Abstract
Long non-coding RNAs (lncRNAs) were considered to be involved in vascular complications in diabetes mellitus, but still only limited knowledge in this regard has been obtained. Herein, we further explored the roles of lncRNAs and mRNAs in diabetic vasculopathy (DV) through conducting bioinformatics analysis using data set downloaded from GEO database. The differentially expressed lncRNAs and mRNAs were identified by edge package. GO enrichment analysis and KEGG pathway analysis were performed based on clusterprofiler package. The relationship between lncRNA and miRNA was predicted using starBase database, and the potential mRNAs targeted by miRNAs were predicted by TargetScan, miRTarbase and miRDB database. The string database was used to analyze the protein-protein interaction (PPI). As a result, a total of 12 lncRNAs and 711 mRNAs were found to be differentially expressed in the diabetic subdermal endothelial cells compared with normal controls. A ceRNA network was established, which was composed of seven lncRNA nodes, 49 miRNA nodes, 58 mRNA nodes and 183 edges, and MSC-AS1 and LINC01550 may serve as key nodes. GO function enrichment analysis showed enrichments of epithelial cell proliferation, intercellular junction, and cell adhesion molecule binding. KEGG pathway analysis revealed 33 enriched pathways. PPI protein interaction analysis identified 57 potential ceRNA-related proteins. Overall, this study suggests that multiple lncRNAs, specifically MSC-AS1 and LINC01550, may play an important role in DV development and they are like to be developed as the therapeutic targets for DV. However, further experiments in vitro and in vivo should be conducted to validate our results.