Chronic heart failure is associated with adverse remodeling of the heart, leading to cardiac dysfunction and patient morbidity and mortality. There is increasing evidence for a role of non-myocytes in this process. Endothelial cells form the inner lining of the blood vessel system and are thus directly exposed to mechanical and biochemical factors. They regulate the supply of oxygen and nutrients to the tissue as well as the immigration of cells from the bloodstream and thus have an important barrier function. In addition, endothelial cells interact with other cells of the heart via cell-cell contacts and a variety of paracrine mediators and thus influence cardiac homeostasis. The aim of this study was to reveal gene expression in endothelial cells of failing human hearts and to identify disease-relevant gene clusters and corresponding hub genes.

 

To assess differences in gene expression we used left ventricular samples from patients diagnosed with end-stage heart failure obtained during surgery (n=15) and samples from healthy organ donors (n=2). We established a protocol to isolate endothelial cells from snap frozen heart tissue for RNA sequencing. Single cell suspensions were obtained by combined mechanical and enzymatic disruption. Endothelial cells were labelled with FITC-conjugated lectin and PE-conjugated anti-CD144 and isolated by fluorescence assisted cell sorting (FACS) for RNA sequencing. The ratio of marker gene expression for endothelial cells (PECAM1, CDH5) versus cardiomyocyte marker genes (MYH7, PLN, ATP2A2) was 8.725 ± 1.094 (SEM) in isolated endothelial cells and 0.027 ± 0.002 (SEM) in heart tissue RNA sequencing data obtained from a published reference set.

 

We identified 447 genes which were differentially expressed in endothelial cells in heart failure compared to cells from control hearts (q<0.05). Subgroup analysis of patients with different heart failure causes (ischemic vs non-ischemic) or co-morbidities (arterial hypertension, heart valve disease, diabetes mellitus, chronic kidney disease, or lung disease) revealed no impact on differential gene expression. In a weighted coexpression network analysis, we identified 26 gene clusters, of which five clusters showed a significant positive correlation (r range = 0.51 to 0.74; p<0.05) and four clusters displayed a significant negative correlation (r range = -0.5 to -0.76; p<0.05) with the occurrence of heart failure. Genes within those clusters were associated with inflammation, vasoconstriction, steroid and angiotensin hormone receptor pathways, transcription, and DNA methylation. Hierarchical gene ranking identified hub genes as potential upstream regulators of these processes in cardiac endothelial cells.

 

Heart failure is associated with substantial changes in cardiac endothelial cell gene expression. Gene network analysis revealed disease-related gene clusters and associated hub genes that are now explored as potential biomarkers or therapeutic targets.