ABSTRACT
Brief (2-3d) monocular deprivation (MD) during the critical period induces a profound loss of responsiveness within binocular (V1b) and monocular (V1m) regions of rodent primary visual cortex. This has largely been ascribed to long-term depression (LTD) at thalamocortical synapses, while a contribution from intracortical inhibition has been controversial. Here we used optogenetics to isolate and measure feedforward thalamocortical and feedback intracortical excitation-inhibition (E-I) ratios following brief MD. Despite depression at thalamocortical synapses, thalamocortical E-I ratio was unaffected in V1b and shifted toward excitation in V1m, indicating that thalamocortical excitation was not effectively reduced. In contrast, feedback intracortical E-I ratio was shifted toward inhibition in V1m, and a computational model demonstrated that these opposing shifts produced an overall suppression of layer 4 excitability. Thus, feedforward and feedback E-I ratios can be independently tuned by visual experience, and enhanced feedback inhibition is the primary driving force behind loss of visual responsiveness.
Subject(s)
Feedback, Physiological , Neural Inhibition/physiology , Sensation/physiology , Visual Cortex/physiology , Action Potentials , Animals , Excitatory Postsynaptic Potentials , Interneurons/physiology , Long-Term Synaptic Depression , Mice, Inbred C57BL , Pyramidal Cells/physiology , Rats, Long-Evans , Sensory Deprivation/physiology , Synapses/physiology , Thalamus/physiologyABSTRACT
How are neural circuits organized and tuned to achieve stable function and produce robust behavior? The organization process begins early in development and involves a diversity of mechanisms unique to this period. We summarize recent progress in theoretical neuroscience that has substantially contributed to our understanding of development at the single neuron, synaptic and network level. We go beyond classical models of topographic map formation, and focus on the generation of complex spatiotemporal activity patterns, their role in refinements of particular circuit features, and the emergence of functional computations. Aided by the development of novel quantitative methods for data analysis, theory and computational models have enabled us to test the adequacy of specific assumptions, explain experimental data and propose testable hypotheses. With the accumulation of experimental data, theory and models will likely play an even more important role in understanding the development of neural circuits.
