S98
EACR-23 Poster Sessions / European Journal of Cancer 50, Suppl. 5 (2014) S23–S242
PRDM1a, and TGIF1d between normal GC B-cells and N, M, and PB as well as between DLBCL subtypes (except BCL6b). The PRDM1b and TGIF1f transcripts were tumor specific. Finally, analysis of the clinical prognostic impact of isoform-specific RT-qPCR for overall survival identified BCL6a as an independent clinical prognostic variable. Conclusion: We identified and validated 3 TFs with isoform-specific expression between normal B-cell subpopulations and DLBCL subtypes, and identified one of these, BCL6a, as an independent clinical prognostic variable, supporting its impact in the pathogenesis of DLBCL. No conflict of interest. 410 HER2 binds to the genome at specific regulatory sites of key genes in breast cancer A. Redmond1 , J.S. Carroll1 . 1 Cancer Research UK Cambridge Institute, Cambridge, United Kingdom Background: HER2 is traditionally known as a transmembrane tyrosine kinase receptor which heterodimerises with other members of the EGFR family and mediates signaling through various intracellular pathways. In the clinic, it is used as a biomarker to stratify treatment for breast cancer patients and is a therapeutic target for a number of drugs including trastuzumab and lapatinib. Recent work has shown a role for HER2 in the nucleus and has confirmed full-length receptor binding to the DNA at a small number of individual loci. Materials and Methods: We have explored this observation by mapping HER2–DNA interaction sites using chromatin immunoprecipitation sequencing combined with exonuclease digestion (ChIPexo) in three different breast cancer cell lines. RNAseq has been conducted to analyse the expression profiles induced by HER2 activation. In addition, mass spectrometry experiments have been utilized to investigate interacting proteins. Results: Analysis of binding of HER2 to the DNA has been conducted in three different breast cancer cell lines, BT474, SKBR3 and T47D. Binding of HER2 to the DNA is increased by treating cells with the ligand EGF. HER2 binding sites are spread across various genomic sites − promoter regions, exons, introns and intergenic regions. Interestingly, pathway analysis of binding sites in the region of coding genes reveals a propensity for genes involved in the EGFR signaling pathway including binding within its own gene, indicating a possible feedback loop mechanism. RNAseq analysis will be combined with ChIPexo data to determine the direct effect of the binding of HER2 at these genomic locations. Further analysis of HER2 binding in primary breast cancer tissue will be conducted. Mass spectrometry analysis has revealed EGFR and Stat3 as binding partners at the chromatin. ChIPexo analysis of these interacting proteins is underway. Conclusions: This study highlights a role for HER2 on the genome revealing an additional level of complexity and biological function for this important breast cancer factor. No conflict of interest. 411 Epigenetic mechanisms of cadmium urothelial carcinogenesis L. Cowling1 , C. Varley1 , T. Minshull2 , M. Dickman2 , J. Catto3 , J. Southgate1 . 1 University of York, Department of Biology, York, United Kingdom, 2 University of Sheffield, Department of Chemical and Biological Engineering, Sheffield, United Kingdom, 3 University of Sheffield, Department of Oncology, Sheffield, United Kingdom Background: The urothelium, the epithelial lining of the bladder, is exposed to urinary-excreted carcinogens from environmental, occupational and dietary sources. These carcinogens include heavy metals such as cadmium. As cadmium is a weak mutagen, this suggests that genetic mechanisms are not responsible for cadmium-induced carcinogenesis. Non-genotoxic carcinogenesis is relatively poorly understood, however recent advances show that epigenetic dysregulation of gene expression may play an important role in urothelial tumorigenesis. The aim of my research is to investigate epigenetic dysregulation as the candidate mechanism underlying cadmium carcinogenesis of human urothelial cells. Materials and Methods: Normal human urothelial (NHU) cells were cultured as finite cell lines following isolation from surgical specimens. When established in vitro, NHU cells have a highly proliferative phenotype and can be induced to differentiate, forming a functional barrier urothelium. Quantitative RT-PCR, western blotting and immunofluorescence microscopy were used to assess transcript and protein expression of tumour suppressor genes and urothelial differentiation markers in NHU cells following exposure to 10 mM cadmium. Additionally a mass spectrometry approach was utilised to investigate post-translational histone modification changes caused by cadmium exposure. Results: Exposure of NHU cells to cadmium inhibited expression of the tumour suppressor genes, p16, APC and RASSF1A. As dysplasia is axiomatic of carcinoma in situ, the precursor to muscle invasive urothelial carcinoma, the potential of NHU cells to differentiate in the presence of cadmium was investigated. Following exposure to cadmium, there was a failure to upregulate
