- Identification of DEGs between the relapsed and nonrelapsed groups after tamoxifen treatment in HR+/HER2-negative breast cancer patients
To identify gene expression related to tamoxifen resistance, we matched 10 pairs of breast cancer samples from the TCGA database of patients who received tamoxifen therapy according to their age and pathological characteristics (Table 1); all of these patients relapsed in 5 years. Figure 1a shows the process of screening for suitable patients. In total, we identified 647 DEGs between recurrent and nonrecurrent patients with the selection criteria of q value < 0.05 and |log2(Fold Change (FC)) | >1. Among the DEGs, 506 genes were downregulated, and 141 were upregulated. Figure 1b shows a volcano plot to examine the differences in the expression level of the genes in the two groups of samples and the statistical significance of these differences.
- GO and KEGG pathway analysis of the DEGs
The DEGs were annotated in the GO and KEGG databases. GO enrichment analysis of the DEGs was implemented using the Database for Annotation, Visualization and Integrated Discovery (DAVID) website. Figure 2a-c shows the results of the DEGs annotated in the GO database. In the category of biological processes, the DEGs were mainly enriched in epidermal development, cytoskeletal organization, cell-cell signaling, BMP signaling, etc. In the cellular component category, the DEGs were mainly enriched in the extracellular space, extracellular region, integral component of the plasma membrane, plasma membrane, etc. In the category of molecular function, the DEGs were mainly enriched in structural molecule activity, structural constituents of the cytoskeleton, heparin binding, growth factor activity, and inflammatory response. The KEGG pathway analysis (Figure 2d) shows that 50 genes were distributed in metabolic pathways, 16 genes were distributed in cytokine−cytokine receptor interactions, 7 genes were distributed in drug metabolism-cytochrome P450, and 6 genes were distributed in steroid hormone biosynthesis. Herein, the mechanism of tamoxifen therapy failure, whether associated with estrogen metabolism or tamoxifen metabolism, needs further investigation.
- Protein-protein interaction network construction and module analysis
To analyze the genes associated with tamoxifen resistance, the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) website was used to investigate the protein-protein interaction (PPI) network associated with the ER pathway. The PPI network was drawn using Cytoscape, which is an open-source bioinformatics software platform for visualizing molecular interaction networks, and the most significant module in the PPI network was identified by using the cytoHubba plug-in of Cytoscape, which is an APP for clustering a given network based on a topology to find densely connected regions. Figure 3b shows the top ten hub genes—C3, PF4, SAA1, CXCL1, CX3CL1, CXCL13, GAL, GPER1, CXCL2, and CXCL—most of which are inflammatory factors, which implies that tamoxifen resistance may be associated with the inflammatory response, and the differentiated expression of these ten genes was shown in Table 2. A previous study has proven that mediators secreted by cancer-associated fibroblasts (CAFs), in addition to components of the inflammatory response such as cytokines and growth factors, exert an important role in drug resistance.
- Validation of the expression of the top 10 hub genes
Considering that there were few samples from TCGA, we validated the expression of the top 10 hub genes in a different subgroup of breast cancer patients on the GEPIA2 website (http://gepia2.cancer-pku.cn/#index). As shown in Figure 4, the expression of SAA1, CX3CL1, CXCL1, CXCL2, and GPER1 was higher in the adjacent normal breast tissue than in the luminal A subgroup of breast cancer, while the expression of CXCL13 was lower, and the differential expression was statistically significant. The expression of C3, PF4, GAL, and CXCL6 was not significant between the para-cancerous normal breast tissue and luminal A breast cancer tissue, but its general trend was in line with our results, especially for C3, PF4, SAA1, CXCL1, CX3CL1, CXCL13, GAL, GPER1, CXCL2, and CXCL6, most of which are inflammatory factors, thus indicating that tamoxifen resistance may be associated with the inflammatory response. The previous study has proven that mediators secreted by CAFs, in addition to components of the inflammatory response such as cytokines and growth factors, exert an important role in drug resistance.
- Survival analysis
Survival analysis was performed to explore the correlation between the 10 DEGs and patients’ recurrence-free survival (RFS) with the KM-plot website (http://kmplot.com/analysis/). Figure 5(a-j) shows that the increased expression of C3, CX3CL1, CXCL1, CXCL2, CXCL6, GPER, and PF4 was closely related to improved RFS, while this result did not show that the other three genes (CXCL13, GAL, and SAA1) were associated with RFS in luminal A breast cancer.