AAV gene therapy restores ON-bipolar cell function in a canine model of LRIT3-congenital stationary night blindness

Abstract

Purpose : Congenital stationary night blindness (CSNB) is characterized by non-progressing, impaired night vision for which there is no therapy. Mutations in 18 genes, many of which are involved in the rod to ON-bipolar cell (ON-BC) signaling are known. We previously characterized the only canine model of Schubert-Bornschein type complete CSNB, and more recently, identified the causal mutation in LRIT3, a gene associated with CSNB in mice and patients. While AAV gene therapy has successfully targeted the outer retina of dogs and humans, studies targeting the mid-retinal cells such as ON-BC are limited. Reasons include physical barriers of the retina architecture, need for cell-specific AAV and promoters for greater efficiency, and until recently, lack of appropriate large animal models. We used the newly established canine LRIT3-CSNB for therapeutic targeting of ON-BC function by AAV gene therapy.

Methods : Canine LRIT3 cDNA driven by ON-BC-specific mGluR6 promoters was packaged into mutant AAV capsids previously identified through a directed-evolution approach. Four unaffected control animals received subretinal and/or intravitreal AAV-LRIT3 injections to study safety and transgene expression. Seven CSNB affected dogs received AAV-LRIT3 subretinal and/or intravitreal injections. Functional recovery was assessed by ERG and standardized obstacle-avoidance course. Expression of LRIT3 and associated markers was examined by IHC.

Results : No adverse effects were observed in controls injected with AAV-LRIT3. Of the CSNB animals, eyes that received AAV K9#4, optimized for subretinal injection, combined with a shorter mGluR6 promoter resulted in dramatic recovery of the scotopic b-wave. Further, obstacle course performance under dim light was improved in these eyes. Notably, combined subretinal and intravitreal injections resulted in higher b-wave recovery than subretinal injection alone, whereas intravitreal injection alone had no therapeutic effect. Functional rescue was sustained for up to 12 months post-injection.

Conclusions : We have demonstrated safety and long-term efficacy of ON-BC functional recovery following AAV-LRIT3 therapy. With intact retinal architecture and validation of reliable output measures, the canine LRIT3-CSNB model has great potential for treating CSNB in patients as well as serving as a platform to test therapeutic targeting of ON-BC function.

This is a 2020 ARVO Annual Meeting abstract.