Introduction: Carotid atherosclerotic plaque is an inflammatory lesion associated with a significantly increased risk for stroke. The formation of carotid plaque involves complex processes, such as endothelial injury and activation and recruitment of immuno-inflammatory cells. Although much progress has been made in the understanding of the role of these processes in carotid plaque development and stroke, a comprehensive view of the regulatory interactions among them at the molecular level is lacking.
Hypothesis: We hypothesize that perturbations of gene regulatory networks in carotid vascular tissues lead to dysfunctional molecular pathways promoting atherosclerotic plaque formation, and these networks are governed by central regulators that likely play more significant roles in pathogenesis.
Methods: We collected gene-expression profiles of 11 publically available gene expression datasets of carotid plaque samples (n = 1,546) and reconstructed gene coexpression networks, comprised of modules of co-regulated genes. The preservation of the coexpression networks was assessed across datasets. To evaluate the relevance of the coexpression modules to stroke, we integrated the network models with genome-wide association studies of ischaemic stroke from International Stroke Genetics Consortium and expression quantitative trait loci using a Marker Set Enrichment Analysis from the Mergeomics package. The coexpression modules demonstrating genetic association with stroke were then integrated with Bayesian gene regulatory networks to identify central regulators of the stroke-related networks using a Key Driver Analysis.
Results: We identified preserved networks of carotid plaque across studies. A total of 17 coexpression modules were enriched for genetic risks of stroke. Intriguingly, we found both well-known processes (such as toll-like receptor pathway, and homocysteine metabolism) and relatively novel processes (such as phagosome formation and maturation). Using Bayesian models, we identified central regulators in the networks such as F2, APOH, and AMBP.
Conclusion: By integrating multi-omics datasets, our network modeling provides molecular insights into the pathogenesis and regulators of carotid plaque and stroke.