Targeted optogenetic tools for visual restoration – a comparison

Abstract

Purpose : Basic levels of visual restoration have been achieved following delivery of light sensitive proteins to the surviving cells of the degenerate retina (optogenetics). However, a diversity of cellular light responses naturally exists in the healthy retina and it remains unclear which optogenetic tool is best suited to replicate these native responses. Here we use a mouse model of retinal degeneration devoid of both canonical and ipRGC photoreception (Pde6b^rd1/rd1,Opn4^-/-) to test the hypothesis that three leading optogenetic tools, each targeted to ON-bipolar cells, would produce characteristic retinal ganglion cell (RGC) light responses.

Methods : Retina-degenerate mice of both sexes, lacking native melanopsin and expressing Cre recombinase in retinal ON-bipolar cells (L7.Cre,Pde6b^rd1/rd1,Opn4^-/-) received intravitreal injections (at P45) of saline or adeno-associated virus (AAV2.2) containing either melanopsin (OPN4), rhodopsin (RHO), or red-shifted channelrhodopsin (ReaChR) genes driven by a “floxed” promoter (n=8 per group). Multiple electrode array (MEA) recordings of retinal ganglion cell (RGC) light responses were performed from intact retina explants 8 weeks post injection with a range of intensities of light. Comparisons were performed with 2-way ANOVA with Dunn’s correction.

Results : Melanopsin treated retinae demonstrated slow onset but long-lived RGC light responses that persisted beyond the end of the light stimulus. By comparison, ReaChR responses showed fast onset ‘sustained’ type responses which terminated on stimulus removal, and rhodopsin treated retina showed fast onset but transient “ON” type responses that terminated before light offset. Overall, melanopsin driven responses were most sensitive (EC₅₀=13.63±0.21 log₁₀photonscm^-2s^-1) and ReaChR (EC₅₀=14.21±0.06log₁₀photonscm^-2s^-1) least sensitive (p=0.021).

Conclusions : This direct comparison of leading optogenetic tools (the first of its kind in the literature in the absence of intrinsic light responses) shows that three optogenetic tools drive RGC light responses with different properties (temporal kinetics and sensitivity). Each is evocative of a different type of native RGC response and confirms ON-bipolar targeting of multiple tools as a powerful approach to restore naturalistic neuro-retinal responses, central to optimising optogenetics as an approach to clinical vision restoration.

This is a 2020 ARVO Annual Meeting abstract.