ, Fukuda, N, Hu, WY, and Kanmatsuse, K (2000).Ribozyme to human TGF-beta1 mRNA inhibits theproliferation of human vascular smooth muscle cells.Biochem Biophys Res Commun 278, 401-407.Suzuki, E, Evans, T, Lowry, J, Truong, L, Bell, DW, Testa, JR,and Walsh, K (1996). The human GATA-6 gene: structure,chromosomal location and regulation of expression by tissuespecificand mitogen-responsive signals. Genomics 38, 283-290.Suzuki, J-I, Isobe, M, Morishita, R, Aoki, M, Horie, S, Okubo,Y, Kaneda, Y, Sawa, Y, Matsuda, H, Ogihara, T, andSekiguchi, M (1997). Prevention of graft coronaryarteriosclerosis by antisense cdk2 kinase oligonucleotide.Nat Med 3, 900-903.Sylvester, AM, Chen, D, Krasinski, K, and Andrés, V (1998).Role of c-fos and E2F in the induction of cyclin Atranscription and vascular smooth muscle cell proliferation. JClin Invest 101, 940-948.Tanaka, K, Zou, JP, Takeda, K, Ferrans, VJ, Sandford, GR,Johnson, TM, Finkel, T, and Epstein, SE (1999). Effects ofhuman cytomegalovirus immediate-early proteins on p53-mediated apoptosis in coronary artery smooth muscle cells.Circulation 99, 1656-1659.Tanner, FC, Boehm, M, Akyürek, LM, San, H, Yang, Z-Y,Tashiro, J, Nabel, GJ, and Nabel, EG (2000). Differentialeffects of the cyclin-dependent kinase inhibitors p27 Kip1 ,p21 Cip1 , and p16 Ink4 on vascular smooth muscle cellproliferation. Circulation 101, 2022-2025.Tanner, FC, Yang, Z-Y, Duckers, E, Gordon, D, Nabel, GJ, andNabel, EG (1998). Expression of cyclin-dependent kinaseinhibitors in vascular disease. Circ Res 82, 396-403.Ueno, H, Masuda, S, SNishio, S, Li, JJ, Yamamoto, H, andTakeshita, A (1997a). Adenovirus-mediated transfer ofcyclin-dependent kinase inhibitor p21 suppresses neointimalformation in the balloon-injured rat carotid arteries in vivo.Ann N Y Acad Sci 811, 401-411.Ueno, H, Yamamoto, H, Ito, S-i, Li, J-J, and Takeshita, A(1997b). Adenovirus-mediated transfer of a dominantnegativeH-ras suppresses neointimal formation in ballooninjuredarteries in vivo. Arterioscler Thromb Vasc Biol 17,898-904.van Vlijmen, BJ, Gerritsen, G, Franken, AL, Boesten, LS,Kockx, MM, Gijbels, MJ, Vierboom, MP, van Eck, M, vanDe Water, B, van Berkel, TJ, and Havekes, LM (2001).Macrophage p53 deficiency leads to enhancedatherosclerosis in APOE*3- Leiden transgenic mice. CircRes 88, 780-786.Veinot, JP, Ma, X, Jelley, J, and O'Brien, ER (1998). Preliminaryclinical experience with the pullback atherectomy catheterand the study of proliferation in coronary plaques. Can JCardiol 14, 1457-1463.Villa, AE, Guzman, LA, Poptic, EJ, Labhasetwar, V, D'Souza, S,Farrell, CL, Plow, EF, Levy, RJ, DiCorleto, PE, and Topol,EJ (1995). Effects of antisense c-myb oligonucleotides onvascular smooth muscle cell proliferation and response tovessel wall injury. Circ Res 76, 505-513.Wang, W, Chen, HJ, Schwartz, A, Cannon, PJ, Stein, CA, andRabbani, LE (1996). Sequence-independent inhibition of invitro vascular smooth muscle cell proliferation, migration,and in vivo neointimal formation by phosphorothioateoligodeoxynucleotides. J Clin Invest 98, 443-450.Wei, GL, Krasinski, K, Kearney, M, Isner, JM, Walsh, K, andAndrés, V (1997). Temporally and spatially coordinatedexpression of cell cycle regulatory factors after angioplasty.Circ Res 80, 418-426.Weir, L, Chen, D, Pastore, C, Isner, JM, and Walsh, K (1995).Expression of GAX, a growth-arrest homeobox gene, israpidly down-regulated in the rat carotid artery during theproliferative response to balloon injury. J Biol Chem 270,5457-5461.Winston, JT, Coats, SR, Wang, Y-Z, and Pledger, WJ (1996).Regulation of the cell cycle machinery by oncogenic ras.Oncogene 12, 127-134.Yamamoto, K, Morishita, R, Tomita, N, Shimozato, T,Nakagami, H, Kikuchi, A, Aoki, M, Higaki, J, Kaneda, Y,and Ogihara, T (2000). Ribozyme oligonucleotides againsttransforming growth factor-beta inhibited neointimalformation after vascular injury in rat model: potentialapplication of ribozyme strategy to treat cardiovasculardisease. Circulation 102, 1308-1314.Yang, Z, Oemar, BS, Carrel, T, Kipfer, B, Julmy, F, and Lüscher,TF (1998). Different proliferative properties of smoothmuscle cells of human arterial and venous bypass vessels:role of PDGF receptors, mitogen-activated protein kinase,and cyclin-dependent kinase inhibitors. Circulation 97, 181-187.Yang, Z-Y, Simari, RD, Perkins, ND, San, H, Gordon, D, Nabel,GJ, and Nabel, EG (1996). Role of p21 cyclin-dependentkinase inhibitor in limiting intimal cell proliferation inresponse to arterial injury. Proc Natl Acad Sci USA 93,7905-7910.88

Gene Therapy and Molecular Biology Vol 7, page 89Yonemitsu, Y, Kaneda, Y, Tanaka, S, Nakashima, Y, Komori, K,Sugimachi, K, and Sueishi, K (1998). Transfer of wild-typep53 gene effectively inhibits vascular smooth muscle cellproliferation in vitro and in vivo. Circ Res 82, 147-156.Zhou, YF, Leon, MB, Waclawiw, MA, Popma, JJ, Yu, ZX,Finkel, T, and Epstein, SE (1996). Association between priorcytomegalovirus infection and the risk of restenosis aftercoronary atherectomy. N Engl J Med 335, 624-630.Zhou, YF, Yu, ZX, Wanishsawad, C, Shou, M, and Epstein, SE(1999). The immediate early gene products of humancytomegalovirus increase vascular smooth muscle cellmigration, proliferation, and expression of PDGF betareceptor.Biochem Biophys Res Commun 256, 608-613.Zhu, NL, Wu, L, Liu, PX, Gordon, EM, Anderson, WF, Starnes,VA, and Hall, FL (1997). Downregulation of cyclin G1expression by retrovirus-mediated antisense gene transferinhibits vascular smooth muscle cell proliferation andneointima formation. Circulation 96, 628-635.Zou, X, Rudchenko, S, Wong, K-k, and Calame, K (1997).Induction of c-myc transcription by the v-Abl tyrosine kinaserequires Ras, Raf1, and cyclin-dependent kinases. GenesDev 11, 654-662.Zou, Y, Hu, Y, Metzler, B, and Xu, Q (1998). Signaltransduction in arteriosclerosis: mechanical stress-activatedMAP kinases in vascular smooth muscle cells (review). Int JMol Med 1, 827-834.89

Gascón-Irún et al: Gene therapy antiproliferative strategies against cardiovascular disease90

Gene Therapy and Molecular Biology Vol 7, page 91Gene Ther Mol Biol Vol 7, 91-97, 2003Regulation of the Sp/KLF-family of transcriptionfactors: focus on post-transcriptional modificationand protein-protein interaction in the context ofchromatinReview ArticleToru Suzuki 1,2* , Masami Horikoshi 3,4 and Ryozo Nagai 11Department of Cardiovascular Medicine, 2 Department of Clinical Bioinformatics, Graduate School of Medicine, TheUniversity of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan, 3 Laboratory of Developmental Biology, Instituteof Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan, 4Horikoshi Gene Selector Project, Exploratory Research for Advanced Technology (ERATO), Japan Science andTechnology Corporation, 5-9-6 Tokodai, Tsukuba, Ibaraki 300-2635 Japan__________________________________________________________________________________*Correspondence:Toru Suzuki, MD, PhD, Department of Cardiovascular Medicine, Department of Clinical Bioinformatics, GraduateSchool of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan; Tel: 81-3-3815-5411; Fax: 81-3-5800-8824; e-mail: torusuzu-tky@umin.ac.jpKey words: transcription factors, gene regulation, chromatin, Sp1, acetyltransferase, nucleosome remodelingReceived: 25 June 2003; Accepted: 10 July 2003; electronically published: July 2003SummaryThe Sp1- and Krüppel-like zinc finger transcription factor family is a rapidly expanding and highlighted group offactors given important biological roles. Understanding specific regulation is important to dissect individualfunctions. In this collective review, the regulation of this family of transcription factors with a particular focus onpost-transcriptional modification and protein-protein interaction in the context of chromatin will be discussed.Studies by ourselves and others show that the zinc finger DNA-binding domain region of these factors mediatesimportant regulatory interactions and modifications which may explain at least in part their specific regulation.Their possible implications in gene therapy are discussed.I. IntroductionThe zinc finger motif (paired cysteine and histidinetype) was discovered approximately two decades ago(Diakun et al, 1986). Since then, we have learnt that this isone of the major motifs for proteins in the cell rangingfrom enzymes to transcription factors. Recent analysis ofthe human genome showed that transcription factors withthis zinc finger motif have evolved in cascadingmagnitude as shown by their increased genomiccomplexity in eukaryotes (Tupler et al, 2001). At present,the paired-cysteine and histidine-type (C 2 H 2 -type) zincfinger transcription factors are thought to be one of themost important type of regulatory transcription factor inthe eukaryotic cell. Among these factors, the Sp/KLF (forSp1- and Krüppel-like factor) family of transcriptionfactors has received recent attention due to important rolesin development, differentiation, and oncogenic processes(Philipsen and Suske, 1999; Turner and Crossley, 1999;Dang et al, 2000; Bieker, 2001; Black et al, 2001;Bouwman and Philipsen, 2002; Kaczynski et al, 2003).DNA-binding activators/repressors bind in asequence-specific manner to their cognate binding sites inenhancers/silencers and core promoter regions andactivate/repress transcription of genes throughcombinatorial effects with the general transcriptionmachinery (Horikoshi et al. 1988a, b; Zawel and Reinberg1995). The DNA-binding transcription factor has beenclassically shown to possess modular functional regionsconsisting of an activation/regulatory domain whichregulates transcription through interactions with basaltranscription machinery and the DNA-binding domain(DBD) which specifies the target promoter gene (Ptashneand Gann, 1990; Zawel and Reinberg, 1995).The DNA-binding transcription factor is regulated atmultiple steps. Presence as dictated by spatial expression(e.g. ubiquitous versus restricted expression) in addition totemporal regulation (e.g. constitutive versus inducibleexpression) plays a primary regulatory role. Sequence-91

Suzuki et al: Regulation of the Sp/KLF-family of transcription factorsspecific DNA-binding is further critically important fordictating gene-specific actions. DNA-binding transcriptionfactors with common DNA-binding domains often bindsimilar DNA sequences (e.g. basic helix-loop-helixproteins bind E-boxes, homeoproteins bind A/T-rich sites)but additional regulatory steps must be present as thecomplexity of these factors in undertaking specificfunctions cannot be readily explained by their expressionpatterns and sequence-specific DNA binding propertiesalone. Regulation through differential protein-proteininteractions and/or chemical modifications (e.g.phosphorylation, acetylation) further contribute to theirdifferential functions. In the present review, the regulationof the Sp/KLF-family of transcription factors with aparticular focus on post-transcriptional modification andprotein-protein interactions in the context of chromatinwill be discussed.II. Basic classification of Sp/KLFfactorsThe Sp/KLF family of zinc-finger transcriptionfactors are comprised of over 20 mammalian familymembers which have in common three contiguous C 2 H 2 -type zinc fingers at the carboxyl-terminus whichcomprises the DNA-binding domain (Philipsen and Suske,1999; Turner and Crossley, 1999; Dang et al, 2000;Bieker, 2001; Black et al, 2001; Bouwman and Philipsen,2002; Kaczynski et al, 2003). Sp/KLF family memberscan be classified into Sp- and KLF-subsets based on theirsimilarities. The Sp-subtype is based on the foundingubiquitous factor Sp1 (Dynan and Tjian, 1983), and theKLF-subtype is based on the Drosophila Krüppel gene(Preiss et al, 1985). The first systematic classification usedto distinguish mammalian Krüppel-like factors wasdemonstrated in a distinction with the GLI subgroup,which defined the consensus amino acid finger sequencefor the Krüppel subgroup to be[Y/F]XCX2CX3FX5LX2HXRXHTGEKP (Ruppert et al,1988). The Sp subgroup is based on similarity to thefounding factor Sp1. Among the KLFs are erythroiddifferentiation factor EKLF/KLF1 (Miller and Bieker,1993) and the tumor suppressor geneKLF6/GBF/Zf9/COPEB which we and others identified asa cellular factor possibly involved in HIV-1 transcription(Koritschoner et al, 1997; Suzuki et al, 1998; Narla et al,2001). We have recently shown by gene knockout studiesthat the protooncogene KLF5/BTEB2/IKLF (Sogawa et al,1993; Shi et al, 1999) is important for cardiovascularremodeling in response to stress (Shindo et al, 2002).At present, the annotation of this family of factorsuses a numbering system in order of identification inaccordance with an international collaboration to unify thenomenclature. Factors of the Sp-subset have six to eightmembers, whereas the KLF-subset have approximately 15members, and are still increasing in numbers. Contrary toinitial expectations that this family of factors would likelyhave redundant functions, they in fact have importantindividual biological functions as shown by gene knockoutstudies (e.g. EKLF/KLF1, LKLF/KLF2, KLF5). However,the underlying mechanisms governing their specificfunctions and regulation are poorly understood.III. Differential regulation of Sp/KLFfactorsThe mechanisms underlying specificity of this familyof factors have been the topic of great interest amongconcerned researchers to understand the basis for theirindividual functions. As the paired cysteine-histidine typezinc finger is a DNA-binding motif, initial studies beganby investigations of DNA-binding characteristics. One ofthe hallmark features of the Sp/KLF factors is that theybind to similar GC-rich sites and/or CACC-boxes. Wellstudied crystal structure analyses of DNA-binding zincfinger transcription factors have allowed the prediction ofthe cognate DNA-binding sequence from the primaryamino acid structure (Klevit, 1991; Suzuki et al, 1994).Amino acids which contact DNA reside in the α-helicalregion of the zinc finger. As these critical amino acids arehighly conserved in Sp/KLF zinc finger transcriptionfactors, it is tempting to assume that they likely sharesimilar DNA binding properties.Closer examination of this zinc finger region,however, shows discrete yet distinct differences. Forinstance, the third amino acid critical for DNA binding ofthe third zinc finger, and in the amino acids N-terminaladjacent to the first amino acid critical for DNA bindingand the third amino acid critical for DNA binding in eachof the zinc fingers differ (Suzuki et al, 1998). Therelevance of these differences in the context of DNAbindingspecificity or affinity remains to be clarified. Theoptimal cognate binding sequence of selected factors havebeen shown experimentally which showed that Sp1 bindsthe sequence 5'-GGGGCGGGGT-3' (Thiesen et al, 1990)and KLF4/GKLF binds the sequence 5'-G/AG/AGGC/TGC/T-3' (Shields and Yang, 1998) whichis a derivative of the CACC-box and BTE-element (whichis a GC-rich site which binds BTEB1). Collectively, it isgenerally thought that this family of factors bind similarGC-rich sequences in a sequence-specific manner with abinding selectivity which does not allow individual factorsto be clearly discriminated based on their DNA-bindingcharacteristics alone.It is important to note here, however, that DNAbindingcharacteristics likely differ in the context ofchromatin DNA as separate from the naked DNA-stateoften used for biochemical experiments. One importantexample using transgenic mice showed that EKLF/KLF1preferentially binds the beta-globin locus site in vivowhich had been shown to bind both EKLF and Sp1 inbiochemical studies (Gillemans et al, 1998).We too had been interested in understanding whetherthere is specific binding of factors to GC-rich sites in vivowhich are not reflected in biochemical studies in vitro. Forthis, we used a yeast one-hybrid assay using the GC-richsites of the HIV-1 core promoter which have been shownto bind Sp1 to investigate what factors actually bind thissite. The binding site probe used for the assay wasintegrated into the yeast genome to better reflect cellular92

Gene Therapy and Molecular Biology Vol 7, page 93conditions. Although a mammalian environment was notused and as there was limitation by overexpression offactors, we believed that the yeast environment would bebetter reflective of the eukaryotic intracellularenvironment as compared to the traditional southwesternfilter hybridization or affinity chromatography techniques.Our studies interestingly resulted in the isolation ofKLF6/GBF, a novel KLF factor which shows similar GCrichbinding properties as Sp1 (Suzuki et al, 1998). Thiswas the only Sp/KLF factor identified in our screen thussuggesting the possibility that distinct factors may bindGC-rich sites in the cellular environment. Therefore, atpresent, while biochemical studies do show that Sp/KLFfactors bind similar GC-rich sites, the actual intracellularenvironment especially in the context of chromatin mayallow for preferential binding of different factors. Thisissue on effect of intracellular context remains to befurther explored.IV. Regulation through chemicalmodifications and/or differential proteinproteininteractionsRegulation through differential protein-proteininteractions and/or chemical modifications (e.g.acetylation) are further likely to contribute to thedifferential functions of Sp/KLF factors. We have focusedour attention on the role of the DNA-binding domain(DBD) because it is most reasonable, if not optimal, forregulating DNA-associated events such as promoter accessand topological changes given its ability and activity tobind DNA (Figure 1). Amino acid differences are evidentin the zinc finger DNA-binding domain of Sp/KLF factors,although there is extensive conservation overall. Asidefrom the likelihood of affecting DNA-binding properties,these differences in primary structure and quite possibly inthe overall conformation of the folded protein may have aprofound effect on post-translational modifications inaddition to protein-protein interactions.A. Regulation by chemical modificationFocusing on the regulatory role of acetylation onSp/KLF transcription factors, we have shown differentialregulation through interaction and acetylation on theDNA-binding domain by the coactivator/acetylase p300(Suzuki et al, 2000). Acetylation is an important nuclearregulatory signal which regulates transcriptionalprocesses, importantly with biological implications whichinclude regulation of development, differentiation andoncogenesis (Brownell and Allis, 1996; Cheung et al,2000; Nakatani 2001; Freiman and Tjian, 2003) whichclosely resembles the roles of Sp/KLF family members.We thought that the Sp/KLF-factors might bedifferently regulated by acetylation and showed that thecoactivator/acetylase p300 but not the MYST-typeacetylase Tip60 specifically interacts and acetylates Sp1but not KLF6 through the zinc finger DNA-bindingdomain, and further that DNA binding inhibits thisinteraction and acetylation (Suzuki et al, 2000). Interactionof p300 acetyltransferase region and the Sp1 zinc fingerDNA-binding domain stimulates the DNA-binding activityof the latter, while acetylation per se has only marginaleffects. While much is known of acetylation in general, itsregulation and implications are still poorly understood.A similar mechanism has been shown forKLF13/FKLF2. KLF13 is acetylated both by PCAF andCBP, as well as interact through the zinc finger DNAbindingdomain of KLF13. The acetyltransferase regionsof PCAF and CBP stimulate KLF13 binding to its cognateDNA-binding site. These findings suggest and furthersupport that acetyltransferase interaction with the zincfinger DNA-binding domain of at least KLFs affectsDNA-binding activity (Song et al, 2002). Acetylation ofKLF13 by CBP has been further shown to inhibit KLF13DNA-binding activity, and that PCAFFigure 1. Regulation of DNA-binding transcription factors in general. Note that there are modular activation and DNA-binding domains.Regulation through interaction and modification of DNA-binding domains is poorly understood. We have focused our studies on the roleof the zinc finger DNA-binding domain for Sp/KLF factors. The active role of the DNA-binding domain is suggested in DNA-bindingprocesses not only for naked DNA but also in the context of nucleosomal DNA.93

Suzuki et al: Regulation of the Sp/KLF-family of transcription factorsblocks CBP acetylation and its disruption of DNA binding(Song et al, 2003).Our findings on Sp1 and further those on KLF13provide an attractive model of promoter access bycooperative action of DNA-binding activator withcoactivator/acetyltransferase. Important here is that thereis a concerted interaction between these two factors whichfacilitates promoter access (Figure 2). The regulatory andactivation domains likely play an additional role. This is incontrast to the extant model of recruitment ofcoactivator/acetyltransferase to the DNA-binding activatorinvolving specific binding by the latter to its cognatebinding site with subsequent recruitment of the former tothe promoter (Ogryzko et al, 1996). Our interpretation andmodel explains one of the limitations of this prior modelon how the DNA-binding activator accesses its cognatesite or how interaction with coactivator/acetyltransferasesaffects this reaction which were issues which remainedunclear.Other Sp/KLF factors are also acetylated in the zincfinger DNA-binding domain. EKLF/KLF1 is acetylated byp300 and its homologue CBP at two lysine residues, oneresiding in the DNA-binding zinc finger domain and theother in the transactivation domain. The mutation of thezinc finger acetylated residue does not affect DNAbindingactivity and the individual role of its acetylation isunclear, but mutation of the transactivation domain lysineresidue results in decreased transactivation and acetylationcollectively increased affinity for the SWI/SNF chromatinremodeling factors (Zhang and Bieker, 1998; Zhang et al,2001). Sp3 is acetylated in its inhibitory domain lyingbetween the glutamine-rich activation domain and zincfinger DNA-binding domain. Acetylation of this lysineresidue regulates transcriptional activity (Braun et al,2001).There are other modifications such asphosphorylation, methylation, glycosylation,ubiquitination, and SUMOylation (SUMO; smallubiquitin-related modifier) among others. From theperspective of the DNA-binding domain, cell-cycledependent phosphorylation by a putative kinase has beenreported for Sp1 (Black et al, 1999). Casein kinase II alsophosphorylates the second zinc finger of Sp1 resulting in areduction in DNA-binding activity (Armstrong et al,1997). PKC-zeta also binds and phosphorylates the zincfinger region of Sp1 which is suggested to result intranscriptional activation (Pal et al, 1998). Sp1 is alsoglycosylated (Jackson and Tjian, 1988). Much of ourknowledge on the regulatory mechanisms of the Sp/KLFfactors at present are centered on Sp1 as it was one of thefirst eukaryotic DNA-binding regulatory transcriptionfactors ever identified and serves as an excellent molecularmodel to dissect and understand mechanisms oftranscriptional activation.A recent report has further shown that Sp3 isSUMOylated at the same residue that is acetylated(Sapetschnig et al, 2002). While we still have much tolearn on post-transcriptional modifications, cross-talk andco-regulation of signaling pathways not only for lysinemodifications but also for coupling of pathways such as aphosphorylation-acetylation cascade will likely show thecomplex nature of regulation by chemical modifications.B. Regulation by protein-proteininteractionThe zinc finger DBD motif, while binding DNA, isalso an interface for protein-protein interaction such ashomo- and hetero-dimerization in addition to proteinproteininteractions with heterologous proteins (MacKayand Crossley 1998)Figure 2. Model of promoter access as mediated by interaction betweeen the zinc finger DNA-binding domain (DBD) of the Sp/KLFtranscription factor and catalytic region of acetyltransferase (HAT) (e.g. p300 for Sp1 and PCAF for KLF13). Interaction between theactivation domain (AD) of the DNA-binding factor and regulatory domain (RD) of the acetyltransferase is unknown but is likely to playan additional role to retain the DNA-binding factor and HAT on the promoter.94

Gene Therapy and Molecular Biology Vol 7, page 95which results in specific regulation. In general, whilemuch research on transcription factors has focused on therole of the activation domain to mediate regulation (e.g.activation, repression, ligand-dependent modulation, etc.)(Horikoshi et al, 1988a,b; Roeder, 1996; Lemon and Tjian,2000), functions of the DBD other than its DNA-bindingactivity have received little attention (Wagner and Green,1994). Here the discussion will focus on the fact thatnumerous chromatin remodeling factors and other factorswhich act on transcription at the level of higher-orderDNA interact and regulate through the zinc finger DBD(Figure 1).As mentioned in the above section on acetylation,Sp1 and KLF13 catalytically interact withacetyltransferase (e.g. p300 with Sp1, and PCAF and CBPwith KLF13). Importantly, they also stably interactthrough the zinc finger DBD which results in stimulationof DNA-binding activity of the DNA-binding transcriptionfactor. These findings allow for the model of promoteraccess as shown in Figure 1. While we assume a priorithat DNA-binding factors recruit acetyltransferase andother chromatin remodeling factors to DNA after they arepre-bound to DNA, these results suggest that they in factshow interaction in solution and that DNA binding isinhibitory to interaction. This suggests that interactionpromotes access of the DNA-binding factor to DNA but isreleased once bound to DNA.Deacetylases also bind Sp/KLF factors through thezinc finger DNA-binding domain. Both Sp1 andEKLF/KLF1 have been shown to associate with HDAC1.Both Sp1 and EKLF bind HDAC1 through the zinc fingerDNA-binding domain. Interaction of Sp1 and HDAC1 isthought to be repressive on Sp1 transcription becausecoexpression of E2F1, which interferes with HDAC1binding to Sp1, abolishes Sp1-mediated transcriptionalrepression (Doetzlhofer et al, 1999). EKLF also bindsHDAC1 through its zinc finger DNA-binding domainwhich results in transcriptional regulation (Chen andBieker, 2001). From within the HDAC-associatedcorepressor complex, sin3A also binds EKLF through thezinc finger DNA-binding domain.Further, the zinc finger DNA-binding domains ofSp1 and that of EKLF interact with the ATP-dependentnucleosome remodeling enzyme Swi/Snf (Kadam et al,2000). Two SWI/SNF subunits (BRG1 and BAF155) arerequired for targeted chromatin remodeling andtranscriptional activation by EKLF in vitro. Remodeling isachieved with only the BRG1-BAF155 minimal complexand the EKLF zinc finger DBD, whereas transcriptionadditionally requires an activation domain.We have recently shown that the zinc finger DNAbindingdomain of Sp1 mediates interaction with thehistone chaperone TAF-I (template activatingfactor)(Suzuki et al, 2003). Interaction is specific, asdifferent subsets of DNA-binding factors do not bindTAF-I and as other ATP-independent nucleosomeremodeling enzymes do not bind Sp1. TAF-I negativelyregulates Sp1 activity by inhibiting DNA binding, andlikely as a consequence of this, regulates Sp1-mediatedpromoter activation.Based on these findings, the Sp1 DBD interacts withall three major chromatin-related factors consisting ofchemical modification enzymes (e.g. acetyltransferasep300), ATP-dependent nucleosome assembly factor (e.g.SWI/SNF) and histone chaperone (e.g. TAF-I)(Figure 3).This finding is of particular interest because it implicatesthe DBD to play a likely role in mediating transcriptionalregulatory processes in eukaryotes at the chromatin level.Although interaction with individual chromatinremodeling factors has been documented for numerousproteins, as interaction with all three chromatinremodeling factors has only been reported previously forhistones, the DNA-transcription factor, and importantly itsDNA-binding domain, may, therefore, represent a vitaltarget for chromatin-related transcriptional processesFigure 3. Model (deducted from Sp1 interactions) explaining how the DNA-binding domain of the transcription factor (DBP) interactswith all three classes of chromatin remodeling enzymes which has only been known for histones. Interactions include the chemicalmodification enzyme acetyltransferase (HAT)(Suzuki et al, 2000), the ATP-independent nucleosome remodeling enzyme histonechaperone (HC)(Suzuki et al, 2003), and the ATP-dependent nucleosome remodeling enzyme (ATPase)(Kadam et al, 2000).95

Suzuki et al: Regulation of the Sp/KLF-family of transcription factorsthrough cooperative interaction with chromatinremodelingfactors.The zinc finger transcription factors are the mostwidely evolved family of transcription factors ineukaryotes. Given that this biological diversification wascoupled with the evolution of nuclear structure ineukaryotes, it is conceivable that regulation of chromatinis a necessary process to further allow for efficient use andaccess of factors to the tightly packaged DNA geneticinformation. Important mechanisms of transcriptionalregulation in the context of chromatin have been shown asdiscussed in this review. The mechanism that the DBDmediates important regulation of the DNA-bindingtranscription factors through interaction and modificationwith chromatin factors can certainly be generalized toDNA-binding transcription factors other than thedescribed zinc finger factors. Selectivity may be foundbetween interaction of subsets for chromatin factors andDBD motifs. Furthermore, although only three types ofchromatin factors were described including modificationenzymes (e.g. acetyltransferase), ATP-independent (e.g.histone chaperones) and ATP-dependent (Swi/snf) factors,other chromatin factors are likely also to participate inregulatory interactions. Understanding the hierarchy andnetwork of regulation among DNA-binding transcriptionfactors and chromatin factors will likely play an importantrole in understanding the complexity of eukaryotictranscriptional regulation. As the Sp/KLF factors are a keyfamily important in mammalian biological processesranging from development, differentiation, to oncogenicprocesses, further studies aimed at understanding thetemporospatial regulation of chromatin centered onSp/KLF factors will surely advance our understanding ofeukaryotic transcriptional mechanisms of chromatinactivation in a biological context. Future gene therapyapproaches could use strategies of expressing suchactivator, modifier or factor genes individually or incomplexed form to facilitate regulation of therapeuticallyimportant genes at the physiologically relevant chromatinDNA level.AcknowledgementsThis study was supported by grants from the NewEnergy and Industrial Technology DevelopmentOrganization, Ministry of Health, Labour and Welfare,Ministry of Education, Culture, Sports, Science andTechnology, Japan Science and Technology Corporation,Sankyo Life Science Foundation, Takeda MedicalResearch Foundation, and the Applied EnzymeAssociation.ReferencesArmstrong SA, Barry DA, Leggett RW and Mueller CR (1997)Casein kinase II-mediated phosphorylation of the C terminusof Sp1 decreases its DNA binding activity. J Biol Chem 272,13489-3495.Bieker, JJ (2001) Krüppel-like factors: three fingers in manypies. J Biol Chem 276, 34355-34358.Black AR, Jensen D, Lin SY and Azizkhan JC (1999)Growth/cell cycle regulation of Sp1 phosphorylation. J BiolChem 274, 1207-1215.Black, AR, Black, JD and Azizkhan-Clifford, J (2001) Sp1 andkrüppel-like factor family of transcription factors in cellgrowth regulation and cancer. J Cell Physiol 188, 143-160.Braun H, Koop R, Ertmer A, Nacht S and Suske G (2001)Transcription factor Sp3 is regulated by acetylation. NucleicAcids Res 29, 4994-5000.Brownell JE, Allis CD (1996) Special HATs for specialoccasions: linking histone acetylation to chromatin assemblyand gene activation. Curr Opin Genet Dev 6, 176-184.Bouwman P and Philipsen S (2002) Regulation of the activity ofSp1-related transcription factors. Mol Cell Endocrinol 195,27-38.Chen X and Bieker JJ (2001) Unanticipated repression functionlinked to erythroid Krüppel-like factor. Mol Cell Biol 21,3118-3125.Cheung WL, Briggs SD and Allis CD (2000) Acetylation andchromosomal functions. Curr Opin Cell Biol 12, 326-333.Dang DT, Pevsner J and Yang VW (2000) The biology of themammalian Krüppel-like family of transcription factors. IntJ Biochem Cell Biol 32, 1103-1121.Diakun GP, Fairall L and Klug A (1986) EXAFS study of thezinc-binding sites in the protein transcription factor IIIA.Nature 324, 698-699.Doetzlhofer A, Rotheneder H, Lagger G, Koranda M, Kurtev V,Brosch G, Wintersberger E and Seiser C (1999) Histonedeacetylase 1 can repress transcription by binding to Sp1.Mol Cell Biol 19, 5504-5511.Dynan, WS and Tjian R (1983) The promoter-specifictranscription factor Sp1 binds to upstream sequences in theSV40 early promoter. Cell 35, 79-87.Freiman, RN and Tjian R (2003) Regulating the regulators:lysine modifications make their mark. Cell 112, 11-17.Gillemans N, Tewari R, Lindeboom F, Rottier R, de Wit T,Wijgerde M, Grosveld F and Philipsen S (1998) AlteredDNA-binding specificity mutants of EKLF and Sp1 showthat EKLF is an activator of the beta-globin locus controlregion in vivo. Genes Dev 12, 2863-2873.Horikoshi M, Hai T, Lin YS, Green MR and Roeder RG (1988a)Transcription factor ATF interacts with the TATA factor tofacilitate establishment of a preinitiation complex. Cell 54,1033-1042.Horikoshi M, Carey MF, Kakidani H and Roeder RG (1988b)Mechanism of action of a yeast activator: direct effect ofGAL4 derivatives on mammalian TFIID-promoterinteractions. Cell 54, 665-669.Jackson SP, Tjian R (1988) O-glycosylation of eukaryotictranscription factors: implications for mechanisms oftranscriptional regulation. Cell 55, 125-33.Kaczynski J, Cook T and Urrutia R (2003) Sp1- and Krüppel-liketranscription factors. Genome Biol 4, 206.Kadam S, McAlpine GS, Phelan ML, Kingston RE, Jones KAand Emerson BM (2000) Functional selectivity ofrecombinant mammalian SWI/SNF subunits. Genes Dev 14,2441-2451.Klevit RE (1991) Recognition of DNA by Cys2, His2 zincfingers. Science 253, 1367.Koritschoner NP, Bocco JL, Panzetta-Dutari GM, Dumur CI,Flury A and Patrito LC (1997) A novel human zinc fingerprotein that interacts with the core promoter element of aTATA box-less gene. J Biol Chem 272, 9573-9580.96

Gene Therapy and Molecular Biology Vol 7, page 97Lemon B and Tjian R (2000) Orchestrated response: a symphonyof transcription factors for gene control. Genes Dev 14,2551-2569.Mackay JP and Crossley M (1998) Zinc fingers are stickingtogether. Trends Biochem Sci 23, 1-4.Miller IJ and Bieker JJ (1993) A novel, erythroid cell-specificmurine transcription factor that binds to the CACCC elementand is related to the Krüppel family of nuclear proteins. MolCell Biol 13, 2776-2786.Narla G, Heath KE, Reeves HL, Li D, Giono LE, KimmelmanAC, Glucksman MJ, Narla J, Eng FJ, Chan AM, Ferrari AC,Martignetti JA and Friedman S (2001) KLF6, a candidatetumor suppressor gene mutated in prostate cancer. Science294, 2563-2566.Nakatani Y (2001) Histone acetylases--versatile players. GenesCells 6, 79-86.Ogryzko VV, Schiltz RL, Russanova V, Howard BH andNakatani Y (1996) The transcriptional coactivators p300 andCBP are histone acetyltransferases. Cell 87, 953-959.Pal S, Claffey KP, Cohen HT and Mukhopadhyay D (1998)Activation of Sp1-mediated vascular permeabilityfactor/vascular endothelial growth factor transcriptionrequires specific interaction with protein kinase C zeta. JBiol Chem 273, 26277-26280.Philipsen S and Suske G (1999) A tale of three fingers: thefamily of mammalian Sp/XKLF transcription factors.Nucleic Acids Res 27, 2991-3000.Preiss A, Rosenberg UB, Kienlin A, Seifert E and Jackle H(1985) Molecular genetics of Krüppel, a gene required forsegmentation of the Drosophila embryo. Nature 313, 27-32Ptashne M and Gann AA (1990) Activators and targets. Nature346, 329-331.Roeder RG (1996) The role of general initiation factors intranscription by RNA polymerase II. Trends Biochem Sci 9,327-335.Ruppert JM, Kinzler KW, Wong AJ, Bigner SH, Kao FT, LawML, Seuanez HN, O'Brien SJ and Vogelstein B (1998) TheGLI-Krüppel family of human genes. Mol Cell Biol 8, 3104-3113.Sapetschnig A, Rischitor G, Braun H, Doll A, Schergaut M,Melchior F and Suske G. (2002) Transcription factor Sp3 issilenced through SUMO modification by PIAS1. EMBO J21, 5206-15.Shi H, Zhang Z, Wang X, Liu S, and Teng CT (1999) Isolationand characterization of a gene encoding human Krüppel-likefactor 5 (IKLF): binding to the CAAT/GT box of the mouselactoferrin gene promoter. Nucleic Acids Res 27, 4807-4815.Shindo T, Manabe I, Fukushima Y, Tobe K, Aizawa K,Miyamoto S, Kawai-Kowase K, Moriyama N, Imai Y,Kawakami H, Nishimatsu H, Ishikawa T, Suzuki T, MoritaH, Maemura K, Sata M, Hirata Y, Komukai M, KagechikaH, Kadowaki T, Kurabayashi M, and Nagai R (2002)Krüppel-like zinc-finger transcription factor KLF5/BTEB2 isa target for angiotensin II signaling and an essential regulatorof cardiovascular remodeling. Nat Med 8, 856-863.Shields JM and Yang VW (1998) Identification of the DNAsequence that interacts with the gut-enriched Krüppel-likefactor. Nucleic Acids Res 26, 796-802.Sogawa K, Kikuchi Y, Imataka H and Fujii-Kuriyama Y (1993)Comparison of DNA-binding properties between BTEB andSp1. J Biochem 114, 605-609.Song CZ, Keller K, Murata K, Asano H and StamatoyannopoulosG (2002) Functional interaction between coactivatorsCBP/p300, PCAF, and transcription factor FKLF2. J BiolChem 277, 7029-7036.Song CZ, Keller K, Chen Y and Stamatoyannopoulos G (2003)Functional Interplay between CBP and PCAF in Acetylationand Regulation of Transcription Factor KLF13 Activity. JMol Biol 329, 207-215.Suzuki M, Gerstein M and Yagi N (1994) Steriochemical basisof DNA recognition by Zn fingers. Nucleic Acids Res 22,3397-3405Suzuki T, Yamamoto T, Kurabayashi M, Nagai R, Yazaki Y andHorikoshi M (1998) Isolation and initial characterization ofGBF, a novel DNA-binding zinc finger protein that binds tothe GC-rich binding sites of the HIV-1 promoter. J Biochem124, 389-395.Suzuki T, Kimura A, Nagai R and Horikoshi M (2000)Regulation of interaction between the acetyltransferaseregion of p300 and the DNA-binding domain of Sp1 on andthrough DNA binding. Genes Cells 5, 29-41.Suzuki T, Muto S, Miyamoto S, Aizawa K, Horikoshi M andNagai R (2003) Functional interaction of the DNA-bindingtranscription factor Sp1 through its DNA-binding domainwith the histone chaperone TAF-I. J Biol Chem 278, 28758-28764Thiesen HJ and Bach C (1990) Target Detection Assay (TDA): aversatile procedure to determine DNA binding sites asdemonstrated on SP1 protein. Nucleic Acids Res 18, 3203-3209.Tupler R, Perini G and Green MR (2001) Expressing the humangenome. Nature 409, 832-833.Turner J and Crossley M (1999) Mammalian Krüppel-liketranscription factors: more than just a pretty finger. TrendsBiochem Sci 24, 236-40.Wagner S and Green MR (1994) DNA-binding domains: targetsfor viral and cellular regulators. Curr Opin Cell Biol 6, 410-414.Zawel L and Reinberg D (1995) Common themes in function ofeukaryotic transcription complexes. Annu Rev Biochem 64,533-561.Zhang W and Bieker JJ (1998) Acetylation and modulation oferythroid Krüppel-like factor (EKLF) activity by interactionwith histone acetyltransferases. Proc Natl Acad Sci USA 95,9855-9860.Zhang W, Kadam S, Emerson BM and Bieker JJ (2001) Sitespecificacetylation by p300 or CREB binding proteinregulates erythroid Krüppel-like factor transcriptionalactivity via its interaction with the SWI-SNF complex. MolCell Biol 21, 2413-2422Dr. Toru Suzuki97

Suzuki et al: Regulation of the Sp/KLF-family of transcription factors98

Gene Therapy and Molecular Biology Vol 7, page 99Gene Ther Mol Biol Vol 7, 99-102, 2003Detection of MET oncogene amplification inhepatocellular carcinomas by comparative genomichybridization on microarraysResearch ArticleW.L. Robert Li 1 , Nagy A. Habib¨* , Steen L. Jensen¨* , Paul Bao 2 , Diping Che 3 ,Uwe R. Müller 2 Vysis Inc., Downers Grove, Illinois, USA, ¨Liver Surgery Section, Imperial College School of Medicine, HammersmithHospital Campus, London, UK.1 Pharmacia Corporation, 700 Chesterfield Parkway North, Chesterfield, MO 63198, 2 Corning Incorporated, SP-FR-01,Corning, NY 14831, 3 Illumina, Inc., 9390 Towne Center Drive, Suite 200, San Diego, CA 92121, USA__________________________________________________________________________________*Correspondence: Nagy A. Habib, ChM FRCS, Head of Liver Surgery Section, Imperial College London, Faculty of Medicine,Hammersmith Hospital Campus, Du Cane Road, London W12 ONN, UK; tel: +44-20-8383-8574, fax: +44-20-8383-3212, e-mail:nagy.habib@imperial.ac.ukKey words: MET oncogene, amplification, hepatocellular carcinoma, microarrays, comparative genomic hybridizationAbbreviations: HCC, hepatocellular carcinoma; CLM, colorectal liver metastases; FISH, fluorescent in situ hybridization; P1, phage P1;PAC, P1-derived artificial chromosome; BAC, bacterial artificial chromosome; CCD, charge coupled device.Received: 26 June 2003; Accepted: 10 July 2003; electronically published: July 2003SummaryThe oncogene MET localized on human chromosome 7q21-31 encodes a transmembrane protein with tyrosinekinase activity and is believed to be implicated in progression of colorectal cancer. The aims of the study were todetermine whether overexpression and amplification of the MET oncogene confers a selective growth advantage tohepatocellular carcinomas. Comparative genomic hybridization on microarrays was used in the analysis of DNAfrom 32 liver tumors (6 hepatocellular carcinoma; 16 colorectal liver metastases; 3 cholangiocarcinomas; 2adenomas; 2 fibrolamellar; 3 unclassified) to screen for sequence copy number changes. The results revealed aMET gene amplification in hepatocellular carcinoma, cholangiocarcinoma, and colorectal liver metastases tumors.Moreover, one of the patients with hepatocellular carcinoma showed MET amplifications in both tumor and nontumorsamples, with the tumor having approximately 12.8 copies of the MET target locus per cell. These findingssuggest that amplifications in the MET gene may play an important role in hepatocarcinogenesis.I. IntroductionThe oncogene MET, localized on humanchromosome 7q21-31 by in situ hybridization (Dean et al,1985), encodes a transmembrane protein with tyrosinekinase activity (Dean et al, 1985; Park et al 1996). It wasshown that this protein is the receptor of hepatocytegrowth factor (HGF)/ Scatter factor (Giordano et al, 1989;Bottaro et al, 1991), and the signals of HGF are transducedthrough the receptor tyrosine kinase encoded by the METproto-oncogene. The MET gene can be activated by theformation of a chimeric gene through fusing thetranslocated promoter region (TPR) on chromosome 1 tothe N-terminally truncated MET kinase domain (Park et al,1996). Gene amplification and mutation may be anotherpath to MET proto-oncogene activation, since MET geneamplifications have been reported in human gastriccarcinomas (Soman et al, 1990; Ponzetto et al, 1991) andgliomas (Fischer et al 1995). Furthermore, MET geneamplification and the resulting over-expression arebelieved to be involved in progression of colorectal cancer(Di Renzo et al, 1995).Human hepatocellular carcinoma (HCC) is one of themost common and devastating cancers with a poorprognosis. It has been widely considered that hepatitis Bvirus (HBV) and environmental agents such as aflatoxinB1 are major risk factors. However, the molecularmechanism of hepatocarcinogenesis is poorly understood.Loss of heterozygosity (LOH) has been reported forseveral genomic loci, such as the region surrounding RB1on 13q (Nishida et al, 1992; Zhang et al, 1994), orsequences on 11p (Rogler et al, 1985), and 6q (De Souza99

Li et al: MET amplification in liver tumorset al, 1995). Mutation of the p53 gene was detected inapproximately 36% of advanced HCC (Murakami et al,1991) and was also implicated in tumor progression(Teramoto et al, 1994). Overexpression was reported forseveral oncogenes such as N-ras, c-myc and fos(Arbuthnot et al, 1991). However, oncogene amplificationappears to be rarely the underlying mechanism of cancerdevelopment in these cases. Amplifications associatedwith HCC have been found on 11q13 (Nishida et al,1994), involving both INT2 and cyclin D1. Nishida andcolleagues (1994) showed that the cyclin D1 gene wasamplified 3 to 16 fold in about 11% of HCC samplesanalyzed, with a concomitant 6 to 10 fold overexpression.Based on this finding they suggested that amplificationand overexpression of the cyclin D1 gene might beresponsible for rapid growth of a subset of HCC.The rapid emergence of microarray technology hasallowed new approaches to tumor analysis. The mostcommon application of this technology has focused on theuse of cDNA arrays for the large-scale analysis of geneexpression to monitor tumor progression (Sgroi et al,1999) or for cancer typing (Anbazhagan et al, 1999).Oligonucleotide arrays have enabled rapid re-sequencingfor genotyping or point mutation analysis, such as p53mutation detection (Hacia, 1999). Applying ComparativeGenomic Hybridization (CGH) to microarrays of largegenomic clones has also been successful, allowing thedetection of gross chromosomal abnormalities that resultin copy number changes for a given sequence, such asgene amplifications or deletions (e.g. LOH), (Solinas-Toldo et al, 1997; Pinkel et al, 1998; Muller, 2001). Suchsub-chromosomal aneuploidies are known to befundamental causes of cancer and many other humandiseases, often leading to the over- or under-expression ofgenes.II. Materials and methodsWe have developed a CGH-based microarray system(Genosensor System) and a microarray to specifically detectabnormalities of 52 genomic loci that have been associated withformation of various human solid tumors (Müller et al, 2002).The arrays consist of 3 repeats each of 52 P1, PAC or BAC cloneDNAs that are arrayed on a chromium-coated glass surface. Forhybridization to this array, genomic DNA samples were extractedfrom human liver tumors or from histopathologically non-tumorliver sections from the same patient. After purification (GentraKits, Gentra Systems, Inc., Minneapolis, MN), the genomic DNAsamples were then labeled by nick translation (NicktranslationKit, Vysis, Inc., Downers Grove, IL) in the presence ofSpectrum-Green dUTP (green fluorophore). Genomic DNA froma normal human male donor was chemically labeled with a redfluorophore (Vysis, Inc., Downers Grove, IL), and served as areference. The test probe (green) and reference probe (red) werethen mixed with unlabeled human cot-1 DNA and co-hybridizedto the microarrays. After removal of un-hybridized probes, thearray was imaged by a multi-color CCD based image analysissystem, and fluorescence intensities were determined for eachtarget spot. Under the assumption that the hybridization kineticsfor a given sequence are equal for the test and reference DNA,the signal intensity is proportional to the copy number of thatsequence in the hybridization mixture. The test/referenceintensity ratio for each target genomic locus (average of 3 spots)was normalized by dividing with the average ratio of all"normal" targets, resulting in an estimate for the copy numberchange of that specific sequence compared to the rest of thegenome.III. ResultsAs shown in Figure 1, the DNA extracted from thetumor tissue of HCC patient #21 was found to have anaverage normalized ratio of 4.2 ± 1 by Genosensoranalysis for the MET target locus (average of 5experiments), and 6.4 ± 0.8 by Southern analysis (3experiments; see below). Since the reference sample usedhere was from a normal human male and has 2 copies ofthe MET sequence, this ratio suggests that there are onaverage between 8.4 to 12.8 copies per cell (4.2 or 6.4 x 2)of the MET gene in the HCC tumor sample. Thisamplification is considered a significant finding, as it isthe first time to be reported in this type of cancer.Since microarray or Southern analysis yields anestimate for the copy number of a sequence averaged overall cells from which the DNA was extracted, the METamplification level was confirmed by fluorescent in situhybridization (FISH). The tumor tissue from the sameHCC patient was formalin-fixed, paraffin-embedded, andsectioned. FISH was performed with SpectrumGreenlabeled DNA from a BAC clone containing the MET gene.SpectrumOrange labeled CEP 7 DNA (containingchromosome 7-specific centromere DNA sequences;Vysis) was co-hybridized as a control. The signal for both,the MET gene and chromosome 7 were counted under afluorescent microscope after counterstaining with DAPI.As expected, the majority (60%) of the cells contained 2copies of chromosome 7 per nucleus, while approximately40 % of cells have an average of 25 copies of MET(Figure 2). Since the remaining 60% of cells have only 2copies of the MET gene, the DNA extracted from thistumor section should have 11 copies of the MET gene,which is in good agreement with the microarray andSouthern data.For further confirmation and comparisons additionalSouthern blot analyses were carried out with EcoR1-digested DNA from 32 tumor samples including 6 HCC,16 colorectal liver metastasis (CLM), 3cholangiocarcinomas, 2 adenomas, 2 fibrolamellar (HCCvariant), and 3 unclassified liver tumors. Normal humangenomic DNA (control) and DNA from the non-tumorliver tissue of HCC patient #21 were also included in theSouthern blot analyses. A 360bp DNA fragment (1) wasamplified by polymerase chain reaction (PCR) in thepresence of the following pair of primers, primer H1: 5'-TCTTGATTACCTGCATTTGC-3' and primer H2: 5'-TGGGGCAAGAAGGCCTCTCT-3' from a BAC clonecontaining the entire MET gene. The 360bp MET probewas labeled by PCR in the presence of 32 P-labelled dCTPand hybridized to the Southern blot. A probe generatedfrom a genomic clone on 11q13 was re-hybridized to thesame Southern blot for normalization, after the MET probewas stripped from the blot.100

Gene Therapy and Molecular Biology Vol 7, page 101Figure 1: Genosensor and Southern analysis of HCC samples. Genomic DNA (8 µg for DNA from tumor tissue and 8 µg for DNA fromnormal tissue) was digested with Eco R1, run on agarose gels and blotted. Southern hybridization was performed with a P32 labeled 360bp MET probe as described in the text. A composite image (red, green and blue) of a Gneosensor oncogene array after hybridizationwith a mixture of sample 21T DNA (green) and normal refernce DNA (red) is shown after counterstaining with DAPI.The level of MET gene amplification was determinedusing a PhosphoImager (Molecular Dynamics). Some ofthe results are shown in Figure 1. Among the 6 HCCsamples analysed, 2 MET gene amplifications wereobserved (6.4 and 2.5 fold after normalization). MET geneamplifications were also observed in thecholangiocarcinoma and CLM samples. Two of the threecholangiocarcinoma patients had MET gene amplificationsin their tumour specimens at a level of 6.5 fold and 1.6fold, respectively. Of the 16 patients with CLM, three hadMET gene amplifications of 2.3, 2.1 and 1.8 fold,respectively. Of specific interest is the finding that both,the tumor as well as non-tumor tissues from the sameHCC patient (No. 21) showed a similar level of METamplification (6.4 fold and 6.1 fold, respectively),suggesting that MET amplification may precede malignanthistopathological changes. This patient developed HCC inthe background of a cirrhotic liver complicating hepatitisC infection. Liver cirrhosis provides a pre-malignant fieldchange for HCC development.of the MET oncogene in hepatocellular carcinomasstrongly suggests a role here as well.This finding in combination with multiple otherreports of cancer associated gene amplificationsunderscores the need for a rapid, quantitative detectionmethod for such genetic changes. The microarray basedmethod described here is consistent (within a factor of 2)with other established methods (FISH, Southern blotting),and therefore suitable for the screening of geneamplifications. Since this method is non-radioactive,simpler, faster, and more economical than either FISH orSouthern, especially when the mutated genetic locus is notknown, it lends itself to applications in clinicaldiagnostics.IV. DiscussionHepatocyte growth factor (HGF) plays an importantrole in the growth, progression and angiogenesis ofvarious tumors and is known to specifically promotehepatocyte proliferation and liver regeneration. Inaddition, it may also be involved in tumor invasion andprogression (Tamatani et al, 1999). Overexpression andamplification of the HGF receptor (MET gene) have beenimplicated in progression of colorectal cancer (Di Renzoet al, 1995), by a mechanism where the elevated level ofthe MET gene product confers a selective growthadvantage to tumor cells (Di Renzo et al, 1991). In thecontext of this information, our finding of amplificationsFigure 2: FISH on interphase nuclei of patient #21. FISH wasperformed on formalin-fixed, de-parafinized tumor tissuesections. A BAC clone containing the MET gene was labeledwith SpectrumGreen by nick translation and used as a probe. ASpectrumOrange labeled chromosome 7-specific centromereprobe (CEP7; Vysis Inc.) was co-hybridized as reference.101

Li et al: MET amplification in liver tumorsAcknowledgmentsWe thank Ragai Mitry, Teresa Ruffalo and AnnaLublinsky for their excellent technical support. We wouldalso like to thank The Pedersen Family CharitableFoundation for their financial support with this research.ReferencesAnbazhagan R., Tihan T, Bornman DM, Johnston JC, Saltz JH,Weigering A, Piantadosi S, Gabrielson E. (1999)Classification of small cell lung cancer and pulmonarycarcinoid by gene expression profiles. Cancer Res 59, 5119-5122.Arbuthnot P, Kew M, Fitschen W. (1991) c-fos and c-myconcoprotein expression in human hepatocellular carcinoma.Anticancer Res 11, 921-924.Bottaro DP, Rubin JS, Faletto DL, Chan AM-L, Kmiecik TE,Vande Woude GF, Arronson SA. (1991) Identification of thehepatocyte growth factor receptor as the met proto-oncogeneproduct. Science 251, 802-804.De Souza AT, Hankins GR, Washington MK, Fine RL, OrtonTC, Jirtle RL. (1995) Frequent loss of heterozygosity on 6qat the mannose 6-phosphate/insulin-like growth factor IIreceptor locus in human hepatocellular tumors. Oncogene10, 1725-1729.Dean M, Park M, Le Beau MM, Robins TS, Diaz MO, RowleyJD, Blair DG, Vande Woude GF. (1985) The human metoncogene is related to the tyrosine kinase oncogenes. Nature318, 385-388.Di Renzo MF, Narsimhan RP, Olivero M, Bretti S, Giordano S,Medico E, Gaglia P, Zara P, Comoglio PM. (1991)Expression of the met/HGF receptor in normal and neoplastichuman tissues. Oncogene 6, 1997-2003.Di Renzo MF, Olivero M, Giacomini A, Porte H, Chastre E,Mirossay L, Nordlinger B, Bretti S, Bottardi S, Giodarno S.(1995) Overexpression and amplification of the met/HGFreceptor gene during the progression of colorectal cancer.Clin Cancer Res 1, 147-154.Fischer U, Muller H-W, Sattler H-P, Feiden K, Zang KD, MeeseE. (1995) Amplification of the met gene in glioma. GenesChromosom Cancer 12, 63-65.Giordano S, Ponzetto C, Di Renzo MF, Cooper CS, ComoglioPM. (1989) Tyrosine kinase receptor indistinguishable fromthe c-met protein. Nature 339, 155-156.Hacia JG. (1999) Resequencing and mutational analysis usingoligonucleotide microarrays. Nat Genet 21 (1 suppl), 42-47.Müller U, Bao P, Che D, et al. (2002) DNA-Chips in the analysisof genetic aberrations. Springer Laboratory Manual FISHTechnology ed. Rautenstrasse BW, Liehr T, Springer Verlag,Heidelberg, Germany.Murakami Y, Hayashi K, Hirohashi S, et al. (1991) Aberrationsof the tumor suppressor p53 and retinoblastoma genes inhuman hepatocellular carcinoma. Cancer Res 52, 5520-5525.Nishida N, Fukuda Y, Kokuryu H, Sadamoto T, Isowa G, HondaK, Yamaoka Y, Ikenaga M, Imura H, Ishizaki K. (1992)Accumulation of allelic loss of arms of chromosomes 13q,16q and 17p in the advanced stages of human hepatocellularcarcinoma. Int J Cancer 51, 862-868.Nishida N, Fukuda Y, Komeda T, Kita R, Sando T, Furukawa M,Amenomori M, Shibagaki I, Nakao K, Ikenaga M. (1994)Amplification and overexpression of the cyclin D1 gene inagrressive human hepatocellular carcinoma. Cancer Res 54,3107-3110.Park M, Dean M, Cooper CS, Schmidt M, O’Brien SJ, Blair DG,,Vande Woude GF. (1986) Mechanism of met oncogeneactivation. Cell 45, 895-904.Pinkel D, Segraves R, Sudar D, Clark S, Poole I, Kowbel D,Collins C, Kuo W-L, Chen C, Zhai Y, Dairkee S, Ljung B-M, Gray JW, Albertson DG. (1998) High resolution analysisof DNA copy number variation using comparative genomichybridization to microarrays. Nat Genet 20, 207-211Ponzetto C, Giordano S, Peverali F, Della Valle G, Abate ML,Vaula G, Comoglio PM. (1991) c-met is amplified but notmutated in a cell line with an activated met tyrosine kinase.Oncogene 6, 553-559.Rogler CE, Sherman M, Su CY, Shafritz DA, Summers J, ShowsTB, Henderson A, Kew M. (1985) Deletion in chromosome11p associated with a hepatitis B integration site inhepatocellular carcinoma. Science 230, 319-322.Sgroi D, Teng S, Robinson G, LeVanglie R, Hudson JR, Jr,Elkahloun AG. (1999) In vivo gene expression profileanalysis of human breast cancer progression. Cancer Res 59,5656-5661.Solinas-Toldo S, Lampel S, Stilgenbauer S, Nickolenko J,Benner A, Dohner H, Crmer T, Lichter P. (1997) Matrixbasedcomparative genomic hybridization: biochips to screenfor genomic imbalances. Genes Chromosom Cancer 20,399-407.Soman NR, Wogan GN, Rhim JS. (1990) TPR-MET oncogenicrearrangement: detection by polymerase chain reactionamplification of the transcript and expression in humantumor cell lines. Proc Natl Acad Sci USA 87, 739-742.Tamatani T., Hattori K., Iyer A., Tamatani K, Oyasu R. (1999)Hepatocyte growth factor is an invasion/migration factor ofrat urothelial carcinoma cells in vitro. Carcinogenesis 20,957-962.Teramoto T, Satonaka K, Kitazawa S, Fujimori T, Hayashi K,Maeda S. (1994) p53 gene abnormalities are closely relatedto hepatovirus infections and occur at a late stage ofhepatocarcinogenesis. Cancer Res 54, 231-235.Zhang X, Xu H-J, Murakami Y, Sachse R, Yashima K,Hirohasha S, Hu S-X, Benedict WF, Sekiya T. (1994)Deletions of chromosome 13q, mutations in Retinoblastoma1, and retinoblastoma protein state in human hepatocellularcarcinoma. Cancer Res 54, 4177-4182.102

Gene Therapy and Molecular Biology Vol 7, page 103Gene Ther Mol Biol Vol 7, 103-111, 2003HMG-CoA-reductase inhibition-dependent and -independent effects of statins on leukocyte adhesionResearch ArticleTriantafyllos Chavakis 1,2* , Thomas Schmidt-Wöll 2 , Peter. P. Nawroth 1 , Klaus T.Preissner 2 , Sandip M. Kanse 21 Department of Internal Medicine I, University Heidelberg and 2 Institute for Biochemistry, Justus-Liebig-Universität,Giessen, Germany__________________________________________________________________________________*Correspondence: Dr. T. Chavakis, Department of Internal Medicine I, University Heidelberg, Bergheimer Strasse 58, D-69115Heidelberg, Germany; tel.: ++49 6221 56 4776; fax: ++49 6221 56 ; email: triantafyllos.chavakis@med.uni-heidelberg.deKey words: leukocyte, adhesion, β2-integrins, urokinase-receptor, statins, lovastatin, HMG-CoA reductaseAbbreviations: BSA, bovine serum albumin, FBG, fibrinogen, HMG-CoA, 3-hydroxy-3-methylglutaryl coenzyme-A, ICAM-1,intercellular cell adhesion molecule-1, PBS, phosphate buffered saline, uPA, urokinase-type plasminogen activator , uPAR, urokinasetypeplasminogen activator receptor, VN, vitronectinReceived: 1 July 2003; Accepted: 10 July 2003; electronically published: July 2003SummaryStatins are inhibitors of 3-hydroxy-3-methylglutaryl coenzyme-A (HMG-CoA) reductase, a key enzyme forcholesterol biosynthesis and isoprenoid intermediates. Increasing evidence suggests that statins might affectinflammatory processes including leukocyte recruitment, yet, the underlying mechanisms are not defined. In thisstudy two different pathways for inhibition of leukocyte adhesion by statins are described. (i) Coincubation withlovastatin inhibited adhesion of LFA-1 (CD11a/CD18, αLβ2)-transfected K562 cells to ICAM-1 and of p150.95(CD11c/CD18, αXβ2)-transfected K562 cells to both ICAM-1 and fibrinogen (FBG), whereas adhesion of Mac-1(CD11b/CD18, αMβ2)-transfected K562 cells was not affected. Moreover, only LFA-1-mediated adhesion to ICAM-1 but not Mac-1-mediated adhesion to FBG or urokinase-receptor (uPAR)-me di ate d adhesion to vitronectin (VN)of myelo-monocytic U937 cells was blocked by coincubation with lovastatin. The antiadhesive effect of lovastatinwas independent of HMG-CoA-reductase inhibition, as it was not reversible in the presence of mevalonate, farnesylpyrophosphateor geranyl-pyrophosphate. In purified systems, lovastatin only blocked the ICAM-1/LFA-1interaction but not the ICAM-1/Mac-1, FBG/Mac-1 or the VN/uPAR interactions. (ii) In contrast, preincubation ofU937 cells for up to 18 h with lovastatin completely abrogated LFA-1-, Mac-1- and uPAR-dependent cell adhesionto the respective ligands. This anti-adhesive function of lovastatin was dependent on HMG-CoA reductaseinhibition, since mevalonate or the isoprenoid intermediates restored adhesion, while no downregulation of integrinoruPAR-expression was observed. Thus, two distinct pathways, involving a direct interaction with LFA-1 andp150.95 and an indirect inhibition of cell adhesion through disruption of cholesterol and/or isoprenoid metabolitebiosynthesis are induced by statins. These functions can explain at least in part the inhibition of leukocyte adhesionand the associated antiinflammatory role of statinsI. IntroductionWhen leukocytes emigrate from the blood-streaminto sites of inflammation or injury, they undergo acomplex sequence of adhesion and locomotion stepsrequiring the expression and upregulation of variousadhesion receptors on the surface of leukocytes andvascular cells. During their transmigration phaseleukocytes adhere to provisional matrix substrates such asfibrinogen (FBG), fibronectin or vitronectin (VN) at sitesof increased vascular permeability or damage. Theprominent adhesion receptors on leukocytes are integrins,such as VLA-4 (α4β1), that can bind to fibronectin,whereas adhesion to FBG is mediated by the β2 integrinsMac-1 (CD11b/CD18, αMβ2, CR3) and p150.95(CD11c/CD18). Mac-1 together with LFA-1(CD11a/CD18, αLβ2) also provide firm adhesion to andtransmigration through the endothelium by recognition oftheir counter-receptor ICAM-1 on endothelial cells;evidence exists that p150.95 binds ICAM-1 as well(Springer, 1994; Carlos and Harlan, 1994; Stewart et al,1995; Blackford et al, 1996; Gahmberg, 1997). Thefunctional properties of integrins in general can bemodulated by lateral (cis) interaction with integrin103

Chavakis et al: Leukocyte adhesion and statinsassociated protein (CD47), members of the tetraspaninfamily, syndecans, caveolin-1 or urokinase typeplasminogen activator receptor (uPAR) (CD87), leading tothe formation of transient multireceptor complexes thatfacilitate the dynamic recruitment of signaling moleculesto sites of cellular contacts or focal adhesions (Ossowskiand Aguirre-Ghiso, 2000; Preissner et al, 2000). Besidesits ability to regulate integrin-dependent adhesionphenomena, uPAR can also directly mediate leukocyteadhesion to matrix-associated VN (Wei et al, 1994; Sitrinet al, 1996; May et al, 1998).Recently, attention has been drawn to the role ofmicrodomain structures of the plasma membrane, denotedlipid rafts, in cell adhesion. Lipid rafts are enriched inglycosphingolipids, cholesterol, transmembrane proteinsand signaling molecules. GPI-anchored proteins maybecome sequestered into the microdomains as well, whichhave a lower fluidity than the surrounding membraneallowing the formation of multireceptor adhesioncomplexes. On epithelial cells, caveolin is a unique raftcomponent, that has the intrinsic propensity to oligomerizeand, thereby, contribute to formation of membraneinvaginations termed caveolae (Horejsi et al, 1999;Kurzchalia and Parton, 1999; Smart et al, 1999; Simonsand Toomre, 2000). Although leukocytes lack caveolinexpression, they still contain lipid rafts that may facilitatethe formation of adhesion complexes. The possibility thatlipid rafts might regulate leukocyte adhesion bymodulating integrin avidity has already been suggested(Krauss and Altevogt, 1999).Statins inhibit the key enzyme of cholesterolbiosynthesis 3-hydroxy-3-methylglutaryl coenzyme-Areductase (HMG-CoA reductase). In addition to loweringplasma cholesterol, increasing evidence suggests thatstatins play a pleiotropic role in the vascular system byeffects on nitric oxide synthesis, smooth muscle cellproliferation, fibrinolysis or the immune system (Soma etal, 1993; Aikawa et al, 1998; Essig et al, 1998; Guisarro etal, 1998; Laufs and Liao, 1998; Laufs et al, 1998, 1999;Kwak et al, 2000; Diomede et al, 2001; Kwak and Mach,2001). In particular, statins could inhibit leukocyterecruitment by regulating the expression of monocytechemoattractant protein-1 (Romano et al, 2000) and ofadhesion receptors (Weber et al, 1997; Ganne et al, 2000;Yoschida et al, 2001) or they might modulate integrinaffinity by preventing geranyl-geranylation o f RhoA (Liuet al, 1999). Cholesterol depletion by statins might alsodisrupt lipid rafts and, thereby, affect cell adhesion (Krausand Altevogt, 1999; Simons and Toomre, 2000). Finally, arecent report suggested that different statins selectivelybind to LFA-1, thereby blocking LFA-1 mediatedleukocyte adhesion (Kallen et al, 1999; Weitz-Schmidt etal, 2001).These observations prompted us to investigate inmore detail the role of lovastatin in β2-integrin- anduPAR-mediated leukocyte interactions. Two distinctmechanisms, a HMG-CoA reductase-dependent and an–independent, for inhibition of leukocyte adhesion aredescribed, which further help to understand theantiinflammatory role of statins.II. Materials and methodsA. ReagentsTwo-chain high molecular weight urokinase typeplasminogen activator (uPA) was from American Diagnostica(Bergstrasse, Germany). VN was purified from human plasmaand converted to the multimeric form as previously described(Chavakis et al, 1998). FBG and fibronectin were purchasedfrom Sigma (Munich, Germany). Vitamin D3 was from Biomol(Hamburg, Germany), transforming growth factor-β was from R& D Systems (Boston, MA), and interleukin-3 was from PBH(Hannover, Germany). Phorbol 12-myristate 13-acetate (PMA)was from Gibco (Paisley, Scotland,UK). The blockingmonoclonal antibody against human CD18, 60.3, was kindlyprovided by Dr. J. Harlan (University of Washington, Seattle,WA), the blocking monoclonal antibody against human CD11a,L15, was a generous gift from Dr. C. Figdor (University ofNijmegen, The Netherlands) and anti-uPAR monoclonalantibodies R3 and R4 (Chavakis et al, 1999) were given by Dr.G. Hoyer-Hansen (The Finsen Laboratory, Copenhagen,Denmark). Monoclonal antibodies K20 against β1-integrins(CD29), 6.5B5 against ICAM-1, 2LPM19c against CD11b,KB90 against CD11c, MHM24 against CD11a and polyclonalrabbit-anti-FBG were from Dako (Hamburg, Germany). IsolatedMac-1, LFA-1 and ICAM-1 were kindly obtained from Dr. S.Bodary (Genentech, San Francisco, CA). Recombinant solubleuPAR was kindly provided by Dr. D. Cines (University ofPennsylvania, Philadelphia, PA). Lovastatin, mevalonate,farnesyl-pyrophosphate and geranyl-pyrophosphate were fromSigma (Munich, Germany). Peroxidase-conjugated secondaryanti-mouse and anti-rabbit immunoglobulins were from DAKO(Hamburg, Germany).B. Cell cultureMyelomonocytic cells (U937) obtained from AmericanType Culture Collection (ATCC) (Rockville, MD) were culturedin RPMI-1640 medium containing 10% (vol/vol) fetal calfserum. K562 cells transfected with Mac-1 were kindly providedby Dr. M. Robinson (Celltech Ltd, Slough, England) and K562cells transfected with LFA-1 or p150.95 were a generous giftfrom Dr. Y. van Kooyk (University of Nijmegen, TheNetherlands) and were cultivated in a mixture of 75% RPMIcontaining 10% fetal calf serum and 25% ISCOVE´s mediumcontaining 5% fetal calf serum. Expression of the respective β2-integrins was tested by FACS analysis (see below). All culturemedia were from Gibco (Eggenstein, Germany), and the cellculture plastic was from Nunc (Rocksilde, Denmark).C. Cell adhesion assaysCell adhesion to VN, ICAM-1 and FBG coated plates (andto BSA-coated wells as control) was tested according topreviously described protocols (Chavakis et al, 1999, 2000, 2001,2002). Briefly, multiwell plates were coated with 5 µg/ml ICAM-1, FBG or 2 µg/ml VN (dissolved in bicarbonate buffer, pH 9.6),respectively, and blocked with 3% (wt/vol) BSA. U937 cells,which had been differentiated for 24 h with vitamin D3 (100 nM)and transforming growth factor-β (2 ng/ml), or K562 cells werewashed in serum-free RPMI and plated onto the precoated wellsfor 60-90 min at 37°C in the absence or presence of competitorsin serum-free RPMI as indicated in the figure legends. Whereindicated, U937 cells were preincubated for various time periodswithout or together with lovastatin in the absence or presence ofmevalonate, farnesyl-pyrophosphate or geranyl-pyrophosphate.Following the incubation period for the adhesion assay, the wellswere washed and the number of adherent cells was quantified by104

Gene Therapy and Molecular Biology Vol 7, page 105crystal violet staining at 590 nm.D. Analysis of uPAR and integrin expressionby flow cytometryAfter incubation for 18 h in the absence or presence oflovastatin differentiated U937 cells were washed twice withHEPES-buffered saline and were incubated with saturatingconcentrations of primary antibody (10 µg/ml) for 60 min at 4°C.Cells were washed again, resuspended in HEPES buffer andphycoerythrin-conjugated F(ab , )2 fragment of goat anti-rabbit (ormouse) IgG (Dianova, Hamburg, Germany) was added insaturating concentrations for 60 min at 4°C. After washing andresuspension, mean fluorescence of 10,000 cells was measured ina flow cytometer (Beckton Dickinson, Heidelberg, Germany).Nonspecific fluorescence was determined using control speciesandisotype-matched primary antibody.inhibitory effect of lovastatin on ICAM-1 adhesion wasunchanged in the presence of the isoprenoid metabolitesmevalonate, farnesyl-pyrophosphate, or geranylpy ro p h o s p h a t e ( F i gu r e 1B). None of these threemetabolites alone could affect U937 cell adhesion toICAM-1 (not shown).U937 cells engage both Mac-1 and LFA-1 forICAM-1-dependent adhesion; however, the lack ofinhibitory activity of lovastatin on Mac-1-related adhesionto FBG indicated that lovastatin interacts only with LFA-1directly.E. ELISA for ligand-receptor interactionsMaxisorp plates (high binding capacity; Nunc) were coatedwith Mac-1 or LFA-1 (5 µg/ml) dissolved in 20 mM HEPES,150 mM NaCl, 1 mM Mn 2+ , pH 7.2 and then blocked with 3%(wt/vol) bovine serum albumin (BSA) in the same buffer.Binding of FBG (10 µg/ml) or ICAM-1 (10 µg/ml) to theimmobilized integrin was performed in a final volume of 50 µl ofthe same buffer as above together with 0.05% (wt/vol) Tween-20and 0.1 % (wt/vol) BSA in the absence or presence of differentcompetitors as indicated in the figure legends. After incubationfor 2 h at 22°C and a washing step, bound ligands were detectedby the addition of polyclonal rabbit anti-FBG or monoclonalmouse anti-ICAM-1 followed by the addition of 1:1000 dilutedperoxidase-conjugated antibody against rabbit or mouseimmunoglobulins, respectively. The conversion of the substrate2,2-azino-di(3-ethly)benzthiazoline sulphate (Boehringer,Mannheim, Germany) was monitored at 405 nm in a Thermomaxmicrotitre plate reader (Molecular Devices, Menlo Park, CA).Nonspecific binding to BSA-coated wells was used as blank andwas subtracted to calculate the specific binding. The sameprotocol was used when binding of multimeric VN (2 µg/ml) toimmobilized uPAR (5 µg/ml, dissolved in bicarbonate buffer, pH9.6) was tested, except that the binding buffer was TBScontaining 0.05 % (wt/vol) Tween-20 0.1 % (wt/vol) BSA.Bound VN was detected with the anti-VN monoclonal antibodyVN7 and additional steps of quantitation were the same asmentioned above.III. ResultsA. HMG-CoA reductase independentregulation of leukocyte adhesion by lovastatinAs previously established, the adhesion of myelomonocyticU937 cells [differentiated with TGFβ (2 ng/ml)and vitamin D3 (100 nM) for 24 h] to immobilized FBG ispredominantly mediated by Mac-1, whereas both Mac-1and LFA-1 mediate adhesion to immobilized ICAM-1.U937 cell adhesion to FBG and ICAM-1 is enhanced byMn 2+ or phorbol ester (PMA). Moreover, U937 celladhesion to VN is uPAR-dependent; uPA can stimulateadhesion, as it increases the affinity of the uPAR/VNinteraction(Chavakis et al, 2000, 2001 Preissner et al,2000). In the presence of lovastatin, adhesion of U937cells to ICAM-1was markedly reduced, whereas adhesionto FBG or VN was not affected at all (Figure 1A). TheFigure 1. U937 cell adhesion to ICAM-1, FBG and VN. (A)PMA (50 ng/ml)-stimulated U937 cell adhesion to immobilizedICAM-1 (5 µg/ml) and FBG (5 µg/ml) or uPA (50 nM)-stimulated U937 cell adhesion to immobilized VN (2 µg/ml) wasstudied in the absence (open bars) or the presence of lovastatin(100 µM, filled bars) or the following blocking antibodies(hatched bars): anti-CD18 (15 µg/ml) for ICAM-1- and FBGmediatedadhesion, anti-uPAR (10 µg/ml) for VN-dependentadhesion. (B) PMA (50 ng/ml)-stimu l a t e d U 937 cell adhesion toimmobilized ICAM-1 (5 µg/ml) was studied in the absence (-) orpresence of a blocking anti-CD18 antibody (15 µg/ml), ablocking anti-LFA-1 (CD11a) antibody (15 µg/ml), lovastatinalone (100 µM), or in combination with mevalonate (100 µM,MEV), farnesyl-pyrophosphate (100 µM, FP), or geranylpyrophosphate(100 µM, GP). Cell adhesion is expressed aspercent of control, which is represented by the adhesion in thepresence of PMA (or uPA, where adhesion to VN is shown) andin the absence of any competitor. Data are mean ± SEM (n=3) ofa typical experiment; similar results were obtained in at leastthree separate experiments.105

Chavakis et al: Leukocyte adhesion and statinsIn order to test this hypothesis in detail, theinhibitory capacity of lovastatin was tested in two furthersystems: (i) In a purified system, lovastatin inhibited onlybinding of ICAM-1 to LFA-1, whereas the binding ofICAM-1 to immobilized Mac-1, the binding of FBG toMac-1 or the binding of VN to immobilized uPAR werenot affected at all (Figure 2). (ii) The effect of lovastatinon adhesion of differently transfected erythroleukemicK562 cells was studied: While non-transfected K562 cellsdid not adhere to FBG or ICAM-1, respectively, cellsbecame adherent to both substrates upon transfection withMac-1 or p150.95, whereas LFA-1 transfected cells onlyadhered to ICAM-1 (not shown). As expected, adhesion ofMac-1 transfected cells to ICAM-1 and FBG was notchanged in the presence of lovastatin, whereas adhesion ofLFA-1 transfected cells was completely inhibited bylovastatin with an IC50 of approximately 20 µM.Interestingly, adhesion of p150.95 transfected cells to bothFBG and ICAM-1 was partially blocked by lovastatin withan IC50 of about 70 µM (Figure 3A and Figure 3B). Theantiadhesive effect of lovastatin on adhesion of both LFA-1- and p150.95- transfected cells was not abolished in thepresence of mevalonate, farnesyl-pyrophosphate orgeranyl-pyrophosphate (Figure 3C and Figure 3D).Taken together, these data indicate that lovastatinselectively interacts with LFA-1 and with a lower potencywith p150.95 but not with Mac-1. Lovastatin thereby canblock LFA-1-mediated cell adhesion to ICAM-1 and to alower extent p150.95-mediated adhesion to FBG andICAM-1 in a manner independent of inhibition of HMG-CoA reductase.Figure 2: Influence of lovastatin on different ligand receptorinteractions. The binding of ICAM-1 (10 µg/ml) to immobilizedMac-1 (open squares) or to immobilized LFA-1 (filled triangles),the binding of FBG (10 µg/ml) to immobilized Mac-1 (filledsquares) or the binding of VN to immobilized uPAR (opencircles) is analyzed in the absence or presence of increasingconcentrations of lovastatin. Specific binding is expressed aspercent of control, which is represented by the binding of theligand to the respective immobilized receptor in the absence oflovastatin. Data are mean ± SEM (n=3) of a typical experiment;similar results were obtained in at least three separateexperiments.Figure 3: Influence of lovastatin coincubation on the adhesion ofK562 cells. PMA (50 ng/ml)-stimulated adhesion of Mac-1-transfected K562 cells (filled squares), p150.95-transfected K562cells (open circles) and LFA-1-transfected K562 cells (filledtriangles) to immobilized ICAM-1 (5 µg/ml) (A) and PMA (50ng/ml)-stimulated adhesion of Mac-1-transfected K562 cells(filled squares) and p150.95-transfected K562 cells (open circles)to immobilized FBG (5 µg/ml) (B) was studied in the presence ofincreasing concentrations of lovastatin. PMA (50 ng/ml)-stimulated adhesion of Mac-1-transfected K562 cells, p150.95-transfected K562 cells and LFA-1-transfected K562 cells toimmobilized ICAM-1 (5 µg/ml) (C) and PMA (50 ng/ml)-stimulated adhesion of Mac-1-transfected K562 cells andp150.95-transfected K562 cells to immobilized FBG (5 µg/ml)(D) was studied in the absence (open bars) or presence oflovastatin alone (100 µM, filled bars), or in combination withmevalonate (100 µM, hatched bars), farnesyl-pyrophosphate(100 µM, dotted bars), or geranyl-pyrophosphate (100 µM,vertical lines). Cell adhesion is shown as percent of control,which is represented by the adhesion of cells in the absence ofany competitor. Data are mean ± SEM (n=3) of a typicalexperiment; similar results were obtained in at least threeseparate experiments.106

Gene Therapy and Molecular Biology Vol 7, page 107B. HMG-CoA reductase-dependentregulation of leukocyte adhesion by lovastatinIn contrast to the described direct antiadhesive effectof lovastatin on cells during coincubation, a completelydifferent pattern of inhibition was observed whenlovastatin was preincubated with leukocytes for up to 18 hfollowed by removal of excess reagent prior to the celladhesion experiment. In particular, lovastatin preincubatedfor 18 h with U937 cells dose-dependently inhibited theiradhesion to ICAM-1, FBG or VN. The inhibitory capacitywas almost identical in all three systems (IC50 of about 1-2 µM) (Figure 4). Furthermore, the following differenceswere observed between U937 cell adhesion to ICAM-1and adhesion to FBG and VN: In the time course, after 5 hof incubation with lovastatin about 30 % inhibition ofU937 cell adhesion to FBG and VN was observed andinhibition was almost complete after 12 h. At all timepoints the effect of lovastatin was restored by mevalonate.Farnesyl-pyrophosphate or geranyl-pyrophosphate as wellcould completely reverse the antiadhesive effect oflovastatin on cell adhesion to FBG and VN (Figure 5). Incontrast, already after 2 h of lovastatin preincubationadhesion to ICAM-1 was inhibited by 50% but could notbe restored by mevalonate. Again, after 12 h lovastatinpreincubation U937 cell adhesion to ICAM-1 wascompletely abolished However, this effect was onlypartially (50% of initial adhesion) reversed in the presenceof mevalonate reaching a cell adhesion level that wascomparable to cell adhesion after 2 h lovastatinpreincubation (Figure 5). Thus, the action of lovastatinpreincubation on U937 cell adhesion to ICAM-1 consistsof two components, a HMG-CoA reductase-independentdirect blocking effect on LFA-1 and a HMG-CoAreductase-dependent effect.The HMG-CoA reductase-dependent a n ti a dh e s ivee f fe c t of lo va s ta ti n pr e i nc u ba tio n mig ht result from adownregulation of the expression of respective adhesionreceptors, namely β2-integrins or uPAR. However,lovastatin preincubation for 18 h did not affect theexpression level of uPAR, β2-integrins (no change inCD11a, CD11b and CD18 expression) or β1 integrins(CD29) (Table 1). The CD11c chain was not detected onU937 cells, explaining the lack of inhibition of U937 celladhesion to FBG by coincubation with lovastatin (Figure1).In conclusion, these findings indicate that lovastatinpreincubation can regulate both β2-integrin and uPARmediatedleukocyte adhesion in a cholesterol biosynthesisdependentmanner without changing the expression levelof β2-integrins or uPAR.Figure 4: Influence of lovastatin preincubation on U937 celladhesion. Following preincubation for 18 h in the absence orpresence of increasing concentrations of lovastatin, adhesion ofPMA (50 ng/ml)-stimulated U937 cells to immobilized ICAM-1(5 µg/ml) (filled triangles), to immobilized FBG (5 µg/ml) (opensquares) or uPA (50 nM)-stimulated U937 cell adhesion toimmobilized VN (2 µg/ml) (open circles) was studied. Celladhesion is expressed as percent of control, which is representedby the adhesion in the presence of PMA (or uPA, where adhesionto VN is shown) and in the absence of lovastatin. Data are mean± SEM (n=3) of a typical experiment; similar results wereobtained in three separate experiments.Figure 5: Influence of preincubation of lovastatin and isoprenoidmetabolites on U937 cell adhesion. Following preincubation forvarious time periods as indicated, PMA (50 ng/ml)-stimulatedU937 cell adhesion to (A) immobilized ICAM-1 (5 µg/ml), to(B) immobilized FBG (5 µg/ml) or (C) uPA (50 nM)-stimulatedU937 cell adhesion to immobilized VN (2 µg/ml) was studied inthe absence (vertical lines) or presence of lovastatin (20 µM)alone (open bars) or in combination with mevalonate (100 µM,filled bars). In the 18 h preincubation setting lovastatin was alsoreacted together with farnesyl-pyrophosphate (100 µM, hatchedbars) or geranyl-pyrophosphate (100 µM, dotted bars). Celladhesion is expressed as percent of control, which is representedby the adhesion in the presence of PMA (or uPA, where adhesionto VN is shown) and in the absence of any competitor. Data aremean ± SEM (n=3) of a typical experiment; similar results wereobtained in three separate experiments.107

Chavakis et al: Leukocyte adhesion and statinsTable 1: Influence of lovastatin on integrin and uPAR expression.Receptors Control LovastatinCD11a 100+8.2 92.6+3.1CD11b 100+6.4 97.3+5.3CD18 100+12.9 110.7+9.1CD29 100+8.4 106.7+4.5uPAR 100+8.9 97.9+1.7The expression of CD11a, CD11b, CD18, CD29 and uPAR on U937 cells that were preincubated for 18 h in the absence or presence oflovastatin (40 µM) as measured by FACS-analysis is shown. The expression of the various integrins or uPAR is presented as percent ofcontrol, which relates to the expression of the respective adhesion molecule in the absence of lovastatin. Data are mean ± SEM (n=3) ofa typical experiment; similar results were obtained in three separate experiments.IV. DiscussionLeukocyte activation and adhesion to theendothelium and the subsequent transendothelial migrationare pivotal steps in the recruitment of cells to theinflammatory /injured tissue. This highly coordinatedmultistep process requires tight regulation of adhesiveevents (Carlos and Harlan, 1994; Springer, 1994)including the induction of genes coding for participatingadhesion receptors including integrins, their change inavidity as well as the modification of ligand-bindingproperties (Porter and Hogg, 1998; Woods and Couchman,2000). Conversely, in pathological situations associatedwith organ transplantation, atherosclerosis andischemia/reperfusion injury, arthritis and psoriasis theantagonism of these adhesive leukocytic interactions maybecome a promising therapeutic appproach (Nahakura etal, 1996; Issekutz, 1998; Kruegeret al, 2000; Martin et al,2000; Poston et al, 2000). In this respect, recent evidencepoints to an immunomodulatory role of statins (Katznelsonand Kobashigawa, 1995; Maron et al, 2000; Kwak andMach, 2001) which are commonly used to reduce plasmacholesterol levels in order to decrease the risk ofcardiovascular disease (Corsini et al, 1995). In this studywe define the direct and indirect role of statins inleukocyte adhesion and the possible underlyingmechanisms. Two distinct pathways, a HMG-CoAreductase-dependent and an –independent weredistinguished and appear to be relevant for theantiadhesive effects of statins.In particular, coincubation of monocytes withlovastatin resulted in a dramatic reduction of LFA-1-dependent cell adhesion to ICAM-1, but not of Mac-1-dependent adhesion to FBG or uPAR-dependent adhesionto VN. This direct antiadhesive effect of lovastatin wasunrelated to HMG-CoA reductase inhibition, as it was notreversed by mevalonate or other isoprenoid metabolites.Rather, it was attributed to the direct inhibition of theLFA-1/ICAM-1 interaction by lovastatin as corroboratedin a purified system. Whereas Mac-1 binding to its ligandsICAM-1 and FBG as well as uPAR interaction with VNwere not directly affected by lovastatin, binding of anotherβ2-integrin, p150.95, to FBG and ICAM-1 was partiallyblocked directly by lovastatin. Our data are in accordancewith and extend a recent report showing that statins inhibitLFA-1 by binding to an allosteric L-site located within theI-domain of the α chain (Weitz-Schmidt et al, 2001).Thus, lovastatin binds to LFA-1 as well as with loweraffinity to p150.95, but not to Mac-1, thereby directlyaffecting leukocyte adhesion.When lovastatin was preincubated with monocytesfor up to 18 h, a different inhibition profile was observed:Lovastatin completely blocked all three adhesive events,namely LFA-1/Mac-1-dependent adhesion to ICAM-1,Mac-1-dependent adhesion to FBG and uPAR-dependentadhesion to VN. Inhibition of ICAM-1-related adhesioncould be partially attributed to the direct LFA-1 bindingproperty of lovastatin, as (i) a significant inhibition by50% occured already after 2 h, and was not reversed bymevalonate and (ii) complete inhibition was observed afterlonger preincubation times (12-18 h) and could bepartially reversed by mevalonate up to the adhesion levelobtained after 2 h preincubation with lovastatin. Incontrast, both Mac-1- and uPAR-dependent cell adhesionwere partially inhibited after 6 h preincubation withlovastatin and were completely blocked after 12-18 h. Thiseffect of lovastatin was dependent on HMG-CoAreductase inhibition, as it was completely reversible in thepresence of mevalonate. Interestingly, the IC50 of theHMG-CoA reductase-dependent effect of lovastatin wasapproximately 1 µM, which is about 20 times (LFA-1) or70 times (p150.95) lower than the IC50 of the HMG-CoAreductase-independent direct abrogation of both integrinmediatedadhesion reactions. Thus, the antiinflammatoryaction of statins implied in clinical studies are very likelyattributable to the HMG-CoA reductase-dependentpathway, as the higher concentrations of statins requiredfor the direct inhibition of the LFA-1/ICAM-1-, thep150.95/FBG- and the p150.95/ICAM-1-interactions maynot be reached with the standard doses (nanomolar range)of approved statin drugs (Frenette, 2001). Indeed, a recentreport demonstrated that mevalonate-derived isoprenoidmetabolites mediate the antiinflammatory activity ofstatins in the in vivo air-pouch model of localinflammation (Diomede et al, 2001). Finally, the antiinflammatorycapacity of statins may vary dependent ontheir individual structure (Weitz-Schmidt et al, 2001).While direct binding to LFA-1 and p150.95sufficiently explains the HMG-CoA reductaseindependentantiadhesive effect of lovastatin, differentmechanisms might be involved in the HMG-CoAreductase-dependent anti-adhesive property of lovastatin:(i) Lowering the plasma membrane cholesterol content canaffect cell adhesion by disrupting lipid raft integrity(Krauss and Altevogt, 1999; Simons and Toomre, 2000).Recently, the assembly of adhesion complexes containing108

Gene Therapy and Molecular Biology Vol 7, page 109adhesion receptors as well as signaling molecules such asfocal adhesion kinase or src kinases has been proposed tobe confined to glycosphingolipid- and cholesterol-rich,detergent insoluble microdomains of the cell membrane.The antiadhesive effect of lovastatin preincubationpresented here could very well be due to raft disruption bycholesterol depletion, as other approaches to disrupt thesemembrane microdomains result in a very similardownregulation of β2-integrin and uPAR mediatedleukocyte adhesion (Chavakis et al., unpublishedobservations). Moreover, as lipid rafts have beenimplicated in T-cell receptor-, EGF-receptor-, insulinreceptor-, H-Ras-, eNOS- and integrin-dependentsignalling phenomena (Simons and Toomre, 2000), thepotential modulatory role of HMG-CoA-reductaseinhibitors on raft integrity and associated vital cellularfunctions renders these drugs very attractive for severaltherapeutic interventions in vascular medicine. (ii)Although conflicting results have been reported as to theinfluence of statins on the cell type specific integrin anduPAR expression (Weber et al, 1997; Liu et al, 1999;Wojeiak-Stothard, 1999; Yoschida et al, 2001), our dataare in accordance with these reports showing no change inintegrin expression in e.g. myelo-monocytic U937 cells bylovastatin (Weber et al, 1995; Liu et al, 1999). (iii) It hasbeen demonstrated that protein geranyl-geranylation isrequired for β1-integrin-dependent adhesion of leukocytes.It is thus conceivable that statin treatment may affectintegrin-dependent leukocyte adhesion via inhibition ofthe geranyl-geranylation of RhoA, which is thought to beone of the most important effectors involved in regulationof the cytoskeleton network, including the clustering ofadhesion molecules during monocyte adherence (Liu et al,1999; Wojciak-Stothard et al, 1999; Kwak and Mach,2001; Yoshida et al, 2001). The possibility that statintreatment could thereby directly inhibit RhoA activationand disrupt actin polymerization leading to failure ofintegrin clustering is a likely interpretation of thepresented data, since isoprenoid metabolites could reversethe antiadhesive effect of lovastatin pretreatment.Together, our findings help to decipher the mechanismsunderlying the postulated antiinflammatory effects ofstatins, which, besides atherothrombosis, may prove to bebeneficial in arthritis, organ transplantation or psoriasis.AcknowledgmentsThis work was supported in part by a grant from theNovartis Foundation for Therapeutical Research to TC andKTP (Nürnberg, Germany), by a grant from the DeutscheForschungsgemeinschaft to TC (CH279/1-1) and by agrant from Vascular Genomics-Kerckhoff Klinik GmbH toKTP (Bad Nauheim, Germany). We acknowledge thegenerous gift of reagents from Drs. D.B. Cines(Philadelphia, PA), G. Hoyer-Hansen and N. Behrendt(Copenhagen, Denmark), S. Bodary (San Francisco, CA)and J. Harlan (Seattle, WA) and Ms M. Economopouloufor help during manuscript preparation.ReferencesAikawa M, Rabkin E, Okada Y, Voglic SJ, Clinton SK,Brinckerhoff CE, Sukhova GK, Libby P (1998) Lipidlowering by diet reduces matrix metalloproteinase activityand increases collagen content of rabbit atheroma: a potentialmechanism of lesion stabilization. Circulation 97,2433–2444Blackford J, Reid HW, Pappin DJ, Bowers FS, Wilkinson JM(1996) A monoclonal antibody, 3/22 to rabbit CD11c whichinduces homotypic T cell aggregation: evidence that ICAM-1is a ligand for CD11c/CD18. Eur J Immunol 26, 525-531Carlos TM, Harlan JM (1994) Leukocyte-endothelial adhesionmolecules. Blood 84, 2068-2101Chavakis T, Kanse SM, Yutzy B, Lijnen HR, Preissner KT(1998) Vitronectin concentrates proteolytic activity on thecell surface and extracellular matrix by trapping solubleurokinase receptor-urokinase complexes. Blood 91, 2305-2312Chavakis T, May AE, Preissner KT, Kanse SM (1999) Molecularmechanisms of zinc-dependent leukocyte adhesion involvingthe urokinase receptor and β2-integrins. Blood 93, 2976-2983Chavakis T, Kanse SM, Lupu F, Hammes HP, Muller-Esterl W,Pixley RA, Colman RW, Preissner KT (2000) Differentmechanisms define the antiadhesive function of highmolecular weight kininogen in integrin- and urokinasereceptor-dependent interactions. Blood 96, 514-522Chavakis T, Kanse SM, Pixley RA, May AE, Isordia-Salas I,Colman RW, Preissner KT (2001) Regulation of leukocyterecruitment by polypeptides derived from high molecularweight kininogen. FASEB J 15 2365-2376Chavakis T, Hussain M, Kanse SM, Peters G, Bretzel RG, FlockJI, Herrmann M, Preissner KT (2002) Staphylococcus aureusextracellular adherence protein (Eap) serves as antiinflammatoryfactor by inhibiting the recruitment of hostleukocytes. Nature Medicine 8, 687-693Corsini A, Maggi FM, Catapano AL (1995) Pharmacology ofcompetitive inhibitors of HMG-CoA reductase. PharmacolRes 31, 9–27Diomede L, Albani D, Sottocorno M, Polentarutti N, Donati MB,Bianchi M, Fruscella P, Salmona M (2001) The in vivo antiinflammatoryeffect of statins is mediated by nonsterolmevalonate products. Arterioscler Thromb Vasc Biol 21,1327–1332Essig M, Nguyen G, Prie D, Escoubet B, Sraer JD, Friedlander G(1998) 3-Hydroxy-3-methylglutaryl coenzyme A reductaseinhibitors increase fibrinolytic activity in rat aorticendothelial cells: role of geranylgeranylation and Rhoproteins. Circ Res 83, 683–690Frenette PS (2001) Locking a leukocyte integrin with statins. NEngl J Med 345, 1419-1421Gahmberg CG (1997) Leukocyte adhesion: CD11/CD18integrins and intercellular adhesion molecules. Curr OpinCell Biol 9, 643-650Ganne F, Vasse M, Beaudeux JL, Peynet J, Francois A, MishalZ, Chartier A, Tobelem G, Vannier JP, Soria J, Soria C(2000) Cerivastatin, an inhibitor of HMG-CoA reductase,inhibits urokinase/urokinase-receptor expression and MMP-9secretion by peripheral blood monocytes-a possibleprotective mechanism against atherothrombosis. ThrombHaemost 84, 680-688Guijarro C, Blanco-Colio LM, Ortego M, Alonso C, Ortiz A,Plaza JJ, Diaz C, Hernandez G, Edigo J (1998) 3-Hydroxy-3-109

Chavakis et al: Leukocyte adhesion and statinsmethylglutaryl coenzyme A reductase and isoprenylationinhibitors induce apoptosis of vascular smooth muscle cellsin culture. Circ Res 83, 490 –500Horejsi V, Drbal K, Cebecauer M, Cerny J, Brdicka T,Angelisova P, Stockinger H (1999) GPI-microdomains: arole in signalling via immunoreceptors. Immunol Today 20,356-361Issekutz AC (1998) Adhesion molecules mediating neutrophilmigration to arthritis in vivo and across endothelium andconnective tissue barriers in vitro. Inflamm Res 47,S123–S132Kallen J, Welzenbach K, Ramage P, Geyl D, Kriwacki R, LeggeG, Cottens S, Weitz-Schmidt G, Hommel U (1999)Structural basis for LFA-1 inhibition upon lovastatin bindingto the CD11a I-domain. J Mol Biol 292, 1-9Katznelson S, Kobashigawa JA (1995) Dual roles of HMG-CoAreductase inhibitors in solid organ transplantation: Lipidlowering and immunosuppression. Kidney Int 48,S112–S115Krauss K, Altevogt P (1999) Integrin leukocyte functionassociatedantigen-1-mediated cell binding can be activatedby clustering of membrane rafts. J Biol Chem 274, 36921-36927Krueger J, Gottlieb A, Miller B, Dedrick R, Garovoy M, WalickeP (2000) Anti-CD11a treatment for psoriasis concurrentlyincreases circulating T-cells and decreases plaque T-cells,consistent with inhibition of cutaneous T-cell trafficking. JInvest Dermatol 115, 333Kurzchalia TV, Parton RG (1999) Membrane microdomains andcaveolae. Curr Opin Cell Biol 11, 424-431Kwak BR, Mach F (2001) Statins inhibit leukocyte recruitment.New evidence for their anti-inflammatory properties.Arterioscler Thromb Vasc Biol 21, 1256-1258Kwak B, Mulhaupt F, Myit S, Mach F (2000) Statins as a newlyrecognized type of immunomodulator. Nat Med 6, 1399-1403Laufs U, La Fata V, Plutzky J, Liao JK (1998) Upregulation ofendothelial nitric oxide synthase by HMG CoA reductaseinhibitors. Circulation 97, 1129–1135Laufs U, Liao JK (1998) Post-transcriptional regulation ofendothelial nitric oxide synthase mRNA stability by RhoGTPase. J Biol Chem 273, 24266–24271Laufs U, Marra D, Node K, Liao JK (1999) 3-Hydroxy-3-methylglutaryl-CoA reductase inhibitors attenuate vascularsmooth muscle proliferation by preventing rho GTPaseinduceddown-regulation of p27(Kip1). J Biol Chem 274,21926 –21931Liu L, Moesner P, Kovach NL, Bailey R, Hamilton AD, SebtiSM, Harlan JM (1999) Integrin-dependent leukocyteadhesion involves geranylgeranylated protein(s). J BiolChem 274, 33334–33340Maron DJ, Fazio S, Linton MF (2000) Current perspectives onstatins. Circulation 101, 207–213Martin X, Da Silva M, Virieux SR, Hadj Aissa A, Buffet R,Tiollier J, Dubernard JM (2000) Protective effect of an anti-LFA 1 monoclonal antibody (odulimomab) on renal damagedue to ischemia and kidney autotransplantation. TransplantProc 32, 481May AE, Kanse SM, Lund LR, Gisler RH, Imhof BA, PreissnerKT (1998) Urokinase receptor (CD87) regulates leukocyterecruitment via β2-integrins in vivo. J Exp Med 188, 1029-1037Nakakura EK, Shorthouse RA, Zheng B, McCabe SM, JardieuPM, Morris RE (1996) Long-term survival of solid organallografts by brief anti-lymphocyte function-associatedantigen-1 monoclonal antibody monotherapy.Transplantation 62, 547–552Ossowski L, Aguirre-Ghiso JA (2000) Urokinase-receptor andintegrin partnership: coordination of signaling for celladhesion, migration and growth. Curr Opin Cell Biol 12,613-620Porter JC, Hogg N (1998) Integrins take partners: cross-talkbetween integrins and other membrane receptors. TrendsCell Biol 8, 390-396Poston RS, Robbins RC, Chan B, Simms P, Presta L, Jardieu P,Morris RE (2000) Effects of humanized monoclonalantibody to rhesus CD11a in rhesus monkey cardiac allograftrecipients. Transplantation 69, 2005–2013Preissner KT, Kanse SM, May AE (2000) Urokinase receptor: amolecular organizer in cellular communication. Curr OpinCell Biol 12, 621-628Romano M, Diomede L, Sironi M, Massimiliano L, SottocornoM, Polentarutti N, Guglielmotti A, Albani D, Bruno A,Fruscella P, Salmona M, Vecchi A, Pinza M, Mantovani A(2000) Inhibition of monocyte chemotactic protein-1synthesis by statins. Lab Invest 80, 1095–1100Simons K, Toomre D (2000) Lipid rafts and signal transduction.Nat Rev Mol Cell Bio 1, 31-39Sitrin RG, Todd RF, Petty HR, Brock TG, Shollenberger SB,Albrecht E, Gyetko MR (1996) The urokinase receptor(CD87) facilitates CD11b/CD18-mediated adhesion ofhuman monocytes. J Clin Invest 97, 1942-1951Smart EJ, Graf GA, McNiven MA, Sessa WC, Engelman JA,Scherer PE, Okamoto T, Lisanti MP (1999) Caveolins,liquid-ordered domains, and signal transduction. Mol CellBiol 19, 7289-7304Soma MR, Donetti E, Parolini C, Mazzini G, Ferrari C,Fumagalli R, Paoletti R (1993) HMG CoA reductaseinhibitors: in vivo effects on carotid intimal thickening innormocholesterolemic rabbits. Arterioscler Thromb 13,571–578Springer TA (1994) Traffic signals for lymphocyte recirculationand leukocyte emigration: The multistep paradigm. Cell 76,301-314Stewart M, Thiel M, Hogg N (1995) Leukocyte integrins. CurrOpin Cell Biol 7, 690-696Weber C, Erl W, Weber KS, Weber PC (1997) HMG-CoAreductase inhibitors decrease CD11b expression and CD11bdependentadhesion of monocytes to endothelium and reduceincreased adhesiveness of monocytes isolated from patientswith hypercholesterolemia. J Am Coll Cardiol 30,1212–1217Weber C, Erl W, Weber PC (1995) Lovastatin inducesdifferentiation of Mono Mac 6 cells. Cell Biochem Funct13, 273-277Wei Y, Waltz DA, Rao N, Drummond RJ, Rosenberg S,Chapman HA (1994) Identification of the urokinase receptoras an adhesion receptor for vitronectin. J Biol Chem 269,32380-32388Weitz-Schmidt G, Welzenbach K, Brinkmann V, Kamata T,Kallen J, Bruns C, Cottens S, Takada Y, Hommel U (2001)Statins selectively inhibit leukocyte function antigen-1 bybinding to a novel regulatory integrin site. Nat Med 7, 687-692110

Gene Therapy and Molecular Biology Vol 7, page 111Wojciak-Stothard B, Williams L, Ridley AJ (1999) Monocyteadhesion and spreading on human endothelial cells isdependent on Rho-regulated receptor clustering. J Cell Biol145, 1293–1307Woods A, Couchman JR (2000) Integrin modulation by lateralassociation. J Biol Chem 275, 24233-24236Yoshida M, Sawada T, Ishii H, Gerszten RE, Rosenzweig A,Gimbrone MA Jr, Yasukochi Y, Numano F (2001) HMG-CoA reductase inhibitor modulates monocyte endothelialinteraction under physiological flow condition in vitro:involvement of Rho GTPase-dependent mechanism.Arterioscler Thromb Vasc Biol 21, 1165–1171111

Chavakis et al: Leukocyte adhesion and statins112

Gene Therapy and Molecular Biology Vol 7, page 113Gene Ther Mol Biol Vol 7, 113-133, 2003Current progress in adenovirus mediated genetherapy for patients with prostate carcinomaReview ArticleAhter D. Sanlioglu 1,3 , Turker Koksal 2,3 , Mehmet Baykara 2,3 , Guven Luleci 1,3 , BahriKaracay 4 and Salih Sanlioglu 1,3, *1 Departments of Medical Biology and Genetics, 2 Department of Urology and 3 The Human Gene Therapy Unit of AkdenizUniversity, Faculty of Medicine, Antalya, Turkey, 07070; 4 Department of Pediatrics, University of Iowa, College ofMedicine, Iowa City, IA, 52240, USA__________________________________________________________________________________*Correspondence: Salih Sanlioglu V.M.D., Ph.D., Director of The Human Gene Therapy Unit of Akdeniz University, Faculty ofMedicine, B- Block, 1 st floor, Campus, Antalya, 07070 Turkey; Phone: (90) 242-227-4343/ext: 44359, Fax: (90) 242-227-4482; e-mail:sanlioglu@akdeniz.edu.trKey words: Prostate cancer, adenovirus, gene therapy, immunomodulation, apoptosis, inducible promotersReceived: 1 July 2003; Accepted: 11 July 2003; electronically published: July 2003SummaryProstate cancer is the most frequently diagnosed male cancer in the world. Like all cancers, prostate cancer is adisease of uncontrolled cell growth. In some cases tumors are slow growing and remain local, but in others they mayspread rapidly to the lymph nodes, other organs and especially bone. Although surgery and radiation can cure earlystages of organ confined prostate carcinoma (stages I and II), there is no curative therapy at this time for locallyadvanced or metastatic disease (stages III and IV). The likelihood of postsurgical local recurrence increases withcapsular penetration as detected in 30 % of the patients at the time of radical prostatectomy. Moreover, 10-15 % ofpatients have metastatic cancer at the time of diagnosis. Considering the fact that 60 % local recurrence is observedin patients receiving radiation therapy with or without adjuvant hormonal ablation therapy, it is generally believedthat androgen ablation therapy simply delays the progression of prostate carcinoma to a more advanced stage. Inaddition, the overall ten-year survival rate of patients with locally recurrent prostate cancer is only around 35 %;thus; the ultimate progression into androgen independent prostate carcinoma appears to be inevitable. Genetherapy arose as a novel treatment modality with the potential to decrease the morbidity associated withconventional therapies. Therefore, gene therapy is expected to lower the incidence of tumor recurrence and finallyimprove the outcome of patients with recurrent and androgen independent prostate carcinoma. Viral vectors aremost commonly used for the purpose of gene therapy. Currently, there are a total of 40 clinical trials beingconducted using viral vectors for the treatment of prostate carcinoma. 22 out of 40 clinical protocols (55 %)approved for the treatment of prostate cancer utilize adenovirus vectors. Most of these adenovirus mediatedtherapeutic approaches employ either selectively replicating adenoviruses or suicide gene therapy approaches. Inthis review, we mainly concentrated on the progress in adenovirus mediated gene therapy approaches for prostatecancer. Analysis of the death ligand mediated gene therapy approach was also discussed in detail, while our novelfindings were incorporated as an example for up-to-date approaches used for adenovirus mediated gene therapyagainst prostate carcinoma.I. IntroductionProstate cancer is the second leading cause of deathin men from cancer following lung carcinoma with anannual mortality rate of 38,000 (Yeung and Chung, 2002).There are 200,000 newly diagnosed cases of prostatecarcinoma every year in the United States alone (Boring etal, 1994; Greenlee et al, 2001). As a result, prostatecarcinoma is claimed to be the most frequently diagnosedmale cancer in the United States (Powell et al, 2002).Despite the fact that there has been a considerable effortfor screening and early detection of prostate cancer inrecent years, the lifetime risk of being diagnosed withprostate cancer is still reported to be 1 in 5 (Grumet andBruner, 2000). Several hundred clinical studies usingexperimental or approved chemotherapeutics failed toimprove survival rates of patients with prostate cancer(Devi, 2002). Because prostate cancer is a heterogeneous113

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomadisease, treating patients with prostate cancer still remainsa formidable task. In addition, the molecular mechanismresponsible for the onset of the disease is poorlyunderstood. However, earlier detection of prostate cancerhas been associated with an improved outcome (Perrotti etal, 1998). Thus, the detection of prostate cancer at anearlier stage remains to be the most realistic chance fortherapy.For this purpose, different molecular screeningmethods (Ross et al, 2002a, 2002b) have been employed,but the most effective method is yet to be established. Themost commonly used screening assays are based on thedetection of up-regulated prostate specific markers such asprostate specific antigen (PSA). Currently, prostatespecific antigen, (Farkas et al, 1998) when it is used inconjunction with other markers such as Gleason Scoring(Koksal et al, 2000) and TNM grading (Schroder et al,1992), is considered to be a valuable tool to evaluate thehistological grade of prostate carcinomas (Xess et al,2001). As a result, patients were provided with varioustreatment options based on the results obtained with theseparameters. These treatment options included but were notlimited to operation, (Klotz, 2000b) radiotherapy, (Do etal, 2002) chemotherapy (Wang and Waxman, 2001) andhormone therapy (Klotz, 2000a; Smith et al, 2002).Regrettably, these conventional treatment modalities couldnot decrease the casualties from prostate cancer (Hsiehand Chung, 2001). Hence, there is a great need fordevelopment of novel treatment modalities to fight againstprostate cancer. These remorseful facts ignited theinitiation of gene therapy trials for prostate carcinoma(Sanda, 1997). So far, various viral vectors includinglentivirus (Yu et al, 2001a), herpes simplex virus(Jorgensen et al, 2001), adeno-associated virus (Vieweg etal, 1995) and adenovirus (Loimas et al, 2001) were testedas carriers for therapeutic genes against prostate cancer.Other types of viruses such as Semliki Forest virus andSindbis virus were also tested for gene delivery to prostatecancer cells (Loimas et al, 2001), but these viruses wereunable to transduce prostate cells efficiently. Due to itsantigenic properties and tissue transduction characteristics,adenovirus arose as a favored transporter vector. Theexploitation of the tissue specific promoter in gene therapyespecially eased adenovirus use in clinical trials (Lu andSteiner, 2000). In this review, we mainly highlighted theprogress in adenovirus mediated prostate cancer genetherapy within the last three years with a particularemphasis in death ligand mediated gene therapy approach.II. ImmunomodulationTumors exhibit some degree of immunogenicity andthe human immune system responds to these tumorspecific antigens by mounting humoral and cellularresponses, which are essential for the eradication oftumors. Adenovirus is commonly used for the delivery ofgenes encoding tumor-associated antigens in order toaugment tumor-specific immune responses. However,antiviral immunity against adenovirus is a big concern,challenging its application in gene therapy. Variousmethods were employed in order to get around theantiviral immunity barrier to increase the efficacy ofadenovirus mediated gene delivery. One of these methodsinvolves the testing of a collagen-based matrix (Gelfoam)(Siemens et al, 2001). Coinjection of Gelfoam withadenovirus vectors carrying prostate-specific antigen(Ad5-PSA) into mice naive to PSA but immune toadenovirus, relinquished the inhibitory effects ofadenoviral immunity on CTL activation. Viral vectors arealso being tested to deliver tumor specific peptides intodendritic cells (DCs) to evoke an immune response. Thedegree of immune response generated relies on thefunctionality of DCs following viral transduction. Toprove this, adenovirus and retrovirus vectors werecompared on the basis of their influence on thefunctionality of DCs (Lundqvist et al, 2002a). Adenovirustransducedmonocyte-derived DCs (MO-DCs) stimulatedallogenic lymphocytes and produced high levels of TNFand IL12. In addition, the expression of NF-κB andantiapoptotic molecules such as Bcl-X(L) and Bcl-2(Lundqvist et al, 2002b) were also increased inadenovirus-transduced MO-DCs. Consequently, thesecells became more resistant to spontaneous as well as Fasmediatedcell death. In contrast, retroviruses failed even totransduce MO-DCs. Although CD34(+) cell-derived DCswere transducable with retroviruses to a lesser extent, theywere less potent in their ability to stimulate allogeniclymphocytes in comparison to nontransduced DCs. Theseresults suggest that adenovirus transduction of DCsincreased the survival and the potency of DC mediatedactivation of the immune system. This might be importantfor prolonging the antigen presentation to generate agreater degree of immune response.Cytokine stimulated tumor infiltrating macrophagesalso play a major role in the generation of the cellularimmune response against the tumor. The role of tumorinfiltratingmacrophages in IFN-β-induced host defenseagainst prostate cancer was revealed using xenograft micemodels injected with adenovirus carrying IFN-β gene(Zhang et al, 2002a). Injection of an adenoviral vectorencoding murine IFN-β (AdIFN-β) directly into the tumorsuppressed the growth of PC-3MM2 tumors as well asprevented metastasis and prolonged the survival of tumorbearingmice. Based on immunohistochemical staining,AdIFN-β infection resulted in the reduction of microvesseldensity of the tumor and increased apoptotic cell death(Cao et al, 2001). On the contrary, macrophage-selectiveanti-Mac-1 and anti-Mac-2 antibodies significantlyreduced the antitumor effect of AdIFN-β induced therapy.Therefore, it was concluded that tumor-infiltratingmacrophages must be involved in IFN-β inducedsuppression of tumor growth and metastasis.III. Suicide Gene TherapySuicide strategy is a combined treatment modalityinvolving chemotherapy and the gene transfer technology.The underlying principle is to limit the cytotoxicity of adrug to the local area of the tumor. To achieve this, thecDNA of a prodrug-converting enzyme is delivered intothe tumor using viral vectors followed by regional orsystemic application of the corresponding prodrug. As114

Gene Therapy and Molecular Biology Vol 7, page 115soon as the prodrug reaches the tumor, it is taken up andconverted to a cytotoxic drug by tumor cells expressingthe prodrug-converting enzyme. For example, 5-Fluorouracil (5-FU) is widely used as a chemotherapeuticagent for the treatment of various malignancies. Althoughclinical trials have been conducted, so far 5-FU manifesteda poor therapeutic index, which drastically limited itsclinical use for cancer therapy. It is still not knownwhether the lack of success was due to problemsassociated with drug delivery or inherent insensitivity ofcancer cells to this metabolite. However, adenovirus (Ad)vector-mediated cytosine deaminase (CD)/5-fluorocytosine (5-FC) gene therapy had the potential toovercome pharmacokinetic issues associated with systemic5-FU administration. Escherichia coli cytosine deaminaseconverts the prodrug 5-FC to the cytotoxic product 5-FU.Adenovirus encoding cytosine deaminase (AdCD) genewas injected into the prostate cancer cells transplantedorthotopically on mice followed by the systemic use of 5-FC in order to investigate the antitumor and antimetastaticeffects of this approach (Zhang et al, 2002c).An effective inhibition on tumor growth and metastasiswas observed through in situ injection of AdCD followedby systemic use of 5-FC in the xenograft mouse model ofprostate cancer. The use of E. coli uracilphosphoribosyltransferase (UPRT), a pyrimidine salvageenzyme, which modifies 5-FU into 5-fluorouridinemonophosphate, improved the activity of AdCD throughenhancing the anti-tumoral effect of 5-FU. In order toassess the efficacy of the combined suicide gene therapyapproach, two separate adenovirus constructs expressingeither the E. coli CD or E. coli UPRT genes were infectedinto androgen refractory prostate cancer cell line DU145bearing mice. This combined gene therapy approachdrastically regressed the growth of tumors in these animalsbetter than what was achieved with AdCD alone (Miyagiet al, 2003).The most commonly used prodrug-convertingenzyme for clinical approaches is the herpes simplex virusthymidine kinase gene (HSV-tk). The enzyme thymidinekinase phosphorylates the prodrug ganciclovir (GCV) toganciclovir monophosphate, which is then furtherphosphorylated by cellular enzymes to ganciclovirtriphosphate, a toxic metabolite and inhibitor of DNApolymerase. The efficacy of this approach was evaluatedin an extended phase I/II study involving 36 prostatecancer patients with local recurrence after radiotherapy.These patients received single or repeated cycles ofreplication-deficient adenoviral mediated HSV-tk plusGCV in situ gene therapy (Miles et al, 2001). The studyconcluded that the repeated cycles of in situ HSV-tk plusGCV gene therapy can safely be administered to patientswith prostate cancer who failed radiotherapy and have alocalized recurrence. The therapeutic parameters such asPSA doubling time (PSADT), the mean PSA reduction(PSAR), and return to initial PSA (TR-PSA) values wereall increased as a response to the treatment, indicating atherapeutic effect. A combined gene therapy approachusing a recombinant adenovirus containing a fusion geneof CD and HSV-tk controlled by a cytomegalovirus(CMV) enhancer-promoter was designed to explore newfrontiers in prostate cancer gene therapy (Lee et al,2002b). Both of the prostate carcinoma cell lines tested(DU-145 or PC-3 cells) were effectively transduced andkilled by this replication-incompetent adenovirus encodingCD-TK fusion protein in the presence of prodrugs. Theeffect of radiation and heat treatment was also tested usingthis vector system. Interestingly, heat treatment not onlyincreased the expression of CD-TK but sensitized prostatecancer cells to radiation as well. These results suggestedthat combining heat treatment with radiation therapyimproved the efficacy of the adenovirus mediated suicidegene therapy approach for prostate carcinoma. The CD-TK fusion fragment was also cloned into a lytic,replication-competent adenovirus (Ad5-CD/TKrep) andadministered into patients with prostate carcinoma in aPhase I trial. This was the first gene therapy study inwhich a replication-competent virus was used to deliver atherapeutic gene to humans (Freytag et al, 2002a). Thisstudy demonstrated that intraprostatic administration ofthe replication-competent Ad5-CD/TKrep virus followedby 2 weeks of 5-fluorocytosine and ganciclovir prodrugtherapy led to the destruction of tumor cells in patientswithout safety concerns. In addition, the efficacy and thetoxicity of replication-competent adenovirus-mediateddouble suicide gene therapy (AdCD-TK) combined withan external beam radiation therapy (EBRT) approach wastested as a trimodal treatment modality in a preclinicalstudy (Freytag et al, 2002b). Animals bearing prostatetumors were first injected with the lytic, replicationcompetentAd5-CD/TKrep virus, then received 1 week of5-fluorocytosine + ganciclovir (GCV) prodrug therapysupplemented with EBRT. The results from this studysuggested that replication-competent adenovirus-mediateddouble suicide gene therapy combined with EBRT is veryeffective in eliminating tumors and reducing metastasis inan orthotropic mouse model of prostate carcinoma.The efficacy of another gene-directed enzyme prodrugtherapy based on the Escherichia coli enzyme purinenucleoside phosphorylase (PNP) was tested in androgenindependentprostate cancer cells. PNP modifies theprodrug fludarabine to 2-fluoroadenine (Voeks et al,2002). In this study, a recombinant ovine adenovirusvector (OAdV220) with a different receptor choice thanthat of human adenovirus type 5 carrying the PNP geneunder the control of RSV promoter was used for functionalstudies. OAdV220 manifested a higher transgeneexpression compared to human Ad5 vector in infectedmurine RM1 prostate cancer cells during in vitro studies.Furthermore, the OAdV220 construct dramaticallyinhibited subcutaneous tumor growth when fludarabinephosphate was administered systemically inimmunocompetent mice. Similar results were obtainedusing human PC3 xenografts in mice. PNP is also knownto convert the prodrug 6MPDR to a toxic purine (6MP)causing cell death. In order to assess the efficacy of thisapproach for prostate cancer, replication-deficient humantype-5 adenovirus (Ad5) carrying the PNP gene (Ad5-SVPb-PNP) was directly injected into PC3 tumors(Martiniello-Wilks et al, 2002). The specificity and thelevel of transgene expression from this recombinantadenoviral vector were controlled by the promoter from115

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomathe androgen-dependent, prostate-specific rat probasin(Pb) gene hooked up to the SV40 enhancer (SVPb).Unexpectedly, the SVPb element confirmed substantialprostate specificity even in the absence of androgens.Intratumoral delivery of Ad5-SVPb-PNP followed by6MPDR administration significantly suppressed thegrowth of human prostate tumors in nude mice. Theseresults suggested that Ad5-SVPb-PNP has therapeuticpotential even in the absence of androgens for thetreatment of prostate carcinoma.Another non-toxic prodrug, CB1954, which isconverted to a toxic metabolite by the Escherichia colinitroreductase gene (NTR), was tested as a suicide genetherapy approach for prostate cancer. Adenovirus vectorexpressing NTR (CTL102) was injected into subcutaneousprostate cancer xenografts followed by systemic CB1954administration (Djeha et al, 2001). A clear anti-tumoreffect of the approach was observed. In addition to all themethods mentioned above, a novel approach inspired fromradioiodine therapy for thyroid cancer was developedusing sodium iodide symporter (NIS). NIS is normallyexclusively expressed in thyroid glands. Adenoviruscarrying the NIS gene (AdCMVNIS) was constructed andtested for the treatment of prostate cancer following 131 Iadministration (Spitzweg et al, 2001). Injection ofAdCMVNIS construct to prostate cancer xenograftsmanifested highly active radioiodine uptake resulting in adrastic reduction in the tumor size following 131 Iadministration in nude mice. This new approachrepresented an effective and potentially curative modalityleading to the accumulation of therapeutically effectiveradioiodine in prostate.Diphtheria toxin (DT) is known to be a potentinhibitor of protein synthesis. The fact that a singlemolecule of DT can result in cell death complicated theutilization of DT as a suicide gene for cancer therapy.Thus, the feasibility of using DT gene therapy wouldgreatly be influenced by tissue specific gene expression.Adenovirus vector carrying the catalytic domain (A chain)of DT under the control of the prostate-specific antigen(PSA) promoter (Ad5PSE-DT-A) induced apoptosis inPSA-positive prostate cancer cells in the presence ofexogenous androgen (R1881) (Li et al, 2002a). In addition,Ad5PSE-DT-A injection regressed the growth of a PSApositiveLNCaP xenograft in nu/nu mice. Non-PSAsecretingDU-145 cells did not manifest the same effectdue to the lack of activation of PSA promoter in thesecells. Therefore, the Ad5PSE-DT-A viral gene therapyapproach might be a viable alternative in the treatment ofPSA-secreting androgen-dependent prostate carcinoma.IV. Joint approaches involvingimmunomodulation-hormonal orradiation therapy in combination withsuicide gene approachAdHSV-tk suicide gene therapy was coupled toadenovirus-mediated IL-12 delivery as a combined genetherapy approach in order to enhance NK activity inducedby HSV-tk gene expression and ganciclovir (GCV)treatment (Hall et al, 2002). This dual treatment generatedradical local and systemic growth suppression in ametastatic model of mouse prostate cancer (RM-1). Theunification of AdHSV-tk/GCV + Ad.mIL-12 gene therapyapproaches resulted in the induction of apoptosis due toincreased expression of Fas and FasL and improved antimetastaticactivity secondary to a strong NK effect. Intratumoralinjection of AdHSV-tk vector followed bysystemic ganciclovir or local radiation therapy or thecombination of gene and radiation therapy wasadministered to subcutaneously transplanted mouseprostate tumors (Chhikara et al, 2001). The combinedtreatment reduced tumor growth by 61% compared to 38%obtained by single therapy modalities. Combined therapyalso increased the mean survival time. In order to analyzesystemic anti-tumor activity, lung metastases weregenerated by tail vein injection of RM-1 prostate cancercells. While radiotherapy alone had no effect on themetastatic growth, the number of lung nodules wasreduced by 37% following treatment with AdHSV-tk. Thecombinational therapy led to an additional 50% reductionin lung colonization. This was the first studydemonstrating a significant systemic effect of AdHSV-tkadministration combined with radiation. A Phase I/II studyof radiotherapy and in situ gene therapy(adenovirus/herpes simplex virus thymidine kinasegene/valacyclovir) in combination with or withouthormonal therapy in the treatment of prostate cancer wasconducted recently (Teh et al, 2001). Based on thepreliminary results, no serious side effect of the combinedtherapy was observed. This was reported as the first trialof its kind in the field of prostate cancer, and is expectedto enlarge the curative index of radiotherapy by merging insitu gene therapy.V. Molecular signaling pathwaysmodulating the efficacy of adenovirusmediated therapeutic gene deliveryExpression of certain hormone and growth factorreceptors as well as cytokines and related downstreammolecules can affect the efficacy of adenovirus-mediatedgene therapy for prostate cancer. For example,gonadotrophin-releasing hormone (GnRH) restrains cellgrowth of reproductive tissue via gonadotrophin-releasinghormone receptors (GnRH-Rs) expressed in most cancersof reproductive tissues like that of prostate. Unfortunately,endogenous GnRH-R expression was not detected in PC3cells, indicating that the cells are insensitive to GnRH.Exogenous expression of high affinity GnRH-R usingadenovirus vectors (AdGnRH-R) facilitatedantiproliferative effects of GnRH agonists in prostatecancer cells (Franklin et al, 2003). In addition, most of theprostate cancer cell lines overexpress fibroblast growthfactors (FGFs). FGF signaling controls cell proliferationand inhibits cell death. A recombinant adenovirusexpressing a dominant-negative FGF receptor(AdDNFGFR-1) was created in order to determine thebiological significance of altered FGF signaling in human116

