Endoplasmic reticulum stress-related genes drive M1 macrophage polarization in preeclampsia via modulating metabolic reprogramming: a bioinformatic study

BACKGROUND: Preeclampsia (PE), a serious obstetric complication impacting maternal and fetal health, still lacks reliable biomarkers owing to limited sensitivity, specificity, and disease heterogeneity. METHODS: Differentially expressed genes (DEGs) were identified from placental samples in GSE114691 (20 PE, 21 controls), and module genes were determined via weighted gene co-expression network analysis (WGCNA). These genes were intersected with ERS genes. Feature genes were selected using LASSO, RF, and SVM-RFE, and a diagnostic model was constructed. Single-gene gene set enrichment analysis (GSEA), interaction network analysis (including PPI, miRNA/TF-target gene, chemical-gene interaction), immune infiltration, and molecular docking were performed. RESULTS: Using analyses performed in R, along with WGCNA and ML algorithms, four ERS-related feature genes (HTRA1, GBA1, KL, PC) were identified, which showed high discriminatory power (AUC = 0.846). GSEA linked these genes to metabolic reprogramming, involving central energy metabolism, amino acid metabolism, short-chain fatty acid metabolism, and redox homeostasis. Immune infiltration analysis showed KL and PC negatively correlated with M1 but positively with M2 macrophages, opposite to HTRA1 and GBA1. Molecular docking showed stable binding between aspirin and the signature gene-encoded proteins. CONCLUSIONS: HTRA1, GBA1, KL, and PC may serve as diagnostic biomarkers for PE, potentially influencing M1 polarization of placental macrophages via metabolic reprogramming.