Circadian rhythms are everywhere in living matter from protozoans to human beings. Every cell, organ, and biological system works on an endogenous 24-hour rhythm. However, in patients with Alzheimer’s disease (AD), a neurodegenerative disorder affecting memory, this rhythm is dampened. In addition to cognitive decline, many AD patients (~50%) also exhibit agitation and increased anxiety in the evening or nocturnal hours, often referred to as sundowning. Here, we aim to dissect the neural circuitry underlying sleep disturbances, circadian rhythm abnormalities, and increased anxiety-like behavior representing sundowning and how it relates to cognitive decline and AD pathology. We will utilize behavioral assays, optogenetics, whole-brain microscopy, and in vivo Ca2+ imaging. Our preliminary evidence suggests that AD mice exhibit increased anxiety-like (e.g., sundowning) behavior compared to control (Ctrl) mice immediately before their sleep cycle but not before their wake cycle. Therefore, in Aim 1A, we will use the PiezoSleep mouse behavioral tracking system to monitor sleep/wake cycles and circadian homecage locomotor activity in Ctrl and AD mice at different ages in order to better characterize AD mice throughout their wake/sleep cycle. We will also measure anxiety-like behavior at various points throughout their sleep/wake cycles. Using an activity-dependent tagging mouse, the ArcCreERT2 x enhanced yellow fluorescent protein (EYFP) mice crossed with the APP/PS1 (AD) model, we will then identify whole-brain neural ensembles and therefore, neural networks that contribute to sundowning in Aim 1B. This mouse line allows for the indelible labeling of cells expressing the immediate early gene (IEG) Arc/Arg3.1 and allows for a comparison between the cells that are activated during one experience and those that are activated during a second experience. Here, we will compare and contrast the neural ensembles within subjects that are active during anxiety-like behaviors prior to the sleep and wake cycles. In Aim 2A, after identifying which neuronal ensembles are altered during sundowning, we will use nVoke minimicroscopes from Inscopix, to image Ca2+ transients in freely moving AD x ArcCreERT2 mice, injected with a GCaMP6f virus, during the dark versus light cycle. Here, we will be able increase resolution and the time points at which we can correlate neural activity with increased anxiety-like (e.g., sundowning) behavior. In Aim 2B, we will then test the hypothesis that optogenetic modulation of these neural ensembles drives/inhibits sundowning behavior. The neural circuitry underlying sundowning has not yet been elucidated, but this project will inform us of novel brain regions and the neural circuitry affected by sundowning and how we can manipulate these systems to rescue this phenotype in AD.