Gene Therapy and Molecular Biology Vol 7, page 117prostate cancer (Ozen et al, 2001). AdDNFGFR-1infection of LNCaP and DU145 prostate cancer cellsinduced extensive cell death within 48 hours. Some of theprostate cancer cell lines are androgen dependent (LNCaP)whereas some are androgen independent (DU145 or PC3).Androgen ablation therapy, surgery, and radiation therapyare relatively effective in treating androgen dependentprostate carcinoma. However these treatments wereineffective for androgen-insensitive prostate carcinoma.Upregulation of IL6 cytokine induced by the constitutiveNF-κB and Jun D activation is one of the distinctiveparameters of androgen independent cell lines (Giri et al,2001). IL6 is known to function as a proliferation anddifferentiation factor for prostate carcinoma. The infectionwith adenovirus vectors encoding either the dominantnegative form of IκBα gene or Jun D reduced IL6 geneexpression, leading to growth suppression of prostatecancer cells (Zerbini et al, 2003). Some but not all prostatecancer cells respond to vitamin D treatment. 1α, 25-Dihydroxyvitamin D(3) (1α, 25-(OH)(2)D(3)) is known tohave significant antiproliferative effects on certainprostatic carcinoma (PC) cell lines. 1α, 25-(OH)(2)D(3)inhibited cell growth and upregulated p21 expression inPC cell lines such as ALVA-31 and LNCaP (Moffatt et al,2001). Stable transfection with a p21 antisense constructabolished the growth inhibition of ALVA-31 cells withoutaltering vitamin D receptor expression. On the contrary,adenovirus-mediated expression of a sense p21 cDNAsignificantly reduced the proliferation of 1α, 25-(OH)(2)D(3) unresponsive TSU-Pr1 and JCA-1 prostatecancer cell lines. Therefore, Adp21 gene therapy may beuseful even for prostate cancer patients not responding tovitamin D treatment.Molecular signaling pathways are also altered incancer cells. For instance, highly metastatic tumor celllines display increased activity for focal adhesion kinase(FAK). The role of FAK in regulating migration ofprostate carcinoma cell lines with increasing metastaticpotential was studied in detail (Slack et al, 2001). Highlytumorigenic PC3 and DU145 cells displayed intrinsicmigratory capacity correlating with an increased FAKexpression and activity. On the contrary, poorlytumorigenic LNCaP cells required a stimulus to migrate.Inhibiting the FAK/Src signal transduction pathway byoverexpressing FRNK (Focal adhesion kinase-RelatedNon-Kinase), an inhibitor of FAK activation, significantlyinhibited migration of prostate carcinoma cells.Modulation of phosphatidylinositol 3'-kinase (PI3'-kinase),leading to Akt activation, frequently occurs in prostatecancer and disrupts apoptotic signaling induced by variouscytokines such as tumor necrosis factor TNF and TNFrelatedapoptosis-inducing ligand (TRAIL). Two prostatecancer cell lines with constitutively activated PI3'-kinasecascades (LNCaP and PC-3) were examined in order tostudy the role of PI3' phosphorylation in cellular responseto TNF or TRAIL alone. Both TNF and TRAIL failed toactivate apoptosis in either LNCaP or PC-3 cells.Interestingly, downregulation of PI3'-kinase/Akt signalingsignificantly enhanced the apoptotic activity of both TNFand TRAIL in LNCaP cells but not in PC-3 cells. Infectionwith adenovirus delivered PTEN/MMAC1 (phosphataseand tensin homologue/mutated in multiple advancedcancers) reduced Akt activation, activated apoptosis andsensitized cells to TNF but not to TRAIL in LNCaP cellline (Beresford et al, 2001). Therefore, it was concludedthat although PI3'-kinase signaling inhibits both TNF andTRAIL mediated apoptosis, this may only represent one ofthe several apoptotic resistance mechanisms in signalingpathways.Selenium compounds are known to be potentialchemotherapeutic agents for prostate cancer. NF-κB hasbeen categorized as the key antiapoptotic signalingmolecule often activated in transformed cells. Testing ofselenium compounds on DU145 and JCA1 prostatecarcinoma cells revealed that these compounds inducedapoptosis through the inhibition of NF-κB pathways inthese cell lines (Gasparian et al, 2002b). Increased IKKactivity was blamed for constitutive NF-κB activationresponsible for survival of androgen independent prostatecarcinoma cell lines (Gasparian et al, 2002a).60-80 % of prostate cancers acquire the PTENmutation during tumorigenesis. This results in theconstitutive activation of the PI3'-kinase pathway andprostatic cell proliferation. The loss of PTEN activity isalso correlated with the loss of activity of the FOXOfamily of forkhead transcription factors such as FKHRL1and FKHR. Interestingly, these transcription factors areshown to control the expression of apoptosis inducingligand TRAIL. Not surprisingly, the expression of TRAILwas also reduced in PTEN-lacking prostate cancer cells,leading to decreased apoptosis. Restoration of TRAILexpression using adenovirus-mediated overexpression ofthese transcription factors in LAPC4 prostate cancer cellline induced apoptosis (Modur et al, 2002).VI. Apoptosis ModulatorsA. The exploitation of death ligands toinduce apoptosis in cancer cellsApoptosis, known as programmed cell death (Reed,2000) is defined as cell’s preferred form of death underhectic conditions (Sears and Nevins, 2002). In reality, it isalso a key mechanism for homeostasis throughoutembryonic and adult life. Genetic aberrations disruptingprogrammed cell death underpin tumorigenesis and drugresistance. Therefore, the specific activation of apoptosiswithin tumor cells could be a highly effective therapeuticintervention for prostate cancer. Currently, chemotherapy(Stein et al, 2002) and radiotherapy (Wang et al, 2002) areamong the most commonly used treatment modalitiesagainst prostate cancer. The tumor suppressor gene, p53, isrequired in order for both of these treatment methods towork as anti-tumor agents (Levine, 1997). However, morethan half of the human tumors acquire p53 mutationsduring tumorigenesis (Horowitz, 1999; Zeimet et al,2000). As a result, tumors lacking p53 display resistanceto both chemotherapy and radiotherapy (Obata et al,2000). Intriguingly, death ligands induce apoptosisindependent of p53 status of the cells (Ehlert andKubbutat, 2001; Norris et al, 2001). Thus, these methodsconstitute somewhat of a complementary treatmentmodality to currently employed conventional treatments.117

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaAt present, death ligands are being evaluated as potentialcancer therapeutic agents (Herr and Debatin, 2001).Previously, several studies using external Fas agonists,anti-Fas antibodies and membrane-bound FasL failed toinduce Fas L mediated apoptosis in prostate cancer cells.Although the down regulation of c-FLIP expressionthrough the use of anti-sense oligonucleotides sensitizedDU145 cells to an anti-Fas monoclonal antibody (Hyer etal, 2002), efficient cell killing was not observed by thisapproach. However, intracellular expression of FasL usingadenoviruses efficiently killed 70-90% of various humanprostate cancer cell lines tested (Hyer et al, 2000).Furthermore, part of this cell killing was attributed to thebystander effect mediated by FasL carried within theapoptotic bodies and cellular debris (Hyer et al, 2003).Despite the fact that human prostate cancer cells expressapoptotic FasL, some of the cell lines, such as LNCaP, areresistant to Fas L mediated cell death. Even so, priorexposure to IFNγ sensitized orthotropic prostate primarytumors to recombinant adenovirus mediated FasL delivery(Selleck et al, 2003). Despite the fact that tumor necrosisfactor (TNF) (Terlikowski, 2001) and FasL (Nagata, 1997)have been studied extensively and were shown toeffectively induce apoptosis in cancer cells, their systemicuse in cancer gene therapy is not recommended due to thesystemic toxicity.With the discovery of a novel death ligand,TRAIL/Apo2L, (Wiley et al, 1995; Pitti et al, 1996) a newera emerged for the deployment of death ligands forcancer gene therapy (Nagane et al, 2001). The fact thatTRAIL does not cause any harm to normal cells but canselectively induce apoptosis in cancer cells brought up thepossibility of TRAIL testing for systemic use (Griffith andLynch, 1998). Five different receptors were identified tointeract with TRAIL; TRAIL-R1, TRAIL-R2, TRAIL-R3,TRAIL-R4 and osteoprotegrin (Abe et al, 2000; Sheikhand Fornace, 2000). TRAIL-R1 and TRAIL-R2 functionas authentic death receptors inducing apoptosis whileTRAIL-R3 and TRAIL-R4 are unable to induce suchsignaling but can serve as decoy receptors (Meng et al,2000). However even today, no single mechanism hasbeen found to account for TRAIL resistance observed innormal cells. The soluble form of TRAIL has successfullybeen tested and no toxicity due to systemic use wasobserved in animal models. However, large quantities ofTRAIL were needed in order to suppress the tumorgrowth. A replication-deficient adenovirus encodinghuman TRAIL (TNFSF10; Ad5-TRAIL) was generated asan alternative to recombinant, soluble TRAIL protein(Griffith and Broghammer, 2001). Ad5-TRAIL infectioninto TRAIL-sensitive prostate tumor cells inducedapoptosis through the activation of Caspase 8 pathways.Normal prostate epithelial cells were not harmed by Ad5-TRAIL infection. Moreover, in vivo Ad5-TRAILadministration suppressed the outgrowth of humanprostate tumor xenografts in SCID mice. Eight prostatecancer cell lines (CWR22Rv1, Du145, DuPro, JCA-1,LNCaP, PC-3, PPC-1, and TsuPr1) and primary culturesof normal prostate epithelial cells (PrEC) were tested forsensitivity to soluble TRAIL induced cell death in anotherstudy (Voelkel-Johnson et al, 2002). 100 ng/mL of solubleTRAIL administration did not induce apoptosis in Du145,DuPro, LNCaP, TsuPr1, and PrEC. Interestingly,treatment with the chemotherapeutic agent doxorubicinsensitized almost all prostate cancer cells to TRAILinducedcell death. On the other hand, an adenoviral vectorexpressing full-length TRAIL (AdTRAIL-IRES-GFP)killed prostate cancer cell lines and, unexpectedly, PrECas well, independent of doxorubicin cotreatment. Thisstudy suggested that the AdTRAIL-IRES-GFP genetherapy approach, complemented with tissue-specificpromoters, would be useful for the treatment of prostatecarcinoma. However, the mechanism of TRAIL resistancein normal cells is not understood and some prostate cancercells appeared to be TRAIL-resistant (Nesterov et al,2001). In one study, ALVA-31, PC-3, and DU 145 celllines were highly sensitive to apoptosis induced byTRAIL, while TSU-Pr1 and JCA-1 cell lines weremoderately sensitive, and the LNCaP cell line wasresistant (Nesterov et al, 2001). Due to the lack of activelipid phosphatase PTEN, LNCaP cells demonstrated aconstitutive Akt activity. Akt is a negative regulator of thephosphatidylinositol (PI)3-kinase/Akt pathway. PI3-kinaseinhibitors sensitized LNCaP prostate cancer cells toTRAIL. In addition, adenovirus expressing a constitutivelyactive Akt reversed the ability of wortmannin to potentiateTRAIL-induced BID cleavage. This suggested thatconstitutive Akt activity inhibits TRAIL-mediatedapoptosis (Nesterov et al, 2001).B. NF-κB inhibiting approaches used tobreakdown TRAIL resistance in prostatecancer cellsThe mechanism of TRAIL induced apoptosis andresistance is outlined in Figure 1. So far, at least twodifferent hypotheses that may partly explain TRAILresistance are asserted. The first hypothesis advocates thatnormal cells carry decoy receptors (TRAIL-R3, TRAIL-R4), which compete with apoptosis inducing TRAILreceptors (TRAIL-R1, TRAIL-R2) for binding to TRAIL(Pan et al, 1997; Sheridan et al, 1997). In this hypothesis,it is believed that decoy receptors either function to diluteout TRAIL ligands (like TRAIL-R3) or supply antiapoptoticsignals (like TRAIL-R4) to cells. As reportedpreviously, TRAIL-R4 binding activates the anti-apoptoticNF-κB signaling pathway, leading to the blockade ofTRAIL induced apoptosis (Degli-Esposti et al, 1997). Inaddition, the expression of decoy receptors is downregulatedin cancer cells through promoterhypermethylation leading to differential sensitivity toTRAIL (van Noesel et al, 2002). However, the linkbetween TRAIL resistance and the expression of decoyreceptors has not been clearly established in human cells(Griffith and Lynch, 1998). Interestingly, activation ofdeath receptors such as TRAIL-R1 and TRAIL-R2 alsostimulated the NF-κB pathway (Chaudhary et al, 1997;Schneider et al, 1997). Under these circumstances, thereason(s) for cells undergoing apoptosis despite theinduction of anti-apoptotic pathways through the samedeath receptors is not fully understood.118

Gene Therapy and Molecular Biology Vol 7, page 119Figure 1: A gene therapy strategy to block anti-apoptotic NF-κB signaling pathway to induce TRAIL sensitivity in prostate cancer cells.Activation of TRAIL receptor 1 (R1) or 2 (R2) by trimeric TRAIL ligands leads to the recruitment of Fas associated death domainprotein (FADD) to the membrane. Then, FADD recruits procaspase 8 to form death inducing signaling complex (DISC). DISC inducedsignaling activates caspase pathway inducing cells into apoptosis. TRAIL receptor 3 (R1) and 4 (R4) serve as decoy receptors. R4activates NF-κB signaling pathways as well. In addition, NF-κB pathway is also activated by R1 and R2 via TNFR-associated deathdomain protein (TRADD) and receptor interacting protein (RIP). Consequently, NF-κB activation augments expressions of various antiapoptoticgenes such as cIAP, BclxL and cFlip in addition to R3. c-Flip, a procaspase 8 homologue, competes with procaspase 8 forbinding to FADD. Thereby it inhibits apoptotic signaling. The expression of adenovirus delivered IKKβKA mutant prevented theactivation of anti-apoptotic NF-κB signaling. This method sensitized prostate cancer cells to TRAIL.The second hypothesis claims the presence of apoptosisinhibitory substances in these cells. Such a molecule,cFLIP (FLICE Inhibitory Protein), a caspase 8 homologue,has been shown to obstruct death ligand induced apoptosis(Irmler et al, 1997; Griffith et al, 1998). Intriguingly, NFκBactivating agents up-regulated cFLIP synthesis (Kreuzet al, 2001). Furthermore, the NF-κB pathway has beenproven to increase TRAIL-R3 synthesis, a decoy receptorfor TRAIL, (Bernard et al, 2001) and the expression ofapoptosis inhibitor Bcl-xL (Hatano and Brenner, 2001;Ravi et al, 2001) resulting in the obstruction of TRAILmediated apoptosis. Apoptosis inhibitors such as cIAP arealso activated by NF-κB pathways (Mitsiades et al, 2002).Based on these results, we can clearly state that the activeNF-κB signaling pathway may provide cells with TRAILresistance by at least four different ways (Figure 1).Additionally, it has been reported that a novel tumorsuppressor gene, PTEN/MMAC1 (Steck et al, 1997;Simpson and Parsons, 2001) negatively regulated TNFinduced NF-κB activity (Ozes et al, 1999; Mayo et al,2002) through the IKK complex (Gustin et al, 2001). Theobservation in which IKK activity is required for PI3K-Akt induced NF-κB activation (Burow et al, 2000;Demarchi et al, 2001) confirmed this report (Madrid et al,2001; Sizemore et al, 2002). Due to a negative correlationbetween the expression of PTEN and the progression ofprostate cancer, advanced prostate cancer cells might haveintrinsically higher NF-κB activity due to the progressiveloss of PTEN. Absence of PTEN function may result inincreased Akt activity induced by PI3K. Since NF-κB is adownstream target for Akt, (Kane et al, 1999;Romashkova and Makarov, 1999; Andjelic et al, 2000;Jones et al, 2000) TRAIL resistance would ultimately beensured in cells by way of the NF-κB pathway. Inagreement with this hypothesis, PTEN sensitized prostatecancer cells to TRAIL induced apoptosis (Yuan andWhang, 2002). Thus, these possible scenarios make NFκBinhibiting vectors such as Ad.IKKβKA (Sanlioglu etal, 2001a) or Ad.IκBαSR (Batra et al, 1999; Sanlioglu andEngelhardt, 1999) ideal candidates for overcoming theTRAIL resistance in PTEN mutant prostate cancer cells. Ina similar manner, TNF induced apoptosis can also beprevented by NF-κB activation as reported (Beg andBaltimore, 1996; Van Antwerp et al, 1996). Previously,NF-κB inhibiting approaches such as adenovirus mediatedtransfer of IKKβ (Ad.IKKβKA) (Sanlioglu et al, 2001a,2001b) or IκBα (Ad.IκBαSR) (Batra et al, 1999;Sanlioglu and Engelhardt, 1999) dominant negativemutants were successfully deployed in order to sensitizelung cancer cells to TNF. Since some tumor cells haveintrinsically high NF-κB activity, which might beresponsible for TRAIL resistance, NF-κB blocking agentscan potentially be useful to overcome TRAIL resistance.For example, a constitutive NF-κB activation wasobserved in renal carcinoma (Oya et al, 2001). Notsurprisingly, melanoma cells having a constitutive NF-κB119

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaactivity exhibit TRAIL resistance (Franco et al, 2001).Resistant melanoma cells were sensitized to TRAIL eitherwith proteasome inhibitors or transfections with plasmidsencoding degradation resistant IκBα protein (Franco et al,2001). In accordance with these studies, we have tested ifadenovirus mediated NF-κB inhibiting approach wouldsensitize prostate cancer cells to TRAIL. Consequently,adenovirus mediated delivery of IKKβKA mutant(Ad.IKKβKA) sensitized PTEN mutant prostate cancercells (PC3) to TRAIL as shown in Figure 2. At first, PC3cells appeared to be relatively resistant to pro-apoptoticeffects of TRAIL when cells were infected withadenovirus vector encoding hTRAIL (Ad.hTRAIL) evenat an MOI of 1000 DNA particles/cell (Figure 2 Panel A).Infection with Ad.IKKβKA vector alone did not yield anycell death either (Figure 2, Panel B). However, when thedose of Ad.hTRAIL vector was kept constant at an MOI of1000 DNA particles/cell, increasing the amount ofAd.IKKβKA construct sensitized PC3 cells to TRAILmediated apoptosis (Figure 2, Panel C).C. Intracellular proapoptotic regulatorsAlthough caspases are the effector mediators of apoptosis,the expression of proapoptotic molecules such asprocaspase 3 or 7 using adenovirus constructs did notinduce apoptosis in prostate cancer cells due to theinability of these caspases to undergo autocatalyticactivation (Li et al, 2001). A novel suicide gene therapyapproach was developed using chemically inducibleeffector caspases to trigger apoptosis in prostate cancercells. Cell death was mediated by replication-deficientadenoviral vector expressing conditional caspase-1 (Ad-G/iCasp1) or caspase-3 (Ad-G/iCasp3) and the caspaseactivation was achieved by nontoxic, lipid-permeable,chemical inducers of dimerization (CID) (Shariat et al,2001). Aggregation and activation of these recombinantcaspases occurred, leading to rapid apoptosis only aftervector transduction followed by CID administration inboth human (LNCaP and PC-3) and murine (TRAMP-C2and TRAMP-C2G) prostate cancer cell lines.Subcutaneous TRAMP-C2 tumors displayed focal butextensive apoptosis following direct injection of Ad-G/iCasp1 in vivo. In order to express caspase 9exclusively in prostate, a recombinant adenovirus carryingiCaspase-9 was constructed with two copies of theandrogen response region (ARR) placed upstream of theprobasin promoter elements (ADV.ARR(2)PB-iCasp9)(Xie et al, 2001b). AP20187 is a chemical dimeric ligand,which causes dimerization and thereby activation ofiCaspase-9 leading to rapid apoptosis in both dividing andnondividing cells. Testing of ADV.ARR(2)PB-iCasp9construct in LNCaP tumor xenografts demonstrated thatthis construct induces apoptosis in prostate cancer cellsonly in the presence of AP20187.The proapoptotic members of Bcl- 2 protein familyincluding Bax, Bak, Bad, and Bik also mediate apoptosis.Apoptosis-inducing proteins were cloned into adenovirusconstructs and shown to induce apoptosis in prostatecancer cell lines previously.Figure 2. Adenovirus mediated IKKβKA expression sensitized PC3 cells to TRAIL mediated apoptosis. PC3 cells were infected withincreasing MOIs of either Ad5hTRAIL (Panel A) or Ad.IKKβKA (Panel B). In panel C, the dose of Ad.IKKβKA vector was increasedgradually (stated just above each panel) while the amount of Ad5hTRAIL was kept constant (as indicated under the panel). Cell deathwas detected using molecular probe’s Live and Death Cellular viability and toxicity kit 48 hours following infection. Numbers indicateviral doses as MOI values of DNA particles/cell.120

Gene Therapy and Molecular Biology Vol 7, page 121However, overexpression of proapoptotic genes withoutthe use of tissue specific promoters could result inunwanted apoptosis even in normal cells. In order toprovide tissue specificity, an adenoviral construct wasgenerated containing Bax cDNA under control of theprobasin promoter that included two androgen responseelements (Av-ARR2PB-Bax). Av-ARR2PB-Bax constructdrove Bax overexpression in an androgen-dependent wayin androgen receptor (AR)-positive cell lines of prostaticorigin but not in others. The androgen dihydrotestosteroneactivated apoptosis in LNCaP cells infected with Av-ARR2PB-Bax but not in those infected with controlvectors. These results demonstrated that Av-ARR2PB-Baxinduced apoptosis was androgen dependent and limited toAR positive cells of prostatic epithelium. On the otherhand, using a binary co-transfection strategy involvingAd/GT Bax and Ad/PGK-GV16; overexpression ofproapoptotic Bax protein induced apoptosis both inandrogen-insensitive (DU145 and PC3), and androgensensitive(LNCaP) cell lines (Honda et al, 2002). The samebinary approach was tested to assess the consequences ofBcl-2 overexpression in the progression of prostatecarcinoma leading to apoptosis-resistant and androgenindependentphenotype in DU145, PC3 and LNCaP celllines which represent models of advanced prostatecarcinoma. Bax expression generated by the adenoviralco-transfection system induced apoptosis even in theseBcl-2 overexpressing cell lines. These results suggest thatthe Ad/GT Bax and Ad/PGK-GV16 combined expressionsystem might represent a powerful gene therapy strategyfor the treatment of androgen-independent and apoptosisresistantprostate carcinoma. Moreover, monogene andpolygene approaches were compared in an experimentalprostate cancer model using apoptotic genes bad and baxdriven by a prostate specific promoter (ARR(2)PB) in anadenovirus construct (Zhang et al, 2002b). The ARR(2)PBis a dihydrotestosterone (DHT)-inducible third-generationprobasin-derived promoter. In this study, animals bearingtumors of prostatic origin responded better to combinedbad and bax therapy than either of the vectors alone.Therefore, it was concluded that polygene therapyinvolving more than one apoptotic molecule is moreeffective in xenograft models of androgen-dependent orindependent prostate cancer than monogene therapy alone.It is also known that overexpression of anti-apoptoticgenes such as Bcl-2 in prostate carcinoma providesresistance to radiation therapy and androgen ablation. Asecond-generation adenoviral vector (ARR2PB.Bax.GFP)was constructed with the modified prostate-specificprobasin promoter (ARR2PB) directing the expression of aHA-tagged Bax gene in order to restore the balance ofBcl-2 family members to induce apoptosis in prostatecancer cells (Lowe et al, 2001). ARR2PB.Bax.GFP vectorinduced significant levels of apoptosis in LNCaP cells 48hours following infection even in the presence of highlevels of Bcl-2 protein. No toxicity in liver, lung, kidney,and spleen was detected by systemic administration ofARR2PB.Bax.GFP in nude mice. Therefore, a secondgenerationadenovirus-mediated, prostate-specific Baxgene therapy appeared to be a very safe and efficientapproach for the treatment of prostate cancer. Anothermember of the proapoptotic Bcl-2 family, namely "Bik",was cloned into adenovirus vectors to explore itstherapeutic potential. AdBik infection also inducedapoptosis and suppressed the growth of PC-3 xenograftsestablished in nude mice (Tong et al, 2001).Several other genes were also tested for their abilityto induce apoptosis in prostate tumor cell lines as well asin xenograft models. The antiapoptotic protein CLN3negatively regulates endogenous ceramide production, aninducer of apoptotic cell death. CLN3 protein isoverexpressed in most of the cancer cell lines testedincluding those of prostate (Du145, PC-3, and LNCaP).An adenovirus-expressing antisense CLN3 (Ad-AS-CLN3) blocked CLN3 protein expression in prostatecancer cell lines as demonstrated by Western Blotting(Rylova et al, 2002). Ad-AS-CLN3 infection resulted inthe inhibition of cell growth and reduction in cell viabilityof cancer cells through elevation of endogenous ceramideproduction. This study revealed CLN3 as a novel target toinduce apoptosis in prostate cancer cells. A recombinantadenovirus containing pHyde cDNA gene (AdpHyde), anovel gene cloned from Dunning rat prostate cancer cells,was constructed in order to study its function (Zhang et al,2001). Surprisingly, the AdpHyde construct inhibited thegrowth of human prostate cancer cells and inducedapoptosis involving the caspase-3 pathway in humanprostate cancer tumor xenografts in nude mice. Ionicmovement also influences apoptosis. For instance, K +efflux is an early event in apoptosis, which is regulated byK + channel-associated protein (KChAP). A recombinantadenovirus encoding KChAP (Ad/KChAP) wasconstructed in order to determine if KChAP expressioncould induce apoptosis in prostate cancer cells (Wible etal, 2002). The LNCaP cell line displayed a reduction incell size upon infection with Ad/KChAP. The Ad/KChAPconstruct also induced apoptosis in DU145 cells in a p53independent manner. In addition, infection withAd/KChAP prevented growth of DU145 and LNCaPtumor xenografts in nude mice.VII. Tumor suppressor genesAberrations in the expression of tumor suppressorgenes have been one of the key factors affecting theoutcome of cancer therapy. Several studies examined thepossible use of tumor suppressor genes as therapeuticagents for prostate cancer. Doxorubicin (Dx) is acommonly used chemotherapeutic agent in recurrentprostate cancer and is a strong inducer of p53 expressionleading to p21(CIP1/WAF1) transactivation. As suggestedby previous reports, p21 plays a role in the modulation ofchemotherapy-induced apoptosis, prostate cancerprogression and androgen regulation. Two androgenregulatedhuman prostate cancer cell lines (MDA PCa 2band LNCaP) were exposed to Dx and growth factorwithdrawal in order to investigate if p21 plays a role in thesurvival of prostate cancer cells under stress (Martinez etal, 2002). Infection with adenovirus vectors encoding theantisense strand of p21 reduced p21 levels, sensitizedprostate cancer cells to Dx and facilitated apoptosis inresponse to growth factor withdrawal. These resultssuggest that modulation of p21 pro-survival gene121

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaexpression via adenovirus constructs sensitizes prostatecancer cells to chemotherapeutics and androgenwithdrawal. Another tumor suppressor protein, p27, alsoknown as cyclin-dependent kinase inhibitor (CDKI), isnormally expressed in human prostate. However, themajority of human prostate cancers have reduced levels ofp27. The down regulation of this putative tumorsuppressor gene through proteolysis is mediated bySCFSKP2 ubiquitin ligase complex. Adenovirus-mediatedoverexpression of SKP2 induced ectopic down-regulationof p27 in LNCaP prostate carcinoma cells (Lu et al, 2002).This observation confirmed that SKP2 activity was themajor determinant of p27 levels in human prostate cancercells. Based on in vitro studies, it is believed that theoverexpression of SKP2 might be one of the mechanismsallowing prostate cancer cells to escape growth controlmediated by p27. Therefore, knocking out SKP2 functionwould be a logical novel approach to fight prostate cancer.In another study, an adenovirus construct carrying p27coding sequences Adp27(Kip1) was generated to assesswhether the overexpression of p27 has any affect on theprostatic tumor growth in vivo (Katner et al, 2002).Injection of Adp27(Kip1) vector reduced the growth ofLNCaP tumor xenografts in mice. This study supportedthe idea that Adp27(Kip1) can serve as a potentialtherapeutic vector for the treatment of prostate carcinoma.p14(ARF), encoded by the human INK4a gene locus,is another tumor suppressor protein which is frequentlyinactivated in human cancer. p14(ARF) has recently beenimplicated in p53-independent cell cycle regulation andapoptosis. A replication-deficient adenoviral constructcarrying p14(ARF) coding sequence (Ad-p14(ARF)) wasgenerated in order to explore the pro-apoptotic function ofp14(ARF) in relationship to p53 function (Hemmati et al,2002). Ad-p14(ARF) construct induced apoptosis inp53/Bax-mutated DU145 prostate cancer cells andHCT116 cells lacking functional Bax expression. Thisstudy demonstrated that overexpression of p14 throughadenovirus vectors is sufficient to induce apoptosis in p53-and bax-deficient prostate cancer cells. Prostate carcinomawith p53 mutant phenotype represents a clear obstacle forirradiation therapy. Ionizing radiation (IR) and adenoviralp53 gene therapy (Ad5CMV-p53) were utilizedindividually as well as in combination in order to assessthe effectiveness of combined therapy for prostate cancer(Sasaki et al, 2001). In this study, IR alone did not inducesignificant levels of apoptotic cell death in DU145 andPC-3 cells. However, after combined therapy, theproportion of apoptotic cells was greatly amplified in bothof the cell lines tested. Therefore, it was concluded that theobserved synergistic effect might be useful for thetreatment of radio-resistant prostate carcinoma.The loss of MMAC/PTEN tumor suppressor geneexpression is frequently detected in human tumors.Survival signaling through the phosphatidylinositol-3kinase/Akt pathway is constitutively activated in cellslacking functional PTEN expression. Therefore, thefunctional effect of MMAC/PTEN expression wasexamined in LNCaP cells, which are devoid of afunctional PTEN product (Davies et al, 1999). Infectionwith an adenovirus construct driving the expression ofMMAC/PTEN resulted in a specific inhibition of Akt/PKBactivation. This is consistent with the phosphatidylinositolphosphatase activity of MMAC/PTEN. Compared toadenovirus delivered p53 expression, MMAC/PTENexpression induced apoptosis in LNCaP cells to a lesserextent. Interestingly, the growth suppression properties ofMMAC/PTEN were significantly greater than thoseaccomplished with p53. Moreover, Bcl-2 overexpressionin LNCaP cells blocked both the adenovirus mediatedMMAC/PTEN- and p53-induced apoptosis, but it did notaffect the growth-suppressive properties of MMAC/PTEN. This is consistent with the fact that MMAC/PTENmay play multiple roles in the cell. Prostate cells wereinfected with adenovirus vector carrying PTEN codingsequence in order to determine if supplying PTENfunction would sensitize these cells to various apoptoticstimuli (Yuan and Whang, 2002). As predicted,adenovirus-mediated PTEN delivery sensitized LNCaPprostate cancer cells to apoptosis through the inhibition ofconstitutive Akt activation. Since PTEN G129E mutantlacking lipid phosphatase activity was unable to sensitizecells to apoptosis, it was concluded that the lipidphosphatase activity of PTEN was required for apoptosis.The therapeutic effect of adenoviral delivery ofMMAC/PTEN was tested on both the in vitro and in vivogrowth of PC3 human prostate cancer cells (Davies et al,2002). The in vitro growth of PC3 cells was repressed byadenovirus expression of MMAC/PTEN via blocking ofcell cycle progression. Although this approach did notinhibit the tumor progression of orthotopically implantedPC3 cells, a significant reduction was observed in thetumor size in vivo, in addition to complete inhibition ofmetastases. Therefore, it was suggested thatMMAC/PTEN might play a role mostly in the regulationof the metastatic potential of prostate cancer.A considerable fraction of prostate tumors display analteration of Mxi1 expression, an antagonist to c-Myc.This was confirmed by transgenic approaches in whichprostatic hyperplasia was observed in mice deficient forMxi1. Mxi1-expressing adenovirus (AdMxi1) wasgenerated to study the ability of Mxi1 to act as a growthsuppressor in prostate tumor cells (Taj et al, 2001).Overexpression of Mxi1 using adenovirus vectors in theDU145 prostate carcinoma cell line resulted in growtharrest and decreased colony formation on soft agar. Allthese studies emphasize that the modulation of tumorsuppressor gene function might be necessary for anoptimum therapeutic response to fight against prostatecancer.VIII. Cell adhesion molecules and antiangiogenicapproachesCell adhesion molecules play major roles especiallyin metastasis of cancer cells. Therefore, aberrantexpression patterns of cell adhesion molecules arefrequently associated with poor prognosis. For instance,the expression of a well-known cell adhesion molecule, C-CAM1, is downregulated during the early stages ofprostate carcinoma in an animal model (TRAMP) (Pu etal, 1999). C-CAM1 was cloned into an adenovirus122

Gene Therapy and Molecular Biology Vol 7, page 123construct and its efficacy was tested both in vitro and invivo using PC3 xenograft murine model (Lin et al, 1999).AdC-CAM1 construct manifested a strong antitumoralactivity on PC3 tumor cells grown in nude mice.Therefore, selective use of cell adhesion molecules mightbe beneficial for the treatment of prostate carcinoma.Moreover, combining C-CAM1-based therapy with TNP-470, a potent angiogenesis inhibitor, induced greatergrowth suppression on DU145 tumor xenografts than byeither Ad-C-CAM1 or TNP-470 application alone (Pu etal, 2002).Vascularization of a solid tumor is required forcancer growth. Recently, preventing vascularizationthrough inhibition of angiogenesis was a popular target forcancer gene therapy. For example, a 16-kDa prolactinprotein (PRL) has previously been shown to possess anantiangiogenic activity (Galfione et al, 2003). Notsurprisingly, adenovirus delivery of PRL proteinmanifested a significant antitumoral activity in vivo (Kimet al, 2003). In addition, vascular endothelial growth factor(VEGF) receptor signaling is another relevant pathway,which modulates the vascularization of newly growingtumors. Interfering with such a signaling pathway mightbe valuable in controlling the tumor growth. In fact, whenfused to an Fc domain and cloned into the recombinantadenovirus construct, the ligand-binding ectodomain ofVEGF receptor 2 (Flk1) manifested a considerablereduction in tumor growth induced by a drastic decline inthe microvessel density in SCID mice carrying humanLNCaP xenografts (Becker et al, 2002).Growth factors are needed for survival of cancer cellsand molecular chaperones are required for functionalproduction of these molecules. A new member of the heatshock protein family functioning as a molecular chaperonein the endoplasmic reticulum was recently discovered andnamed as 150-kDa oxygen-regulated protein (ORP150).Since prostate cancer cells exhibited an upregulation ofORP150 protein and VEGF, adenovirus delivery of anantisense ORP150 cDNA approach was used to reduceangiogenicity and tumorigenicity through inhibition ofVEGF secretion. This approach indeed suppressed thegrowth of DU145 prostate carcinoma cell line in axenograft model (Miyagi et al, 2002).IX. Replication competent adenovirusvectorsReplication competent adenoviral vectors providepowerful means to kill cancer cells through cell lysis.Since they only replicate in tumor cells, the therapeuticrange is limited to cancer cells. Two replication-competentadenoviruses, CV706 and CV787, were generated in orderto selectively destroy PSA producing prostate cancer cells.It has been demonstrated earlier that prostate-specificantigen (PSA)-selective replication-competent adenovirusvariant CV706 specifically eliminated tumors in humanprostate cancer xenografts in preclinical models(Rodriguez et al, 1997). Since adenovirus E1A is known tobe a potent inducer of chemosensitivity andradiosensitivity through p53-dependent and independentmechanisms, the potential radiosensitizing effects ofCV706 on prostate cancer cells were evaluated (Chen etal, 2001). The CV706 construct demonstrated a synergisticantitumoral effect both on irradiated human prostatecancer cells and tumor xenografts. Moreover, in order toinvestigate the safety and the functionality of intraprostaticdelivery of CV706 for the treatment of patients withlocally recurrent prostate cancer following radiationtherapy, a Phase I dose-escalation study was conducted(DeWeese et al, 2001). Results from this study suggestedthat even at high doses, intraprostatic delivery of theCV706 was relatively safe for patients and CV706construct demonstrated high therapeutic activity asreflected by the reduction in serum PSA. This was the firstclinical trial of a prostate-specific, replication-restrictedadenovirus for the treatment of prostate cancer. Anotherprostate-specific replication-competent adenoviruscarrying not one, but two, cell type specific promoters(CV787) was constructed. This construct contained E1Bgene driven by the human prostate-specificenhancer/promoter and the adenovirus type 5 (Ad5) theE1A gene under the control of prostate-specific ratprobasin promoter. The Ad5 E3 region was also conservedin the vector to improve the efficacy. A single tail veininjection of CV787 eliminated LNCaP xenografts within 4weeks in nude mice (Yu et al, 1999). When the prostatecancer-specific adenovirus CV787 was combined withchemotherapeutic agents like taxanes (paclitaxel anddocetaxel), a synergistic antitumoral effect was observedin mice carrying human prostate cancer xenografts (Yu etal, 2001b).Heat-inducible gene expression is another approachused in the context of suicide gene therapy. A recombinantadenovirus containing the CD-TK fusion gene controlledby the human inducible heat shock protein 70 promoter(Ad.HS-CDTK) was generated for this purpose. Heatapplication at 41 o C for 1 hour induced therapeutic geneexpression from this vector. Despite the fact that theAd.HS-CDTK construct induced CD-TK expression inhuman prostate cancer cells, a therapeutic benefit was notobserved due to lower transduction efficiency of tumors invivo. Instead, a replication-competent, E1B-attenuatedadenoviral vector containing the hsp70 promoter-drivenCD-TK gene (Ad.E1A + HS-CDTK) was generated toincrease CD-TK gene expression to achieve a therapeuticeffect (Lee et al, 2001). Contrary to replicationincompetent Ad.HS-CDTK, replication competentAd.E1A + HS-CDTK construct yielded severe cytotoxicityand greater levels of therapeutic index in the presence ofprodrugs. This approach revealed the beneficial effects ofusing replication competent virus complemented with aheat inducible suicide gene therapy approach for prostatecarcinoma.X. Adenovirus vectors with cell typespecific and inducible promotersEven though adenovirus-mediated HSVTK suicidegene therapy approach manifested a satisfactory toxicityprofile in Phase I clinical trials, the toxicity studies usingadenovirus vectors were very restricted in numbers.123

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaHowever, it was known that the promoter of choice mightinfluence the level of toxicity. In order to study thepromoter effect on adenovirus mediated toxicity the mousecaveolin 1 promoter was cloned into the adenovirus HSVtkvector (Adcav-1tk) because this promoter was highlyactive in metastatic and androgen-resistant prostate cancercells (Pramudji et al, 2001). The efficacy of this vector forsuicide gene therapy was compared to those of AdHSV-tkvectors carrying either cytomegalovirus (AdCMV-tk) orrous sarcoma virus (AdRSV-tk) promoters in micetransplanted with mouse prostate cancer cells. FollowingGCV administration, all the HSV-tk expressing vectorsregressed the tumor growth in situ. Interestingly, theefficacy of Adcav-1tk vector was much greater in terms ofinducing necrosis and microvessel density. In order toevaluate the toxicity profile of adenovirus vectors carryingCMV, RSV or mouse caveolin promoter-driven HSV-tktransgenes, these vectors were also injected systemicallyinto mice (Ebara et al, 2002). Adenovirus vectors withCMV and RSV promoters, but not caveolin promoter,exhibited significant levels of liver damage. These resultssuggested that the promoter selection greatly influencesthe toxicity profile of adenovirus-mediated suicide genetherapy approach. In order to increase the number ofpromoters available for prostate specific gene expression,transgenic mice were generated expressing a reporter gene(SV40 Tag) directed by prostate secretory protein of 94amino acids (PSP94) (Gabril et al, 2002). PSP94 genepromoter/enhancer region directed SV40 Tag expressionexclusively in prostate leading to prostatic intraepithelialneoplasia and eventually to high-grade prostate carcinoma.These studies suggested that this PSP94 genepromoter/enhancer strategy could be employed for thetreatment of prostate carcinoma.One conventional way to limit the toxicity of virusmediated suicide gene therapy is to use cell type specificpromoters as suggested above. Although adenovirusvectors with the native PSA enhancer and promoter(PSAP) provided prostate-specific expression, lowertranscriptional activity observed in prostate challenged itsuse in prostate-targeted gene therapy. To improve theactivity and specificity of the prostate-specific PSAenhancer for gene therapy, various studies were carried outby exploring the properties of the natural PSA controlregions. Chimeric PSA enhancer constructs weregenerated with tandem copies of the proximal AREelements and then inserted into adenovirus constructs (Ad-PSE-BC-luc) (Wu et al, 2001). This construct was highlyinducible with androgens as shown by systemicadministration into SCID mice carrying LAPC-9 humanprostate cancer xenografts while retaining prostate specificgene expression. Furthermore, the CreLoxP system wasalso utilized to enhance the activity of PSAP. CD suicidegene therapy approach using adenoviral vectors withCRELoxP augmented PSAP activity effectively inhibitedsubcutaneous LNCaP tumor growth in nude mice(Yoshimura et al, 2002). In addition, hormone refractoryprostate cancer cells retain the expression of prostatespecificmembrane antigen (PSMA) and prostate-specificantigen (PSA). An adenovirus construct with an artificialchimeric enhancer (PSES) composed of two modifiedregulatory elements of PSA and PSMA genes (Ad-PSESluc)was generated and tested for its promoter activity forthe treatment of prostate cancer (Lee et al, 2002a).Systemic injection of Ad-PSES-luc construct into miceproduced very low levels of reporter gene expression inmajor organs. However, when injected directly intoprostate, only the prostate but not other tissues producedhigh levels of reporter gene expression. These resultsencouraged the use of PSES for the treatment of androgenindependentprostate carcinoma. Even though prostatespecificantigen (PSA/hK3) provided prostate specificgene expression, its expression displayed an inversecorrelation with prostate cancer grade and stage, givingreason to doubt its effectiveness for advanced stage ofprostate carcinoma. A new approach was developed inorder to generate gene therapy vectors targeting highergrades especially of prostate carcinoma. The humanglandular kallikrein 2 (hK2) is upregulated in an advancedform of prostate cancer with a higher grade. Therefore thehK2 promoter was cloned into adenovirus construct incombination with EGFP reporter gene (ADV.hK2-E3/P-EGFP) in order to obtain preferential expression of EGFPin prostate cancer (Xie et al, 2001a). Indeed ADV.hK2-E3/P-EGFP injection led to a robust but tumor-restrictedEGFP expression in subcutaneously generated LNCaPtumors. These results showed that adenovirus constructswith the hk2 multienhancer/promoter driven therapeuticgenes might be a powerful tool for gene therapy ofadvanced prostate cancer.Previous studies have shown that the bone matrixprotein osteocalcin is predominantly expressed in prostatecancer epithelial cells, fibromuscular stromal cells andosteoblasts. A conditional replication competentadenovirus vector carrying the osteocalcin promoterdriven early E1A gene (AdOCE1A) was generated to cotargetboth prostate cancer cells and their surroundingstromal cells (Matsubara et al, 2001). Both PSA-producing(LNCaP) and non-producing (DU145 and PC3) humanprostate cancer cell lines as well as human stromal cellsand osteoblasts were effectively killed by this recombinantvirus in vitro. In addition a single systemic intravenousinjection of the AdOCE1A construct significantlydestroyed prostate tumor cells transplanted in SCID mice.This co-targeting strategy appeared to have a broadereffect compared to other recombinant constructs tested onthe preclinical models of human prostate cancer. Thesepromising results initiated first gene therapy trial (phase I)in which adenoviruses carrying the osteocalcin promoterdriven HSV-tk gene (AdOCHSVTK) were directlyinjected into prostate cancer lymph node and bonemetastasis (Kubo et al, 2003). The results of this trialsuggested that adenoviruses did not display any adverseeffects and the treatment was well tolerated in all patients.In addition, 63 % of the patients had local cell death intreated lesions. Further studies are suggested in order toassess the efficacy of this approach for androgenindependentprostate carcinoma. A new treatmentmodality to enhance adenoviral replication by vitamin D 3in androgen-independent human prostate cancer cells andtumors was tested using a novel replication-competentadenoviral vector, Ad-hOC-E1, carrying the human124

Gene Therapy and Molecular Biology Vol 7, page 125osteocalcin (hOC) promoter to drive both the early viralE1A and E1B genes (Hsieh et al, 2002). While thereplication properties of Ad-hOC-E1 vector wererestricted to OC-expressing cells, vitamin D 3 exposurefurther enhanced viral replication by 10 fold. The growthof both androgen-dependent and androgen-independentprostate cancer cells was suppressed by Ad-hOC-E1infection, irrespective of the cells’ androgenresponsiveness and PSA status. This is in contrast to AdsPSA-E1vector, which only replicated in PSA-expressingcells with androgen receptor (AR). Ad-hOC-E1 injectioninhibited the growth of DU145 (an AR and PSA-negativecell line) tumor xenografts in mice. Consequently, vitaminD 3 -enhanced Ad-hOC-E1 viral replication represented analternative for the treatment of localized or osseousmetastatic prostate cancer. Prostate specific antigenpromoter (PSAP) and rat probasin (rPB) promoter arecurrently employed to drive the therapeutic transgeneexpression in prostate cancer cells. However, since thesepromoters require the binding of androgen to androgenreceptor for activation, they were only functional inandrogen-dependent prostate carcinoma cells. Becauseandrogen refractory prostate carcinoma cells lose theexpression of androgen receptor along the way, constructswith PSAP or rPB promoters are not useful for treatingpatients with androgen-independent prostate carcinoma. Inorder to circurment this problem, prostate specificpromoters were modified so that they were activated inresponse to the retinoids-retinoid receptor complex inplace of the androgen-AR complex. As a result, retinoidtreated androgen-independent prostate cancer cells weresensitized to HSVTK-ganciclovir gene therapy usingpromoters responding to retinoids (Furuhata et al, 2003).Apart from promoters providing tissue specific geneexpression, expression inducible promoters were clonedinto adenovirus constructs to control the onset and theduration of gene expression. Tetracycline-inducibleadenovirus vectors expressing the cytokine interleukin-12were successfully tested in an immunotherapy model forprostate cancer (Nakagawa et al, 2001). Thus, recombinantadenovirus vectors with tetracycline-inducible geneexpression opened up new avenues while improving thesafety of viral vector administration for cancer genetherapy. Limitation of cytotoxic gene expression only totumor cells is very much desired in adenovirus-mediatedgene therapy approach for cancer. Unfortunately, theexpression levels of many tumor and tissue-specificpromoters are much lower than the constitutively activepromoters. A complex adenoviral vector was generated byfusing the tetracycline transactivator gene to a prostatespecificARR2PB promoter while placing a mouse FASL-GFP fusion gene under the control of the tetracyclineresponsive promoter. This allowed the joining of cell-typespecificity with high-level regulation of transgeneexpression (Rubinchik et al, 2001). The doxycyclineregulated, ARR2PB driven FASL-GFP vector generatedhigher levels of prostate-specific FASL-GFP expressionthan FASL-GFP expression directed with ARR2PB alone,leading to apoptosis in LNCaP cells. Systemic delivery ofboth the prostate-specific and the prostate-specific/tetregulatedvectors was well tolerated in animals at dosesthat were lethal for adenovirus vectors with CMV-drivenFASL-GFP expression. This approach improved the safetyand efficacy of adenovirus-mediated cytotoxic genedelivery for the treatment of prostate carcinoma.The prostate-specific adenovirus gene expressiontechnology can also be used for the identification ofmetastatic lesions of prostate cancer through the use ofnon-invasive imaging. A prostate-specific adenovirusvector expressing a luciferase reporter gene (AdPSE-BCluc)and a charge-coupled device-imaging system wereemployed for this purpose (Adams et al, 2002). A robustexpression from AdPSE-BC-luc construct was found inthe prostate, especially in the androgen-independenttumors. Furthermore, metastatic lesions in the lung andspine with prostatic origin were identified successfullythrough repetitive imaging over a three-week period afterAdPSE-BC-luc injection into tumor-bearing mice. Theseresults demonstrate that adenovirus gene delivery specificto the prostate can be coupled to a non-invasive imagingmodality for therapeutic and diagnostic strategies forprostate cancer.XII. Adenovirus vectors forvaccination and adjuvant gene therapyCAR receptors and MHC class I heavy chains areimportant mediators of adenovirus entry into tumor cells.Contrary to the cell lines derived from other malignancies,down regulation of CAR or MHC class I expression isrelatively rare in both human and murine prostatecarcinoma cells. This brought the possibility of developingvaccine strategies for prostate cancer based on themodification of prostate cancer cells using recombinantadenovirus vectors (Pandha et al, 2003). The expression ofprostate-specific antigen (PSA) is highly restricted toprostatic epithelial cells. In fact, 95 % of patients withprostate carcinoma express PSA, making this antigen agood candidate for targeted immunotherapy. Arecombinant PSA adenovirus type 5 (Ad5-PSA) wasgenerated in order to activate PSA-specific T-cell responsewith the potential of eliminating prostate cancer cells(Elzey et al, 2001). Ad5-PSA immunized mice displayed aPSA-specific cellular immunity involving CD8 + Tlymphocytes. This approach deterred subcutaneous tumorformation with RM11 prostate cancer cells expressingPSA (RM11psa). However, this did not affect the growthof existing RM11psa tumors. On the contrary, Ad5-PSAadministration followed by intratumoral injection ofrecombinant canarypox viruses (ALVAC) encodinginterleukin-12 (IL-12), IL-2, and tumor necrosis factor-αeffectively eliminated established RM11psa tumors.Surgery is one of the conventional treatmentmodalities used against solid tumors. Due to the fact thatminor residual tumors following surgical operation mayresult in local recurrence, surgery is neither efficient norplausible for the treatment of metastatic disease. AlthoughAdHSV-tk gene therapy followed by gancicloviradministration has been evaluated extensively as apotential treatment modality for numerous tumors, it hasnot yet been proven to achieve a complete cure on its own.125

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaProstate-derived tumor models were used to evaluate theeffects of AdHSV-tk gene therapy as an adjuvant tosurgery (Sukin et al, 2001). Lung nodules of prostatecancer cells were generated by intravenous injection oftumor cells in order to evaluate systemic effects.Following resection of subcutaneous tumors, AdHSV-tkwas delivered to the resection site. Toxicity, local tumorrecurrence, survival, and lung nodule formation wereevaluated in animals; increased survival and decreasedrecurrence accompanied by no systemic toxicity wereobserved. Adjuvant AdHSV-tk gene therapy resulted in asignificant reduction in lung nodules as well. This studysuggested that AdHSV-tk gene therapy might be beneficialas an adjuvant for patients undergoing surgical treatmentof cancer.XIII. Current progress to overcomerate-limiting steps in adenovirus-mediatedgene therapy for prostate carcinomaThe success of adenovirus mediated gene therapy forprostate carcinoma is effected by several factors includingthe level of expression of the receptor which facilitates theentry of the viral vectors into the cells, penetration oftransgenes to surrounding tissues, and finally theexpression of the delivered gene. Enhancing these factorshas been the focus of many laboratories working onadenovirus-mediated gene therapy for prostate carcinoma.Although a limited number of studies have beencompleted regarding these issues, effectiveness of prostatecancer gene therapy will certainly benefit from theprogress in this field.A. Receptor abundanceThe presence of the coxsackie adenovirus cellsurface receptor, CAR, is required for an effectiveadenovirus infection of target cells. CAR expressionpatterns of normal prostate and prostate carcinoma werecompared using immunohistochemical approaches in orderto assess the feasibility of adenovirus mediated genetherapy for prostate cancer (Rauen et al, 2002). While arobust membrane staining for CAR was detected in themetastatic prostate specimens with higher Gleason scores,just lumenal and lateral cell membrane staining weredetected in the benign prostate epithelia. Therefore,adenovirus mediated gene delivery should be moreeffective for aggressive prostate tumors than it is forbenign cases.B. Penetration of hybrid therapeutictransgenes to the surrounding tissueDespite the fact that adenovirus could transduce cellsvery efficiently in vitro, adenovirus mediated genedelivery is restricted by the inefficient transduction ofsurrounding cells for a given tumor. In order to overcomethis obstacle, an important intercellular transport proteinnamed VP22, was first fused to the therapeutic transgeneof interest (p53 gene) and then cloned into adenovirusvector (Roy et al, 2002). Infection of p53 negative humanprostate cancer cells (LNCaP) by this approach generatedvery efficient gene delivery of p53, inducing apoptosis notonly in the infected cells but also in the surroundinguninfected cells.C. Enhancement of transgene expressionthrough transcriptional regulationAlthough the use of prostate specific promoters isnecessary to limit the transgene toxicity, the low level oftransgene expression directed by these promotersrepresents a barrier to gene therapy. The observation,which led to the idea that chemotherapeutics enhanced thetransgene expression from viral promoters, represented anew approach to overcome this barrier. Two recombinantadenovirus constructs were used to deliver p21WAF-1/CIP1 and p53 protein c-DNA under the control ofcytomegalovirus promoter to the metastatic androgenindependent prostate cancer cells treated withchemotherapeutic agents docetaxel or paclitaxel (Li et al,2002b). Both chemotherapeutics appeared to enhanceadenovirus mediated transgene expression in androgenindependent prostate cancer cell lines. This increase intransgene expression was attributed to the enhancement ofCMV promoter activity rather than the increased viraluptake. Therefore, the observed synergy of gene therapywith these chemotherapeutics may become useful whenthe transgene expression is a limiting factor for thetreatment of the metastatic androgen independent prostatecancer. The possible use of other chemotherapeutic agentsand their effect on prostate specific promoters should alsobe explored.XIV. Summary of clinical trialsThere are 636 clinical protocols involving 3496patients employed in gene therapy worldwide as reportedto the Journal of Gene Medicine website by the year 2002.403 clinical studies (63.4 %) with regard to gene therapyfor cancer were tested on 2392 (68.5 %) patients.Adenovirus was the vector of choice in 171 of theseprotocols (27 %), and 644 patients (18.4 %) received theadenovirus vector for gene therapy. 22 out of 171 clinicalprotocols were engaged in adenovirus mediated genetherapies targeting the prostate only as summarized inTable 1. 13 of these were reported to be in Phase I, 3 trialsin Phase II and the rest (5) were in Phase I/II. There is noPhase III clinical study reported using adenovirus vectorstargeting prostate yet. Some of the adenovirus mediatedgene therapy approaches were complemented either withradiotherapy or radical prostatectomy. The percentage ofthe choice of gene therapy modalities targeting prostate isprovided in Figure 3. The use of selectively replicatingadenovirus constructs leads other approaches followed bysuicide gene therapy. This is partly because not long agoastonishing results were obtained with selectivelyreplicating adenovirus constructs in the preclinical animalmodels. It is also interesting to note that two of theseclinical trials utilize suicide gene therapy in combinationwith the selectively replicating adenovirus approach126

Gene Therapy and Molecular Biology Vol 7, page 127(Figure 3). No clinical studies have been carried out usingthe death ligand-mediated gene therapy approach andadenovirus vectors up to date. However we should not besurprised if such trials are being initiated and we encountersome of these in the near future. Although preliminaryresults are very encouraging from these clinicalinvestigations, clear conclusions can be drawn only uponcompletion of these studies.Considering all these preclinical and clinical studies,we concluded that great progress in adenovirus mediatedgene therapy for prostate carcinoma has been made withinthe last 3 years. While the molecular mechanismsresponsible for prostate carcinoma are not fullyunderstood, the effectiveness of gene therapy is still quiteamazing. As more data become available on theunderstanding of prostate carcinoma, we anticipate thatmore effective treatment modalities will be developedusing adenovirus to target prostate cancer.Table 1. A summary of ongoing clinical trials of adenovirus mediated gene therapy targeting prostate as of 2002. The datawas collected from the Journal of Gene Medicine web site (www.wiley.co.uk/genmed/clinical) and published with thepermission from ©John Wiley and Sons 2002.Country Investigator Mode of Therapy PhaseCanada A. K. Stewart Immunotherapy (IL-2) ICanada J. Dancey Immunotherapy (IL-2) IUSA Peter T. Scardino Suicide gene therapy (HSV-tk) + radiotherapy IUSA Simon J. Hall Neo-adjuvant suicide gene therapy (HSV-tk) + radical prostatectomy IUSA Arie Belldegrun Tumor suppressor gene therapy (p53) IUSA Christopher J.LogothetisTumor suppressor gene therapy (p53)USA Dov Kadmon Neo-adjuvant suicide gene therapy (HSV-tk) + radical prostatectomy IUSA Jonathan W. Simons Selectively replicating adenovirus (CN706) II/IIUSA Thomas A. Gardner Suicide gene therapy (HSV-tk) IUSA Jae Ho Kim Suicide gene therapy (CD/Tk) with selectively replicating adenovirus +radiotherapyUSA E. Brian Butler Suicide Gene Therapy (HSV-tk) + radiotherapy I/IIUSA Jeffrey R. Gingrich Neo-adjuvant CDK inhibitor (p16) + radical prostatectomy IUSA Martha K. Terris Selectively replicating adenovirus (CV787) + Radiotherapy I/IIUSA George Wilding Selectively replicating adenovirus (CV787) I/IIUSA Alan Pollack Tumor suppressor gene therapy (p53) + radiotherapy IIUSA Thomas A. Gardner Selectively replicating adenovirus with osteocalcin promoter (Ad-OC-E1A)USA David M. Lubaroff Immunotherapy (PSA) IUSA Brian J. Miles Immunotherapy (IL-12) + radiotherapy IUSA Theodore L.DeWeeseSelectively replicating adenovirus (CV706)IIUSA Eric J. Small Selectively replicating adenovirus (CV787) + chemotherapy IIUSA Svend O. Freytag Neo-adjuvant suicide gene therapy (CD/Tk) with selectively replicating Iadenovirus + RadiotherapyUSA John M. Corman Selectively replicating adenovirus (CG7060) + radiotherapy I/IIII127

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaFigure 3. Adenovirus mediated clinical gene therapy modalities for prostate. The types of clinical gene therapy modalities for prostateare represented as percentages in a pie graph in order to better appreciate the contribution of each treatment modality.AcknowledgmentsThis work is supported by Akdeniz UniversityScientific Research Project Administration DivisionGrants (#2002.01.0122.06, #2002.01.0122.07 and#2002.01.0200.005 to Dr. Salih Sanlioglu).ReferencesAbe K, Kurakin A, Mohseni-Maybodi M, Kay B, and Khosravi-Far R (2000) The complexity of TNF-related apoptosisinducingligand. Ann N Y Acad Sci 926, 52-63.Adams JY, Johnson M, Sato M, Berger F, Gambhir SS, Carey M,Iruela-Arispe ML, and Wu L (2002) Visualization ofadvanced human prostate cancer lesions in living mice by atargeted gene transfer vector and optical imaging. Nat Med8, 891-897.Andjelic S, Hsia C, Suzuki H, Kadowaki T, Koyasu S, and LiouHC (2000) Phosphatidylinositol 3-kinase and NF-κ B/Rel areat the divergence of CD40-mediated proliferation andsurvival pathways. J Immunol 165, 3860-3867.Batra RK, Guttridge DC, Brenner DA, Dubinett SM, BaldwinAS, and Boucher RC (1999) IκBα gene transfer is cytotoxicto squamous-cell lung cancer cells and sensitizes them totumor necrosis factor-α-mediated cell death. Am J RespirCell Mol Biol 21, 238-245Becker CM, Farnebo FA, Iordanescu I, Behonick DJ, Shih MC,Dunning P, Christofferson R, Mulligan RC, Taylor GA, KuoCJ, and Zetter BR (2002) Gene therapy of prostate cancerwith the soluble vascular endothelial growth factor receptorflk1. Cancer Biol Ther 1, 548-553.Beg AA, and Baltimore D (1996) An essential role for NF-κB inpreventing TNF-α-induced cell death. Science 274, 782-784.Beresford SA, Davies MA, Gallick GE, and Donato NJ (2001)Differential effects of phosphatidylinositol-3/Akt-kinaseinhibition on apoptotic sensitization to cytokines in LNCaPand PCc-3 prostate cancer cells. J Interferon Cytokine Res21, 313-322.Bernard D, Quatannens B, Vandenbunder B, and Abbadie C(2001) Rel/NF-κB transcription factors protect against tumornecrosis factor (TNF)-related apoptosis-inducing ligand(TRAIL)-induced apoptosis by up-regulating the TRAILdecoy receptor DcR1. J Biol Chem 276, 27322-27328.Boring CC, Squires TS, Tong T, and Montgomery S (1994)Cancer statistics, 1994. CA Cancer J Clin 44, 7-26.Burow ME, Weldon CB, Melnik LI, Duong BN, Collins-BurowBM, Beckman BS, and McLachlan JA (2000) PI3-K/AKTregulation of NF-κB signaling events in suppression of TNFinducedapoptosis. Biochem Biophys Res Commun 271,342-345.Cao G, Su J, Lu W, Zhang F, Zhao G, Marteralli D, and Dong Z(2001) Adenovirus-mediated interferon-β gene therapysuppresses growth and metastasis of human prostate cancerin nude mice. Cancer Gene Ther 8, 497-505.Chaudhary PM, Eby M, Jasmin A, Bookwalter A, Murray J, andHood L (1997) Death receptor 5, a new member of the TNFRfamily, and DR4 induce FADD- dependent apoptosis andactivate the NF-κB pathway. Immunity 7, 821-830.Chen Y, DeWeese T, Dilley J, Zhang Y, Li Y, Ramesh N, Lee J,Pennathur-Das R, Radzyminski J, Wypych J, Brignetti D,Scott S, Stephens J, Karpf DB, Henderson DR, and Yu DC(2001) CV706, a prostate cancer-specific adenovirus variant,in combination with radiotherapy produces synergisticantitumor efficacy without increasing toxicity. Cancer Res61, 5453-5460Chhikara M, Huang H, Vlachaki MT, Zhu X, Teh B, Chiu KJ,Woo S, Berner B, Smith EO, Oberg KC, Aguilar LK,Thompson TC, Butler EB, and Aguilar-Cordova E (2001)Enhanced therapeutic effect of HSV-tk+GCV gene therapyand ionizing radiation for prostate cancer. Mol Ther 3, 536-542.Davies MA, Kim SJ, Parikh NU, Dong Z, Bucana CD, andGallick GE (2002) Adenoviral-mediated expression ofMMAC/PTEN inhibits proliferation and metastasis of humanprostate cancer cells. Clin Cancer Res 8, 1904-1914.Davies MA, Koul D, Dhesi H, Berman R, McDonnell TJ,McConkey D, Yung WK, and Steck PA (1999) Regulation ofAkt/PKB activity, cellular growth, and apoptosis in prostatecarcinoma cells by MMAC/PTEN. Cancer Res 59, 2551-2556.Degli-Esposti MA, Dougall WC, Smolak PJ, Waugh JY, SmithCA, and Goodwin RG (1997) The novel receptor TRAIL-R4induces NF-κB and protects against TRAIL-mediatedapoptosis, yet retains an incomplete death domain.Immunity 7, 813-820.Demarchi F, Verardo R, Varnum B, Brancolini C, and SchneiderC (2001) Gas6 anti-apoptotic signaling requires NF-κ Bactivation. J Biol Chem 276, 31738-31744.Devi GR (2002) Prostate cancer: status of current treatments andemerging antisense- based therapies. Curr Opin Mol Ther4, 138-148.DeWeese TL, van der Poel H, Li S, Mikhak B, Drew R,Goemann M, Hamper U, DeJong R, Detorie N, Rodriguez R,Haulk T, DeMarzo AM, Piantadosi S, Yu DC, Chen Y,128

Gene Therapy and Molecular Biology Vol 7, page 129Henderson DR, Carducci MA, Nelson WG, and Simons JW(2001) A phase I trial of CV706, a replication-competent,PSA selective oncolytic adenovirus, for the treatment oflocally recurrent prostate cancer following radiation therapy.Cancer Res 61, 7464-7472Djeha AH, Thomson TA, Leung H, Searle PF, Young LS, KerrDJ, Harris PA, Mountain A, and Wrighton CJ (2001)Combined adenovirus-mediated nitroreductase gene deliveryand CB1954 treatment: a well-tolerated therapy forestablished solid tumors. Mol Ther 3, 233-240Do LV, Do TM, Smith R, and Parker RG (2002) PostoperativeRadiotherapy for Carcinoma of the Prostate: Impact on BothLocal Control and Distant Disease-Free Survival. Am J ClinOncol 25, 1-8.Ebara S, Shimura S, Nasu Y, Kaku H, Kumon H, Yang G, WangJ, Timme TL, Aguilar-Cordova E, and Thompson TC (2002)Gene therapy for prostate cancer: toxicological profile offour HSV-tk transducing adenoviral vectors regulated bydifferent promoters. Prostate Cancer Prostatic Dis 5, 316-325Ehlert JE, and Kubbutat MH (2001) Apoptosis and its relevancein cancer therapy. Onkologie 24, 433-440.Elzey BD, Siemens DR, Ratliff TL, and Lubaroff DM (2001)Immunization with type 5 adenovirus recombinant for atumor antigen in combination with recombinant canarypoxvirus (ALVAC) cytokine gene delivery induces destructionof established prostate tumors. Int J Cancer 94, 842-849Farkas A, Schneider D, Perrotti M, Cummings KB, and WardWS (1998) National trends in the epidemiology of prostatecancer, 1973 to 1994: evidence for the effectiveness ofprostate-specific antigen screening. Urology 52, 444-448;discussion 448-449.Franco AV, Zhang XD, Van Berkel E, Sanders JE, Zhang XY,Thomas WD, Nguyen T, and Hersey P (2001) The role ofNF-κB in TNF-related apoptosis-inducing ligand (TRAIL)-induced apoptosis of melanoma cells. J Immunol 166, 5337-5345.Franklin J, Hislop J, Flynn A, and McArdle CA (2003)Signalling and anti-proliferative effects mediated bygonadotrophin- releasing hormone receptors after expressionin prostate cancer cells using recombinant adenovirus. JEndocrinol 176, 275-284.Freytag SO, Khil M, Stricker H, Peabody J, Menon M,DePeralta-Venturina M, Nafziger D, Pegg J, Paielli D,Brown S, Barton K, Lu M, Aguilar-Cordova E, and Kim JH(2002a) Phase I study of replication-competent adenovirusmediateddouble suicide gene therapy for the treatment oflocally recurrent prostate cancer. Cancer Res 62, 4968-4976Freytag SO, Paielli D, Wing M, Rogulski K, Brown S,Kolozsvary A, Seely J, Barton K, Dragovic A, and Kim JH(2002b) Efficacy and toxicity of replication-competentadenovirus-mediated double suicide gene therapy incombination with radiation therapy in an orthotopic mouseprostate cancer model. Int J Radiat Oncol Biol Phys 54,873-885Furuhata S, Ide H, Miura Y, Yoshida T, and Aoki K (2003)Development of a prostate-specific promoter for genetherapy against androgen-independent prostate cancer. MolTher 7, 366-374.Gabril MY, Onita T, Ji PG, Sakai H, Chan FL, Koropatnick J,Chin JL, Moussa M, and Xuan JW (2002) Prostate targeting:PSP94 gene promoter/enhancer region directed prostatetissue-specific expression in a transgenic mouse prostatecancer model. Gene Ther 9, 1589-1599Galfione M, Luo W, Kim J, Hawke D, Kobayashi R, Clapp C,Yu-Lee LY, and Lin SH (2003) Expression and purificationof the angiogenesis inhibitor 16-kDa prolactin fragment frominsect cells. Protein Expr Purif 28, 252-258.Gasparian AV, Yao YJ, Kowalczyk D, Lyakh LA, Karseladze A,Slaga TJ, and Budunova IV (2002a) The role of IKK inconstitutive activation of NF-κB transcription factor inprostate carcinoma cells. J Cell Sci 115, 141-151.Gasparian AV, Yao YJ, Lu J, Yemelyanov AY, Lyakh LA, SlagaTJ, and Budunova IV (2002b) Selenium compounds inhibit IκB kinase (IKK) and nuclear factor- κB (NF-κB) in prostatecancer cells. Mol Cancer Ther 1, 1079-1087.Giri D, Ozen M, and Ittmann M (2001) Interleukin-6 is anautocrine growth factor in human prostate cancer. Am JPathol 159, 2159-2165.Greenlee RT, Hill-Harmon MB, Murray T, and Thun M (2001)Cancer statistics, 2001. CA Cancer J Clin 51, 15-36.Griffith TS, and Broghammer EL (2001) Suppression of tumorgrowth following intralesional therapy with TRAILrecombinant adenovirus. Mol Ther 4, 257-266.Griffith TS, Chin WA, Jackson GC, Lynch DH, and Kubin MZ(1998) Intracellular regulation of TRAIL-induced apoptosisin human melanoma cells. J Immunol 161, 2833-2840.Griffith TS, and Lynch DH (1998) TRAIL: a molecule withmultiple receptors and control mechanisms. Curr OpinImmunol 10, 559-563.Grumet SC, and Bruner DW (2000) The identification andscreening of men at high risk for developing prostate cancer.Urol Nurs 20, 15-18, 23-14, 46.Gustin JA, Maehama T, Dixon JE, and Donner DB (2001) ThePTEN tumor suppressor protein inhibits tumor necrosisfactor- induced nuclear factor κB activity. J Biol Chem 276,27740-27744.Hall SJ, Canfield SE, Yan Y, Hassen W, Selleck WA, and ChenSH (2002) A novel bystander effect involving tumor cellderivedFas and FasL interactions following Ad.HSV-tk andAd.mIL-12 gene therapies in experimental prostate cancer.Gene Ther 9, 511-517.Hatano E, and Brenner DA (2001) Akt protects mousehepatocytes from TNF-α- and Fas-mediated apoptosisthrough NK-κB activation. Am J Physiol GastrointestLiver Physiol 281, G1357-1368.Hemmati PG, Gillissen B, von Haefen C, Wendt J, Starck L,Guner D, Dorken B, and Daniel PT (2002) Adenovirusmediatedoverexpression of p14(ARF) induces p53 and Baxindependentapoptosis. Oncogene 21, 3149-3161.Herr I, and Debatin KM (2001) Cellular stress response andapoptosis in cancer therapy. Blood 98, 2603-2614.Honda T, Kagawa S, Spurgers KB, Gjertsen BT, Roth JA, FangB, Lowe SL, Norris JS, Meyn RE, and McDonnell TJ (2002)A recombinant adenovirus expressing wild-type Bax inducesapoptosis in prostate cancer cells independently of their Bcl-2 status and androgen sensitivity. Cancer Biol Ther 1, 163-167Horowitz J (1999) Adenovirus-mediated p53 gene therapy:overview of preclinical studies and potential clinicalapplications. Curr Opin Mol Ther 1, 500-509.Hsieh CL, and Chung LW (2001) New prospectives of prostatecancer gene therapy: molecular targets and animal models.Crit Rev Eukaryot Gene Expr 11, 77-120.Hsieh CL, Yang L, Miao L, Yeung F, Kao C, Yang H, Zhau HE,and Chung LW (2002) A novel targeting modality toenhance adenoviral replication by vitamin D(3) in androgenindependenthuman prostate cancer cells and tumors. CancerRes 62, 3084-3092.Hyer ML, Sudarshan S, Kim Y, Reed JC, Dang JY, SchwartzDA, and Norris JS (2002) Downregulation of c-FLIPSensitizes DU145 Prostate Cancer Cells to Fas- MediatedApoptosis. Cancer Biol Ther 1, 401-406.Hyer ML, Sudarshan S, Schwartz DA, Hannun Y, Dong JY, andNorris JS (2003) Quantification and characterization of thebystander effect in prostate cancer cells following129

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaadenovirus-mediated FasL expression. Cancer Gene Ther10, 330-339.Hyer ML, Voelkel-Johnson C, Rubinchik S, Dong J, and NorrisJS (2000) Intracellular Fas ligand expression causes Fasmediatedapoptosis in human prostate cancer cells resistant tomonoclonal antibody-induced apoptosis. Mol Ther 2, 348-358.Irmler M, Thome M, Hahne M, Schneider P, Hofmann K, SteinerV, Bodmer JL, Schroter M, Burns K, Mattmann C, RimoldiD, French LE, and Tschopp J (1997) Inhibition of deathreceptor signals by cellular FLIP. Nature 388, 190-195.Jones RG, Parsons M, Bonnard M, Chan VS, Yeh WC,Woodgett JR, and Ohashi PS (2000) Protein kinase Bregulates T lymphocyte survival, nuclear factor κBactivation, and Bcl-X(L) levels in vivo. J Exp Med 191,1721-1734.Jorgensen TJ, Katz S, Wittmack EK, Varghese S, Todo T,Rabkin SD, and Martuza RL (2001) Ionizing radiation doesnot alter the antitumor activity of herpes simplex virus vectorG207 in subcutaneous tumor models of human and murineprostate cancer. Neoplasia 3, 451-456Kane LP, Shapiro VS, Stokoe D, and Weiss A (1999) Inductionof NF-κB by the Akt/PKB kinase. Curr Biol 9, 601-604.Katner AL, Hoang QB, Gootam P, Jaruga E, Ma Q, Gnarra J, andRayford W (2002) Induction of cell cycle arrest andapoptosis in human prostate carcinoma cells by arecombinant adenovirus expressing p27(Kip1). Prostate 53,77-87.Kim J, Luo W, Chen DT, Earley K, Tunstead J, Yu-Lee LY, andLin SH (2003) Antitumor activity of the 16-kDa prolactinfragment in prostate cancer. Cancer Res 63, 386-393.Klotz L (2000a) Hormone therapy for patients with prostatecarcinoma. Cancer 88, 3009-3014.Klotz L (2000b) Intraoperative cavernous nerve stimulationduring nerve sparing radical prostatectomy: how and when?Curr Opin Urol 10, 239-243.Koksal IT, Ozcan F, Kadioglu TC, Esen T, KiliCaslan I, andTunc M (2000) Discrepancy between Gleason scores ofbiopsy and radical prostatectomy specimens. Eur Urol 37,670-674.Kreuz S, Siegmund D, Scheurich P, and Wajant H (2001) NF-κBinducers upregulate cFLIP, a cycloheximide-sensitiveinhibitor of death receptor signaling. Mol Cell Biol 21, 3964-3973.Kubo H, Gardner TA, Wada Y, Koeneman KS, Gotoh A, YangL, Kao C, Lim SD, Amin MB, Yang H, Black ME,Matsubara S, Nakagawa M, Gillenwater JY, Zhau HE, andChung LW (2003) Phase I dose escalation clinical trial ofadenovirus vector carrying osteocalcin promoter-drivenherpes simplex virus thymidine kinase in localized andmetastatic hormone-refractory prostate cancer. Hum GeneTher 14, 227-241Lee SJ, Kim HS, Yu R, Lee K, Gardner TA, Jung C, Jeng MH,Yeung F, Cheng L, and Kao C (2002a) Novel prostatespecificpromoter derived from PSA and PSMA enhancers.Mol Ther 6, 415-421.Lee YJ, Galoforo SS, Battle P, Lee H, Corry PM, and Jessup JM(2001) Replicating adenoviral vector-mediated transfer of aheat-inducible double suicide gene for gene therapy. CancerGene Ther 8, 397-404.Lee YJ, Lee H, and Borrelli MJ (2002b) Gene transfer intohuman prostate adenocarcinoma cells with an adenoviralvector: Hyperthermia enhances a double suicide geneexpression, cytotoxicity and radiotoxicity. Cancer GeneTher 9, 267-274.Levine AJ (1997) p53, the cellular gatekeeper for growth anddivision. Cell 88, 323-331.Li X, Marani M, Yu J, Nan B, Roth JA, Kagawa S, Fang B,Denner L, and Marcelli M (2001) Adenovirus-mediated Baxoverexpression for the induction of therapeutic apoptosis inprostate cancer. Cancer Res 61, 186-191.Li Y, McCadden J, Ferrer F, Kruszewski M, Carducci M, SimonsJ, and Rodriguez R (2002a) Prostate-specific expression ofthe diphtheria toxin A chain (DT-A): studies of inducibilityand specificity of expression of prostate- specific antigenpromoter-driven DT-A adenoviral-mediated gene transfer.Cancer Res 62, 2576-2582.Li Y, Okegawa T, Lombardi DP, Frenkel EP, and Hsieh JT(2002b) Enhanced transgene expression in androgenindependent prostate cancer gene therapy by taxanechemotherapeutic agents. J Urol 167, 339-346.Lin SH, Pu YS, Luo W, Wang Y, and Logothetis CJ (1999)Schedule-dependence of C-CAM1 adenovirus gene therapyin a prostate cancer model. Anticancer Res 19, 337-340.Loimas S, Toppinen MR, Visakorpi T, Janne J, and Wahlfors J(2001) Human prostate carcinoma cells as targets for herpessimplex virus thymidine kinase-mediated suicide genetherapy. Cancer Gene Ther 8, 137-144.Lowe SL, Rubinchik S, Honda T, McDonnell TJ, Dong JY, andNorris JS (2001) Prostate-specific expression of Baxdelivered by an adenoviral vector induces apoptosis inLNCaP prostate cancer cells. Gene Ther 8, 1363-1371.Lu L, Schulz H, and Wolf DA (2002) The F-box protein SKP2mediates androgen control of p27 stability in LNCaP humanprostate cancer cells. BMC Cell Biol 3, 22.Lu Y, and Steiner MS (2000) Transcriptionally regulatedadenoviruses for prostate-specific gene therapy. World JUrol 18, 93-101.Lundqvist A, Choudhury A, Nagata T, Andersson T, Quinn G,Fong T, Maitland N, Pettersson S, Paulie S, and Pisa P(2002a) Recombinant adenovirus vector activates andprotects human monocyte- derived dendritic cells fromapoptosis. Hum Gene Ther 13, 1541-1Lundqvist A, Nagata T, Kiessling R, and Pisa P (2002b) Maturedendritic cells are protected from Fas/CD95-mediatedapoptosis by upregulation of Bcl-X(L). Cancer ImmunolImmunother 51, 139-144.Madrid LV, Mayo MW, Reuther JY, and Baldwin AS, Jr. (2001)Akt stimulates the transactivation potential of the RelA/p65Subunit of NF-κB through utilization of the IκB kinase andactivation of the mitogen-activated protein kinase p38. J BiolChem 276, 18934-18940Martinez LA, Yang J, Vazquez ES, Rodriguez-Vargas Mdel C,Olive M, Hsieh JT, Logothetis CJ, and Navone NM (2002)p21 modulates threshold of apoptosis induced by DNAdamageand growth factor withdrawal in prostate cancercells. Carcinogenesis 23, 1289-129Martiniello-Wilks R, Tsatralis T, Russell P, Brookes DE,Zandvliet D, Lockett LJ, Both GW, Molloy PL, and RussellPJ (2002) Transcription-targeted gene therapy for androgenindependentprostate cancer. Cancer Gene Ther 9, 443-452.Matsubara S, Wada Y, Gardner TA, Egawa M, Park MS, HsiehCL, Zhau HE, Kao C, Kamidono S, Gillenwater JY, andChung LW (2001) A conditional replication-competentadenoviral vector, Ad-OC-E1a, to cotarget prostate cancerand bone stroma in an experimental model of androgenindependentprostate cancer bone metastasis. Cancer Res 61,6012-6019Mayo MW, Madrid LV, Westerheide SD, Jones DR, Yuan XJ,Baldwin AS, Jr., and Whang YE (2002) PTEN blocks tumornecrosis factor-induced NF-κB-dependent transcription byinhibiting the transactivation potential of the p65 subunit. JBiol Chem 277, 1111Meng RD, McDonald ER, 3rd, Sheikh MS, Fornace AJ, Jr., andEl-Deiry WS (2000) The TRAIL decoy receptor TRUNDD130

Gene Therapy and Molecular Biology Vol 7, page 131(DcR2, TRAIL-R4) is induced by adenovirus-p53overexpression and can delay TRAIL-, p53-, andKILLER/DR5-dependent colon cancer apoptosis. Mol Ther1, 130-144.Miles BJ, Shalev M, Aguilar-Cordova E, Timme TL, Lee HM,Yang G, Adler HL, Kernen K, Pramudji CK, Satoh T, GdorY, Ren C, Ayala G, Wheeler TM, Butler EB, Kadmon D, andThompson TC (2001) Prostate-specific antigen response andsystemic T cell activation after in situ gene therapy inprostate cancer patients failing radiotherapy. Hum GeneTher 12, 1955-1967Mitsiades N, Mitsiades CS, Poulaki V, Chauhan D, RichardsonPG, Hideshima T, Munshi N, Treon SP, and Anderson KC(2002) Biologic sequelae of nuclear factor-κB blockade inmultiple myeloma: therapeutic applications. Blood 99, 4079-4086.Miyagi T, Hori O, Koshida K, Egawa M, Kato H, Kitagawa Y,Ozawa K, Ogawa S, and Namiki M (2002) Antitumor effectof reduction of 150-kDa oxygen-regulated protein expressionon human prostate cancer cells. Int J Urol 9, 577-585.Miyagi T, Koshida K, Hori O, Konaka H, Katoh H, Kitagawa Y,Mizokami A, Egawa M, Ogawa S, Hamada H, and NamikiM (2003) Gene therapy for prostate cancer using the cytosinedeaminase/uracil phosphoribosyltransferase suicide system. JGene Med 5, 30-37.Modur V, Nagarajan R, Evers BM, and Milbrandt J (2002)FOXO proteins regulate tumor necrosis factor-relatedapoptosis inducing ligand expression. Implications for PTENmutation in prostate cancer. J Biol Chem 277, 47928-47937.Moffatt KA, Johannes WU, Hedlund TE, and Miller GJ (2001)Growth inhibitory effects of 1α, 25-dihydroxyvitamin D(3)are mediated by increased levels of p21 in the prostaticcarcinoma cell line ALVA-31. Cancer Res 61, 7122-7129.Nagane M, Huang HJ, and Cavenee WK (2001) The potential ofTRAIL for cancer chemotherapy. Apoptosis 6, 191-197.Nagata S (1997) Apoptosis by death factor. Cell 88, 355-365.Nakagawa S, Massie B, and Hawley RG (2001) Tetracyclineregulatableadenovirus vectors: pharmacologic properties andclinical potential. Eur J Pharm Sci 13, 53-60.Nesterov A, Lu X, Johnson M, Miller GJ, Ivashchenko Y, andKraft AS (2001) Elevated AKT activity protects the prostatecancer cell line LNCaP from TRAIL-induced apoptosis. JBiol Chem 276, 10767-10774.Norris JS, Hyer ML, Voelkel-Johnson C, Lowe SL, Rubinchik S,and Dong JY (2001) The use of Fas Ligand, TRAIL and Baxin gene therapy of prostate cancer. Curr Gene Ther 1, 123-136.Obata A, Eura M, Sasaki J, Saya H, Chikamatsu K, Tada M, IggoRD, and Yumoto E (2000) Clinical significance of p53functional loss in squamous cell carcinoma of theoropharynx. Int J Cancer 89, 187-193.Oya M, Ohtsubo M, Takayanagi A, Tachibana M, Shimizu N,and Murai M (2001) Constitutive activation of nuclearfactor-κB prevents TRAIL-induced apoptosis in renal cancercells. Oncogene 20, 3888-3896.Ozen M, Giri D, Ropiquet F, Mansukhani A, and Ittmann M(2001) Role of fibroblast growth factor receptor signaling inprostate cancer cell survival. J Natl Cancer Inst 93, 1783-1790.Ozes ON, Mayo LD, Gustin JA, Pfeffer SR, Pfeffer LM, andDonner DB (1999) NF-κB activation by tumour necrosisfactor requires the Akt serine- threonine kinase. Nature 401,82-85.Pan G, Ni J, Wei YF, Yu G, Gentz R, and Dixit VM (1997) Anantagonist decoy receptor and a death domain-containingreceptor for TRAIL. Science 277, 815-818.Pandha HS, Stockwin LH, Eaton J, Clarke IA, Dalgleish AG,Todryk SM, and Blair GE (2003) Coxsackie B andadenovirus receptor, integrin and major histocompatibilitycomplex class I expression in human prostate cancer celllines: implications for gene therapy strategies. ProstateCancer Prostatic Dis 6, 6-11Perrotti M, Rabbani F, Farkas A, Ward WS, and Cummings KB(1998) Trends in poorly differentiated prostate cancer 1973to 1994: observations from the Surveillance, Epidemiologyand End Results database. J Urol 160, 811-815.Pitti RM, Marsters SA, Ruppert S, Donahue CJ, Moore A, andAshkenazi A (1996) Induction of apoptosis by Apo-2 ligand,a new member of the tumor necrosis factor cytokine family.J Biol Chem 271, 12687-12690.Powell IJ, Dey J, Dudley A, Pontes JE, Cher ML, Sakr W,Grignon DJ, and Wood DP (2002) Disease-free survivaldifference between African Americans and whites afterradical prostatectomy for local prostate cancer: amultivariable analysis. Urology 59, 907-Pramudji C, Shimura S, Ebara S, Yang G, Wang J, Ren C, YuanY, Tahir SA, Timme TL, and Thompson TC (2001) In situprostate cancer gene therapy using a novel adenoviral vectorregulated by the caveolin-1 promoter. Clin Cancer Res 7,4272-4279.Pu YS, Do KA, Luo W, Logothetis CJ, and Lin SH (2002)Enhanced suppression of prostate tumor growth bycombining C-CAM1 gene therapy and angiogenesis inhibitorTNP-470. Anticancer Drugs 13, 743-749.Pu YS, Luo W, Lu HH, Greenberg NM, Lin SH, and Gingrich JR(1999) Differential expression of C-CAM cell adhesionmolecule in prostate carcinogenesis in a transgenic mousemodel. J Urol 162, 892-896.Rauen KA, Sudilovsky D, Le JL, Chew KL, Hann B, WeinbergV, Schmitt LD, and McCormick F (2002) Expression of thecoxsackie adenovirus receptor in normal prostate and inprimary and metastatic prostate carcinoma: potentialrelevance to gene therapy. Cancer Res 62, 3812-3818.Ravi R, Bedi GC, Engstrom LW, Zeng Q, Mookerjee B, GelinasC, Fuchs EJ, and Bedi A (2001) Regulation of death receptorexpression and TRAIL/Apo2L-induced apoptosis by NF-κB.Nat Cell Biol 3, 409-416.Reed JC (2000) Mechanisms of apoptosis. Am J Pathol 157,1415-1430.Rodriguez R, Schuur ER, Lim HY, Henderson GA, Simons JW,and Henderson DR (1997) Prostate attenuated replicationcompetent adenovirus (ARCA) CN706: a selective cytotoxicfor prostate-specific antigen-positive prostate cancer cells.Cancer Res 57, 2559-25Romashkova JA, and Makarov SS (1999) NF-κB is a target ofAKT in anti-apoptotic PDGF signalling. Nature 401, 86-90.Ross JS, Sheehan CE, Dolen EM, and Kallakury BV (2002a)Morphologic and molecular prognostic markers in prostatecancer. Adv Anat Pathol 9, 115-128.Ross JS, Sheehan CE, Fisher HA, Kauffman RA, Dolen EM, andKallakury BV (2002b) Prognostic markers in prostate cancer.Expert Rev Mol Diagn 2, 129-142.Roy I, Holle L, Song W, Holle E, Wagner T, and Yu X (2002)Efficient translocation and apoptosis induction by adenovirusencoded VP22-p53 fusion protein in human tumor cells invitro. Anticancer Res 22, 3185-3189.Rubinchik S, Wang D, Yu H, Fan F, Luo M, Norris JS, and DongJY (2001) A complex adenovirus vector that delivers FASL-GFP with combined prostate-specific and tetracyclineregulatedexpression. Mol Ther 4, 416-426.Rylova SN, Amalfitano A, Persaud-Sawin DA, Guo WX, ChangJ, Jansen PJ, Proia AD, and Boustany RM (2002) The CLN3gene is a novel molecular target for cancer drug discovery.Cancer Res 62, 801-808.131

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinomaSanda MG (1997) Biological principles and clinical developmentof prostate cancer gene therapy. Semin Urol Oncol 15, 43-55.Sanlioglu S, and Engelhardt JF (1999) Cellular redox state altersrecombinant adeno-associated virus transduction throughtyrosine phosphatase pathways. Gene Ther 6, 1427-1437.Sanlioglu S, Luleci G, and Thomas KW (2001a) Simultaneousinhibition of Rac1 and IKK pathways sensitizes lung cancercells to TNFα-mediated apoptosis. Cancer Gene Ther 8,897-905.Sanlioglu S, Williams CM, Samavati L, Butler NS, Wang G,McCray PB, Jr., Ritchie TC, Hunninghake GW, Zandi E, andEngelhardt JF (2001b) Lipopolysaccharide induces Rac1-dependent reactive oxygen species formation and coordinatestumor necrosis factor-α secretion through IKK regulation ofNF-κB. J Biol Chem 276, 30188-30198.Sasaki R, Shirakawa T, Zhang ZJ, Tamekane A, Matsumoto A,Sugimura K, Matsuo M, Kamidono S, and Gotoh A (2001)Additional gene therapy with Ad5CMV-p53 enhanced theefficacy of radiotherapy in human prostate cancer cells. Int JRadiat Oncol Biol Phys 51, 1336-1345.Schneider P, Thome M, Burns K, Bodmer JL, Hofmann K,Kataoka T, Holler N, and Tschopp J (1997) TRAIL receptors1 (DR4) and 2 (DR5) signal FADD-dependent apoptosis andactivate NF-κB. Immunity 7, 831-836.Schroder FH, Hermanek P, Denis L, Fair WR, GospodarowiczMK, and Pavone-Macaluso M (1992) The TNMclassification of prostate cancer. Prostate Suppl 4, 129-138.Sears RC, and Nevins JR (2002) Signaling networks that link cellproliferation and cell fate. J Biol Chem 22, 22.Selleck WA, Canfield SE, Hassen WA, Meseck M, Kuzmin AI,Eisensmith RC, Chen SH, and Hall SJ (2003) IFN-gammasensitization of prostate cancer cells to fas-mediated death: agene therapy approach. Mol Ther 7, 185-192.Shariat SF, Desai S, Song W, Khan T, Zhao J, Nguyen C, FosterBA, Greenberg N, Spencer DM, and Slawin KM (2001)Adenovirus-mediated transfer of inducible caspases: a novel"death switch" gene therapeutic approach to prostate cancer.Cancer Res 61, 2562-2Sheikh MS, and Fornace AJ, Jr. (2000) Death and decoyreceptors and p53-mediated apoptosis. Leukemia 14, 1509-1513.Sheridan JP, Marsters SA, Pitti RM, Gurney A, Skubatch M,Baldwin D, Ramakrishnan L, Gray CL, Baker K, Wood WI,Goddard AD, Godowski P, and Ashkenazi A (1997) Controlof TRAIL-induced apoptosis by a family of signaling anddecoy receptors. Science 277, 8Siemens DR, Elzey BD, Lubaroff DM, Bohlken C, Jensen RJ,Swanson AK, and Ratliff TL (2001) Cutting edge:restoration of the ability to generate CTL in mice immune toadenovirus by delivery of virus in a collagen-based matrix. JImmunol 166, 731-735.Simpson L, and Parsons R (2001) PTEN: life as a tumorsuppressor. Exp Cell Res 264, 29-41.Sizemore N, Lerner N, Dombrowski N, Sakurai H, and Stark GR(2002) Distinct roles of the IκB kinase α and β subunits inliberating nuclear factor κB (NF-κB) from κB and inphosphorylating the p65 subunit of NF-κB. and inphosphorylating the p65 subunit of NF-κB. J Biol Chem277, 3863-3869Slack JK, Adams RB, Rovin JD, Bissonette EA, Stoker CE, andParsons JT (2001) Alterations in the focal adhesionkinase/Src signal transduction pathway correlate withincreased migratory capacity of prostate carcinoma cells.Oncogene 20, 1152-1163.Smith MR, Finkelstein JS, McGovern FJ, Zietman AL, FallonMA, Schoenfeld DA, and Kantoff PW (2002) Changes inBody Composition during Androgen Deprivation Therapyfor Prostate Cancer. J Clin Endocrinol Metab 87, 599-603.Spitzweg C, Dietz AB, O'Connor MK, Bergert ER, Tindall DJ,Young CY, and Morris JC (2001) In vivo sodium iodidesymporter gene therapy of prostate cancer. Gene Ther 8,1524-1531.Steck PA, Pershouse MA, Jasser SA, Yung WK, Lin H, LigonAH, Langford LA, Baumgard ML, Hattier T, Davis T, FryeC, Hu R, Swedlund B, Teng DH, and Tavtigian SV (1997)Identification of a candidate tumour suppressor gene,MMAC1, at chromosome 10q23.3 that is mutated in multipleadvanced cancers. Nat Genet 15, 356-362.Stein R, Chen S, Reed L, Richel H, and Goldenberg DM (2002)Combining radioimmunotherapy and chemotherapy fortreatment of medullary thyroid carcinoma. Cancer 94, 51-61.Sukin SW, Chhikara M, Zhu X, Ayala G, Aguilar LK, O'BrianSmith E, Miles BJ, Thompson TC, Kadmon D, and Aguilar-Cordova E (2001) In vivo surgical resection plus adjuvantgene therapy in the treatment of mammary and prostatecancer. Mol Ther 3, 500-506.Taj MM, Tawil RJ, Engstrom LD, Zeng Z, Hwang C, Sanda MG,and Wechsler DS (2001) Mxi1, a Myc antagonist, suppressesproliferation of DU145 human prostate cells. Prostate 47,194-204.Teh BS, Aguilar-Cordova E, Kernen K, Chou CC, Shalev M,Vlachaki MT, Miles B, Kadmon D, Mai WY, Caillouet J,Davis M, Ayala G, Wheeler T, Brady J, Carpenter LS, LuHH, Chiu JK, Woo SY, Thompson T, and Butler EB (2001)Phase I/II trial evaluating combined radiotherapy and in situgene therapy with or without hormonal therapy in thetreatment of prostate cancer--a preliminary report. Int JRadiat Oncol Biol Phys 51, 605-613.Terlikowski SJ (2001) Tumour necrosis factor and cancertreatment: a historical review and perspectives. Rocz AkadMed Bialymst 46, 5-18.Tong Y, Yang Q, Vater C, Venkatesh LK, Custeau D, ChittendenT, Chinnadurai G, and Gourdeau H (2001) The pro-apoptoticprotein, Bik, exhibits potent antitumor activity that isdependent on its BH3 domain. Mol Cancer Ther 1, 95-102.Van Antwerp DJ, Martin SJ, Kafri T, Green DR, and Verma IM(1996) Suppression of TNF-α-induced apoptosis by NF-κB.Science 274, 787-789.Van Noesel MM, van Bezouw S, Salomons GS, Voute PA,Pieters R, Baylin SB, Herman JG, and Versteeg R (2002)Tumor-specific down-regulation of the tumor necrosis factorrelatedapoptosis-inducing ligand decoy receptors DcR1 andDcR2 is associated with dense promoter hypermethylation.Cancer Res 62, 2157-2161.Vieweg J, Boczkowski D, Roberson KM, Edwards DW, PhilipM, Philip R, Rudoll T, Smith C, Robertson C, and Gilboa E(1995) Efficient gene transfer with adeno-associated virusbasedplasmids complexed to cationic liposomes for genetherapy of human prostate cancer. Cancer Res 55, 2366-2372.Voeks D, Martiniello-Wilks R, Madden V, Smith K, Bennetts E,Both GW, and Russell PJ (2002) Gene therapy for prostatecancer delivered by ovine adenovirus and mediated by purinenucleoside phosphorylase and fludarabine in mouse models.Gene Ther 9, 759-7Voelkel-Johnson C, King DL, and Norris JS (2002) Resistance ofprostate cancer cells to soluble TNF-related apoptosisinducingligand (TRAIL/Apo2L) can be overcome bydoxorubicin or adenoviral delivery of full-length TRAIL.Cancer Gene Ther 9, 164-172.Wang J, and Waxman J (2001) Chemotherapy for prostatecancer. Clin Oncol 13, 453-454.Wang L, Yorke E, Desobry G, and Chui CS (2002) Dosimetricadvantage of using 6 MV over 15 MV photons in conformal132

Gene Therapy and Molecular Biology Vol 7, page 133therapy of lung cancer: Monte Carlo studies in patientgeometries. J Appl Clin Med Phys 3, 51-59.Wible BA, Wang L, Kuryshev YA, Basu A, Haldar S, andBrown AM (2002) Increased K+ efflux and apoptosisinduced by the potassium channel modulatory proteinKChAP/PIAS3β in prostate cancer cells. J Biol Chem 277,17852-17862.Wiley SR, Schooley K, Smolak PJ, Din WS, Huang CP, NichollJK, Sutherland GR, Smith TD, Rauch C, Smith CA, and etal. (1995) Identification and characterization of a newmember of the TNF family that induces apoptosis.Immunity 3, 673-682.Wu L, Matherly J, Smallwood A, Adams JY, Billick E,Belldegrun A, and Carey M (2001) Chimeric PSA enhancersexhibit augmented activity in prostate cancer gene therapyvectors. Gene Ther 8, 1416-1426.Xess A, Singh M, Raghwendra KH, Sharma HP, and Shahi SK(2001) Prostate specific antigen as tumor marker:relationship with histologic grading. Indian J PatholMicrobiol 44, 261-264.Xie X, Zhao X, Liu Y, Young CY, Tindall DJ, Slawin KM, andSpencer DM (2001a) Robust prostate-specific expression fortargeted gene therapy based on the human kallikrein 2promoter. Hum Gene Ther 12, 549-561.Xie X, Zhao X, Liu Y, Zhang J, Matusik RJ, Slawin KM, andSpencer DM (2001b) Adenovirus-mediated tissue-targetedexpression of a caspase-9-based artificial death switch for thetreatment of prostate cancer. Cancer Res 61, 6795-6804.Yeung F, and Chung LW (2002) Molecular basis of co-targetingprostate tumor and stroma. J Cell Biochem Suppl Suppl,65-72.Yoshimura I, Ikegami S, Suzuki S, Tadakuma T, and HayakawaM (2002) Adenovirus mediated prostate specific enzymeprodrug gene therapy using prostate specific antigenpromoter enhanced by the Cre-loxP system. J Urol 168,2659-2664.Yu D, Chen D, Chiu C, Razmazma B, Chow YH, and Pang S(2001a) Prostate-specific targeting using PSA promoterbasedlentiviral vectors. Cancer Gene Ther 8, 628-635.Yu DC, Chen Y, Dilley J, Li Y, Embry M, Zhang H, Nguyen N,Amin P, Oh J, and Henderson DR (2001b) Antitumorsynergy of CV787, a prostate cancer-specific adenovirus, andpaclitaxel and docetaxel. Cancer Res 61, 517-525.Yu DC, Chen Y, Seng M, Dilley J, and Henderson DR (1999)The addition of adenovirus type 5 region E3 enables calydonvirus 787 to eliminate distant prostate tumor xenografts.Cancer Res 59, 4200-4203.Yuan XJ, and Whang YE (2002) PTEN sensitizes prostate cancercells to death receptor-mediated and drug-induced apoptosisthrough a FADD-dependent pathway. Oncogene 21, 319-327.Zeimet AG, Riha K, Berger J, Widschwendter M, Hermann M,Daxenbichler G, and Marth C (2000) New insights into p53regulation and gene therapy for cancer. Biochem Pharmacol60, 1153-1163.Zerbini LF, Wang Y, Cho JY, and Libermann TA (2003)Constitutive Activation of Nuclear Factor κB p50/p65 andFra-1 and JunD Is Essential for Deregulated Interleukin 6Expression in Prostate Cancer. Cancer Res 63, 2206-2215.Zhang F, Lu W, and Dong Z (2002a) Tumor-infiltratingmacrophages are involved in suppressing growth andmetastasis of human prostate cancer cells by INF-β genetherapy in nude mice. Clin Cancer Res 8, 2942-2951.Zhang X, Steiner MS, Rinaldy A, and Lu Y (2001) Apoptosisinduction in prostate cancer cells by a novel gene product,pHyde, involves caspase-3. Oncogene 20, 5982-5990.Zhang Y, Yu J, Unni E, Shao TC, Nan B, Snabboon T, Kasper S,Andriani F, Denner L, and Marcelli M (2002b) Monogeneand Polygene Therapy for the Treatment of ExperimentalProstate Cancers by Use of Apoptotic Genes bax and badDriven by the Prostate- Specific Promoter ARR(2)PB. HumGene Ther 13, 2051-2064.Zhang Z, Yin L, Zhang Y, and Zhao F (2002c) In situtransduction of cytosine deaminase gene followed bysystemic use of 5-fluorocytosine inhibits tumor growth andmetastasis in orthotopic prostate cancer mouse models. ChinMed J (Engl) 115, 227-231.Dr. Salih Sanlioglu133

Sanlioglu et al: Adenovirus mediated gene therapy for prostate carcinoma134

Gene Therapy and Molecular Biology Vol 7, page 135Gene Ther Mol Biol Vol 7, 135-151, 2003Gene therapy for vascular diseasesReview ArticleSarah J. George 1 , Filomena de Nigris 2 , Andrew H. Baker 3 , Claudio Napoli 4,51 Bristol Heart Institute, University of Bristol, Bristol, BS2 8H, UNITED KINGDOM; 2 Department of PharmacologicalSciences, University of Salerno, 84084 Italy; 3 Division of Cardiovascular and Medical Sciences, University of Glasgow,Western Infirmary, Glasgow G11 6NT, UNITED KINGDOM; 4 Departments of Medicine and Clinical Pathology,University of Naples, Naples 80131, Italy; 5 Department of Medicine-0682, University of California San Diego, CA92093,USASJ George and F de Nigris contributed equally to this review.__________________________________________________________________________________*Correspondence: Claudio Napoli, MD, PhD, FACA, PO BOX 80131, Naples, Italy, e-mail: claunap@tin.itKey words: Atherosclerosis, gene therapy, adenoviruses, vascular diseases.Received: 2 July 2003; Accepted: 18 July 2003; electronically published: July 2003SummaryCurrently, successful pharmacological treatments are unavailable for many vascular diseases. Many patientsundergo surgical interventions and then present with recurrence of symptoms. Recently, gene therapy using bothnon-viral and viral delivery has emerged as a novel tool to treat patients with vascular diseases. Here we discuss therequirement to develop suitable gene delivery vectors for vascular diseases. Our expanding knowledge of thepathogenesis of vascular diseases has allowed the identification of several gene therapy strategies and manycandidate genes. Gene therapy using both gene knockout and gene overexpression has been considered. In preclinicalstudies, antisense and decoy oligonucleotides have been successfully employed to knockout the expression ofstimulatory genes such as cell cycle promoters and growth factors. Furthermore, overexpression of inhibitory genessuch as cell cycle inhibitors and nitric oxide and overexpression of genes to promote therapeutic angiogenesis havebeen shown potential in animal models. The progress of pre-clinical studies to treat vein graft failure, restenosis,myocardial and peripheral ischemia and hypertension and the development of clinical trials will be discussed.Despite the quite promising findings with clinical trials, particularly with therapeutic angiogenesis, improved genetransfer vectors and methods for safe long-term gene transfer are still required to bring gene therapy to clinicalpractice.I. IntroductionGene therapeutics have been proposed as a potentialnovel therapy for a host of diverse disease that encompassacquired conditions such as cancer, cardiovascular diseaseand arthritis as well as monogenic diseases through genereplacement strategies. In theory the concept has seemedrelatively simply; in practice, however, gene therapy isextremely complex, both technically and clinically. Itrequires a multifaceted approach involving identificationof suitable therapeutic gene(s), identification of a suitablegene delivery vehicle together with the availability ofsatisfactory pre-clinical models in which to evaluate thepotential benefit of the gene therapeutic approach,particularly against alternative pharmacological therapies,if available. The issue of long-term safety of gene therapyapproaches is still unclear. To date, major progress at theclinical level has been made in defined areas, particularcancer, cystic fibrosis, haemophilia and some vasculardiseases. These advances have not been without majordrawbacks. Tragic events involving high dose delivery ofadenoviral vectors to a patient on a gene therapy clinicaltrial for ornithine transcarbamylase (OTC) deficiency aswell as the evolution of leukaemia in severe combinedimmunodeficiency (SCID) patients involving retroviralvectors (Cavazzana-C et al, 2000; Somia et al, 2000; Fox2003) have highlighted safety issues relating to genedelivery vectors. In vascular diseases, successful genetherapy will require the following:Identification of the optimal transgene cassette.Expression systems vary considerably for different genetherapy applications. Traditionally strong viral promotershave been used to provide maximal levels of expression ina multitude of recipient cell types. However, it isbecoming increasing important to supply expressionselectively to individual cell types or in a regulatedmanner through inducible promoters (such as tetracyclinsystem (Gossen et al, 1992; Vigna et al, 2002) thuscircumventing potentially deleterious effects of transgeneexpression in non-target cell types. Additionally, viralpromoters, particularly the cytomegalovirus immediate135

George et al: Gene therapy for vascular diseasesearly promoter (CMV IEP) is prone to host-mediatedsilencing in vivo (De Geest et al, 2000) leading to a shutdown in transgene expression, an effect not observed withcell-specific promoters. Further optimisation of expressioncassettes can be made through incorporation of introns andenhancers to elevate promoter activity as well as posttranscriptionalmodifications including the Woodchuckpost-transcriptional regulatory element (WPRE) which isthought to act through promoting mRNA stability (Loeb etal, 1999; Zufferey et al, 1999).Optimisation and evaluation of the gene deliveryvehicle. At present the repertoire of gene delivery vectorsavailable for human gene therapy is limited. Traditionally,non-viral vectors such as naked DNA and liposome DNAcomplexes provide low efficiency gene transfer and arerestricted to the delivery of highly potent biologicalagents, such as angiogenic gene therapy (see below).Improvements in the efficiency of non-viral vectors, suchas inclusion of targeting peptides into DNA liposomecomplexes (Hart et al, 1997; Parkes et al, 2002) have beenrealised but are still someway from the efficiency of viralvectors. Certain viruses, by virtue of evolution, infecthuman cells with high efficiency resulting in high potencygene transfer and overexpression of candidate therapeuticgenes. For gene delivery to vascular tissues the currentarmoury of viral vectors includes adenoviruses (Ad),adeno-associated viruses (AAV), lentiviruses and Semlikiforest viruses.Efficient modalities for gene delivery to the targetsite. Certain vascular diseases, such as vein grafting areoptimal for gene therapyapy since the target tissue (i.e. thevein to be grafted) is harvested and is available ex vivo forgene delivery prior to grafting within a clinically relevanttime window (approximately 30 minutes). This enablesdelivery of genes in a safe and efficient manner (Baker etal, 1997; Tamirisa et al, 2002). Due to the short timeframe, however, efficient vectors are required. Adenoviralvectors have proven particularly suited for this application(Channon et al, 1997; George et al, 2000; Tamirisa et al,2002). Conversely, gene delivery to blood vessels in vivorequires the use of devices to allow localised in vivo genedelivery. Specific catheter systems have been developedand utilised with high efficiency for post-angioplasty andin-stent restenosis in a variety of animal species and bloodvessels (French et al, 1994; Klugherz et al, 2000, 2002).Additionally, local delivery technology has been appliedfor gene therapyapy aimed at the myocardium. Differentapplications, such as atherosclerosis or hypertensionrequire alternate delivery systems and often rely onintravenous vehicle administration.Together, a combined approach to optimise the geneexpression system, the delivery vehicle and the route ofdelivery are required for successful gene therapy. Anumber of key areas within vascular diseases havesuccessfully exploited this and advanced to clinical trialswhile other areas have been severely limited due todeficiencies in one or more of the above requirements.Here, we discuss a number of these applications.There is no doubt that gene therapy may offeradvantages above traditional pharmacological therapies incertain respects. Delivery of gene can be achieved locallyin the vasculature thereby increasing the selectivity and,potentially, the safety. This would be particularlyimportant when the therapy may have an adverse effect ifcontact to non-target tissue in vivo occurred. Since manyof the strategies that have been designed to be effective invascular disease may be deleterious if exposed to nontargettissue, this advantage becomes very important. Forexample, in development of gene therapy for vein graftfailure (see later) pro-apoptotic genes are highly effectivebut clearly their expression in other tissues such as theliver, may be detrimental. Likewise, in restenosis postangioplasty(cytotoxic or cytostatic strategies) andangiogenesis gene therapy can be delivered locally and is apre-requisite for clinical translation. A second (and equallyimportant) advantage of gene therapy might be therequirement for only a single administration compared tothe requirement for multiple administrations ofconventional drugs, often daily for the lifetime of thepatient. Again, this depends largely on the application andis to date unproven. Evidence suggests that beneficialeffects of gene therapy for hypertension, vein grafting andrestenosis can be elicited in the long term from singleadministrations (see later). This provides ample preclinicalevidence to support these concepts.In the following review, we discuss gene therapy forsome vascular diseases and its progression in differentexperimental and clinical applications.II. Local gene delivery to the vesselwallIt has been known for over a decade that genedelivery to the vessel wall can result in alterations in cellbehaviour (Nabel et al, 1993 a, b, c) thereby initiating aplethora of studies that have evaluated and optimised genedelivery to the vessel wall. Although the first studiesrevealed that non-viral gene delivery could lead tophenotypic modulation of cell behaviour, it soon becameclear that adenoviral vectors provided the most efficientmeans to achieve high-level gene delivery to the vesselwall in vivo (Lemarchand et al, 1993; French et al, 1994).Pioneering studies by Lemerchand and colleagues (1993)and French et al, (1994) showed that local exposure ofhigh titre adenoviral vectors to normal and diseased bloodvessels in vivo led to high-level transduction, in sheep andrabbit models, respectively. Catheter systems were rapidlydeveloped and optimised for gene delivery postangioplastyresulting in transgene expression throughoutthe vessel wall in a geographical localisation defined bythe mode of vector delivery by the catheter utilised. Thisinitiated a host of studies and led to the use of adenoviralvectors as the most commonly used modality throughwhich to deliver genes to the vessel wall in vivo.However, this is not without limitations since adenoviralmediatedgene delivery was found to evoke aninflammatory response in the vessel wall leading totoxicity and endothelial cell activation (Newman et al,1995). Furthermore, the use of these first-generation Advectors only resulted in transient gene expression lasting7-14 days. Unlike other tissues, second generation vectors(that contained modifications of the Ad genome to reduce136

Gene Therapy and Molecular Biology Vol 7, page 137expression of Ad-related genes) did not lead to sustainedtransgene expression in the vessel wall in vivo (Engelhardtet al, 1994; Wen et al, 2000). Other vector systems haverecently been tested including improved non-viral systemssuch as peptide-targeted DNA/liposome complexes (Hartet al, 1997; Parkes et al, 2002), HVJ-modified liposomes(Morishita et al, 1995; Von Der Leyen et al, 1995; Dzau etal, 1996) and ultrasound-enhanced systems (Lawrie et al,1999; Taniyama et al, 2002). Likewise, other viral vectors(including adeno-associated viruses (Maeda et al, 1997;Richter et al, 2000), Semliki-forest viruses (Lundstrom etal, 2001) and lentiviruses (Dishart et al, 2003) have beenutilised. Modified viral systems in particular provideopportunities to modify the longevity of transgeneexpression as well as the principle cell type transduced. Asan example, adeno-associated viruses (AAV) transducesmooth muscle cells in the vessel wall, even in thepresence of an intact endothelial layer (Richter et al,2000). This is in direct comparison to Ad-mediate deliverysince endothelial transduction is high when an intactendothelium is present and represents a barrier totransduction (Lemarchand et al, 1993). This finding mayin part be due to different physical sizes of Ad and AAVand due to different vector tropisms of each, which isdictated by host expression of viral receptors and coreceptors(Wickham et al, 1993; Bergelson et al, 1997;Tomko et al, 1997; Summerford et al, 1998; Qing et al,1999; Summerford et al, 1999; Dishart et al, 2003). Hence,these systems have provided researchers with a diverserange of vectors through which to evaluate the phenotypiceffects of overexpression of candidate therapeutic genes inthe vessel wall in vivo.III. Gene therapy and vein graftfailureThe failure of vein bypass grafts in the coronary orlower extremity circulation is a common clinicaloccurrence that incurs significant morbidity and mortality.Despite the very common use of saphenous vein grafts totreat coronary and lower extremity occlusions the failurerate is extremely high, approximately 50% and 70% ofvein grafts fail within 5-10 years after surgery,respectively (Angelini 1992; Conte et al, 2001). To date,pharmacological approaches to prolong vein graft patencyhave produced very limited results. Consequently, geneticapproaches to modulate bypass grafts are actively beingstudied both in vitro and in vivo and are progressing toclinical trials. Vein grafts are uniquely amenable tointraoperative genetic modification because of the abilityto manipulate the tissue ex vivo with controlledconditions. We will describe how both geneoverexpression and gene blockade strategies have beentested, and how the latter is now in clinical trials (see alsoFigure 1 for schematic summary of gene therapystrategies).AngioplastyIntimal proliferationConstrictive remodellingRestenosisStent PlacementIntimal proliferationEarly FailureThrombosisVein Graft FailureLate FailureIntimal proliferationConstrictive remodellingGene Therapy StrategiesAnti-VSMC proliferation:-Cytostatic: cell cycleinhibitors, antisense cell cyclegenes& growth factors-Cytotoxic: tk, p53,Anti-thrombotic: uPA, tPA,NORe-endothelialization:VEGF*Anti-VSMC migration &matrix remodelling: TIMPsGene Therapy StrategiesAnti-VSMC proliferation:-Cytostatic: cell cycleinhibitors, antisense cell cyclegenes& growth factorsCytotoxic: tk, p53,Anti-thrombotic: uPA,tPA, NORe-endothelialization:VEGFGene Therapy StrategiesAnti-thrombotic: nonetestedRe-endothelialization: C-type natriuretic peptideGene Therapy StrategiesAnti-VSMC proliferation:-Cytostatic: cell cycleinhibitors, antisense cell cyclegenes, transcription factors(E2F)* & growth factorsRe-endothelialization:VEGF*Anti-VSMC migration &matrix remodelling: TIMPsFigure 1: Gene therapy strategies for the treatment of restenosis and vein graft failure. Many preclinical studies have been utilised todetermine the potential of these various strategies * indicates those that have progressed to clinical trials.137

George et al: Gene therapy for vascular diseasesA. Biological processes involved inrestenosis and molecular targets in vein graftfailureA complex series of biological events is initiated inthe vein immediately after implantation into the arterialcirculation. Within the first few days after implantationmany vein grafts fail due to thrombosis, stimulated byendothelial injury (Bryan et al, 1994). Furthermore, in thefirst 24 hours vein grafts undergo a period of ischemiafollowed by reperfusion, which leads to the generation ofsuperoxide and other reactive oxygen species that triggerscytoxicity of endothelial and smooth muscle cells (Shi etal, 2001; West et al, 2001). The grafted vein is thentargeted by an acute inflammatory response involvingneutrophil and mononuclear cell recruitment and oxidativestress persists (West et al, 2001). In the first week afterimplantation matrix remodelling and migration of smoothmuscle cells into the intima takes place; once in the intimathe smooth muscle cells proliferate contributing further tothe intimal thickening (Newby et al, 1996). Each of theseprocesses offers a set of potential molecular targets forgene therapyapy.B. Anti-thrombotic and accelerated reendothelializationstrategiesAnti-thrombotic strategies have been investigated asa relevant target for gene transfer to reduce thrombosis invarious models of arterial injury and thrombosisformation. Thrombosis is dramatically reduced usingnatural anti-thrombotic, anti-aggregatory, and fibrinolyticpathways such as overexpression of thrombomodulin(Waugh et al, 1999), tissue factor pathway inhibitor(Nishida et al, 1999; Zoldhelyi et al, 2000), CD39(Gangadharan et al, 2001) and tissue plasminogenactivator (Waugh et al, 1999). Despite their provensuccess, the potential of these anti-thrombotic strategieshas not been widely tested in vein graft models perhapsdue to the availability of pharmacological treatments.However, acceleration of re-endothelialization by genetransfer of C-type natriuretic peptide in rabbit jugular veingrafts reduced both thrombosis and intimal thickening(Ohno et al, 2002). This illustrates that promoting reendothelializationand reducing thrombosis is a promisingstrategy to circumvent vein graft failure.C. Anti-proliferative strategyIn an attempt to inhibit VSMC proliferation in veingrafts both overexpression of cell cycle inhibitory proteinsand inhibition of cell cycle promontory genes usingantisense has been investigated in arterial injury and veingraft models. In fact it is thought that strategies targetingmultiple cell cycle genes offer greater potential than singletargets. Rabbit vein grafts treated simultaneously withantisense oligonucleotides to proliferating cell nuclearantigen (PCNA) and cell division cycle-2 kinase showedreduced intimal thickening and diet inducedatherosclerosis (Mann et al, 1995).Recently, transfection of cis-element double-strandedoligonucleotides (decoy ODNs) has been reported as anew powerful tool in a new class of anti-gene strategiesfor gene therapyapy. Transfection of double-strandedODN corresponding to the cis sequence will result inattenuation of the authentic cis-trans interaction, leading toremoval of trans-factors from the endogenous cis-elementswith subsequent modulation of gene expression. A decoyto E2F, which induces the coordinated expression of anumber of critical cell cycle genes, including PCNA,cyclin-dependent kinase-1, cell division cycle-2 kinase, c-myc, c-myb, was used successfully. This E2F decoy ODNnot only almost completely inhibited intimal thickeningafter balloon injury of the rat carotid at two weeks afterinjury (Morishita et al, 1995), but sustained inhibition wasobserved after eight weeks. This inhibition of intimalthickening was also observed using a porcine coronaryartery model (Nakamura et al, 2002). Furthermore, asingle intraoperative pressure-mediated delivery of E2Fdecoy effectively provided vein grafts with long-term (upto 6 months) resistance to intimal thickening andatherosclerosis (Ehsan et al, 2001). Interestingly, it hasbeen demonstrated that although E2F decoy ODNtreatment of vascular grafts inhibits VSMC proliferationand activation, it spares the endothelium, thereby allowingnormal endothelial healing (Ehsan et al, 2002). A clinicaltrial (PREVENT) using intraoperative delivery of E2Fdecoy ODN to infrainguinal arterial bypass graftsdemonstrated fewer graft occlusions, revisions, or criticalstenoses in the E2F-treated group (Mann et al, 1999).Recently, a corporate-sponsored (Corgentech, Inc, PaloAlto, Calif) phase II trial of E2F decoy treatment ofcoronary vein grafts was completed (SoRelle 2001). Thisstudy, which involved 200 patients revealed a 30%reduction in critical stenosis and has formed the basis fordesign of a phase III trial in coronary bypass grafting.Furthermore, on the basis of this combination ofpreclinical and phase I/II clinical data, a phase III trial ofE2F decoy ODN for the prevention of lower extremityvein graft failure involving 1400 patients was initiated inDecember 2001.D. Pro-apoptotic strategyIn addition to the above-mentioned cytostaticapproaches, cytotoxic strategies have also beenconsidered. Delivery of TIMP-3, which in addition toinhibiting MMP activity and VSMC migration promotesVSMC apoptosis significantly reduced intimal thickeningin a porcine vein graft model (George et al, 2000).Adenoviral delivery of wild type p53 which promotesVSMC apoptosis has also been studied in humansaphenous vein in vitro studies (George et al, 2001).Induction of VSMC apoptosis by overexpression of p53,without a detectable reduction in VSMC proliferation, ledto a significant reduction, >70%, in intimal thickening(George et al, 2001). Studies using a porcine arteriovenousbypass model are currently been underway todetermine if this cytostatic strategy reduces intimalthickening in vivo. Despite initial concerns, this proapoptoticstrategy with TIMP-3 and p53 did not lead to a138

Gene Therapy and Molecular Biology Vol 7, page 139loss of VSMC density or thinning of the graft wall thatmay lead to aneurysm (George et al, 2000, 2001).E. Anti-migration/matrix remodellingCell migration is critical to intimal thickening andrequires remodelling of the matrix by proteolytic enzymessuch as matrix-degrading metalloproteinases (MMPs) andplasmin. The tissue inhibitors of matrix-degradingmetalloproteinases (TIMPs) regulate the proteolyticactivity of MMPs whilst the balance of plasminogenactivators and plasminogen activator inhibitor-1 (PAI-1)regulate plasmin. Increased MMP activity has beendemonstrated both in vitro (George et al, 1997) and in vivo(Southgate et al, 1999) models of vein graft failure. Localoverexpression of TIMPs (1, 2 and 3) reduced intimalthickening in a human in vitro model of vein graft failure(George et al, 1998a,b, 2000). Furthermore, ex-vivodelivery of TIMP-3 gene reduced MMP activity andintimal thickening in a porcine vein graft model (George etal, 2000), (Figure 2). Using the recently establishedmouse model of vein grafting the potential of gene therapyof TIMPs was further illustrated (Hu et al, 2001).Inhibition of plasminogen activators also inhibits intimalthickening in a human in vitro model of vein graft failure(Quax et al, 1997). Intimal thickening after balloon injuryof the rat carotid was reduced by 35% at 4 weeks afteradenoviral delivery of a hybrid protein which consists ofthe amino-terminal fragment of urokinase plasminogenactivator linked to bovine pancreas trypsin inhibitor, apotent inhibitor of plasmin (Lamfers et al, 2001). Genetransfer of TIMPs has not been used yet in adversingcerebral ischemia (Napoli, 2002).F. Anti-ischemia/reperfusion, oxidativestress, inflammationMolecular therapies targeted at scavenging the excessof reactive oxygen species generated locally or protectingresident cells from their downstream effects may be usefulin the prevention of vein graft failure. Gene therapy usingnaturally occurring cytoprotective and anti-oxidantmechanisms including heat shock protein-70 (Jayakumaret al, 2000), and scavenging enzymes such as catalase(Danel et al, 1998), superoxide dismutase (Li et al, 2001),and heme oxygenase-1 (Yang et al, 1999) have provenefficacy in models of arterial and lung injury and cardiacreperfusion but to date have not been used in vein grafts.Similarly, gene transfer of TIMPs has not been used incerebral ischemia (Napoli, 2002). Although pre-treatingthe vein with anti-oxidant gene therapy is an attractivestrategy it may be difficult in practice because of theimmediate onset of reperfusion after implantation and thetime delay before adequate transgene expression.However, antioxidant gene therapy might be advantageousfor later stages of graft healing, as oxidative stress is aconsequence of inflammation (West et al, 2001). Possibleanti-inflammatory strategies include overexpression ofnitric oxide synthase (NOS), soluble adhesion moleculesand CC-chemokine blockade. By far the most progress hasbeen made with NOS overexpression, probably since italso inhibits thrombosis formation and VSMCproliferation (Cable et al, 1997). Ex vivo gene transfer ofendothelial (e)NOS to canine ipsilateral femoral veingrafts (Matsumoto et al, 1998) and inducible (i)NOS toporcine jugular (Kibbe et al, 2001) and intraoperative genetransfer of neuronal (n)NOS to jugular vein grafts inrabbits (West 2001) significantly reduced (30% to 50%)intimal thickening. However, only in the latter study was areduction in inflammation observed. A current clinicaltrial (Cardion, Inc, Cambridge, Mass) is examining theeffects of liposome-mediated iNOS gene transfer tocoronary arteries after angioplasty for the prevention ofrestensosis but no such trials are currently examining thepotential for prevention of vein graft failure. Despitedemonstration of the ability to overexpress a soluble formof the vascular adhesion molecule in vein grafts andhighlighting the potential for reducing vein graft failure(Chen et al, 1994), its efficacy has not been demonstrated.Furthermore, the ability of overexpression of 35K, a CCchemokineinactivator, to inhibit inflammation has onlybeen demonstrated in the peritoneum of mice (Bursill et al,2003).Figure 2: Adenoviral-mediated gene transfer of TIMP-3 reduced intimal thickening in vein grafts. Transverse sections stained for α-smooth muscle cell actin illustrate that intimal thickening was dramatically reduced in porcine arterio-venous vein grafts at one month byAd-mediated over-expression of TIMP-3 compared to controls (AdlacZ). White dotted line indicates the intimal/medial boundary.139

