Neural circuits for visual feature detection

A significant percentage of people in the US suffer from disabilities resulting from traumatic injury, stroke, or degenerative disease which cause deficits in visual perception. Therefore, an understanding of the basic circuit neurobiology and neuromodulation of feature-based visual perception will sharpen our understanding of the mechanism of visual processing, and should facilitate the development of treatments for these disabilities. One form of visual attention is targeted to salient features of the visual scene, during which a subject detects optical disparities that distinguish a salient object from the cluttered visual surroundings. The cell-circuit mechanism for how this occurs is not well understood. This renewal project will capitalize on recent discoveries and significant experimental advantages of the fruit fly Drosophila to explore the elementary neural circuitry required for feature-based visual attention. The fly has a numerically simple nervous system, with which highly advanced genetic techniques can be used to identify, manipulate, and repeatedly record the activity of individual neurons, as well as their upstream and downstream network partners. The fly also displays robust feature-based visual perception, even under stimulus conditions that defeat classical models of motion vision, for which similar processes have been localized to cortical pathways in humans and non-human primates. The PI hypothesizes that flies detect and discriminate the higher-order features of visual objects with specialized circuits that integrate first-order elementary motion signals retinotopically, which are further enhanced by the action of inhibitory neurotransmitters. The PI will perform two-photon Ca2+ imaging of candidate cellular pathways in response to stimuli the PI has discovered to elicit robust feature detection by flies within a virtual reality flight simulator. Armed with physiological receptive fields, the PI will use live imaging to ‘read’ and optogenetics to ‘write’ activity patterns in a behaving fly to directly observe input-output functions of feature detection processing on visual behavior. Finally, the PI will study how biogenic amines modulate the functional properties of feature detecting neurons to enable plasticity required for visual feature detection in switching behavioral contexts such as the transition from stationary quiescence to active locomotion.