DOI: 10.7717/peerj.21591 ISSN: 2167-8359
Identifying miR-431-5p as a promising candidate biomarker for early-onset severe preeclampsia through integrated bioinformatic and clinical analyses: a retrospective study
Jianxin Zhang, Hongwei Li, Yu Yan, Xiaoxiao Li, Linlin Zhang, Yangnan Ding, Jun Ma, Enwu Yuan, Xin Zhao
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
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.