Vision restoration in animal models of human blindness using natural and engineered light-gated receptors

Most inherited forms of human blindness are caused by mutations that lead to photoreceptor cell death, but spare the inner retina, providing an opportunity for treatment. With the help of azobenzene-derived chemicals that we call ‘photoswitches’, I engineered light gated receptors to be applied as therapeutics towards vision restoration in animal models of human blindness. The first part of my thesis describes this work.Next, I targeted expression of natural and engineered light-gated proteins to the remaining neurons of the retina, using viruses as gene delivery vehicles. I asked if these cells would then function as the new photoreceptors and if they would be able to drive visual responses. The quick answer is: yes, they do. In blind mice, I compared the ability of different target cells to act as new photoreceptors. Installing light sensors downstream in retinal ganglion cells lead to robust and uniform responses, whereas expression upstream in bipolar cells lead to more diverse activity patterns in response to light. I characterized mammalian proteins as optical actuators and found that light gated ion channels drive fast responses but require very high light intensities whereas G-protein coupled receptors are about 1000x more sensitive to light but at the cost of slow kinetics. I then further extended our studies to dogs and were able to show that our treatment restored light responses in blind rcd1 dog retinas in vitro and was safe and well tolerated in vivo. My results in both large and small animal models of photoreceptor degeneration provide a path to clinical translation. These findings are summarized in the second part of my thesis.Finally, I explored non-invasive approaches to restore a sense of ‘space’ and enable navigation for the blind. Towards this end, I helped design and prototype a sensory substitution device, the ‘sonic eye’. This device is inspired by bats and human echolocators and it allows users to ‘see with sound’. This work is described in the last part of my thesis