George et al: Gene therapy for vascular diseasesIV. Gene therapy and restenosisTreatment of symptomatic coronary arteryatherosclerotic plaques by angioplasty leads to vascularresponses including intimal thickening and constrictiveremodelling causing restenosis in approximately 30% ofinitially successfully treated patients. Although stentsprevent constrictive vascular remodelling, they inducevascular injury eventually leading to intimal thickeningand thereby restenosis. Gene therapy has been perceivedas attractive to treat restenosis as it can be deliveredlocally and appears to be able to treat excessive vascularcell proliferation.To date, a number of small (rat, mice) or large sizeanimal modes (rabbit, pig) have been used to evaluate thepotential of many gene therapy approaches for restenosis.The gene therapy strategies for treatment of restenosis aresummarized below and also in Figure 1. However, despitethe successful use of gene therapy to treat animalrestenosis by various approaches, application of genetherapy to prevent restenosis in man has only been carriedout using a re-endothelialization strategy with VEGF.Before further clinical trials are initiated a betterunderstanding of vascular biology, gene expression, vectordesign, and catheter-tissue interactions is required. It mustalso be mentioned that the efficacy of sirolimus(rapamycin) for the treatment of in-stent restenosis(Serruys et al, 2002; Sousa et al, 2003) has reduced theimpetus for designing gene therapy for in-stent restenosis.A. Biological processes involved inrestenosis and molecular targets in restenosisThe two major components that lead to restenosis areintimal thickening and negative (constrictive) remodelling.Intimal thickening following experimental injury involvesa combination of many processes, including VSMC andadventitial cell migration, proliferation, and matrixdeposition. Negative remodelling, which only occurs afterangioplasty and not after stent placement may also arisefrom many processes, including VSMC apoptosis, medialand adventitial fibrosis and matrix remodelling. However,restenosis, both in the absence and in the presence ofstents, is primarily due to VSMC accumulation. Sincemural thrombi may aggravate restenosis by contributingdirectly to cell proliferation, anti-thrombotic strategieshave received attention. Finally, strategies that acceleratere-endothelialization of the injury artery have beeninvestigated.B. Inhibition of VSMC proliferationCytotoxic strategies have been tested based on theexpression of enzymes capable of converting nucleosideanalogues into toxic metabolites that impair DNAreplication and consequently cause death of transducedcells entering S phase. Adenoviral delivery of thymidinekinase (tk), a gene from herpes simplex virus (HSV),followed by ganciclovir treatment led to death of tkexpressingcells and reduced intimal thickening afterinjury of rat and rabbit arteries (Guzman et al, 1994;Simari et al, 1996). Similarly, expression of cytosinedeaminase in the presence of 5-fluorocytosine caused a45% reduction of stenosis (Harrell et al, 1997).Endogenous inducers of cell death have also been utilized.Delivery of the tumour suppressor p53 to injured ratcarotid arteries reduced intimal thickening (Yonemitsu etal, 1998), as did gene transfer of FasL (Luo et al, 1999).Some caution has been applied to the use of cytotoxicgene therapy for restenosis, since VSMC viability isessential for the integrity of the lesion, particularly thefibrous cap, and thereby the stability of atheroscleroticplaques. In addition, promotion of apoptosis in injuredvessels may increase intimal thickening, sinceoverexpression of fortilin, a recently characterised,negative regulator of apoptosis reduced intimal thickeningin injured rat arteries (Tulis et al, 2003).It has been well documented that cytostatic geneticstrategies using antisense oligonucleotides (ODN), decoyODN and gene transfer of cell cycle inhibitory genes (Li etal, 1999) limit VSMC proliferation and inhibit intimalthickening following experimental injury. Despiteencouraging results using antisense ODN to immediateearly genes such as c-myb (Simons et al, 1992) and c-myc(Shi et al, 1994) and promoters of cell cycle such as cyclinB and CDK-2 (Morishita et al, 1994), where intimalthickening was inhibited between 40 and 84% to in rat andin some cases also porcine injured arteries some years ago,this strategy appears to have made little progress recently.This is despite the observation that co-transfection ofcombinations of these antisense resulted in furtherinhibition (Morishita et al, 1994). Transfer ofretinoblastoma protein (Rb) to restrict the cell cycle, intorat and porcine injured arteries prevented intimalthickening (Chang et al, 1995). Similarly, overexpressionof the CDK inhibitors p21 and p27 resulted in reduction ofintimal thickening both in rat and porcine injured arteries(Chang et al, 1995; Yang et al, 1996; Chen et al, 1997).Furthermore, overexpression of a mutated form of p21 wasable to reduce restenosis in hypercholesterolemic mice byenhancing vascular apoptosis and reducing VSMCproliferation (Condorelli et al, 2001). A further strategythat has been examined is the inhibition of signallingmolecules. H-ras, a key protein in signal transduction,mediates mitogenic signals, therefore blocking this earlysignal transduction. Application of an adenoviral dominantnegative H-ras and Gβγ-binding peptide affecteddownstream signalling events and reduced intimalthickening by 70-80% (Ueno et al, 1997; Iaccarino et al,1999). Targeting of transcription factors by gene therapy isalso a strategy of interest. Inhibition of NFκB and E2F,cytoplasmic transcription factor using antisense ODNs inballoon-injured rat carotid arteries reduced intimalthickening by approximately 70% (Autieri et al, 1995;Morishita et al, 1995). Overexpression of the growth arresthomeobox gene (GAX) reduced intimal thickening by 50-70% in rat and rabbit injury models (Maillard et al, 1997;Smith et al, 1997). Although the use of transcriptionfactors as targets for gene therapyapy in restenosisappeared promising, it should be noted that thesetranscription factors are also involved in severalmechanisms regulating vascular wall homeostasis.140

Gene Therapy and Molecular Biology Vol 7, page 141Control of VSMC proliferation has also beenattempted by inhibition of growth factor expression andoverexpression of inhibitory growth factors and cytokines.Delivery of basic fibroblast growth factor (bFGF) (Hannaet al, 1997) as well as platelet-derived growth factor-β(PDGF-β) (Deguchi et al, 1999) antisense ODN and TGFβribozyme ODN (Yamamoto et al, 2000) inhibitedintimal thickening by 60-90% in injured rat carotidarteries. Similarly, adenoviral delivery of the extracellularregion of the PDGF-β receptor and of endovascularPDGF-β receptor antisense ODN reduced intimalthickening in injured rat arteries (Sirois et al, 1997;Noiseux et al, 2000). Activin, a TGF-β-like factor thatinduces a contractile phenotype in VSMCs, reducedintimal thickening by more that 70% in injured mousefemoral arteries (Engelse et al, 2002). The inhibitorycytokine interferon-g delivery by Ad-mediated genetherapy reduced intimal thickening in a porcine model ofarterial injury (Stephan et al, 1997).C. Cell migration and matrix remodellingConstrictive (negative) remodelling plays a veryimportant in human restenosis particularly in the absenceof a stent (Mintz et al, 1996), therefore gene therapystrategies aimed at reducing intimal thickening alone areunlikely to be successful in humans following angioplasty.Post injury intimal thickening is also reliant on VSMCmigration, which requires remodelling of the extracellularmatrix that surrounds the VSMC. Adenoviral gene transferof tissue inhibitor of metalloproteinase-1 (TIMP-1) andTIMP-2 reduced intimal thickening (Cheng et al, 1998;Furman et al, 2002). A combination of anti-proliferativeand anti-migratory approaches may therefore be useful.D. Anti-thrombotic strategyA number of studies have focused on seeding stentswith genetically modified endothelial cells with increasedfibrinolytic of anticoagulant activity (Dichek et al, 1989,1996; Dunn et al, 1996). Although seeding stented vesselswith endothelial cells overexpressing tPA and uPAproduced anti-thrombotic activity (Dichek et al, 1996),overexpression of tPA was associated with increaseddetachment of seeded cells (Dunn et al, 1996).Another strategy to prevent thrombosis as well asintimal thickening is to inhibit platelet activation oraggregation or to increase nitric oxide (NO). NO isvasoprotective by inhibiting platelet and leukocyteadhesion, inhibiting VSMC proliferation and migrationand promoting endothelial cell survival and proliferation(Li et al, 1999); therefore, nitric oxide synthase (NOS) thatincreases NO production was proposed as a suitablecandidate to treat restenosis. Delivery of endothelial(e)NOS by non-viral methods (von der Leyen et al, 1995)and adenoviruses (Chen et al, 1998; Janssen et al, 1998;Varenne et al, 1998) reduced intimal thickening by 37-70% in rat and pig injured arteries. Interestingly,adenoviral delivery of inducible (i)NOS by adenovirusesto rat injured arteries almost completely (95%) inhibitedintimal thickening, whilst reduced it by only 50% inporcine injured arteries (Shears et al, 1998), illustratingthat the degree of response differs greatly betweendifferent animal models. Furthermore, administration ofthe iNOS Ad could not mediate regression of establishedintimal thickening.D. Re-endothelializationAs regeneration of the endothelium is associatedwith reduction in thrombotic and proliferative processes inthe vessel wall it has been seen as a potential strategy ofgene therapy for restenosis. Local intravascular andextravascular expression of vascular endothelial growthfactor (VEGF), a potent endothelium specific angiogenicfactor, using plasmid DNA accelerated reendothelializationand decreased intimal thickening afterarterial injury in rabbit models (Asahara et al, 1996;Laitinen et al, 2000), and reduced in-stent restenosis by50% (Van Belle et al, 1997).The feasibility of this approach was tested in a smallclinical trial, in which VEGF plasmid/liposome genetransfer after angioplasty was seen to be safe and welltolerated (Laitinen et al, 2000). A recently published largerclinical trial was designed to test the feasibility,tolerability and efficacy of VEGF gene therapy to preventrestenosis after stenting (Hedman et al, 2003). The overallrestenosis rate in this study was surprisingly low (6%),virtually precluding the detection of a difference amongtreatments. Nevertheless, the results establish feasibilityand provide safety data on the used of naked DNA and Adto express VEGF. This strategy is perceived attractive as itis trying to mimic nature’s inhibitory strategy to limitintimal thickening, but we await clinical evidence of itssuccess. The use of VEGF is also attractive as it should beendothelial cell specific; however, there are safetyconcerns in respect to tumour growth as VEGF isinvolved in induction and progression (Huang et al, 2003).V. Gene therapy for hypertensionGene therapy for essential hypertension represents isan enormous challenge due to the complex polygenic traitthat underlies human essential hypertension. Gene therapyis however attractive since it offers the opportunity to treatthe disease with a single administration rather than dailydrug regimens. Essential hypertension is associated withendothelial dysfunction and contributes significantly tocardiovascular risk. Gene therapy would, therefore, targetspecific systems with the explicit aim of lowering bloodpressure and reducing end organ damage. Unlike otherdisease targets discussed above, gene therapy forhypertension requires the use of strategies to provide longtermeffects on blood pressure. These have includedantisense/ribozyme strategies to block systems thatregulate blood pressure as well as vasodilator strategiesusing overexpression of pro-vasodilator genes.Preclinical studies on gene therapy for hypertensionhave taken two main approaches (Phillips, 2002). First,extensive studies on gene transfer to increase vasodilatorproteins (kallikrein, atrial natriuretic peptide,adrenomedullin, and endothelin nitric oxide synthase)141

George et al: Gene therapy for vascular diseaseshave been carried out in different rat models (Lin et al,1995; Chao et al, 1996, 1997; Lin et al, 1997; Chao et al,1998 a, b; Yayama et al, 1998; Alexander et al, 1999;Dobrzynski et al, 1999; Lin et al, 1999; Alexander et al,2000; Dobrzynski et al, 2000; Wolf et al, 2000; Zhang etal, 2000; Wang et al, 2001; Emanueli et al, 2002). Usingthese approaches, blood pressure can be lowered for 3-12weeks with the expression of these genes. Second, anantisense approach, which began by targetingangiotensinogen and the angiotensin type 1 (AT1)receptor, has now been tested independently by severaldifferent groups in multiple models of hypertension(Katovich et al, 1999; Tang et al, 1999; Wang et al, 2000;Kimura et al, 2001). Other genes targeted include the β1-adrenoreceptor, TRH, angiotensin gene activatingelements, carboxypeptidase Y, c-fos, and CYP4A1(Gardon et al, 2000; Phillips, 2001; Tomita et al, 2002).There have been two methods of delivery antisense, shortODNs, and full-length DNA in viral vectors. All thestudies show a decrease in blood pressure lasting severaldays to weeks or months. ODNs are safe and particularnon-toxic. The decreased hypertension after systemicadeno-associated virus delivery antisense to AT1 receptorsin adult rodents for up to 6 months, may constitute a goodincentive for testing the antisense ODNs first and later theAAV (Kimura et al, 2001; Phillips 2001).Hypertension is also the presenting feature of someof these disorders, such as congenital adrenal diseases, andadrenal and pituitary tumors. Preclinical data indicate thatgene transfer to both the adrenal gland and the pituitary isnot only feasible but also quite efficient (Alesci et al,2002).A. Inhibition of vasoconstrictor genesThis has been achieved using antisenseoligonucleotides to block the renin-angiotensin system.For example, Wielbo et al (1996) used DNA/liposomescomplexes containing angiotensinogen antisense andlowered mean arterial pressure, angiotensinogen andangiotensin II levels in adult spontaneously hypertensiverats following systemic administration. These highlyeffective results are somewhat surprising when it isrealised that the in vivo uptake of DNA/liposomecomplexes into the vasculature and organs is very poorwhen delivery intravenously. Not surprisingly viral vectorsystems have also been engineered to deliver antisense.Using a retroviral system to deliver antisense against theangiotensin type-1 receptor to young (5 day old)hypertensive and normotensive animals, blood pressurewas significantly lowered selectively in the hypertensiveanimals (Lu et al, 1996). Interestingly, the effect of theantisense was sustained for 90 days while losartan had theexpected transient effect of less than 24 hours. This doeshighlight the clinical relevance of such technology toprovide sustained benefit compared to traditionalpharmacological regimens. However, in the light of recentclinical experience using retroviral vectors withdevelopment of leukaemia on phase I trial (Cavazzana etal, 2000), the use of retroviral vectors is unlikely to bedeveloped in this disease. Other studies have alsohighlighted the benefit of viral delivery of antisense(Wang C et al, 1995; Martens et al, 1998; Reaves et al,1999; Tang et al, 1999; Wang H et al, 1999).B. Vasodilator overexpressionThere are a number of candidate genes foroverexpression that may provide therapeutic benefit ofdifferent aspects of hypertension. These include kallikrein,adrenomedullin, nitric oxide synthase and superoxidedismutase. Kallikrein cleaves kininogen producing kininpeptide, which in turn stimulates the release of thevasodilators prostacyclin, endothelium-derivedhyperpolarising factor and nitric oxide. Based on thisprinciple, infusion of naked DNA expressing kallikreinreduced blood pressure for 6 weeks (Wang et al, 1995).Comparative studies showed that naked DNA plasmidsand adenoviral vectors both proved effective (Chao et al,1997). Kallikrein delivery using viruses has also beenestablished as an anti-hypertensive strategy in differentmodels demonstrating the potential benefit of this strategyand the potency of the transgene (Dobrzynski et al, 1999;Wolf et al, 2000).Adrenomedullin also causes vasodilation.Adenoviral-mediated overexpression of adrenomedullin inhypertensive rats led to a blood pressure drop of 41 mmHg 9 days after tail vein injection (Dobrzynski et al,2000). This lasted nearly 20 days. Again, proof of thisstrategy was realised when other studies gained similarfindings in different labs and models of hypertension(Zhang et al, 2000; Wang et al, 2001).Targeting endothelial dysfunction is highly attractivefor gene therapyapy. Endothelial dysfunction ischaracterised by reduced nitric oxide (NO)-mediatedvasodilation and a reduction in available NO. The loss ofNO leads to deleterious effects on platelet aggregation andadhesion, smooth muscle proliferation, inflammation andincreased oxidative stress in the vessel wall. Improving thebioavailability of NO, therefore, is a highly logicalstrategy to improve a number of key processes that areintegral to vessel wall homeostasis in order to reduceblood pressure. This can be achieved by increasing NOproduction itself through nitric oxide synthase (NOS) genedelivery or by preventing NO degradation by superoxidedismutase (SOD) gene transfer. A number of studies haveaddressed these issues. An early study established such aconcept by systemic delivery of naked DNA encoding theendothelial form of NOS (eNOS) with a significantreduction in blood pressure that lasted for at least 12weeks (Lin et al, 1997). Again, such effects with nakedDNA are astonishing since little uptake was achieved invivo and the majority was sequestered to the liver. It isimportant to note that targeting gene delivery to theendothelium is extremely difficult using currentlyavailable vector systems when the delivery mode isintravenously. The liver sequesters the vast majority of allcommonly used vector systems with relatively little uptakeby the endothelium itself. This has restricted studies tolocal applications of gene delivery to selected bloodvessels in vivo. Adenoviral delivery of eNOS or SOD3,142

Gene Therapy and Molecular Biology Vol 7, page 143but not SOD-1 or –2 are able to improve endothelialfunction in carotid arteries in the spontaneouslyhypertensive stroke-prone (SHRSP) rats (Alexander et al,1999, 2000; Fennell et al, 2002).VI. Therapeutic angiogenesisTherapeutic angiogenesis represents a novel strategyfor the treatment of vascular insufficiency. It is based onsupplementation with angiogenic growth factors toenhance native angiogenesis in critical myocardial orperipheral ischaemia. Angiogenic growth factors havebeen delivered both as protein and by way of gene transferand have demonstrated positive results (Yla-Herttuala etal, 2003). The recent insights in the molecular basis ofangiogenesis have resulted in great interest in the genetherapy field. However, because of the rapid evolution andenthusiasm in the field, angiogenic molecules have beentested without a complete understanding of theirmechanism of action. Among the angiogenic growthfactors used in pre-clinical studies, VEGF165 andVEGF121, FGF1, FGF2 and hepatocyte growth factor(HGF) have all shown significant improvement of nativeangiogenic response to ischemia, resulting in acceleratedrate of perfusion, (see reviews by Hammond et al, 2001)(Emanueli et al, 2001; Manninen et al, 2002). Besidesgrowth factors a number of other substances have beeninvestigated, such as human tissue kallikrein (Emanueli etal, 2001), angiopoietin (Shyu et al, 1998), leptin(Bouloumie et al, 1998) and thrombopoietin (Brizi et al,1999).Although difficulties have been encountered in thefield of gene therapy, great progress has been made in thefield of pro-angiogenic gene therapy. It has been suggestedthat this is because the long-term gene expression is notrequired for therapeutic vascular growth and the currentgene therapy vectors induce at least some physiologicalimprovement (Yla-Herttuala et al, 2003). Over 23 clinicaltrials have been initiated; approximately half are forperipheral disease and the other half for coronary heartdisease. The first set of clinical trials involved pioneeringattempts to overexpress VEGF165 with naked DNA (Isneret al, 1996; Baumgartner et al, 1998; Losordo et al, 1998)and adenoviruses (Rosengart et al, 1999). The secondphase of trials were small, uncontrolled trials using nakedDNA and adenoviruses to overexpress VEGF165 andVEGF121; many of these had positive results (Symes etal, 1999; Laitinen et al, 2000; Rajagopalan et al, 2001).Only recently, the third set of clinical trials has begun totest the potential of this gene therapy fully. Theserandomised, controlled and blinded trials have involvedlarger numbers of patients and defined primary andsecondary endpoints (Grines et al, 2002; Makinen et al,2002; Stewart et al, 2002; Hedman et al, 2003;Rajagopalan et al, 2003). Several of these have beenjudged positive according to primary and secondaryendpoints but it has been suggested that this may not betransferable to a clear-cut clinical benefit (Yla-Herttuala etal, 2003).Critically ischaemic lower limbs from diabetes thatare not suitable candidates for surgical endovascularapproaches may be amenable to gene therapy fortherapeutic angiogenesis. Diabetes impairs endogenousneovascularization of ischaemic tissues due to a reducedexpression of VEGF (Rivard et al, 1999) and HGF(Taniyama et al, 2001). Consequently Ad-mediatedoverexpression of VEGF and plasmid HGF restoredneovascularization in mouse and rat models of diabetes,respectively (Rivard et al, 1999; Taniyama et al, 2001).Enhanced angiogenesis by such strategies also improvesneuropathy both when growth factors including VEGF, aregiven alone (Rissanen et al, 2001) or in conjunction withthe prostacyclin synthase gene (Koike et al, 2003).Furthermore, a small clinical trial which included 6diabetic patients with critical leg ischaemia, observedneurologic improvement and therapeutic angiogenesisafter plasmid injections of VEGF165 in the muscles of theischaemic limb (Simovic et al, 2001). Inhibition ofangiogenesis may also have therapeutic potential for thetreatment of retinopathy, since lentiviral delivery ofangiostatin inhibited neovascularization in a murineproliferative retinopathy model (Igarashi et al, 2003).Although, this strategy has made great progress inthe last decade there are still some unresolved issues. Forexample is administration of a single angiogenic moleculesufficient? Will administration of VEGF lead to toxiceffects such as oedema? Will an angiogenic factor besuitable for myocardial and peripheral angiogenesis? Sincethe same adenoviral VEGF121 gave positive effects in themyocardium (Stewart et al, 2002) but failed in peripheralvascular disease (Rajagopalan et al, 2003), will VEGF beproven clinically benefial? Some caution has been cast onthe potential of VEGF gene therapy by the observationthat VEGF enhances atherosclerotic plaque progression inboth mice and rabbits (Celletti et al, 2001). Are otherVEGF homologues safer options? Increasedlymphogenesis and reduced oedema is observed withVEGFC and VEGFD (Yla-Herttuala et al, 2003).VII. Future directionsRecent advances through preclinical studies haveraised the profile of gene therapy in some vasculardiseases, particularly with respect to angiogenic genetherapy in the myocardium and peripheral vasculature aswell as in vein graft disease. These studies, presently inphase II, highlight the potential of the technology forrelieving symptoms of human vascular diseases.Despite the lack of dramatic cures, a decade ofclinical trials has provided important news about thestrengths and weaknesses of current vectors. Bothadenoviruses and liposomal vectors have been shown to beable to transduce transgenes in patients with a variety ofdisorders. From this work, it is now extremely clear thatthe expression is temporary and is associated with aninflammatory response. However, there are someimportant points to consider. First, with respect tomyocardial and peripheral vascular gene transfer clinicaltrials, these have been performed with single proangiogenicgenes with gene delivery using sub-optimalvector systems (e.g. naked DNA/adenoviral vectors). With143

George et al: Gene therapy for vascular diseasesrespect to the former, angiogenic gene therapy may besignificantly more therapeutic with respect to collateralvessel formation with a combination of therapeutic genesrather than single gene therapy strategies. With recentadvances in adenoviral vector technology [e.g. using"gutted" adenoviral vectors (Kochanek et al, 1996; Parkset al, 1996)] the cloning capacity required for such studiesis now available. Equally, the gutted adenoviral vectorsystems are less immunogenic in vivo and would allowlonger term overexpression of transgenes that in turn maypromote sustained angiogenic effects. It is known thatvascular cell uptake by these vectors (all based on serotype5 adenoviruses) is extremely poor in comparison to othercells, such as hepatocytes in the liver (Nicklin et al, 2001).Indeed, pre-clinical experiments have shown that localdelivery of adenoviruses serotype 5 vectors to thevasculature leads to virion dissemination, not only to theliver but also to testes and other organs posing additionalsafety concerns (Hiltunen et al, 2000; Baker, 2002).Given the limited ability of liposomes andadenoviruses to enable long-term gene expression, andgiven the poor in vivo performance of retroviruses, theAAV vectors are being developed. This virus is smallerthan the adenovirus and has a relatively low-capacity size.However, it allows for long-term gene expression (ie,months to years) with only minimal induction ofinflammation or antiviral immune responses. A betterunderstanding of the life cycle of this virus, along withimproved production techniques, has allowed investigatorsto conduct clinical trials with AAV in diseases such ashemophilia and cystic fibrosis (see http://www.wiley.co.uk/wileychi/genmed/clinical/). Preclinical data in miceinjected intramuscularly with an AAV-human alpha-1-antitrypsin (1AT) vector are encouraging (Xiao et al,1998; Phillips et al, 2002).To date, the major problem in gene therapy remainsthe relative inefficiency of current vectors. Currently, thisinefficiency, coupled with a relatively poor specificity ofmost vectors, requires the delivery of large doses ofvector. This is both expensive and more likely to lead toside effects. Pathophysiological questions still remainabout which and how many cells need to be transduced toobtain a clinical response. One new and very exciting areaof gene therapy that has not yet reached clinical trials isthe "gene correction" (Gamper et al, 2000; Metz et al,2002). It is possible to design oligonucleotides that bind toareas of single-nucleotide changes that are associated withabnormal functions and to catalyze corrections of thenucleotide errors. This concept clearly has beendemonstrated to work in cell cultures and in animalmodels, although the efficiency is still quite low. With thedevelopment of better oligonucleotides and improveddelivery methods, this approach will likely be tested firstin diseases such as hemophilia and 1AT.When it is considered that angiogenic gene therapyshould be highly localised due to potential side effects[including potentiation of atherosclerosis (Celletti et al,2001) and development of cancer (Lee et al, 2000)] othervector systems should now be considered. The choice ofpotential new vectors is broad and must be consideredwith caution and evaluated based on current knowledge ofexisting systems (de Nigris et al, 2003). Additionalevidence now suggests that the vast majority of AAVgenomes remain in a non-integrative capacity withininfected cells (Nakai et al, 2001; Schnepp et al, 2003)further supporting the safety of this vector system. Ofequal potential are adenoviral vectors originating fromdifferent serotypes. Previous pre-clinical data support ofthe notion that novel vector systems can be isolated for thecapacity to efficiently infect an individual tissue type(Havenga et al, 2001, 2002). For example, adenovirusesbased on serotype 16 have a high propensity to transduceboth endothelial cells and smooth muscle cells thanserotype 5 vectors (Havenga et al, 2001). Again, likeAAV-2, this may provide a system through which tooptimise gene delivery for defined gene therapeuticapplications. The use of cell selective promoters (tissuespecificexpression) to drive transgene expression will adda further level of selectivity to such systems. Thecombined use of vectors and immuno-suppressors may bealso reasonable.Gene therapy remains the key link between advancesin genetics and genomics and the translation of thisknowledge into useful outcomes for patients. Althoughprogress has been slower than hoped for, clear advancesare being made; gene therapy will probably find a numberof key therapeutic niches. Together, these modificationswill enhance the utility and safety of gene therapy astransition from pre-clinical to clinical gene therapyproceeds for the vascular system and its diseases.ReferencesAlesci S Chrousos GP, Pacak K (2002) Genomic medicine:exploring the basis of a new approach to endocrinehypertension. Annals of the New York Academy ofSciences 970, 177-92Alexander M.Y, Brosnan M, Hamilton C (2000) Gene transfer ofendothelial nitric oxide synthase but not Cu/Zn superoxidedismutase restores nitric oxide availability in the SHRSP.Cardiovasc Res 47, 609-617.Alexander M.Y, Brosnan M, Hamilton CA, Downie P, Devlin A,Dowell F, Martin W, Prentice H, O'Brien T, Dominiczak,A.F (1999) Gene transfer of endothelial nitric oxide synthaseimproves nitric oxide-dependent endothelial function in ahypertensive rat model. Cardiovasc Res 43, 798-807.Angelini G (1992) Saphenous vein graft failure: etiologicconsiderations and strategies for prevention. CurrentOpinion in Cardiology 7, 939-944.Asahara T, Chen D, Tsurumi Y, Kearney, M, Rossow, S, PasseriJ, Symes J.F, Isner J (1996) Accelerated restitution ofendothelial integrity and endothelium-dependent functionafter phVDGF165 transfer. Circulation 94, 3291-3302.Autieri M.V, Yue T, Ferstein G.Z, Ohlstein E (1995) Antisenseoligonucleotides to the p65 subunit of NF-kB inhibit humanvascular smooth muscle cell adherence and proliferation andprevent neointima formation in rat carotid arteries. BiochemBiophys Res Commun 213, 827-836.Baker A, Mehta D, George S, Angelini G (1997) Prevention ofvein graft failure: potential applications for gene therapyapy.Cardiovasc Res 35, 442-450.144

Gene Therapy and Molecular Biology Vol 7, page 145Baker A (2002) Development and use of gene transfer fortreatment of cardiovascular disease. J Card Surg. 17, 543-548.Baumgartner I, Pieczek, A, Manor O, Blair R, Kearney, M,Walsh, K, Isner J (1998) Constitutive expression of phVEGF(165) after intramuscular gene transfer promotes collateralvessel development in patients with critical limb ischemia.Circulation 97, 1114-1123.Bergelson J, Cunningham, JA, Droguett G, Kurt-Jones EA,Krithivas A, Hong J, Horwitz, M, Crowell R, Finberg R(1997) Isolation of a common receptor for coxsackie Bviruses and adenoviruses 2 and 5. Science 275, 1320-1323.Bouloumie A, Drexler H, Lafontan M, Busse R (1998) Leptin theproduct of Ob gene promotes angiogenesis. Circ Res 83,1059-1066.Brizi M, Battaglia E, Montrucchio G, Dentelli P, Del Sorbo L,Garbarino G, Pegoraro L, Camussi G (1999) Thrombopoietinstimulates endothelial cell motility and neoangiogenesis by aplatelet-activating factor-dependent mechanism. Circ Res84, 785-796.Bryan A, Angelini G (1994) The biology of saphenous vein graftocclusion: etiology and strategies for prevention. CurrentOpinion in Cardiology 9, 641-649.Bursill C, Cai S, Channon K, Greaves D (2003) Adenoviralmediateddelivery of a viral chemokine binding proteinblocks CC-chemokine activity in vitro and in vivo.Immunobiology 207, 187-196.Cable D, O'Brien T, Schaff H, Pompili V (1997) Recombinantveins: Gene Ther to augment nitric oxide production inbypass conduits. Circulation 96 (suppl 9), II173-II178.Cavazzana-C M, Hacein-Bey, S, Saint-Basile C, Gross F, YvonE, Nusbaum, P, Selz, F, Hue C, Certain S, Casanova J.-L,Bousso P, Deist F, Fischer A (2000) Gene Therapy of humansevere combined immunodeficiency (scid)-1x Disease.Science 288, 669 - 672.Celletti F, Waugh, J, Amabile P, Brendolan A, Hilfiker P, DakeM (2001) Vascular endothelial growth factor enhancesatherosclerotic plaque progression. Nat Med 7, 425-429.Chang M, Barr E, Lu M, Barton K, Leiden J (1995) Adenovirusmediatedover-expression of the cyclin cyclin-dependentkinase inhibitor p21 inhibits vascular smooth muscle cellproliferation and neointima formation in the rat carotid arterymodel of balloon angioplasty. J Clin Invest 96, 2260-2268.Chang M, Barr E, Seltzer J, Jiang Y, Nabel G, Nabel E,Parmacek, M, Leiden J (1995) Cytostatic gene-therapy forvascular proliferative disorders with a constitutively activeform of the retinoblastoma gene- product. Science 267(5197), 518-522.Channon K, Fulton G, Gray, J, Annex, B, Shetty, GA, BlazingMA, Peters K, Hagen P.O, George S.E (1997) Efficientadenoviral gene transfer to early venous bypass grafts:comparison with native vessels. Cardiovasc Res 35, 505-513.Chao J, Chao L (1997) Experimental kallikrein Gene Therapy inhypertension cardiovascular and renal diseases.Pharmacological Research 35, 517-522.Chao J, Jin L, Chen L, Chen V Chao L (1996) Systemic andportal vein delivery of human kallikrein gene reduces bloodpressure in hypertensive rats. Human Gene Ther 7, 901-911.Chao J, Yang Z.R, Jin L, Lin K.F, Chao L (1997) KallikreinGene Therapy in newborn and adult hypertensive rats.Canadian Journal of Physiology and Pharmacology 75,750-756.Chao J, Zhang J, Lin K.F, Chao L (1998a) Human kallikreingene delivery attenuates hypertension cardiac hypertrophy,and renal injury in Dahl salt-sensitive rats. Human GeneTher 9, 21-31.Chao J, Zhang J, Lin K.F, Chao L (1998b) Adenovirus-mediatedkallikrein gene delivery reverses salt-induced renal injury inDahl salt-sensitive rats. Kidney International 54, 1250-1260.Chen D, Krasinski K, Sylvester A, Chen J, Nisen P, Andres V(1997) Downregulation of cyclin-dependent kinase 2 activityand cyclin A promoter activity in vascular smooth musclecells by p27 (KIP1), an inhibitor of neointima formation inthe rat carotid artery. J Clin Invest 99 (10), 2334-2341.Chen L, Daum, G, Forough, R, Clowes M, Walter U, Clowes A(1998) Overexpression of human endothelial nitric oxidesynthase in rat vascular smooth muscle cells and in ballooninjuredcarotid artery. Circ Res 82, 862-870.Chen S, Wilson J, Muller D (1994) Adenovirus-meidated genetransfer of soluble vascular cell adhesion molecule to porcineinterposition vein grafts. Circulation 89, 1922-1928.Cheng L, Mantile G, Pauly, R, Nater C, Felici A, Monticone R,Bilato C, Gluzband YA, Crow, M Stetler-Stevenson W,Capogrossi M.C (1998) Adenovirus-mediated gene transferof the human tissue inhibitor of metalloproteinase-2 blocksvascular smooth muscle cell invasiveness in vitro andmodulates neointimal development in vivo. Circulation 98,2195-2201.Condorelli GL, Aycock J, Frati G, Napoli C (2001) Mutatedp21/WAF/CIP transgene overexpression reduces smoothmuscle cell proliferation, macrophage deposition, oxidationsensitivemechanisms and restenosis in hypercholesterolemicapolipoprotein-E knockout mice. FASEB Journal 15, 2162-2170.Conte M, Belkin M, GR U, Mannick, J, Whittemore A,Donaldson M (2001) Impact of increasing comorbidity oninfrainguinal reconstructions: a 20-year perspective. AnnSurg 233, 445-452.Danel C, Erzurum, S, Prayssac P, Eissa N, Crystal R, Herve P,Baudet B, Mazmanian M, Lemarchand P (1998) Genetherapy for oxidant injury-related diseases: adenovirusmediatedtransfer of superoxide dismutase and catalasecDNAs protects against hyperoxia but not against ischemiareperfusionlung injury. Hum Gene Ther 9 (10), 1487-1496.De Geest B, Van Linthout S, Lox, M, Collen D, Holvoet P(2000) Sustained expression of human apolipoprotein A-Iafter adenoviral gene transfer in C57BL/6 mice: role ofapolipoprotein A-I promoter apolipoprotein A-I introns andhuman apolipoprotein E enhancer. Hum Gene Ther 11, 101-12.Deguchi J, Namba T, Hamada H, Nakaoka T, Abe J, Sato O,Miyata T, Makuuchi M, Kurokawa K, Takuwa Y (1999)Targeting endogenous platelet-derived growth factor B-chainby adenovirus-mediated gene transfer potently inhibits invivo smooth muscle proliferation after arterial injury. GeneTher 6, 956-965.de Nigris F, Sica V, Herrmann J, Condorelli G, Chade A, TajanaG, Lerman A, Lerman LO, Napoli C (2003) c-Myconcoprotein: cell cycle-related events and new therapeuticchallenges in cancer and cardiovascular diseases. Cell Cycle2, 325-328.Dichek DA, Neville R.F, Zwiebel JA, Freeman S, Leon M.B,Anderson W.F (1989) Seeding of intravascular stents withgenetically engineered endothelial cells. Circulation 80,1347-1353.145

George et al: Gene therapy for vascular diseasesDichek DA, Anderson J, Kelly, A.B, Hanson S, Harker L (1996)Enhanced in vivo antithrombotic effects of endothelial cellsexpressing recombinant plasminogen activators transducedwith retroviral vectors. Circulation 93, 301-309.Dishart K, Denby, L, George S, Nicklin S, Yendluri S, Tuerk, M,Kelley, M, Donahue B, Newby, A, Harding T, Baker A(2003) Third-generation lentivirus vectors efficientlytransduce and phenotypically modify vascular cells:implications for gene therapyapy. J Mol Cell Cardiol 35,739-748.Dobrzynski E, Yoshida H, Chao J (1999) Adenovirus-mediatedkallikrein gene delivery attenuates hypertension and protectsagainst renal injury in deoxycorticosterone-salt rats.Immunopharmacology 44, 57-65.Dobrzynski E, Wang C, Chao J (2000) Adrenomedullin genedelivery attenuates hypertension cardiac remodeling andrenal injury in deoxycorticosterone acetate-salt hypertensiverats. Hypertension 36, 995-1001.Dunn P, Newman K.D, Jones M, Yamada I, Shayani V, VirmaniR, Dicheck, D (1996) Seeding of vascular grafts withgenetically modified endothelial cells. Secretion ofrecombinant TPA results in decrease seeded cell retention invitro and in vivo. Circulation 93, 1439-1446.Dzau V, Mann M, Morishita R, Kaneda Y (1996) Fusigenic viralliposome for gene therapy in cardiovascular diseases. ProcNatl Acad Sci USA 93, 11421-11425.Ehsan A, Mann M, Dell'Acqua G, Dzau V (2001) Long-termstabilization of vein graft wall architecture and prolongedresistance to experimental atherosclerosis after E2F decoyoligoneucleotide Gene Ther. J Thorac Cardiovasc Surg121, 714-722.Ehsan A, Mann M, Dell'Acqua G, Tamura K, Braun-Dullaeus R,Dzau V (2002) Endothelial healing in vein grafts.Proliferative burst unimpaired by genetic therapy ofneointimal disease. Circulation 105 (1686-1692).Emanueli C, Madeddu P (2001) Angiogenesis Gene Ther torescue ischaemic tissues: achievements and furturedirections. Br J Pharm 133, 951-958.Emanueli C, Salis M, Pinna A, Stacca T, Milia A, Spano A, ChaoJ, Chao L, Sciola L, Madeddu P (2002) Prevention ofdiabetes-induced microangiopathy by human tissue kallikreingene transfer. Circulation 106, 993 - 999.Engelhardt J.F, Litzky, L, Wilson J (1994) Prolonged transgeneexpression in cotton rat lung with recombinant adenovirusesdefective in E2a. Human Gene Ther 5, 1217-1229.Engelse MA, Lardenoye JP, Neele J, Grimbergen J, de VriesM.R, Lamfers M, Pannekoek, H, Quax, P, deVries C M(2002) Adenoviral activin A expression prevents intimalhyperplasia in human and murine blood vessels bymaintaining the contractile smooth muscle cell phenotype.Circ Res 90, 1128-1134.Fennell J.P, Brosnan M, Frater A, Hamilton CA, Alexander M.Y,Nicklin SA, Heistad D.D, BAker A, Dominiczak, A.F (2002)Adenovirus-mediated overexpression of extracellularsuperoxide dismutase improves endothelial dysfunction in arat model of hypertension. Gene Ther 9, 110-117.Fox, J (2003) US authorities uphold suspension of SCID GeneTher. Nat Biotechnol 21, 217-217.French, B, Mazur W, Ali N, Geske R, finnigan J, Rodgers G,Roberts R, Raizner A (1994a) Percutaneous transluminal invivo gene transfer by recombinant adenovirus in normalporcine coronary arteries atherosclerotic arteries and twomodels of coronary restenosis. Circulation 90, 2402-2413.French, BA, Mazur W, Geske R, Bolli R (1994b) Direct in vivogene transfer into procine myocardium using replicationdeficientadenoviral vectors. Circulation 90, 2414-2424.Furman C, Luo Z, Walsh, K, Duverger N (2002) Sytemic tissueinhibitor of metalloproteinase-1 gene delivery reducesneointimal hyperplasia in balloon-injured rat carotid artery.FEBS Letters 531, 122-126.Gangadharan S, Imai M, Rhynhart K, Sevigny, J, Robson S,Conte M (2001) Targeting platelet aggregation: CD39 genetransfer augments nucleoside triphosphatediphosphohydrolase activity in injured rabbit arteries.Surgery 130, 296-303.George S, Capogrossi M Angelini G, Baker A (2001) Wild typep53 gene transfer inhibits neointima formation in humansaphenous vein by modulation of smooth muscle cellmigration and induction of apoptosis. Gene Ther 8, 668-676.George S, Johnson J, Angelini G.D, Newby, A, Baker A (1998)Adenovirus-mediated gene transfer of the human TIMP-1gene inhibits SMC migration and neointima formation inhuman saphenous vein. Human Gene Ther 9, 867-877.George S, Lloyd C Angelini G.D, Newby, A, Baker A (2000)Inhibition of late vein graft neointima formation in humanand porcine models by adenovirus-mediated overexpressionof tissue inhibitor of metalloproteinase-3. Circulation 101,296-304.George S, Zaltsman A.B, Newby, A.C (1997) Surgicalpreparative injury and neointima formation increase MMP-9expression and MMP-2 activation in human saphenous vein.Cardiovasc Res 33, 447-459.George S, Baker A, Angelini G, Newby, A.C (1998) Genetransfer of tissue inhibitor of metalloproteinase-2 inhibitsmetalloproteinase activity and neointima formation in humansaphenous veins. Gene Ther 5, 1552-1560.Gossen M, Bujard H (1992) Tight control of gene expression inmammalian cells by tetracycline-responsive promoters. ProcNatl Acad Sci USA 89, 5547-5551.Grines C, Watkins M, Helmer G, Penny, W, Brinker J, JD M,West A, Rade J, Marrott P, Hammond H, Engler R (2002)Angiogenic Gene Therapy (AGENT) trial in patients withstable angina pectoris. Circulation 105, 1291-1297.Guzman R, Hirschowitz, EA, Brody, S, Crystal R, Epstein S.E,Finkel T (1994) In-vivo suppression of injury-inducedvascular smooth-muscle cell accumulation using adenovirusmediatedtransfer of the herpes-simplex virus thymidinekinase gene. Proc Natl Acad Sci USA 91 (22), 10732-10736.Hammond H.K, McKirman M (2001) Angiogenic Gene Therapyfor heart disease: a review of animal studies and clinicaltrials. Cardiovasc Res 49, 561-567.Hanna A.K, Fox, J Neschis D, Safford S.D, Swain J, Golden M(1997) Antisense basic fibroblast growth factor gene transferreduces neointimal thickening after arterial injury. J VascSurg 25, 320-325.Harrell R, Rajanayagam, S, Doanes A, Guzman R, Hirschowitz,EA, Crystal R, Epstein S.E, Finkel T (1997) Inhibition ofvascular smooth muscle cell proliferation and neointimalaccumulation by adenovirus-mediated gene transfer ofcytosine deaminase. Circulation 96, 621-627.Hart S, Collins L, Gustafsson K, Fabre J (1997) Integrinmediatedtransfection with peptides containing arginineglycine-asparticacid domains. Gene Ther 4, 1225-1230.Havenga M E, Lemckert A C, Grimbergen J, Vogels R, HuismanL, Valerio D, Bout A, Quax, P.H (2001) Improved146

Gene Therapy and Molecular Biology Vol 7, page 147adenovirus vectors for infection of cardiovascular tissues. J.Virol. 75, 3335-3342.Havenga M, Lemckert A, Ophorst O, Meijer M, Germeraad W,Grimbergen J, Doel M, Vogels R, Duetekom, J, Janson A,Bruijn J, Uytdehaag F, Quax, P, Logtenberg T, Mehtali M,Bout A (2002) Exploiting the natural diversity in adenovirustropism for therapy and prevention of disease. Virology 76,4612 - 4620.Hedman M, Hartikainen J, Syvanne M, Stjernvall J, Hedman A,Kivela A, Vanninen E, Mussalo H, Kauppila E, S S,Narvanen O, Rantala A, Peuhkurinen K, Nieminen M,Laakso M, Yla-Herttuala S (2003) Safety and feasibility ofcatheter-based local intracoronary vascular endothelialgrowth factor gene transfer in the prevention ofpostangioplasty and in-stent restenosis and in the treatmentof chronic myocardial ischemai. Phase II results of theKupoio Angiogenesis Trial (KAT) Circulation 107, 2677-2683.Hiltunen M.O, Turunen M.P, Turunen A.-M, Rissanen TLaitinen M, Kosma V.-M, Yla-Herttuala S (2000)Biodistribution of adenoviral vector to nontarget tissues afterlocal in vivo gene transfer to arterial wall using intravascularand periadventitial gene delivery methods. FASEB Journal14, 2230-2236.Hu Y, Baker A, Zou Y, Newby, A, Xu Q (2001) Local genetransfer of tissue inhibitor of metalloproteinase-2 influencesvein graft remodeling in a mouse model. ArteriosclerThromb Vasc Biol 21, 1275-1280.Huang J, Frischer J, Serur A, Kadenhe A, Yokoi A, McCruddenK, New, T, O'Toole K, Zabski S, Rudge J, Holash, J,Yancopoulos G, Yamashiro D, Kandel J (2003) Regressionof established tumors and metastases by potent vascularendothelial growth factor blockade. Proc Natl Acad SciUSA 100, 7785-7790.Iaccarino G, Smithwick, L, Lefkowitz, R, Koch, W (1999)Targetting Gbeta gamma signaling in arterial vascularsmooth muscle proliferation: a novel strategy to limitrestenosis. Proc Natl Acad Sci USA 96, 3945-3950.Igarashi T, Miyake K, Kato K, Watanabe A, Ishizaki M, OharaK, Shimada T (2003) Lentivirus-mediated expression ofangiostatin efficiently inhibits neovascularization in a murineproliferative retinopathy model. Gene Ther 10, 219-226.Isner J, Pieczek, A, Schainfeld R, Blair R, Haley, L, Asahara T,Rosenfield K, Razvi S, Walsh, E, Symes J.F (1996) Clinicalevidence of angiogenesis after arterial gene transfer ofphVEGF (165) in patient with ischaemic limb. Lancet 348,370-374.Janssen S, Flaherty, D, Nong Z, Varenne O, Van Pelt N,Haustermans C, Zoldhelyi P, Gerard R, Collen D (1998)Human endothelial nitric oxide synthase gene transferinhibits vascular smooth muscle cell prolifeation andneointima formation after balloon injury in rats. Circulation97, 1274-1281.Jayakumar J, Suzuki K, Khan M, Smolenski R, Farrell A, LatifN, Raisky, O, Abunasra H, Sammut I, Murtuza B, AmraniM, Yacoub M (2000) Gene therapy for myocardialprotection: transfection of donor hearts with heat shockprotein 70 gene protects cardiac function against ischemiareperfusioninjury. Circulation 102 (19 suppl 3), III302-III306.Katovich, M, Gelband C, Reaves P (1999) Reversal ofhypertension by angiotensin II type 1 receptor antisense genetherapy in the adult SHR. Am J Physiol 277, H1260-H1264.Kibbe M.R, Tzeng E, Gleixner S, Watkins S, Kovesdi L,Lizonova A, Makaroun M, Billiar T, Rhee R (2001)Adenovirus-meidated gene transfer of human inducible nitricoxide synthase in porcine vein grafts inhibits intimalhyperplasia. J Vasc Surg 34, 156-165.Kimura B, Mohuczy, D, Tang X, Phillips I (2001) Attenuation ofhypertension and heart hypertrophy by adeno-associatedvirus delivering angiotensinogen antisense. Hypertension37, 376 - 380.Klugherz, B, Song C, Defelice S, Cui X, Lu Z, Connolly, J,Hinson J, Wilensky, R, Levy, R (2002) Gene delivery to pigcoronary arteries from stents carrying antibody-tetheredadenovirus. Human Gene Ther 13, 443 - 454.Klugherz, B.D, Jones P, Cui X, Chen W, Meneveau N.F,DeFelice S, Connolly J, Wilensky, R, Levy, R (2000) Genedelivery from a DNA controlled-release stent in porcinecoronary arteries. Nat Biotechnol 18 (11), 1181-4.Kochanek, S, Clemens P.R, Mitani K, Chen H.-H, Chan S,Caskey, C. (1996) A new adenoviral vector: replacement ofall viral coding sequences with 28kb of DNA independentlyexpressing both full-length dystrophin and B-galactosidase.Proc Natl Acad Sci USA 93, 5731-5736.Koike H, Morishita R, Iguchi S, Aoki M, Matsumoto K,Nakamura T, Yokoyama C, Tanabe T, Ogihara T, Kaneda Y(2003) Enhanced angiogenesis and improvement ofneuropathy by cotransfection of human hepatocyte growthfactor and prostacyclin synthase gene. FASEB Journal 17,779-781.Laitinen M, Hartikainen J, Hiltunen M.O, Eranen J, KiviniemiM, Narvanen O, Makinen K, Manninen H, Syvanne M,Martin J.F, Laakso M, Yla-Herttuala S (2000) Cathetermediatedvascular endothelial growth factor gene transfer tohuman coronary arteries after angioplasty. Human GeneTher 11, 263-270.Lamfers M, Lardenoye J, de Vries M.R, Aalders M, EngelseMA, Grimbergen J, van Hinsbergh, V, Quax, P (2001) Invivo suppression of restenosis in balloon-injured rat carotidartery by adenovirus-mediated gene transfer of the cellsurface-directed plasmin inhibitor ATF.BPTI. Gene Ther 8,534-541.Lawrie A, Brisken A, Francis S, Tayler D, Chamberlain J,Crossman D, Cumberland D, Newman C (1999) UltrasoundEnhances Reporter Gene Expression After Transfection ofVascular Cells In Vitro. Circulation 99, 2617-2620.Lee R, Springer M, Blanco-Bose W.E, Shaw, R, Ursell P Blau H(2000) VEGF Gene Delivery to Myocardium: DeleteriousEffects of Unregulated Expression. Circulation 102, 898-901.Lemarchand P, Jones M, Yamada I, Crystal R (1993) In vivogene transfer and expression in normal uninjured bloodvessels using replication-deficient recombinant adenovirusvectors. Circ Res 72, 1132-1138.Li JM, Brooks G (1999) Cell cycle regulatory molecules (cyclinscyclin-dependent kinases and cyclin dependent kinaseinhibitors) and the cardiovascular system. Potential targetsfor therapy? Eur Heart J 20, 406-420.Li Q, Bolli R, Qiu Y, Tang X, Guo Y, French, B (2001) Genetherapy with extracellular superoxide sismutase protectsconscious rabbits against myocardial infarction. Circulation103, 1893-1898.Lin K.F, Chao J, Chao L (1995) Human atrial natriuretic peptidegene delivery reduces blood pressure in hypertensive rats.Hypertension 26, 847-853.Lin K.F, Chao L, Chao J (1997) Prolonged reduction of highblood pressure with human nitric oxide synthase genedelivery. Hypertension 30, 307-313.147

George et al: Gene therapy for vascular diseasesLin K.F, Chao J, Chao L (1999) Atrial natriuretic peptide genedelivery reduces stroke-induced mortality rate in Dahl saltsensitiverats. Hypertension 33, 219-224.Loeb J.E, Cordier W, Harris M.E, Weitzman M, Hope T (1999)Enhanced expression of transgenes from adeno-associatedvirus vectors with the woodchuck hepatitis virusposttranscriptional regulatory element: implications for genetherapy. Hum Gene Ther 10, 2295-305.Losordo D, Vale P.R, Symes J.F, Dunnington C, Esakof D.D,Maysky, M, Ashare A.B, Lathi K, Isner J (1998) Genetherapy for myocardial angiogenesis - Initial clinical resultswith direct myocardial injection of phVEGF (165) as soletherapy for myocardial ischemia. Circulation 98, 2800-2804.Lu L, Bonham, CA, Chambers F, Watkins S Hoffman RA,Simmonds R, Thomson A (1996) Induction of nitric oxidesynthase in mouse dendritic cells by IFN-gamma endotoxinand interaction with allogeneic T Cells. Immunology 157,3577-3586.Lundstrom, K, Schweitzer C, Rotmann D, Hermann D,Schneider E, Ehrengruber M (2001) Semliki forest virusvectors: efficient vehicles for in vitro and in vivo genedelivery. FEBS Letters 504, 99 - 103.Luo Z, Sato M, Nguyen T, Kaplan J, Akita G, Walsh, K (1999)Adenovirus-mediated delivery of fas ligand inhibits intimalhyperplasia after balloon injury in immunologically primedanimals. Circulation 99, 1776-1779.Maeda Y, Ikeda U, Ogasawara Y, Urabe M, Takizawa T, SaitoT, Colosi T, Kurtzman G, Shimada K, Ozawa K (1997) Genetransfer into vascular cells using adeno-associated virus(AAV) vectors. Cardiovasc Res 35, 514-521.Maillard L, Van Belle E, Smith, R, Le Roux, A, Denefle P, StegG, Barry, J, Branellec D, Isner J, Walsh, K (1997)Percutaneous delivery of the gax gene inhibits vessel stenosisin a rabbit model of balloon angioplasty. Cardiovasc Res 35,536-546.Makinen K, Manninen H, Hedman M, Matsi P, Mussalo H,Alhava E, Yla-Herttuala S (2002) Increased vascularitydetected by digital subtraction angiography after VEGF genetransfer to human lower limb artery: a randomized placebocontrolleddoubl-blinded phase II study. Mol Ther 6, 127-133.Mann M, Gibbons G, Kernoff R, Diet F, Tsao P, Cooke J,Kaneda Y, Dzau V (1995) Genetic engineering of vein graftsresistant to atherosclerosis. Proc Natl Acad Sci USA 92,4502-4506.Mann M, Whittemore A.D, Donaldson M Belkin M, Conte M,Polak, J.F, Orav, E, Ehsan A, Dell'Acqua G, Dzau V (1999)Ex-vivo gene therapy of human vascular bypass grafts withE2F decoy: the PREVENT single-centre randomisedcontrolled trial. Lancet 354, 1493-1498.Manninen H, Makinen K (2002) Gene therapy techniques forperipheral arterial disease. Cardiovasc Intervent Radiol 25,98-108.Martens J.R, Reaves P.Y, Lu D (1998) Prevention ofrenovascular and cardiac pathophysiological changes inhypertension by angiotensin II type 1 receptor antisense geetherapy. Proc Natl Acad Sci USA 95, 2664-2669.Matsumoto T, Komori K, Yonemitsu Y, Morishita R, Sueishi K,Kaneda Y, Sugimachi K (1998) Hemagglutinating virus ofJapan-liposome-mdiated gene transfer of endothelial cellnitric oxide synthase inhibits intimal hyperplasia of caninegrafts under conditions of poor runoff. J Vasc Surg 27, 135-144.Mintz, G, Popma J, Pichard A, Kent K, Satler L, Wong C.Y,Hong M, Kovach, J, Leon M (1996) Arterial remodelingafter coronary angioplasty: a serial intravascular ultrasoundstudy. Circulation 94, 35-43.Morishita R, Gibbons G, Kaneda Y, Ogihara T, Dzau V (1994)Pharmacokinetics of antisense oliogdeoxyribonucleotides(cyclin B1 and CDC 2 kinase) in the vessel wall in vivo:enhanced therapeutic utility for restenosis by HVJ-liposomedelivery. Gene 149, 13-19.Morishita R, Gibbons G, Horiuchi M, Ellison K, Nakama M,Zhang L, Kaneda Y, Ogihara T, Dzau V (1995) A genetherapy strategy using a transcription factor decoy of the E2Fbinding site inhibits smooth muscle proliferation in vivo.Proc Natl Acad Sci USA 92, 5855-5859.Nabel E, Shum, L, Pompili V, Yang ZY, San H, Shu H.B,Liptay, S, Gold L, Gordon D, Derynck, R (1993) Directtransfer of transforming growth factor b1 gene into arteriesstimulates fibrocellular hyperplasia. Proc Natl Acad SciUSA 90, 10759-10763.Nabel E, Yang Z, Liptay, S, San H, Gordon D, Haudenschild CNabel G (1993a) Recombinant platelet-derived growth factorB expression in porcine arteries induces intimal hyperplasiain vivo. J Clin Invest 91, 1822-1829.Nabel E, Yang Z, Plautz, G, Forough, R, Zhan X, HaudenschildC Maciag T, Nabel G (1993b) Recombinant fibroblastgrowth factor-1 promotes intimal hyperplasia andangiogenesis in arteries in vivo. Nature 362, 844-846.Nakai H, Yant S, Storm, T, Fuess S, Meuse L, Kay, M (2001)Extrachromosomal recombinant adeno-associated virusvector genomes are primarily resposible for stable livertransduction in vivo. Virology 75, 6969 - 6976.Nakamura T, Morishita R, Asai T, Tsuboniwa N, Aoki M,Sakonjo H, Yamasaki K, Hashiya N, Kaneda Y, Ogihara T(2002) Molecular strategy using cis-element 'decoy' of E2Fbinding site inhibits neointimal formation in porcine ballooninjuredcoronary artery model. Gene Ther 9, 488-494.Napoli C (2002) MMP-inhibition and the development ofcerebrovascular atherosclerosis: the road ahead. Stroke 33,2864-2865.Newby A, George S (1996) Proliferation migration matrixturnover and death of smooth muscle cells in native coronaryand vein graft atherosclerosis. Curr Opin Cardiol 11, 574-582.Newman K.D, Dunn P.F, Owens J, Schulick A, Virmani R,Sukhova G, Libby, P, Dichek, D (1995) Adenovirusmediatedgene transfer into normal rabbit arteries results inprolonged vascular cell activation inflammation andneointimal hyperplasia. J Clin Invest 96, 2955-2965.Nicklin SA, Von Seggern D, Work, L, Pek, D.C.K, Dominiczak,A.F, Nemerow, G, Baker A (2001) Ablating adenovirus type5 fiber-CAR binding and HI loop insertion of the SIGYPLPpeptide generate an endothelial cell-selective adenovirus.Mol Ther 4, 534-542.Nishida T, Ueno H, Atsuchi N, Kawano R, Asada Y, NakaharaY, Kamikubo Y, Takeshita A, Yasui H (1999) Adenovirusmediatedlocal expression of human tissue factor pathwayinhibitor eliminates shear stress-induced redurrentthrombosis in the injured carotid artery of the rabbit. CircRes 84, 1446-1452.Noiseux, n, boucher C, Cartier R, Sirois M (2000) Bolusendovascular PDGFR-beta antisense treatment suppressedintimal hyperplasia in a rat carotid injury model. Circulation102 (11), 1330-1336.148

Gene Therapy and Molecular Biology Vol 7, page 149Ohno N, Itoh, H, Ikeda T, Ueyama K, Yamahara K, Doi K,Yamashita J, Inoue M, Masatsugu K, Sawada N, FukunagaY, Sakaguchi S, Sone M, Yurugi T, Kook, H, Komeda M,Nakano K (2002) Accelerated reendothelialization withsuppressed thrombogenic property and neointimalhyperplasia of rabbit jugular vein grafts by adenovirusmediatedgene transfer of C-type natriuretic peptide.Circulation 105, 1623-1626.Parkes R, Meng Q, Siapati K, McEwan J, Hart S (2002) Highefficiency transfection of porcine vascular cells in vitro witha synthetic vector system. J. Gene Medicine 4, 292-299.Parks R, Chen L, Anton M, Sankar U, Rudnicki M, Graham, F(1996) A helper-dependent adenovirus vector system:Removal of helper virus by Cre-mediated excision of theviral packaging signal. Proc Natl Acad Sci USA 93 (24),13565-13570.Phillips M.I (2001) Gene therapy for Hypertension: ThePreclinical Data. Hypertension 38, 543-548.Qing K, Mah, C, Hansen J, Zhou S, Dwarki V, Srivastava A(1999) Human fibroblast growth factor receptor 1 is a coreceptorfor infection by adeno-associated virus 2. Nat Med5, 71-7.Quax, PA, Lamfers M, Grimbergen J, Verheijen J, vanHinsbergh, V (1997) Inhibition of neointima formation incultured human saphenous vein segments by an advenovirusexpressing an urokinase receptor binding plasmin inhibitor.Circulation 96 (Suppl), I-669.Rajagopalan S, Shah, M, Luciano A, Crystal R, Nabel E (2001)Adenovirus-mediated gene transfer of VEGF121 improveslower extremity endothelial function and flow reserve.Circulation 104, 753-755.Rajagopalan S, Mohler E, Lederman R, Saucedo J, MendelsohnF, Olin J, Bleba J, Goldman C, Trachtenberg J, Pressler M,Rasmussen H, Annex, B, Hirsch, A, trial R W.V.E.G.F(2003) Regional angiogenesis with vascular endothelialgrowth factor (VEGF) in peripheral arterial disease: Designof the RAVE trial. Am Heart J 145, 1114-8.Reaves P.Y, Gelband C, Wang H, Yang H, Lu D, Berecek, K,Katovitch, M, Raizada M.K (1999) Permanentcardiovascular protection from hypertension by the AT1receptor antisense gene therapy in hypertensive rat offspring.Circ Res 85, e45-e50.Richter M, Iwata A, Nyhuis J, Nitta Y, Miller A, Halbert L,Allen M (2000) Adeno-associated virus vector transductionof vascular smooth muscle cells in vivo. Physiol. Genomics2, 117-127.Rissanen T, Vajanto I, Yla-Herttuala S (2001) Gene therapy fortherapeutic angiogenesis in critically ischaemic lower limb -on the way to the clinic. Eur J Clin Invest 31, 651-666.Rivard A, Silver M, Chen D, Kearney, M, Magner M, Annex, B,Peters K, Isner J (1999) Rescue of diabetes-relatedimpairment of angiogensis by intramuscular gene therapywith adeno-VEGF. Am J Pathol 154, 355-363.Rosengart T.K, Lee L.Y, Patel S.R, Sanborn TA, Parikh, M,Bergman G, Hachamovitch, R, Szulc M, Kligfield P.D, OkinP, Hahn R Devereux, R.B, Post M.R, Hackett N.R, Foster T,Grasso T, Lesser M, Isom, O, Crystal R (1999) AngiogenesisGene Therapy - Phase I assessment of direct intramyocardialadministration of an adenovirus vector expressing VEGF121cDNA to individuals with clinically significant severecoronary artery disease. Circulation 100, 468-474.Schnepp, B, Clark, K, Klemanski D, Pacak, C, Johnson P (2003)Genetic fate of recombinant adeno-associated virus vectorgenomes in muscle. J Virol 77, 3495-3504.Serruys P, Degertekin M, Tanabe K, Abizaid A, Sousa J.E,Colombo A, Guagliumi G, Wijns W, Lindeboom, W,Ligthart J, de Feyter P, Morice M, Group, R (2002)Intravascular ultrasound findings in the multicenterrandomized double-blind RAVEL (RAndomized study withthe sirolimus-eluting VElocity balloo-expandable stent in thetreatment of patients with de novo native coronary arteryLesions) trial. Circulation 106, 798-803.Shears L, Kibbe M, Murdock, A, Billiar T, Lizonova A, KovesdiI, Watkins S, Tzeng E (1998) Efficient inhibition of intimalhyperplasia by adenovirus-mediated inducible nitric oxidesynthase gene transfer to rats and pigs in vivo. J Am CollSurg 187, 295-306.Shi Y, Fard A, Galeo A, Hutchinson H, Vermani P, Dodge G,Hall D, Shaheen F, Zalewski A (1994) Transcatheter deliveryof c-myc antisense oligomers reduces neointimal formationin a porcine model of coronary artery balloon injury.Circulation 90, 944-951.Shi Y, Patel S, Davenpeck, K, Niculescu R, Rodriguez, E,Magno M, Ormont M, Mannion J, Zalewski A (2001)Oxidative stress and lipid retention in vascular grafts:comparison between venous and arterial conduits.Circulation 15 (103), 19.Shyu K, Manor O, Magner M, Yancopoulos G, Isner J (1998)Direct intramuscular injection of plasmid DNA encodingangiopoietin-1 but not angiopoietin-2 augmentsrevascularization in the rabbit ischemic hindlimb.Circulation 98, 2081-2087.Simari R.D, San H, Rekhter M, Ohno T, Gordon D, Nabel G,Nabel E (1996) Regulation of cellular proliferation andintimal formation following balloon injury in atheroscleroticrabbit arteries. J Clin Invest 98, 225-235.Simons M, Edelman E.R, DeKeyser J.-L, Langer R, Rosenberg R(1992) Antisense c-myb oligonucleotides inhibit intimalarterial smooth muscle cell accumulation in vivo. Nature359, 67-70.Simovic D, Isner J, Ropper A, Pieczek, A, Weinberg D (2001)Improvement in chronic ischemic neuropathy afterintramuscular phVEGF165 gene transfer in patients withcritical limb ischemia. Arch Neurol 58, 761-768.Sirois M, Simons M, Edelman E (1997) Antisenseoligonucleotide inhibition of PDGFR-beta receptor subunitexpression directs suppression of intimal thickening.Circulation 95, 669-676.Smith, R, Branellec D, Gorski D, Guo K, Perlman H, Dedieu J,Pastore C, Mahfoudi A, Denefle P, Isner J, Walsh, K (1997)p21CIP1-mediated inhibition of cell proliferation byoverexpression of gax homeodomain gene. Genes Dev 11,1674-1689.Somia N, Verma I (2000) Gene therapy: trials and tribulations.Nat Rev Genet 1, 1-9.SoRelle R (2001) Late-breaking clinical trials at the AmericanHeart Association's Scientific Session 2001. Circulation104, c9046.Sousa J.E, Costa MA, Sousa A, Abizaid A, Seixas A, Abizaid A,Feres F, Mattos L, Falotico R, Jaeger J, Popma J, Serruys P(2003) Two-year angiographic and intravascular ultrasoundfollow-up after implantation of sirolimus-eluting stents inhuman coronary arteries. Circulation 107, 381-383.Southgate K, Mehta D, Izzat M.B, Newby, A, Angelini G (1999)Increased secretion of basement membrane-degradingmetalloproteinases in pig saphenous vein into carotid arteryinterposition grafts. Arterioscler Thromb Vasc Biol 19,1640-1649.149

George et al: Gene therapy for vascular diseasesStephan D, San H, Yang Z, Gordon D, Goelz, S, Nabel G, NabelE (1997) Inhibiton of vascular smooth muscle cellproliferation and intimal hyperplasia by gene transfer of betainterferon.Mol Med 3, 593-599.Stewart D, investigators o.b.o.ts (2002) A phase 2, randomizedmulticenter 26-week study to assess the efficacy and safetyof BIOBYPASS (AdGVVEGF121.10) delivered throughminimally invasive surgery versus maximum medicaltreatment in patients with severe angina advanced coronaryartery disease and no options for revascularization.Circulation 106, 23-26.Summerford C, Samulski R (1998) Membrane-associatedheparan sulfate proteoglycan is a receptor for adenoassociatedvirus yype 2 virions. J. Virol. 72, 1438-1445.Summerford C, Bartlett J, Samulski R (1999) aVb5 integrin: aco-receptor for adeno-associated virus type 2 infection. NatMed 5, 78-82.Symes J.F, Losordo D, Vale P.R, Lathi K, Esakof D.D, MayskiyM, Isner J (1999) Gene therapy with vascular endothelialgrowth factor for inoperable coronary artery disease. Annalsof Thoracic Surgery 68, 830-836.Tamirisa K.P, Mukherjee D (2002) Gene therapy incardiovascular diseases. Curr Gene Ther. 2:427-435Tang X, Mohuczy, D, Zhang Y.C (1999) Intravenousangiotensinogen antisense in AAV-based vector decreaseshypertension. Am J Physiol 277, H2392-H2399.Taniyama Y, Morishita R, Hiraoka K, Aoki M, Nakagami H,Yamasaki K, Matsumoto K, Nakamura T, Kaneda Y,Ogihara T (2001) Therapeutic angiogenesis induced byhuman hepatocyte growth factor gene in rat diabetic hindlimb ischemia model: molecular mechanisms of delayedangiogenesis in diabetes. Circulation 104, 2344-2350.Taniyama Y, Tachinaba K, Hiroaka K, Namba T, Yamasaki K,Hashiya N, Aoki M, Ogihara T, Yasufumi K, Morishita R(2002) Local delivery of plasmid DNA into rat carotid arteryusing ultrasound. Circulation 105, 1233-1239.Tomko R.P, Xu R, Philipson L (1997) HCAR and MCAR: Thehuman and mouse cellular receptors for subgroup Cadenoviruses and group B coxsackieviruses. Proc Natl AcadSci USA 94, 3352-3356.Tulis DA, Mynjoyan Z, Schiesser R, Shelat H, Evans A,Zoldhelyi P, Fujise K (2003) Adenoviral gene transfer offortilin attenuates neointima formation through suppressionof vascular smooth muscle cell proliferation and migration.Circulation 107, 98-105.Ueno H, Yamamoto H, Ito S.I, Li J, Takeshita A (1997)Adenovirus-mediated transfer of a dominant-negative H-rassuppresses neointimal formation in balloon-injured arteries invivo. Arterioscler Thromb Vasc Biol. 17, 898-904.Van Belle E, Tio F, Chen D, Maillard L, Chen D, Kearney, M,Isner J (1997) Passivation of metallic stents after arterialgene transfer of phVEGF165 inhibits thrombus formationand intimal thickening. J Am Coll Cardiol 29, 1371-1379.Varenne O, Pislaru S, Gillijns H, Van Pelt N, Gerard R.D,Zoldhelyi P, Van de Werf F, Collen D, Janssens S.P (1998)Local adenovirus-mediated transfer of human endothelialnitric oxide synthase reduces luminal narrowing aftercoronary angioplasty in pigs. Circulation 98 (9), 919-926.Vigna E, Cavalieri S, Ailles L, Guena M, Loew, R, Buyjard H,Naldini L (2002) Robust and efficient regulation of transgeneexpression in vivo by improved tetracycline-dependentlentiviral vectors. Mol Ther 5, 252 -257.von der Leyen H.E, Gibbons G, Morishita R, Lewis N, Zhang L,Nakajima M, Kaneda Y, Cooke J, Dzau V (1995) Genetherapy inhibiting neointimal vascular lesion: in vivo transferof endothelial cell nitric oxide synthase gene. Proc NatlAcad Sci USA 92, 1137-1141.Wang C, Chao L, Chao J (1995) Direct gene delivery of humantissue kallikrein reduces blood pressure in spontaneouslyhypertensive rats. J Clin Invest 95, 1710-1716.Wang C, Dobrzynski E, Chao J (2001) Adrenomedullin genedelivery attenuates renal damage and cardiac hypertrophy inGoldblatt hypertensive rats. Am J Physiol Renal Physiol280, F964-F971.Wang H, Katovich, M, Gelband C, Reaves P.Y, Phillips M.I,Raizada M.K (1999) Sustanined inhibition of angiotensin I-converting enzyme (ACE) expression and long-termantihypertensive action by virally mediated delivery of ACEantisense cDNA. Circ Res 85, 614-622.Wang H, Reaves P.Y, Gardon M (2000) Angiotensin I-converting enzyme antisense gene therapy causes permanentantihypertensive effects in the SHR. Hypertension 35, 202-208.Waugh, J, Kattash, M, Li J, Yuksel E, Kuo M.D, Lussier M,Weinfeld A.B, Saxena R, Rabinovsky, E.D, Thung S, WooS.L Shenaq, S (1999a) Gene therapy to promotethromboresistance: local overexpression of tissueplasminogen activator to prevent arterial thrombosis in an invivo rabbit model. Proc Natl Acad Sci USA 96, 1065-1070.Waugh, J, Yuksel E, Li J, Kuo M.D, Kattash, M, Saxena R,Geske R, Thung S.N, Shenaq, S, Woo S (1999b) Localoverexpression of thrombomodulin for in vivo prevention ofarterial thrombosis in a rabbit model. Circ Res 84, 84-92.Wen S, Schneider D, Driscoll R, Vassalli G, Sassani A, Dichek,D (2000) Second-generation adenoviral vectors do notprevent rapid loss of transgene expression amd vector DNAfrom the arterial wall. Arterioscler Thromb Vasc Biol. 20,1452 - 1458.West N, Guzik, T, Black, E, Channon K (2001) Enhancedsuperoxide production in experimental venous bypass graftintimal hyperplasia. Role of NAD (P)H oxidase.Arterioscler Thromb Vasc Biol 21 (189-194).Wickham, T, Mathias P, Cheresh, D, Nemerow, G (1993)Integrins avb3 and avb5 promote adenovirus internalizationbut not virus attachment. Cell 73, 309-319.Wielbo D, Simon A, Phillips M.I (1996) Inhibition ofhypertension by peripheral administration of antisenseoligodeoxynucleotides. Hypertension 28, 147-151.Wolf W Yoshida H, Agata J (2000) Human tissue kallikrein genedelivery attenuates hypertension renal injury, and cardiacremodeling in chronic renal failure. Kidney Int 58, 730-739.Yamamoto K, Morishita R, Tomita N, Shimozato T, NakagamiH, Kikuchi A, Aoki M, Higaki J, Kaneda Y, Ogihara T(2000) Ribozyme oligonucleotides against transforminggrowth factor-b inhibited neointimal formation after vascularinjury in rat model. Circulation 102, 1308-1314.Yang L, Quan S, Abraham, N (1999) Retrovirus-mediated HOgene transfer into endothelial cells protects against osidantinducedinjury. Am J Physiol 277 (1 pt 1), L127-L133.Yang Z, Simari R, ND P, H, S, Gordon D, GJ, N, Nabel E (1996)Role of the p21 cyclin-dependent kinase inhibitor in limitingintimal cell proliferation in response to arterila injury. ProcNatl Acad Sci USA 93 (15), 7905-7910.Yayama K, Wang C, Chao L (1998) Kallikrein gene deliveryattenuates hypertension and cardiac hypertrophy andenhances renal function in Goldblatt hypertensive rats.Hypertension 31, 1104-1110.150

Gene Therapy and Molecular Biology Vol 7, page 151Yla-Herttuala S, Alitalo K (2003) Gene transfer as a tool toinduce therapeutic vascular growth. Nat Med 9, 694-701.Yonemitsu Y, Kaneda Y, Tanaka S, Nakashima Y, Komori K,Sugimachi K, Sueishi K (1998) Transfer of wild-type p53gene effectively inhibits vascular smooth muscle cellproliferation in vitro and in vivo. Circ Res 82, 147-156.Zhang J, Yoshida H, Chao L (2000) Human adrenomedullin genedelivery protects against cardiac hypertrophy, fibrosis andrenal damage in hypertensive dahl salt-sensitive rats. HumGene Ther 11, 1817-1827.Zoldhelyi P, McNatt J, Shelat H, Yamamoto Y, Chen Z,Willerson J. (2000) Thromboresistance of balloon-injuredporcine carotid arteries after local gene transfer of humantissue factor pathway inhibitor. Circulation 101, 289-295.Zufferey, R, Donello J.E, Trono D, Hope T (1999) Woodchuckhepatitis virus posttranscriptional regulatory elementenhances expression of transgenes delivered by retroviralvectors. J Virol. 73, 2886-2892.151

George et al: Gene therapy for vascular diseases152

Gene Therapy and Molecular Biology Vol 7, page 153Gene Ther Mol Biol Vol 7, 153-165, 2003.Angiogenic gene therapy for improving islet graftvascularizationReview ArticleNan Zhang 1 , Karen Anthony 1 , Katsunori Shinozaki 1 , Jennifer Altomonte 1 ,Zachary Bloomgarden 2 and Hengjiang Dong 1,3 *1 Carl Icahn Institute for Gene Therapy and Molecular Medicine, 2 Department of Medicine, 3 Division of ExperimentalDiabetes and Aging, Department of Geriatrics, Mount Sinai School of Medicine, New York, NY 10029.__________________________________________________________________________________*Correspondence: Hengjiang Dong, Ph.D., Mount Sinai School of Medicine, Box 1496, One Gustave L. Levy Place, New York, NY10029; tel: 212-241-3662; fax: 212-241-0738; email: hengjiang.dong@mssm.edu.Key words: Type 1 diabetes, islet transplantation, islet revascularization, VEGF, gene transfer.Received: 3 July 2003; Accepted: 19 August, 2003; electronically published: August 2003SummaryClinical islet transplantation is considered a curative treatment for type 1 diabetes, but long-term survival andfunction of implanted islets is greatly compromised by a number of adverse events. In addition to immune rejectionand recurrent autoimmunity, the survival and function of islets is determined by the rate and degree of isletrevascularization, an essential process termed angiogenesis that is required for the development of new vesselswithin islet grafts to derive blood from the host vasculature. Rapid and adequate revascularization is crucial forislet survival and function. Delay in islet revascularization can deprive islets of oxygen and nutrients, resulting inislet cell death and early graft failure. There is evidence that despite the infusion of sufficiently large amounts ofislets (~11,000 islets/kg body weight) per diabetic recipient, less than 30% of islet mass becomes stably engraftedpost transplantation. In this article, we will review the molecular basis of islet revascularization and highlight theimportance of developing novel therapeutic strategies to stimulate angiogenesis within islet grafts and enhance isletgraft vascularization post transplantation. Such strategies, when applied in conjunction with islet transplantation,are expected to improve the viability of transplanted islets and provide long-term survival of functional islet masspost transplantation, thereby increasing the overall success rate of islet transplantation.I. IntroductionA. Type 1 diabetesType 1 diabetes is a metabolic disorder that is causedby insulin deficiency due to autoimmune destruction of βcells, leading to chronic elevation of blood sugar levels.Because of its onset in children and young adolescents,type 1 diabetes was previously referred to as juvenilediabetes or insulin-dependent diabetes. Prior to thediscovery and isolation of insulin for therapeutic use,patients with type 1 diabetes survived only for a period ofmonths, with death caused primarily by the accumulationof ketones in the body, leading to diabetic ketoacidosis.Over the past century, the prevalence of type 1 diabeteshas increased in a variety of populations with an incidencerate ranging from 1-3 per 100,000 children per year in theUS at the beginning of the 20th century to 4-7 per 100,000in Scandinavian countries between 1930-1950, and toapproximately 20 per 100,000 in Scandinavia over the pasttwo decades (Bloomgarden, 1998; Gale, 2002). Currently,there are about 1.7 million patients with an overall annualincidence of about 15 per 100,000 children in the US alone(Karvonen et al, 2000). This poses a tremendous burdenon patients and healthcare economies.B. Insulin therapy and limitationsType 1 diabetes is commonly treated with twicedailyinjection of a mixture of delayed and short-actinginsulin. Delayed-acting insulin is provided to maintain arelatively constant background level of plasma insulin forthe basal requirement, on which short-acting insulin isimposed to meet the postprandial demand of insulin aftermeals. Nevertheless, such conventional insulin therapytypically leads to inadequate blood sugar control as mosttreated patients experience to a lesser or greater extentelevated blood sugar levels between meals and during thenight, the cumulative effect of which can result in thedevelopment of diabetic complications at a late stage.There is clinical evidence that more than half of diabeticpatients have eyes affected by diabetic retinopathy(Bloomgarden, 1998), with additional effects on the153

Zhang et al: Angiogenic Gene Therapy for Improving Islet Graft Vascularizationkidneys by diabetic nephropathy (Chaturvedi et al, 2000)and on nerves by diabetic neuropathy, together with about4- and 10-fold lifetime increase in rates of cardiovascularmortality among men and women, respectively (Laing etal, 2003). To improve glycemic control, a number ofinsulin analogs, such as short-acting insulin lispro andaspart (Plank et al, 2002), as well as delayed-acting insulinglargine (Murphy et al, 2003) and detimir (Vague et al,2003) have been developed. Nevertheless, implementationof treatment regimens with insulin analogs in differentformulations to strive for normoglycemic control withoutrisk of hypoglycemia can be very challenging and requiresextraordinary efforts from both health care providers anddiabetic patients (Bloomgarden et al, 2002).sources by generating insulin-producing cells throughgenetic engineering of embryonic stem cells (Lumelsky etal, 2001; Soria et al, 2001). In addition, limited progresshas been made to induce graft tolerance using immunemodulation or allorecognition (Cote et al, 2001). An indepthdiscussion of these two outstanding issues inrelation to the optimal clinical outcome of islettransplantation, which is beyond the scope of this article,has been reviewed elsewhere (Waldmann, 2002; Lechleret al, 2003; Lechner and Habener, 2003). Here we wouldlike to highlight a third limiting factor, namely isletrevascularization, which appears to play an important rolein determining the long-term survival and optimalperformance of functional islet mass post transplantation.C. Islet transplantationOf alternative insulin replacement therapiesdeveloped, islet transplantation offers the prospect ofproviding a curative treatment for type 1 diabetes withoutthe need for exogenous insulin. The protocol of islettransplantation developed by Shapiro and colleagues at theUniversity of Alberta at Edmonton, Canada, known as theEdmonton protocol, is relatively simple and minimallyinvasive, which is carried out under local anestheticswithout surgery. Using fluoroscopic guidance, isolatedhuman islets are implanted intraportally to a diabeticrecipient, such that islets are engrafted in the liver andfunction to provide near physiological insulin release froman ectopic site. The success of this protocol has largelybeen attributed to technical advances in isolating highqualityhuman islets in relatively large quantities and theapplication of more potent and less toxic non-steroidalimmunosuppressants (Shapiro et al, 2000). Using theEdmonton protocol, long-term excellent glycemic controlhas been achieved with sustained freedom from insulininjection in type 1 diabetic patients (Shapiro et al, 2000).Currently, this protocol is being rigorously tested inclinical trials at multiple clinical centers to evaluate thesafety and efficacy of islet transplantation and assess thebenefit and risk ratio associated with long-term use ofimmunosuppressive drugs (Boker et al, 2001).Although promising for providing a curative optionfor type 1 diabetes, the Edmonton protocol is limited bytwo major factors: the lack of a sufficiently large source ofislets due to the scarcity of cadaveric pancreas donors, andthe presence of persistent immune rejection as well as thepotential for recurrence of autoimmunity. Recent followupstudies indicate that even with the rigorous applicationof steroid-free immunosuppressive regimens, there is stilla slow and progressive loss of insulin production fromtransplanted islets in diabetic recipients over time, asevidenced by reports that 30-40% of islet recipients mayexperience recurrence of autoimmune diabetes with reacquisitionof insulin dependence one to two years posttransplantation (Shapiro et al, 2000; Boker et al, 2001;Ryan et al, 2001, 2002). To overcome these limitations,attempts have been made to develop alternative islet1. Islet revascularization posttransplantationa. Re-establishment of islet microvasculature.Native islets in the pancreas have a rich glomerularlikevascular system that consists of fine capillariessupplied by one to five feeding arterioles and drained bycoalescing into an efferent plexus exiting the islet via oneto five venules (Menger et al, 2001; Mattson et al, 2002).Such a rich microvasculature in pancreatic islets serves toprovide efficient delivery of oxygen and nutrients to isletcells, and at the same time ensure rapid dispersal ofpancreatic hormones to the circulation. However, isolatedislets are avascular in both structural and functionalentities, such that after transplantation, the survival andfunction of islets must rely on the re-establishment of newvessels in the grafts to derive blood flow from the hostvessel system (Boker et al, 2001; Vasir et al, 2001). Thereis evidence that freely transplanted islets are associatedwith significantly reduced islet revascularization incomparison to native islets in the pancreas and thisproblem occurs irrespective of whether islets aretransplanted intraportally in the liver, retrogradely into thespleen, or under the kidney capsule (Figure 1) (Mattson etal, 2002).What are the likely consequences of delayed orinsufficient islet revascularization post islettransplantation? To answer this question, let us take aquantitative view of the relative partitioning of blood flowto islets vs. exocrine tissue in the pancreas. Using amodified microsphere technique, it has been shown thatislets take up more than 10% of the total pancreatic bloodflow despite their collectively comprising only about 1%of the tissue mass of the pancreas (Jansson and Carlsson,2002). Thus, it is critically important to maintain adequatemicrovascular perfusion to islet cells for oxygen andnutrient supplies. While islets are transplanted either assingle entities or as aggregated islet clusters under thekidney capsule or intraportally in the liver, adequatemicrovascular perfusion to islet cells does not resumeimmediately after islet transplantation.154

Gene Therapy and Molecular Biology Vol 7, page 155Figure 1. Intra-islet microvasculature. A. Microvasculature in the mouse pancreas, as visualized by immunostaining for the endotheliummarker CD-31, also known as the platelet endothelial cell adhesion molecule-1 (PECAM-1). B. Microvasculature in engrafted isletsunder the renal capsule of a diabetic mouse following 16 days of islet transplantation. Islet grafts are indicated by arrows. Bar, 50 µm.Instead, it can take up to three to five days for theformation of intra-graft microvessels to occur post islettransplantation and the re-establishment of intra-graftblood perfusion can take even longer time (>14 days)(Vasir et al, 2001, Jansson and Carlsson, 2002). This delayin the re-establishment of a functional microvasculature innewly grafted islets can starve islet cells of oxygen andnutrients. Indeed, several studies have shown that newlytransplanted islets are hypoxic, causing islet cells toundergo apoptosis and/or necrosis, which attributes to theloss of functional β-cell mass post transplantation (Vasir etal, 2001; Jansson and Carlsson, 2002).Consistent with this interpretation, it has been shownthat despite the administration of a large number of islets(11,000 islets/kg body weight) per diabetic recipient, onlyabout 30% of transplanted islets become stably engrafted,corresponding to a total loss of about 70% of thefunctional islet mass in the early post transplantation phase(Boker et al, 2001). In addition, recent clinical dataindicate that even when fasting blood glucose levels arerestored to the physiological range post islettransplantation, the optimal performance of engraftedislets in terms of glucose-inducible insulin secretion isabnormal. In response to intravenous glucose infusion, theamplitude of the first phase insulin secretion is only about20% of normal, which coincides with relatively slowglucose disposal rates following an oral glucose load inpost-transplant subjects (Ryan et al, 2002). Although thereis no direct proof suggesting that this observed suboptimalperformance of transplanted islets in glycemic control isassociated with insufficient vascularization, there isgeneral agreement that impaired islet revascularizationdoes adversely affect the optimal function of islets posttransplantation. Recent preclinical studies have shown thateven after transplanted islets are stably engrafted, theextent of vascularization, defined as microvascular densityin transplanted islets is significantly lower than that innative islets in the pancreas (Jansson and Carlsson, 2002).In addition, engrafted islets in all three of the differenttransplantation organs (kidney cortex, liver and spleen)also exhibit markedly low oxygen tension, in comparisonto native islets in the pancreas, which is associated with aconcomitant reduction in intra-graft blood perfusion(Carlsson et al, 2000, 2001). Currently, the extent to whichthis observed low oxygen tension and reduced bloodperfusion in islet grafts, as a result of insufficient isletrevascularization, adversely affect the long-term survivaland optimal performance of functional islet mass andcontribute to early graft failure is not known. Anadditional factor that might contribute to the metabolicabnormality in glucose tolerance in diabetic recipients isislet graft reinnervation post transplantation. However,little is currently known about its molecular basis inrelation to islet revascularization and the optimalperformance of islet function in glycemic control posttransplantation.b. Mechanism of islet graft vascularizationTo date, the molecular mechanism of isletrevascularization post islet transplantation remains poorlyunderstood. In general, tissue graft vascularizationdepends on a coordinated process of angiogenesis andvasculogenesis, which are functionally governed by twokey protein factors, vascular endothelial growth factor(VEGF) and angiopoietin-1 (Ang-1). These twoangiogenic/vasculogenic factors play separate butcomplementary roles in the de novo formation of bloodvessels during embryonic development (vasculogenesis) aswell as in the formation of new blood vessels from preexistingones (angiogenesis) (Yancopoulos et al, 2000).VEGF acts in the early phase to stimulate the formation ofprimitive vascular networks by vasculogenesis andangiogenic sprouting, whereas Ang-1 functionssubsequently for remodeling and maturation of theprimary vascular system by integrating the endothelialcells of vessels with surrounding matrix and supportingcells (smooth muscle cells and pericytes) (Thurston et al,1999). Thus, in terms of their specific roles inangiogenesis/vasculogenesis, VEGF seems to be a critical"driver" for initiating vascular formation, whereas Ang-1works as a "stabilizer" to ensure subsequent maturation155

Zhang et al: Angiogenic Gene Therapy for Improving Islet Graft Vascularizationand stability of the newly formed blood vessels. These twofactors act synergistically to ensure new blood vesselformation, growth and maturation.VEGF has four different isoforms in humans,consisting of 121, 165, 189 and 206 amino acid residues,all of which are generated by alternative splicing of asingle gene. Rodents have only three isoforms, namelyVEGF120, VEGF164 and VEGF188, each polypeptideone amino acid shorter than their corresponding humanhomologues (Kim et al, 2000; Vasir et al, 2000, 2001).The most abundant and widely distributed form isVEGF165 in humans (or VEGF164 in rodents). In concertwith their respective functions in angiogenesis /vasculogenesis, the receptors for both VEGF (VEGFR-1/Flt1 and VEGFR-2/Flk-1/KDR) and Ang-1 (Tie2) areselectively expressed in the vascular endothelium (Ferraraand Davis-Smyth, 1997; Otani et al, 1999; Kim et al,2000). In addition, both VEGF and Ang-1 are expressed inthe pancreas, suggesting their functional importance inpancreatic tissue angiogenesis / vasculogenesis (Vasir etal, 2001). However, due to limited data in the literature,little is known about the functional interplay betweenVEGF and Ang-1 in islet revascularization posttransplantation.c. Genes involved in islet revascularizationOf the genes whose functions are involved inangiogenesis, VEGF seems to play a crucial role in isletrevascularization. Recent studies by Vasir and colleagues(2000, 2001) indicate that VEGF expression in islet cellsis transiently induced, followed by significant decline twothreedays post transplantation. This impaired expressionof VEGF is further pronounced in the presence ofprevailing hyperglycemia, which coincides with delayedexpression profiles of VEGF receptor molecules, Flk-1/KDR and Flt-1, in islet grafts post transplantation indiabetic animals (Hellerstrom et al, 1898; Korsgren andJansson, 1989; Mattson et al, 2002). These results reflectto some extent an impaired angiogenesis of islet grafts inthe diabetic milieu, which is contributable to the lack ofadequate islet revascularization under hyperglycemicconditions.In addition to VEGF, there are a number of otherangiogenic molecules whose expression in islet cells alsoseems to affect islet revascularization, including fibroblastgrowth factor (FGF), hepatic growth factor (or scatterfactor) (HGF/SF) and its receptor c-Met, transforminggrowth factor-α (TGF-α) and -β (TGF-β), and urokinaseplasminogen activator (uPA) and its receptor uPAR. LikeVEGF, FGF appears to be a positive regulator ofangiogenesis, as it has been shown to induce endothelialcell proliferation, migration and angiogenesis (Bikfalvi etal, 1997; Vasir et al, 2000, 2001, Kawakami et al, 2001).Regarding the function of TGF in angiogenesis, TGF-αhas been shown to stimulate the growth of microvascularendothelial cells (Tokuda et al, 2003). In addition, TGF-αis also a potent inducer of VEGF (Gille et al, 1997; Li etal, 2003). On the other hand, TGF-β is found to stimulatewound healing and regulate differentiation of certain celltypes (Chegini, 1997; Asplin et al, 2001; Li et al, 2003).Although FGF and TGF have been implicated to playimportant roles in angiogenesis (Vasir et al, 2000;Kawakami et al, 2001), their functional contributions toislet revascularization remain unknown.HGF/SF is a mitogen that acts to stimulate celldivision and proliferation of a variety of cell types,including smooth muscle cells and pericytes that arefunctionally involved in blood vessel formation (Bussolinoet al, 1992; Ahmet et al, 2003; Ding et al, 2003; Senguptaet al, 2003). In addition, it has recently been shown thatelevated HGF production in islet grafts significantlyimproves the outcome of marginal islet transplantation dueto its proliferative effect on islet cells (Garcia-Ocana et al,2003). c-Met is a tyrosine kinase receptor of HGF/SF,which is expressed in endothelial cells. In concert with theaction of HGF/SF, the islet-specific expression of c-Metfunctions to mediate the mitogenic effect of HGF/SF onislet cell growth and proliferation (Weidner et al, 1993;Rosen et al, 1997). Vasir and colleagues (2000) showedthat the expression of HGF/SF together with its receptor innewly transplanted islets is profoundly delayed in diabeticanimals (Laing et al, 2003), which correlates with reducedislet graft vascularization. Nevertheless, its specific role inislet revascularization has not been defined.The urokinase plasminogen activator system,consisting of uPA and uPAR, plays a pivotal role inangiogenic sprouting. uPA binds to its cell surfacereceptor uPAR and converts plasminogen to plasmin, aserine protease with a broad specificity that functions tocatalyze the degradation of extracellular matrix/basementmembrane, an essential process that is required forclearing a path to facilitate endothelial cell migration andtissue remodeling in an angiogenic cascade (Saksela andRifkin, 1988; Bacharach et al, 1992; Pepper et al, 1993).Consistent with their roles in angiogenesis, both uPA anduPAR expression are stimulated by VEGF and HGF/SF(Pepper et al, 1992; Mandriota et al, 1995). Like otherangiogenic molecules, the expression of uPA and uPAR innewly engrafted islets is significantly delayed (Vasir et al,2000). It has been suggested that impaired uPA and uPARexpression in newly transplanted islets also contributes toinsufficient islet revascularization under diabeticconditions.4. Factors affecting islet revascularizationAs discussed above, islet revascularization is animportant determinant for the clinical outcome of islettransplantation. Unfortunately, transplanted islets areinvariably associated with markedly reducedrevascularization no matter whether islets are transplantedin the renal, splenic or hepatic subcapsular space (Janssonand Carlsson, 2002). What are the factors that adverselyaffect islet revascularization?.One potential factor that affects isletrevascularization is the presence of prevailinghyperglycemia in diabetic recipients. Data in support ofthis view have been obtained by Vasir et al. (2000, 2001),who showed that the expression of several key angiogenic156

Gene Therapy and Molecular Biology Vol 7, page 157proteins and their respective receptor molecules in newlyengrafted islets is significantly delayed in diabeticrecipient mice, compared to that in nondiabetic recipientmice. These results suggest that islets transplanted underthe renal capsule in a diabetic environment fare less wellin terms of graft vascularization than those transplanted ina normoglycemic subject. In contrast, a different view ofthe possible impact of prevailing hyperglycemia on isletrevascularization is provided by Menger et al, (1992), whoshowed that the relative microvascular blood perfusion isequivalent in islets engrafted in the striated skin muscle inhyperglycemic and normoglycemic Syrian goldenhamsters. Unfortunately, there is no quantitative dataregarding the functional vascular density in islet grafts inrelation to the presence or absence of persistenthyperglycemia provided in these studies. Thus, whetherand to what extent prevailing hyperglycemia affects isletrevascularization still remain an issue of debate.A second factor that may potentially influence isletrevascularization is the use of immunosuppressive agentsassociated with islet transplantation. One outstandingconcern is that immunosuppressive agents are commonlyassociated with anti-proliferative activity and their clinicalapplication in conjunction with islet transplantation mayadversely affect islet revascularization. Theimmunosuppressants, sirolimus and tacrolimus, are shownto inhibit angiogenesis in a dose-dependent manner in bothin vitro and in vivo angiogenesis assays (Eckhard et al,2003). In the same sensitive assays, cyclosporine andprednisolone are also found to retain anti-angiogenicactivities in counteracting the proliferative effect of FGFin angiogenesis (Eckhard et al, 2003), although it has beenpreviously reported that the application of cyclosporin-Adoes not seem to alter microvascular perfusion to isletgrafts (Mendola et al, 1997; Vajkoczy et al, 1999). Theseresults raise a great deal of concern that clinicalapplication of immunosuppressive drugs, which isintended to prevent islet graft loss, may actuallycompromise the viability of newly transplanted islets byhampering the process of islet revascularization.A third limiting factor for islet revascularization isthe presence of contaminating exocrine cells in isolatedislets, including macrophage, dendritic cells (DC) andendothelial cells. It has been suggested that exocrine cellsperturb angiogenesis and islet revascularization (Heuser etal, 2000; Jansson and Carlsson, 2002). Consistent with thisidea is the observation that culturing of islets prior totransplantation tends to improve the outcome of islettransplantation, as culturing helps eliminate contaminatingcells, in particular, the antigen presenting cells (APC) inislet preparation (Gaber et al, 2001; Kuttler et al, 2002).However, culturing of freshly isolated islets also results inthe loss of endothelium in islets. Interestingly, recentstudies show that intra-islet endothelial cells serve asintegrated components in angiogenesis and functiontogether with recipient endothelium to facilitate the overallislet graft vascularization (Brissova et al, 2003; Linn et al,2003). These results suggest that transplantation of freshlyisolated islets may be favorable for islet viability becauseof the functional contribution of intra-islet endothelialcells to islet revascularization post transplantation(Jansson and Carlsson, 2002).Finally, a less well-characterized factor that mightaffect islet revascularization is islet cryopreservation. Thisprocess is necessary as it can afford a great deal offlexibility and additional advantages to clinical islettransplantation. Cryopreservation allows pooling ofmarginal islets and subsequent distribution of islets todifferent islet transplantation centers/hospitals. It alsoallows sufficient time for pre-transplantation qualitycontrol testing of isolated islets to ensure islet cell viabilityand microbiological sterility prior to transplantation. Inaddition, cryopreservation also allows for geneticmodification of islets by introducing angiogenic,cytoprotective or immunomodulatory genes via genetransfer to islets prior to islet transplantation to improvethe clinical outcome of islet transplantation in the future.However, recovery of functional islets aftercryopreservation has been technically challenging, asfreezing and thawing can significantly reduce the viabilityof islet cells (Kuo et al, 2002). Up to 50% of functionalislet loss has been reported after cryopreservation (Lakeyet al, 2001). Furthermore, the extent to whichcryopreservation affects islet revascularization remains tobe determined.B. Enhancing islet revascularization1. Angiogenic gene transfer to enhance isletrevascularizationAs discussed above, rapid and sufficient isletrevascularization is crucial for long-term survival andfunction of islet grafts post transplantation. Delayed andinadequate revascularization of newly transplanted isletscan deprive islet cells of oxygen and nutrients, resulting inislet cell death and premature graft failure. Given the factthat successful islet transplantation depends on theinfusion of sufficiently large amounts of islets, whichusually requires at least two cadaveric pancreata perrecipient, increased islet revascularization is expected toreduce the number of islets and improve the pancreasdonor to recipient ratio required for transplantation. Inaddition, rapid and adequate islet revascularization willprotect islet grafts from hypoxia-induced inflammationand necrosis, thereby improving long-term graft survivaland providing better preservation of functional islet mass.However, only limited efforts have been made in the pastin this aspect of islet transplantation.VEGF is known to play a pivotal role inangiogenesis / vasculogenesis. To investigate itsangiogenic effect on islet revascularization, Sigrist andcolleagues (2002) have applied collagen-immobilizedVEGF protein in encapsulated islets, followed bytransplantation into the peritoneal cavity of streptozotocininduceddiabetic mice. Blood glucose and plasma insulinlevels were determined and animals were sacrificed twoweeks post transplantation. It was found that isletstransplanted in the presence of collagen-immobilizedVEGF protein show significantly increased angiogenesisand microvasculature in islet grafts, which associated withincreased insulin production and improved glycemic157

Zhang et al: Angiogenic Gene Therapy for Improving Islet Graft Vascularizationcontrol, in comparison to control islets that aretransplanted in the absence of VEGF protein. These resultssuggest that local VEGF delivery to islet grafts improvesthe outcome of islet transplantation by enhancing isletrevascularization (Sigrist et al, 2002).To improve islet graft vascularization, we havedelivered the human vascular endothelial growth factor(hVEGF) cDNA by adenoviral-gene transfer to mouseislets, followed by transplantation under the renal capsulein streptozotocin-induced diabetic mice (Zhang et al,2003). We showed that all the renal capsules containingthe hVEGF vector-transduced islets (250 islets) displayedsignificantly higher functional islet mass, as measured byinsulin immunostaining, and greater vascular density, asdetermined by immunostaining of CD31, the plateletendothelial cell adhesion molecule-1 (PECAM-1)(Watanabe et al, 2000). As a result, diabetic micereceiving the hVEGF vector-treated islets exhibitednormoglycemia with improved glucose tolerance. Incontrast, diabetic mice receiving an equivalent islet massthat were pre-transduced with a control vector maintainedmoderate hyperglycemia with impaired glucose tolerance.These results provide the proof-of-principle thatangiogenic gene transfer to islets prior to islettransplantation allows local production of VEGF in isletgrafts, which in turn stimulates graft angiogenesis andaugments islet revascularization (Zhang et al, 2003).While therapeutic angiogenesis, so called biobypass,has been considered an alternative modality for treatingcoronary and peripheral artery diseases, based on theefficacy and safety of plasmid- or adenoviral vectormediatedVEGF delivery in angiogenesis in a number ofpreclinical studies and clinical trials (Isner, 2002;Koransky et al, 2002; Mercadier and Logeart, 2002;Rasmussen et al, 2002; Sylven, 2002, Khan et al, 2003;Kusumanto et al, 2003), our view is that a similarangiogenic strategy should be explored to accelerate isletgraft angiogenesis, allowing rapid and adequate isletrevascularization post transplantation. Such an approach,when used in conjunction with islet transplantation, hasthe potential for improving the success rate and clinicaloutcome of islet transplantation with long-term glycemiccontrol at a reduced cost of islets.2. Ex vivo gene delivery to isletsThe rationale for enhancing islet graftvascularization by angiogenic gene transfer is as follows:islets are transduced in culture with a vector expressingangiogenic molecules, such as VEGF, followed bytransplantation into a diabetic subject, as illustratedschematically in Figure 2. Using an adenoviral-mediatedgene delivery system, we have validated this concept byshowing that VEGF production in newly transplantedislets significantly improves islet revascularization andfunctional islet mass (Zhang et al, 2003). It is noteworthythat adenoviral vectors are associated withimmunogenecity. In addition, islets are terminallydifferentiated post-mitotic cells, which poses a greatchallenge for ex vivo gene delivery to islets by vectorswhose transduction depends on cell division (Ito andKedes, 1997; Robbins and Ghivizzani, 1998). However,recent advances in both viral and nonviral vectordevelopment have made it feasible to transfer genes tointact islets ex vivo at reasonable efficiencies withoutadversely affecting the architecture and function of islets.Below is a focused review of a number of vector systemsthat are currently in use for ex vivo gene transfer toisolated islets.Figure 2. Schematic representation of angiogenic gene transfer in conjunction with islet transplantation. Islets are isolated and incubatedin culture media in the presence of a gene vector that expresses angiogenic molecules. After transduction, islets are transplantedintraportally into the liver of a diabetic subject.158

Gene Therapy and Molecular Biology Vol 7, page 159a. Adenovirus-mediated gene transfer to isletsAdenovirus is the most commonly used vectorsystem in preclinical studies due to its relatively hightransduction efficiency for both dividing and nondividingcell types. Adenovirus is capable of accommodating largeDNA inserts and can be produced in a large quantity andat a relatively high titer. Although adenoviral vectors havebeen shown to efficiently transduce islets without alteringglucose-inducible insulin secretion from β cells (Newgard,1994; Csete et al, 1995; O'Brien et al, 1999), recent studiesindicate that adenoviral-mediated transduction of isletsinduces the production of a number of chemokines andtheir respective receptors, resulting in subsequentrecruitment of inflammatory cells to islet grafts. This maypotentially impair islet engraftment (Zhang et al, 2003).b. rAAV-mediated gene delivery to isletsRecombinant adeno-associated virus (rAAV) hasbecome the vector of choice for gene transfer to a varietyof cell types because of its ability to mediate long-termtransgene expression in the absence of cytotoxicity (Flotteet al, 2001; Kapturczak et al, 2002; Mah et al, 2002;Vizzardelli et al, 2002). The most commonly used rAAVis derived from AAV-2, an AAV serotype that belongs toa group of non-pathogenic human parvoviruses. AAV-2contains a 4.7-kb single-stranded genome encoding viralreplication (rep) and capsid (cap) genes flanked byinverted terminal repeat sequences (ITRs) (Srivastava etal, 1994). Productive replication of AAV-2 depends onadenoviral or herpes viral helper functions, in the absenceof which, AAV2 establishes a "rep-dependent" latentinfection by integrating its genome site-specifically intothe AAVS1 site in human chromosome 19 (Kotin et al,1992; Rabinowitz and Samulski, 1998). In rAAV-2vectors, the entire viral coding sequences are replaced withthe therapeutic gene of interest (insertion size

Zhang et al: Angiogenic Gene Therapy for Improving Islet Graft VascularizationFigure 3. Ex vivo transduction of murine and human islets by rAAV. Prior to exposure to rAAV, islets were incubated with a helperadenovirus (Adv-5) at an MOI of 5 pfu/cell for 2 h in CMRL-1066 medium (Sigma-Aldrich, St. Louis, MO) in a 37 _C incubator with5% CO 2 . Subsequently, islets were transduced with the rAAV-GFP vector expressing the green fluorescent protein at an MOI of 1,000pfu/cell and visualized in a fluorescent microscope. One islet contains about 1,000 cells on average. Shown are murine islets that weremock-treated (A) and rAAV1-GFP transduced (B), as well as human islets that were mock-treated (C) and rAAV2-GFP transduced (D)...Figure 4. Lentiviral-mediated transduction of islets. Freshly isolated mouse islets were mock-transduced (A) and transduced with theFIV-LacZ vector at an MOI of 100 transducing units/cell (B) and stained for β-gal after 24 h of incubation in the CMRL-1066 medium.In addition, after transduction with the FIV-LacZ vector, islets were paraffin-embedded and thin-sections of embedded islets wereimmuno-stained for insulin (C, brown) and stained with X-gal for β-gal (D, blue). Bar, 25 µm.d. Nonviral vector-mediated gene transfer to isletsIn addition to viral-mediated gene delivery systems,nonviral systems such as liposome-mediated transfectionhave been used to deliver genes to a variety of cells bothin vitro and in vivo (Ledley et al, 1995). Cationicliposomes are artificial membrane vesicles that cancomplex with DNA. The resulting liposome-DNAcomplex is thought to fuse with the negatively chargedplasma membrane (Felgner and Ringold, 1991) or becomeendocytosed (Zhou and Huang, 1994), resulting in genedelivery to the nucleus.It has been shown that islet cells in a monolayerderived from dispersed islets or intact islets can beeffectively transduced using the monoliposomal reagentLipofectin or the polycationic liposome Lipofectamine oradenovirus-polylysine (AdpL) DNA complexes (Welsh etal, 1990; Welsh and Andersson, 1994; Saldeen et al, 1996;Benhamou et al, 1997). Recently, Mahato and colleagues(Mahato et al, 2003) reported that human islets transducedwith the hVEGF gene by nonviral-mediated gene transferresulted in sustained hVEGF production for up to 10 dayspost transduction. Although nonpathogenic, nonviralmediatedgene transfer is in general associated with arelatively low efficiency and short duration of transgeneexpression (Lakey et al, 2001). It has been suggested thatafter liposome-mediated endocytosis, a vast majority oflipid-DNA particles are retained in the perinuclear areaand subsequently degraded (Zabner et al, 1995). Thus, the160

Gene Therapy and Molecular Biology Vol 7, page 161failure of DNA to leave the endosomal compartmentrepresents a major hurdle to liposome-mediated genetransfer. Nonviral-mediated gene transfer systems are of apreferred choice when persistent transgene expression isnot desirable.Recently, a novel system, known as proteintransduction, is being developed. Unlike gene transfersystems, this protein transduction system allows selectivedelivery of proteins into cells, when linked to a specificprotein transduction domain (PTD). PTD is a smallpeptide domain that can freely cross the cytoplasmicmembrane through a receptor-mediated process, which isindependent of ATP (Hawiger et al, 1999; Schwarze et al,2000). In particular, a PTD designated PTD-5, which isoriginally selected from an M13 phage peptide displaylibrary, has been reported to successfully transduce bothhuman and mouse islets without significant effects on isletfunction (Mi et al, 2000; Rehman et al, 2003). Likewise,Embury et al, (2001) also showed that a small peptide of11 amino acid residues that constitute the PTD of theHIV/TAT protein, when fused to β-galactosidase, is ableto transduce rat islets ex vivo with the fusion protein in adose-dependent manner at a relatively high efficiency.However, such a protein transduction system is normallyassociated with a transient effect, depending on therelative stability of the fusion protein. In addition, fortherapeutic protein delivery, caution should be taken toascertain that the fusion of a PTD does not adverselyaffect the proper folding and compromise the function ofthe therapeutic protein..III. ConclusionRapid re-establishment of an appropriatemicrovascular system in newly transplanted islets iscrucial for survival and function of islet grafts.Unfortunately, islets implanted at ectopic sites, such asunder the renal capsule or in the liver and spleen, areinvariably associated with markedly reducedvascularization, in comparison with native islets in thepancreas (Beger et al, 1998; Mattson et al, 2002). Thisimpairment in islet revascularization accounts at least inpart for the demand of sufficiently large quantities of isletmass for restoration of normoglycemia in type 1 diabeticsubjects. In addition, delayed and inadequate islet graftvascularization can deprive islets of oxygen and nutrients,causing islet cells to undergo cellular apoptosis andsubsequent cell death, particularly in the core of largeislets or in the center of aggregated islet clusters posttransplantation. Moreover, a lack of sufficient isletrevascularization may also compromise the optimalperformance of transplanted islets. Indeed, there areclinical data indicating that even after postabsorptiveblood glucose homeostasis is restored to normal post islettransplantation, implanted islets do not seem to function atoptimal levels, as reflected in their significantly impairedglucose tolerance in diabetic recipients in response tointravenous glucose challenge (Ryan et al, 2001, 2002).Thus, it is of great significance to define the molecularmechanism of islet revascularization and developtherapeutic angiogenesis approaches to enhance theprocess of islet revascularization. Such approaches areexpected to ensure adequate microvascular perfusion toislet cells and protect implanted islet cells from hypoxiainducedinflammation and necrosis, which will ultimatelyimprove the outcome of islet transplantation by reducingthe donor/recipient ratio thus increasing the success rate ofislet transplantation.AcknowledgementWe would like to thank Marcia Meseck for criticalreading of this manuscript. This project is supported partlyby the Juvenile Diabetes Research Center at Mount SinaiSchool of Medicine.ReferencesAhmet I, Sawa Y, Yamaguchi T, Matsuda H (2003) Genetransfer of hepatocyte growth factor improves angiogenesisand function of chronic ischemic myocardium in canineheart. Ann Thorac Surg 75, 1283-1287.Asplin IR, Wu SM, Mathew S, Bhattacharjee G, Pizzo SV (2001)Differential regulation of the fibroblast growth factor (FGF)family by alpha(2)-macroglobulin: evidence for selectivemodulation of FGF-2-induced angiogenesis. Blood 97, 3450-3457.Bacharach E, Itin A, Keshet E (1992) In vitro patterns ofexpression of urokinase and its inhibitor PA-1 suggest aconcerted role in regulating physiological angiogenesis. ProcNatl Acad Sci U S A. 89, 10686-10690.Beger C, Cirulli V, Vajkoczy P, Halban PA, Menger MD (1998)Vascularization of purified pancreatic islet-like cellaggregates (pseudoislets) after syngerneic transplantation.Diabetes 47, 559-565.Benhamou PY, Moriscot C, Prevost P, Rolland E, Hallmi S,Chroboczek J (1997) Standardization of procedure forefficient ex vivo gene transfer into porcine pancreatic isletswith cationic liposomes. Transplantation 63, 1798-1803, .Bikfalvi A, Klein S, Pintucci G, Rifkin DB (1997) Biologicalroles of fibroblast growth factor-2. Endocrine Reviews 18,26-45.Bloomgarden ZT (1998) The Epidemiology of Complications.Diabetes Care 25, 924-932.Bloomgarden ZT (2002) Treatment Issues in Type 1 Diabetes.Diabetes Care 25, 230-238.Boker A, Rothenberg L, Hernandez C, Kenyon NS, Ricordi C,Alejandro R (2001) Human islet transplantation: update.World J Surg 25, 481-486.Brissova M, Fowler M, Shiota M, Radhika A, Shostak A, WiebeP, Gannon M, Powers A (2003) Intra-islet endothelial cellscontributor to revascularization of transplanted pancreaticislets. Diabetes 52 (suppl. 1), A13.Bussolino F, Di Renzo MF, Ziche M, Bocchietto E, Olivero M,Naldini L, Gaudino G, Tamagnone L, Coffer A, ComoglioPM (1992) Hepatocyte growth factor is a potent angiogenicfactor which stimulates endothelial cell motility and growth.J Cell Biol 119, 629-64.Carlsson P-O, Palm F, Andersson A, Liss P (2000) Chronicallydecreased oxygen tension in rat pancreatic islets transplantedunder the kidney capsule. Transplantation 69, 761-766.Carlsson P-O, Palm F, Andersson A, Liss P (2001) Markedlydecreased oxygen tension in transplanted rat pancreatic isletsirrespective of implantation sites. Diabetes 50, 489-495.161

Zhang et al: Angiogenic Gene Therapy for Improving Islet Graft VascularizationChang LJ, Gay EE (2001) The molecular genetics of lentiviralvectors-current and future perspectives. Curr Gene Ther 1,237-251.Chaturvedi N, Bandinelli S, Mangili R, Penno G, Rottiers RE,Fuller JH (2000) Microalbuminuria in type 1 diabetes: rates,risk factors and glycemic threshold. Kidney Int 60,219–227.Chegini N (1997) The role of growth factors in peritonealhealing: transforming growth factor beta (TGF-beta). Eur JSurg Suppl 577, 17-23 .Cote I, Rogers NJ, Lechler RI (2001) Allorecognition. TransfusClin Biol. 8, 318-323.Csete ME, Benhamou PY, Drazan KE, Wu L, McIntee DF, AfraR, Mullen Y, Busuttil RW, Shaked A (1995) Efficient genetransfer to pancreatic islets mediated by adenoviral vectors.Transplantation 59, 263-269.Curran MA, Kaiser SM, Achacoso PL, Nolan GP (2000)Efficient transduction of nondividing cells by optimizedfeline immunodeficiency virus vectors. Mol Ther 1, 31-38.Curran MA, Ochoa MS, Molano RD, Pileggi A, Inverardi L,Kenyon NS, Nolan GP, Ricordi C, Fenjves ES (2002)Efficient transduction of pancreatic islets by felineimmunodeficiency virus 1 vectors. Transplantation 74, 299-306.Ding S, Merkulova-Rainon T, Han ZC, Tobelem G (2003) HGFreceptor up-regulation contributes to the angiogenicphenotype of human endothelial cells and promotesangiogenesis in vitro. Blood 101, 4816-4822.Djalilian HR, Tsuboi Y, Ozeki M, Tekin M, Djalilian AR,Obritch W, Lin J (2002) Feline immunodeficiency virusmediatedgene therapy of middle ear mucosa cells. AurisNasus Larynx 29, 183-186.Eckhard M, Erb D, Bretzel R, Brendel M, Linn T (2003)Inhibition of angiogenesis by immunosuppressive agentsused in clinical transplantation. Diabetes 52 (supppl. 1),A14.Embury J, Klein D, Pileggi A, Ribeiro M, Jayaraman S, MolanoRD, Fraker C, Kenyon N, Ricordi C, Inverardi L, Pastori RL(2001) Proteins linked to a protein transduction domainefficiently transduce pancreatic islets. Diabetes 50, 1706-1713.Felgner PL, Ringold GM (1991) Cationic liposome-mediatedtransfection. Nature 337, 387-388.Ferrara N, Davis-Smyth T (1997) The biology of vascularendothelial growth factor. Endocrine Rev 18, 4-25, .Flotte T, Agarwal A, Wang J, Song S, Fenjves ES, Inverardi L,Chesnut K, Afione S, Loiler S, Wasserfall C, Kapturczak M,Ellis T, Nick H, Atkinson M (2001) Efficient ex vivotransduction of pancreatic islet cells with recombinant adenoassociatedvirus vectors. Diabetes 50, 515-520.Gaber AO, Fraga DW, Callicutt CS, Gerling IC, Sabek OM,Kotb MY (2001) Improved in vivo pancreatic islet functionafter prolonged in vitro islet culture. Transplantation 72,1730-1736.Gale EAM (2002) The Rise of Childhood Type 1 Diabetes in the20th Century. Diabetes 51, 3353–3361.Gao GP, Alvira MR, Wang L, Calcedo R, Johnston J, Wilson JM(2002) Novel adeno-associated viruses from rhesus monkeysas vectors for human gene therapy. Proc Natl Acad Sci U SA. 99, 11854-11859.Garcia-Ocana A, Takane KK, Reddy VT, Lopez-Talavera JC,Vasavada RC, Stewart AF (2003) Adenovirus-mediatedhepatocyte growth factor expression in mouse isletsimproves pancreatic islet transplant performance and reducesbeta cell death. J Biol Chem 278, 343-351.Giannoukakis N, Mi Z, Gambotto A, Eramo A, Ricordi C,Trucco M, Robbins PD (1999) Infection of intact humanislets by a lentiviral vector. Gene Ther 6, 1545-1551.Gille J, Swerlick RA, Caughman SW (1997) Transforminggrowth factor-alpha-induced transcriptional activation of thevascular permeability factor (VPF/VEGF) gene requires AP-2-dependent DNA binding and transactivation. EMBO J 16,750-759 .Girod A, Ried M, Wobus C, Lahm H, Leike K, Kleinschmidt J,Deleage G, Hallek M (1999) Genetic capsid modificationsallow efficient re-targeting of adeno-associated virus type 2.Nat Med 5, 1052-1056.Hansen J, Qing K, Srivastava A. (2001) Adeno-associated virustype 2-mediated gene transfer: altered endocytic processingenhances transduction efficiency in murine fibroblasts. JVirol 75, 4080-4090.Hawiger J (1999) Noninvasive intracellular delivery offunctional peptides and proteins. Curr. Opin. Chem. Biol. 3,89-94.Hellerstrom C, Andersson A, Korsgren O, Jansson L, S. S (1898)Aspects of pancreatic islet transplantation in diabetesmellitus. Baillieres Clin Gastroenterol. 3, 851-863.Heuser M, Wolf B, Vollmar B, Menger MD (2000) Exocrinecontamination of islolated islets of Langerhans deterioratesthe process of revascularization after free transplantation.Transplantation 69, 756-761.Hughes SM, Moussavi-Harami F, Sauter SL, Davidson BL(2002) Viral-mediated gene transfer to mouse primary neuralprogenitor cells. Mol Ther 5, 16-22.Isner JM (2002) Myocardial gene therapy. Nature 415, 234-239.Ito M, Kedes L (1997) Two-step delivery of retroviruses topostmitotic, terminally differentiated cells. Hum Gene Ther8, 57-63.Jansson L, Carlsson P-O (2002) Graft vascular functioin aftertransplantation of pancreatic islets. Diabetologia 45, 749-763.Ju Q, Edelstein D, Brendel MD, Brandhorst D, Brandhorst H,Bretzel RG, Brownlee M (1998) Transduction of nondividingadult human pancreatic beta cells by an integratinglentiviral vector. Diabetologia. 41, 736-739.Kapturczak M, Zolotukhin S, Cross J, Pileggi A, Molano RD,Jorgensen M, Byrne B, Flotte TR, Ellis T, Inverardi L,Ricordi C, Nick H, Atkinson M, Agarwal A (2002)Transduction of human and mouse pancreatic islet cells usinga bicistronic recombinant adeno-associated viral vector. MolTher 5, 154-160.Karvonen M, Viik-Kajander M, Moltchanova E, Libman I,LaPorte R, Tuomilehto J (2000) Incidence of childhood type1 diabetes worldwide. Diabetes Mondiale (DiaMond) ProjectGroup. Diabetes Care. 23, 1516-1526.Kawakami Y, Iwata H, Gu YJ, Miyamoto M, Murakami Y,Balamurugan AN, Imamura M, Inoue K (2001) Successfulsubcutaneous pancreatic islet transplantation using anangiogenic growth factor-releasing device. Pancreas 23,375-381.Kay MA, Glorioso JC, Naldini L (2001) Viral vectors for genetherapy: the art of turning infectious agents into vehiches oftherapeutics. Nat Med 7, 33-40.Khan TA, Sellke FW, Laham RJ (2003) Gene therapy progressand prospects: therapeutic angiogenesis for limb andmyocardial ischemia. Gene Ther 10, 285-291.162

