Project Summary R01 EY022638 Dissecting neural circuit computations in the peripheral visual system PI: Thomas R. Clandinin Vision provides critical sensory inputs that guide our routine behaviors; as a result, blindness represents perhaps the most devastating deprivation we can experience. To connect perception to action in this context, complex visual scenes must be efficiently represented in the neural activities of relatively small groups of cells; from these signals, particularly salient cues are extracted, integrated with behavioral goals, and linked to the appropriate responses. These neural processes can be broken down into the individual actions of relatively simple microcircuits, small groups of neurons that perform elementary operations that are widespread in the brain, but which subserve distinct purposes in different contexts. This proposal develops the Drosophila visual system as a model in which the functions of these microcircuits can be dissected at the molecular, cellular and behavioral level, and combines techniques drawn from genetics and systems neuroscience to derive new understanding. This proposal focuses on three computations that are central to vision. First, one fundamental circuit process in the visual system transforms the intensity of a light signal into an estimate of contrast, the change in light level relative to a previous intensity. This transformation corresponds to taking the mathematical derivative of an input, an operation that is performed in many circuits, but one whose circuit and molecular implementation is unknown. The first goal of this proposal is to determine how this operation is implemented at the circuit and molecular level. Second, the ability to detect motion is probably the most critical visual signal extracted by the brain, providing information central to guiding movement and navigation. The emergence of this direction-selectivity in the brain represents a long-standing, paradigmatic neural computation with rich theoretical underpinnings. However, the circuit and molecular implementations of these theories are only incompletely understood. The second goal of this proposal is to identify and dissect the microcircuits that first extract motion signals. Third, the tuning of visual neurons for oriented edges is central to representing the spatial structure of the world. Again, the mechanisms that allow neurons to become tuned for these features are only incompletely understood. The third goal of this proposal is to determine the structure and functional architecture of orientation selective circuitry. These studies will broadly inform our understanding of retinal function in health and disease. As the development of retinal prostheses that directly stimulate specific circuit elements represents an important treatment possibility for blindness, understanding how these circuits can encode behaviorally-relevant visual information represents a important goal.