Weighted Gene Coexpression Network Analysis (WGCNA) Modeling of Multiorgan Dysfunction Syndrome after Mechanical Circulatory Support Therapy

Abstract

Purpose MOD is a major cause of mortality after MCSD implantation in heart failure patients. We hypothesized that WGCNA can model the MOD inflammatory response. Methods and Materials We analyzed peripheral blood mononuclear cell gene expression of 29 MCSD patients between 3/2010 and 3/2011, and 8 age matched controls. MOD was defined by SOFA score obtained at median 8 days (25-75% IQR 7 - 14) postoperatively. SOFA risk was defined as low (≤ 4), medium (5-11) and high (≥12). Purified total RNA was amplified and hybridized on Illumina Whole Transcriptome microarrays. Statistical analysis was conducted on GeneSpring GX 12 using Kruskal-Wallis and adjusted for multiple testing (FDR Results Mean age was 57±15 years. 2444 gene transcripts were differentially expressed between controls and SOFA risk groups. Unsupervised WGCNA (Figure) identified major gene expression modules ( Fig 1 A) with different degrees of correlation to SOFA score. Comparison of module-characteristic eigengenes ( Fig 1 B) showed SOFA score was best correlated with the black module, composed of 125 genes (entities). Gene ontology and pathway analysis revealed α-defensins (DEFA1, DEFA3, DEFA4) as the most relevant members of this module, which are multifunctional mediators of innate and adaptive immunity. Conclusions WCGNA can facilitate identification of key biomarkers of the immune response triggered by MCSD implantation, including a close association between α-defensin pathways and MOD.