Gene Therapy and Molecular Biology Vol 7, page 163Kim I, Kim HG, Moon SO, Chae SW, So JN, Koh KN, Ahn BC,Koh GY (2000) Angiopoietin-1 induces endothelial cellsprouting through the activation of focal adhesion kinase andplasmin secretion. Circ Res 86, 952-959.Koransky ML, Robbins RC, Blau HM (2002) VEGF genedelivery for treatment of ischemic cardiovascular disease.Trends Cardiovasc Med 12, 108-114.Korsgren O, Jansson L, AA (1989) Effects of hyperglycemia onfunction of isolated mouse pancreatic islets transplantedunder kidney capsule. Diabetes 38, 510-515.Kotin RM, Linden RM, Berns KI (1992) Characterization of apreferred site on human chromosome 19q for integration ofadeno-associated virus DNA by non-homologousrecombination. EMBO J. 11, 5071-5078.Kuo CH, Juang JH, Hsu BR, Lu WT, Yao NK (2002) In vitroand in vivo studies on the cryopreserved pancreatic islets.Transplant Proc 34, 2693-2695.Kusumanto YH, Hospers GA, Mulder NH, Tio RA (2003)Therapeutic angiogenesis with vascular endothelial growthfactor in peripheral and coronary artery disease: a review. IntJ Cardiovasc Intervent. 5, 27-34.Kuttler B, Hartmann A, Wanka H (2002) Long-term culture ofislets abrogates cytokine-induced or lymphocyte-inducedincrease of antigen expression on beta cells.Transplantation 74, 440-445.Laing SP, Swerdlow AJ, Slater SD, Burden AC, Morris A,Waugh NR, Gatling W, Bingley PJ, Patterson CC (2003)Mortality from heart disease in a cohort of 23,000 patientswith insulin-treated diabetes. Diabetologia 760-765.Lakey JR, Anderson TJ, Rajotte RV (2001) Novel approaches tocryopreservation of human pancreatic islets.Transplantation 72, 1005-1011.Lakey JRT, Young ATL, Pardue D, Calvin S, Albertson TE,Jacobson L, Cavanagh TJ (2001) Non-viral transfection ofintact pancreatic islets. Cell Transplant 10, 697-708.Lechler RI, Garden OA, Turka LA (2003) The complementaryroles of deletion and regulation in transplantation tolerance.Nat Rev Immunol 2, 147-158.Lechner A, Habener JF (2003) Stem/progenitor cells derivedfrom adult tissues: potential for the treatment of diabetesmellitus. Am J Physiol Endocrinol Metab 284, E259-266.Ledley FD (1995) Nonviral gene therapy: the promise of genesas pharmaceutical products. Hum Gene Ther 6, 1129-1144.Leibowitz G, Beattie GM, Kafri T, Cirulli V, Lopez AD, HayekA, Levine F (1999) Gene transfer to human pancreaticendocrine cells using viral vectors. Diabetes 48, 745-753.Li J, Zhang YP, Kirsner RS (2003) Angiogenesis in woundrepair: angiogenic growth factors and the extracellularmatrix. Microsc ResTech 60, 107-114.Linn T, Schneider K, Hammes HP, Preissner KT, Brandhorst D,Morgenstern E, Kiefer F, Bretzel RG (2003) Angiogeniccapacity of endothelial cells in islets of Langerhans. FASEBJ. 17, 881-883.Lumelsky N, Blondel O, Laeng P, Velasco I, Ravin R, McKay R(2001) Differentiation of embryonic stem cells to insulinsecretingstructures similar to pancreatic islets. Science 292,1389-1394.Mah C, Byrne BJ, Flotte TR (2002) Virus-based gene deliverysystems. Clin Pharmacokinet 41, 901-911.Mahato RI, Henry J, Narang AS, Sabek O, Fraga D, Kotb M,Gaber AO (2003) Cationic lipid and polymer-based genedelivery to human pancreatic islets. Mol Ther 7, 89-100.Mandriota SJ, Seghezzi G, Vassalli J-D, Ferrara N, Wasi S,Mazzieri R, Mignatti P, Pepper MS (1995) Vascularendothelial growth factor increases urokinase receptorexpression in vascular endothelial cells. J Biol Chem 270,9709-9716.Mattson G, Jansson L, Carlsson P-O (2002) Decreased vasculardensity in mouse pancreatic islets after transplantation.Diabetes 51, 1362-1366.Mendola JF, Goity C, Esmatjes E, Seanz A, Fernandez-Cruz L,Gomis R (1997) Cyclosporine does not inhibit the process ofvascularization of pancreatic islet transplantation. CellTransplant 6, 69-76 .Menger MD, Vajkoczy P, Leiderer R, Jager S, Messmer K(1992) Influence of experimental hyperglycemia onmicrovascular blood perfusion of pancreatic islet isografts. JClin Invest 90, 1361-1369.Menger MD, Yamauchi J-I, Vollmar B (2001) Revascularizationand microcirculation of freshly grafted islets of Langerhans.World J Surg 25, 509-515.Mercadier JJ, Logeart D (2002) Mycardial gene therapy. Arch.Mal. Coeur. Vaiss. 95, 197-203.Mi Z, Mai J, Lu X, Robbins PD (2000) Characterization of aclass of cationic peptides able to facilitate efficient proteintransduction in vitro and in vivo. Mol Ther 2, 339-347.Mingozzi F, Schuttrumpf J, Arruda VR, Liu Y, Liu YL, HighKA, Xiao W, Herzog RW (2002) Improved hepatic genetransfer by using adeno-associated virus serotype 5 vector. JVirol 76, 10497-10502.Murphy NP, Keane SM, Ong KK, Ford-Adams M, Edge JA,Acerini CL, Dunger DB (2003) Randomized Cross-OverTrial of Insulin Glargine Plus Lispro or NPH Insulin PlusRegular Human Insulin in Adolescents With Type 1 Diabeteson Intensive Insulin Regimens. Diabetes Care 26, 799-804.Newgard CB (1994) Cellular engineering and gene therapystrategies for insulin replacement in diabetes. Diabetes 43,341-350.O'Brien T, Karlsen AE, Andersen HU, Mandrup-Poulsen T,Nerup J (1999) Absence of toxicity associated withadenoviral-mediated transfer of the β-galactosidase reportergene to neonatal rat islets in vitro. Diab Res & Clin Pract44, 157-163.Otani A, Takagi H, Oh H, Koyama S, Matsumura M, Honda Y(1999) Expressions of angiopoietins and Tie2 in humanchoroidal neovascular membranes. Invest Ophthalmol VisSci 40, 1912-1920.Pepper MS, Matsumoto K, Nakamura T, Orci L, Montesano R(1992) Hepatocyte growth factor increases urokinase-typeplasminogen activator (uPA) and u-PA receptor expressionon Madin-Darby canine kidney epithelial cells. J Biol Chem267, 20493-20496 .Pepper MS, Sappino A-P, Stocklin R, Montesano R, Orci L,Vassalli J-D (1993) Upregulation of urokinase receptorexpression on migrating endothelial cells. J Cell Biol 122,673-684 .Plank J, Wutte A, Brunner G, Siebenhofer A, Semlitsch B,Sommer R, Hirschberger S, Pieber TR (2002) A directcomparison of insulin aspart and insulin lispro in patientswith type 1 diabetes. Diabetes Care. 25, 2053-2057.Qing K, Li W, Zhong L, Tan M, Hansen J, Weigel-Kelley KA,Chen L, Yode MC, Srivastava A (2003) Adeno-associatedvirus type 2-mediated gene transfer: role of cellular T-cellprotein tyrosine phosphatase in transgene expression inestablished cell lines in vitro and transgenic mice in vivo. JVirol 77, 2741-2746.163

Zhang et al: Angiogenic Gene Therapy for Improving Islet Graft VascularizationRabinowitz JE, Rolling F, Li C, Conrath H, Xiao W, Xiao X,Samulski RJ (2002) Cross-packaging of a single adenoassociatedvirus (AAV) type 2 vector genome into multipleAAV serotypes enables transduction with broad specificity. JVirol 76, 791-801.Rabinowitz JE, Samulski J (1998) Adeno-associated virusexpression systems for gene transfer. Curr. Opin.Biotechnol. 9, 470-475.Rasmussen HS, Rasmussen CS, Macko J (2002) VEGF genetherapy for coronary artery disease and peripheral vasculardisease. Cardiovasc. Radiot. Med. 3, 114-117.Rehman KK, Bertera S, Bottino R, Balamurugan AN, Mai JC,Mi Z, Trucco M, Robbins PD (2003) Protection of islets byin situ peptide-mediated transduction of the IκB kinaseinhibitor nemo-binding domain peptide. J Biol Chem 278,9862-9868.Robbins PD, Ghivizzani SC (1998) Viral vectors for genetherapy. Pharmacol Ther. 80, 35-47.Rosen EM, Lamszuz K, Laterra J, Polverini PJ, Rubin JS,Goldberg ID (1997) HGF/SF in angiiogenesis. Ciba FoundSymp 212, 215-226, .Ryan EA, Lakey JR, Paty BW, Imes S, Korbutt GS, KnetemaNM, Bigam D, Rajotte RV, Shapiro AM (2002) Successfulislet transplantation: continued insulin reserve provides longtermglycemic control. Diabetes 51, 2148-2157.Ryan EA, Lakey JR, Rajotte RV, Korbutt GS, Kin T, Imes S,Rabinovitch A, Elliott JF, Bigam D, Kneteman NM,Warnock GL, Larsen I, Shapiro AM (2001) Clinicaloutcomes and insulin secretion after islet transplantation withthe Edmonton protocol. Diabetes 50, 710-179.Saksela O, Rifkin DB (1988) Cell-associated plasminogenactivation: regulation and physiologic functions. Ann RevCell Biol 4, 93-126, .Saldeen J, Curiel DT, Eizirik DL, Andersson A, Strandell E,Buschard K, Welsh N (1996) Efficient gene transfer todispersed human pancreatic islet cells in vitro usingadenovirus-polylysine/DNA complexes or polycationicliposomes. Diabetes 45, 1197-1203.Schwarze SR, Hruska KA, Dowdy SF (2000) Proteintransduction: unrestricted delivery into all cells? Trends CellBiol 10, 290-295.Sengupta S, Gherardi E, Sellers LA, Wood JM, Sasisekharan R,Fan TP (2003) Hepatocyte growth factor/scatter factor caninduce angiogenesis independently of vascular endothelialgrowth factor. Arterioscher Thromb Vasc Biol 23, 69-75.Shapiro AM, Lakey JR, Ryan EA, Korbutt GS, Toth E, WarnockGL, Kneteman NM, Rajotte RV (2000) Islet transplantationin seven patients with type 1 diabetes mellitus using aglucocorticoid-free immunosuppressive regimen. N Engl JMed 343, 230-238.Sigrist S, Mechine A, Leenders H, Calanda V, Laurence K(2002) Influence of VEGF on the viability of encapsulatedpancreatic rat islets after transplantation in diabetic mice.Diabetes 51 (suppl. 2), A10.Soria B, Roche E, Berma G, Leon-Quinto T, Reig J, Martin F(2001) Insulin-secreting cells derived from embryonic stemcells normalize glycemia in streptozotocin-induced diabeticmice. Diabetes 49, 157-162.Srivastava A (1994) Parvovirus-based vectors for human genetherapy. Blood Cells 20, 531-536.Srivastava A (2002) Obstacles to human hematopoietic stem celltransduction by recombinant adeno-associated virus 2vectors. J Cell Biochem (Suppl.) 38, 39-45.Sylven C (2002) Angiogenic gene therapy. Drugs Today 38,819-827.Thurston G, Suri C, Smith K, McClain J, Sato TN, YancopoulosGD, McDonald DM (1999) Leakage-resistant blood vesselsin mice transgenically overexpressing angiopoietin-1.Science 286, 2511-2514.Tokuda H, Hatakeyama D, Akamatsu S, Tanabe K, Yoshida M,Shibata T, Kozawa O (2003) Involvement of MAP kinases inTGF-beta-stimulated vascular endothelial growth factorsynthesis in osteoblasts. Arch Biochem Biophys 415, 117-125.Vague P, Selam JL, Skeie S, De Leeuw I, Elte JW, Haahr H,Kristensen A, Draeger E (2003) Insulin detemir is associatedwith more predictable glycemic control and reduced risk ofhypoglycemia than NPH insulin in patients with type 1diabetes on a basal-bolus regimen with premeal insulinaspart. Diabetes Care 26, 590-596.Vajkoczy P, Vollmar B, Wolf B, Menger MD (1999) Effects ofcyclosporine A on the process of vascularization of freelytransplanted islets of Langerhans. J Mol Med 77, 111-114.Valle D, Santoro D, Bates P, Scarpa L (2001) Italian multicentrestudy of intensive therapy with insulin lispro in 1184 patientswith Type 1 diabetes. Diabetes Nutr. Metab. 14, 126-132.Vasir B, Jonas JC, Steil GM, Hollister-Lock J, Hasenkamp W,Sharma A, S. B-W, Weir GC (2001) Gene expression ofVEGF and its receptors Flk-1/KDR and Flt-1 in cultured andtransplanted rat islets. Transplantation 71, 924-935.Vasir B, Reitz P, Xu G, Sharma A, Bonner-Weir S, Weir GC(2000) Effects of diabetes and hypoxia on gene markers ofangiogenesis (HGF, cMet, uPA adn uPAR, TGF-α, TGF-β,bFGF and vimentin) in cultured and transplanted rat islets.Diabetologia 43, 763-772.Vigna E, Naldini L (2000) Lentiviral vectors: excellent tools forexperimental gene transfer and promising candidates forgene therapy. J. Gene Med. 2, 308-316.Vizzardelli C, Molano RD, Pileggi A, Berney T, Cattan P,Fenjves ES, Peel A, Fraker C, Ricordi C, Inverardi L (2002)Neonatal porcine pancreatic cell clusters as a potential sourcefor transplantation in humans: characterization ofproliferation, apoptosis, xenoantigen expression and genedelivery with recombinant AAV. Xenotransplantation 9,14-24 .Waldmann H (2002) Reprogramming the immune system.Immunol Rev 185, 227-235.Walsh CE (2003) Gene therapy Progress and Prospects: Genetherapy for the hemophilias. Gene Ther 10, 999-1003.Wang G, Slepushkin V, Zabner J, Keshavjee S, Johnston JC,Sauter SJ, Jolly DJ, Dubensky TWJ, Davidson BL, McCrayPBJ (1999) Feline immunodeficiency virus vectorspersistently transduce nondividing airway epithelia andcorrect the cystic fibrosis defect. J Clin Invest 104, R55-62.Watanabe H, Sumi S, Urushihata T, Kitamura Y, Iwasaki S, XuG, Yano S, Nio Y, Tamura K (2000) Immunohistochemicalstudies on vascular endothelial growth factor and plateletendothelial cell adhesion molecule-1/CD-31 in islettransplantation. Pancreas 21, 165-173.Weidner KM, Hartmann G, Sachs M, Birchmeier W (1993)Properties and functions of scatter factor/hepatocyte growthfactor and its receptor c-Met. Am J Respir Cell Mol Biol 8,229-237, .Welsh M, Andersson A (1994) Transplantation of transfectedpancreatic islets. Stimulation of beta cell DNA synthesis bythe src oncogene. Transplantation 57, 297-299.164

Gene Therapy and Molecular Biology Vol 7, page 165Welsh N, Oberg C, Hellerstrom C, Welsh M (1990) Liposomemediated in vitro transfection of pancreatic islet cells.Biomed Biochim Acta 12, 1157-1164.Wu P, Xiao W, Conlon T, Hughes J, Agbandje-McKenna M,Ferkol T, Flotte T, Muzyczka N (2000) Mutational analysisof the adeno-associated virus type 2 (AAV2) capsid gene andconstruction of AAV2 vectors with altered tropism. J Virol74, 8635-8647.Yancopoulos GD, Davis S, Gale NW, Rudge JS, Wiegand SJ,Holash J (2000) Vascular-specific growth factors and bloodvessel formation. Nature 407, 242-248.Zabner J, Fasbender AJ, Moninger T, Poellinger KA, Welsh MJ(1995) Cellular and molecular barriers to gene transfer by acationic lipid. J Biol Chem, 18997-19007.Zhang N, Richter A, Suriawinata J, Altomonte J, Meseck M,Dong H (2003) Elevated VEGF production in islet cellsenhances islet graft vascularization and improves functionalislet mass post transplantation. Diabetes 52 (suppl. 1), A14.Zhang N, Schroppel B, Chen D, Fu S, Hudkins KL, Zhang H,Murphy BM, Sung RS, Bromberg JS (2003) Adenovirustransduction induces expression of multiple chemokines andchemokine receptors in murine β cells and pancreatic islets.Am. J. Transplantation. In press..Zhang N, Suriawinata J, Meseck M, Woo SLC, Dong H (2002)Effective ex vivo transduction of murine islets by felineimmunodeficiency virus vectors. Adv. Islet Cell Biol: Fromstem cell differentiation to clinical transplantation. Anaheim,CA., p67.Zhou X, Huang L (1994) DNA transfection mediated by cationicliposomes containing lipopolylysine: characterization andmechanism of action. Biochem Biophys Acta 1189, 195-203..Dr. Hengjiang Dong165

Zhang et al: Angiogenic Gene Therapy for Improving Islet Graft Vascularization166

Gene Therapy and Molecular Biology Vol 7, page 167Gene Ther Mol Biol Vol 7, 167-172, 2003G-CSF Receptor-mediated STAT3 activation andgranulocyte differentiation in 32D cellsResearch ArticleRuifang Xu 1 , Akihiro Kume 1 , Yutaka Hanazono 1 , Kant M. Matsuda 1 , YasujiUeda 2 , Mamoru Hasegawa 2 , Fumimaro Takaku 1,3 and Keiya Ozawa 1,31Division of Genetic Therapeutics, Center for Molecular Medicine, Jichi Medical School, 3311-1 Yakushiji,Minamikawachi, Tochigi 329-0498, Japan, 2 DNAVEC Research Inc., 1-25-11 Kannondai, Tsukuba, Ibaraki 305-0856,Japan, 3 Division of Hematology, Department of Medicine, Jichi Medical School, 3311-1 Yakushiji, Minamikawachi,Tochigi 329-0498, Japan__________________________________________________________________________________*Correspondence: Akihiro Kume, M.D., Ph.D.; Division of Genetic Therapeutics, Center for Molecular Medicine, Jichi MedicalSchool, 3311-1 Yakushiji, Minamikawachi, Tochigi 329-0498, Japan; Phone: +81-285-58-7402; Fax: +81-285-44-8675; E-mail:kume@jichi.ac.jpKey words: STAT3, G-CSF receptor, granulocyte differentiation, estrogen binding domain, selective amplifier geneReceived: 3 July 2003; Accepted: 20 August 2003; electronically published: August 2003SummaryGranulocyte colony-stimulating factor (G-CSF) receptor (GcR) mediates growth and differentiation signals in thegranulocyte/monocyte lineage of hematopoietic cells. To investigate the differentiation signal via GcR, a conditionalreceptor activation system was constructed. Wild-type and mutant GcRs were controlled by fusion to a molecularswitch derived from the hormone binding domain of the estrogen receptor (ER). GcR-associated signalingmolecules were analyzed in 32D progenitor cells that possess a potential of granulocyte differentiation. While thewild-type GcR-ER fusion molecule induced a granulocyte differentiation in 32D cells, a substitution ofphenylalanine for tyrosine 703 (Y703F) in GcR resulted in a differentiation block. The activation of the JAK1 andJAK2 kinases was indistinguishable between the cells expressing the wild-type fusion and the Y703F mutant, andphosphorylation of the STAT5 transcription factor was comparable, too. On the other hand, tyrosinephosphorylation of STAT3 was significantly decreased following activation of the Y703F mutant compared to thewild-type GcR fusion. The results suggested that tyrosine 703 was responsible, at least in part, for transmitting adifferentiation signal via STAT3 in 32D. The fusion system with the estrogen binding domain provides a valuabletool to analyze mutant effector proteins in the natural cellular milieu while bypassing the endogenous counterparts.I. IntroductionRecent advances in stem cell biology, together withgene transfer technology, have led to the prospect of a newgeneration of cell therapy. However, many obstacles mustbe overcome before this vision becomes a reality. Onemajor hurdle is to control transplanted cells in therecipient’s body, in particular, to expand the desired cellsubsets so that they exhibit therapeutic benefit. We havedeveloped a novel system for selective expansion ofgenetically modified cells to supplement current genetransfer vectors (Ito et al, 1997; Kume et al, 2002). In thissystem, the target cells are harnessed with a ‘selectiveamplifier gene (SAG)’ which encodes a fusion proteincomprising the granulocyte colony-stimulating factor (G-CSF) receptor (GcR) and the hormone binding domain(HBD) of the estrogen receptor (ER). The ER-HBD worksas a molecular switch so that the fusion protein generates aGcR-derived growth signal upon binding to estrogen(Mattioni et al, 1994). Besides the prototype SAGencoding a chimera of the full-length GcR and ER-HBD(GcRER), a series of derivative fusion receptors wereconstructed to attain altered ligand specificity and signalcharacteristics. The modifications include a deletion of theG-CSF binding site (ΔGcR) (Ito et al, 1997), replacementof the ER with a mutant specific for 4-hydroxytamoxifen(TmR) (Xu et al, 1999), and the substitution ofphenylalanine for the most proximal tyrosine residue inthe GcR cytoplasmic domain (Y703FGcR) (Matsuda et al,1999a).The Y703F mutant is of particular interest becausethis amino acid substitution apparently led to adifferentiation block in myeloid progenitor 32D cells(Matsuda et al, 1999a). To explore the mechanisms ofgranulocyte differentiation in 32D cells, we examined167

Xu et al: G-CSF receptor-mediated STAT3 activationJAK-STAT pathways involved in GcR signaling, andidentified reduced STAT3 phosphorylation associated withthe Y703F mutation.II. Materials and methodsA. Plasmids and cellsBicistronic vector plasmids were constructed with the pMXretrovirus backbone and the encephalomyocarditis virus(EMCV)-derived internal ribosome entry site (IRES; nucleotides259-833 of EMCV-R genome) (Duke et al, 1992; Onishi et al,1996). pMX/ΔGcRER-IRES-CD8a encodes a fusion protein ofΔGcR and ER-HBD, and murine CD8a as a selectable marker(Fukunaga et al, 1991; Koike et al, 1987; Nakauchi et al, 1985).The Y703F mutation in the GcR part was introduced into thisplasmid as previously described (pMX/ΔY703FGcRER-IRES-CD8a) (Matsuda et al, 1999a). The recombinant DNAexperiments were carried out following the National Institutes ofHealth guidelines and approved by the Jichi Medical SchoolRecombinant DNA Research Advisory Board.The murine myeloid progenitor line 32D and its derivativeswere maintained in RPMI-1640 medium (Invitrogen, GrandIsland, NY) supplemented with 10% fetal bovine serum(Bioserum, Victoria, Australia) and 0.5% conditioned medium ofC3H10T1/2 cells transfected with a murine IL-3 expressionplasmid pBMG-hph-IL-3 (Valtieri et al, 1987; Matsuda et al,1999a; Xu et al, 1999).B. Immunoprecipitation and western blotting32D cells were deprived of serum and IL-3 for 3 hours at adensity of 5 x 10 5 cells/ml, and incubated in RPMI mediumcontaining 1 mM Na 3 OV 4 for an additional 1 hour at 1 x 10 7cells/ml. After starvation, cells were stimulated with either 10 -7M E 2 (Sigma, St. Louis, MO) or 10 -9 M recombinant human G-CSF (provided by Chugai Pharmaceuticals, Tokyo, Japan) forgiven periods, then washed with ice-cold phosphate-bufferedsaline (PBS) containing 100 µM Na 3 OV 4 . Subsequently, cellswere solubilized in lysis buffer (1% NP-40, 20 mM Tris-HCl [pH7.4], 137 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 50µg/ml aprotinin and 2 mM Na 3 OV 4 ) on ice for 30 minutes, andcentrifuged for 10 minutes. The soluble proteins were measuredby Protein Assay (Bio-Rad, Hercules, CA).For immunoprecipitation, the cell lysate containing 1 mg ofprotein was incubated with one of the following antibodies for 8hours at 4°C: anti-JAK1 (Upstate Biotechnology, Lake Placid,NY), anti-JAK2 (Upstate Biotechnology), anti-STAT3 (C-20;Santa Cruz Biotechnology, Santa Cruz, CA) and anti-STAT5 (C-17; Santa Cruz Biotechnology). The immune complexes wereabsorbed by protein G-Sepharose beads (Sigma) for 2 hours at4°C. The beads were washed with the lysis buffer and boiled insample buffer (60 mM Tris-HCl [pH 6.8], 2% sodium dodecylsulfate [SDS], 10% glycerol and 5% 2-mercaptoethanol) for 3minutes. After centrifugation, the supernatants were subjected toSDS-7.5% polyacrylamide gel electrophoresis and blotted ontopolyvinylidene fluoride membranes (Immobilon-P; Millipore,Yonezawa, Japan). After blocking treatment with 5% bovineserum albumin (Fraction V; Roche Diagnostics, Mannheim,Germany), the membranes were incubated with an antiphosphotyrosineantibody (4G10; Upstate Biotechnology) for 1hour at room temperature. Immunoreactive proteins werevisualized by enhanced chemiluminescence (ECL; AmershamPharmacia Biotech, Little Chalfont, UK). In some instances,membranes were stripped by incubation in denaturing buffer(62.5 mM Tris-HCl [pH 6.7], 2% SDS and 100 mM 2-mercaptoethanol) for 30 minutes at 50°C and reprobed withanother antibody.III. ResultsA. Construction of conditionallyactivated G-CSF receptorsStructures of the chimeric receptors used in thisstudy are shown in Figure 1. The fusion protein system isbased on the fact that ER-HBD functions as an estrogenspecificmolecular switch to control heterologous effectorproteins, in our case, GcR (Mattioni et al, 1994). GcRbelongs to the type I cytokine receptor superfamily, and itscytoplasmic domain comprises functionally distinctsubdomains: the membrane-proximal region is sufficientfor mitogenic signaling, and the membrane-distal portionis essential for granulocyte maturation (Dong et al, 1993;Fukunaga et al, 1993; Avalos, 1996; Koay and Sartorelli,1999). All of the four conserved tyrosine residues in thecytoplasmic domain of GcR (at positions 703, 728, 743and 763 in the murine GcR) are in the membrane-distalregion and phosphorylated upon G-CSF stimulation.Among these, the tyrosine at position 703 (Y703) wasmost prominently phosphorylated and involved ingranulocyte differentiation (Yoshikawa et al, 1995).However, previous studies on functional domains of GcRwere carried out with ectopically expressed wild-type andmutant molecules in receptor-negative cells. It may bemore informative if mutant receptors are analyzed in thenatural intracellular environment where the endogenousmolecule functions. From this viewpoint, the ER-HBDfusion system provides a valuable experimental tool.Estrogen specifically activates the introduced GcRER (andits derivatives) without influencing the endogenous GcR inthe same cell, and the downstream events can be studiedindependently.Figure 1. Structures of the chimeric receptors involved in thisstudy. GcRER is a fusion of the full-length murine granulocytecolony-stimulating factor (G-CSF) receptor (GcR) and thehormone binding domain (HBD) of rat estrogen receptor (ER).ΔGcRER is a derivative of GcRER deleted of the G-CSF bindingsite (amino acids 5-195). ΔY703FGcRER carries a substitutionof phenylalanine for a cytoplasmic tyrosine at position 703(Y703F) in GcR. Ext, extracellular domain; G, G-CSF bindingsite; TM, transmembrane domain; Cyt, cytoplasmic domain; TA,transactivation domain; DNA, DNA binding domain; YYYY,conserved tyrosine residues in GcR cytoplasmic domain; FYYY,Y703F mutation in GcR.168

Gene Therapy and Molecular Biology Vol 7, page 169In our previous report, the biological response to theΔGcRER- and ΔY703FGcRER-mediated signal wasevaluated in murine myeloid progenitor 32D cells (Δdesignates a deletion of amino acids 5-195 required for G-CSF binding; Matsuda et al, 1999a). Parental 32D cells aredependent on interleukin-3 (IL-3) for continuous growth,and switching from IL-3 to G-CSF makes the cellsdifferentiate into morphologically mature neutrophils(Valtieri et al, 1987). By retrovirus-mediated genetransfer, stable clones expressing ΔGcRER(32D/ΔGcRER) or ΔY703FGcRER (32D/ΔY703FGcRER)were established and stimulated by estrogen. Whileestrogen-treated 32D/ΔGcRER cells underwentgranulocyte differentiation indistinguishable from thatseen in G-CSF-treated cells, 32D/ΔY703FGcRER cellsshowed a distinct phenotype. Estrogen supported a longtermproliferation of 32D/ΔY703FGcRER withmyeloblastic appearance, indicating that the Y703Fmutation abrogated the differentiation signal (Matsuda etal, 1999a). This observation prompted us to characterizesignaling molecules downstream of GcR in more detail.Following ligand-induced homodimerization, GcRinduces a wide array of intracellular signaling events(Avalos, 1996). Like many other cytokine receptors, GcRhas no intrinsic kinase activity; instead, it recruits andactivates other cytoplasmic kinases such as Janus kinases(JAKs), signal transducer and activation of transcription(STAT) proteins, Src family kinases and components ofthe mitogen-activated protein kinase pathway. Theactivation of JAKs is one of the earliest events in the GcRsignaling cascade, followed by the tyrosinephosphorylation of STATs and GcR itself (Nicholson et al,1994; Dong et al, 1995). Since the signal transduction forgranulocyte differentiation has been ascribed to the JAK-STAT pathway, we focused on these molecules inΔGcRER and ΔY703FGcRER cells.B. Estrogen-induced phosphorylation ofJAK1 and JAK2 via fusion receptorsFirst, we examined the tyrosine phosphorylation ofJAK1 and JAK2. As shown in Figure 2, these kinaseswere not tyrosine-phosphorylated in resting 32D/ΔGcRERand 32D/ΔY703FGcRER cells. Addition of G-CSF rapidlyinduced phosphorylation of JAK1 and JAK2; this eventwas induced by dimerization of the endogenous GcR, andmaximal activation was observed within 10 minutes (datanot shown). Similarly, 10 -7 M 17β-estradiol (E 2 ) inducedtyrosine phosphorylation of JAK1 and JAK2 in these cells(Figure 2). The estrogen-induced activation of JAK1 andJAK2 was mediated by chimeric receptors, at a slower ratethan the activation mediated by the endogenous GcR; themaximal phosphorylation was observed 60 minutes afterE 2 addition (time course not shown). The difference inkinetics of JAK1/JAK2 phosphorylation may be due todifferent mechanisms of receptor activation. While G-CSFdirectly crosslinks GcR at the extracellular domain, theactivation of ER-HBD fusion receptors is a ligand-inducedderepression that involves other proteins such as HSP90(Mattioni et al, 1994). Nevertheless, the levels ofJAK1/JAK2 phosphorylation were comparable whetherthe cells were stimulated with G-CSF or estrogen. Asshown in Figure 2, the levels of estrogen-inducedJAK1/JAK2 phosphorylation in 32D/ΔY703FGcRER cellswere comparable to those seen in 32D/ΔGcRER cells.Reprobing of the blots with anti-JAK1 and anti-JAK2antibodies showed that approximately equal amounts ofthe kinases were loaded on these lanes (not shown). Thus,we concluded that the Y703F mutation had little, if any,effect on the tyrosine phosphorylation of JAK1 and JAK2.Considering that JAK1 and JAK2 are constitutivelyassociated with the membrane-proximal region of GcRwhich is sufficient to activate them (Nicholson et al, 1994;Dong et al, 1995; Avalos, 1996), it is conceivable that thekinases were not affected by the GcR mutation in themembrane-distal region.C. Comparable STAT5 phosphorylationfollowing fusion receptor activationNext, we investigated the activation of STATproteins in 32D/ΔGcRER and 32D/ΔY703FGcRER cells.It was shown that G-CSF-induced signaling involvesSTAT1, STAT3 and STAT5 (Tian et al, 1994; de Koninget al, 1996; Tian et al, 1996; Shimozaki et al, 1997; Donget al, 1998; Chakraborty et al, 1999; Ward et al, 1999).Since the membrane-distal cytoplasmic region of GcR wasnot required for STAT1 activation (de Koning et al.,1996), we addressed whether the phosphorylation ofSTAT5 and STAT3 is affected by the Y703F mutation.Figure 3 shows the time course of STAT5 activation in32D/ΔGcRER and 32D/ΔY703FGcRER cells (upperpanel). STAT5 was not tyrosine-phosphorylated inunstimulated 32D cells, and addition of 10 -9 M G-CSFinduced a rapid phosphorylation of this molecule throughcrosslinking of the endogenous GcR. On the other hand,10 -7 M of E 2 induced a slower and less extensivephosphorylation of STAT5.Figure 2. Tyrosine phosphorylation of JAK1 and JAK2. Serumandcytokine-starved 32D/ΔGcRER and 32D/ΔY703FGcRERcells were harvested before (0’) and after 60 minutes (60’) ofincubation with 10 -7 M of estradiol (E 2 ). Lysates from32D/ΔGcRER and 32D/ΔY703FGcRER cells wereimmunoprecipitated (IP) with either an anti-JAK1 (αJAK1;upper panel) or an anti-JAK2 (αJAK2; lower panel) antibody.Immunoblotting (IB) was carried out with an antiphosphotyrosineantibody (αPY).169

Xu et al: G-CSF receptor-mediated STAT3 activationThe estrogen-induced STAT5 activation was comparablein 32D/ΔGcRER and 32D/ΔY703FGcRER cells at 60minutes after stimulation, and reprobing of the blot withan anti-STAT5 antibody showed that approximately equalamounts of STAT5 were loaded (Figure 3, lower panel).The delay in STAT5 phosphorylation may be associatedwith a slower JAK1/JAK2 activation through estrogeninduceddimerization of the chimeric receptors. The reasonfor the reduced STAT5 phosphorylation in the E 2 -stimulated cells is currently unknown; we speculate thatthe linking of ER-HBD to the C-terminal of GcR mighthinder STAT proteins from freely accessing themembrane-distal region of the receptor. In any case,STAT5 appeared to be phosphorylated to the same extentin 32D/ΔGcRER and 32D/ΔY703FGcRER cells. Othersdemonstrated that STAT5 was activated even when themembrane-distal region of GcR was deleted or thereceptor tyrosine phosphorylation was abrogated(Shimozaki et al, 1997; Tian et al, 1996). Taken togetherwith our observation that JAK1 and JAK2 were activatedin both 32D/ΔGcRER and 32D/ΔY703FGcRER cells(Figure 2), we concluded that the Y703F mutation did notaffect the tyrosine phosphorylation of STAT5.D. Decrease in STAT3 Activation byY703F G-CSF Receptor MutantFinally, we addressed whether the Y703F mutationin GcR affects tyrosine phosphorylation of STAT3. Aftercytokine starvation, 32D/ΔGcRER and32D/ΔY703FGcRER clones were incubated with 10 -7 M ofE 2 for 60 minutes. While estrogen induced a significanttyrosine phosphorylation of STAT3 in 32D/ΔGcRER, onlya slight activation of STAT3 was detected in32D/ΔY703FGcRER clones (Figure 4, upper panel,arrow). Reprobing of the membrane with an anti-STAT3antibody revealed an even loading of STAT3 in theselanes (Figure 4, lower panel).Repeated experiments constantly demonstrated adecreased STAT3 phosphorylation in32D/ΔY703FGcRER. Consistent with this observation,Tian et al showed that the G-CSF-induced STAT3activation was greatly abrogated in UT-7epo celltransfectants by deleting a membrane-distal part includingY703 from GcR (Tian et al, 1996). We thereforeconcluded that Y703 in GcR was involved in STAT3activation, and that the event is crucial to granulocytedifferentiation in 32D cells.IV. DiscussionThe phosphotyrosine residues in GcR create potentialdocking sites for the recruitment of signaling moleculessuch as STATs that contain a Src homology 2 (SH2)domain. STAT3 is recruited via the interaction of its SH2domain with receptor tyrosine residues that are present in atyrosine-X-X-glutamine (YXXQ) sequence (Stahl et al,1995). Among four conserved tyrosine residues in thecytoplasmic region of GcR, only Y703 provides a YXXQmotif, accounting for the reduced STAT3 activation by theY703F mutant. However, there was a residual level ofSTAT3 activation in ΔY703FGcRER and other GcRmutants devoid of this motif, which suggested thepresence of another STAT3 binding site in GcR or somebridging molecule (Avalos, 1996; Chakraborty et al,1999). We observed a few additional phosphorylatedproteins coimmunoprecipitated with STAT3 including a130 kDa species (Figure 4, upper panel, arrowheads).These proteins are yet to be identified; at least they did notreact with an antibody against GcR in a subsequentreprobing (data not shown).Figure 3. Tyrosine phosphorylation of STAT5. Starved32D/ΔGcRER and 32D/ΔY703FGcRER cells were harvestedbefore (0’) and after 10, 30, and 60 minutes (10’, 30’, 60’) ofincubation with 10 -9 M of G-CSF or 10 -7 M of estradiol (E 2 ).Lysates were immunoprecipitated (IP) with an anti-STAT5antibody (αSTAT5) and immunoblotted (IB) with an antiphosphotyrosineantibody (αPY; upper panel). The blot wasreprobed with the anti-STAT5 antibody to confirm the equalloading of STAT5 (lower panel).Figure 4. Tyrosine phosphorylation of STAT3. Starved32D/ΔGcRER and 32D/ΔY703FGcRER (clone 1 and clone 2)cells were harvested before (0’) and after 60 minutes (60’) ofincubation with 10 -7 M of estradiol (E 2 ). Lysates wereimmunoprecipitated (IP) with an anti-STAT3 antibody(αSTAT3) and immunoblotted (IB) with an anti-phosphotyrosineantibody (αPY; upper panel). The blot was reprobed with theanti-STAT3 antibody to confirm the equal loading of STAT3(lower panel). Besides STAT3 (92 kDa, arrow), severalphosphoproteins including a 130 kDa species (arrowheads) werecoimmunoprecipitated.170

Gene Therapy and Molecular Biology Vol 7, page 171A consensus has been reached that tyrosinephosphorylation of GcR and activation of STAT3 iscrucial to granulocyte differentiation, but there remainssome controversy over the relative contribution of eachtyrosine residue depending on the cells used (Tian et al,1994, 1996; de Koning et al, 1996; Shimozaki et al, 1997;Chakraborty et al, 1999; Ward et al, 1999). Previousreports employed either GcR-negative cells to examine thefunction of the receptor and associated molecules, oroverexpression of dominant-negative forms of GcR toelucidate the mechanisms for growth and differentiation.By using ER-HBD fusion proteins to bypass endogenousGcR, we herein provided additional data suggesting themajor involvement of Y703 in STAT3 activation. It is ofparticular note that the cells retained the expression ofwild-type GcR and downstream signaling molecules,thereby rapidly undergoing granulocyte differentiation inresponse to G-CSF, indistinguishable from the parent 32Dcells (Matsuda et al, 1999a).Contrary to its promoting function in myeloid celldifferentiation, STAT3 was shown to play a central role inthe maintenance of the pluripotent phenotype ofembryonic stem cells (Matsuda et al, 1999b; Niwa et al,1998). STAT3 appears to dictate widely divergentinstructions such as differentiation and proliferationdepending on the cell type. Thus, it is crucial to set up anappropriate venue to study the physiological molecularinteraction involving a promiscuous molecule such asSTAT3. The HBD fusion system provides a powerful toolto examine the behavior of mutated proteins controlled byspecific ligands, in the exact milieu where the wild-typemolecules coexist but remain unstimulated.AcknowledgmentsWe are grateful to Chugai Pharmaceuticals forproviding G-CSF. This work was supported by grantsfrom the Ministry of Education, Culture, Sports, Scienceand Technology, and the Ministry of Health, Labor andWelfare, JapanReferencesAvalos BR (1996) Molecular analysis of the granulocyte colonystimulatingfactor receptor. Blood 88, 761-777.Chakraborty A, Dyer KF, Cascio M, Mietzner TA and TweardyDJ (1999) Identification of a novel Stat3 recruitment andactivation motif within the granulocyte colony-stimulatingfactor receptor. Blood 93, 15-24.de Koning JP, Dong F, Smith L, Schelen AM, Barge RMY, vander Plas DC, Hoefsloot LH, Löwenberg B and Touw IP(1996) The membrane-distal cytoplasmic region of humangranulocyte colony-stimulating factor receptor is required forSTAT3 but not STAT1 homodimer formation. Blood 87,1335-1342.Dong F, van Buitenen C, Pouwels K, Hoefsloot LH, LöwenbergB and Touw IP (1993) Distinct cytoplasmic regions of thehuman granulocyte colony-stimulating factor receptorinvolved in induction of proliferation and maturation. MolCell Biol 13, 7774-7781.Dong F, van Paassen M, van Buitenen C, Hoefsloot LH,Löwenberg B and Touw IP (1995) A point mutation in thegranulocyte colony-stimulating factor receptor (G-CSF-R)gene in a case of acute myeloid leukemia results in theoverexpression of a novel G-CSF-R isoform. Blood 85, 902-911.Dong F, Liu X, de Koning JP, Touw IP, Henninghausen L,Larner A and Grimley PM (1998) Stimulation of Stat5 bygranulocyte colony-stimulating factor (G-CSF) is modulatedby two distinct cytoplasmic regions of the G-CSF receptor. JImmunol 161, 6503-6509.Duke GM, Hoffman MA and Palmenberg AC (1992) Sequenceand structural elements that contribute to efficientencephalomyocarditis virus RNA translation. J Virol 66,1602-1609.Fukunaga R, Ishizaka-Ikeda E, Pan C-X, Seto Y and Nagata S(1991) Functional domains of the granulocyte colonystimulatingfactor receptor. EMBO J 10, 2855-2865.Fukunaga R, Ishizaka-Ikeda E and Nagata S (1993) Growth anddifferentiation signals mediated by different regions in thecytoplasmic domain of granulocyte colony-stimulating factorreceptor. Cell 74, 1079-1087.Ito K, Ueda Y, Kokubun M, Urabe M, Inaba T, Mano H,Hamada H, Kitamura T, Mizoguchi H, Sakata T, HasegawaM and Ozawa K (1997) Development of a novel selectiveamplifier gene for controllable expansion of transducedhematopoietic cells. Blood 90, 3884-3892.Koay DC and Sartorelli AC (1999) Functional differentiationsignals mediated by distinct regions of the cytoplasmicdomain of the granulocyte colony-stimulating factorreceptor. Blood 93, 3774-3784.Koike S, Sakai M and Muramatsu M (1987) Molecular cloningand characterization of rat estrogen receptor cDNA. NucleicAcids Res 15, 2499-2513.Kume A, Hanazono Y, Mizukami H, Okada T and Ozawa K(2002) Selective expansion of transduced cells forhematopoietic stem cell gene therapy. Int J Hematol 76,299-304.Matsuda KM, Kume A, Ueda Y, Urabe M, Hasegawa M andOzawa K (1999a) Development of a modified selectiveamplifier gene for hematopoietic stem cell gene therapy.Gene Ther 6, 1038-1044.Matsuda T, Nakamura T, Nakao K, Arai T, Katsuki M, Heike Tand Yokota T (1999b) STAT3 activation is sufficient tomaintain an undifferentiated state of mouse embryonic stemcells. EMBO J 18, 4261-4269.Mattioni T, Louvion J-F and Picard D (1994) Regulation ofprotein activities by fusion to steroid binding domains.Methods Cell Biol 43, 335-352.Nakauchi H, Nolan GP, Hsu C, Huang HS, Kavathas P andHerzenberg LA (1985) Molecular cloning of Lyt-2, amembrane glycoprotein marking a subset of mouse Tlymphocytes: molecular homology to its human counterpart,Leu-2/T8, and to immunoglobulin variable regions. ProcNatl Acad Sci USA 82, 5126-5130.Nicholson SE, Oates AC, Harpur AG, Ziemiecki A, Wilks AFand Layton JE (1994) Tyrosine kinase JAK1 is associatedwith the granulocyte-colony-stimulating factor receptor andboth become tyrosine-phosphorylated after receptoractivation. Proc Natl Acad Sci USA 91, 2985-2988.Niwa H, Burdon T, Chambers I and Smith A (1998) Self-renewalof pluripotent embryonic stem cells is mediated viaactivation of STAT3. Genes Dev 12, 2048-2060.Onishi M, Kinoshita S, Morikawa Y, Shibuya A, Phillips J,Lanier LL, Gorman DM, Nolan GP, Miyajima A andKitamura T (1996) Applications of retrovirus-mediatedexpression cloning. Exp Hematol 24, 324-329.Shimozaki K, Nakajima K, Hirano T and Nagata S (1997)Involvement of STAT3 in the granulocyte colony-stimulatingfactor-induced differentiation of myeloid cells. J Biol Chem272, 25184-25189.171

Xu et al: G-CSF receptor-mediated STAT3 activationStahl N, Farruggella TJ, Boulton TG, Zhong Z, Darnell JEJr andYancopoulos GD (1995) Choice of STATs and othersubstrates specified by modular tyrosine-based motifs incytokine receptors. Science 267, 1349-1353.Tian S-S, Lamb P, Seidel HM, Stein RB and Rosen J (1994)Rapid activation of the STAT3 transcription factor bygranulocyte colony-stimulating factor. Blood 84, 1760-1764.Tian S-S, Tapley P, Sincich C, Stein RB, Rosen J and Lamb P(1996) Multiple signaling pathways induced by granulocytecolony-stimulating factor involving activation of JAKs,STAT5, and/or STAT3 are required for regulation of threedistinct classes of immediate early genes. Blood 88, 4435-4444.Valtieri M, Tweardy DJ, Caracciolo D, Johnson K, Mavilio F,Altmann S, Santoli D and Rovera G (1987) Cytokinedependentgranulocytic differentiation: regulation ofproliferative and differentiative responses in a murineprogenitor cell line. J Immunol 138, 3829-3835.Ward AC, Smith L, de Koning JP, van Aesch Y and Touw IP(1999) Multiple signals mediate proliferation, differentiation,and survival from the granulocyte colony-stimulating factorreceptor in myeloid 32D cells. J Biol Chem 274, 14956-14962.Xu R, Kume A, Matsuda KM, Ueda Y, Kodaira H, OgasawaraY, Urabe M, Kato I, Hasegawa M and Ozawa K (1999) Aselective amplifier gene for tamoxifen-inducible expansionof hematopoietic cells. J Gene Med 1, 236-244.Yoshikawa A, Murakami H and Nagata S (1995) Distinct signaltransduction through the tyrosine-containing domains of thegranulocyte colony-stimulating factor receptor. EMBO J 14,5288-5296.Dr. Akihiro Kume172

Gene Therapy and Molecular Biology Vol 7, page 173Gene Ther Mol Biol Vol 7, 173-179, 2003.Calcium induces apoptosis and necrosis inhematopoetic malignant cells: Evidence for caspase-8 dependent and FADD-autonomous pathwayResearch ArticleChristof J. Burek † , Malgorzata Burek † , Johannes Roth # , and Marek Los †¨†Institute of Experimental Dermatology, University of Münster, D-48149 Münster; # Institute of Molecular Medicine,University of Düsseldorf, D-40225 Düsseldorf, Germany; ¨ Manitoba Institute of Cell Biology, CancerCare Manitoba,Winnipeg, Canada.__________________________________________________________________________________*Correspondence: Marek Los, MD/PhD, Institute of Experimental Dermatology, University of Münster, Röntgenstrasse 21, D-48149Münster, Germany; Phone: 49-251-83-52943; Fax: 49-251-83-56549; e-mail: los@uni-muenster.deKey Words: A23187, apoptosis, Bcl-2, caspase-8, FADD, necrosisAbbreviations: propidium iodide (PI), Fas-associated death domain protein (FADD), endoplasmic reticulum (ER), mitochondrialpermeability transition (MPT), apoptosis-inducing factor (AIF)Received: 1 September 2003; Accepted: 18 September 2003; electronically published: September 2003SummaryOne of the killing mechanisms employed by Natural Killer (NK) cells and Lymphokine-Activated Killer (LAK) cellsis the perforation of the cellular membrane that causes the increase of cytoplasmic calcium concentration anddisturbs further the homeostasis of other ions. Cytoplasmic calcium influx, exceeding the tolerated physiologicthreshold in cell signaling events, can induce either apoptosis or necrosis depending on its final concentration.Despite several years of intensive research and identification of some molecular targets of action like e.g. calpains,calcineurin or calreticulin, the exact mechanism of calcium-induced cell death is not known in detail. We show herethat death pathways triggered by calcium rely on a novel, caspase-8-dependent and Bcl-2-inhibitable pathway thatis FADD-adaptor molecule -independent. This is shown in a leukemic cell model. The experimental system employseither cells that lack the expression of casapase-8 or cells genetically modified to overexpress, Bcl-2, or a FADDdominantnegative mutant (FADD-DN).I. IntroductionCalcium is one of the most versatile and powerfulsmall molecules applied by a cell to regulate its biologicfunctions. It can either protect from or induce cell death,depending on concentration and cell type (Franklin andJohnson, 1992; Barros et al, 2002). Although themechanism of calcium triggered death has beeninvestigated for years, the exact mechanism(s) responsiblefor this process are not known in detail. Dying cells entereither apoptosis, necrosis or an intermediate form of celldeath, depending on the death stimulus, its intensity andthe level of intracellular ATP (Leist and Jaattela, 2001;Los et al, 2002). In accordance, calcium can induce bothforms of cell death as well as an intermediate process,depending on available intracellular concentration and celltype (Gwag et al, 1999; Barros et al, 2002). Calciumrelatedcell death is best described in neurones (Gwag etal, 1999; Xu et al, 2001), however, detailed studies inlymphatic tissue, from recent date are scarce. Calciumionophores, such as ionomycin or A-23187 are frequentlyapplied to manipulate intracellular Ca 2+ concentration andthus to mimic signaling events or to induce cell death(Errasfa and Stern, 1994; Nakamura, 1996). Severalauthors provide observations that various tumor cell linesexposed to A-23187 or ionomycin undergo either nonapoptoticdegeneration (Duke et al, 1994; Kressel andGroscurth, 1994), or classical apoptosis (Ojcius et al,1991; Ning and Murphy, 1993).Caspases (cysteine-dependent aspartases) are crucialapoptotic executioner proteases (Los et al, 1995; Herr andDebatin, 2001). They are members of the C14 proteasefamily according to the Barrett and Rawlings classification(Los et al, 1999; Barrett and Rawlings, 2001). All caspasesare characterized by a nearly absolute specificity forsubstrates containing aspartic acid in the P1 cleavageposition and a cysteine in the active center of the enzyme(Stennicke et al, 2002). There are currently 12 knowncaspases in humans. Caspases-1, -4 and -5 mainly play arole in the regulation of inflammatory response, byproteolytic activation of inflammatory cytokines (Cassenset al, 2003). Caspases-2, -3, -6, -7, -8, -9 and -10 are173

Burek et al: Calcium induced cell deathconsidered to be involved predominantly in apoptoticsignalling (Sadowski-Debbing et al, 2002). In addition tothe role in apoptosis and inflammation, an involvement ofcaspases in other processes, like cell cycle regulation,hematopoesis and signal transduction in the immunesystem have been proposed (Denis et al, 1998; Los et al,2001). All caspases are synthesized as inactive zymogensthat are activated through proteolytic cleavage. Among thecaspase activation pathways, the best described ones arethe death-receptor dependent signalling cascades, withFADD adaptor molecule and caspase-8 as the key players,and the mitochondria/apoptosome dependent pathway thatrelies on Apaf-1 and caspase-9 (Krammer, 2000; Walczakand Krammer, 2000; Zheng and Flavell, 2000; Renz et al,2001). Both pathways are interconnected, thusamplification loops may take place (Sadowski-Debbing etal, 2002). The mitochondrial pathway is largely controlledby Bcl-2 family members. Bcl-2 family proteins exert itspro-and antiapoptotic action partially by influencingcalcium homeostasis of mitochondria and endoplasmicreticulum (ER) (reviewed in Hajnoczky et al, 2003).The family comprises both antiapoptotic andproapoptotic proteins. All antiapoptotic family members(e.g. Bcl-2, Bcl-X L ) share three or four Bcl-2 homology(BH) regions, and they localize to the cytoplasmic side ofintracellular membranes (Bouillet and Strasser, 2002). Theproapoptotic Bcl-2 family members can be further dividedinto two subgroups. Members of the first subgroup, bestrepresented by Bax and Bak (reviewed in Bouillet andStrasser, 2002) have two or three BH regions and appearto be structurally similar to their prosurvival relatives(Suzuki et al, 2000). The second subgroup of proapoptoticBcl-2-related proteins, (e.g. Bad, Bid, Bim) share only theshort BH3 region (reviewed in Bouillet and Strasser,2002). The exact mechanism of apoptosis regulation byBcl-2 family members is not fully understood (Strasser etal, 2000). It is widely believed that Bcl-2 functions topreserve the mitochondrial membrane integrity,mitochondrial and ER calcium homeostasis and preventthe release of cytochrome c and other proapoptoticmolecules from the mitochondria. BH3-only proteinsappear to sense stimuli that cause cellular stress andinitiate the death cascade. Proapoptotic Bax and Bak areessential for cell killing governed by BH3-only proteins,and this form of cell death is antagonized byoverexpresion of Bcl-2 (reviewed in Hajnoczky et al,2003; Marsden and Strasser, 2003).To gain insight into the mechanisms that governcalcium triggered cell death we have used a T-cellleukemiabased model and calcium ionophores asmodulators of intracellular Ca 2+ level. We show here thatthe calcium activated apoptotic pathway rely on yet-to-bedefined,caspase-8-dependent and Bcl-2-inhibitablepathway. Interestingly, the pathway does not rely onFADD-adaptor molecule. Thus, we provide furtherevidences for an intrinsic (death receptor-independent)death pathway that relies on caspase-8.II. Materials and methodsA. Materials and cell cultureAll cell lines were grown in 5% CO 2 at 37°C using aRPMI-1640 medium supplemented with 10% heat-inactivatedfetal calf serum and antibiotics (GIBCO, Eggenstein, Germany).A23187 was purchased from Sigma (Deisenhofen, Germany).The caspase inhibitor zVADfmk (benzyloxycarbonyl-Val-Ala-Asp-fluoro-methylketone) was purchased from Enzyme SystemsProducts (Dublin, CA), and staurosporine from RocheBiochemicals (Mannheim, Germany). All other chemicals werefrom Merck KG (Darmstadt, Germany) or Roth (Karlsruhe,Germany). Stable transfectants of Jurkat cells overexpressingBcl-2 and Jurkat clone that was deficient in caspase-8 were akind gift of Dr. J. Blenis, (Harvard Medical School, Boston,Massachusetts, USA).B. Cell extracts and immunoblottingThe proteolytic processing of caspase-3 and caspase-8 wasdetected by immunoblotting. Briefly, 5 x 10 5 cells were seeded in6-well plates and treated with the apoptotic stimuli. After theindicated time, cells were washed in cold PBS and lysed in 1%Triton X-100, 50 mM Tris-HCl, pH 7.6 and 150 mM NaClcontaining 3 µg/ml aprotinin, 3 µg/ml leupeptin, 3 µg/mlpepstatin A and 2 mM phenylmethylsulfonyl fluoride (PMSF).Subsequently, the proteins were separated under reducingconditions by 12 % sodium dodecyl sulfate-polyacrylamide gelelectrophoresis and electroblotted to a polyvinylidene difluoridemembrane (Amersham, Braunschweig, Germany). The equalloading of protein was controlled by measuring the proteinconcentration using the Bradford assay (BioRad, Munich,Germany). Membranes were blocked for 1 h with 5% non-fat drymilk powder in TBS and then incubated for 1 h with murinemonoclonal antibodies directed against caspase-3 (TransductionLaboratory, Heidelberg, Germany). Membranes were washedfour times with TBS/0.02% Triton X-100 and incubated with therespective peroxidase-conjugated affinity-purified secondaryantibody for 1 h. Following extensive washing, the reaction wasdeveloped by enhanced chemiluminescent staining using ECLreagents (Amersham).C. Fluorimetric assay of caspase activity -DEVD-ase assayCytosolic cell extracts were prepared by lysing cells in abuffer containing 0.5% NP-40, 20 mM HEPES pH 7.4, 84 mMKCl, 10 mM MgCl 2 , 0.2 mM EDTA, 0.2 mM EGTA, 1 mMDTT, 5 µg/ml aprotinin, 1 µg/ml leupeptin, 1 µg/ml pepstatinand 1 mM PMSF. Caspase activity was determined by theincubation of cell lysates with 50 µM of the fluorogenic substrateDEVD-AMC (N-acetyl-Asp-Glu-Val-Aspaminomethylcoumarin,Bachem, Heidelberg, Germany) in 200 µl buffer containing 50mM HEPES pH 7.3, 100 mM NaCl, 10% sucrose, 0.1% CHAPSand 10 mM DTT. The release of aminomethylcoumarin wasmeasured by fluorometry using an excitation wavelength of 360nm and an emission wavelength of 475 nm.D. Measurement of cell death and apoptosisCell death was measured either by the detection ofhypodiploid nuclei (Nicoletti method) (Renz et al, 2001) or bythe uptake of propidium iodide (PI) (Stroh et al, 2002). Briefly,for the measurement of hypodiploid DNA, nuclei were preparedby lysing 10 4 cells in 100 µl of hypotonic lysis buffer (1%sodium citrate, 0.1% Triton X-100, and 50 µg/ml PI). The nuclei174

Gene Therapy and Molecular Biology Vol 7, page 175were subsequently analyzed by flow cytometry, using aFACScalibur (Becton Dickinson, Heidelberg, Germany) andCellQuest analysis software. To assess PI uptake, cells wereharvested after the indicated times and incubated with PI (2µg/ml). The uptake of PI into nonfixed cells was measured byflow cytometry, using the FSC/FL2 profile.III. ResultsA. Calcium influx induces apoptotic andnecrotic cell death in a dose dependentmannerIn order to get insight into the mechanism(s) ofcalcium induced cell death we have performed time-, andconcentration- kinetic studies. Jurkat human T-leukemiacells were treated with increasing concentrations of theA23187 calcium ionophore. A23187 induces cell death ina dose dependent manner (Figure 1). Higherconcentrations of intracellular calcium induce cell deathwith faster kinetics. At the concentration of 800 ng/mlA23187 induces a maximum cell death at 18 h, whereaslower concentrations of the ionophore show slowerkinetics. The assessment of data obtained by themeasurement of PI uptake and apoptosis-specificmeasurement by the detection of hypodiploid nuclei(“Nicoletti” method) indicates that higher concentrationsinduce not only apoptotic, but also necrosis in theexperimental system (Figure 1C). Since contrary tonecrosis the apoptotic cell death relies on caspases, werepeated the series of experiments employing the broadspectrumcaspase inhibitor zVADfmk (Figure 2). ThuszVAD-fmk inhibitable cell death represents the apoptoticfraction. The zVADfmk based approach largely confirmsthe data obtained by the combination of the PI-uptakebased- and the “Nicoletti” method (Figure 1C). Unlike theNicoletti method that detects (lack of) the intactness ofnuclear DNA (hypodiploidy), PI-uptake stains cells withpermeable cell membranes (necrotic and late apoptoticcells). zVADfmk inhibits the proteolytic caspase activityand, therefore, it blocks the apoptotic fraction of celldeath. The experiments involving the caspase inhibitorindicate the highest zVADfmk-independent (presumablynecrotic) fraction of cell death upon the treatment withintermediate (200 ng/ml) concentrations of A23187calcium ionophore (Figure 2C). These method-relateddifferential results are explained in detail in thediscussion-part of the paper.B. Caspase-8 deficiency impairs calciuminduced cell deathThe broad-spectrum caspase inhibitor zVADfmk waslargely protective against calcium induced cell death. Toexamine further the role of caspases in the deathmechanism triggered by calcium we have employed aJurkat cell clone that lacks caspase-8 activity. Calciuminduced cell death measured by PI uptake wassignificantly impaired in cells lacking caspase-8 activity(Figure 3A). The observed effect could be detected atseveral time points and it was most pronounced after 18 h.Figure 1. Induction of cell death by calcium ionophore inJurkat cells. (A and B) show parallel-, time-kinetic experimentsevaluated either by PI-uptake, a method that unspecificallydetects cell death (A), or by apoptosis-specific “Nicoletti”method that measures hypodiploid, apoptotic nuclei (B). Thestandard deviation of four independent experiments shown heredid not exceeded 11 %. The percentage representation of bothdeath modes, that occurred after 18 h are visualized in the panel(C).175

Burek et al: Calcium induced cell deathFigure 2. Delineation of caspase-dependent (zVADfmkinhibitable)and caspase-independent components of calciuminducedcell death. Jurkat cells were treated with differentconcentrations of A23187 as indicated in A and B. Theapplication of zVADfmk, the broad-spectrum caspase inhibitorhas significantly, but only partially blocked cell death events (B).The panel (B) shows data from four independent experiments.The standard deviation did not exceeded 9 %. The “zVADfmk”resistant cell death component is depicted in the panel (C). Celldeath was measured by PI-uptake.Figure 3. Caspase-8 activity deficiency protects from necroticcomponent, but not from the apoptotic constituent ofcalcium-induced cell death. Jurkat cells were induced to die bythe addition of 200 ng/ml of A23187. Cell death was measured inparallel by PI-uptake (A), and by the assessment of nuclearhypodiploidy that corresponds to apoptotic cell death (B). To getthe confirmation of the data, we conducted a kinetic study usingincreasing concentrations of the calcium ionophore A23187 (C).The cell death was measured by PI-uptake. The activation ofcaspase cascade was assessed by Western blot detection ofcaspase-3 cleavage (D).176

Gene Therapy and Molecular Biology Vol 7, page 177Interestingly, despite having a strong effect on cell death(Figure 3A), caspase-8 deficient cells were equallysensitive towards the apoptotic form of cell death (Figure3B), measured by the “Nicoletti” method. To furtherconfirm the observation, the A23187 concentrationkinetics at 18 h was performed (Figure 3C). Similarly asin Figure 3A, here the cell death was measured by PIuptakethat cannot discriminate well between apoptosisand necrosis. Also this data fully confirmed theobservations that caspase-8 deficiency significantlyprotects from the ionophore-triggered death. To get furtherinsight into the death mechanisms induced by calciuminflux we have examined caspase-3 cleavage (activation)by Western blot (Figure 3D). To our surprise a significantportion of caspase-3 was cleaved unspecificaly, yieldingnon-active proteolytic fragments. The subsequentenzymatic measurement of caspase-3 (DEVDase) activityfully confirmed the Western blot data, showing only avery moderate increase in activity (data not shown).C. Calcium induced cell death is FADDindependent,and it is inhibitable by Bcl-2Since caspase-8 deficiency was largely protectiveagainst calcium-induced cell death in our experimentalsystem, we next tested the effect of FADD, the adaptormolecule that is necessary for caspase-8 recruitment todeath receptors. In addition, we examined the possibleinvolvement of apoptosome/mitochondrial death pathwayemploying Jurkat cells overexpressing Bcl-2 proteins.Cells overexpressing a mutated form of the FADDmolecule, that lack the death effector domain required forthe interaction with caspase-8, were as equally sensitive asthe control Jurkat cell line (Figure 4A). Thus, althoughcaspase-8 deficiency significantly impairs death triggeredby calcium, the adaptor molecule FADD plays no role inthe system. Whereas, Bcl-2 overexpression was fullyprotective against low concentrations (200 ng/ml) of thecalcium ionophore A23187 (Figure 4B). Higherconcentrations of A23187 (e.g. 400 ng/ml) partiallyovercame the Bcl-2 protective effect, but still about 50 %more of the Jurkat-Bcl-2 cells survived the forced calciuminflux as compared to the control Jurkat clone.IV. DiscussionThe presented study identifies a novel, caspase-8dependent, calcium-triggered pathway involved in thepropagation of cell death. The pathway differssignificantly from the classical, death receptor-triggeredapoptotic signaling cascades since it is FADDindependent.Caspase-8 requires adaptor molecules for itsactivation. This requirement can be fulfilled by the ERlocalizedprotein Bap31 that binds caspase-8(Breckenridge et al, 2002; Ducret et al, 2003). Theobserved sensitivity towards overexpression of Bcl-2 maybe indicative for the involvement ofmitochondrial/apoptosome-dependent signaling events.The Bcl-2 sensitivity of the pathway can also be explainedalternatively. It has been described previously (Foyouzi-Youssefi et al, 2000; Vanden Abeele et al, 2002) that someFigure 4. The effect of FADD death receptor adaptormolecule and Bcl-2 on calcium triggered apopotosis. FADDnegative-and control (J16) cells were treated with A23187 (400ng/ml) over different time points and cell death was measured byPI-uptake (A). To examine the effect of Bcl-2 on calciuminduced death we have used a Jurkat cell clone that overexpressthe protein. Time kinetics were done with two differentconcentrations of A23187. Bcl-2 almost completely inhibited celldeath induced by 200 ng/ml of A23187 (B), and it was about 40-50 % protective upon treatment with 400 ng/ml of the ionophore(C). Cell death was measured by PI-uptake.177

Burek et al: Calcium induced cell deathantiapoptotic Bcl-2 family members including Bcl-2 itselfand Bcl-X L , protect cells from calcium by lowering theCa 2+ -storage capacity of ER. Thus, the death stimuli thatcause the release of calcium from ER will be less efficientin elevating the cytoplasmic calcium concentration andtherefore, will less effectively activate the calciumdependentsignaling pathways.The death inducted by the calcium ionophore A23187was a mixture of necrosis and apoptosis. A critical factorthat influences the form of cell death (apoptotic ornecrotic) is the cellular ATP content. Stimuli that undernormal condition induce apoptosis will cause classicalnecrotic cell death if the cellular concentration of ATPdrops below 10-15 % of the normal level (Nieminen et al,1994; Los et al, 2002). One of the mechanisms that causesevere ATP depletion is the uncoupling of phosphorylativeoxidation and ATP production caused by mitochondrialpermeability transition (MPT). MPT may be triggered by arising Ca 2+ level and the subsequent activation of thehypothetical permeability transition pore componentcyclophilin D. Once the pH and electrical gradient acrossthe inner mitochondrial membrane collapses the finalenzyme of the mitochondrial respiratory chain, the F 1 F 0 -ATPase, that normally converts ADP to ATP, reverses andconsumes ATP while trying to restore the gradient. Thismechanism is among the strongest depletors of cellularATP, since it also consumes ATP produced by thecompensatory, glycolytic pathway (reviewed in Lemasterset al, 2002; Hajnoczky et al, 2003). The above mechanismpermits both necrotic- and apoptotic death. A strongincrease of Ca 2+ concentration would cause a significantportion of mitochondria to collapse, massive ATPdepletion would follow, thus, cells would die by necrosis.A less pronounced rise of calcium concentration wouldresult in a slow and asynchronous MPT occurrence.Affected mitochondria would release proapoptoticmolecules like cytochrome c, AIF and endonuclease G.While the depletion of ATP would not be significant, thecell would have enough energy to die in an orderly,apoptotic fashion. This is exactly what we observed in ourexperimental system. While low concentrations of thecalcium ionophore A23187 induce apoptosis, intermediateand higher concentrations of it cause substantial necrosis.In summary, we are presenting here evidence for anew caspase-8-dependent calcium-induced death pathway.Since it is FADD-independent, we hypothesize that theBap31 ER-localized adaptor molecule is involved in thepathway. In addition to the ER-compartment, themitochondrial death pathways are important mediators ofdeath induced by an elevated cellular calcium level.AcknowledgementsThis work was supported by grants from “DeutscheKrebschilfe” (10-1893), DFG (Lo 823/1-1 and Lo 823/3-1), and by IZKF-Muenster, (E-8).ReferencesBarrett AJ, and Rawlings ND (2001) Evolutionary lines ofcysteine peptidases. Biol Chem 382, 727-733.Barros LF, Castro J, and Bittner CX (2002) Ion movements incell death: from protection to execution. Biol Res 35, 209-214.Bouillet P, and Strasser A (2002) BH3-only proteins -evolutionarily conserved proapoptotic Bcl-2 family membersessential for initiating programmed cell death. J Cell Sci115, 1567-1574.Breckenridge DG, Nguyen M, Kuppig S, Reth M, and Shore GC(2002) The procaspase-8 isoform, procaspase-8L, recruitedto the BAP31 complex at the endoplasmic reticulum. ProcNatl Acad Sci U S A 99, 4331-4336.Cassens U, Lewinski G, Samraj AK, von Bernuth H, Baust H,Khazaie K, and Los M (2003) Viral modulation of cell deathby inhibition of caspases. Arch Immunol Ther Exp 51, 19-27.Denis F, Rheaume E, Aouad SM, Alam A, Sekaly RP, andCohen LY (1998) The role of caspases in T cell developmentand the control of immune responses. Cell Mol Life Sci 54,1005-1019.Ducret A, Nguyen M, Breckenridge DG, and Shore GC (2003)The resident endoplasmic reticulum protein, BAP31,associates with gamma-actin and myosin B heavy chain. EurJ Biochem 270, 342-349.Duke RC, Witter RZ, Nash PB, Young JD, and Ojcius DM(1994) Cytolysis mediated by ionophores and pore-formingagents: role of intracellular calcium in apoptosis. Faseb J 8,237-246.Errasfa M, and Stern A (1994) Melittin inhibits epidermal growthfactor-induced protein tyrosine phosphorylation: comparisonwith phorbol myristate acetate and calcium ionophoreA23187. Biochim Biophys Acta 1222, 471-476.Foyouzi-Youssefi R, Arnaudeau S, Borner C, Kelley WL,Tschopp J, Lew DP, Demaurex N, and Krause KH (2000)Bcl-2 decreases the free Ca2+ concentration within theendoplasmic reticulum. Proc Natl Acad Sci U S A 97, 5723-5728.Franklin JL, and Johnson EM, Jr. (1992) Suppression ofprogrammed neuronal death by sustained elevation ofcytoplasmic calcium. Trends Neurosci 15, 501-508.Gwag BJ, Canzoniero LM, Sensi SL, Demaro JA, Koh JY,Goldberg MP, Jacquin M, and Choi DW (1999) Calciumionophores can induce either apoptosis or necrosis incultured cortical neurons. Neuroscience 90, 1339-1348.Hajnoczky G, Davies E, and Madesh M (2003) Calciumsignaling and apoptosis. Biochem Biophys Res Commun304, 445-454.Herr I, and Debatin KM (2001) Cellular stress response andapoptosis in cancer therapy. Blood 98, 2603-2614.Krammer PH (2000) CD95's deadly mission in the immunesystem. Nature 407, 789-795.Kressel M, and Groscurth P (1994) Distinction of apoptotic andnecrotic cell death by in situ labelling of fragmented DNA.Cell Tissue Res 278, 549-556.Leist M, and Jaattela M (2001) Four deaths and a funeral: fromcaspases to alternative mechanisms. Nat Rev Mol Cell Biol2, 589-598.Lemasters JJ, Qian T, He L, Kim JS, Elmore SP, Cascio WE, andBrenner DA (2002) Role of mitochondrial inner membranepermeabilization in necrotic cell death, apoptosis, andautophagy. Antioxid Redox Signal 4, 769-781.Los M, Mozoluk M, Ferrari D, Stepczynska A, Stroh C, Renz A,Herceg Z, Wang Z-Q, and Schulze-Osthoff K (2002)Activation and caspase-mediated inhibition of PARP: amolecular switch between fibroblast necrosis and apoptosisin death receptor signaling. Mol Biol Cell 13, 978-988.Los M, Stroh C, Janicke RU, Engels IH, and Schulze Osthoff K(2001) Caspases: more than just killers? Trends Immunol22, 31-34.178

Gene Therapy and Molecular Biology Vol 7, page 179Los M, van de Craen M, Penning CL, Schenk H, Westendorp M,Baeuerle PA, Dröge W, Krammer PH, Fiers W, and Schulze-Osthoff K (1995) Requirement of an ICE/Ced-3 protease forFas/Apo-1-1mediated apoptosis. Nature 371, 81-83.Los M, Wesselborg S, and Schulze Osthoff K (1999) The role ofcaspases in development, immunity, and apoptotic signaltransduction: lessons from knockout mice. Immunity 10,629-639.Marsden VS, and Strasser A (2003) Control of Apoptosis in theImmune System: Bcl-2, BH3-Only Proteins and More. AnnuRev Immunol 21, 71-105.Nakamura J (1996) Calcium ionophore, A23187, alters the modeof cAMP formation in wild-type S49 murine lymphomacells. Biochim Biophys Acta 1313, 6-10.Nieminen AL, Saylor AK, Herman B, and Lemasters JJ (1994)ATP depletion rather than mitochondrial depolarizationmediates hepatocyte killing after metabolic inhibition. Am JPhysiol 267, C67-74.Ning ZQ, and Murphy JJ (1993) Calcium ionophore-inducedapoptosis of human B cells is preceded by the inducedexpression of early response genes. Eur J Immunol 23,3369-3372.Ojcius DM, Zychlinsky A, Zheng LM, and Young JD (1991)Ionophore-induced apoptosis: role of DNA fragmentationand calcium fluxes. Exp Cell Res 197, 43-49.Renz A, Berdel WE, Kreuter M, Belka C, Schulze-Osthoff K,and Los M (2001) Rapid extracellular release of cytochromec is specific for apoptosis and marks cell death in vivo.Blood 98, 1542-1548.Sadowski-Debbing K, Coy JF, Mier W, Hug H, and Los M(2002) Caspases – their role in apoptosis and otherphysiological processes as revealed by knock-out studies.Arch Immunol Ther Exp 50, 19-34.Stennicke HR, Ryan CA, and Salvesen GS (2002) Reprievalfrom execution: the molecular basis of caspase inhibition.Trends Biochem Sci 27, 94-101.Strasser A, O'Connor L, and Dixit VM (2000) Apoptosissignaling. Annu Rev Biochem 69, 217-245.Stroh C, Cassens U, Samraj AK, Sibrowski W, Schulze-OsthoffK, and Los M (2002) The role of caspases in cryoinjury:caspase inhibition strongly improves the recovery ofcryopreserved hematopoietic and other cells. FASEB J 16,1651-1653.Suzuki M, Youle RJ, and Tjandra N (2000) Structure of Bax:coregulation of dimer formation and intracellularlocalization. Cell 103, 645-654.Vanden Abeele F, Skryma R, Shuba Y, Van Coppenolle F,Slomianny C, Roudbaraki M, Mauroy B, Wuytack F, andPrevarskaya N (2002) Bcl-2-dependent modulation of Ca(2+)homeostasis and store-operated channels in prostate cancercells. Cancer Cell 1, 169-179.Walczak H, and Krammer PH (2000) The CD95 (APO-1/Fas)and the TRAIL (APO-2L) apoptosis systems. Exp Cell Res256, 58-66.Xu K, Tavernarakis N, and Driscoll M (2001) Necrotic cell deathin C. elegans requires the function of calreticulin andregulators of Ca(2+) release from the endoplasmic reticulum.Neuron 31, 957-971.Zheng TS, and Flavell RA (2000) Divinations and surprises:genetic analysis of caspase function in mice. Exp Cell Res256, 67-73.Marek Los, MD, PhD179

Burek et al: Calcium induced cell death180

Gene Therapy and Molecular Biology Vol 7, page 181Gene Ther Mol Biol Vol 7, 181-209, 2003The current status and future direction of fetal genetherapyReview ArticleAnna L David 1 , Michael Themis 2 , Simon N Waddington 2 , Lisa Gregory 2 , SuzanneMK Buckley 2 , Megha Nivsarkar 2 , Terry Cook 3 , Donald Peebles 1 , Charles HRodeck 1 , Charles Coutelle 21 Department of Obstetrics and Gynaecology, Royal Free and University College London Medical School, London WC1E6HX2 Gene Therapy Research Group, Section of Cell and Molecular Biology, Division of Biomedical Sciences, ImperialCollege School of Medicine, London SW7 2AZ3 Department of Histopathology, Imperial College School of Medicine, London W12 0HS__________________________________________________________________________________*Correspondence: Dr A.L. David, Room 212, 2 nd floor, Department of Obstetrics and Gynaecology, Royal Free and University CollegeMedical School, 86-96 Chenies Mews, London, WC1E 6HX, UK. Telephone: +44-20-7679-6059; Fax: +44-20-7383-7429; e-mail:a.david@ucl.ac.ukKey words: fetal gene therapy; adenovirus; retrovirus; lentivirus; adeno-associated virus; Sendai virus; liposomeAbbreviations: Cystic fibrosis (CF), Cystic Fibrosis Transmembrane Regulator (CFTR), and ornithine transcarbamylase (OTC),lysosomal storage disorders (LSDs), cerebrospinal fluid (CSF), Duchenne muscular dystrophy (DMD), Spinal muscular atrophy (SMA),survival motor neuron gene 1 (SMN 1), adeno-associated viral (AAV), severe combined immunodeficiency disorders (SCID), recessiveadenosine deaminase deficiency (ADA), bone marrow transplantation (BMT), dystrophic form of epidermolysis bullosa (DEB),congenital diaphragmatic hernia (CDH), Intrauterine growth restriction (IUGR)Received: 18 September 2003; Accepted: 29 October 2003; electronically published: November 2003SummaryApplication of gene therapy in utero has been considered as a strategy for treatment or even prevention of earlyonset genetic disorders such as cystic fibrosis and Duchenne muscular dystrophy. Prenatal gene transfer may targetrapidly expanding stem cell populations that are inaccessible after birth, permit induction of immune toleranceagainst vector and transgene and allow permanent gene transfer by use of integrating vector systems. Application ofthis therapy in the fetus must be safe, reliable and cost-effective. Recent developments in the understanding ofgenetic disease, vector design, and minimally invasive delivery techniques have brought fetal gene therapy closer toclinical practice. Prenatal studies in animal models are being pursued in parallel with adult studies of gene therapy,but they remain presently at the experimental stage.I. IntroductionGene therapy uses the intracellular delivery ofgenetic material for the treatment of disease. A wide rangeof diseases including cancer, vascular andneurodegenerative disorders and inherited genetic diseasesare being considered as targets for this therapy in adults.Application of gene therapy in utero has been consideredas a strategy for treatment or even prevention of earlyonset genetic disorders such as cystic fibrosis andDuchenne muscular dystrophy (Coutelle et al, 1995). Genetransfer to the developing fetus may target rapidlyexpanding stem cell populations that are inaccessible afterbirth and may allow permanent gene transfer by use ofintegrating vector systems. The functionally immaturefetal immune system may permit induction of immunetolerance against vector and transgene, and therebyfacilitate repeated treatment after birth. Finally, and mostimportantly for clinicians, fetal gene therapy would give athird choice to parents following prenatal diagnosis ofinherited disease, where currently termination ofpregnancy or acceptance of an affected child have been theonly options. Application of this therapy in the fetus mustbe safe, reliable and cost-effective. Recent developmentsin the understanding of genetic disease, vector design, andminimally invasive delivery techniques have brought fetalgene therapy closer to clinical practice. Prenatal studies inanimal models are being pursued in parallel with adultstudies of gene therapy, but they remain presently at theexperimental stage. This review explores the latestdevelopments in the field of in utero gene therapy andtheir implications for its future clinical application.181

David et al: Current status and future direction of fetal gene therapyTable 1: Examples of candidate diseases for fetal gene therapyDisease Therapeutic gene product Target cells/organCystic fibrosis (CF) CF transmembrane regulator airway and intestinal epithelial cellsMetabolic disorders:Ornithine transcarbamylase deficiency Ornithine transcarbamylase hepatocytesGlycogen storage disorders:Pompe disease α1,4-glucosidase hepatocytes, myocytes and neuronsSphingolipid storage disorders:Tay-Sachs disease β-N-acetylhexosaminidase fibroblasts, neuronsMucopolysaccharide storage disorders:Sly disease β-glucuronidase hepatocytes, neuronsMuscular dystrophies:Duchenne dystrophin myocytesNeurological disorders:Spinal muscular atrophy survival motor neuron protein motor neuronsHaemophilias:Haemophilia B human factor IX clotting factor hepatocytesHaemoglobinopathies:β o -thalassemia β-globin chains of haemoglobin haematopoietic precursor cellsImmunodeficiency disorders:X-linked severe combinedγc cytokine receptorhaematopoietic precursor cellsimmunodeficiencySkin disorders:Dystrophic epidermolysis bullosa type VII collagen keratinocytesNon-inherited perinatal diseases:Hypoxia-ischaemia neurotrophic factors cortical neuronsInfectious diseases:Herpes simplex herpes DNA oral mucosaPlacental disorderSevere pre-eclampsia nitric oxide synthase trophoblastsII. The candidate diseasesFetal gene therapy has been proposed to beappropriate for life-threatening disorders, in whichprenatal gene delivery maintains a clear advantage overcell transplantation or postnatal gene therapy and forwhich there are currently no satisfactory treatmentsavailable (Wilson and Wivel 1999). Some of the diseasesthat may be suitable for in utero treatment are listed inTable 1 and are discussed as examples for conditions withsimilar manifestations and/or target tissues.A. Cystic fibrosisCystic fibrosis (CF) appears to be an ideal candidatefor treatment with in utero gene therapy. Firstly it is themost common lethal autosomal recessive disorder inCaucasians with an incidence of 1 in 2000 livebirths inWestern Europe and North America. Several mutations ofthe Cystic Fibrosis Transmembrane Regulator (CFTR)gene encoding the CFTR protein have been identified andthe resulting disease is characterized by abnormalelectrolyte transport in the epithelia of the airways, theducts of the sweat glands and exocrine pancreas, and theintestine. The main sites of CFTR expression in the non-CF human bronchi are the submucosal glands (Engelhardtet al, 1992). In vitro studies where normal and CF airwaycells were mixed, suggest that as few as 6-10% of cellsexpressing normal CFTR are required to correct thechloride transport defect of an epithelial cell monolayer(Johnson et al, 1992); thus, successful gene therapy mayrequire only relatively low level epithelial airwaytransduction.Phase I gene therapy trials directed towardspulmonary disease in CF have shown equivocal resultsand highlight the problems of present gene therapyapproaches in adults (Bigger and Coutelle 2001). Thelungs may already be severely damaged or obstructed,even in young adult patients, limiting delivery of genetherapy to the airway epithelium. Fluorocarbon liquidssuch as perflubron have recently been shown to improvedistribution of adenoviral vectors and gene expression innormal and diseased adult lungs (Weiss et al, 1999a,2001). Pretreatment of airways with detergents (Parsons etal, 1998) or the fatty acid sodium caprate (Gregory et al,2002) or EGTA (Wang et al, 2000) also improvesadenovirus-mediated airways transduction. A comparisonof agents to modulate paracellular permeability showedthat pretreatment of adult murine airways with sodiumcaprate had a good safety profile, and enhancedadenovirus-mediated gene transfer to the trachea moreefficiently than sodium laurate, another fatty acid sodiumsalt or EGTA, a calcium chelator (Johnson et al, 2003).Immune responses to the vector, particularly in the case ofadenoviral vectors, limit the dose that may be safelyadministered, and reduce the duration of expression.The CFTR gene has been proposed to play animportant, albeit still unknown, physiological role innormal fetal development (Gaillard et al, 1994; Tizzano etal, 1994). Furthermore the cystic fibrosis disease processappears to begin during development of CF fetuses sinceby the mid-trimester a pro-inflammatory state exists infetal CF airways (Hubeau et al, 2001) and there areabnormalities of the pancreas and small bowel (Boué et al,182

Gene Therapy and Molecular Biology Vol 7, page 1831986). Prenatal diagnosis is usually performed bydetection of the CFTR mutation in placental tissue, fetalskin cells or blood after chorionic villus biopsy,amniocentesis or cordocentesis respectively.Submucosal gland development has been studied inthe rhesus monkey fetus (Plopper et al, 1986) and althoughnot characterized in the human fetal airways, submucosalgland progenitors have been identified in the human adultlung (Engelhardt et al, 1995). Gene transfer to a humanfetal lung xenograft model in SCID mice was efficientlyachieved using adenoviral vectors (Peault et al, 1994) andlong-term expression in the surface epithelial andsubmucosal gland cells was observed up to 4 weeks and 9months after administration of adeno-associated andlentiviral vectors respectively (Lim et al, 2002, 2003).The early disease manifestation and poor results fromgene therapy treatment of adults with CF has led toresearch on in utero gene therapy for this disease in animalmodels. Despite the multiorgan manifestation of CF, firstapproaches are directed towards gene delivery to the fetalairways, which has been achieved by intra-amnioticapplication and, in larger animals by intratracheal injection(see chapter IV). Other genetic diseases which couldbenefit from progress achieved in pulmonary genedelivery are α1-antitrypsin deficiency (Stecenko andBrigham 2003) and surfactant protein B deficiency (Coleet al, 2003).B. Metabolic disordersInherited inborn errors of metabolism can affect anumber of metabolic pathways. For example the ureacycle disorders are caused by defects in genes encodingenzymes or membrane transporters in ureagenesis. Theirprevalence is approximately 1:30,000 births and ornithinetranscarbamylase (OTC) deficiency is one of the mostsevere of these conditions (Summar and Tuchman, 2001).OTC deficiency is transmitted as a partially dominant X-linked trait. In patients with partial OTC deficiency, suchas hemizygous males and heterozygous females, the firstclinical episode is delayed for months or years with lesssevere hyperammonemia. However, patients withcomplete OTC deficiency present with life-threateninghyperammonemia within one week of birth and despitemedical therapy to reduce the ammonia levels, 50% of thechildren are dead by the age of 4, and of those surviving,the mean IQ is less than 50 (Maestri et al, 1999). Since theurea cycle is principally sited in the liver, gene therapydirected towards hepatocytes has the potential to correctthe metabolic abnormality. Indeed the success of orthopticliver transplantation in long-term treatment of thiscondition supports the concept (Lee and Goss, 2001).Adenoviral vectors have been shown to transientlycorrect OTC deficiency in the sparse fur murine modelafter neonatal and adult treatment (Stratford-Perricaudet etal, 1990; Ye et al, 1996). In a phase I human clinical trialin patients with partial OTC deficiency, adenoviral vectorsexpressing the human OTC–cDNA were administered.There was evidence of dose-related toxicity to theadenovirus and the last patient treated suffered a systemicinflammatory response syndrome that lead to his death(Raper et al, 2002).Because of its early onset, severity and presentdifficulties in postnatal gene therapy, OTC deficiency is aninteresting candidate for in utero gene application targetedto the fetal liver (see chapter IV). Prenatal diagnosis forOTC deficiency by detection of the genetic mutation infetal DNA is available in families with a known congenitalabnormality. In non-informative families, deficiency ofOTC enzyme can be detected in the fetal liver after liverbiopsy (Holzgreve and Golbus, 1986). Other seriousgenetic diseases that would primarily require hepatocytedirected gene transfer are amino acid disorders (e.g.phenylketonuria, tyrosinaemia), carbohydrate disorders(e.g. galactosaemia) and fatty acid oxidation disorders(e.g. long-chain acyl-CoA dehydrogenase deficiency)(Preece and Green 2002).C. Storage disordersThe lysosomal storage disorders (LSDs) are a groupof congenital deficiencies of one or more lysosomalenzymes. In mucopolysaccharidosis type VII (MPS typeVII) a deficiency of β-glucuronidase activity leads toaccumulation of undegraded glycosaminoglycans inlysosomes. Clinically, patients develop hepatosplenomegaly,mental and growth retardation, hearing and visiondefects, skeletal deformities and die of cardiac failure.Many of the LSDs present already during fetal life withhydrops fetalis and prenatal diagnosis can be performed bydetection of β-glucuronidase deficiency in chorionic villior fetal blood (Geipel et al, 2002). Although individuallyrare, as a group they occur in approximately 1 in 7500 livebirths and are one of the more prevalent groups ofinherited diseases in humans (Wraith, 2002). Bone marrowtransplantation and enzyme replacement therapy are beingdeveloped for many of the mucopolysaccharidoses.However, the short half-life of lysosomal enzymes in thecirculation means that patients need biweekly parenteraladministration which increases the risk of an immuneresponse to the infused enzyme. In addition, systemicallyadministered enzyme is unable to cross the blood-brainbarrier and can therefore not be used to treat centralnervous system disease manifestation.The LSDs are considered to be good candidates forgene therapy and the liver may be the ideal site for genetransfer. Newly synthesized lysosomal enzymes aresecreted into the systemic circulation and are recapturedby distant cells. Based on the observed enzyme levels inpatients with mild late-onset disease, the amount ofenzyme needed to correct the deficiency may only be 1-10% of normal levels (Cheng and Smith, 2003). Genetransfer to naturally occurring animal models of MPS typeVII has been investigated using adeno-associated virus(Daly et al, 1999), adenovirus (Kamata et al, 2003) andlentivirus (McCray Jr et al, 2001). Intravenousadministration of retroviral vectors containing canine β-glucuronidase to neonatal MPS type VII dogs preventedsome bone and joint abnormalities, corneal clouding andheart valve defects that commonly occur in this animal183

David et al: Current status and future direction of fetal gene therapymodel (Ponder et al, 2002). Some aspects of bone diseasewere not prevented however, which may be due toabnormal bone formation in utero. There was also concernthat systemic gene therapy administration may not reachthe brain even in neonatal dogs when the blood-brainbarrier is still forming. The immature blood-brain andblood –cerebrospinal fluid (CSF) barrier is morepermeable to small proteins than in mature brains andthere is a developmentally regulated mechanism thatselectively transfers some larger proteins from the blood tothe CSF (Dziegielewska et al, 2001). Thus a prenatal genetransfer approach may be more effective and alsoapplicable to other disorders that affect the brain, such asthe glycosphingolipid lysosomal storage diseases (Gaucherand Tay-Sachs disease) (Jeyakumar et al, 2002).D. Muscular dystrophiesDuchenne muscular dystrophy (DMD) is thecommonest form of muscular dystrophy, a group ofcongenital disorders characterised by muscle wasting andweakness. This X-linked recessive disease has anincidence of 1 in 3500 live male births. Affected boys areusually diagnosed aged 3-4 years and characteristically,skeletal muscle degeneration after repeated rounds ofnecrosis is followed by the onset of fibrosis that eventuallyleads to muscle weakness and death (Emery, 1993).Patients are usually confined to a wheelchair by age 11years, and although improved nursing care and positivepressure ventilation to aid breathing allows some patientsto reach the 3rd decade, respiratory or cardiac failure is thecommon cause of death (Simonds et al, 2000). Prenataldiagnosis is available for almost all muscular dystrophiesincluding Duchenne (Emery, 2002). Current treatmentincludes supportive measures such as surgery forcorrection of contractures and prevention of respiratoryinfections. The disease is caused by mutations in the DMDgene that encodes the 427kDA protein dystrophin,associated with the sarcolemma in muscle. Skeletal andcardiac muscle biopsies from DMD patients arecharacterized by absent or abnormal dystrophin. Genetransfer into muscle cells has been explored usingnaturally occurring animal models of muscular dystrophythat involve mutations in the DMD gene (Wells and Wells,2000). The large size of dystrophin cDNA (14kb)precludes insertion into conventional vectors with theexception of gutless adenovirus. Consequently themajority of viral constructs incorporate mini ormicrodystrophin cassettes based on a 6.3kb truncateddystrophin gene resulting from a large inframe deletion inthe rod domain which was isolated from a Beckermuscular dystrophy patient with very mild symptoms.Adenoviral transfer of minidystrophin results in goodtransduction of neonatal mdx mouse muscle with reduceddegeneration and improved muscle mechanics (Deconincket al, 1996; Vincent et al, 1993). In the neonatal and adultmdx mouse, injection of an adeno-associated viruscontaining a minidystrophin into the leg muscle led tonormal myofiber histology and protected membraneintegrity (Wang B et al, 2000). The early onset of thisdisease, which begins to be visible histologically by the18th-20th week of gestation (Vassilopoulos and Emery,1977; Turkel et al, 1981) and presents clinically between2-4 years of age, complicates postnatal gene therapy. Thusa prenatal approach to treatment might prevent the diseaseprocess.Prenatal gene transfer may offer advantages overneonatal or adult treatment. Efficient gene delivery toseveral affected muscles groups is technically difficult andthe alternative may be efficient gene transfer to a largepercentage of existing and rapidly expanding muscle cellsin utero. Postnatal gene delivery is also complicated by therisk of cellular immune responses against the transgenicproteins as demonstrated in the dystrophin-deficient mdxmouse model by loss of transgenic dystrophin-expressingfibres following dystrophin gene transfer (Wells andWells, 2000; Chamberlain 2002). In contrast in utero genetransfer may avoid the development of immune reactionsto the vector or transgene product and enable repeatinjection postnatally. Furthermore immune responses havebeen reported in several adenovirus-mediated genetransfer studies although it was not possible to determinethe relative contribution of the immune response to thevector or transgene. In most DMD patients, there is a lackof dystrophin expression which could lead to a functionalcopy of the dystrophin protein being recognised as aforeign antigen. Gene transfer during fetal life could leadto immunological tolerance to the dystrophin or allowrepeated injection post-natally. Similar conditions such asthe congenital Emery-Dreifuss and Fukuyama musculardystrophies (Emery, 2002) could also potentially betreated using a prenatal gene transfer approach.E. Neurological disordersSpinal muscular atrophy (SMA) is one of the mostcommon inherited causes of childhood mortality, with anincidence of 1 in 10,000 live births. It is characterized byprogressive degeneration of alpha motor neurons withinthe spinal cord and results in proximal, symmetrical limband trunk muscle paralysis that leads to death (Crawfordand Pardo, 1996). SMA is caused by homozygous loss ormutation in the survival motor neuron gene 1 (SMN 1)which is telomeric. Humans and primates also have acentromeric copy called the SMN 2 gene but this fails toprovide sufficient full-length SMN protein to maintainmotor neurons. Evidence from family studies and animalmodels of SMA suggest that the number of copies of theSMN 2 gene may modify the severity of the disease. Genetherapy strategy would have to provide and express afunctional copy of the SMN gene in the relevant neuronalcells. Efficient expression of the SMN gene wasdemonstrated recently after adenovirus-mediated deliveryof the SMN gene to human primary fibroblasts from SMApatients in vitro (DiDonato et al, 2003). Intraspinal orintramuscular application of a vector targeting neuronalcells will be required for in vivo therapy and other diseasesrequiring this targeting include amyotrophic lateralsclerosis.Immunohistochemical analysis of normal fetal tissuehas demonstrated that the expression of SMN protein isrelatively high in skeletal muscle, heart and brain and184

Gene Therapy and Molecular Biology Vol 7, page 185undergoes a marked drop in the postnatal period. Incontrast, SMN protein is greatly reduced in all tissuesfrom fetuses affected with SMA (Burlet et al, 1998). Theseobservations suggest that SMN protein may be requiredduring embryo-fetal development and as such, prenatalgene transfer may be more effective than adult treatment.Prenatal diagnosis is available using deletion analysis ofthe SMN 1 gene (Matthjis et al, 1998).F. HaemophiliasThe haemophilias A and B are also particularlysuitable for gene therapy in utero. Both are X-linkedhereditary haemorrhagic disorders which occur in 1 in10,000 and 1 in 25,000 males respectively and are causedby the absence or dysfunction of the respective humanfactor VIII (hFVIII) or IX (hFIX) clotting factors (Furie etal, 1994). Current treatment uses replacement therapy withhFVIII or hFIX. Unfortunately, a number of patientsdevelop antibodies to therapy leading to ineffectivetreatment and occasional anaphylaxis (Lusher, 2000).Indeed, the complications of haemophilia treatment havein some cases been far worse than the diseases themselves,increasing their morbidity and mortality (Soucie et al,2000).As the coagulation factors are required in the bloodand can be secreted functionally from a variety of tissues,the actual site of production is not so important as long astherapeutic plasma levels are realized. Adult gene therapystrategies have therefore concentrated on application to themuscle or the liver. Successful delivery and expression ofFIX has been achieved in adult animal models ofhaemophilia B following portal intravascularadministration of adenoviral (Kay et al, 1994) andretroviral vectors (Kay et al, 1993). Sustained FIXexpression was also observed after intramuscular injectionof adult haemophiliac dogs with adeno-associated viral(AAV) vectors expressing canine FIX (Chao et al, 1999;Herzog et al, 1999) and after intravascular injection ofadult haemophiliac mice with AAV vectors expressinghFIX (Snyder et al, 1999). These results have culminatedin the first clinical trial in humans that shows promisingresults although only low level hFIX expression has so farbeen observed (Kay et al, 2000). Successful delivery andexpression of therapeutic hFIX without formation ofantibodies has been achieved following administration ofretroviral vectors in neonatal animal models (Xu et al,2003). Prenatal gene therapy could be applied to the fetusvia a number of routes including muscle, peritoneal,hepatic, intravascular or skin application. More recentlyour group has demonstrated that in utero application canprovide long-term postnatal correction of the haemophiliacphenotype in FIX deficient mice (Waddington et al,submitted). Prenatal diagnosis is available early inpregnancy (Ljung, 1999).G. Haematopoietic diseases1. The thalassaemiasThe thalassaemias are inherited anaemias caused byover 200 mutations and globally are the commonestmonogenic disorders. They are most prevalent in theMediterranean region, the Middle East, the Indiansubcontinent and South-East Asia where gene frequenciesreach 3-10% of the population (Weatherall and Clegg,1996). β-thalassaemia is characterized by insufficientproduction of the β-globin peptide by erythroid cellswhich results in low levels of the major form of adulthaemoglobin, HbA, made up of two α- and two β-globinchains. The excess α-globin chains then precipitate in theerythroid cells, impair their maturation and this leads tohaemolysis and anaemia. Homozygotes or compoundheterozygotes suffer with the most severe form of thedisease, β-thalassaemia major. Similarly α-thalassaemiaresults in excess β-globin chains due to different degreesof α-globin chain deficiency. In the most severe form, α o -thalassaemia, all four α-globin chains are defective orabsent which leads to hydrops fetalis and intrauterinedeath. Patients with thalassaemia require regular lifelongblood transfusions to survive although this leads to ironoverload that affects the liver, heart and endocrine organs.Prevention of iron overload with iron-chelating therapysuch as parenteral deferoxamine is the mainstay of currentpatient management. Therapies aimed to increase theproduction of fetal haemoglobin have had disappointingresults (Olivieri and Weatherall, 1998). Allogeneichaematopoietic stem cell replacement offers the onlydefinitive cure and has been successful in over 1000patients worldwide (Olivieri, 1999). Outcomes depend onwhether the patient has hepatomegaly, portal fibrosis andhas effective chelating therapy before transplantation. The3 year disease-free survival falls from over 90% to 60% inchildren with the above risk factors.Gene therapy approaches have aimed to stablyintroduce a regulated human globin gene intohaemopoietic stem cells. Recently high expression oferythropoietin was found to improve the anaemia of β-thalassaemia in a mouse model by induction of high levelsof HbF synthesis (Johnston et al, 2003). Expression oftransgenic globin sequences would need to be sustained,finely regulated and at high levels since haemoglobinsynthesis represents 95% of all protein synthesis inreticulocytes. Initial attempts at gene therapy using the β-globin gene and a minimal locus control region (LCR)incorporated into a retroviral vector showed low levels andshort-term expression of β-globin after transplantation oftransduced haematopoietic stem cells into lethallyirradiated mice (Raftopoulos et al, 1997; Sadelain 2002).More recently lentiviral vectors containing the β-globingene and larger LCR elements have been used to transfectbone marrow from β-thalassaemic mice. This was thentransplanted into β o -thalassaemic heterozygote mice andresulted in therapeutically relevant levels of circulatinghaemoglobin (May et al, 2000). An advantage of prenatalgene therapy application in this context could be theaccess to rapidly dividing stem cell populations. Prenataldiagnosis for haemoglobinopathies can be done byassessment of globin-chain synthesis in fetal blood or bydirect analysis of fetal DNA obtained by chorionic-villussampling or amniocentesis.Sickle cell disease, another inherited disorder ofhaemoglobin may also be amenable to prenatal genetherapy. In this condition missense mutations in the β-185

David et al: Current status and future direction of fetal gene therapyglobin gene lead to haemoglobin polymerization causingthe red blood cells to become deformed or ‘sickled’. Theability of gene therapy to correct the pathophysiology hasbeen demonstrated in a study in transgenic sickle Hbmouse models. Bone marrow transduced with lentiviralvectors containing a β A globin gene variant that preventshaemoglobin polymerization was transplanted into twomouse sickle cell disease models resulting in therapeuticcorrection of the disease (Pawliuk et al, 2001).2. Immunodeficiency disordersThe greatest success of gene therapy so far has beenin the treatment of congenital severe combinedimmunodeficiency disorders (SCID). These represent themost severe form of primary immunodeficiencies and theyoccur in approximately 1 in 75,000 births. The mostcommon types of SCID are X-linked (Xl-SCID) and theautosomal recessive adenosine deaminase deficiency(ADA) found in 50% and 15% of sufferers respectively. Inboth conditions the genetic defect causes a profound blockin T cell differentiation which leads to absent T cell andhumoral responses. Xl-SCID is due to a deficiency of theγc chain, an essential component of cytokine receptorswhich is necessary for T cell and natural killer celldevelopment. In ADA deficiency there is selectiveaccumulation of the toxic metabolite deoxyATP in T cells.Clinically the patients present with chronic diarrhoea andfailure to thrive with recurrent respiratory andopportunitstic infections leading to death within the firstyear of life (Cavazzana-Calvo et al, 2001).Histocompatible bone marrow transplantation(BMT) has been used to treat both conditions with somesuccess. Survival after transplantation with HLA-identicalbone marrow is over 90% but matched sibling donors areusually not available. Haploidentical BMT with T-celldepletion is commonly performed instead, with survivalrates of up to 78% although many patients require lifelongimmunoglobulin replacement therapy because ofinadequate humoral activity (Buckley RH et al, 1999). Inutero haematopoietic stem cell transplantation has beenachieved in fetuses with Xl-SCID by ultrasound guidedintraperitoneal or intravenous injection (Flake et al, 1996;Touraine 1992; Wengler et al, 1996; Westgren et al,2002). A selective T-cell and natural killer cellreconstitution can be achieved but B cell engraftment hasnot been detected. In ADA deficiency, a long-circulatingform of bovine ADA conjugated with polyethylene glycol(PEG-ADA) has been used to correct the metabolicabnormalities and prevent life-threatening opportunisticinfections.The strategy for gene therapy of SCID is based onthe concept that genetically corrected autologous T-cellprecursors should have a selective survival advantage overnon-corrected cells. In addition, patients are unable tomount an effective immune response to the transgenewhich has proved to be a major problem in gene therapytreatment of other genetic diseases. In ADA-SCID, clinicaltrials have used infusion of autologous peripheral T-cells,CD34 + bone marrow or umbilical cord blood cellstransduced with a retroviral vector containing ADAcDNA. The earlier trials did not use conditioning of thebone marrow and PEG-ADA treatment was continued inall patients during and after treatment which made itdifficult to evaluate immune function (Blaese et al, 1995;Bordignon et al, 1995; Kohn et al, 1995). Some patientsshowed long term persistence of the transduced cellsalthough at low level. A more recent trial was performedin two infants with nonmyeloablative conditioning usingbusulfan and without concurrent PEG-ADA treatment.Both patients showed sustained engraftment of geneticallycorrected haematopoietic stem cells with differentiationinto multiple lineages and improvement in their clinicalcondition (Aiuti et al, 2002).In a similar way Xl-SCID has been treated usingautologous transplantation of CD34 + bone marrowtransduced ex vivo with retroviral vectors containing the γcgene. Fifteen patients have now been treated and effectiveimmune reconstitution has been achieved in thirteenpatients (Friedmann, 2003). Unfortunately because of aserious adverse event in two of the patients, all genetherapy trials involving retroviral vectors inhaematopoietic stem cells were initially halted in the US(Gansbacher and European Society of Gene Therapy2003) (see VI Ethical and safety issues) and have nowbeen restricted to case by case reviewed permission(Friedmann, 2003). Nevertheless this study has shown theability of gene therapy to cure such conditions. Because ofthe survival advantage of genetically corrected cells andthe ineffective immune response in SCID patients, it isunlikely that prenatal gene transfer would provide aparticular benefit over postnatal treatment of thiscondition.H. Skin disordersFetal gene delivery into the amniotic cavity mayhave unique benefits for treatment of inherited skindisorders. Epidermolysis bullosa is a group of inheritedblistering diseases characterized by epidermal-dermalseparation resulting from mutations that affect the functionof critical components of the basement membrane zone.The dystrophic form of epidermolysis bullosa (DEB) isdue to mutations in COL7A1, the gene encoding type VIIcollagen and has a prevalence of up to 2.4 per 100,000population (Horn and Tidman, 2002). The clinicalpresentation varies from a mild dominantly inheriteddisease characterized by skin and oral blisters and naildystrophy to a severe recessive subtype in which patientssuffer from contractures, severe dental caries, dysphagia,anal fissures and squamous cell carcinoma. Currenttherapy involves management of the diseasemanifestations with proper wound care, surgical release ofskin contractures, balloon dilatation of oesophagealstrictures and graft skin therapy (Pai and Marinkovich,2002).Easy accessibility and visualization of skin make itan attractive target for gene therapy. Gene delivery can bein vivo by direct introduction to the skin by injection,electroporation or a ‘gene gun’. Alternatively a skinsample could be removed from the patient, and epidermalkeratinocytes cultured and transduced ex vivo to insertgenetic material and the genetically engineered cells186

Gene Therapy and Molecular Biology Vol 7, page 187returned in the form of a skin graft (Uitto and Pulkkinen,2000). Preliminary studies show keratinocytes andfibroblasts from patients with DEB can be successfullytransduced using lentiviral vectors containing theCOL7A1 transgene in vitro resulting in long-termexpression and synthesis of type VII collagen (Chen et al,2002). In a canine animal model of DEB, transduction ofkeratinocytes with a retrovirus containing the collagentype VII cDNA corrected the observable defects in in vitroreconstructed skin (Baldeschi et al, 2003). A non-viralgene transfer approach has been used for junctionalepidermolysis bullosa (JEB) in which there is severelaminin-5 deficiency. Integration of an attB-containinglaminin 5 β3 expression plasmid using φC31 integrase intohuman keratinocytes from JEB patients produced skintissue with no histological evidence of subepidermalblistering when regenerated on SCID mice (Ortiz-Urda etal, 2003).Epidermolysis bullosa however, is a generalizeddisorder affecting the entire skin and the extracutaneoustissues. Prenatal therapy delivered into the amniotic fluidwould bathe the entire skin surface and reach thegastrointestinal system by fetal swallowing. Injection intothe amniotic cavity can be performed safely at relativelyearly gestation, but the timing of intra-amniotic deliverywill be important from developmental considerations.Even at 20 weeks gestation, the fetal epidermis isincompletely keratinized and this would aid gene tranfer.However there is a high rate of apotosis in fetalkeratinocytes and therefore the ideal strategy would be totarget stem cells (Haake and Cooklis, 1997). Prenataldiagnosis for epidermolysis bullosa can now be performedwith a 98% success rate in at risk families, paving the wayfor preliminary studies into prenatal treatment (Pfendner etal, 2003). Disorders of defective keratinisation such asharlequin ichthyosis, an autosomal recessive severe andusually fatal congenital ichthyosis (Akiyama, 1998), mayalso be amenable to prenatal gene transfer.I. Perinatal diseasePulmonary hypoplasia is another important cause ofneonatal morbidity and mortality. In this condition, thefetal lungs fail to develop resulting in respiratoryinsufficiency at birth. Current neonatal management issupportive and involves surfactant replacement, carefulmechanical ventilation avoiding barotrauma and treatmentof pulmonary hypertension. Pulmonary hypoplasia canoccur when there is reduced or no liquor surrounding thefetus (oligo or anhydramnios) prior to 22 weeks gestation,most commonly because of preterm premature rupture ofthe membranes (PPROM). Serial amnioinfusion has beenused for the prevention of pulmonary hypoplasia withsome success but has a high complication rate (Tan et al,2003). Space occupying lesions that compress the lungswithin the chest cavity also result in pulmonaryhypoplasia. Examples of such conditions include pleuraleffusion associated with congenital cardiac defects andcongenital diaphragmatic hernia (CDH) in which thebowel herniates through the diaphragmatic defect. Fetalinterventions such as drainage of pleural effusions can beused to treat the underlying cause of the pulmonaryhypoplasia. Temporary occlusion of the trachea with anexpandable balloon for treatment of CDH results inimpressive expansion of the hypoplastic lung with trachealfluid. However ‘plugging’ has yet to be shown to improveoutcome in the long term (Harrison et al, 1998). Studiessuggest that pulmonary hypoplasia in CDH begins duringembryogenesis as an abnormality in growth factorsignalling and actually precedes the development of theanatomical defect (Jesudason 2002). Prenatal gene therapycould be envisaged in the future to enhance antenatal lunggrowth and maturation by the targeted delivery of growthfactors at specific times during lung development.J. Infectious diseaseInfectious diseases with pathogens such as Group Bstreptococcus, human immunodeficiency virus, hepatitis Bvirus and herpes simplex virus are a major cause ofneonatal morbidity and mortality. Transmission of thesediseases from mother to infant often occurs shortly before,during, or after birth by early rupture of the amnioticmembranes or direct contact with infectious secretionsduring labor and delivery. Delivery by caesarean section toprevent such contact, and antibiotic and maternal antiviraltreatments have been used with some success, particularlyin the prevention of vertical HIV transmission.Immunisation of the fetus with DNA vaccines in latepregnancy has been proposed as an alternative approach toprevent neonatal infection (Gerdts et al, 2000; Sarzotti etal, 1996; Watts et al, 1999). The mucosal surfaces of theeyes, respiratory and gastrointestinal tract are the primarysite of entry for infectious agents during birth and theneonatal period. Thus intra-amniotic or intra-oral deliveryof antigen would probably provide the best diseaseprotection. Studies in the fetal mouse (Sarzotti et al, 1996),sheep (Gerdts, et al, 2000) and baboon (Watts et al, 1999)have shown that fetal immunisation can induce activeimmunity in the newborn. In particular, in the fetal sheep,intra-oral administration of hepatitis B surface antigenDNA resulted in a higher protective antibody titre than anintramuscular injection of the recombinant protein vaccine(Gerdts, et al, 2003). The timing of such an intervention iscrucial since exposure of the fetus to the antigen beforeimmune competence is reached may result in tolerance. Inaddition a single in utero injection may not be sufficient tomaintain immunity. At present there is no clinicalindication for such a prenatal immunization strategy.K. Placental disordersPre-eclampsia/eclampsia is one of the leading causesof maternal and fetal morbidity and mortality. Theunderlying defect is believed to be inadequate deepplacentation that fails to transform the spiral arteries intouteroplacental vessels and thus limits placental blood flow(Brosens et al, 2002). Secondary damage such as fibrindeposition and thrombosis then limit placental perfusionfurther and there is also widespread activation of thematernal vascular endothelium leading to decreasedformation of vasodilators such as nitric oxide (Walker,2000). Gene therapy could be used to improve uteroplacentalperfusion by for example, temporary expressionof nitric oxide synthase or placental growth factor. This187

David et al: Current status and future direction of fetal gene therapycould prolong the pregnancy until fetal maturity wasattained and reduce the likelihood of long-termcomplications in the mother and fetus.Intrauterine growth restriction (IUGR) affects up to8% of all pregnancies. It commonly occurs in pregnanciescomplicated by pre-eclampsia but can also arise innormotensive pregnancy. As well as leading to neonatalproblems, the long-term consequences are serious sinceIUGR infants exhibit higher rates of coronary heartdisease, type 2-diabetes, hypertension and stroke as adults(Barker et al, 1993). Abnormalities in placentaldevelopment are believed to adversely affect placentalfunction and deprive the fetus of the nutrients required foroptimal growth. Transport of amino acids and essentialfatty acids across the placenta is altered in IUGR fetusesand impaired oxygenation and acid base balance may beseen in severe cases (Pardi et al, 2002). Prenatal genetherapy could target placental transport mechanisms andincrease the availability of essential nutrients to the fetus.III. Vectors for in utero gene deliveryThe development of efficient vector systems iscrucial for the success of gene therapy. The ideal vectorfor fetal somatic gene therapy would introduce atranscriptionally regulated therapeutic gene into all organsrelevant to the genetic disorder by a single safeapplication. Although none of the present vector systemsmeet all these criteria, many of them have characteristicsthat may be beneficial to the fetal approach.A. Non-viral vectorsCationic liposome/DNA complexes have theadvantage of being relatively non-toxic and nonimmunogenicbut are still very inefficient in vivo. Anotherdrawback with these vehicles is that the DNA introducedas plasmid molecules remains episomal and will be lostover time following cell division. This is a particulardisadvantage in the fetus where cell populations arerapidly dividing. However, short term transgeneexpression has been shown to be a promising approach tomaintain a patent ductus arteriosus prior to surgery forcongenital heart defects in neonates (Mason et al, 1999).Liposomes containing plasmid expressing a decoy RNAdesigned to sequester fibronectin mRNA binding proteinwere delivered to the ductus arteriorus in fetal sheep at 90days of gestation, prior to the onset of intimal cushionformation at 100 days of gestation. Fibronectin synthesiswas inhibited resulting in a 60% reduction in intimalthickness and increased ductal patency at term.More recently, non-viral systems have beendeveloped that integrate into the host genome and couldthus in principle provide long term gene expression, butthese vectors are still at an early stage of experimentaldesign (Olivares et al, 2002).B. Viral vectorsStudies of in utero gene therapy have thereforeconcentrated on viral vectors, many of which have beendesigned to deliver reporter genes such as the β-galactosidase gene (lacZ). These allow tracking of thetransduced cells and to define tissue expression bybiochemical staining assays. Alternatively, use of vectorscarrying therapeutic genes allows the assessment ofpotentially curative levels of the expressed protein and, inanimal models of disease, even the observation ofphenotype correction. The hFIX gene for instance, can beused both as a marker gene, allowing the analysis of bloodlevels of the hFIX protein over time in non-haemophiliacanimals, and to study the correction of the blood clottingparameters in animal models of haemophilia. Postnatalreadministration of hFIX protein or the hFIX vector tofetally treated animals can be used to examine whetherimmune tolerance has been achieved.1. RetrovirusVectors that are able to integrate into the hostgenome such as retroviruses, lentiviruses and to a lesserextent adeno-associated viruses, may offer the possibilityof permanent gene delivery. Although only fairly lowvirus titres can be produced, virus gene transfer may beimproved by complexing vectors with cationic agents,(Themis et al, 1998) or by the administration of retrovirusproducer cells in vivo to allow localised gene deliveryclose to the site of cell transfer (Douar et al, 1997; Russelet al, 1995).Retroviruses require dividing cells for gene transfer(Miller DG et al, 1990) which suggests that they may bebetter suited for use in fetal tissues where cells are rapidlydividing rather than in adult applications. Other problemsinclude reports of premature promoter shutdown (Palmeret al, 1991; Challita and Kohn 1994) leading totranscriptional shutoff. Human serum can almostcompletely inactivate some retroviral particles (Welsh etal, 1975) which limits their use in vivo although increasedresistance to serum inactivation can be achieved bygenerating retroviruses from particular human packagingcells (Cosset et al, 1995) or by pseudotyping, whichreplaces the natural envelope of the retrovirus with aheterologous envelope (Engelstädter et al, 2001). Aparticular problem with in utero application is thatamniotic fluid has also been shown in vitro to have a mildinhibitory effect on retrovirus infection (Douar et al,1996). A further difficulty is the relatively short half-lifeof the retroviral particles in vivo which may hindertransduction because fetal cell division is nonsynchronizedand only those cells undergoing cell divisionat the time of infection will become transduced.Retroviruses were used in the first successful genetherapy trial, where bone marrow stem cells transduced exvivo with retroviral vectors expressing the correct cDNAwere delivered to infants suffering from an X-linked formof severe combined immunodeficiency (SCID)(Cavazzana-Calvo et al, 2000). The infants were able toleave protective isolation, discontinue treatment andappear to be developing normally (Hacein-Bey-Abina etal, 2002). However two of the fifteen patients treated forX-linked SCID have developed leukemia which has beenshown to involve insertional mutagenesis. An expandedclonal population of T-cells was demonstrated to be188

Gene Therapy and Molecular Biology Vol 7, page 189carrying the transgene inserted at 11p13 in the region ofLMO2, an oncogene frequently overexpressed in T cellleukemias (Marshall 2002). Insertional mutagenesis is anacknowleged potential complication with retroviralmediated gene transfer because gene integration occursrandomly into the genome. This is the first report ofmalignant change in humans following retroviral genetherapy and only one example has been found in extensiveanimal studies using this vector (Li et al, 2002).Investigations are ongoing to determine whether any otherfactor contributed to the development of insertionalmutagenesis and clonal expansion in these particularpatients (Friedmann 2003).2. LentivirusBecause of the limitation of infection to dividingcells by retroviruses, alternative vectors such aslentiviruses have been developed to circumvent thisrestriction. Significant progress has been made in recentyears in the development of lentiviral vectors, a retroviralsub-group based on the Human Immunodeficiency Virus(HIV) (Trono, 2000) or Equine Infectious Anaemia Virus(EIAV) (Mitrophanous et al, 1999). HIV vectors arecapable of transferring genes into nondividing cells suchas neurons (Naldini et al, 1996) and quiescenthaematopoietic progenitor cells, (Case et al, 1999) whichwill be particularly useful for these tissue targets.Lentiviral vectors integrate into the genome randomly andare therefore theoretically able to cause insertionalmutagenesis.Lentiviruses can be made more stable bypseudotyping which allows virus titres to be improved byultracentrifugation. This offers the opportunity of infectinga greater number of cells in vivo and different envelopesallow targeted gene transfer to specific tissues, forexample to the nervous system (Mazarakis et al, 2001) andairways (Kobinger et al, 2001). Both the EIAV vector, avector derived from non-primate animal lentiviruses,(Mitrophanous et al, 1999) and Feline ImmunodeficiencyVirus (FIV) (Wang, et al, 1999) have been developed in anattempt to create vectors for use in human treatment whichare not associated with any human pathology. Our recentwork has shown that high level sustained transgeneexpression can be achieved in a variety of tissues using theEAIV vector in fetal mice after intravascularadministration (Figure 1) (Waddington et al, 2003).3. Adeno-associated viral vectorsAdeno-associated virus (AAV) is also a promisingnovel vector system. It is a common human parvovirusthat is not associated with any human pathology. AAVnaturally requires co-infection with adenovirus as a helpervirus, but the latest AAV vectors circumvent the need foradenovirus and therefore make the production of pureAAV particles easier (Xiao et al, 1998). AAV is also ableto infect non-dividing cells and to achieve long-lastinggene correction in vitro and in vivo (Herzog et al, 1999;Wang et al, 1999; Kay et al, 2000). The basis for longtermtransgene expression is not quite clear. Integration ofthe wild type virus is predominantly at an apparentlyspecific functionally unimportant location on humanchromosome 19 reducing the theoretical risk of insertionalmutagenesis; however recombinant vector appears tointegrate at low levels and non-specifically (Monahan andSamulski, 2000). AAV vectors have a limited capacity forthe insertion of foreign genes that is about 4.7kb, althoughrecently 'split AAV vectors' have been designed wherelarge genes are split between two AAV genomes toincrease AAV packaging capacity. Afterconcatemerisation of these genomes in the host cellmRNA, splicing allows the removal of intervening ITRsequences and restoration of the split coding sequence toyield wild-type functional protein (Sun et al, 2000).Because the extent of AAV integration is still in question,this vector system may not give the permanent geneexpression ideal for in utero gene therapy without repeattreatment, although long term transgene expression afterintraperitoneal delivery in mice has recently been reported(Lipshutz et al, 2003). Some caution has also beenexpressed as AAV integration appears to inducechromosome deletions (Nakai et al, 2003).4. AdenovirusAdenoviral vectors have been used as attractivevectors for proof of principle studies in fetal gene therapysince they have continually achieved highly efficient genetransfer in vivo. The adenoviral coding sequencesnecessary for viral replication are deleted, rendering themreplication defective. They are relatively stable and can beobtained at high titre making systemic administration inhumans and large animal models feasible. The adenovirusgenome replicates outside the chromosome, which avoidsthe risk of insertional mutagenesis but results in onlytransient gene expression. Their broad host range andtropism to most cells of the human body, including therespiratory epithelium has made them very useful in initialpathfinder studies on vector delivery and transgeneexpression. They are particularly useful for exploringdifferent technical approaches to fetal gene therapy.Factors that determine the kinetics of transgeneexpression include vector elimination, since adenovirus isnot an integrating vector, and promoter shutdown.Adenoviral vectors are also highly immunogenic. Majorconcerns about the safety of adenoviral vectors wereraised following the death of Jesse Gelsinger from asystemic inflammatory response to a first generationadenovirus vector used for a phase I clinical trial towardsgene therapy of the inherited metabolic disorder, ornithinetranscarbamylase deficiency (Lehrman, 1999). Even fetaladministration of adenoviral vectors has been associatedwith an immune response (McCray, et al, 1995)particularly after postnatal repeat exposure to the vector(Iwamoto et al, 1999). Attempts to reduce theimmunogenicity and toxicity of the vector and to increaseits insert capacity have led to the generation of the socalled ‘gutless vectors’ in which essentially all adenoviralcoding sequences have been eliminated (Chen et al, 1997;Schiedner, et al, 1998).189

David et al: Current status and future direction of fetal gene therapyFigure 1. Upper panel. Representative sections of fetal livers harvested at 72h, 7, 14, 28, 79, 168 days and 1 year after yolk sacinjection of high titre titre EAIV SMART2Z (equine infectious anaemia virus vector expressing the β-galactosidase gene driven by theCMV promoter) lentiviral vector (n=1, 1, 3, 1 and 1, respectively). Uniform hepatocyte staining is observed after 72 h followed by theemergence of clusters of β-galactosidase-stained hepatocytes to day 79. Macroscopic appearance of liver sections (top row, x 10).Microscopic analyses (bottom row, x 400). Age matched noninfected control livers of 3 day old and 1-year-old animals are shown in thelower panel. Lower panel. Representative sections of fetal tissues harvested at 72 h, 7, 14, 79 days and 1 year after yolk sac injection ofhigh titre EAIV SMART2Z lentiviral vector (n…1, 1, 3 and 1, respectively). High-level staining is observed after 72 h and 79 days inbrain, 7, 14 and 79 days in heart and 14 and 79 days in skeletal muscle. Low-level expression is shown in lung and kidney at 79 dayspostinjection. Macroscopic appearance of tissues (left columns, x 10). Microscopic analysis (right column, x 400). (Waddington et al2003). Republished with permission from Nature Publishing Group.190

Gene Therapy and Molecular Biology Vol 7, page 191Because adenoviruses provide highly efficient genetransfer yet transient expression, novel hybrid vectors havebeen developed to take advantage of adenovirus infectivityand the permanent nature of integrative vectors such asretroviruses and lentiviruses (Murphy et al, 2002; Kuboand Mitani, 2003). Hybrid vectors may offer efficient geneexpression to fetal organs such as the lung in which it hasso far proved difficult to achieve high level gene transferwith integrating vectors.5. Sendai virusRecently, the negative strand RNA cytoplasmicallyreplicating Sendai virus, a member of the paramyxovirusfamily was developed as a gene transfer vector. Earlyvectors still capable of self-propagation, were found toprovide very high levels of marker gene expression in awide range of tissues including bronchial epithelium(Yonemitsu et al, 2000), skeletal muscle (Shiotani et al,2001) and vascular endothelium (Masaki et al, 2001).Second generation vectors, although still capable of intracytoplasmicreplication of the RNA genome, are incapableof intercellular propagation. In these vectors, genesencoding surface glycoproteins including thehaemaglutinin-neuraminidase (HN) protein or the fusion(F) protein, which are responsible for cell binding andinfection, have been deleted from the viral genome (Inoueet al, 2003). Injection of F-deficient Sendai virus vectorinto the fetal mouse via various routes including intravascular,intra-amniotic, intra-muscular, intra-peritonealand intra-spinal resulted in expression of marker gene ingut wall, lung, muscle, peritoneal mesothelia and dorsalroute ganglia respectively. Further optimisation will beneeded to develop these first generation constructs intoclinically applicable vectors (Waddington et al,submitted).IV. Fetal gene therapy studiesSince the initial attempts in the early 1990s, in uterogene therapy has been investigated in a range of differentanimals using a variety of techniques. The possible routesof administration are illustrated in Figure 2.Figure 2. Routes of administration of gene therapy to the fetus. Routes in italics have not yet been applied in a fetal animal model usingultrasound guided injection.191

