Abstract
INTRODUCTION: Osteosarcoma (OS) is the most common primary bone malignancy, predominantly affecting children and adolescents. Current treatment approaches have limited efficacy, with a 5-year survival rate of approximately 60%. Epithelial-mesenchymal transition (EMT) plays a key role in the onset, progression, and metastasis of OS, potentially influencing patient prognosis. METHODS: We screened EMT-related genes from multiple transcriptomic datasets of OS and performed unsupervised consensus clustering of EMT-related gene sets. Key EMT-related genes were identified using weighted gene co-expression network analysis (WGCNA) and intersected with differentially expressed genes (DEGs) between OS and normal tissue samples. The least absolute shrinkage and selection operator (LASSO) algorithm was applied to screen candidate genes for developing a prognostic model. Single-cell RNA-Seq (scRNA-Seq) analysis was conducted on OS samples to identify cell populations expressing model genes. Functional validation was performed using si-GPC3 in the MG-63 cell line. RESULTS: The EMT-based prognostic model demonstrated strong predictive capacity across several validation cohorts. The model effectively predicted immune-related features and immunotherapy responses in high-risk and low-risk patient groups. Seven primary cell types were identified from scRNA-Seq data of OS samples, with the osteoblast population showing the highest proportion of cells positive for model genes. The OS_C3 subpopulation exhibited significantly higher scores and included nine gene modules associated with metabolism, structural integrity, proliferation, differentiation, adhesion, migration, immune responses, inflammatory reactions, and signal transduction. The model genes also demonstrated prognostic value across various cancer types. Knockdown of GPC3 in MG-63 cells resulted in decreased proliferation and migration ability. CONCLUSION: This study provides new insights into the potential mechanisms of EMT in OS and its impact on the tumor immune microenvironment and response to immunotherapy. These findings may pave the way for novel personalized treatment strategies for OS patients.