Abstract
Our study proposed a novel SiaM-based classification that identified three metabolic subtypes with distinct characteristics of tumor microenvironment and clinical outcomes in GC.
Introduction
Gastric cancer (GC) remains a leading cause of cancer-related mortality worldwide, largely due to its high degree of heterogeneity. While molecular classifications like those from The Cancer Genome Atlas (TCGA) have improved our understanding, they often overlook metabolic drivers of tumor progression. Sialylation—the addition of sialic acid to cell-surface proteins—is a critical metabolic process. Hypersialylation creates a "sugar coat" that promotes immune evasion by binding to SIGLEC receptors and facilitates metastasis by reducing cell adhesion. This study aims to map the sialic acid metabolism (SiaM) landscape in GC to identify prognostic subtypes and therapeutic targets.
Methodology and Data Integration
The researchers employed a comprehensive multi-omics approach, integrating data from public and institutional cohorts totaling over 1,000 patients. The workflow involved unsupervised consensus clustering of 31 SiaM-related genes, followed by functional annotation and predictive modeling.
The study utilized five public datasets (including TCGA-STAD and the ACRG cohort) and two institutional cohorts (ZJUGC-A and ZJUGC-B) for proteomic and qPCR validation. Single-cell RNA sequencing (scRNA-seq) was also used to explore the tumor microenvironment (TME) at a high resolution.
Identification of Metabolic Subtypes
Unsupervised clustering identified three distinct SiaM subtypes: Clusters A, B, and C.
- Cluster C emerged as the most aggressive subtype, characterized by high SiaM activity, diffuse tumor structure, and epithelial-to-mesenchymal transition (EMT) features.
- Immune Landscape: Cluster C exhibited an "immune-excluded" phenotype. Hypersialylation in these tumors likely drives immune resistance by polarizing macrophages toward an M2-like immunosuppressive state and overexpressing SIGLECs that inhibit T-cell activity.
- Clinical Outcomes: Patients in Cluster C had the worst overall survival (OS) and the highest rate of peritoneal metastasis (63.0%) compared to Clusters A (22.2%) and B (14.8%).
Predictive Modeling for Peritoneal Metastasis
To translate these findings into clinical tools, the authors developed a 6-gene risk score (ARHGAP6, ST3GAL1, ADAM28, C7, PLCL1, TTC28) using LASSO regression.
This model demonstrated robust performance in predicting peritoneal metastasis across multiple independent cohorts, with Area Under the Curve (AUC) scores reaching up to 0.800. High SiaM scores were consistently associated with poor prognosis across all tested datasets, including the TCGA-STAD and ZJUGC-A cohorts.
ST3GAL1 as a Therapeutic Target
The study identified ST3GAL1 as a hub gene driving the aggressive phenotype of Cluster C. Experimental validation showed that ST3GAL1 overexpression promotes cell migration and metastasis in vitro and in vivo. Crucially, treatment with a sialyltransferase inhibitor (STi), 3Fax-Peracetyl Neu5Ac, significantly reduced peritoneal metastasis in mouse models, suggesting a potential new therapeutic strategy for advanced GC.
Conclusion
This SiaM-based classification provides a new framework for understanding GC heterogeneity. By identifying high-risk metabolic subtypes and a druggable target in ST3GAL1, this research paves the way for more personalized treatment strategies, particularly for patients at high risk of peritoneal metastasis. Future prospective clinical trials are needed to validate the efficacy of sialyltransferase inhibitors in human patients.
Summary updated