Project Summary Nothing is more central to the human experience than memory, and no human capacity is more devastating when it is lost. Memory has been intensely studied in basic and disease research for many decades but to my knowledge we do not have any physical explanation of how memories alter complex behavior. This is in large part because we do not have a mechanistic explanation of how most behaviors are executed in the first place, but also because memory formation is best understood in brain regions like the hippocampus, which are far removed from action control centers, which themselves are not well understood in the mammalian brain. Drosophila courtship behavior stands out from other complex behaviors because it is exceptionally well understood at the circuit level. Here I describe a new and robust paradigm for courtship learning in which the memories appear to be formed, stored, modified, implemented, and erased within core courtship circuitry. In both insects and mammals, dopamine has long been known to translate memory-relevant information from the outside world into an internal teaching signal. In our new courtship learning paradigm we find that dopamine signaling is both necessary and, under the appropriate circumstances, sufficient for memory formation. I propose to test the hypothesis that dopamine-induced memory is written directly within courtship circuitry, likely within a set of command neurons called P1. We will identify the specific dopaminergic neurons required to write the memory, as well as localize the neurons that receive the dopamine signal. We will then use our established activity monitoring and behavioral assays, together with the courtship wiring diagram, to understand how memory alters the flow of information through behavioral circuitry. The results will provide new insights into how memories directly impact behavioral decision-making, a goal that has been unachievable in systems lacking detailed behavioral circuit maps. The conserved role of dopamine in memory writing suggests that our findings will generalize to mammals. Learning usually requires repetition or extreme events. This prevents the over-generalization that might occur with single-trial learning. By analyzing the contributions of various known components of the courtship circuitry to memory formation, we find a striking and novel role for a group of neurons, called mAL, in setting a threshold for memory formation. mAL stimulation can even erase old memories, but only while they are being recalled. I believe this is the first circuit-level manipulation that has been shown to disrupt memory reconsolidation. I outline experiments designed to leverage this new behavioral paradigm and its circumscribed neuronal populations to understand the nature of memory thresholding, maintenance, and removal in behavioral circuitry at the molecular, circuit, and physiological levels. We will work to establish the new and potentially transformative hypothesis that wit...