Visual processing mode switching regulated by VIP cells

Abstract

The responses of neurons in mouse primary visual cortex (V1) to visual stimuli depend on behavioral states. Specifically, surround suppression is reduced during locomotion. Although locomotion-induced vasoactive intestinal polypeptide positive (VIP) interneuron depolarization can account for the reduction of surround suppression, the functions of VIP cell depolarization are not fully understood. Here we utilize a firing rate model and a computational model to elucidate the potential functions of VIP cell depolarization during locomotion. Our analyses suggest 1) that surround suppression sharpens the visual responses in V1 to a stationary scene, 2) that depolarized VIP cells enhance V1 responses to moving objects by reducing self-induced surround suppression and 3) that during locomotion V1 neuron responses to some features of the moving objects can be selectively enhanced. Thus, VIP cells regulate surround suppression to allow pyramidal neurons to optimally encode visual information independent of behavioral state.Visual perception, an internal model of external environment, does not merely reflect exogenous stimuli. Instead, it depends on various endogenous contexts. Consider a number of striking studies in mouse visual cortex that suggest that contextual information originating from other cortical areas modulates primary visual cortex (V1) neuron responses by way of vasoactive intestinal polypeptide positive (VIP) interneurons [1][2][3][4] . For instance, the cingulate area (Cg), which modulates the gain of V1 neurons, induces excitatory postsynaptic potentials in VIP cells 1 as it occurs during locomotion 2 . Thus, it is imperative to comprehend how VIP cells contribute to contextual modulation of V1 neuron responses.VIP cells, one of the major inhibitory cell types in neocortex 5,6 , are commonly found in superficial layers 7 . They preferentially inhibit somatostatin positive (SST) cells that mediate surround suppression 8,9 . That is, depolarized VIP cells disinhibit pyramidal (Pyr) cells by lowering surrounding suppression. This disinhibition, in fact, accounts for the reduction of surround suppression during locomotion 2,10 . However, it remains unclear why surround suppression is reduced during locomotion. When an animal moves forward, the entire scene, including all objects, appears to move backward (optical flow). When the image of an object moves over the retina, it stimulates multiple receptive fields. As the center of one receptive field constitutes the surround of nearby receptive fields, this motion can induce surround suppression among these cells, a phenomenon we refer to as self-induced surround suppression. Thus, the responses of visual-selective neurons to object motion will depend on the strength of self-induced surround suppression.During locomotion, surround suppression in V1 can become too strong for V1 neurons to respond properly to visual stimuli, as all objects are in relative motion. Thus, we hypothesize that VIP cells are depolarized to reduce such surround suppression which may be ...

Cortical circuits implement optimal context integration

“…This could be accomplished by a third form of inhibition mediated by VIP interneurons, in which there is increased activity of VIP interneurons and reduced activity of SST cells. This would effectively result in dis-inhibition of the surround, thereby allowing better processing of movement [9,17]. When the non-linearity g is taken to be identity (g = 1), it also predicts that inputs from lateral connections should combine multiplicatively with the feedforward input to drive a neuron's activity (thus affecting its gain).…”

Amplifying the redistribution of somato-dendritic inhibition by the interplay of three interneuron types

“…The present study provides a hypothesis for one candidate role, which may well be only one in a broad repertoire of functions performed in parallel. Computational models may offer a useful resource to understanding this functional repertoire (Hayut et al, 2011;Litwin-Kumar et al, 2016;Yang et al, 2016;El-Boustani and Sur, 2014;Lee and Mihalas, 2017), given that they offer a degree of control over the circuit that is hard to achieve in experimental setups.…”