Cross-tissue regulatory gene networks in coronary atherosclerosis

Abstract

Coronary artery disease (CAD) is the underlying cause of myocardial infarction and stroke that together are responsible for nearly 30% of all global deaths. CAD is a common complex disease caused by the interactions of multiple genetic and environmental risk factors acting across several metabolic and vascular tissues. Owing to the complexity of these interactions, systems genetics is an increasingly recognized path to a better understanding of complex diseases. In this thesis, we applied systems genetics by integrating the analysis of genotype (DNA) and global gene expression (RNA) data from metabolic and vascular tissues with phenotype data from the clinically well-characterized subjects in the Stockholm Atherosclerosis Gene Expression (STAGE) study. We validated the initial findings using genome-wide association studies (GWAS) and several gene expression datasets from mice and cell models. As a result, we for the first time inferred regulatory gene networks (RGNs) with key drivers of CAD, several of its main risk factors and atherosclerosis regression.

In paper I, we designed a computational pipeline to reconstruct RGNs with key drivers in CAD using the STAGE study. Then, by integrating expression quantitative traits (eQTLs) of these RGNs with genotype data from several GWAS, 30 CAD-causal RGNs interconnected in blood, vascular and metabolic tissues were identified. Twelve of these RGNs were further validated in gene expression and phenotype data from the Hybrid Mouse Diversity Panel. As proof of concept, by targeting the key drivers AIP, DRAP1, POLR2I, and PQBP1 in a cross-species-validated, arterial-wall RGN involving RNA-processing genes, we re-identified this RGN in THP-1 foam cells and independent gene expression data from CAD macrophages and carotid lesions.

In paper II, we developed a cross-tissue weighted gene co-expression network analysis (X-WGCNA) method (used in Paper I) that reliably captures gene activities both within and across tissues. X-WGCNA is implemented as a package in R and is available online.

In paper III, we inferred transcription factor (TF) RGNs from three plasma cholesterol lowering (PCL)-responsive gene sets causally related to regression of early, mature, and advanced mouse atherosclerosis. We then used THP-1 cells in an in vitro atherosclerosis regression model to successfully validate 3 key drivers in these RGNs driving regression in early (PPARG), mature (MLL5), and advanced (SRSF10/XRN2) atherosclerosis.

In paper IV, we inferred the STAGE eQTLs (used in papers I and II) and identified subsets with gene regulatory effects across multiple tissues that according to GWAS were highly enriched in association with CAD. To better understand the pathophysiological role of these multi-tissue eQTLs, we identified and analyzed a number of associated gene sets.

A key result of this thesis is a repository of RGNs with key drivers for CAD, CAD risk factors, and atherosclerosis regression. This repository together with the computational pipeline including X-WGCNA should be useful in future studies that aim to go beyond genetic loci identified by GWAS and provide opportunities for novel diagnostics and therapies.