Severe photoreceptor cell death in retinal degenerative diseases leads to partial or complete blindness. Optogenetics is a promising strategy to treat blindness. The feasibility of this strategy has been demonstrated through the ectopic expression of microbial channelrhodopsins (ChRs) and other genetically encoded light sensors in surviving retinal neurons in animal models. A major drawback for ChR-based visual restoration is low light sensitivity. Here, we report the development of highly operational light-sensitive ChRs by optimizing the kinetics of a recently reported ChR variant, Chloromonas oogama (CoChR). In particular, we identified two CoChR mutants, CoChR-L112C and CoChR-H94E/L112C/K264T, with markedly enhanced light sensitivity. The improved light sensitivity of the CoChR mutants was confirmed by ex vivo electrophysiological recordings in the retina. Furthermore, the CoChR mutants restored the vision of a blind mouse model under ambient light conditions with remarkably good contrast sensitivity and visual acuity, as evidenced by the results of behavioral assays. The ability to restore functional vision under normal light conditions with the improved CoChR variants removed a major obstacle for ChR-based optogenetic vision restoration.

In axon-bearing neurons, action potentials conventionally initiate at the axon initial segment (AIS) and are important for neuron excitability and cell-to-cell communication. However in axonless neurons, spike origin has remained unclear. Here we report in the axonless spiking AII amacrine cell of the mouse retina a dendritic process sharing organizational and functional similarities with the AIS. This process was revealed through viral-mediated expression of channelrhodopsin-2-GFP (ChR2-GFP) with the AIS-targeting motif of sodium channels (NavII-III). The AII-processes showed clustering of voltage-gated Na+ channel 1.1 (Nav1.1) as well as AIS markers ankyrin-G and neurofascin. Furthermore, NavII-III targeting disrupted Nav1.1 clustering in the AII-process which drastically decreased Na+ current and abolished the ability of the AII amacrine cell to generate spiking. Our findings indicate that despite lacking an axon, spiking in the axonless neuron can originate at a specialized AIS-like process.

The ectopic expression of microbial opsin-based optogenetic sensors, such as channelrhodopsin-2 (ChR2) in surviving inner retinal neurons, is a promising approach to restoring vision after retinal degeneration. However, a major limitation in using native ChR2 as a light sensor for vision restoration is the low light sensitivity of its expressing cells. Recently, two ChR2 mutations, T159C and L132C, were reported to produce higher photocurrents or have ultra light sensitivity. In this study, we created additional ChR2 mutants at these two sites to search for more light responsive ChR2 forms and evaluate their suitability for vision restoration by examining their light responsive properties in HEK cells and mouse retinal ganglion cells. We found additional ChR2 mutants at these two sites that showed a further increase in current amplitude at low light levels in the cells expressing these mutants, or operational light sensitivity. However, the increase in the operational light sensitivity was correlated with a decrease in temporal kinetics. Therefore, there is a trade-off between operational light sensitivity and temporal resolution for these more light responsive ChR2 mutants. Our results showed that for the two most light responsive mutants, L132C/T159C and L132C/T159S, the required light intensities for generating the threshold spiking activity in retinal ganglion cells were 1.5 and nearly 2 log units lower than wild-type ChR2 (wt-ChR2), respectively. Additionally, their ChR2-mediated spiking activities could follow flicker frequencies up to 20 and 10 Hz, respectively, at light intensities up to 1.5 log units above their threshold levels. Thus, the use of these more light responsive ChR2 mutants could make the optogenetic approach to restoring vision more feasible.