Minimally invasive routes of AAV administration to treat GM1 gangliosidosis

Abstract

GM1 gangliosidosis is a lysosomal storage disease caused by a deficiency of lysosomal β-galactosidase (βgal), which results in the accumulation of GM1 ganglioside and fatal neurodegeneration. In the human type II (late infantile / juvenile) disease, GM1 gangliosidosis is a rapidly progressive disease and is fatal by 4 years of age. Outside of supportive and palliative care, no effective treatments exist. The feline model of GM1 closely mimics the human type II disease and provides a vital resource for developing therapeutics and testing safety prior to initiating human studies. After intracranial injection of adeno-associated viral (AAV) gene therapy, GM1 cats show a dramatic increase in length and quality of life. While the efficacy of intracranial AAV treatment is clear, a less invasive option must be considered, especially with the goal of treating children with severe brain disease. Previous intracranial administration of AAV into GM1 cats utilized bilateral injection of the thalamus and deep cerebellar nuclei (DCN). In order to decrease the invasiveness of the treatment, but maintain efficacy, GM1 cats were injected bilaterally in the thalamus (Thal) and in the left lateral ventricle. The intracerebroventricular (ICV) injection has the added benefit of delivering AAV to the cerebrospinal fluid (CSF), which surrounds the central nervous system (CNS) and could theoretically increase treatment distribution. Additionally, two methods of density gradient purification of AAV vectors, cesium chloride (CsCl) and iodixanol, were studied, to evaluate if there is a clinical difference between the two methods. The Thal+ICV, CsCl GM1 animals survived 14.8 ± 3.6 months and Thal+ICV, iodixanol animals survived 31.4 ± 8.4 months. Compared to untreated GM1 cats, which survive 8.0 ± 0.6 months, there are significant increases in lifespans in both cohorts. However, the Thal+ICV, iodixanol cohort had a significantly longer average lifespan than that of the CsCl group. Similarly, normalization of most biomarkers was superior in the iodixanol cohort versus the cesium chloride cohort, further supporting the advantages of iodixanol purification. Furthering the evaluation of less invasive routes of AAV administration, AAV was delivered directly to the CSF through the cisterna magna (CM). The CM is ventral to the cerebellum and requires no brain penetration to access. Two commonly used AAV serotypes, AAV9 and AAVrh10, were evaluated in this study as well, to determine whether one serotype is more beneficial for treatment of neurological disease. Both the AAV9 and AAVrh10 serotypes led to a significant improvement in average lifespan in comparison to untreated GM1 animals, with the AAV9 cohort surviving 13.9 ± 1.8 months and the AAVrh10 cohort surviving 11.3 ± 0.5 months. All animals had a delay, but not halt, of disease progression. There was normalization of myelin integrity in the cerebellum of the AAV9 cohort but not AAVrh10 cohort. In both CM cohorts, there was recovery of βgal activity in the cerebellum and spinal cord but this recovery did not penetrate deep brain structures. In an attempt to increase systemic and CNS biodistribution of βgal, AAV9 was delivered via an intravenous injection. With IV treatment alone, GM1 cats survived to approximately 3.5 years. The IV-treated animals had minimal clinical symptoms that were corroborated by other in life studies. Additionally, post-mortem analysis showed increased levels of βgal throughout the CNS, including deep brain structures, and into the periphery. All antemortem and postmortem biomarkers were improved compared to untreated GM1 animals and those treated by Thal+ICV or CM injection. Taken together, these studies show the therapeutic potential of AAV gene therapy for the treatment of GM1 gangliosidosis and indicate that IV gene therapy is the most effective in the feline model.