Integrative epigenomic profiling of early breast carcinogenesis

Abstract

Epigenetic dysregulation in cancer is well documented, however, much of the detail remains unclear. Whilst studies in cancer cell lines and clinical samples utilising next generation sequencing are delivering insight into the cancer epigenome, it is unknown which changes drive carcinogenesis or the influence of the temporal nature of epigenetic change during cancer progression. Therefore, it is important to find novel systems for the study of epigenetic events that occur during early breast carcinogenesis. Human mammary epithelial cells (HMEC) provide an in vitro model of early breast carcinogenesis. In culture HMEC are able to temporarily escape senescence and acquire a pre-malignant breast cancer like phenotype (variant or vHMEC). The transition from HMEC to vHMEC recapitulates the epigenomic progression of normal to pre-malignant epithelial transformation. I have performed detailed epigenomic profiling of the HMEC model system to elucidate the role of epigenomic aberrations in early breast carcinogenesis. Epigenomic profiling of the HMEC model system included analysis of mRNA and non-coding miRNA expression, chromatin modifications and DNA methylation. Profiling was performed by multiple genome-wide technologies including gene expression microarrays, mRNA sequencing and chromatin profiling by a combination of chromatin immunoprecipitation and promoter tiling microarray and/or next generation sequencing. Profiling of DNA methylation was performed using the novel methodology MBD.Cap-seq. MBD.Cap-seq utilises methylated DNA enrichment followed by DNA sequencing. The data generated exhibited high correlation indicating that it is robust and reflective of the true epigenomic change occurring as vHMEC escape senescence. Analysis of the individual epigenetic layers identified widespread inhibition of the p53 tumour suppressive pathway, activation of the proto-oncogene MYC and deregulation of genes controlled by the polycomb group family of proteins. Via an integrative approach I identified functional chromatin states in the HMEC that correlated with publicly available data, indicating that they are biologically relevant. Overlap of these functional chromatin states with DNA methylation revealed a pattern of hypomethylation at enhancer loci in vHMEC. This pattern indicates a potentially crucial role for hypomethylation at non-repetitive intergenic regions during early carcinogenesis that differs from that observed in late stage disease.