Restoration of Vision with Ectopic Expression of Human Rod Opsin

Abstract

Many retinal dystrophies result in photoreceptor loss, but the inner retinal neurons can survive, making them potentially amenable to emerging optogenetic therapies. Here, we show that ectopically expressed human rod opsin, driven by either a non-selective or ON-bipolar cell-specific promoter, can function outside native photoreceptors and restore visual function in a mouse model of advanced retinal degeneration. Electrophysiological recordings from retinal explants and the visual thalamus revealed changes in firing (increases and decreases) induced by simple light pulses, luminance increases, and naturalistic movies in treated mice. These responses could be elicited at light intensities within the physiological range and substantially below those required by other optogenetic strategies. Mice with rod opsin expression driven by the ON-bipolar specific promoter displayed behavioral responses to increases in luminance, flicker, coarse spatial patterns, and elements of a natural movie at levels of contrast and illuminance (≈50-100 lux) typical of natural indoor environments. These data reveal that virally mediated ectopic expression of human rod opsin can restore vision under natural viewing conditions and at moderate light intensities. Given the inherent advantages in employing a human protein, the simplicity of this intervention, and the quality of vision restored, we suggest that rod opsin merits consideration as an optogenetic actuator for treating patients with advanced retinal degeneration.

Keywords: rhodopsin.

Figure 1

Ectopic Expression of Human Rod Opsin Restores Light Responses in rd¹ Mouse Retina (A) Schematic of the DNA expression cassette delivered by AAV2/2 vector to the retina. A human rod opsin coding sequence (RHO) is driven by a hybrid CMV enhancer/chickenβ-actin (CAG) promoter. The sequence is flanked by inverted terminal repeats (ITRs) and stabilized by a polyadenylation signal sequence (polyA) and a woodchuck hepatitis posttranscriptional regulatory element (WPRE). (B and C) Exemplar images of a section through an rd¹ mouse retina >4 months after intravitreal delivery of vector in (A) in conjunction with glycosidic enzymes. Expression of human rod opsin in cells of the ganglion cell layer (GCL) and inner nuclear layer (INL) and processes in the inner plexiform layer (IPL) are revealed by staining with an α-hRho antibody (red) and counterstaining of nuclei with DAPI (blue) to aid orientation (B). A monochrome version of α-hRho antibody staining in (B) in which rod opsin expression appears in white is shown in (C). Calibration bar = 50 μm. (D and E) Perievent rasters and associated perievent firing rate histograms (PSTHs) for eight representative single units isolated from multi-electrode array (MEA) recordings of rd¹-CAG-RHO retinas without (D) and with (E) exogenous 9-cis-retinal. Each set of rasters depicts spiking activity for 20 sequential presentations of a 2-s white light flash (4 × 10¹⁴ rod photons/cm²/s; interstimulus interval 20 s) starting at time 0. PSTHs below depict mean firing rate in 100-ms epochs across all 20 repeats. In both conditions, units show increases in firing associated with light presentation (from 0 to 2 s), but these are most pronounced for the first few trials (lower traces in raster) in (D), indicating bleaching, while inclusion of 9-cis-retinal (E) renders them repeatable across many trials. (F and G) Heatmap representations of mean firing rate across at least 20 presentations of 2-s stimulus (ON at time 0) for 104 units from 5 rd¹-CAG-RHO mice (F) and six units from three control rd¹-CAG-GFP mice (G) meeting an objective criterion of stimulus-associated change in firing. Color code represents normalized firing rate (−1 and 1 being minimum and maximum firing rate for that unit, respectively). Traces are ordered according to response latency. (H) Population mean (±SEM) normalized firing rate profiles for rd¹-CAG-RHO units grouped according to response latency (horizontal white lines in F delineate extent of clusters). (I) Mean ± SEM normalized firing rate (mean firing rate from −2 s to 6 s was normalized to maximum and minimum, and the normalized pre-stimulus firing rate (−2 to 0 s) was then subtracted) for all light-responsive units exposed to 2-s pulses (starting at 0 s) at 4 × 10¹⁴, 4 × 10¹³, and 4 × 10¹² rod photons/cm²/s. (J and K) Distribution of response amplitudes (J; mean change in firing rate) and latencies (K; mean time at which mean firing rate first fell outside 2 SDs of baseline firing) for units in (F) responding with increases (excit’n) or decreases (inhib’n) in firing at 4 × 10¹⁴, 4 × 10¹³, and 4 × 10¹² rod photons/cm²/s. (L) Perievent rasters for three single units showing firing of three units across multiple repeats of a 2-s light pulse (4 × 10¹⁴ rod photons/cm²/s) without (above) and with (below; shaded in green) application of GABA receptor antagonists (TPMP 25 μM and picrotoxin 50 μM).