David et al: Current status and future direction of fetal gene therapyA. Animal modelsSmall animals are the most commonly used becausethey offer a number of advantages. Transgenic mousemodels exist for many genetic diseases such as cysticfibrosis and haemophilia and this allows the therapeuticeffect of the gene therapy to be studied. Small animals arealso cheaper to maintain and have short breeding cycleswith large litters which permit studies over severalgenerations e.g. on germline transmission. However, theirsize precludes their use for the development of minimallyinvasive techniques for gene therapy delivery as requiredin human application.Studies in large animals have mainly used sheep, sincethey are well established as an animal model relevant tohuman fetal physiology, have a good tolerance to in uteromanipulations and a consistent gestation period of 145days, which is approximately half that of the human. Thereare some differences between ovine and human biology(Newnham and Kelly 1993). In late gestation the fetalgrowth rate in sheep is over double that in humans(Fowden, 1995) and the placental weight declines from 90days gestation while it remains static in the human(Barcroft and Barron 1946). However the major differenceis in the structure of the placenta. In sheep thesynepitheliochorial placenta consists of six tissue layers,three from the mother and three from the fetus, and it isthe most complete barrier possible (Benirschke andKaufmann 1990). The maternofetal interdigitations(placentomes) are spread throughout the uterine cavity andmay be difficult to avoid during ultrasound-guided uterineinterventions. In humans, there is only a single discoidplacenta and there is extensive invasion of theendometrium by the trophoblast that removes the threematernal tissue barriers and results in a hemomonochorialplacenta at term. Probably as a result of these structuraldifferences, γ-globulin does not pass from the mother tothe fetus in the sheep, but is able to cross the placenta inhumans.Nonhuman primates are close physiologically tohumans with menstrual cycles of similar length andhormonal control, comparable cellular and endocrineprocesses of implantation, and similar timetables ofprenatal development. The placental structure in somenonhuman primates is also the same, for example in therhesus monkey the placenta is hemomonochorial andbidiscoidal (Benirschke and Kaufmann 1990). For thisreason they are used as an animal model in studies ofteratology, developmental biology, infertility andcontraception (Hendrickx and Peterson 1997). Ultrasoundguided injection techniques as used in fetal medicine havealso been applied extensively in the fetal nonhumanprimate with comparable results (Tarantal, 1990).However nonhuman primates are more costly than sheepand are difficult to maintain.The rabbit has been studied in some prenatal genetherapy studies. Minimally-invasive percutaneousultrasound guided injection and fetoscopic procedures arealso being developed (Brandt et al, 1997; Papadopulos etal, 1999). Because of the small size of the fetus and litternumber however, technically this is only possible fromlate gestation. The guinea pig has the same placentalstructure as humans but they are not commonly used inprenatal gene therapy studies because of the small fetalsize and lack of transgenic models of disease.There are unfortunately few large animal models ofhuman genetic disease available for testing of genetherapy. Efforts to produce transgenic domestic animalsare continuing particularly in the pig, sheep and cow(Piedrahita 2000). There are however, some dog modelsincluding mucopolysaccharidosis type VII, Duchennemuscular dystrophy and haemophilia B, which are usefulfor investigating the therapeutic effect of gene therapy.The dog is also a suitable model for minimally invasivedelivery techniques and studies on prenatal gene transferhave used ultrasound guided intraperitoneal or yolk sacinjection through the exposed uterus (Lutzko et al, 1999;Meertens et al, 2002).B. Application routes in fetal medicineInvasive surgical techniques such as maternallaparotomy or hysterotomy must be performed to accessthe fetus in small animal models, but have also beenapplied in large animal studies such as in the sheep (Tranet al, 2000; Vincent et al, 1995). Surgery carries a highmorbidity from wound infection and haemorrhage and therisk of mortality is significant.Minimally invasive procedures with fibreoptictelescopes are currently in use in fetal medicine and arebeing adapted for application of gene therapy in largeanimal fetuses. Fetoscopy was developed in the late 1970sfor examination of 2 nd trimester fetuses and for fetal bloodsampling (Rodeck, 1980). The morbidity from fetoscopy issignificant however, because of the relatively largerdiameter of the puncture site in the fetal membranes whichleads to premature rupture of the membranes and pretermlabour and its associated problems. With the improvementin ultrasound technology in the 1990s, more detailedanatomical survey of the fetus could be performed andfetal blood sampling by ultrasound guided injectionbecame routine practice. Operative fetoscopy has recentlyre-emerged for use together with ultrasound in endoscopicfetal surgery for conditions such as twin reversed-arterialperfusionsequence (Quintero et al, 1994), severe feto-fetaltransfusion syndrome (Ville et al, 1997) and congenitaldiaphragmatic hernia (Harrison et al, 1998).Percutaneous ultrasound-guided injection is the leastinvasive technique for accessing the fetus and is usedfrequently in the clinical setting. Coelocentesis usesultrasound to guide a needle into the extraembryoniccoelom in the early first trimester. It has a success rate of>95% at 6-11 weeks of gestation, and has been suggestedas a possible technique for stem cell engraftment in earlygestation (Wilson and Wivel 1999). It may be of little use,however for in utero gene therapy because of the limitedtransfer from the extraembryonic coelom via the amnioticmembrane to the amniotic cavity (Jauniaux and Gulbis2000). Studies on the risk of miscarriage in ongoingpregnancies beyond the 1 st trimester followingcoelocentesis gave controversial results (Makrydimas et al,1997; Ross et al, 1997; Santolaya-Forgas et al, 1998).192

Gene Therapy and Molecular Biology Vol 7, page 193Amniocentesis is mainly used clinically for prenataldiagnosis. Although it is one of the safest intrauterineprocedures, intra-amniotic application of vectors may beonly of limited use in fetal gene therapy because of vectordilution by the large volume of amniotic fluid, although itwould be the ideal application route for in utero genetherapy of skin diseases.Accessing the systemic circulation has greaterpotential. In fetal medicine, fetal blood can be obtained inthe second trimester under ultrasound guidance either fromthe placental cord insertion, the fetal heart or more safelyfrom the intrahepatic umbilical vein (Chinnaiya et al,1998). The procedure has a good success rate clinically, islow risk and is used commonly for rapid karyotyping orfetal blood transfusion (Nicolini et al, 1990). From 12weeks of gestation ultrasound-guided intracardiacpuncture for fetal blood sampling has been performed onpatients undergoing surgical termination of pregnancy(Jauniaux et al, 1999). Similarly, radiolabelled fetal livercells were successfully injected into the heart of 13 weekold fetuses under ultrasound guidance (Westgren et al,1997) prior to prostaglandin termination of pregnancy. Nofetal heart rate abnormalities were detected and all fetuseswere alive at least 6 hours after the procedure.Intraperitoneal injection has been applied for in utero stemcell transplantation in humans from 14 weeks of gestation(Touraine 1999; Muench et al, 2001) and is an alternativeroute for blood transfusion before 18 weeks of gestation(Rodeck and Deans 1999). Ultrasound guidedintramuscular injection has been used to delivercorticosteroid therapy for maturation of preterm infantlungs and vitamin K to the fetus (Larsen et al, 1978;Ljubic et al, 1999).C. Direct targeting of the fetal circulationDelivery of vectors to the systemic fetal circulationappears to be a highly effective route for targeting genetherapy to a range of fetal tissues and particularly to theliver for treatment of diseases such as the haemophiliasand the metabolic and storage disorders. This can beaccomplished in small animals such as the mouse byintracardiac injection (Christensen et al, 2000; Wang et al,1998) or by injection into the yolk sac vessels (Schachtneret al, 1996). Indeed, yolk sac vessel injection of adenoviralvectors containing the hFIX gene into fetal mice resultedin therapeutic levels of hFIX expression (Waddington etal, 2002). Long-term transgene expression was observed inthe liver, heart, brain and muscle up to a year afterdelivery of lentiviral vectors containing the β-galactosidase gene into yolk sac vessels of fetal mice(Waddington et al, 2003) and was then used to achievecorrection of the haemophilic phenotype in factor IXdeffcient mice (Waddington, submitted).In larger animals such as in the sheep, intravasculardelivery can be achieved by injection via the umbilicalvein (Yang et al, 1999). Adenoviral vectors containing thelacZ or hFIX genes were delivered into the umbilical veinof late gestation fetal sheep using ultrasound-guidedpercutaneous injection from 102 days gestation (term =145 days) (Themis et al, 1999). Positive lacZ expressionwas seen in about 30% of fetal hepatocytes, and hFIXexpression in fetal and neonatal plasma by ELISA analysisreached therapeutic levels within a week of delivery in twoanimals.In early gestation, delivery of adenoviral vectors intothe umbilical vein of fetal sheep at 60 days of gestation viahysterotomy resulted in widespread transduction of fetaltissues (Yang et al, 1999). Our group has attemptedultrasound-guided umbilical vein injection of adenoviralvectors in fetal sheep at the earlier time of 53 days ofgestation but this was unsuccessful due to procedurerelatedmortality (David et al, 2003a).Ultrasound-guided intracardiac injection has beenused to deliver adenoviral vectors to the late gestation fetalrabbit (Wang et al, 1998). Transgene expression wasobserved in up to 40% of fetal hepatocytes and wastransient as expected. A fetal immune response to thevector and transgene was detected. Unfortunately theprocedure also had a 25-40% mortality rate, comparable toother studies on fetal blood sampling in rabbits (Moise etal, 1992). Although technically straightforward,ultrasound-guided intracardiac delivery of adenoviralvectors to fetal sheep in early gestation resulted in 100%mortality due to haemorrhage (David et al, 2003a).D. Alternative routes for targeting thefetal circulation and liverDue to the peculiarities of the fetal anatomy, vectordelivery via the umbilical vein or yolk sac vessels willpreferentially target the liver, which is an important organfor treatment of many genetic diseases. However in earlypregnancy this not been technically possible andalternative approaches to reach the liver and thecirculation have been tried.1. Intrahepatic injectionFetal intrahepatic injection has been performed inmice using adenoviral vectors (Lipshutz et al, 1999a, b,2000; Mitchell et al, 2000), adeno-associated vectors(Mitchell et al, 2000; Sabatino et al, 2002) and lentiviralvectors (MacKenzie et al, 2002). In these studies, highlevels of transgene expression in fetal hepatocytes wereobserved as well as gene transfer to other organs such asthe heart, spleen, lung, intestine and brain suggestinghaematogenic spread.Ultrasound guided intrahepatic injection has beenperformed in a few large animal models. In the lategestation fetal rabbit, X-gal staining of the fetalhepatocytes was seen 2 days after ultrasound guidedintrahepatic injection of adenoviral vectors containing theβ-galactosidase gene in late-gestation fetal rabbits(Baumgartner et al, 1999). Similarly, strong expression oftransgenic enhanced green fluorescent protein wasobserved in hepatocytes one month after ultrasoundguidedintrahepatic delivery of adeno-associated viralvectors to the late-gestation rhesus monkey (Lai et al,2002). Ultrasound guided intrahepatic injection in earlygestation sheep fetuses has also been performed with fetalsurvival rates of 81% (David et al, 2003a). Only low level193

David et al: Current status and future direction of fetal gene therapyhepatocyte transduction however was observed afteradenoviral and retroviral mediated gene transfer into fetalsheep (David et al, 2003a) and primates (Tarantal et al,2001b).2. Intraperitoneal injectionIntraperitoneal injection has also been used forsuccessful gene transfer to multiple tissues including theliver in fetal mice (Lipshutz et al, 1999b, c) rats(Hatzoglou et al, 1990, 1995) and sheep (Tran et al, 2000).Persistent peritoneal expression was observed 18 monthsafter intraperitoneal injection of adeno-associated virusserotype 2 (AAV2) vectors containing the luciferase genein fetal mice (Lipshutz et al, 2001). Recent studies in thefetal mouse have shown that transgene expression couldbe increased by intraperitoneal injection of AAV5serotype vectors rather than AAV2 serotype vectors andby changing from the elongation factor 1α or CMVpromoter to the woodchuck hepatitis virusposttranscriptional regulatory element (Lipshutz et al,2003).In large animal models, retroviral vectors containingthe α-L-iduronidase gene were delivered by ultrasoundguided injection after exteriorisation of the uterus into theperitoneal cavity or yolk sac of mid-gestation fetal dogswith canine α-L-iduronidase deficiency (mucopolysaccharidosistype 1). Low level tissue transduction wasobserved but expression of the transgene did not persistbeyond the neonatal period (Meertens et al, 2002). In earlygestation fetal primates, ultrasound guided intraperitonealinjection of Moloney murine leukemia virus amphotrophicand vesicular stomatitis virus-G protein (VSV-G)pseudotyped retrovirus and VSV-G pseudotyped HIV-1lentiviral vectors resulted in only low level tissuetransduction (Tarantal et al, 2001b). In contrast long-termtransduction of hematopoietic stem cells in the bonemarrow and blood could be demonstrated 5 yearsfollowing delivery of retroviral vectors into the peritonealcavity of early gestation fetal sheep at laparotomy (Poradaet al, 1998). Delivery of adenoviral vectors containing thehFIX gene to early gestation fetal sheep by ultrasoundguided intraperitoneal injection had good fetal survival of77% and therapeutic hFIX production was achieved, albeittransiently (Figure 3) (David et al, 2003a).Immunohistochemical analysis after delivery of adenoviralvectors containing the lacZ gene showed positivetransgene expression on the surface of the umbilical cord,in the fetal small bowel serosa and in the hepatocytesbeneath the fetal liver capsule following intraperitonealinjection (Figure 4 A-C). The intraperitoneal route alsogave the most comprehensive spread of vector to fetaltissues as determined by PCR analysis but no vector wasdetectable by sensitive PCR analysis in the germline oflambs born after each route of administration (David et al,2003a).E. Intramuscular injectionThe main aim of intramuscular injection is to targetthe muscle for treatment of muscular dystrophies but thisroute may also be used for ectopic production of proteinssuch as hFIX in the treatment of haemophilias. In the fetalmouse, injection of adenoviral vectors containing the β-galactosidase gene into the shoulder or hindlimbmusculature resulted in persistent muscle and livertransgene expression for 16 and 8 weeks respectively afterinjection (Yang et al, 1999). Intramuscular injection oflentiviral vectors led to transduction of myocytes andcardiomyocytes indicating systemic spread of the virusfrom the site of injection (MacKenzie et al, 2002).Our group successfully achieved in vivo expressionof hFIX after injection of adenovirus and AAV hFIXvectors in adult and fetal mice (Schneider et al, 2002). Arecent study using EIAV lentivirus containing the lacZgene combined intrathoracic, supracostal, intraperitonealand intramuscular injection of three limbs and a singleflank in the fetal mouse. This resulted in widespread geneexpression in all injected muscles and also the diaphragmand heart which are the essential muscle groups to bereached for successful gene therapy of DMD (Gregory etal, 2003).Finally, delivery of adenoviral vectors into thehindlimb musculature by ultrasound guided injection hasbeen explored in one study in the early gestation fetalsheep. Fetal survival was 91% and therapeutic levels ofhFIX were also obtained after injection of adenovirushFIX vector (Figure 3).Figure 3. Time course of transgene expression after ultrasoundguided intraperitoneal, intramuscular, intrahepatic or intraamnioticdelivery of an adenoviral vector containing the humanfactor IX gene to early gestation sheep fetuses. Concentrations ofhuman factor IX in fetal or lamb plasma were determined byELISA analysis. Fetal samples were collected at post mortem(David et al 2003a). Republished with permission from MaryAnn Liebert Inc, Publishers.194

Gene Therapy and Molecular Biology Vol 7, page 195Immunohistochemistry for β-galactosidase showed strongstaining of the hindlimb musculature and occasionalpositively stained hepatocytes after injection of adenoviruslacZ vector. PCR analysis of vector presence in fetaltissues confirmed that broad haematogenic spread ofvector had occurred (David et al, 2003a).Figure 4A-C. Expression of β-galactosidase byimmunohistochemistry 2 days after intraperitoneal or intraamnioticdelivery of an adenoviral vector containing the β-galactosidase gene to early gestation fetal sheep. Originalmagnifications are as indicated. Intraperitoneal injection at 52days of gestation, positive staining is seen in (A) fetal smallbowel serosa, (B) surface of umbilical cord and (C) fetalsubcapsular hepatocytes.F. Targeting the fetal airways1. Intra-amniotic injectionIntra-amniotic application has been investigatedextensively in small animal models. Adenoviral vectorsexpressing the lacZ gene have been delivered to the fetalrat (Sekhon and Larson, 1995), mouse (Holzinger et al,1995; Sekhon and Larson, 1995; Douar et al, 1997; Larsonet al, 1997; Larson et al, 2000a; Mitchell et al, 2000) andguinea pig (Senoo et al, 2000) while adeno-associatedviral vectors have been applied to the fetal mouse(Mitchell et al, 2000). In general, transgene expression ismaximal in those tissues in contact with the amniotic fluid,namely the amniotic membranes and the fetal skin withless transduction of the gut and the mucosae. Indeed,therapeutic plasma concentrations of hFIX were achievedin fetal mice after intra-amniotic injection of adenoviralvectors carrying the hFIX gene (Schneider et al, 1999) andthe transgenic protein remained detectable after birth.Intra-amniotic delivery of retroviral producer cells to thefetal mouse resulted in only low level transduction of theamniotic membranes and fetal skin and no airways or guttransduction (Douar et al, 1997).In larger animals such as the fetal sheep, ultrasoundguided intra-amniotic injection of an amphotropicretroviral producer cell line encoding the lacZ generesulted in inefficient tissue transduction (Galan et al,2002). Amniotic fluid was found to have an inhibitoryeffect on retroviral mediated tissue transduction, and thiseffect increased as gestational age progressed (Bennett etal, 2001). Better results have been obtained withadenoviral vectors. Low level transgene expression wasseen in the fetal oesophagus and trachea after injection ofadenoviral lacZ vectors at laparotomy in late gestationfetal sheep (Holzinger et al, 1995). Attempts to deliveradenoviral vectors into the amniotic cavity of fetal sheepusing catheters placed at laparotomy had high mortality(Iwamoto et al, 1999). Ultrasound-guided intra-amnioticdelivery of adenoviral vectors containing the lacZ or hFIXgenes has been achieved in the early gestation fetal sheep(33 - 39 days of gestation, term = 145 days) equivalent to8 – 10 weeks gestation in humans with 86% fetal survival(David et al, 2003a). Therapeutic plasma concentrations ofhFIX were detectable up to 11 days after injection (Figure3) and immunohistochemical analysis showed positiveexpression of β-galactosidase in the fetal skin and nasalcavities (Figure 4 D-F). This suggests that transduction ofkeratinocytes in utero may be able to deliver proteins tothe circulation as well as to treat hereditary skin diseasesuch as epidermolysis bullosa.Gene transfer to the fetal airways is important for inutero treatment of cystic fibrosis.However, no significant airway or gastrointestinaltissue transduction was seen after ultrasound-guided intra-195

David et al: Current status and future direction of fetal gene therapyamniotic delivery of adenoviral vectors to early gestationfetal sheep (David et al, 2003a). Similarly ultrasoundguidedintra-amniotic injection of adenoviral vectors inmid-trimester rhesus macaque fetuses resulted insignificant transgene spread to tissues coming into contactwith amniotic fluid but low level transgene expression inthe fetal airways and intestine (Larson et al, 2000b).Similar findings were observed in fetal rabbits(Boyle et al, 2001). Low levels of airway transduction areprobably due to dilution of the vector by the relativelylarger volume of the amniotic fluid as well as the lack offetal breathing movements or fetal swallowing at this earlygestation. It may be possible to enhance fetal breathingmovements in later gestation using agents such astheophylline (Moss and Scarpelli, 1981) that lead to anintake of amniotic fluid to the lungs against the continuousoutflow of tracheal fluid (Badalian et al, 1993; Kalache etal, 2000). Indeed increased intake of marker dye and someenhancement of adenovirus mediated marker geneexpression was observed in mouse fetuses aftertheophylline administration. However other still unknownfactors appear to influence the level of gene transfer to thefetal airways more effectively (Buckley, in preparation).Recent work in our laboratory aimed to reproduce theiconoclastic report by Larson et al, (1997) that the CFphenotypein CFTR-knockout mice can be cured by shorttermprenatal expression of CFTR from an adenovirusvector, could not substantiate this claim (Buckley et al,2003). We are, therefore, constructing integratingexpression vector systems under tissue specific promotercontrol to achieve long-term postnatal CFTR-geneexpression after in utero gene delivery.2. Direct lung parenchymal injectionDirect injection of the lung parenchyma has beenattempted to access the fetal airways but with poor results.In mid-gestation fetal primates, ultrasound guidedinjection of lentiviral vectors into the lung resulted in lowlevel transgene expression in the fetal airways (Tarantal etal, 2001a). However, in the mid-gestation sheep fetus,ultrasound-guided delivery of an adenoviral vector to thelung parenchyma elicited only localized gene transfer andno spread within the airways could be detected(unpublished results).Figure 4 D-F. Intra-amniotic injection at 33 days of gestation,positive staining is seen in (D) surface of umbilical cord, (E)fetal nasal cavity and (F) fetal skin (David AL et al 2003a).Republished with permission from Mary Ann Liebert Inc,Publishers3. Tracheal injectionDirect instillation of vector into the trachea has beenmore successful. Placement of catheters in the tracheae offetal sheep can be performed by highly invasivetechniques at laparotomy (McCray et al, 1995; Pitt et al,1995; Vincent et al, 1995) or fetoscopically (Sylvester etal, 1997; Yang et al, 1999). Low level transduction of theproximal airways can be achieved using adenoviral orretroviral vectors, and occlusion of the trachea with aballoon improves distal airway transduction. Thesetechniques however, carry a significant morbidity andmortality.Recently a percutaneous transthoracic route ofinjection of the fetal trachea has been developed in midgestationsheep using ultrasound guidance to target the196

Gene Therapy and Molecular Biology Vol 7, page 197fetal airways as illustrated in Figure 5 (David et al,2003b). Using this technique we achieved good transgeneexpression in the fetal trachea and airways followingintratracheal delivery of an adenovirus containing the β-galactosidase gene (Peebles et al, 2003). Transgeneexpression was enhanced by pretreatment of the fetalairways with sodium caprate, a fatty acid that opens thetight junctions between airways epithelial cells. Thisallows the vector to reach the basolateral surface where thecoxsackie-adenovirus receptor (CAR receptor) responsiblefor binding adenovirus is located. Further enhancement oftransgene expression was achieved by complexing theadenoviral vector with DEAE dextran, a polycation thatneutralizes the negative charge on the vector, improvingvector binding to the CAR receptor (Figure 6 and Figure7).Instillation of perflubron, an inert fluorocarbon, resulted ina redistribution of expression from the upper to theperipheral airways and is most likely due to flushing of thevector solution further down the airways by the waterimmiscible perflubron (Weiss et al, 1999b). These resultsshow proof of principle for the relatively safe andminimally invasive in utero delivery of a gene therapyvector to the fetal airways that resulted in levels oftransgene expression in the airway epithelia that may berelevant to a therapeutic application in cystic fibrosis genetherapy.G. Targeting the fetal gutIntrapharyngeal delivery has been attempted once infetal rabbits at laparotomy to target the fetalgastrointestinal system as a model for the treatment ofmeconium ileus due to cystic fibrosis (Wu et al, 1999).Gene transfer to the small bowel enterocytes was achievedbut there was significant maternal and fetal loss related toanaesthesia and the invasive surgery used. Ultrasoundguidedinjection of barium into the fetal stomach of rabbitshas been performed successfully (Brandt et al, 1997) andthis technique could be extended to deliver gene to thefetal gut. Gene delivery to the gut of fetal mice has beenobserved after intra-amniotic vector application and wasmost likely a result of fetal swallowing (Douar et al,1997).H. Delivery to the placentaTargeting the placenta could be used in the treatmentof placental disorders such as pre-eclampsia or intrauterinegrowth restriction. Low level gene transfer to theplacenta has been achieved using angiographically guidedinjection of non-viral vectors into the uterine artery(Heikkilä et al, 2001). The intraplacental route has beenattempted in mice, rats, guinea pigs and rabbits. Somaticgene transfer to the fetal heart and liver was achieved insome studies using mice (Woo et al, 1997; Türkay et al,1999), but others have found little or no fetal gene transferin mice and guinea pigs (Senoo et al, 2000) or rats (Xinget al, 2000). Commonly, the placenta showed the mosttransfection, but maternal tissues also demonstratedtransgene expression, which although not unexpected, isundesirable in therapy aimed at the fetus.Figure 5: (A) Ultrasonogram and (B) diagram of sheep fetus at114 days of gestation in longitudinal section. A 20 Gauge spinalneedle is inserted into the fetal thorax between the 3rd and 4thrib, penetrates the lung parenchyma and enters the fetal tracheajust proximal to the carina (David et al 2003b). Republished withpermission from S Karger AG, Basel.V. Development of the fetal immunesystemA major restriction in adult gene therapy is theimmune response to vector and/or transgene. In uteroapplication, on the other hand, aims to circumvent this bytreatment before maturity of the functional immune systemand this depends critically on the time at which fetaltolerance might be induced. The human immune systemdevelops progressively through the first trimester and isnot fully functional until 1-2 years after birth (Riley,1998). Lymphoid cells appear first in the fetal liver from 8weeks of gestation, with B lymphocytes and natural killercells predominating over T cells (Pahal et al, 2000). Tlymphocytes increase in number in the fetal liver andcirculation from 12 weeks of gestation.197

David et al: Current status and future direction of fetal gene therapyFigure 6: Na-caprate stimulation of DEAE dextran complexed adenovirus mediated airway transduction. Panel 1: Examples of stainingin the peripheral lungs after virus alone (a) and DEAE complexed virus (b) and of the trachea after Na-caprate pre-treatment anduncomplexed virus administration (c) in fetal sheep injected between 102 and 109 days of gestation. Panel 2a: Na-caprate pre-treatmentfollowed by DEAE dextran complexed virus in a 108 day sheep. Widespread gene expression was seen in the small (a), medium (b) andlarge (c) airways and also the main bronchi (d) and trachea (e). Panel 2b: Similar results were observed in a fetus injected at 81 days ofgestation. Expression was seen in the airways (a & b) and trachea (c) Panel 3: Na-caprate pre-treatment followed by DEAE dextrancomplexed virus followed by perflubron. Staining of the peripheral airways in transverse sections (a & b) and longitudinal sectionshowing gene expression was limited to the terminal branches of the bronchial tree (c). Some staining of the bronchioles (d) and trachea(e) was also observed, although less than in the absence of perflubron. Scale bar = 5mm in all cases. (Peebles et al 2003).198

Gene Therapy and Molecular Biology Vol 7, page 199Although they are not capable of producing a definitivecytotoxic response until 18 weeks of gestation (Mackenzieand Maclean, 1980) natural killer cells and some T celllines may provide a limited immune response earlier ingestation (Miyagawa et al, 1992; Phillips et al, 1992). Thefetal lamb is able to produce detectable circulatingantibodies in response to some antigenic stimuli from 66days of gestation (Silverstein et al, 1963) and to reject skingrafts after 77 days of gestation (Silverstein et al, 1964).This would suggest a 'window of opportunity' in the firstthird to half of pregnancy during which time introductionof foreign genetic material may not produce an immuneresponse. No humoral immune response to the transgenewas observed in early gestation fetal sheep, althoughantibodies to the adenoviral vector were detected for eachroute of injection (David et al, 2003a). Similarly, umbilicalvein injection of adenoviral vectors into fetal sheep at 60days of gestation via hysterotomy resulted in widespreadtransduction of fetal tissues with no humoral immuneresponse to the adenoviral vector (Yang et al, 1999).Expression of a foreign antigen during early fetaldevelopment may also result in its recognition as “self”where exposure of the fetus to foreign antigen ismaintained (Billingham et al, 1956; Binns, 1967) thusallowing development of tolerance. Evidence to supportinduced tolerance has been reported after in uterointraperitoneal delivery of retroviral vectors in fetal sheep(Tran et al, 2001).Induction of tolerance to transgene in adults althoughpossible, is expensive, therefore, prenatal induction oftolerance may provide an excellent alternative. Forexample, a single injection of adenovirus expressing thefactor IX gene into the fetal mouse was shown to providelong term, albeit diminishing expression over five months.Furthermore, 56% of these adult mice remained tolerant torepeated challenges with hFIX protein (Figure 8). Incontrast, a group of mice which received adenovirus forthe first time as adults developed high levels of anti-hFIXantibodies (Waddington et al, 2002). This provides proofof principle that gene therapy applicaton in utero mayallow induction of immune tolerance.However the paradigm of self/non-self immunetolerance and sensitisation has been recently challenged bythe hypothesis of Matzinger (2002). This suggests thatimmunity arises as a consequence of cellular alarm signalsfrom distressed or injured cells stimulating antigenpresenting cells. A recent study examined the idea that thefetus is particularly susceptible to induction of tolerance;the study concluded that, rather than being due toignorance, timing-based tolerance or properties of naïve Tcells in early life, tolerance induction in fetus may arisefrom differences in fetal antigen presentation; this remainsto be identified (Anderson, et al, 2001).VI. Ethical and safety issuesThere are various ethical issues in relation to in uterogene therapy that need to be addressed before such therapycould be applied clinically (Fletcher and Richter, 1996;Recombinant DNA Advisory Committee 2000). Onemajor concern is that fetal gene therapy has potentialadverse effects such as injury, infection, severe immunereactions or preterm labour on the fetus as well as on themother. Furthermore, many parents decide to terminate anaffected pregnancy, and therefore the option of in uterotreatment must be at least as safe for the mother, andshould also reliably treat the disease (Coutelle andRodeck, 2002).There is a theoretical risk that the therapeutic geneproduct or vector that is required at a certain stage duringfetal development could cause oncogenesis. In addition,insertion of vector sequences may cause developmentalaberrations to occur.While one of the aims of prenatal gene therapy is toachieve immune tolerance to the transgene and deliverysystem, vectors must be designed to be sufficientlydifferent to the wild type so that the immune systemremains able to mount an effective immune responseagainst wild-type virus infection.The problem of insertional mutagenesis as apotential risk of retroviral gene therapy has been debatedfor some years. This serious adverse event has now beenidentified in a trial of gene therapy for X-linked severecombined immunodeficiency syndrome in which CD34 +haemopoietic stem cells were transduced ex vivo with theγc gene using retroviral vectors. Two patients out offifteen treated developed acute lymphoblastic leukemia(ALL) three years after successful gene therapy treatment.Analysis of the lymphocytes showed that the transgenehad been inserted adjacent to an oncogene, LMO2, theproduct of which has been implicated in the pathogenesisof ALL (Juengst, 2003). Further work is needed to addressthis issue and to devise strategies to determine andpossibly direct integration sites.Germline transmission is another risk that raisesethical concerns. Fetal somatic gene therapy does not aimto modify the genetic content of the germ-line butinadvertent gene transfer to the germ-line could occur.Compartmentalisation of the primordial germ cells in thegonads is complete by 7 weeks of gestation in humans andit is unlikely therefore that any therapy applied after thistime would result in germ-line transduction. Examinationof germ cells after delivery of retroviral vectors (Porada etal, 1998; Tran et al, 2000) or adenoviral vectors to earlygestation fetal sheep has not shown any detectabletransmission (David et al, 2003a). Following intravascularadministration of adenoviral vectors to late gestation fetalsheep, vector DNA was detectable by PCR in the gonads,but extensive investigation by RT-PCR could not detectany gene expression. A similar risk of germlinetransduction occurs with AAV that can integrate into thegenome. No AAV sequences were detectable in thegermline tissues of fetal mice receiving injection of AAVvectors via the intraperitoneal route nor the tissues of theirprogeny (Lipshutz et al, 2001). Many of these issues arenot confined to in utero or even adult gene therapy andconcerns regarding germ-line transmission can be raised inparticular for chemotherapy and infertility treatment(Schneider and Coutelle 1999).Finally there is the concernthat fetal gene therapy research poses special challenges toinformed consent (Burger and Wilfond 2000).199

David et al: Current status and future direction of fetal gene therapyFigure 7: Na-caprate stimulation of DEAE dextran complexed adenovirus mediated β-galactosidase expression. Panel 1: Na-capratepre-treatment followed by DEAE dextran complexed virus. Widespread X-gal staining (a-c) and immunohistochemical localisation (d-f)of β-galactosidase expression in the trachea (a & d), bronchial epithelium (e) and airway epithelium (b,c & f). Panel 2: Na-caprate pretreatmentfollowed by DEAE dextran complexed virus followed by perflubron. X-gal staining (a-c) and immunohistochemicallocalisation (d-f) of β-galactosidase expression in the peripheral airways. All fetuses were injected between 102 and 116 days. Scale bar= 5mm in all cases. (Peebles D et al 2003).200

Gene Therapy and Molecular Biology Vol 7, page 201Figure 8: Durability of expression and tolerance of exogenous and expressed hFIX. Prenatal and adult mice were injected intravenouslywith adenoviral vectors expressing the hFIX gene (AdhFIX) and repeatedly rechallenged, as adults, with intraperitoneal hFIX proteinthen intravenous AdhFIX while hFIX concentrations were measured. The y axis shows blood hFIX concentrations (µg/ml) after in uteroor adult injection of AdhFIX (Phase I), repeated injection of hFIX protein to the adult mice (Phase II) and repeated injection of AdhFIXto the adult mice (Phase III). The x-axis shows the experimental time course in days. Arrows indicate injection points. Groups I and IIare mice initially injected in utero with AdhFIX at days 15 and 17 of gestation, respectively. Group III contains mice initially injectedintravenously with AdhFIX as adults. Group IV did not receive prior injection of AdhFIX. Group V received neither prior injections ofAdhFIX or hFIX protein. A line representing a therapeutic threshold of 40 ng/ml hFIX is included. Points are mean±S.D. (Waddingtonet al 2002). Reprinted with permission from the American Society of Hematology.201

David et al: Current status and future direction of fetal gene therapyThe decision to participate in a fetal gene therapy trialwould occur close to the time of prenatal diagnosis of thecondition. The parents may hear information in a highlybiased way and not consider the risk to future pregnancies.It will be important to ensure that parents are adequatelycounselled and understand these issues before agreeing totake part in any future research. The general publicremains concerned that ethical discussion about issuessuch as gene therapy, cloning and the Human GenomeProject are falling behind the technology (Brown, 2000). Itis therefore important to provide adequate informationwhich will allow the public to understand the risks andbenefits of these novel techniques and to enable aneducated involvement in the decision-making processalong with health professionals. This will also helpindividuals to give informed consent as these proceduresbecome used in clinical practice.VII. ConclusionsFetal gene therapy offers the potential forobstetricians and gene therapists not only to diagnose butalso to treat inherited genetic disease. However, for thetreatment to be acceptable, it must offer advantages overpostnatal gene therapy, be safe for both mother and fetus,and preferably avoid germ-line transmission. Currently, inutero gene therapy remains an experimental procedure.But in the future, better understanding of the developmentof genetic disease in the fetus, and improvements in vectordesign and targeting of fetal tissues should allow thistechnology to move into clinical practice.ReferencesAiuti A, Slavin S, Aker M, Ficara F, Deola S, Mortellaro A,Morecki S, Andolfi G, Tabucchi A, Carlucci F, Marinello E,Cattaneo F, Vai S, Servida P, Miniero R, Roncarolo MG,Bordignon C (2002) Correction of ADA-SCID by stem cellgene therapy combined with nonmyeloablative conditioning.Science 296, 2410-2413.Akiyama M (1998) Severe congenital ichthyosis of the neonate.Int J Dermatol 37, 722-728.Anderson CC, Carroll JM, Gallucci S, Ridge JP, Cheever AW,Matzinger P (2001) Testing time-, ignorance-, and dangerbasedmodels of tolerance. J Immunol 166, 3663-3671.Badalian SS, Chao CR, Fox HE, Timor TI (1993) Fetalbreathing-related nasal fluid flow velocity in uncomplicatedpregnancies. Am J Obstet Gynecol 169, 563-567.Badawi N, Kurinczuk JJ, Keogh JM, Alessandri LM, O'SullivanF, Burton PR, Pemberton PJ, Stanley FJ (1998) Intrapartumrisk factors for newborn encephalopathy: the WesternAustralian case-control study. BMJ 317, 1549-1553.Baldeschi C, Gache Y, Rattenholl A, Bouille P, Danos O,Ortonne JP, Bruckener-Tuderman L, Meneguzzi G (2003).Genetic correction of canine dystrophic epidermolysisbullosa mediated by retroviral vectors. Hum Mol Genet 12,1897-1905.Barcroft J, Barron (1946) Observations upon the form andrelations of the maternal and fetal vessels in the placenta ofthe sheep. Anat Rec 94, 569-595.Barker DJ, Gluckman PD, Godfrey KM, Harding JE, Owens JA,Robinson JS (1993) Fetal nutrition and cardiovasculardisease in adult life. Lancet 341, 938-941.Baumgartner TL, Baumgartner BJ, Hudon L, Moise KJ (1999)Ultrasonographically guided direct gene transfer in utero:Successful induction of β-galactosidase in a rabbit model.Am J Obstet Gynecol 181, 848-852.Benirschke K, Kaufmann P (1990) Placental types. Pathology ofthe human placenta. Springer-Verlag, New York.Bennett M, Galan H, Owens G, Dewey R, Banks R, Hobbins J,Accurso F, Schaack J (2001) In utero gene delivery byintraamniotic injection of a retroviral vector producer cellline in a nonhuman primate model. Hum Gene Ther 12,1857-1865.Bigger B, Coutelle C (2001) Perspectives on gene therapy forcystic fibrosis airway disease. Biodrugs 15, 615-634.Billingham RE, Brent L, Medawar PB (1956) Quantitativestudies on tissue transplantation immunity III Activelyacquired tolerance. Phil Trans R Soc (London) B239, 357-369.Binns R (1967) Bone marrow and lymphoid cell injection of thepig fetus resulting in transplantation tolerance or immunity,and immunoglobulin production. Nature 214, 179-180.Blaese RM, Culver KW, Miller AD, Carter CS, Fleisher T,Clerici M, Shearer G, Chang L, Chiang Y, Tolstoshev P,Greenblatt JJ, Rosenberg SA, Klein H, Berger M, MullenCA, Ramsey WJ, Muul L, Morgan RA, French Anderson W(1995) T lymphocyte-directed gene therapy for ADA- SCID:initial trial results after 4 years. Science 270, 475-480.Bordignon C, Notarangelo LD, Nobili N, Ferrari G, Casorati G,Panina P, Mazzolari E, Maggioni D, Rossi C, Servida P,Ugazio AG, Mavilio F (1995) Gene therapy in peripheralblood lymphocytes and bone marrow for ADAimmunodeficientpatients. Science 270, 470-475.Boué A, Muller F, Nezelof C, Oury JF, Duchatel F, Dumez Y,Aubry MC, Boue J (1986) Prenatal diagnosis in 200pregnancies with a 1-in-4 risk of cystic fibrosis. Hum Genet74, 288-297.Boyle MP, Enke RA, Adams RJ, Guggino WB, Zeitlin PL (2001)In utero AAV-mediated gene transfer to rabbit pulmonaryepithelium. Mol Ther 4, 115-121.Brandt ML, Moise KJJ, Eckert JW, Johnson L, Waltrip T, SaadeG, Wu Y, Finegold MJ (1997) Transuterine puncture of thefetal stomach provides access to the small bowel in therabbit. J Invest Surg 10, 41-46.Brosens JJ, Pijnenborg R, Brosens IA ( 2002) The myometrialjunctional zone spiral arteries in normal and abnormalpregnancies: a review of the literature. Am J ObstetGynecol 187, 1416-1423.Brown P (2000) Regulations not keeping up with developmentsin genetics, says poll. BMJ 321, 1369.Buckley RH, Schiff SE, Schiff RI, Markert L, Williams LW,Roberts JL, Myers LA, Ward FE (1999) Hematopoietic stemcelltransplantation for the treatment of severe combinedimmunodeficiency. New Engl J Med 340, 508-516.Buckley SMK, Waddington SN, Jezzard S, Themis M, ColledgeWH, Coutelle C (2003) Intra-amniotic application of CFTRexpressingadenovirus does not reverse cystic fibrosisphenotype in inbred Cftr-knockout mice. Mol Ther 7, S200.Burger IM, Wilfond BS (2000) Limitations of informed consentfor in utero gene transfer research: implications forinvestigators and institutional review boards. Hum GeneTher 11, 1057-1063.Burlet P, Huber C, Bertrandy S, Ludosky MA, Zwaenepoel I,Clermont O, Roume J, Delezoide AL, Cartaud J, Munnich A,Lefebvre S (1998) The distribution of SMN protein complexin human fetal tissues and its alteration in spinal muscularatrophy. Hum Mol Genet 7, 1927-1933.Case SS, Price MA, Jordan CT, Yu XJ, Wang L, Bauer G, HaasDL, Xu D, Stripecke R, Naldini L, Kohn DB, Crooks GM(1999) Stable transduction of quiescent CD34(+)CD38(-)human hematopoietic cells by HIV-1-based lentiviralvectors. Proc Natl Acad Sci 96, 2988-2993.202

Gene Therapy and Molecular Biology Vol 7, page 203Cavazzana-Calvo M, Hacein-Bey S, de Saint Basile G, Gross F,Yvon E, Nusbaum P, Selz F, Hue C, Certain S, Casanova J,Bousso P, Le Deist F, Fischer A (2000) Gene therapy ofhuman severe combined immunodeficiency (SCID)-X1disease. Science 288, 669-672.Cavazzana-Calvo M, Hacein-Bey S, Yates F, De Villartay JP, LeDeist F, Fischer A (2001) Gene therapy of severe combinedimmunodeficiencies. J Gene Med 3, 201-206.Challita PM, Kohn DB (1994) Lack of expression from aretroviral vector after transduction of murine hematopoieticstem cells is associated with methylation in vivo. Proc NatlAcad Sci 91, 2567-2571.Chamberlain JS (2002) Gene therapy of muscular dystrophy.Hum Mol Genet 11, 2355-2362.Chao H, Samulski RJ, Bellinger D, Monahan PE, Nichols T,Walsh CE (1999) Persistent expression of canine factor IX inhemophilia B canines. Gene Ther 6, 1695-1704.Chen HH, Mack LM, Kelly R, Ontell M, Kochanek S, ClemensPR (1997) Persistence in muscle of an adenoviral vector thatlacks all viral genes. Proc Natl Acad Sci 94, 1645-1650.Chen M, Kasahara N, Keene DR, Chan L, Hoeffler WK, FinlayD, Barcova M, Cannon PM, Mazurek C, Woodley DT (2002)Restoration of type VII collagen expression and function indystrophic epidermolysis bullosa. Nat Genet 32, 670-675.Cheng SH, Smith AE (2003) Gene therapy progress andprospects: gene therapy of lysosomal storage disorders. GeneTher 10, 1275-1281.Chinnaiya A, Venkat A, Dawn C, Chee WY, Choo KB, GoleLA, Meng CT (1998) Intraheptatic vein fetal blood sampling:current role in prenatal diagnosis. J Obstet Gynaecol Res24, 239-246.Christensen G, Minamisawa S, Gruber PJ, Wang Y, Chien KR(2000) High-efficiency, long-term cardiac expression offoreign genes in living mouse embryos and neonates.Circulation 101, 178-184.Cole FS, Hamvas A, Nogee LM (2003) Genetic disorders ofneonatal respiratory function. Ped Res 50, 157-162.Cosset FL, Takeuchi Y, Battini JL, Weiss RA, Collins MK(1995) High-titer packaging cells producing recombinantretroviruses resistant to human serum. J Virol 69, 7430-7436.Coutelle C, Douar A-M, Colledge WH, Froster U (1995) Thechallenge of fetal gene therapy. Nat Med 1, 864-866.Coutelle C, Rodeck C (2002) On the scientific and ethical issuesof fetal somatic gene therapy. Gene Ther 9, 670-673.Crawford TO, Pardo CA (1996) The neurobiology of childhoodspinal muscular atrophy. Neurobiol Dis 3, 97-110.Daly TM, Vogler C, Levy B, Haskins ME, Sands MS (1999)Neonatal gene transfer leads to widespread correction ofpathology in a murine model of lysosomal storage disease.Proc Natl Acad Sci 96, 2296-2300.David AL, Cook T, Waddington S, Peebles D, Nivsarkar M,Knapton H, Miah M, Dahse T, Noakes D, Schneider H,Rodeck C, Coutelle C, Themis M (2003a) Ultrasound guidedpercutaneous delivery of adenoviral vectors encoding β-galactosidase and human factor IX genes to early gestationfetal sheep in utero. Hum Gene Ther 14, 353-364.David AL, Peebles DM, Gregory L, Themis M, Cook T, CoutelleC, Rodeck C, (2003b) Percutaneous ultrasound-guidedinjection of the trachea in fetal sheep: a novel technique totarget the fetal airways. Fetal Diagn Ther 18, 385-390.Deconinck N, Ragot T, Marechal G, Perricaudet M, Gillis JM(1996) Functional protection of dystrophic mouse (mdx)muscles after adenovirus-mediated transfer of a dystrophinminigene. Proc Natl Acad Sci 93, 3570-3574.DiDonato CJ, Parks RJ, Kothary R (2003) Development of agene therapy strategy for the restoration of survival motorneuron protein expression: implications for spinal muscularatrophy therapy. Hum Gene Ther 14, 179-188.Douar A-M, Adebakin S, Themis M, Pavirani A, Cook T,Coutelle C (1997) Foetal gene delivery in mice by intraamnioticadministration of retroviral producer cells andadenovirus. Gene Ther 4, 883-890.Douar A-M, Themis M, Sandig V, Friedmann T, Coutelle C(1996) Effect of amniotic fluid on cationic lipid mediatedtransfection and viral infection. Gene Ther 3, 789-796.Dziegielewska KM, Ek J, Habgood MD, Saunders NR (2001)Development of the choroid plexus. Microsc Res Tech 52,5-20.Emery AEH (1993) Duchenne muscular dystrophy. OxfordUniversity Press.Emery AEH (2002) The muscular dystrophies. Lancet 359, 687-695.Engelhardt JF, Schlossberg H, Yankaskas JR, Dudus L (1995)Progenitor cells of the adult human airway involved insubmucosal gland development. Development 121, 2031-2046.Engelhardt JF, Yankaskas JR, Ernst SA, Yang Y, Marino CR,Boucher RC, Cohn JA, Wilson JM (1992) Submucosalglands are the predominant site of CFTR expression in thehuman bronchus. Nat Genet 2, 240-248.Engelstädter M, Buchholz CJ, Bobkova M, Steidl S, Merget-Millitzer H, Willemsen RA, Stitz J, Cichutek K (2001)Targeted gene transfer to lymphocytes using murineleukaemia virus vectors pseudotyped with spleen necrosisvirus envelope proteins. Gene Ther 8, 1202-1206.Flake AW, Roncarolo MG, Puck JM, Almeida-Porada G, EvansMI, Johnson MP, Abella EM, Harrison DD, Zanjani ED(1996) Treatment of X-linked severe combinedimmunodeficiency by in utero transplantation of paternalbone marrow. N Engl J Med 335, 1806-1810.Fletcher JC, Richter G (1996) Human fetal gene therapy: moraland ethical questions. Hum Gene Ther 7, 1605-1614.Fowden AL (1995) Nutrient requirements for normal fetalgrowth and metabolism. In: Hanson MA, Spencer JAD,Rodeck CH (Eds.), Fetus and Neonate: Physiology andClinical Applications: Growth. Cambridge University Press,Cambridge, pp. 31-56.Friedmann T (2003) Gene therapy's new era: a balance ofunequivocal benefit and unequivocal harm. Mol Ther 8, 5-7.Furie B, Limentani SA, Rosenfield CG (1994) A practical guideto the evaluation and treatment of hemophilia. Blood 84, 3-9.Gaillard D, Ruocco S, Lallemand A, Dalemans W, Hinnrasky J,Puchelle E (1994) Immunohistochemical localization ofcystic fibrosis transmembrane conductance regulator inhuman fetal airway and digestive mucosa. Ped Res 36, 137-143.Galan HL, Bennett ML, Tyson RW, Owens G, Ragnault TR,Accurso F, Robbins JC, Schaack J (2002) Inefficienttransduction of sheep in utero after intra-amniotic injectionof retroviral producer cells. Am J Obstet Gynecol 187, 469-474.Gansbacher B, European Society of Gene Therapy (2003) Reportof a second serious adverse event in a clinical trial of genetherapy for X-linked severe combined immune deficiency(X-SCID). Position of the European Society of GeneTherapy (ESGT). J Gene Med 5, 261-262.Geipel A, Berg C, Germer U, Krapp M, Kohl M, Gembruch U(2002) Mucopolysaccharidosis VII (Sly disease) as a causeof increased nuchal translucency and non-immune fetalhydrops: study of a family and technical approach to prenataldiagnosis in early and late pregnancy. Prenat Diagn 22, 493-495.203

David et al: Current status and future direction of fetal gene therapyGerdts V, Babiuk LA, van Drunen Littel-van den Hurk, GriebelPJ (2000) Fetal immunization by a DNA vaccine deliveredinto the oral cavity. Nat Med 6, 929-932.Gerdts V, Snider M, Brownlie R, Babiuk LA, Griebel PJ (2003)Oral DNA vaccination in utero induces mucosal immunityand immune memory in the neonate. J Immunol 168, 1877-1885.Gregory L, Waddington SN, Holder M, Mitrophanous K,Buckley SMK, Bigger B, Ellard FM, Walmsley LE,Lawrence L, Cook T, Al-Allaf F, Kingsman S, Coutelle C,Themis M (2003) Highly efficient in utero gene transfer andpersistent gene expression in heart and respiratory and limbmusculature of mice demonstrate applicability of EIAVlentivirus for gene therapy of Duchenne/Becker MuscularDystrophies. Hum Gene Ther submitted.Gregory LG, Harbottle RP, Lawrence L, Knapton HJ, Themis M,Coutelle C (2002) Enhancement of adenovirus-mediatedgene transfer to the airways by DEAE dextran and sodiumcaprate in vivo. Mol Ther 7, 1-8.Haake AR, Cooklis M (1997) Incomplete differentiation of fetalkeratinocytes in the skin equivalent leads to the defaultpathway of apoptosis. Exp Cell Res 231, 83-95.Hacein-Bey-Abina S, Le Deist F, Carlier F, Bouneaud C, Hue C,De Villartay JP, Thrasher AJ, Wulffraat N, Sorensen R,Dupuis-Girod S, Fischer A, Davies EG, Kuis W, Leiva L,Cavazzana-Calvo M (2002) Sustained correction of X-linkedsevere combined immunodeficiency by ex vivo gene therapy.N Engl J Med 346, 1185-1193.Harrison MR, Mychaliska GB, Albanese CT, Jennings RW,Farrell JA, Hawgood S, Sandberg P, Levine AH, Lobo E,Filly RA (1998) Correction of congenital diaphragmatichernia in utero IX: fetuses with poor prognosis (liverherniation and low lung-to-head ratio) can be saved byfetoscopic temporary tracheal occlusion. J Pediatr Surg 33,1017-1022.Hatzoglou M, Lamers W, Bosch F, Wynshaw-Boris A, ClappDW, Hanson RW (1990) Hepatic gene transfer in animalsusing retrovirus containing the promoter from the gene forphosphoenolpyruvate carboxykinase. J Biol Chem 265,17285-17293.Hatzoglou M, Moorman A, Lamers W (1995) Persistentexpression of genes transferred in the fetal rat liver viaretroviruses. Somat Cell Mol Genet 21, 265-278.Heikkilä A, Hiltunen MO, Turunen MP, Keski-Nisula L,Turunen A-M, Räsänen H, Rissanen TT, Kosma V-M,Manninen H, Heinonen S, Ylä-Herttuala S (2001)Angiographically guided utero-placental gene transfer inrabbits with adenoviruses, plasmid/liposomes andplasmid/polyethyleneimine complexes. Gene Ther 8, 784-788.Hendrickx AG, Peterson PE (1997) Perspectives on the use ofthe baboon in embryology and teratology research. HumanReprod Update 3, 575-592.Herzog RW, Yang EY, Couto LB, Hagstrom JN, Elwell D,Fields PA, Burton M, Bellinger DA, Read MS, BrinkhousDM, Podsakoff GM, Nichols TC, Kurtzman GJ, High KA(1999) Long-term correction of canine hemophilia B by genetransfer of blood coagulation factor IX mediated by adenoassociatedviral vector. Nat Med 5, 56-63.Holzgreve W, Golbus MS (1986) Prenatal diagnosis of ornithinetranscarbamylase deficiency utilizing fetal liver biopsy. AmJHuman Genet 36, 320-328.Holzinger A, Trapnell BC, Weaver TE, Whitsett JA, IwamotoHS (1995) Intraamniotic administration of an adenoviralvector for gene transfer to fetal sheep and mouse tissues. PedRes 38, 844-850.Horn HM, Tidman MJ (2002) The clinical spectrum ofdystrophic epidermolysis bullosa. Br J Dermatol 146, 267-274.Hubeau C, Puchelle E, Gaillard D (2001) Distinct pattern ofimmune cell population in the lung of human fetuses withcystic fibrosis. J Allergy Clin Immunol 108, 524-529.Inoue M, Tokusumi Y, Ban H, Kanaya T, Tokusumi T, Nagai Y,Iida A, Hawegawa M (2003) Nontransmissible virus-likeparticle formation by F-deficient Sendai virus is temperaturesensitive and reduced by mutations in M and HN proteins. JVirol 77, 3238-3246.Iwamoto HS, Trapnell BC, McConnell CJ, Daugherty C,Whitsett JA (1999) Pulmonary inflammation associated withrepeated, prenatal exposure to an E1, E3-deleted adenoviralvector in sheep. Gene Ther 6, 98-106.Jauniaux E, Gulbis B (2000) Fluid compartments of theembryonic environment. Hum Reprod Update 6, 268-278.Jauniaux E, Gulbis B, Gerloo E (1999) Free amino acids inhuman fetal liver and fluids at 12 - 17 weeks of gestation.Hum Reprod 14, 1638-1641.Jesudason EC (2002) Challenging embryological theories oncongenital diaphragmatic hernia: future therapeuticimplications for paediatric surgery. Ann R Coll Surg Engl84, 252-259.Jeyakumar M, Butters TD, Dwek RA, Platt FM (2002)Glycosphingolipid lysosomal storage diseases: therapy andpathogenesis. Neuropathol Appl Neurobiol 28, 343-357.Johnson LG, Olsen JC, Sarkadi B, Moore KL, Swanstrom R,Boucher RC (1992) Efficiency of gene transfer forrestoration of normal airway epithelial function in cysticfibrosis. Nat Genet 2, 21-25.Johnson LG, Vanhook MK, Coyne CB, Haykal-Coates N, GavettSH (2003) Safety and efficiency of modulating paracellularpermeability to enhance airway epithelial gene transfer invivo. Hum Gene Ther 14, 729-747.Johnston J, Tazelaar J, Rivera VM, Clackson T, Gao GP, WilsonJM (2003) Regulated expression of erythropoietin from anAAV vector safely improves the anemia of beta-thalassemiain a mouse model. Mol Ther 7, 493-497.Juengst ET (2003) What next for human gene therapy? BMJ326, 1410-1411.Kalache KD, Chaoui R, Marcks B, Nguyen-Dobinsky TN,Wernicke KD, Wauer R, Bollmann R (2000) Differentiationbetween human fetal breathing patterns by investigation ofbreathing-related tracheal fluid flow velocity using Dopplersonography. Prenat Diagn 20, 45-50.Kamata Y, Tanabe A, Kanaji A, Kosuga M, Fukuhara Y, Li XK,Suzuki S, Yamada M, Azuma N, Okuyama T (2003) Longtermnormalization in the central nervous system, ocularmanifestations, and skeletal deformities by a single systemicadenovirus injection into neonatal mice withmucopolysaccharidosis VII. Gene Ther 10, 406-414.Kay MA, Landen CN, Rothenberg SR, Taylor LA, Leland F,Wiehle S, Fang B, Bellinger D, Finegold M, Thompson AR,Read M, Brinkhous KM, Woo SLC (1994) In vivo hepaticgene therapy: Complete albeit transient correction of factorIX deficiency in hemophilia B dogs. Proc Natl Acad Sci 91,2353-2357.Kay MA, Manno CS, Ragni MV, Larson PJ, Couto LB,McClelland A, Glader B, Chew AJ, Tai SJ, Herzog RW,Arruda V, Johnson F, Scallan C, Skarsgard E, Flake AW,High KA (2000) Evidence for gene transfer and expressionof factor IX in haemophilia B patients treated with an AAVvector. Nat Genet 24, 257-261.Kay MA, Rothenberg SR, Landen CN, Bellinger DA, Leland F,Toman C, Finegold M, Thompson AR, Read MS, BrinkhousKM, Woo SLC (1993) In vivo gene therapy of hemophilia B:204

Gene Therapy and Molecular Biology Vol 7, page 205sustained partial correction in factor IX-deficient dogs.Science 262, 117-119.Kobinger GP, Weiner DJ, Yu QC, Wilson JM (2001) Filoviruspseudotypedlentiviral vector can efficiently and stablytransduce airway epithelia in vivo. Nat Biotechnol 19, 225-230.Kohn DB, Weiner CP, Nolta JA, Heiss IN, Lenarsky C, CrooksGM, Hanley ME, Annett G, Brooks JS, El-Khoureiy A,Lawrence K, Wells S, Moen RC, Bastian J, Williams-Herman DE, Elder M, Wara D, Bowen T, Hershfield MS,Mullen CA, Blaese RM (1995) Engraftment of genemodifiedcells from umbilical cord blood in neonates withadenosine deaminase deficiency. Nat Med 1, 1017-1023.Kubo S, Mitani K (2003) A new hybrid system capable ofefficient lentiviral vector production and stable gene transfermediated by a single helper-dependent adenoviral vector. JVirol 77, 2964-2971.Lai L, Davison BB, Veazey RS, Fisher KJ, Baskin GB (2002) Apreliminary evaluation of recombinant adeno-associatedvirus biodistribution in rhesus monkeys after intrahepaticinoculation in utero. Hum Gene Ther 13, 2027-2039.Larsen JF, Jacobsen B, Holm HH, Pedersen JF, Mantoni M(1978) Intrauterine injection of vitamin K before the deliveryduring anticoagulant therapy of the mother. Acta ObstetGynecol Scand 57, 227-230.Larson JE, Delcarpio JB, Farberman MM, Morrow SL, Cohen JC(2000a) CFTR modulates lung secretory cell proliferationand differentiation. Am J Physiol 279, L333-L341.Larson JE, Morrow SL, Delcarpio JB, Bohm RP, Ratterree MS,Blanchard JL, Cohen JC (2000b) Gene transfer into the fetalprimate: evidence for the secretion of transgene product. MolTher 2, 631-639.Larson JE, Morrow SL, Happel L, Sharp JF, Cohen JC (1997)Reversal of cystic fibrosis phenotype in mice by genetherapy in utero. Lancet 349, 619-620.Lee B, Goss J (2001) Long-term correction of urea cycledisorders. J Pediatr 138, S62-S71.Lehrman S (1999) Virus treatment questioned after gene therapydeath. Nature 401, 517-518.Li Z, Düllmann J, Schiedlmeier B, Schmidt M, von Kalle C,Meyer J, Forster M, Stocking C, Wahlers A, Frank O,Ostertag W, Kühlcke K, Eckert H-G, Fehse B, Baum C(2002) Murine leukemia induced by retroviral gene marking.Science 296, 497.Lim F-Y, Kobinger GP, Weiner DJ, Radu A, Wilson JM,Crombleholme TM (2003) Human fetal trachea-SCID mousexenografts: efficacy of vesicular stomatitis virus-Gpseudotyped lentiviral-mediated gene transfer. J PediatrSurg 38, 834-839.Lim F-Y, Martin B, Radu A, Crombleholme TM (2002) Adenoassociatedvirus (AAV)-mediated fetal gene transfer inrespiratory epithelium and submucosal gland cells in humanfetal tracheal organ culture. J Pediatr Surg 37, 1051-1057.Lipshutz GS, Flebbe-Rehwaldt L, Gaensler KML (1999a)Adenovirus-mediated gene transfer in the midgestation fetalmouse. J Pediatr Surg 84, 150-156.Lipshutz GS, Flebbe-Rehwaldt L, Gaensler KML (1999b)Adenovirus-mediated gene transfer to the peritoneum andhepatic parenchyma of fetal mice in utero. Surgery 126,171-177.Lipshutz GS, Flebbe-Rehwaldt L, Gaensler KML (2000)Reexpression following readministration of an adenoviralvector in adult mice after initial in utero adenoviraladministration. Mol Ther 2, 374-380.Lipshutz GS, Gruber CA, Cao Y, Hardy J, Contag CH, GaenslerKML (2001) In utero delivery of adeno-associated viralvectors: intraperitoneal gene transfer produces long-termexpression. Mol Ther 3, 284-292.Lipshutz GS, Sarkar R, Flebbe-Rehwaldt L, Kazazian H,Gaensler KML (1999c) Short-term correction of factor VIIIdeficiency in a murine model of hemophilia A after deliveryof adenovirus murine factor VIII in utero. Proc Natl AcadSci 96, 13324-13329.Lipshutz GS, Titre D, Brindle M, Bisconte AR, Contag CH,Gaensler KM (2003) Comparison of gene expression afterintraperitoneal delivery of AAV2 or AAV5 in utero. MolTher 8, 90-98.Ljubic A, Cvetkovic M, Sulovic V, Radunovic N, Antonovic O,Vukolic D, Popovic B, Petkovic S (1999) New technique forartificial lung maturation. Direct intramuscular fetalcorticosteroid therapy. Clin Exp Obstet Gynecol 26, 16-19.Ljung RCR (1999) Prenatal diagnosis of haemophilia.Haemophilia 5, 84-87.Lusher JM (2000) Inhibitors in young boys with haemophilia.Baillieres Best Pract Res Clin Haematol 13, 457-468.Lutzko C, Omori F, Abrams-Ogg AC, Shull R, Li L, Lau K,Ruedy C, Nanji S, Gartley C, Dobson H, Foster R, Kruth S,Dube ID (1999) Gene therapy for canine alpha-L-iduronidasedeficiency: in utero adoptive transfer of genetically correctedhematopoietic progenitors results in engraftment but notamelioration of disease. Hum Gene Ther 10, 1521-1532.Mackenzie IZ, Maclean DA (1980) Pure fetal blood from theumbilical cord obtained at fetoscopy: experience with 125consecutive cases. Am J Obstet Gynecol 138, 1214-1218.MacKenzie TC, Kobinger GP, Kootstra NA, Radu A, Sena-Esteves M, Bouchard S, Wilson JM, Verma IM, Flake AW(2002) Efficient transduction of liver and muscle after inutero injection of lentiviral vectors with differentpseudotypes. Mol Ther 6, 349-358.Maestri NE, Clissold D, Brusilow SW (1999) Neonatal onsetornithine transcarbamylase deficiency: A retrospectiveanalysis. J Pediatr 134, 268-272.Makrydimas G, Georgiou I, Kranas V, Zikopoulos K, Lolis D(1997) Prenatal diagnosis of beta-thalassaemia bycoelocentesis. Mol Hum Reprod 3, 729-731.Marshall E (2002) Gene therapy a suspect in leukemia-likedisease. Science 298, 34-35.Masaki I, Yonemitsu Y, Komori K, Ueno H, Nakashima Y,Nakagawa K, Fukumura M, Kato A, Hasan MK, Nagai Y,Sugimachi K, Hasegawa M, Sueishi K (2001) RecombinantSendai virus-mediated gene transfer to vasculature: a newclass of efficient gene transfer vector to the vascular system.FASEB J 15, 1294-1296.Mason CA, Bigras JL, O'Blenes SB, Zhou B, McIntyre B,Nakamura N, Kaneda Y, Rabinovitch M (1999) Genetransfer in utero biologically engineers a patent ductusarteriosus in lambs by arresting fibronectin-dependentneointimal formation. Nat Med 5, 176-182.Matthjis G, Devriendt K, Fryns JP (1998) The prenatal diagnosisof spinal muscular atrophy. Prenat Diagn 18, 607-610.Matzinger P (2002) The danger model: a renewed sense of self.Science 296, 301-305.May C, Rivella S, Callegari J, Heller G, Gaensler KM, LuzzattoL, Sadelain M (2000) Therapeutic haemoglobin synthesis inbeta-thalassaemic mice expressing lentivirus-encoded humanbeta-globin. Nature 406, 86.Mazarakis ND, Azzouz M, Rohll JB, Ellard FM, Wilkes FJ,Olsen AL, Carter EE, Barber RD, Baban DF, Kingsman SM,Kingsman A, O'Malley K, Mitrophanous K (2001) Rabiesvirus glycoprotein pseudotyping of lentiviral vectors enablesretrograde axonal transport and access to the nervous systemafter peripheral delivery. Hum Mol Genet 10, 2109-2121.McCray Jr PB, Stein CS, Kang Y, Sauter SL, Townsend K,Staber P, Derksen TA, Martins I, Qian J, Davidson BL,McCray PB (2001) In vivo treatment of hemophilia A and205

David et al: Current status and future direction of fetal gene therapymucopolysaccharidosis type VII using nonprimate lentiviralvectors. Mol Ther 3, 850-856.McCray PB, Armstrong K, Zabner J, Miller DW, Koretzky GA,Couture L, Robillard JE, Smith AE, Welsh MJ (1995)Adenoviral-mediated gene transfer to fetal pulmonaryepithelia in vitro and in vivo. J Clin Invest 95, 2620-2632.Meertens L, Zhao Y, Rosic-Kablar S, Li L, Chan K, Dobson H,Gartley C, Lutzko C, Hopwood J, Kohn D, Kruth S, HoughMR, Dube ID (2002) In utero injection of alpha-Liduronidase-carryingretrovirus in caninemucopolysaccharidosis type I: infection of multiple tissuesand neonatal gene expression. Hum Gene Ther 13, 1809-1820.Miller DG, Adam MA, Miller AD (1990) Gene transfer byretrovirus vectors occurs only in cells that are activelyreplicating at the time of infection. Mol Cell Biol 10, 4239-4242.Mitchell M, Jerebtsova M, Batshaw ML, Newman K, Ye X(2000) Long-term gene transfer to mouse fetuses withrecombinant adenovirus and adeno-associated virus (AAV)vectors. Gene Ther 7, 1986-1992.Mitrophanous K, Yoon S, Rohll J, Patil D, Wilkes F, Kim V,Kingsman S, Kingsman A, Mazarakis N (1999) Stable genetransfer to the nervous system using a non-primate lentiviralvector. Gene Ther 6, 1808-1818.Miyagawa Y, Matsuoka T, Baba A, Nakamura T, Tsuno T,Tamura A, Agematsu K, Yabuhara A, Uehara Y, Kawai H(1992) Fetal liver T cell receptor gamma/delta+ T cells ascytotoxic T lymphocytes specific for maternal alloantigens. JExp Med 176, 1-7.Moise KJ, Hesketh DE, Belfort MM, Saade G, Van den VeyerIB, Hudson KM, Rodkey LS (1992) Ultrasound-guidedblood sampling of rabbit fetuses. Lab Anim Sci 42, 398-401.Monahan PE, Samulski RJ (2000) Adeno-associated virusvectors for gene therapy: more pros than cons? Mol MedToday 6, 433-440.Moss IR, Scarpelli EM (1981) Stimulatory effect of theophyllineon regulation of fetal breathing movements. Ped Res 15,870-873.Muench MO, Rai J, Barcena A, Leemhuis T, Farrell J, HumeauL, Maxwell-Wiggins JR, Capper J, Mychaliska GB,Albanese CT, Martin T, Tsukamoto A, Curnutte JT, HarrisonMR (2001) Transplantation of a fetus with paternal Thy-1(+)CD34(+) cells for chronic granulomatous disease. BoneMarrow Transplant 27, 355-364.Murphy SJ, Chong H, Bell S, Diaz RM, Vile RG (2002) Novelintegrating adenoviral/retroviral hybrid vector for genetherapy. Hum Gene Ther 13, 745-760.Nakai H, Montini E, Fuess S, Storm TA, Grompe M, Kay MA(2003) AAV serotype 2 vectors preferentially integrate intoactive genes in mice. Nat Genet 34, 297-302.Naldini L, Blomer U, Gallay P, Ory D, Mulligan R, Gage FH,Verma IM, Trono D (1996) In vivo gene delivery and stabletransduction of nondividing cells by a lentiviral vector.Science 272, 263-267.Newnham JP, Kelly RW (1993) Ultrasound for research withfetal sheep. In: Neilson JP, Chambers SE (Eds.), ObstetricUltrasound I. Oxford Medical Publications, pp. 203-222.Nicolini U, Nicolaidis P, Fisk NM, Tannirandorn Y, Rodeck C(1990) Fetal blood sampling from the intrahepatic vein:analysis of safety and clinical experience with 214procedures. Obstet Gynaecol 76, 47-53.Olivares EC, Hollis RP, Chalberg TW, Meuse L, Kay MA, CalosMP (2002) Site specific genomic integration producestherapeutic Factor IX levels in mice. Nat Biotechnol 20,1124-1128.Olivieri NF (1999) The beta-thalassaemias. N Engl J Med 341,99-109.Olivieri NF, Weatherall DJ (1998) The therapeutic reactivationof fetal haemoglobin. Hum Mol Genet 7, 1655-1658.Ortiz-Urda S, Thyagarajan B, Keene DR, Lin Q, Calos MP,Khavari PA (2003) _C31 Integrase-Mediated NonviralGenetic Correction of Junctional Epidermolysis Bullosa.Hum Gene Ther 14, 923-928.Pahal GS, Jauniaux E, Kinnon C, Thrasher AJ, Rodeck C (2000)Normal development of human fetal hematopoiesis betweeneight and seventeen weeks' gestation. Am J Obstet Gynecol183, 1029-1034.Pai S, Marinkovich MP (2002) Epidermolysis bullosa: new andemerging trends. Am J Clin Dermatol 3, 371-380.Palmer TD, Rosman GJ, Osborne WRA, Miller D (1991)Genetically modified skin fibroblasts persist long aftertransplantation but gradually inactivate introduced genes.Proc Natl Acad Sci 88, 1330-1334.Papadopulos NA, Dumitrascu I, Ordonez JL, Decaluwe H, LerutTE, Barki G, Deprest JA (1999) Fetoscopy in the pregnantrabbit at midgestation. Fetal Diagn Ther 14, 118-121.Pardi G, Marconi AM, Cetin I (2002) Placental-fetalinterrelationship in IUGR fetuses--a review. Placenta 23suppl A, S136-S141.Parsons DW, Grubb BR, Johnson LG, Boucher RC (1998)Enhanced in vivo airway gene transfer via transientmodification of host barrier properties with a surface-activeagent. Hum Gene Ther 9, 2661-2672.Pawliuk R, Westerman KA, Fabry ME, Payen E, Tighe R,Bouhassira EE, Acharya SA, Ellis J, London IM, Eaves CJ,Humphries RK, Beuzard Y, Nagel RL, Leboulch P (2001)Correction of sickle cell disease in transgenic mouse modelsby gene therapy. Science 294, 2368-2371.Peault B, Tirouvanziam R, Sombardier MN, Chen S, PerricaudetM, Gaillard D (1994) Gene transfer to human fetalpulmonary tissue developed in immunodeficient SCID mice.Hum Gene Ther 5, 1131-1137.Peebles D, Gregory LG, David A, Themis M, Waddington S,Knapton HJ, Miah M, Cook T, Lawrence L, Nivsarkar M,Rodeck C, Coutelle C (2003) Widespread and efficientmarker gene expression in the airway epithelia of fetal sheepafter minimally invasive tracheal application of recombinantadenovirus in utero. Gene Ther in press.Peebles DM, Wyatt JS (2002) Synergy between antenatalexposure to infection and intrapartum events in causation ofperinatal brain injury at term. BJOG 109, 737-739.Pfendner EG, Nakano A, Pulkkinen L, Christiano AM, Uitto J(2003) Prenatal diagnosis for epidermolysis bullosa: a studyof 144 consecutive pregnancies at risk. Prenat Diagn 23,447-456.Phillips JH, Hori T, Nagler A, Bhat N, Spits H, Lanier LL (1992)Ontogeny of human natural killer (NK) cells: fetal NK cellsmediate cytolytic function and express cytoplasmic CD3epsilon, delta proteins. J Exp Med 175, 1055-1066.Piedrahita JA (2000) Targeted modification of the domesticanimal genome. Theriogeniology 53, 105-116.Pitt BR, Schwarz MA, Pilewski JM, Nakayama D, Mueller GM,Robbins PD, Watkins SA, Albertine KH, Bland RD (1995)Retrovirus-mediated gene transfer in lungs of living fetalsheep. Gene Ther 2, 344-350.Plopper CG, Weir AJ, Nishio SJ, Cranz DL, St George JA (1986)Tracheal submucosal gland development in the rhesusmonkey, Macaca mulatta: ultrastructure and histochemistry.Anat Embryol 174, 167-178.Ponder KP, Melniczek JR, Xu L, Weil MA, O'Malley TM,O'Donnell PA, Knox VW, Aguirre GD, Mazrier H,Ellinwood NM, Sleeper M, Maguire AM, Volk SW, MangoRL, Zweigle J, Wolfe JH, Haskins ME (2002) Therapeuticneonatal hepatic gene therapy in mucopolysaccharidosis VIIdogs. Proc Natl Acad Sci 99, 13102-13107.206

Gene Therapy and Molecular Biology Vol 7, page 207Porada CD, Tran N, Eglitis M, Moen RC, Troutman L, FlakeAW, Zhao Y, Anderson WF, Zanjani ED (1998) In uterogene therapy: transfer and long-term expression of thebacterial neo r gene in sheep after direct injection of retroviralvectors into preimmune fetuses. Hum Gene Ther 9, 1571-1585.Preece MA, Green A (2002) Pregnancy and inherited metabolicdisorders: maternal and fetal complications. Ann ClinBiochem 39, 444-455.Quintero RA, Reich H, Puder KS, Bardicef M, Evans MI,CottonDB, Romero R (1994) Umbilical-Cord Ligation of anAcardiac Twin by Fetoscopy at 19 Weeks of Gestation. NEngl J Med 330, 469-471.Raftopoulos H, Ward M, Leboulch P, Bank A (1997) Long-termtransfer and expression of the human _-globin gene in amouse transplant model. Blood 90, 3414-3422.Raper SE, Yudkoff M, Chirmule N, Gao GP, Nunes F, HaskalZJ, Furth EE, Propert KJ, Robinson MB, Magosin S, SimoesH, Speicher L, Hughes J, Tazelaar J, Wivel NA, Wilson JM,Batshaw ML (2002) A pilot study of in vivo liver-directedgene transfer with an adenoviral vector in partial ornithinetranscarbamylase deficiency. Hum Gene Ther 13, 163-175.Recombinant DNA Advisory Committee (2000) Prenatal genetransfer; Scientific, medical, and ethical issues. Hum GeneTher 11, 1211-1229.Riley RL (1998) Neonatal immune response. In: Delves PJ, RoittIM (Eds.), Encyclopedia of Immunology. Academic Press,London, pp. 1818-1821.Rodeck C (1980) Fetoscopy guided by real-time ultrasound forpure fetal blood samples, fetal skin samples, and examinationof the fetus in utero. BJOG 87, 449-456.Rodeck CH, Deans A (1999) Red cell alloimmunisation. In:Rodeck CH, Whittle MJ (Eds.), Fetal medicine: basicscience and clinical practice. Churchill Livingstone,London, pp. 785-804.Ross J, Jurkovic D, Nicolaides KH (1997) Coelocentesis: a studyof short-term safety. Prenat Diagn 17, 913-917.Russel DW, Berger MS, Miller AD (1995) The effects of humanserum and cerebrospinal fluid on retroviral vectors andpackaging cell lines. Hum Gene Ther 6, 635-641.Sabatino DE, MacKenzie TC, Campagnoli C, Liu Y-L, High KA,Flake AW (2002) Long-term expression of factor IX after inutero administration of adeno-associated viral vectors inhemophilia B mice. Blood 100 (11), 3421Sadelain M (2002) Globin gene transfer for the treatment ofsevere hemoglobinopathies: a paradigm for stem cell-basedgene therapy. J Gene Med 4, 113-121.Santolaya-Forgas J, Vengalil S, Kushwaha A, Bieniarz A,Fortman J (1998) Assessment of the risk of fetal loss afterthe coelocentesis procedure using a baboon model. FetalDiagn Ther 13, 257-260.Sarzotti M, Robbins DS, Hoffman PM (1996) Induction ofprotective CTL responses in newborn mice by a murineretrovirus. Science 271, 1726-1728.Schachtner SK, Buck CA, Bergelson JM, Baldwin HS (1996)Temporally regulated expression patterns following in uteroadenovirus-mediated gene transfer. Gene Ther 6, 1249-1257.Schiedner G, Morral N, Parks RJ, Wu Y, Koopmans SC,Langston C, Graham FL, Beaudet AL, Kochanek S (1998)Genomic DNA transfer with a high-capacity adenovirusvector results in improved in vivo gene expression anddecreased toxicity. Nat Genet 18, 180-183.Schneider H, Adebakin S, Themis M, Cook T, Douar A-M,Pavirani A, Coutelle C (1999) Therapeutic plasmaconcentrations of human factor IX in mice after genedelivery into the amniotic cavity: a model for the prenataltreatment of haemophilia B. J Gene Med 1, 424-432.Schneider H, Coutelle C (1999) In utero gene therapy: The casefor. Nat Med 5, 256-257.Schneider H, Mühle C, Douar A-M, Waddington S, Jiang Q-J,von der Mark K, Coutelle C, Rascher W (2002) Sustaineddelivery of therapeutic concentrations of human clottingfactor IX - a comparison of adenoviral and AAV vectorsadministered in utero. J Gene Med 4, 46-53.Sekhon HS, Larson JE (1995) In utero gene transfer into thepulmonary epithelium. Nat Med 1, 1201-1203.Senoo M, Matsubara Y, Fujii K, Nagasaki Y, Hiratsuka M, KureS, Uehara S, Okamura K, Yajima A, Narisawa K (2000)Adenovirus-mediated in utero gene transfer in mice andguinea pigs: tissue distribution of recombinant adenovirusdetermined by quantitative TaqMan-polymerase chainreaction assay. Mol Genet Metab 69, 269-276.Shiotani A, Fukumura M, Maeda M, Hou X, Inoue M, KanamoriT, Komaba S, Washizawa K, Fujikawa S, Yamamoto T,Kadono C, Watabe K, Fukuda H, Saito K, Sakai Y, Nagai Y,Kanzaki J, Hawegawa M (2001) Skeletal muscleregeneration after insulin-like growth factor 1 gene transferby recombinant Sendai virus vector. Gene Ther 8, 1043-1050.Silverstein AM, Prendergast RA, Kraner KL (1964) Fetalresponse to antigenic stimulus IV. Rejection of skinhomografts by the fetal lamb. J Exp Med 119, 955-964.Silverstein AM, Uhr JW, Lukes RJ (1963) Fetal response toantigenic stimulus II. Antibody production by the fetal lamb.J Exp Med 117, 799-812.Simonds AK, Ward S, Heather S, Bush A, Muntoni F (2000)Outcome of paediatric domiciliary mask ventilation inneuromuscular and skeletal disease. Eur Respir J 16, 476-481.Snyder RO, Miao C, Meuse L, Tubb J, Donahue BA, Lin H-F,Stafford DW, Patel S, Thompson AR, Nichols T, Read MS,Brinkhous DM, Kay MA (1999) Correction of hemophilia Bin canine and murine models using recombinant adenoassociatedviral vectors. Nat Med 5, 64-70.Soucie JM, Nuss R, Evatt B, Abdelhak A, Cowan L, Jill H,Kolakoski M, Wilber N (2000) Mortality among males withhemophilia: relations with source of medical care. TheHemophilia Surveillance System Project Investigators. Blood96, 437-442.Stecenko AA, Brigham KL (2003) Gene therapy progress andprospects: alpha-1 antitrypsin. Gene Ther 10, 95-99.Stratford-Perricaudet L, Levrero M, Chasse J-F, Perricaudet M,Briand P (1990) Evaluation of the transfer and expression inmice of an enzyme encoding gene using a human adenovirusvector. Hum Gene Ther 1, 241-256.Summar M, Tuchman M (2001) Proceedings of a consensusconference for the management of patients with urea cycledisorders. J Pediatr 138, S6-S10.Sun L, Li J, Xiao X (2000) Overcoming adeno-associated virusvector size limitation through viral DNA heterodimerization.Nat Med 6, 599-602.Sylvester KG, Yang EY, Cass DL, Crombleholme TM, ScottAdzick N (1997) Fetoscopic gene therapy for congenital lungdisease. J Pediatr Surg 7, 964-969.Tan LK, Kumar S, Jolly M, Gleeson C, Johnson P, Fisk N (2003)Test amnioinfusion to determine suitability for serialtherapeutic amnioinfusion in midtrimester premature ruptureof membranes. Fetal Diagn Ther 18, 183-189.Tarantal AF (1990) Interventional ultrasound in pregnantmacaques: embryonic/fetal applications. J Med Primatol 19,47-58.Tarantal AF, Lee CI, Ekert JE, McDonald R, Kohn DB, PlopperCG, Case SS, Bunnell BA (2001a) Lentiviral vector genetransfer into fetal rhesus monkeys (Macaca mulatta): lungtargetingapproaches. Mol Ther 4, 614-621.207

David et al: Current status and future direction of fetal gene therapyTarantal AF, O'Rourke JP, Case SS, Newbound GC, Li J, Lee CI,Baskin CR, Kohn DB, Bunnell BA (2001b) Rhesus monkeymodel for fetal gene transfer: studies with retroviral-basedvector systems. Mol Ther 3, 128-138.Themis M, Forbes SJ, Chan L, Cooper RG, Etheridge CJ, MillerAD, Hodgson HJ, Coutelle C (1998) Enhanced in vitro andin vivo gene delivery using cationic agent complexedretrovirus vectors. Gene Ther 5, 1180-1186.Themis M, Schneider H, Kiserud T, Cook T, Adebakin S,Jezzard S, Forbes S, Hanson M, Pavirani A, Rodeck C,Coutelle C (1999) Successful expression of β-galactosidaseand factor IX transgenes in fetal and neonatal sheep afterultrasound-guided percutaneous adenovirus vectoradministration into the umbilical vein. Gene Ther 6, 1239-1248.Tizzano EF, O'Brodovich H, Chitayat D, Benichou JC,Buchwald M (1994) Regional expression of CFTR indeveloping human respiratory tissues. Am J Respir CellMol Biol 10, 355-362.Touraine JL (1992) In-utero transplantation of fetal liver stemcells into human fetuses. Hum Reprod 7, 44-48.Touraine JL (1999) Induction of transplantation tolerance inhumans using stem cell transplants prenatally or postnatally.Transplant Proc 31, 2735-2737.Tran ND, Porada CD, Almeida-Porada G, Glimp HA, AndersonWF, Zanjani ED (2001) Induction of stable prenataltolerance to β-galactosidase by in utero gene transfer intopreimmune sheep fetuses. Blood 97, 3417-3423.Tran ND, Porada CD, Zhao Y, Almeida-Porada G, AndersonWF, Zanjani ED (2000) In utero transfer and expression ofexogenous genes in sheep. Exp Hematol 28, 17-30.Trono D (2000) Lentiviral vectors: turning a deadly foe into atherapeutic agent. Gene Ther 7, 20-23.Turkel SB, Howell R, Iseri AL, Chui L (1981) Ultrastructure ofmuscle in fetal Duchenne's dystrophy. Arch Pathol LabMed 105, 414-418.Türkay A, Saunders TL, Kurachi K (1999) Intrauterine genetransfer: gestational stage-specific gene delivery in mice.Gene Ther 6, 1685-1694.Uitto J, Pulkkinen L (2000) The genodermatoses: candidatediseases for gene therapy. Hum Gene Ther 11, 2267-2275.Vassilopoulos D, Emery AEH (1977) Muscle nuclear changes infetuses at risk for Duchenne muscular dystrophy. J MedGenet 14, 13-15.Ville Y, Van Peborgh P, Gagnon A, Frydman R, Fernandez H(1997) Surgical treatment of twin-to-twin transfusionsyndrome: coagulation of anastomoses with a Nd:YAG laser,under endosonographic control. Forty four cases. J GynecolObstet Biol Reprod (Paris) 26, 175-181.Vincent MC, Trapnell BC, Baughman RP, Wert SE, Whitsett JA,Iwamoto HS (1995) Adenovirus-mediated gene transfer tothe respiratory tract of fetal sheep in utero. Hum Gene Ther6, 1019-1028.Vincent N, Ragot T, Gilgenkrantz H (1993) Long-termcorrection of mouse dystrophic degeneration by adenovirusmediatedtransfer of a minidystrophin gene. Nat Genet 5,130-134.Waddington SN, Buckley SMK, Nivsarkar M, Jezzard S,Schneider H, Tucker N, Dallman MJ, Themis M, Coutelle C(2002) In utero gene transfer of human factor IX to fetalmice can induce postnatal tolerance of the exogenous clottingfactor. Blood 101, 1359-1366.Waddington SN, Mitrophanous KA, Ellard F, Buckley SMK,Nivsarkar M, Lawrence L, Cook HT, Al-Allaf F, Bigger B,Kingsman S, Coutelle C, Themis M (2003) Long-termtransgene expression by administration of a lentivirus-basedvector to the fetal circulation of immuno-competent mice.Gene Ther 10, 1234-1240.Walker JJ (2000) Pre-eclampsia. Lancet 356, 1260-1265.Wang B, Li J, Xiao X (2000) Adeno-associated virus vectorcarrying human minidystrophin genes effectively amelioratesmuscular dystrophy in mdx mouse model. Proc Natl AcadSci 97, 13714-13719.Wang G, Slepushkin V, Zabner J, Keshavjee S, Johnston JC,Sauter SL, Jolly DJ, Dubensky TW, Davidson BL, McCrayPB (1999) Feline immunodeficiency virus vectorspersistently transduce nondividing airway epithelia andcorrect the cystic fibrosis defect. J Clin Invest 104, R55-R62.Wang G, Williamson R, Mueller G, Thomas P, Davidson BL,McCray Jr PB (1998) Ultrasound-guided gene therapy tohepatocytes in utero. Fetal Diagn Ther 13, 197-205.Wang G, Zabner J, Deering C, Launspach J, Shao J, Bodner M,Jolly DJ, Davidson BL, McCray Jr PB (2000) Increasingepithelial junction permeability enhances gene transfer toairway epithelia in vivo. Am J Respir Cell Mol Biol 22,129-138.Wang L, Takabe K, Bidlingmaier SM, Ill CR, Verma IM (1999)Sustained correction of bleeding disorder in hemophilia Bmice by gene therapy. Proc Natl Acad Sci 96.Watts AM, Stanley JR, Shearer MH, Hefty PS, Kennedy RC(1999) Fetal immunization of baboons induce a fetal-specificantibody response. Nat Med 5, 427-430.Weatherall DJ, Clegg JB (1996) Thalassaemia - a global publichealth problem. Nat Med 2, 847-849.Weiss DJ, Bonneau L, Liggitt D (2001) Use ofperfluorochemical liquid allows earlier detection of geneexpression and use of less vector in normal lung andenhances gene expression in acutely injured lung. Mol Ther3, 734-745.Weiss DJ, Strandflord TP, Liggitt D, Clark JG (1999a)Perflubron enhances adenovirus-mediated gene expression inlungs of transgenic mice with chronic alveolar filling. HumGene Ther 10, 2287-2293.Weiss DJ, Strandjord TP, Jackson JC, Clark JG, Liggitt D(1999b) Perfluorochemical liquid-enhanced adenoviralvector distribution and expression in lungs of spontaneouslybreathing rodents. Exp Lung Res 25, 317-333.Wells DJ, Wells KE (2000) Gene transfer studies in animals:what do they really tell us about the prospects for genetherapy in DMD? Neuromuscul Disord 12 supplement 1,S11-S22.Welsh RJ, Cooper NR, Jensen FC, Oldstone MB (1975) Humanserum lyses RNA tumor viruses. Nature 257, 612-614.Wengler GS, Lanfranchi A, Frusca T (1996) In-uterotransplantation of parental CD34 haematopoietic progenitorcells in a patient with X-linked severe immunodeficiency(SCIDXI). Lancet 348, 1484-1487.Westgren M, Ek S, Bui T, Jansson B, Kjaeldgaard A, MarklingL, Nennesmo I, Seiger A, Sarby B, Thornström S, Ringden O(1997) Tissue distribution of transplanted fetal liver cells inthe human fetal recipient. Am J Obstet Gynecol 176, 49-53.Westgren M, Ringden O, Bartmann P, Bui T, Lindton B, MattsonJ, Uzunel M, Zetterquist H, Hansmann M (2002) Prenatal T-cell reconstitution after in utero transplantation with fetalliver cells in a patient with X-linked severe combinedimmunodeficiency. Am J Obstet Gynecol 187, 475-482.Wilson JM, Wivel NA (1999) Report on the potential use of genetherapy in utero. Gene Therapy Advisory Committee. HumGene Ther 10, 689-692.Woo YJ, Raju GP, Swain JL, Richmond ME, Gardner TJ,Balice-Gordon RJ (1997) In utero cardiac gene transfer viaintraplacental delivery of recombinant adenovirus.Circulation 96, 3561-3569.Wraith JE (2002) Lysosomal disorders. Semin Neonatol 7, 75-83.208

Gene Therapy and Molecular Biology Vol 7, page 209Wu Y, Liu J, Woo S, Finegold MJ, Brandt ML (1999) Prenatalorogastric gene delivery results in transduction of the smallbowel in the fetal rabbit. Fetal Diagn Ther 14, 323-327.Xiao X, Li J, Samulski RJ (1998) Production of high-titerrecombinant adeno-associated virus vectors in the absence ofhelper adenovirus. J Virol 72, 2224-2232.Xing A, Boileau P, Caüzac M, Challier J-C, Girard J, Hauguelde-MouzonS (2000) Comparative in vivo approaches forselective adenovirus-mediated gene delivery to the placenta.Hum Gene Ther 11, 167-177.Xu L, Gao C, Sands MS, Cai SR, Nichols TC, Bellinger DA,Raymer RA, McCorquodale S, Ponder KP (2003) Neonatalor hepatocyte growth factor-potentiated adult gene therapywith a retroviral vector results in therapeutic levels of caninefactor IX for hemophilia B. Blood 101, 3924-3932.Yang EY, Cass DL, Sylvester KG, Wilson JM, Adzick NS(1999) Fetal gene therapy: efficacy, toxicity, andimmunologic effects of early gestation recombinantadenovirus. J Pediatr Surg 34, 235-241.Yang EY, Kim HB, Shaaban AF, Milner R, Adzick NS, FlakeAF (1999) Persistent postnatal transgene expression in bothmuscle and liver after fetal injection of recombinantadenovirus. J Pediatr Surg 34, 766-772Ye X, Robinson MB, Batshaw ML, Furth EE, Smith I, WilsonJM (1996) Prolonged metabolic correction in adult ornithinetranscarbamylase-deficient mice with adenoviral vectors. JBiol Chem 271, 3639-3646.Yonemitsu Y, Kitson C, Ferrari S, Farley R, Griesenbach U,Judd D, Steel R, Scheid P, Zhu J, Jeffery PK, Kato A, HasanMK, Nagai Y, Masaki I, Fukumura M, Hasegawa M, GeddesDM, Alton EW (2000) Efficient gene transfer to airwayepithelium using recombinant Sendai virus. Nat Biotechnol18, 970-973.Dr. Anna David209

David et al: Current status and future direction of fetal gene therapy210

Gene Therapy and Molecular Biology Vol 7, page 211Gene Ther Mol Biol Vol 7, 211-219, 2003The role of EBV and genomic sequences in geneexpression from extrachromosomal gene therapyvectors in mouse liverResearch ArticleStephanie M. Stoll 1 , Leonard Meuse 2§ , Mark A. Kay 1,2 , and Michele P. Calos* 1Departments of 1 Genetics and 2 Pediatrics, Stanford University School of Medicine, Stanford, CA 94305-5120__________________________________________________________________________________*Correspondence: Michele P. Calos, phone 650-723-5558, fax 650-725-1534, e-mail calos@stanford.edu§ Present address: University of Washington, Department of Neurology, Box 357720, Seattle, WA 98195-7720.Key words: Epstein-Barr virus (EBV); extrachromosomal gene therapy, SERPINA1 sequence, α1-antitrypsin (AAT)Received: 18 September 2003; Accepted: 29 October 2003; electronically published: November 2003SummaryA plasmid vector containing Epstein-Barr virus (EBV) sequences and the full genomic SERPINA1 locus encodingthe gene for α 1 -antitrypsin is capable of providing long-term, high-level expression when transfected into mouseliver. It was unclear which viral and genomic sequences were required for efficient expression of this transgene invivo. We tested here the requirement for EBV sequences for retention and expression of plasmid DNA in normaland replicating liver in vivo. The results showed that EBV sequences provided increased retention and expression ofan extrachromosomal vector containing the full SERPINA1 transgene, in addition to the expression provided by thefull gene alone. We also minimized the SERPINA1 sequence and determined which portions were necessary forpersistent, high expression levels. Finally, we demonstrated that the SERPINA1 sequence can act to enhanceexpression of a heterologous gene cloned within it. Expression from a factor IX minigene was increased ~50-foldwhen it was expressed from within the SERPINA1 sequence, compared to a vector containing the factor IXminigene alone. The results presented here demonstrate that a significant amount of genomic sequence may berequired for persistent, high levels of expression in vivo and that the persistence of plasmid DNA in dividing tissuesand expression levels are enhanced by inclusion of EBV sequences on the vector.I. IntroductionThe ability to achieve persistent, regulated, highlevels of transgene expression in vivo is often necessaryfor the success of a gene therapy vector. Unfortunately,with most gene therapy vectors used to date, expression istemporary, often falling to non-therapeutic or undetectablelevels within a few weeks after treatment. For viralvectors, transience may be due to the immunogenicity ofthe vector, resulting in loss of transfected cells with aconcurrent reduction in transgene expression. In the caseof non-integrating vectors, viral or non-viral, transiencecan result from vector loss as the cells divide.For both integrating and non-integrating systems,decreased transgene expression may also be attributable toDNA silencing. For example, when mouse hepatocyteswere transfected in vivo with naked plasmid DNAencoding the AAT cDNA under control of thecytomegalovirus (CMV) promoter, day 1 expression levelsof 500 µg/ml were observed. These levels fell to 300 µg/ml in vivo that persisted atthese high levels for > 9 months. However, similarconstructs carrying the AAT cDNA driven by the RSVpromoter gave equivalent day 1 expression levels, but theexpression dropped >100-fold within two weeks. Again,Southern analysis showed that plasmid DNA wasmaintained extrachromosomally in these relativelyquiescent liver cells. In addition to the SERPINA1 locus,the successful genomic AAT vector also possessed211

Stoll et al: The role of EBV and genomic sequences in gene expressionsequences from Epstein-Barr virus (EBV) that can aid inextrachromosomal plasmid maintenance and expression.Epstein-Barr virus (EBV) is a human herpes virusthat is capable of maintaining its genomeextrachromosomally in dividing primate cells.Maintenance is accomplished by the viral latent origin ofreplication, oriP, and the EBV nuclear antigen 1, EBNA1,which act together to replicate the viral genome and retainit in the nucleus (Yates et al, 1984, 1985; Reisman et al,1985). Plasmids containing EBNA1 and a truncated oriPcarrying only the tandem array of 21 EBNA1 binding sites(family of repeats) from oriP for retention, but lacking theoriP dyad symmetry element for replication, are retainedin the nucleus of the cells, but can replicate efficientlyonly if the plasmid also contains a functional mammalianorigin of replication, such as the 19 kb SERPINA1sequence ( Krysan et al, 1989; Heinzel et al, 1991; Stoll etal, 2001). These same EBV components that providereplication and retention functions are also associated withtranscriptional enhancer and anti-silencing activity(Reisman and Sugden, 1986; Kaneda et al, 2000).Furthermore, in addition to the replication function of thegenomic SERPINA1 sequence demonstrated in ourprevious experiments (Stoll et al, 2001), the full AAT genewas also able to provide more stable expression in vivothan its equivalent cDNA sequence, which may be subjectto silencing.Silencing of cDNA vectors may occur because thetransgenes are often driven by viral promoters. It has beenobserved that many common viral promoters, such asthose from cytomegalovirus (CMV), simian virus 40(SV40), and Rous sarcoma virus (RSV) often exhibitmarkedly decreased activity in mammalian cells in vivowithin a few weeks of transfection, a phenomenon that hasbeen attributed to inhibition by various cytokines (Paillard,1997). Gill (2001) recently demonstrated that the use ofthe cellular elongation factor 1α (EF1α) and ubiquitin C(UbC) mammalian promoters gave increased persistenceand ~10-fold higher expression levels of a luciferasereporter gene in lungs, compared to a control constructthat expressed luciferase from the CMV promoter.Quantitative PCR analysis of plasmid vector in the lungtissue revealed that there were no significant differences inplasmid copy number in the CMV versus EF1α or UbCpromoter vectors.In addition to the reduced transgene silencingobserved when mammalian promoters are used, genomicsequences may provide additional benefits that lead toincreased transgene expression. Studies in transgenic micehave indicated that introns are essential for stable, highlevels of transgene expression. In comparing transgenicmice generated with cDNA constructs versus full genomicsequences, the intronless constructs resulted in a lowerfrequency of transgenic mice expressing rat growthhormone (rGH), mouse metallothionein I (mMTI), orhuman β-globin (hBG) reporter genes, as well asdecreased expression levels in those mice that did haveobservable expression (Brinster et al, 1988). Similarresults have been observed for AAT and β-lactoglobulinexpression constructs in mammary cells of transgenic mice(Whitelaw et al, 1991). It is possible that genomic intronscontain transcriptional enhancer sequences that may act ontheir own to increase transgene expression or may act inconcert with upstream/promoter sequences. These intronicsequences may also act to help the transgene attain anopen chromatin configuration, making it more accessibleto transcription factors. This idea is supported byobservations that deletion of intronic sequences makestransgenes more susceptible to chromosomal positioneffects in vivo than their full genomic counterparts(Webster et al, 1997).Unfortunately, the large size of most full genes oftenprecludes their use in vectors. In order to obtain theexpression advantages of intronic sequences, while stillminimizing transgene size, heterologous introns andgenomic minigenes have been developed. Palmiter (1991)found that including only select introns, specifically thefirst one, in the rGH gene resulted in transgenicfrequencies and expression levels comparable to thoseachieved when the full rGH was used. Heterologousintrons inserted between promoter and cDNA gave similarresults (Palmiter et al, 1991). This strategy has beenapplied to the construction of expression vectors fortherapeutically relevant genes. Miao (2000) constructed ahuman factor IX minigene, which included the ApoEhepatic locus control region (HCR), the hepatocytespecificAAT promoter, and the human factor IX cDNA,with its intron A and 3' untranslated region (UTR). This6.1 kb minigene was reported to provide therapeutic serumlevels of factor IX (0.5 - 2 µg/ml) that were sustained forat least 225 days, whereas a 2.0 kb AAT promoter-factorIX cDNA construct gave transient expression that fell to

Gene Therapy and Molecular Biology Vol 7, page 213heterologous gene, by inserting a human factor IXminigene (Miao et al, 2000) upstream of the first exon inSERPINA1. Here we report the successful reduction in sizeof the AAT gene, without loss of expression, as well as thesuccessful stabilization of factor IX expression in vivowhen expressed from within the AAT gene.II. Materials and methodsA. Vector constructionPlasmids pEF, pEF-AAT, pEF-cAAT, and pcAAT havebeen described previously (Stoll et al, 2001). Removal of theEBV components (EBNA1 and family of repeats) from plasmidpEF-AAT was achieved by digestion with ClaI and vectorreligation, creating the plasmid pAAT.Vectors pcAAT2 and pEFcAAT2 are identical to pcAATand pEF-cAAT, respectively, except that the RSV promoter hasbeen replaced with the endogenous, hepatocyte-specific AATpromoter. 471 bp containing the AAT promoter was PCRamplified from vector pF9 (Sclimenti et al, 2003) and cloned intothe vector pCR2.1 (Invitrogen, Carlsbad, CA), making the vectorpTA-AATpro. From there it was subcloned as an MluI-HinDIIIfragment into the MluI-HinDIII sites in the plasmid pcAAT,replacing the RSV promoter. This step created the vectorpcAAT2. From that plasmid, the 2.1 kb AATpro-cDNA fragmentwas liberated as an XhoI fragment and cloned into the SalI site inplasmid pEF, creating plasmid pEF-cAAT2.Minimization of the AAT genomic sequence began withremoval of some of the flanking sequences. From the 5' upstreamsequence, 3.7 kb was removed by digesting pAAT with ClaI andMfeI and religating the vector, creating the vector pAAT(-5').From the 3' flanking sequence in pAAT, 970 bp was removed bydeleting an AflII-SalI fragment, generating pAAT(-3'). Thisplasmid was further reduced by digestion with ClaI and MfeI andreligation, generating the double deletion plasmid pAATΔ, whichis reduced by a total of 4.7 kb, compared to pAAT.Cloning of the AAT minigene vector was performed asfollows. The AAT cDNA was removed from pcAAT as a 1.3 kbBamHI-PstI fragment, and cloned into the BamHI-PstI sites ofpBCSK+ (Stratagene, La Jolla, CA), creating the vectorpBcAAT. An AflII linker was cloned into its BglI-PstI sites,generating vector pBcAATii. A 1529 bp fragment of AAT 3'sequence was PCR amplified and cloned into vector pCR2.1,creating pTA-AAT3', from which the AAT3' sequence wassubcloned as an EcoRI fragment into the EcoRI sites ofpBcAAT, creating pBcAAT-AAT3'. From this plasmid, a 622 bpAvaI-PstI fragment was cloned into the AvaI-PstI sites ofpBcAATii, generating the vector pBcAATiii. Next, a 499 bp PstIfragment from pBcAAT-AAT3' was cloned into pBcAATiii,creating plasmid pBcAATiv. A 552 bp StuI-AflII fragment frompBcAAT-AAT3' was cloned into pBcAATiv, generatingpBcAATv. Next, a 771 bp BamHI fragment from pAAT wascloned into the BamHI site of pBcAATv, to create the plasmidpBcAATvi. The last step in creating the minigene was to clonethe 2.9 kb minigene from pBcAATvi, as an AflII-SnaBIfragment, into the AflII-SnaBI sites of vector pAAT. The result isvector pAATmg, which lacks introns 2-4, resulting in a totalreduction of 3.5 kb in gene size.Vector pAAT-fIXmg was constructed as follows. A BglIIlinker was ligated into the SacI site of pF9, generating pfIXmgi.Next, a 2.6 kb fragment of the AAT first intron was PCRamplified, digested with BamHI and BglII, and cloned into theBglII site of pfIXmgi, creating pfIXmgii. The AAT promoter wasreleased from pTA-AATpro as a SapI-ClaI fragment and clonedinto the SapI-ClaI sites of pfIXmgii, generating the vectorpfIXmgiii. Finally, the 7.9 kb fIXmg was released from plasmidpfIXmgiii by digestion with BglII and cloned into the BglII sitesin plasmid pAAT, resulting in the final vector pAAT-fIXmg.B. In vivo delivery and analysis.C57BL/6 mice were injected over a period of 6 – 9 s with25 µg of DNA in 1.8 ml of 0.9% NaCl, by hydrodynamic tailveininjection (Liu et al, 1999; Zhang et al, 1999). Serumsamples were periodically obtained by retro-orbital bleed. Allanimal procedures were performed under the guidelines set forthby Stanford University and the National Institutes of Health. Thesamples were assayed by a polyclonal capture ELISA assay tomeasure serum AAT (Song et al, 1998; Yant et al, 2000) and/orserum fIX (Sclimenti et al, 2003). Some animals were subjectedto partial, surgical hepatectomy (PH) (Chen et al, 2001) and thenallowed to recover. Serum samples were obtained periodically byretro-orbital bleed and analyzed by ELISA. At 51 d post-PH,treated and control animals were sacrificed and the livers wereremoved for analysis of plasmid DNA. The liver tissue that wasremoved during PH surgery was also subjected to analysis. Livertissue was diced and total DNA was prepared using the Bloodand Cell Culture DNA Maxi kit (Qiagen, Valencia, CA).Southern blot analysis was performed as follows. DNA wasdigested with ScaI and separated on a 0.65% agarose gel. The gelwas depurinated in 0.25 M HCl, denatured in 0.5 M NaOH,neutralized in 0.5M Tris-HCl (pH 7.0) and transferred to an S&SNytran blotting membrane (Schleicher & Shuell, Keene, NH) in20X SSC transfer buffer. The membrane was probed with a 484bp fragment from the hygromycin resistance gene common toboth plasmids, labeled with alkaline phosphatase enzyme, usingthe AlkPhos Direct DNA labeling kit (Amersham PharmaciaBiotech). Hybridization occurred at 55°C for 16 h inhybridization buffer provided with the kit. Membranes werewashed according to protocol and incubated with CDP-StarDetection Reagent for 4 min. The membrane was then exposed toHyperfilm ECL (Amersham Pharmacia Biotech). Southern blotquantification was performed using Kodak 1D software.III. ResultsA. Role of EBV sequences inextrachromosomal plasmid expression in vivoOur previous research showed that the plasmid pEF-AAT, bearing the 19 kb SERPINA1 genomic locusencoding human α 1 -antitrypsin on a vector with the EBVEBNA1 and family of repeats retention sequences, wascapable of maintaining long-term high expression levelsafter transfection into mouse liver (Stoll et al, 2001). Wefirst wanted to determine what role the EBV componentson this plasmid played in sustaining these high expressionlevels. EBNA1 and the EBV family of repeats wereremoved, creating plasmid pAAT. Naked plasmid DNA ofpEF-AAT and pAAT, along with pEF-cAAT and pcAATcontrol cDNA plasmids with and without EBV retentionsequences (Figure 1A), were injected into the mouse livervia hydrodynamic tail-vein injection (Liu et al, 1999;Zhang et al, 2000). Serum AAT levels were determined byELISA. The results are shown in Figure 1B. Bothgenomic vectors were able to provide expression levels~1000-fold greater than the cDNA control vectors, whichfell to ≤ 100 ng/ml within 3 weeks post-injection. Whilethe non-EBV plasmid pAAT was still capable of providingstable, high-levels of expression of AAT (151 µg/ml at 6months), the levels provided by pEF-AAT were ~4-fold213

Stoll et al: The role of EBV and genomic sequences in gene expressionhigher (585 µg/ml at 6 months). This result indicated thatthe EBV components provided some expression advantageto the plasmid pEF-AAT.The liver is a relatively quiescent tissue, suggestingthat cell division-induced vector loss should not be asignificant problem. This conjecture is supported byobservations that plasmid DNA is stably maintainedextrachromosomally in liver cells (Zhang et al, 2000; Stollet al, 2001). In order to determine if there was indeed aDNA retention advantage provided by the EBV sequencesin pEF-AAT compared to pAAT, partial hepatectomy(PH) was performed on sample mice injected with each ofthese plasmids. Partial surgical hepatectomy involvedremoval of two-thirds of the liver, stimulating division andregeneration of the remaining liver tissue. Under thesecircumstances, extrachromosomal plasmid DNA is lostfrom the cells, unless it contains retention sequences suchas those provided by EBNA1 and the EBV family ofrepeats (Krysan et al, 1989). Following PH, serum levelsof AAT were monitored, and the results are shown inFigure 2A. Liver genomic DNA was also analyzed in PHmicebefore and after surgery and in control mice (Figure2B).The effects of partial hepatectomy on AATexpression and DNA retention are summarized in Table 1.In mice injected with the EBV vector pEF-AAT, partialhepatectomy resulted in a 1.7-fold decrease in totalplasmid DNA in the liver and a 4.6-fold decrease in serumAAT levels. Mice injected with the non-EBV plasmidpAAT showed a 3.5-fold decrease in total liver plasmidDNA and a 7.0-fold decrease in serum AAT levels. Theseresults suggested that the EBV sequences were playing arole in enhancing the retention and expression of plasmidDNA in vivo in the presence of cell division. When in nondividingtissue, the non-EBV pAAT plasmid appeared tobe capable of maintaining stable, high-levels ofexpression. For this reason, and also to eliminate anyeffects on expression that the EBV sequences may have,all further experiments were conducted using a non-EBVextrachromosomal vector.B. Minimization of AAT sequencesrequired for expression in vivoWhile the 19 kb SERPINA1 sequence was notprohibitively large for use in extrachromosomal genetherapy vectors, it was still a cumbersome size. It waslikely that some of the 8.5 kb of flanking sequences werenot necessary for regulation or enhancement of geneexpression. In order to address this question and possiblyminimize the functional size of this genomic sequence, webegan by deleting segments of flanking sequence andanalyzing the in vivo expression of these deleted vectors.Using convenient restriction enzyme sites, 3.7 kb of 5'flanking sequence were removed from pAAT to generatepAAT(-5'), 921 bp were removed from the 3' flankingsequence to create pAAT(-3'), and both deletions weregenerated in the same plasmid to produce plasmidpAATΔ. These plasmids, along with the full gene vectorpAAT and a cDNA vector pcAAT, were injectedhydrodynamically into the tail-vein of mice and serumAAT levels were monitored by ELISA. The results areshown in Figure 3. All genomic vectors were able toprovide high expression levels of AAT, >190 µg/ml at twomonths post-injection, though vectors pAAT(-5') andpAATΔ gave ~2-fold higher expression levels than pAATand pAAT(-3').Figure 1. Effect on AAT expression of removal of EBV sequences. (A) pEF-AAT contains the 19 kb AAT genomic region on a plasmidwith the EBV family of repeats (FR) and EBNA1 gene, (B) pAAT contains the 19 kb AAT genomic region (encoded by SERPINA1) in anon-EBV backbone, (C) pEF-cAAT contains the AAT cDNA driven by an RSV promoter, on a plasmid containing the EBV FR andEBNA1 retention sequences, (D) pcAAT contains the AAT cDNA driven by an RSV promoter in a non-EBV backbone. (E) Groups of 5mice were injected with 25 µg of pEF-AAT (), pAAT ( ), pEF-cAAT (), or pcAAT (). Blood was sampled periodically, andserum AAT levels were determined by ELISA. Error bars represent the standard error from five mice.214

Gene Therapy and Molecular Biology Vol 7, page 215Figure 2. Effect of partial hepatectomy (PH) on plasmid DNA and AAT expression in vivo. (A) AAT expression, as determined byELISA on day of and following PH, in mice injected with pEF-AAT () or pAAT (). Solid lines are control (no PH) mice; dashedlines underwent PH on Day 0. (B) DNA was extracted from mouse livers harvested during PH and at 51 d post-PH. DNA was digestedand probed with a labeled fragment from the hygromycin resistance gene.Figure 3. Effect on AAT expressionof removal of flanking sequences inSERPINA1. Groups of five mice wereinjected with 25 µg of pAAT (),pAAT(-5’) ( ), pAAT(-3’) (),pAATΔ (), or pcAAT (). Bloodwas periodically sampled, and serumAAT levels were determined byELISA. Error bars represent thestandard error from five mice.Plasmids are described in the text.Table 1: Effect of partial hepatectomy (PH) on retention of plasmid DNA in liver and AAT expression in vivoAAT expression analysis aSouthern Analysis bDecrease inAATexpressionAvg Decreasein expressionper group4.6Fold decreasein expression inPH vs non-PH groupsDecrease inplasmid DNAin liverAvg Decrease inplasmid DNA inliver per groupMouse 1 PH 3.5 1.78Mouse 25.61.671.73.3pEF- Mouse 3 No PH 1.3 1.4n/aAAT Mouse 41.5n/an/aMouse 1 PH 7.8 7.0ndMouse 26.13.483.57.0pAAT Mouse 3 No PH 0.8 1.0n/aMouse 41.1n/an/aaAAT expression was determined by ELISA, as reported in Figure 2A.bPlasmid DNA presence in liver was determined by southern blot, as shown in Figure 2B. Quantification of bandintensities was performed using Kodak 1D software. Mice are labeled 1-4 as they are labeled in Figure 2B.Control cDNA vectors pcAAT and pEF-cAAT(Figure 1A) were both expressed from the Rous sarcomavirus (RSV) promoter. It was possible that the AATgenomic sequence was not necessary at all, and that thehigh expression was simply the result of the strength of theAAT promoter. To address this possibility, the RSVpromoter was replaced with the AAT promoter inplasmids pcAAT and pEF-cAAT, generating the vectorspcAAT2 and pEF-cAAT2, respectively (Figure 4A).Following liver transfection via tail-vein injection, in vivoexpression was monitored by ELISA (Figure 4B). TheAAT promoter was able to provide some transient stabilityto cDNA expression in vivo. While the RSV-cDNAvectors fell to nearly undetectable levels within 2 weeks215

Stoll et al: The role of EBV and genomic sequences in gene expressionpost-injection, the AAT promoter-cDNA vectors did notfall to these levels until ~5 weeks post-injection. However,there was no significant difference in expression levelsfrom the two promoters. Ultimately, expression was lostfrom all four cDNA vectors, whereas the genomicsequence in pAAT maintained expression at >280 µg/mlfor the duration of the experiment.Miao (2000) created a well-expressed human factorIX minigene that retained only a portion of the first intronof that gene. We wondered if construction of a similarminigene for AAT would be equally successful. Wecreated the plasmid pAATmg, which contained the full 5'and 3' flanking sequences of pAAT, the AAT promoter,and all exons, but retained only the 6.1 kb first intron(Figure 4A). Serum AAT levels were monitored byELISA following tail-vein injection of plasmid DNA intomice, with the results shown in Figure 4C. The 15.5 kbAAT minigene was not capable of providing stable, highexpression levels of AAT in vivo. While day 1 levels ofexpression were within 3-fold of each other, expressionfrom pAATmg dropped ~100-fold below that of pAATwithin 3 weeks post-injection, and continued to drop,approaching cDNA expression levels by day 85.C. Expression of a heterologous genefrom within SERPINA1Our unsuccessful attempt to create an AAT minigeneindicated that not all genes could be easily minimized to asize convenient for use in gene therapy vectors, transgenicmice, or numerous other applications, while stillmaintaining high expression. Due to the high expressionlevels of AAT in vivo in these and previous experiments(Stoll et al, 2001), we hypothesized that the full-genomicSERPINA1 might contain sequences that could enhanceexpression of a heterologous gene cloned within it. Todemonstrate this idea, a human factor IX minigene (Miaoet al, 2000) was cloned into the first exon of the AATgene. Miao (2000) reported high levels of expression forthis factor IX minigene that were maintained in vivo for ≥225 days. However, Sclimenti (2003) observed onlytransiently high expression levels for this minigene thatstabilized at ≤100 ng/ml after 2-3 months post-injection.We therefore chose this factor IX minigene (fIXmg),rather than a factor IX cDNA, to clone within SERPINA1for expression studies, creating the vector pAAT-fIXmg(Figure 5A). Plasmid DNA was transfected into mouseliver by hydrodynamic tail-vein injection, and serum fIXlevels were monitored by ELISA. In addition to pAATfIXmg,we also tested expression from control plasmidspfIX and pDY-fIX, which contained the fIXmg on aplasmid with EBV retention sequences (Sclimenti et al,2003). Results are shown in Figure 5B. As observed bySclimenti (2003), the factor IX minigene alone plasmid,pfIX, had high day 1 levels (13 µg/ml) that fell to

Gene Therapy and Molecular Biology Vol 7, page 217Figure 5. Effect of genomic sequence on expression of aheterologous fIX minigene. (A) Plasmid maps. (B) Factor IXexpression levels in mice. Groups of 5 mice were injected with25 µg of pfIX (), pDY-fIX (), or pAAT-fIXmg (). Bloodwas sampled periodically, and serum fIX levels were determinedby ELISA. Error bars represent the standard error.was able to provide long-term high-level expression ofAAT in vivo. The AAT genomic sequence was capable ofacting as a mammalian origin of replication for thisextrachromosomal plasmid (Stoll et al, 2001), and theEBV sequences can act to retain the plasmid in cells.However, these features were not likely to be relevant invivo because these experiments were conducted in nondividingliver cells.In order to determine whether the EBV sequenceswere providing any advantage in vivo in expression orretention of the vector, we constructed the genomic/non-EBV vector pAAT. Vectors with and without EBVsequences both provided stable, high expression levels invivo (>150 µg/ml), but the pEF-AAT vector with EBVsequences was ~4-fold higher (Figure 1B). Followingpartial hepatectomy in representative mice, we observedan average 3.3-fold decrease in AAT expression in pEF-AAT-injected mice and a 7.0-fold decrease in serum AATin pAAT-injected mice (Figure 2A; Table 1). Southernanalysis revealed that partial hepatectomy-inducedhepatocyte replication resulted in loss of plasmid DNA inboth the presence and absence of EBV retentionsequences. However, pEF-AAT mice retained twice asmuch plasmid DNA as did pAAT mice, which directlycorrelated with the observed serum AAT levels in thesetwo different groups of mice. These results indicated thatthe EBV sequences were providing increased retention tothese extrachromosomal vectors in dividing tissue in vivo.The ~4-fold higher expression levels in pEF-AATmice vs. pAAT mice pre-PH suggested that the EBVsequences were also providing expression enhancementactivity. Enhancer activity has previously been attributedto the action of these EBV sequences (Kenney et al, 1998;Langle-Rouault et al, 1998; Sclimenti et al, 2003). Thisenhancer function appears to rely on the basal expressionability of the transgene construct. The EBV sequencesprovided a ~4-fold boost in expression from the full AATgene, but were unable to prolong or increase expressionfrom AAT cDNA constructs. Even when the viral RSVpromoter driving the cDNA was replaced with the humanAAT promoter, expression was only minimally prolonged,with or without EBV sequences (Figure 4B). This resultsuggests that either the AAT promoter alone was not ableto maintain persistent expression in vivo, or that thesecDNA constructs lacked structural or regulatory sequencesthat contribute to the persistence of expression in vivo, aswas observed for the full-genomic construct pAAT.Using the pAAT vector, we attempted to minimizethe size of the SERPINA1 locus required for highexpression levels in vivo. We observed an ~2-fold increasein expression when 3.7 kb of 5' upstream sequence wasdeleted from the vector, either alone or in combinationwith 921 bp deleted from the 3' end (Figure 3). It ispossible that the 5' flanking sequence may contain arepressor sequence that results in enhanced expressionwhen it is removed from the SERPINA1 genomicsequence. However, expression levels were stillremarkably high and stable for all four of these genomicvectors, indicating that a reduced genomic sequence of14.4 kb was sufficient for attaining good expression ofAAT in vivo.Further minimization of the genomic sequence,through creation of an AAT minigene deleted of all but thefirst intron, resulted in an ~100-fold decrease in expressionin vivo (Figure 4C). However, an alternative AATminigene, retaining all introns except the first one andflanked by β-lactoglobulin sequences, has been shown toexpress AAT at high levels in mice transgenic for theconstruct (Whitelaw et al, 1991; Clark et al, 1993).Considering these two different findings together suggeststhat there are sequences within the last three exons that arenecessary for efficient AAT expression in vivo, though thenature of these expression modulation sequences isunclear. It is possible that one or more of the deletedintrons contain transcriptional enhancers. Alternatively,the first intron may contain transcriptional repressorsequences, which are naturally modulated by sequences inthe downstream exons, removal of which in our minigeneresulted in suppression of expression. Also, it may be thatthe additional exons are required to allow stabilizationand/or proper processing of the mRNA. While several labshave observed high expression levels from minigenescontaining only a first intron (Palmiter et al, 1991; Miao etal, 2000), our results suggest that not all genes areamenable to such simple minimization. Creation ofgenomic minigenes may require analysis of multiple introncombinations to identify a minigene that provides asuitable expression profile in vivo.While SERPINA1 is 19 kb long, thus prompting ourattempts to minimize the amount of sequence sufficient forhigh expression levels in vivo, it is still a relatively small217

Stoll et al: The role of EBV and genomic sequences in gene expressiongene. Some therapeutically relevant genes are much larger,such as factor VIII (~185 kb), cystic fibrosistransmembrane regulator (~230 kb), and dystrophin (~2.4Mb). Unfortunately, cDNA constructs often provide onlytransient expression in vivo. It has been observed that acDNA sequence cloned within another full-length genecan benefit from the genomic sequence in cis, allowing forstable expression of the cDNA (Palmiter et al, 1987;Palmiter et al, 1991). To test this idea for the AAT gene,we cloned the human factor IX minigene (Miao et al,2000) into SERPINA1 between the promoter and theinitiation codon. We observed an ~50-fold increase in fIXexpression when the fIX minigene was placed within theAAT gene, compared to a vector containing only theminigene (Figure 5B). The AAT gene enhancedexpression of the fIX minigene to levels ~9-fold higherthan EBV sequences enhanced fIX expression (pDY-fIX;Figure 5B). It therefore appears that the AAT genomicsequence is capable of increasing and maintainingexpression of a heterologous minigene. It seems likely thatthese AAT sequences will be capable of exerting asimilarly beneficial effect on other cDNA or minigeneconstructs. Since AAT is expressed only in the liver, it ispossible that this effect will be limited to expression fromthat tissue. However, if the benefit is the result of a tissueindependentenhancer activity or a structurally openchromatin configuration of the genomic sequence, it maybe extendable to other tissues, provided that a suitablepromoter drives the heterologous cDNA.While we have demonstrated that some minimizationof SERPINA1 genomic size did not affect its in vivoexpression, there was a limit to the amount and choice ofsequences that could be removed without affectingexpression. For example, the last three introns ofSERPINA1 appeared to be much more important than thefirst intron alone, as evidenced by the difference in in vivoexpression from these two different minigene constructs(Figure 4C) (Whitelaw et al, 1991), though our pAATmgconstruct actually retained 1.8 kb more genomic sequence.A factor IX minigene containing only the first intron hasbeen shown to provide adequate expression levels in vivo(Miao et al, 2000). It is therefore important to note that thesequences required for efficient expression in vivo arelikely to differ from gene to gene. These resultsdemonstrate the value of utilizing genomic sequences forin vivo expression of a transgene in a gene therapy system.AcknowledgmentsThis work was supported by NIH grants HL69737 toMPC and HL64274 to MAK.ReferencesBrinster, RL, Allen, JM, Behringer, RR, Gelinas, RE andPalmiter, RD (1988) Introns increase transcriptionalefficiency in transgenic mice. Proc Natl Acad Sci U S A 85,836-840.Chen, ZY, Yant, SR, He, CY, Meuse, L, Shen, S and Kay, MA(2001) Linear DNAs concatemerize in vivo and result insustained transgene expression in mouse liver. Mol Ther 3,403-410.Clark, AJ, Archibald, AL, McClenaghan, M, Simons, JP,Wallace, R and Whitelaw, CB (1993) Enhancing theefficiency of transgene expression. Philos Trans R SocLond B Biol Sci 339, 225-232.Gill, DR, Smyth, SE, Goddard, CA, Pringle, IA, Higgins, CF,Colledge, WH and Hyde, SC (2001) Increased persistence oflung gene expression using plasmids containing the ubiquitinC or elongation factor 1alpha promoter. Gene Ther 8, 1539-1546.Heinzel, SS, Krysan, PJ, Tran, CT and Calos, MP (1991)Autonomous DNA replication in human cells is affected bythe size and the source of the DNA. Mol Cell Biol 11, 2263-2272.Kaneda, Y, Saeki, Y, Nakabayashi, M, Zhou, WZ, Kaneda, MWand Morishita, R (2000) Enhancement of transgeneexpression by cotransfection of oriP plasmid with EBNA-1expression vector. Hum Gene Ther 11, 471-479.Kenney, S, Ge, JQ, Westphal, EM and Olsen, J (1998) Genetherapy strategies for treating Epstein-Barr virus-associatedlymphomas: comparison of two different Epstein-Barr virusbasedvectors. Hum Gene Ther 9, 1131-1141.Krysan, PJ, Haase, SB and Calos, MP (1989) Isolation of humansequences that replicate autonomously in human cells. MolCell Biol 9, 1026-1033.Langle-Rouault, F, Patzel, V, Benavente, A, Taillez, M,Silvestre, N, Bompard, A, Sczakiel, G, Jacobs, E and Rittner,K (1998) Up to 100-fold increase of apparent geneexpression in the presence of Epstein-Barr virus oriPsequences and EBNA1: implications of the nuclear import ofplasmids. J Virol 72, 6181-6185.Liu, F, Song, Y and Liu, D (1999) Hydrodynamics-basedtransfection in animals by systemic administration of plasmidDNA. Gene Ther 6, 1258-1266.Miao, CH, Ohashi, K, Patijn, GA, Meuse, L, Ye, X, Thompson,AR and Kay, MA (2000) Inclusion of the hepatic locuscontrol region, an intron, and untranslated region increasesand stabilizes hepatic factor IX gene expression in vivo butnot in vitro. Mol Ther 1, 522-532.Miao, CH, Thompson, AR, Loeb, K and Ye, X (2001) Long-termand therapeutic-level hepatic gene expression of humanfactor IX after naked plasmid transfer in vivo. Mol Ther 3,947-957.Paillard, F (1997) Promoter attenuation in gene therapy: causesand remedies. Hum Gene Ther 8, 2009-2010.Palmiter, RD, Behringer, RR, Quaife, CJ, Maxwell, F, Maxwell,IH and Brinster, RL (1987) Cell lineage ablation intransgenic mice by cell-specific expression of a toxin gene.Cell 50, 435-443.Palmiter, RD, Sandgren, EP, Avarbock, MR, Allen, DD andBrinster, RL (1991) Heterologous introns can enhanceexpression of transgenes in mice. Proc Natl Acad Sci U S A88, 478-482.Reisman, D and Sugden, B (1986) trans activation of an Epstein-Barr viral transcriptional enhancer by the Epstein-Barr viralnuclear antigen 1. Mol Cell Biol 6, 3838-3846.Reisman, D, Yates, J and Sugden, B (1985) A putative origin ofreplication of plasmids derived from Epstein-Barr virus iscomposed of two cis-acting components. Mol Cell Biol 5,1822-1832.Sclimenti, CR, Neviaser, AS, Baba, EJ, Meuse, L, Kay, MA andCalos, MP (2003) Epstein-Barr Virus Vectors ProvideProlonged Robust Factor IX Expression in Mice. BiotechnolProg 19, 144-151.Song, S, Morgan, M, Ellis, T, Poirier, A, Chesnut, K, Wang, J,Brantly, M, Muzyczka, N, Byrne, BJ, Atkinson, M andFlotte, TR (1998) Sustained secretion of human alpha-1-218

