Effective Limb Transduction and Phenotypic Correction after Injection of rAAV8-U7 snRNA in GRMD Dogs

HAL Id: hal-01191233

https://hal.archives-ouvertes.fr/hal-01191233

Submitted on 3 Jun 2020

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Effective Limb Transduction and Phenotypic Correction

after Injection of rAAV8-U7 snRNA in GRMD Dogs

Caroline Le Guiner, Marie Montus, Laurent Servais, Luis Garcia, Yves Fromes,

Jean-Yves Hogrel, Pierre Carlier, Yan Cherel, Philippe Moullier, Thomas Voit

To cite this version:

Caroline Le Guiner, Marie Montus, Laurent Servais, Luis Garcia, Yves Fromes, et al.. Effective

Limb Transduction and Phenotypic Correction after Injection of rAAV8-U7 snRNA in GRMD Dogs.

ASGCT Translational Science Training Course, American Society of Cell and Gene Therapy (ASGCT).

USA., Mar 2011, Seattle, United States. pp.332. �hal-01191233�

Molecular Therapy Volume 19, Supplement 1, May 2011 _S1

A

O

DNA V

V

I: V

E

Adenovirus and Other DNA Virus Vectors I:

Virus Entry

1.

Retargeting Parvovirus B19 to

HER2/neu-Positive Cells for Breast Cancer Treatment

Harald G. P. Messer,1,2 _{Yiwen Xiang,}1 _{Mavis Agbandje-McKenna,}2 Kirsten A. K. Weigel-Van Aken.1

1_{Pediatrics, University of Florida, Gainesville, FL;} 2_Biochemistry and Molecular Biology, University of Florida, Gainesville, FL.

Human epidermal growth factor receptor 2 (HER2), which is over-expressed in breast cancers, is associated with poor differentiation, high rates of proliferation, lymph node involvement, resistance to certain types of chemotherapy, and poor prognosis. A humanized anti-HER2 monoclonal antibody, rhu 4D5 (trastazumab, Herceptin®), has been developed and tested in clinical trials for breast cancer treatment. It was shown to bind to cell surface-expressed HER2, inhibit HER2 kinase activity, down-modulate HER2, and sensitize tumor cells to apoptosis when used in conjunction with radiation or chemotherapy. Previously, an anti-HER2/neu-binding peptidomimetic (AHNP), derived from the structure of Herceptin’’s CDR-H3 loop has been developed to avoid side effects associated with whole antibody treatments and ef cient HER2 interaction was demonstrated in vitro, albeit with reduced HER2 down-modulation. Parvovirus B19 (B19), a small ssDNA-containing virus, is restricted in its replication to erythroid progenitor cells in the human bone marrow. The virus binds to these cells through interactions with the blood group P antigen (globoside), and uses activated α5β1 integrin and Ku80 co-receptors for entry. We thus rationalized that due to its bone marrow tropism, B19 could be an attractive candidate for targeted gene delivery therapies to bone marrow-resident cells, such as metastatic HER2-positive, breast cancer cells, particularly if its primary cellular receptor binding site is replaced by the HER2 cancer cell-targeting peptide, AHNP. We have replaced the endogenous P antigen binding site on the capsid surface of B19 with the AHNP peptide and demonstrate retargeting of chimeric vectors to HER2-positive tumor cells. Structural modeling was used to aid the design of chimeric AHNP-B19 VP2 vectors based on the crystal structures of the B19 capsid VP2 and the CDR-H3 loop of rhu 4D5. Comparative transduction studies with the wt B19 VP2 (wt capsid-B19-scGFP) and AHNP-VP2 (AHNP-B19-scGFP) vectors in murine MMTV-HER2/neu mammary tumor epithelial and human medulloblastoma cells showed signi cant enhancement in transduction of HER2-overexpressing cells with the AHNP-B19-scGFP vector compared to wt capsid-AHNP-B19-scGFP vector. This transduction improvement was significantly reduced following exposure to a HER2 kinase domain-derived peptide that interfered with tyrosine 1248 phosphorylation of the receptor, indicating that HER2 kinase activity is required for AHNP-B19-scGFP vector entry. On the contrary, preincubation of AHNP-B19-scGFP virions with P antigen had no effect on transduction ef ciency, corroborating the successful replacement of the P antigen binding epitope on B19 capsids and introduction of binding to HER2 as the receptor. Importantly, HER2-binding AHNP-B19 vectors still used β1 integrins as co-receptors for infection. This novel engineered vector serves as a promising vehicle for the targeted delivery of cytotoxic transgenes to HER2-positive metastatic cancer cells for the eradication of dormant breast cancer cells in niches, such as the bone marrow, which are highly intractable to most treatment strategies.

2.

Identi cation of the Receptor for

Adenovirus Serotypes 3, 7, 11, and 14: Structural

Analysis of Adenovirus –– Receptor Interaction

Human adenoviruses (Ads) have been classi ed into six species (A to F) currently containing 55 serotypes. Most Ad serotypes utilize the coxsackie-adenovirus-receptor (CAR) as a primary attachment receptor. Group 1 of species B Ads (Ad16, 21, 35, 50) nearly exclusively utilize CD46 as a receptor; Group 2 (Ad3, Ad7, 14) share a common, unidenti ed receptor/s, which is not CD46 and which was tentatively named receptor X; Group 3 (Ad11) preferentially interacts with CD46, but also utilizes receptor X if CD46 is blocked. Recently, using Ad3 virons and recombinant Ad3 PtDds as a probe for receptor X, we identi ed Desmoglein 2 (DSG2) as a high-af nity receptor for AdB-2/3 serotypes. Loss- and gain-of function studies were then performed on cell lines to validate DSG2 as a crucial receptor for Ad binding and infection. First structural studies of Ad3-DSG2 interaction were performed. Ad3 binding involved the third and fourth extracellular domain of DSG2. Competition and surface plasmon resonance studies demonstrated that the DSG2 interacting domain(s) within Ad3 are formed by the  ber only in the spatial constellation that is present in viral particles. This clearly widens our understanding of Ad attachment mechanisms, which, so far, were thought to involve only a high af nity interaction between the  ber knob and the cellular receptor, i.e. CAR or CD46. Chimeric Ad5 vectors with Ad3  bers also bound to DSG2 indicating that the only function of penton is to constrain Ad3  bers as multimers. We then generated a series of N-terminal His-tagged Ad3  ber knob protein containing different numbers of shaft repeats, and found that all these  ber proteins were unable to ef ciently block Ad3 infection. However, cross-linking of  bers with anti-His antibodies increased the blocking ef ciency signi cantly. To further investigate if dimerization is required, K-coil and E-coil heterodimerization domains were included into Ad3  ber proteins. CryoEM and Biacore studies showed  ber dimerization/ multimerization. Ad3  ber dimers ef ciently and speci cally blocked Ad3 infection. Incubation of epithelial cells with Ad3 dimers triggered the same DSG2-mediated opening of epithelial junctions as seen with Ad3 particles.

3.

Multi-Year Transgene Expression

in Nonhuman Primates Following Hepatic

Transduction with Helper-Dependent Adenoviral

Vectors

Nicola Brunetti-Pierri,1 _{Gary Stapleton,}2 _{Thomas Ng,}3 _David Iannitti,3 _{William Ciof ,}3 _{Mark Law,}2 _{John Breinholt,}2 _Donna Palmer,4 _{Nathan Grove,}4 _{Milton Finegold,}5 _{Arthur Beaudet,}4 Charles Mullins,2 _{Philip Ng.}4

1_{Telethon Institute of Genetics and Medicine, Naples, Italy;} 2_{Cardiology, Baylor College of Medicine, Houston, TX;} 3_Surgery, Brown Medical School, Providence, RI; 4_{Molecular and Human} Genetics, Baylor College of Medicine, Houston, TX; 5_Pathology, Baylor College of Medicine, Houston, TX.