Figure 2

Rod Opsin Expression Driven by the Ubiquitous CAG Promoter Restores Light Responses in Blind rd¹ Mouse Thalamus (A) Schematic of recording apparatus allowing presentation of separate light stimuli to each eye and insertion of silicone multi-channel recording electrode probes to the dorsal lateral geniculate nuclei (dLGNs) in either hemisphere. Representative histological sections through the left and right dLGN with DiI tracks (in red) showing path of insertion for recording probes. (B and C) Heatmap representations of mean firing rate across multiple presentations of 2-s stimulus (ON at time 0) to rd¹-CAG-RHO (B) and control rd¹-CAG-GFP (C) eyes of units showing a significant change in firing associated with stimulus presentation (n = 31 units downstream of 5 treated eyes and n = 10 units downstream of 5 control eyes). Color code represents normalized firing rate (−1 and 1 being minimum and maximum firing rate for that unit, respectively). Traces are ordered according to response latency. (D) Sensitivity response profile (perievent rasters and associated perievent firing rate histograms) for two representative dLGN single units isolated from (B) at three different retinal irradiances: 8 × 10¹³, 8 × 10¹², and 8 × 10¹¹ rod-equivalent photons/cm²/s. (E) Light-adapted responses (perievent rasters and associated perievent firing rate histograms) for two representative dLGN units from rd¹-CAG-RHO eyes recorded under light-adapted conditions (retinal irradiance 8 × 10¹³ rod-equivalent photons/cm²/s and Michelson contrast 96%). (F and G) Distribution of response latencies (F; time at which mean firing rate first fell outside 2 SDs of baseline for units responding within 2.5 s of lights on) and amplitude (G; mean change in firing rate) for units in (B) responding with increases (excit’n) or decreases (inhib’n) in firing. CAG is a hybrid CMV enhancer/chickenβ-actin promoter. RHO is human rod opsin coding sequence.

Figure 3

Selective Expression of Rod Opsin Using a Cell-Specific grm6 Promoter Restores Visual Responses in the dLGN of rd¹ Mice (A) Schematic of the DNA expression cassette delivered by AAV2/2 vector to the retina, comprising RHO under the ON-bipolar cell-specific (grm6) promoter flanked by ITRs and stabilized by polyA and WPRE. (B) Exemplar image of a section through an rd¹ mouse retina >4 months after intravitreal delivery of viral vector in (A) in conjunction with glycosidic enzymes. Expression of human rod opsin in cells of the INL and processes in the IPL are revealed by staining (red) with an α-hRho antibody and counterstaining of nuclei with DAPI (blue). Calibration bar = 50μm. (C) Heatmap representations of mean firing rate across multiple presentations of 2-s stimulus (ON at time 0) for 30 single retinal units from two rd¹-grm6-RHO mice showing a significant change in firing associated with stimulus presentation. Color code represents normalized firing rate (−1 and 1 being minimum and maximum firing rate for that unit, respectively). Traces are ordered according to response latency. (D and E) Distribution of response latencies (D; time at which mean firing rate fell outside 2 SDs of baseline for units responding within 2.5 s of lights on) and amplitude (E; mean change in firing rate) for units in (C) responding with increases (excit’n) in firing. (F) Sensitivity response profile (perievent rasters and associated perievent firing rate histograms) for two representative retinal single units isolated from (C) at two different retinal irradiances: 4 × 10¹⁴and 4 × 10¹² rod-equivalent photons/cm²/s. (G) Perievent rasters for two single units showing inhibition of excitatory responses after application of GABA receptor antagonists (TPMP 25 μM and picrotoxin 50 μM; lower part of raster plots shaded in green). (H) Heatmap representations of mean firing rate across multiple presentations of 2-s stimulus (ON at time 0) for 73 single dLGN units from rd¹-grm6-RHO eyes showing a significant change in firing associated with stimulus presentation. Color code represents normalized firing rate (−1 and 1 being minimum and maximum firing rate for that unit, respectively). Traces are ordered according to response latency. (I and J) Distribution of response latencies (I; time at which mean firing rate fell outside 2 SDs of baseline for units responding within 2.5 s of lights on) and amplitude (J; mean change in firing rate) for units in (C) responding with increases (excit’n) or decreases (inhib’n) in firing. (K) Sensitivity response profile (perievent rasters and associated perievent firing rate histograms) for representative dLGN single units isolated from (H) at three different retinal irradiances: 8 × 10¹³, 8 × 10¹², and 8 × 10¹¹ rod-equivalent photons/cm²/s.

