Fast pathway-specific reorganization of barrel cortex underlying rapid goal-directed sensorimotor learning
Neuroscience ePoster by Anthony Renard, Georgios Foustoukos, Maya Iuga, Sylvain Crochet and Carl Petersen
Part of FENS Forum 2024 posters
Abstract
Animals exhibit flexible and adaptive behavior in novel situations, in part relying on reward-based learning. Although the reinforcement of positive outcomes plays a central role in shaping behavior, the neural circuit mechanisms of goal-directed sensorimotor transformation remain incompletely understood. Common behavioral paradigms for studying such mechanisms are hindered by the prolonged training period needed to reach expert performance, impeding the real-time observation of plastic changes. Here, we developed a rapid whisker-based detection task where thirsty mice learn to associate a novel stimulus with licking of a water reward spout in the course of a few minutes within a single behavioral session. Inactivation experiments via pharmacology and optogenetics demonstrated the necessity of barrel cortex activity in this learning process. Performing longitudinal two-photon calcium imaging of barrel cortex before, during and after the learning session revealed network reorganization with a long-lasting increase in the whisker-evoked responses of a subpopulation of L2/3 excitatory neurons after learning. To assess how these changes might contribute to reward-based learning by routing sensory information differentially to distinct downstream cortical regions responsible for task execution, we further investigated the involvement of L2/3 barrel cortex neurons projecting to the secondary somatosensory (S2p) and the primary motor (M1p) cortices. Consistent with previous work, S2p neurons exhibited a significant increase in activity after learning, while M1p neurons remained unchanged. Taken together, these results indicate that our rapid-learning task constitutes a promising paradigm to uncover the plasticity rules and synaptic mechanisms governing reward-based learning.