PROJECT SUMMARY Stressful events, or acute stressors, are an inescapable part of daily life and can precipitate the onset of mental health disorders. One powerful mechanism by which acute stress contributes to the development of mental health problems is through altering the structure and function of the hippocampus, a crucial brain structure for learning. However, this relationship is complex: acute stress can both enhance and impair hippocampal learning, making target mechanisms for treatment unclear. Research in rodents has shown that these contradictory findings can be partly explained by distinct stress actions on different pathways and subregions within the hippocampus. Whether these mechanisms extend to humans is unknown. To mitigate acute stress actions on learning that contribute to psychopathology and design targeted interventions, there is a pressing need to understand the mechanisms by which stress alters hippocampal learning in humans. This exploratory R21 proposal leverages recent advances in cognitive neuroscience to develop innovative functional neuroimaging and behavioral protocols targeting distinct hippocampal pathways in order to translate stress findings from animal models and uncover the mechanisms by which stress biases different types of learning in humans. We will test the novel hypothesis that episodic encoding, which involves the trisynaptic pathway (shown to be impaired by stress in rodent models: entorhinal cortex, dentate gyrus, cornu ammonis [CA] 3, and CA 1) will be impaired by acute stress, whereas statistical learning, which involves the monosynaptic pathway (shown to be spared or enhanced by stress: entorhinal cortex, CA1) will be enhanced by acute stress. This work marks the first investigation of stress effects on statistical learning, a ubiquitous learning process recently implicated in mental illness and treatment outcomes. Preliminary data indicate that a single behavioral paradigm can provide indices of episodic encoding and statistical learning that map distinct hippocampal correlates. In Aim 1.1, we will optimize this behavioral paradigm for detecting stress effects, and in Aim 1.2 we will determine the consequences of an acute stress induction for learning and delayed retrieval of these episodic and statistical representations. In Aim 2, we will use functional neuroimaging (fMRI) together with hippocampal subfield segmentation and sophisticated univariate, multivariate, and connectivity analyses, to quantify the neural mechanisms by which stress modulates these distinct learning processes. Successful completion of these aims lays the foundation for translating neurobiological mechanisms of stress actions from rodents to humans, providing critical information to reveal novel targets for interventions that mitigate the risk of negative mental health outcomes resulting from stress.