Integrating single-cell and bulk transcriptomes to reveal prognostic and immunological features of ecDNA-related genes in osteosarcoma.
The role of extrachromosomal DNA (ecDNA)-related genes in osteosarcoma remains largely unexplored. The aim of this study is to investigate the association between ecDNA-related genes and prognosis and tumor microenvironment (TME) in osteosarcoma.
Differential gene expression analysis of GEO datasets was conducted to identify ecDNA-related genes in osteosarcoma. Based on bulk RNA-seq data, a novel ecDNA-related Gene Prognostic Score Model (EGPSM) was developed using an integrated framework of 101 machine learning algorithms, which was validated in training, testing, and external cohorts. The associations between risk scores, prognosis, and TME characteristics were comprehensively evaluated. Single-cell RNA sequencing (scRNA-seq) data were further analyzed to elucidate the relationship between EGPSM, pro-tumor behaviors, and immune modulation in osteosarcoma, as well as to identify key prognostic genes involved in tumor progression. Lastly, we conducted in vitro and in vivo assays to characterize the biological roles of MTDH and to elucidate its regulatory effects on CD8⁺ T cell function.
A robust EGPSM was constructed, demonstrating superior predictive accuracy with a maximum C-index of 0.803. High-risk patients exhibited poorer survival, higher metastatic potential, and an "immune-cold" TME characterized by diminished CD8⁺ T/NK cell infiltration and impaired effector functions. Single-cell analysis confirmed the enrichment of malignant cells and depletion of T/NK populations with lower effector scores in the high-risk group. MTDH was identified as a key driver; functional assays showed it promotes proliferation and invasion while inhibiting apoptosis. Notably, MTDH knockdown potentiated CD8⁺ T-cell cytotoxicity by increasing the levels of granzyme B, IFN-γ, and perforin.
The newly developed EGPSM represents an effective tool for prognostic assessment and therapeutic stratification in osteosarcoma. MTDH may serve as a promising prognostic biomarker and therapeutic target.
Differential gene expression analysis of GEO datasets was conducted to identify ecDNA-related genes in osteosarcoma. Based on bulk RNA-seq data, a novel ecDNA-related Gene Prognostic Score Model (EGPSM) was developed using an integrated framework of 101 machine learning algorithms, which was validated in training, testing, and external cohorts. The associations between risk scores, prognosis, and TME characteristics were comprehensively evaluated. Single-cell RNA sequencing (scRNA-seq) data were further analyzed to elucidate the relationship between EGPSM, pro-tumor behaviors, and immune modulation in osteosarcoma, as well as to identify key prognostic genes involved in tumor progression. Lastly, we conducted in vitro and in vivo assays to characterize the biological roles of MTDH and to elucidate its regulatory effects on CD8⁺ T cell function.
A robust EGPSM was constructed, demonstrating superior predictive accuracy with a maximum C-index of 0.803. High-risk patients exhibited poorer survival, higher metastatic potential, and an "immune-cold" TME characterized by diminished CD8⁺ T/NK cell infiltration and impaired effector functions. Single-cell analysis confirmed the enrichment of malignant cells and depletion of T/NK populations with lower effector scores in the high-risk group. MTDH was identified as a key driver; functional assays showed it promotes proliferation and invasion while inhibiting apoptosis. Notably, MTDH knockdown potentiated CD8⁺ T-cell cytotoxicity by increasing the levels of granzyme B, IFN-γ, and perforin.
The newly developed EGPSM represents an effective tool for prognostic assessment and therapeutic stratification in osteosarcoma. MTDH may serve as a promising prognostic biomarker and therapeutic target.