Studies of the visual system face a number of challenges, two of which are the intricacy of the cell types
and synaptic connections that comprise the nervous system, and the complexity of the computational
processes that underlie vision. Although the retina is one of the most characterized and well understood neural circuits of the visual system, it nonetheless has a great diversity of cell types, connections and computations. The normal function of the retina is to convey information about natural visual scenes, which have complex spatial and temporal structure. The processing of natural scenes has the greatest relevance towards a fundamental understanding retinal function, and the greatest clinical relevance. Yet most studies of retinal visual processing and circuitry focus on responses to simple artificial stimuli rarely encountered normally, such as flashing spots, drifting stripes and flickering checkerboards. With respect to retinal cell types greatest diversity lies in a class of inhibitory interneurons known as amacrine cells. These cells make extensive lateral and feedback connections, and although they form stereotyped connections between each other, excitatory bipolar cells, and ganglion cells that transit signals in the optic nerve, the functional effects of nearly all of these cell types are poorly understood. This proposal aims towards a direct characterization of the functional effects of amacrine cells under ethologically relevant stimuli, including natural scenes. We combine approaches of perturbation and recording using electrical and optical methods as well as computational modeling to characterize the specific contributions of amacrine cells to stimuli that include the representation of moving objects. We take advantage of recently developed computational approaches that
can simultaneously capture the retinal response to a broad range of stimuli including natural scenes, capture a wide range of phenomena previously characterized only with artificial stimuli, and that have internal units highly correlated with retinal interneurons. Our goals are to 1) Create a quantitative understanding of the functional contributions of a class of sustained amacrine cells in the salamander retina for specific stimuli including those that represent moving objects and natural scenes, and test hypotheses related to dynamic effects on visual sensitivity and sensory features generated by those amacrine cells 2) Use molecularly defined amacrine cells in the mouse to quantitatively characterize the functional contribution of specific amacrine cell types to specific stimuli including artificial moving objects and natural scenes. These studies create a new way to generate and test hypotheses related to the quantitative effect of any interneuron on retinal output under any visual stimulus. Understanding how retinal circuitry creates visual processing under natural scenes is critical to our understanding of retinal mechanisms and diseases involving the degeneration of the retinal circuitry. In addition, the computational descriptions of retinal responses will be directly useful in the design of electronic retinal prosthesis systems.