359 novel, shuffled, human specific rAAV vectors

359 Novel, Shuffled, Human Specific rAAV Vectors Molecular Therapy Volume 20, Supplement 1, May 2012 Copyright © The American Society of Gene & Cell Therapy S140 AAV VECTORS II AAV Vectors II 357 Regr[.]

AAV VECTORS II AAV Vectors II 357 Regression of Schwannomas Induced by AAV-Mediated Delivery of Caspase-1 Shilpa Prabhakar,1 Mehran Taherian,1 Davide Gianni, Thomas J Conlon, Anat S Rachamimov,1 Miguel S Esteves, Xandra O Breakefield,1 Gary J Brenner.1 Neurology, Massachusetts General Hospital, Charlestown, MA; Anaesthesiology, MGH, Charlestown, MA; 3Neurology, UMASS Medical School, Worcester, MA; 4Pediatrics, University of Florida, Gainesville, FL; 5Pathology, MGH, Charlestown, MA; 6Neurology, UMASS Medical School, Worcester, MA; 7Neurology, MGH, Charlestown, MA; 8Anaesthesiology, MGH, Charlestown, MA Schwannoma tumors form from proliferation of dedifferentiated Schwann cells within peripheral nerves In the human diseases- NF type and schwannomatosis, these tumors are composed purely of Schwann-lineage cells Although typically non-malignant these tumors can cause disruption of sensory and motor function, including extreme pain and loss of hearing due to nerve compression Currently, the only treatment for these patients is surgical resection of symptomatic tumors, which poses the risk of nerve damage and by inaccessibility due to tumor location We have explored gene therapy for these tumors using a xenograft model in which immortalized human schwannoma cells (HEI-193) expressing a fluorescent protein and luciferase are implanted in the sciatic nerve of nude mice Our strategy has been to express the apoptotic protein, caspase-1 (ICE) under the P0 promoter, which is selectively active in Schwann cells during development using an adeno-associated vector (AAV1) In this mouse model of peripheral nerve schwannomas, we can prevent development of tumors and cause regression of well-established tumors through direct intra-tumoral injection of the AAV1-P0-ICE vector, which had no effect on sensory (using von Frey filament withdrawal threshold of the hind-paw ipsilateral to the tumor or motor (using an accelerating rotarod) functions In order to test whether intra-tumoral ICE injection could reverse schwannoma-related pain behavior we developed a new model in which HEI-193 cells were implanted proximally in the sciatic nerve of nude mice in a location where tumor growth is constrained by pelvic structures, thus leading to nerve compression and associated mechanical sensitivity The injection of ICE into these tumors caused concurrent tumor regression and resolution of mechanical (‘pain’) sensitivity To evaluate any potential for vector-associated nerve damage, neuropathology and behavioral testing was conducted in both nude and immunocompetent mice following injections of ICE While there was an expected transient decrease in the von Frey withdrawal threshold associated with the surgical manipulation (injection), neither mechanical sensitivity nor rotarod performance was otherwise affected Neuropathological evaluation of sciatic nerves of these mice following ICE injections revealed virtually no demyelination, axonal degeneration or inflammation These preclinical studies support the potential efficacy of this AAV1-P0-ICE vector for the clinical treatment of schwannomas by direct injection into tumors Because schwannomas are typically non-malignant and slow-growing, reduction of tumor mass responsible for dysfunction due to nerve compression, rather than a need for complete ablation - would be expected to provide substantial benefit 358 Directed Evolution of AAV for Enhanced Evasion of Human Neutralizing Antibodies Melissa A Bartel,1 Bum-Yeol Hwang,1 Daniel Stone,1 James T Koerber,1 Linda Couto,4 Federico Mingozzi,4 Katherine A High,4,5,6 David V Schaffer.1,2,3 Chemical and Biomolecular Engineering, University of California, Berkeley, CA; 2Bioengineering, University of California, Berkeley, CA; 3Helen Wills Neuroscience Institute, University of California, Berkeley, CA; 4Center for Cellular and Molecular Therapeutics, Children’s Hospital, Abramson Pediatric Research Center, Philadelphia, PA; 5Department of Hematology, Children’s Hospital, University of Pennsylvania, Department of Pediatrics, Philadelphia, PA; 6Howard Hughes Medical Institute, Philadelphia, PA Gene delivery vectors based on adeno-associated viruses (AAV) have demonstrated promise in both preclinical models and human clinical trials for several disease targets, including hemophilia and Leber’s congenital amaurosis However, the high prevalence of anti-capsid neutralizing antibodies, due to widespread exposure to numerous AAV variants and serotypes in the human population, decrease the efficacy of AAV gene therapy This pre-existing immunity, as well as immunity due to prior vector administration, can impede the broader implementation of AAV gene therapy and must be addressed to build upon successful AAV results in immune privileged sites Directed evolution has proven to be a powerful approach to generate AAV vectors with novel capabilities, and our results show that AAV can evolve to significantly