Psychiatric conditions such as phobia, panic and post-traumatic stress disorder (PTSD) are associated with heightened fear-related behavior. This has been attributed in part to increased acquisition and expression of threat memory, which occurs when an innocuous stimulus encountered under threatening circumstances becomes a potent trigger for defensive emotional reactions. Research indicates that this form of memory engages the prelimbic cortex, a rodent brain region that is analogous to an area of pathological hyperactivity in the human medial prefrontal cortex (mPFC). Prelimbic excitatory neurons are thought to mediate memory expression by signaling to key downstream brain regions. However, the circuit and synaptic mechanisms underlying memory-specific recruitment of these cells remain poorly understood. The purpose of this project is to identify prelimbic synaptic changes associated with threat memory formation in mice, to elucidate the role of neural circuits where these modifications occur, and to define specific prelimbic output connections that mediate threat-related behaviors. Based on preliminary experiments, this work will focus sparse populations of inhibitory interneurons that express the peptide marker somatostatin and exhibit increased excitatory synaptic activity after threat learning. Our central hypothesis is that these cells undergo an increase in stimulus-related activity after learning and that they function to suppress parvalbumin-containing interneurons and thereby relieve a major brake on excitatory neuronal firing. In Aim 1, we will use brain slice electrophysiology to examine synapses mediating elevated activity of somatostatin interneurons as well as their interaction with parvalbumin interneurons. In Aim 2, we will use optogenetics and calcium-based imaging in freely behaving animals to test the role of somatostatin and parvalbumin interneurons in threat memory expression. Finally, in Aim 3, we will identify specific excitatory populations mediating behavioral outcomes of somatostatin interneuron manipulations as well as establish downstream brain regions to which these neurons provide synaptic input. This project represents the first attempt to establish how emotional learning alters complex inhibitory stimulus processing to support mPFC memory functions, and will identify potential targets for attenuating pathological activity and associated behaviors.