Identification of hub genes in AR-induced tamoxifen resistance in breast cancer based on weighted gene co-expression network analysis

Background Approximately 30% of patients with oestrogen receptor (ER)-positive breast cancer (BC) exhibit intrinsic or recurrent resistance to tamoxifen (TAM) adjuvant endocrine therapy. The androgen receptor (AR) is expressed in about 90% of ER-positive patients. Our previous studies found that BC patients with an AR:ER expression ratio >= 2.0 are more susceptible to TAM resistance. However, the specific mechanism by which a high AR:ER ratio promotes TAM resistance remains unknown. Methods RNA sequencing was performed on 10 cases of BC tissues with AR:ER ratios >= 2.0 and 3 cases with AR:ER ratios < 2.0. We then compared our data with the screened TAM-resistant and TAM-sensitive cases from the TCGA BC database. Bioinformatics methods were used to screen differentially expressed genes (DEGs) and to perform gene enrichment analysis. Weighted correlation network analysis (WGCNA) was used to screen hub genes in the AR-induced TAM resistance process. Results PAM50 analysis showed that the molecular phenotype of BC patients with AR:ER ratios >= 2.0 was similar to that of triple-negative breast cancer (TNBC), whereas the BC samples with AR:ER ratios < 2.0 were classified as the luminal subtype. Among the AR:ER ratio >= 2.0 and AR:ER < 2.0 BC tumours, 1855 DEGs were identified. Gene enrichment analysis showed that DEGs were enriched mainly in proliferation-related molecular pathways, such as the cell cycle, necroptosis, metabolic pathways and DNA replication. WGCNA analysis showed that SEC14L2, RIIAD1, STC2 and MAGEA6 served as hub genes in AR-induced TAM resistance and were associated with BC survival prognosis in the TCGA cohort. Conclusions A high AR:ER expression ratio is a biomarker for patients who might develop TAM resistance, and AR expression seems to be a possible mechanism of resistance to endocrine therapy.