Integrating machine learning and experimental validation identifies a post-translational modification gene signature for prognosis and treatment response in breast cancer.

Breast cancer (BC) is the most prevalent malignancy among women, and the steadily increasing disease burden has garnered considerable global attention. Post-translational modifications (PTMs) are critical in the initiation and progression of BC. This study aimed to elucidate the associations between diverse PTMs and the prognosis of patients with BC. We collected genes associated with multiple PTMs and evaluated the activity of each PTM using GSVA. We aggregated PTM scores to derive the PTMS and identified differentially expressed genes between the high- and low-PTMS groups. A PTM-related gene signature (PTMRS) was developed based on the optimal combination among 117 machine learning models, and its predictive performance was benchmarked against other published signatures. In addition, we investigated the associations between PTMRS, tumor immunity, and treatment response. Gene expression across different cell types was evaluated using single-cell and spatial transcriptomic analyses. Gene expression levels in cancerous and paired adjacent noncancerous tissues were validated by PCR. The results of GSVA showed that most of the PTMs were dysregulated in cancer. Tumor immunity levels were elevated in the low-PTMS group compared with the high-PTMS group. The PTMRS comprised five genes: SLC27A2, TNFRSF17, PEX5L, FUT3, and COL17A1. The predictive performance of the PTMRS exceeded that of the clinical profile and 14 other published gene signatures. Patients in the high-PTMRS group exhibited poorer prognosis and reduced anti-tumor immunoreactivity. In addition, patients in the low-PTMRS group showed improved responses to chemotherapy and immune checkpoint inhibitors. Spatial transcriptomics analysis revealed that SLC27A2 exhibited higher expression in malignant spots, whereas COL17A1 and TNFRSF17 showed lower expression in malignant spots. SLC27A2 mRNA expression was elevated in tumor tissues relative to adjacent noncancerous tissues, whereas the mRNA expression levels of the other four genes were decreased. This study reveals the important role of PTMs in BC prognosis and provides new perspectives for the prognostic assessment of BC patients as well as personalized treatment.