UC Berkeley

The retina in our eyes is full of tiny neuronal computers, cells with diversely elaborate dendritic arbors that transform, sort, and integrate information about light bouncing off of objects in the world. The focus of this dissertation was to understand how dendrites of one type of cell in the retina, called the starburst amacrine cell, perform computations on visual information. The starburst amacrine cell is a part of the retina’s “direction selective circuit”, a model system for studying the cellular and synaptic implementation of a neural computation. Within this circuit, the starburst amacrine cells are responsible for detecting the direction of motion in the environment—right, left, up, or down—and cueing downstream cells, which pass on this information about motion direction to the brain. To begin, Chapter I presents the recent advances in our understanding of how the direction selective circuit works. Then, the experimental investigations presented in this dissertation attempt to clarify three issues about starburst amacrine cells. First, I examined the precise role of starburst amacrine cells in the direction selective circuit by altering their response properties and investigating how the performance of the circuit changes (Chapter II). I found that perturbations to the excitatory inputs to starburst amacrine cells using either visual adaptation of the retina or a chemo-genetic manipulation substantially altered the computational output of the entire circuit. Second, I described how the cells’ excitatory synaptic inputs are organized and contribute to the cells’ computations (Chapter III). I discovered that the precise distribution of excitatory inputs to the starburst amacrine cells shapes the cells’ ability to detect motion direction. Because of the elaborate structure of the dendrites, this precise arrangement is necessary for the electrical signals coming at different times and onto different dendritic locations to be integrated together in a direction-dependent fashion. Third, I investigated the development of the excitatory inputs to starburst amacrine cells and starburst amacrine cells’ velocity tuning (Chapter IV). These preliminary experiments provide a framework for further inquiry in these areas. Together, these experiments reveal how anatomical wiring and electrical properties of neurons in the retina produce exquisite computational abilities.