State-dependent changes in perception and coding in the mouse somatosensory cortex

Abstract

An animals behavioral state is reflected in the dynamics of cortical population activity and its capacity to process sensory information. To better understand the relationship between behavioral states and information processing, mice are trained to detect varying amplitudes of whisker-deflection under two-photon calcium imaging. Layer 2/3 neurons (n=1436) in the vibrissal primary somatosensory cortex are imaged across different behavioral states, defined based on detection performance (low to high-state) and pupil diameter. The neurometric curve in each behavioral state mirrors the corresponding psychometric performance, with calcium signals predictive of the animals choice outcome. High behavioral states are associated with lower network synchrony, extending over shorter cortical distances. The decrease of correlations in variability across neurons in the high state results in enhanced information transmission capacity at the population level. The observed state-dependent changes suggest that the coding regime within the first stage of cortical processing may underlie adaptive routing of relevant information through the sensorimotor system. HighlightsO_LINetwork synchrony and pupil diameter are coupled to changes in behavioral state. C_LIO_LIHigh behavioral state results in enhanced information transmission capacity at the population level, with neurometric curve in each behavioral state mirroring the corresponding psychometric performance C_LIO_LIBehavioral state and calcium signal in primary somatosensory cortex predict choice outcome. C_LI eTOCO_ST_ABSIn BriefC_ST_ABSLee et al. investigates the relationship between behavioral states and information processing in the primary somatosensory cortex. They demonstrate increases in behavioral state results in decrease cortical variability, enhanced information transmission capacity and stimulus encoding at the population level.