Dissecting the effect of sphingolipid metabolism gene in progression and microenvironment of osteosarcoma to develop a prognostic signature

Abstract

Sphingolipid metabolism (SM) fuels tumorigenesis and the malignant progression of osteosarcoma (OS), which leads to an unfavorable prognosis. Elucidating the molecular mechanisms underlying SM in osteosarcoma and developing a SM-based prognostic signature could be beneficial in the clinical setting. This study included 88 frozen OS samples to recognize the vital SM-relevant genes in the development of OS utilizing univariate Cox regression. The Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis was conducted on the SM- relevant genes to minimize the risk of overfitting. The prognostic signature was generate utilizing the multivariable Cox regression analysis and was verified in the validation cohort. Moreover, cellular and molecular mechanisms associated with SM have an unfavorable prognosis for OS patients and have been widely studied. Resultantly, an SM-based prognostic risk model was established according to critical prognostic genes (CBS, GLB1, and HACD1), which had an excellent ability to predict the prognosis of OS patients (AUC for the train cohort was 0.887 and AUC for validation cohort was 0.737). The high-risk OS patients identified based on this prognostic signature had significantly poor immune microenvironment, indicated by significantly low immune score (mean=216.290 ± 662.463), reduced infiltrations of 25 immune cells, including NK cells (LogFC= -0.3597), CD8+T cells ((LogFC=-0.2346), Cytolytic activity ((LogFC=-0.1998), etc. The immunosuppressive microenvironment could be due to dysregulated SM of glycolipids. Further, a nomogram was constructed by integrating the SM-based prognostic signature and clinical paraments to facilitate clinical application. The nomogram could accurately predict the prognosis of OS invalids. Collectively, this study clarified the function of SM in the development of OS and helped develop a tool for risk stratification based on SM-related genes with application in clinical settings. The results of our study will aid in identifying high-risk patients and provide individualized treatments.