Integrative genomic and transcriptomic analysis identifies microglia-expressed genes with causal links to Alzheimer's disease risk

Abstract

Background

Microglia play a central role in Alzheimer's disease (AD) pathogenesis, yet it remains unclear whether microglia-related gene expression changes contribute causally to disease risk or reflect downstream responses to neurodegeneration.

Objective

This study aimed to systematically identify microglia-expressed genes with genetically regulated expression associated with AD risk and to characterize their functional relevance through integrative genomic analyses.

Methods

We compiled 2454 microglia-associated genes from four transcriptomic studies and five expression databases. Using brain cis-eQTL data from PsychENCODE (n = 1387), we performed summary-data-based Mendelian randomization (SMR) and HEIDI tests with AD GWAS data (111,326 cases, 677,663 controls). Significant SMR signals (FDR < 0.05) were evaluated using Bayesian colocalization. Regulatory context was assessed via overlap with PU.1-associated chromatin features, and differential expression was examined using single-nucleus RNA-seq from AD brains.

Results

SMR identified 16 genes whose genetically predicted expression was associated with AD risk (FDR < 0.05), including TSPAN14 (OR = 1.10, 95% CI: 1.07–1.13) and GRN (OR = 0.90, 0.86–0.95). Colocalization supported shared causal variants at eight loci (PPH4 > 0.8; e.g., BLNK, SIGLEC11, CASS4). Most prioritized genes overlapped with PU.1-associated regulatory regions. Single-nucleus RNA-seq confirmed dysregulation of several candidates (e.g., USP6NL, MS4A4A) in AD microglia.

Conclusions

We identify microglia-expressed genes with genetic evidence consistent with a potential causal role in AD. These genes are enriched in microglia-specific regulatory elements and exhibit transcriptional alterations in AD, highlighting microglial pathways as potential therapeutic targets.