Cellular and circuit mechanisms underlying the generation, regulation, and flexibility of value representations in mouse medial prefrontal cortex

About This Grant

Project Summary/Abstract Most sensory stimuli acquire value through learning and experience. While progress has been made in deline- ating the circuits involved in sensory value-guided behaviors, the cellular and circuit mechanisms that underlie how distinct value representations are formed during learning, regulated during behavior, and modified to ac- commodate changes in cue-outcome contingencies remain unclear. Resolving these open questions is a funda- mental challenge in neuroscience and will advance our understanding of how synaptic plasticity induced by reinforcement learning is maintained, yet behaviors remain adaptive. The overarching goal of this proposal is to seek a mechanistic understanding of the circuits underlying value-guided learning and action in the mouse me- dial prefrontal cortex (mPFC), a region essential for the imposition of value on sensory stimuli to guide behavior. Using an olfactory-based appetitive classical trace conditioning task in combination with high-density electro- physiological recordings and optogenetics, I will focus on how functionally distinct yet spatially intermingled pop- ulations of neurons encoding rewarded (conditioned stimuli, CS+) and unrewarded (CS-) odors interact within the local network to drive stable yet flexible reward-seeking behaviors. I will test the hypothesis that the CS+ and CS- populations form a mutually inhibitory circuit that underlies their behaviorally-opposing roles in reward-seek- ing (Aim 1, K99). I will investigate how the existence of a CS- population may represent a circuit mechanism to permit the maintenance of prior reinforcement despite the extinction of learned behavior, thus uncovering a pre- viously unappreciated role for the CS- ensemble in reinforcement learning (Aim 2, K99). In my R00 phase, I will interrogate the mechanisms by which CS+ and CS- representations arise during learning, testing the hypothesis that distinct inputs to the mPFC drive the reinforcement of odors predicting the presence or absence of reward (Aim 3a,b). Further, I will study how the CS+ and CS- representations are updated when cue-outcome contin- gencies are reversed (Aim 3c). Finally, I will model the mPFC circuit to reveal fundamental principles on value- guided learning, which will inform future experiments and also extend the insights that could be gained beyond that which is feasible by experimentation (Aim 3d). This work will have broad implications for various neuropsy- chiatric conditions in which adaptive value-guided learning is disrupted, including addiction and depression. Can- didate and Career Goals. I aim to establish an independent research program investigating the neural basis by which value representations are generated and flexibly regulated to support context-dependent behaviors. I have extensive experience in molecular and systems neuroscience and have developed foundational tools and in- sights for studying value learning. Career Development Plan. I will be trained by my mentors Drs. Richard Axel and Larry Abbott. Dr. Axel is a foremost leader in sensory neuroscience, who will guide me on all aspects of experimental design. Dr. Abbott is a world-renowned theorist who will provide training in the analysis and mod- eling of complex datasets. All mentors will provide career development training for my transition to independence.