Advantages of human opsins in optogenetic visual restoration

Mihai Teodor Bica^1*Jasmina Cehajic-Kapetanovic^1,2

• ¹Oxford University Hospitals NHS Trust, Oxford, United Kingdom
• ²Nuffield Laboratory of Ophthalmology, Oxford, United Kingdom

Optogenetic vision restoration has progressed from proof-of-concept to early clinical testing, yet most programmes rely on microbial channels that demand high irradiance and offer limited adaptation. This review synthesises preclinical and clinical evidence comparing microbial actuators with human opsins (rhodopsin, cone opsins, melanopsin) and outlines vector and safety considerations for translation. Human opsins activate G-protein–coupled cascades, providing intrinsic signal amplification and operation at room-light levels (~1011–1012 photons·cm⁻²·s⁻¹), in contrast to the ≥1015 photons·cm⁻²·s⁻¹ typically needed for channelrhodopsins. Rhodopsin and MW cone opsin preserve photopic-range sensitivity (rhodopsin > cone opsin) while delivering millisecond-scale kinetics and adaptation across backgrounds, enabling patterned retinal responses without optical intensification devices; clinical validation without external intensification is pending. Such mammalian pigments also confer bleaching-based light adaptation, whereas microbial tools are photocyclic and can desensitise under steady illumination, limiting sustained contrast encoding. Bistable melanopsin enables durable irradiance coding but with slow dynamics; chimeric designs (e.g., melanopsin–mGluR6, Gloeobacter–human rhodopsin) aim to combine amplification with favourable reset properties. In contrast to human opsins, microbial channels warrant safety considerations including light-dose budgeting (particularly at short wavelengths), potential cytotoxicity from proton or calcium loads, and vector-related ocular inflammation; red-shifted actuators improve photochemical safety margins. Targeting opsins to ON bipolar (ON-BP) cells retains inner-retinal computations (centre–surround, ON/OFF segregation, temporal filtering). Engineered adeno-associated virus (AAV) capsids (e.g., AAV2-7m8 intravitreally; AAV8.BP2 subretinally) paired with GRM6 or L7 promoters achieve broad ON-BP expression in rodents but a much more limited expression profile in non-human primates. First clinical studies report acceptable early ocular safety with emerging efficacy signals. We propose accelerating phase I safety human trials of human-opsin vectors with prospectively defined light-exposure budgets and low vision functional endpoints such as navigation, face and object recognition, temporal contrast sensitivity, alongside work on chromophore support, cascade integrity in late degeneration, and scalable vector–promoter solutions. Pharmacological noise suppression in degenerating retinas (e.g., gap-junction blockers or retinoic-acid pathway modulators) may further enhance signal-to-noise without altering opsin biochemistry. Together, these steps can move human-opsin optogenetics from experimental promise to clinically meaningful restoration of light sensitivity.