Proposal Summary Neuronal circuits maintain a delicate balance of excitatory drive and inhibitory regulation to execute high order functions, such as learning and memory, and maintain network stability which is severely compromised in temporal lobe epilepsy (TLE). With a better understanding of mechanisms of memory formation and how TLE disrupts the processes, we gain insights that can be used to improve memory deficits in disease. The hippocampal dentate gyrus (DG) acts as a functional gate into the hippocampal trisynaptic circuit and plays a key role in learning and memory. Formation of memories is believed to be coded by activity of a distinct collection of neurons which represent a memory or experience known as an engram. Sparse activity in dentate granule cells (GCs) has been shown to be involved in engram formation; however, the circuit mechanism that underlie formation of these neuronal activity patterns are not fully understood. The DG is a circuit with low spontaneous activity and robust inhibition of the projection neurons, the GCs, by local inhibitory neurons (IN). Recent studies have found that a sparse subtype of dentate projection neurons, semilunar granule cell (SGC) are preferentially recruited in engrams. SGCs differ from GCs in their wide dendritic arbors, molecular layer axon collaterals and persistent firing and have been proposed to support feedback inhibition of GCs. However, circuit connectivity and functional effects of SGCs are not known. My objective is to better understand SGC’s role in information processing as well as their involvement in microcircuit changes related to epilepsy. I hypothesize SGCs differ from GCs in their input integration and SGCs that outputs directly activate a subset of GCs involved memory engrams and further refine GC engrams by engaging feedback inhibition of surrounding “non-engram” GCs. In acquired epilepsy, I propose that SGC’s support of the DG inhibitory gate is compromised and SGC dependent excitation increased resulting learning deficits. Aim 1 will identify differences in afferent inputs to GCs and SGCs using virally mediated pathway specific expression of channelrhodopsin to activate distinct DG inputs and adopt morphometric computational modeling to test the effect of dendritic structure on input integration in SGCs and GCs. Aim 2 will use the inducible cFOS TRAP2 system coupled to fluorescent reporters to label neurons active during a specific memory task followed by electrophysiology to determine how SGC output influences activity of GCs within and outside the shared engram. Finally, in Aim 3, will examine how engram stability, SGC activity and its influence on GC activity are altered in the pilocarpine model of experimental epilepsy. Together these studies will provide novel fundamental insights into dentate circuit function and memory processing and how these are altered in epilepsy and enable future development of circuit-based therapies to improve memory function.