Identification of biomarkers associated with endoplasmic reticulum stress-related cell death in osteoporosis based on bulk and single-cell transcriptomic analyses and experimental validation

Osteoporosis (OP) is a metabolic bone disease characterized by low bone mineral density (BMD), and its pathogenesis involves endoplasmic reticulum (ER) stress-related cell death. This study aimed to identify diagnostic biomarkers associated with ER stress-related cell death in OP and explore their underlying mechanisms. The training dataset (GSE56815), validation dataset (GSE56814), and single-cell RNA sequencing (scRNA-seq) dataset (GSE147287) were downloaded. Differentially expressed genes (DEGs) between OP patients and controls were identified. Candidate genes were obtained by intersecting DEGs with ER stress-related genes and programmed cell death (PCD)-related genes. Machine learning was used to screen intersection genes, and biomarkers were determined via expression level analysis. Gene set enrichment analysis (GSEA), immune cell infiltration analysis, drug prediction and molecular docking, scRNA-seq analysis, key cell screening, cell communication analysis, and pseudotime analysis were performed. Finally, reverse transcription quantitative polymerase chain reaction (RT-qPCR) were further conducted. A total of 28 candidate genes were obtained by intersection. CAMKK2 and DAPK3 were confirmed as biomarkers, and were consistently down-regulated in both datasets and verified by RT-qPCR. GSEA analysis revealed that biomarkers were enriched in cytokine-cytokine receptor interaction. Correlations between biomarkers and activated dendritic cells were found via immune cell infiltration analysis. Preliminary computational analyses indicated that drugs including calcitriol and danazol may potentially interact with the biomarkers in a stable manner. Bone marrow-derived mesenchymal stem cells (BM-MSCs) were identified as potential key cells via scRNA-seq analysis. Complex interactions involving BM-MSCs, such as ANGPTL4-CDH11 mediating BM-MSC self-communication, were revealed by cell communication analysis. Dynamic expression of biomarkers during BM-MSC differentiation was shown by pseudotime analysis: CAMKK2 fluctuated with differentiation stages, while DAPK3 shifted from high to low then high expression. CAMKK2 and DAPK3 were confirmed as diagnostic biomarkers for OP, providing insights into OP diagnosis and potential therapeutic targets.