Identification of glycogene signature as a tool to predict the clinical outcome and immunotherapy response in breast cancer

BACKGROUND: Breast cancer is one of the most important diseases in women around the world. Glycosylation modification correlates with carcinogenesis and roles of glycogenes in the clinical outcome and immune microenvironment of breast cancer are unclear. METHODS: A total of 1297 breast cancer and normal cases in the TCGA and GTEx databases were enrolled and the transcriptional and survival information were extracted to identify prognostic glycogenes using Univariate Cox, LASSO regression, Multivariate Cox analyses and Kaplan-Meier method. The immune infiltration pattern was explored by the single sample gene set enrichment method. The HLA and immune checkpoint genes expression were also compared in different risk groups. The expressions of a glycogene MGAT5 as well as its products were validated by immunohistochemistry and western blotting in breast cancer tissues and cells. RESULTS: A 19-glycogene signature was identified to separate breast cancer patients into high- and low-risk groups with distinct overall survival rates (P < 0.001). Compared with the high-risk group, proportion of naive B cells, plasma cells and CD8(+) T cells increased in the low-risk group (P < 0.001). Besides, expressions of HLA and checkpoint genes, such as CD274, CTLA4, LAG3 and TIGIT3, were upregulated in low-risk group. Additionally, highly expressed MGAT5 was validated in breast cancer tissues and cells. Downstream glycosylation products of MGAT5 were all increased in breast cancer. CONCLUSIONS: We identified a 19-glycogene signature for risk prediction of breast cancer patients. Patients in the low-risk group demonstrated a higher immune infiltration and better immunotherapy response. The validation of MGAT5 protein suggests a probable pathway and target for the development and treatment of breast cancer.