SUMMARY In day-to-day life, we often must select actions by envisioning the likely outcomes of our behaviors, and we adjust our behavior if prospective outcomes become more or less valuable. This capacity for prospective decision making can be imperiled by addictive drugs, such that individuals suffering from substance use disorders may seek addictive drugs despite threats to social ties, employment, etc. The ability to select actions that will result in valued rewards requires the medial orbitofrontal cortex (MO), particularly when those rewards are not immediately delivered (are unobservable) and must be anticipated. Addictive drugs like cocaine cause MO atrophy and impede MO function in both rodents and humans, emphasizing the need to fully understand the manner by which the MO coordinates prospective action selection. The neurotrophic factor, Brain-Derived Neurotrophic Factor (BDNF), and its high-affinity receptor tropomyosin receptor kinase B (TrkB) control MO function. BDNF presence in the MO is necessary and sufficient for mice to appropriately direct behaviors towards rewards of higher value relative to low-value options. Further, locally overexpressing a truncated isoform of TrkB (TrkB.t1), which acts as a dominant negative, or locally ablating TrkB occludes value-based action, particularly when potential rewards are unobservable. We hypothesize that neurotrophin signaling stabilizes neural connections essential to adaptive choice. Which neural connections are necessary for choice behavior? We recently discovered that ventral hippocampal (vHC) inputs to the MO are necessary for value memory updating – and particularly, the integration of new value information into instrumental response strategies. Meanwhile, projections from the MO to the basolateral amygdala (BLA) are necessary for value memory retrieval (and not updating). Thus, vHC inputs to the MO appear to update value memory, while outputs retrieve new memories to execute adaptive action. The goals of this proposal are to: (1) determine whether functional vHC-to-MO and MO-to-BLA connections require TrkB-mediated signal propagation on excitatory MO neurons. (2) We will next determine whether MO neurons form stable value memory traces necessary for adaptive choice, and whether trace formation requires i.) vHC inputs, ii.) TrkB, and iii.) local dendritic spine plasticity. (3) The capacity for prospective goal-directed action improves throughout adolescence. Because of this, mechanistic insights can be gained by studying adolescent development. We will delineate the adolescent development of vHC-to-MO projections, the morphological maturation of projection-defined layer V MO neurons, and the stimulation of TrkB+ MO neurons during decision-making tasks in mice. We will test the hypotheses that TrkB controls typical development, and that TrkB+ MO neurons are increasingly stimulated during tests of behavioral flexibility as mice mature, ultimately forming stable memory traces attune...