Analysis of multi-tissue transcriptomes reveals candidate genes and pathways influenced by cerebrovascular diseases

Abstract

Cerebrovascular diseases (CVD) are a group of medical conditions that impair circulation of blood to the brain, including stroke, transient ischemic attack (TIA), embolism, aneurysm, and other circulatory disorders affecting the brain. Here, we investigated the effects of having CVD history on the molecular signature of brain regions by comparing gene expression profiles from several brain tissues between cohorts with and without CVD history. We first merged tissue samples from GTEx RNA-Seq dataset into clusters based on the overall gene expression similarity. Then we performed differential expression (DE) analyses for each cluster using a linear mixed model that controls covariates and the individual random effect. Cross-region DE genes were ranked by the combined q-values derived from the mixed model using Fisher's method. Functional enrichment analyses were performed using Gene Set Enrichment Analysis (GSEA) program. We identified hundreds of DE genes, and many of them are related to endothelial or brain functions and associated diseases. We found that STAB1 was highly overexpressed across brain regions in the CVD cohort, and the upregulation of STAB1 in brain tissues may contribute to weaker selfdefense mechanisms against lesions in the brain. Our results suggest a list of candidate genes and pathways that may be dysregulated in the brains of people with CVD history, implying that suffering from CVD could pose potential hazard to the brain.