Abstract
Breast cancer (BRCA) cells typically exist in nutrient-deficient microenvironments and quickly adapt to states with fluctuating nutrient levels. The tumor microenvironment of starvation is intensely related to metabolism and the malignant progression of BRCA. However, the potential molecular mechanism has not been thoroughly scrutinized. As a result, this study aimed to dissect the prognostic implications of mRNAs involved in the starvation response and construct a signature for forecasting the outcomes of BRCA. In this research, we investigated how starvation could affect BRCA cells' propensities for invasion and migration. The effects of autophagy and glucose metabolism mediated by starved stimulation were examined through transwell assays, western blot, and the detection of glucose concentration. A starvation response-related gene (SRRG) signature was ultimately generated by integrated analysis. The risk score was recognized as an independent risk indicator. The nomogram and calibration curves revealed that the model had excellent prediction accuracy. Functional enrichment analysis indicated this signature was significantly enriched in metabolic-related pathways and energy stress-related biological processes. Furthermore, phosphorylated protein expression of the model core gene EIF2AK3 increased after the stimulus of starvation, and EIF2AK3 may play an essential role in the progression of BRCA in the starved microenvironment. To sum up, we constructed and validated a novel SRRG signature that could accurately predict outcomes and may be developed as a therapeutic target for the precise treatment of BRCA.