UCLA

Genetic engineering in mammals is furthest developed in Mus musculus and has facilitated great strides towards understanding the molecular and cellular mechanisms underlying human biology and disease. Despite the advantages afforded through genetic manipulations, studies involving retinal photoreceptors have been largely constrained to rods due to the technical challenges of isolating cones. This dissertation describes the methodology I have developed to reliably identify unlabeled mouse cone somata, and using whole-cell patch clamp, record their conductance and voltage changes in response to diverse stimuli.

I made highly resolved measurements of cone dark current, membrane capacitance, and resting membrane potentials. Photoresponses were recorded with brief and steady-light stimulation protocols, and I characterized parameters describing response sensitivity and kinetics. Wild-type cones showed evidence of lateral electrical coupling to rod photoreceptors (i.e. the rod secondary pathway). The loss of calcium- sensitive proteins affected waveform of the responses to brief and steady-light stimulation. Notably, when compared to controls, the loss of guanylyl cyclase accelerating proteins caused cones to be more sensitive to a given light stimulus and to reopen fewer light-sensitive channels.

The ability to manipulate the cone membrane potential enabled the biophysical characterization of multiple inner segment ion conductances. Synaptic calcium and the hyperpolarization-activated rectifying conductances were isolated with pharmacology. By combining voltage stimulation with light flashes, I also studied the reversal potential of the light response for the first time in any mammalian species. The inner segment calcium- and calcium-activated anion conductances were significantly large in the mouse cones and had to be blocked in order to isolate the light-sensitive conductance.

Unlike rods, cones must remain active in brighter ambient light. In addition to calcium-dependent adaptation that adjusts the phototransduction machinery, the cone must replenish large fractions of photopigment. We identified a light-driven, non- enzymatic pathway in which all-trans-retinal does not even leave the photoreceptor. We show that when bound to a retinyl phosopholipid-complex in the membrane and exposed to blue light, all-trans-retinal can be preferential photoconverted back to 11-cis- retinal.

All the mechanisms employed to keep cones functional and out of saturation are for naught if well defined synaptic connections are not made properly. We show that cones lacking a synaptic adaptor protein LRIT1 responded to light with normal characteristics. Cone bipolar cells, however, responded to light flashes with altered sensitivities. Thus, visual perception relies not only on high-fidelity encoding of light stimuli, but also on precise circuitry and signal transmission.