Abstract
BACKGROUND: Sialylation is a crucial glycosylation modification of eukaryotic cell surface proteins. Tumor cell growth, immune evasion, and drug resistance are driven by excessive sialylation. However, the expression levels of genes associated with sialylation, prognostic value, and impact on the response to immunotherapy in brain tumors remain unclear. This study hypothesized that sialylation-related genes could serve as grouping genes to identify 20 significant genes for predicting the survival outcomes of patients with brain tumors and their responsiveness to immune checkpoint inhibitors. METHODS: Using 83 genes related to sialylation, we classified the cohort into two distinct groups with marked differences in survival outcomes and immune cell infiltration. After identifying the differential genes, we subsequently performed unsupervised clustering, yielding two groups with high concordance. Principal component analysis (PCA) was used for dimensionality reduction in the two groups, with the PC1 and PC2 components defined as the PCA score. RESULTS: A positive correlation was observed between the PCA score and immune cell infiltration, immune checkpoint expression, as well as chemokine expression, suggesting enhanced immunotherapy efficacy. We also validated the expression levels of KCNIP3 in patient tissues and found that this gene was overexpressed in U87 and LN229 cells, with low endogenous expression. These results demonstrate that KCNIP3 inhibits tumor cell proliferation, migration, and invasion. Moreover, intracranial xenograft results in mice were consistent with the in vitro findings. CONCLUSION: We developed a robust model to predict patient prognosis in low-grade gliomas and glioblastomas, emphasizing the prognostic significance of differentially expressed genes associated with sialylation.