257 characteristic of responsive CD8+ t cells to AAV 28 and ad c7 HIV gag vector vaccines

257 Characteristic of Responsive CD8+ T Cells to AAV 2/8 and Ad C7 HIV Gag Vector Vaccines antibodies to AIAT On the other hand, an AlAT specific CDS+ ''''I'''' cell response was observed This partial CTL s[.]

antibodies to AIAT On the other hand, an AlAT-specific CDS+ 'I' cell response was observed This partial CTL suppression effect therefore seems likely to be responsible for the observed reduction in AIAT expression, This strain-specific dominant effect ofAAVS to suppress COST and B cell responses is relevant for the use of AAV in therapeutic gene transfer as well as viral immune evasion The mechanisms of active suppression including the involvement of regulatory 'I' cells are discussed 255 Cross Presentation of AAV2 Capsids Activates Cytotoxic T Cells and Hinders MuscleDirected Gene Transfer but Does Not Render Hepatocytes Effective Cytolytic Targets Lili Wang,1 Joanita Figueredo; Roberto Caicedo; Jianping Lin; James M Wilson} [Gene Therapy Program Department ofPathology and Laboratory Medicine University ofPennsylvania Philadelphia, PA Liver toxicity observed in a recent clinicaltrial ofAAV2 delivered systemically to patients with hemophilia was ascribed to killing of vector transduced hepatocytes by capsid specific 'I' cells This study evaluated the biology ofT eell activation to AAV capsids in murine models COST cell epitopes were mapped to capsids from AAV2, and S A tetramer was generated to a dominant capsid epitope in Balb/C mice shared between these AAVserotypes, Administration ofAAV2vectors resulted in the activation of capsid CD8 'I' cells as evidenced by binding to tetramer and production ofcapsid induced IFN g and TNF a expression; this was not observed with AAV7or AAVSvectors.COST cells toAAV2capsids demonstratefunctional cytolyticactivity in vivo to peptide loadedtargetcells.The frequency ofcapsid 'I' cells was much higher in liver than blood or spleen The pre-existing capsid 'I' cells induced by an adenovirus expressing the AAV capsid (Ad.Cap) were able to decrease the muscle-directed gene transfer efficiencyofAAV2.cFIXby 2-fold compared to mice pre-injected with an empty adenovirus However, the performance ofliver directed AAVmediated gene transfer was not diminished in animals with high levelsofpre-existing capsid 'I' cells Weconclude that cross presentation of AAVcapsids does result in activation of CTL in a serotype specific manner and the pre-existing capsid-specific'l-cells arc capable to hinder the muscle-directedgene transfer, however there is no evidence that vector transduced hepatocytesare targets for CTL effector activity 256 Ultrasound Guided Gene Transfer to the Fetal Thymus Results in Thymic Epithelial Cell Transduction and Deletion of Transgene Specific T Cells Masayuki Endo, William H Peranteau, Phillip W Zoltick, Antoneta Radu, Nidal Muvarak, Alan W Flake [The Children's Center for Fetal Research, The Children's Hospital ofPhiladelphia Philadelphia PA Fetal immune ontogeny provides a unique opportunity for manipulationof tolerancefor prenataland anticipated postnatalgenetic and cellular therapies Postnatal intra-thymic cell or gene transfer has been shown to result in antigen specifictolerance in rodents, but these results have not been successfully translated to large animals Whereas in rodents the thymus plays a key role in determination of T-cellrepertoireand generationofT-regulatorypopulationswell into adult life, in higher mammals this role is confinedto early gestation Our aim in this study was to develop a murine model for thymic antigen presentation, which would be analogous to the anticipated timing required for effective thymic presentation of antigen, in a large animal system We hypothesized that thymic presentation of antigen early in gestation would result in deletion of antigen specific 'l-cells, To test this hypothesis, we developed two differS9S ent ultrasound guided approaches for achieving intrathyrnic gene transfer in fetal mice First, we injected lentiviral vector carrying the green fluorescent protein (OFP) reporter gene into the murine amniotic cavity from post coital day (E8) to Ell, when the thymic primordiumconsists of pharyngeal pouch epithelium and is in direct contact with the amniotic fluid Second, we injected vector directly into the fetal thymus at E17 We analyzed the injected mice under fluorescence stereornicroscopyat sequential time points after birth and identifiedOFP expressing cell types by immunohistochemistry Furthermore, we assessed deletion of specific TCR bearing 'l-cells after early intra-amniotic gene transfer (lAGT) with Icntiviral vector encoding an exogenous mammary tumor virus oncogene (rntv), Balb/c (MHC II I-E+, mtv7-) fetal mice underwent IAOTat E9 with either Lenti-mtv7(experimentalgroup), Lenti-mtv9(irrelevant mtv vector) or no vector.The peripheral blood (PB), thymus and spleen of experimental and control mice were analyzed at intervals after birth by flow cytometry for Tscells expressing VJ36 TCR which are normally deleted in mtv?", 1-£+ strains of mice Sustained OFP expression was observed in the thymus after lAO'I' from E8 to ElO and after direct intra-thymic injection at E17.Histological analysis confirmed the expression of OFP in cortical and medullar thymic epithelial cells Expressionofmtv7 in the thymus resulted in partial deletion of relevant VJ36 TCR bearing T-cells in the PB and spleen Experimentalanimals maintaineda significantreduction in the level oftransgene specific 'I' cells for the duration ofthe study, i.e beyond months of life, The current study supports the ability of in utero thymic gene transfer to I) transduce thymic cortical and medullary antigen presenting cells and 2) induce deletion oftransgene specific T cells This model system should be valuable for testing strategies for specific tolerance induction that may translate to large animal systems 257 Characteristic of Responsive CD8+ T Cells to AAV 2/8 and Ad C7 HIV Gag Vector Vaccines Jianping Lin,' Joanita M Figueredo,' Yan Zhi,' Roberto Caicedo,1 Lili Wang,1 Zhe Zhang,' James M Wilson.' 