Gene Therapy and Molecular Biology Vol 7, page 219antitrypsin from murine muscle transduced with adenoassociatedvirus vectors. Proc Natl Acad Sci U S A 95,14384-14388.Stoll, SM, Sclimenti, CR, Baba, EJ, Meuse, L, Kay, MA andCalos, MP (2001) Epstein-Barr virus/human vector provideshigh-level, long-term expression of alpha1-antitrypsin inmice. Mol Ther 4, 122-129.Webster, J, Donofrio, G, Wallace, R, Clark, AJ and Whitelaw,CB (1997) Intronic sequences modulate the sensitivity ofbeta-lactoglobulin transgenes to position effects. Gene 193,239-243.Whitelaw, CB, Archibald, AL, Harris, S, McClenaghan, M,Simons, JP and Clark, AJ (1991) Targeting expression to themammary gland: intronic sequences can enhance theefficiency of gene expression in transgenic mice. TransgenicRes 1, 3-13.Yant, SR, Meuse, L, Chiu, W, Ivics, Z, Izsvak, Z and Kay, MA(2000) Somatic integration and long-term transgeneexpression in normal and haemophilic mice using a DNAtransposon system. Nat Genet 25, 35-41.Yates, J, Warren, N, Reisman, D and Sugden, B (1984) A cisactingelement from the Epstein-Barr viral genome thatpermits stable replication of recombinant plasmids in latentlyinfected cells. Proc Natl Acad Sci U S A 81, 3806-3810.Yates, JL, Warren, N and Sugden, B (1985) Stable replication ofplasmids derived from Epstein-Barr virus in variousmammalian cells. Nature 313, 812-815.Zhang, G, Budker, V and Wolff, JA (1999) High levels offoreign gene expression in hepatocytes after tail veininjections of naked plasmid DNA. Hum Gene Ther 10,1735-1737.Zhang, G, Song, YK and Liu, D (2000) Long-term expression ofhuman alpha1-antitrypsin gene in mouse liver achieved byintravenous administration of plasmid DNA using ahydrodynamics-based procedure. Gene Ther 7, 1344-1349.219

Stoll et al: The role of EBV and genomic sequences in gene expression220

Gene Therapy and Molecular Biology Vol 7, page 221Gene Ther Mol Biol Vol 7, 221-228, 2003Site-specific kidney-targeted plasmid DNA transferusing nonviral techniquesReview ArticleHiroki Maruyama 1 , Noboru Higuchi 1 , Shigemi Kameda 1 , Gen Nakamura 1 , JunichiMiyazaki 2 , and Fumitake Gejyo 11 Division of Clinical Nephrology and Rheumatology, Niigata University Graduate School of Medical and Dental Sciences,1-757 Asahimachi-dori, Niigata 951-8120, Japan2 Division of Stem Cell Regulation Research, G6, Osaka University Medical School, 2-2 Yamadaoka, Suita 565-0871,Japan__________________________________________________________________________________*Correspondence: Hiroki Maruyama, MD, Division of Clinical Nephrology and Rheumatology, Niigata University Graduate School ofMedical and Dental Sciences, 1-757 Asahimachi-dori, Niigata 951-8120, Japan; Phone: +81-25-227-2194; Fax: +81-25-227-0775; E-mail: hirokim@med.niigata-u.ac.jpKey words: renal artery, renal vein, ureter, pelvis, glomerulus, tubules, interstitium, lipoplex, polyplex, hydrodynamic-basedtransfection, electroporation, ultrasound-microbubbleAbbreviations: hemagglutinating virus of Japan, (HVJ); anionic artificial viral envelope, (AVE); chloramphenicol acetyltransferase,(CAT); Epstein-Barr virus, (EBV); EBV nuclear antigen-1, (EBNA-1); transforming growth factor, (TGF)- β1; platelet-derived growthfactor, (PDGF)-B; N[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethylammonium chloride, (DOTMA); cytomegalovirus, (CMV);dioleoylphosphatidyl ethanolamine, (DOPE); 2,3-dioleyloxy-N-[2(sperminecarboxamido)ethyl]-N,N-dimethyl-1-propanaminiumtrifluoroacetate, (DOSPA), 3β-[N-(N,N-dimethylaminoethane)carbamoyl] cholesterol, (DAC-Chol); 3-dimethyl-hydroxyethylammoniumbromide, (DMRIE); Polyethyleneimine, (PEI); heat shock protein, (HSP); peritubular capillaries, (PTC); erythropoietin,(Epo); carbonic anhydrase, (CA); liposome-polycation-DNA complexes, (LPD)Received: 18 September 2003; Revised: 19 November 2003;Accepted: 25 November 2003; electronically published: November 2003SummaryKidney-targeted gene transfer has the potential to be one of the most important tools for broadening ourunderstanding of renal disease processes and for revolutionizing the treatment of renal diseases. We reviewed theliterature on kidney-targeted plasmid DNA transfer using nonviral techniques: naked plasmid DNA, cationiclipid/DNA complex, hemagglutinating virus of Japan (HVJ)-liposome complex, anionic artificial viral envelope-typeHVJ-liposome complex, cationic polymer/DNA complex, electroporation, ultrasound-microbubble, andhydrodynamics-based transfection. Gene-transfer methods using nonviral techniques are administered via renalarterial, renal venous, pelvic, ureteric, or direct routes into the glomerulus, tubules, or interstitial fibroblasts. Genetransfer can be achieved with varying degrees of transfection efficiency and duration of gene expression. Thus, wecan select the most effective gene transfer technique to deliver the appropriate therapeutic gene to the particularsites involved in various renal diseasesI. IntroductionWe can access the kidney by four different localroutes: the renal artery, anterogradely; the renal vein,retrogradely; the urinary tract (ureter or pelvis),retrogradely; and by direct injection.Viruses (adenovirus, adeno-associated virus,retrovirus and others) are used as a vector to transferexogenous genes into kidneys. Although these vectors canachieve high gene transfer efficiency, it is still possiblethat the protein encoded by viral genes induceimmunological responses. On the other hand, nonviraltechniques have several advantages over transfer usingviral vectors. A large quantity of highly purified plasmidDNA is easily and inexpensively obtained. Because thereare fewer size constraints than with current viral vectors,plasmid vectors can carry larger genes. Plasmid vectorsare less likelihood of inducing an immunological responsethan with viral vectors.The local route to the kidney and the kind of nonviraltechniques used principally determine the transgeneexpression sites (Tomita et al, 1992; Isaka et al, 1993;Akami et al, 1994; Arai et al, 1995; Boletta et al, 1997; Laiet al, 1997, 1998; Tsujie et al, 2000; Foglieni et al, 2000;Tsujie et al, 2001a, 2001b; Maruyama et al, 2002a; Lan etal, 2003; Kita et al, 2003). Tubules have been targeted bythe injection of naked plasmid DNA into the intra-renalparenchyma (Lai et al, 1997), cationic lipid/DNA complex221

Maruyama et al: Kidney-targeted plasmid DNA transfervia the pelvic route (Lai et al, 1997, 1998) or renal artery(Lai et al, 1997), anionic artificial viral envelope (AVE)-type hemagglutinating virus of Japan (HVJ)-liposomecomplex via the renal artery (Kita et al, 2003), or cationicpolymer/DNA complex via the renal artery (Boletta et al,1997; Foglieni et al, 2000), or by ultrasound-Optisonmediatednaked plasmid DNA transfer via the renal artery(Lan et al, 2003).The glomerulus has been targeted by delivering anHVJ-liposome complex via the renal artery (Tomita et al,1992; Isaka et al, 1993; Akami et al, 1994; Arai et al,1995), an AVE-type HVJ-liposome complex via the renalartery (Tsujie et al, 2001a), by electroporation-mediatednaked plasmid DNA transfer via the renal artery (Tsujie etal, 2001b), and by ultrasound-Optison-mediated nakedplasmid DNA transfer via the renal artery (Lan et al,2003).Interstitial fibroblasts have been targeted by injectingan AVE-type HVJ-liposome complex retrogradely via theureter (Tsujie et al, 2000), by the retrograde injection ofnaked plasmid DNA via the renal vein usinghydrodynamics-based transfection (Maruyama et al,2002a), or by ultrasound-Optison-mediated naked plasmidDNA transfer via the renal artery (Lan et al, 2003).Most nonviral techniques of kidney-targeted genetransfer have resulted in transient gene expression forrelatively short periods of time: the HVJ-liposome (Tomitaet al, 1992; Isaka et al, 1993; Arai et al, 1995),electroporation-mediated naked plasmid DNA transfer(Tsujie et al, 2001b), and ultrasound-Optison-mediatednaked plasmid DNA transfer (Lan et al, 2003) methodshave resulted in expression for 1 week or less. Thecationic polymer/DNA complex (Boletta et al, 1997) andAVE-type HVJ-liposome delivered retrogradely via theureter (Tsujie et al, 2000) have resulted in 2 weeks ofexpression. Genes delivered by cationic lipid/DNAcomplex were expressed for 5 weeks (Lai et al, 1997,1998), and AVE-type HVJ-liposome complex deliveredvia the renal artery for 8 weeks (Tsujie et al, 2001a).Recently, the technique of retrograde injection via therenal vein has achieved long-term gene expression, lastingmore than 24 weeks (Maruyama et al, 2002a).Thus, the recent history of kidney-targeted plasmidDNA transfer shows the rapid development of nonviraltechniques for use in gene therapy. In this article, wereview kidney-targeted plasmid DNA transfer usingnonviral techniques via several routes into different sites,their transfection efficiency, preclinical studies using thesemethods, safety concerns, and future clinical applications.II. Kidney-targeted gene transfer viathe renal arteryTable 1 provides an overview of studies that haveused kidney-targeted plasmid DNA transfer via the renalartery.A. Liposomes1 HVJ-liposomeTomita et al, (1992) first reported that the HVJliposomeis efficient for glomerulus-targeted gene transfer.Its expression is transient. Four days after the injection ofHVJ-liposome, SV40 large T antigen is detectedimmunohistochemically in 15% of the glomerular cells inthe kidney. Isaka et al, (1993) demonstrated a 35%transfection efficiency of the glomerulus with plasmidscarrying the gene for chloramphenicol acetyltransferase(CAT), transforming growth factor (TGF)-β, and plateletderivedgrowth factor (PDGF)-B genes, driven by thechicken β actin promoter. The blood urea nitrogen andplasma creatinine concentrations are not changedsignificantly by this procedure (Tomita et al, 1992; Isakaet al, 1993). Proteinuria is not observed in the rats treatedwith the HVJ-liposome method (Tomita et al, 1992; Isakaet al, 1993). Isaka et al (1993) demonstrated that theintroduction of either the TGF-β1 or PDGF-B gene intothe kidney induces glomerulosclerosis. Akami et al, (1994)reported the HVJ-liposome mediated gene transfer of ahuman CD59 expression plasmid, pSRαCD59, into thecanine kidney. Human CD59 was detectedimmunohistochemically in the canine glomerular cells for7 days with a maximum on day 2. Yamada et al (1995)identified a silencer in the mouse renin gene promoterregion by transferring renin gene promoter-CAT fusionconstructs into the kidney using the HVJ-liposometechnique. Arai et al, (1995) demonstrated human reninexpression in approximately 30% of the glomeruli on day3 after the introduction of the human renin andangiotensinogen genes. In addition, on day 7 after thetransfection of the renin and angiotensinogen genes, theextracellular matrix had expanded in the glomeruli and α-smooth muscle was expressed in the mesangial cells.These findings suggest that the locally activated reninangiotensin system induces glomerular sclerosis and aphenotypic change in mesangial cells.The mechanism of the HVJ-liposome technique thatresults in glomerulus-specific gene expression is unknown.A possible explanation is that the intrarenal hydrodynamicstate facilitates the contact of HVJ-liposomes withglomerular cell membranes, and another is that thephagocytotic activity of glomerular mesangial cellsfacilitates the uptake of HVJ-liposomes (Tomita et al,1992).3. Cationic lipid/DNA complexesCationic lipid/DNA complexes have been used forkidney-targeted gene transfer via the renal artery.3.1 LipofectinLipofectin (Invitrogen, San Diego, CA) contains twolipid species, a cationic lipid, N[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethylammonium chloride (DOTMA), and aneutral phospholipid, dioleoylphosphatidyl ethanolamine(DOPE). After intrarenal-arterial injection of pCMV-βgal-Lipofectin,lacZ was expressed mainly in thecytoplasm of renal tubular cells in the outer medulla withsome staining in the cortex (Lai et al, 1997). It is not clearwhy there was no transgene expression in the innermedulla. The authors speculate that the high salt and ureacontent in the inner medulla might have affected the entryof the plasmid DNA-Lipofectin complex.222

Gene Therapy and Molecular Biology Vol 7, page 223Table 1: Gene transfer via renal arteryHVJ, HVJ lipospme; pAct, gene expression vector driven by the chicken β_actin promoter (Fregien and Davidson, 1986); CAT,chloramphenicol acetyltransferase; TGF-β1, transforming growth factor-β1; PDGF-B, platelet-derived growth factor-B; SD rat, Sprague-Dawley rat; pSRα, cDNA expression vector composed of the simian virus 40 early promoter and the R-U5’ segment of human T-cellleukemia virus type 1 long terminal repeat (Tatebe et al, 1988); pUtkCAT, mammalian expression vector encoding CAT under thecontrol of a truncated herpes simplex virus thymidine kinase (tk) promoter; pCAGGS, gene expression vector driven by the CAG(cytomegalovirus(CMV) immediate-early enhancer/chicken β-actin hybrid) promoter (Niwa et al, 1991); Lipofectin (Invitrogen),contains two lipid species, a cationic lipid, N[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethylammonium chloride (DOTMA), and a neutralphospholipid, dioleoylphosphatidyl ethanolamine (DOPE); pCMV, gene expression vector driven by CMV promoter; β-gal, Escherichiacoli β-galactosidase; DOGS, dioctadecylamidoglycyl spermine; P. pyralis Luc, firefly Photinus pyralis Luc; AVE-type HVJ, artificialviral envelope-type HVJ liposome; pEBAct, the EBV replicon vector, which was constructed by cloning the Epstein-Barr virus nuclearantigen-1 (EBNA-1) and oriP sequences in the same plasmid; Luc, luciferase; Lipofectamine (Invitrogen), contains two lipid species, acationic lipid, 2,3-dioleyloxy-N-[2(sperminecarboxamido)ethyl]-N,N-dimethyl-1-propanaminium trifluoroacetate (DOSPA), and aneutral phospholipid DOTMA; DAC-Chol, conposed of 3β-[N-(N,N-dimethylaminoethrane)carbamoyl] cholesterol and DOPE; DMRIE,3-dimethyl-hydroxy-ethylammonium bromide; pBAG, β-gal-transducing vector driven by Moloney murine leukemia virus promoter(Price et al, 1987); Renin FURIN , the human renin cDNA fragment is modofied by insertion of a furin cleavage site at the end of the reninprosegment, redering renin active following furin cleavage; 25-kD PEI, branched 25-kD cationic polymer polyethylenimine; β-galNuc,lacZ gene equipped with a nuclear localization signal; pcDNA3 (Invitrogen), expression vector driven by thecytomegalovirusimmediate-early enhancer/promoter; pTRE (Clontech), tetracycline-inducible vector driven by tetracyclin responseelement/minCMV promoter (CMV promoter deleted enhancer; pEFpurop-Tet-on (Clontech), improved pTet-on vector driven by theCMV promoter; m2Smad7, Smad7 cDNA with a flag tag (m2) at its NH 2 terminus; UUO model, unilateral ureteral obstruction model;pRL-CMV (Promega), Renilla luciferase expression vector driven by the cytomegalovirusimmediate-early enhancer/promoter; HSP70,heat shock protein 70. Note that in cases where no incubation time is given, this information was not reported in the cited study.223

Maruyama et al: Kidney-targeted plasmid DNA transferIt is also unclear how the plasmid DNA-Lipofectincomplex entered the tubular cells from the vasculature.The basement membranes do not preclude the transfectionby plasmid DNA-Lipofectin of tubular epithelial cells viathe vascular route (Lai et al, 1997). The major limitation ofthe liposomes tested so far for efficient in vivo genetransfer is their toxicity upon intrarenal arterialadministration.Lai et al, (1997) reported severe ischemic changeswith necrosis and fibrosis at the time of mouse kidneyharvest after the intrarenal arterial injection of Lipofectinliposomes. Thus, the high incidence of renal ischemicinjury reported by Lai et al, (1997) indicates that there areserious safety concerns with this technique.3.2. OthersMadry et al, (2001) reported that neitherLipofectamine (Invitrogen), which contains two lipidspecies, a cationic lipid, 2,3-dioleyloxy-N-[2(sperminecarboxamido)ethyl]-N,N-dimethyl-1-propanaminium trifluoroacetate (DOSPA), and a neutralphospholipid, DOTMA, nor DAC-Chol liposomes,composed of 3β-[N-(N,N-dimethylaminoethane)carbamoyl] cholesterol (DAC-Chol), nor DOPE coulddeliver β-galactosidase or human renin FURIN to the kidney.Cationic liposome-mediated gene transfer via the renalartery was accompanied by nephrotoxicity and did notresult in marker gene expression (Madry et al, 2001).Neither the monocationic lipid DOTAP (N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulphate) (Boletta et al, 1997; Foglieni et al, 2000) nor thepolycationic lipid dioctadecylamidoglycyl spermine(DOGS, Transfectam, Promega, Madison, WI) (Boletta etal, 1997) was effective for kidney-targeted gene transfer invivo. When 3-dimethyl-hydroxy-ethylammonium bromide(DMRIE) liposomes or Lipofectamine liposomes wereused, the treated kidneys showed pericapsular hematomas,hemorrhages, and necrotic areas consistent with infarction(Madry et al, 2001).B. Cationic polymer/DNA complexCationic polymer/DNA complexes have been used asnonviral vectors for gene transfer.1. 25-kD Polyethyleneimine (PEI)Boletta et al, (1997) demonstrated that efficientreporter gene expression was achieved by injecting DNAcomplexed to the branched 25-kD cationic polymerpolyethylenimine (PEI 25k) into the kidney. The productof the reporter gene, β-galactosidase, was localized almostexclusively to the proximal tubules. The transfectionefficiency of the branched PEI 25k was significantlyhigher than that of the linear form of PEI 22k, PEI 800k.A critical parameter in nonviral DNA delivery is theoverall charge of the DNA-containing particles. Whereasa net positive charge is desired for particle interaction withthe cellular plasma membrane and its entry into the cell, anexcessive positive charge might favor entrapment of theDNA complexes by the extracellular matrix (Boletta et al,1997).To investigate whether the specific transfection ofthe proximal tubules involves glomerular filtration of theDNA-containing particles, Foglieni et al (2000) preparedfluorescent PEI 25k polyplexes containing fluoresceinatedpoly-L-lysine (mean diameter, 93 nm). This allowedvisualization of the route of the particles into the kidney.Foglieni et al (2000) demonstrated that polyplexes that cantransfect proximal tubular cells have access to these cellsthrough glomerular filtration. Conversely, fluorescentlipoplexes containing the cationic lipid DOTAP (meandiameter, 160 nm) were never observed in tubular cells.The size of the transfecting particles is a key parameter inthis process, as expected by the constraints imposed by theglomerular filtration barrier. To date, PEI 25k has beenthe most useful polymer for kidney-targeted gene transfer.Cationic polymer/DNA complexes injected via the leftrenal artery are not confined to the left kidney and are alsoentrapped by other organs (Boletta et al, 1997).C. ElectroporationTsujie et al, (2001b) first demonstrated that in vivoelectroporation provides an efficient approach forglomerulus-targeted gene transfer. Expression vectors in0.5 ml of balanced salt solution were injected into the leftrenal artery via a catheter in a one-shot manner and the leftrenal vein was clamped immediately after the injection.The left kidney was then sandwiched between a pair ofoval-shaped tweezer-type electrodes and electroporated.Four days after the transfection of pCAGGS-lacZ, β-galactosidase expression was observed in 75% ofglomeruli from the injected kidney. Thus, in vivoelectroporation with intrarenal-arterial DNA injection wasmore effective than the HVJ-liposome method.For this study, Tsujie et al, (2001b) used a luciferaseexpression plasmid, pEBAct-Luc, and transferred it byelectroporation with 25, 50, 75, or 100 V. Luciferaseactivities in glomeruli did not change significantly atvoltages of 25 to 100 V. The authors observed fewharmful effects on the treated kidneys, except for smallburns on the surface of the kidney in contact withelectrodes. No histologic damage was seen in theglomeruli or tubular epithelial cells. A possibleexplanation of the glomerulus-specific gene expression isthat, as the glomerular capillaries have abundantfenestrations, a large volume of DNA solution could alsoenter the mesangial area through them (Tsujie et al,2001b). Electroporation-mediated gene transfer candeliver naked plasmid DNA into the entire area in contactwith the DNA within the electric field. Therefore, theexact mechanism of the glomerulus-specific geneexpression requires further study. We speculate that ahydrodynamics-based transfection by the injection of alarge volume of DNA solution in a one-shot manner couldachieve the desired gene transfer.D. Ultrasound-microbubble (Optison)Lan et al, (2003) first demonstrated that theultrasound-mediated disruption of gas-filled microbubbles224

Gene Therapy and Molecular Biology Vol 7, page 225could be used effectively to transfer naked plasmid DNAinto the kidney. The mechanism by which ultrasoundmicrobubbles enhance transgene expression in all celltypes within the kidney may largely be attributed toultrasound-mediated microbubble cavitation. It is possiblethat the cavitation may largely increase the permeability ofcapillary and tubular basement membranes, which allowsthe locally released DNA to cross through these basementmembranes and enter cells, including glomerular,interstitial, and tubular epithelial cells.The ultrasound-Optison-mediated m2Smad7/Tet-onplasmid transfer resulted in transgene expression in morethan 90% of glomerular, tubular, and interstitial cells.Gene transfer of inducible Smad7 using the ultrasoundmicrobubblesystem inhibited renal fibrosis in the ratunilateral ureteral obstruction model. The ultrasoundtreatment did not cause any abnormal histologic orfunctional changes, as evidenced by normal urinaryprotein excretion, normal glomerular and tubulointerstitialmorphology, and the lack of cellular and interstitial edemaand of local inflammation.E. Naked plasmid DNAIt has not been possible to express transgenes in thekidney by injecting naked plasmid DNA via the renalartery (Lai et al, 1997, Boletta et al, 1997). Lai et al (1997)and Boletta et al (1997) performed intrarenal arterialinjection of naked plasmid DNA while the renal arterialblood flow was interrupted but the renal venous bloodflow was not. The hydrodynamics-based transfectionmechanism did not work in either of these studies.F. AVE-type HVJ-liposome ex vivoThe AVE-type HVJ-liposome method effectivelyinduced the heat shock protein (HSP) 70 or the bcl-2 genein kidney grafts. This was effective even when the HVJliposomevector was mixed with a cold-preservationsolution (Ringer’s lactated or University of Wisconsinsolution) and infused into the renal artery just prior tostorage of the kidney at 4°C for 24 to 48 h in the samepreservation solution (Kita et al, 2003). The induction ofthe HSP70 or bcl-2 gene reduced the occurrence ofprimary non-function of grafted rat kidneys after longtermpreservation. The transgene expression in the kidneywas limited to the tubules. In contrast, Tsujie et al (2001a),as described above, observed expression exclusively in theglomeruli when they introduced plasmid DNA using theAVE-type HVJ-liposome by renal arterial injection. Thediscrepancy between the two studies may be due to thedifference in plasmid vectors used, or for other, as yetunknown, reasons.III. Kidney-targeted gene transfer viarenal veinTable 2 gives an overview of studies that have usedkidney-targeted naked plasmid DNA transfer via the renalvein. Kidney-targeted gene transfer via the renal vein canbe achieved exclusively by hydrodynamics-basedtransfectionA. Hydrodynamics-based transfectionMaruyama et al, (2002a) have developed a techniquefor transferring naked plasmid DNA into the kidney ofnormal rats by hydrodynamics-based transfection byretrograde injection of the DNA into the renal vein. Whenthis technique was performed using a lacZ expressionplasmid as the reporter gene, lacZ expression was detectedexclusively in the interstitial fibroblasts near theperitubular capillaries (PTC) of the injected kidney, asassessed by immunoelectron microscopic analysis. Nonephrotoxicity attributable to gene transfer was apparenteither by histological or functional examinations of theinjected kidney. These authors also used this method totransfect rat kidney with erythropoietin (Epo) gene.Maximal Epo expression was obtained when the vector,pCAGGS-Epo, was injected in Ringer’s solution within 5sec, and with a volume of 1.0 ml. The transgene-derivedEpo mRNA was detected by RT-PCR only in the targetedkidneys and without aberrant expression in nontargetorgans. After an injection of 100 µg of pCAGGS-Epo, theserum Epo levels peaked at 208.3 ± 71.8 mU/ml at week 5,and gradually decreased to 116.2 ± 38.7 mU/ml at week24. Transgene-derived Epo secretion resulted in significanterythropoiesis. The preparation of naked plasmid DNA issimple, compared with the preparation involved for othernonviral techniques. Moreover, the most long-term stablegene expression in the kidney has been obtained using thistechnique.The mechanism underlying the transfer of the nakedplasmid DNA into fibroblasts is unclear. The negativelycharged (Dworkin et al, 2000) PTC could be refractory tothe transfer of negatively charged naked plasmid DNAinto the endothelium.However, the PTC wall consists of an extremely thinendothelium, and fifty percent of the PTC endothelium isfenestrated, and thus highly permeable to water and smallsolutes (Lemley and Kriz, 1994). No incubation time isrequired for this technique; therefore, hydrostatic pressure(Liu et al, 1999) may be the mechanism underlying thetransfer of the naked plasmid DNA. Efficient expressionmay depend on the elevated intravascular hydrostaticpressure caused by the rapid injection of a sufficientvolume of fluid, which leads to the transfer of the nakedplasmid DNA through the PTC endothelium, despite thenegative charge. The sharp vascular resistance gradientbetween the efferent arteriole and the PTC (Lemley andKriz, 1994) probably plays a major role in the blockade ofthe retrograde stream of injected naked plasmid DNAsolution and in the dilation of the PTC, which are the mostexpandable sites in this route, resulting in transgeneexpression in the fibroblasts. Moreover, these cells mayhave the ability to take up the naked plasmid DNA.Recently, Shimizu et al, (2003) demonstrated that thetransfer of a kidney-targeted naked plasmid encoding 7ND(anti-monocyte chemoattractant protein-1) into the kidneyinterstitial cells using the technique of retrograde injectioninto the renal vein (Maruyama et al 2002a) attenuates the225

Maruyama et al: Kidney-targeted plasmid DNA transferTable 2. Gene transfer via renal veinEpo, erythropoietin; MCP-1, Anti-monocyte chemoattractanttubulointerstitial renal injury induced by protein-overloadproteinuria.IV. Kidney-targeted gene transfer viathe urinary tractTable 3 provides an overview of studies of kidneytargetedplasmid DNA transfer in a retrograde manner viathe urinary tract. Two kinds of liposomes have been usedin retrograde gene transfer via the pelvis or ureter. Usingthem, investigators can target tubules and other interstitialfibroblasts.A. Lipofectin via the pelvisAfter the intrarenal pelvic injection of pCMV-β-gal-Lipofectin, lacZ was expressed mainly in the cytoplasm ofrenal tubular cells in the outer medulla with some stainingin the cortex (Lai et al, 1997). The transgene expressionsite accessed by this route was similar to that accessed bythe intrarenal arterial route, as described above. Lai et al(1997) speculated that the difference in themicroenvironments of the inner and outer medulla mightaffect the interactions between the DNA and the liposome,therefore preventing gene uptake by the papillary cells,and resulting in a lack of staining in the inner medulla.Alternatively, the transgene taken up by the papillary cellsmay not have been expressed properly, owing to factorssuch as limited promoter efficiency and specificity, earlydegradation of the DNA, and altered half-life of themRNA and/or protein synthesized from transgene.Intrarenal pelvic injection is more feasible than intra-renalarterialinjection in mice, because it is difficult to inject anadequate amount of plasmid DNA-Lipofectin complexinto the renal artery, and because intrarenal-arterialinjection has a high incidence of renal ischemic injury (Laiet al, 1997).Lai et al, (1998) also demonstrated a transientphysiological effect of gene delivery into the kidney by theretrograde injection of a cationic liposome complexed witha carbonic anhydrase (CA) II chimeric gene, pCMV-CAII-Lipofectin into the renal pelvis of CAII-deficient mice.The CAII-deficiency was produced by introducing a pointmutation into the CAII gene, and it manifested as renaltubular acidosis. The delivery of the CAII gene correctedthe renal tubular acidosis of the CAII-deficient mousemodel. Immunohistochemistry studies using anti-CAIIantibodies showed that CAII was expressed in the tubularcells of the outer medulla and corticomedullary junction.The gene therapy was not associated with nephrotoxicity,as assessed by evaluating the blood urea nitrogen levelsand the renal histology. This is the first successful genetherapy of a genetic renal disease.B. Naked plasmid DNA via pelvisIt has not been possible to express transgenes in thekidney by injecting naked plasmid DNA via the pelvis(Lai et al, 1997).C. AVE-type HVJ-liposome viaTo examine the pattern of transgene-expressing cells,Tsujie et al (2000) used the N-lacF fragment containingthe SV40 nuclear transport signal at the N terminus of thelacZ gene, which limits the β-galactosidase proteinlocalization to the nucleus of transfected cells and can bedistinguished from endogeneous β-galactosidase. pEBAct-NlacF was introduced into the kidney of normal ratsretrogradely via the ureter using the AVE-type HVJliposomemethod (Tsujie et al, 2000). Nuclear β-galactosidase activity was observed in interstitialfibroblasts for 2 weeks. After the introduction of thepEBActN-lacF expression vector into kidneys with anobstructed ureter, Tsujie et al, (2000) observed a greatlyexpanded interstitium with marked cell proliferation ondays 7 and 14. In contrast, no pathological changes norincreased numbers of infiltrated cells were observed in thetreated kidneys 1 day after the transfection, suggesting thatthere was no interstitial damage from the HVJ-liposomesolution, but that the damage was caused by ureteralobstruction. It has been speculated that the AVE-typeHVJ-liposome may reach interstitial fibroblasts byslipping between either tubular epithelial cells or papillaepithelial cells; gene transfer could then occur via thefusion activity of the HVJ-liposome.Gene transfer targeting interstitial fibroblasts might beexplained by the possibility that glycol-type sialic acids,the receptors for HVJ, are abundant in fibroblasts but rarein endothelial or epithelial cells (Tsujie et al, 2000).However, the precise mechanism remains to be elucidated.226

Gene Therapy and Molecular Biology Vol 7, page 227Table 3 Gene transfer via urinary tractCAII, carbonic anhydrase II; Nlac F, containing nuclear transport signal, which resulted in the specific and exclusive nuclear localizationof the protein, was inserted into the plasmid at the N terminus of the LacZ geneV. Direct injectionIntrarenal parenchymal injection of a plasmid DNA-Lipofectin complex did not result in gene transfer into thekidney (Lai et al, 1997). Direct injection caused focalinfiltration of inflammatory cells. On the other hand, theintrarenal parenchymal injection of naked plasmid DNAled to gene expression in the tubules (Lai et al, 1997). Thegene expression was limited to small areas near or deeperthan the injection site. The transfection efficiency of nakedplasmid DNA via intrarenal parenchymal injection ismuch less than that of liposome-mediated plasmid DNAtransfer via intrarenal pelvic or intrarenal arterialinjections.VI. SystemicAlthough the kidney can be targeted by a systemicintravenous route (tail vein injection), the gene transferefficiency is not sufficient for the specific transfection ofthe kidney. The plasmid DNA:cationic liposome(DOTMA:DOPE) complex becomes trapped in thepulmonary tissue and vascular endothelial cells (Zhu et al,1993), resulting in the highest expression in the lungs.Compared with plasmid DNA:cationic liposome 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP)complexes, liposome-polycation-DNA complexes (LPD),such as DOTAP-protamine-plasmid DNA complexes,offer better protection of the plasmid DNA fromenzymatic digestion and give higher gene expression inthe mouse kidney following tail vein injection (Li andHuang, 1997). When Li and Huang (1997) used aluciferase reporter gene with this technique, they foundgene expression in all the tissues examined, includinglungs, heart, spleen, liver, and kidney, with the highestexpression in the lungs.These organs can also be transfected byhydrodynamics-based naked plasmid DNA transfectionvia tail vein injection. In this case, the highest level oftransgene expression observed is in the liver (Liu et al,1999; Zhang et al, 1999; Maruyama et al, 2002b; Higuchiet al, 2003).Thus, the transgene expression that results from thesystemic injection of a gene is not confined to the kidney.VII. Future clinical applicationsAlthough the precise mechanisms of the genetransfer into cells induced by the nonviral techniques hasnot yet been clarified, recent progress in kidney-targetedgene transfer is promising for future clinical applications.High transfection efficiency and long-term transgeneexpression are required for clinical applications.Nephrotoxicity attributable to gene transfer is obviouslyundesirable. An ischemic period is caused by theinterruption of renal blood flow during the injection andthe following incubation time, in the case of sometechniques. The shortest ischemic duration possible isdesirable because of safety concerns. For clinical uses, thenonviral techniques should confine transgene expressionto the injected kidney, without aberrant expression innontarget organs. Urinary tract infection attributable togene transfer via the retrograde urinary tract route isundesirable.A. In vivo catheter-mediated genetherapyPutting the catheter technique to practical use, theabove-mentioned gene transfer techniques via threedifferent routes, renal artery anterogradely, renal veinretrogradely, the urinary tract (ureter or pelvis)retrogradely, can deliver therapeutic genes to the kidneyless invasively than techniques that do not use cathetersand therefore require an abdominal incision. Therefore, thedevelopment of catheters that are exclusively for genetransfer is key to the progress of catheter-mediated genetherapy. Similarly, ureterorenoscopes can be used todeliver therapeutic genes and specific adaptations tooptimize their use should be undertaken (Lai et al, 1998).B. Ex vivo gene therapy for kidneytransplantationTo obtain successful outcomes following kidneytransplantation, we need to overcome the common227

Maruyama et al: Kidney-targeted plasmid DNA transferproblems of post-transplant ischemic injury and rejection.Some of the above-described techniques (Tsujie et al,2001b; Maruyama et al, 2002a; Lan et al, 2003; Kita et al,2003) could be useful for ex vivo plasmid DNA deliveryduring kidney transplantation. The kidney to be used forgrafting is readily accessible for gene transfer. In addition,the transgene expression in the graft that is induced by exvivo nonviral techniques is, because of the nature of exvivo transfection, confined to the injected kidney.AcknowledgmentsThis work was supported by a grant from theMinistry of Education, Science, Sports and Culture.ReferencesAkami T, Arakawa K, Okamoto M, Akioka K, Fujiwara I, NakaiI, Mitsuo M, Tomita N, Kaneda Y, Tanaka K, and Oka T(1994) Introduction and expression of human CD59 gene inthe canine kidney. Transplant Proc 26, 1315-1317.Arai M, Wada A, Isaka Y, Akagi Y, Sugiura T, Miyazaki M,Moriyama T, Kaneda Y, Naruse K, Naruse M, OritaY, AndoA, Kamada T, Ueda N, and Imai E (1995) In vivotransfection of genes for renin and angiotensinogen into theglomerular cells induced phenotypic change of the mesangialcells and glomerular sclerosis. Biochem Biophys ResCommun 206, 525-532.Boletta A, Benigni A, Lutz J, Remuzzi G, Soria MR, andMonaco L (1997) Nonviral gene delivery to the rat kidneywith polyethylenimine. Hum Gene Ther 8, 1243-1251.Dworkin LD, Sun AM, and Brenner BM (2000) The renalcirculations. In Brenner & Rector’s The kidney.B.M.Brenner, ed. (W.B.Saunders, Philadelphia, USA) pp.277-318.Foglieni C, Bragonzi A, Cortese M, Cantu L, Boletta A,Chiossone I, Soria MR, and Monaco L (2000) Glomerularfiltration is required for transfection of proximal tubular cellsin the rat kidney following injection of DNA complexes intothe renal artery. Gene Ther 7, 279-285.Higuchi N, Maruyama H, Kuroda T, Kameda S, Iino N, KawachiH, Nishikawa Y, Hanawa H, Tahara H, Miyazaki J, andGejyo F (2003) Hydrodynamics-based delivery of the viralinterleukin-10 gene suppresses experimental crescenticglomerulonephritis in Wistar-Kyoto rats. Gene Ther 10,1297-1310.Isaka Y, Fujiwara Y, Ueda N, Kaneda Y, Kamada T, and Imai E(1993) Glomerulosclerosis induced by in vivo transfection oftransforming growth factor-beta or platelet-derived growthfactor gene into the rat kidney. J Clin Invest 92, 2597-2601.Kita J, Kobayashi E, Hishinuma A, and Kaneda Y (2003)Genetic modification of cold-preserved renal grafts usingHSP70 or bcl-2 HVJ-liposome method. TransplantImmunol 11, 7-14.Lai L-W, Moeckel GW, and Lien Y-HH (1997). Kidney-targetedliposome-mediated gene transfer in mice. Gene Ther 4, 426-431.Lai L-W, Chan DM, Erickson RP, HsuSJ, and Lien YH (1998)Correction of renal tubular acidosis in carbonic anhydrase IIdeficientmice with gene therapy. J Clin Invest 101, 1320-1325.Lan HY, Mu W, Tomita N, Huang XR, Li JH, Zhu HJ, MorishitaR, and Johnson RJ. (2003) Inhibition of renal fibrosis bygene transfer of inducible Smad7 using ultrasoundmicrobubblesystem in rat UUO model. J Am Soc Nephrol14, 1535-1548.Lemley, KV, and Kriz, W (1994) Structure and function of therenal vasculature with clinical and functional correlations. InRenal pathology. C.C. Tisher, and B.M. Brenner, eds.(J.B.Lippincott, Philadelphia, USA) pp. 981-1026.Li S, and Huang L (1997) In vivo gene transfer via intravenousadministration of cationic lipid-protamine-DNA (LPD)complexes. Gene Ther 4, 891-900.Liu F, Song Y, and Liu D (1999) Hydrodynamics-basedtransfection in animals by systemic administration of plasmidDNA. Gene Ther 6, 1258-1266.Madry H, Reszka R, Bohlender J, and Wagner J (2001) Efficacyof cationic liposome-mediated gene transfer to mesangialcells in vitro and in vivo. J Mol Med 79, 184-189.Maruyama H, Higuchi N, Nishikawa Y, Hirahara H, Iino N,Kameda S, Kawachi H, Yaoita E, Gejyo F, and Miyazaki J(2002a) Kidney-targeted naked DNA transfer by retrograderenal vein injection in rats. Hum Gene Ther 13, 455-468.Maruyama H, Higuchi N, Nishikawa Y, Kameda S, Iino N,Kazama JJ, Takahashi N, Sugawa M, Hanawa H, Tada N,Miyazaki J, and Gejyo F (2002b) High-level expression ofnaked DNA delivered to rat liver via tail vein injection. JGene Med 4, 333-341.Niwa H, Yamamura K, and Miyazaki J (1991) Efficient selectionfor high-expression transfectants with a novel eukaryoticvector. Gene 108, 193-199.Price J, Turner D, Cepko C (1987) Lineage analysis in thevertebrate nervous system by retrovirus-mediated genetransfer. Proc Natl Acad Sci USA 84, 156-160.Shimizu H, Maruyama S, Yuzawa Y, Kato T, Miki Y, Morita Y,Maruyama H, Egashira K, and Matsuo S (2003) Antimonocytechemoattractant protein-1 gene therapy attenuatesrenal injury induced by protein-overload proteinuria. J AmSoc Nephrol 14, 1496-1505.Takebe Y, Seiki M, Fujisawa J, Hoy P, Yokota K, Arai K,Yoshida M, and Arai N (1988) SRα promoter: an efficientand versatile mammalian cDNA expression systemcomposed of the simian virus 40 early promoter and the R-U5 segment of human T-cell leukemia virus type 1 longterminal repeat. Mol Cell Biol 8, 466-472.Tomita N, Higaki J, Morishita R, Kato K, Mikami H, Kaneda Y,and Ogihara T (1992) Direct in vivo gene introduction intorat kidney. Biochem Biophys Res Commun 186, 129-134.Tsujie M, Isaka Y, Ando Y, Akagi Y, Kaneda Y, Ueda N, ImaiE, and Hori M (2000) Gene transfer targeting interstitialfibroblasts by the artificial viral envelope-typehemagglutinating virus of Japan liposome method. KidneyInt 57, 1973-1980.Tsujie M., Isaka Y., Nakamura H., Kaneda Y., Imai E., and HoriM. (2001a) Prolonged transgene expression in glomeruliusing an EBV replicon vector system combined with HVJliposomes. Kidney Int 59,1390-1396.Tsujie M., Isaka Y., Nakamura H., Imai E., and Hori M (2001b)Electroporation-mediated gene transfer that targetsglomeruli. J Am Soc Nephrol 12, 949-954.Yamada T, Horiuchi M, Morishita R, Zhang L, Pratt RE, andDzau VJ (1995) In vivo identification of a negativeregulatory element in the mouse renin gene using direct genetransfer. J Clin Invest 96, 1230-1237.Zhang G, Budker V, and Wolff JA (1999) High levels of foreigngene expression in hepatocytes after tail vein injections ofnaked plasmid DNA. Hum Gene Ther 10, 1735-1737.Zhu N, Liggitt D, Liu Y, and Debs R (1993) Systemic geneexpression after intravenous DNA delivery into adult mice.Science 261, 209-211.228

Gene Therapy and Molecular Biology Vol 7, page 229Gene Ther Mol Biol Vol 7, 229-238, 2003Hepatocyte-targeted delivery of Sleeping Beautymediates efficient gene transfer in vivoResearch ArticleBetsy T. Kren, 1 Siddhartha S. Ghosh, 2,3 Cheryle L. Linehan, 1,4 Namita Roy-Chowdhury, 2,3 Perry B. Hackett, 4 Jayanta Roy-Chowdhury, 2,3 and Clifford J.Steer 1,4Departments of 1 Medicine and 4 Genetics, Cell Biology and Development, University of Minnesota Medical School,Minneapolis, MN 554552 Departments of Medicine and Molecular Genetics, and 3 Marion Bessin Liver Research Center, Albert Einstein College ofMedicine, Bronx, NY 10461.__________________________________________________________________________________*Correspondence: Clifford J. Steer, M.D., Department of Medicine, Mayo Mail Code 36, Mayo Building, Room A536, University ofMinnesota Medical School, 420 Delaware Street S.E., Minneapolis, MN 55455. Telephone (612) 624-6648; fax: (612) 625-5620, e-mail:steer001@tc.umn.eduKey words: asialoglycoprotein receptor, gene therapy, genomic integration, polyethyleneimine, Sleeping Beauty transposonAbbreviations: Sleeping Beauty (SB); green fluorescent protein (GFP); partial hepatectomy (PH); asialoglycoprotein receptor (ASGR);inverted repeats/direct repeats (IR/DRs); chicken β-actin/rabbit globin intron (CAGGS); elongation factor (EF)-1α; human embryonickidney (HEK293)Received: 22 September 2003; Revised: 19 November 2003;Accepted: 21 November 2003; electronically published: November 2003SummaryCurrently, most gene therapy studies utilize viral vectors that can potentially produce immunological and toxic sideeffects. To circumvent these limitations, we evaluated the efficiency of nonviral hepatocyte-targeted in vivo deliveryof plasmids that mediate stable genomic integration of transgenes via the Sleeping Beauty (SB) transposon system.We constructed plasmids that express a reporter green fluorescent protein (GFP) transposon and the SBtransposase, required for transgene insertion into genomic DNA, from either a single plasmid (cis) or two differentplasmids (trans). The constructs were compacted to an average diameter of < 50 nm with lactosylatedpolyethyleneimine, a polycation, for targeting to the hepatocyte asialoglycoprotein receptor. Intravenousadministration of the cis plasmid resulted in greater efficiency of transgene integration in mouse liver compared totransposase expression from a separate plasmid. Furthermore, by western blot analysis and fluorescencemicroscopy, delivery of the cis plasmid to rat livers resulted in transgene expression that persisted for months evenafter regeneration from partial hepatectomy. Southern blot analysis of the regenerated livers indicated that SBmediated genomic integration of the GFP transgene at random sites, and this correlated with disappearance of SBtransposase. In conclusion, receptor-mediated targeted delivery of a transposon system capable of transgeneintegration and stable expression provides an attractive alternative to viral vectors for gene therapy to the liver.I. IntroductionRecombinant viral vectors are the current mainstay ofgene therapy for inherited metabolic disorders (Kay et al,2001). However, clinical trials have achieved only modestsuccess, in part, because of the limitations set by viralvectors. For example, adenovirus-based vectors do notintegrate into host chromosomes (Harui et al, 1999) andtheir immunogenicity precludes repeated gene transfer.Furthermore, in contrast to the highly efficient genetransfer to livers of laboratory animals, clinical trials withadenovirus have produced low levels of transgeneexpression in human liver (Raper et al, 2002).Recombinant adeno-associated viral vectors also do notintegrate efficiently in liver (Hillgenberg et al, 2001),resulting in progressive loss of the episomal DNA (Nakaiet al, 2001; Ehrhardt and Kay, 2002). Moreover, the lowlevel integration appears to occur preferentially into activegenes and is associated with chromosomal deletions at thesite (Nakai et al, 2003). Although oncoretroviral vectorsintegrate into the host genome, the process is veryinefficient in non-replicating cells such as hepatocytes invivo (Kalpana, 1999). Lentiviruses, which appear topartially overcome this (Pfeifer et al, 2001; Follenzi et al,2002), are difficult to generate in quantities adequate for229

Kren et al: Hepatocyte-targeted delivery of Sleeping Beautyhuman therapy. Moreover, despite removal of the viralgenes, potential safety concerns persist. Thus,development of efficient non-viral methods for long-termgene transfer would be important for gene therapy.Plasmid-based non-viral gene transfer has beenattempted by direct injection into liver, with limited levelsof transgene expression. A “hydrodynamic” method thatrelies on rapidly injecting plasmids in large volumesintravenously has been used to transfer nucleic acids to thelivers of rodents (Zhang et al, 1999; Maruyama et al,2002). An elegant alternative employs targeted delivery ofnucleic acids to hepatocytes via the asialoglycoproteinreceptor (ASGR) (Wu and Wu, 1988). Unfortunately, thedelivery of naked plasmids to hepatocytes results in littleor no integration of the transferred DNA into the hostgenome (Zhang et al, 1999; Maruyama et al, 2002). Apotential solution to this problem arises from the discoverythat the Sleeping Beauty (SB) transposon systemdeveloped from fish can mediate the transposition of DNAinto chromosomes for a broad range of vertebrates,including humans (Ivics et al, 1997; Izsvák et al, 2000).The SB transposon system functions by a cut-andpastemechanism catalyzed by binding of the SBtransposase to the inverted repeats/direct repeats (IR/DRs)of the transposons. It excises the transposon at the outsideends of the IR/DRs and inserts the element into a new TAdinucleotide site. The hydrodynamic delivery of twoseparate plasmids in mice, one expressing SB transposaseand another comprising a transgene flanked by theIR/DRs, resulted in long-term gene expression in the livereven after partial hepatectomy (PH) (Yant et al, 2000,2002; Montini et al, 2002). This gene transfer methodreproducibly transduced up to 5% of hepatocytes.However, although useful for delivery of naked DNA inmice (Nakai et al, 2001; Yant et al, 2000, 2002; Montini etal, 2002), and rats (Maruyama et al, 2002), the rapidhydrodynamic delivery of large volumes may poseconsiderable restrictions for clinical use.In this study, we determined the efficiency oftransposition in liver using a single plasmid, containingboth a transposon with a transgene and SB transposase,targeted for delivery to hepatocytes via the ASGR. Ourresults indicated that the SB complex efficiently deliveredgreen fluorescent protein (GFP) genes in vivo tohepatocytes of mice and rats. Long-term gene expressionoccurred only in animals that received both the transposonand transposase. In addition, transposition was increasedwhen the GFP transgene and SB were delivered in cis,rather than in trans as separate plasmids.II. Materials and methodsA. Construction of transposon vectorsTwo different GFP reporter transposons were constructedusing either the elongation factor (EF)-1α promoter (Johnson andKrieg, 1994) (pT/GFP), or the hybrid CMV enhancer chicken β-actin/rabbit globin intron (CAGGS) promoter (Okabe et al, 1997)(pT2/GFP). pT/GFP was flanked by the original IR/DRs (Ivics etal, 1997) while pT2/GFP, constructed by cloning the EcoR V-Sma I coding sequence of pT/GFP into the EcoR I site of theCAGGS vector, was flanked by alternate IR/DRs (Cui et al,2002). For the cis SB constructs, the 2 kb SB10 transgene wasremoved from pCMVSB10 using EcoR I and Xba I (Ivics et al,1997) and inserted outside the IR/DRs at either the unique Nar I(pT/GFP//SB10) or Xho I site (pT2/GFP//SB10). ThepT2/CAGGS//DsRed2 (pT2/DsRed2//SB10) construct containsthe DsRed2 fluorescent protein gene (BD Biosciences Clonetech,Palo Alto, CA) driven by the CAGGS promoter and the same 2kb CMVSB10 transgene inserted in the unique BsaA I site. Allplasmids were prepared using Qiagen TM (Valencia, CA)endotoxin free plasmid isolation kits according to standardprotocols.B. Cell culture, transfection and cloning oftransduced cellsTo validate transposase expression, primary rat hepatocytesor HuH-7 cells (Bandyopadhyay et al, 1998) at ~ 40% confluentwere transfected with 1 µg of the cis vector constructs as well asthe initial pCMVSB10 plasmid using the same L-PEI amine(N):DNA phosphate (P) ratios as in vivo. Cells were harvested byscraping hepatocytes 48 h or HuH-7 cells 2 to 10 days aftertransfection. HEK293 cells seeded on a 10 cm 2 plate weretransfected at ~ 60% confluence with 2 µg of the cis pT2/DsRed2construct using Lipofectamine TM (Invitrogen), After 72 h, thecells were transferred to a 75 cm 2 plate and grown to confluence.Subsequently, the cells were split 1:3 and passaged 4 times.Finally, ~ 100 cells from the fourth passage were plated on a 75cm 2 plate. The positive clones were picked after a week using 8mm cloning cylinders (Bellco Glass, Inc., Vineland, NJ) andcultured in DMEM with 10% fetal calf serum.C. Electron microscopyThe size of the cis transposon:L-PEI complexes wasdetermined by electron microscopy. The complexes in 5%dextrose were applied onto glow-discharged formvar-carboncoated 300 mesh grids (Polysciences Inc., Warrington, PA) for ~2 min. PEI complexes were negatively stained with aqueous 1%uranyl acetate and were visualized using a JEOL100-CX electronmicroscope.D. In vivo administrationAll animal studies were reviewed and approved by theInstitutional Animal Care and Use Committee at the Universityof Minnesota and Albert Einstein College of Medicine accordingto the NIH Guidelines for Animal Care. The plasmids werecomplexed using primary amine lactosylated 25 kDa branchedPEI (L-25) (Aldrich, Milwaukee, WI) (Kren et al, 2002) and 10kDa branched PEI (L-10) (Polysciences, Inc.) at a ratio of 1.5:1(L-25:10) in 5% dextrose. The amine (N) to DNA phosphate (P)ratio was 6:1 (Bandyopadhyay et al, 1998). C57B16 gus -/- mice(10 g) received a single tail vein injection of 400 µl containing2.5 or 5 µg of pCMVSB10 and/or pT/GFP, or cis pT/GFP//SB10.Animals were sacrificed at 1, 2 and 8 weeks post-injection andliver tissue removed for analysis. For rats, the complexes wereprepared identically except the concentration was increased to100 µg/ml of transposons. The ~ 200 g Wistar rats received 500µg/kg bw as a single bolus injection into the tail vein. Livertissue was sampled at 1, 2 or 4 days by biopsy. PHs of 70%(Higgins and Anderson, 1931) were performed at 1, 2 or 3 weeksafter injection. The animals were sacrificed at least 2 weeks post-PH and liver tissue removed for analysis.E. Protein detectionTissue for microscopic analysis was fixed in 4%paraformaldhyde in PBS, pH 7.4 at 4°C for 1 h prior to OCT.230

Gene Therapy and Molecular Biology Vol 7, page 231Frozen sections of 6 µm were viewed using a Nikon, Diaphot(Melville, NY) fluorescent microscope or post-fixed for 10 minprior to examination with a BioRad MRC1000 ConfocalMicroscope (Hercules, CA). For western blot analysis, 100 to150 µg protein/lane of a 10% (w/vol) homogenate of either tissueor cells in 0.25 M Tris acetate, pH 7.8, 0.25 M sucrose, 0.2 mMEDTA and complete EDTA-free protease inhibitor cocktail(Roche Molecular Biochemicals, Indianapolis, IN) was separatedby SDS 10% PAGE, and electrophoretically transferred tonitrocellulose membranes. GFP and SB were detected by ECL(Pierce Super Signal, Rockford, IL) with monoclonal anti-GFP(SC-9996; Santa Cruz Biotechnology, Santa Cruz, CA) andrabbit polyclonal anti-SB (Geurts et al, 2003), respectively (Krenet al, 1999).F. Southern blot and PCR analysisDNA for Southern blots and PCR was isolated from frozentissue using DNAzol, High Pure PCR template kit (RocheMolecular Biochemicals) or a DNAeasy Tissue kit (Qiagen)according to the manufacturers’ protocol. For plasmid copynumber, samples were spiked with known amounts of pT/GFPprior to extraction of the DNA. For Southern blots, the sampleswere digested with Afl II or Bsa I and incubated with a 32 P-labeled 757 bp EcoR V Sma I GFP probe isolated from pT/GFP.To detect genomic transposition of GFP, DNA was isolated 4weeks post-PH and digested with EcoR V or EcoR V and Sma I,prior to incubation with the 32 P-labeled probe. For PCR, primersF5 GGTGATGTTAATGGGCACA and B3GGGATCTTTCGAAAGGGCA were used to amplify a 535 bpregion of the GFP gene using 95°C 5 min (54°C 20 sec, 72°C 45sec, 95°C 45 sec) x 40 cycles with Expand Hi-Fidelitypolymerase (Roche Molecular Biochemicals). Under the sameconditions, primers SBF GGACCACGCAGCCGTCATAC andSBR CCTGTTTCCTCCAGCATCTTCAC amplified a 136 bpregion of the SB gene; and primers ApoBFCGTGGCTCCAGCATTCTA and ApoBRTCACCAGTCATTTCTGCCTTG were used to amplify a 72 bpregion of the apoB gene. The PCR products were analyzed using1% agarose gels and visualized using ethidium bromide stainingand UV light. Quantitation was performed using NIH image1.62; and statistical significance determined by two tailedunequal variance T-tests.transfected cells showed an accelerated loss compared topCMVSB10 (data not shown).B. Transfection and integration ofpT2/DsRed2//SB10 in HEK293 cellsTo evaluate a second reporter construct in a differentcell line, human embryonic kidney HEK293 cells weretransfected with the cis transposon pT2/DsRed2//SB10 toexpress both SB and the DsRed2 protein. After 48 h, ~30% of the cells expressed DsRed2 (Figure 2C, A-C).Following 4 passages, 6 DsRed2-positive clones werederived from single cells by dilutional cloning. The clonesremained homogeneously positive for DsRed2, indicativeof host genome integration (Figure 2C, D-F). Forcomparison, we transfected the HEK293 cells with aplasmid containing the DsRed2 transposon but without theSB transgene (pT2/DsRed2). This also resulted in theexpression of DsRed2 in ~ 30% of the cells at 48 h.However, the transgene expression disappeared after 1 or2 passages (data not shown).III. ResultsA. Co-expression of the SB10 Transgenein cisWe constructed cis plasmids carrying both thetransposon and SB transposase (Figure 1). Primary rathepatocytes were transfected with the original pCMVSB10plasmid (Ivics et al, 1997) (SB expression only), pT/GFP(no SB), or the cis plasmid, pT/GFP//SB10. Both SBconstructs resulted in similar levels of transposase proteinexpression after 2 days (Figure 2A). We then transfectedplasmids into HuH-7 cells to determine the duration of SBexpression for pCMVSB10 and pT/GFP//SB10. There wasno difference in the initial expression of SB aftertransfection of the two plasmids (Figure 2B), with peaklevels being reached in 2 days. Interestingly, by day 5 adecrease in SB expression was observed in cells thatreceived the cis construct relative to pCMVSB10 alone(Figure 2B, lanes 4, 5). At 10 days, SB expression wasundetectable in both cases. By PCR analysis, theabundance of SB coding sequence in the cis pT/GFP//SB10Figure 1 The cis and trans Sleeping Beauty vector systems. (A)To construct pT/GFP//SB10, CMVSB10 transposase (right) wasinserted at a Nar I site outside the transposon EF-1α drivenpT/GFP (left). (B) The CAGGS driven GFP (pT2/GFP) reportertransposon and cis counterpart pT2/GFP//SB10, andpT2/DsRed2//SB10 are shown. The location and orientation ofthe CMV driven transposase expression cassette are indicatedrelative to the reporter transgene as well as the direction oftranscription (arrows). Amp, bla gene for plasmid selection;black regions, introns; EF-1α, elongation factor-1αenhancer/promoter; CMV, cytomegalovirus immediate-earlygene promoter; p(A), polyadenylation signal; CAG, hybrid CMVenhancer, chicken β-actin/rabbit globin intron (CAGGS)promoter; R, IR/DRs.231

Kren et al: Hepatocyte-targeted delivery of Sleeping BeautyFigure 2 Characterization of the cis SB10 transposons. (A) Western blot analysis of SB10 transposase expression in primary rathepatocytes. Hepatocytes were transiently transfected with pT/GFP (lane 1), pCMVSB10 (lanes 2) or pT/GFP//SB10 (lanes 3) for 48 h.Total protein from duplicate transfections was analyzed by immunoblot using anti-SB10 rabbit polyclonal antibodies. Transposaseexpression was detectable only in cells that received plasmids encoding SB10 (lanes 2,3). The transposase produced from transientexpression co-migrated with purified SB10 protein (lane 4) at its predicted molecular size of 39.5 kDa (Geurts et al., 2003). (B) Westernblot analysis of SB10 transposase expression in HuH-7 cells. HuH-7 cells were transiently transfected with pT/GFP (lane 1) pCMVSB10(lanes 2,4,6) or pT/GFP//SB10 (lanes 3,5,7). Total protein from the cultures harvested at the indicated times was analyzed byimmunoblot using anti-SB10 rabbit polyclonal antibodies. (C) Transposition of DsRed2 genes into HEK293 cells. Cells transientlytransfected with pT2/DsRed2//SB10 were examined by (a) phase contrast and (b) fluorescence microscopy. (c) The overlay of (a) and (b)indicate that ~ 30% of the cells expressed DsRed2 fluorescent protein. Clonal isolation of DsRed2 positive cells following limitingdilution and expansion visualized by (d) phase contrast, (e) fluorescence and (f) both. Original magnification x 20. (D) Transmissionelectron microscopy of pT/GFP//SB10:L-PEI (lactosylated polyethyleneimine) complexes. A representative micrograph of negativelystained L-PEI:pT/GFP//SB10 complexes formed at a 1:6 (N:P) ratio in 5% dextrose showing their small size and monodisperse nature.Bar, 100 nm. N, PEI amine; P, DNA phosphate.C. Size determination of the plasmidvectorcomplexes.We determined the size of the transposon constructscomplexed with branched L-PEI to insure that they wereable to pass through the ~ 100 nm fenestrae into the Spaceof Disse (Hara et al, 1997). In 5% dextrose, thepT/GFP//SB10 construct formed monodisperse particleswith an average diameter of ~ 50 nm at a 6:1 PEI amine toDNA phosphate ratio (Figure 2D).D. Injection of GFP reporter transposonsinto mice and ratsMice received a single tail vein injection of eitherpCMVSB10, pT/GFP, equal amounts of pCMVSB10 andpT/GFP (trans), or cis pT/GFP//SB10. Fluorescencemicroscopy of liver sections showed GFP expression in allthe animals that received the GFP transposon after 1 week(Figure 3, A-D). In contrast, only those mice that alsoreceived the transposase in either cis or trans expressedGFP at 8 weeks (Figure 3, E-H).Adult rats also received either pT/GFP orpT/GFP//SB10 complexed with L-PEI by a single tail veininjection. We then performed 70% PH 2 weeks postinjectionto induce hepatocyte replication. Fluorescencemicroscopy of the removed tissue showed that 10-35% ofthe hepatocytes expressed GFP in rats that received eitherpT/GFP or cis pT/GFP//SB10 (Figure 3, I, J, M, N). After3 weeks, the fully regenerated livers were harvested andanalyzed by fluorescence microscopy. Rats that receivedthe cis pT/GFP//SB10 showed GFP expression in single orin small clusters of hepatocytes at the same frequency aswas initially observed (Figure 3, O, P). In contrast, < 1%of the hepatocytes remained positive after liverregeneration in rats that received pT/GFP (Figure 3, K,L).GFP protein was detected by western blot analysis inmouse liver homogenates 1 and 2 weeks after injection ofpT/GFP or cis pT/GFP//SB10 (Figure 4A). However, onlymice that received SB in trans or cis continued to expressGFP after 8 weeks. By 2 weeks post-injection, GFPexpression was 28% and 8% greater in animals that hadreceived cis and trans constructs, respectively, relative topT/GFP alone. At 8 weeks, GFP expression in the transgroup was ~ 2-fold less than that observed in the cisanimals (p < 0.05).232

Gene Therapy and Molecular Biology Vol 7, page 233Figure 3 Expression of GFP fluorescent protein in rodent liver. Representative sections of liver tissue isolated 1 (a-d) or 8 (e-h) weeksafter tail vein injection of 5 µg of (a,e) pCMVSB10, (b,f) pT/GFP, (c,g) 5 µg each of pT/GFP and pCMVSB10, or (d,h) 5 µg ofpT/GFP//SB10 complexed with L-PEI. Original magnification x 20. Representative sections of resected rat liver 2 weeks post-injection(i,j,m,n) and regenerated liver 3 weeks post-PH (k,l,o,p) from animals injected with pT/GFP (i-l) or pT/GFP//SB10 (m-p) in complexwith L-PEI; (i,k,m) phase contrast. Original magnification i-o x 4; p x 40.Figure 4 Immunoblot analysis of GFP proteinexpression. (A) Total protein was isolated frommouse livers 1 week (2-7), 2 weeks (8-10) and 8weeks (11-13) after tail vein injection with 5%dextrose (lane 2), L-PEI only (lane 3), 5_g of L-PEI complexed pCMVSB10 (lane 4), pT/GFP(lanes 5,8,11), 5 µg each of pT/GFP andpCMVSB10 (lanes 6,9,12) or 5 µg ofpT/GFP//SB10 (lanes 7,10,12). Immunopositive30 kDa GFP (lane 1) was identified using amonoclonal anti-GFP antibody. (B) Immunoblotanalysis of using anti-GFP monoclonal antibodyof total protein isolated from rat liver 4 daysafter injection of 100 µg of L-PEI:pT/GFP//SB10 (lanes 2-4) or L-PEI:pTGFP(lanes 8-10). Six weeks after injection and 3weeks post-PH, GFP expression in theregenerated livers of the same animals showedsubstantial GFP expression only in rats that hadreceived pT/GFP//SB10 (lanes 5-7) rather thanthe transposon only (lanes 11-13). A nonreactiveprotein and alternate plasmid vectorwas used as control (lane 1).233

Kren et al: Hepatocyte-targeted delivery of Sleeping BeautyIn rats, the GFP levels observed 4 days after injection withGFP transposon alone or the cis construct were similar(Figure 4B). Only the animals that received cispT/GFP//SB10 expressed high levels of GFP after PH, andthose were unchanged from the original livers (lanes 2-7).E. Integration into the host genomeWe extracted total DNA from mouse livers harvested1 or 8 weeks after injection. Ampicillin-resistant coloniesfrom transformed electrocompetent E. coli were recoveredwith DNA isolated from the 1 week samples, but not after8 weeks (data not shown). Southern blot analysis showedthat non-integrated plasmids were present at less than asingle copy per cell after 1 week (Figure 5A), and no freeplasmid was detectable at 8 weeks. In mice treated withpT/GFP alone, PCR amplification of the GFP codingregion showed the persistence of a small but detectableamount of the transgene at 8 weeks (Figure 5B, lane 2),suggesting that spontaneous integration of plasmidsoccurred at a very low level, as previously reported(Montini et al, 2002; Yant et al, 2000; 2002). Livers fromcontrol mice that received pCMVSB10 alone showed noGFP amplicons (lane 7). In contrast, mice that received theSB-transgene either in cis or trans showed persistent GFPcoding sequences by PCR. Samples with cis constructsgenerated ~ 45% (p < 0.05) more amplicons (lanes 4,6)than those with pT/GFP plus pCMVSB10 in trans (lanes3,5), and correlated with GFP expression by confocalmicroscopy and western blot analysis. SemiquantitativePCR using apoB as a genomic control indicated that genetransfer efficiency of the cis construct at 8 weeks was ~ 1copy per genome (Figure 5C, lane 5). Delivery in transresulted in significantly lower (p < 0.05) GFP copynumber (lane 4).We also examined the persistence of the SB codingregion by semiquantitative PCR in mice that received thetransposase either alone (Figure 5D, lanes 1,4), in cis(lanes 3,6) or in trans (lanes 2,5). No difference in SBamplicon levels between groups was observed 1 weekpost-injection (lanes 1-3). However, by 2 weeks a greaterdecrease in SB coding sequences was seen in animals thathad received the cis construct (lane 6), compared withthose that had received pCMVSB10 alone or in trans(lanes 4,5).We then examined the loss of the plasmid andpersistence of the GFP transgene in regenerating rat liverpost-PH by PCR amplification. There was considerableloss of the plasmid vectors in the first week (Figure 6A,lanes 2-4) but the GFP coding sequence persisted in thegenomic DNA of animals that received cispT2/GFP//SB10 (lanes 5,7-9). Those rats that were givenpT2/GFP alone retained no detectable GFP DNA by 3weeks post-PH (lane 6). The data also suggested that theposition and orientation of the SB10 expression cassettemight influence the efficiency of transposition. Lowertransgene levels were observed in animals that receivedthe cis constructs in which transcription of the SB10 andGFP genes were in opposite directions (lane 5).Figure 5 Analysis of the GFP coding sequence in mousegenomic DNA. (A) Southern blot detection of plasmidpersistence in total DNA isolated from mouse livers 1 or 8 weeksafter injection with L-PEI complexed with pT/GFP (lane 1);pT/GFP and pCMVSB10 (lanes 2,3), or pT/GFP//SB10 (lanes4,5). The predicted plasmid size is indicated at left (arrow). (B)PCR amplification was used to detect GFP sequence in mouseDNA isolated 8 weeks post-injection of 5 µg of pT/GFP (lane 2),2.5 µg (lane 3) or 5 _g (lane 5) each of pT/GFP and pCMVSB10,or 2.5 µg (lane 4) or 5 µg (lane 6) of pT/GFP//SB10; SB10control liver (lane 7), and no DNA (lane 8). A 500 bp DNAstandard (lane 1) and the predicted 535 bp amplification productare indicated (arrow). (C) Semiquantiative PCR was used todetermine the efficiency of GFP transfer 8 weeks post-injectionof either 5 µg pCMVSB10 (lane 2), pT/GFP (lane 3), 5 µg eachof pT/GFP and pCMVSB10 (lane 4), or 5 µg of pT/GFP//SB10(lane 5); and no DNA (lane 6). The 535 bp GFP and 72 bp apoBamplicons are indicated at right (arrows). DNA 100 to 600 bpladder (lane 1). (D) PCR analysis was used to detect SB10 genein mouse DNA isolated 1 week (lanes 1-3) or 2 weeks (lanes 4-7)post-injection of 5 µg of pCMVSB10 (lanes 1,4), 5 µg each ofpT/GFP and pCMVSB10 (lanes 2,5), 5 µg of pT/GFP//SB10(lanes 3,6) or pT/GFP (lane 7); and no DNA (lane 8). The 136 bpamplification product is indicated at right (arrow). DNAstandards of 100 and 200 bp (lane 9).234

Gene Therapy and Molecular Biology Vol 7, page 235We also investigated the persistence of pT/GFP andcis pT/GFP//SB10 in rats. By Southern blot analysis, therewas a significant loss of plasmids by 1 week (Figure 6B).Interestingly, the loss of plasmid appeared to be morerapid in the animals that received cis constructs suggestingthat excision of the transposon might accelerate vectordegradation. We did additional Southern blot analysisusing a GFP probe to compare the transgene abundance inrats that received either pT/GFP or pT/GFP//SB10. At 4days post-injection, GFP was undetectable in the highmolecular weight region in DNA samples from rats thatreceived pT/GFP (Figure 6C, lanes 1-3). In contrast,pT/GFP//SB10 delivery showed high molecular weightreactivity, consistent with integration of the transgene(lanes 4-6). As expected, DNA from regenerated livers ofpT/GFP rats at 6 weeks did not contain detectable GFPsequence (lanes 7-9). In rats that received the cis construct,GFP transgene levels remained essentially unchangedfrom the original samples (lanes 10-12). Southern analysisof DNA extracted from regenerated livers of rats that hadreceived cis pT/GFP//SB10 using a GFP probe showedthat the transgene was detectable only in the highmolecular weight DNA band (Figure 6D, lane 3). Whendigested with EcoR V, hybridization with the GFP probegenerated a smear consistent with a large number ofdifferent integration sites (lane 4). This finding alsoexcluded the presence of concatemers of episomallinearized plasmid DNA (Chen et al, 2001) or theintegration of plasmid concatemers, both of which wouldhave generated more distinct bands. The GFP sequencewas released from the integrated transposon after digestionwith both EcoR V and Sma I (lane 5).To distinguish between spontaneous integration andSB-mediated transposition of the GFP sequences, weextracted total DNA from post-PH regenerated livers ofrats that had received cis pT/GFP//SB10. PCR wasperformed using two sets of equal size amplimers of (a)both sense and antisense primers corresponding to thecoding region of GFP; and (b) a sense primercorresponding to the plasmid sequence immediately 5’ tothe 5’ DR and an antisense primer corresponding to theGFP coding region.Figure 6 GFP coding sequence analysis in rat genomic DNA. (A) PCR amplification of GFP transgene in rat liver pre- and post-PH.Animals received either pT2/GFP (lane 6), pT2/GFP//SB10A (lanes 2-5), or pT2/GFP//SB10B (lanes 7-9) and liver tissue was removedfor DNA isolation and PCR amplification of GFP. Standard DNA ladder (lane 1); DNA from livers 24, 48 and 96 h, respectively, postinjectionof pT2/GFP//SB10A (lanes 2-4); DNA from regenerated liver 3 weeks post-PH (lanes 5,6,8); DNA isolated 1 week (lane 7) or 6months (lane 9) post-PH, after injection of pT2/GFP//SB10B; control rat liver DNA (lane 10). The 535 bp GFP amplicon is indicated atright (arrow). (B) Southern blot analysis of plasmid disappearance in rat liver. At 24 h and 1 week after tail vein injection the transposonplasmid:L-PEI complexes, liver tissue was removed and total DNA isolated. The predicted size of the plasmid bands after Afl II (2.8 kb)or Bsa I (1.8 kb) digestion is indicted at left (arrows). (C) Southern blot analysis of the integrated GFP transgene (arrow) from genomicDNA of rat livers after PH. Liver lobes were resected 4 days after injection from 6 different rats that received only pT/GFP (lanes 1-3) orpT/GFP//SB10 (lanes 4-6), and 6 weeks post-PH for pT/GFP (lanes 7-9) or pT/GFP//SB10 (lanes 10-12). (D) Representative Southernblot (n > 3 for each of the groups) of DNA isolated from regenerated livers of rats that received cis pT/GFP//SB10. DNA standards (lane1); undigested DNA from liver treated with an alternate plasmid vector (lane 2); the cis construct in a regenerated liver (lane 3). DNAfrom the regenerated liver was digested with EcoR V alone (lane 4) or both EcoR V and Sma I (lane 5) to release the GFP codingsequence (arrow).235

Kren et al: Hepatocyte-targeted delivery of Sleeping BeautyAn ampicon of predicted size was obtained when bothprimers corresponded to the GFP coding region, but noamplification product was observed when the sense primerwas upstream to the direct repeat (data not shown). Theresult indicated that spontaneous integration of the plasmidwas rare or absent, strongly suggesting that the integrationhad occurred at the transposon DR.IV. DiscussionWe have shown that a single cis transposon plasmidthat expresses SB and carries a transgene can effectivelypromote long-term gene expression in the liver of miceand rats. Using hepatocyte-targeted in vivo delivery, thecis transposon system was almost 2-fold more efficient byPCR and western blot analysis than the SB transposasedelivered in trans. This finding is in contrast to a recentstudy using hydrodynamic delivery of a cis SB transposonin a murine model of tyrosinemia type I (Montini et al,2002). The authors reported reduced transposition usingthe cis construct, and concluded that this had resulted fromoverproduction of the transposase in mouse liver (Yant etal, 2000, 2002; Montini et al, 2002). Therefore, wecompared the efficiency of the cis and the trans systemsby using equal amounts of the transposase and transposonplasmid vectors when co-delivering the two plasmids intrans. In fact, both systems produced similar levels oftransposase protein immediately after transfection,suggesting that inhibition by transposase overproductionwould be equivalent. In contrast, other mouse studies(Yant et al, 2000, 2002; Montini et al, 2002) used a 1 to 25ratio of transposase construct to transposon. Interestingly,the optimal transposon/transposase ratio appears to beinfluenced by the amount of transposon (Geurts et al,2003). The lower doses were optimal at a 1:3 transposonto transposase construct ratio, while increasing thetransposon levels 5-fold decreased the optimal ratio to1:0.2. Thus, the reduced amounts of transposon in ourstudy may, in part, account for the observed increasedtransposition in cis. Additionally, the more rapid loss ofthe cis construct may have reduced levels of transposase,thereby increasing transposition. By targeted delivery, thecis plasmid showed greater efficacy for GFP transpositionin mice. A potential advantage of using the cis system isthat the transposition can result in self-destruction of thecis plasmid, eliminating the possibility of repeatedtransposition from the effect of any persisting SBexpression.In the present study, the DNA delivery systemprovides a potentially useful application to human trials. Inprevious studies, naked plasmids were delivered to theliver by rapid high volume intravenous administration thatcauses transient congestive heart failure and hepatic stasis,thereby enhancing DNA uptake by liver cells (Budker etal, 2000). This method, although useful in animal studies,is unlikely to find clinical application. In contrast, weachieved targeted delivery of DNA to the liver viahepatocyte-specific ASGR-mediated endocytosis (Wu andWu, 1998; Wu et al, 2002). The fate of plasmid DNAdelivered to the murine liver by the hydrodynamic methoddiffers dramatically from that with PEI (Oh et al, 2001).Using the branched 25 kDa polycation, 50% of theplasmid initially delivered to the liver was still present upto 10 days after transfer. In contrast, 50% of the nakedplasmid DNA delivered to the liver was lost within 15 minof administration, and there was no detectable plasmid at 3days. Although delivery of branched PEI to the lung hasbeen associated with a systemic immune response(Regnstrom et al, 2003), branched PEI-DNA complexesshowed no significant liver toxicity (Oh et al, 2001).Moreover, it efficiently transfects quiescent cells such ashepatocytes (Pollard et al, 1998), protects the DNA fromnuclease degradation (Boussif et al, 1995), and promotesefficient endosomal disruption (Behr, 1997). L-PEI mayalso enhance nuclear uptake of DNA by binding to alectin-like protein with galactose specificity in the nuclearpore complex (Klink et al, 2001). Finally, in contrast tocationic lipids, PEI does not appear to inhibit transgeneexpression (Pollard et al, 1998). Safety, efficacy andhepatocyte-specificity of this endocytosis-based DNAdelivery system makes it attractive for potential use ingene therapy.This study is the first report of transposon-mediatedgene transfer in a mammal other than the mouse. In the ratliver, a single dose of the cis transposon, or pT/GFP aloneresulted in stable transgene expression, althoughsignificantly less homogeneous than in mice. For the cisplasmid, the expression observed in the intact liver wassimilar to that after regeneration post-70% PH. In contrastto the results reported with transposon delivery to themouse liver using adenovectors, PH in the rats markedlyreduced the transgene content and expression in theanimals that received pT/GFP alone (Yant et al, 2002).One explanation for this finding is the differentregenerative response of hepatocytes post-PH in the twospecies (Fausto, 2000; Higgins and Anderson, 1931). Also,there is increased persistence of the adenoviral vectorsrelative to plasmids during cell cycling in vivo (Ehrhardt etal, 2003). Finally, the method of DNA delivery may havecontributed to the observed differences.The finding of small clusters of cells expressing GFPafter PH in the cis transposon rats suggested that theintegration event preceded hepatocyte replication.Integration of the transgene was confirmed by Southernblot analysis and like the mouse studies (Dupuy et al,2001; Fischer et al, 2001; Horie et al, 2001; Yant et al,2000, 2002) SB-mediated gene transfer in rats alsooccurred randomly within the genome. Interestingly, onaverage a single copy of the transposon was observed perdiploid liver genome when clonal selection for transgeneexpression occurs in vivo (Montini et al, 2002), yet singlecopies of a transposon were not associated with transgeneexpression in other in vivo mouse systems (Dupuy et al,2002; Horie et al, 2001). Thus, the observed transgeneexpression may underestimate the overall transpositionfrequency. Expression of randomly inserted transgenes canbe variable because of the known positional effects (Ivicset al, 1997; Izsvák et al, 2000; Yant et al, 2000, 2002;Dupuy et al, 2001, 2002; Horie et al, 2001; Montini et al,2002). Insertion of insulator sequences flanking thetransgene carried by transposons might abrogate thepositional variation of transgene expression and time-236

Gene Therapy and Molecular Biology Vol 7, page 237related gene silencing (Pikaart et al, 1998). The efficiencyof transposition by the cis transposon will most likely beincreased using different promoters to regulate transposaseexpression (Mikkelsen et al, 2003).In summary, our data indicate that by using areceptor-mediated DNA delivery system and equivalentinitial levels of transposase expression, the cis delivery oftransposons is more efficient than trans for transgeneintegration into the liver. Furthermore, we havedemonstrated that the combination of a cis construct and anonviral DNA delivery system could achieve stabletransgene expression at levels required to potentially treatmany inherited metabolic disorders of the liver. SBpromises to play an important role in the gene therapy ofhuman genetic diseases.AcknowledgmentsWe thank Phillip Y.P. Wong, L. Xiaoming Ma, JoelFrandsen and Stefan Kren for excellent technicalassistance. This work was supported by National Institutesof Health grants P01 HD32652 to B.T.K. and P.B.H.,RO1-DK46057 to J.RC., and P01-HL65578 and P01-HL55552 to C.J.S.ReferencesBandyopadhyay P, Kren BT, Ma X and Steer CJ (1998)Enhanced gene transfer into HuH-7 cells and primary rathepatocytes using targeted liposomes and polyethylenimine.BioTechniques 25, 282-292.Behr JP (1997) The proton sponge: a trick to enter cells theviruses did not exploit. Chimia 51, 34-36.Boussif O, Lezoualc'h F, Zanta MA, Mergny MD, Scherman D,Demeneix B and Behr JP (1995) A versatile vector for geneand oligonucleotide transfer into cells in culture and in vivo:polyethylenimine. Proc Natl Acad Sci USA 92, 7297-7301.Budker V, Budker T, Zhang G, Subbotin V, Loomis A and WolffJA (2000) Hypothesis: naked plasmid DNA is taken up bycells in vivo by a receptor-mediated process. J Gene Med 2,76-88.Chen ZY, Yant SR, He CY, Meuse L, Shen S and Kay MA(2001) Linear DNAs concatemerize in vivo and result insustained transgene expression in mouse liver. Mol Ther 3,403-410.Cui Z, Geurts AM, Liu G, Kaufman CD and Hackett PB (2002)Structure-function analysis of the inverted terminal repeats ofthe Sleeping Beauty transposon. J Mol Biol 318, 1221-1235.Dupuy AJ, Clark K, Carlson CM, Fritz S, Davidson AE, MarkleyKM, Finley K, Fletcher CF, Ekker SC, Hackett PB, Horn Sand Largaespada DA (2002) Mammalian germ-linetransgenesis by transposition. Proc Natl Acad Sci U S A 99,4495-4499.Dupuy AJ, Fritz S and Largaespada DA (2001) Transpositionand gene disruption in the male germline of the mouse.Genesis 30, 82-88.Ehrhardt A and Kay MA (2002) A new adenoviral helperdependentvector results in long-term therapeutic levels ofhuman coagulation factor IX at low doses in vivo. Blood 99,3923-3930.Ehrhardt A, Xu H and Kay MA (2003) Episomal persistence ofrecombinant adenoviral vector genomes during the cell cyclein vivo. J Virol 77, 7689-7695.Fausto N (2000) Liver regeneration. J Hepatol 32, 19-31.Fischer SEJ, Wienholds E and Plasterk RHA (2001) Regulatedtransposition of a fish transposon in the mouse germ line.Proc Natl Acad Sci U S A 98, 6759-6764.Follenzi A, Sabatino G, Lombardo A, Boccaccio C and Naldini L(2002) Efficient gene delivery and targeted expression tohepatocytes in vivo by improved lentiviral vectors. HumGene Ther 13, 243-260.Geurts AM, Yang Y, Clark KJ, Liu G, Cui Z, Dupuy AJ, Bell JB,Largaespada DA and Hackett PB (2003) Gene transfer intogenomes of human cells by the sleeping beauty transposonsystem. Mol Ther 8, 108-117.Hara T, Tan Y and Huang L (1997) In vivo gene delivery to theliver using reconstituted chylomicron remnants as a novelnonviral vector. Proc Natl Acad Sci USA 94, 14547-14552.Harui A, Suzuki S, Kochanek S and Mitani K (1999) Frequencyand stability of chromosomal integration of adenovirusvectors. J Virol 73, 6141-6146.Higgins GM and Anderson RM (1931) Experimental pathologyof the liver. I. Restoration of the liver of the white ratfollowing partial surgical removal. Arch Pathol 12, 186-202.Hillgenberg M, Schnieders F, Loser P and Strauss M (2001)System for efficient helper-dependent minimal adenovirusconstruction and rescue. Hum Gene Ther 12, 643-657.Horie K, Kuroiwa A, Ikawa M, Okabe M, Kondoh G, Matsuda Yand Takeda J (2001) Efficient chromosomal transposition ofa Tc1/mariner- like transposon Sleeping Beauty in mice.Proc Natl Acad Sci U S A 98, 9191-9196.Ivics Z, Hackett PB, Plasterk RH and Izsvák Z (1997) Molecularreconstruction of Sleeping Beauty, a Tc1-like transposonfrom fish, and its transposition in human cells. Cell 91, 501-510.Izsvák Z, Ivics Z and Plasterk RH (2000) Sleeping Beauty, awide host-range transposon vector for genetic transformationin vertebrates. J Mol Biol 302, 93-102.Johnson AD and Krieg PA (1994) pXeX, a vector for efficientexpression of cloned sequences in Xenopus embryos. Gene147, 223-226.Kalpana GV (1999) Retroviral vectors for liver directed genetherapy. Semin Liver Dis 19, 27-37.Kay MA, Glorioso JC and Naldini L (2001) Viral vectors forgene therapy: the art of turning infectious agents intovehicles of therapeutics. Nat Med 7, 33-40.Klink DT, Chao S, Glick MC and Scanlin TF (2001) Nucleartranslocation of lactosylated poly-L-lysine/cDNA complex incystic fibrosis airway epithelial cells. Mol Ther 3, 831-841.Kren BT, Bandyopadhyay P, Roy Chowdhury N, RoyChowdhury J and Steer CJ (2002) Oligonucleotide-mediatedsite-directed gene repair. Meth Enzymol 346, 14-31.Kren BT, Parashar B, Bandyopadhyay P, Roy Chowdhury N,Roy Chowdhury J and Steer CJ (1999) Correction of theUDP-glucuronosyltransferase gene defect in the Gunn ratmodel of Crigler-Najjar syndrome type I with a chimericoligonucleotide. Proc Natl Acad Sci USA 96, 10349-10354.Maruyama H, Higuchi N, Nishikawa Y, Kameda S, Iino N,Kazama JJ, Takahashi N, Sugawa M, Hanawa H, Tada N,Miyazaki J and Gejyo F (2002) High-level expression ofnaked DNA delivered to rat liver via tail vein injection. JGene Med 4, 333-341.Mikkelsen JG, Yant SR, Meuse L, Huang Z, Xu H and Kay MA(2003) Helper-independent Sleeping Beauty transposontransposasevectors for efficient nonviral gene delivery andpersistent gene expression in vivo. Mol Ther 8, 654-665.Montini E, Held PK, Noll M, Morcinek N, Al-Dhalimy M,Finegold M, Yant SR, Kay MA and Grompe M (2002) Invivo correction of murine tyrosinemia type I by DNAmediatedtransposition. Mol Ther 6, 759-769.Nakai H, Montini E, Fuess S, Storm TA, Grompe M and KayMA (2003) AAV serotype 2 vectors preferentially integrateinto active genes in mice. Nat Genetics 34, 297-302.237

Kren et al: Hepatocyte-targeted delivery of Sleeping BeautyNakai H, Yant SR, Storm TA, Fuess S, Meuse L and Kay MA(2001) Extrachromosomal recombinant adeno-associatedvirus vector genomes are primarily responsible for stableliver transduction in vivo. J Virol 75, 6969-6976.Oh YK, Kim JP, Yoon H, Kim JM, Yang JS and Kim CK (2001)Prolonged organ retention and safety of plasmid DNAadministered in polyethylenimine complexes. Gene Ther 8,1587-1592.Okabe M, Ikawa M, Kominami K, Nakanishi T and NishimuneY (1997) 'Green mice' as a source of ubiquitous green cells.FEBS Lett 407, 313-319.Pfeifer A, Kessler T, Yang M, Baranov E, Kootstra N, ChereshDA, Hoffman RM and Verma IM (2001) Transduction ofliver cells by lentiviral vectors: analysis in living animals byfluorescence imaging. Mol Ther 3, 319-322.Pikaart MJ, Recillas-Targa F and Felsenfeld G (1998) Loss oftranscriptional activity of a transgene is accompanied byDNA methylation and histone deacetylation and is preventedby insulators. Genes Dev 12, 2852-2862.Pollard H, Remy JS, Loussouarn G, Demolombe S, Behr JP andEscande D (1998) Polyethylenimine but not cationic lipidspromotes transgene delivery to the nucleus in mammaliancells. J Biol Chem 273, 7507-7511.Raper SE, Yudkoff M, Chirmule N, Gao GP, Nunes F, HaskalZJ, Furth EE, Propert KJ, Robinson MB, Magosin S, SimoesH, Speicher L, Hughes J, Tazelaar J, Wivel NA, Wilson JMand Batshaw ML (2002) A pilot study of in vivo liverdirectedgene transfer with an adenoviral vector in partialornithine transcarbamylase deficiency. Hum Gene Ther 13,163-175.Regnström K, Ragnarsson EGE, Köping-Höggård M,Torstensson E, Nyblom H and Artursson P (2003) PEI - apotent, but not harmless, muscosal immuno-stimulator ofmixed T-helper cell response and FasL-mediated cell deathin mice. Gene Therapy 10, 1575-1583.Wu CH, Sapozhnikov E and Wu GY (2002) Evaluation ofmulticomponent non-viral vectors for liver directed genedelivery. J Drug Target 10, 105-111.Wu GY and Wu CH (1988) Receptor-mediated gene deliveryand expression in vivo. J Biol Chem 263, 14621-14624.Yant SR, Ehrhardt A, Mikkelsen JG, Meuse L, Pham T and KayMA (2002) Transposition from a gutless adeno-transposonvector stabilizes transgene expression in vivo. NatBiotechnol 20, 999-1005.Yant SR, Meuse L, Chiu W, Ivics Z, Izsvák Z and Kay MA(2000) Somatic integration and long-term transgeneexpression in normal and haemophilic mice using a DNAtransposon system. Nat Genet 25, 35-41.Zhang G, Budker V and Wolff JA (1999) High levels of foreigngene expression in hepatocytes after tail vein injections ofnaked plasmid DNA. Hum Gene Ther 10, 1735-1737.Dr. Clifford J. Steer238

Gene Therapy and Molecular Biology Vol 7, page 239Gene Ther Mol Biol Vol 7, 239-243, 2003.PRL-3 as a target for cancer therapyResearch ArticleKoh Vicki, Fu Jianlin, Guo Ke, Lip Kuo Ming, Li Jie and Zeng Qi*Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609, Singapore__________________________________________________________________________________*Correspondence: Zeng Qi, Ph.D., Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609, Singapore Tel: 65-6874-3752; Fax: 65-6779-1117; e-mail: mcbzengq@imcb.nus.edu.sgKey Words: PRL-1, -2 and -3, prenylated phosphatase, cancer metastatasis, phosphatase inhibitor, therapeutic target and cancer therapyAbbreviations: phosphatase of regenerating liver (PRL); protein tyrosine phosphatases (PTPs); Chinese Hamster Ovary (CHO) cells;cells; serial analysis of gene expression (SAGE) technology; human embryonic kidney fibroblasts HEK293; restriction enzyme mediatedintegration, (REMI)Received: 24 October 2003; Revised: 3 December 2003;Accepted: 4 December 2003; electronically published: December 2003SummaryA group of protein tyrosine phosphatases (PTPs), the PRL family, has been implicated in the contribution andprogression of cancer metastasis. The PRL family consists of three members: PRL-1, -2 and -3. This small (20kD)class of prenylated protein tyrosine phosphatase contains a PTP signature motif (VHCXAGXXR) at their activesites and a catalytic domain, similar to dual specificity phosphatases. The three closely related PRLs share 76-87%identities in their amino acid sequences, with a unique C-terminal prenylation motif and with significant sequencehomology to Cdc14p, a mitotic regulator, and PTEN/MMAC1, the tumor suppressor (Zeng et al, 1998a). PRL-1 wasidentified as an immediate-early gene which was induced in mitogen-stimulated cells and regenerating liver. PRL-3,along with PRL-2, was identified subsequently by using PRL-1 sequence to search mouse EST database. Recently,PRLs have been implicated in the process of oncogenic transformation and cancer metastasis. We suggest thatPRLs might be important modulators in the process of cancer metastasis and are therefore potential targets fortherapeutic intervention of cancer.I. IntroductionPRL proteins are conserved from human to C.elegans (Figure 1). We suggested that they have commoncritical physiological functions, but each might has its owndistinct target substrates. PRL-1 is expressed at high levelsin proliferating cells and a number of human tumor celllines, including HeLa (Diamond et al, 1994; Wang et al,2002). Overexpression of PRL-1 and PRL-2 in epithelialcells results in a transformed phenotype in culture andtumor growth in nude mice (Cates et al, 1996; Zeng et al,2003). By PCR-based subtractive hybridization, PRL-2was shown to be upregulated 1.8, 2.7 and 4-fold inadvanced prostate cancer cell lines LNCaP, PC-3 and DU-145 respectively, in comparison with normal epithelialcells. The data suggest that PRL-2 is associated withprostate tumor progression (Wang et al, 2001). Recently,by using serial analysis of gene expression (SAGE)technology, researchers (Bert Vogelstein and hiscolleagues) at the Johns Hopkins Medical Institutesreported that among 144 upregulated genes detected inmetastatic colorectal liver samples, PRL-3 was the onlygene specifically overexpressed in all 18 metastaticcolorectal cancers examined (Saha et al, 2001). We foundthat prenylated PRL-1 and -3 are enriched on the plasmamembrane (Zeng et al, 2000, 2003). Overexpression ofPRL-1 and -3 in Chinese Hamster Ovary (CHO) cellspromotes cell migration, invasion and metastasis.Moreover, PRL-1 and -3 overexpressing CHO cells arecapable of inducing metastatic tumor formation in nudemice (Zeng et al, 2003). Overexpression of PRL-3 inhuman embryonic kidney fibroblasts HEK293 cells hasalso been found to enhance growth rates versus nontransfectedcells (Matter et al, 2001). Taken together, thesedata suggest that the PRL-1, -2 and -3 might be associatedwith cancer metastasis and act as major players inoncogenic and metastatic processes. How does PRL-3 spurcolon cancer metastasis? How might PRLs be involved inthe pathways of signal transduction related to cancerdevelopment and metastasis? Much remains to be learnedfrom these striking discoveries to answers. Here, wesuggest that PRL-3 might serve as one of the importantmarkers to track the events leading to colon cancermetastasis. We speculate that an excess of PRLphosphatases activity could bring about key alterationsthat would contribute to the acquisition of metastaticproperties in tumor cells. PRLs may therefore be potentialtargets for new cancer therapeutic strategies.239

Zeng: PRL-3 as a target for cancer therapyFigure 1. Alignment of amino acid sequence of C. elegans PRL (C_PRL), Drosophila PRL, (Dro._PRL), mouse PRLs (mPRL-1, mPRL-2 and mPRL-3) and human PRLs (hPRL-1, hPRL-2 and hPRL-3) with GENE Doc. Program. Identical residues among all these PRLsare shaded in yellow.II. Materials and methodsA. Establishment of a stable cell lineexpressing myc-PRL-3 and a stable cell poolexpressing EGFP-PRL-3The pStar-myc-PRL-3 and pEGFP-PRL-3 plasmids wereconstructed and described in our early studies (Zeng et al, 2000and 2003). CHO-K1 cell line was obtained from the AmericanType Culture Collection (Manassas, VA) to generate CHO cellline stably expressing PRL-3 (clone 36) in pStar vector (Zeng etal, 1998b) and to generate CHO stable pooled cells expressingPRL-3 in pEGFP-C1 vector (Clontech: http://www.clontech.com/index.shtml). Briefly, the cells were transfected with the pStar-Myc-PRL-3 or pEGFP-PRL-3 respectively, cultured in RPMI1640 medium supplemented with 10% fetal bovine serum andselected in 1mg/ml of G418 to establish a stable clonal cell linefor pStar-myc-PRL-3 or stable pooled transfectants for pEGFP-PRL-3.B. Experimental metastasis andimmunoperoxidase 10-week old female nude mice(Jackson Labs, USA) were each injected via the tail vein withmyc-PRL-3 expressing cells (5x10 5 ). Mice were sacrificed 25days after the tail vein injection and all tissues were examined formetastasis. Lungs with metastasis were performed withimmunoperoxidase labeling by using a Vectastain ABC kit fromVector Laboratories (Burlingame, CA) according to theprocedure provided by the supplier. The c-Myc antibody (9E10)was from Santa Cruz Biotechnology (Santa Cruz, CA).C. Confocal microscopyCells stably expressing EGFP-PRL-3 were seeded ontoglass coverslips and grew for 24 hours. Cells were washed twicewith PBSCM (PBS containing 1mM of MgCl 2 and 1mM ofCaCl 2 ) and then fixed in 3% paraformaldehyde for 20 min atroom temperature. After three more washes with PBSCM, cellswere mounted onto a glass slide with one drop of anti-fadereagent in PBS glycerol (Biomedia Corp, Foster City, CA), andkept at 4 o C in the dark until analysis. Confocal imaging wasperformed using a laser scanning head (MRC 1024, Bio-RadLaboratories, GB).D. Production and analysis of transgenictadpolesA modified restriction enzyme mediated integration(REMI) method was used to generate transgenic Xenopus laevistadpoles (Duncan B et al, 2000). Briefly, 2.5 µl containing 100-250 ng of the construct was incubated with 2.5 µl containing1x10 5 /µl of sperm nuclei for 10 min at room temperature. Theincubation mixture was then diluted to a concentration of onesperm per 10-12 µl and injected into de-jellied eggs usingconstant flow injector (Harvard Apparatus PHD 2000 Infusion)at a flow rate of 0.7 µl/min. Normal-looking embryos werecollected and incubated. The tadpoles were then screened for theexpression of EGFP. Pictures were taken by a digital camera(Nikon Coolpix 995) attached to a fluorescent microscope (Zeiss,M2 Bio Quad).III. PRL-3 might help tumorformation in foreign territoriesMetastasis is a complex process involving manychanges in cell/tissue physiology and gene expression. Themajor events in the process of metastasis include theability of tumor cells to leave their sites of primary tumorand enter the circulation to extravagate into a new tissue,begin and maintain growth in this tissue to formpreangiogenic micrometastases, and then develop a bloodsupply to enable the formation of macroscopic tumors(Weiss, 2000; Chambers et al, 2002; Takeda et al, 2002).It is metastasis that defeats oncologists to cure theirpatients. What invokes tumor cells to dissociate from theprimary tumor and migrate to distant tissues is largelyunclear. Overexpression of PRL-3 has been found in 100%of 18 colorectal cancer liver metastases examined (Saha etal, 2001). To carry out this process in vivo and to study therole of PRL-3 in the pathways of metastasis, cells from astable CHO clone overexpressing of PRL-3 were injectedinto the tail veins of 10-week-old female nude mice, thusintroducing these cells directly into circulatory bloodsystem of the animals. We found rapid metastasis in ouranimal experimental metastasis model (Zeng et al, 2003).The tumor was performed with fresh frozen sections andchecked for PRL-3 protein expression level by240

Gene Therapy and Molecular Biology Vol 7, page 241Figure 2. Expression of PRL-3 protein in small tumor area (S) but not in large tumor area (L). A cryosection of lung tissue was obtainedfrom 13-week-old nude mouse after 25 days injection with clonal Chinese Hamster Ovary cells overexpression of myc-PRL-3 by the tailvein. The section was precessed for immunoperoxidase labeling using c-Myc antibody (9E10) which was from Santa CruzBiotechnology (Santa Cruz, CA). Bar represents 20 µm.immunohistochemistry method (ABC). Interestingly, inmost cases, PRL-3 was detected in small tumors (Figure2) but not in large tumors. The process of metastasis byoverexpression of PRL-3 likely does not require anyfurther inputs from its upstream. These results give usclues that PRL-3 might act as an initiator for the tumorimplantation (or perhaps for blood vessel formation) inforeign territory and then pass the job to its downstreameffectors.IV. PRL-3 might have a role incardiac hypertrophyHuman PRL-3 has been reported to play a physiologicalrole of preventing the tyrosine phosphorylation of p130 casand has an effect on the mobilizing of intracellular calciumin response to Angiotensin-II in HEK 293 cells (Matter etal, 2001). The PRL-3 was preferentially expressed in adultand fetal hearts at all parts except aorta (Li and Zeng,unpublished data). Transgenic mice overexpressing PRL-3in the heart show overt cardiac hypertrophy and reducedcardiac function associated with impaired calciumhandling (Kadambi et al, 2000). Cardiac hypertrophyinvolves an increase in size and mass of individualcardiomyocytes without an increase in cell number. Wediscovered that overexpression of the PRL-3 in CHO cellscaused dramatically enlargement in sizes of sometransfected cells (Figure 3). The studies suggest that PRL-3 may have a physiological role in maintaining normalfunctions of the heart. Overexpression of the PRL-3 mightbe involved in the progression of cardiac hypertrophy.Figure 3. A giant cell overexpressing the EGFP-PRL-3. A CHOcell overexpressionof PRL-3 dramatically enlarged in size. Bar,20 µm241

Zeng: PRL-3 as a target for cancer therapyV. PRL-3 might play an essential rolein embryonic developmentThe protein-tyrosine phosphatases (PTP) superfamilyconsists of a large group of enzymes that play critical rolesin the regulation of cellular growth and differentiation.The balance between phosphorylation anddephosphorylation is a major regulatory mechanismaffecting the functions of diverse proteins that participatein many aspects of cellular, physiological, pathogenicprocesses and embryogenesis. In order to have a quickexamination of the possible functions of PRL-3 duringembryonic development, we generated transgenic frogembryos overexpressing EGFP-PRL-3 and compared themwith control embryos overexpressing EGFP to investigatethe outcomes of the animals. Xenopus embryos areexcellent for analyzing defects in early development(Amaya et al, 1999; Duncan et al, 2000). Overexpressionof EGFP-PRL-3 in transparent Xenopus embryos allowsus to easily trace the green fluorescent protein marker withbetter spatial and temporal control. A large amount ofembryos can be easily monitored and manipulated at thesame time. Statistic data can be obtained shortly.Disturbances in PRL-3 expression in transgenic Xenopusembryos showed severe and diverse defects duringembryonic development at very early stages. Theexamples were showed at day 5 (Figure 4). The poorembryonic survival rate of the transgenic animals is due totheir death, abnormal growth and retardation. From day 1to day 5, the total survival rate of the transgenic EGFP-PRL-3 is 8% which is 7.5 times lower than controltransgenic EGFP (53%). Our preliminary results thereforesupport a major physiological role for the PRL-3 duringembryogenesis.VI. Summary and future perspectivesIn summary, PRL-3 has been so far the onlyphosphatase linked to colorectal cancer metastasis (Saha etal, 2001). Whether PRL-3, perhaps PRL-1 and -2, is alsoinvolved in other types of cancer and how these PRLsinitiate and maintain the process of metastasis remain asimportant questions to be solved. The reason that makesscientists believe PRL-3 gene might provide a newtherapeutic target for colorectal cancer is thatFigure 4. Overexpression of the EGFP-PRL-3 disrupts Xenopus embryogenesis at early stages. The EGFP-PRL-3 positive embryoswere examined and compared with EGFP control positive embryos on day 5 (stages 41).242

Gene Therapy and Molecular Biology Vol 7, page 243most of the previously described genetic alterations incolorectal cancers involve inactivation of tumorsuppressor genes. It is difficult to target with drugs forinactive or absent genes in the cancers (Saha et al, 2001).In contrast, PRL-3 is elevated in cancer cells, whichprovides an excellent target for drug discovery purposes.By NMR resonance, the PRL-2 and -3 structures havebeen investigated (Kozlov et al, 2002; Zhou et al, 2003).Understanding of its structure, especially the catalyticdomain, may help us to search specific inhibitors for PRL-3 in the treatment of the colorectal metastases. We studieda catalytically inactive PRL-3 (C104S) mutant and foundthat it reduced effect on promoting cell migration (Zeng etal, 2003). Here, we suggest that the consensus phosphatasemotif might potentially be a therapeutic target. We alsosuggest that specific antibodies against each PRL areneeded to distinguish the expression and investigate theindividual roles of these closely related PTPs, which inturn may lead to new insights into cancer metastasis. Suchantibodies may also provide a platform for noveldiagnostic, prognostic or therapeutic approaches. Thefunctions of the PRLs-PTP family in cancer metastasisundoubtedly cannot be ignored.AcknowledgmentsI am very grateful to Yuen WH and Loh CB inTransgenic Frog Unit; Gan, BQ and Tang JP. in HistologyUnit from our Institute to provide excellent facilities tothis related work. I appreciate Drs. Tang BL and Yin XJfor critical reading of the article.ReferencesAmaya E, and Kroll KL (1999) A method for generatingtransgenic frog embryos. Methods Mol Biol 97, 393-414Cates CA, Michael RL, Stayrook KR, Harvey KA, Burke YD,Randall SK, Crowell PL, Crowell DN (1996) Prenylation ofoncogenic human PTPCAAX protein tyrosine phosphatases.Cancer Letters 110, 49-55Chambers AF, Groom AC, and MacDonal IC (2002)Dissemination and growth of cancer cells in metastatic sites.Nat Rev Cancer 2, 563-572Diamond RH, Cressman DE, Laz TM, Abrams CS, and Taub R(1994) PRL-1, a unique nuclear protein tyrosine phosphatase,affects cell growth. Mol Cell Biol 14, 3752-3762Duncan B, Sparrow BL and Tim JM (2000) A simplified methodof generating transgenic Xenopus. Nucleic Acids Res 28,NO.4 E12-e12Kadambi VJ, Lorenz JN, Stagliano NE, Matter WF, Wang XS,Bloem L, Acton S, Rodrigue-Way AC, Breitbart RE, VlahosCJ (2000) Imparired ventricular relaxation resulting fromcardiac-specific overexpression of a human prenylatedprotein tyrosine phosphatase. Circulation (Suppl) 102, II-268. (abstract 1312)Kozlov G, Cheng J, Lievre C, Banville D , Gehring K, Ekiel I(2002) Letter to the Editor, 1 H, 13 C and 15 N resonanceassignments of the human phosphatase PRL-3. J Biom NMR24, 169-170Li and Zeng et al, unpublished dataMatter WF, Estridge T, Zhang C, Belagaje R, Stancato L, DixonJ, Johnson B, Bloem L, Pickard T, Donaghue M, Acton S,Jeyaseelan R, Kadambi V, Vlahos CJ (2001) Role of PRL-3,a Human Muscle-Specific Tyrosine Phosphatase, inAngiotensin-II Signaling. Biochem Biophys Res Commun283, 1061-1068Saha S, Bardelli A, Buckhaults P, Velculescu VE, Rago C, CroixBS, Romans KE, Choti MA, Lengauer C, Kinzler KW,Vogelstein B (2001) A phosphatase associated withmetastasis of colorectal cancer. Science 294, 1343-1346Takeda A, Stoeltzing O, Ahmad SA, Reinmuth N, Liu W, ParikhA, Fan F, Akagi M., and Ellis LM (2002) Role ofangiogenesis in the development and growth of livermetastasis. Ann Surg Oncol 9, 610-616Wang J, Kirby CE, Herbst R (2002) The tyrosine phosphatasePRL-1 localizes to the endoplasmic reticulum and the mitoticspindle and is required for normal mitosis. J. Biol Chem277, 46659-46668Wang Q, Holms DIR, Powell SM, Lu QL, Waxman J (2001)Analysis of stromal-epithelial interactions in prostate canceridentifies PTPCAAX2 as a potential oncogene. CancerLetters 175, 63-69Weiss L (2000) Metastasis of cancer, a conceptual history fromantiquity to the 1990s. Cancer Metastasis Rev 19, 193-383Zeng Q, Hong W, Tan YH (1998a) Mouse PRL-2 and PRL-3,two potentially prenylated protein tyrosine phosphataseshomologous to PRL-1. Biochem Biophys Res Commun244, 421-427Zeng Q, Tan YH and Hong WJ (1998b) A single plasmid vector(pStar) mediating efficient tetracycline-induced geneexpression. Anal. Biochem. 259, 187-194Zeng Q, Si XN, Horstmann H., Xu Y., Hong WJ, Pallen CJ(2000) Prenylation-dependent Association of ProteintyrosinePhosphatases PRL-1, -2, and -3 with the PlasmaMembrane and the Early Endosome. J Biol Chem 275,21444-21452Zeng Q, Dong JM, Guo K, Li J, Tan HX, Koh V, Pallen CJ,Manser E, Hong WJ (2003) PRL-3 and PRL-1 promote cellmigration, invasion, and metastasis. Cancer Res 63,2716-2722Zhou H, Gallina M, Mao H, Nietlispach D, Betz SF, Fetrow JS,Domaille PJ (2003) 1 H, 13 C and 15 N resonance assignmentssecondary structure of the human protein phosphatase, PRL-2. J Biom NMR 27, 397-8Dr. Qi Zeng243

Zeng: PRL-3 as a target for cancer therapy244

Gene Therapy and Molecular Biology Vol 7, page 245Gene Ther Mol Biol Vol 7, 245-254, 2003Protective effect of heat shock proteins:potential for gene therapyReview ArticleDavid S. LatchmanInstitute of Child Health, 30 Guilford Street, London WC1N 1EH & Birkbeck, University of London Malet Street, LondonWC1E 7HX__________________________________________________________________________________*Correspondence: David S. Latchman, Tel (+44) 20 7631 6274; Fax (+44) 20 7631 6259; e-mail: d.latchman@bbk.ac.ukKey words: heat shock proteins (hsps); Drosophila protein, gene therapyAbbreviations: heat shock proteins, (hsps); herpes simplex virus, (HSV); heat shock transcription factor, (HSF-1); cytokinecardiotrophin-1, (CT-1)Received: 24 September 2003; Accepted: 30 September 2003; electronically published: December 2003Contributed by Dr. LatchmanSummaryThe heat shock proteins (hsps) are expressed in normal cells but their expression is enhanced by a number ofdifferent stresses including heat and ischaemia. They play important roles in chaperoning the folding of otherproteins and in protein degradation. In the heart and the brain, a number of studies have shown that priorinduction of the hsps by a mild stress has a protective effect against a more severe stress. Moreover, over-expressionof an individual hsp in cardiac or neuronal cells in culture and in the intact heart or brain of either transgenicanimals or using virus vectors, also produces a protective effect, directly demonstrating the ability of the hsps toproduce protection. These findings indicate the potential importance of developing procedures for elevating hspexpression in a safe and efficient manner in human individuals using either pharmacological or gene therapyprocedures.I. IntroductionHeat shock proteinsIt is now over forty years since Ritossa observed thatexposure of the larval salary gland of Drosophila toelevated temperature resulted in the appearance of newpuffs in the giant chromosomes of these cells (Ritossa,1962). It is now clear that these puffs represent thetranscriptional induction of specific genes which encode agroup of proteins known as the heat shock proteins (forreview see Lindquist and Craig, 1988; Parsell andLindquist, 1993).Although originally demonstrated in Drosophila, theinduction of a small number of heat shock proteins byelevated temperature, is observed in all organisms studiedranging from prokaryotic bacteria to mammals includingman. Moreover, this evolutionary conservation extends notonly to the existence of the heat shock response indifferent organisms but also to the induced proteinsthemselves which are very similar to one another in verydifferent organisms. Thus, the best characterised hsps,hsp90, hsp70, hsp65 and hsp27 (each hsp is namedaccording to its mass in kilodaltons) are induced inresponse to heat in all organisms studied from bacteria toman and are highly conserved between different species,for example, the hsp90 protein from mammals shows 60%amino acid identity with the corresponding yeast proteinand 78% with the Drosophila protein (Rebbe et al, 1987).The various hsps and their characteristics are listed inTable 1.Although originally identified on the basis of theirinduction by elevated temperature and therefore named theheat shock proteins, these proteins are in fact induced by awide range of stimuli which are potentially damaging tothe cell. Such inducers include infections with a widevariety of different viruses (Collins and Hightower, 1982;Khandijan and Turler, 1983; La Thangue and Latchman,1988), treatment with ethanol (Plesset et al, 1982), steroidhormones (Norton and Latchman, 1989), and amino acidanalogues (Li and Laszlo, 1985). Interestingly, hsps arealso induced by processes which may occur during humandisease, notably exposure of specific cells such as cardiacor neuronal cells to ischaemia or elevated levels of freeradicals (Nowak, 1985; Polla, 1988).The strong evolutionary conservation of the hsps orstress proteins which was discussed above and theirinduction by a variety of stressful stimuli, indicates thatthey are likely to have some critical function in the cellular245

Latchman: Protective effect of heat shock proteinsresponse to stress. Interestingly, however, many of theseproteins are also synthesised by normal unstressed cellswith their synthesis being further enhanced upon exposureto stress. For example, hsp90 is one of the most abundantproteins in unstressed cells, constituting approximately 1%of the total protein in mammalian cells even prior toexposure to stress. This has led to the idea that thefunction of the hsps is one which is required in normalcells but is needed to an even greater extent in stressedcells. This idea is in accordance with the detailedfunctional studies of individual hsps which, as shown inTable 1, have indicated that a number of them have a rolein ensuring the correct protein folding of other proteinswithin the cells, acting as so-called “molecularchaperones” (for review see Ellis, 1990). Thus, forexample, hsp90 associates with the steroid receptors, suchas the glucocorticoid receptor and keeps them in aninactive form located in the cytoplasm prior to exposure tosteroid. Upon steroid treatment, hsp90 dissociates from thereceptor which then can move to the nucleus and activatesteroid responsive genes.Clearly, correct protein folding is of importance innormal cells but factors which aid this process will berequired at higher level in stressed cells, when forexample, stimuli such as elevated temperature result in anincreased level of denatured or partially denaturedproteins. This idea is also in agreement with findingswhich indicate that hsps can be induced by treatment ofcells with amino acid analogues, which again wouldinduce the formation of abnormally folded proteins (Liand Laszlo, 1985).Both in normal cells and in stressed cells there willalso be a need to degrade proteins which have becomeabnormally folded and cannot be rescued by the action ofchaperone proteins. It is therefore of interest that ubiquitinwhich plays a critical role in protein turnover by beinglinked to proteins marked for degradation, is also inducedby elevated temperature and is therefore a heat shockprotein (see Table 1). A further link between the hsps andprotein degradation is provided by the observation thatinhibition of hsp70 synthesis enhances the cell deathwhich is induced by inhibiting the proteosome whichmediates the degradation of ubiquitinated proteins(Robertson et al, 1999).The idea of the hsps as proteins which are ofimportance in normal cells but which assume a greatersignificance in stressed cells, leads logically to the ideathat the induction of these proteins by a stressful stimulusis of itself important in assisting the cell to protect itselffrom stress. In turn, this leads to the idea that the priorinduction of the hsps by a mild stress or by some othernon-stressful procedure, would be protective againstsubsequent more severe stress. This idea obviously hasconsiderable medical importance and has therefore beenintensively investigated.Over the years, this work has effectively proceededin three stages. Firstly, the demonstration that exposure tomildly stressful stimuli which can induce hsp expression,can in turn protect cells against exposure to a more severestress. Clearly, such findings implicate the hsps as beingprotective but do not prove this, since the protective effectcould be due to some other action of the mildly stressfultreatment, other than its ability to induce the hsps. Thisidea leads directly to the second stage of theseinvestigations, namely, the use of gene constructs to overexpressthe hsps in cultured cells and then demonstrate aprotective effect against subsequent exposure to stress.Finally, more recently, these experiments in cultured cellshave been complemented by experiments over-expressingthe hsps in an intact animal and again demonstrating aprotective effect. In subsequent sections of this review, Iwill discuss these three stages of work on the protectiveeffect of heat shock proteins, focusing on studies involvingneuronal or cardiac cells in culture or in the intact brainand heart, because of the key medical importance of theseorgans.Table 1. Major eukaryotic hspsFamily Members ProkaryoticHomologueFunctionalRoleCommentsHsp90Hsp100,Hsp90Grp94C62.5 (E. coli)Maintenance ofproteins such assteroid receptor.Src. in aninactive formuntil appropriateDrosophila andyeast homologues ofhsp90 are knownas hsp83Hsp70Grp78 (= Bip)Hsp72, Hsp73Hsx70dna K (E. coli)Protein foldingand unfolding:assembly ofmultimericcomplexesHsx - 70 only inprimatesHsp65Hsp65gro EL (E. coli)Mycobacterial65 kd antigenProtein foldingand unfolding:organelletranslocationMajor antigen ofmany bacteria andparasites which infectman246

Gene Therapy and Molecular Biology Vol 7, page 247Hsp56Hsp56-Protein folding,component of steroidreceptorcomplexBinds FK506(tarcolimus) and isalso known asFKBP56Hsp32Hsp32-Cleaves heme to yieldcarbon monoxide andthe protective antioxidantmolecule,biliverdinAlso know as hemeoxygenase-1Hsp27Hsp27,Hsp26, etc.Mycobacterial18 kd antigenUnclearVery variable in sizeand number indifferent organismsUbiquitinUbiquitinNoneProtein degradationAlso conjugated tohistone H2A in thenucleus leading topotential role in generegulationII. Protective effect of stimuli whichinduce hsp synthesisDuring the 1980s, a very large number of studiesdemonstrated that, in cells in culture, stimuli whichinduced hsp synthesis such as a mild stress resulted inprotection against subsequent exposure to a more severestress. Moreover, it was also demonstrated that the levelsof the hsps induced by such mildly stressful procedures,generally correlated with the level of protection which wasobserved against the subsequent more severe stress (forreview see Lindquist and Craig, 1988; Parsell andLindquist, 1993).Following such early studies primarily carried out incell lines of fibroblast origin, this work was extended alsoby carrying out similar studies both in cell lines ofneuronal origin and in primary neuronal cells. Thus, forexample, primary neuronal cells cultured in vitro areprotected by exposure to mild heat or ischaemic stressfrom a subsequent more severe heat or ischaemic stress orexposure to the excitotoxin glutamate (Lowenstein et al,1991; Rordorf et al, 1991; Amin et al, 1995). Indeed, inour own studies, the degree of protection afforded by aninitial mild stress correlated with the amount of hspinduced, rather than the nature of the subsequent stress.Thus, a mild heat stress produced a better protective effectagainst subsequent severe heat stress or severe ischaemia,than was observed for a mild ischaemic stress correlatingwith the greater degree of hsp induction produced by themild heat stress (Amin et al, 1995).These in vitro studies examining the protective effectof a mild stress against a more severe stress, have alsobeen supplemented by examining the protective effect ofsuch mild stresses against exposure to stimuli whichinduce apoptosis (programmed cell death). Thus, it hasbeen shown that prior mild heat shock can protected theND7 neuronal cell line against apoptosis induced by serumwithdrawal and addition of retinoic acid (Mailhos et al,1993). Similarly, primary neonatal dorsal root ganglioncells are protected by mild heat shock against subsequentwithdrawal of nerve growth factor, which inducesapoptosis in these neurones (Mailhos et al, 1994).These in vitro observations were subsequentlyextended by studies in the intact animal in vivo. Forexample, prior exposure of the animal to a mild heat stressis sufficient to protect retinal neurones against asubsequent in vivo exposure to either light damage (Barbeet al, 1988) or ischaemia (Chopp et al, 1989). Similarly,exposure to a mild cerebral ischaemia protectshippocampal neurones against subsequent exposure to amore severe ischaemia (Kitagawa et al, 1990).Similar studies have also been carried out in thewhole heart either perfused ex vivo or in the intact animalin vivo, demonstrating that stimuli which result in hspinduction can protect the heart against subsequentexposure to a more severe stress. This was firstdemonstrated by Currie et al, (1988) who exposed rats toelevated temperature and then removed their hearts andexposed them to ischaemia on a Langendorff perfusionapparatus. They demonstrated that the hearts from ratswhich had been exposed to an elevated temperatureshowed improved recovery of contractile functionfollowing subsequent ischaemia and reperfusion comparedto control hearts. Furthermore, the reperfusion damage, asmeasured by creatine kinase release was significantlyreduced in the heat shock hearts. These findings thereforedemonstrated for the first time that a stimulus whichinduced hsp induction in the intact heart was able to247

Latchman: Protective effect of heat shock proteinsproduce a protective effect against subsequent exposure toischaemia/reperfusion. These results have subsequentlybeen extended both by examining other parameters ofheart function and by using other species such as the rabbit(Karmazyn et al, 1990; Yellon et al, 1992) (for review seeYellon and Latchman, 1992).These studies demonstrating a protective effect in theheart on a Langerdorff perfusion apparatus following priorexposure to elevated temperature in vivo, lead naturally tothe question of whether a similar protective effect wouldbe observed in hearts exposed to myocardial ischaemiawithin the intact animal following a prior exposure to heatshock.Donnelly et al, (1992) demonstrated that this wasindeed the case with an effective reduction of infarct sizebeing observed when rat hearts were exposed to 35minutes of left coronary artery occlusion in the intactanimal following exposure to heat shock. Moreover, thisprotective effect is not confined to the use of heat shockitself to induce the hsps. Thus, Marber et al, (1993) wereable to demonstrate that four brief periods (5 minuteseach) of cardiac ischaemia were able to induce hspsynthesis and were also able to reduce infarct size whenthe hearts were subsequently exposed to 30 minutes ofischaemia in the intact animal.Hence, stimuli which result in hsp induction in theintact heart in vivo can produce a protective effect againstsubsequent exposure of the heart to ischaemia/reperfusioneither on a perfusion apparatus or within the intact animal.In addition, a number of studies have demonstrated thatthe protective effect correlates with the amount of heatshock protein which is induced. Thus, for example,Marber et al, (1994) showed a correlation between theamount of hsp70 produced by heat stress of papillarymuscle and the muscle’s ability to recover functionfollowing a period of hypoxia. Similarly, Hutter et al,(1994) demonstrated a similar correlation between theamount of hsp70 and the ability to limit infarct sizefollowing exposure of the heart to ischaemia andsubsequent reperfusion.These studies indicate therefore, that neuronal andcardiac cells can be protected by prior exposure to a mildstress sufficient to induce hsp over-expression. Moreover,the correlation between the amount of hsp induced and thedegree of protection observed, suggests that it is theinduction of the hsp rather than some other effect of themild stress, which produces the protective effect againstthe more severe stress. However, such studies areessentially only correlative and to prove that hsps can havea protective effect, it is necessary to over-expressindividual hsps in neuronal or cardiac cells. Such studiesare discussed in the next section.III. Protective effect of individual hspsin neuronal and cardiac cells in vitroIn a number of cases, it has been possible to showthat over-expression of an individual hsp can provide aprotective effect against damaging stimuli, in the samemanner as a mild hsp-inducing stress. Thus, for example,dorsal root ganglion neurones can be protected againstthermal or ischaemic stress by over-expression of eitherhsp70 or hsp90 (Uney et al, 1993; Amin et al, 1996; Wyattet al, 1996) and a similar effect of hsp70 and hsp90 hasbeen observed in the ND7 neuronal cell line (Mailhos etal, 1994). Similarly, Fink et al, (1997) were able to protectcultured hippocampal neurones against subsequent heatshock using a herpes simplex virus (HSV)-derivedamplicon vector expressing hsp70, indicating that thiseffect applies to neurones derived from both the centraland the peripheral nervous systems.Similar studies on the protective effect of the hsps incardiac cells, initially focused on hsp70 and utilised theH9c2 cell line which was derived initially from the ratheart. In 1994, two groups reported the results ofexperiments in which stable transfection was used toproduce clonal cell lines derived from H9c2 whichconstitutively over-expressed hsp70 (Heads et al, 1994;Mestril et al, 1994). These cells were shown to beprotected against subsequent exposure to thermal orischaemic stress compared to control cells which did notover-express hsp70. These studies were subsequentlyextended by Cumming et al, (1996b) who demonstratedthat similar protective effects against heat stress orsimulated ischaemia could be observed when hsp70 wasover-expressed by transfection of primary rat cardiacmyocyte cultures, demonstrating that this protective effectcould be observed both in primary cardiac cells and in celllines derived from them. A similar protective effect wasalso observed when hsp70 was over-expressed bytransfection in coronary endothelial cells (Suzuki et al,1998) indicating that hsp70 can protect these cells as wellas cardiac myocytes. This is of particular interest since ithas been shown that when the heart is exposed to elevatedtemperature in vivo, hsp70 induction occurs primarily inendothelial cells rather than in cardiac myocytes (Amraniet al, 1998; Leger et al, 2000).To extend these experiments to other hsps,transfection methods were used to over-express hsp90,hsp65, or hsp56 either in the H9c2 cell line (Heads et al,1995) or in cultured primary cardiac cells (Cumming et al,1996a,b). In these experiments, hsp90 over-expression wasable to protect the cells against subsequent thermal stressbut not against subsequent simulated ischaemia whereashsp65 or hsp56 had no protective effect. Since hsp70 overexpressionprotected against both thermal or simulatedischaemic stress in these experiments, these studiesindicate that different hsps can have different protectiveeffects and need to be tested individually for theirprotective effect in any specific situation.This idea is reinforced by findings in neuronal cellswhere the over-expression of an individual hsp does notalways reproduce the protective effect of a mild hspinducingstress. Thus, over-expression of hsp70 or hsp90in ND7 cells (Mailhos et al, 1994) or DRG neurones(Wyatt et al, 1996) does not reproduce the protectiveeffect of mild heat shock against subsequent apoptoticstimuli. Similarly, in the experiments of Fink et al, (1997)over-expression of hsp70 with an hsp vector did notprotect hippocampal neurones against glutamate toxicitywhich may act by inducing apoptosis (Kure et al, 1991),despite the fact that previous studies demonstrated a clear248

