Abstract
Effective strategies for the neuroprotection and preservation of retinal ganglion cells (RGCs) remain elusive in the management of glaucoma. A spontaneous genetic model of glaucoma has been identified in cats and extensively characterized as a viable translational model, with eye size and anatomy similar to humans. In this study we sought to establish initial proof of concept for gene delivery to feline RGCs via intravitreal injection of AAV2 in normal cats. Pre-retinal, posterior vitreal injection of AAV2/2-CMV-GFP, was performed overlying the area centralis in 5 adult cats. Immunosuppressive oral prednisolone was administered perioperatively and gradually tapered over 6-10wks post-injection. Ophthalmic examination was performed pre- and post-injection. The GFP reporter expression and morphological effects of viral transduction on the retina were monitored in vivo using confocal scanning laser ophthalmoscopy (cSLO) and optical coherence tomography (OCT), respectively (Spectralis OCT-HRA, Heidelberg), at 1-2wk intervals over 6-10wks. Full-field electroretinograms (ERG) and visual evoked potentials (VEP) were recorded at baseline and post-injection. Retinas were examined by histology and immunolabeling for the RGC marker RBPMS and Müller cell and astrocyte marker SOX9, and GFP expression was examined in the retina, optic nerve (ON), optic tract and lateral geniculate nucleus (LGN). GFP+ retinal cells and RGC axons were visualized by cSLO at 1-2 weeks post-injection. No retinal morphological changes were observed by OCT in vivo but 3/5 eyes exhibited mild retinal inflammation on histology. Retinal and ON function were preserved in injected eyes compared to baseline and untreated eyes. GFP expression was predominantly identified in RBPMS+ RGC cells as well as SOX9+ Müller cells. GFP fluorescence was observed throughout RGC nerve fiber tract in the central visual pathway. Peak transduction in RGCs (up to ∼ 20 %) was observed in the regions with high GFP expression, but < 1 % of RGCs expressed GFP across the whole retina. Our data provide proof of concept that pre-retinal injection of AAV2/2 may represent a feasible platform for gene delivery to feline RGCs in vivo but highlight a need for further refinement to improve RGC transduction efficiency and control low-grade retinal inflammation.