Current Projects

1. At the first synaptic stage, a cone photoreceptor makes synapses with 10 - 12 cone bipolar cell types, each with distinct visual responses. For example, a cell may respond to either light increments (‘ON’) or decrements (‘OFF’), either briefly (‘transient’) or continuously (‘sustained’). This functional diversity is critical for the formation of about twenty distinct visual representations at the level of the retinal ganglion cells, which selectively encode contrast polarity, size, and color, the presence of edges, and visual motion. While the response properties of bipolar cells depend in part on the glutamate receptors expressed on their dendrites (for example, ON bipolar cells express mGluR6 receptors, whereas OFF bipolar cells express kainate receptors), increasing evidence suggests that bipolar cell properties strongly depend on interactions at the other end of the cell – the axon terminal - where bipolar cells receive inhibitory inputs from amacrine cells. Our goal is to determine which bipolar cell properties are generated at the level of the dendrites in the outer retina, and which at the axon terminal in the inner retina. We use advanced imaging with genetically targeted fluorescent biosensors during visual stimulation of the retina addresses this question. An important new insight from these imaging studies is that parallel pathways are not strictly parallel: ON-type bipolar cells, through cross-over inhibition, strongly influence the OFF-type bipolar cells. The goal of the current experiments is to understand the extend of this cross-over signaling, and the properties that it bestows on OFF bipolar cell pathways.

2. More than fifty years ago, Horace Barlow and colleagues discovered that the mammalian retina contains ganglion cells that respond selectively to visual motion in a particular direction. Solving the neural mechanisms underlying this direction selectivity has been the focus of intense study, not only as a key example of retinal signal processing, but also more generally, as an example of detection of spatio-temporal patterns - a task solved in neural circuits throughout the brain. The origin of direction selectivity has been located unambiguously to the dendrites of a particular amacrine cell type, the starburst amacrine cell (SAC). SACs come in two types ('ON', activated by light increments, and 'OFF', activated by light decrements) and are directly presynaptic to the direction selective ganglion cells. Now the question is: what makes SAC dendrites directionally selective? While a dendrite-dependent mechanism has been proposed for the ON SACs, a recent study based on EM reconstruction predicts a dendrite-independent mechanism for the OFF SACs. We use two-photon fluorescence imaging and targeted electrophysiology to explore the spatial organization of synaptic inputs onto ON and OFF SACs and their selective responses to visual motion under a variety of conditions, to test these two alternative models for direction selectivity at the level of the SAC dendrites, and to explain how a specific computation is performed within a defined retinal neural circuit.