Immunological significance of prognostic markers for breast cancer based on alternative splicing.

Objectives: Breast cancer (BC) currently has the highest incidence rate. Epigenetic regulation could alter gene expression and is closely related to BC initiation. This study aimed to develop an alternative splicing (AS)-based prognostic signature and clarify its relevance to the tumor immune microenvironment (TIME) status and immunotherapy of BC.

Methods: Cox regression analysis was conducted to screen for prognosis-related AS events. Thereafter, LASSO with Cox regression analyses was designed to construct a prognostic signature model. Kaplan-Meier survival analysis, receiver operating characteristic curves, and proportional hazard model were then utilized to confirm the prognostic value. Multiple methods were employed to reveal the context of TIME in BC. QPCR, western blotting and immunofluorescence microscopy were carried out to detect myc-associated zinc finger protein (MAZ) expression in different cell lines, and BC and paracancerous tissues.

Results: A total of 1,787 prognosis-related AS events were screened. Eight AS prognostic signatures were constructed with robust predictive accuracy based on the splicing subtypes. Furthermore, the establishment of a quantitative prognostic nomogram and consolidated signature was significantly correlated with TIME diversity and immune checkpoint blockade-related genes. MAZ was detected to be upregulated in BC. Finally, a newly constructed splicing regulatory network model revealed the potential functions of splicing factors.

Conclusions: AS-splicing factor networks may enable a new approach to investigating potential regulatory mechanisms. Moreover, pivotal players in AS events with regards to TIME and efficiency of immunotherapy were uncovered and could facilitate clinical decision-making and individual determination of BC prognosis.