Neural Circuits That Process Visual Information

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Two powerful inhibitory networks in the visual thalamus converge on relay cells and influence every spike that travels downstream. Local interneurons provide feedforward inhibition to relay cells and each other. The thalamic reticular nucleus receives input from relay cells and inhibits them in return. Work in fixed tissue or brain slices has provided insight into the pharmacology, cellular physiology and anatomy of these circuits. It is patently necessary apply these ex vivo results to function in vivo. To bridge this gap, we record from inhibitory cells directly and monitor the inhibition they generate in relay cels during vision. Our strategy updates classical comparative anatomical and physiological approaches by combining whole-cell recording and intracellular labeling in vivo with theory and computational techniques. Aim 1) Exploring the integration of On and Off pathways in the LGN. Relay cells have receptive fields made of concentric On and Off subregions with a push-pull layout of excitation and inhibition; e.g. where bright stimuli excite, dark inhibit. Retina supplie the push (excitation). We propose that the pull (inhibition to stimuli of the reverse sign) comes from local interneurons with receptive fields like those of their postsynaptic partners, but with te opposite preference for stimulus polarity. It is difficult, however, to map connectivity between and On and OFF cells because these cannot be anatomically distinguished in most mammals. Thus, we will test our hypothesis by using the ferret, where On and Off cells occupy different sublaminae in the LGN. Aim 2) Model systems to explore inhibitory mechanisms in higher animals. Genetic approaches make rodent a popular subject for studying vision. However, the cortical organization of carnivore vs. rodent is vastly different, from the level of the functional architecture to properties of single cells. We ask where these differences emerge by quantitatively comparing the synaptic structure of receptive fields in carnivore vs. rodent LGN. Preliminary studies suggest that basic principles of processing in the LGN are conserved. Thus, we will probe push-pull using mutants lacking an On channel. Further, interneurons and relay cells in cat process their inputs in quantitatively different ways that optimize information transmission; we will dissect the bases for these differences in rodent. Aim 3) Inhibitory contributions to processing stimulus contrast. We hypothesize that push-pull and same-sign inhibition (inhibition to the preferred stimulus polarity) expand the range of sensitivity to stimuus contrast and improve feature detection at high contrasts. We will explore extra-retinal mechanisms of contrast gain by comparing retinal input to thalamic output patterns in relay cells and by recording from interneurons. Push- pull vs. same-sign inhibition will be separated empirically by silencing the On channel, and computationally with conductance based models. In addition, we will ask how inhibition contributes to feature selectivity by assessing changes in the relative weights of push and pull at different contrasts.