Machine Learning-Based Sialylation-Associated Gene Signature Predicts Prognosis and Immune Landscape in Hepatocellular Carcinoma: Validation via Multi-Omics Analysis and in vitro Assays.

Background: Sialylation, an important post-translational modification crucial for protein activity, affects tumor development and spread by altering immune response. Nevertheless, the roles they play in the microenvironment of Hepatocellular Carcinoma (HCC) and their clinical implications remain unclear. The purpose of this research was to investigate the function of genes involved in sialylation concerning tumor immunity and their clinical implications in HCC. Methods: We intended to build a prognostic prediction model called the sialylation score through the application of sialylation-associated genes and a machine learning integrative approach. Results: Our findings show that the sialylation score independently affects overall survival of HCC patients, with trustworthy and stable outcomes. Sialylation scores are notably more accurate than conventional clinical variables and previously published signatures. Moreover, patients with low sialylation scores had significant immune infiltration. Further analysis of single-cell cohorts indicates that patients with high sialylation scores have an immunosuppressive microenvironment, where T and NK cells improve their interactions with myeloid cells via signaling pathways like MHC-II, CLEC and COLLAGEN pathways. To validate the biological significance of this signature, we targeted the key gene ST3GAL4 in vitro, revealing that its knockdown significantly reduces the proliferation, invasion, and migration capabilities of HCC cells. Conclusion: The sialylation score could serve as a dependable method for anticipating immune response, thereby improving clinical results in patients with HCC.