Summary Over 90% of Alzheimer's disease (AD) patients suffer from behavioral and psychological symptoms of dementia (BPSD) including agitation, aggression, depression, apathy and psychosis. BPSD can present at almost any stage of AD, and in some patients, these symptoms can even appear before dementia develops. The severity of BPSD increases significantly with disease progression, and affects the quality of life of both patients and their caregivers. In many patients, BPSD is the main reason for institutionalization. However, the mechanisms underlying BPSD are not known, and there is no specific treatment strategy available. Although BPSD presents differently in each patient, the presence of certain symptoms in a patient make the co-occurrence of other symptoms more likely. In an ongoing collaboration with Rush Alzheimer's Disease Center, we have developed a method for clustering the symptoms of BPSD into four domains (affective, hyperactivity/disinhibition, psychosis and apathy). Based on these domains, we then conducted an RNA-seq and found different gene expression profiles in AD patients with and without BPSD. This evidence supports the notion that distinct molecular pathways may be involved in the appearance of BPSD. In this proposal, we hypothesize that individual BPSD domains in patients with AD are due to definable perturbations in molecular pathways and that these pathways can be analogized in AD mouse models, allowing for a causal investigation of the relationship between specific pathway alterations and domain behaviors. We will test this hypothesis through both human study and animal work. For the human study, 1) we will expand on our behavioral analyses by increasing subjects for pre-mortem clinical assessments and defining BPSD trends over time in AD patients. 2) Within each behavioral domain, we will employ RNA-seq to investigate gene expression patterns in different brain sub-regions that are unique to each BPSD domain and the gene expression pattern will be compared across normal, MCI and AD subjects. 3) Finally, we will identify which pathways are most clearly associated with each of the BPSD domains using bioinformatics and biochemical analyses. For the animal model work, 1) we will characterize how mouse behaviors analogous to human BPSD symptoms evolve during AD-like neuropathgenesis progression 2) We will identify the most promising molecular candidates for intervention from our RNA-seq findings using these AD/BPSD models. 3) Finally, we will determine whether altering these pathways leads to changes in BPSD-like behavior using virally mediated genetic manipulations (AAV9/CRISPR-Cas9). Overall, this project will establish a translational pipeline by associating BPSD symptom domains with molecular alterations in human AD patients, and by demonstrating that manipulations of these pathways can cause BPSD-like behaviors in transgenic mouse models of AD. These data-driven approaches will lead to a better understanding of the mole...