Abstract
BACKGROUND: Glucose metabolism in breast cancer has a potential effect on tumor progression and is related to the immune microenvironment. Thus, this study aimed to develop a glucose metabolism-tumor microenvironment score to provide new perspectives on breast cancer treatment. METHOD: Data were acquired from the Gene Expression Omnibus and UCSC Xena databases, and glucose-metabolism-related genes were acquired from the Gene Set Enrichment Analysis database. Genes with significant prognostic value were identified, and immune infiltration analysis was conducted, and a prognostic model was constructed based on the results of these analyses. The results were validated by in vitro experiments with MCF-7 and MCF-10A cell lines, including expression validation, functional experiments, and bulk sequencing. Single-cell analysis was also conducted to explore the role of specific cell clusters in breast cancer, and Bayes deconvolution was used to further investigate the associations between cell clusters and tumor phenotypes of breast cancer. RESULTS: Four significant prognostic genes (PMAIP1, PGK1, SIRT7, and SORBS1) were identified, and, through immune infiltration analysis, a combined prognostic model based on glucose metabolism and immune infiltration was established. The model was used to classify clinical subtypes of breast cancer, and PMAIP1 was identified as a potential critical gene related to glucose metabolism in breast cancer. Single-cell analysis and Bayes deconvolution jointly confirmed the protective role of the PMAIP1+ luminal cell cluster.