A major barrier to effective treatment of the central nervous system (CNS) in most inherited lysosomal storage diseases (LSD) is that the metabolic defect in all brain cells results in widespread pathology. The therapeutic principle for most LSDs is to transfer a normal enzyme cDNA into a subset of diseased cells, thereby correcting both vector-transduced cells and neighboring non-transduced cells that take up the secreted therapeutic enzyme. To achieve global correction, transduced cells must be dispersed 3-dimensionally throughout the brain so that secreted vector-encoded therapeutic enzyme can reach all non-transduced cells. Although intravascular injection of certain AAV vector serotypes can cross the blood-brain barrier and deliver genes widely in rodent brains, AAV vector distribution in large mammals occurs mostly in the lower brain and spinal cord, limiting the potential for treatment. Thus, optimization of vector distribution in large animal models of human diseases is critically needed to facilitate translation into clinical usage. In the prior grant period, we have shown that a novel AAV serotype (AAV.hu32) mediates widespread gene delivery in a cat model of alpha-mannosidosis (AMD) caused by a mutation in the gene encoding lysosomal a-mannosidase (MANB) that recapitulates the severe form of AMD. Notably, high doses of AAV.hu32 encoding MANB resulted in global correction of the storage lesions and improvements in disease parameters. Although this appears promising for translation into clinical usage, the vector doses required would be near the limits of production for clinical grade AAV vector when scaled up to human patients and there have been safety concerns raised in recent clinical trials for use of such high vector doses. Finally, new preliminary data shows that vector doses that are completely effective when AMD cats are treated in the early stages of the disease do not mediate complete brain correction when treatment is initiated at a more advanced stage of disease, providing additional impetus to improve global brain delivery at lower vector doses. In exciting new studies, we have found that injection via carotid artery was even more effective than intravenous injection, suggesting that passage of concentrated virus through the vascular bed of the brain improves uptake. Surprisingly, we also found that while direct injection of AAV.hu32 into brain parenchyma transduced oligodendrocytes, astrocytes and neurons, its vascular delivery resulted in the almost exclusive transduction of neurons. Thus, current studies will focus on reducing the required vector dose by determining the route the vector takes between blood and brain to specifically transduce neurons, increasing the levels of MANB produced by the genetically corrected neurons, and increasing the number of transduced cells by an alternative dosing regimen that will maximize vector uptake in the vascular bed of the brain. We will then determine if an optimized combination...