Helper-dependent adenoviral (HDAd) vectors hold tremendous potential for liver-directed gene therapy because they mediate long-term transgene expression without chronic toxicity. One obstacle to clinical application of HDAd is that high and thus potentially toxic doses are required to achieve ef cient hepatocyte transduction due, in part, to extrahepatic vector uptake. We previously addressed this obstacle by developing 3 vector delivery methods in baboons to permit

A

O

DNA V

V

I: V

E

preferential hepatocyte transduction by minimizing extrahepatic vector uptake. Method 1 involved HDAd injection directly into the surgically isolated liver via the portal vein (PMID16610927). Method 2 mimics hydrodynamic injection and was based on our observation that hydrodynamic injection in mice resulted in preferential hepatocyte transduction and reduced extrahepatic vector uptake (PMID17285138). Method 3 involved in ating a balloon occlusion catheter in the inferior vena cava to occlude hepatic venous out ow followed by vector injection directly into the occluded liver via the hepatic artery (PMID 19050700). These methods signi cantly improved hepatocyte transduction, up to 80-fold, compared to systemic injection, thus permitting high level transgene expression using low sub-toxic doses. Although achieving high level transgene expression using low vector doses is important, equally important for risk:benefit assessment and successful gene therapy is the duration of transgene expression. Thus, in this study, we provide an update on the duration of transgene expression following delivery of HDAd expressing the reporter alpha-fetoprotein (AFP) by the 3 aforementioned methods. Our results revealed that all 3 methods resulted in sustained transgene expression to date. Speci cally, HDAd injected into baboons (n=3) by Method 1 resulted in up to 6.7 years of transgene expression. Method 2 (n=2) resulted in at least 6 years of transgene expression. Method 3 (n=7) resulted in up to 4.6 years of transgene expression. In all cases, transgene expression declines over time but with a very slow rate.

No chronic toxicity has been observed in any animal. These unprecedented results clearly demonstrate strong multi-year transgene expression in nonhuman primates following a single administration of HDAd and support their potential long-term ef cacy in clinical trials.

4.

The Re nement of the Molecular Model

of Adenovirus Binding to Blood Coagulation

Factors and the Role of This Interaction in Virus

Recognition by the Host

Konstantin Doronin,1 _{Justin Flatt,}2 _{Steffen Lindert,}2,3 _{Dewight R.} Williams,2 _{Phoebe L. Stewart,}2 _{Dmitry M. Shayakhmetov.}1 1_{Division of Medical Genetics, Department of Medicine, University} of Washington, Seattle, WA; 2_{Department of Molecular Physiology} and Biophysics, Vanderbilt University School of Medicine, Nashville, TN; 3_{Department of Molecular Physiology and} Biophysics, Vanderbilt University School of Medicine, Nashville, TN.

Viral vectors based on human adenovirus serotype 5 (Ad5) are highly ef cient tools in gene transfer studies both in vitro and in vivo. However, upon intravascular administration in mice, the bulk of the virus particles is trapped in the liver, leading to clinically signi cant hepatotoxicity. Ad5 hexon was recently shown to be the principal capsid protein mediating virus entry into hepatocytes via interaction with vitamin K-dependent blood coagulation factors, including FX, FVII, or FIX. These factors form the bridge between Ad virion surface and heparan sulfate proteoglycans at the hepatocyte plasma membrane. Using cryo-electron microscopy, in this study, we determined the structure of Ad5 in complex with human blood factors VII and X at a subnanometer (<10Å) resolution. Both Ad5/factor structures show density for the Gla domain of the factor interacting

with the surface loops of hexons in the virus capsid. Modeling of the interaction between the coagulation factor Gla domain and virus hexon indicates that the predicted TET amino acid sequence in the hexon hyper-variable loop 7 (HVR7) is an important site for this interaction. The structure and modeling results also offer a possible explanation for the lower af nity of factor VII (5nM) compared to that of factor X (200pM) for Ad5. Furthermore, we constructed a set of Ad5 vectors possessing single amino acid substitutions in the HVR7 of the hexon protein. The biological properties of hexon-mutated vectors are currently being analyzed both in vitro and in vivo. Our studies are critical for re ning the molecular model of blood coagulation factor interaction with Ad5-based vectors and will aid in de ning the functional role of this interaction for adenovirus recognition by the innate immunity of the host.

5.

Suppression of Protein Phosphatase

2A Activity Enhances Ad5/F35 Adenovirus

Transduction Ef ciency in Human Normal B

Lymphocytes and Cell Lines

Marie-Pierre Cayer,1 _{Mélanie Samson,}1,2 _{Claudia Bertrand,}1,2 _Nellie Dumont,1 _{Mathieu Drouin,}1 _{Daniel Jung.}1,2

1_{Department of Cell Engineering, Héma-Québec, Québec, QC,} Canada; 2_{Department of Biochemistry and Microbiology, Laval} University, Québec, QC, Canada.

Adenoviral vectors have the particularity of ef ciently and rapidly transduce both proliferating and quiescent cells and do not integrate into the host genome, allowing a reproducible gene expression level. We previously demonstrated that adenoviral vector Ad5/F35, which utilizes the ubiquitous CD46 as cellular receptor, displays differences in transduction ef ciency between normal human B lymphocytes and B lymphocytes cell lines as determined by EYFP expression. We have observed that normal B lymphocytes were poorly transduced compared to some B lymphocytes cell lines with less than 50% EYFP+ cells, despite the fact that these cells express a similar level of CD46. In contrast normal plasma cells or plasma cell lines were easily transduced by Ad5/F35 with a rate of up to 95% EYFP+ cells. Here to further investigate the differences in transduction ef ciency we analyzed intracellular traf cking of Ad5/F35 labelled with CY3 and  uorescently tagged endocytic markers by confocal microscopy. We demonstrated that Ad5/F35 failed to escape early endosomes and were redirected to recycling vesicles or late endosomes and lysosomes in B lymphocytes whereas in plasma cell lines Ad5/ F35 easily escaped early endosomes to migrate toward the nucleus. Since phosphastases are well known to in uence endocytosis and intracellular traf cking, B lymphocytes were treated with phosphatase inhibitors speci c for PP1 or PP2A, prior adenoviral infection. Inhibition of PP1 by Tautomycin or PP2B by cyclosporine A had no effects on adenoviral transduction ef ciency whereas inhibition of PP2A with Cantharidin or PP1 and PP2A with Okadaic acid substantially increase transduction ef ciency. Moreover, inhibition of PP2A results in a rapid increase of AKT, ERK1/2 and MEK1/2 phosphorylation. Importantly, confocal microscopy analyses revealed that inhibition of PP2A abolishes the recycling of adenovirus and enhances their escape from early endosome to the nucleus. Finally, no increase of adenoviral entry was observed following PP2A inhibition. Our results are in accord with reports indicating that PP2A is involved in the formation of recycling vesicles and may improve gene transfer in the context of gene therapy.

Molecular Therapy Volume 19, Supplement 1, May 2011 _S3

DNA V

6.

Basic Residues in the Heparin-Binding

Exosite of the Factor X Serine Protease Domain

Mediate Ad5/FX Complex Binding to Hepatocytes

1_{BHF Glasgow Cardiovascular Research Centre, University of} Glasgow, Glasgow, United Kingdom; 2_{Thrombosis Research} Institute, Emmanuel Kaye Building, Manresa Road, London, United Kingdom.

Recent evidence shows that hepatocyte transduction by adenovirus 5 (Ad5) is mediated by a high af nity interaction between coagulation factor X (FX) and the Ad5 hexon proteins. Whilst the FX gamma-carboxylated glutamic acid (Gla) domain mediates binding with the Ad5 capsid, the FX serine protease (SP) domain bridges the Ad5/ FX complex to the hepatocyte surface through an interaction with heparan sulfate proteoglycan receptors (HSPGs). Pharmacological blockade of the heparin-binding exosite in the serine protease domain of FX prevents cell binding and gene transfer mediated through FX1_. Previous studies have indicated that the residues Arg-93, Lys-96, Arg-125, Arg-165, Lys-169, Lys-236 and Arg-240 (chymotrypsin numbering system) of FX constitute this exosite2_{. The aim of this} study was to identify the speci c residues in the SP domain of FX which mediate Ad5 attachment to HSPGs. We therefore generated a FX mutant construct (SP rFX) to assess the importance of this exosite. A human FX plasmid construct with a preprothrombin propeptide (to enhance gamma carboxylation3_{) was generated. Site directed} mutagenesis was carried out at all seven basic residues in the FX SP domain and con rmed by sequence analysis. Stable cell lines were generated to constitutively produce the wildtype and mutant recombinant FX in the presence of vitamin K. The rFX generated was biologically active and had the predicted molecular weight. Enzyme-linked immunosorbent assay was used to quantify rFX production. Surface plasmon resonance analysis showed the SP mutations had no effect on rFX binding to the Ad5 hexon, as was expected, thus any effect on gene transfer was as a result of inef cient cell binding.

