Efficient in vivo editing of CEP290 IVS26 by EDIT-101 as a novel therapeutic for treatment of Leber Congenital Amaurosis 10

Abstract

Purpose : Leber Congenital Amaurosis 10 (LCA10) is an early-onset retinal degenerative disease caused by mutations in the CEP290 gene. LCA10 is not amenable to AAV-mediated gene replacement therapy because the large size of the CEP290 cDNA (~7.5kb) exceeds the packaging capacity of AAV. Therefore, we developed EDIT-101, which is an AAV5 vector expressing SauCas9 driven by the photoreceptor-specific GRK1 promoter and gRNAs selectively targeting the common CEP290 IVS26 c.2991+1655 A>G mutation. Based on existing clinical observations, we hypothesize that correction of CEP290 mRNA splicing defect in ≥10% of photoreceptors of LCA10 patients would lead to clinically meaningful visual improvements.

Methods : To support the clinical dose selection, we evaluated the dose, exposure and editing efficiency following subretinal delivery of EDIT-101 in human CEP290 IVS26 knock-in (KI) mice and a surrogate AAV5 vector targeting non-human primate (NHP) CEP290 gene in primates. A human retinal explant model was also developed to demonstrate that EDIT-101 can produce up to 49% CEP290 gene editing in the therapeutic target cell type. The specificity of the gRNAs was demonstrated through orthogonal methods including in silico bioinformatics, GUIDE-Seq and Digenome, and verified by targeted next generation sequencing (NGS) in human cell lines and retinal explants.

Results : Subretinal delivery of EDIT-101 and NHP vector achieved efficient vector exposure in ~30% of the neural retina, and close to 50% productive CEP290 gene edits was observed in transduced cells in HuCEP290 KI mice and NHP at a vector dose of 1E+12 vg/mL. The productive CEP290 gene editing is detectable as early as on Day 7 post dosing in HuCEP290 KI mice, correlated with the levels of Cas9 mRNA and gRNAs and sustained in the 6 months study. The minimal therapeutic target of 10% productive edits was readily achieved by 3E+11 vg/mL of EDIT-101 in HuCEP290 KI mice. In NHP, the pre-existing immunity to AAV5 capsid or SauCas9 did not prevent the transduction and on-target editing in photoreceptor cells.

Conclusions : Thus, the PK/PD profile of EDIT-101 and the surrogate NHP vector support the translation of the dosing regimen from preclinical models to humans. The in vivo CRISPR approach may have broad application to other inherited retinal diseases with significant unmet medical needs.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.