Combining genetics with real-world patient data enables ancestry-specific target identification and drug discovery in Alzheimer’s disease

Although high-throughput DNA/RNA sequencing technologies have generated massive genetic and genomic data in human disease, these findings have not been translated into new patient treatments. To address this problem, we utilized Mendelian randomization (MR) and large patient-level genetic and functional genomic data to evaluate druggable targets using Alzheimer’s disease (AD) as a prototypical example. Specifically, we applied the genetic instruments from 9 expression quantitative trait loci (eQTL) and 3 protein quantitative trait loci (pQTL) datasets across five human brain regions from three human brain biobanks and performed MR in 7 genome-wide association study (GWAS) datasets of European ancestry (EA) and African ancestry (AA), with AD cases and controls. We identified 19 drug targets, including the inflammatory target of epoxide hydrolase 2 (EPHX2) as a potent AD target. We demonstrated that a genome-wide significant and protective variant of p.Arg287Gln in EPHX2 (β = -0.096, PGWAS = 1.08 × 10-11) significantly reduced level of phosphorylated-tau (p-tau181) and the ratio of p-tau181/total tau and increased neuron clump size in patient induced Pluripotent Stem Cells (iPSC)-derived neurons, mechanistically supporting MR results. Pharmacologic inhibition of EPHX2 significantly improved cognitive behaviors in two AD transgenic rodent models (5xFAD and TgF344-AD). We further identified that 12 drugs (i.e., trazodone [ADRA1A] and baclofen [GABBR1]) harboring MR-supported targets are significantly associated with reduced incidence of AD in 111,680 mild cognitive impairment (MCI) patients from the Optum database. Using a new user active-comparator design, we found that usage of trazodone was significantly associated with 22% reduced incidence of AD (hazard ratio [HR] = 0.78, P=6.44x10-6) in people with MCI in the MarketScan database. In summary, combining genetics and real-world patient data identifies ancestry-specific therapeutic targets and medicines for AD and other neurodegenerative diseases if broadly applied.