In vitro assays in both low and high CAR expressing cell lines

demonstrated that the exosite mutations blocked the ability of FX to enhance cell surface binding (1.75 ± 0.25x104 _{SP mutant rFX vs.} 10.28 ± 0.276x104 _{WT rFX vector genomes, p<0.01) and transduction} (4.12 ± 0.23x104 _{SP mutant rFX vs. 37.3 ± 2.72x10}4 _{WT rFX RLU/} mg of protein, p<0.01) of Ad5 thus con rming the importance of this exosite in HSPG engagement. These results show that modi cation of the heparin-binding exosite of the FX serine protease domain ablates HSPG engagement by Ad5/FX complexes. [1] Waddington 2008 Cell 132, 397-409. [2] Rezaie 2000 JBC 132, 3320-3327. [3] Camire 2000 Biochemistry 39, 14322-14329

7.

GSK3beta-Mediated Regulation of the

Coxsackie-Adenovirus Receptor In uences

Adenovirus Infection

Priyanka Sharma,1 _{Kathleen Frondorf,}1 _{Katherine J. D. A.} Excoffon.1

1_{Biological Sciences, Wright State University, Dayton, OH.} Viruses strike a  ne balance as both pathogens and important medical resources for genetic and immunological interventions. Adenovirus is both an emerging pathogen and a viral vector currently in clinical trials. One major limitation for adenovirus as a vector is that the coxsackievirus and adenovirus receptor (CAR) is primarily hidden on the basolateral surface of polarized epithelia. We have discovered an alternative CAR isoform (CAR-Ex8) that appears on the apical surface of human airway epithelia. This  nding provides the opportunity to directly alter viral susceptibility by either increasing or decreasing CAR-Ex8 expression. Glycogen synthase kinase 3 beta (GSK3β) is a central target of several signaling pathways, including the Wnt, Akt,

and MAPK pathways, and a gate-keeper to transcriptional changes that dictate numerous cellular pathways important for cell differentiation, proliferation, apoptosis, and development. Sequence analysis of CAR-Ex8 has revealed a series of potential GSK3β consensus sites. Additionally, the promoter for the CAR gene (CXADR) is known to be regulated by transcription factors that are GSK3β regulated. Thus, we hypothesize that CAR-Ex8 expression and localization are regulated by GSK3β. Treatment of Madin-Darby canine kidney cells (MDCK) with GSK3β inhibitors increased the expression of CAR-Ex8 speci c mRNA, as measured by quantitative RT-PCR, protein levels, measured by Western blot, and altered the localization of CAR-Ex8, as observed by immunohistochemistry in polarized cells. Treatment with GSK3β inhibitors also increased adenovirus infection of MDCK cells, suggesting that the levels of CAR are critical for adenovirus susceptibility. Therapeutic interventions that transiently augment or decrease CAR expression to enhance adenoviral-mediated gene therapy or protect susceptible populations from viral infections, respectively, will enhance our ability to treat or prevent a variety of diseases and may change our view of viral infections. Moreover, with GSK3β modulating molecules in clinical trials for conditions such as Alzheimer’’s disease and diabetes mellitus, it is important to consider how these molecules may affect viral infection and the health of patients they are meant to treat.

8.

Mining the Adeno-““Virome”” for Effective

Oncolytics Against Solid Tumors and Leukemias

1_{Infectious Disease, Mayo Clinic, Rochester, MN;} 2_{Organic Acid} Research, National Human Genome Research Institute, Bethesda, MD; 3_{Virology and Gene Therapy, Mayo Clinic, Rochester, MN;} 4_{Infectious Disease, Molecular Medicine, Mayo Clinic, Rochester,} MN.

Adenovirus serotype 5 (Ad5) has demonstrated oncolytic potential in a number of human tumor xenografts. Unfortunately, 25-100% of people from various world populations have pre-existing immunity to Ad5 which would compromise the ef cacy of Ad5 in the clinic. To look for a less seroprevalent alternative to Ad5, 12 other lower seroprevalence human adenoviruses from species B, C, D, and E were tested for ef cacy against an array of solid tumor cell lines, B-cell cancer cell lines, and primary cells from cancer patients. These viruses were also tested in vivo in both immunode cient mice and immunocompetent Syrian hamster tumor models. Against solid tumors, species C Ad6 and species B Ad11 performed similarly to Ad5. Because Ad6 and Ad11 are less seroprevalent than Ad5, these data suggest they may be more viable options against solid tumors. Species D adenoviruses consistently killed B-cell cancers more effectively than other Ad species while sparing non-malignant patient cells. When tested in vivo in mouse xenografts of human lymphoma RL, species D viruses also delayed tumor growth after single intratumoral injection. These data suggest species D adenoviruses may hold promise for clinical translation against B cell cancers.

DNA Vectorology

9.

Chromatin Structure Effects on Integration

of Gammaretroviruses MLV and XMRV

Shoshannah L. Roth,1 _{Nirav Malani,}1 _{Troy Brady,}1 _{Charles Berry,}2 Frederic Bushman.1

1_{Microbiology, University of Pennsylvania, Philadelphia, PA;} 2_{Department of Family/Preventive Medicine, University of} California, San Diego School of Medicine, San Diego, CA.

The use of gammaretroviruses as vectors in gene therapy trials has led to adverse events, increasing interest in the mechanism

DNA V

behind gammaretrovirus integration site selection. Early studies suggested that open chromatin might be favorable for integration by MLV, and recent deep sequencing experiments show that MLV favors integration near DNaseI hypersensitive sites in chromatin. For HIV, previous reports indicate that outward-facing major grooves of DNA in nucleosomes are a favored site of HIV integration in primary human T cells, suggesting that integration in vivo is taking place on nucleosome-wrapped target DNA. In this study we assessed nucleosome wrapping of target DNA during gammaretroviral integration. We infected primary human CD4+ _{T lymphocytes with} two gammaretroviruses, an MLV-based retroviral vector or xenotropic murine leukemia virus-related virus (XMRV). Human T-cells were chosen for study because particularly deep annotation is available for epigenetic modi cations and chromatin-bound proteins. We isolated integration sites with ligation-mediated PCR and analyzed the products with 454 pyrosequencing, yielding 47,758 unique integration sites. We correlated gammaretrovirus integration sites to mapped epigenetic features including histone modi cations and chromatin bound proteins, and employed nucleosome prediction software to predict nucleosome positions surrounding the recovered gammaretrovirus integration sites. This analysis showed that gammaretroviruses integrate in a de ned pattern on outward-facing major grooves on nucleosomes, similar to the integration pattern of HIV. This result is surprising, indicating that favored integration near DNaseI hypersensitive sites does not restrict integration to nucleosome free regions. Further, the pattern of favored integration near sites of bound proteins and histone modi cation implies that gammaretrovirus integration may be guided by tethering interactions with chromatin components.

10.

Elimination of Gene Targeted Chromosome

in Trisomic Cells

Li Li,1 _{Kai-Hsin Chang,}1 _{Roli Hirata,}1 _{Gaoying Ren,}1 _Andy Herman,1 _{Thalia Papayannopoulou,}1 _{David Russell.}1

1_{Hematology, University of Washington, Medical Center, Seattle,} WA.

