Somatic or germline ABCA4 editing to generate a pig model of Stargardt disease type 1

Abstract

Purpose : Stargardt disease type 1 (STGD1), due to mutations in the ABCA4 gene, is the most common form of inherited macular degeneration. Abca4-/- knockout (KO) mice are currently used as animal models of STGD1, however they recapitulate only some of the features of the disease and fail to develop the severe photoreceptor degeneration observed in STGD1 patients. This might be due to the structure of the mouse retina, which largely differs from that of humans. The development of an adequate animal model is thus required to both better understand STGD1 mechanisms as well as testing novel potential therapeutic strategies. The porcine eye shares many similarities with the human eye both in terms of size and retinal architecture, including the presence of a large streak-like region in the retina, which is similar to the human fovea and where the cone/rod ratio reaches 1:3/1:5. Thus, we have planned to generate a pig model of STGD1.

Methods : To this aim, we explored either: i. somatic cell nuclear transfer (SCNT) from primary fibroblasts edited using CRISPR/Cas9 technology, to generate ABCA4 knock-out (KO) pigs; or ii. photoreceptor somatic gene transfer of CRISPR/Cas9 with adeno-associated viral (AAV) vectors in adult pigs.

Results : Screening of potential CRISPR/Cas9 cutting sites in the pig ABCA4 gene led to the selection of one gRNA in exon 2 used to edit fibroblasts for SCNT, and 3 gRNAs in exon 2, 5 and 10 for somatic gene editing. Effective editing of the ABCA4 gene was confirmed in both ABCA4 KO pigs and in retinal lysates from adult pigs 6 months after subretinal delivery of AAV vectors encoding for Cas9 and the combination of the 3 gRNAs. Notably, ABCA4 gene modification resulted in undetectable levels of ABCA4 protein production in ABCA4 KO pigs and significant, although variable, reduction in the ABCA4 levels in AAV-Cas9-injected pigs. Interestingly, we found increased autofluorescence presumably due to lipofuscin accumulation in the retinas of both ABCA4 KO pigs and AAV-Cas9-injected pigs.

Conclusions : We show that ABCA4 editing in pigs results in retinal pigmented epithelium lipofuscin accumulation, as expected. Further characterization of the retinal phenotype is in progress. If successful, the generation of an animal model of STGD1 will provide unique tools for both studying the mechanism of STGD1 rod and cone cell death as well as testing new therapies.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.