Project Summary / Abstract Depression is by definition a fundamentally episodic form of mental illness featuring discrete symptomatic periods, interposed between periods of apparent wellness. The neurobiological mechanisms driving the induction, remission, and recurrence of depressive episodes over time are not well understood, especially at the circuit level, but converging evidence indicates that synaptic remodeling in prefrontal cortical (PFC) circuits plays an important role. Still, despite decades of pioneering work in this area, a mechanistic understanding of how postsynaptic dendritic spine remodeling contributes to changes in PFC circuit function and depression- related behaviors over time remains elusive. To date, most studies have relied on cross-sectional comparisons of spine density in fixed tissue at a single time point, obscuring dynamic effects on spine formation, stabilization, and pruning—distinct processes with differing implications for identifying new treatment targets. How stress affects dynamic spine remodeling processes differently in the female PFC is also unclear, despite the fact that sex is a critical risk factor for stress-related psychiatric disorders. Perhaps most importantly, whether spine remodeling causes or merely correlates with behavioral changes and altered function in specific PFC circuits is unknown. This proposal will investigate how stress-induced spine remodeling in the PFC contributes to anhedonia, a core feature of depression. PFC circuits support reward-seeking behavior by mediating action valuation computations, which integrate information about the magnitude of an anticipated reward and the expected effort required to obtain it. Leveraging newly developed optogenetic and 2P imaging methods for visualizing and manipulating spine dynamics and defining their effects on circuit function, we will investigate how spine remodeling in topologically defined projection neuron subtypes contributes to the induction, remission, and recurrence of anhedonic behavioral states. We will use a two-hit stress model, whereby early life stress (ELS) induces heightened stress sensitivity and sex-specific effects on HPA axis reactivity in adulthood, imaging PFC projection neurons before and after chronic stress and longitudinally during recovery. Reward-seeking behavior will be quantified in a 2P imaging-compatible action valuation task, in which we can independently manipulate anticipated reward magnitude and expected effort. We will test the hypothesis that stress disrupts action valuation by selectively eliminating dendritic spines and disrupting multicellular ensemble activity in PFC projections that play a critical role in encoding reward predictive cues. Next, we will test pharmacological and circuit-based strategies for promoting stress resilience and recovery.