No abstract

Epigenetic gene inactivation through promoter hypermethylation is an important aberration involved in the silencing of tumor-associated genes in cancer. Here we identified the apoptosis associated tyrosine kinase (AATK) as an epigenetically downregulated tumor related gene. We analyzed the epigenetic regulation of AATK in several human cancer cell lines and normal tissues by methylation and expression analysis. Hypermethylation of AATK was also analyzed in 25 primary lung tumors, 30 breast cancers and 24 matching breast tissues. In normal tissues the AATK CpG island promoter was unmethylated and AATK was expressed. Hypermethylation of AATK occurred frequently in 13 out of 14 (93%) human cancer cell lines. Methylation was reversed by 5-aza-2′-deoxycytidine treatment leading to re-expression of AATK in cancer cell lines. Aberrant methylation of AATK was also revealed in primary lung (40%) and breast (53%) cancers, but was found to be significantly less methylated in matching normal breast tissues (17%; p<0.01). In addition, we observed that AATK is epigenetically reactivated through the chromatin regulator CTCF. We further show that overexpression of Aatk significantly suppresses colony formation in cancer cell lines. Our findings suggest that the apoptosis associated tyrosine kinase is frequently inactivated in human cancers and acts as a tumor suppressive gene.

BackgroundDual specificity phosphatases are a class of tumor-associated proteins involved in the negative regulation of the MAP kinase pathway. Downregulation of the dual specificity phosphatase 2 (DUSP2) has been reported in cancer. Epigenetic silencing of tumor suppressor genes by abnormal promoter methylation is a frequent mechanism in oncogenesis. It has been shown that the epigenetic factor CTCF is involved in the regulation of tumor suppressor genes.MethodsWe analyzed the promoter hypermethylation of DUSP2 in human cancer, including primary Merkel cell carcinoma by bisulfite restriction analysis and pyrosequencing. Moreover we analyzed the impact of a DNA methyltransferase inhibitor (5-Aza-dC) and CTCF on the epigenetic regulation of DUSP2 by qRT-PCR, promoter assay, chromatin immuno-precipitation and methylation analysis.ResultsHere we report a significant tumor-specific hypermethylation of DUSP2 in primary Merkel cell carcinoma (p = 0.05). An increase in methylation of DUSP2 was also found in 17 out of 24 (71 %) cancer cell lines, including skin and lung cancer. Treatment of cancer cells with 5-Aza-dC induced DUSP2 expression by its promoter demethylation, Additionally we observed that CTCF induces DUSP2 expression in cell lines that exhibit silencing of DUSP2. This reactivation was accompanied by increased CTCF binding and demethylation of the DUSP2 promoter.ConclusionsOur data show that aberrant epigenetic inactivation of DUSP2 occurs in carcinogenesis and that CTCF is involved in the epigenetic regulation of DUSP2 expression.Electronic supplementary materialThe online version of this article (doi:10.1186/s12885-016-2087-6) contains supplementary material, which is available to authorized users.

Pheochromocytomas (PCCs) are rare neuroendocrine tumors that arise from the medulla of the adrenal gland or the sympathetic ganglia and are characterized by the secretion of catecholamines. In 30–40% of patients, PCCs are genetically determined by susceptibility genes as various as RET, VHL, and NF1. We have analyzed the Ras-association domain family members (RASSFs) in PCCs regarding their inactivating promoter hypermethylation status. Previously, we reported a promoter methylation in PCC for the first family member RASSF1A. Promoter hypermethylation of CpG islands leads to the silencing of the according transcript and is a common mechanism for inactivation of tumor suppressors. In this study, we observed inactivating DNA modifications for the RASSF members RASSF2, RASSF5A, RASSF9, and RASSF10, but not for the members RASSF3, RASSF4, RASSF5C, RASSF6, RASSF7, and RASSF8. The degree of promoter methylation was 19% for RASSF2, 67% for RASSF5A, 18% for RASSF9, and 74% for RASSF10. Interestingly, the degree of hypermethylation for RASSF10 in hereditary PCCs was 89 vs. 60% in sporadic PCCs. A similar but less dramatic effect was observed in RASSF5A and RASSF9. Including all RASSF members, we found that of 25 PCCs, 92% show promoter methylation in at least in one RASSF member. In 75% of the hereditary PCC samples, we found two or more methylated RASSF promoters, whereas in sporadic PCCs only 46% were observed. In summary, we could show that in PCC several RASSF members are strongly hypermethylated in their promoter regions and methylation of more than one RASSF member occurs in the majority of PCCs. This adds the inactivation of genes of the RASSF tumor suppressor family to the already known deregulated genes of PCC.