overcome neutralization by anti-AAV antibodies, both in vitro and in vivo An AAV2 cap gene containing point mutations isolated from previous selections using individual human serum samples was further diversified through saturation mutagenesis at additional immunogenic positions Following three rounds of evolution, novel AAV variants derived from the saturation mutagenesis library and a shuffled AAV library required higher neutralizing antibody titers (using human IVIG) than wild-type AAV in vitro One mutant (SM 10.2) that contained wild type residues at the saturation mutagenesis sites, yet acquired new mutations in other regions, required a 3.4-fold higher in vitro IVIG concentration for neutralization vs AAV1 (6-fold vs AAV2) and displayed similar in vitro tropism to AAV1/AAV6, despite being derived from AAV2 A shuffled mutant (Sh 100.3) composed of AAV1, 2, 3, 4, 6, 7, and required a 20-fold higher in vitro IVIG concentration for neutralization vs AAV1 (35-fold vs AAV2) and displayed similar in vitro tropism to AAV1/AAV6 SM10.2 and Sh100.3 also showed enhanced in vitro transduction in the presence of serum samples from individuals excluded from a hemophilia B clinical trial The antibody neutralization properties also translate to enhanced transduction in vivo SM 10.2 displayed similar in vivo tropism to AAV2, but with higher heart transduction and significantly lower liver transduction in naïve mice In IVIG-immunized mice, SM 10.2 had significantly higher transduction of heart, liver, and muscle compared to AAV2 In vivo tropism of Sh 100.3 is currently under evaluation The isolation of such novel clones resistant to anti-AAV antibodies may enable the future treatment of patients with high antibody titers that are currently ineligible for AAV gene therapy 359 Novel, Shuffled, Human-Specific rAAV Vectors Leszek Lisowski,1 Kirk Chu,1 Mark A Kay.1 Department of Pediatrics and Genetics, Stanford University School of Medicine, Stanford, CA Recombinant adeno-associated viral (rAAV) vectors are one of the most promising vectors for use in human gene therapy The proviral genomes are episomally maintained in quiescent cells Most importantly, rAAVs can be pseudotyped with a large number of available capsids that can have significant effects on the transduction S140 Molecular Therapy Volume 20, Supplement 1, May 2012 Copyright © The American Society of Gene & Cell Therapy AAV VECTORS II properties including cell specificity and immune responses To this end, we used a shuffled AAV library approach to select a chimeric rAAV variant that would specifically transduce human hepatocytes and not be recognized by pre-existing humoral anti-AAV antibodies Since the transduction of primary hepatocytes in culture not correlate well with what is observed in vivo, we hypothesized that in order to select a vector specific for human hepatocytes in vivo, the library selection would also need to be performed on human cells in vivo For this reason, we performed AAV selection in Fah-/-/ Rag2-/-/IL2rg-/- (FRG) mice repopulated with more than 60% human hepatocytes A new AAV capsid library was generated from ten naturally occurring AAV serotypes, in a manner similar to that previously published (Grimm et al., 2008) After four rounds of selection, a number of variants showed significant enrichment in the isolated pool We selected 19 novel AAV variants to make vectors for further in vitro and in vivo analysis Interestingly the new capsids contained significant contributions from AAV-1, AAV-3, AAV-8, and AAV-9, while sequences of AAV-2, AAV-4 and AAV-5 capsids were underrepresented We have performed a detailed analysis of the packaging efficiencies (titers) of the new isolates, as well as transduction efficiencies on number of mouse and human cells in culture One of the isolates, AAV-LK03, was shown to be very efficient at transducing number of human cells, including hepatocytes, endothelial cells, fibroblasts, primary keratinocytes and T-ALL cells, with virtually no detectable transduction of murine cells, indicating that this vector might be human specific We are currently testing this and other top candidates for their potential to evade IVIG (pooled human antibodies) neutralization Due to the predicted human-specific character of AAV-LK03, we are using the humanized FRG mice to determine the relative transduction efficiency of these new vectors in human vs mouse hepatocytes in vivo Currently we are repeating the in vivo selection with novel improved AAV libraries using more stringent conditions at each selection cycle to enhance the selection process We believe this approach will allow us to select for novel AAV capsids with new and superior transduction properties for use in clinical studies 360 Enhanced Alpha-1 Antitrypsin (AAT) Expression and Decrease Immune Response with Isolated Limb Perfusion (ILP) Delivery of rAAV1CB-hAAT Jeffrey D Chulay,1 Guo-jie Ye,1 David R Knop,1 Chrystal L Montgomery,2 Jerry R Mendell,2 Louise R Rodino-Klapac.2 Applied Genetic Technologies Corp., Alachua, FL; 2The Ohio State University/Nationwide Children’s Hospital, Columbus, OH Recombinant adeno-associated virus (rAAV) vectors offer