Construction of a ceRNA network reveals functional lncRNAs in left ventricular noncompaction through weighted gene co-expression network analysis

Background Left ventricular noncompaction (LVNC) is a rare genetic cardiomyopathy. Lack of biomarkers and insights into epigenetic mechanism limits its screening, diagnosis and therapy. This study aimed to delineate the lncRNA-miRNA-mRNA interaction underlying LVNC through weighted gene co-expression network analysis (WGCNA) and construction of a competing endogenous RNA (ceRNA) network. Methods By analyzing microarray data from GSE178582, we acquired lncRNA and mRNA expression profiling in LVNC. Differentially expressed lncRNAs and mRNAs were identified using R package “limma”. Gene set enrichment analysis (GSEA) was conducted to detect biological function alteration in LVNC. We performed WGCNA to pick cardinal gene modules in LVNC and supplemented enrichmsent annotation based on mRNA modules. Interactions of lncRNA-miRNA and miRNA-mRNA were respectively predicted by LncBase and miRTarBase. In comprehensive consideration of expression change, gene clustering results and database prediction, we constructed a ceRNA network and visualized the lncRNA-miRNA-mRNA interaction. Then, a subnetwork centered on hub genes was extracted. Results 2452 mRNAs (1275 upregulated and 1177 downregulated) and 1415 lncRNAs (839 upregulated and 576 downregulated) were differentially expressed between LVNC patients and healthy controls. Enriched pathways involved acute myeloid leukemia, bile secretion, thyroid hormone synthesis etc. A ceRNA network with 3 lncRNAs, 5miRNAs and 106 mRNAs was constructed. Furthermore, we identified 9 hub mRNAs and visualized their lncRNA-miRNA-mRNA subnetwork. Conclusions The ceRNA regulatory network participates in the pathogenesis of LVNC, and identified hub genes may be potential targets for intervention and biomarkers to screen LVNC patients.