Mechanisms of Context-Dependent Reward Seeking in Hippocampus and Prefrontal Cortex

PROJECT SUMMARY Relapse is a common barrier to the effective treatment of substance use disorders (SUD). Clinical studies have found that relapse can be triggered by exposure to contexts that have become associated with drug use. Drug contexts may be directly observable, such as a location of prior drug use. However, drug contexts may be more abstract, such as an expected time of day. Such contexts, which must be inferred due to their lack of observable cues, are common in real-world settings. However, inferred contexts are not typically studied in animal models of addiction. As such, a greater understanding of context-dependent relapse in SUDs requires the study of contextual inferences and how contextual encoding influences action selection. Prior studies of context-dependent behaviors have implicated the hippocampus (HPC) and medial prefrontal cortex (mPFC) in context encoding, and have implicated the mPFC and secondary motor area (MOs) in action selection. Recent theoretical work has sought to unify hippocampal processes under the general function of state inference, in which the HPC encodes environmental states and creates maps over related states. State representations in HPC could then be communicated to mPFC to support context-based action selection. However, abstract contextual inference has yet to be tested directly in HPC, the physiology of mPFC subregions is understudied in relation to context, and the mechanisms of coordination between HPC, PFC, and MOs in context-dependent action selection are poorly understood. This project aims to use cutting-edge experimental techniques to study how inferred contexts are encoded and inform action selection. Using experiments in which mice must infer their current context to choose reward-maximizing actions, I will test the hypothesis that in context-dependent behaviors, HPC encodes inferred contexts with distinct representations that coordinate with representations of action selection through theta synchronization between HPC, mPFC, and MOs. Electrophysiological recordings performed in CA1, mPFC, and MOs of behaving mice will be analyzed for encoding properties and communication between brain regions. The results of this project will provide a mechanistic foundation for addiction studies of context-dependent drug seeking and relapse.