Contribution of multiple sensory modalities to texture discrimination in head-fixed mice

Decision-making tasks involving the somatosensory system, such as texture discrimination, have been extensively used to investigate the cellular and circuit mechanisms underlying the processing of sensory information in rodents and are believed to rely primarily on whisker-mediated inputs. However, when whiskers are used to explore the environment, a process called “whisking”, mice may combine tactile inputs with those from other sensory modalities (e.g., olfaction and vision). In fact, whisking is a stereotyped behavior that occurs in strict coordination with nose and head movements. Moreover, the whisking frequency can be phased-locked to the respiration rate and olfactory inputs have been recently shown to modulate the whisker-dependent responses of a subset of neurons in the barrel cortex. However, the impact of olfactory signals on complex whisker-dependent behaviors remains poorly understood. Here, we tested the hypothesis that sensory modalities other than somatosensation influence the behavior of head-fixed mice in a Go/No-Go texture discrimination task. In expert animals, we found that the proportion of correct choices was not affected by whisker trimming and that reaction times tended to be longer upon whisker trimming, indicating that sensory modalities other than somatosensation contributed to the animal’s performance in the task. We then observed that the behavioral responses of the mice were unchanged in total darkness, suggesting that the animals did not rely on visual inputs to solve the task. In contrast, experimental manipulations interfering with the processing of olfactory signals significantly decreased the fraction of correct choices both in the presence and in the absence of whiskers. Using two-photon calcium imaging in the superficial layers of the primary somatosensory cortex (S1) in combination with information theory analysis, we observed the encoding of task-related information in S1 both in the presence of whiskers and after whisker trimming. Finally, manipulations perturbing olfaction cancelled the encoding of task-related information in S1. However, when the encoding of the sensory stimuli and the choice signals were disentangled, S1 activity during the task primarily reflected decision-related processes. Altogether, these findings demonstrate that in head-fixed mice, olfactory cues alone can be sufficient to behaviorally discriminate textures. These results further suggest that S1 neuronal activity integrates sensory inputs signals originating from different sources (e.g., movement) associated with specific task requirements.