Exploration of potential mechanisms and biomarkers related to ERS-associated RCD in steroid-induced osteonecrosis of the femoral head based on bioinformatics, with experimental validation

Background The pathogenesis of osteonecrosis of the femoral head (ONFH) is associated with endoplasmic reticulum stress (ERS) and regulated cell death (RCD), but reliable linking biomarkers and their mechanisms remain unclear. This study aimed to screen SONFH biomarkers related to ERS/RCD and explore their mechanisms. Methods Using datasets from public databases on ONFH, ERS-related genes (ERGs), and RCD-related genes (RRGs), candidate genes were isolated by intersecting the differentially expressed genes (DEGs) with the ERG and RRG sets. We identified biomarkers by integrating a protein-protein interaction (PPI) network, four computational algorithms, and machine learning, with results refined by the Wilcoxon rank-sum test. The diagnostic power of these biomarkers was assessed using receiver operating characteristic (ROC) curve analysis. A subsequent comprehensive investigation included constructing and evaluating a nomogram, along with analyses of pathway enrichment, immune infiltration, drug prediction, and molecular docking, culminating in experimental validation. Results Initially, the 100 candidate genes were identified. After multi-step screening, 2 biomarkers (TGFB1, MCL1) were finally determined. Both were significantly upregulated in the ONFH group (p < 0.05), and the area under the ROC curve (AUC) of both was greater than 0.85. The constructed nomogram showed good predictive accuracy (AUC = 0.897, Hosmer-Lemeshow test p = 0.171) and had a net benefit greater than zero, indicating good clinical practicality. These biomarkers were enriched in pathogen-related pathways, correlated with immune cells such as plasmacytoid dendritic cells, and exhibited good molecular docking effects with domperidone (Vina score ≤ -7.7 kcal/mol). RT-qPCR and western blot validation in GC-treated cell models (BMSC, BMEC, THP-1) confirmed significant upregulation of both TGFB1 and MCL1, consistent with the bioinformatics findings. Conclusion TGFB1 and MCL1 could serve as potential SONFH biomarkers related to ERS/RCD. The nomogram, as well as the two biomarkers’ regulatory and immune associations, provided a reference for ONFH diagnosis. Domperidone shows potential as a candidate drug for the targeted therapy of ONFH, although its clinical application remains pending in-vivo efficacy and clinical cohort validation.