Identifying miR-431-5p as a promising candidate biomarker for early-onset severe preeclampsia through integrated bioinformatic and clinical analyses: a retrospective study
Background Early-onset severe preeclampsia (EOSPE) is a serious pregnancy complication associated with adverse maternal and neonatal outcomes. This study aimed to identify potential diagnostic biomarkers for EOSPE and to investigate their clinical significance. Methods MicroRNAs (miRNAs) expression datasets from the Gene Expression Omnibus (GEO) database, specifically from placental tissue (GSE103542) and plasma (GSE234611), were screened to identify miRNAs consistently dysregulated in early-onset preeclampsia (EOPE) and to discover robust biomarker candidates. Weighted gene co-expression network analysis (WGCNA) was employed to identify key modules associated with preeclampsia (PE). The differentially expressed miRNAs (DEmiRNAs) were intersected with miRNAs from WGCNA-identified modules to prioritize candidates. Functional enrichment analysis predicted target genes and involved pathways. Clinical samples (serum and placental tissues from EOSPE patients and normal pregnant women) and a lipopolysaccharide (LPS)-induced preeclampsia rat model were utilized to validate the expression of DEmiRNAs via reverse transcription quantitative polymerase chain reaction (RT-qPCR). The diagnostic value of these miRNAs was evaluated using receiver operating characteristic (ROC) curve analysis, and their correlations with clinical parameters were assessed. Results Bioinformatics analysis identified miR-431-5p as a key molecule, with its expression was significantly upregulated in both serum and placental tissues of patients with EOSPE. Circulating miR-431-5p demonstrated promising diagnostic potential for EOSPE (AUC = 0.803). Its expression level positively correlated with systolic blood pressure (r = 0.395, P = 0.017) and diastolic blood pressure (r = 0.41, P = 0.013), uric acid (r = 0.389, P = 0.02), lactate dehydrogenase (r = 0.399, P = 0.018), and the umbilical artery S/D ratio (r = 0.457, P = 0.01). Conversely, it negatively correlated with gestational age at delivery (r = −0.562, P < 0.001) and neonatal birth weight (r = −0.503, P = 0.02). Multivariable regression analysis showed that after adjusting for confounding factors, miR-431-5p remained significantly positively associated with the occurrence of EOSPE. Animal experiments confirmed similar upregulation of miR-431-5p in the LPS-induced preeclampsia rat model. Conclusions In this study, miR-431-5p was significantly upregulated in EOSPE and demonstrated preliminary diagnostic potential. Its expression level was potentially associated with disease severity and adverse perinatal outcomes. These findings suggest that miR-431-5p may serve as a promising candidate biomarker for EOSPE.