PROJECT SUMMARY When making a decision we can use our understanding of available action-outcome relationships to prospectively evaluate the consequences of our potential actions and choose the one that is currently most beneficial. This goal- directed strategy is, thus, quite flexible, allowing us to readily adapt our behavior when circumstances change. But we don’t always think about the consequences of our behavior. Usually this is fine. Such habits are a way for our brain to efficiently execute routine behaviors. A balance between goal-directed and habitual control allows behavior to be adaptive when needed, but efficient when appropriate. But disrupted goal-directed learning and overreliance on habit can cause inadequate consideration of consequences, inflexibility, a lower threshold for compulsivity, and disrupted decision making. This can contribute to aspects of numerous diseases, including substance use disorder (SUD), obsessive-compulsive disorder, obesity, schizophrenia, depression, anxiety, and autism. An obstacle to mechanistic understanding is the dearth of information on the neuronal circuits that support action-outcome and habit learning. Thus, our broad goal is to expose neuronal pathways regulating both forms of learning. Midbrain dopamine neurons have long been implicated in learning. Canonically, they have been thought to signal the reward prediction errors that support habit learning. But emerging evidence indicates that midbrain dopamine has a much broader role in learning than originally thought, including contributions to goal-directed learning. How can dopamine support both these, often opposing, forms of learning? Our solution is simple: dopamine’s different functions in learning are achieved through its different projections. We will reveal the pathways that enable dopamine function in both habit formation and action-outcome learning. Dopamine may mediate action-outcome and habit learning via projections to the amygdala. The basolateral amygdala (BLA) is critical for action-outcome learning. By contrast, the central nucleus of the amygdala (CeA) mediates habit learning. Midbrain dopamine projections to the BLA and CeA have long been known to exist, but little is known of their function. We will reveal a function for each pathway in instrumental learning here. Our working hypothesis is that midbrain dopamine projections to the BLA support action-outcome learning and dopamine projections to the CeA support habit learning. We will test this hypothesis in two aims using a suite of modern systems neuroscience tools including fluorescent sensor-based dopamine monitoring and cell-type and pathway- specific, bidirectional, optogenetic manipulation coupled with theory-driven behavioral assessment of action- outcome and habit learning with outcome-specific devaluation and omission contingency tests. This will provide a critical basic science foundation for our long-term goal of mechanistic understanding of the causes of disrupted decision making and maladaptive habits in pathological states.

PROJECT NARRATIVE This project will uncover the distinct neuronal pathways through which dopamine drives both action-outcome and habit learning. Given that substance use disorders and other psychiatric diseases are characterized by ill-informed decisions that lack consideration of future consequences, resulting from disrupted action-outcome learning and maladaptive habits, in the longer term this work could lead to novel targets for treating pathological decision making.