circRNA-miRNA-mRNA networks reveal a proangiogenic action of circNPHP1 in human ischemic heart disease

Abstract

BackgroundIschemic heart disease (IHD) is characterized by insufficient myocardial blood flow and aggravated by diabetes. IHD is often treated by coronary revascularization. Complementary therapies that increase the microvascular flow by promoting endothelial homeostasis and repair are still lacking and require the identification of new therapeutic targets. The expression of protein coding genes is regulated by multimodal interactions between different classes of non-coding RNAs. This study aimed to 1) unravel circular RNA (circRNA) -microRNA (miRNA) - messenger RNA (mRNA) networks across the human ischemic and control hearts; 2) identify and functionally test the sub-network with highest potential to control endothelial cell (EC) function.MethodsTwo intraoperative left ventricle biopsies were prospectively collected from type-2 diabetic and non-diabetic people with IHD and from a control group of non-IHD, non-diabetic patients (N=12 patients/group; ARCADIA study) and used for long transcriptome bulk-sequencing (circRNAs and mRNAs datasets) and small RNA-sequencing (miRNAs datasets). Differentially expressed (DE) RNAs were identified with a customized bioinformatics pipeline. We identified potential “sponging associations” between circRNAs and miRNAs and between miRNAs and mRNAs. Next, we created circRNA-miRNA-mRNA networks and retained the subnetworks where individual RNAs were expressed in ECs (using datasets from GEO database) and of potential relevance to endothelial function based on pathway analyses. Next, we screened top differentially expressed circRNAs in cultured ECs exposed to disease-mimic conditions vs control. The circRNA-miRNA sponging interactions of the top emerging circRNA candidate were confirmed after circRNA pulldown in ECs and the functional relevance was mechanistically tested in EC cell biology assays.ResultsWe created circRNA-miRNA-mRNA networks in IHD (with and without T2DM) and controls indicating distinct interactions in each. We unravelled a novel putative proangiogenic subnetwork, circNPHP1/miR-221-3p/VEGFA/BCL2, in the human ischemic heart specially enriched in IHD and T2DM. CircNPHP1 pull-down assay confirmed it’s binding to miR-221-3p. CircNPHP1 silencing in ECs reduced angiogenesis and decreased the expression of miR-221-3p target genes (VEGFA and BCL2). The proangiogenic action of circNPHP1/miR-221-3p was confirmed in ECs by combining circNPHP1 silencing with either miR-221-3p inhibition or overexpression in ECs.ConclusionsCircNPHP1 supports endothelial function through sponging of miR-221-3p in IHD and holds potential as novel therapeutic target in IHD.