The functions of the claustrum in flexible decision making.

PROJECT SUMMARY: Cognitive flexibility is disrupted in many neuropsychiatric diseases, but how the brain generates flexible behavior is not fully understood. Learning from recent actions requires neural mechanisms that maintain information about relevant decision variables to influence future decisions. Persistent activity in cortical areas required for cognitive flexibility, such as prefrontal cortex (PFC), connect an animal's choices and recent outcomes. Neural mechanisms for generating and maintaining this persistent activity involve excitatory loops between the neocortex and subcortical structures. Recent studies suggest that the claustrum, a poorly understood subcortical nucleus that forms reciprocal connections with the neocortex, is particularly highly interconnected with cortical areas strongly implicated in behavioral flexibility. We propose to test whether the claustrum contributes to generating flexible behaviors using two tasks, dynamic foraging and reversal learning. These tasks require medial prefrontal cortex (mPFC) and lateral orbitofrontal cortex (lOFC) function respectively, two areas likely influenced by claustrum activity based on recent studies of claustro-cortical- claustral loops. First, we will test the hypothesis that the activity of claustrum neurons encodes decision variables in these two tasks. Second, we will determine whether claustrocortical neurons are required for dynamic decision making and reversal learning, respectively. In addition, we will test whether claustrocortical projections contribute to generating persistent cortical activity and influence the encoding of decision variables by cortical neurons. We will further test whether claustrocortical neurons with projections biased to two different cortical areas, Cla→mPFC and Cla→lOFC neurons, form distinct functional modules within the claustrum and whether the claustrocortical projections to mPFC and lOFC influence cortical activity using similar cellular mechanisms. Third, we will test whether claustrocortical projections are required for learning at different timescales, both during and after the acquisition of a reversal learning task. We predict that claustrocortical loops contribute to generating and maintaining persistent activity in the cortex required for task performance and learning at multiple timescales. Together, these data will represent the first studies of claustrum neurons in well established, carefully controlled, decision-making tasks combined with quantitative models using normative theory. Furthermore, these experiments will directly test effects of claustrocortical inputs on an animal's responses and on cortical activity during flexible behavior, enabling the integration of the claustrocortical system into models of flexible decision making and cognitive control in health and disease.