Identification and Validation of Biomarkers Associated With Mitochondrial Dynamics in Preeclampsia

Despite studies highlighting mitochondrial dynamics in preeclampsia (PE), the regulatory mechanisms of mitochondrial dynamics-related genes (MD-RGs) in PE remain unclear. This study aims to explore these mechanisms.

PE-related datasets (GSE10588 and GSE98224) and 23 MD-RGs were used. Differential expression, protein-protein interaction (PPI) network, machine learning (ML), gene expression, and receiver operating characteristic (ROC) analyses were conducted to screen for PE biomarkers. A nomogram was established to predict PE occurrence. Gene set enrichment analysis (GSEA) and immunoinfiltration analysis were subsequently performed.

Reverse transcription quantitative PCR (RT-qPCR) was used to measure biomarker expression, identifying GLUL, GPX3, and KCND2 as PE biomarkers. GPX3 and KCND2 were upregulated, whereas GLUL was downregulated in PE. These biomarkers had area under the curve (AUC) values of 0.7-0.9 in the datasets, and the integrated nomogram reached 0.973, indicating precision in PE prediction. GSEA revealed that ribosome and E. coli coenriched KCND2 and GLUL. Immunoinfiltration analysis revealed that Tregs were significantly correlated with KCND2 and GLUL among differential immune cells in PE patients versus controls. Finally, expression validation revealed that GLUL was significantly upregulated and expressed in the control group, whereas GPX3 and KCND2 were highly expressed in PE patients, with KCND2 reaching significant levels, which is consistent with the bioinformatics analysis.

In total, 3 mitochondrial dynamics-related biomarkers (GLUL, GPX3, and KCND2) were identified, and the biomarker-based nomogram was used to predict the onset risk of PE patients. The mining of these biomarkers contributed to an in-depth understanding of the pathogenesis of MD-RGs in PE.