Project Summary There is extensive preclinical evidence for anticonvulsant effects of neuroactive steroids, which are thought to be mediated through their actions as positive allosteric modulators (PAMs) on GABAA receptors (GABAARs). These studies have relied on the administration of exogenous neuroactive steroids, although, neurosteroids can also be synthesized in the brain. However, we know very little about the function of endogenous neurosteroids due to the lack of tools necessary to investigate their function. The rate-limiting enzymes involved in neurosteroid synthesis, 5α-reductase 1 or 2, are expressed in the hippocampus and given their actions as PAMs at GABAARs, they are well-suited to constrain network excitability in this region. To investigate the function of endogenous neurosteroids, we generated novel mouse models (floxed Srd5a1- IRES-GFP and floxed Srd5a2-IRES-tdTomato) enabling visualization and quantification of changes in the expression of these enzymes associated as well as the ability to knockout these enzymes in a cell type- and brain region-specific manner. Here we propose to utilize these novel tools to test the hypothesis that endogenous neurosteroids constrain neuronal excitability and that deficits in 5α-reductase 1 and 2 expression in the hippocampus contributes to seizure susceptibility. We will determine whether 5α-reductase 1 and/or 2 expression is altered in chronically epileptic mice (Specific Aim 1) and if loss of 5α-reductase 1 and/or 2 expression in the hippocampus increases network excitability and acute seizure susceptibility (Specific Aim 2) or increases seizure frequency in chronically epileptic mice (Specific Aim 3).