Trisomy, three copies of particular chromosomes, is caused by non-disjunction during cell division and causes genetic abnormalities, such as Down syndrome (trisomy 21). In cell culture, trisomy 12 and 17 exist among pluripotent stem cells and compromise the use of these stem cells in research and therapy. Yet, no method has been described to remove the extra chromosome from trisomic cells. We hypothesize that a spontaneous mitotic non-disjunction event that happens to a trisomic 21 chromosome during cell division will produce a tetrasomic and a disomic daughter cell. The disomic cells could be selected via negative selection, if one of the trisomic chromosomes is  rst engineered to contain a negative selection marker. Down syndrome  broblasts were transduced with lentiviral vectors carrying NANOG, SOX2, OCT4 and LIN28 and several independent iPS clones were generated. Karyotyping showed that 47, XX, +21 lines were established. A positive-negative selection marker was then introduced into one copy of chromosome 21 at the amyloid precursor protein (APP) locus in trisomic 21 iPS cells by AAV-mediated gene targeting, a homologous recombination process that we have shown works ef ciently in human pluripotent stem cells. The selection marker, a fusion of the herpesvirus thymidine kinase (TK) and neomycin resistance genes allowed us to use G418 resistance for positive selection and GCV (ganciclovir) resistance for negative selection. The AAV vector was a promoter trap design with an IRES element and the selection marker  anked by 5’’ and 3’’ homology arms respectively.

Three different trisomic 21 iPS clones were infected with AAV to produce 0.11% to 0.24% G418-resistant cells, 80% of which were targeted at APP based on Southern blots. These targeted cells were then cultured in GCV to select for loss of the TK gene. Approximately 1-4[/sup]x10-4 _{of total cells became GCV-resistant,} and 50% to 90% of these clones had lost the targeted chromosome 21. Cytogenetic analysis con rmed that these cells had reverted to a normal karyotype. A comparison of the hematopoietic differentiation of these iPSCs showed that the trisomic 21 iPSCs had the expected increased hematopoietic potential, which was lost upon conversion to disomy.

Table. Disomic 21 cells arising from trisomic 21 iPSCs Targeted Tri21

iPSC Total CFU platedGCV

R

Foci Frequency of GCVR _colonies Di21/GCVR colonies

C1-1 158,400 13 8x10-5 _50%

C2-1 15,000 5 3.3x10-4 _60%

C3-1 48,000 21 4.4x10-4 _90%

These corrected iPS cells may prove useful for treating Down syndrome-related leukemias by autologous hematopoietic cell transplantation with the hematopoietic progeny of disomic iPSCs. Moreover, our method could be employed to repair other types of chromosome trisomy as well, and we will present the results of ongoing experiments to remove trisomy 17 from pluripotent stem cells.

11.

Towards Eliciting the Mechanism of

Plasmid DNA Silencing In Vivo

Jiamiao Lu,1 _{Mark Kay.}1

1_{Department of Pediatrics and Genetics, Stanford University,} Stanford, CA.

A major disadvantage that holds up the usage of non-viral plasmid vectors for human gene therapy is the short duration of therapeutic transgene expression. Standard plasmid DNA transfection into mouse liver results in an initial high level of transgene expression but is slowly silenced beginning several days later. Our lab has demonstrated that minicircle DNA vectors devoid of the plasmid DNA backbone (BB) persistently expresses the therapeutic transgene at 10 to 1000 times that of the same expression cassette in routine plasmid in

vivo even though the amount of transfected plasmid or minicircle

DNA remaining in the liver is the same (Cheng et al. 2003 Journal). Furthermore our studies have shown that covalent attachment of the plasmid BB to the expression cassette is required for plasmid DNA silencing (Zhi-ying Cheng et al. 2003). While histone modi cations are associated with the persistence of minicircle-directed transgene expression (RIu et al. 2007 Mol Therapy), the enriched CpG content in bacterial plasmid DNA and CpG methylation status of the BB DNA did not in uence transgene silencing (Chen et al., Mol Therapy 2008). We hypothesized that perhaps the distance between the promoter and poly adenylation site and not the plasmid BB DNA itself was responsible for silencing. Thus, we replaced the bacterial plasmid DNA sequences with varied non-bacterial non-coding sequences and tested these new constructs for transgene expression in mice. Two minicircle expression systems were used in this study, human α1-antitrypsin (hAAT) and human factor IX (hFIX) driven by the RSV-LTR and EF1α promoter, respectively. Using random or fragmented DNA sequences of 500bp or less, these two transgenes were expressed similarly to standard minicircle DNA as they were not silenced. In contrast, with 1kb or longer spacers, the transgene

Molecular Therapy Volume 19, Supplement 1, May 2011 _S5

DNA V

expression was silenced as with the standard plasmid vector. These results suggested that spacer DNA regardless of its source >1kb caused the minicircle constructs to be silenced even in the absence of plasmid BB. To further establish this idea, 500bp of plasmid BB was used as spacer to make new constructs. These constructs also showed persistent expression suggesting the size of the non-transcribed insert and not the source of the spacer DNA was the critical factor for DNA silencing. To better understand the mechanism of plasmid silencing, we attempted to determine if leaky transcriptional termination across the spacer sequences was an important parameter for silencing. To do this, we added a robust transcriptional terminator between the poly A site and DNA spacer. Even with an additional terminator, only spacers of > 1kb silenced the transgene. Our data suggest that the distance between the poly A site and promoter region in uences DNA silencing. These results have important implications for designing DNA plasmid vectors for gene therapy.

12.

Development of a High-Throughput

Quantitative Method for Assessing Plasmid

Nuclear Import

Holly M. Reynolds,1 _{David A. Dean.}1

1_{Pediatrics, University of Rochester, Rochester, NY.}

One area of interest in gene therapy is designing strategies that can allow DNA to be delivered only to a speci c cell type or tissue. Historically, these strategies have included: application of the agent directly to desired tissue, exploitation of receptor-ligand interactions, and using cell-speci c promoters to drive transcription. Our lab has developed a novel and robust method for cell-speci c delivery that allows for nuclear import in non-dividing cells by a sequence speci c process that requires the cytoplasmic binding of transcription factors, these factors provide transport across the nuclear membrane. Namely, this exploits the fact that promoter sequences bind transciption factors and in some cases, when the Nuclear Localization Signal is oriented in the correct way, allow for taxiing across the nuclear membrane. To date, we have discovered DNA sequences that provide cell-speci c import in smooth muscle cells, osteoblasts, endothelial cells and alveolar type 2 epithelial cells. Identifying such plasmid sequences, however, is a tedious and time-consuming task, requiring cytoplasmic microinjection of individual plasmids containing potential sequences into hundreds to thousands of individual cells and determining nuclear localization of the plasmids by in situ hybridization and/ or  uorescence microscopy. We have developed a simpler, higher-throughput assay that allows us to screen a large number of potential sequences for cell-speci c nuclear entry in a large panel of cell types. Using a 96-well electroporator with optimized  eld strengths, we can induce cell uptake of pools of  uorescently-labeled plasmids containing various DNA sequences. After incubation, these cells are  xed and run, in a 96-well format, on an AMNIS Imagestream with a nuclear stain. This instrument is essentially a  ow cytometer that images each cell in up to 8  uorescent channels with a 40X or 60X objective as it moves through the stream. The instrument is capable of collecting data from up to 5000 cells per minute and images can then be analyzed with software to assess and quantify subcellular localization of the plasmids. Using this approach, we have shown that plasmids carrying the ubiquitously active SV40 DNA nuclear targeting sequence localize to the nuclei of multiple cell types while isogenic plasmids lacking this sequence remain in the cytoplasm unless incubation times are greatly increased to allow for cell division (and concomitant breakdown of the nuclear envelope). This approach should greatly enhance our ability to screen DNA sequence libraries in a wide variety of cell types to identify nuclear localizing DNA sequences for cell-speci c gene therapy.

13.

Robust Long-Term Gene Expression

in Mouse Liver Via the Novel hAT Transposon

TcBuster

Lauren E. Woodard,1 _{Aparna Kaja,}1 _{Robert H. Hice,}2 _{Peter W.} Atkinson,2 _{Nancy L. Craig,}3 _{Matthew H. Wilson.}1

1_{Department of Medicine, Division of Nephrology, Baylor College} of Medicine, Houston, TX; 2_{Department of Entomology &} Institute for Integrative Genome Biology, University of California, Riverside, Riverside, CA; 3_{Howard Hughes Medical Institute,} Department of Molecular Biology and Genetics, Johns Hopkins School of Medicine, Baltimore, MD.