'Gene Therapy Program Department ofPathology and Laboratory Medicine, University ofPennsylvania, Philadelphia, PA Adeno-associated viruses (AAV) and adenoviruses (Ad) have been developed as vectors for gene therapy and genetic vaccines In this study, we characterized the peak and long term COST cell responses elicited by recombinant AAV2/S and simian derived AdC7 vectors expressing HIV-I gag in CB6Fl hybrid mice The gag specific COST cell responses were evaluated by monitoring the frequency, phenotype and functional properties of gag specific 'I' cells in mouse spleen and liver Intramuscular administration of lEI I OC of AAV2/S gag resulted in peak frequencies of gag tetramer (H2-Kd AMQMLKETI) equal to 24 and 60% of all CDS 'I' cells in spleen and liver, respectively.About 50% of these cells were found to persist as long as 140 days post immunization A 10fold reduction in dose of AAV2/S vector still produced high levels of tetramer positive COST cells (6 and 2S% in spleen and liver, respectively).These tetramer positive 'I' cells were further classified as effector, effector memory and central memory subsets based on the expression of CD62L and CDI27 cell surface markers The majority of the gag tetramer specific COST cells (93%) following AAV2/S administrationwere effectorcells which is in contrast to the response achieved with AdC7 vector where both effector and effector memory gag specific cells were elicited at the peak of response and a small percentage of central memory cells were observed as long as day 127 post immunization In addition to phenotyping for memory markers, 'I' cell responses were characterized with respect to antigen induced cytokine secretion At the peak, AAV2/S gag induced an equal proportion of CDS T cells secreting IFN-g alone and IFN-glTNF-a in all tissues studied In the case of AdC7gag, Molecular Therapy Volume 15.Supplement Iã.\br W07 Copyright â "1111; Arncricm $Ol;(f,:ty of Gene Therapy the majority of the gag specific CD8 T cells secreted both IFN-g and TNF-a in multiple organs at the peak of response Long term evaluation revealed 20% of gag specific CD8 T cells secreting IFN-g, TNP-a and IL-2 in AdC7 immunized mice Moreover, both AAV2/8 and AdC7 gag immunized mice produced gag specific CD8 T cells expressing granzyrne B In vivo CTL assays demonstrated that both AAV and Ad immunizations can elicit the production of CTLs with the capability of lysing in vivo antigen loaded target cells In conclusion, AAV2/8 gag immunization induces a robust and relatively long-lasting antigen specific effector T cell response, while AdC7 gag immunization resulted in the generation ofeffector memory and central memory T cells 258 Lentiviral-Mediated Gene Transfer to Lung Activates Transgene and Vector Specific T Cells Maria L Limberis, I Christie Bell, I Roberto Caicedo, I Luk 1-1 Vandenberghe; Deirdre Mclvlenamin,' Regina Munden; Di WU,1 Julie C Johnston, I Peter Bell, I James M Wilson.' 'Gene Therapy Program, Department ofPathology and Laboratory Medicine, University ofPennsylvania, Philadelphia, PA Gene transfer vectors based on HIV-I (Ientiviral vectors) are attractive gene transfer agents for airway genetic diseases such as cystic fibrosis due to their inherent ability to transduce quiescent cells in a stable manner Since less than I% of the airway epithelium is actively dividing at any given time, the use of these vectors is warranted To improve targeting efficiency of lentiviral vectors in airway epithelium much effort has been placed on modifying envelopes from viruses with known affinity for airway One of these viruses is the Ebola Zaire virus, the glycoprotein of which has been mutated (NTDL6) to successfully pseudotype a Ientivirus vector [Medina et aI., 2003, Mol Ther; (5) 777-89] and significantly improve its transduction efficiency in rodent and non-human primate airway (Kobinger, unpublished) To further characterize the potential of a lentivirus vector for lung-directed gene transfer we studied the likelihood ofT and B cell activation to the vector and encoded transgene product following delivery to the lung which may compromise the therapeutic potential ofthis approach in vivo For dosing, vectors wcrc delivered in 50 ml (4E+8 transducing units) in either C57B1I6 or BalbC mice (n=41 group) For delivery to lung, mice were subjected to an intratracheal (IT) instillation of NTDL6-pseudotyped lentivirus vector expressing GPP under the control of'the CMV promoter As negative control for gene transfer, a VSVG-pseudotyped lentivirus vector expressing GFP was used To activate T cells, groups of mice were also injected intramuscularly (1M) with the same vectors (i.e., NTDL6- or VSVG- pscudotyped HIV-GFP vector) Spleens were