854. Optimization of AAV Serotype and Promoter for Increased Distribution of Expression in Both Somatic and Central Nervous Systems of Mice

The characterization of adeno-associated virus (AAV) vector biodistribution, as measured by transgene expression, may be affected by many factors including the presence of viral receptor molecules on cell surfaces, cell-type related intracellular trafficking, the ability of the cell to form double-stranded DNA from the single-stranded AAV genome, and the cell-type specificity of the promoter. The optimal evaluation of biodistribution eliminates or controls as many of these variables as possible. We have approached this by using self-complementary AAV vectors (scAAV) and combiniations of promoters with activity in a very broad range of cell-types, thus narrowing the variable to the intrinsic capacity of the cells to take up and traffic the vector to the nuclei. Self-complementary AAV serotype 1 (AAV1) or 2 (AAV2) vectors (11011 DNase resistant particles) expressing green fluorescent protein (GFP) from either the human cytomegalovirus (CMV) promoter, or a murine small nuclear RNA promoter (U1a), were delivered into adult mice by an intravenous (IV) injection, 10 minutes after an IV infusion of mannitol. The mannitol enhances delivery of vector across the blood brain barrier. Transgene expression was detected by immunofluorescence in multiple somatic and CNS tissues using a polyclonal antibody against GFP, at 1 and 3 months postinjection. Both AAV1 and AAV2 vectors mediated transgene expression in liver, kidney, heart, spleen, intestine, skeletal muscles, and lung, as well as a global distribution of GFP-positive neuronal and non-neuronal cells in the brain and spinal cord. In the CNS, major targets of AAV1 were glial cells, while AAV2 targeted mainly neurons. Endothelial cells of capillaries in both somatic and CNS tissues were efficiently transduced by AAV1, but not AAV2. In kidney, AAV2, but not AAV1, targeted medullar tubule cells. In the white pulp of spleen, AAV1 trasduced lymphatic sheaths while AAV2 infected B-lymphocytes. Promoter activity also contributed to the observed biodistribution of AAV vector expression. The CMV promoter was more efficient than U1a in neurons of the CNS and in endothelial cells in all tissues. The U1a promoter functioned more efficiently in hepatocytes in liver and tubule cells in kidney. The U1a promoter, but not CMV, was active in lymphatic tissues in spleen, while CMV, but not U1a, functioned in glial cells in the CNS. Our observations provide a basis for matching the specific cell-type tropism of different AAV serotypes with the specific transcriptional activity of different promoters to control biodistribution. A cocktail of serotypes containing optimal promoters will afford the broadest vector expression for the treatment of conditions such as lysosomal storage diseases with widespread somatic and CNS pathology.