Search

Retina represents a highly complex, interconnected neural circuit that performs a variety of computations starting from the first photoreceptor synapse. Retinal ganglion cells (RGCs) are commonly described as output retinal neurons that relay visual information to the brain. As they do not form chemical synapses in the retina, RGCs are viewed as upstream signal integrators. Yet, they form electrical synapses among each other and with upstream amacrine cells. To investigate these connections and their function, we incorporated gap junctional couplings into a biologically inspired cascade modeling framework that faithfully fits RGC responses to light. This model structure is based on the knowledge of retinal anatomy and physiology. Though electrical synapses convey excitatory signals between neurons, apart from excitation, our model predicted inhibitory connections among RGCs in both salamander and mouse retinas. Experimental results confirmed that such a negative feedback effect involved gap junctions and amacrine cells. As well, modeling results suggested that inhibition between RGCs modulates their response gain without affecting their visual feature selectivity. Such gain modulation was also confirmed by the experiments. Together, our finding showed that RGCs actively participate in the visual information processing by sending feedback signals into the inner retina. To foster further investigations of retinal circuit processing capabilities, we have been also developing a recording technique that will allow to monitor signal flow in the retinal circuitry. A greater understanding of neural computation can ultimately help develop retina prosthetics. As a side project, we have also contributed to the investigations of the visual loss therapeutics. A chapter is devoted to describe our on-going collaborative efforts to examine the functional role of nerve growth factor in retinal ganglion cell survival.