harvested from each mouse (treated either IT or 1M) as well as the draining superficial cervical lymph nodes from mice treated IT Isolated lymphocytes were assayed by IFN-g ELISPOT following stimulation with the entire GFP library (to assay transgene-specific responses) or the p24/p 17 library (to assay vector-specific responses) NTDL6-pseudotyped H1V-GFP vector resulted in significant transduction of the alveolar and less so of the conducting airway epithelium VSV-G pseudotyped HIV vector resulted in no GFP transduction in lung Immunologic studies were performed in all animals dosed with pseudotyped HIV vectors Following 1M injection, there was significant T cell activation to GPr, and B cell responses to both GPP and p24 in C57BI/6 and BalbC mice, with responses being more pronounced in the latter POl' [T delivery, responses were lower than those achieved by 1M injection but appeared to be sustained up to wks post injection There was a trend in the decrease of GFP-cxpressing cells in lung with time possibly reflecting immuno-toxicity We arc currently investigating the impact of the OFP-specific T cells to decipher Molecular Therapy Volume15.Suppkmen.l ~br Copyright © The American So(,.;C!;}"of Gene Therapj- 2007 whether the decrease in gene expression in lung is due to the normal cell turnover of the epithelium or due to the cytotoxic nature of the OPP-specific CD8 T cells 259 Characterization of AAV Capsid Responses in Humans Using Normal Donor Splenocytes Daniel J Hui,' Federico Mingozzi,' Etiena Basner-Tschkaraian,' Shyrie Edmonson, I Katherine A High.l-' I Hematology, Children s Hospital ofPhiladelphia, Philadelphia, PA; lHemat%gy, Howard Hughes Medica/Institute, Philadelphia, PA Although adeno-associated viral (AAV) vectors remain a promising gene therapy approach for the treatment of genetic disorders, concerns over potential immune responses are a major obstaele facing successful implementation in a clinical setting It has recently been reported that liver-directed gene transfer of factor IX using AAV serotype to correct hemophilia B activated a T cell response against AAV capsid, thereby causing destruction of transduced hepatocytes leading to a transient transaminitis (Manno et al Nat Med 2006) In order to further elucidate T cell responses against AAV capsid and to help predict potential immune responses in future subjects undergoingAAV-mediated gene transfer, a novel approach using normal donor human splenocytcs as a source for T cells was developed Splenocytes offer two major advantages over a conventional peripheral blood mononuclear cell (PBMC)-based approach First, cells are harvested directly from the spleen, a lymphatic compartment, as opposed to the peripheral circulation; and second, splenocytes are harvested in much higher cell numbers than PBMCs, which allows multiple experiments to be performed using cells from a single donor To date, a total of 43 donor splenocyte samples have been collected from normal adults or children undergoing splenectomy Once harvested, splenocytes are utilized forT cell functional assays such as intracellular cytokine staining (ICCS) or Enzyme-Linked Immunospot (ELI Spot) assays To improve detection of low-frequency T cell responses, an expansion strategy was used in which splenocytes were incubated with candidate cpitopes in culture prior to testing Ofthe 19 samples that wcre expandcd, 12 had detectable T cell responses by ELiSpot assay representing a much higher percentage (63%) than previous PBMC data (Blood 2006; 108( [ I): [38a) In conjunction with bioinformatics prediction algorithms, new epitopes are reported here for MI-IC Class I HLA haplotypes, including of the most frequently occurring haplotypes in the normal population Although most subjects recognized a single AAV epitope, of these subjects recognized or more cpitopes Expanded splenocytes wcrc also successfully used as effector cells in cytotoxicity (CTL) assays against HLAmatched AAV-transduced human hepatocytes, further verifying functionality of T cell epitopes There was no correlation between the magnitude ofthe T cell response and anti-AAV IgO levels from ELISA assays performed on matched serum samples for [2 of the donors screened, although all donors did have detectable levels of anti-AAV antibodies While responses were seen in donors as young as years ofage, more robust responses were detected in donors >20 years old, which has implications for selecting candidates for future clinical trials Current studies are focused on utilizing splenocytes to examine cross-reactive T cell responses among AAV serotypes and phenotyping memory responses In conclusion, the cumulative results from these splenocyte studies validate this approach as being more informative for characterizing T cell responses than current PBMC-based methods S99 ... vector for lung-directed gene transfer we studied the likelihood ofT and B cell activation to the vector and encoded transgene product following delivery to the lung which may compromise the therapeutic... reported that liver-directed gene transfer of factor IX using AAV serotype to correct hemophilia B activated a T cell response against AAV capsid, thereby causing destruction of transduced hepatocytes... long-lasting antigen specific effector T cell response, while AdC7 gag immunization resulted in the generation ofeffector memory and central memory T cells 258 Lentiviral-Mediated Gene Transfer to