Identification of diagnostic biomarkers of gestational diabetes mellitus based on transcriptome gene expression and alternations of microRNAs

Abstract

<abstract><sec> <title>Background</title> <p>Gestational diabetes mellitus (GDM), characterized by glucose intolerance during pregnancy, poses substantial health risks for both mothers and infants due to the interplay of insulin resistance and β-cell dysfunction. Molecular biomarkers, including SNPs, microRNAs (miRNAs), and proteins, have been linked to GDM development during pregnancy. Notably, miRNA-mediated regulation of gene expression holds pivotal roles in metabolic disorders. This study aims to identify diagnostic biomarkers for GDM and establish a diagnostic model.</p> </sec><sec> <title>Methods</title> <p>Firstly, gene expression data from GDM samples (N = 9) and normal samples (N = 9) were sourced from the Gene Expression Omnibus (GEO) database. Subsequently, the limma package was employed to discern differentially expressed genes (DEGs), with subsequent functional and enrichment analyses executed using the clusterProfiler package. A comprehensive exploration of genes significantly correlated with GDM was undertaken via weighted gene co-expression network analysis (WGCNA). The construction of a protein-protein interaction (PPI) network was facilitated by STRING, while visualization of hub genes was achieved through Cytoscape. Moreover, the miRNA-mRNA network was established using StarBase. Concurrently, immune infiltration significantly correlated with hub genes was identified.</p> </sec><sec> <title>Results</title> <p>In this study, 209 DEGs between normal and GDM samples were identified, and these genes were associated with collagen containing extracellular matrix heparin binding and axon guidance, etc. Then, 18 modules were identified by WGCNA and the brown module including 212 genes had a significantly negative correlation with GDM (r = −0.66, P = 0.003). Additionally, five low gene expressions (CXCL12, MEF2C, MMP2, SOX17 and THBS2) and two high gene expressions (BMP4 and SFRP5) were identified as GDM related hub genes. Moreover, hub genes regulated by alternations of miRNAs were established and three hub genes (CXCL12, MEF2C and THBS2) were negatively correlated with activated Natural Killer (NK) cells while two hub genes (BMP4 and SFRP5) were positively correlated with activated NK cells.</p> </sec><sec> <title>Conclusions</title> <p>This study offers novel hub genes that could contribute to the diagnostic approach for GDM, potentially shedding light on the intricate mechanisms underpinning GDM's developmental pathways.</p> </sec></abstract>