Scholarlink

Background <p>The high recurrence and metastasis rates of oral squamous cell carcinoma (OSCC) pose major therapeutic challenges, with tumor stem cell properties playing a pivotal role. This study aims to systematically explore the tumor stem cell characteristics of OSCC through multi-omics data and establish a reliable prognostic risk assessment model.</p> Methods <p>The Cancer Genome Atlas (TCGA)-OSCC dataset was obtained from UCSC-Xena and combined with the Gene Expression Omnibus (GEO) database data for multi-omics analysis. Tumor stemness index (mRNAsi) was calculated using the One-Class Logistic Regression (OCLR) algorithm, and gene modules significantly correlated with mRNAsi were identified via weighted gene co-expression network analysis (WGCNA). Univariate Cox and LASSO regression models were constructed to develop mRNAsi-associated risk scores. Immunological microenvironment characteristics and immunotherapy response potential across risk groups were further analyzed using ESTIMATE, MCP-counter, single-sample gene set enrichment analysis (ssGSEA), and tumor immune dysfunction and exclusion (TIDE) algorithms. Finally, a cellular atlas was constructed from OSCC single-cell RNA&#xa0;sequencing data to explore stemness features in epithelial cell subpopulations. Wet experiments were conducted to verify the expression and function of genes.</p> Results <p>Based on WGCNA, we identified gene modules significantly positively correlated with the tumor stem cell index (mRNAsi), primarily enriched in pathways such as cell cycle and chromosome segregation. Subsequently, we identified eight key genes (<i>RNF128, KIAA0513, MPP4, TYMS, CIB2, CELSR3, HRG, CORO2A</i>) and constructed a risk scoring model. This model demonstrated robust prognostic predictive capability across the training set, test set, and external validation cohort, with patients exhibiting high-risk scores showing poorer outcomes. Immune cell infiltration analysis revealed that tumors in the low-risk group exhibited higher levels of immune cell infiltration (such as T cells, neutrophils, and fibroblasts) and demonstrated superior potential for immune therapy response. Single-cell transcriptomic analysis revealed the cellular composition landscape of OSCC, with epithelial cells constituting the largest proportion. Furthermore, seven epithelial cell clusters were identified within this cell population, among which cluster_2 exhibited significantly higher stem cell characteristics. Finally, in vitro functional experiments demonstrated that knocking down <i>CORO2A</i> effectively suppressed the migration and invasion capabilities of OSCC cells.</p> Conclusion <p>We systematically elucidated tumor stemness-related molecular features in OSCC, established an mRNA-based risk model, and clarified its relationship with the immune microenvironment and epithelial stemness subpopulations, providing a theoretical basis for OSCC prognosis assessment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrating multi-omics analysis and functional validation to identify key gene in the stemness regulatory network of oral squamous cell carcinoma

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月29日

查看全文
  • Yi Lu,
  • Baojun Lu

摘要

Background

The high recurrence and metastasis rates of oral squamous cell carcinoma (OSCC) pose major therapeutic challenges, with tumor stem cell properties playing a pivotal role. This study aims to systematically explore the tumor stem cell characteristics of OSCC through multi-omics data and establish a reliable prognostic risk assessment model.

Methods

The Cancer Genome Atlas (TCGA)-OSCC dataset was obtained from UCSC-Xena and combined with the Gene Expression Omnibus (GEO) database data for multi-omics analysis. Tumor stemness index (mRNAsi) was calculated using the One-Class Logistic Regression (OCLR) algorithm, and gene modules significantly correlated with mRNAsi were identified via weighted gene co-expression network analysis (WGCNA). Univariate Cox and LASSO regression models were constructed to develop mRNAsi-associated risk scores. Immunological microenvironment characteristics and immunotherapy response potential across risk groups were further analyzed using ESTIMATE, MCP-counter, single-sample gene set enrichment analysis (ssGSEA), and tumor immune dysfunction and exclusion (TIDE) algorithms. Finally, a cellular atlas was constructed from OSCC single-cell RNA sequencing data to explore stemness features in epithelial cell subpopulations. Wet experiments were conducted to verify the expression and function of genes.

Results

Based on WGCNA, we identified gene modules significantly positively correlated with the tumor stem cell index (mRNAsi), primarily enriched in pathways such as cell cycle and chromosome segregation. Subsequently, we identified eight key genes (RNF128, KIAA0513, MPP4, TYMS, CIB2, CELSR3, HRG, CORO2A) and constructed a risk scoring model. This model demonstrated robust prognostic predictive capability across the training set, test set, and external validation cohort, with patients exhibiting high-risk scores showing poorer outcomes. Immune cell infiltration analysis revealed that tumors in the low-risk group exhibited higher levels of immune cell infiltration (such as T cells, neutrophils, and fibroblasts) and demonstrated superior potential for immune therapy response. Single-cell transcriptomic analysis revealed the cellular composition landscape of OSCC, with epithelial cells constituting the largest proportion. Furthermore, seven epithelial cell clusters were identified within this cell population, among which cluster_2 exhibited significantly higher stem cell characteristics. Finally, in vitro functional experiments demonstrated that knocking down CORO2A effectively suppressed the migration and invasion capabilities of OSCC cells.

Conclusion

We systematically elucidated tumor stemness-related molecular features in OSCC, established an mRNA-based risk model, and clarified its relationship with the immune microenvironment and epithelial stemness subpopulations, providing a theoretical basis for OSCC prognosis assessment.