Exploration of metastasis-related signatures in osteosarcoma based on tumor microenvironment by integrated bioinformatic analysis

Background: The present study aims to explore the metastasis-related signatures in connection with tumor microenvironment (TME), revealing new molecular targets promising in improving osteosarcoma (OS) patients' outcomes.

Methods: The high-throughput sequencing data was downloaded from the TARGET database and performed the ESTIMATE algorithm. Metastasis-related information was obtained from the GSE21257 dataset. Differentially expressed genes (DEGs) associated with the stromal and immune cell infiltration patterns were identified. DEGs with similar biological functions were grouped into the same module by Gene Ontology (GO) analysis and MCODE analysis. Prognostic DEGs were selected in two datasets through survival analysis. Weighted gene co-expression network analysis (WGCNA) was performed to find metastasis-related modules and genes. RT-PCR was utilized to evaluate the expression of the key prognostic DEGs associated with metastasis in OS patients.

Results: The median scores of the stromal and immune groups of OS samples were 58 and -416, and a total of 200 overlapping DEGs were identified. These DEGs basically played fundamental roles in immune response relevant GO terms and were clustered into 9 different modules. Among them, 24 metastasis-related DEGs were selected from the GSE21257 dataset which contains the stromal and immune cell infiltration patterns. Finally, IRF8, HLA-DMA, and HLA-DMB were proved to exhibit significant higher expression levels in cancerous tissues than in para-cancerous tissues for OS patients.

Conclusion: We identified three principal genes as promising signatures for predicting the survival the prognosis of OS patients. Exploration of metastasis-related signatures in TME may be valuable for enhancing treatment strategies for OS.