promise for gene therapy for AAT deficiency, but higher transgene expression than has been attained in clinical trials to date will be required to achieve therapeutic serum AAT concentrations We compared intramuscular (IM) and ILP delivery of rAAV1-CB-hAAT (1 x 1011 vg = x 1012 vg/kg) in C57BL/6 mice (n=7 per group) For IM injections two sites (tibialis anterior and gastrocnemius) in one leg received 50 μL ILP delivery to one leg was performed using published methods (J Transl Med 2007;5:45) Serum was obtained before and at 4, and 12 weeks after vector administration for measurement of human AAT (hAAT) and antibodies to hAAT by ELISA At each time point, the average serum hAAT concentration was 6.6- to 8.9-fold higher after vector administration by ILP compared to IM injections Serum anti-hAAT (U/mL, mean ± SD) and ratio of ILP:IM values IM injections ILP delivery ILP:IM Week 17 ± 7.6 0.7 ± 0.9 0.04 Week ± 3.9 0.7 ± 1.1 0.08 Seek 12 ± 2.2 2.2 ± 3.4 0.52 These results provide support for future clinical evaluations of ILP delivery of rAAV vectors expressing AAT in patients with AAT deficiency 361 Arsenic Trioxide Treatment Increases the Transduction of Adeno-Associated Virus Serotype (AAV2) Vectors Both In Vitro and In Vivo Angela M Mitchell,1,2 Richard Jude Samulski.1,3 Gene Therapy Center, University of North Carolina at Chapel Hill, NC; 2Department of Microbiology and Immunology, University of North Carolina at Chapel Hill, NC; 3Department of Pharmocology, University of North Carolina at Chapel Hill, NC Adeno-associated virus (AAV) vectors are popular vectors for gene therapy and have thus far been used in over seventy clinical trials to treat both monogenetic diseases, such as hemophilia and muscular dystrophy, and acquired diseases, such as heart disease and prostate cancer Although these trials have demonstrated increasing success in reaching their efficacy goals, especially in restricted sites such as the retina and the CNS, many trials are still hampered by low initial transgene expression or gradual loss of transgene expression over time as observed in a recent hemophilia clinical trial (Nathwani et al, N Engl J Med, 2011 365: 2357) In an effort to increase transgene expression from AAV vectors, we have investigated the effects of chemotherapeutic agents on the transduction of AAV serotype (AAV2) vectors Arsenic trioxide (As2O3) is currently used as a chemotherapeutic drug in the treatment of acute promyelocytic leukemia resulting from a fusion of the promyelocytic leukemia protein (PML) with the retinoic acid receptor (RARα) Using AAV2 carrying an EGFP reporter and 293 cells, we have determined that pretreatment with As2O3 1) increases the percentage of cells transduced in a dose dependent manner, 2) increases transduction three to four fold with relatively little toxicity, and 3) causes increases the number of vector genomes found per cell that correlates with the increase in transduction In exploring the universality of this effect, we have determined increased transduction can be observed 1) at several time points after transduction, 2) at a wide range of vector doses, 3) in several human cell lines, and 4) in cell lines from several other species Furthermore, we have extended these results in vivo We have determined that, in mice, a short course of As2O3 treatment centering on the time of transduction with AAV2-CBA –luciferase vectors can increase transduction 1) as early as days post-transduction, 2) continuing to more than eight weeks post transduction, and 3) from AAV delivery methods including intravenous, intraperitoneal and intramuscular injection Thus, As 2O can increase AAV2 transduction both in vitro in human cell lines, as well as cell lines of other species, and in vivo through multiple AAV2 delivery methods Currently, we are investigating the mechanism behind the increase in AAV2 transduction due to As2O3 treatment, which, unlike several other chemotherapeutic agents, includes a possible role for the PML protein The mechanistic studies will be discussed in more detail at the meeting These studies should help to elucidate mechanisms by which AAV transduction can be increased without increasing vector doses Serum hAAT (μg/mL, mean ± SD) and ratio of ILP:IM values IM injections ILP delivery ILP:IM Week 20 ± 5.6 133 ± 81 6.6 Week 15 ± 4.2 101 ± 64 6.7 Week 12 10 ± 2.9 91 ± 55 8.9 Despite the higher serum hAAT levels, antibodies to hAAT were lower after vector administration by ILP Molecular Therapy Volume 20, Supplement 1, May 2012 Copyright © The American Society of Gene & Cell Therapy S141 ... might be human specific We are currently testing this and other top candidates for their potential to evade IVIG (pooled human antibodies) neutralization Due to the predicted human- specific character...AAV VECTORS II properties including cell specificity and immune responses To this end, we used a shuffled AAV library approach to select a chimeric rAAV variant that would specifically... we hypothesized that in order to select a vector specific for human hepatocytes in vivo, the library selection would also need to be performed on human cells in vivo For this reason, we performed