Figure 4

Ectopic Expression of Rod Opsin Restores Visual Behavior in Blind rd¹ Mice (A) Open box activity plots for freely moving mice with LCD screens switched from “black” to “white” at time 5 min (illuminance 40 lux; estimated retinal irradiance 1 × 10¹² rod-equivalent photons/cm²/s). (B) Open box activity plot for rd¹-grm6-RHO mice exposed to 4-Hz flicker starting at 5 min (illuminance 20 lux; estimated retinal irradiance 8 × 10¹¹ rod-equivalent photons/cm²/s. (C and D) Histograms of activity for rd¹-grm6-RHO mice showing distance traveled in 30 s before (black bars) and 30 s after (white bars) presentation of “white” screen at different flicker frequencies (C) and at 4-Hz flicker at different contrast ratios (D). (E) Representative movement trajectories for a wild-type and two different rd¹-grm6-RHO mice in the open field box in the 30 s before (left) and 30 s after (right) presentation of gratings. (F) Histogram of activity for wild-type mice showing distance traveled in 30 s before (black bars) and 30 s after (white bars) presentation of drifting squarewave gratings (contrast ratio 1:8) at different spatial frequencies. (G) Histogram of change in activity in response to two different spatial frequencies (0.04 and 0.08 cpd) for rd¹-grm6-RHO mice. Sample sizes for data in (A)–(D) are five wild-type, six rd¹-CAG-GFP, six rd¹-CAG-RHO, and five rd¹-grm6-RHO mice; in (F) eight wild-type; in (G) nine rd¹-grm6-RHO. In all panels, activity is represented by mean ± SEM of the mean distance traveled by each animal in a 30-s time bin; time in min since introduction to testing arena. Two-tailed paired t tests comparing activity before and after stimulus appearance (^∗p < 0.05, ^∗∗p < 0.01). For Figures 4B and 4C, two-way RM ANOVA; p < 0.0001 for interaction between flicker frequency and gray versus flicker, post hoc Bonferroni correction p < 0.05 for gray versus flicker at 4 and 10 Hz. For Figure 4F, two-way RM ANOVA; p < 0.01 for gray versus gratings, post hoc Bonferroni correction p < 0.05 at 0.1 and 0.4 cpd.

Figure 5

Rod Opsin Restores Visual Behavior in Response to Natural Scenes (A and B) Perievent rasters and associated perievent firing rate histograms for a dLGN unit to multiple presentations of a 30-s naturalistic movie (mice moving in an open arena in horizontal view; mean estimated retinal irradiance 1 × 10¹³ rod-equivalent photons/cm²/s) to an rd¹-grm6-RHO eye. (A) and (B) show presentations of the high-contrast movie (HCM; black:white contrast ratio ≈ 1:100) and low-contrast movie (LCM; contrast ratio reduced 1:50), respectively. Horizontal line on histograms shows the 99% confidence interval for firing rate across the movie presentation; note the increase in firing above this line at the same time point for both movie presentations. (C) Firing pattern of a representative dLGN unit from a wild-type mouse exposed to the HCM is presented for comparison. (D) Example frames from a naturalistic movie featuring a swooping owl presented to mice in a behavioral arena. (E) Open box activity plots for rd¹-grm6-RHO mice presented with a naturalistic swooping owl movie starting at 5 min (shaded in green; estimated retinal irradiance 8 × 10¹¹ rod-equivalent photons/cm²/s). (F) Histogram of activity (mean ± SEM distance traveled by each animal) for rd¹-CAG-GFP (n = 6), rd¹-CAG-RHO (n = 6), rd¹-grm6-RHO (n = 5), and wild-type (n = 10) mice showing distance traveled in 30 s before (black bars) and after (white bars) presentation of the swooping owl movie. Two-tailed paired t tests comparing activity before and after stimulus appearance (^∗∗p < 0.01).