archetypical differentiation-associated genes, including uroplakin 1a and 2, and cytokeratins 13 and 20. These changes were reversible using trichostatin A, a histone deacetylase inhibitor. Mass spectroscopy data showed that there was a change in histone modification marks present in NHU cells which failed to upregulate differentiation markers. An increase in repressive histone marks such as H3K9me2 and H3K9me3 was found alongside a decrease in active marks such as acetylation at H3K18 and H3K23. Conclusions: This study presents evidence that cadmium exposure changes the epigenome of NHU cells, and leads to compromised urothelial differentiation and downregulation of tumour suppressor genes, both of which may promote cadmium carcinogenesis. No conflict of interest. 412 DNA methylation levels of OPCML and SFRP1 as diagnostic biomarkers in cholangiocarcinoma T. Limpaiboon1 , R. Amornpisutt1 , S. Proungvitaya1 , P. Jearanaikoon1 . 1 CMDL Khon Kaen University, Khonkaen, Thailand Background: Cholangiocarcinoma (CCA) is a malignancy of the bile duct epithelium and is the second most common primary liver cancer by which the incidence and mortality rates of CCA are increasing worldwide. The cause of CCA in northeast Thailand shows a strong association with liver fluke (Opisthorchis viverrini) infection. The clinical symptoms of CCA reveal as the disease progresses to advanced stage. Thus, specific diagnostic biomarkers are important for this fatal disease. We established Methylation-sensitive high resolution melting (MS-HRM) to quantify DNA methylation levels of OPCML and SFRP1, and evaluated the sensitivity, specificity and accuracy of the test. Material and Methods: Genomic DNA was extracted from 73 primary CCA and 10 adjacent normal tissues. Leukocyte DNA was used as an unmethylated control and DNA treated with SssI methyltransferase was used as a methylated control. All DNA samples were modified by sodium bisulfite and subjected to MS-HRM to quantify DNA methylation levels of OPCML and SFRP1. The diagnostic performance of the test was evaluated using Receiver operating characteristic (ROC) curve analysis. Results: The median methylation level of OPCML in CCA was 38.7% (ranged from 0 to 82.2%) and of SFRP1 was 31.5% (ranged from 0 to 86.2%). Methylation cut-off values of OPCML and SFRP1 derived from adjacent normal tissue were 6.90% and 10.44%, respectively. With these cut-off values, the area under curve (AUC) of OPCML was 0.932 (95% CI: 0.878–0.986) and of SFRP1 was 0.951 (95% CI: 0.905–0.996). The sensitivity, specificity and accuracy of OPCML were 89.04, 100 and 90.36%, respectively, and of SFRP1 were 83.56, 100 and 85.54%, respectively. Conclusions: We have established MS-HRM to quantify DNA methylation levels of OPCML and SFRP1 in CCA. The methylation levels of these loci could be potential biomarkers for diagnosis of CCA with high specificity, sensitivity and accuracy. Further validation in non-invasive samples such as serum or plasma is warranted for clinical applicability, especially as early diagnostic biomarkers. No conflict of interest. 413 TET1 was an important factor for DNA demethylation to regulate the expression of MUC1 and MUC4 in lung cancer S. Yokoyama1 , S. Kitamoto1 , M. Higashi1 , H. Tsutsumida1 , J. Wakimoto2 , S. Yonezawa1 . 1 Kagoshima University, Human Pathology, Kagoshima, Japan, 2 Minami kyushu national hospital, Clinical laboratory, Kagoshima, Japan Introduction: Lung cancer is still a disease of high mortality, despite advanced diagnostic techniques. Mucins (MUC) play crucial roles in carcinogenesis and tumor invasion in lung neoplasms. Our immunohistochemistry (IHC) studies have shown a consensus position on mucin expression profiles in lung neoplasms as follows: A high MUC1/SP-A ratio is strongly associated with a poor outcome in patients with small-size lung adenocarcinoma; and High MUC4 expression in lung adenocarcinoma patients with a poor outcome. We also showed that these mucin genes (MUC1 and MUC4) expression in cancer cell lines was regulated by DNA methylation. Material and Method: We have developed a novel ‘methylation-specific electrophoresis (MSE)’ method to analyze the DNA methylation status of MUC1 and MUC4 by high sensitivity and resolution. By using the MSE method, we evaluated lung tissue samples from 33 patients with various lung lesions. The results were compared with expression level of MUC1 and MUC4 in IHC. They were also compared with expression levels of mucin mRNAs and several DNA demethylation factors (e.g. Ten-eleven translocation: TET, activationinduced cytidine deaminase: AID and glial cells missing: GCM) using real time PCR analysis. Results and Discussion: The results indicated that the DNA methylation status of MUC1 and MUC4 in lung tissue matched with the expression level of mucin mRNA. In addition, the expression level of TET1 mRNA showed significant correlation with these mucin mRNAs expression level as well as DNA methylation status. Furthermore, the treatment of lung cancer cell line