Identification and experimental validation of biomarkers associated with the Wnt signaling pathway in preeclampsia

Preeclampsia (PE) is a pregnancy disorder associated with hypertension. The Wnt signaling pathway regulates human cell proliferation, migration, and apoptosis. This study aimed to develop a blood-based (non-invasive) diagnostic biomarker panel for PE based on Wnt signaling pathway-related genes (WSPRGs). Placental transcriptomic datasets were integrated to substantiate the pathological foundation of these circulating biomarkers in placental dysfunction. A total of 11 differentially expressed WSPRGs (DE-WSPRGs) were identified through overlap analysis between WSPRGs and differentially expressed genes (DEGs) screened from GSE48424 and GSE114691. Subsequently, three biomarkers (MAPK8, CSNK1E, and NOTUM) were identified using two machine learning algorithms and receiver operating characteristic curves. The diagnostic nomogram constructed based on these biomarkers demonstrated excellent diagnostic performance (area under the curve, AUC = 0.920). Additionally, CD56 dim natural killer cells correlated significantly with MAPK8 and CSNK1E (|correlation coefficient (cor)| > 0.3; P < 0.05). MAPK8 was associated with 11 diseases. Notably, the drug with the highest interaction score was 4-(3-(pyridin-2-yl)-1H-pyrazol-4-yl)quinoline, which was associated with CSNK1E. Moreover, MAPK8 corecruited six miRNAs. Experimental validation showed that CSNK1E and NOTUM expression levels were significantly elevated in the PE group ( P < 0.05), consistent with the reverse transcription quantitative polymerase chain reaction results, whereas MAPK8 was significantly downregulated in PE clinical samples ( P < 0.05). In summary, through a retrospective bioinformatic pipeline and pilot-level RT-qPCR validation, this study identified three candidate Wnt pathway-related biomarkers (MAPK8, CSNK1E, and NOTUM) exhibiting diagnostic associations with established PE. However, before their clinical applicability can be assessed, prospective validation in adequately powered and phenotypically diverse cohorts is required.