Identification of a PANoptosis-related Gene Signature for Predicting the Prognosis, Tumor Microenvironment and Therapy Response in Breast Cancer.

Breast cancer (BC) is the most prevalent malignancy among women worldwide. Mounting evidence suggests that PANoptosis participates in cancer development and therapy. However, the role of PANoptosis in BC remains unclear. In this study, we identified ten PANoptosis-related genes using Cox regression analysis, random forest (RF) algorithm and least absolute shrinkage and selection operator (LASSO) algorithm. A PANoptosis-related score (PRS) was calculated based on the coefficient of LASSO. Notably, we divided the patients into high- and low-risk groups according to the PRS and revealed a negative correlation between PRS and overall survival. Next, a nomogram model was constructed and validated to improve the clinical application of PRS. Functional enrichment analyses and the Bayesian network demonstrated that differentially expressed genes between high- and low-risk groups were mainly enriched in immune-related pathways. Besides, we found significant differences in tumor mutation burden and tumor immune microenvironment between patients in these two groups using bulk-RNA and single-cell RNA sequencing data. Furthermore, charged multivesicular body protein 2B (CHMP2B) was identified as the hub gene by combining LASSO, weighted gene co-expression network analysis, RF and eXtreme Gradient Boosting. Importantly, using immunohistochemistry analysis based on our tissue microarray, we found that CHMP2B was highly expressed in tumor tissue, and CD4 and CD8 were more likely to be positive in the CHMP2B-negative group. Survival analyses revealed that CHMP2B adversely impacted the survival of BC patients. In conclusion, we not only constructed a highly accurate predictive model based on PRS, but also revealed the importance of PANoptosis-related gene signature in the modulation of the tumor microenvironment and drug sensitivity in BC.