TcBuster is a recently discovered mobile DNA element from

the hAT superfamily. Colony count assays in HEK 293 cells using varying amounts of transposon and transposase indicate that maximal transposition occurred using 500 ng of pTCBNeo and 100 ng of pCMV-TCB. Out of 48 TcBuster integration sites sequenced from these cells, 62.5% were in genes and 4.2% were in exons, suggesting a preference for transcriptionally active areas. Comparing different transposons by colony count assay in HEK 293 cells, the transposition of TcBuster was approximately equivalent to hAT family member Tol2 and about half that of piggyBac, depending on the ratio of transposon to transposase plasmids. When an equal amount of transposase and transposon plasmids were used, all three systems were approximately equal. However, when an excess of transposase (200 ng of transposon and 800 ng of transposase plasmids) or an excess of transposon (900 ng of transposon and 100 ng of transposase plasmids) were provided,

piggyBac outperformed both hAT systems, indicating the signi cant

 exibility of the piggyBac system. As for the hAT systems, when the amount of transposase plasmid was higher, the Tol2 system was superior to TcBuster, whereas when the amount of transposase plasmid was low, the TcBuster system was superior to Tol2. After transfection with a construct carrying an HA-tagged version of the TcBuster transposase, staining revealed localization to unique intranuclear rodlets. We hypothesize that the rodlets may serve a protective role, limiting free transposase and thus DNA damage. In support of this hypothesis, as the amount of pCMV-HA-TCB is increased beyond 100 ng, the number of rodlets per nuclei increased exponentially as the number of colonies formed decreased. To compare TcBuster with

piggyBac for integration into mammalian chromosomes in vivo, adult

female FVB mice were given hydrodynamic injections containing luciferase transposon plasmids with or without the plasmid encoding the transposase. Mice were imaged using the IVIS system to determine the levels of liver-speci c luciferase expression at timepoints out to 24 weeks. At the later timepoints, the normalized luciferase values for groups given either TcBuster or piggyBac transposase were both statistically signi cantly higher than groups given transposon alone (p<0.0005 and p<0.01, respectively, by two-tailed t-test). An equivalent t-test comparingTcBuster with piggyBac revealed that they were not signi cantly different from one another. Therefore, we conclude that TcBuster is an exciting new transposon that is highly capable of providing long-term gene expression in vivo.

14.

Immunosuppression with

Cyclophosphamide Signi cantly Prolongs

High-Level Expression of IDUA Mediated by the

Sleeping Beauty Transposon System

Elena L. Aronovich,1 _{Jason B. Bell,}1 _{Kelly M. Podetz-Pedersen,}1 Brenda L. Koniar,2 _{R. Scott McIvor,}1 _{Perry B. Hackett.}1

1_{Genetics, Cell Biology and Development, University of Minnesota,} Minneapolis, MN; 2_{Research Animal Resources, University of} Minnesota, Minneapolis, MN.

Mucopolysaccharidosis type I (MPS I) is a systemic disease caused by de ciency of the lysosomal enzyme α-L-iduronidase (IDUA). The goal of gene therapy for MPS I is to achieve sustained expression

DNA V

of an IDUA transgene. We have reported the effectiveness of the

Sleeping Beauty (SB) transposon system for gene therapy of MPS

I in adult immunode cient mice. However, in immunocompetent mice IDUA expression is curtailed mainly by immune responses to the transgenic IDUA polypeptide. Although more than 99% of transgene expression comes from the liver following hydrodynamic delivery, restoration of de cient enzyme activity in other organs can be achieved through metabolic cross-correction whereby IDUA enters the circulation and is distributed to other tissues. The IDUA gene must therefore be regulated by a strong promoter that will drive high-level expression, i.e. ≥ 100-fold the wild-type (WT) level. We have previously shown that a strong liver-speci c promoter ApoEhAAT has several advantages over using the strong ubiquitously expressed CAGGS promoter to drive IDUA including: 1) increased IDUA expression level at 5-6 weeks post-infusion and 2) absence of a gender bias that we characterized using the CAGGS promoter for expression in MPS I mice. We hypothesized that a combination of immunosuppression and liver-speci c expression would increase the levels of IDUA maintained in immunocompetent MPS I mice. We hydrodynamically injected 5 µg pT2/ApoEhAAT-IDUA SB transposons into immunocompetent C57BL/6 mice with or without pCMV-SB100 transposase (+ or -SB). Preliminary comparison of transposition ef ciency by real-time qPCR excision assay showed that SB100 was at least 30 times more effective than SB11 in mouse liver. For long-term studies, mice (n=10 mice /cohort) were treated with or without cyclophosphamide (50 mg/kg i.p. on -1 day, -5 h, +1 day, and +2 days with respect to hydrodynamic injection). IDUA activity in plasma was 200-300-fold greater than WT in all mice 1-day post-injection (p.i.). Without cyclophosphamide, plasma IDUA declined slightly over the  rst two weeks but thereafter deteriorated in most mice (+ and -SB) to background levels by six weeks p.i. In a few mice that were not immunosuppressed, IDUA activity remained consistent for 6 - 12 weeks before declining to background over a period of 2 weeks. In contrast, IDUA expression levels were maintained in all except 5 of the 19 cyclophosphamide-treated mice and at 12 weeks p.i. IDUA expression was 217±159-fold WT (+SB) and 75±60 (-SB). Thus, the combination of transient immunosuppression and a liver-speci c promoter resulted in sustained IDUA expression at levels that, with SB100, were therapeutically relevant for treating MPS I.

15.

Evaluating the Safety of PiggyBac

Mediated Gene Transfer in Human Cells

Sunandan Saha,1,2 _{Aparna Kaja,}1 _{Cliona M. Rooney,}2,3 _{Matthew H.} Wilson.1,2

1_{Department of Medicine- Nephrology Divison, Baylor College of} Medicine, Houston, TX; 2_{Program in Translational Biology and} Molecular Medicine, Baylor College of Medicine, Houston, TX; 3_{Department of Pediatrics, Baylor College of Medicine, Houston,} TX.

PiggyBac has been successfully used for modi cation of primary human cells and cell lines with transgene(s) of interest and holds promise as an effective non-viral gene delivery method. Currently, limited information exists about the safety of the piggyBac system for the modi cation and generation of clinical grade human cells. In this study we began to evaluate the safety of piggyBac for modi cation of human cells. PiggyBac works through a ““cut and paste”” mechanism thereby delivering a transgene of interest  anked by inverted repeat sequences into the genome. We found that piggyBac leads to stable transgene integration and transposase expression is undetectable after 7 days in a mixed population of human cells. Although there are no sequences in the human genome with complete similarity to the 17bp terminal repeat sequence (TRS) of the piggyBac transposon inverted repeats (IR), there are sequences with 16, 15 or 14 bp similarities all of which end in the tetranucleotide TTAA required for transposition. We subsequently tested the ability of piggyBac to mobilize these

genomic elements. The potential of these sequences mimicking the TRS was assessed by replacing the TRS of the 5’’Inverted repeat with these genomic sequences and looking at transposition ef ciency using a colony count assay. None of the 14 tested sequences were able to effectively act as a terminal repeat sequence. Nor did they mediate excision of transposons in presence of the transposase. To assess the safety of delivering the transposase and transposon from a single vector, we isolated clones derived from transfections using the transposase and the transposon cassettes in cis (on the same plasmid vector) and found that all clones had residual transposase expression. In contrast, stable clones generated with transposase delivered in trans from a separate non-integrating plasmid showed no residual transposase expression. Studies are ongoing to further evaluate the potential genotoxicity of piggyBac in human cells. Thus, our data suggests that piggyBac transposase expression is short lived (7 days) in a mixed population of human cells, the piggyBac transposase appears incapable of mobilizing TRS-like-sequences in the human genome, and delivering transposase in trans should be safer than delivery in cis when modifying human cells. piggyBac appears to be a promising non-viral gene delivery system for therapuetic genetic modi cation of human cells.

16.

Development of a Sleeping Beauty

Transposon-Based Integration System for Gene

Transfer in Human Epithelial Cells

Giandomenico Turchiano,1 _{Maria Carmela Latella,}1 _Zsuzsanna Izsvak,2 _{Zoltan Ivics,}2 _{Fulvio Mavilio,}1 _{Alessandra Recchia.}1 1_{Center for Regenerative Medicine, University of Modena} and Reggio Emilia, Modena, Italy; 2_{Max Delbruck Center for} Molecular Medicine, Berlin, Germany.

