Comprehensive analysis based on DNA methylation and RNA-seq reveals hypermethylation of the up-regulated WT1 gene with potential mechanisms in PAM50 subtypes of breast cancer.

BACKGROUND

Breast cancer (BC), one of the most widespread cancers worldwide, caused the deaths of more than 600,000 women in 2018, accounting for about 15% of all cancer-associated deaths in women that year. In this study, we aimed to discover potential prognostic biomarkers and explore their molecular mechanisms in different BC subtypes using DNA methylation and RNA-seq.

METHODS

We downloaded the DNA methylation datasets and the RNA expression profiles of primary tissues of the four BC molecular subtypes (luminal A, luminal B, basal-like, and HER2-enriched), as well as the survival information from The Cancer Genome Atlas (TCGA). The highly expressed and hypermethylated genes across all the four subtypes were screened. We examined the methylation sites and the downstream co-expressed genes of the selected genes and validated their prognostic value using a different dataset (GSE20685). For selected transcription factors, the downstream genes were predicted based on the Gene Transcription Regulation Database (GTRD). The tumor microenvironment was also evaluated based on the TCGA dataset.

RESULTS

We found that Wilms tumor gene 1 (), a transcription factor, was highly expressed and hypermethylated in all the four BC subtypes. All the methylation sites exhibited hypermethylation. The methylation levels of the TSS200 and 1stExon regions were negatively correlated with expression in two BC subtypes, while that of the gene body region was positively associated with expression in three BC subtypes. Patients with low expression had better overall survival (OS). Five genes including , , , , and were predicted as the downstream genes of . Those five genes were dysregulated in the four BC subtypes. Patients with a favorable 6-gene signature (low expression of and its five predicted downstream genes) exhibited better OS than that with an unfavorable 6-gene signature. We also found a correlation between and tamoxifen using STITCH. Higher infiltration rates of CD8 T cells, plasma cells, and monocytes were found in the lower quartile group and the favorable 6-gene signature group. In conclusion, we demonstrated that is hypermethylated and up-regulated in the four BC molecular subtypes and a 6-gene signature may predict BC prognosis.