INTERACTIONS BETWEEN THE CEREBELLUM AND MEDIAL PREFRONTAL CORTEX IN RULE-BASED CATEGORY LEARNING AND RULE-SWITCHING IN RATS

Patients with posterior cerebellar lesions exhibit impaired executive function. In rodents, the posterior cerebellum communicates bidirectionally with the frontal cortex via the lateral cerebellar nucleus (LCN). Rule-based (RB) category learning requires discrimination along relevant physical features, and generalization along irrelevant features. We previously showed that medial prefrontal cortex (mPFC) NMDA lesions disrupt RB category learning. Additionally, chemogenetic inhibition of the LCN impairs, while excitation facilitates, this learning. Therefore, we hypothesize that LCN–mPFC communication is essential for RB categorization. We investigated these interactions using in-vivo electrophysiology in rats. Using a touchscreen-equipped operant chamber, rats performed a categorization task with circular stimuli that varied along two dimensions (spatial frequency and orientation), requiring discrimination of the stimuli along a relevant dimension, while ignoring the other (e.g., high frequency: category A, and low frequency: category B). Rats were trained to 75% accuracy on both categories, after which they received a multi-channel hyperdrive implant, with tetrode bundles targeting the LCN and mPFC. Neuronal activity was recorded from both areas while rats performed the task. After 7 sessions, the relevant dimension (rule) changed, and recordings continued until 75% accuracy was regained. We found that neurons in both regions show differential firing to the categories (i.e., category selectivity). LCN neurons show category selectivity during the first rule, which increases when the rule is changed. Similar results were observed for the mPFC, consistent with previous findings. These results support the role of the cerebellum’s involvement in category learning and suggest communication with the mPFC.