Molecular subtyping and therapeutic drug prediction in esophageal squamous cell carcinoma based on manganese-metabolism-related genes

Abstract

Background and study aims

Immunotherapy offers promising prospects for esophageal squamous cell carcinoma (ESCC), a highly fatal malignant tumor. Given the association between manganese metabolism and tumor immunity, this study explored the prognostic value and role of manganese-metabolism-related genes (MMRGs) in the ESCC immune microenvironment to uncover their clinical potential.

Patients and methods

Transcriptomic and clinical data of ESCC were retrieved from TCGA and the GEO as training and validation cohorts, respectively. Patients were clustered and subtyped based on MMRGs, with survival compared. A prognostic risk model was constructed using differential analysis, PPI network, and Cox regression; its relationship with immune characteristics, immunotherapy response, and drug sensitivity was evaluated. SLC40A1 was knocked down in vitro, and its effects on cellular function and drug sensitivity were assessed via qRT-PCR, Western blot, colony formation, Transwell, and CCK-8 assays.

Results

ESCC patients were stratified into two MMRG-defined subtypes. Cluster 2 showed significantly worse overall survival (OS) than Cluster 1. An 8-gene prognostic model was established and patients were assigned to RiskScore_high and RiskScore_low groups. The high-risk group, characterized by worse OS, displayed an immunosuppressive microenvironment with abundant M2 macrophages and high immune-checkpoint expression (PDCD1, CTLA4, TIGIT), along with a lower TIDE score, suggesting potentially greater benefit from immune-checkpoint blockade. The RiskScore_high group was more sensitive to Gemcitabine and Oxaliplatin, whereas the RiskScore_low group responded better to BI-2536 and NU7441. Cellular functional assays confirmed high expression of SLC40A1 in ESCC cells. SLC40A1 knockdown significantly inhibited cell proliferation, migration, and invasion. Additionally, cells exhibited greater sensitivity to Gemcitabine than to BI-2536.

Conclusion

MMRG-based subtyping and the reliable prognostic risk score model provide novel insights for predicting prognosis and developing personalized therapy in ESCC.