All types of cancer show wide-spread aberrant epigenetic changes, leading to changes in expression of genes involved in all the classic hallmarks of cancer, as well as genes involved in how tumors will respond to chemotherapy. These epigenetic changes can be reversed, at least temporarily, using small molecule inhibitors of maintenance of the epigenetic state. DNA demethylating agents and histone deacetylase inhibitors have shown activity against certain haematological malignancies, however their activity in solid tumors remains more uncertain (1). For successful treatment of solid tumors with epigenetic therapies, major challenges remain in the delivery of epigenetic therapy, the maintenance of a pharmacodynamic response and the achievement of a therapeutic index. In addition, the development of robust predictive biomarkers linked to an understanding of the underlying biology will be key to improved epigenetic therapy approaches. Defining the epigenetic profile of individual cancer subtypes may allow epigenetic therapies to be targeted to those groups of patients who are most likely to benefit from these types of treatment. Furthermore, targeting drug resistant or tumor sustaining subpopulations with epigenetic therapies may be vital, especially given the key role of epigenetic mechanisms in maintenance of a stem cell state.

Ovarian cancer is the most lethal gynaecological cancer. Although at least 70% of patients respond to platinum-based chemotherapy, the majority of patients eventually relapse. We have shown that methylation of multiple promoter CpG islands (CGIs) at genes in the Wnt pathways are associated with progression-free survival (PFS) of epithelial ovarian cancer (EOC) patients (2). We have taken this approach further by systematically profiling DNA methylation at promoter CGIs of 9 further key pathways/families (AKT/mTOR, BRCA1/2, Redox, p53, FA families, igLON families, HR, NHEJ and MMR) in EOCs to identify methylation biomarkers predicting early disease relapse and response to platinum-based chemotherapy. Late-stage EOCs prospectively collected through a SGCTG cohort study were analysed by differential methylation hybridisation (DMH) (n=150) and association with PFS evaluated by Cox model using DMH ratios as a continuous variable. Methylation of 30 loci were correlated with PFS (p<0.01, FDR<10%) and independent from conventional prognostic factors of age, stage, grade and histology (p<0.05). Combining the 30 loci from the current study with 7 loci previously identified in the Wnt pathways (2), we constructed a multivariate Cox model which only incorporates NKD1, PRD×2 and VEGFB. Validation of the prognostic values of loci identified were then sought in an independent patient cohort of high-grade serous ovarian tumors (n=311) collected by The Cancer Genome Atlas (TCGA) study and methylation analysed using Illumina Infinium assay. Among the 30 loci identified by DMH as significantly prognostic, only 19 loci were evaluable in the TCGA cohort. Of these 19 loci, promoter methylation of 10 loci remain significantly prognostic in this independent cohort and using this alternative method of methylation analysis. Further analysis using logistic regression model found patients with higher methylation of VEGFB, GP×4 and RAD54L were more likely to have poor response to platinum-based chemotherapy (p<0.05). These data demonstrate the importance of methylation of multiple promoter CGIs for predicting clinical outcome in EOCs and their potential as stratification markers in future molecularly targeted clinical studies.

We have shown that subpopulations of tumor sustaining cells can be isolated from human ovarian tumors and cell lines which have stem cell like properties, including expression of stem cell markers, growth as spheroids in anchorage independent manner and as xenografts in NOD/SCID mice (3). Such subpopulations are more resistant to chemotherapeutic drugs such as carboplatin and appear to be enriched for following carboplatin-based chemotherapy in patients. This highlights the importance of drugs that can target this drug resistant subpopulation. Ovarian tumor sustaining cells over-express EZH2 (3), a histone lysine methyltransferase (HKMT) and key component of the polycomb repressive complex, PRC2. SiRNA knock-down of EZH2 alone or in combination with EZH1, leads to reduced spheroid and tumor growth of ovarian tumor sustaining cells. We have used a cell based assay to identify small molecules which are able to re-express endogenous genes where EZH2 has been shown to be involved in the epigenetic silencing process. We have identified compounds that up-regulate EZH2 target genes, but not DNA methylated, EZH2 target genes. Chromatin immunoprecipitation (ChIP) experiments verified a decrease in silencing marks (H3K27me3, H3K9me3) and importantly an increase in active chromatin marks (e.g. H3K4me2, H3K4me3) at promoter regions.

In ES cells, bivalent chromatin domains, containing H3K4me3 and H3K27me3 marks, silence developmental genes while keeping them poised for later activation following differentiation (4). We have identified gene sets associated with H3K27me3 and H3K4me3 marks at transcription start sites in a high grade ovarian serous tumor to test for correlations with epigenetic silencing and malignancy characteristics. This revealed novel silenced bivalent marked genes not described previously for ES cells, which are significantly enriched for the PI3K (p<10−7) and the TGF-beta signalling pathways (p<10−5). We matched histone marked gene sets to gene expression sets of 8 normal fallopian tube and 499 high grade serous malignant ovarian samples which revealed a significant decrease in gene expression for the H3K27me3 and bivalent gene sets, which occurred largely independent of DNA methylation. Next, we correlated H3K27me3 and bivalent gene sets to gene expression data of ovarian tumor stem cell like sustaining cells versus non-sustaining cells, which showed a significantly lower expression for the H3K27me3 and bivalent gene sets in the tumor sustaining cells. Similarly, comparison of matched chemo-sensitive and chemo-resistant ovarian cell lines showed again a significantly lower expression of H3K27me3/bivalent marked genes in the chemo-resistant compared to the chemo-sensitive cell line. Our analysis supports the hypothesis that bivalent marks are associated with epigenetic silencing in ovarian cancer (5), but suggests that additional tumor specific bivalent marks to those known in ES cells are present in tumor at presentation that can potentially influence the subsequent acquisition of drug resistance and tumor progression.


1. Graham, J, Kaye SB and Brown R (2009) The promises and pitfalls of epigenetic therapies in solid tumors Eur J Cancer, 45: 1129–1136, 2009.

2. Dai W, Teodoridis JM, Zeller C, Graham J, Hersey J, Flanangan JM, Stronach E, Siddiqui N, Paul J, Brown R (2011) Systematic CpG Islands Methylation Profiling of Genes in the Wnt Pathway in Epithelial Ovarian Cancer Identifies Biomarkers of Progression-Free Survival Clin Cancer Res, 17:4052–62.

3. Rizzo S, Hersey JM, Mellor P, Dai W, Santos-Silva A, Liber D, Luk L, Titley I, Carden CP, Box G, Hudson DL, Kaye SB, and Brown R (2011) Ovarian Cancer Stem Cell-Like Side Populations Are Enriched Following Chemotherapy and Overexpress EZH2 Mol Cancer Therapeutics, 10:325–335.

4. Bernstein, B. E., T. S. Mikkelsen, et al. (2006). “A bivalent chromatin structure marks key developmental genes in embryonic stem cells.” Cell, 125: 315–26.

5. Widschwendter, M., Fiegl, H., Egle, D., Mueller-Holzner, E., Spizzo, G., Marth, C., Weisenberger, D. J., Campan, M., Young, J., Jacobs, I., and Laird, P. W. (2007) Epigenetic stem cell signature in cancer. Nat Genet, 39: 157–158.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr PL05-04.