In-Silico Analysis of Key Regulatory Pathways and Hub Genes From Peripheral Blood in Type 2 Diabetes Mellitus

AbstractBackground: Type II diabetes mellitus (T2DM), characterized by fasting hyperglycemia and impaired insulin secretion and action, is a global health burden. Despite the advances in this field, the mechanism underlying T2DM is far from clear. Objective: The present study sheds light upon a systematic evaluation of the genes, pathways, and interaction networks underlying T2DM with the aid of bioinformatics. Methods: Two Gene Expression Omnibus microarray datasets: GSE148961 and GSE26168 were selected for this study. The common differentially expressed genes (DEG) were sorted by the cutoff |logFC|≥1.0 for the first dataset and |logFC|≥0.263 for the second. Gene Ontology (GO), functional enrichment, and protein-protein interaction (PPI) network were analyzed in Search Tool for the Retrieval of Interacting Genes (STRING). The MCODE and CytoHubba plugins in Cytoscape (v3.7.2) were used to identify gene clusters and top hub genes, respectively. Top 10 nodes were ranked in CytoHubba according to MCC, DMNC, MNC, Degree, and EPC methods, and genes common in at least 3 methods were selected as top nodes. Results: 88 common DEGs were identified by Venn diagram (http://bioinformatics.psb.ugent.be/cgi-bin/liste/Venn/calculate_venn.htpl). GO analysis had 91 significantly enriched biological processes, including regulated exocytosis, secretion, vesicle mediated transport, antibacterial humoral response, and neutrophil degranulation. 4 molecular functions- fibrinogen binding, fibronectin binding, lipopolysaccharide binding, and Extracellular matrix binding; and 38 cellular components, including secretory vesicle, endomembrane system, and adherens junction were significant. The PPI network was highly significant (p-value < 0.001) at medium confidence (0.400) with 88 nodes, 140 edges, and an average node degree of 3.18. The MCODE plugin revealed two clusters, the former with 14 nodes, 73 edges, and a score of 9.733, and the latter with 6 nodes, 14 edges, and a score of 4.000. 9 candidate genes: ELANE, DEFA4, BPI, MPO, LTF, CAMP, OLFM4, LCN2, and VCL were identified, amongst which ELANE, LCN2, and MPO are associated with T2DM pathogenesis, while BPI and LTF have protective effects. OLFM4 deletion has been observed to improve glucose tolerance in mice models. Conclusion: This study provides a comprehensive analysis of genes, pathways, and functions which may be pivotal in T2DM pathogenesis and may represent potential therapeutic targets.