Identification of GPX3 as a key biomarker of placental ferroptosis in gestational diabetes mellitus via bioinformatics and clinical analysis.

Gestational diabetes mellitus (GDM) is characterized by glucose intolerance during pregnancy, and emerging evidence implicates dysregulated iron metabolism as a critical modulator of its pathogenesis. Ferroptosis, an iron-mediated cell death, has recently been studied in GDM, with research beginning to unravel the connection between iron-induced oxidative stress and placental dysfunction. In this study, we employed datasets from the Gene Expression Omnibus database to identify markers of ferroptosis that are associated with GDM. A total of 57 differentially expressed genes related to ferroptosis were identified. Feature selection was performed using machine learning approaches, including Boruta, Random Forest, and LASSO regression, to pinpoint the most critical genes. Among them, GPX3 emerged as the central biomarker linked to ferroptosis in GDM. We further validated GPX3 expression across various placental cell types using single cell RNA sequencing data. Further CIBERSORT analysis determined a significant association between GPX3 and several immune cell populations, including macrophages, B cells, monocytes, and T cells. Finally, mRNA expression of GPX3 was experimentally validated in placental samples from GDM patients, where it was found to correlate with a reduced sTFR/ferritin ratio, suggesting disrupted iron homeostasis. In conclusion, GPX3 is identified as a crucial immuno-ferroptotic biomarker in GDM, with potential diagnostic value. Integrating bioinformatics, machine learning, and clinical validation, this study highlights the role of GPX3 at the intersection of immune infiltration and iron metabolism, offering new insights for future diagnostic and therapeutic strategies in GDM.