The brain multi-omic approach to identify key molecular drivers of neuropsychiatric symptoms in Alzheimer's dementia

Approximately 65% of individuals with mild cognitive impairment (MCI), Alzheimer's disease (AD), or AD related dementias (ADRD) experience neuropsychiatric symptoms (NPS). These debilitating symptoms include depression, anxiety, apathy, delusions, hallucinations, agitation, sleep disturbances and are associated with faster disease progression, greater functional impairment, higher caregiver burden, and earlier institutionalization. Current treatments for NPS in MCI/dementia have limited efficacy but high rates of adverse side effects, including higher mortality. Therefore, safe and effective treatments for NPS are urgently needed. However, we have limited insights into molecular mechanisms of NPS in MCI/dementia to nominate therapeutic targets. To address this knowledge gap, we aim to elucidate the genetic and molecular mechanisms underlying NPS in MCI/dementia using two complementary but independent approaches. In the first approach, we will leverage the substantial genetic basis of NPS (heritability of ~61%) to identify novel genes involved in NPS. First, we will perform genome-wide association studies (GWAS) of NPS in ~32,000 individuals with MCI/AD/ADRD from across the United States. We will perform a GWAS of each NPS domain as well as a multivariate GWAS of all NPS collectively. Results of each GWAS will be integrated with reference human brain transcriptomic and proteomic profiles to identify individual brain transcripts and proteins that are consistent with being causal in NPS. We have applied these integrative approaches to AD and identified 11 genes that regulate their brain protein abundances to predispose to AD (Wingo et al. 2021, Nat Genet). Unlike GWAS results, these findings point to specific brain proteins that are consistent with a causal role in AD. Importantly, this approach does not solely depend on genome-wide significant signals, and the causal inference of this integrative strategy has been experimentally tested and found to be robust. Thus, we expect that findings from this approach will provide novel insights into the pathogenesis of NPS. Our second approach aims to identify brain transcripts, proteins, and co-expression networks associated with NPS. We will perform transcriptomic and proteomic sequencing of post-mortem brain tissues from 400 individuals with MCI/AD and longitudinal NPS assessment, as existing brain omics data do not have robust NPS measures. Then we will examine NPS domains individually as well as collectively in our differential expression and network analyses. We anticipate that molecular alterations at the brain transcript and protein levels identified with this approach will provide insights into NPS pathogenesis and progression. Studying both transcripts and proteins would provide complementary information and a more complete molecular picture. We view understanding proteins as particularly important since they are the vast majority of drug targets. The proposed project will likely lead to new insights into molecular mechanisms and novel therapeutic targets for NPS and thereby have an important and sustained impact on public health.