Gene Therapy and Molecular Biology Vol 7, page 249protective effect of mild heat stress against subsequentexposure to glutamate (Lowenstein et al, 1991; Rordorf etal, 1991).However, in further experiments we were able toshow that the protective effect of a mild heat stress againstapoptotic stimuli in neuronal cells could be reproduced byover-expressing the small heat shock protein hsp27. Thus,over-expression of hsp27 using an herpes simplex virus(HSV-)based vector was able to protect both ND7 cellsand DRG neurones against apoptosis induced bywithdrawal of serum or nerve growth factor, whereas suchprotection was not observed when hsp70 was overexpressedwith a similar vector (Wagstaff et al, 1999)(Figure 1a). As expected, over-expression of either hsp27or hsp70 by this means was able to protect the neuronalcells against subsequent exposure to heat shock orischaemia, paralleling the results obtained with plasmidconstructs for hsp70 and extending this to hsp27 (Wagstaffet al, 1999).Interestingly, when these experiments with HSVvectors over-expressing individual hsps were used todetermine their protective effect in primary cardiac cells(Brar et al, 1999), we confirmed our earlier results thathsp70 over-expression can protect cardiac cells againstsimulated ischaemia or thermal stress, whereas overexpressionof hsp56 has no such protective effect.Moreover, we were able to extend these studies byshowing firstly, that hsp70 can protect against theinduction of apoptosis (programmed cell death) in cardiaccells by exposure to ceramide, whereas hsp56 has noprotective effect and secondly, to demonstrate that overexpressionof hsp27 (which we had not previously tested)similarly protects cardiac cells against subsequentexposure to thermal or ischaemic stress or to ceramide(Figure 1b). Hence, in cardiac cells both hsp70 and hsp27can protect against apoptosis whereas in neuronal cellsonly hsp27 has this protective effect. This reinforces theneed to study individual hsps for their protective effectagainst specific stimuli and in specific cell types.Figure 1. (A) Number of DRG neurones undergoing apoptosis (as assayed by TUNEL staining) after NGF withdrawal following priorinfection with the indicated virus. Values are the mean of three determinations whose standard error is shown by the bars. Significantenhancement of survival (p < 0.05) was observed only with hsp27-expressing virus. (B) Percentage of apoptotic cells (as assayed byTUNEL staining) in cardiomyocytes pre-infected with the indicated viruses or left untreated (C) and then 24 hours after infection eitherleft untreated or treated for six hours with 25µM ceramide. The data represent the means of two independent experiments whosestandard error is indicated by the bars. Both hsp27 and hsp70-expressing viruses significantly reduced the number of apoptotic cellscompared to uninfected cells (C) or cells infected with a control virus expressing green fluorescent protein (GFP) (p < 0.05).249

Latchman: Protective effect of heat shock proteinsThese findings also suggest that hsp27 may be asprotective as hsp70 in cardiac cells whilst potentiallybeing more protective in neuronal cells. Similar resultswere also obtained by Martin et al, (1997; 1999) who usedan adenovirus vector to over-express hsp27 or the relatedprotein αB-crystallin in cardiac cells. They were able todemonstrate that both these proteins were able to protectcardiac myocytes from the effect of simulated ischaemiaand that decreasing the level of endogenous hsp27 usingan antisense approach enhanced the damaging effects of asubsequent ischaemic stimulus.Taken together therefore, these results demonstratethat several hsps can play an important role in protectingcells in culture from the effects of damaging stimuli. Theytherefore raise the possibility that the ultimate use of hspsin therapeutic procedures (see below) may be optimised bystimulating the over-expression of more than one hsp toproduce an optimal protective effect. This is reinforced bythe findings of Lau et al (1997) who were unable todemonstrate any protective effect of over-expressinghsp60 or hsp10 individually in cultured cardiac cells butdid observe a protective effect when both proteins wereover-expressed together.Hence, these studies do clearly demonstrate that overexpressionof individual hsps can protect cultured neuronalor cardiac cells against different death-inducing stimuli,extending the results with prior mild hsp-inducing stresses.Moreover, they demonstrate that the protective effect ofindividual hsps may vary with the cell type beinginvestigated and the nature of the stress being used andalso, provide the first suggestion that hsp27 may have amore potent protective effect than hsp70 in the nervoussystem. These findings reinforce the need to extend thesein vitro studies on over-expression of the hsps to theirover-expression in the intact animal. Such studies arediscussed in the next section.IV. Protective effect of hsps in vivoIn keeping with the strong focus in the hsp field onthe major inducible hsp, hsp70, transgenic animals overexpressingthis hsp were reported by several groups in1995-96 (Marber et al, 1995; Plumier et al, 1995; Radfordet al, 1996). In their initial analysis of these mice, all thesegroups focused primarily on the potential protective effectof hsp70 in the heart. In all cases, they were able todemonstrate that such over-expression of hsp70 was ableto protect the heart against the damaging effects ofischaemia using a variety of assays such as infarct size,creatine kinase release, recovery of high-energy phosphatestores and correction of metabolic acidosis. Moreover, in asubsequent study, it was demonstrated that such aprotective effect could also be observed againstmyocardial dysfunction caused by a brief ischaemia whichwas insufficient to produce an infarct (Trost et al, 1998).These studies thus establish for the first time, that theover-expression of a single hsp in vivo in the intact animalis sufficient to protect a specific organ, namely the heart,against the damaging effects of a stressful stimulus. Inview of the clear evidence demonstrating that there is alsoa protective effect for the hsps in neuronal cells in vitro(see above), it is not surprising that these hsp70 transgenicanimals were rapidly used in attempts to demonstrate thatover-expression of hsp70 would also protect the brain ofthe intact animal from specific stresses. In general,however, the results of these studies have been far moreequivocal than the corresponding studies in the heart, withprotection being observed against some insults but notagainst others (Plumier et al, 1997; Rajdev et al, 2000; Leeet al, 2001). Thus, for example, Lee et al, (2001) foundthat one of the hsp70 transgenic mouse strains showed noreduction of infarct size or enhanced survival of neuronalcells following cerebral ischaemia and similar results werealso obtained using a different strain by Plumier et al,(1997) in terms of infarct size and striatal neuronesurvival, although they did observe enhanced survival ofhippocampal neurones.In view of these results and the apparent highlypotent protective effect of hsp27 in cultured neuronal cells,we have recently prepared the first transgenic mice overexpressinghsp27 in the brain and other tissues (Akbar etal, 2003). Most interestingly, the hsp27 over-expressingtransgenic animals showed very clear protective effectswhen treated with kainate, which normally inducesconsiderable cell death in the CA3 region of thehippocampus. Thus, whilst control wild type litter-matesshowed a 33% cell loss in this region when treated withkainate compared to control untreated animals, nosignificant cell death was observed in the kainate-treatedtransgenic animals compared to untreated transgenicanimals. Moreover, this effect at the cellular level wasaccompanied by a significant effect on reducing mortalityin the hsp27 over-expressing animals, with on average17% of hsp27 transgenic mice dying following kainatetreatment compared to 38% of wild type mice, similarlytreated.As well as these effects on survival, both at thewhole animal and the individual neurone level, the hsp27mice also showed much milder seizures throughout thefour-hour observation period following kainate treatment,compared with the wild type animals (Figure 2).Figure 2. Reduced seizure activity in hsp27 transgenic mice (tg)compared to wild type mice (wt) following kainate treatment.250

Gene Therapy and Molecular Biology Vol 7, page 251Hence, hsp27 over-expression has a clear potent protectiveeffect in the nervous system in terms of kainate toxicity,both at the level of whole animal seizure activity andsurvival and at the level of vulnerable neurones within thehippocampus (Akbar et al, 2003). It will evidently be ofconsiderable interest to extend this study by determiningwhether hsp27 can have protective effects against otherdamaging stimuli in the nervous system including, forexample, cerebral ischaemia. It will also be of interest toconduct a detailed study of the hsp27 and hsp70 overexpressingtransgenic animals, to compare the potency ofthese different heat shock proteins against variousdamaging stimuli in both the heart and the brain and toconfirm or deny the suggestion that hsp27 may be morepotent than hsp70 in protecting the nervous system.It is already clear however, that hsp27 does have a clearprotective effect in the nervous system in vivo as well as invitro. This reinforces the need to conduct studies usingdifferent individual hsps and investigating their protectiveeffects, rather than simply focusing on the major heatshock protein hsp70.V. Therapeutic potential of hspsThe experiments described in earlier sections, clearlysuggest that procedures which elevate hsp levels in theheart or the brain may be of significant benefit, forexample, during reperfusion following a period ofischaemia, or in patients undergoing neuronal cell loss dueto neurodegenerative diseases such as Alzheimer’s orParkinson’s diseases. Similarly, elevation of hsp levelsmay be beneficial during cardiac bypass or to preservedonor heart function prior to transplantation. Indeed, inview of the use of cold storage during transportation priorto transplantation, it is of particular interest that reduced aswell as elevated temperature induces hsp expression in theheart (Laios et al, 1997). Moreover, a mild heat treatmentprior to hypothermic storage has been shown to enhancesubsequent recovery of the heart (Gowda et al, 1998a).Such temperature-based manipulations of the hspsmay thus have a role to play in cardiac transplantationprocedures. Similarly, it has been shown that a protectiveeffect across the whole heart can be obtained by using athermal probe to produce local heating of the heart (Goudaet al, 1998b) suggesting that a similar procedure could beused therapeutically.The induction of the hsps by stressful stimuli such aselevated temperature (for review see Morimoto, 1998) orischaemia (Nishizawa et al, 1999) is mediated by the heatshock transcription factor (HSF-1). Thus followingexposure to elevated temperature, the cytoplasmic HSF-1monomer, forms a trimer and moves to the nucleus whereit binds to its target sites (known as heat shock elements)in the regulatory regions of the hsp genes and, followingHSF-1 phosphorylation, it induces hsp gene expression.Interestingly, the induction of the hsps by stressfulstimuli diminishes with age in a variety of tissuesincluding the heart and this has been shown to be due toimpaired activation of HSF-1 by stress in the aged heart(Locke and Tanguay, 1996). Moreover, this effect isassociated with a reduced protective effect of mild heatshock or ischaemia against a subsequent severe ischaemicstress in aged hearts (Locke and Tanguay, 1996; Fenton etal, 2000). As many of the situations where the protectiveeffect of hsps would be valuable involve elderlyindividuals, this suggest that other procedures notinvolving stressful stimuli or HSF-1 may be required forthe therapeutic induction of the hsps.Obviously, such an approach would be dependent onover-expression of the hsps actually having a protectiveeffect in aged cells. To test this, we have recently used ourHSV vectors to over-express individual hsps in neuronesfrom aged rats or peripheral blood lymphocytes from agedhumans. These experiments (Alsbury et al, submitted)clearly showed that hsps do have a protective effect incells from aged humans or animals, if they aresuccessfully over-expressed. Hence, it is indeedappropriate to try and identify procedures which result inover-expression of the hsps in a non-stressful manner.Such procedures can be divided into pharmacological andgene therapy procedures.A. Pharmacological methodsAlthough the hsps were identified on the basis oftheir induction by stressful procedures, they are alsoinduced naturally by specific non-stressful procedures (forreview see Latchman, 1998). For example, the cytokinesinterleukin-6 (Stephanou et al, 1997) and interleukin-10(Ripley et al, 1999) can induce hsp gene expression in anon-stressful manner. It has been shown that theseinducers do not act via HSF-1 but activate hsp expressionvia other transcription factors such as NF-IL6 and STAT3(for review see Stephanou and Latchman, 1999). Based onthe use of these inducers in non-cardiac cells, wedemonstrated that the interleukin-6-like cytokinecardiotrophin-1 (CT-1) was able to induce hsp synthesis incultured cardiac cells and that such treatment protectsthem against subsequent exposure to severe thermal orischaemic stress (Stephanou et al, 1998). Similar inductionof the hsps in cultured cardiac cells and protection againstsubsequent severe stress is also observed with the tyrosinekinase inhibitor herbimycin A (Morris et al, 1996; Condeet al, 1997).These protective effects in cardiac cells in culturehave also been extended to the intact heart. For example,Vigh et al, (1997) showed that bimoclomol, a novelhydroxylamine derivative was able to induce hsp synthesisin the intact perfused heart ex vivo and to produce aprotective effect against a subsequent ischaemia.Similarly, Meng et al, (1996) demonstrated thatnorepinephrine treatment of an intact rat resulted in hspinduction in the heart and protection against ischaemiawhen the heart was subsequently perfused ex vivo.Several compounds thus exist which can induce hspsand produce a protective effect, although it should be notethat in no case has this protective effect been directlyshown to be due to the ability to induce hsps. Before theprotective effect of any of these compounds could beexploited clinically, it is evidently necessary to investigatewhether their protective effect in the heart can be achievedwithout any significant side-effects. For example, CT-1was originally identified on the basis of its ability toinduce cardiac hypertrophy (Pennica et al, 1995) whilst251

Latchman: Protective effect of heat shock proteinsherbimycin as a tyrosine kinase inhibitor is likely to havesignificant effects on cellular growth and division.Similarly, none of these compounds has as yet beentested to see whether, unlike stressful procedures, they caninduce hsps in cells from aged animals or humans. This isof particular importance in the case of bimoclomol, whichappears to be non-toxic but which has recently been shownto act by targeting HSF-1 (Hargitai et al, 2003).Nonetheless, the identification of compounds able toinduce the hsps without inducing a full stress response ishighly promising and suggests that pharmacologicalinduction of hsp synthesis may be a viable therapy in thenot too distant future.B. Gene therapy proceduresClearly any potential side-effects of compoundswhich can induce hsps could be avoided if hsp genes couldbe efficiently delivered to the heart or the brain in vivo.Since transgenic procedures are evidently not applicable inhumans, this will require the development of proceduresable to safely and efficiently deliver hsp genes to the heartof individual patients.Encouragingly, it has been shown that the hsp70gene within a plasmid vector can be delivered to the heartvia intra-coronary infusion of liposome particlescontaining it. The elevated expression of hsp70 producedby this means confers effective protection againstsubsequent ischaemia-reperfusion (Suzuki et al, 1997) orendotoxin-induced cardiac damage (Meldrum et al, 1999).These experiments are of considerable importance sincethey demonstrate that hsp70 over-expression can have aprotective effect not only in a transgenic animal but also ina situation directly relevant to the human case where hspover-expression is produced in the adult heart byintroduction of the gene construct.As well as these experiments in the heart, similarsuccessful delivery of hsps to the intact nervous systemhas been reported. Indeed, Benn et al, (2002) have used anadenovirus vector over-expressing hsp27 to rescue sciaticmotor neurones in the intact animal from the cell deathwhich normally follows nerve transection. Similarly, usingour HSV vector over-expressing hsp27, we havedemonstrated a protective effect of inoculation of the brainwith this virus against kainate-induced cell death (Kalwyet al, 2003), reproducing the protective effect of hsp27 intransgenic animals.If these reports of the successful in vivo delivery andprotective effect of hsps can be followed up using viral ornon-viral vectors which combine sufficient efficiency withthe safety required for human use, then it may be possibleto contemplate gene therapy-type procedures with hsps.The results presented in this review however,indicate that it is necessary to develop such proceduresalways bearing in mind the fact that the protective effectof hsps can differ in different cell types and in differentsituations. Any attempt to use the hsps therapeuticallytherefore, has to be preceded by careful studies identifyingthe optimal hsp or combination of hsps to produce aprotective effect in a particular situation. This isillustrated, for example, by our experiments using aconstitutively active form of the heat shock transcriptionfactor, HSF1, which should induce a range of hsps. Whenan HSV vector expressing this HSF1 mutant was used toinfect neuronal cells, it induced accumulation of hsp70 butnot of hsp27 and therefore, did not result in protectionagainst apoptosis, although it was able to produceprotection against heat shock and ischaemia (Wagstaff etal, 1998).Nonetheless, the clear protective effect of the hsps inthe cardiac and nervous systems offers hope fortherapeutic procedures which enhance endogenous orexogenous hsp expression.ReferencesAkbar MT, Lundberg AM. C Liu K, Vidyadaran S, Wells, K. E,Dolatshad H, Wynn S, Wells DJ, Latchman DS and deBelleroche J, (2003) The neuroprotective effects of heatshock protein 27 overexpression in transgenic animalsagainst Kainate-induced seizures and hippocampal cell death.J Biol Chem 278, 19956-19965.Amin V, Cumming DVE and Latchman DS (1996) Overexpressionof heat shock protein 70 protects neuronal cellsagainst both thermal and ischaemic stress but with differentefficiencies. Neurosci Lett 206, 45-48.Amin V, Cumming DVE, Coffin RS and Latchman DS (1995)The degree of protection provided to neuronal cells by a preconditioningstress correlates with the amount of heat shockprotein it induces and not with the similarity of thesubsequent stress. Neurosci Lett Neurosci Lett 200, 85-88.Amrani M, Latif N, Morrison K, Gray C, Jayakumar J, Corbett J,Goodwin AT, Dunn MJ and Yacoub MH, (1998) Relativeinduction of heat shock protein in coronary endothelial cellsand cardiomyocytes, implications for myocardial protection.J Thorac Cardiovasc Surg 115, 200-209.Barbe MF, Tytell M, Gower DJ and Welch, WJ, (1988)Hyperthermia protects against light damage in the rat retina.Science 241, 1817-1820Benn SC Perrelet, D, Kato AC Scholz J, Decosterd I, MannionRJ, Bakowska JC and Woolf C. J, (2002) Hsp27Upregulation and phosphorylation is required for injuredsensory and motor neuron survival. Neuron 36, 45-56Brar, BS, Stephanou A, Wagstaff MJD, Coffin RS, Marber MS,Engelman, G and Latchman DS (1999) Heat shock proteinsdelivered with a virus vector can protect cardiac cells againstapoptosis as well as against thermal or ischaemic stress. JMol Cell Cardiol 31, 135-146.Chopp, M Chen H, Ho, K-L, Dereski MO, Brown, E, Hetzel, F.W and Welch, K. M, (1989) Transient hyperthermia protectsagainst subsequent forebrain ischemic cell damage in the rat.Neurology 39, 1396-1398.Collins, P. and Hightower LE, (1982) Newcastle disease virusstimulates the cellular accumulation of stress (heat shock)mRNAs and proteins. J Virol 44, 703-707.Conde AG, Lau SS, Dillmann, H and Mestril R, (1997) Inductionof heat shock proteins by tyrosine kinase inhibitors in ratcardiomyocytes and myogenic cells confers protectionagainst simulated ischemia. J Mol Cell Cardiol 26, 1927-1938.Cumming DVE, Heads RJ Coffin RS, Yellon DM and LatchmanDS (1996a) Pharmacological preconditioning of primary ratcardiac myocytes by FK506. Basic Res Cardiol 91, 367-373.Cumming DVE, Heads RJ, Watson A, Latchman DS and YellonDM, (1996b) Differential protection of primary ratcardiocytes by transfection of specific heat stress proteins. JMol Cell Cardiol 28, 2343-2349.252

Gene Therapy and Molecular Biology Vol 7, page 253Currie RW, Karmazyn, M and KloC M, (1988) Heat-shockresponse is associated with enhanced postischemicventricular recovery. Circ Res 63, 543-549.Donnelly, T. J, Sievers RE, Vissern, F. L. J, Welch, W. J andWolfe C. L, (1992) Heat shock protein induction in rathearts. A role for improved myocardial salvage afterischemia and reperfusion? Circulation 85, 769-778.Ellis RJ (1990) Molecular Chaperones. Semin Cell Biol 1, 1-72.Fenton RA, Dickson EW, Meyer TE and Dobson JG, (2000)Ageing reduces the cardioprotective effect of ischemicpreconditioning in the rat heart. J Mol Cell Cardiol 32,1371-1375.Fink SL Chang LK, Ho DY and Sapolsky RM, (1997) Defectiveherpes simplex virus vectors expressing the rat brain stressinducibleheat shock protein 72 protect cultured neuronsfrom severe heat shock. J Neurochem 68, 961-969.Gowda A, Yang C Asimakis GK, Rastegar, S and Motamedi M,(1998a) Heat shock improves recovery and providesprotection against global ischemia after hypothermic storage.Ann Thorac Surg 66, 1991-1997.Gowda A, Yang C-j Asimakis GK, Ruef J, Rastegar S, RungeMS and Motamedi M, (1998b) Cardioprotection by localheating, improved myocardial salvage after ischemia andreperfusion. Ann Thorac Surg 65, 1241-1247.Hargitai J, Lewis H, Boros I, Racz T, Fiser A, Kurucz I,Benjamin I, Vigh L, Penzes, Z Csermely, P and LatchmanDS (2003) Bimoclomol a heat shock protein co-inducer Actsby the prolonged activation of heat shock factor-1. BiochemBiophys Res Commun 307, 689-695.Heads RJ, Latchman DS and Yellon DM, (1994) Stable highlevel expression of a transfected human HSP70 gene protectsa heart-derived muscle cell line against thermal stress. J MolCell Cardiol 26, 695-699.Heads RJ, Yellon DM and Latchman DS (1995) Differentialcytoprotection against heat stress or hypoxia followingexpression of specific stress protein genes in myogenic cells.J Mol Cell Cardiol 27, 1669-1678.Hutter MM, Sievers RE, Barbosa V and Wolfe C. Heat shockprotein induction in rat hearts. A direct correlation betweenthe amount of heat shock protein induced and the degree ofmyocardial protection. Circulation 1994, 89, 353-360.Kalwy S, Akbar MT Coffin S, de Belleroche J and Latchman DS(2003) Heat shock protein 27 delivered via a herpes simplexvirus vector can protect neurones of the hippocampus againstkainic acid-induced cell loss. Mol Brain Res 111, 91-103.Karmazyn M, Mailer, K and Currie RW, (1990) Acquisition anddecay of heat-shock enhanced postischemic ventricularrecovery. Am J Physiol 259, H421-H431.Kh and ijan, E. W. and Turler H, (1983) Simian virus 40 andpolyoma virus induce synthesis of heat shock proteins inpermissive cells. Mol Cell Biol 3, 1-8.Kitagawa K, Matsumoto M, Tagaya M, Hata R, Keda H NinobeM, Handa N, Fukunaga R, Kimura K, Mikshiba, K andKamada T, (1990) 'Ischaemic tolerance' phenomenon foundin the brain. Brain Res 528, 21-24.Kure S, Tuminaga T, Yoshimoto T, Tada, K and Narisawa K,(1991) Glutamate triggers internucleosomal DNA cleavagein neuronal cells. Biochem Biophys Res Commun 179, 39-45.La Thangue NB. and Latchman DS (1988) A cellular proteinrelated to heat shock protein 90 accumulates during herpessimplex virus infection and is over-expressed in transformedcells. Exp Cell Res 178, 169-179.Laios E, Rebeyka, IM and Prody CA, (1997) Characterization ofcold-induced heat shock protein expression in neonatal ratcardiomyocytes. Mol Cell Biochem 173, 153-159.Latchman DS (1998) Heat shock proteins, protective effect andpotential therapeutic use. Int J Mol Med 2, 375-381.Lau S, Patnaik N, Sayen MR and Mestril R, (1997) Simultaneousoverexpression of two stress proteins in rat cardiomyocytesand myogenic cells confers protection against ischemia.Circulation 96, 2287-2294.Lee JE, Yenari MA, Sun GH, Xu L, Emond MR Cheng, D,Steinberg GK and Giffard RG, (2001) Differentialneuroprotection from human heat shock protein 70 overexpressionin in vitro and in vivo models of ischaemia andischaemia-like conditions. Exp Neurol 170, 129-139.Leger JP, Smith, F. M and Currie RW, (2000) Confocalmicroscopic localisation of constitutive and heat shockinducedproteins HSP70 and HSP27 in the rat heart.Circulation 102, 1703-1709.Li GC. and Laszlo A, (1985) Amino acid analogues whileinducing heat shock protein sensitive cells to thermaldamage. J Cell Phys 122, 91-97.Lindquist S and Craig EA, (1988) The heat shock proteins. AnnuRev Genet 22, 631-677.Locke M and Tanguay RM, (1996) Diminished heat shockresponse in the aged myocardium. Cell Stress Chaperones1, 251-260.Lowenstein DH Chan, P. H and Miles MF, (1991) The stressprotein response in cultured neurons Characterization andevidence for a protective role in excitotoxicity. Neuron 7,1053-1060.Mailhos C, Howard MK and Latchman DS (1994) Heat shockproteins hsp90 and hsp70 protect neuronal cells from thermalstress but not from programmed cell death. J Neurochem 63,1787-1795.Mailhos C, Howard MK and Latchman DS (1993) Heat shockprotects neuronal cells from programmed cell death byapoptosis. Neuroscience 55, 621-627.Marber MS, Latchman DS, Walker JM and Yellon DM, (1993)Cardiac stress protein elevation 24 hours after briefischaemia or heat stress is associated with resistance tomyocardial infarction. Circulation Res 88, 1264-1272.Marber MS, Mestril R Chi SH and Sayen MR, (1995)Overexpression of the rat inducible 70kDa heat shock proteinin a transgenic mouse increases the resistance of the heart toischemic injury. J Clin Invest 95, 1446-1456.Marber MS, Walker JM, Latchman DS and Yellon DM, (1994)Myocardial protection following whole body heat stress inthe rabbit is dependent on metabolic substrate and is relatedto the amount of the inducible 70kb Dalton heat shockprotein. J Clin Invest 93, 1087-1094.Martin JL, Hickey, E, Weber LA, Dillmann WH and Mestril R,(1999) Influence of phosphorylation and oligomerization onthe protective role of the small heat shock protein 27 in ratadult cardiomyocytes. Gene Expr 7, 349-355.Martin JL, Mestril R, Hilal-Dandan R, Brunton LL and DilmannWH, (1997) Small heat shock proteins and protection againstischaemic injury in cardiac myocytes. Circulation 96, 4343-4348.Meldrum DR, Meng X, Shames BD, Pomerantz B, DonnahooKK, Banerjee A and Harken AH, (1999) Liposomal deliveryof heat-shock protein 72 into the heart prevents endotoxininducedmyocardial contractile dysfunction. Surgery 126,135-141.Meng, X, Brown JM Ao L, Banerjee A and Harken AH, (1996)Norepinephrine induces cardiac heat shock protein 70 anddelayed cardioprotection in the rat through α1 adrenoceptors.Cardiovasc Res, 32, 374-383.Mestril R Chi SH, Sayen R, O'Reilly, K and Dilmann WH,(1994) Expression of inducible stress protein 70 in heartmyogenic cells confers protection against simulatedischaemia-induced injury. J Clin Invest 93, 759-767.Morimoto RI, (1998) Regulation of the heat shock transcriptionalresponse Cross talk between a family of heat shock factors,253

Latchman: Protective effect of heat shock proteinsmolecular chaperones and negative regulators. Genes Dev12, 3788-3796.Morris SD, Cumming DVE, Latchman DS and Yellon DS,(1996) Specific induction of the 70kDa heat stress proteinsby the tyrosine kinase inhibitor herbimycin-A protects ratneonatal cardiomyocytes A new pharmacological route tostress protein expression. J Clin Invest 97, 706-712.Nishizawa J Nakai A, Matsuda K, Komeda M, Ban, T andNagata K, (1999) Reactive oxygen species play an importantrole in the activation of heat shock factor 1 in ischemicreperfusedheart. Circulation 99, 934-941.Norton, P. M. and Latchman DS (1989) Levels of the 90kd heatshock protein and resistance to glucocorticoid mediated cellkilling in a range of human and murine lymphocyte celllines. Genes Dev 33, 149-154..Nowak, T. S. J, (1985) Synthesis of a stress protein followingtransient ischaemia in the gerbil. J Neurochem 45, 1635-1641.Parsell DA and Lindquist S, (1993) The functional of heat-shockproteins in stress tolerance, Degradation and reactivation ofdamaged proteins. Annu Rev Biochem 27, 437-496.Pennica D, King, KL, Shaw, K. J, Luis, E, Rullamas J, Luoh SM,Darbonne WC, Knutzon DS, Yen R, Chien KR, Baker JBand Wood WI, (1995) Expression cloning of cardiotrophin 1a cytokine that induces cardiac myocyte hypertrophy.Proc.Natl.Acad.Sci.USA, 92, 1142-1146.Plesset J, Palm C, and McLaughlin CS, (1982) Induction of heatshock proteins and thermotolerance by ethanol inSaccharomyces cerevisiae. Biochem Biophys Res Commun108, 1340-1345.Plumier JCL, Krueger, AM, Currie RW, Kontoyiannis G, KolliasG and Pagoulatos GN, (1997) Transgenic mice expressingthe human inducible hsp70 have hippocampal neuronsresistant to ischemic injury. Cell Stress Chaperones 2, 162-167.Plumier JCL, Ross, B. M Currie RW Angelidis C. E, Kazlaris H,Kollias G and Pagoulatos GN, (1995) Transgenic miceexpressing the human heat shock protein 70 have improvedpostischemic myocardial recovery. J Clin Invest 95, 1854-1860.Polla BS, (1988) A role for heat shock proteins in inflammation?Immunol Today 9, 134-137.Radford NB, Fina M, Benjamin IJ, Moreadith RW, Graves KH,Zhao, P, Gavva S, Wiethoff A, Sherry AD, Malloy C. R andWilliams RS, (1996) Cardioprotective effects of 70kDa heatshock protein in transgenic mice. Proc.Natl.Acad.Sci.USA93, 2339-2342.Rajdev S, Hara K, Kokubo Y, Mestril R, Dillmann W, WeinsteinPR and Sharp FR, (2000) Mice overexpressing rat heat shockprotein 70 are protected against cerebral infarction. AnnNeurol 47, 782-791.Rebbe NF, Ware J, Bertina RM, Modrich P and Stafford DW,(1987) Nucleotide sequence of a cDNA for a member of the90kDa heat shock protein family. Gene 53, 235-245.Ripley, B. J. M, Stephanou A, Isenberg DA and Latchman DS(1999) Interleukin-10 activates heat shock protein 90β geneexpression. Immunology 97, 226-231.Ritossa, F, (1962) A new puffing pattern induced andtemperature shock and DNP in Drosophila. Experientia 18,571-573.Robertson JD, Datta K, Biswal SS and Kehrer JP, (1999) Heatshockprotein 70 antisense oligomers enhance proteasomeinhibitor-induced apoptosis. Biochem J 344, 477-485.Rordorf G, Koroshetz, W. J and Bonventre JV, (1991) Heatshock protects cultured neurons from glutamate toxicity.Neuron 7, 1043-1052.Stephanou A and Latchman DS Transcriptional regulation of theheat shock protein genes by STAT family transcriptionfactors. Gene Expr 1999, 7, 311-319.Stephanou A Amin V, Isenberg DA Akira S, Kishimoto, T andLatchman DS (1997) Interleukin-6 activates heat shockprotein 90β gene expression. Biochem J 321, 103-106.Stephanou A, Brar B, Heads R, Knight RD, Marber MS, Pennica,D and Latchman DS (1998) Cardiotrophin-1 induces heatshock protein accumulation in cultured cardiac cells andprotects them from stressful stimuli. J Mol Cell Cardiol 30,849-855.Suzuki K, Sawa, Y, Kaneda, Y, Ichikawa H, Shirakura, R andMatsuda H, (1997) In vivo gene transfection with heat shockprotein 70 enhances myocardial tolerance to ischemiareperfusioninjury in rat. J Clin Invest 99, 1645-1650.Suzuki K, Sawa Y, Kaneda Y, Ichikawa H, Shirakura R andMatsuda H, (1998) Overexpressed heat shock protein 70attenuates hypoxic injury in coronary endothelial cells. JMol Cell Cardiol 30, 1129-1136.Trost SU, Omens JH, Karlon, W. J, Meyer M, Mestril R CovellJW and Dillmann, W. H, (1998) Protection againstmyocardial dysfunction after a brief ischemic period intransgenic mice expressing inducible heat shock protein 70. JClin Invest 101, 855-862.Uney JB, Kew C. N. N, Staley K, Tyers, P and Sofroniew MV,(1993) Transfection mediated expression of human hsp70iprotects rat dorsal root ganglion neurones and glia from heatshock. FEBS Letters 334, 313-317.Vigh L, Literati, P. N, Horvath I, Torok Z, Balogh G, Glatz A,Kovacs E, Boros I, Ferdinandy P, Farkas B, Jaszlits L,Jednakovits A, Koranyi, L and Maresca B, (1997)Bimoclomol A nontoxiC hydroxylamine derivative withstress protein-inducing activity and cytoprotective effects.Nature Med 3, 1150-1154.Wagstaff MJD, Collaco-Moraes Y, Smith J, de Belleroche J,Coffin RS and Latchman DS (1999) Protection of neuronalcells from apoptosis by hsp27 delivered with a herpessimplex virus-based vector. J Biol Chem 274, 5061-5069.Wagstaff MJD, Smith J, Collaco-Moraes Y, de Belleroche JS,Coffin RS and Latchman DS (1998) Delivery of aconstitutively active form of the heat shock factor using avirus vector protects neuronal cells from thermal orischaemic stress but not from apoptosis. Eur J Neurosci 10,3343-3350.Wyatt S, Mailhos C and Latchman DS (1996) Trigeminalganglion neurons are protected by the heat shock proteinshsp70 and hsp90 from thermal stress but not fromprogrammed cell death following NGF withdrawal. MolBrain Res 39, 52-56.Yellon DM. and Latchman DS (1992) Stress proteins andmyocardial protection. J Mol Cell Cardiol 24, 113-124.Yellon DM, Pasini E Cargnoni A, Marber MS, Latchman DS andFerrari R, (1992) The protective role of heat stress in theischaemic and reperfused rabbit myocardium. J Mol CellCardiol 24, 895-907.254

Gene Therapy and Molecular Biology Vol 7, page 255Lung cancer gene therapyReview ArticleGene Ther Mol Biol Vol 7, 255-272, 2003Kexia Cai 1 , Mai Har Sham 2 , Paul Tam 3 , Wah Kit Lam 4 and Ruian Xu 1*1 Gene Therapy Laboratory, IMB, 2 Department of Biochemistry and 3 Department of Surgery and 4 Department of Medicine,The University of Hong Kong, Hong Kong__________________________________________________________________________________*Correspondence: RA Xu, Gene Therapy Laboratory, IMB, The University of Hong Kong, Hong Kong; Tel: (852) 22990757; Fax:(852) 2817 9488; e-mail: rxua@hkucc.hku.hkKey words: lung cancer, gene therapy, tumor suppressor genes, growth factor pathway targets, suicide gene therapy, angiogenesis,immunotherapyAbbreviations: small cell lung cancer (SCLC); non-small cell lung cancer (NSCLC); intratumoral injection of Ad-p53 (INGN 201);murine double minute-2 (MDM2); cisplatin (CDDP); active metabolite of irinotecan (CPT-11); O(6)-methylguanine-DNAmethyltransferase (MGMT); retinoblastoma (RB); fragile histidine triad (FHIT); tumor necrosis factor-related apoptosis-inducing ligand(TRAIL); triplex forming oligonucleotides (TFO); melanoma differentiation associated gene-7 (mda-7); carcinoembryonic antigen(CEA); replication-deficient adenovirus vector, Ad-mda7 (INGN 241); MAPK-activated kinases (Rsks); ). Insulin-like growth factorbinding proteins (IGFBPs); cyclooxygenase (COX)-2; ganglioside G(D2); Herpes simplex virus 1 (HSV); thymidine kinase (tk);ganciclovir (GCV); gastrin-releasing peptide (GRP); neuron specific enolase (NSE); Cre recombinase(Cre)/loxP; hypoxanthine-guaninephosphoribosyl transferase (HGPRT); Trypanosoma brucei (Tb); sodium iodide symporter (NIS); thyroperoxidase (TPO); matrixmetalloproteinase (MMP); secret form of human platelet facter 4 (Spf4); vascular endothelial growth factor (VEGF); soluble flt-1 (sFLT-1); Tie2-expressing mononuclear (TEM); dendritic cells (DCs); Lewis lung carcinoma (LLC); natural killer (NK) cells; tumor necrosisfactor receptor (TNF-R); 1,3) Galactosyl epitopes (·Gal); proliferin-related protein (PRP); interferon-inducible protein 10 (IP10);interferon (IFN)Received: 29 October 2003; Accepted: 1 December 2003;Revised: 23 December 2003; electronically published: December 2003SummaryLung cancer is the most lethal cancer worldwide. Although progress has been made in prevention, early detection,and treatment, mortality from this disease is still increasing. Current treatments in clinical trials have yielded onlyvery limited results, and it is therefore necessary to develop new therapeutic strategies. Gene therapy is a novel fieldof medicine that may signal a more promising future for patients with lung cancer. Several studies on lung cancertherapy have held out the promise of treatment methods, including the alteration of intracellular molecular defects,the introduction of suicide genes, the inhibition of angiogenesis, and the augmentation of specific antitumorimmunity. Various methods have been used to achieve specific gene transduction and effective gene expression.Clinical trials indicate that a combination of different treatment modalities is needed to obtain better results in lungcancer therapy. This review will summarize and discuss some recent advances and the potential future applicationsof gene therapy approaches in lung cancer.I. IntroductionLung cancer is the most common cause of death bycancer in both men and women, accounting for 18% of allcancer cases around the world. The average worldwideincidence of lung cancer is 37.5 per 100,000 persons,though this number varies greatly by country. Theincidence is highest in Eastern Europe and lowest inAfrica. The 5-year survival rate for lung cancer is 11%worldwide. In most countries, mortality from this diseaseis still increasing, especially in Southern and EasternEurope (Parkin et al, 1999). The American Cancer Societyestimates that 171,900 new cases of lung cancer will bediagnosed in the United States in 2003 alone. About157,200 people will die of this disease: 88,400 men and68,800 women, accounting for 28% of all cancer deaths.More people die of lung cancer than of colon, breast, andprostate cancers combined (Jemal et al, 2003).Lung cancer is divided into two main histologicgroups: small cell lung cancer (SCLC) and non-small celllung cancer (NSCLC). Approximately 80% of lung cancercases are NSCLC, with small cell lung cancer (SCLC)accounting for the remaining 20%. Lung cancer arisesfrom a series of morphological and molecular changes inwhich a normal epithelium transforms into an invasivecancer. To date, no efficient and safe therapy has yet beenintroduced for lung therapy.Gene therapy, although still a comparatively youngdiscipline, has made rapid strides in the past decade (Xu etal., 2003). Considerable efforts have been made toimprove protocols for human gene therapy. Four mainstrategies for the treatment of cancer have been reported:alteration of mutated genes; introduction of suicide genes;255

Cai et al: Lung cancer gene therapyantiangiogenic gene therapy; and immunotherapy. Clinicaltrials have already been initiated. The number of approvedprotocols in clinical trials has increased, and at least 50%are designed for cancer (Folkman, 1998). Nervethless, acentral challenge is perfecting methods for deliveringtherapeutic genes to the appropriate cells. The ideal genetransfer systems should be tailored to the specific tissue orcells requiring modification, to the needed duration ofgene action and to the desired physiological effect of thegene product. Practical and theoretical limitationscurrently exist for the application of gene therapy incancer patients, Most of these approaches have yet to passeven the most preliminary clinical tests demonstratingtheir overall safety and efficacy, but these ideas may leadto better cancer treatments in the future.The molecular changes and underlying mechanismsof lung cancer have been continuously identified. Theaccumulation of the progresses may actually offer athorough understanding of the disease and clinical context,and offer many targets for gene therapy. Theimprovements in gene transfer systems offer promise forthe development of an efficient, specific-targeted andnontoxic gene delivery system, and thus there is very goodreason to believe that greater success will be achieved inthe near future.II. Alteration of mutated genesGene alteration therapy is potentially a very powerfultool, targeting intracellular mutant genes of lung tumors.These gene products are specific molecular mediators ofcancer development and progression.A. Tumor suppressor genes1. p53p53 mutations, with frequencies up to 50% inNSCLC and 80% in SCLC, are the most common geneticlesions observed in lung cancers (Salgia et al, 1998).Mitsudomi et al, (2000) have shown by meta-analysis thatp53 mutation or overexpression was an indicator of poorprognosis, especially in patients with adenocarcinoma.Roth et al, (1996) first reported the use of the strategy ofreplacing p53 in the treatment of nine lung cancer patientsby local injection of retroviral vectors encoding wild typep53. Tumor regression was noted in three patients, andtumor growth stabilized in three other patients. In thesecond Phase I trial performed by this group, anadenoviral vector was used. Repeated intratumoralinjections of Ad-p53 appeared to be well tolerated,resulted in transgene expression of wild-type p53, andmediated antitumor activity in a subset of patients withadvanced NSCLC (Swisher et al, 1999). Because thesecompleted studies have demonstrated only modestresponse rates, several protocols have been developed thatcombine the p53 gene transfer approach with othertreatment modalities. No enhanced radiosensitivity ofnormal cells was noted when the ability of Ad-p53 (INGN201) in NSCLC cell lines and human fibroblast cells wascompared (Kawabe et al, 2001).Recently, the same group reported the results of theirPhase II study on 19 patients with NSCLC. The groupfound that intratumoral injection of Ad-p53 (INGN 201) incombination with radiation therapy was well tolerated, andalso demonstrated evidence of tumor regression in primaryinjected tumors. Additionally, they found BAK expressionwas significantly increased 24h after injection of Ad-p53(INGN 201), providing the first demonstration ofinduction of an apoptotic pathway by tumor suppressorgene expression in actual human cancers (Swisher et al,2003). Schuler et al, (1998) reported the results of anotherPhase II trial, in which Ad-p53 gene therapy appeared toprovide no additional benefit in patients receiving first-linechemotherapy for advanced NSCLC. To elucidate thecombined effects of p53 gene transfer, chemotherapy, andradiation therapy on lung cancer growth in vitro and invivo, Nishizaki et al, (2001) evaluated the synergistic,additive, or antagonistic efficacy of these therapeuticagents in four human NSCLC cell lines at the ID50 andID80 levels. Synergistic inhibitory effects on tumor cellgrowth were noted both in an in vitro and a murine modelwith H1299 and A549 xenografts. Using two humancancer cell lines, H157 and H1299, Osaki et al, (2000)evaluated several anticancer agents, and suggested thatcisplatin (CDDP) and an active metabolite of irinotecan(CPT-11) would be suitable candidates for a combinationof chemotherapy and gene therapy for NSCLC.Srivenugopal et al, (2001) demonstrated that enforcedexpression of wild-type p53 curtailed the transcription ofO(6)-methylguanine-DNA methyltransferase (MGMT), aDNA repair protein that confers tumor resistance on manyanticancer alkylating agents. This finding suggests that acombination of MGMT-directed alkylators with the p53gene should achieve improved antitumor efficacy. Severalstudies (see below) have indicated the benefits ofcombination therapy on lung cancer, as one of thefunctions of p53 is to keep the cell from progressingthrough the cell cycle if there is damage to DNA present(Lowe et al, 1993).Since wild type p53 reconstitution was notcompletely effective in all cases, mutants of p53 wereexplored for their ability to prevent p53 inactivation. Ap53 derivative vector, in which the p53 domains bound byits inhibitor (murine double minute-2, MDM2) werereplaced, was significantly more efficient than the p53vector in tumor models overexpressing MDM2. Both invitro and in vivo, a higher inhibition of tumor growth withthe mutant p53 vector correlated with a higher induction ofapoptosis (Bougeret et al, 2000).2. RBAbnormalities of retinoblastoma (RB), consisting ofthe tumor suppressor pRb/p105 and related protein p107and pRB2/p130, are detected in more than 90% of SCLCsand in 15% to 30% of NSCLCs (Forgacs et al, 2001).Immunohistochemical studies of the expression patterns ofthe Rb family members in 235 specimens of lung cancersuggest an independent role for pRB2/p130 in thedevelopment and/or progression of human lung carcinoma(Baldi et al, 1996; 1997). Loss of pRb2/p130 expression is256

Gene Therapy and Molecular Biology Vol 7, page 257also associated with an unfavorable clinical outcome inlung cancer (Caputi et al, 2002). The effects of expressingpRB2/p130 in a human lung adenocarcinoma cell line H23have been analyzed, and it has been reported thatretrovirus-mediated delivery of wild-type RB2/p130 toH23 potently inhibits tumorigenesis in vitro and in vivo.When tested in established tumors in nude mice, thisapproach reduced tumor mass twelve times moreeffectively than the control viruses (Claudio et al, 2000).These results offer promise for the potential future use ofRB2/p130 in lung cancer gene therapy.3. p16In many instances, p53 and Rb are activated topromote senescence by the two products of p16 gene,protein p16(INK4a) and protein p14(ARF) (Lowe et al,2003). p16(Ink4a) engages the Rb pathway by inhibitingcyclin D-dependent kinases that would otherwisephosphorylate and inactivate Rb. p14(ARF), on the otherhand, increases the growth suppressive function of p53 byinterfering with its negative regulator, MDM2. Clinicalstudies suggest that p16(INK4a) is a positive prognosticmarker for NSCLC (Gessner et al, 2002). Several studieshave suggested that polygene therapy with the p16 andp53/Rb gene may contribute to a greater antitumor effect(Kawabe et al, 2000; Tango et al, 2002). In vitro studiesusing adenoviral vector have demonstrated that p-16(INK4a)-mediated cytotoxicity is closely associatedwith the presence of functional pRb. Kawabe et al, (2000)also used adenoviral delivery systems to show thatp16(INK4a) mediated radiosensitization of tumor cellsdepended on intracellular p53 status. Coinfection of Adp14(ARF)and Ad-p53 in human lung cancer cells resultedin a significantly higher in vitro cytotoxicity than Ad-p53infection alone, coupled with an increase in expression ofp53-inducible genes. Intratumoral injection of these twovectors significantly inhibited tumor growth in vivo(Tango et al, 2002). These results suggest that the p16gene should be considered for possible applications inhuman lung cancer therapy.4. FHIT geneAlteration of the FHIT (fragile histidine triad) geneoccurs as an early and frequent event in lungcarcinogenesis (Sozzi et al, 1998). Small cell lung tumors(80%) and non-small cell lung cancers (40%) have shownabnormalities in RNA transcripts of FHIT, and 76% of thetumors exhibited loss of FHIT alleles (Sozzi et al, 1996).FHIT-negative patients tend to correlate with a worseprognosis (Pavelic et al, 2001). Seven lung cancer celllines and three cervical cancer cell lines showed inductionof apoptosis in all Fhit-negative cell lines, together withactivation of caspase-8 by adenovirus vector-mediatedFHIT gene expression (Roz et al, 2002). Consistently,increased level of BAK in FHIT-reexpressing cells linkedthe tumor-suppressor activity of FHIT to its proapoptoticfunction (Sard et al, 1999). In vivo reintroduction of wildtype FHIT not only suppressed the tumorigenicity of lungcancer cells in nude mice (Ji et al, 1999), but also inhibitedtumor development in heterozygous Fhit(+/-) knockoutmice, which were prone to tumor development aftercarcinogen exposure (Dumon et al, 2001). With animproved liposome vector, successful treatment of primaryand disseminated murine tumors and human lung tumorxenografts was achieved. This treatment suppressedtumor growth and prolonged animal survival with minimaltoxicity (Ramesh et al, 2001). Further studies on thisinteresting gene are required, but FHIT gene therapy mayeventually offer a promising clinical approach for theprevention and treatment of lung cancer.5. p27p27(Kip1), a member of the Cip/Kip family ofcyclin-dependent kinase inhibitors, may also function as apotential tumor suppressor gene. Significantly reducedp27(Kip1) expression is frequent in NSCLC, and isassociated with shortened patient survival (Esposito et al,1997; Yatabe et al, 1998). p27(Kip1) might play a distinctbiological role in SCLC as a CDK inhibitor, conferring onSCLC cells the ability to escape from apoptosis underconditions unfavorable for cell growth (Masuda et al,2001). The transfer of full-length human p27 cDNA by anadenoviral vector into lung cancer cell lines showed thatinduction of growth arrest and apoptosis by overexpressionof p27 required expression of pRB (Naruse etal, 2000). With two adenoviruses expressing wild-type p27(Ad-p27wt) and mutant p27(Ad-p27mt), Park et al, (2001)demonstrated the anti-tumor effects of p27 in vitro and invivo in nude mice, and demonstrated that Ad-p27mt,which was believed to bind cyclin E/CDK2 more stably,had more potent anti-tumor effects than Ad-p27wt.B. Apoptotic signaling checkpoints inresponse to DNA damageDefects in apoptosis underpin both tumorigenesisand drug resistance, and because of these defectschemotherapy often fails (Johnstone et al, 2002). Tumorresponse to radiotherapy is regulated by endothelial cellapoptosis (Garcia-Barros et al, 2003). SCLC and NSCLCrepresent the two major categories of lung cancer, andthey differ in their sensitivity to apoptosis (Joseph et al,1999). It is therefore important to understand themolecular events that contribute to drug- and radiationinducedapoptosis, and how tumors evade apoptotic death,as it may be possible to harness this knowledge for noveltherapeutic approaches.1. BCL-2 familyThe BCL-2 family of proteins, consisting of bothantagonists (e.g. BCL-2, BCL-XL) and agonists (e.g. Bax,Bak) that regulate apoptosis and compete throughdimerization (Reed 1994), are among the most closelystudied apoptotic molecules in lung cancer. p53 is aregulator of bcl-2 and Bax gene expression in vitro and invivo (Miyashita et al, 1994), and Bax acts as a tumorsuppressor and as a component of the p53-mediatedapoptotic response (Yin et al, 1997). Tumors harboring a257

Cai et al: Lung cancer gene therapyBcl2-mediated apoptotic block undergo a drug-inducedcytostasis involving the accumulation of p53, p16(INK4a), and typically acquire p53 or INK4a mutationsupon progression to a terminal stage (Schmitt et al, 2002).Bax (Kagawa et al, 2000) and Bak (Pataer et al, 2000)retained an impressive antitumor ability in the absence ofchemotherapeutic drugs, and were able to effectively killboth p53-sensitive and p53-resistant tumors in vitro and invivo. To avoid their toxicity to the packaging cell line, abinary adenoviral vector system was used. Usui et al,(2003) used the Cre-loxP system to propagateadenoviruses expressing the N-terminally truncated Bax(ΔN Bax), which was not blocked by Bcl-2 or Bcl-xl, andintratumoral injection into nude mice showed asignificantly stronger suppression of tumor growth (74%)than full-length Bax (25%). The synergic effects of Baxand tumor necrosis factor-related apoptosis-inducingligand (TRAIL) were evaluated by human telomerasereverse transcriptase promoter-driven and adenovirusmediatedgene expression in vitro and in vivo, and it wasfound that combined Bax and TRAIL therapy producedmore profound cell killing in human lung cancer lineH1299 and prolonged survival in mice with ovarian cancerxenograft (Huang et al, 2002). As these are strongproapoptotic genes, targeted expression of the genes ishighly desirable when they are used as a therapeutic agent.When Bax was expressed under the control of humanvascular endothelial growth factor (VEGF) promoter,adenovirus-mediated overexpression of Bax resulted inapoptosis in human lung cancer cells and also in normalhuman bronchial epithelial cells (Kaliberov et al, 2002).Like Bax, BID also counters the protective effect ofBCL2. Sax et al, (2002) suggested that BID was a p53-responsive ‘chemosensitivity gene’ that may enhance celldeath response to chemotherapy. Fukazawa et al, (2003)noted adenoviral Bid overexpression could induceapoptosis in NSCLC cell lines and enhancechemosensitivity in the absence of p53. The function ofBCL2 could also be blocked by silencing this gene withtriplex forming oligonucleotides (TFO) (Shen et al,2003a), or by down-regulation of its transcripts usingantisense oligonucleotides (Buck et al, 2002).2. p21 and MycActivation of the tumor suppressor p53 by DNAdamage induces either cell cycle arrest or apoptotic celldeath. The cytostatic effect of p53 is mediated bytranscriptional activation of the cyclin-dependent kinase(CDK) inhibitor p21(Cip1) (Bunz et al, 1998). In vitroexperiments have suggested that p21 could serve as amarker for biological response to p53 gene therapy (Tangoet al, 2002; Choi et al, 2000; Dubrez et al, 2001). A similarresult was later obtained from biopsy examinations: p21expression was up-regulated in NSCLC patients aftertreatment, especially when injections of higher doses ofp53-expressing adenovirus were combined withsimultaneous chemotherapy (Boulay et al, 2000). Joshi etal, (1998) have provided preliminary evidence for growthinhibition of NSCLC by p21WAF1 adenoviral genetransfer in vitro and in vivo. Myc was involved in thisapoptotic signaling in response to DNA damage byselectively inhibiting bound p53 from activatingp21(Cip1) transcription (Seoane et al, 2002).Downregulating c-myc expression by the combinationtreatment of c-myc antisense DNA with all-trans-retinoicacid resulted in inhibition of cell proliferation of small celllung cancer in vitro (Akie et al, 2000). In a Lewis lungsyngeneic drug-resistant murine tumor model,chemotherapeutic drugs in combination with c-Mycinhibition (which was specifically achieved by using nontoxicantisense DNA chemistry) suppressed tumor growthdramatically, but only with a regimen in which cisplatin ortaxol treatment preceded the antisense compound (Knappet al, 2003).3. mda-7It has been reported that adenoviral-mediatedoverexpression of the mda-7 gene exhibited cancer cellspecificgrowth inhibition irrespective of the status ofother tumor suppressor genes, such as p53, RB, and p16(Mhashilkar et al, 2001). When this attractive gene wasused in lung cancer, similar results were noted in NSCLCcells in which the product of the transgene induced G2/Mcell cycle arrest and an increase of Bax and Bak (Saeki etal, 2000). The induction of apoptosis was associated withactivation of specific caspase cascades (Saeki et al, 2000;Pataer et al, 2002). In vivo studies correlated well with invitro inhibition of lung tumor cell proliferation andendothelial cell differentiation mediated by Ad-mda7.Besides its proapoptotic properties, Ad-mda7 alsodemonstrated antiangiogenic abilities (Saeki et al, 2002).As a potent radiosensitizer, Ad-mda7 has been shown toenhance the radiation sensitivity of NSCLC cells, but notof normal human lung fibroblast lines (Kawabe et al,2002). A Phase I/II dose-escalation trial of intratumoralinjection with a replication-deficient adenovirus vector,Ad-mda7 (INGN 241), will be performed in combinationwith radiation therapy in patients with locally recurrentbreast cancer (http: //www4.od.nih.gov/oba/rac/PROTOCOL.pdf).4. Fas/Fas ligandThe interaction between Fas and Fas ligand (FasL) isinvolved in the apoptotic death of a number of cells,including lymphocytes. Hahne et al, (1996) proposed thatFasL-expressing melanoma cells might induce apoptosisof Fas-sensitive tumor infiltrating cells. Human lungcancer cells have been shown to express FasL, enablingthem to destroy T lymphocytes expressing Fas (Niehans etal, 1997). Moreover, apoptotic FasL-expressing tumorcells suppressed antitumor immunity, in contrast to thepotent tumor-specific protective immunity generated byviable FasL-expressing tumors (Tada, 2003). Direct invivo transfection of antisense FasL produced a systemicdecrease in soluble FasL, and reduced tumor growth andinvasion (Nyhus et al, 2001). However, membrane-boundFasL had opposite effects. Tada et al, (2002) demonstratedthat forced expression of membrane-bound FasL in murinelung carcinoma cells produced anti-tumor effects through258

Gene Therapy and Molecular Biology Vol 7, page 259an apoptotic mechanism by Fas/FasL interaction.Adenoviral infection with the Fas-associated death domainprotein gene in lung cancer cell lines resulted in activationof caspase-8 and dose-dependent apoptosis (Kim et al,2003). Shin et al, (2002) noted that the inactivatingmutations of the genes in the pathway of Fas-mediatedapoptosis were associated with nodal metastasis inNSCLC. Using adenoviral vectors to restore wild-type p53function in a human lung cancer cell line, Thiery et al,(2003) reported that this restored not only Fas expressionbut also the Fas-mediated apoptotic pathway, andsuggested that the wt p53-induced optimization of tumorcell killing by specific CTL may involve at least in part aFas-mediated pathway via induction of Fas expression bytumor cells. Wt p53-dependent Fas-mediated apoptosis hasbeen reconfirmed in human cancer cells expressing atemperature-sensitive p53 mutant (Li et al, 2003).C. Growth factor pathway targetsContinuous growth of tumors depends on the alteredregulation of the cell cycle, which is in turn modulated bysignals from growth factors and their receptors, whichprovide the therapeutic targets.Growth factors directly inactivating a criticalcomponent of the cell-intrinsic death machinery may resultin continuous tumor growth. Bad, a heterodimeric partnerfor Bcl-XL and Bcl-2, displaces Bax and promotes celldeath (Yang et al, 1995). It links p53 pathways with AKTand MAPK pathways, as phosphorylation of Bad by AKTor MAPK-activated kinases (Rsks) blocks pro-apoptoticactivity to promote cell survival (Datta et al, 1997; Bonniet al, 1999). An in vitro model of variant differentiation inSCLC, which was chemo- and radio-resistant, elevatedactivation of AKT and MAP kinase associated withincreased levels of phosphorylated BAD and activated NFκB(Kraus et al, 2002). Therapeutic modalities thatovercome the antiapoptotic function of AKT and Rsks areexpected to be a novel strategy for lung cancer treatment.A combination of Bad with Bax resulted in a successfultreatment in experimental tumor models (Zhang et al,2002). IκBα, a specific inhibitor of NF-κB, has also beenshown to be able to increase cytotoxicity in lung cancercells (Batra et al, 1999). In addition, reduction of NF-κBactivation in lung cancer cells was induced by TNF-α(Batra et al, 1999; Jiang et al, 2001). Evidence has beenaccumulated that IκBα is responsible for strong negativefeedback that allows for a fast turn-off of the NF-κBresponse, whereas IκBβ and -ε function to reduce thesystem’s oscillatory potential and stabilize NF-κBresponses during longer stimulations (Hoffmann et al,2002). IκBβ appeared to block the IGF-1 signalingpathway in IκBβ-expressing lung adenocarcinoma cells,and metastatic growth of such cells in the lungs of nudemice was significantly inhibited (Jiang et al, 2001).Besides activating the AKT pathway to blockapoptosis, IGF-IR (the type 1 receptor for insulin-likegrowth factor) activates other two signaling pathways tophosphorylate BAD protein and suppress apoptosis, one ofwhich involves ras-mediated activation of the map kinasepathway. IGF-IR mediates cell survival and growth inresponse to its ligands IGF-I and IGF- II. Blockade ofIGF-I and IGF-IR demonstrated antitumor effects on lungcancer (Hochscheid et al, 2000; Sueoka et al, 2000;Pavelic et al, 2002; Lee et al, 2003). Antisenseoligodeoxynucleotides to IGF-IR and IGF- II wererecruited to suppress the proliferation of lung cancer celllines in vitro, and concomitant treatment inhibited growthup to 80% (Pavelic et al, 2002). Dominant negative IGF-IR has also shown potential for gene-based cancer therapy.Two kinds of defective IGF-IR expressed by adenoviruseseffectively blocked IGF-I-induced Akt kinase activationand significantly suppressed growth in lung cancerxenografts (Lee et al, 2003). Insulin-like growth factorbinding proteins (IGFBPs) are another promisingcandidate (Hochscheid et al, 2000; Sueoka et al, 2000).Ad-IGFBP6 reduced NSCLC cells growth in vitro and invivo in xenografts through activation apoptosis (Sueoka etal, 2000). Damage of downstream target IGF-IR-regulatedgene, such as ras, may be an alternative solution toinducing apoptosis. The antitumor effect has beendemonstrated in human lung tumor xenografts using ananti-K-ras ribozyme adenoviral vector (Zhang et al, 2000).D. New targets and approachesThe list of potential therapeutic genes promises toexpand considerably with the identification of additionalgenes related to human lung cancer.1. SurvivinA high level expression of survivin, a novelapoptosis inhibitor, has been noted in lung and breastcancers (Shen et al, 2003b). RT-PCR assay on tumorsamples from a group of 83 NSCLC patients demonstratedthat the survivin gene was expressed in samples from 71patients who showed poorer overall survival than the other12 patients (Monzo et al, 1999). Down-regulation ofsurvivin by a targeted antisense oligonucleotide (Olie et al,2000) or a TFO (Monzo et al, 1999) induced apoptosis inhuman lung cancer cells. Although further studies arerequired, this gene might provide promising clinicalbenefit in patients overexpressing survivin.2. Cyclooxygenase-2An increase in cyclooxygenase (COX)-2 expression,which is an important biomarker for biologicallyaggressive disease in NSCLC (Khuri et al, 2001;Brabender et al, 2002), may be associated with thedevelopment of human lung cancers and enhanced tumorinvasiveness (Hida et al, 1998). Tumor COX-2-dependentinvasion seems to be mediated by a number of factors(Dohadwala et al, 2001; 2002). Recently, Heuze-Vourc’hrevealed a novel mechanism that, due to the deficiency ofIL-10Rα on the surface of NSCLC cells and theunresponsiveness of COX-2 to IL-10 (known to potentlysuppress COX-2 in normal cells), contributes to themaintenance of elevated COX-2 and its product in the lung259

Cai et al: Lung cancer gene therapytumor environment (Heuze-Vourc’h et al, 2003). Thesefindings suggest the potential efficacy of COX-2 targetedgene therapy, and offer new targets for the furtherdevelopment of prevention and therapy.3. Galectin-3Galectin-3, a member of the β-galactoside-bindinganimal lectins, was recently identified as a key factor intumor metastasis in NSCLC cancer (Yoshimura et al,2003). Galectin-3 has been implicated in tumor invasionand metastasis (Inohara et al, 1998). Compared withhealthy individuals, Galectin-3 serum levels in patientswith lung cancer and some other cancers weresignificantly elevated, especially in patients withmetastatic disease (Iurisci et al, 2000). In vitroexperiments have suggested that Galectin-3 expressionmay play a role in NSCLC cell motility, invasion, andmetastasis (O’Driscoll et al, 2002). A population (10/30)of the NSCLC samples from cell lines and biopsy tissuewere found to overexpress the Galectin-3 protein at levelsthree times higher than those of normal epithelial cells(Yoshimura et al, 2003). Accordingly, Galectin-3 mayrepresent a novel target molecule in NSCLC therapy.Multiple genes are implicated in lung cancerdevelopment and progression to malignancy. Preliminarystudies have proven the tumor suppressor activity of thesenew candidates, such as ganglioside G(D2) (Yoshida et al,2001; Chen et al, 2003), uteroglobin (Lee et al, 2003) andseveral genes in the human chromosome 3p21.3 (Ji et al,2002). However, further investigation is necessary toresolve a number of uncertainties before human trials canbegin.III. Suicide gene therapyA. HSV-tkAlthough the Herpes simplex virus 1 (HSV)thymidine kinase (tk) suicide gene together withganciclovir (GCV) have been successfully used for the invivo treatment of various solid tumors in recent clinicaltrials, a careful assessment and improvement of theefficacy and safety of such a strategy in different tissues inanimal models of human lung cancer is essential beforethey can be used clinically. With the aim of establishing aneffective therapy for pleural metastasis of lung cancer,liposome-mediated transfer of HSK-tk was performed in anude mice model. Direct eradication together with abystander effect contributed to a therapeutic outcome(Nagamachi et al, 1999). Using an orthotopic lung cancermodel employing immunocompetent mice, Fukunaga et al,(2002) have assessed the therapeutic potential ofadenovirus-mediated HSV-tk. Prolonged survival rateswere obtained in mice treated with adenovirally HSV-tktransfectedtumor cells, and were related to genetransduction efficiencies.In order to obtain the specific transduction of HSVtkinto human lung cancer cells, several tumor-specificpromoters have been evaluated. In vitro and ex vivoexperiments have demonstrated the specific expression ofusing gastrin-releasing peptide (GRP) promoter in SBC5human SCLC cell line, in which GRP mRNA expressionwas detected (Inase et al, 2000). However, anotherexperiment on the same cell line showed that neuronspecific enolase (NSE) was not optimal for use in suicidegene transfer to SCLC cells, although NSE mRNA wasexpressed more abundantly in the SBC3 human SCLC cellline than in other cancer cell lines (Tanaka et al, 2001).Myc-Max response element demonstrated potential forspecific expression of HSV-tk in any myc- overexpressingSCLC cells (Kumagai et al, 1996; Nishino et al, 2001). Invivo injections with Ad-MycTK followed by GCVadministration selectively and markedly suppressed thegrowth of myc-overexpressing tumors established in thesubcuties or in the peritoneal cavity of athymic mice; andin contrast to treatment with Ad-CATK, which conferredstrong but nonspecific expression of HSV-tk, no apparentside effects were observed (Nishino et al, 2001). Theseresults emphasis the importance of cell type-specificpromoter selection to target different subpopulations.Carcinoembryonic antigen (CEA) promoter isanother practical choice to reduce toxicity to normal cells,because CEA is found in lung and other cancers (Konishiet al, 1999; Goto et al, 2001). Goto et al, (2001) exploiteda Cre recombinase(Cre)/loxP system consisting of twoadenoviral vectors (one expressing the Cre gene under thecontrol of the CEA promoter (Ad.CEA-Cre), and the otherthe herpes simplex virus thymidine kinase (HSV-TK)gene) to provide a sutilized Cre recombinase(Cre)/loxPsystem to enhance antitumor effects together with minimaladverse reactions in HSV-tk gene therapy againstdisseminated CEA-producing cancer cells in the peritonealcavity of mice. This provided an effective tool againstdisseminated cancer cells without significant side effects.Modification of the HSV-tk gene itself or theprodrug should offer a practical way of improving thistherapeutic system. Delivery of the HSV-TK mutant TK30in a VSV-G pseudotyped retroviral vector, which wasfound to enhance the efficacy of prodrug therapy, provideda therapeutic efficacy after subsequent GCV application inhuman NSCLC cell lines in a preclinical murinexenotransplant model (Kurdow et al, 2002). Recently, twoHSV-tk mutants transferred by adenoviral vector showedmore tumor growth inhibition than the wild-type whentested in several cell lines, including human lung cancerand in their flank tumor models (Wiewrodt et al, 2003).On the other hand, a novel guanosin analog A-5021, whichcan be used more safely than GCV, demonstratedcytotoxic activity as potent as that of GCV in response toretroviral mediated HSV-tk-transduced human lung cancercell lines, but did not exhibit a inhibitory effect on bonemarrow progenitor cells and colony formation (Hasegawaet al, 2000).B. New targets and approaches1. Hypoxanthine-guanine phosphoribosyltransferaseLike HSV-tk, the newly-discovered enzymehypoxanthine-guanine phosphoribosyl transferase(HGPRT), expressed by the parasite Trypanosoma brucei260

Gene Therapy and Molecular Biology Vol 7, page 261(Tb), can serve as a suicide gene, as it convertsallopurinol, a purine analogue, to a cytotoxic metabolite.Retrovirus-mediated TbHGPRT expression can sensitizefive NSCLC cell lines to allopurinol to levels 2.1 to 7.6higher than control values, and represents a practicalapproach in lung cancer therapy (Trudeau et al, 2001).2. Thyroperoxidase-mediated retention ofradioiodideIn much the same way as such gene-prodrugtreatment strategy, the sodium iodide symporter (NIS)gene, that allows rapid internalization of iodide into cells,can be used to obtain radionuclide accumulation byradioactive iodide administration for tumor cell killing. Acombination of the NIS gene and the thyroperoxidase(TPO) gene, which can catalyze iodination of protein,resulted in an augmentation of radioiodide uptake andretention and subsequent effective tumor cell death intransfected NSCLC cell lines (Huang et al, 2001).Although there have so far been few reports on thetreatment of lung cancer with NIS gene, it promises to bean effective approach for cancer treatment.IV. AntiangiogenesisTargeting angiogenesis is an attractive strategy totreat cancer. As progressive growth and metastasis of solidtumors is dependent on the formation of new blood vessels(Folkman, 1971), antiangiogenic therapy is a broadspectrum treatment for cancer. Two strategies used in thedevelopment of antiangiogenic agents involve therapywith endogenous inhibitors of angiogenesis as well as theinhibition of proangiogenic factors.A. Endogenous inhibitors of angiogenesis1. EndostatinEndostatin, a 20-kDa C-terminal fragment ofcollagen XVIII (O’Reilly et al, 1997), is the leadingmember of a class of physiologic inhibitors ofangiogenesis with potent antitumor activity. Boehm et al,(1997) have also reported that when three different mousetumors were subjected to chronic, intermittent therapywith endostatin, there were no traces of acquiredresistance. To establish a constant therapeuticconcentration of circulating endostatin, investigations intoendogenetic expression by a gene therapy approach havebeen prompted. Many viral vectors are actively understudy in endostatin delivery. After systemic administrationof a recombinant adenovirus to nude mice, persistent highserum levels of murine endostatin were achieved. Theendostatin vector treatment not only resulted in significantreduction of the growth rates and volumes of Lewis lungcarcinoma, but also completely prevented the formation ofpulmonary micrometastases (Sauter et al, 2000).Intramuscular injection of adeno-associated viral vectorexpressing human endostatin led to a sufficient level ofserum endostatin to inhibit angiogenesis and tumor growth(Shi et al, 2002). High-level endostatin was also detectedin the vasculature of mice in which hematopoietic stemcells were implanted after being transduced by retrovirusencoding a secretable form of endostatin (Pawliuk et al,2002). In addition, Lentiviral vector (Shichinohe et al,2001) and Semliki Forest viral vector (Yamanaka et al,2001) have been developed to express endostatin, andwere first evaluated in T24 human bladder cancer cells andmice bearing B 16 brain tumor respectively. Some othernonviral transgene delivery approaches also involveendostatin transfer. Utilizing cationic vector, Nakashima etal, (2003) found that intravenous endostatin gene deliverysignificantly inhibited murine lung metastases.Intramuscular injection of polymerized endostatin plasmidinhibited syngeneic tumor growth and lung metastases inmice (Blezinger et al, 1999), and was also shown to inhibitmurine metastatic brain tumor growth (Oga et al, 2003).When electroporation was used to enhance endostatin genetransfer into muscle tissues, the electrotransfer resulted inreduced numbers of experimental melanoma metastases inthe lungs, while intratumoral electrotransfer significantlyinhibited tumor growth (Cichon et al, 2002). Recently,engineered Bifidobacterium, a type of nonpathogenicanaerobic bacterial vector, was applied to bear endostatinby Li X et al, (2003), who demonstrated that vectorscentered in tumors only, and inhibited local tumor growthafter delivery by tail vein injection.2. AngiostatinAngiostatin is another specific endogenous inhibitorof endothelial cell proliferation. It is an internal fragmentof plasminogen, isolated from the urine of mice bearingLewis lung carcinoma (LLC) (O’Reilly et al, 1994).Tanaka et al, (1998) have demonstrated that retroviral andadenoviral vectors transducing angiostatin cDNA can beused to inhibit endothelial cell growth in vitro andangiogenesis in vivo. In a pulmonary metastatic breastcancer model, the delivery of Ad-angiostatin (1x10 9 pfu)to the lung significantly delayed tumor growth, asmeasured by the number of visible surface tumor nodules(Gyorffy et al, 2001). Intratumoral injection of a high-titerAAV-angiostatin vector effectively suppressed tumors andresulted in long-term survival in 40% of a group of treatedrats, whereas the control AAV-GFP vector had notherapeutic benefits (Ma et al, 2002a). As angiostatin is anendogenous internal fragment of plasminogen, effectivesystemic gene therapy could be obtained by angiostatingene transfer. Studies on liposome-coated plasmidcarrying murine and human angiostatin showed that repeatintraperitoneal vector injection resulted in tumor growthsuppression and delay in the onset of tumor growth to thesame degree as intratumoral injection in a nude micemelanoma xenograft model (Rodolfo et al, 2001). Genetransfer of AAV-angiostatin via the portal vein led tosignificant suppression of the growth of both nodular andmetastatic EL-4 lymphoma tumors established in the liver,and prolonged the survival time of the mice (Xu et al,2003). Similar long-term therapeutic effects have alsobeen demonstrated by Ma et al, (2002b), who used a singlei.m. injection of AAV-angiostatin to effectively suppresshuman glioma growth in the brain of nude mice. Thegeneration of angiostatin from endogenous plasminogen261

Cai et al: Lung cancer gene therapyby delivery of protease gene, such as mouse macrophagemetalloelastase (Gorrin-Rivas et al, 2000; 2001) andporcine pancreatic elastase 1 (Matsuda et al, 2000), havebeen demonstrated as an effective alternative in differentcancers.3. TIMPsTIMP-1, TIMP-2, and TIMP-3 are natural matrixmetalloproteinase (MMP) inhibitors that prevent thedegradation of extracellular matrix proteins (Anand-Apteet al, 1997; Moses et al, 1990; Takigawa et al, 1990). Thein vivo efficiency of TIMP-2 has been evaluated in murinelung cancer LLC, and colon cancer C51 in syngeneic miceas well as in human breast cancer in athymic mice (Li etal, 2001). A single intratumoral injection of Ad-TIMP-2significantly reduced tumor growth rates by 60-80% andtumor-associated angiogenesis index by 25-75%, and wasaccompanied by significantly prolonged survival. Lungmetastasis of LLC tumor was inhibited by >90%.Pulmonary metastasis was significantly reduced in amurine melanoma metastasis model following 4 weeks ofintramuscular injection with plasmid encoding TIMP-1compared to controls treated with the plasmid DNA vectoralone. Further therapeutic effects were realized bycombination treatment with IL-2 (Shi et al, 2002). Genetransfer based on nontoxic cationic cholesterol derivativesindicated potent antitumor efficiency of TIMP-2 andTIMP-3 in HCC xenograft in nude mice (Tran et al, 2003).However, if TIMPs are to be utilized in antiangiogenesistherapy, close consideration should be given to a studysuggesting an angiostatin-producing role for MMP-9(Pozzi et al, 2002).4. Combination strategiesMany of the endogenous inhibitors involved incancer gene therapy succeed merely in slowing tumorgrowth, and need to be used in combination therapy forgreater effectiveness (Shi et al., 2003). A combinationapproach has been attempted with tricistronic retroviralvectors encoding two inhibitors of angiogenesis expressedin a rat glioblastoma cell line: N-terminal fragment of ratprolactin and a secret form of human platelet factor 4(Spf4). The results suggested that, in order to successfulcounteract tumor progression, antiangiogenetic strategyshould be combined with other strategies (Ciafre et al,2002). Another multigene therapy presented dormant anderadicated tumors by inhibition of angiogenesis usingendostatin gene together with cytotoxic HSV-tk genetherapy (Pulkkanen et al, 2002). Adeno-associated virusmediatedgene transfer, when combined with ionizingradiation, enhanced inhibition of tumor growth (Shi et al,2003). When assessing antitumor immune responseagainst the recombinant protein of angiostatin andendostatin, Li et al, (2001, 2001) demonstrated that thehost’s immune response may potentiate the antitumoreffects of antiangiogenic agents. Angiostatin gene therapypreceded by an in situ gene transfer of T-cell costimulatorB7.1 eradicates pre-established tumors and a systemicchallenge of cancer cells (Sun et al, 2001). More than anendogenous inhibitor, IL-12 is stronglyimmunomodulatory. When multigene therapy usingangiostatin plus IL-2 was performed, a synergistictherapeutic effect was noted (Wilczynska et al, 2001).B. Inhibition of proangiogenic factors1. Endothelial cell-specific ligand/receptortyrosine kinase systemsKeeping tumors from proangiogenic stimuli andinterrupting the resultant angiogenesis can be achieved bygene therapy to damage endothelial cell-specificligand/receptor tyrosine kinase systems. One of thesesystems consists of vascular endothelial growth factor(VEGF) and its two receptors flt1 and flk1/KDR, andanother consists of angiopoietin-1 and its receptor tie2.The antisense strategy to inhibit transcription of VEGF(Im et al, 1999) and angiopoietin-1 (Shim et al, 2001)produced controlled tumor growth in vivo by inhibitingtumor angiogenesis.The possibility of blocking VEGF and angiopoietin-1 function by gene delivery to produce a soluble form oftheir receptors has recently attracted attention. Hoshida etal, (2002) have demonstrated that the intratumoraladministration of adenovirus-mediated soluble flt-1(sFLT-1) gene results in a regional tumor suppressioneffect. Using intramuscular injection of adenoviral vectorsexpressing sFLT-1, they demonstrated subcutaneousgrowth inhibition in five out of six human lung carcinomacell lines tested in nude mice (Takayama et al, 2000). Asimilar strategy was used by Mahasreshti et al, (2001),who showed that adenovirus-mediated sFLT-1 genetherapy inhibited s.c. ovarian tumor growth, and i.p.injection increased survival in a murine model of ovariancarcinoma. Mori et al, (2000) demonstrated that repeatedintraperitoneal transduction of a soluble flt-1 gene usingHVJ-cationic liposomes suppressed peritoneal metastasesof some cancers, thereby contributing to a longer survivalperiod.In vivo studies of the soluble form of flk-1 (sFLK-1)showed that the growth of neuroblastoma cells wasinhibited by retroviral mediated expression of sFLK-1(Davidoff et al, 2001) or by inoculation with fibroblastwhich produced retroviral vectors encoding sFLK-1(Davidoff et al, 2000). Tseng et al, (2002) evaluated theantitumor effects of the in vivo administration of anadenovirus vector encoding sFLK1 in 3 murine models ofpancreatic adenocarcinoma. Intravenous injection of AdsFLK1resulted in smaller tumor volumes in subcutaneoustumor models both in immunocompetent and SCID mice.The treatment also contributed to longer survival in themetastatic model. A recent investigation employed anAAV vector to transfer the sFLK1 gene. Intraportalinjection of this vector preceded the intrarenal ororthotopic renal tumor implant, and resulted in growthrestriction of tumors or tumor rejection (Davidoff et al,2002).After generating an adenoviral vector encodingsoluble Tie2 gene, Lin et al, demonstrated that i.v.injection of this vector significantly inhibited the growthof subcutaneous primary tumors, as well as experimentalor spontaneously occurring lung metastases (Lin et al,262

Gene Therapy and Molecular Biology Vol 7, page 2631998). Hangai et al, (2001) produced a high plasma levelof soluble Tie2 in mice by a single intramuscular injectionof adenovirus expressing soluble Tie2. This treatmentinhibited intraocular neovascularization, providing apotential approach to treat metastatic cancer using anangiogenesis inhibitor gene.However, not all attempts to target the angiogenesisof cancer using gene therapy strategy have been effective.Some studies have produced negative results even whencontinuous, high levels of protein were produced (Eistereret al, 2002; Pawliuk et al, 2002). Kuo et al, (2001)generated adenoviral vectors encoding angiostatin,endostatin, and the ligand-binding ectodomans of Flk1,Flt1, and neuropilin, and evaluated them in severaldifferent preexisting murine tumor models by systemicdelivery. Ad-Flk1 and Ad-Flt1 resulted in approximately80% inhibition of preexisting tumor growth in murine andhuman tumors. By contrast, adenoviruses encodingangiostatin, endostatin, or neuropilin were significantlyless effective. Regulier et al, (2001) compared theadenoviral delivery of endostatin, proliferin-related protein(PRP), and interferon-inducible protein 10 (IP10) genes ina murine B16F10 melanoma model in immunocompetentmice. Ad-PRP or Ad-IP10 was significantly more efficientthan Ad-endostatin, leading to complete tumor rejectionand prolonged survival in a high proportion of treatedanimals.2. Endothelial progenitor cells targetsThe modification of bone marrow-derived cells withtherapeutic genes has recently provided long-term targetedangiogenesis inhibition. Davidoff et al, (2001) transducedmurine bone marrow cells with a retroviral vectorencoding sFlk1. Tumor growth in mice challenged 3months after transplantation with tsFlk-1-expressing bonemarrow cells was significantly inhibited. De Palma et al,(2003) showed that when tumors were grown in mice intowhich bone marrow progenitors transduced with lentiviralvectors expressing genes from transcription-regulatoryelements of Tie2/Tek gene were transplanted, these Tie2-expressing mononuclear (TEM) cells had a distinguishablephenotype and were present selectively at angiogenic sites.An HSV-tk & GVC approach targeting TEM cells resultedin substantial inhibition of angiogenesis and slower tumorgrowth without systemic toxicity. This experimentdemonstrated that targeting exogenous genes to tumorangiogenesis could be achieved by transplantation ofgenetically-modified hematopoietic stem cells.V. ImmunotherapyA. DNA vaccine1. Tumor-associated genesCarcinoembryonic antigen (CEA) is a cell surfacetumor marker present in a variety of cancers, includinglung cancer. The antitumor effects of an oral DNA vaccineencoding human CEA were obtained in mice, whenboosted with an antibody-IL2 fusion protein. This vaccinebroke peripheral T-cell tolerance toward CEA expressedby Lewis lung carcinoma stably transduced with CEA,resulting in eradication of subcutaneous tumors in 100%of mice and prevention of experimental pulmonarymetastases in 75% of experimental animals in prophylacticsettings (Niethammer et al, 2001). Song et al, (2000)demonstrated that intramuscular injection of a CEAplasmid without coinjection of IL-12 plasmid could notachieve complete resistance to a tumor challenge inwildtype mice by CEA-positive Lewis lung carcinomacells, while injection of the IL- 12 plasmid alone was notprotective. Luo et al, (2003) improved naked CEA DNAvaccine by absorbing it onto cationic microparticles,which are more immunogenic. Boosting with GM-CSFplasmid increased the vaccine’s efficacy, resulting in acomplete rejection of tumor cells in 50% of mice.Utilizing conventional and transgenic mice, Grosenbach etal, (2001) demonstrated that the use of cytokines anddiversified prime and boost regimens could be combinedwith the use of recombinant pox virus vectors expressingsignal 1, such as B7.1, and multiple costimulatorymolecules to further amplify T-cell responses toward moreeffective vaccine strategies. Three different costimulatorymolecule transgenes (B7-1, ICAM-1, and LFA-3) wereused, and the two unique vectors rV-CEA-TRICOM(recombinant vaccinia vector) and rF-CEA-TRICOM(recombinant fowlpox vector). A similar conclusion wasreached by Aarts et al, (2002), who evaluated a diversifiedvaccination protocol consisting of rV-CEA/TRICOM andrF-CEA/TRICOM on transgenic mice. A Phase I clinicaltrial on colorectal cancer using naked DNA immunizationagainst the CEA showed that the vaccination was toleratedwell. The success of the treatment, which has proved to beeffective in a number of patients treated solely byimmunizations, clearly depends on the stage of thedisease. The treatment is most efficient in patients withminimal disease or no metastases (Mincheff et al, 2001).In patients with metastatic carcinoma, clinical study hasshown that ALVAC-CEA B7.1, a canarypox virusencoding the gene for CEA and for B7.1, is safe andstabilizes the disease for up to 13 months (von Mehren etal, 2001). This approach may be a promising strategy forlung cancer vaccines, as immunofluorescence assayshowed that no cell surface expression of CD80 proteinwas detected at all in 31 human NSCLC cell lines(Wroblewski et al, 2001).MUC1 is a cell surface glycoprotein, expressed inmost epithelial tissues and normal lung tissue, and hasbeen shown to be preserved in most NSCLC cell lines andtumors. However, it is not expressed in normal lymphnodes. Vaccination of mice with naked DNA of MUC1produced long-term tumor growth suppression (Johnen etal, 2001), and also suppression of the development of lungmetastases, in which natural killer cells are the majoreffector cells (Kamata et al, 2002). When a similar vaccinewas given in a tumor-bearing mice model, it wasinsufficient to suppress tumor growth. However, theaddition of activated but nonprimed dendritic cells (DCs)obtained from syngeneic mice markedly suppressed tumorgrowth, and prolonged survival time (Kontani et al, 2002).2. Tumor vasculature targets263

Cai et al: Lung cancer gene therapyThe development of vaccines targeting tumorvasculature is a new strategy for cancer immunotherapy.Recently Niethammer et al, (2002) presented an oral FLK-1 DNA vaccine that targets stable, proliferatingendothelial cells in the tumor vasculature, whicheffectively protected mice from lethal challenges withmelanoma, colon carcinoma and lung carcinoma cells, andreduced the growth of established metastases in atherapeutic setting. Angiogenesis in the tumor vasculaturewas suppressed without impairment of fertility,neuromuscular performance or hematopoiesis, thoughthere was a slight delay in wound healing. Theinvestigation of a cross-reaction between microvessels insolid tumors and xenogeneic endothelial cells has shedlight on DNA vaccine for cancer therapy (Wei et al, 2000).Several xenogeneic molecules identified as involved inthis cross-reaction were explored to treat cancer in avaccine formulation, including chicken homologousmatrix metalloproteinase-2 (Su et al, 2003), ligand-bindingdomain of chicken homologous integrin β3 (Lou et al,2002), Xenopus homologous vascular endothelial growthfactor (Wei et al, 2001), and xenogeneic epidermal growthfactor receptor (Lu et al, 2003). These have alldemonstrated potential for antitumor therapy in vivo.B. Tumor cell-based immune modulation1. Cytokines and co-stimulatory moleculesGene therapy with cytokine and lymphocyte surfacemolecules (B7.1 and CD40 ligand) has been applied inclinical studies of tumors.In a spontaneous metastasis model of LLC-f5 model,particle-mediated IL-12 gene transfer into skin distantfrom the tumor site elicited antimetastatic effectsequivalent to local gene transfer, although its effect onprimary tumors was not as evident (Oshikawa et al, 2001).Interleukin-12-transduced Lewis lung carcinoma(LLC/IL12) cells were found to have diminishedtumorigenicity in syngeneic C57BL/6 mice, depending ontheir level of IL-12 production, and both CD4+, CD8+ Tcells and natural killer (NK) cells were involved. Inaddition, LLC/IL12 apparently had a much strongerantitumor effect against established LLC/wt tumors thanLLC transduced with B7-1 or GM-CSF cDNA (Sumimotoet al, 1998). On the other hand, it has been reported thatcostimulatory molecule B7.1 is required for initial tumorsensitivity to IL-12 gene therapy (Heise et al, 2001). Thisobservation may offer the prospect of developing aneffective multiple cytokine gene therapy. Dietrich et al,(2003) demonstrated antitumoral and antimetastatic effectsof continuous particle-mediated cytokine gene (IL-12, IL2,IFN-γ/B7.1) therapy in an LLC model, but a significantlyenhanced survival and reduced tumor growth wasdependent on the sequence and order. To presentsynergistic activities, hetero-dimeric IL-12 could beexpressed either in a single-chain form, or maintained as aheterodimer in which the p40 subunit is fused to IL-2.Gillies et al, (2002) showed that IL12/IL2 bi-functionalcytokine fusion protein induced extremely high levels ofinterferon-γ, similar to the synergy normally seen with thecombined application of the individual cytokines. Inaddition, these bifunctional molecules have been shown tohave striking anti-tumor activity as either gene therapy oras a fusion protein. A comparison of the antitumor effectsof IFN-α and IL-12 revealed that interferon-α inducestumor-specific immune responses while interleukin-12stimulates non-specific killing (Eguchi et al, 2003).Kusumoto et al, conducted a Phase 1 clinical trial todetermine the safety and antitumor activity of anautologous GM-CSF-secreting (granulocyte-macrophagecolony-stimulating factor) melanoma cell vaccine that wasengineered ex vivo with recombinant replicationincompetentadenovirus harboring a human GM-CSFgene. One of the 9 enrolled patients responded to thevaccination by an apparent reduction in the size of ametastatic tumor in the lung. It was shown that infiltrationof inflammatory cells, such as T cells (CD3+, CD8+),macrophages and dendritic cells (CD83+), were involvedin the activation of antitumor immune response(Kusumoto et al, 2001). Several studies on animal modelsalso demonstrated that autologous tumor cell vaccinesecreting GM-CSF is effective in preventing and treatingestablished and metastatic tumors (Nagai et al, 1998; Leeet al, 2000; Kinoshita et al, 2001; Maini et al, 2003). Itsefficiency could also be enhanced by the cosecretion ofIL-6 (Kinoshita et al, 2001) and IL-2 (Lee et al, 2000).Maini et al, (2003) showed, in a murine renal cellcarcinoma (RCC) model, that lung irradiation plusvaccination with autologous tumor cells producingrecombinant interleukin-2 (IL-2), interferon-γ (IFN-γ) andgranulocyte-macrophage colony-stimulating factor (GM-CSF) reduced the number of lung metastases by over 90%.It appears that NK cells and granulocytes arepredominantly involved in the antitumor action. Mostrecently, a Phase I clinical trial was conducted by Salgia etal, (2003), which demonstrated that vaccination withirradiated autologous tumor cells engineered to secretegranulocyte-macrophage colony-stimulating factoraugmented antitumor immunity in some patients withmetastatic non-small-cell lung carcinoma.CD40 is a member of the tumor necrosis factorreceptor (TNF-R) family of cell surface proteins expressedin B cells, dendritic cells, human thymic epithelial cells,human endothelial cells, and several carcinoma cell lines.Interaction between CD40 and CD40 ligand (CD40L;CD154) is important for cross talking between T cells andB cells, an essential requirement for B-cellimmunoglobulin class switching (Banchereau et al, 1994)Imaizumi et al, (1999) demonstrated that stimulation ofCD40 molecules on the surface of alveolar macrophageswith CD40L-expressing clones of Lewis lung cancer cellsenhanced the production of NO, TNF-α, and IL-12, andalso improved tumoricidal activity under the stimulation ofIFN-γ. Noguchi et al, (2001) showed that murine lungcancer cells (3LLSA) transduced with the CD40L gene(3LLSA-CD40L) were rejected in syngeneic C57BL/6mice, but grew in CD40-deficient mice to the same extentas control tumor cells. Coinoculation of interferon (IFN)-γ-transduced 3LLSA with 3LLSA-CD40L cells enhancedantitumor immunity efficiently in vivo. Tada et al, (2003)have shown that the expression of CD40L in tumors inmurine lung carcinoma (A11) cells could produce264

Gene Therapy and Molecular Biology Vol 7, page 265antitumor effects by facilitating the interaction betweenDCs and tumors, enhancing the maturation of DCs,inducing secretion of cytokines, and consequentlyproducing T-cell-dependent systemic immunity. Thesefindings suggest that CD40L gene therapy approaches forthe treatment of lung cancer should be pursued.2. (1,3) Galactosyl epitopes (·Gal)The role of (1,3) Galactosyl epitopes (·Gal) inexnograft rejection has been closely studied (Sandrin andMcKenzie 1994). Unfer et al, (2003) have demonstratedthat immunity to ·Gal provided protection in mice againstchallenge with genetically modified colon cancer cellsexpressing ·Galactosyl-transferase. These resultsdemonstrate the potential for a cancer gene therapy thatuses the innate immunity to Gal antibody in humans todestroy tumors as xenografts.3. Dendritic cell-based vaccineAntigen presentation by dendritic cells (DC) iscrucial for the induction of primary T cell-mediatedimmune responses in vivo. To further augment a cellularimmune response against tumor antigens, attempts havebeen made to increase antigen presentation capacity bygenetically modifying DCs with cytokine genes or tumorassociatedantigen genes (Sharma et al, 2003; Eppler et al,2002). In two murine lung cancer models adenoviral IL-7-transduced DCs (DC-AdIL-7) were administratedintratumorally. Compared with other intratumor therapiessuch as AdIL-7, DC-AdIL-7 therapy was more effective inachieving systemic antitumor responses and enhancingimmunogenicity, and in induction of splenocyte GM-CSFand IFN-γ, although both treatments resulted in completetumor eradication (Miller et al, 2000). Its potential is nowbeing evaluated in clinical trials. In a metastatic livercancer model, local delivery of DCs transduced with theIL-12 gene was able not only to inhibit colorectal tumorgrowth in vivo, but also to mount systemic antitumorimmune responses, evidenced by enhanced production ofIFN-γ by T lymphocytes isolated from both spleen anddraining lymph nodes (Satoh et al, 2002). Liu et al, (2002)demonstrated that DCs transfected with AdV-CD40L(DC(CD40L)) could stimulate enhanced allogeneic T-cellproliferation and Mut1-specific CD8(+) cytotoxic T-cellresponses in vitro. Vaccination of Mut1 peptide-pulsedAdV-CD40L-transfected DC (CD40L) induced anaugmented antitumor immunity in vivo by completeprotection of mice (8/8) from challenge of both low andhigh doses of Lewis lung carcinoma cells. However, moreinvestigation into the role of DC maturation, as well as itstiming and sequence, is needed before it can be used inclinical applications.VI. ConclusionFor successful gene therapy to lung cancer or othercancers, gene delivery systems play a key role. It is wellrecongized that at current developing stage of cancer genetherapy, gene delivery technology is still a major obstacleto success of the cancer therapy, although majorimprovements in all areas of vector development havebeen achived. Further work on technology issues isnecessary. Much has yet to be learned before safe,efficient, stable, economic, convenient gene deliverysystems with an appropriate regulation system eithertargeting specific tissues or cells to obtain long-term geneexpression or targeting tumor directly is developed.As the molecular biology of lung cancerpathogenesis and progression becomes increasinglyunderstood, and as techniques for gene cloning andidentification improve, a number of possible approaches tolung cancer gene therapy are emerging, which havedemonstrated promise in pre-clinical tests. Only some ofthese approaches have been mentioned here. Clinical trialsindicate that different types of combined modalities mayhave to be tailored to deal with specific sub-populations orindividuals. In other words, an optimal outcome willprobably depend on a combination of several genes orcombination of gene therapy and conventional treatments.The crux is how to best combine these novel approachesso that they produce such an optimal outcome. The diversenature of lung cancer suggests that molecular staging ofindividual cases will provide the best direction forcombined modality treatment. Most importantly, althoughthey are not always a reliable indicator of clinicaloutcome, carefully tested and controlled studies on animalmodels should be conducted to optimize the protocolsbefore clinical trials are made.AcknowledgmentsThis study was supported by grants awarded byHKU Research Committee and the PRC’s Ministry ofScience and Technology to R.A. Xu. We would also liketo thank Dr David Wilmshurst for his manuscriptcomment on this review.ReferencesAarts WM, Schlom J, Hodge JW (2002) Vector-basedvaccine/cytokine combination therapy to enhance inductionof immune responses to a self-antigen and antitumor activityCancer Res 62, 5770-5777Akie K, Dosaka-Akita H, Murakami A, Kawakami Y (2000) Acombination treatment of c-myc antisense DNA with alltrans-retinoicacid inhibits cell proliferation bydownregulating c-myc expression in small cell lung cancer.Antisense Nucleic Acid Drug Dev 10, 243-249Anand-Apte B, Pepper MS, Voest E, Montesano R, Olsen B,Murphy G, Apte SS, Zetter B (1997) Inhibition ofangiogenesis by tissue inhibitor of metalloproteinase-3.Invest Ophthalmol Vis Sci 38, 817–823.Baldi A, Esposito V, De Luca A, Howard CM, Mazzarella G,Baldi F, Caputi M, Giordano A (1996) Differentialexpression of the retinoblastoma gene family memberspRb/p105, p107, and pRb2/p130 in lung cancer. ClinCancer Res 2, 1239-1245Baldi A, Esposito V, De Luca A, Fu Y, Meoli I, Giordano GG,Caputi M, Baldi F, Giordano A (1997) Differentialexpression of Rb2/p130 and p107 in normal human tissuesand in primary lung cancer. Clin Cancer Res 3, 1691-1697Banchereau J, Bazan F, Blanchard D, Briere F, Galizzi JP, vanKooten C, Liu YJ, Rousset F, Saeland S (1994) The CD40antigen and its ligand. Annu Rev Immunol 12, 881-922265

Cai et al: Lung cancer gene therapyBatra RK, Guttridge DC, Brenner DA, Dubinett SM, BaldwinAS, Boucher RC (1999) IκBα gene transfer is cytotoxic tosquamous-cell lung cancer cells and sensitizes them to tumornecrosis factor-α-mediated cell death. Am J Respir CellMol Biol 21, 238-245Blezinger P, Wang J, Gondo M, Quezada A, Mehrens D, FrenchM, Singhal A, Sullivan S, Rolland A, Ralston R, Min W(1999) Systemic inhibition of tumor growth and tumormetastases by intramuscular administration of the endostatingene. Nat Biotechnol 17, 343-348Boehm T, Folkman J, Browder T, O’Reilly MS (1997)Antiangiogenic therapy of experimental cancer does notinduce acquired drug resistance. Nature 390, 404-407.Bonni A, Brunet A, West AE, Datta SR, Takasu MA, GreenbergME (1999) Cell survival promoted by the Ras-MAPKsignaling pathway by transcription-dependent and -independent mechanisms Science 286, 1358-1362Bougeret C, Virone-Oddos A, Adeline E, Lacroix F, Lefranc C,Ferrero L, Huet T (2000) Cancer gene therapy mediated byCTS1, a p53 derivative: advantage over wild-type p53 ingrowth inhibition of human tumors overexpressing MDM2.Cancer Gene Ther 7, 789-798Boulay JL, Perruchoud AP, Reuter J, Bolliger C, Herrmann R,Rochlitz C (2000) P21 gene expression as an indicator for theactivity of adenovirus-p53 gene therapy in non-small celllung cancer patients. Cancer Gene Ther 7, 1215-1219Brabender J, Park J, Metzger R, Schneider PM, Lord RV,Holscher AH, Danenberg KD, Danenberg PV (2002)Prognostic significance of cyclooxygenase 2 mRNAexpression in non-small cell lung cancer Ann Surg 235, 440-443Buck AC, Shen C, Schirrmeister H, Schmid-Kotsas A, MunzertG, Guhlmann A, Mehrke G, Klug N, Gross HJ, Bachem M,Reske SN (2002) Liposomal delivery of antisenseoligonucleotides for efficient downregulation of Bcl-2 andinduction of apoptosis. Cancer Biother Radiopharm 17,281-289Bunz F, Dutriaux A, Lengauer C, Waldman T, Zhou S, BrownJP, Sedivy JM, Kinzler KW, Vogelstein B (1998)Requirement for p53 and p21 to sustain G2 arrest after DNAdamage. Science 282, 1497-1501Caputi M, Groeger AM, Esposito V, De Luca A, Masciullo V,Mancini A, Baldi F, Wolner E, Giordano A (2002) Loss ofpRb2/p130 expression is associated with unfavorable clinicaloutcome in lung cancer. Clin Cancer Res 8, 3850-3856Chen HH, Fukumoto S, Furukawa K, Nakao A, Akiyama S,Urano T, Furukawa K (2003) Suppression of lung metastasisof mouse Lewis lung cancer P29 with transfection of theganglioside GM2/GD2 synthase gene. Int J Cancer 103,169-176Choi JH, Ahn KS, Kim J, Hong YS (2000) Enhanced inductionof Bax gene expression in H460 and H1299 cells with thecombined treatment of cisplatin and adenovirus mediated wtp53gene transfer. Exp Mol Med 32, 23-28Ciafre SA, Barillari G, Bongiorno-Borbone L, Wannenes F,Izquierdo M, Farace MG (2002) A tricistronic retroviralvector expressing natural antiangiogenic factors inhibitsangiogenesis in vitro, but is not able to block tumorprogression in vivo. Gene Ther 9, 297-302Cichon T, Jamrozy L, Glogowska J, Missol-Kolka E, Szala S(2002)Electrotransfer of gene encoding endostatin intonormal and neoplastic mouse tissues: inhibition of primarytumor growth and metastatic spread. Cancer Gene Ther 9,771-777.Claudio PP, Howard CM, Pacilio C, Cinti C, Romano G, MinimoC, Maraldi NM, Minna JD, Gelbert L, Leoncini L, Tosi GM,Hicheli P, Caputi M, Giordano GG, Giordano A (2000)Mutations in the retinoblastoma-related gene RB2/p130 inlung tumors and suppression of tumor growth in vivo byretrovirus-mediated gene transfer. Cancer Res 60, 372-382Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y,Greenberg ME (1997) Akt phosphorylation of BAD couplessurvival signals to the cell-intrinsic death machinery. Cell 91,231-241Davidoff AM, Leary MA, Ng CY, Vanin EF (2001) Genetherapy-mediated expression by tumor cells of theangiogenesis inhibitor flk-1 results in inhibition ofneuroblastoma growth in vivo J Pediatr Surg 36, 30-36.Davidoff AM, Leary MA, Ng CY, Vanin EF (2000) Retroviralvector-producer cell mediated angiogenesis inhibitionrestricts neuroblastoma growth in vivo. Med Pediatr Oncol35, 638-640.Davidoff AM, Nathwani AC, Spurbeck WW, Ng CY, Zhou J,Vanin EF (2002) rAAV-mediated long-term liver-generatedexpression of an angiogenesis inhibitor can restrict renaltumor growth in mice. Cancer Res 62, 3077-3083.Davidoff AM, Ng CY, Brown P, Leary MA, Spurbeck WW,Zhou J, Horwitz E, Vanin EF, Nienhuis AW (2001) Bonemarrow-derived cells contribute to tumor neovasculatureand, when modified to express an angiogenesis inhibitor, canrestrict tumor growth in mice. Clin Cancer Res 7, 2870-2879.De Palma M, Venneri MA, Roca C, Naldini L (2003) Targetingexogenous genes to tumor angiogenesis by transplantation ofgenetically modified hematopoietic stem cells Nat Med 9,789-795Dietrich A, Kraus K, Brinckmann U, Stockmar C, Muller A,Liebert UG, Schonfelder M (2003) Antitumoral andantimetastatic effects of continuous particle-mediatedcytokine gene therapy. Recent Results Cancer Res 162,157-168.Dohadwala M, Luo J, Zhu L, Lin Y, Dougherty GJ, Sharma S,Huang M, Pold M, Batra RK, Dubinett SM (2001) Non-smallcell lung cancer cyclooxygenase-2-dependent invasion ismediated by CD44. J Biol Chem 276, 20809-20812Dohadwala M, Batra RK, Luo J, Lin Y, Krysan K, Pold M,Sharma S, Dubinett SM (2002) Autocrine/paracrineprostaglandin E2 production by non-small cell lung cancercells regulates matrix metalloproteinase-2 and CD44 incyclooxygenase-2-dependent invasion. J Biol Chem 277,50828-50833Dubrez L, Coll JL, Hurbin A, de Fraipont F, Lantejoul S, FavrotMC (2001) Cell cycle arrest is sufficient for p53-mediatedtumor regression. Gene Ther 8, 1705-1712.Dumon KR, Ishii H, Fong LY, Zanesi N, Fidanza V, Mancini R,Vecchione A, Baffa R, Trapasso F, During MJ, Huebner K,Croce CM (2001) FHIT gene therapy prevents tumordevelopment in Fhit-deficient mice. Proc Natl Acad Sci U SA 98, 3346-3351.Eguchi J, Hiroishi K, Ishii S, Mitamura K (2003) Interferon-αand interleukin-12 gene therapy of cancer: interferon-αinduces tumor-specific immune responses while interleukin-12 stimulates non-specific killing. Cancer ImmunolImmunother 52, 378-386.Eisterer W, Jiang X, Bachelot T, Pawliuk R, Abramovich C,Leboulch P, Hogge D, Eaves C (2002) Unfulfilled promiseof endostatin in a gene therapy-xenotransplant model ofhuman acute lymphocytic leukemia. Mol Ther 5, 352-359.Eppler E, Horig H, Kaufman HL, Groscurth P, Filgueira L(2002) Carcinoembryonic antigen (CEA) presentation andspecific T cell-priming by human dendritic cells transfectedwith CEA-mRNA. Eur J Cancer 38, 184-193Esposito V, Baldi A, De Luca A, Groger AM, Loda M, GiordanoGG, Caputi M, Baldi F, Pagano M, Giordano A (1997)Prognostic role of the cyclin-dependent kinase inhibitor p27in non-small cell lung cancer. Cancer Res 57, 3381-3385266

Cai et al: Lung cancer gene therapyJohnstone RW, Ruefli AA, Lowe SW (2002) Apoptosis: a linkbetween cancer genetics and chemotherapy. Cell 108, 153-64Joseph B, Ekedahl J, Sirzen F, Lewensohn R, Zhivotovsky B(1999) Differences in expression of pro-caspases in smallcell and non-small cell lung carcinoma. Biochem BiophysRes Commun 262, 381-387Joshi US, Chen YQ, Kalemkerian GP, Adil MR, Kraut M, SarkarFH (1998) Inhibition of tumor cell growth by p21WAF1adenoviral gene transfer in lung cancer Cancer Gene Ther5, 183-191Kagawa S, Gu J, Swisher SG, Ji L, Roth JA, Lai D, Stephens LC,Fang B (2000) Antitumor effect of adenovirus-mediated Baxgene transfer on p53-sensitive and p53-resistant cancer lines.Cancer Res 60, 1157-1161Kaliberov SA, Buchsbaum DJ, Gillespie GY, Curiel DT, ArafatWO, Carpenter M, Stackhouse MA (2002) Adenovirusmediatedtransfer of BAX driven by the vascular endothelialgrowth factor promoter induces apoptosis in lung cancercells. Mol Ther 6, 190-198Kamata M, Denda-Nagai K, Kubota N, Aida S, Takeda K,Irimura T (2002) Vaccination of mice with MUC1 cDNAsuppresses the development of lung metastases. Clin ExpMetastasis 19, 689-696Kawabe S, Munshi A, Zumstein LA, Wilson DR, Roth JA, MeynRE (2001) Adenovirus-mediated wild-type p53 geneexpression radiosensitizes non-small cell lung cancer cellsbut not normal lung fibroblasts Int J Radiat Biol 77, 185-194Kawabe S, Nishikawa T, Munshi A, Roth JA, Chada S, Meyn RE(2002) Adenovirus-mediated mda-7 gene expressionradiosensitizes non-small cell lung cancer cells via TP53-independent mechanisms. Mol Ther 6, 637-644.Kawabe S, Roth JA, Wilson DR, Meyn RE (2000) Adenovirusmediatedp16INK4a gene expression radiosensitizes nonsmallcell lung cancer cells in a p53-dependent mannerOncogene 19, 5359-5366Khuri FR, Wu H, Lee JJ, Kemp BL, Lotan R, Lippman SM, FengL, Hong WK, Xu XC (2001) Cyclooxygenase-2overexpression is a marker of poor prognosis in stage I nonsmallcell lung cancer. Clin Cancer Res 7, 861-867Kim PK, Park SY, Koty PP, Hua Y, Luketich JD, Billiar TR(2003) Fas-associating death domain protein overexpressioninduces apoptosis in lung cancer cells. J Thorac CardiovascSurg 125, 1336-1342.Kinoshita Y, Kono T, Yasumoto R, Kishimoto T, Wang CY,Haas GP, Nishisaka N (2001) Antitumor effect on murinerenal cell carcinoma by autologous tumor vaccinesgenetically modified with granulocyte-macrophage colonystimulatingfactor and interleukin-6 cells. J Immunother 24,205-211Knapp DC, Mata JE, Reddy MT, Devi GR, Iversen PL (2003)Resistance to chemotherapeutic drugs overcome by c-Mycinhibition in a Lewis lung carcinoma murine modelAnticancer Drugs 14, 39-47Konishi F, Maeda H, Yamanishi Y, Hiyama K, Ishioka S,Yamakido M. (1999) Transcriptionally targeted in vivo genetherapy for carcinoembrionic antigen-producingadenocarcinoma. Hiroshima J Med Sci 48, 79-89.Kontani K, Taguchi O, Ozaki Y, Hanaoka J, Tezuka N, Sawai S,Inoue S, Fujino S, Maeda T, Itoh Y, Ogasawara K, Sato H,Ohkubo I, Kudo T (2002) Novel vaccination protocolconsisting of injecting MUC1 DNA and nonprimed dendriticcells at the same region greatly enhanced MUC1-specificantitumor immunity in a murine model. Cancer Gene Ther9, 330-337Kraus AC, Ferber I, Bachmann SO, Specht H, Wimmel A, GrossMW, Schlegel J, Suske G, Schuermann M (2002) In vitrochemo- and radio-resistance in small cell lung cancercorrelates with cell adhesion and constitutive activation ofAKT and MAP kinase pathways. Oncogene 21, 8683-8695Kumagai T, Tanio Y, Osaki T, Hosoe S, Tachibana I, Ueno K,Kijima T, Horai T, Kishimoto T (1996) Eradication of Mycoverexpressingsmall cell lung cancer cells transfected withherpes simplex virus thymidine kinase gene containing Myc-Max response elements. Cancer Res 56, 354-358Kuo CJ, Farnebo F, Yu EY, Christofferson R, Swearingen RA,Carter R, von Recum HA, Yuan J, Kamihara J, Flynn E,D’Amato R, Folkman J, Mulligan RC (2001) Comparativeevaluation of the antitumor activity of antiangiogenicproteins delivered by gene transfer Proc Natl Acad Sci U SA 98, 4605-4610.Kurdow R, Boehle AS, Haye S, Boenicke L, Schniewind B,Dohrmann P, Kalthoff H (2002) Ganciclovir prodrug therapyis effective in a murine xenotransplant model of human lungcancer. Ann Thorac Surg 73, 905-910Kusumoto M, Umeda S, Ikubo A, Aoki Y, Tawfik O, Oben R,Williamson S, Jewell W, Suzuki T (2001) Phase 1 clinicaltrial of irradiated autologous melanoma cells adenovirallytransduced with human GM-CSF gene. Cancer ImmunolImmunother 50, 373-381.Lee CT, Park KH, Adachi Y, Seol JY, Yoo CG, Kim YW, HanSK, Shim YS, Coffee K, Dikov MM, Carbone DP (2003)Recombinant adenoviruses expressing dominant negativeinsulin-like growth factor-I receptor demonstrate antitumoreffects on lung cancer. Cancer Gene Ther 10, 57-63Lee JC, Park KH, Han SJ, Yoo CG, Lee CT, Han SK, Shim YS,Kim YW (2003) Inhibitory effect of adenovirus-uteroglobintransduction on the growth of lung cancer cell lines CancerGene Ther 10, 287-293Lee SG, Heo DS, Yoon SJ, Jee YS, Kang JO, Kim K, Kim CD,Sung MW, Kim NK (2000) Effect of GM-CSF and IL-2 coexpressionon the anti-tumor immune response AnticancerRes 20, 2681-2686Li H, Lindenmeyer F, Grenet C, Opolon P, Menashi S, Soria C,Yeh P, Perricaudet M, Lu H (2001) AdTIMP-2 inhibitstumor growth, angiogenesis, and metastasis, and prolongssurvival in mice. Hum Gene Ther 12, 515-526Li M, Huang X, Zhu Z, Wong M, Watkins S, Zhao Q,Herberman R, Gorelik E (2001) Immune response against3LL Lewis lung carcinoma potentiates the therapeuticefficacy of endostatin. J Immunother 24, 472-481.Li M, Huang X, Zhu Z, Zhao Q, Wong M, Gorelik E (2002) Thetherapeutic efficacy of angiostatin against weakly- andhighly-immunogenic 3LL tumors. In Vivo 16, 577-582Li X, Fu GF, Liu WH, Liu XJ, Wang JJ, Xu GX (2003)Bifidobaxterium adolescentis as a delivery system ofendostatin for cancer gene therapy: selective inhibitor ofamgiogenesis and hypoxic tumor growth. Cancer GeneTher 10, 105-111Li Y, Raffo AJ, Drew L, Mao Y, Tran A, Petrylak DP, Fine RL(2003) Fas-mediated apoptosis is dependent on wild-typep53 status in human cancer cells expressing a temperaturesensitivep53 mutant alanine-143. Cancer Res 63, 1527-1533.Lin P, Buxton JA, Acheson A, Radziejewski C, MaisonpierrePC, Yancopoulos GD, Channon KM, Hale LP, DewhirstMW, George SE, Peters KG (1998) Antiangiogenic genetherapy targeting the endothelium-specific receptor tyrosinekinase Tie2. Proc Natl Acad Sci U S A 95, 8829-8834Liu Y, Zhang X, Zhang W, Chen Z, Chan T, Ali K, Jia Z, XiangJ (2002) Adenovirus-mediated CD40 ligand gene-engineereddendritic cells elicit enhanced CD8 (+) cytotoxic T-cellactivation and antitumor immunity. Cancer Gene Ther 9,202-208Lou YY, Wei YQ, Yang L, Zhao X, Tian L, Lu Y, Wen YJ, LiuF, Huang MJ, Kang B, Xiao F, Su JM, He QM, Xie XJ, Mao268

Gene Therapy and Molecular Biology Vol 7, page 269YQ, Lei S, Liu JY, Lou F, Zhou LQ, Peng F, Jiang Y, Hu B(2002) Immunogene therapy of tumors with a vaccine basedon the ligand-binding domain of chicken homologousintegrin β3. Immunol Invest 31, 51-69Lowe SW, Ruley HE, Jacks T, Housman DE (1993) p53-dependent apoptosis modulates the cytotoxicity of anticanceragents. Cell 74, 957-967Lowe SW, Sherr CJ (2003) Tumor suppression by Ink4a-Arf:progress and puzzles. Curr Opin Genet Dev 13, 77-83Lu Y, Wei YQ, Tian L, Zhao X, Yang L, Hu B, Kan B, Wen YJ,Liu F, Deng HX, Li J, Mao YQ, Lei S, Huang MJ, Peng F,Jiang Y, Zhou H, Zhou LQ, Luo F (2003) Immunogenetherapy of tumors with vaccine based on xenogeneicepidermal growth factor receptor. J Immunol 170, 3162-3170Luo Y, O’Hagan D, Zhou H, Singh M, Ulmer J, Reisfeld RA,James Primus F, Xiang R (2003) Plasmid DNA encodinghuman carcinoembryonic antigen (CEA) adsorbed ontocationic microparticles induces protective immunity againstcolon cancer in CEA-transgenic mice. Vaccine 21, 1938-1947Ma HI, Lin SZ, Chiang YH, Li J, Chen SL, Tsao YP, Xiao X(2002a) Intratumoral gene therapy of malignant brain tumorin a rat model with angiostatin delivered by adeno-associatedviral (AAV) vector Gene Ther 9, 2-11.Ma HI, Guo P, Li J, Lin SZ, Chiang YH, Xiao X, Cheng SY(2002b) Suppression of intracranial human glioma growthafter intramuscular administration of an adeno-associatedviral vector expressing angiostatin Cancer Res 62, 756-763Mahasreshti PJ, Navarro JG, Kataram M, Wang MH, Carey D,Siegal GP, Barnes MN, Nettelbeck DM, Alvarez RD,Hemminki A, Curiel DT (2001) Adenovirus-mediatedsoluble FLT-1 gene therapy for ovarian carcinoma ClinCancer Res 7, 2057-2066Maini A, Nishisaka N, Kinoshita Y, Jones RF, Wang CY, HaasGP (2003) Combination of radiation and vaccination withautologous tumor cells expressing IL-2, IFN-γ and GM-CSFfor treatment of murine renal carcinoma In Vivo 17, 119-123Masuda A, Osada H, Yatabe Y, Kozaki K, Tatematsu Y,Takahashi T, Hida T, Takahashi T, Takahashi T (2001)Protective function of p27(KIP1) against apoptosis in smallcell lung cancer cells in unfavorable microenvironments. AmJ Pathol 158, 87-96Matsuda KM, Madoiwa S, Hasumi Y, Kanazawa T, Saga Y,Kume A, Mano H, Ozawa K, Matsuda M (2000) A novelstrategy for the tumor angiogenesis-targeted gene therapy:generation of angiostatin from endogenous plasminogen byprotease gene transfer. Cancer Gene Ther 7, 589-596Mhashilkar AM, Schrock RD, Hindi M, Liao J, Sieger K,Kourouma F, Zou-Yang XH, Onishi E, Takh O, Vedvick TS,Fanger G, Stewart L, Watson GJ, Snary D, Fisher PB, SaekiT, Roth JA, Ramesh R, Chada S (2001) Melanomadifferentiation associated gene-7 (mda-7): a novel anti-tumorgene for cancer gene therapy. Mol Med 7, 271-82Miller PW, Sharma S, Stolina M, Butterfield LH, Luo J, Lin Y,Dohadwala M, Batra RK, Wu L, Economou JS, Dubinett SM(2000) Intratumoral administration of adenoviral interleukin7 gene-modified dendritic cells augments specific antitumorimmunity and achieves tumor eradication. Hum Gene Ther11, 53-65Mincheff M, Altankova I, Zoubak S, Tchakarov S, Botev C,Petrov S, Krusteva E, Kurteva G, Kurtev P, Dimitrov V,Ilieva M, Georgiev G, Lissitchkov T, Chernozemski I,Meryman HT (2001) In vivo transfection and/or crossprimingof dendritic cells following DNA and adenoviralimmunizations for immunotherapy of cancer--changes inperipheral mononuclear subsets and intracellular IL-4 andIFN-γ lymphokine profile. Crit Rev Oncol Hematol 39,125-132Mitsudomi T, Hamajima N, Ogawa M, Takahashi T (2000)Prognostic significance of p53 alterations in patients withnon-small cell lung cancer: a meta-analysis. Clin CancerRes 6, 4055-4063Miyashita T, Krajewski S, Krajewska M, Wang HG, Lin HK,Liebermann DA, Hoffman B, Reed JC (1994) Tumorsuppressor p53 is a regulator of bcl-2 and bax geneexpression in vitro and in vivo. Oncogene 9, 1799-1805Monzo M, Rosell R, Felip E, Astudillo J, Sanchez JJ, Maestre J,Martin C, Font A, Barnadas A, Abad A (1999) A novel antiapoptosisgene: Re-expression of survivin messenger RNAas a prognosis marker in non-small-cell lung cancers. J ClinOncol 17, 2100-2104Monzo M, Rosell R, Sanchez JJ, Lee JS, O’Brate A, Gonzalez-Larriba JL, Alberola V, Lorenzo JC, Nunez L, Ro JY, MartinC (1999) Paclitaxel resistance in non-small-cell lung cancerassociated with β-tubulin gene mutations. J Clin Oncol 17,1786-1793.Mori A, Arii S, Furutani M, Mizumoto M, Uchida S, FuruyamaH, Kondo Y, Gorrin-Rivas MJ, Furumoto K, Kaneda Y,Imamura M (2000) Soluble Flt-1 gene therapy for peritonealmetastases using HVJ-cationic liposomes. Gene Ther 7,1027-1033.Moses MA, Sudhalter J, Langer R (1990) Identification of aninhibitor of neovascularization from cartilage. Science 248,1408–1410.Nagai E, Ogawa T, Kielian T, Ikubo A, Suzuki T (1998)Irradiated tumor cells adenovirally engineered to secretegranulocyte/macrophage-colony-stimulating factor establishantitumor immunity and eliminate pre-existing tumors insyngeneic mice. Cancer Immunol Immunother 47, 72-80Nagamachi Y, Tani M, Shimizu K, Yoshida T, Yokota J (1999)Suicidal gene therapy for pleural metastasis of lung cancerby liposome-mediated transfer of herpes simplex virusthymidine kinase gene. Cancer Gene Ther 6, 546-553.Nakashima Y, Yano M, Kobayashi Y, Moriyama S, Sasaki H,Toyama T, Yamashita H, Fukai I, Iwase H, Yamakawa Y,Fujii Y (2003) Endostatin gene therapy on murine lungmetastases model utilizing cationic vector-mediatedintravenous gene delivery. Gene Ther 10, 123-130.Naruse I, Hoshino H, Dobashi K, Minato K, Saito R, Mori M(2000) Over-expression of p27kip1 induces growth arrestand apoptosis mediated by changes of pRb expression inlung cancer cell lines. Int J Cancer 88, 377-383Niehans GA, Brunner T, Frizelle SP, Liston JC, Salerno CT,Knapp DJ, Green DR, Kratzke RA (1997) Human lungcarcinomas express Fas ligand. Cancer Res 57, 1007-1012.Niethammer AG, Primus FJ, Xiang R, Dolman CS, RuehlmannJM, Ba Y, Gillies SD, Reisfeld RA (2001) An oral DNAvaccine against human carcinoembryonic antigen (CEA)prevents growth and dissemination of Lewis lung carcinomain CEA transgenic mice. Vaccine 20, 421-429Niethammer AG, Xiang R, Becker JC, Wodrich H, Pertl U,Karsten G, Eliceiri BP, Reisfeld RA (2002) A DNA vaccineagainst VEGF receptor 2 prevents effective angiogenesis andinhibits tumor growth. Nat Med 8, 1369-1375.Nishino K, Osaki T, Kumagai T, Kijima T, Tachibana I, Goto H,Arai T, Kimura H, Funakoshi T, Takeda Y, Tanio Y,Hayashi S (2001) Adenovirus-mediated gene therapy specificfor small cell lung cancer cells using a Myc-Max bindingmotif. Int J Cancer 91, 851-856Nishizaki M, Meyn RE, Levy LB, Atkinson EN, White RA, RothJA, Ji L (2001) Synergistic inhibition of human lung cancercell growth by adenovirus-mediated wild-type p53 genetransfer in combination with docetaxel and radiation269

