Project Summary Over 150 million people are projected to be living with dementia worldwide by 2050. Alzheimer’s disease (AD) is the most common form of dementia, responsible for ~70% of dementia cases. A hallmark pathological feature of Alzheimer’s disease (AD) is progressive neurodegeneration, which is thought to initiate and track progressive cognitive decline in AD patients. While increasingly sensitive biochemical markers are available to diagnose AD in individuals, treatments to stall the disease in its early stages remain elusive. Increasing evidence in AD patients models indicates that significant accumulation of these biomarkers may be preceded by early circuit dysfunction. A large plurality of AD patients display subclinical epilepsy. Furthermore, circuit hyperexcitability has been observed before plaque formation in several familial AD mouse models as well, with similar findings in mouse models of sporadic AD. Cellular evidence from these studies suggests that circuit dysregulation is due to altered circuit inhibition from GABAergic interneurons. In particular, parvalbumin-expressing (PV) interneurons appear to be prone to changes in their action potential (AP) firing, and potentially neurotransmission, across distinct familial and sporadic AD mouse models. Neurodegeneration in AD is often thought to progress through well- defined brain regions, and interestingly, hyperexcitable circuits may accelerate this pathology. Whether physiological changes to PV interneurons emerge first in regions of high vulnerability in AD is unclear. Our central hypothesis is that PV interneurons will develop dysfunctional physiological deficits first in vulnerable brain regions during early AD, which may then progress to other brain areas in a Braak-esque fashion. To evaluate this hypothesis, which will require electrophysiological recordings from thousands of individual neurons, we will use the PatcherBot, our robotic platform capable of performing high-throughput, automated electrophysiology of neurons in brain slices; however, in this work, we will augment the PatcherBot’s machine vision capabilities with fluorescence imaging to specifically target PV-expressing interneurons in brain slices. The rationale and feasibility of this proposal are shown in preliminary work, demonstrating (1) fully automated patch clamping of florescent-targeted interneurons using the PatcherBot, (2) brain-wide introduction of PV specific labeling and optogenetic methods in AD mice in vivo, and (3) early-stage PV firing and neurotransmission deficits in a prodromal FAD mouse model. Here, we will address our hypothesis across three major regions (entorhinal cortex, hippocampus, isocortex) in 3 distinct models (APOE4, hAPP-KI, 5xFAD) and 3 relevant developmental timepoints. Findings from this proposal will yield wide-ranging advances, including high-throughput cell-type- specific physiology, to information regarding the potential circuit-seeding of cognitive dysfunction in early ...