Transplantation of autologous, genetically corrected epidermal stem cells (EpSC) was successfully used to treat junctional epidermolysis bullosa (EB), a genetic skin adhesion disorder. The dystrophic forms of EB is caused by mutations in the type-VII collagen gene (COL7A1) Delivering the >9 kb COL7A1 cDNA by a retroviral vector is problematic, due to the large size and highly repeated nature of its sequence, which induce genetic rearrangements during reverse transcription and integration. We tested the feasibility of using a non-viral vector system based on Sleeping Beauty (SB)-derived transposons, taking advantages of the recently developed, high-capacity ““sandwich”” version of the SB transposon and the ““hyperactive”” SB 100X transposase, which showed high transposition ef ciency in human stem cells. We tested the system in HeLa cells and in a keratinocyte cell line (HaCaT), which were co-transfected with the SB 100X transposase and either the normal or the sandwich version of the SB transposon containing a small-size reporter gene (Venus or GFP) expression cassette. In both cell lines, transposition was obtained in up to 80% of the transfected cells with the sandwich transposon, compared to ∼50% obtained with the older version. Transposed HaCaT cells were cloned and analysed for integration events. Individual clones carried one to 14 copies of the integrated transposon of the predicted size. High-throughput sequencing is under way to analyze the sandwich SB transposon integration characteristics. We then tested a transposon carrying an 8.8-kb cassette, which again showed up to 80% transposition ef ciency in transfected cells. Finally, we generated sandwich transposons containing an expression cassette for the COL7A1 cDNA under the control of a constitutive (PGK) or a keratinocyte-speci c (K14) promoter, which are currently being tested for integration in HaCaT cells. The SB-based gene delivery system will  nally be tested in human primary keratinocyte cultures.

Molecular Therapy Volume 19, Supplement 1, May 2011 _S7

I

D

V

Infectious Diseases and Vaccines

17.

Induction of T Cell Dysfunction by

AAV-Mediated Hepatitis B Virus Infection

Mi-Hua Tao,1,2 _{Ya-Hui Huang,}1,2 _{Ho-Yuan Chou,}1,3 _Cheng-Chieh Fang.1

1_{Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan;} 2_{Graduate Institute of Life Sciences, National Defense Medical} Center, Taipei, Taiwan; 3_{Graduate Institute of Microbiology,} National Taiwan University, Taipei, Taiwan.

The complex virus-host interactions responsible for dysfunctions of hepatitis B virus (HBV)-speci c T cells in the early phase and the subsequent development of severe liver diseases several decades later are poorly understood and have been dif cult to study due to the absence of appropriate small animal models that are susceptible to infection. Mice are not naturally infected by HBV, presumably due to a lack of HBV receptors on mouse hepatocytes. To bypass this entry step of HBV infection, we used adeno-associated viral vectors to introduce the HBV genome (AAV/HBV) into the liver of immunocompetent mice. We con rmed the presence of HBV virions and proteins in the serum and liver of all AAV/HBV-transduced mice in all four mouse strains tested. The AAV/HBV-transduced mice were negative for HBV surface (HBs) and e protein (HBe)-speci c antibodies, but positive for HBV core (HBc)-speci c antibody, a pro le similar to that observed in chronic HBV patients. AAV/HBV transduction induced HBs- but not HBc-speci c IFN-g-producing CD8+ _{T cells, preferentially enriched in the liver. The number of} HBs-reacting CD8+ _{T cells decreased with time and their effector functions} were severely impaired as suggested by their failure to respond to HBs immunization. Interestingly, PD1 expression on hepatic CD8+ _{T cells increased gradually after AAV/HBV transduction,} suggesting an important role of PD-1 in constraining T cell activity and maintaining HBV chronicity. Nevertheless, six months after transduction these AAV/HBV-transduced mice manifested evidence of hepatic in ammation and regeneration. Conclusion: Our AAV/ HBV-transduced murine model is a unique instrument for studying the immunological events in human chronic HBV and induction of PD1 expression is correlated with dysfunction of hepatic CD8+ _{T cells.}

18.

Desmoglein 2 Is a Receptor for Adenovirus

Serotypes 3, 7, 11, and 14: Implications for

Pathogenesis of Adenovirus Infections and Cancer

Treatment

Hongjie Wang,1 _{ZongYi Li,}1 _{Jonas Persson,}1 _{Ines Beyer,}1 _Robert Strauss,1 _{Akseli Hemminki,}2 _{Pascal Fender,}3 _{Andre Lieber.}1 1_{Division of Medical Genetics, University of Washington, Seattle,} WA; 2_{Cancer Gene Therapy Group, University of Helsinki} & Helsinki University Central Hospital, Heskinki, Finland; 3_{Unit of} Virus Host Cell Interaction, EMBL, Grenoble, France.

We have identi ed desmoglein 2 (DSG2) - a calcium-binding transmembrane glycoprotein and component of the epithelial cell-cell adhesion structure -as the primary high-af nity receptor used by adenovirus (Ad) serotypes Ad3, Ad7, Ad11, and Ad14. These serotypes represent important human pathogens causing respiratory tract infections. Loss- and gain-of-function studies con rmed DSG-2 to be crucial for the binding, infection and spread of these serotypes. While Ad binding CAR and CD46 involves only a high af nity interaction between the  ber knob domain and the cellular receptor, binding of Ad3 to DSG2 requires multimerization of the  ber knob domain. We therefore designed two recombinant Ad3  ber knob domains containing heterodimerization domains. Both proteins can be can be produced in E.coli and puri ed by af nity chromatography. Upon mixing they dimerize (Ad3K-K dimer). Ad3K-K ef ciently and speci cally blocked Ad3 infection. We also studied the consequences

of adenoviral-DSG-2 interaction, following exposure to Ad3 virions or Ad3K-K, using epithelial cancer cell lines. Pathways involved in epithelial-to-mesenchymal transition were found to be activated, leading to transient intercellular junction opening. This  nding has implications for Ad spread in epithelial tissues and for cancer therapy as intercellular junctions represent physical obstacles for access and intratumoral dissemination of anti-cancer therapeutics. We demonstrated that pre-incubation of cancer cell lines with Ad3K-K increased the cytotoxicity of trastuzumab and cetuximab - monoclonal antibodies which bind to the Her2/neu receptor and EGFR respectively. In vivo, in mice carrying Her2/neu-positive breast cancer and EGFR1-positve lunger cancer xenografts, pre-injection of Ad3K-K signi cantly increased the therapeutic ef cacy of trastuzumab and cetuximab. Ad3-DSG2 interactions potentially have consequences beyond opening epithelial junctions as we also detected Ad3 binding to DSG2 on platelets and granulocytes. We are currently performing a systemic biodistribution study of DSG2 expression in humans and non-human primates and data will be presented. Together, these studies will likely contribute to the design of novel antiviral therapies and may also have important implications for cancer therapy.

19.

Increased Mucosal CD4

_{T Cell Activation}

Following Vaccination with a Species-Speci c

Adenovirus Vector in Rhesus Macaques

Carnathan,1 _{Rebecca Grant,}2 _{Sarah J. Rutcliffe,}3 _{James M. Wilson,}2 Michael R. Betts.1

1_{Department of Microbiology, University of Pennsylvania,} Philadelphia; 2_{Department of Pathology & Laboratory Medicine,} University of Pennsylvania, Philadelphia; 3_{Department of} Biostatistics, University of Pennsylvania, Philadelphia.