Cai et al: Lung cancer gene therapytherapeutics in vitro and in vivo Clin Cancer Res 7, 2887-2897Noguchi M, Imaizumi K, Kawabe T, Wakayama H, Horio Y,Sekido Y, Hara T, Hashimoto N, Takahashi M, Shimokata K,Hasegawa Y (2001) Induction of antitumor immunity bytransduction of CD40 ligand gene and interferon-γ gene intolung cancer. Cancer Gene Ther 8, 421-429Nyhus JK, Wolford C, Feng L, Barbera-Guillem E (2001) Directin vivo transfection of antisense Fas-ligand reduces tumorgrowth and invasion. Gene Ther 8, 209-214.O’Driscoll L, Linehan R, Liang YH, Joyce H, Oglesby I, ClynesM (2002) Galectin-3 expression alters adhesion, motility andinvasion in a lung cell line (DLKP), in vitro. Anticancer Res22, 3117-3125Oga M, Takenaga K, Sato Y, Nakajima H, Koshikawa N, OsatoK, Sakiyama S (2003) Inhibition of metastatic brain tumorgrowth by intramuscular administration of the endostatingene. Int J Oncol 23, 73-79Olie RA, Simoes-Wust AP, Baumann B, Leech SH, Fabbro D,Stahel RA, Zangemeister-Wittke U (2000) A novel antisenseoligonucleotide targeting survivin expression inducesapoptosis and sensitizes lung cancer cells to chemotherapy.Cancer Res 60, 2805-2809.O’Reilly MS, Holmgren L, Shing Y, Chen C, Rosenthal RA,Moses M, Lane WS, Cao Y, Sage EH, Folkman J (1994)Angiostatin: a novel angiogenesis inhibitor that mediates thesuppression of metastases by a Lewis lung carcinoma. Cell79, 315-328.O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS,Flynn E, Birkhead JR, Olsen BR, Folkman J (1997)Endostatin: an endogenous inhibitor of angiogenesis andtumor growth. Cell 88, 277-285.Osaki S, Nakanishi Y, Takayama K, Pei XH, Ueno H, Hara N(2000) Alteration of drug chemosensitivity caused by theadenovirus-mediated transfer of the wild-type p53 gene inhuman lung cancer cells. Cancer Gene Ther 7, 300-307Oshikawa K, Rakhmilevich AL, Shi F, Sondel PM, Yang N,Mahvi DM (2001) Interleukin 12 gene transfer into skindistant from the tumor site elicits antimetastatic effectsequivalent to local gene transfer. Hum Gene Ther 12, 149-160Park KH, Seol JY, Kim TY, Yoo CG, Kim YW, Han SK, ShimYS, Lee CT (2001) An sdenovirus expressing mutant p27showed more potent antitumor effects than adenovirus-p27wild type Cancer Res 61, 6163-6169Parkin DM, Pisani P, Ferlay J (1999) Global cancer statistics CACancer J Clin 49, 33-64, 1Pataer A, Fang B, Yu R, Kagawa S, Hunt KK, McDonnell TJ,Roth JA, Swisher SG (2000) Adenoviral Bak overexpressionmediates caspase-dependent tumor killing. Cancer Res 60,788-792Pataer A, Vorburger SA, Barber GN, Chada S, Mhashilkar AM,Zou-Yang H, Stewart AL, Balachandran S, Roth JA, HuntKK, Swisher SG (2002) Adenoviral transfer of the melanomadifferentiation-associated gene 7 (mda7) induces apoptosis oflung cancer cells via up-regulation of the double-strandedRNA-dependent protein kinase (PKR). Cancer Res 62,2239-2243Pavelic K, Krizanac S, Cacev T, Hadzija MP, Radosevic S, CrnicI, Levanat S, Kapitanovic S (2001) Aberration of FHIT geneis associated with increased tumor proliferation anddecreased apoptosis-clinical evidence in lung and head andneck carcinomas. Mol Med 7, 442-453Pavelic J, Pavelic L, Karadza J, Krizanac S, Unesic J, Spaventi S,Pavelic K (2002) Insulin-like growth factor family andcombined antisense approach in therapy of lung carcinoma.Mol Med 8, 149-157Pawliuk R, Bachelot T, Zurkiya O, Eriksson A, Cao Y, LeboulchP (2002) Continuous intravascular secretion of endostatin inmice from transduced hematopoietic stem cells. Mol Ther 5,345-351.Pozzi A, LeVine WF, Gardner HA (2002) Low plasma levels ofmatrix metalloproteinase 9 permit increased tumorangiogenesis. Oncogene 21, 272-281Pulkkanen KJ, Laukkanen JM, Fuxe J, Kettunen MI, Rehn M,Kannasto JM, Parkkinen JJ, Kauppinen RA, Pettersson RF,Yla-Herttuala S (2002) The combination of HSV-tk andendostatin gene therapy eradicates orthotopic human renalcell carcinomas in nude mice. Cancer Gene Ther 9, 908-916Ramesh R, Saeki T, Templeton NS, Ji L, Stephens LC, Ito I,Wilson DR, Wu Z, Branch CD, minna JD, Roth JA (2001)Successful treatment of primary and disseminated human lungcancer by systemic delivery of tumor suppressor genes usingan improved liposome vector Mol Ther 337-350Reed JC (1994) Bcl-2 and the regulation of programmed celldeath. J Cell Biol 124, 1-6Regulier E, Paul S, Marigliano M, Kintz J, Poitevin Y, LedouxC, Roecklin D, Cauet G, Calenda V, Homann HE (2001)Adenovirus-mediated delivery of antiangiogenic genes as anantitumor approach Cancer Gene Ther 8, 45-54Rodolfo M, Cato EM, Soldati S, Ceruti R, Asioli M, Scanziani E,Vezzoni P, Parmiani G, Sacco MG (2001) Growth of humanmelanoma xenografts is suppressed by systemic angiostatingene therapy. Cancer Gene Ther 8, 491-496Roth JA, Nguyen D, Lawrence DD, Kemp BL, Carrasco CH,Ferson DZ, Hong WK, Komaki R, Lee JJ, Nesbitt JC, PistersKM, Putnam JB, Schea R, Shin DM, Walsh GL, DolormenteMM, Han CI, Martin FD, Yen N, Xu K, Stephens LC,McDonnell TJ, Mukhopadhyay T, Cai D (1996) Retrovirusmediatedwild-type p53 gene transfer to tumors of patientswith lung cancer. Nat Med 2, 985-991Roz L, Gramegna M, Ishii H, Croce CM, Sozzi G (2002)Restoration of fragile histidine triad (FHIT) expressioninduces apoptosis and suppresses tumorigenicity in lung andcervical cancer cell lines. Proc Natl Acad Sci U S A 99,3615-3620.Saeki T, Mhashilkar A, Chada S, Branch C, Roth JA, Ramesh R(2000) Tumor-suppressive effects by adenovirus-mediatedmda-7 gene transfer in non-small cell lung cancer cell invitro Gene Ther 7, 2051-2057Saeki T, Mhashilkar A, Swanson X, Zou-Yang XH, Sieger K,Kawabe S, Branch CD, Zumstein L, Meyn RE, Roth JA,Chada S, Ramesh R (2002) Inhibition of human lung cancergrowth following adenovirus-mediated mda-7 geneexpression in vivo Oncogene 21, 4558-4566Salgia R, Lynch T, Skarin A, Lucca J, Lynch C, Jung K, HodiFS, Jaklitsch M, Mentzer S, Swanson S, Lukanich J, BuenoR, Wain J, Mathisen D, Wright C, Fidias P, Donahue D, CliftS, Hardy S, Neuberg D, Mulligan R, Webb I, Sugarbaker D,Mihm M, Dranoff G (2003) Vaccination with irradiatedautologous tumor cells engineered to secrete granulocytemacrophagecolony-stimulating factor augments antitumorimmunity in some patients with metastatic non-small-celllung carcinoma. J Clin Oncol 21, 624-630Salgia R, Skarin AT (1998) Molecular abnormalities in lungcancer J Clin Oncol 16, 1207-1217Sandrin MS and McKenzie IF (1994) Gal· (1,3)Gal, the majorxenoantigen(s) recognized in pigs by human naturalantibodies. Immunol Rev 141, 169–190Sard L, Accornero P, Tornielli S, Delia D, Bunone G, CampiglioM, Colombo MP, Gramegna M, Croce CM, Pierotti MA,Sozzi G (1999) The tumor-suppressor gene FHIT is involvedin the regulation of apoptosis and in cell cycle control. ProcNatl Acad Sci U S A 96, 8489-8492270

Gene Therapy and Molecular Biology Vol 7, page 271Satoh Y, Esche C, Gambotto A, Shurin GV, Yurkovetsky ZR,Robbins PD, Watkins SC, Todo S, Herberman RB, LotzeMT, Shurin MR (2002) Local administration of IL-12-transfected dendritic cells induces antitumor immuneresponses to colon adenocarcinoma in the liver in mice. JExp Ther Oncol 2, 337-349Sauter BV, Martinet O, Zhang WJ, Mandeli J, Woo SL (2000)Adenovirus-mediated gene transfer of endostatin in vivoresults in high level of transgene expression and inhibition oftumor growth and metastases Proc Natl Acad Sci U S A 97,4802-4807Sax JK, Fei P, Murphy ME, Bernhard E, Korsmeyer SJ, El-DeiryWS (2002) BID regulation by p53 contributes tochemosensitivity. Nat Cell Biol 4, 842-849Schuler M, Rochlitz C, Horowitz JA, Schlegel J, Perruchoud AP,Kommoss F, Bolliger CT, Kauczor HU, Dalquen P, FritzMA, Swanson S, Herrmann R, Huber C (1998) A Phase Istudy of adenovirus-mediated wild-type p53 gene transfer inpatients with advanced non-small cell lung cancer. HumGene Ther 9, 2075-2082Schmitt CA, Fridman JS, Yang M, Lee S, Baranov E, HoffmanRM, Lowe SW (2002) A senescence program controlled byp53 and p16INK4a contributes to the outcome of cancertherapy. Cell 109, 335-346Seoane J, Le HV, Massague J (2002) Myc suppression of thep21(Cip1) Cdk inhibitor influences the outcome of the p53response to DNA damage. Nature 419, 729-734Sharma S, Yang SC, Batra RK, Dubinett SM (2003) Intratumoraltherapy with cytokine gene-modified dendritic cells inmurine lung cancer models. Methods Mol Med 75, 711-722Shen C, Rattat D, Buck A, Mehrke G, Polat B, Ribbert H,Schirrmeister H, Mahren B, Matuschek C, Reske SN (2003a)Targeting bcl-2 by Triplex-Forming Oligonucleotide-APromising Carrier for Gene-Radiotherapy Cancer BiotherRadiopharm 18, 17-26Shen C, Buck A, Polat B, Schmid-Kotsas A, Matuschek C, GrossHJ, Bachem M, Reske SN (2003b) Triplex-formingoligodeoxynucleotides targeting survivin inhibit proliferationand induce apoptosis of human lung carcinoma cells. CancerGene Ther 10, 403-410.Shi J, Zheng D, Man K, Fan ST, Xu RA (2003) A potentialagent for cancer therapy Curr. Mol. Med. 3, 727-736Shi W, Teschendorf C, Muzyczka N, Siemann DW (2003) Genetherapy delivery of endostatin enhances the treatmentefficacy of radiation. Radiother Oncol 66, 1-9Shi Y, Parhar RS, Zou M, Al-Mohanna FA, Paterson MC (2002)Gene therapy of melanoma pulmonary metastasis byintramuscular injection of plasmid DNA encoding tissueinhibitor of metalloproteinases-1. Cancer Gene Ther 9, 126-132Shichinohe T, Bochner BH, Mizutani K, Nishida M, Hegerich-Gilliam S, Naldini L, Kasahara N (2001) Development oflentiviral vectors for antiangiogenic gene delivery. CancerGene Ther 8, 879-889.Shin MS, Kim HS, Lee SH, Lee JW, Song YH, Kim YS, ParkWS, Kim SY, Lee SN, Park JY, Lee JH, Xiao W, Jo KH,Wang YP, Lee KY, Park YG, Kim SH, Lee JY, Yoo NJ(2002) Alterations of Fas-pathway genes associated withnodal metastasis in non-small cell lung cancer. Oncogene 21,4129-4136Shim WS, Teh M, Mack PO, Ge R (2001) Inhibition ofangiopoietin-1 expression in tumor cells by an antisenseRNA approach inhibited xenograft tumor growth inimmunodeficient mice. Int J Cancer 94, 6-15Song K, Chang Y, Prud’homme GJ (2000) IL-12 plasmidenhancedDNA vaccination against carcinoembryonicantigen (CEA) studied in immune-gene knockout mice. GeneTher 7, 1527-1535Sozzi G, Veronese ML, Negrini M, Baffa R, Cotticelli MG,Inoue H, Tornielli S, Pilotti S, De Gregorio L, Pastorino U,Pierotti MA, Ohta M, Huebner K, Croce CM (1996) TheFHIT gene 3p14.2 is abnormal in lung cancer. Cell 85, 17-26Sozzi G, Pastorino U, Moiraghi L, Tagliabue E, Pezzella F,Ghirelli C, Tornielli S, Sard L, Huebner K, Pierotti MA,Croce CM, Pilotti S (1998) Loss of FHIT function in lungcancer and preinvasive bronchial lesions. Cancer Res 58,5032-5037Srivenugopal KS, Shou J, Mullapudi SR, Lang FF Jr, Rao JS,Ali-Osman F (2001) Enforced expression of wild-type p53curtails the transcription of the O(6)-methylguanine-DNAmethyltransferase gene in human tumor cells and enhancestheir sensitivity to alkylating agents. Clin Cancer Res 7,1398-1409Su JM, Wei YQ, Tian L, Zhao X, Yang L, He QM, Wang Y, LuY, Wu Y, Liu F, Liu JY, Yang JL, Lou YY, Hu B, Niu T,Wen YJ, Xiao F, Deng HX, Li J, Kan B (2003) Activeimmunogene therapy of cancer with vaccine on the basis ofchicken homologous matrix metalloproteinase-2. CancerRes 63, 600-607Sueoka N, Lee HY, Wiehle S, Cristiano RJ, Fang B, Ji L, RothJA, Hong WK, Cohen P, Kurie JM (2000) Insulin-likegrowth factor binding protein-6 activates programmed celldeath in non-small cell lung cancer cells. Oncogene 19,4432-4436Sumimoto H, Tani K, Nakazaki Y, Tanabe T, Hibino H, Wu MS,Izawa K, Hamada H, Asano S (1998) Superiority ofinterleukin-12-transduced murine lung cancer cells to GM-CSF or B7-1 (CD80) transfectants for therapeutic antitumorimmunity in syngeneic immunocompetent mice. CancerGene Ther 5, 29-37.Sun X, Kanwar JR, Leung E, Lehnert K, Wang D, KrissansenGW (2001) Angiostatin enhances B7.1-mediated cancerimmunotherapy independently of effects on vascularendothelial growth factor expression. Cancer Gene Ther 8,719-727Swisher SG, Roth JA, Nemunaitis J, et al (1999) Adenovirusmediatedp53 gene transfer in advanced non-small-cell lungcancer. J Natl Cancer Inst 91, 763-771Swisher SG, Roth JA, Komaki R, Gu J, Lee JJ, et al. (2003)Induction of p53-regulated genes and tumor regression inlung cancer patients after intratumoral delivery of adenoviralp53 (INGN 201) and radiation therapy. Clin Cancer Res 9,93-101Tada Y, O-Wang J, Seimiya M, Takiguchi Y, Tatsumi K,Kuriyama T, Tagawa M (2002) Antitumor effects areproduced by forced expression of membrane-bound but notsoluble Fas ligand in murine lung carcinoma cells.Anticancer Res 22, 831-836.Tada Y, O-Wang J, Wada A, Takiguchi Y, Tatsumi K, KuriyamaT, Sakiyama S, Tagawa M (2003) Fas ligand-expressingtumors induce tumor-specific protective immunity in theinoculated hosts but vaccination with the apoptotic tumorssuppresses antitumor immunity. Cancer Gene Ther 10, 134-140.Tada Y, O-Wang J, Yu L, Shimozato O, Wang YQ, Takiguchi Y,Tatsumi K, Kuriyama T, Takenaga K, Sakiyama S, TagawaM (2003) T-cell-dependent antitumor effects produced byCD40 ligand expressed on mouse lung carcinoma cells arelinked with the maturation of dendritic cells and secretion ofa variety of cytokines. Cancer Gene Ther 10, 451-456Takayama K, Ueno H, Nakanishi Y, Sakamoto T, Inoue K,Shimizu K, et al (2000) Suppression of tumor angiogenesisand growth by gene transfer of a soluble form of vascularendothelial growth factor receptor into a remote organ.Cancer Res 60, 2169–2177271

Cai et al: Lung cancer gene therapyTakigawa M, Nishida Y, Suzuki F, Kishi J, Yamashita K,Hayakawa T (1990) Induction of angiogenesis in chick yolksacmembrane by polyamines and its inhibition by tissueinhibitors of metalloproteinases (TIMP and TIMP-2).Biochem Biophys Res Commun 171:1264–1271.Tanaka M, Inase N, Miyake S, Yoshizawa Y (2001) Neuronspecific enolase promoter for suicide gene therapy in smallcell lung carcinoma. Anticancer Res 21, 291-294Tanaka T, Cao Y, Folkman J, Fine HA. (1998) Viral vectortargetedantiangiogenic gene therapy utilizing an angiostatincomplementary DNA Cancer Res 58, 3362-3369Tango Y, Fujiwara T, Itoshima T, Takata Y, Katsuda K, Uno F,Ohtani S, Tani T, Roth JA, Tanaka N (2002) Adenovirusmediatedp14ARF gene transfer cooperates with Ad5CMVp53to induce apoptosis in human cancer cells. Hum GeneTher 13, 1373-1382Thiery J, Dorothee G, Haddada H, Echchakir H, Richon C,Stancou R, Vergnon I, Benard J, Mami-Chouaib F, ChouaibS (2003) Potentiation of a tumor cell susceptibility toautologous CTL killing by restoration of wild-type p53function. J Immunol 170, 5919-5926Tran PL, Vigneron JP, Pericat D, Dubois S, Cazals D, Hervy M,DeClerck YA, Degott C, Auclair C (2003) Gene therapy forhepatocellular carcinoma using non-viral vectors composedof bis guanidinium-tren-cholesterol and plasmids encodingthe tissue inhibitors of metalloproteinases TIMP-2 andTIMP-3. Cancer Gene Ther 10, 435-444Trudeau C, Yuan S, Galipeau J, Benlimame N, Alaoui-JamaliMA, Batist G (2001) A novel parasite-derived suicide genefor cancer gene therapy with specificity for lung cancer cells.Hum Gene Ther 12, 1673-1680.Tseng JF, Farnebo FA, Kisker O, Becker CM, Kuo CJ, FolkmanJ, Mulligan RC (2002) Adenovirus-mediated delivery of asoluble form of the VEGF receptor Flk1 delays the growth ofmurine and human pancreatic adenocarcinoma in mice.Surgery 132, 857-865.Unfer RC, Hellrung D, Link CJ Jr (2003) Immunity to the α(1,3)galactosyl epitope provides protection in micechallenged with colon cancer cells expressing α(1,3)galactosyl-transferase: a novel suicide gene for cancergene therapy. Cancer Res 63, 987-993.Usui K, Saijo Y, Narumi K, Koyama S, Maemondo M, KikuchiT, Tazawa R, Hagiwara K, Ishibashi Y, Ohta S, Nukiwa T(2003) N-terminal deletion augments the cell-death-inducingactivity of BAX in adenoviral gene delivery to nonsmall celllung cancers. Oncogene 22, 2655-2663von Mehren M, Arlen P, Gulley J, Rogatko A, Cooper HS,Meropol NJ, Alpaugh RK, Davey M, McLaughlin S, BeardMT, Tsang KY, Schlom J, Weiner LM (2001) The influenceof granulocyte macrophage colony-stimulating factor andprior chemotherapy on the immunological response to avaccine (ALVAC-CEA B7.1) in patients with metastaticcarcinoma. Clin Cancer Res 7, 1181-1191Wei YQ, Huang MJ, Yang L, Zhao X, et al (2001) Immunogenetherapy of tumors with vaccine based on Xenopushomologous vascular endothelial growth factor as a modelantigen. Proc Natl Acad Sci U S A 98, 11545-11550.Wei YQ, Wang QR, Zhao X, Yang L, et al (2000)Immunotherapy of tumors with xenogeneic endothelial cellsas a vaccine Nat Med 6, 1160-1166Wiewrodt R, Amin K, Kiefer M, Jovanovic VP, Kapoor V, ForceS, Chang M, Lanuti M, Black ME, Kaiser LR, Albelda SM(2003) Adenovirus-mediated gene transfer of enhancedHerpes simplex virus thymidine kinase mutants improvesprodrug-mediated tumor cell killing. Cancer Gene Ther 10,353-364Wilczynska U, Kucharska A, Szary J, Szala S (2001) Combineddelivery of an antiangiogenic protein (angiostatin) and animmunomodulatory gene (interleukin-12) in the treatment ofmurine cancer. Acta Biochim Pol 48, 1077-1084Wroblewski JM, Bixby DL, Borowski C, Yannelli JR (2001)Characterization of human non-small cell lung cancer(NSCLC) cell lines for expression of MHC, co-stimulatorymolecules and tumor-associated antigens Lung Cancer 33,181-194Xu RA, Sun X, Tse LY, Li H, Chan PC, Xu S, Xiao W, KungHF, Krissansen GW, Fan ST (2003) Long-term expression ofangiostatin suppresses metastatic liver cancer in mice.Hepatology 37, 1451-1460.Xu RA, Li H., Tse, LY, Kung, H., Lu H, Lam SLDiabetes GeneThreapy: Potential and Challenges (2003) Curr Gene Ther 365-83Yamanaka R, Zullo SA, Ramsey J, Onodera M, Tanaka R,Blaese M, Xanthopoulos KG (2001) Induction of therapeuticantitumor antiangiogenesis by intratumoral injection ofgenetically engineered endostatin-producing Semliki Forestvirus Cancer Gene Ther 8, 796-802Yang E, Zha J, Jockel J, Boise LH, Thompson CB, Korsmeyer SJ(1995) Bad, a heterodimeric partner for Bcl-XL and Bcl-2,displaces Bax and promotes cell death. Cell 80, 285-291Yatabe Y, Masuda A, Koshikawa T, Nakamura S, Kuroishi T,Osada H, Takahashi T, Mitsudomi T, Takahashi T (1998)p27KIP1 in human lung cancers: differential changes insmall cell and non-small cell carcinomas. Cancer Res 58,1042-1047Yin C, Knudson CM, Korsmeyer SJ, Van Dyke T (1997) Baxsuppresses tumorigenesis and stimulates apoptosis in vivo.Nature 385, 637-640Yoshida S, Fukumoto S, Kawaguchi H, Sato S, Ueda R,Furukawa K (2001) Ganglioside G(D2) in small cell lungcancer cell lines: enhancement of cell proliferation andmediation of apoptosis. Cancer Res 61, 4244-4252Yoshimura A, Gemma A, Hosoya Y, Komaki E, Hosomi Y,Okano T, Takenaka K, Matuda K, Seike M, Uematsu K,Hibino S, Shibuya M, Yamada T, Hirohashi S, Kudoh S(2003) Increased expression of the LGALS3 (Galectin 3)gene in human non-small-cell lung cancer. GenesChromosomes Cancer 37, 159-164Zhang Y, Yu J, Unni E, Shao TC, Nan B, Snabboon T, Kasper S,Andriani F, Denner L, Marcelli M (2002) Monogene andPolygene Therapy for the Treatment of ExperimentalProstate Cancers by Use of Apoptotic Genes bax and badDriven by the Prostate-Specific Promoter ARR(2)PB. HumGene Ther 13, 2051-2064Zhang YA, Nemunaitis J, Scanlon KJ, Tong AW (2000) Antitumorigeniceffect of a K-ras ribozyme against human lungcancer cell line heterotransplants in nude mice Gene Ther 7,2041-2045Dr. Ruian Xu272

Gene Therapy and Molecular Biology Vol 7, page 273Gene Ther Mol Biol Vol 7, 273-289, 2003Advances in cationic lipid-mediated gene deliveryReview ArticleBenjamin Martin 1 , Abderrahim Aissaoui 2 , Matthieu Sainlos 1 , Noufissa Oudrhiri 2 ,Michelle Hauchecorne 2 , Jean-Pierre Vigneron 1 , Jean-Marie Lehn 1 and PierreLehn 2*1Laboratoire de Chimie des Interactions Moléculaires, CNRS UPR 285, Collège-de-France, 11 Place Marcelin Berthelot,75005 Paris, France.2INSERM U458, Hôpital Robert Debré, 48 Boulevard Sérurier, 75019 Paris, France__________________________________________________________________________________*Correspondence: Pierre Lehn, INSERM U458, Hôpital Robert Debré, 48 Boulevard Sérurier, 75019 Paris, France. Phone:33(0)140031932, Fax: 33(0)140031903, E-mail: lehn@idf.inserm.frKey words: gene therapy, gene delivery, transfection, cationic lipid, synthetic vector, lipoplexAbbreviations: N-(1-(2,3-dioleyloxy)propyl)-N,N,N-trimethyl ammonium chloride, (DOTMA); dioctadecylamido-glycylspermine,(DOGS); 3‚-[_N-(N’,N’-dimethylaminoethyl)carbamoyl]-cholesterol, (DC-Chol); dioleoyl phosphatidylethanolamine, (DOPE);dimethyldioctadecyl ammonium bromide, (DDAB); 1,2-dioleoyloxy-3-[trimethylammonio]-propane, (DOTAP); N 1 -[2-((1S)-1-[3-aminopropyl)amino]-4-[di(3aminopropyl)amino]butylcarboxamido)ethyl]-3,4-di(oleyloxybenzamide), (MVL5); 3β–[6’-kanamycincarbamoyl]cholesterol,(KanaChol); bis-guanidinium-spermidine-cholesterol, (BGSC); bis-guanidinium-tren-cholesterol, (BGTC); 1,2-dimyristyloxypropyl-3-dimethyl-hydroxyethyl ammonium bromide, (DMRIE); N’-octadecylsperminecarboxamide hydrofluoroacetate,(C 18 Sper 3+ ); O-(2R-1,2-di-O-(1 Z, 9 Z-octadecadienyl)-glycerol)-3-N-(bis-2-aminoethyl)-carbamate, (BCAT); 1,2-dioleoyl-sn-glycero-3-succinyl-2-hydroxyethyl disulfide ornithine, (DOGSDSO); cholesteryl hemidithiodiglycolyl tris(aminoethyl)amine, (CHDTAEA);gamma-interferon-inducible lysosomal thiol reductase, (GILT); small-angle x-ray scattering, (SAXS); dioleoyl phosphatidylcholine,(DOPC); polyethylenimine, (PEI); nuclear pore complexes, (NPCs); glucocorticoid receptors, (GRs); peptide nucleic acid, (PNA);polyethyleneglycol, (PEG)Received: 3 November 2003; Accepted: 1 December 2003; electronically published: December 2003SummaryOver previous years, problems associated with virus-mediated gene delivery have stimulated the synthesis andbiological evaluation of non-viral vectors as a possible alternative for gene therapy applications. Of the various nonviralvectors, cationic lipids have come forward as effective gene delivery agents, although it is clear that theirtransfection efficiency must be increased in order for them to become of real therapeutic value. This can beachieved by overcoming both the intracellular and extracellular barriers they encounter while conveying thetransgene towards the nucleus of the target cells. The purpose of this review is to highlight the advances made todate in facing these challenges by paying particular attention to the design of the cationic lipid itself and thecomplexes (termed lipoplexes) formed on interacting with DNA. Because the structures of all three parts of acationic lipid – the cationic headgroup, the hydrophobic moiety and the connecting linker – are importantdeterminants of transfection efficiency, each will be considered here in turn, with special attention focused on recentstudies including our own work. In addition, the stability of the lipoplex in the extracellular medium and thefeatures of its intracellular trafficking towards the cell nucleus will be assessed from both chemical and biologicalviewpoints. In conclusion, the future will probably see the development of sophisticated modular self-assemblinggene delivery systems incorporating various functional elements to face the various biological barriers encountered.Such vectors can be envisaged as ‘virus-like’ systems which share the levels of gene delivery efficiency of their viralcounterparts, but coupled with the safety of purpose-made organic molecules.I. IntroductionThe use of genes as medicines remains both acaptivating goal and a formidable challenge. By deliberateintroduction of either a functional gene or a sequencecapable of interfering with the functioning of a cellulargene, a wide variety of diseases of inherited and acquiredorigin are open to treatment in a most fundamental sense(Mulligan, 1993; Anderson, 1998). Those working in thefield of gene delivery have much to learn from viruseswhich achieve efficient levels of gene transductioncommensurate both with their need to deliver their geneticmaterial into host cells for the purpose of reproduction and273

Martin et al: Advances in cationic lipid-mediated gene deliverywith the evolutionary time-scale upon which these abilitieshave been honed. Although it seemed therefore natural toharness viruses (among them adenoviruses, adenoassociatedviruses and retroviruses) for therapeutic genedelivery (Mulligan, 1993; Kootstra and Verma, 2003), itcan be contested that much of their inconveniences are yetto be discarded. Problems with immunogenicity andtoxicity remain, added to the practical issues of large scaleproduction and quality control.Focus has therefore shifted to a de novo approach invector design, where synthetic organic molecules are usedto bind the transgene and facilitate its passage across thesignificant extracellular and intracellular barriers thatseparate it from the cell nucleus where expression takesplace via the cellular transcription machinery (Crystal,1995; Lehn et al, 1998). Such carriers are termed non-viralvectors and generally take the form of cationic lipids orcationic polymers. In addition to avoiding problemsassociated with the use of recombinant viruses, anadvantage of using synthetic vectors is that there is nolimit on the size of DNA to be delivered. A large numberand wide variety of synthetic non-viral vectors have beenprepared and their transfection efficiency assessed notonly in in vitro and in vivo experimental studies, butfurther, into the clinical setting for treatment in particularof cancer (Roth and Cristiano, 1997; Hersh and Stopeck,1998) and cystic fibrosis (Alton et al, 1999; Boucher,1999; Griesenbach et al, 1999; Davies et al, 2001). Anexhaustive list of clinical gene therapy trials is available atwww.wiley.co.uk/genmed/clinical.Despite some positive results, the overall outcomeindicates that a critical requirement for successful genetherapy is the use of more efficient gene delivery systems,i.e. systems leading to a higher percentage of transfectedcells or an increased amount of transgene protein in thetransfected cells according to the given experimental orclinical situation (Crystal, 1995; Aissaoui et al, 2002;Miller, 2003). This review aims to highlight the recentadvances in improving cationic lipid-mediated genedelivery in terms of overcoming both intracellular andextracellular barriers to gene transfer. This will be dealtwith by firstly surveying the progress made in vectordesign at the molecular level. Structure and functionalityof the cationic lipid/DNA complexes will then bedescribed with special focus placed on our own work withnovel lipids. Finally, the stability of the lipoplex in theextracellular medium and the features of its intracellulartrafficking towards the cell nucleus will be discussed, aswell as the proposal of creating sophisticated modular selfassemblinggene delivery systems incorporating variousfunctional elements to face the barriers encountered. Inshort, the goal is the development of ‘virus-like’ systems,which share the levels of gene delivery efficiency of viralcounterparts, but coupled with the safety of purpose-madeorganic molecules.II. Basic principlesThe first stage in the preparation of particles suitablefor gene delivery is the condensation of the large DNAmolecules by the vectors. The general structure of acationic lipid vector is shown in Figure 1. The cationicnature of the amphiphilic vector drives an electrostaticinteraction in the presence of negatively charged DNA,spontaneously self-assembling into nanometricvector/DNA complexes termed lipoplexes (stage 1, Figure2). This initial compaction step enables protection of theDNA from nucleases which are found in the extracellularmedium. Use of an excess of cationic vector (quantified bythe lipid/DNA ratio resulting in a mean theoretical chargeratio of the lipoplex (+/-)) conveniently decorates the outersurface of the lipoplex with a net positive charge which isgenerally considered to facilitate subsequent cellularuptake by interaction with negative cell surface residuessuch as proteoglycans (Friend et al, 1996; Labat-Moleur etal, 1996). Non-specific endocytosis ensues, encapsulatingthe lipoplex in intracellular vesicular compartments(Zabner et al, 1995) (stage 2), though fusion-based uptakecannot be entirely ruled out (Gao and Huang, 1995).Internalisation achieved, the DNA must avoid degradationin the late endosome and lysosome (barred arrow) byescaping the endosome to the cytoplasm (stage 3) (Zabneret al, 1995; Mukherjee et al, 1997). Trafficking of theDNA to the perinuclear region precedes passage across thenuclear membrane (stage 4) and subsequent expression ofthe transgene (stage 5). When localised within the nucleus,the DNA is already separated from its vector (Hasegawa etal, 2001) and it has been shown by microinjectionexperiments (Zabner et al, 1995) that gene expression doesnot occur if the complex remains intact.Cationic lipids designed to achieve the ambitioustask of gene delivery were first introduced by Felgner etal, whose work was founded on initial attempts to transfernucleic acids via encapsulation into classical liposomes(Nicolau and Sene, 1982; Nicolau et al, 1983).The lipid DOTMA (N-(1-(2,3-dioleyloxy)propyl)-N,N,N-trimethyl ammonium chloride) resulted, consistingof a quaternary amine connected to two unsaturatedaliphatic hydrocarbon chains via ether groups (Felgner etal, 1987) (Figure 3).Figure 1: Schematic representation of a cationic lipid: lipid moiety, linker and cationic headgroup274

Gene Therapy and Molecular Biology Vol 7, page 2751 Spontaneous self-assembly of vector/DNA complex (lipoplex)2 Endocytosis – encapsulation of lipoplex into endosome3 Endosomal escape before degradation of DNA4 Trafficking of DNA to perinuclear region and nuclear ingress5 Nuclear expression of the transfected DNAFigure 2: Schematic representation of lipoplex formation and trafficking to the target cell nucleus.A multivalent lipid soon followed in the form of thelipopolyamine DOGS (dioctadecylamido-glycylspermine)(Behr et al, 1989) and use of cholesterol as thehydrophobic portion was subsequently validated by thevector DC-Chol (3‚-[_N-(-(N’,N’-dimethylaminoethyl)carbamoyl] cholesterol) (Gao andHuang, 1991) (Figure 3). It is of particular importancethat the transfection efficiencies of many cationic lipidscan be enhanced by their formulation as stable cationicliposomes (Farhood et al, 1995). This is achieved bymixing cationic lipids, especially those which areincapable of forming bilayers alone, with the neutralcolipid DOPE (dioleoyl phosphatidylethanolamine) priorto complexation with DNA. For example, DOTMA/DOPEliposomes are commercially available as Lipofectin(Gibco BRL). Also, the structural analogues DDAB(dimethyldioctadecyl ammonium bromide) (Gibco BRL)and DOTAP (1,2-dioleoyloxy-3-[trimethylammonio]-propane, where ester groups replace ethers of DOTMA)(Boehringer Mannheim) (Figure 4), are commerciallyavailable alone, as well as formulated with DOPE.In addition to stabilisation properties, DOPE is alsothought to have fusogenic properties which are expected toplay a role in endosomal membrane disruption and soenhance escape of the lipoplexes into the cytoplasm(Ellens et al, 1986; Farhood et al, 1995; Vidal andHoekstra, 1995). However, because a selection oflipoplexes formed in the absense of DOPE are also able toescape the endosome (Behr et al, 1989; Vigneron et al,1996), these cationic lipids are also credited with intrinsicmembrane destabilisation properties.Since the initial ‘proof of principle’ period whichconfirmed the ability of cationic lipids to protect, transferand release DNA for cellular expression, a challengingperiod has followed. Indeed slow progress has been madein improving the level of transfection efficiency up to thatrequired for the potential therapeutic use of non-viralvectors, and this is largely attributed to an unclearstructure-activity relationship in vector design. Thus, thedevelopment of novel vectors is justified as the highlycomplex series of steps that connect the DNA outside thecell to its expression in the nucleus are not fullyunderstood and so a novel cationic lipid may not just be a‘me too’ addition to an already extended list, but ratheropen new possibilites for differently influencing thosesteps (Lehn, 1999). As a result, a diverse library of vectorsexists, representing a wide variety in structures and thusnumerous potential mechanisms by which bettertransfection levels might be obtained. Highlights of itscontents will be discussed in the following section.III. Design of the basic domains of acationic lipidAll cationic lipids are positively charged amphiphilescontaining the three following functional domains: i) apolar hydrophilic headgroup which is positively charged,generally via the protonation of one (monovalent lipid) orseveral (multivalent lipid) amino groups; ii) a linker whoselength and nature may influence the stability and thebiodegradability of the vector; and iii) a hydrophobicportion composed of alkyl chains (saturated orunsaturated) or of a steroid (Figure 1).275

Martin et al: Advances in cationic lipid-mediated gene deliveryFigure 3: Structure of the cationic lipids DOTMA, DOGS, DC-Chol and of the neutral colipid DOPE.Progress has been made in the design of each ofthese components. In particular, the choice of headgrouphas expanded into the use of natural architectures andfunctional groups with recognised DNA binding modes.Linkers have been developed which are sensitive tobiological stimuli, inducing DNA release from thelipoplex at defined time-points.Finally, modifications of the hydrophobic portionhave revealed that optimal vector design is oftendependant on this moiety. Accordingly, the followingsections will deal with the advances made to the threefunctional domains.A. Headgroup design1. Monovalent amine headgroups and thehydration issueThe purpose of the headgroup is to sustain a positivecharge for binding of the DNA sequence to be transferred.The charge is often located on ammonium groups(exceptions include phosphonium and arsoniumheadgroups (Guenin et al, 2000)) and a relationshipbetween the degree of hydration of the ammoniumheadgroup and the transfection efficiency has beenelucidated (Bennett et al, 1997). According to thiscorrelation, the greater the imbalance between the crosssectionalarea of the headgroup (small end) andhydrophobic domain (large end) - that is the more coneshapedthe cationic lipid - the more unstable the resultinglipid assembly and so the greater the likeliness to undergofusion with anionic vesicles. Instability of the lipoplex isthought to be related to improved transfection efficienciesbecause a fusion event between the cationic lipoplex andthe endosomal membrane is associated with DNA releaseinto the cytoplasm (Gao and Huang, 1995; Xu and Szoka,1996; Hasegawa et al, 2002). A decrease in headgrouphydration can be achieved by incorporation of ahydroxyalkyl chain which is capable of hydrogen-bondingto neighbouring cationic headgroups and therefore reducesthe space available to associated water and thus the crosssectionalarea of the headgroup. Accordingly, genedelivery by the vectors DOTMA (Figure 3) and the esterlinkedvariant DOTAP (Figure 4) was improved byincorporation of a hydroxyethyl group into their structuresto give vectors DORIE (1,2-dioleoyl-3-dimethylhydroxyethylammonium bromide) and DORI (1,2-dioleoyloxypropyl-3-dimethyl-hydroxyethyl ammoniumchloride) respectively (Figure 4) ((Felgner et al, 1994;Bennett et al, 1997). In complement, the hydration of thehydrophobic domain of both vectors was increased byincorporating alkyl chains that contain cis-unsaturatedbonds (e.g. oleoyl rather than myristoyl) which lead toreduced packing with neighbouring lipids, thereby leavingspace for water molecules and increasing the cross-276

Gene Therapy and Molecular Biology Vol 7, page 277sectional area of this end of the molecule (Bennett et al,1997). Nakanishi and coworkers found that replacement ofthe dimethylamino headgroup of DC-Chol (Figure 3) bydiethylamino and diisopropylamino groups led to reducedgene delivery (Takeuchi et al, 1996). This again is inaccordance with the relationship between headgroup sizeand transfection efficiency as the alkyl chains are likely tocause steric repulsion between neighbouring vectorheadgroups. The same research group went on to showthat lipid I (cholesteryl-3β-carboxyamido ethylene-Nhydroxyethylamine)(Figure 4) was better at gene deliverythan its non-hydroxyethylated dimethyl tertiary aminohomologue and demonstrated by means of FRETexperiments that lipoplexes formulated from lipid I wereparticularly unstable in the presence of anionic liposomes(Hasegawa et al, 2002). It was therefore suggested that theability of lipid I to transfect efficiently was related tolipoplex instability in the endosomes leading to subsequentDNA release.However, other results have not confirmed thefavourable effect of the incorporation of a hydroxyethylgroup on transfection efficiency. An ether-linkedcholesterol conjugate with a dimethyl hydroxyethylheadgroup (cholest-5-en-3β-oxyethane-N,N-dimethyl-N-2-hydroxyethyl ammonium bromide) was found to be lessefficient at gene delivery than its trimethyl nonhydroxyethylatedhomologue (Ghosh et al, 2000).However, it should be taken into account that in anotherstudy, the transfection efficiency of the ester analoguelipid II (cholesteryl-3β-carboxy-ethylene-N,N-dimethyl-N-2-hydroxyethyl ammonium iodide) (Figure 4), wasfound to be dependant on the lipoplex charge ratio (whichwas not assessed for the ether analogue). At charge ratiosup to 5, the hydroxyethylated vector lipid II was lessefficient compared to its dimethyl tertiary amino andtrimethyl quaternary amino homologues, but efficiencywas seen to overtake the methylated homologues at ratiosabove 7 at which the dimethyl and trimethyl homologueswere inefficient (Fichert et al, 2000; Ghosh et al, 2000).Finally, the nature of the counter-ion has beenidentified as a determinant of transfection efficiency. Byvarying the counter-ion of DOTAP (Figure 4) using ionexchangechromatography, it has been shown that thetransfection efficiency in vitro and in vivo varies accordingto the known ability of the counter-ion to either structurewater or shield the cationic charge (Aberle et al, 1996).Indeed, anions such as bisulphate and iodide were found toconvey better transfection efficiencies than acetate andchloride anions. Again, the reduced hydration of theheadgroups thought to occur when using bisulphate oriodide counter-ions is expected to lessen the distancebetween neighbouring headgroups, leading to liposomes orlipoplexes formed from more conical vectors and thereforeprone to undergo the non-bilayer lipid reorganisationsrequired for membrane fusion.2. Novel headgroups with known nucleic acidbinding modesMultivalent cationic lipids are expected to formliposomes with a greater surface charge density thanmonovalent equivalents, and as such are generallyconsidered better than the latter at DNA binding anddelivery to the target cells. A logical step in moving frommonovalent to multivalent species was the incorporationof natural polyamines such as spermidine and spermine,which have the further benefit of a pre-characterisedability to interact with the minor groove of B-DNA(Schmid and Behr, 1991). Incorporation of the triaminespermidine into cholesteryl-spermidine (Moradpour et al,1996) (available commercially as Transfectall (ApollonInc)) (Figure 5), and of the tetraamine spermine into thelipid DOGS (Behr et al, 1989) (Figure 3) are earlyexamples. In addition, in a lipid such as DOGS, it ispossible that the presence of protonation sites withdifferent pKa values may lead to buffering of theendosomal acidification, thereby protecting the DNA fromdegradation and providing a possible endosome escapemechanism (Demeneix and Behr, 1996). More recently,the importance of the length of, and charge distribution on,the polyamine chain have been investigated. Ohwada andcoworkers found that additions of amino groups separatedby methylene portions to the end of a linear polyaminechain did not automatically enhance gene delivery by aseries of polyamine-steroid conjugates, regardless of theextra protonation sites (Fujiwara et al, 2000). Withreference to molecular modelling data, the authorshighlight the importance of the flexibility of the polyaminechain which can adopt increasingly folded conformationson increasing length. It was suggested that the foldedconformations may disfavour interactions with DNA. Withthe aim of designing a polyamine headgroup that had anoptimised interaction specifically with DNA, Blagbroughand coworkers have shown that the central tetramethyleneportion of the polyamine spermine is crucial in conferringhigh transfection activity in a series of cholesterolpolyamineconjugates (Geall et al, 1999). Indeed, thetetramethylene portion of spermine may be able to bridgebetween the complementary strands of DNA, whereas apolyamine with a trimethylene central spacer would onlyinteract with adjacent phosphate groups on the same DNAstrand. The central tetramethylene portion may equally befound in branched polyamino headgroups such as that ofthe multivalent lipid MVL5 (N 1 -[2-((1S)-1-[3-aminopropyl)amino]-4-[di(3aminopropyl)amino]butylcarboxamido)ethyl]-3,4-di[oleyloxybenzamide) (Figure 5) whichcan afford the inclusion of additional protonation sites asthe problems of linear polyamine folding encountered byOhwada and coworkers are conveniently avoided (Byk etal, 1998; Ewert et al, 2002). Thus the resulting lipoplexescan achieve the same charge density with lesser amountsof the cationic lipid in the formulation. Accordingly, theuse of small quantities of multivalent cationic lipids isproposed as a simple solution to lessen the problem ofcationic lipid-associated cytotoxicity.The incorporation of natural moieties in headgroupdesign has recently been extended by ourselves to thefamily of aminoglycoside antibiotics. This group ofnatural compounds is characterised by oligosaccharidesdecorated with up to six amino groups as well asnumerous hydroxyl groups, thus providing a versatilepolycationic framework (Umezawa and Hooper, 1982).277

Martin et al: Advances in cationic lipid-mediated gene deliveryFigure 4: Modification to the headgroup moiety: inclusion of a hydroxyethyl group.Aminoglycosides have the further advantage of a knowninteraction with rRNA sequences (Moazed and Noller,1987; Hendrix et al, 1997). Thus we have recentlyreported acylation of the 6’ amine of kanamycin A,resulting in a series of transfection vectors typified byKanaChol (3β–[6’-kanamycin-carbamoyl]cholesterol)whose structure is shown in Figure 5 (Sainlos et al, 2003).KanaChol may be triply charged at physiological pH andwas found to be highly effective for in vitro genetransfection into a variety of mammalian cell lines,especially when formulated with DOPE (Belmont et al,2002). Gene transfer into the airway epithelium is anactive area of research as it could offer treatment for lungdiseases both inherited, such as cystic fibrosis, andacquired, such as asthma. Investigation into the usefulnessof KanaChol/DOPE liposomes to deliver reporter genesinto the mouse airways, confirmed its significant ability totransfect the respiratory epithelium. The applicability oflipid-derivatives of other members of the aminoglycosidefamily to gene delivery is currently under investigation.Guanidinium groups are also able to bind DNAphosphate anions, the interaction in this case beingcharacterised by a pair of parallel hydrogen bonds whichprovide binding strength by their structural organization.In addition, guanidinium groups are highly basic andevidence also exists for hydrogen-bonding to nucleic bases(in particular guanine). Their DNA binding ability shouldtherefore be relatively insensitive to environmentalmodifications (pH, proximity of other positively chargedgroups). Further, the group is found naturally in arginineamino acid residues which play a key role in DNA-bindingproteins such as histones and protamines. Two bisguanidiniumcholesterol derivatives have been synthesisedand tested within our group: BGSC (bis-guanidiniumspermidine-cholesterol)and BGTC (bis-guanidinium-trencholesterol)(Vigneron et al, 1996). The structure ofBGTC is shown in Figure 5. Both vectors were found tobe highly efficient for gene transfection in vitro into avariety of mammalian cell lines when formulated ascationic liposomes with DOPE. In addition, as BGTC wasfound to be soluble in aqueous medium, it could besuccessfully used for transfection when formulatedwithout DOPE. Further studies confirmed the ability ofBGTC/DOPE liposomes to efficiently transfect airwayepithelial cells in vivo (Oudrhiri et al, 1997). Interestinglyhowever, per-guanylation of the amino groups ofKanaChol caused a significant drop in transfection activity(Belmont et al, 2002).278

Gene Therapy and Molecular Biology Vol 7, page 279Finally, the ability of the four natural cationic aminoacids to function as headgroups has been assessed (Heyeset al, 2002). Tryptophan as headgroup caused aggregationduring formulation and would not form liposomes,whereas the histidine headgroup yielded moderate levelsof transfection, with the lysine and arginine headgroupsleading to highly efficient gene delivery. However, notonly poly-L-lysine (Gao and Huang, 1993; Monsigny et al,1994; Wagner, 1998) but also histidine rich polymers andpeptides have been identified as efficient gene deliveryagents (Pichon et al, 2001; Ihm et al, 2003) and recentstudies with divalent cationic lipids incorporating a singlehistidine residue suggest that their transfection efficiencyis due to histidine-mediated membrane fusion activitywithin the endosome (Kumar et al, 2003). Becausemembrane fusion was only induced at acidic pH, amechanism whereby protonation of the imidazole group ofthe histidine led to an increased interaction between thevector and the phospholipids of the anionic endosomalmembrane was proposed.B. Lipid moiety designIt has been shown that the length and type ofaliphatic chain incorporated into cationic lipids cansignificantly affect transfection efficiency and so vectorsare often prepared in a series differing in theirhydrophobic portion. Results obtained with the vectorDMRIE (Felgner et al, 1994) (1,2-dimyristyloxypropyl-3-dimethyl-hydroxyethyl ammonium bromide, an analogueof DOTMA), glycine betaine derivatives (Floch et al,1998), alkyl acyl carnitine esters (Wang et al, 1998), lacticacid derivatives (Laxmi et al, 2001) and bis-ether lipidsFigure 5: Advances made in headgroup design: multivalent lipids with recognised modes of nucleic acid binding.279

Martin et al: Advances in cationic lipid-mediated gene deliveryrelated to DOTAP (Heyes et al, 2002), have shown that acomparison of vectors based solely on the lengths of thetwo saturated aliphatic chains led to identify the orderC 14 >C 16 >C 18 in transfection efficiency evaluated in termsof transgene expression. As explanation, it was suggestedthat a shorter chain length may facilitate intermembranemixing, an important factor in endosomal escape (Felgneret al, 1994). A common modification is the use of cismonounsaturatedalkyl chains such as the oleoyl group(C 18 ) which appears to lead to higher levels of transfectionthan saturated stearyl derivatives (C 18 ) likely related toissues of hydrophobic moiety hydration (see above)(Felgner et al, 1994; Wang et al, 1998). It should be notedhere that because of the wide variety of vectors that havebeen synthesized, it is difficult to make generalassumptions based solely on a single aspect of the vectordesign. For example, the transfection efficiency of thelipopolyamines prepared by Behr and coworkers was seento be independent of chain saturation (oleoyl vs stearyllipopolyamines) (Remy et al, 1994), and when comparingsaturated chains, Scherman and coworkers found C 18chains to be optimal with decreased efficiency on gradualshortening (Byk et al, 1998).The use of two aliphatic chains of different length,has been assessed and suggests that C 12 /stearyl orC 12 /oleoyl combinations may convey the vector with amore fusogenic character advantageous for endosomalescape (Balasubramaniam et al, 1996; Heyes et al, 2002).Micelles are expected to form when using vectorscontaining either two short alkyl chains of equal length(such as the C 8 chain homologue of the C 12 lipopolyamineDOGS (Figure 3)), or just a single aliphatic chain (such asthe lipopolyamine C 18 Sper 3+ (N’-octadecylsperminecarboxamide hydrofluoroacetate))(Remy et al, 1994). However, such micellar vectors werenot found to transfect mammalian cell lines. Ongoingresearch focuses on the use of single-chained cationicdetergents capable of dimerisation via oxidation (Lleres etal, 2001; Zuber et al, 2003).Finally, because of the rigidity of cholesterol, as wellas its endogenous biodegradability and fusion activity, ithas been often used as an alternative to aliphatic chains,especially when lipoplexes with a high degree of physicalstability are required as for aerosol delivery. Covalentlybound cholesterol was first included as the hydrophobicportion of the vector DC-Chol by the Huang group (Gaoand Huang, 1991) (Figure 3), and then subsequently withour own lipids BGTC and KanaChol and their analogues(see above). Finally, cholesterol can be used as a neutralcolipid alternative to DOPE in the formulation of cationicliposomes.C. Linker designStable linking of the hydrophobic and hydrophilicportions of cationic lipids is commonly achieved usingcarbamate, amide, ester or ether bonds with no particulargroup emerging as consistently optimal in structureactivitystudies across different vector types. A balancedchoice must be made between the stability of the vectorsmediated by the linking bond, and their toxicity whichmay be related to the half-life of the vector in the cell.Although inconsistencies exist, it is generally agreed thatether linked vectors (Ghosh et al, 2000) are particularlystable, but as such are expected to be more toxic than esterlinked lipids which may be more easily cleaved within thecell and so correspondingly less toxic (Gao and Huang,1995). Carbamates in particular are thought to achieve asuitable balance between stability and toxicity, and as suchare often used in vector design (Gao and Huang, 1991;Vigneron et al, 1996; Aissaoui et al, 2002).Of emerging interest however is the use of linkersincorporating functional groups that are cleavable onshorter time scales and under specific stimuli such thatDNA release may be facilitated by a triggereddecomplexation mechanism. Cleavable vectors have thusbeen designed that are sensitive to stimuli such as decreasein pH, change in redox potential and, recently, photosensitivity(Nagasaki et al, 2003). Clearly biologicalstimuli occurring post-internalisation of the lipoplexes areof special interest and therefore only the advances in linkerdesign that incorporate pH and redox sensitivity will bediscussed here. The incorporation of unstable linkers intothe neutral colipid rather than the cationic (vectoring) lipidis a complementary approach that will not be covered,though a comprehensive review has recently beenpublished which covers this topic (Guo and Szoka, 2003).Before detailing the methods by which cationic lipids canbe made to facilitate intracellular release of DNA, itshould be stressed that the use of degradable amphiphilesmay be associated with reduced cellular toxicity whencompared with lipids with a more chemically stable linker.1. pH-sensitive linkersIt is generally agreed that lipoplexes are taken up bycells via an endosomal pathway (stage 2, Figure 2).Evidence suggests that the poor levels of transfectionactivity attained by non-viral vectors are in part due to aninefficiency in escaping the endosome (stage 3, Figure 2)before degradation of the DNA by nucleases in the lateendosomes and lysosomes (barred arrow in Figure 2)(Zabner et al, 1995). Upon internalisation, the pH of theendosome, which is initially that of the extracellular fluids(pH 7.2-7.4), is lowered to approximately 5.0 by the actionof ATP-dependent proton pumps located in the endosomalmembrane (Mukherjee et al, 1997). By incorporating anacid-sensitive functional group into the linker between thehydrophobic and hydrophilic moieties, it may therefore behoped that the pH drop will act as a trigger, cleavinghydrophobic and hydrophilic portions of the lipoplex, andthereby destabilising the lipoplex structure. Thus, if theresulting DNA decomplexation would be concomitantwith endosomal membrane destabilisation (by remainingintact cationic lipids or a colipid such as DOPE), release ofthe DNA into the cytosol should be enhanced andconsequently transfection efficiency might be improved.Boomer et al, have reported the synthesis of thecationic lipid BCAT (O-(2R-1,2-di-O-(1 Z, 9Zoctadecadienyl)-glycerol)-3- N- (bis-2-aminoethyl)carbamate) incorporating acid-sensitive vinyl ether groups(Boomer et al, 2002) (Figure 6). The vector undergoes280

Gene Therapy and Molecular Biology Vol 7, page 281complete hydrolysis in acidic solution and was found toeffect higher levels of transgene expression than a nonhydrolysablecontrol cationic lipid. The hydrolytically lessstable ortho ester bond has also been integrated into thestructure of a cationic lipid by Zhu et al, who reported thesuccessful gene delivery by the hydrolysabletrioxabicyclo[2,2,2]octane containing cationic lipid III ofFigure 6 (Zhu et al, 2000).The acid-sensitive acylhydrazone group has beenmuch used as a linker between antineoplastic drugs(doxorubicin, daunorubicin) and carriers (antibodies,serum albumin, transferrin, polyethylene glycol) with theaim of reducing unwanted drug toxicity (Mueller et al,1990; Kaneko et al, 1991; Kratz et al, 1997). We haverecently introduced an acylhydrazone group between asteroid moiety and a bis-guanidinium headgroup for thepurposes of gene delivery. The vector undergoeshydrolytic cleavage in acidic solution and was found to berelatively stable at physiological pH. Transfection activityhas been confirmed in vitro and the vector showed lowcytotoxicity likely due to the unstable bond. Further, ouracylhydrazone cationic lipid was found to be tolerant toserum and showed significant gene transfection efficiencyinto mouse lungs (Aissaoui et al, submitted forpublication).2. Redox-sensitive linkersAnother class of triggerable vectors are the redoxpotential-sensitive lipids. These vectors work on theprinciple that the lipoplex, once internalised into the cells,is presented with a relatively high concentration ofreductive substances, such as glutathione present inconcentrations of up to 10 mM (Meister and Anderson,1983), and reducing enzymes including thioredoxin andglutaredoxin (Saito et al, 2003). On incorporation ofdisulphide bonds into the vector structure, cleavage of thegroup is expected to coincide with exposure of thelipoplex to the reducing environment of the cytoplasm,destabilising the complex and leading to DNA release inmuch the same way as pH-sensitive systems. Tang andHughes reported the synthesis of the disulphide-containingornithine conjugate DOGSDSO (1,2-dioleoyl-sn-glycero-3-succinyl-2-hydroxyethyl disulfide ornithine), shown inFigure 7, which can be cleaved by dithiothreitol withconcomitant release of plasmid DNA (Tang and Hughes,1998). In vitro testing demonstrated a 6 to 15-fold increasein transfection compared to DOTAP, depending on the cellline used, and up to 50-fold enhanced transfectioncompared to a non-cleavable analogue. Increasedsensitivity of the disulphide linker was achieved usingdithiodiglycolic acid to tether the polar and hydrophobicdomains, such that the less reducing but endogenousglutathione could induce cleavage of CHDTAEA(cholesteryl hemidithiodiglycolyl tris(aminoethyl)amine)(Figure 7) (Tang and Hughes, 1999). Interestingly, theincreased sensitivity to cleavage rendered the lipid noncytotoxic.Redox-sensitive cationic lipids are a developingbranch of triggerable non-viral vector as the mechanism ofdisulphide reduction is as yet not fully understood. Forexample, endosomal cleavage has only recently beenrecognised with the discovery of the reducing enzymeGILT (gamma-interferon-inducible lysosomal thiolreductase) which is the first to be primarily located in theendosomal pathway (Phan et al, 2000).Environment sensitive or ‘triggerable’ cationic lipidsrepresent the first generation of a new approach to genedelivery by non-viral vectors. With the incorporation of a‘triggerable’ function, direct parallels can be drawn withviral vectors which themselves exploit the acidification ofthe endosome and the reducing environment of thecytoplasm (Goff, 2001; Meier and Greber, 2003).Connections are indeed being established between viraland non-viral gene therapy with the inclusion of the virusderivedEALA and GALA fusogenic peptides intomultimodular formulations with cationic lipids.Endosomal membrane destabilisation by these systems hasbeen seen to occur in response to the drop in pH, aconformational change from random coil to α-helix beinginduced in the protein (Parente et al, 1990; Parente et al,1990; Gottschalk et al, 1996; Vogel et al, 1996). However,regardless of whether dealing with viral or non-viralmediated gene transfer, it is clear that the environmentresponsivefunction must be sensitive enough to betriggered at the correct time during the trafficking processwith avoidance of premature or late responses. If morethan one trigger is included, as is certainly the case withviruses, then the sequence of transitions must take placechronologically (Lehn et al, 1998). Thus the developmentof highly ‘sophisticated’ cationic lipid-based gene deliverysystems may be viewed as ‘programmed supramolecularsystems’ obtained via a defined plan, the informationnecessary for the assembly process to take place and theFigure 6: The use of acid-sensitive linkers in cationic lipid design.281

Martin et al: Advances in cationic lipid-mediated gene deliveryFigure 7: The use of redox-sensitive linkers in cationic lipid design.algorithm that it follows being stored in the componentsand operating via selective molecular recognition events(Lehn, 1993).IV. Structural features ofsupramolecular assemblies formed bylipoplexesAs gene transfection by cationic lipids involves thespontaneous formation of discrete lipid/DNA particles, itmay be expected that the transfection efficiency of a givencationic lipid system depends not only on the properties (atthe molecular level) of the cationic lipid itself, but also onthe structural and functional characteristics of the selfassembledsupramolecular assemblies formed by thelipoplexes. Because of the extensive characterisation thatwe have undertaken on the BGTC-based lipoplexesprepared by our group (Pitard et al, 1999), the structuralfeatures of these assemblies will be summarised here andcompared with the results obtained by others when usingother cationic lipid systems.The structural features of BGTC-based lipoplexeswere visualised by cryotransmission electron microscopy(cryo-TEM) which has the advantage of imagingbioassemblies close to their native state. With theadditional perspective given by data from synchrotronsmall-angle x-ray scattering (SAXS), precise structuraldetails could be resolved. The results obtained indicatedthe presence of highly ordered multilamellar domains witha regular spacing of 70 _ and 68 _ in BGTC/DOPE/DNAand BGTC/DNA lipoplexes, respectively (Pitard et al,1999). It can be assumed from this data that the DNA isintercalated between the lipid bilayers. Interestingly, DNAlipoplexes with similar lamellar patterns were alsodetected inside transfected HeLa cells by conventionalelectron microscopy and as such may be the “active”lipoplexes in the process of trafficking. Further, theselamellar lipoplexes were mainly detected in intracellularvesicles, a finding suggesting that endocytosis was themajor route of uptake.Characterisation of the structural features oflipoplexes used for gene transfection has always been anactive area of research. A hypothetical “bead on thestring” model in which unmodified cationic liposomeswere distinctly attached to the DNA was originallyproposed (Felgner and Ringold, 1989). Over the years,various electron microscopy techniques have then beenused to visualise the structures of the lipoplexes formed bydifferent cationic lipids. These studies essentiallysuggested that the DNA was entrapped in condensedstructures formed by interrelated lipid fusion and DNAcollapse for which thermodynamic models have beenproposed in order to facilitate analysis and understandingof the process of particle formation (Gershon et al, 1993;Ahearn and Malone, 1999). These condensed structureswere found to exhibit various ordered patterns ofsupramolecular organisation, including multilamellarstructures and direct or inverted hexagonal packing(Gustafsson et al, 1995; Labat-Moleur et al, 1996; Lasic etal, 1997; Lasic et al, 1998). Of particular note is the studyof DC-Chol/DOPE/DNA lipoplexes where in addition toconcentric lamellar structures, tubular “spaghetti-like”structures consisting of DNA rods coated by a single lipidbilayer were observed (Sternberg et al, 1994). As concernsx-ray diffraction studies, lamellar domains with aperiodicity similar to that found in BGTC lipoplexes wereobserved with DOTAP/DOPC (dioleoylphosphatidylcholine, a structural analogue of DOPE),DDAB and lipopolyamines (Lasic et al, 1997; Pitard et al,1997; Radler et al, 1997; Safinya and Koltover, 1999).However, a study has shown that DOTAP/DOPE/DNA282

Gene Therapy and Molecular Biology Vol 7, page 283lipoplexes have an inverted hexagonal organization whichwas attributed to the presence of DOPE (as a multilamellarordering was found when DOPC was used instead ofDOPE) and it was further suggested that such an invertedhexagonal arrangement favoured endosomal escape of thelipoplexes via its bilayer-destabilising activity (Koltover etal, 1998).In conclusion, it should be emphasised here thatelucidation of the structural and functional features of the“active” lipoplexes may require the determination not onlyof their initial features but also of the structural changesthat may be induced in the lipoplex by the environmentsthat are encountered during trafficking to the nucleus. Forexample, it has been demonstrated that an interaction withthe microtubules is involved in the cellular trafficking ofthe lipoplex. Fluorescently tagged lipid I/DOPE lipoplexes(Figure 4) were found to be co-localised withmicrotubules using laser scanning confocal microscopy incells expressing fluorescently tagged tubulin (Hasegawa etal, 2001). Equally, live-cell real-time multiple particletracking has been used to observe the trafficking of thecationic polymer polyethylenimine (PEI) (Boussif et al,1995) through the cell, presumably by the same process(Suh et al, 2003). These results provide a striking parallelwith virus-mediated uptake which is known to be‘directed’ through the cytoplasm (Tang et al, 1999; Goff,2001; Seisenberger et al, 2001) rather than relying onrandom diffusion, which was expected to be the onlymode of transport available to non-viral vector/DNAcomplexes.V. Inclusion of ligands for nuclearuptakeAs vectors have become increasingly sophisticated,it has become clearer that significant advances can only bemade if cationic lipid design also addresses the mostconsiderable barrier to gene delivery: the passage acrossthe nuclear membrane. Indeed, experiments comparingtransgene expression after microinjection of plasmid DNAeither into the cytoplasm or directly into the nucleus haveidentified the nuclear membrane as being a severelylimiting step in transgene delivery (Capecchi, 1980;Zabner et al, 1995; Escriou et al, 1998). Accordingly, thesuccess of the current non-viral vectors may at least partlyrely on the disappearance of the nuclear membrane duringmitosis, as transfection is generally found to depend on theproliferation status of the cells (Fasbender et al, 1997;Oudrhiri et al, 1997; Mortimer et al, 1999; Escriou et al,2001). As such it is suggested that gene delivery duringinterphase is likely to be low. When adding to this therelatively instability of plasmid DNA in the cytoplasm(probably due to degradation by Ca 2+ -sensitive nucleases(Lechardeur et al, 1999; Pollard et al, 2001), the need foran efficient method of crossing the nuclear membranebecomes evident.The nuclear membrane consists of two concentriclipid bilayers which make contact at numerous points,forming aqueous pores through the two membranes, whichare termed the nuclear pore complexes (NPCs) and havean internal diameter of roughly 9 nm (Mazzanti et al,2001). These are the only known route for direct exchangeof substances between the cytoplasm and the intranuclearcompartment. Consequently, the transport of largemolecules such as plasmid DNA is non-passive andrequires signal mediation by shuttling molecules(importins) (Boulikas, 1997; Escriou et al, 2003). Thebasic peptide from the SV40 large tumor antigencharacterised by a PKKKRL amino acid sequence acts as anuclear localisation signal (NLS) and has been used inattempts to induce nuclear uptake of plasmid DNA.Binding of multiple NLSs to plasmid DNA (Collas et al,1996; Ciolina et al, 1999; Neves et al, 1999) was found tobe bettered by attachment of just a single NLS peptide toone end of a capped linear DNA fragment (Zanta et al,1999). However, although transfection (mediated by PEIor DOGS) was enhanced by an impressive 10-1000 foldwith the single NLS-DNA conjugate, the techniqueremains only a proof of principle as preparation of such aDNA fragment requires complex engineering and istherefore relatively impractical for gene therapyapplications.An alternative approach to the use of NLSs fornuclear uptake of plasmids is steroid-mediated genedelivery. This technique relies on binding of thetransfected plasmid to the glucocorticoid receptors (GRs),which thus stimulated, actively transport the boundplasmid into the nucleus. The steroid dexamethasone wasused as a GR-binding motif and so attached to a DNAbinder (psoralen) via a short spacer (Rebuffat et al, 2001).The small increase in transfection activity seen incomparison to lipofection of unmodified DNA in dividingcells, became more pronounced (15-40 fold) in nondividingcells. Importantly, increasing the number of GRtargeting molecules attached to the DNA (via psoralencoupling and thus without regiospecificity) led to loss oftransgene expression probably due to covalent damage tothe reporter gene sequence in the plasmid. These studieswere therefore extended, and the preparation of a peptidenucleic acid (PNA) ‘clamp’ followed, which wasdemonstrated to link dexamethasone to the DNA at adefined position where transgene expression remainedunhindered (Rebuffat et al, 2002).VI. Extracellular barriers andrequirements of in vivo lipofectionThe applicability of cationic lipids for in vivo genedelivery was investigated following their provenefficiency for in vitro gene transfection into a greatdiversity of established cell lines and primary cell cultures.Although some promising results have been reported, suchas the transfection of the airway epithelium via the airwaypassages, numerous studies in animals yielded much lesssatisfactory results, especially as regards systemicadministration (Gao and Huang, 1995; Miller, 1998).Accordingly, in the first gene therapy clinical trials in manwith cationic lipids, the lipoplexes were applied to thepatients via in situ administration such as intranasalinstillation or direct intratumoral injection (Hersh and283

Martin et al: Advances in cationic lipid-mediated gene deliveryStopeck, 1998; Alton et al, 1999; Boucher, 1999;Griesenbach et al, 1999; Davies et al, 2001).Identifying the numerous environments that thelipoplex is to encounter on passage to the target cells andhighlighting the transfection-limiting barriers therein hasbeen a particular priority. Firstly, serum-associatedinhibition of lipofection has been reported (Li and Huang,1996). Indeed, binding of the positive lipoplexes to thenegatively charged molecules found in the serum may leadto their neutralisation, thereby hindering the non-specificelectrostatic interaction of the lipoplex with the plasmamembrane of the target cells. However, serum sensitivityof lipofection was found to be dependent on the type ofcationic lipid, as different observations were noted bydifferent investigators including serum resistance andlipoplex charge ratio and size dependence of the inhibitoryeffect of serum (Brunette et al, 1992; Yang and Huang,1997; Turek et al, 2000; Almofti et al, 2003). Our ownresults with guanidinium-based cationic lipids have shownfor example that the in vitro transfection activity of lipidBGTC, of a bisguanidinylated diacetylene lipid and of anacylhydrazone linked bisguanylated lipid were notdecreased in the presence of serum (Patel et al, 2001;Aissaoui et al, submitted for publication).The second barrier to lipofection in vivo is thepresence of opsonins in the blood stream which may bindto the positive lipoplexes and trigger their rapid clearancefrom the blood via uptake by the macrophages of thereticuloendothelial system, thereby leading to a decreasedcirculation of the lipoplexes and hindering DNA uptake bythe target tissues (Gao and Huang, 1995; Li and Huang,1996; Aissaoui et al, 2002). The components of thecomplement system are likely to be involved inopsonisation of intravenously administered lipoplexes(Plank et al, 1996). The degree of complement activationhas been shown to depend on the type of cationic lipidused, with monovalent lipids being weak activators.However, incubation of the cationic vectors with DNA toform complexes was found to reduce complementactivation, as was coating the DNA complexes withpolyethyleneglycol (PEG) (Plank et al, 1996). Thus bylimiting the half-life and the targetability of the lipoplexes,complement activation appears to be a potential barrier forintravenous gene delivery, although it may be minimisedby appropriate vector formulation.A third barrier to in vivo lipofection is theextracellular matrix which is present between cells andprotects the plasma membrane of the target cells (Felgner,1999). The extracellular matrix contains negativelycharged glycosaminoglycans capable of interacting with,and limiting the diffusion of, positive lipoplexes. It is alsonoteworthy that DNases present in the serum and theextracellular space could rapidly degrade unprotectedDNA.Finally, it should be stressed here that in vivolipofection usually necessitates the administration ofhighly concentrated solutions of lipoplexes which shouldideally be electrically neutral and carry a specific ligandfor receptor-mediated uptake when cell targeting isrequired. Unfortunately, it is common knowledge that thepreparation of such lipoplexes is problematic, as highDNA concentrations and electroneutrality lead to colloidalinstability resulting in the flocculation and precipitation ofthe lipoplexes (Lee and Huang, 1997; Lasic et al, 1998).We and others have reported the use of lipophilic PEGderivatives to sterically stabilise lipoplexes formed at highDNA concentration (Hong et al, 1997; Wheeler et al,1999; Pitard et al, 2001). Indeed, it has been reported thatincorporation of PEG derivatives into liposomes (therebycreating so-called ‘stealth’ liposomes) resulted inprolonged circulation times in blood, the PEG polymerforming an exclusion barrier around the liposome whichhinders its aggregation as well as opsonisation in thepresence of serum components (Klibanov et al, 1990;Lasic et al, 1991; Hong et al, 1997; Martin and Boulikas,1998; Zhang et al, 1999). However, although PEGcoatings have been undoubtably useful for the preparationof stabilised particles, recent reports state that theprotection offered by the polymer can inhibit both bindingto the target cell and, upon internalisation, release of thelipoplex from the endosomal compartment causing PEGconcentration-dependent decreases in levels of transfection(Harvie et al, 2000; Shi et al, 2002; Song et al, 2002;Keller et al, 2003). Although careful choice of PEG lengthis likely to help avoid such inhibition (Mori et al, 1991),alternative options are also being sought such as thepreparation of stabilised liposomes which are able to losetheir protective coat, in particular after their accumulationat tumor sites (Martin and Boulikas, 1998). DisuphidelinkedPEG chains have been reported, which, oncecleaved intracellularly, have been shown to rapidly releasethe liposomal contents and re-enable membrane fusionproperties (Kirpotin et al, 1996). Recovered transfectionactivity has also been achieved by varying the length ofthe acyl chain connecting a PEG polymer to a ceramidegroup used to anchor it to the lipsosome, short acyl chainsbeing found to cause rapid dissociation of the PEG coating(Wheeler et al, 1999). In addition, entirely differentstabilising agents are proposed based on a singlecarbohydrate linked to cholesterol. These neoglycolipidshave been reported to confer the stabilisation properties ofPEGylated systems, but without impairing transfectionefficiency (Perouzel et al, 2003).Finally, targeting ligands such as monoclonalantibodies (Trubetskoy et al, 1992), the iron-carryingprotein transferrin (Zenke et al, 1990) and sugars (Wagner,1998; Fajac et al, 1999) have been shown to conferselective delivery to target cells. For example,lipopolyamine-condensed DNA particles decorated with atriantennary galactose ligand have allowed targeted genetransfer into hepatoma cells (Remy et al, 1995). Aninteresting concept is use of the PEG coating as a linker toequip the lipoplexes with specific ligands for targetedreceptor-mediated gene delivery. It has been shown thatcoupling of plasminogen to the ends of PEG chains led tolong circulation times and effective target binding of thePEG-modified liposomes (Blume et al, 1993). In anelegant study, Behr and coworkers introduce, as an outerenvelope to a nanometric pre-condensed lipoplex, a PEGchain which acts as linker between the targeting ligandfolic acid at one end and DNA binding agents at the other(Zuber et al, 2003). However, although stability and folate284

Gene Therapy and Molecular Biology Vol 7, page 285targeting was confirmed, this supramolecular assemblywas not efficient. The inclusion of endosomolytic andnuclear uptake agents to the lipid-bearing nanoparticles isexpected to be required for optimisation of such a ‘viruslike’gene delivery system (Zuber et al, 2001).VII. ConclusionsCationic lipid-mediated gene delivery has nowpassed from the in vitro stage of validation onto the in vivostage, and this with early clinical data emerging. It is clearthough that the future of non-viral systems will require anincrease in efficiency of gene delivery such thattherapeutic levels of gene expression can be attained (Liand Huang, 2000). Because the multiple barriers to nonviralgene delivery are only just becoming clear at themolecular level, the advances in vector design,formulation and modular assembly are at present focusedon surmounting single, at most a few, of these barriers inany distinct study. This article was intended to alert thereader to research efforts which have included stabilisationof the lipoplex in the extracellular medium, targeting of aparticular cell type by pendant motifs, decomplexation ofthe transgene in response to the drop in endosomal pH orthe reducing environment of the cytosol, and finallytrafficking of the transgene to the perinuclear region withultimately active passage across the nuclear membrane.The future will require the difficult task of incorporatingthe functions capable of conferring the series of actionsdescribed above into a single system, capable of workingin a chronological order with unwanted inter-reactivity ofthe individual functions avoided. Certainly such amultimodular assembly is imaginable as viruses form suchhighly complex systems. For example, the humanadenovirus is of about 90 nm in diameter and infectsrespiratory epithelial cells. Pendant targeting ligands inthe knob of the viral fiber bind to cellular CAR receptorsand locally activate α V integrins which trigger endocytosis.The virus is then delivered to an intracellularcompartment, whereupon it rapidly escapes on drop in pHvia interactions including the binding of the penton baseprotein to α V β 5 integrin. The virus is subsequentlytransported to the perinuclear region by exploitation ofmicrotubule- and dynein/dynactin-dependent mechanisms,upon which it docks with the nuclear pore receptorCAN/Nup214 and disassembles to allow entry of the 36-kb linear DNA into the nucleus (Meier and Greber, 2003).However, reasonably, the required non-viral gene deliverysystem is more likely to resemble something between alipoplex and such a virus, together in structure, complexityand efficacy. The system will be designed to avoid theimmunological and toxicological responses which impedevirus-mediated transfer and the gap in efficiencies betweenthe two methods may be further narrowed by using agreater number of non-viral transfecting particles ofsomewhat lesser efficiency. Clearly this does not obviatethe advances that need to be made to bring non-viral genetransfer into the therapeutic mainstream, a gradualcombination of the current approaches being anindispensable first step. Finally, one might also predictthere will be no ultimate ‘all purpose’ vector, rather eachvector having to be individualised according to the clinicalsetting.AcknowledgementsThis work was supported by the Marie CurieIndividual Fellowship Program and grants from Vaincre laMucoviscidose (Paris, France) and the AssociationFrançaise contre les Myopathies (Evry, France).ReferencesAberle AM, Bennett MJ, Malone RW, Nantz MH (1996) Thecounterion influence on cationic lipid-mediated transfectionof plasmid DNA. Biochim Biophys Acta 1299, 281-283Ahearn A, Malone R (1999) Models of cationic liposomemediated transfection. Gene Ther Mol Biol 4, 159-170Aissaoui A, Martin B, Kan E, Oudrhiri N, Hauchecorne M,Vigneron J-P, Lehn J-M, Lehn P Novel cationic lipidsincorporating an acid-sensitive acylhydrazone linker:synthesis and transfection properties, submitted forpublication.Aissaoui A, Oudrhiri N, Petit L, Hauchecorne M, Kan E, SainlosM, Julia S, Navarro J, Vigneron JP, Lehn JM, Lehn P (2002)Progress in gene delivery by cationic lipids: Guanidiniumcholesterol-basedsystems as an example. Curr DrugTargets 3, 1-16Almofti MR, Harashima H, Shinohara Y, Almofti A, Li W,Kiwada H (2003) Lipoplex size determines lipofectionefficiency with or without serum. Mol Membr Biol 20, 35-43Alton EWFW, Stern M, Farley R, Jaffe A, Chadwick SL, PhillipsJ, Davies J, Smith SN, Browning J, Davies MG, Hodson ME,Durham SR, Li D, Jeffery PK, Scallan M, Balfour R,Eastman SJ, Cheng SH, Smith AE, Meeker D, Geddes DM(1999) Cationic lipid-mediated CFTR gene transfer to thelungs and nose of patients with cystic fibrosis: a double-blindplacebo-controlled trial. Lancet 353, 947-954Anderson WF (1998) Human gene therapy. Nature 392, 25-30Balasubramaniam RP, Bennett MJ, Aberle AM, Malone JG,Nantz MH, Malone RW (1996) Structural and functionalanalysis of cationic transfection lipids: the hydrophobicdomain. Gene Ther 3, 163-172Behr JP, Demeneix B, Loeffler JP, Perez-Mutul J (1989)Efficient gene transfer into mammalian primary endocrinecells with lipopolyamine-coated DNA. Proc Natl Acad SciU S A 86, 6982-6986Belmont P, Aissaoui A, Hauchecorne M, Oudrhiri N, Petit L,Vigneron J-P, Lehn J-M, Lehn P (2002) Aminoglycosidederivedcationic lipids as efficient vectors for genetransfection in vitro and in vivo. J Gene Med 4, 517-526Bennett MJ, Aberle AM, Balasubramaniam RP, Malone JG,Malone RW, Nantz MH (1997) Cationic lipid-mediated genedelivery to murine lung: correlation of lipid hydration with invivo transfection activity. J Med Chem 40, 4069-4078Blume G, Cevc G, Crommelin MDJA, Bakker-WoudenbergIAJM, Kluft C, Storm G (1993) Specific targeting withpolyethylene glycol-modified liposomes: coupling of homingdevices to the ends of the polymeric chains combineseffective target binding with long circulation times. BiochimBiophys Acta 1149, 180-184Boomer JA, Thompson DH, Sullivan SM (2002) Formation ofPlasmid-Based Transfection Complexes with an Acid-LabileCationic Lipid: Characterization of in Vitro and in VivoGene Transfer. Pharm Res 19, 1292-1301285

Martin et al: Advances in cationic lipid-mediated gene deliveryBoucher RC (1999) Status of gene therapy for cystic fibrosislung disease. J Clin Invest 103, 441-445Boulikas T (1997) Nuclear localization signal peptides for theimport of plasmid DNA in gene therapy (Review). Int JOncol 10, 301-309Boussif O, Lezoualch F, Zanta MA, Mergny MD, Scherman D,Demeneix B, Behr JP (1995) A Versatile Vector for Geneand Oligonucleotide Transfer into Cells in Culture and in-Vivo - Polyethylenimine. Proc Natl Acad Sci U S A 92,7297-7301Brunette E, Stribling R, Debs R (1992) Lipofection does notrequire the removal of serum. Nucleic Acids Res 20, 1151Byk G, Dubertret C, Escriou V, Frederic M, Jaslin G, Rangara R,Pitard B, Crouzet J, Wils P, Schwartz B, Scherman D (1998)Synthesis, Activity, and Structure-Activity RelationshipStudies of Novel Cationic Lipids for DNA Transfer. J MedChem 41, 224-235Capecchi MR (1980) High efficiency transformation by directmicroinjection of DNA into cultured mammalian cells. Cell22, 479-488Ciolina C, Byk G, Blanche F, Thuillier V, Scherman D, Wils P(1999) Coupling of Nuclear Localization Signals to PlasmidDNA and Specific Interaction of the Conjugates withImportin a. Bioconjug Chem 10, 49-55Collas P, Husebye H, Alestrom P (1996) The nuclear localizationsequence of the SV40 T antigen promotes transgene uptakeand expression in zebrafish embryo nuclei. Transgenic Res5, 451-458Crystal RG (1995) Transfer of genes to humans: early lessonsand obstacles to success. Science 270, 404-410Davies JC, Geddes DM, Alton EW (2001) Gene therapy forcystic fibrosis. J Gene Med 3, 409-417Demeneix BA, Behr J-P. The Proton Sponge: A Trick theViruses Did Not Exploit. In: Felgner PL, Heller, H.J., Lehn,P., Behr, J.-P., Szoka, Jr.F.C. Eds (ed). Artificial Self-Assembling Systems for Gene Delivery. American ChemicalSociety, Washington, DC, USA, 1996, pp. 146-151.Ellens H, Bentz J, Szoka FC (1986) Fusion ofphosphatidylethanolamine-containing liposomes andmechanism of La-HII phase transition. Biochemistry 25,4141-4147Escriou V, Carriere M, Bussone F, Wils P, Scherman D (2001)Critical assessment of the nuclear import of plasmid duringcationic lipid-mediated gene transfer. J Gene Med 3, 179-187Escriou V, Carriere M, Scherman D, Wils P (2003) NLSbioconjugates for targeting therapeutic genes to the nucleus.Adv Drug Deliv Rev 55, 295-306Escriou V, Ciolina C, Helbling-Leclerc A, Wils P, Scherman D(1998) Cationic lipid-mediated gene transfer: analysis ofcellular uptake and nuclear import of plasmid DNA. CellBiol Toxicol 14, 95-104Ewert K, Ahmad A, Evans HM, Schmidt H-W, Safinya CR(2002) Efficient synthesis and cell-transfection properties ofa new multivalent cationic lipid for nonviral gene delivery. JMed Chem 45, 5023-5029Fajac I, Briand P, Monsigny M, Midoux P (1999) Sugarmediateduptake of glycosylated polylysines and genetransfer into normal and cystic fibrosis airway epithelialcells. Hum Gene Ther 10, 395-406Farhood H, Serbina N, Huang L (1995) The Role of DioleoylPhosphatidylethanolamine in Cationic Liposome-MediatedGene-Transfer. Biochim Biophys Acta 1235, 289-295Fasbender A, Zabner J, Zeiher BG, Welsh MJ (1997) A low rateof cell proliferation and reduced DNA uptake limit cationiclipid-mediated gene transfer to primary cultures of ciliatedhuman airway epithelia. Gene Ther 4, 1173-1180Felgner JH, Kumar R, Sridhar CN, Wheeler CJ, Tsai YJ, BorderR, Ramsey P, Martin M, Felgner PL (1994) Enhanced genedelivery and mechanism studies with a novel series ofcationic lipid formulations. J Biol Chem 269, 2550-2561Felgner PL. Progress in gene delivery research and development.In: Huang L, Hung, M.C., Wagner, E., Eds (ed). NonviralVectors for Gene Therapy. Academic Press, San Diego,USA, 1999, pp. 25-38.Felgner PL, Gadek TR, Holm M, Roman R, Chan HW, Wenz M,Northrop JP, Ringold GM, Danielsen M (1987) Lipofection:a highly efficient, lipid-mediated DNA-transfectionprocedure. Proc Natl Acad Sci U S A 84, 7413-7417Felgner PL, Ringold GM (1989) Cationic liposome-mediatedtransfection. Nature 337, 387-388Fichert T, Regelin A, Massing U (2000) Synthesis andtransfection properties of novel non-toxic monocationiclipids. Variation of lipid anchor, spacer and head groupstructure. Bioorg Med Chem Lett 10, 787-791Floch V, Legros N, Loisel S, Guillaume C, Guilbot J, BenvegnuT, Ferrieres V, Plusquellec D, Ferec C (1998) Newbiocompatible cationic amphiphiles derivative from glycinebetaine: a novel family of efficient nonviral gene transferagents. Biochem Biophys Res Commun 251, 360-365Friend DS, Papahadjopoulos D, Debs RJ (1996) Endocytosis andintracellular processing accompanying transfection mediatedby cationic liposomes. Biochim Biophys Acta 1278, 41-50Fujiwara T, Hasegawa S, Hirashima N, Nakanishi M, Ohwada T(2000) Gene transfection activities of amphiphilic steroidpolyamineconjugates. Biochim Biophys Acta 1468, 396-402Gao X, Huang L (1991) A novel cationic liposome reagent forefficient transfection of mammalian cells. Biochem BiophysRes Commun 179, 280-285Gao X, Huang L (1993) Cationic liposomes and polymers forgene transfer. J Lip Res 3, 17-30Gao X, Huang L (1995) Cationic liposome-mediated genetransfer. Gene Ther 2, 710-722Geall AJ, Eaton MAW, Baker T, Catterall C, Blagbrough IS(1999) The regiochemical distribution of positive chargesalong cholesterol polyamine carbamates plays significantroles in modulating DNA binding affinity and lipofection.FEBS Lett 459, 337-342Gershon H, Ghirlando R, Guttman SB, Minsky A (1993) Modeof formation and structural features of DNA-cationicliposome complexes used for transfection. Biochemistry 32,7143-7151Ghosh YK, Visweswariah SS, Bhattacharya S (2000) Nature oflinkage between the cationic headgroup and cholesterylskeleton controls gene transfection efficiency. FEBS Lett473, 341-344Goff SP (2001) Intracellular trafficking of retroviral genomesduring the early phase of infection: viral exploitation ofcellular pathways. J Gene Med 3, 517-528Gottschalk S, Sparrow JT, Hauer J, Mims MP, Leland FE, WooSLC, Smith LC (1996) A novel DNA-peptide complex forefficient gene transfer and expression in mammalian cells.Gene Ther 3, 448-457Griesenbach U, Geddes DM, Alton EWFW. Cystic fibrosis genetherapy. In: Huang L, Hung, M.C., Wagner, E., Eds (ed).Nonviral Vectors for Gene Therapy. Academic Press, SanDiego, USA, 1999, pp. 337-356.Guenin E, Herve A-C, Floch V, Loisel S, Yaouanc J-J, ClementJ-C, Ferec C, Des Abbayes H (2000) Cationicphosphonolipids containing quaternary phosphonium andarsonium groups for DNA transfection with good efficiencyand low cellular toxicity. Angew Chem Int Ed 39, 629-631286

Gene Therapy and Molecular Biology Vol 7, page 287Guo X, Szoka FC, Jr. (2003) Chemical Approaches toTriggerable Lipid Vesicles for Drug and Gene Delivery. AccChem Res 36, 335-341Gustafsson J, Arvidson G, Karlsson G, Almgren M (1995)Complexes between cationic liposomes and DNA visualizedby cryo-TEM. Biochim Biophys Acta 1235, 305-312Harvie P, Wong FMP, Bally MB (2000) Use of poly(ethyleneglycol)-lipid conjugates to regulate the surface attributes andtransfection activity of lipid-DNA particles. J Pharm Sci 89,652-663Hasegawa S, Hirashima N, Nakanishi M (2001) Microtubuleinvolvement in the intracellular dynamics for genetransfection mediated by cationic liposomes. Gene Ther 8,1669-1673Hasegawa S, Hirashima N, Nakanishi M (2002) Comparativestudy of transfection efficiency of cationic cholesterolsmediated by liposomes-based gene delivery. Bioorg MedChem Lett 12, 1299-1302Hendrix M, Alper PB, Priestley ES, Wong C-H (1997)Hydroxyamines as a new motif for the molecular recognitionof phosphodiesters: implications for aminoglycoside-RNAinteractions. Angew Chem Int Ed 36, 95-98Hersh EM, Stopeck AT. Cancer gene therapy using nonviralvectors: preclinical and clinical observations. In: KabanovAV, Felgner P.L., Seymour, L.W. Eds (ed). Self-assemblingComplexes for Gene Delivery. John Wiley and Sons,Chichester, UK, 1998, pp. 421-436.Heyes JA, Niculescu-Duvaz D, Cooper RG, Springer CJ (2002)Synthesis of Novel Cationic Lipids: Effect of StructuralModification on the Efficiency of Gene Transfer. J MedChem 45, 99-114Hong K, Zheng W, Baker A, Papahadjopoulos D (1997)Stabilization of cationic liposome-plasmid DNA complexesby polyamines and poly(ethylene glycol)-phospholipidconjugates for efficient in vivo gene delivery. FEBS Lett400, 233-237Ihm J-E, Han K-O, Han I-K, Ahn K-D, Han D-K, Cho C-S(2003) High Transfection Efficiency of Poly(4-vinylimidazole) as a New Gene Carrier. Bioconjug Chem14, 707-708Kaneko T, Willner D, Monkovic I, Knipe JO, Braslawsky GR,Greenfield RS, Vyas DM (1991) New hydrazone derivativesof Adriamycin and their immunoconjugates - a correlationbetween acid stability and cytotoxicity. Bioconjug Chem 2,133-141Keller M, Harbottle RP, Perouzel E, Colin M, Shah I, Rahim A,Vaysse L, Bergau A, Moritz S, Brahimi-Horn C, Coutelle C,Miller AD (2003) Nuclear localisation sequence templatednonviral gene delivery vectors: Investigation of intracellulartrafficking events of LMD and LD vector systems.Chembiochem 4, 286-298Kirpotin D, Hong K, Mullah N, Papahadjopoulos D, Zalipsky S(1996) Liposomes with detachable polymer coating:destabilization and fusion ofdioleoylphosphatidylethanolamine vesicles triggered bycleavage of surface-grafted poly(ethylene glycol). FEBSLett 388, 115-118Klibanov AL, Maruyama K, Torchilin VP, Huang L (1990)Amphipathic polyethylene glycols effectively prolong thecirculation time of liposomes. FEBS Lett 268, 235-237Koltover I, Salditt T, Radler JO, Safinya CR (1998) An invertedhexagonal phase of cationic liposome-DNA complexesrelated to DNA release and delivery. Science 281, 78-81Kootstra NA, Verma IM (2003) Gene therapy with viral vectors.Annu Rev Pharmacol Toxicol 43, 413-439Kratz F, Beyer U, Schumacher P, Kruger M, Zahn H, Roth T,Fiebig HH, Unger C (1997) Synthesis of new maleimidederivatives of daunorubicin and biological activity of acidlabile transferrin conjugates. Bioorg Med Chem Lett 7, 617-622Kumar VV, Pichon C, Refregiers M, Guerin B, Midoux P,Chaudhuri A (2003) Single histidine residue in head-groupregion is sufficient to impart remarkable gene transfectionproperties to cationic lipids: evidence for histidine-mediatedmembrane fusion at acidic pH. Gene Ther 10, 1206-1215Labat-Moleur F, Steffan AM, Brisson C, Perron H, Feugeas O,Furstenberger P, Oberling F, Brambilla E, Behr JP (1996) Anelectron microscopy study into the mechanism of genetransfer with lipopolyamines. Gene Ther 3, 1010-1017Lasic DD, Martin FJ, Gabizon A, Huang SK, Papahadjopoulos D(1991) Sterically stabilized liposomes: a hypothesis on themolecular origin of the extended circulation times. BiochimBiophys Acta 1070, 187-192Lasic DD, Papahadjopoulos D, Podgornik R. Polymorphism oflipids, nucleic acids and their interactions. In: Kabanov AV,Felgner P.L., Seymour, L.W. Eds (ed). Self-assemblingComplexes for Gene Delivery. John Wiley and Sons,Chichester, UK, 1998, pp. 3-26.Lasic DD, Strey H, Stuart MCA, Podgornik R, Frederik PM(1997) The Structure of DNA-Liposome Complexes. J AmChem Soc 119, 832-833Laxmi AA, Vijayalakshmi P, Balagopala Kaimal TN, ChaudhuriA, Ramadas Y, Rao NM (2001) Novel Non-Glycerol-BasedCytofectins with Lactic Acid-Derived Head Groups.Biochem Biophys Res Commun 289, 1057-1062Lechardeur D, Sohn KJ, Haardt M, Joshi PB, Monck M, GrahamRW, Beatty B, Squire J, O'Brodovich H, Lukacs GL (1999)Metabolic instability of plasmid DNA in the cytosol: apotential barrier to gene transfer. Gene Ther 6, 482-497Lee RJ, Huang L (1997) Lipidic vector systems for gene transfer.Crit Rev Ther Drug Carrier Syst 14, 173-206Lehn J-M (1993) Supramolecular chemistry - molecularinformation and the design of supramolecular materials.Makromol Chem, Macromol Symp 69, 1-17Lehn J-M. Preface. In: Huang L, Hung, M.C., Wagner, E., Eds(ed). Nonviral Vectors for Gene Therapy. Academic Press,San Diego, USA, 1999, pp. xxi.Lehn P, Fabrega S, Oudrhiri N, Navarro J (1998) Gene deliverysystems: Bridging the gap between recombinant viruses andartificial vectors. Adv Drug Deliv Rev 30, 5-11Li S, Huang L (1996) Lipidic supramolecular assemblies forgene transfer. J Lip Res 6, 589-608Li S, Huang L (2000) Nonviral gene therapy: promises andchallenges. Gene Ther 7, 31-34Lleres D, Dauty E, Behr JP, Mely Y, Duportail G (2001) DNAcondensation by an oxidizable cationic detergent.Interactions with lipid vesicles. Chem Phys Lipids 111, 59-71Martin F, Boulikas T (1998) The challenge of liposomes in genetherapy. Gene Ther Mol Biol 1, 173-214Mazzanti M, Bustamante JO, Oberleithner H (2001) Electricaldimension of the nuclear envelope. Physiol Rev 81, 1-19Meier O, Greber UF (2003) Adenovirus endocytosis. J GeneMed 5, 451-462Meister A, Anderson ME (1983) Glutathione. Annu RevBiochem 52, 711-760Miller AD (1998) Cationic liposomes for gene therapy. AngewChem Int Ed 37, 1769-1785Miller AD (2003) The problem with cationic liposome/micellebasednon-viral vector systems for gene therapy. Curr MedChem 10, 1195-1211Moazed D, Noller HF (1987) Interaction of antibiotics withfunctional sites in 16S ribosomal RNA. Nature 327, 389-394Monsigny M, Roche A-C, Midoux P, Mayer R (1994)Glycoconjugates as carriers for specific delivery oftherapeutic drugs and genes. Adv Drug Deliv Rev 14, 1-24287

Martin et al: Advances in cationic lipid-mediated gene deliveryMoradpour D, Schauer JI, Zurawski J, Vincent R., Wands JR,Boutin RH (1996) Efficient Gene Transfer into MammalianCells with Cholesteryl-Spermidine. Biochem Biophys ResCommun 221, 82-88Mori A, Klibanov AL, Torchilin VP, Huang L (1991) Influenceof the steric barrier activity of amphipathic polyethyleneglycol and ganglioside GM1 on the circulation time ofliposomes and on the target binding of immunoliposomes invivo. FEBS Lett 284, 263-266Mortimer I, Tam P, MacLachlan I, Graham RW, Saravolac EG,Joshi PB (1999) Cationic lipid-mediated transfection of cellsin culture requires mitotic activity. Gene Ther 6, 403-411Mueller BM, Wrasildo WA, Reisfeld RA (1990) Antibodyconjugates with morpholinodoxorubicin and acid-cleavablelinkers. Bioconjug Chem 1, 325-330Mukherjee S, Ghosh RN, Maxfield FR (1997) Endocytosis.Physiol Rev 77, 759-803Mulligan RC (1993) The basic science of gene therapy. Science260, 926-932Nagasaki T, Taniguchi A, Tamagaki S (2003) Photoenhancementof Transfection Efficiency Using Novel Cationic LipidsHaving a Photocleavable Spacer. Bioconjug Chem 14, 513-516Neves C, Byk G, Scherman D, Wils P (1999) Coupling of atargeting peptide to plasmid DNA by covalent triple helixformation. FEBS Lett 453, 41-45Nicolau C, Le Pape A, Soriano P, Fargette F, Juhel MF (1983) Invivo expression of rat insulin after intravenousadministration of the liposome-entrapped gene for rat insulinI. Proc Natl Acad Sci U S A 80, 1068-1072Nicolau C, Sene C (1982) Liposome-mediated DNA transfer ineukaryotic cells. Dependence of the transfer efficiency uponthe type of liposomes used and the host cell cycle stage.Biochim Biophys Acta 721, 185-190Oudrhiri N, Vigneron JP, Peuchmaur M, Leclerc T, Lehn JM,Lehn P (1997) Gene transfer by guanidinium-cholesterolcationic lipids into airway epithelial cells in vitro and in vivo.Proc Natl Acad Sci U S A 94, 1651-1656Parente RA, Nadasdi L, Subbarao NK, Szoka FC, Jr. (1990)Association of a pH-sensitive peptide with membranevesicles: role of amino acid sequence. Biochemistry 29,8713-8719Parente RA, Nir S, Szoka FC, Jr. (1990) Mechanism of leakageof phospholipid vesicle contents induced by the peptideGALA. Biochemistry 29, 8720-8728Patel M, Vivien E, Hauchecorne M, Oudrhiri N, Ramasawmy R,Vigneron JP, Lehn P, Lehn JM (2001) Efficient genetransfection by bisguanylated diacetylene lipid formulations.Biochem Biophys Res Commun 281, 536-543Perouzel E, Jorgensen MR, Keller M, Miller AD (2003)Synthesis and Formulation of Neoglycolipids for theFunctionalization of Liposomes and Lipoplexes. BioconjugChem 14, 884-898Phan UT, Arunachalam B, Cresswell P (2000) Gammainterferon-induciblelysosomal thiol reductase (GILT).Maturation, activity, and mechanism of action. J Biol Chem275, 25907-25914Pichon C, Goncalves C, Midoux P (2001) Histidine-rich peptidesand polymers for nucleic acids delivery. Adv Drug DelivRev 53, 75-94Pitard B, Aguerre O, Airiau m, Lachages A-M, BoukhnikachviliT, Byk G, Dubertret C, Herviou C, Scherman D, Mayaux J-F, Crouzet J (1997) Virus-sized self-assembling lamellarcomplexes between plasmid DNA and cationic micellespromote gene transfer. Proc Natl Acad Sci U S A 94,14412-14417Pitard B, Oudrhiri N, Lambert O, Vivien E, Masson C, Wetzer B,Hauchecorne M, Scherman D, Rigaud JL, Vigneron JP, LehnJM, Lehn P (2001) Sterically stabilized BGTC-basedlipoplexes: structural features and gene transfection into themouse airways in vivo. J Gene Med 3, 478-487Pitard B, Oudrhiri N, Vigneron JP, Hauchecorne M, Aguerre O,Toury R, Airiau M, Ramasawmy R, Scherman D, Crouzet J,Lehn JM, Lehn P (1999) Structural characteristics ofsupramolecular assemblies formed by guanidiniumcholesterolreagents for gene transfection. Proc Natl AcadSci U S A 96, 2621-2626Plank C, Mechtler K, Szoka FC, Jr., Wagner E (1996) Activationof the complement system by synthetic DNA complexes: Apotential barrier for intravenous gene delivery. Hum GeneTher 7, 1437-1446Pollard H, Toumaniantz G, Amos JL, Avet-Loiseau H, GuihardG, Behr JP, Escande D (2001) Ca2+-sensitive cytosolicnucleases prevent efficient delivery to the nucleus of injectedplasmids. J Gene Med 3, 153-164Radler JO, Koltover I, Salditt T, Safinya CR (1997) Structure ofDNA-cationic liposome complexes: DNA intercalation inmultilamellar membranes in distinct interhelical packingregimes. Science 275, 810-814Rebuffat A, Bernasconi A, Ceppi M, Wehrli H, Verca SB,Ibrahim M, Frey BM, Frey FJ, Rusconi S (2001) Selectiveenhancement of gene transfer by steroid-mediated genedelivery. Nat Biotechnol 19, 1155-1161Rebuffat AG, Nawrocki AR, Nielsen PE, Bernasconi AG,Bernal-Mendez E, Frey BM, Frey FJ (2002) Gene deliveryby a steroid-peptide nucleic acid conjugate. FASEB J 16,1426-1428, 1410.1096/fj.1401-0706fjeRemy J-S, Kichler A, Mordvinov V, Schuber F, Behr J-P (1995)Targeted gene transfer into hepatoma cells withlipopolyamine-condensed DNA particles presentinggalactose ligands: a stage toward artificial viruses. Proc NatlAcad Sci U S A 92, 1744-1748Remy J-S, Sirlin C, Vierling P, Behr J-P (1994) Gene Transferwith a Series of Lipophilic DNA-Binding Molecules.Bioconjug Chem 5, 647-654Roth JA, Cristiano RJ (1997) Gene therapy for cancer: what havewe done and where are we going? J Natl Cancer Inst 89,21-39Safinya CR, Koltover I. I. Self-assembled structures oflipid/DNA nonviral gene delivery systems from synchrotronX-ray diffraction. In: Huang L, Hung, M.C., Wagner, E., Eds(ed). Nonviral Vectors for Gene Therapy. Academic Press,San Diego, USA, 1999, pp. 91-117.Sainlos M, Belmont P, Vigneron J-P, Lehn P, Lehn J-M (2003)Aminoglycoside-Derived Cationic Lipids for GeneTransfection: Synthesis of Kanamycin A Derivatives. Eur JOrg Chem 15, 2764-2774Saito G, Swanson JA, Lee K-D (2003) Drug delivery strategyutilizing conjugation via reversible disulfide linkages: roleand site of cellular reducing activities. Adv Drug Deliv Rev55, 199-215Schmid N, Behr JP (1991) Location of spermine and otherpolyamines on DNA as revealed by photoaffinity cleavagewith poly(amino)benzenediazonium salts. Biochemistry 30,4357-4361Seisenberger G, Ried MU, Endress T, Buening H, Hallek M,Brauchle C (2001) Real-time single-molecule imaging of theinfection pathway of an adeno-associated virus. Science 294,1929-1932Shi F, Wasungu L, Nomden A, Stuart Marc CA, Polushkin E,Engberts Jan BFN, Hoekstra D (2002) Interference ofpoly(ethylene glycol)-lipid analogues with cationic-lipidmediateddelivery of oligonucleotides; role of lipidexchangeability and non-lamellar transitions. Biochem J366, 333-341288

Gene Therapy and Molecular Biology Vol 7, page 289Song LY, Ahkong QF, Rong Q, Wang Z, Ansell S, Hope MJ,Mui B (2002) Characterization of the inhibitory effect ofPEG-lipid conjugates on the intracellular delivery of plasmidand antisense DNA mediated by cationic lipid liposomes.Biochim Biophys Acta 1558, 1-13Sternberg B, Sorgi FL, Huang L (1994) New structures incomplex formation between DNA and cationic liposomesvisualized by freeze-fracture electron microscopy. FEBSLett 356, 361-366Suh J, Wirtz D, Hanes J (2003) Efficient active transport of genenanocarriers to the cell nucleus. Proc Natl Acad Sci U S A100, 3878-3882Takeuchi K, Ishihara M, Kawaura C, Noji M, Furuno T,Nakanishi M (1996) Effect of zeta potential of cationicliposomes containing cationic cholesterol derivatives on genetransfection. FEBS Lett 397, 207-209Tang F, Hughes JA (1998) Introduction of a disulfide bond into acationic lipid enhances transgene expression of plasmidDNA. Biochem Biophys Res Commun 242, 141-145Tang F, Hughes JA (1999) Use of Dithiodiglycolic Acid as aTether for Cationic Lipids Decreases the Cytotoxicity andIncreases Transgene Expression of Plasmid DNA in Vitro.Bioconjug Chem 10, 791-796Tang Y, Winkler U, Freed EO, Torrey TA, Kim W, Li H, GoffSP, Morse HC, III (1999) Cellular motor protein KIF-4associates with retroviral gag. J Virol 73, 10508-10513Trubetskoy VS, Torchilin VP, Kennel SJ, Huang L (1992) Use ofN-terminal modified poly(L-lysine)-antibody conjugate as acarrier for targeted gene delivery in mouse lung endothelialcells. Bioconjug Chem 3, 323-327Turek J, Dubertret C, Jaslin G, Antonakis K, Scherman D, PitardB (2000) Formulations which increase the size of lipoplexesprevent serum-associated inhibition of transfection. J GeneMed 2, 32-40Umezawa H, Hooper IR. Aminoglycoside Antibiotics. Springer-Verlag: New York, Heidelburg, 1982.Vidal M, Hoekstra D (1995) In vitro fusion of reticulocyteendocytic vesicles with liposomes. J Biol Chem 270, 17823-17829Vigneron JP, Oudrhiri N, Fauquet M, Vergely L, Bradley JC,Basseville M, Lehn P, Lehn JM (1996) Guanidiniumcholesterolcationic lipids: Efficient vectors for thetransfection of eukaryotic cells. Proc Natl Acad Sci U S A93, 9682-9686Vogel K, Wang S, Lee RJ, Low PS, Chmielewski J (1996)Peptide-mediated release of folate-targeted liposomecontents from endosomal compartments. J Am Chem Soc118, 1581-1586Wagner E. Polylysine-conjugate based DNA delivery. In:Kabanov AV, Felgner P.L., Seymour, L.W. Eds (ed). SelfassemblingComplexes for Gene Delivery. John Wiley andSons, Chichester, UK, 1998, pp. 309-322.Wang J, Guo X, Xu Y, Barron L, Szoka FC, Jr. (1998) Synthesisand Characterization of Long Chain Alkyl Acyl CarnitineEsters. Potentially Biodegradable Cationic Lipids for Use inGene Delivery. J Med Chem 41, 2207-2215Wheeler JJ, Palmer L, Ossanlou M, MacLachlan I, Graham RW,Zhang YP, Hope MJ, Scherrer P, Cullis PR (1999) Stabilizedplasmid-lipid particles: construction and characterization.Gene Ther 6, 271-281Xu YH, Szoka FC (1996) Mechanism of DNA release fromcationic liposome/DNA complexes used in cell transfection.Biochemistry 35, 5616-5623Yang JP, Huang L (1997) Overcoming the inhibitory effect ofserum on lipofection by increasing the charge ratio ofcationic liposome to DNA. Gene Ther 4, 950-960Zabner J, Fasbender AJ, Moninger T, Poellinger KA, Welsh MJ(1995) Cellular and Molecular Barriers to Gene-Transfer bya Cationic Lipid. J Biol Chem 270, 18997-19007Zanta MA, Belguise-Valladier P, Behr JP (1999) Gene delivery:A single nuclear localization signal peptide is sufficient tocarry DNA to the cell nucleus. Proc Natl Acad Sci U S A96, 91-96Zenke M, Steinlein P, Wagner E, Cotten M, Beug H, BirnstielML (1990) Receptor-Mediated Endocytosis of TransferrinPolycation Conjugates - an Efficient Way to Introduce DNAinto Hematopoietic-Cells. Proc Natl Acad Sci U S A 87,3655-3659Zhang YP, Sekirov L, Saravolac EG, Wheeler JJ, Tardi P, ClowK, Leng E, Sun R, Cullis PR, Scherrer P (1999) Stabilizedplasmid-lipid particles for regional gene therapy: formulationand transfection properties. Gene Ther 6, 1438-1449Zhu J, Munn RJ, Nantz MH (2000) Self-Cleaving Ortho EsterLipids: A New Class of pH-Vulnerable Amphiphiles. J AmChem Soc 122, 2645-2646Zuber G, Dauty E, Nothisen M, Belguise P, Behr JP (2001)Towards synthetic viruses. Adv Drug Deliv Rev 52, 245-253Zuber G, Zammut-Italiano L, Dauty E, Behr J-P (2003) Targetedgene delivery to cancer cells: Directed assembly ofnanometric DNA particles coated with folic acid. AngewChem Int Ed 42, 2666-2669Dr. Benjamin Martin289

Martin et al: Advances in cationic lipid-mediated gene delivery290

Gene Therapy and Molecular Biology Vol 7, page 291Gene Ther Mol Biol Vol 7, 291-298, 2003Unusual chemical hypersensitivity of the d(GA) n•d(TC) n repeat in vivo dependent on functionallactose repressorResearch ArticleGerald L. Buldak and Sergei M. Mirkin*Department of Biochemistry and Molecular Genetics, University of Illinois at Chicago, Chicago, IL 60607__________________________________________________________________________________*Correspondence: Sergei M. Mirkin, Tel: (312)-996-9610; fax: (312)-413-0353; e-mail: mirkin@uic.eduKey Words: d(GA) n •d(TC) n , microsatellites, primers, operator deletions, repressor deletion,Abbreviations: Chloroacetaldehyde, (CAA); Escherichia coli, (E. coli); transcription start site, (TSS); isopropyl-D-thiogalactoside,(IPTG);Received: 17 November 2003; Accepted: 23 December 2003; electronically published: December 2003SummaryA microsatellite, d(GA) n •d(TC) n , was inserted upstream of an inducible promoter in an Escherichia coli plasmidand its structure was probed by chemical footprinting in vivo. Hyper-reactivity to the single-strand DNA specificchemical, chloroacetaldehyde, was observed within the repeat, pointing to a structural transition within it.Surprisingly, hyper-reactivity of the d(GA) n •d(TC) n repeat diminished upon increased negative supercoiling causedby transcription. Furthermore, the fine modification pattern of the repeat was inconsistent with H-DNA or otherknown conformations that it adopts in vitro. Finally, functional lactose repressor appeared to be required forchemical hyper-reactivity of the repeat. We believe, therefore, that unanticipated binding of the lactose repressor tothe d(GA) n •d(TC) n repeat, which is non-homologous to its regular binding sites, leads to elevated chemicalsensitivity of the repeat in vivo.I. IntroductionSimple tandem DNA repeats are of great interestbecause of their prevalence within various genomes (Coxand Mirkin, 1997), associations with various hereditarydisorders (Kunkel, 1993) and ability to form various noncanonicalstructures both in vitro and in vivo (Sinden,1994). Among the microsatellites that have drawnconsiderable interest is the repeat, d(GA) n •d(TC) n . Thissequence is known to comprise approximately 0.4-0.5% ofthe human genome, making it among the most highlyrepresented microsatellites (Manor et al, 1988). Thisrepeat has been found to play roles in important moleculartransactions, most notably in areas associated withtranscriptional regulation (Gilmour et al, 1989; Glaser etal, 1990; Wilkins and Lis, 1997; Leibovitch et al, 2002; Luet al, 2003) and at replication borders (Baran et al, 1987;Rao et al, 1988; Baran et al, 1991; Rao, 1994;Krasilnikova et al, 2001). Destabilization of thed(GA) n •d(TC) n repeats has also been implicated inmelanoma thus demonstrating its importance from theclinical perspective (Chakraborty, 2000).The d(GA) n •d(TC) n repeat is capable of formingvarious non-canonical DNA structures in vitro. Thesestructures include intramolecular triplex DNA structures,i.e. H-DNA (Mirkin, 1999), parallel-stranded DNA(Germann et al, 1998) and GA hairpins (Kalisch et al,1998; Ortiz-Lombardia et al, 1998). The formation ofthese structures is dependent upon negative DNAsupercoiling (Sinden, 1994). H-DNA can form in twopossible ways, depending upon which strand of the repeatserves as the third strand (either the polypurine strand orthe polypyrimidine strand). The H-r conformation, wherethe purine strand is the third strand, is stable atphysiological pH in the presence of divalent cations, e.g.Mg 2+ . The H-y conformation, where the polypyrimidinestrand is the third strand, requires protonation of thecytosine residues, and is, thus, unfavorable underphysiological pH (Mirkin, 1999).Attempts to detect intramolecular triplexes formedby d(G-A) n •d(TC) n repeats DNA structures in vivo weremade using chemical probing of intracellular DNA in amodel Escherichia coli system (Karlovsky et al, 1990;Ussery and Sinden, 1993). The conclusions from thesestudies were that intramolecular triplexes could be formedin E. coli when two requirements were met: (i)superhelicity of intracellular DNA was artificially291

Buldak and Mirkin: Hypersensitivity of the d(GA) n •d(TC) n dependent on lactose repressorincreased, usually by growing cells in the presence ofchloramphenicol, and (ii) ambient conditions of cellgrowth favored H-DNA formation by either the presenceof divalent cations or decreased pH of the media.Another group of studies focused on ad(GA) n •d(TC) n repeat found within the promoter region ofthe hsp26 gene in Drosophila. Both the GA repeat and aprotein factor that recognizes it, GAGA-factor, arerequired for the proper functioning during the heat-shockresponse in Drosophila. While this repeat forms H-DNA invitro, in vivo studies, using mutational analysis andchemical footprinting, indicate that protein binding to thisrepeat rather than its triplex-forming potential, is essentialfor the promoter function (Glaser et al, 1990; Lu et al,2003).One can conclude, therefore, that the formation of H-DNA by a d(GA) n •d(TC) n repeat can be induced in E. colicells in principle, but is not readily detectable in a naturalsetting, such as the Drosophila genome. We decided,therefore, to look at the structure of the d(GA) n •d(TC) nrepeats cloned into an E. coli plasmid under physiologicalgrowth conditions. In this setting, one can elevate negativeDNA supercoiling in the upstream promoter region byinducing transcription (Liu and Wang, 1987; Wu et al,1988). We have previously demonstrated that transcriptiondoes, in fact, elevate DNA supercoiling in vivo inducingcruciform formation at d(AT) n •d(TA) n repeats as far as 1kbp upstream of the promoter (Dayn et al, 1992;Krasilnikov et al, 1999).We inserted the d(GA) n •d(TC) n repeat in twopossible orientations upstream of an inducible promoterand probed its structure by chemical modification withchloroacetaldehyde (CAA) in vivo. While we observedchemical hyperreactivity within the repeat, its pattern wasinconsistent with H-DNA or other known structures.Furthermore, this chemical reactivity was repressed, ratherthan enhanced, upon induction of transcription. Finally,hyper-modification of the d(GA) n •d(TC) n repeat requiredfunctional lactose repressor. We believe, therefore, thatlactose repressor could contribute to a structural alterationwithin the d(GA) n •d(TC) n repeat, leading to its chemicalhyper-reactivity in vivo.II. Materials and methodsA. PrimersSequencing primers used were: Sequencing primer Pr1(5'ACGGTGCACCAATGCTTCTG3') and Pr2(5'CCGGCTCGTATAATGTGT3'). Primers for PCR-mediateddeletion of lactose operators were as follows: del1(5'AAAAAGCTTCACTGCCCGCTTTC3');del1'(5'GCCGTCAACCACCATC3');del2(5'AAAAAGCTTAGCGCGAATTGATC3');del2'(5'CGGATAAAACTTGTGC3');del3(5'AAAACTCGAGTTCCACACATTATAC3');del3'(5'AAAACTCGAGTCACACAGGAAACAGAC3'); del4(5'TAGGTACATTGAGCAAC3');del5(5'TGTGACTCGAGTATTCGCTTGCTTATACGAGCCGGATG3').B. Plasmid constructionPlasmid pTrc99A (Amann et al, 1988) was obtained fromPharmacia. Plasmid pTrcCat/Pst, carrying the unique PstI site atthe -50 position relative to the trc promoter, was described in(Krasilnikov et al, 1999). Plasmid pTrcCat/Cla, carrying theunique ClaI site at the -180 position relative to the trc promoter,was a gift from Dr. Andrey Krasilnikov. Inserts containingd(GA) 30 •d(TC) 30 and d(GA) 37 •d(TC) 37 repeats were obtained bythe EcoRI/HindIII digestion of the pGA30 and pGA37 plasmids,respectively, described in (Krasilnikova et al, 2001). Theserepeats were cloned into either the PstI site of pTrcCat/Pstplasmid, or the ClaI of the pTrcCat/Cla plasmid.C. Operator deletionsThe primary operator site, lacO1, was deleted by using thefollowing PCR-based approach. Plasmid pTrcCat/Pst-GA37 wasused as a template for independent PCR reactions with two pairsof primers. Using primers del1' and del3, a fragment of theplasmid was amplified that included the repeat, and the regionimmediately upstream of the O1 site relative to the transcriptionstart site. A second fragment was generated using primers del3'and del4 that included the region immediately downstream of theO1 site. The two fragments were designed such that an XhoIrestriction site replaced the O1 operator upon XhoI-digestion andco-ligation.Pseudo-operator O3 was deleted by a similar approachusing two PCR fragments from the pTrcCat/Pst-GA37 template.One of the fragments was amplified with primers del1 and del1',corresponding to the region immediately upstream of the O3 site.The other fragment was produced with primers del2 and del4,and contained the region immediately downstream of the O3 site.These fragments were generated such that a HindIII restrictionsite substituted for the O3 operator upon HindIII digestion andco-ligation.To construct the plasmid without both operator sites, ΔO1-ΔO3, we deleted the O3 operator from the ΔO1 plasmid using thePCR approach described above.In all these cases, the ligated PCR products contained NarIand KpnI sites at their 5'- and 3'-ends, respectively. Upondigestion with these two restriction enzymes, they were used toreplace the corresponding NarI-KpnI fragment of the originalpTrcCat/Pst plasmid.D. Repressor deletionDeletion of the lacI q gene was achieved by excising the 0.9kbp BssHII fragment, containing the coding part of the repressor,from the pTrcCat/Pst-GA 37 plasmid.E. Bacterial strainsWe used either an XL1-Blue strain (Invitrogen) or itsderivative XL1-BluES, which was cured of the F'-factor by uspreviously (Krasilnikova et al, 2001).F. Chemical probing of intracellular DNAIntracellular plasmid DNA was modified withchloroacetaldehyde as described in (Krasilnikov et al, 1999).G. Radiolabeling of DNA fragments andMaxam-Gilbert sequencing reactionsLabeling and sequencing reactions were carried outaccording to standard protocols (Sambrook et al, 1989).292

Gene Therapy and Molecular Biology Vol 7, page 293III. ResultsTo look at the structure of d(GA) n •d(TC) n repeatsunder the influence of superhelical stress induced bytranscription, we placed this repeat upstream of theinducible trc promoter in the multicopy pTrc-derivedplasmid. DNA footprinting was carried out by CAA invivo under conditions of both promoter induction andrepression. CAA interacts specifically with base pairingpositions of cytosines and adenines. Consequently, thestability of phosphodiester bonds decreases, and, afterMaxam-Gilbert DNA sequencing, one can see additionalbands corresponding to modified cytosines on the purineladder and to modified adenines on the cytosine ladder(Kohwi and Kohwi-Shigematsu, 1988). Since in doublestranded DNA, these positions are inaccessible, CAAmodification reveals single-standed DNA segments orother significant DNA distortions.The experimental results of chemical modification ofboth strands are presented in Figure 1. One can clearlyobserve modification within both the d(GA) n •d(TC) nrepeat and the trc promoter. The size of the repeat herewas 30 units and its subsequent increase to 37 units did notyield any difference in modification patterns. The schemeof modification within the d(GA) n •d(TC) n insert and thetrc promoter region is presented in Figure 2. As can beseen in the figure, adenine residues within the 3' third ofthe polypurine strand were modified above backgroundwhen the insert was placed into the plasmid in such a waythat the purines were on the top strand as drawn. In eachrepeat length, when IPTG was absent from the growingculture, the 13 3'adenine (A) residues were modified.When IPTG was added, the length of the modificationshrank. No modifications were observed within thepolypyrimidine strand.These results ran contrary to what was expected,based on the previous studies of different structure-pronerepeats. First, the modification was more profound in theabsence of transcription. Second, the modification wasonly evident on one DNA strand. Third, the modificationwas observed in only one orientation of the repeat relativeto the promoter. Furthermore, modification results areinconsistent with H-DNA formation by the d(GA) n •d(TC) nrepeat because of the lack of modification of thepolypyrimidine strand.As far as trc promoter structure, we observedprominent chemical hyperreactivity in both repressed andactivated state. When IPTG was absent, distinctmodifications could be observed at positions -8, -9, -11, -14, -16, -19, -20, -40, -43, -44, -45, -49 and -50 relative tothe transcription start site (TSS), designated +1.Modification patterns in all instances described changeddramatically upon addition of IPTG and promoteractivation. All hypermodifications present from -11 to -50disappeared. At the same time, new modifications thatwere not present in the absence of IPTG appeared atpositions +1 and +2. These results are also summarized inFigure 2.To confirm that these modifications within thepromoter were not occurring as a result of the presence ofa d(GA) n •d(TC) n insert, we performed the same type offootprint analysis on the empty vector, pTrcCat/Pst.Figure 1. CAA modification in vivo of the polypurine (left) andpolypyrimidine (right) strands of a d(GA) 37 •d(TC) 37 repeat in thepTrcCat/Pst plasmid. G, R and C represent Maxam-Gilbertsequencing reactions of plasmid DNA in vitro. IPTG - and +indicate CAA modifications in vivo in the absence or presence ofIPTG, respectively. IPTG 2' is a control in which the plasmidDNA was CAA-modified in vivo for two minutes, instead of thenormal 20 minutes. Arrows indicate modified adenine andcytosine bases.Chemical probing of this plasmid in the presence andabsence of IPTG revealed modification pattern of thepromoter region that was identical to that of repeatcontainingplasmids (Figure 3). Altogether, the promoterdata support the model in which RNA polymeraseinteracts with the DNA at positions from -10 to -50 whenthe promoter is in the repressed state. Upon induction,RNA polymerase moves forward such that RNApolymerase-DNA interaction is now readily seen aroundthe TSS instead of the 5'-part of the promoter.The promoter alone is modified up to position -50relative to the TSS. The d(GA) n •d(TC) n insert was insertedimmediately upstream of this point. It is conceivabletherefore, that modification of the repetitive run is causedby a mere extension of the promoter modification. Toaddress this possibility, repeats were moved back to aposition of -180 bp relative to the TSS. The unusualmodifications within the repeat remained when placed atthis position as well (Figure 4). Furthermore, theinhibitory effect of IPTG remained unchanged.293

Buldak and Mirkin: Hypersensitivity of the d(GA) n •d(TC) n dependent on lactose repressorFigure 2. Summary of CAA modifications observed between positions -70 and +5 relative to the TSS. Modified cytosine and/or adenineresidues shown in red occurred only in the absence of IPTG. Modifications shown in green occurred only in the presence of IPTG.Modifications shown in turquoise occurred in both states of promoter induction.These data, therefore, rule out the possibility that themodification of the repeat results from the deformation atthe promoter region.Though modification of the d(GA) n •d(TC) n insert does notseem to be directly linked to the trc promoter activity, it isdowngraded in the presence of IPTG. Since the onlyknown effect of IPTG is to block DNA-binding activity ofthe lactose repressor, we looked at the role of the lactoserepressor in the d(GA) n •d(TC) n insert modification. Tothis end, we first deleted lacI q gene from our test plasmidand transformed the resultant construct into the lacI -background. The data on chemical probing in vivopresented in Figure 5 show that there are no modificationsof the d(GA) n •d(TC) n insert in this setting. At the sametime, the modifications within the promoter region wereidentical to those seen in previous experiments in thepresence of IPTG, showing that trc promoter was in aconstantly induced state. We conclude, therefore, thatfunctional lactose repressor is necessary for a structuralalteration of the d(GA) n •d(TC) n repeat leading to itschemical hyperreactivity in vivo.Lactose repressor is known to bind to at least twooperator sites in our plasmid. The next question istherefore, whether repressor-operator binding is essentialfor the repeat's modification. To answer this question, wedeleted the operator and pseudo-operator, eitherindividually or together. When the primary operator (O1)was deleted, the modification of the d(GA) n •d(TC) n repeatwas similar to that observed in the presence of the normaloperator site (Figure 6). The modification of the promoterregion, as one would expect, became equivalent to that ofthe induced wild-type promoter even in the absence ofIPTG. When the O3 pseudo-operator was deletedindividually, leaving O1 intact, modifications of both theinsert and of the promoter regions (Figure 7A) wereidentical to those seen in the original plasmid assummarized in Figure 2. Furthermore, when both operatorsites were deleted (ΔO1-ΔO3 construct), the results ofmodification were identical to those seen in the case of theΔO1 (Figure 7B). We conclude, therefore, that lactoserepressor per se, rather than its binding to the operatorsites, is necessary for the d(GA) n •d(TC) n hyperreactivity.Figure 3. CAA modification in vivo of the non-transcribedstrand of pTrcCat/Pst. G, C, Maxam-Gilbert sequencing reactionsof plasmid DNA in vitro. IPTG - and + indicate C reactions for invivo CAA-modified plasmid DNAs. Arrows indicate modifiedadenine and cytosine bases.IV. DiscussionWe found that a d(GA) n •d(TC) n repeat does undergostructural changes, accompanied by chemical modification294

Gene Therapy and Molecular Biology Vol 7, page 295Figure 5. CAA modification in vivo of the polypurine strand of ad(GA) 37 •d(TC) 37 repeat, inserted 50 bp upstream of the trc TSSin pTrcCat/Pst-GA37ΔlacI q . G, C, Maxam-Gilbert sequencingreactions of plasmid DNA in vitro. IPTG - and +, C reactions forplasmid DNAs that were CAA-modified in vivo in the absence orpresence of IPTG, respectively. The left panel shows thed(GA) 37 •(d(TC) 37 repeat with the promoter, while right panelonly the repeat. Arrows indicate modified adenine and cytosineresidues.Figure 4. CAA modification in vivo of the polypurine strand of ad(GA) 37 •d(TC) 37 repeat at position -180 relative to the TSS inpTrcCat/Cla-GA37. G, C, Maxam-Gilbert sequencing reactionsof plasmid DNA in vitro. IPTG - and + indicate CAAmodifications in vivo in the absence or presence of IPTG,respectively. IPTG 2' is a control in which the plasmid DNA wasCAA-modified in vivo for two minutes, instead of the normal 20minutes. Arrows indicate modified adenine residues.of the 3'-part of the polypurine strand, when situated in theupstream promoter area. Contrary to expectations, basedon previous studies by us and others, this modificationpattern is not immediately consistent with previouslydescribed types of non-canonical structures adopted bythis repeat.Based on our previous studies (Dayn et al, 1992;Krasilnikov et al, 1999), we assumed that transcriptionalsupercoiling would stimulate structural transitions withinthe d(GA) n •d(TC) n repeat.295