PROJECT SUMMARY: Thalamic circuits are dominated by two sources of inhibition that have a profound influence on the type and quantity of information that relay cells transmit from eye to brain: These are local interneurons within the dorsal lateral geniculate nucleus of the thalamus (dLGN), and the visual sector of the thalamic reticular nucleus (TRN)— a thin sheet of GABAergic cells that lies nearby. Local interneurons receive retinal input and synapse with relay cells and each other to supply powerful feedforward inhibition. By contrast, relay cells make only sparse connections within the main layers of dLGN. Rather, they contact neurons in TRN, whose dense axonal arbors provide feedback inhibition in return. In addition to receiving ascending information, these three types of cells (local interneurons, relay cells, reticular cells) in the dLGN/TRN complex are embedded in a larger network that involves top-down input from cortex; this arrangement is repeated across primary thalamic nuclei in mammals. Thus, learning how thalamic circuits operate is key to understanding sensory integration and, moreover, serves studies of disorders such as amblyopia or the development of visual prosthetics by providing a blueprint for how healthy brains function. Here we focus on TRN in mouse, a species that has become central to studies of vision because of the many experimental advantages it offers, but whose visual system differs somewhat from those of traditional experimental subjects like carnivore and primate. The design of the project is inspired by a framework voiced by Francis Crick, who theorized that TRN might act as a searchlight that increases thalamic activity in specific regions of interest, or as a thermostat that regulates levels of global activity. We evaluate predictions of each hypothesis to explore visual processing per se, by combining comparative, optogenetic, physiological, anatomical, and computational approaches. The project comprises three interrelated aims, as follows. The searchlight hypothesis suggests that receptive fields in TRN are feature specific and localized and we have shown that this is the case in carnivore; Aim 1 uses physiological and computational approaches to analyze the spatiotemporal features encoded in mouse visual TRN across visual space. These results are then incorporated into a model framework that ties the output of inhibitory cells, including TRN and local interneurons (we have studied these previously) to patterns of inhibition recorded from the relay cell’s receptive field. Aim 2 explores the link between receptive field structures in dLGN and TRN mechanistically by exploring how input from dLGN influences visual response properties in TRN and vice versa; opsins that suppress synaptic transmission and statistical tools that reveal connectivity between neurons are used to approach this topic. Finally, Aim 3 takes the general perspective of the thermostat hypothesis and asks how responses of TRN to global stim...