Identification of paraptosis related genes in osteosarcoma and development of a risk model based on single cell RNA sequencing and bulk RNA sequencing data

Background The induction of paraptosis in osteosarcoma (OS) cells through targeted therapies has the potential to impede tumor growth. Nevertheless, the molecular mechanism of paraptosis-related genes (PRGs) in OS is currently unclear. Hence, this study set out to investigate the related genes and mechanisms of paraptosis in OS. Methods This study incorporated OS-related datasets and PRGs-related gene sets. Prognostic genes were obtained using consensus clustering, differential expression analysis, univariate Cox analysis, and least absolute shrinkage and selection operator (LASSO) regression analysis. Next, the risk model went through additional development and validation. Moreover, Gene Set Enrichment Analysis (GSEA), immune infiltration, and immunotherapy were employed to investigate the molecular mechanisms underlying high-risk group (HRG) and low-risk group (LRG). Ultimately, single-cell RNA sequencing (scRNA-seq) analysis was executed to probe the molecular mechanisms of cells in the development of OS. Results The risk score was calculated through two prognostic genes (C2 and DLGAP1) to classify the OS samples into HRG and LRG. The risk model and nomogram could precisely forecast the prognosis of OS patients. Subsequently, the systemic lupus erythematosus and leishmania infection were significantly enriched in two risk groups. Next, there were 27 notable differences immune cells between HRG and LRG, including activated B cells and activated dendritic cell. Furthermore, there were notable differences in the dysfunction, exclusion and microsatellite instability (MSI) scores between the HRG and LRG. Finally, macrophages were considered as key cells relied on the expression of prognostic genes, together with the pseudotime analysis of key cells revealed that the dynamic expression patterns of C2 during the process of macrophage differentiation. Conclusion This study established a prognostic model associated with paraptosis in OS. This prognostic model was intricately linked to the tumour microenvironment in OS, and had the potential to improve prognostic in OS.