Unveiling prognostic genes and regulatory mechanisms of stress granules in gastric cancers: an integrated analysis of bulk transcriptomics and single-cell RNA sequencing.
[BACKGROUND] Gastric cancer (GC) is often associated with a poor prognosis, and the precise molecular mechanisms driving its pathogenesis are not yet fully characterized.
APA
Kou R, Zhu C, et al. (2026). Unveiling prognostic genes and regulatory mechanisms of stress granules in gastric cancers: an integrated analysis of bulk transcriptomics and single-cell RNA sequencing.. Frontiers in oncology, 16, 1750088. https://doi.org/10.3389/fonc.2026.1750088
MLA
Kou R, et al.. "Unveiling prognostic genes and regulatory mechanisms of stress granules in gastric cancers: an integrated analysis of bulk transcriptomics and single-cell RNA sequencing.." Frontiers in oncology, vol. 16, 2026, pp. 1750088.
PMID
41789004
Abstract
[BACKGROUND] Gastric cancer (GC) is often associated with a poor prognosis, and the precise molecular mechanisms driving its pathogenesis are not yet fully characterized. Stress granules (SGs) are now understood to play a crucial role in tumor progression, yet the prognostic value of SG-related markers in GC remains unclear. This study aimed to identify SG-related prognostic genes, clarify their clinical and biological significance in GC, and validate their potential as predictive indicators for patient overall survival (OS).
[METHODS] Single-cell and transcriptomic data for gastric cancer, along with genes related to stress granules (SGRGs), were acquired from public databases and literature. Candidate genes were identified by intersecting differentially expressed genes (DEGs) with SGRGs. Prognostic genes were identified through univariate Cox regression, and a risk score model was constructed. The model's performance was validated in an independent cohort. Based on risk stratification, functional enrichment analysis, immune cell infiltration pattern assessment, and chemotherapy drug sensitivity analysis were conducted. Cell types expressing the prognostic genes were identified using single-cell RNA sequencing (scRNA-seq), and the related key cell clusters were identified.
[RESULTS] , , , and were identified as prognostic genes. The risk model demonstrated good performance in predicting the survival status of GC patients. GSEA revealed that significantly enriched pathways included neuroactive ligand-receptor interaction and extracellular matrix (ECM)-receptor interaction pathways. CD36, MMRN1, and SERPINE1 demonstrated significant positive correlations with mast cells (correlation coefficients (r) > 0.3, < 0.001). Chemotherapy drugs exhibited greater efficacy in high-risk GC patients. Moreover, endothelial cells were considered key cells and played a critical role in GC. Finally, SERPINE1 expression was associated with clinical features and prognosis in GC.
[CONCLUSION] In summary, we identified a four-gene SG-related signature strongly associated with prognosis in GC and constructed a predictive model with clinical potential. Our integrated analysis identified endothelial cells as a candidate population linked to the expression of these genes. These findings provide associative evidence linking SGs to GC outcomes and highlight potential targets for future mechanistic and therapeutic exploration.
[METHODS] Single-cell and transcriptomic data for gastric cancer, along with genes related to stress granules (SGRGs), were acquired from public databases and literature. Candidate genes were identified by intersecting differentially expressed genes (DEGs) with SGRGs. Prognostic genes were identified through univariate Cox regression, and a risk score model was constructed. The model's performance was validated in an independent cohort. Based on risk stratification, functional enrichment analysis, immune cell infiltration pattern assessment, and chemotherapy drug sensitivity analysis were conducted. Cell types expressing the prognostic genes were identified using single-cell RNA sequencing (scRNA-seq), and the related key cell clusters were identified.
[RESULTS] , , , and were identified as prognostic genes. The risk model demonstrated good performance in predicting the survival status of GC patients. GSEA revealed that significantly enriched pathways included neuroactive ligand-receptor interaction and extracellular matrix (ECM)-receptor interaction pathways. CD36, MMRN1, and SERPINE1 demonstrated significant positive correlations with mast cells (correlation coefficients (r) > 0.3, < 0.001). Chemotherapy drugs exhibited greater efficacy in high-risk GC patients. Moreover, endothelial cells were considered key cells and played a critical role in GC. Finally, SERPINE1 expression was associated with clinical features and prognosis in GC.
[CONCLUSION] In summary, we identified a four-gene SG-related signature strongly associated with prognosis in GC and constructed a predictive model with clinical potential. Our integrated analysis identified endothelial cells as a candidate population linked to the expression of these genes. These findings provide associative evidence linking SGs to GC outcomes and highlight potential targets for future mechanistic and therapeutic exploration.