The possibility that vaccination with Adenoviral vectors increased mucosal T-cell activation remains a central hypothesis to explain the potential enhancement of HIV acquisition within the STEP trial. Modeling this within rhesus macaques is complicated because human Adenoviruses, including Adenovirus type 5 (HAd5), do not naturally infect macaques. Therefore, we created a vector based upon a naturally occurring rhesus macaque Adenovirus (SAdV7) to test whether vaccination with a species-speci c Adenoviral vector enhances mucosal T-cell activation within the natural host. Twelve rhesus macaques were vaccinated 3x intramuscularly with the SAdV7 vector, and PBMC and rectal biopsies were obtained at baseline, multiple times post-prime and post-17 week boost (8x/animal), and post-31 week second boost (1x/animal). Five HAd5-vaccinated animals were included as controls. We assessed rectal mucosal lamina propria and blood for frequency changes of Ad-speci c T-cell responses and T-cell activation levels by measuring IFNγ, TNFα, IL2, CD25, Ki67, CD69, and HLA-DR. Naturally acquired pre-existing SAdV7-speci c CD4+ _{T-cells were detectable in 9/12 macaques within blood} (0.05-0.256%) and/or rectal mucosa (0.3-1.11%). Following SAdV7 vaccination, rectal SAdV7-speci c CD4+ _{T-cell responses increased} above baseline in 9/12 animals (0.3-8.6%) 2-5 weeks post-prime, reaching as high as 10.75% following homologous SAdV7 boost (11/12 animals, 0.16-10.75%). SAdV7-speci c CD4+ T-cells in the rectal mucosa exhibited increased activation levels relative to those found in blood. Importantly, the SAdV7 vaccination resulted in a dramatic perturbation of total memory CD4+ _{T cells, characterized by} heightened expression of the CD25, CD69, and HLA-DR activation markers (up to 80% of memory CD4 cells) and a proportional decrease in naïve CD4+ _{T cells, which peaked at 16 weeks after the prime.} No such effect occurred in HAd5-vaccinated animals. These results indicate that peripheral vaccination with a species-speci c Adenovirus vector can increase the activation state of mucosal CD4+ _T-cells providing a potential mechanism for the increased transmission of HIV following HAd-5 vaccination in the Merck STEP trial.

I

D

V

20.

Most Closely HLA-Matched Allogeneic

Virus Speci c Cytotoxic T-Lymphocytes (CTL)

To Treat Persistent Reactivation or Infection with

Adenovirus, CMV and EBV after Hemopoietic Stem

Cell Transplantation (HSCT)

Ann M. Leen,1 _{Catherine M. Bollard,}1 _{Adam M. Mendizabal,}2 Elizabeth J. Shpall,3 _{Paul Szabolcs,}4 _{Joseph Antin,}5 _{Neena Kapoor,}6 Sung-Yun Pai,7 _{Bambi Grilley,}1 _{Adrian P. Gee,}1 _{Malcolm K.} Brenner,1 _{Cliona M. Rooney,}1 _{Helen E. Heslop.}1

1_{Baylor College of Medicine, Texas Childrens Hospital, The} Methodist Hospital, Houston; 2_{The EMMES Corporation,} Rockville; 3_{MD Anderson Cancer Center, Houston;} 4_Duke University Medical Center, Durham; 5_{Dana Faber Cancer} Institute, Boston, MA; 6_{Children’’s Hospital of Los Angeles, Los} Angeles; 7_{Dana Faber Cancer Institute, Children’’s Hospital} Boston, Boston, MA.

Adoptive transfer of CTLs can reconstitute antiviral immunity to EBV, CMV and adenovirus in allogeneic hemopoietic stem cell transplant (HSCT) recipients. However, the time taken to prepare patient-speci c products and the lack of virus-speci c T-cells in cord blood and seronegative donors restricts application. As part of the NHLBI Specialized Centers for Cell-Based Therapy, we are evaluating in a multicenter setting whether infusion of ““off the shelf”” closely HLA-matched allogeneic CTLs (CHM-CTLs) will overcome this limitation and prove feasible, safe and effective in HSCT recipients with infection refractory to standard therapy. Using the NHLBI Production Assistance for Cellular Therapies (PACT) program, we manufactured and tested >30 multivirus lines, which were polyclonal, comprising CD4+ (median 10%) and CD8+ (median 83%) T-cells with speci city for CMV, EBV and adenovirus (mean 1463, 101, and 227 SFC/1x105 _{cells, respectively). A dose of} 2x107_CHM-CTL/m2 _{was available to subjects with persistent infection} matching at least one HLA antigen. We selected the most suitable line based on target virus speci city through a shared allele and on overall HLA match. To date 61 patients have been screened. A potential line was identi ed for 56, of whom 35 received 1-3 infusions with lines matched at 1-3 HLA antigens. Seven subjects received CHM-CTLs for refractory EBV-PTLD, 15 for persistent CMV, 12 for persistent adenovirus and one for both CMV and adenovirus. There were no immediate adverse effects, and the GVHD incidence was low, with two patients reactivating previous acute skin or chronic GVHD, and two developing transient skin rashes. Two patients developed thrombotic microangiopathy but had other risk factors including sirolimus treatment. Response, de ned as the overall cumulative incidence of  rst CR/PR in the  rst 30 patients by day 42 is 85.7% overall; (92% for CMV, 80% for EBV and 80% for adenovirus). Of note, we obtained complete antiviral responses in 16 patients overall, and even in recipients of lines matching at a single HLA antigen. Despite this objective antiviral activity, immune monitoring studies using IFN-g ELIspot assays showed little or no rise in peripheral blood virus-reactive T cells after infusion. Nevertheless these preliminary results show the feasibility and safety of immunotherapy using banked CHM-CTLs. The antiviral responses we observe are encouraging and suggest that an ““off-the-shelf”” CTL cell line may be a practical strategy for allogeneic stem cell transplant recipients.

21.

Modeling Genetic Protection Against

AIDS: Viral Restriction Factor Transgenesis in the

Domestic Cat

Pimprapar Wongsrikeao,1 _{Dyana T. Saenz,}1 _{Tommy Rinkoski,}1 Takeshige Otoi,2 _{Eric M. Poeschla.}1,3

1_{Department of Molecular Medicine, Mayo Clinic College of} Medicine, Rochester, MN; 2_{Department of Veterinary Medicine,} Yamaguchi University, Yamaguchi, Japan; 3_{Division of Infectious} Diseases, Mayo Clinic College of Medicine, Rochester, MN.

The genome of the domestic cat was recently sequenced at 1.9X coverage and a 10X assembly is imminent. This neurobehaviorally complex, accessible species is valuable in research situations where rodents are unsuited. For example, humans and cats each suffer from pandemic AIDS-causing lentiviruses and the feline eye and nervous system are important model systems. A capability for feline transgenesis is needed to realize the inherent potential. Here we generated multi-gene transgenic domestic cats without nuclear transfer, thus establishing transgenesis by gamete genetic modi cation for the  rst time in a carnivore. The process was highly ef cient, with uniformly transgenic outcomes, 5-15 insertions per animal, widespread organ-pervasive gene expression, germline transmission and no mosaicism. Lymphocytes from cats transgenic for rhesus macaque TRIMCyp, a retroviral restriction factor, had decreased susceptibility to feline immunode ciency virus replication. Feline transgenesis can now be used to investigate fundamentals of lentiviral restriction factor biology at the whole animal level, gauge the potential of these factors in gene therapy for HIV-1/AIDS and create animal models for other diseases, and potentially to protect extremely endangered feline species.

22.

Induction of Strong Cellular and Humoral

Immune Responses Following Immunotherapy of

Post-LEEP CIN 2/3 with HPV 16

& 18 E6/E7 DNA

Vaccines

Jian Yan,1 _{Xuefei Shen,}1 _{Mary Giffear,}1 _{Jessica Lee,}1 _{Amir S.} Khan,1 _{Dawn Harris,}2 _{Panyupa Pankhong,}2 _{Devon Shedlock,}2 _Jean Boyer,2 _{David B. Weiner,}2 _{Mark Bagarazzi,}1 _{Niranjan Y. Sardesai.}1 1_{Inovio Pharmaceuticals, Inc., Blue Bell, PA;} 2_{Pathology and} Laboratory Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA.

