The role of behavioral relevance for sensory processing in the mouse visual system

Abstract

Sensory processing does not only reflect the physical properties of a given stimulus, but can be affected by the behavioral relevance of visual information. These changes in sensory processing can enhance responses to stimuli, which are relevant for behavior, and suppress activity to irrelevant noise. Task-related modulations in sensory processing have been fundamentally investigated in monkey studies, though, more sophisticated experimental methods have pushed towards investigating neural functions of perception in the mouse. An increasing number of studies has documented state-dependent changes in responses of neurons in visual cortex, but how sensory processes change when stimuli become behaviorally relevant is not well understood. In this study I was interested in how the behavioral context of a visual stimulus changes neuronal processing in the early visual system of the mouse. To address this question, I designed a visual foraging task, in which mice learned to discriminate between two stimuli, which provided identical sensory stimulation, but differed in reward contingencies. After mice learned to perform the task, I obtained electrophysiological recordings from the primary visual cortex (V1) to test if behavioral relevance of the stimulus changes firing rates. In a passive viewing condition, I measured sensory responses to the two stimuli outside the context of the task. These measurements were used to exclude from further analyses all neurons that showed any difference in their sensory response to the two stimuli. I found that of the recorded V1 population which did not distinguish between the two stimuli during passive viewing, 25% of neurons signaled stimulus identity during task engagement. To further investigate if these changes arise in cortex or might be inherited from their main input area, the lateral geniculate nucleus (dLGN), I recorded from the dLGN during the same task. As in V1, neurons in dLGN showed changes in firing rates when stimuli became behaviorally relevant. However, the subset of neurons affected was low (10%). Knowing that stimulus context can shape response properties in mouse V1, I further investigated the role of intracortical inhibition. Narrow-spiking, putative inhibitory, and broad-spiking, putative excitatory, neurons exhibited similar strengths of changes in firing rates when stimuli became behaviorally relevant during task engagement. These findings show that the visual foraging task can be used to manipulate behavioral relevance while keeping the sensory input for single neurons constant. These manipulations can affect sensory processing within the mouse early visual system, including inhibition by putative inhibitory interneurons.