Project Summary. A fundamental function of the nervous system is to distinguish between threatening and non- threatening stimuli. For example, a sudden intense sound that indicates danger should trigger an acoustic startle response, but an innocuous sound should not. This type of behavioral threshold is a basic mechanism for sensorimotor filtering, and the importance of setting this threshold appropriately is highlighted by the startle hypersensitivity observed in neuropsychiatric diseases such as autism, anxiety, and schizophrenia. Despite its importance, and in contrast to our knowledge of experience-dependent startle modulation, the molecular and cellular pathways that establish and maintain the innate startle threshold are not well characterized. By developing a more complete understanding of the biological mechanisms that govern the startle threshold, we can generate new hypotheses about the neural bases for these diseases. This project will leverage the powerful larval zebrafish model system to investigate the molecular-genetic and neural circuit bases of the startle threshold. Here a simple, conserved, and genetically accessible circuit drives a stereotyped startle response, with auditory afferents triggering reticulospinal neurons to activate motor neurons and initiate movement. In a recent genome-wide screen, we identified a novel regulator of the innate startle threshold: cytoplasmic Fragile X mental retardation protein (FMRP) interacting protein 2 (cyfip2). cyfip2 mutants are hypersensitive and startle to low intensity sounds that rarely startle wild-types. Cyfip2 acts through FMRP and eIF4E to regulate RNA translation, but it can also control actin polymerization through interactions with Rac1 and the WAVE regulatory complex (WRC). In Aim 1 we will systematically test which of these molecular pathways cyfip2 uses to establish the startle threshold and to maintain it through development. In Aim 2 we will define the cellular basis for cyfip2- mediated threshold control by first locating the site of the primary circuit defect with optogenetic and calcium imaging approaches and then identifying the cell types in which cyfip2 is needed for normal startle sensitivity. Finally, our data show that acute manipulation of the actin cytoskeleton substantially alters the startle threshold while also decreasing the number and size of excitatory synapses in inhibitory glycinergic neurons but not excitatory glutamatergic neurons. In Aim 3 we will test the hypothesis that cyfip2 acts cell-autonomously to maintain excitatory/inhibitory synaptic balance, combining behavioral recording with live imaging of neuronal activity and synaptic scaffolds to define direct links between cyfip2, circuit structure and function, and behavior. Overall, the results of this work will generate a detailed model of molecular and cellular pathways that control the startle behavior threshold and lay a foundation for understanding how these may be affected in human disease.