Photoreceptors (PRs) and retinal pigment epithelium (RPE) are the primary location of mutations underlying many forms of inherited retinal degeneration for which there is currently no treatment. Therefore, a major goal in the field of vision research is to develop improved technology for gene delivery to these cells. This proposal aims to address this problem by creating novel adeno-associated (AAV) vectors capable of targeting PRs and RPE when delivered into the canine vitreous. We have selected the dog as it is a relevant large animal model that shares similar globe volume and retinal structure with humans, as well as naturally- occurring mutations that result in phenotypically similar retinal diseases as those identified in patients. Naturally occurring serotypes of AAVs are unable to penetrate structural barriers of the retina, and therefore therapeutic delivery of DNA to PRs and RPE has required until now that the viral vector be injected into the subretinal space. Subretinal injections entail the passage of a needle through the retina as well as a transient retinal detachment that separates the RPE from the PRs. This microsurgical technique requires general anesthesia, delivers the viral vector to a limited area of the retina/RPE, and has been associated in preclinical safety studies and human clinical trials with retinal thinning under the fovea and the formation of macular holes. Recently, AAV vectors have been optimized for increased targeting of PRs and RPE following intravitreal delivery, which would provide easier, safer and more extensive targeting of the PRs and RPE than vectors delivered via the subretinal route. Successful results have been achieved in mice, however, when tested in large animals these vectors were inefficient due to differences in retinal structure. Based on strong initial results in mice, we will use a similar directed evolution approach to select, from a mixture of highly diverse libraries of mutated AAVs, individual variants with the novel ability to traverse structural barriers and infect the dog outer retina from the vitreous. Resulting vectors will be characterized by packaging GFP and analyzing expression profiles. We will then test whether the genetic features that increase transduction in dogs are also sufficient to improve infection of the non-human primate outer retina. Furthermore, we will use deep sequencing to analyze rounds of selection, in order to elucidate at which steps diversity is reduced and to determine if the final outcome of directed evolution can be predicted from the initial rounds. The identified novel variants will be used to test corrective gene therapies in the rcd1 dog, a model of PR disease caused by a mutation in the rod-specific PDE6? gene, as well as in the RPE65 mutant dog, a model of Leber's congenital amaurosis due to a genetic defect in the RPE. We will then measure whether intravitreal injections provide equivalent or better rescue compared to subretinal injections by evaluating retinal structure and function. Together, these experiments will provide a diverse training experience and valuable preclinical data for novel gene therapy treatments in the retina.
Currently, subretinal injections are required to reach photoreceptors and retinal pigment epithelium (RPE), which are the primary location of mutations underlying many forms of inherited retinal degeneration for which there is currently no treatment. This microsurgical technique requires general anesthesia, delivers the viral vector to a limited area of the retina/RPE, and has been associated in preclinical safety studies and human clinical trials with retinal thinning under the fovea and the formation of macular holes. This proposal aims to create and test novel adeno-associated (AAV) vectors capable of targeting photoreceptors and RPE from the vitreous in the canine, an important large animal model with similar retinal structure to humans as well as naturally occurring mutations resembling those identified in patients.