Visualizing neuronal correlates of sensory prediction in the barrel cortex of mice engaged in a whisker-guided locomotion task

International audienceThe ability of the mammalian central nervous system to constantly adapt motor commands and optimise behaviour according to the context does not only rely on efficient analysis of the incoming sensory flow. The cerebral cortex is indeed thought to compute a dynamic model of our environment allowing anticipation of future sensory inputs. Such computation would imply the production of error signals in case of divergence between actual and predicted sensory inputs. Here, we aim to study the neuronal mechanisms at play in such sensory predictions using the tactile whisker system of mice as a model.To this end, we use a whisker-guided locomotion task where mice have to avoid two obstacles in their path. We wish to analyze the cortical dynamics evoked by the contact of the vibrissae with the obstacles in such familiar environment, but also to reveal the capacity of the somatosensory cortex to produce error signals when this environment is altered by surprise, either by the removal or change in position of an obstacle on the animal's path.To do so, we developed an optical apparatus based on an image guide and motion sensors. It allows us to record mesoscale voltage induced fluorescence changes over the primary somatosensory cortex at a subcolumnar resolution, while the animals are freely navigating and gathering tactile information in an ethologically relevant manner. By simultaneously recording the animal’s path adaptations and whisker movements, we will be able to correlate dynamics of cortical activity, sensory inputs and motor adjustments, at the millisecond timescale