Single-cell transcriptome-wide Mendelian randomization and colocalization analyses uncover cell-specific mechanisms in atherosclerotic cardiovascular disease

Genome-wide association studies (GWAS) have identified numerous genetic loci influencing human disease risk. However, linking these loci to causal genes remains challenging, limiting opportunities for drug target discovery. Transcriptome-wide association studies (TWAS) address this by linking variants to gene expression, but typically rely on bulk RNA sequencing, which lacks cell-specific resolution. Here, we present a single-cell TWAS pipeline combining cis -Mendelian randomization (MR) with colocalization analyses at the single-cell level. As a case study, we examined how genetically proxied gene expression in immune cells influences atherosclerotic cardiovascular disease (ASCVD) risk. We integrated single-cell expression quantitative trait loci (sc-eQTL) for 14 immune cell types with GWAS for coronary artery disease, large artery atherosclerotic stroke, and peripheral artery disease. Single-cell cis -MR analyses revealed 440 gene-outcome associations across cell types, 84% of which were missed by bulk TWAS, despite a considerably smaller sample size of the sc-eQTL dataset. Of these associations, 17 were replicated with external cis -eQTLs and demonstrated colocalization with ASCVD GWAS signals. Notably, genetically proxied expression of LIPA in monocytes was associated with coronary artery disease, large artery atherosclerotic stroke, and subclinical atherosclerosis traits. These findings were confirmed in a phenome-wide association study without evidence of associations with unexpected clinical outcomes. Single-cell RNA sequencing and immunohistochemistry of human carotid plaques revealed high LIPA expression in plaque macrophages. Our pipeline provides a solution for the discovery of cell-specific expression patterns that drive genetic predisposition to human disease, potentially impacting target selection for cell-tailored therapeutics.