Regulatory landscape of Alzheimer's disease variants in human microglia

Genome-wide association studies (GWAS) for Alzheimers disease (AD) predominantly implicate non-coding genetic variants that are presumed to regulate gene expression. Yet, around half of GWAS variants do not coincide with known regulatory variants, such as expression quantitative trait loci (eQTLs). AD GWAS loci are enriched for microglia-specific regulatory regions, but the causal variants and mechanisms remain unresolved. To connect risk variants to microglial function, we evaluated the regulatory activity of AD GWAS variants, microglial eQTLs, and chromatin accessibility QTLs (caQTLs) by screening 11,550 candidate regulatory sequences (CRSs) with a massively parallel reporter assay (MPRA). Whether prioritized by GWAS or eQTLs, variants are equally likely to predict active CRS and expression-modulating variants (emVars) by MPRA, yet there were clear systematic differences, particularly in that predicted target genes of GWAS loci, but not eQTLs, are enriched for AD relevant biology. Notably, only 7% of all emVars were shared between two distinct sources of human microglia cells, human induced pluripotent stem cell (hiPSC)-derived microglia (iMGLs) and immortalized HMC3 cells, underscoring the importance of cellular context in evaluating regulatory activity. Our findings establish the first large-scale microglial-context MPRA, define a microglial regulome underlying AD genetic risk, and provide a functional framework to refine noncoding variant interpretation.