Comprehensive analysis of copper-metabolism-related genes about prognosis and immune microenvironment in osteosarcoma

Abstract

Despite being significant in various diseases, including cancers, the impact of copper metabolism on osteosarcoma (OS) remains largely unexplored. This study aimed to use bioinformatics analyses to identify a reliable copper metabolism signature that could improve OS patient prognosis prediction, immune landscape understanding, and drug sensitivity. Through nonnegative matrix factorization (NMF) clustering, we revealed distinct prognosis-associated clusters of OS patients based on copper metabolism-related genes (CMRGs), showing differential gene expression linked to immune processes. The risk model, comprising 13 prognostic CMRGs, was established using least absolute shrinkage and selection operator (LASSO) Cox regression, closely associated with the OS microenvironment's immune situation and drug sensitivity. Furthermore, we developed an integrated nomogram, combining the risk score and clinical traits to quantitatively predict OS patient prognosis. The calibration plot, timeROC, and timeROC analyses demonstrated its predictable accuracy and clinical usefulness. Finally, we identified three independent prognostic signatures for OS patients: COX11, AP1B1, and ABCB6. This study confirmed the involvement of CMRGs in OS patient prognosis, immune processes, and drug sensitivity, suggesting their potential as promising prognostic signatures and therapeutic targets for OS.

Figure 1

The role of copper metabolism in osteosarcoma. (A) Clustering based on CMRGs (B,C) Prognosis of different clusters; (D––F) volcano plot, GO, and KEGG analyses of DEGs between different clusters. (G) GSEA analysis of different clusters. (H,I) ESTIMATE analysis of different clusters. CMRGs copper-metabolism-related genes, DEGs differentially expressed genes.

Figure 2

Construction and validation of the prognosis risk model. (A–F) Prognostic value in training and testing cohort.

Figure 3

Immune infiltration and functional enrichment of DEGs between high risk and low risk group (A,B) CIBERSORT. (C,D) ESTIMATE. (E) Functional enrichment.

Figure 4

Immune-related analysis of risk score and construction and validation of the integrated nomogram. (A) Cancer immunity cycles. (B) Immunotherapy-predicted pathways. (C,D) Nomogram was constructed based on CMRGs-based risk scores and metastasis. (E) 5-year calibration plots of the nomogram in TARGET-OS cohort. (F) 5-year calibration plots of the nomogram in GSE21257 cohort. (G) The timeROC curve of the nomogram. (H) The timeDCA of the nomogram. (I) Four potential drugs of PI3K/Akt pathway for the treatment of OS. ROC receiver operating characteristic curve, DCA decisions curve analysis.

Figure 5

Expression validation of CMRGs (A) Multivariate Cox regression analysis of prognostic CMRGs. (B) Kaplan–Meier plot of COX11, AP1B1, and ABCB6. (C) Immunohistochemical staining of COX11, AP1B1, and ABCB6.