Epigenomic deregulation in cancer is well established, however, the timing and extent of epigenetic change during carcinogenesis remains unknown. At the time of diagnosis, carcinogenesis is already advanced and a tumour cell contains a cacophony of epigenetic abnormalities, many of which are a consequence of prior cancer-driving genomic aberrations. The extensive nature of epigenomic change confounds studies that utilise traditional cancer-normal comparisons. Such studies are unable to establish the timing of epigenetic change during carcinogenesis and, therefore, cannot isolate cancer driving from passenger events. To this end, we performed a detailed epigenomic analysis of the Human Mammary Epithelial Cell (HMEC) model system, including DNA methylation, gene expression and chromatin modification profiling.
HMEC cells are generated from normal tissue taken from healthy women. When cultured, HMEC undergo partial carcinogenic transformation into a variant cell population (vHMEC) and represent a model of the very earliest stages of breast carcinogenesis. When compared to HMEC, vHMEC exhibit several cancer-associated traits including altered cell morphology, increased rate of cell division and silencing of tumour suppressor genes and/or pathways. For example, the epigenetic silencing of the p16 tumour suppressor gene via hypermethylation of the promoter region is a key event in HMEC transformation.
Our studies have shown that epigenetic aberrations in vHMEC are widespread and impact a range of cancer-associated pathways (including methylation of p53 targets, increased activity of MYC and deregulation of EZH2/polycomb). Comparison of the vHMEC epigenomic profile with public methylation data available from The Cancer Genome Atlas (TCGA) demonstrated that hypermethylation occurring during early carcinogenesis is present in a more advanced cancer state. Furhtermore, these loci may have utility as biomarkers for the diagnosis and/or subtyping of breast cancer. These studies emphasise the powerful nature of the HMEC system as an epigenetic model of breast carcinogenesis.