721. Long-Term (2-8 years) Body-Wide Expression of AAV9-Minidystrophin Gene in Golden Retriever Muscular Dystrophy Dogs After Regional Limb Vein Injection

Abstract

Previously we have shown that AAV9-minidystrophin gene delivery by isolated hindlimb vein injection with pressure was able to obtain widespread gene expression, not only in injected limb but also in muscles body-wide. This phenomenon was attributed to both leaky tourniquet blockade and/or AAV9 vector retained in the vasculature space that later entered systemic circulation. Here we report the results of long-term study up to 8 years. Methods: Five GRMD dogs were treated with the same vector dose of 1 × 1013 v.g./kg. Dog Jelly (2.5 months old; 6.3 kg) was given AAV9-CMV-opti-hMinidys (codon-optimized human minidystrophin gene); Dogs Jasper and Peridot (4 mon old; 12.2 kg & 12.5 kg) were given AAV9-CMV-cMinidys (canine minidystrophin gene). Dogs Rutela and Laredo (14 mon old; 15.5 kg & 15.3 kg) were given AAV-CK-opti-cMinidys. No immuno-suppressant was used. The vectors were injected via the great saphenous with a tourniquet positioned at the proximal pelvic extremity (first 3 dogs) or above the knee (last 2 dogs) to block the blood circulation for a total of 10 minutes. MRI imaging after vector injection in Jelly confirmed vector fluid and muscle enlargement in the injected limb. Muscle biopsy and final necropsy were performed at various time points (from 2 months to 8 years) and analyzed for gene expression and immune responses. Results: Immunofluorescent (IF) staining showed that all 5 dogs obtained long-term minidys expression in a majority of muscle samples examined up to final necropsy. Contractile force measurement showed partial improvement when compared to the untreated dogs. While the percentages of minidys positive myofibers varied among different muscles, certain muscles had greater than 90% of myofibers positive upon necropsy. Importantly, the human minidys expression persisted for 8 years in Jelly despite initial inflammation in injected limb. Overall gene expression was largely stable. For example, positive myofibers in cranial sartorius muscle remained comparable throughout the 5 time points, from 2, 7 months to 1, 4 and 8 years. Minidys was also observed in approximately 20% of the cardiomyocytes. Interestingly, injected limb had lower expression than non-injected limb, suggesting procedure related inflammation and partial CMV promoter shutdown. Jelly remained ambulant throughout the more than 8-year post treatment study and was euthanized due to cardiomyopathy in the final year. DNA sequencing showed that Jelly did not carry the recently reported disease-modifying Jagged 1 mutation found in two phenotypically mild GRMD dogs. However, it is not possible to attribute Jelly's long-term survival to gene therapy because in the past a few GRMD dogs in the same colony survived to 5 years without treatment. For the other 4 dogs who were larger at the age of treatment, the injected limb consistently showed higher expression and no procedure related inflammation throughout the time course of 2 years. Moreover, no tumors were found in any of the treated dogs. Final analyses of necropsy samples on vector distribution and Western blot are underway to evaluate the efficiency of gene transfer and expression. Conclusions: 1) AAV9 can render long-term and significant systemic minidys gene expression without immune suppression upon isolated limb vein injection. 2) The human minidys gene and pressurized perfusion might have caused inflammation and innate immune responses in the injected limb in Jelly. 3) Muscle-specific promoter is preferable for reducing expression in non-muscle cells. 4) These findings support the feasibility of AAV9-CK-minidys gene therapy in human DMD patients.