Background: Cervical cancer due to human papillomavirus (HPV) infection is responsible for 270,000 deaths every year necessitating the development of new therapies. DNA vaccines have potential advantages for cancer immunotherapy although human immunogenicity has disappointed. Codon/RNA optimization, high ef ciency IgE leader sequences, and electroporation (EP) were all employed to enhance the potency of the vaccines in this study. Methods: VGX-3100 is a combination of 2 plasmids, pCon16E6E7 & pCon18E6E7 that encode HPV16 & 18 consensus E6/E7 fusion genes, respectively. HPV-001 is a phase I study evaluating 3 IM doses (0.6, 2, 6 mg) of VGX-3100 using CELLECTRA® _{EP device. Six post-LEEP} CIN2/3 patients were immunized at weeks 0, 4 & 12 in each cohort. Immune responses were measured by ELISA with epitope mapping, Western blot, T-cell ELISpot, and  ow cytometry. Results: VGX-3100 and the vaccination procedure were safe and generally well tolerated. All subjects completed the three dose vaccination regimen and no SAEs or vaccine-related Grade 3 or 4 AEs have been reported. Cellular responses: Overall, in all three doses combined, 13 out of 18 vaccinated subjects (72%) developed signi cant CTL responses, with positive responses ranging from under 100 to over 5000 SFU per million cells. Signi cantly, CTL responses were observed to all four vaccine antigens. In the high dose cohort,  ve of six vaccinated subjects (83%) developed signi cant CTL responses, with average responses of 1362 SFU per million cells after three immunizations.

Molecular Therapy Volume 19, Supplement 1, May 2011 _S9

I

D

V

This was a 118% increase compared to the intermediate dose cohort average of 626 SFU per million cells (four responders out of six) and a 174% increase compared to the low dose cohort average of 497 SFU per million cells (four responders out of six) indicating a vaccine dose effect. Humoral responses: Fifteen out of 18 vaccinated subjects (83%) developed antibody responses to at least one antigen with most subjects developing responses to two or more antigens. Western blot con rmed reactivity to HPV16 & 18 E7. Epitope mapping shows broad responses to E6 & E7 epitopes. Conclusions: Antibodies and ELISpots are higher than previous reports from prior studies of HPV poxviral, peptide or DNA vaccines. Highly optimized HPV vaccines delivered by EP induced robust cellular and humoral responses in humans with as little as 0.3 mg/plasmid. These preliminary results support the further development of immunotherapeutic HPV DNA vaccine strategies in Phase II clinical studies to assess impact on clinical ef cacy.

23.

Enhanced Immunogenicity of Measles

Virus Constructs Encoding Helicobacter pylori

Neutrophil-Activating Protein as Vaccine Platforms

1_{Mayo Clinic College of Medicine, Rochester, MN.}

Background: Helicobacter pylori neutrophil-activating protein

(NAP) is a crucial virulence factor and promising candidate for vaccine development. NAP is TLR2 agonist and potent immunomudulator inducing Th1-biased in ammatory response. Recently, we generated and evaluated the immunogenicity of a panel of attenuated measles virus (MV) strains encoding different forms of NAP antigen. Immunization experiments demonstrated that MV vector engineered to express secretory forms of the antigen induced strong and long-lasting antibody and cellular immunity against NAP. Here we present data on the characterization of the humoral immune response against NAP-expressing MV strains in MV infection permissive type I interferon receptor knockout human CD46 transgenic (Ifnarko-CD46Ge) mice. Results: MV-lambda-NAP and MV-s-MV-lambda-NAP encode secretory MV-lambda-NAP antigen replacing the variable domain of human lambda immunoglobulin chain with (MV-lambda-NAP) or without (MV-s-NAP) the constant lambda domain. We developed a monoclonal antibody based immunoassay for detection of NAP expression in culture supernatants and biological  uids. Vero cells infected with both strains expressed high level (> 1 µg/ml per 106 _{infected cells) of the secretory NAP transgene. On} day 28 post-immunization of Ifnarko-CD46Ge mice, all animals developed antibody titers against NAP above 1:10,000 in ELISA. Characterization of the isotype speci city revealed that the humoral anti-NAP immunity is dominated by the IgG2a and IgG1 sub-isotypes. IgG2b titers were lower and NAP-speci c IgG3 is barely detectable. All mice developed highly protective MV neutralization immunity with an average titer above 1:4,000 in plaque reduction microneutralization test (PNT). Further, we evaluated the NAP-mediated immunostimulatory effect in MV-lambda-NAP immunized mice compared to the immune response in the animal group immunized with the MV-lambda (MV expressing full length unmodi ed human lambda immunoglobulin) control strain (7-9 animals per group). On day 18 all MV-lambda-NAP injected animals had anti-lambda ELISA titers of 1:25,600, signi cantly higher than those of the MV-lambda-immunized control mice, where only less than a half of the group developed titers above 1:1,600. These results demonstrated that ““NAP-tagged”” chimeric lambda is highly immunogenic compared to the unmodi ed protein. This strategy can be applied in vaccination against poor immunogens, such as tumor-associated antigens or poorly immunogenic protective antigens of infectious pathogens.

Conclusions: These data demonstrated the strong immunogenicity of

attenuated MV vaccine strains encoding a single H. pylori protective antigen. MV vector-expressed chimeric NAP constructs can enhance

the immune response to poor immunogens with application in vaccine development, cancer immunotherapy and virotherapy. * This work

was supported by Atwater grant, P50CA116201 and Paul Leibson Memorial Fund.

24.

In Vivo Safety and Ef cacy of a

Combination Anti-HIV Lentiviral Vector in RAG1

Mice for HIV Gene Therapy

Jon Walker,1 _{Brian Fury,}1 _{Jeannine McGee,}1 _{Gerhard Bauer,}1 _{Jan A.} Nolta,1 _{Joseph S. Anderson.}1

1_{Internal Medicine, University of California Davis, Sacramento,} CA.

HIV gene therapy offers a potential alternative to current small molecule antiretroviral treatments which, after prolonged use, can become toxic and allow the generation of escape mutants. A recent hematopoietic stem cell transplant for acute myeloid leukemia in an HIV infected patient was performed utilizing allogeneic cells from an individual homozygous for the delta-32 CCR5 deletion. HIV-1 suppression has been observed in the recipient, to date, even after this individual discontinued antiretroviral therapy. Therefore, HIV gene therapy has the potential to mimic the results of this transplant by inserting anti-HIV genes into hematopoietic progenitor cells for constitutive expression and lifelong protection from further HIV replication. For preclinical analysis of anti-HIV gene therapeutic vectors, it is necessary to utilize an appropriate in vivo animal model capable of demonstrating safety and ef cacy of the novel therapy. The RAG1-IL2r-γ knockout mouse model offers the potential to evaluate multi-lineage human hematopoiesis from intrahepatic injection of transduced human CD34+ hematopoietic progenitor cells (HPCs) into newborn mice. In the current study, preclinical evaluation of a combination anti-HIV lentiviral vector was performed, in vivo, in a humanized RAG1-IL2r-γ knockout mouse model. This combination vector contains a human/rhesus macaque TRIM5alpha isoform, a CCR5 shRNA, and a TAR decoy and has previously displayed strong pre-integration inhibition of HIV-1 infection, in vitro. Here we demonstrate multi-lineage human hematopoiesis from combination lentiviral vector transduced CD34+ HPCs in the peripheral blood and in various lymphoid organs including the thymus, spleen, and bone marrow. Anti-HIV vector transduced cells displayed a normal distribution of immune cells including T cells, B cells, and macrophages compared to control nontransduced and EGFP-alone vector transduced cells. The anti-HIV vector transduced cells also displayed knockdown of cell surface CCR5 due to the expression of the CCR5 shRNA. After in vivo challenge with either an R5-tropic BaL-1 or X4-tropic NL4-3, a selective survival advantage of anti-HIV transduced cells was observed. The total percent of anti-anti-HIV transduced cells (based on CD45+ human cells) increased from the initial engraftment of ∼20% to >50% in the peripheral blood from pre-infection to week 10 post infection. This was in comparison to control EGFP-alone vector transduced cell engrafted mice where the total percent of CD45+ human cells in the peripheral blood remained constant ∼20% from pre-infection engraftment levels to week 10 post-infection. A decrease in the levels of plasma viremia in anti-HIV transduced cell engrafted mice was also observed over the course of infection compared to control cell engrafted mice where the human cells remained unprotected. The data provided here con rm the utility of this combination anti-HIV lentiviral vector and suggests its application in a future human clinical setting.